thamada/

llama-3.1-70b-inst-orig

{'dim': 8192, 'n_layers': 80, 'n_heads': 64, 'n_kv_heads': 8, 'vocab_size': 128256, 'ffn_dim_multiplier': 1.3, 'multiple_of': 4096, 'norm_eps': 1e-05, 'rope_theta': 500000.0, 'use_scaled_rope': True}
000: original/consolidated.00.pth
001: original/consolidated.01.pth
002: original/consolidated.02.pth
003: original/consolidated.03.pth
004: original/consolidated.04.pth
005: original/consolidated.05.pth
006: original/consolidated.06.pth
007: original/consolidated.07.pth

-----------------------------------------------------------------------------
0 params in total.
0 bytes in total.
-----------------------------------------------------------------------------

[1/8]: Loading original/consolidated.00.pth
   0 :  131334144 : tok_embeddings.weight               : [16032, 8192]   : torch.bfloat16
   1 :    8388608 : layers.0.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
   2 :    1048576 : layers.0.attention.wk.weight        : [128, 8192]     : torch.bfloat16
   3 :    1048576 : layers.0.attention.wv.weight        : [128, 8192]     : torch.bfloat16
   4 :    8388608 : layers.0.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
   5 :   29360128 : layers.0.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
   6 :   29360128 : layers.0.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
   7 :   29360128 : layers.0.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
   8 :       8192 : layers.0.attention_norm.weight      : [8192]          : torch.bfloat16
   9 :       8192 : layers.0.ffn_norm.weight            : [8192]          : torch.bfloat16
  10 :    8388608 : layers.1.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  11 :    1048576 : layers.1.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  12 :    1048576 : layers.1.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  13 :    8388608 : layers.1.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  14 :   29360128 : layers.1.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  15 :   29360128 : layers.1.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  16 :   29360128 : layers.1.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  17 :       8192 : layers.1.attention_norm.weight      : [8192]          : torch.bfloat16
  18 :       8192 : layers.1.ffn_norm.weight            : [8192]          : torch.bfloat16
  19 :    8388608 : layers.2.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  20 :    1048576 : layers.2.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  21 :    1048576 : layers.2.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  22 :    8388608 : layers.2.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  23 :   29360128 : layers.2.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  24 :   29360128 : layers.2.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  25 :   29360128 : layers.2.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  26 :       8192 : layers.2.attention_norm.weight      : [8192]          : torch.bfloat16
  27 :       8192 : layers.2.ffn_norm.weight            : [8192]          : torch.bfloat16
  28 :    8388608 : layers.3.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  29 :    1048576 : layers.3.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  30 :    1048576 : layers.3.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  31 :    8388608 : layers.3.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  32 :   29360128 : layers.3.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  33 :   29360128 : layers.3.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  34 :   29360128 : layers.3.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  35 :       8192 : layers.3.attention_norm.weight      : [8192]          : torch.bfloat16
  36 :       8192 : layers.3.ffn_norm.weight            : [8192]          : torch.bfloat16
  37 :    8388608 : layers.4.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  38 :    1048576 : layers.4.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  39 :    1048576 : layers.4.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  40 :    8388608 : layers.4.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  41 :   29360128 : layers.4.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  42 :   29360128 : layers.4.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  43 :   29360128 : layers.4.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  44 :       8192 : layers.4.attention_norm.weight      : [8192]          : torch.bfloat16
  45 :       8192 : layers.4.ffn_norm.weight            : [8192]          : torch.bfloat16
  46 :    8388608 : layers.5.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  47 :    1048576 : layers.5.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  48 :    1048576 : layers.5.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  49 :    8388608 : layers.5.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  50 :   29360128 : layers.5.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  51 :   29360128 : layers.5.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  52 :   29360128 : layers.5.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  53 :       8192 : layers.5.attention_norm.weight      : [8192]          : torch.bfloat16
  54 :       8192 : layers.5.ffn_norm.weight            : [8192]          : torch.bfloat16
  55 :    8388608 : layers.6.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  56 :    1048576 : layers.6.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  57 :    1048576 : layers.6.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  58 :    8388608 : layers.6.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  59 :   29360128 : layers.6.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  60 :   29360128 : layers.6.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  61 :   29360128 : layers.6.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  62 :       8192 : layers.6.attention_norm.weight      : [8192]          : torch.bfloat16
  63 :       8192 : layers.6.ffn_norm.weight            : [8192]          : torch.bfloat16
  64 :    8388608 : layers.7.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  65 :    1048576 : layers.7.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  66 :    1048576 : layers.7.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  67 :    8388608 : layers.7.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  68 :   29360128 : layers.7.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  69 :   29360128 : layers.7.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  70 :   29360128 : layers.7.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  71 :       8192 : layers.7.attention_norm.weight      : [8192]          : torch.bfloat16
  72 :       8192 : layers.7.ffn_norm.weight            : [8192]          : torch.bfloat16
  73 :    8388608 : layers.8.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  74 :    1048576 : layers.8.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  75 :    1048576 : layers.8.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  76 :    8388608 : layers.8.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  77 :   29360128 : layers.8.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  78 :   29360128 : layers.8.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  79 :   29360128 : layers.8.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  80 :       8192 : layers.8.attention_norm.weight      : [8192]          : torch.bfloat16
  81 :       8192 : layers.8.ffn_norm.weight            : [8192]          : torch.bfloat16
  82 :    8388608 : layers.9.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  83 :    1048576 : layers.9.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  84 :    1048576 : layers.9.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  85 :    8388608 : layers.9.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  86 :   29360128 : layers.9.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  87 :   29360128 : layers.9.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  88 :   29360128 : layers.9.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  89 :       8192 : layers.9.attention_norm.weight      : [8192]          : torch.bfloat16
  90 :       8192 : layers.9.ffn_norm.weight            : [8192]          : torch.bfloat16
  91 :    8388608 : layers.10.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
  92 :    1048576 : layers.10.attention.wk.weight       : [128, 8192]     : torch.bfloat16
  93 :    1048576 : layers.10.attention.wv.weight       : [128, 8192]     : torch.bfloat16
  94 :    8388608 : layers.10.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
  95 :   29360128 : layers.10.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
  96 :   29360128 : layers.10.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
  97 :   29360128 : layers.10.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
  98 :       8192 : layers.10.attention_norm.weight     : [8192]          : torch.bfloat16
  99 :       8192 : layers.10.ffn_norm.weight           : [8192]          : torch.bfloat16
 100 :    8388608 : layers.11.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 101 :    1048576 : layers.11.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 102 :    1048576 : layers.11.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 103 :    8388608 : layers.11.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 104 :   29360128 : layers.11.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 105 :   29360128 : layers.11.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 106 :   29360128 : layers.11.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 107 :       8192 : layers.11.attention_norm.weight     : [8192]          : torch.bfloat16
 108 :       8192 : layers.11.ffn_norm.weight           : [8192]          : torch.bfloat16
 109 :    8388608 : layers.12.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 110 :    1048576 : layers.12.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 111 :    1048576 : layers.12.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 112 :    8388608 : layers.12.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 113 :   29360128 : layers.12.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 114 :   29360128 : layers.12.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 115 :   29360128 : layers.12.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 116 :       8192 : layers.12.attention_norm.weight     : [8192]          : torch.bfloat16
 117 :       8192 : layers.12.ffn_norm.weight           : [8192]          : torch.bfloat16
 118 :    8388608 : layers.13.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 119 :    1048576 : layers.13.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 120 :    1048576 : layers.13.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 121 :    8388608 : layers.13.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 122 :   29360128 : layers.13.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 123 :   29360128 : layers.13.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 124 :   29360128 : layers.13.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 125 :       8192 : layers.13.attention_norm.weight     : [8192]          : torch.bfloat16
 126 :       8192 : layers.13.ffn_norm.weight           : [8192]          : torch.bfloat16
 127 :    8388608 : layers.14.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 128 :    1048576 : layers.14.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 129 :    1048576 : layers.14.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 130 :    8388608 : layers.14.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 131 :   29360128 : layers.14.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 132 :   29360128 : layers.14.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 133 :   29360128 : layers.14.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 134 :       8192 : layers.14.attention_norm.weight     : [8192]          : torch.bfloat16
 135 :       8192 : layers.14.ffn_norm.weight           : [8192]          : torch.bfloat16
 136 :    8388608 : layers.15.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 137 :    1048576 : layers.15.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 138 :    1048576 : layers.15.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 139 :    8388608 : layers.15.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 140 :   29360128 : layers.15.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 141 :   29360128 : layers.15.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 142 :   29360128 : layers.15.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 143 :       8192 : layers.15.attention_norm.weight     : [8192]          : torch.bfloat16
 144 :       8192 : layers.15.ffn_norm.weight           : [8192]          : torch.bfloat16
 145 :    8388608 : layers.16.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 146 :    1048576 : layers.16.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 147 :    1048576 : layers.16.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 148 :    8388608 : layers.16.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 149 :   29360128 : layers.16.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 150 :   29360128 : layers.16.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 151 :   29360128 : layers.16.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 152 :       8192 : layers.16.attention_norm.weight     : [8192]          : torch.bfloat16
 153 :       8192 : layers.16.ffn_norm.weight           : [8192]          : torch.bfloat16
 154 :    8388608 : layers.17.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 155 :    1048576 : layers.17.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 156 :    1048576 : layers.17.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 157 :    8388608 : layers.17.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 158 :   29360128 : layers.17.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 159 :   29360128 : layers.17.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 160 :   29360128 : layers.17.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 161 :       8192 : layers.17.attention_norm.weight     : [8192]          : torch.bfloat16
 162 :       8192 : layers.17.ffn_norm.weight           : [8192]          : torch.bfloat16
 163 :    8388608 : layers.18.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 164 :    1048576 : layers.18.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 165 :    1048576 : layers.18.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 166 :    8388608 : layers.18.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 167 :   29360128 : layers.18.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 168 :   29360128 : layers.18.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 169 :   29360128 : layers.18.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 170 :       8192 : layers.18.attention_norm.weight     : [8192]          : torch.bfloat16
 171 :       8192 : layers.18.ffn_norm.weight           : [8192]          : torch.bfloat16
 172 :    8388608 : layers.19.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 173 :    1048576 : layers.19.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 174 :    1048576 : layers.19.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 175 :    8388608 : layers.19.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 176 :   29360128 : layers.19.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 177 :   29360128 : layers.19.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 178 :   29360128 : layers.19.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 179 :       8192 : layers.19.attention_norm.weight     : [8192]          : torch.bfloat16
 180 :       8192 : layers.19.ffn_norm.weight           : [8192]          : torch.bfloat16
 181 :    8388608 : layers.20.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 182 :    1048576 : layers.20.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 183 :    1048576 : layers.20.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 184 :    8388608 : layers.20.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 185 :   29360128 : layers.20.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 186 :   29360128 : layers.20.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 187 :   29360128 : layers.20.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 188 :       8192 : layers.20.attention_norm.weight     : [8192]          : torch.bfloat16
 189 :       8192 : layers.20.ffn_norm.weight           : [8192]          : torch.bfloat16
 190 :    8388608 : layers.21.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 191 :    1048576 : layers.21.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 192 :    1048576 : layers.21.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 193 :    8388608 : layers.21.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 194 :   29360128 : layers.21.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 195 :   29360128 : layers.21.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 196 :   29360128 : layers.21.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 197 :       8192 : layers.21.attention_norm.weight     : [8192]          : torch.bfloat16
 198 :       8192 : layers.21.ffn_norm.weight           : [8192]          : torch.bfloat16
 199 :    8388608 : layers.22.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 200 :    1048576 : layers.22.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 201 :    1048576 : layers.22.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 202 :    8388608 : layers.22.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 203 :   29360128 : layers.22.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 204 :   29360128 : layers.22.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 205 :   29360128 : layers.22.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 206 :       8192 : layers.22.attention_norm.weight     : [8192]          : torch.bfloat16
 207 :       8192 : layers.22.ffn_norm.weight           : [8192]          : torch.bfloat16
 208 :    8388608 : layers.23.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 209 :    1048576 : layers.23.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 210 :    1048576 : layers.23.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 211 :    8388608 : layers.23.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 212 :   29360128 : layers.23.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 213 :   29360128 : layers.23.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 214 :   29360128 : layers.23.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 215 :       8192 : layers.23.attention_norm.weight     : [8192]          : torch.bfloat16
 216 :       8192 : layers.23.ffn_norm.weight           : [8192]          : torch.bfloat16
 217 :    8388608 : layers.24.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 218 :    1048576 : layers.24.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 219 :    1048576 : layers.24.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 220 :    8388608 : layers.24.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 221 :   29360128 : layers.24.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 222 :   29360128 : layers.24.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 223 :   29360128 : layers.24.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 224 :       8192 : layers.24.attention_norm.weight     : [8192]          : torch.bfloat16
 225 :       8192 : layers.24.ffn_norm.weight           : [8192]          : torch.bfloat16
 226 :    8388608 : layers.25.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 227 :    1048576 : layers.25.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 228 :    1048576 : layers.25.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 229 :    8388608 : layers.25.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 230 :   29360128 : layers.25.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 231 :   29360128 : layers.25.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 232 :   29360128 : layers.25.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 233 :       8192 : layers.25.attention_norm.weight     : [8192]          : torch.bfloat16
 234 :       8192 : layers.25.ffn_norm.weight           : [8192]          : torch.bfloat16
 235 :    8388608 : layers.26.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 236 :    1048576 : layers.26.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 237 :    1048576 : layers.26.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 238 :    8388608 : layers.26.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 239 :   29360128 : layers.26.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 240 :   29360128 : layers.26.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 241 :   29360128 : layers.26.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 242 :       8192 : layers.26.attention_norm.weight     : [8192]          : torch.bfloat16
 243 :       8192 : layers.26.ffn_norm.weight           : [8192]          : torch.bfloat16
 244 :    8388608 : layers.27.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 245 :    1048576 : layers.27.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 246 :    1048576 : layers.27.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 247 :    8388608 : layers.27.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 248 :   29360128 : layers.27.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 249 :   29360128 : layers.27.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 250 :   29360128 : layers.27.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 251 :       8192 : layers.27.attention_norm.weight     : [8192]          : torch.bfloat16
 252 :       8192 : layers.27.ffn_norm.weight           : [8192]          : torch.bfloat16
 253 :    8388608 : layers.28.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 254 :    1048576 : layers.28.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 255 :    1048576 : layers.28.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 256 :    8388608 : layers.28.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 257 :   29360128 : layers.28.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 258 :   29360128 : layers.28.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 259 :   29360128 : layers.28.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 260 :       8192 : layers.28.attention_norm.weight     : [8192]          : torch.bfloat16
 261 :       8192 : layers.28.ffn_norm.weight           : [8192]          : torch.bfloat16
 262 :    8388608 : layers.29.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 263 :    1048576 : layers.29.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 264 :    1048576 : layers.29.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 265 :    8388608 : layers.29.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 266 :   29360128 : layers.29.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 267 :   29360128 : layers.29.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 268 :   29360128 : layers.29.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 269 :       8192 : layers.29.attention_norm.weight     : [8192]          : torch.bfloat16
 270 :       8192 : layers.29.ffn_norm.weight           : [8192]          : torch.bfloat16
 271 :    8388608 : layers.30.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 272 :    1048576 : layers.30.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 273 :    1048576 : layers.30.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 274 :    8388608 : layers.30.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 275 :   29360128 : layers.30.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 276 :   29360128 : layers.30.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 277 :   29360128 : layers.30.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 278 :       8192 : layers.30.attention_norm.weight     : [8192]          : torch.bfloat16
 279 :       8192 : layers.30.ffn_norm.weight           : [8192]          : torch.bfloat16
 280 :    8388608 : layers.31.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 281 :    1048576 : layers.31.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 282 :    1048576 : layers.31.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 283 :    8388608 : layers.31.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 284 :   29360128 : layers.31.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 285 :   29360128 : layers.31.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 286 :   29360128 : layers.31.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 287 :       8192 : layers.31.attention_norm.weight     : [8192]          : torch.bfloat16
 288 :       8192 : layers.31.ffn_norm.weight           : [8192]          : torch.bfloat16
 289 :    8388608 : layers.32.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 290 :    1048576 : layers.32.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 291 :    1048576 : layers.32.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 292 :    8388608 : layers.32.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 293 :   29360128 : layers.32.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 294 :   29360128 : layers.32.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 295 :   29360128 : layers.32.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 296 :       8192 : layers.32.attention_norm.weight     : [8192]          : torch.bfloat16
 297 :       8192 : layers.32.ffn_norm.weight           : [8192]          : torch.bfloat16
 298 :    8388608 : layers.33.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 299 :    1048576 : layers.33.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 300 :    1048576 : layers.33.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 301 :    8388608 : layers.33.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 302 :   29360128 : layers.33.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 303 :   29360128 : layers.33.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 304 :   29360128 : layers.33.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 305 :       8192 : layers.33.attention_norm.weight     : [8192]          : torch.bfloat16
 306 :       8192 : layers.33.ffn_norm.weight           : [8192]          : torch.bfloat16
 307 :    8388608 : layers.34.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 308 :    1048576 : layers.34.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 309 :    1048576 : layers.34.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 310 :    8388608 : layers.34.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 311 :   29360128 : layers.34.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 312 :   29360128 : layers.34.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 313 :   29360128 : layers.34.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 314 :       8192 : layers.34.attention_norm.weight     : [8192]          : torch.bfloat16
 315 :       8192 : layers.34.ffn_norm.weight           : [8192]          : torch.bfloat16
 316 :    8388608 : layers.35.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 317 :    1048576 : layers.35.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 318 :    1048576 : layers.35.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 319 :    8388608 : layers.35.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 320 :   29360128 : layers.35.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 321 :   29360128 : layers.35.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 322 :   29360128 : layers.35.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 323 :       8192 : layers.35.attention_norm.weight     : [8192]          : torch.bfloat16
 324 :       8192 : layers.35.ffn_norm.weight           : [8192]          : torch.bfloat16
 325 :    8388608 : layers.36.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 326 :    1048576 : layers.36.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 327 :    1048576 : layers.36.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 328 :    8388608 : layers.36.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 329 :   29360128 : layers.36.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 330 :   29360128 : layers.36.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 331 :   29360128 : layers.36.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 332 :       8192 : layers.36.attention_norm.weight     : [8192]          : torch.bfloat16
 333 :       8192 : layers.36.ffn_norm.weight           : [8192]          : torch.bfloat16
 334 :    8388608 : layers.37.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 335 :    1048576 : layers.37.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 336 :    1048576 : layers.37.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 337 :    8388608 : layers.37.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 338 :   29360128 : layers.37.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 339 :   29360128 : layers.37.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 340 :   29360128 : layers.37.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 341 :       8192 : layers.37.attention_norm.weight     : [8192]          : torch.bfloat16
 342 :       8192 : layers.37.ffn_norm.weight           : [8192]          : torch.bfloat16
 343 :    8388608 : layers.38.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 344 :    1048576 : layers.38.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 345 :    1048576 : layers.38.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 346 :    8388608 : layers.38.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 347 :   29360128 : layers.38.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 348 :   29360128 : layers.38.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 349 :   29360128 : layers.38.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 350 :       8192 : layers.38.attention_norm.weight     : [8192]          : torch.bfloat16
 351 :       8192 : layers.38.ffn_norm.weight           : [8192]          : torch.bfloat16
 352 :    8388608 : layers.39.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 353 :    1048576 : layers.39.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 354 :    1048576 : layers.39.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 355 :    8388608 : layers.39.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 356 :   29360128 : layers.39.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 357 :   29360128 : layers.39.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 358 :   29360128 : layers.39.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 359 :       8192 : layers.39.attention_norm.weight     : [8192]          : torch.bfloat16
 360 :       8192 : layers.39.ffn_norm.weight           : [8192]          : torch.bfloat16
 361 :    8388608 : layers.40.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 362 :    1048576 : layers.40.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 363 :    1048576 : layers.40.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 364 :    8388608 : layers.40.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 365 :   29360128 : layers.40.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 366 :   29360128 : layers.40.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 367 :   29360128 : layers.40.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 368 :       8192 : layers.40.attention_norm.weight     : [8192]          : torch.bfloat16
 369 :       8192 : layers.40.ffn_norm.weight           : [8192]          : torch.bfloat16
 370 :    8388608 : layers.41.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 371 :    1048576 : layers.41.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 372 :    1048576 : layers.41.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 373 :    8388608 : layers.41.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 374 :   29360128 : layers.41.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 375 :   29360128 : layers.41.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 376 :   29360128 : layers.41.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 377 :       8192 : layers.41.attention_norm.weight     : [8192]          : torch.bfloat16
 378 :       8192 : layers.41.ffn_norm.weight           : [8192]          : torch.bfloat16
 379 :    8388608 : layers.42.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 380 :    1048576 : layers.42.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 381 :    1048576 : layers.42.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 382 :    8388608 : layers.42.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 383 :   29360128 : layers.42.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 384 :   29360128 : layers.42.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 385 :   29360128 : layers.42.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 386 :       8192 : layers.42.attention_norm.weight     : [8192]          : torch.bfloat16
 387 :       8192 : layers.42.ffn_norm.weight           : [8192]          : torch.bfloat16
 388 :    8388608 : layers.43.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 389 :    1048576 : layers.43.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 390 :    1048576 : layers.43.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 391 :    8388608 : layers.43.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 392 :   29360128 : layers.43.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 393 :   29360128 : layers.43.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 394 :   29360128 : layers.43.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 395 :       8192 : layers.43.attention_norm.weight     : [8192]          : torch.bfloat16
 396 :       8192 : layers.43.ffn_norm.weight           : [8192]          : torch.bfloat16
 397 :    8388608 : layers.44.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 398 :    1048576 : layers.44.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 399 :    1048576 : layers.44.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 400 :    8388608 : layers.44.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 401 :   29360128 : layers.44.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 402 :   29360128 : layers.44.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 403 :   29360128 : layers.44.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 404 :       8192 : layers.44.attention_norm.weight     : [8192]          : torch.bfloat16
 405 :       8192 : layers.44.ffn_norm.weight           : [8192]          : torch.bfloat16
 406 :    8388608 : layers.45.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 407 :    1048576 : layers.45.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 408 :    1048576 : layers.45.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 409 :    8388608 : layers.45.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 410 :   29360128 : layers.45.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 411 :   29360128 : layers.45.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 412 :   29360128 : layers.45.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 413 :       8192 : layers.45.attention_norm.weight     : [8192]          : torch.bfloat16
 414 :       8192 : layers.45.ffn_norm.weight           : [8192]          : torch.bfloat16
 415 :    8388608 : layers.46.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 416 :    1048576 : layers.46.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 417 :    1048576 : layers.46.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 418 :    8388608 : layers.46.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 419 :   29360128 : layers.46.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 420 :   29360128 : layers.46.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 421 :   29360128 : layers.46.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 422 :       8192 : layers.46.attention_norm.weight     : [8192]          : torch.bfloat16
 423 :       8192 : layers.46.ffn_norm.weight           : [8192]          : torch.bfloat16
 424 :    8388608 : layers.47.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 425 :    1048576 : layers.47.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 426 :    1048576 : layers.47.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 427 :    8388608 : layers.47.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 428 :   29360128 : layers.47.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 429 :   29360128 : layers.47.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 430 :   29360128 : layers.47.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 431 :       8192 : layers.47.attention_norm.weight     : [8192]          : torch.bfloat16
 432 :       8192 : layers.47.ffn_norm.weight           : [8192]          : torch.bfloat16
 433 :    8388608 : layers.48.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 434 :    1048576 : layers.48.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 435 :    1048576 : layers.48.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 436 :    8388608 : layers.48.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 437 :   29360128 : layers.48.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 438 :   29360128 : layers.48.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 439 :   29360128 : layers.48.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 440 :       8192 : layers.48.attention_norm.weight     : [8192]          : torch.bfloat16
 441 :       8192 : layers.48.ffn_norm.weight           : [8192]          : torch.bfloat16
 442 :    8388608 : layers.49.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 443 :    1048576 : layers.49.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 444 :    1048576 : layers.49.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 445 :    8388608 : layers.49.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 446 :   29360128 : layers.49.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 447 :   29360128 : layers.49.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 448 :   29360128 : layers.49.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 449 :       8192 : layers.49.attention_norm.weight     : [8192]          : torch.bfloat16
 450 :       8192 : layers.49.ffn_norm.weight           : [8192]          : torch.bfloat16
 451 :    8388608 : layers.50.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 452 :    1048576 : layers.50.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 453 :    1048576 : layers.50.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 454 :    8388608 : layers.50.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 455 :   29360128 : layers.50.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 456 :   29360128 : layers.50.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 457 :   29360128 : layers.50.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 458 :       8192 : layers.50.attention_norm.weight     : [8192]          : torch.bfloat16
 459 :       8192 : layers.50.ffn_norm.weight           : [8192]          : torch.bfloat16
 460 :    8388608 : layers.51.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 461 :    1048576 : layers.51.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 462 :    1048576 : layers.51.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 463 :    8388608 : layers.51.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 464 :   29360128 : layers.51.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 465 :   29360128 : layers.51.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 466 :   29360128 : layers.51.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 467 :       8192 : layers.51.attention_norm.weight     : [8192]          : torch.bfloat16
 468 :       8192 : layers.51.ffn_norm.weight           : [8192]          : torch.bfloat16
 469 :    8388608 : layers.52.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 470 :    1048576 : layers.52.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 471 :    1048576 : layers.52.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 472 :    8388608 : layers.52.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 473 :   29360128 : layers.52.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 474 :   29360128 : layers.52.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 475 :   29360128 : layers.52.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 476 :       8192 : layers.52.attention_norm.weight     : [8192]          : torch.bfloat16
 477 :       8192 : layers.52.ffn_norm.weight           : [8192]          : torch.bfloat16
 478 :    8388608 : layers.53.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 479 :    1048576 : layers.53.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 480 :    1048576 : layers.53.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 481 :    8388608 : layers.53.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 482 :   29360128 : layers.53.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 483 :   29360128 : layers.53.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 484 :   29360128 : layers.53.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 485 :       8192 : layers.53.attention_norm.weight     : [8192]          : torch.bfloat16
 486 :       8192 : layers.53.ffn_norm.weight           : [8192]          : torch.bfloat16
 487 :    8388608 : layers.54.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 488 :    1048576 : layers.54.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 489 :    1048576 : layers.54.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 490 :    8388608 : layers.54.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 491 :   29360128 : layers.54.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 492 :   29360128 : layers.54.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 493 :   29360128 : layers.54.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 494 :       8192 : layers.54.attention_norm.weight     : [8192]          : torch.bfloat16
 495 :       8192 : layers.54.ffn_norm.weight           : [8192]          : torch.bfloat16
 496 :    8388608 : layers.55.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 497 :    1048576 : layers.55.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 498 :    1048576 : layers.55.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 499 :    8388608 : layers.55.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 500 :   29360128 : layers.55.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 501 :   29360128 : layers.55.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 502 :   29360128 : layers.55.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 503 :       8192 : layers.55.attention_norm.weight     : [8192]          : torch.bfloat16
 504 :       8192 : layers.55.ffn_norm.weight           : [8192]          : torch.bfloat16
 505 :    8388608 : layers.56.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 506 :    1048576 : layers.56.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 507 :    1048576 : layers.56.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 508 :    8388608 : layers.56.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 509 :   29360128 : layers.56.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 510 :   29360128 : layers.56.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 511 :   29360128 : layers.56.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 512 :       8192 : layers.56.attention_norm.weight     : [8192]          : torch.bfloat16
 513 :       8192 : layers.56.ffn_norm.weight           : [8192]          : torch.bfloat16
 514 :    8388608 : layers.57.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 515 :    1048576 : layers.57.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 516 :    1048576 : layers.57.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 517 :    8388608 : layers.57.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 518 :   29360128 : layers.57.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 519 :   29360128 : layers.57.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 520 :   29360128 : layers.57.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 521 :       8192 : layers.57.attention_norm.weight     : [8192]          : torch.bfloat16
 522 :       8192 : layers.57.ffn_norm.weight           : [8192]          : torch.bfloat16
 523 :    8388608 : layers.58.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 524 :    1048576 : layers.58.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 525 :    1048576 : layers.58.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 526 :    8388608 : layers.58.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 527 :   29360128 : layers.58.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 528 :   29360128 : layers.58.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 529 :   29360128 : layers.58.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 530 :       8192 : layers.58.attention_norm.weight     : [8192]          : torch.bfloat16
 531 :       8192 : layers.58.ffn_norm.weight           : [8192]          : torch.bfloat16
 532 :    8388608 : layers.59.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 533 :    1048576 : layers.59.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 534 :    1048576 : layers.59.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 535 :    8388608 : layers.59.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 536 :   29360128 : layers.59.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 537 :   29360128 : layers.59.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 538 :   29360128 : layers.59.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 539 :       8192 : layers.59.attention_norm.weight     : [8192]          : torch.bfloat16
 540 :       8192 : layers.59.ffn_norm.weight           : [8192]          : torch.bfloat16
 541 :    8388608 : layers.60.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 542 :    1048576 : layers.60.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 543 :    1048576 : layers.60.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 544 :    8388608 : layers.60.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 545 :   29360128 : layers.60.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 546 :   29360128 : layers.60.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 547 :   29360128 : layers.60.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 548 :       8192 : layers.60.attention_norm.weight     : [8192]          : torch.bfloat16
 549 :       8192 : layers.60.ffn_norm.weight           : [8192]          : torch.bfloat16
 550 :    8388608 : layers.61.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 551 :    1048576 : layers.61.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 552 :    1048576 : layers.61.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 553 :    8388608 : layers.61.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 554 :   29360128 : layers.61.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 555 :   29360128 : layers.61.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 556 :   29360128 : layers.61.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 557 :       8192 : layers.61.attention_norm.weight     : [8192]          : torch.bfloat16
 558 :       8192 : layers.61.ffn_norm.weight           : [8192]          : torch.bfloat16
 559 :    8388608 : layers.62.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 560 :    1048576 : layers.62.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 561 :    1048576 : layers.62.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 562 :    8388608 : layers.62.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 563 :   29360128 : layers.62.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 564 :   29360128 : layers.62.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 565 :   29360128 : layers.62.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 566 :       8192 : layers.62.attention_norm.weight     : [8192]          : torch.bfloat16
 567 :       8192 : layers.62.ffn_norm.weight           : [8192]          : torch.bfloat16
 568 :    8388608 : layers.63.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 569 :    1048576 : layers.63.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 570 :    1048576 : layers.63.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 571 :    8388608 : layers.63.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 572 :   29360128 : layers.63.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 573 :   29360128 : layers.63.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 574 :   29360128 : layers.63.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 575 :       8192 : layers.63.attention_norm.weight     : [8192]          : torch.bfloat16
 576 :       8192 : layers.63.ffn_norm.weight           : [8192]          : torch.bfloat16
 577 :    8388608 : layers.64.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 578 :    1048576 : layers.64.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 579 :    1048576 : layers.64.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 580 :    8388608 : layers.64.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 581 :   29360128 : layers.64.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 582 :   29360128 : layers.64.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 583 :   29360128 : layers.64.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 584 :       8192 : layers.64.attention_norm.weight     : [8192]          : torch.bfloat16
 585 :       8192 : layers.64.ffn_norm.weight           : [8192]          : torch.bfloat16
 586 :    8388608 : layers.65.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 587 :    1048576 : layers.65.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 588 :    1048576 : layers.65.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 589 :    8388608 : layers.65.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 590 :   29360128 : layers.65.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 591 :   29360128 : layers.65.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 592 :   29360128 : layers.65.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 593 :       8192 : layers.65.attention_norm.weight     : [8192]          : torch.bfloat16
 594 :       8192 : layers.65.ffn_norm.weight           : [8192]          : torch.bfloat16
 595 :    8388608 : layers.66.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 596 :    1048576 : layers.66.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 597 :    1048576 : layers.66.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 598 :    8388608 : layers.66.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 599 :   29360128 : layers.66.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 600 :   29360128 : layers.66.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 601 :   29360128 : layers.66.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 602 :       8192 : layers.66.attention_norm.weight     : [8192]          : torch.bfloat16
 603 :       8192 : layers.66.ffn_norm.weight           : [8192]          : torch.bfloat16
 604 :    8388608 : layers.67.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 605 :    1048576 : layers.67.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 606 :    1048576 : layers.67.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 607 :    8388608 : layers.67.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 608 :   29360128 : layers.67.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 609 :   29360128 : layers.67.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 610 :   29360128 : layers.67.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 611 :       8192 : layers.67.attention_norm.weight     : [8192]          : torch.bfloat16
 612 :       8192 : layers.67.ffn_norm.weight           : [8192]          : torch.bfloat16
 613 :    8388608 : layers.68.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 614 :    1048576 : layers.68.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 615 :    1048576 : layers.68.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 616 :    8388608 : layers.68.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 617 :   29360128 : layers.68.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 618 :   29360128 : layers.68.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 619 :   29360128 : layers.68.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 620 :       8192 : layers.68.attention_norm.weight     : [8192]          : torch.bfloat16
 621 :       8192 : layers.68.ffn_norm.weight           : [8192]          : torch.bfloat16
 622 :    8388608 : layers.69.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 623 :    1048576 : layers.69.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 624 :    1048576 : layers.69.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 625 :    8388608 : layers.69.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 626 :   29360128 : layers.69.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 627 :   29360128 : layers.69.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 628 :   29360128 : layers.69.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 629 :       8192 : layers.69.attention_norm.weight     : [8192]          : torch.bfloat16
 630 :       8192 : layers.69.ffn_norm.weight           : [8192]          : torch.bfloat16
 631 :    8388608 : layers.70.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 632 :    1048576 : layers.70.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 633 :    1048576 : layers.70.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 634 :    8388608 : layers.70.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 635 :   29360128 : layers.70.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 636 :   29360128 : layers.70.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 637 :   29360128 : layers.70.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 638 :       8192 : layers.70.attention_norm.weight     : [8192]          : torch.bfloat16
 639 :       8192 : layers.70.ffn_norm.weight           : [8192]          : torch.bfloat16
 640 :    8388608 : layers.71.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 641 :    1048576 : layers.71.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 642 :    1048576 : layers.71.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 643 :    8388608 : layers.71.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 644 :   29360128 : layers.71.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 645 :   29360128 : layers.71.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 646 :   29360128 : layers.71.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 647 :       8192 : layers.71.attention_norm.weight     : [8192]          : torch.bfloat16
 648 :       8192 : layers.71.ffn_norm.weight           : [8192]          : torch.bfloat16
 649 :    8388608 : layers.72.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 650 :    1048576 : layers.72.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 651 :    1048576 : layers.72.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 652 :    8388608 : layers.72.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 653 :   29360128 : layers.72.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 654 :   29360128 : layers.72.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 655 :   29360128 : layers.72.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 656 :       8192 : layers.72.attention_norm.weight     : [8192]          : torch.bfloat16
 657 :       8192 : layers.72.ffn_norm.weight           : [8192]          : torch.bfloat16
 658 :    8388608 : layers.73.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 659 :    1048576 : layers.73.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 660 :    1048576 : layers.73.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 661 :    8388608 : layers.73.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 662 :   29360128 : layers.73.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 663 :   29360128 : layers.73.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 664 :   29360128 : layers.73.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 665 :       8192 : layers.73.attention_norm.weight     : [8192]          : torch.bfloat16
 666 :       8192 : layers.73.ffn_norm.weight           : [8192]          : torch.bfloat16
 667 :    8388608 : layers.74.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 668 :    1048576 : layers.74.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 669 :    1048576 : layers.74.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 670 :    8388608 : layers.74.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 671 :   29360128 : layers.74.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 672 :   29360128 : layers.74.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 673 :   29360128 : layers.74.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 674 :       8192 : layers.74.attention_norm.weight     : [8192]          : torch.bfloat16
 675 :       8192 : layers.74.ffn_norm.weight           : [8192]          : torch.bfloat16
 676 :    8388608 : layers.75.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 677 :    1048576 : layers.75.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 678 :    1048576 : layers.75.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 679 :    8388608 : layers.75.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 680 :   29360128 : layers.75.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 681 :   29360128 : layers.75.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 682 :   29360128 : layers.75.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 683 :       8192 : layers.75.attention_norm.weight     : [8192]          : torch.bfloat16
 684 :       8192 : layers.75.ffn_norm.weight           : [8192]          : torch.bfloat16
 685 :    8388608 : layers.76.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 686 :    1048576 : layers.76.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 687 :    1048576 : layers.76.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 688 :    8388608 : layers.76.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 689 :   29360128 : layers.76.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 690 :   29360128 : layers.76.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 691 :   29360128 : layers.76.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 692 :       8192 : layers.76.attention_norm.weight     : [8192]          : torch.bfloat16
 693 :       8192 : layers.76.ffn_norm.weight           : [8192]          : torch.bfloat16
 694 :    8388608 : layers.77.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 695 :    1048576 : layers.77.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 696 :    1048576 : layers.77.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 697 :    8388608 : layers.77.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 698 :   29360128 : layers.77.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 699 :   29360128 : layers.77.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 700 :   29360128 : layers.77.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 701 :       8192 : layers.77.attention_norm.weight     : [8192]          : torch.bfloat16
 702 :       8192 : layers.77.ffn_norm.weight           : [8192]          : torch.bfloat16
 703 :    8388608 : layers.78.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 704 :    1048576 : layers.78.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 705 :    1048576 : layers.78.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 706 :    8388608 : layers.78.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 707 :   29360128 : layers.78.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 708 :   29360128 : layers.78.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 709 :   29360128 : layers.78.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 710 :       8192 : layers.78.attention_norm.weight     : [8192]          : torch.bfloat16
 711 :       8192 : layers.78.ffn_norm.weight           : [8192]          : torch.bfloat16
 712 :    8388608 : layers.79.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 713 :    1048576 : layers.79.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 714 :    1048576 : layers.79.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 715 :    8388608 : layers.79.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 716 :   29360128 : layers.79.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 717 :   29360128 : layers.79.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 718 :   29360128 : layers.79.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 719 :       8192 : layers.79.attention_norm.weight     : [8192]          : torch.bfloat16
 720 :       8192 : layers.79.ffn_norm.weight           : [8192]          : torch.bfloat16
 721 :       8192 : norm.weight                         : [8192]          : torch.bfloat16
 722 :  131334144 : output.weight                       : [16032, 8192]   : torch.bfloat16

-----------------------------------------------------------------------------
8820367360 params in total.
17640734720 bytes in total.
5.89 sec, 5.89 sec, 2858.18 MB/s
-----------------------------------------------------------------------------

[2/8]: Loading original/consolidated.01.pth
   0 :  131334144 : tok_embeddings.weight               : [16032, 8192]   : torch.bfloat16
   1 :    8388608 : layers.0.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
   2 :    1048576 : layers.0.attention.wk.weight        : [128, 8192]     : torch.bfloat16
   3 :    1048576 : layers.0.attention.wv.weight        : [128, 8192]     : torch.bfloat16
   4 :    8388608 : layers.0.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
   5 :   29360128 : layers.0.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
   6 :   29360128 : layers.0.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
   7 :   29360128 : layers.0.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
   8 :       8192 : layers.0.attention_norm.weight      : [8192]          : torch.bfloat16
   9 :       8192 : layers.0.ffn_norm.weight            : [8192]          : torch.bfloat16
  10 :    8388608 : layers.1.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  11 :    1048576 : layers.1.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  12 :    1048576 : layers.1.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  13 :    8388608 : layers.1.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  14 :   29360128 : layers.1.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  15 :   29360128 : layers.1.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  16 :   29360128 : layers.1.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  17 :       8192 : layers.1.attention_norm.weight      : [8192]          : torch.bfloat16
  18 :       8192 : layers.1.ffn_norm.weight            : [8192]          : torch.bfloat16
  19 :    8388608 : layers.2.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  20 :    1048576 : layers.2.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  21 :    1048576 : layers.2.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  22 :    8388608 : layers.2.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  23 :   29360128 : layers.2.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  24 :   29360128 : layers.2.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  25 :   29360128 : layers.2.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  26 :       8192 : layers.2.attention_norm.weight      : [8192]          : torch.bfloat16
  27 :       8192 : layers.2.ffn_norm.weight            : [8192]          : torch.bfloat16
  28 :    8388608 : layers.3.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  29 :    1048576 : layers.3.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  30 :    1048576 : layers.3.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  31 :    8388608 : layers.3.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  32 :   29360128 : layers.3.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  33 :   29360128 : layers.3.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  34 :   29360128 : layers.3.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  35 :       8192 : layers.3.attention_norm.weight      : [8192]          : torch.bfloat16
  36 :       8192 : layers.3.ffn_norm.weight            : [8192]          : torch.bfloat16
  37 :    8388608 : layers.4.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  38 :    1048576 : layers.4.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  39 :    1048576 : layers.4.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  40 :    8388608 : layers.4.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  41 :   29360128 : layers.4.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  42 :   29360128 : layers.4.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  43 :   29360128 : layers.4.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  44 :       8192 : layers.4.attention_norm.weight      : [8192]          : torch.bfloat16
  45 :       8192 : layers.4.ffn_norm.weight            : [8192]          : torch.bfloat16
  46 :    8388608 : layers.5.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  47 :    1048576 : layers.5.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  48 :    1048576 : layers.5.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  49 :    8388608 : layers.5.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  50 :   29360128 : layers.5.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  51 :   29360128 : layers.5.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  52 :   29360128 : layers.5.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  53 :       8192 : layers.5.attention_norm.weight      : [8192]          : torch.bfloat16
  54 :       8192 : layers.5.ffn_norm.weight            : [8192]          : torch.bfloat16
  55 :    8388608 : layers.6.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  56 :    1048576 : layers.6.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  57 :    1048576 : layers.6.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  58 :    8388608 : layers.6.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  59 :   29360128 : layers.6.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  60 :   29360128 : layers.6.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  61 :   29360128 : layers.6.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  62 :       8192 : layers.6.attention_norm.weight      : [8192]          : torch.bfloat16
  63 :       8192 : layers.6.ffn_norm.weight            : [8192]          : torch.bfloat16
  64 :    8388608 : layers.7.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  65 :    1048576 : layers.7.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  66 :    1048576 : layers.7.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  67 :    8388608 : layers.7.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  68 :   29360128 : layers.7.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  69 :   29360128 : layers.7.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  70 :   29360128 : layers.7.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  71 :       8192 : layers.7.attention_norm.weight      : [8192]          : torch.bfloat16
  72 :       8192 : layers.7.ffn_norm.weight            : [8192]          : torch.bfloat16
  73 :    8388608 : layers.8.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  74 :    1048576 : layers.8.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  75 :    1048576 : layers.8.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  76 :    8388608 : layers.8.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  77 :   29360128 : layers.8.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  78 :   29360128 : layers.8.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  79 :   29360128 : layers.8.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  80 :       8192 : layers.8.attention_norm.weight      : [8192]          : torch.bfloat16
  81 :       8192 : layers.8.ffn_norm.weight            : [8192]          : torch.bfloat16
  82 :    8388608 : layers.9.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  83 :    1048576 : layers.9.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  84 :    1048576 : layers.9.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  85 :    8388608 : layers.9.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  86 :   29360128 : layers.9.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  87 :   29360128 : layers.9.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  88 :   29360128 : layers.9.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  89 :       8192 : layers.9.attention_norm.weight      : [8192]          : torch.bfloat16
  90 :       8192 : layers.9.ffn_norm.weight            : [8192]          : torch.bfloat16
  91 :    8388608 : layers.10.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
  92 :    1048576 : layers.10.attention.wk.weight       : [128, 8192]     : torch.bfloat16
  93 :    1048576 : layers.10.attention.wv.weight       : [128, 8192]     : torch.bfloat16
  94 :    8388608 : layers.10.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
  95 :   29360128 : layers.10.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
  96 :   29360128 : layers.10.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
  97 :   29360128 : layers.10.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
  98 :       8192 : layers.10.attention_norm.weight     : [8192]          : torch.bfloat16
  99 :       8192 : layers.10.ffn_norm.weight           : [8192]          : torch.bfloat16
 100 :    8388608 : layers.11.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 101 :    1048576 : layers.11.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 102 :    1048576 : layers.11.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 103 :    8388608 : layers.11.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 104 :   29360128 : layers.11.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 105 :   29360128 : layers.11.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 106 :   29360128 : layers.11.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 107 :       8192 : layers.11.attention_norm.weight     : [8192]          : torch.bfloat16
 108 :       8192 : layers.11.ffn_norm.weight           : [8192]          : torch.bfloat16
 109 :    8388608 : layers.12.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 110 :    1048576 : layers.12.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 111 :    1048576 : layers.12.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 112 :    8388608 : layers.12.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 113 :   29360128 : layers.12.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 114 :   29360128 : layers.12.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 115 :   29360128 : layers.12.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 116 :       8192 : layers.12.attention_norm.weight     : [8192]          : torch.bfloat16
 117 :       8192 : layers.12.ffn_norm.weight           : [8192]          : torch.bfloat16
 118 :    8388608 : layers.13.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 119 :    1048576 : layers.13.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 120 :    1048576 : layers.13.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 121 :    8388608 : layers.13.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 122 :   29360128 : layers.13.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 123 :   29360128 : layers.13.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 124 :   29360128 : layers.13.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 125 :       8192 : layers.13.attention_norm.weight     : [8192]          : torch.bfloat16
 126 :       8192 : layers.13.ffn_norm.weight           : [8192]          : torch.bfloat16
 127 :    8388608 : layers.14.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 128 :    1048576 : layers.14.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 129 :    1048576 : layers.14.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 130 :    8388608 : layers.14.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 131 :   29360128 : layers.14.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 132 :   29360128 : layers.14.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 133 :   29360128 : layers.14.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 134 :       8192 : layers.14.attention_norm.weight     : [8192]          : torch.bfloat16
 135 :       8192 : layers.14.ffn_norm.weight           : [8192]          : torch.bfloat16
 136 :    8388608 : layers.15.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 137 :    1048576 : layers.15.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 138 :    1048576 : layers.15.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 139 :    8388608 : layers.15.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 140 :   29360128 : layers.15.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 141 :   29360128 : layers.15.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 142 :   29360128 : layers.15.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 143 :       8192 : layers.15.attention_norm.weight     : [8192]          : torch.bfloat16
 144 :       8192 : layers.15.ffn_norm.weight           : [8192]          : torch.bfloat16
 145 :    8388608 : layers.16.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 146 :    1048576 : layers.16.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 147 :    1048576 : layers.16.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 148 :    8388608 : layers.16.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 149 :   29360128 : layers.16.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 150 :   29360128 : layers.16.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 151 :   29360128 : layers.16.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 152 :       8192 : layers.16.attention_norm.weight     : [8192]          : torch.bfloat16
 153 :       8192 : layers.16.ffn_norm.weight           : [8192]          : torch.bfloat16
 154 :    8388608 : layers.17.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 155 :    1048576 : layers.17.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 156 :    1048576 : layers.17.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 157 :    8388608 : layers.17.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 158 :   29360128 : layers.17.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 159 :   29360128 : layers.17.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 160 :   29360128 : layers.17.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 161 :       8192 : layers.17.attention_norm.weight     : [8192]          : torch.bfloat16
 162 :       8192 : layers.17.ffn_norm.weight           : [8192]          : torch.bfloat16
 163 :    8388608 : layers.18.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 164 :    1048576 : layers.18.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 165 :    1048576 : layers.18.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 166 :    8388608 : layers.18.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 167 :   29360128 : layers.18.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 168 :   29360128 : layers.18.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 169 :   29360128 : layers.18.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 170 :       8192 : layers.18.attention_norm.weight     : [8192]          : torch.bfloat16
 171 :       8192 : layers.18.ffn_norm.weight           : [8192]          : torch.bfloat16
 172 :    8388608 : layers.19.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 173 :    1048576 : layers.19.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 174 :    1048576 : layers.19.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 175 :    8388608 : layers.19.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 176 :   29360128 : layers.19.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 177 :   29360128 : layers.19.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 178 :   29360128 : layers.19.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 179 :       8192 : layers.19.attention_norm.weight     : [8192]          : torch.bfloat16
 180 :       8192 : layers.19.ffn_norm.weight           : [8192]          : torch.bfloat16
 181 :    8388608 : layers.20.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 182 :    1048576 : layers.20.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 183 :    1048576 : layers.20.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 184 :    8388608 : layers.20.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 185 :   29360128 : layers.20.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 186 :   29360128 : layers.20.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 187 :   29360128 : layers.20.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 188 :       8192 : layers.20.attention_norm.weight     : [8192]          : torch.bfloat16
 189 :       8192 : layers.20.ffn_norm.weight           : [8192]          : torch.bfloat16
 190 :    8388608 : layers.21.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 191 :    1048576 : layers.21.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 192 :    1048576 : layers.21.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 193 :    8388608 : layers.21.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 194 :   29360128 : layers.21.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 195 :   29360128 : layers.21.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 196 :   29360128 : layers.21.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 197 :       8192 : layers.21.attention_norm.weight     : [8192]          : torch.bfloat16
 198 :       8192 : layers.21.ffn_norm.weight           : [8192]          : torch.bfloat16
 199 :    8388608 : layers.22.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 200 :    1048576 : layers.22.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 201 :    1048576 : layers.22.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 202 :    8388608 : layers.22.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 203 :   29360128 : layers.22.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 204 :   29360128 : layers.22.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 205 :   29360128 : layers.22.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 206 :       8192 : layers.22.attention_norm.weight     : [8192]          : torch.bfloat16
 207 :       8192 : layers.22.ffn_norm.weight           : [8192]          : torch.bfloat16
 208 :    8388608 : layers.23.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 209 :    1048576 : layers.23.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 210 :    1048576 : layers.23.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 211 :    8388608 : layers.23.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 212 :   29360128 : layers.23.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 213 :   29360128 : layers.23.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 214 :   29360128 : layers.23.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 215 :       8192 : layers.23.attention_norm.weight     : [8192]          : torch.bfloat16
 216 :       8192 : layers.23.ffn_norm.weight           : [8192]          : torch.bfloat16
 217 :    8388608 : layers.24.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 218 :    1048576 : layers.24.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 219 :    1048576 : layers.24.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 220 :    8388608 : layers.24.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 221 :   29360128 : layers.24.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 222 :   29360128 : layers.24.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 223 :   29360128 : layers.24.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 224 :       8192 : layers.24.attention_norm.weight     : [8192]          : torch.bfloat16
 225 :       8192 : layers.24.ffn_norm.weight           : [8192]          : torch.bfloat16
 226 :    8388608 : layers.25.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 227 :    1048576 : layers.25.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 228 :    1048576 : layers.25.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 229 :    8388608 : layers.25.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 230 :   29360128 : layers.25.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 231 :   29360128 : layers.25.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 232 :   29360128 : layers.25.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 233 :       8192 : layers.25.attention_norm.weight     : [8192]          : torch.bfloat16
 234 :       8192 : layers.25.ffn_norm.weight           : [8192]          : torch.bfloat16
 235 :    8388608 : layers.26.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 236 :    1048576 : layers.26.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 237 :    1048576 : layers.26.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 238 :    8388608 : layers.26.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 239 :   29360128 : layers.26.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 240 :   29360128 : layers.26.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 241 :   29360128 : layers.26.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 242 :       8192 : layers.26.attention_norm.weight     : [8192]          : torch.bfloat16
 243 :       8192 : layers.26.ffn_norm.weight           : [8192]          : torch.bfloat16
 244 :    8388608 : layers.27.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 245 :    1048576 : layers.27.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 246 :    1048576 : layers.27.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 247 :    8388608 : layers.27.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 248 :   29360128 : layers.27.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 249 :   29360128 : layers.27.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 250 :   29360128 : layers.27.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 251 :       8192 : layers.27.attention_norm.weight     : [8192]          : torch.bfloat16
 252 :       8192 : layers.27.ffn_norm.weight           : [8192]          : torch.bfloat16
 253 :    8388608 : layers.28.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 254 :    1048576 : layers.28.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 255 :    1048576 : layers.28.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 256 :    8388608 : layers.28.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 257 :   29360128 : layers.28.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 258 :   29360128 : layers.28.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 259 :   29360128 : layers.28.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 260 :       8192 : layers.28.attention_norm.weight     : [8192]          : torch.bfloat16
 261 :       8192 : layers.28.ffn_norm.weight           : [8192]          : torch.bfloat16
 262 :    8388608 : layers.29.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 263 :    1048576 : layers.29.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 264 :    1048576 : layers.29.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 265 :    8388608 : layers.29.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 266 :   29360128 : layers.29.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 267 :   29360128 : layers.29.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 268 :   29360128 : layers.29.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 269 :       8192 : layers.29.attention_norm.weight     : [8192]          : torch.bfloat16
 270 :       8192 : layers.29.ffn_norm.weight           : [8192]          : torch.bfloat16
 271 :    8388608 : layers.30.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 272 :    1048576 : layers.30.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 273 :    1048576 : layers.30.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 274 :    8388608 : layers.30.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 275 :   29360128 : layers.30.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 276 :   29360128 : layers.30.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 277 :   29360128 : layers.30.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 278 :       8192 : layers.30.attention_norm.weight     : [8192]          : torch.bfloat16
 279 :       8192 : layers.30.ffn_norm.weight           : [8192]          : torch.bfloat16
 280 :    8388608 : layers.31.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 281 :    1048576 : layers.31.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 282 :    1048576 : layers.31.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 283 :    8388608 : layers.31.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 284 :   29360128 : layers.31.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 285 :   29360128 : layers.31.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 286 :   29360128 : layers.31.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 287 :       8192 : layers.31.attention_norm.weight     : [8192]          : torch.bfloat16
 288 :       8192 : layers.31.ffn_norm.weight           : [8192]          : torch.bfloat16
 289 :    8388608 : layers.32.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 290 :    1048576 : layers.32.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 291 :    1048576 : layers.32.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 292 :    8388608 : layers.32.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 293 :   29360128 : layers.32.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 294 :   29360128 : layers.32.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 295 :   29360128 : layers.32.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 296 :       8192 : layers.32.attention_norm.weight     : [8192]          : torch.bfloat16
 297 :       8192 : layers.32.ffn_norm.weight           : [8192]          : torch.bfloat16
 298 :    8388608 : layers.33.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 299 :    1048576 : layers.33.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 300 :    1048576 : layers.33.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 301 :    8388608 : layers.33.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 302 :   29360128 : layers.33.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 303 :   29360128 : layers.33.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 304 :   29360128 : layers.33.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 305 :       8192 : layers.33.attention_norm.weight     : [8192]          : torch.bfloat16
 306 :       8192 : layers.33.ffn_norm.weight           : [8192]          : torch.bfloat16
 307 :    8388608 : layers.34.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 308 :    1048576 : layers.34.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 309 :    1048576 : layers.34.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 310 :    8388608 : layers.34.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 311 :   29360128 : layers.34.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 312 :   29360128 : layers.34.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 313 :   29360128 : layers.34.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 314 :       8192 : layers.34.attention_norm.weight     : [8192]          : torch.bfloat16
 315 :       8192 : layers.34.ffn_norm.weight           : [8192]          : torch.bfloat16
 316 :    8388608 : layers.35.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 317 :    1048576 : layers.35.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 318 :    1048576 : layers.35.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 319 :    8388608 : layers.35.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 320 :   29360128 : layers.35.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 321 :   29360128 : layers.35.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 322 :   29360128 : layers.35.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 323 :       8192 : layers.35.attention_norm.weight     : [8192]          : torch.bfloat16
 324 :       8192 : layers.35.ffn_norm.weight           : [8192]          : torch.bfloat16
 325 :    8388608 : layers.36.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 326 :    1048576 : layers.36.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 327 :    1048576 : layers.36.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 328 :    8388608 : layers.36.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 329 :   29360128 : layers.36.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 330 :   29360128 : layers.36.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 331 :   29360128 : layers.36.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 332 :       8192 : layers.36.attention_norm.weight     : [8192]          : torch.bfloat16
 333 :       8192 : layers.36.ffn_norm.weight           : [8192]          : torch.bfloat16
 334 :    8388608 : layers.37.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 335 :    1048576 : layers.37.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 336 :    1048576 : layers.37.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 337 :    8388608 : layers.37.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 338 :   29360128 : layers.37.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 339 :   29360128 : layers.37.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 340 :   29360128 : layers.37.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 341 :       8192 : layers.37.attention_norm.weight     : [8192]          : torch.bfloat16
 342 :       8192 : layers.37.ffn_norm.weight           : [8192]          : torch.bfloat16
 343 :    8388608 : layers.38.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 344 :    1048576 : layers.38.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 345 :    1048576 : layers.38.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 346 :    8388608 : layers.38.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 347 :   29360128 : layers.38.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 348 :   29360128 : layers.38.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 349 :   29360128 : layers.38.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 350 :       8192 : layers.38.attention_norm.weight     : [8192]          : torch.bfloat16
 351 :       8192 : layers.38.ffn_norm.weight           : [8192]          : torch.bfloat16
 352 :    8388608 : layers.39.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 353 :    1048576 : layers.39.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 354 :    1048576 : layers.39.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 355 :    8388608 : layers.39.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 356 :   29360128 : layers.39.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 357 :   29360128 : layers.39.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 358 :   29360128 : layers.39.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 359 :       8192 : layers.39.attention_norm.weight     : [8192]          : torch.bfloat16
 360 :       8192 : layers.39.ffn_norm.weight           : [8192]          : torch.bfloat16
 361 :    8388608 : layers.40.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 362 :    1048576 : layers.40.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 363 :    1048576 : layers.40.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 364 :    8388608 : layers.40.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 365 :   29360128 : layers.40.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 366 :   29360128 : layers.40.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 367 :   29360128 : layers.40.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 368 :       8192 : layers.40.attention_norm.weight     : [8192]          : torch.bfloat16
 369 :       8192 : layers.40.ffn_norm.weight           : [8192]          : torch.bfloat16
 370 :    8388608 : layers.41.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 371 :    1048576 : layers.41.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 372 :    1048576 : layers.41.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 373 :    8388608 : layers.41.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 374 :   29360128 : layers.41.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 375 :   29360128 : layers.41.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 376 :   29360128 : layers.41.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 377 :       8192 : layers.41.attention_norm.weight     : [8192]          : torch.bfloat16
 378 :       8192 : layers.41.ffn_norm.weight           : [8192]          : torch.bfloat16
 379 :    8388608 : layers.42.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 380 :    1048576 : layers.42.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 381 :    1048576 : layers.42.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 382 :    8388608 : layers.42.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 383 :   29360128 : layers.42.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 384 :   29360128 : layers.42.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 385 :   29360128 : layers.42.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 386 :       8192 : layers.42.attention_norm.weight     : [8192]          : torch.bfloat16
 387 :       8192 : layers.42.ffn_norm.weight           : [8192]          : torch.bfloat16
 388 :    8388608 : layers.43.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 389 :    1048576 : layers.43.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 390 :    1048576 : layers.43.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 391 :    8388608 : layers.43.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 392 :   29360128 : layers.43.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 393 :   29360128 : layers.43.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 394 :   29360128 : layers.43.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 395 :       8192 : layers.43.attention_norm.weight     : [8192]          : torch.bfloat16
 396 :       8192 : layers.43.ffn_norm.weight           : [8192]          : torch.bfloat16
 397 :    8388608 : layers.44.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 398 :    1048576 : layers.44.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 399 :    1048576 : layers.44.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 400 :    8388608 : layers.44.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 401 :   29360128 : layers.44.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 402 :   29360128 : layers.44.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 403 :   29360128 : layers.44.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 404 :       8192 : layers.44.attention_norm.weight     : [8192]          : torch.bfloat16
 405 :       8192 : layers.44.ffn_norm.weight           : [8192]          : torch.bfloat16
 406 :    8388608 : layers.45.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 407 :    1048576 : layers.45.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 408 :    1048576 : layers.45.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 409 :    8388608 : layers.45.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 410 :   29360128 : layers.45.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 411 :   29360128 : layers.45.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 412 :   29360128 : layers.45.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 413 :       8192 : layers.45.attention_norm.weight     : [8192]          : torch.bfloat16
 414 :       8192 : layers.45.ffn_norm.weight           : [8192]          : torch.bfloat16
 415 :    8388608 : layers.46.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 416 :    1048576 : layers.46.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 417 :    1048576 : layers.46.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 418 :    8388608 : layers.46.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 419 :   29360128 : layers.46.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 420 :   29360128 : layers.46.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 421 :   29360128 : layers.46.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 422 :       8192 : layers.46.attention_norm.weight     : [8192]          : torch.bfloat16
 423 :       8192 : layers.46.ffn_norm.weight           : [8192]          : torch.bfloat16
 424 :    8388608 : layers.47.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 425 :    1048576 : layers.47.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 426 :    1048576 : layers.47.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 427 :    8388608 : layers.47.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 428 :   29360128 : layers.47.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 429 :   29360128 : layers.47.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 430 :   29360128 : layers.47.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 431 :       8192 : layers.47.attention_norm.weight     : [8192]          : torch.bfloat16
 432 :       8192 : layers.47.ffn_norm.weight           : [8192]          : torch.bfloat16
 433 :    8388608 : layers.48.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 434 :    1048576 : layers.48.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 435 :    1048576 : layers.48.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 436 :    8388608 : layers.48.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 437 :   29360128 : layers.48.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 438 :   29360128 : layers.48.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 439 :   29360128 : layers.48.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 440 :       8192 : layers.48.attention_norm.weight     : [8192]          : torch.bfloat16
 441 :       8192 : layers.48.ffn_norm.weight           : [8192]          : torch.bfloat16
 442 :    8388608 : layers.49.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 443 :    1048576 : layers.49.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 444 :    1048576 : layers.49.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 445 :    8388608 : layers.49.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 446 :   29360128 : layers.49.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 447 :   29360128 : layers.49.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 448 :   29360128 : layers.49.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 449 :       8192 : layers.49.attention_norm.weight     : [8192]          : torch.bfloat16
 450 :       8192 : layers.49.ffn_norm.weight           : [8192]          : torch.bfloat16
 451 :    8388608 : layers.50.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 452 :    1048576 : layers.50.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 453 :    1048576 : layers.50.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 454 :    8388608 : layers.50.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 455 :   29360128 : layers.50.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 456 :   29360128 : layers.50.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 457 :   29360128 : layers.50.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 458 :       8192 : layers.50.attention_norm.weight     : [8192]          : torch.bfloat16
 459 :       8192 : layers.50.ffn_norm.weight           : [8192]          : torch.bfloat16
 460 :    8388608 : layers.51.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 461 :    1048576 : layers.51.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 462 :    1048576 : layers.51.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 463 :    8388608 : layers.51.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 464 :   29360128 : layers.51.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 465 :   29360128 : layers.51.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 466 :   29360128 : layers.51.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 467 :       8192 : layers.51.attention_norm.weight     : [8192]          : torch.bfloat16
 468 :       8192 : layers.51.ffn_norm.weight           : [8192]          : torch.bfloat16
 469 :    8388608 : layers.52.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 470 :    1048576 : layers.52.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 471 :    1048576 : layers.52.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 472 :    8388608 : layers.52.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 473 :   29360128 : layers.52.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 474 :   29360128 : layers.52.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 475 :   29360128 : layers.52.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 476 :       8192 : layers.52.attention_norm.weight     : [8192]          : torch.bfloat16
 477 :       8192 : layers.52.ffn_norm.weight           : [8192]          : torch.bfloat16
 478 :    8388608 : layers.53.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 479 :    1048576 : layers.53.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 480 :    1048576 : layers.53.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 481 :    8388608 : layers.53.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 482 :   29360128 : layers.53.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 483 :   29360128 : layers.53.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 484 :   29360128 : layers.53.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 485 :       8192 : layers.53.attention_norm.weight     : [8192]          : torch.bfloat16
 486 :       8192 : layers.53.ffn_norm.weight           : [8192]          : torch.bfloat16
 487 :    8388608 : layers.54.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 488 :    1048576 : layers.54.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 489 :    1048576 : layers.54.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 490 :    8388608 : layers.54.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 491 :   29360128 : layers.54.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 492 :   29360128 : layers.54.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 493 :   29360128 : layers.54.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 494 :       8192 : layers.54.attention_norm.weight     : [8192]          : torch.bfloat16
 495 :       8192 : layers.54.ffn_norm.weight           : [8192]          : torch.bfloat16
 496 :    8388608 : layers.55.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 497 :    1048576 : layers.55.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 498 :    1048576 : layers.55.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 499 :    8388608 : layers.55.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 500 :   29360128 : layers.55.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 501 :   29360128 : layers.55.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 502 :   29360128 : layers.55.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 503 :       8192 : layers.55.attention_norm.weight     : [8192]          : torch.bfloat16
 504 :       8192 : layers.55.ffn_norm.weight           : [8192]          : torch.bfloat16
 505 :    8388608 : layers.56.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 506 :    1048576 : layers.56.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 507 :    1048576 : layers.56.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 508 :    8388608 : layers.56.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 509 :   29360128 : layers.56.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 510 :   29360128 : layers.56.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 511 :   29360128 : layers.56.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 512 :       8192 : layers.56.attention_norm.weight     : [8192]          : torch.bfloat16
 513 :       8192 : layers.56.ffn_norm.weight           : [8192]          : torch.bfloat16
 514 :    8388608 : layers.57.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 515 :    1048576 : layers.57.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 516 :    1048576 : layers.57.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 517 :    8388608 : layers.57.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 518 :   29360128 : layers.57.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 519 :   29360128 : layers.57.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 520 :   29360128 : layers.57.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 521 :       8192 : layers.57.attention_norm.weight     : [8192]          : torch.bfloat16
 522 :       8192 : layers.57.ffn_norm.weight           : [8192]          : torch.bfloat16
 523 :    8388608 : layers.58.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 524 :    1048576 : layers.58.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 525 :    1048576 : layers.58.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 526 :    8388608 : layers.58.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 527 :   29360128 : layers.58.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 528 :   29360128 : layers.58.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 529 :   29360128 : layers.58.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 530 :       8192 : layers.58.attention_norm.weight     : [8192]          : torch.bfloat16
 531 :       8192 : layers.58.ffn_norm.weight           : [8192]          : torch.bfloat16
 532 :    8388608 : layers.59.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 533 :    1048576 : layers.59.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 534 :    1048576 : layers.59.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 535 :    8388608 : layers.59.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 536 :   29360128 : layers.59.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 537 :   29360128 : layers.59.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 538 :   29360128 : layers.59.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 539 :       8192 : layers.59.attention_norm.weight     : [8192]          : torch.bfloat16
 540 :       8192 : layers.59.ffn_norm.weight           : [8192]          : torch.bfloat16
 541 :    8388608 : layers.60.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 542 :    1048576 : layers.60.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 543 :    1048576 : layers.60.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 544 :    8388608 : layers.60.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 545 :   29360128 : layers.60.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 546 :   29360128 : layers.60.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 547 :   29360128 : layers.60.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 548 :       8192 : layers.60.attention_norm.weight     : [8192]          : torch.bfloat16
 549 :       8192 : layers.60.ffn_norm.weight           : [8192]          : torch.bfloat16
 550 :    8388608 : layers.61.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 551 :    1048576 : layers.61.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 552 :    1048576 : layers.61.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 553 :    8388608 : layers.61.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 554 :   29360128 : layers.61.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 555 :   29360128 : layers.61.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 556 :   29360128 : layers.61.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 557 :       8192 : layers.61.attention_norm.weight     : [8192]          : torch.bfloat16
 558 :       8192 : layers.61.ffn_norm.weight           : [8192]          : torch.bfloat16
 559 :    8388608 : layers.62.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 560 :    1048576 : layers.62.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 561 :    1048576 : layers.62.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 562 :    8388608 : layers.62.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 563 :   29360128 : layers.62.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 564 :   29360128 : layers.62.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 565 :   29360128 : layers.62.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 566 :       8192 : layers.62.attention_norm.weight     : [8192]          : torch.bfloat16
 567 :       8192 : layers.62.ffn_norm.weight           : [8192]          : torch.bfloat16
 568 :    8388608 : layers.63.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 569 :    1048576 : layers.63.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 570 :    1048576 : layers.63.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 571 :    8388608 : layers.63.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 572 :   29360128 : layers.63.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 573 :   29360128 : layers.63.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 574 :   29360128 : layers.63.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 575 :       8192 : layers.63.attention_norm.weight     : [8192]          : torch.bfloat16
 576 :       8192 : layers.63.ffn_norm.weight           : [8192]          : torch.bfloat16
 577 :    8388608 : layers.64.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 578 :    1048576 : layers.64.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 579 :    1048576 : layers.64.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 580 :    8388608 : layers.64.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 581 :   29360128 : layers.64.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 582 :   29360128 : layers.64.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 583 :   29360128 : layers.64.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 584 :       8192 : layers.64.attention_norm.weight     : [8192]          : torch.bfloat16
 585 :       8192 : layers.64.ffn_norm.weight           : [8192]          : torch.bfloat16
 586 :    8388608 : layers.65.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 587 :    1048576 : layers.65.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 588 :    1048576 : layers.65.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 589 :    8388608 : layers.65.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 590 :   29360128 : layers.65.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 591 :   29360128 : layers.65.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 592 :   29360128 : layers.65.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 593 :       8192 : layers.65.attention_norm.weight     : [8192]          : torch.bfloat16
 594 :       8192 : layers.65.ffn_norm.weight           : [8192]          : torch.bfloat16
 595 :    8388608 : layers.66.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 596 :    1048576 : layers.66.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 597 :    1048576 : layers.66.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 598 :    8388608 : layers.66.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 599 :   29360128 : layers.66.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 600 :   29360128 : layers.66.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 601 :   29360128 : layers.66.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 602 :       8192 : layers.66.attention_norm.weight     : [8192]          : torch.bfloat16
 603 :       8192 : layers.66.ffn_norm.weight           : [8192]          : torch.bfloat16
 604 :    8388608 : layers.67.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 605 :    1048576 : layers.67.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 606 :    1048576 : layers.67.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 607 :    8388608 : layers.67.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 608 :   29360128 : layers.67.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 609 :   29360128 : layers.67.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 610 :   29360128 : layers.67.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 611 :       8192 : layers.67.attention_norm.weight     : [8192]          : torch.bfloat16
 612 :       8192 : layers.67.ffn_norm.weight           : [8192]          : torch.bfloat16
 613 :    8388608 : layers.68.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 614 :    1048576 : layers.68.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 615 :    1048576 : layers.68.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 616 :    8388608 : layers.68.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 617 :   29360128 : layers.68.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 618 :   29360128 : layers.68.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 619 :   29360128 : layers.68.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 620 :       8192 : layers.68.attention_norm.weight     : [8192]          : torch.bfloat16
 621 :       8192 : layers.68.ffn_norm.weight           : [8192]          : torch.bfloat16
 622 :    8388608 : layers.69.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 623 :    1048576 : layers.69.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 624 :    1048576 : layers.69.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 625 :    8388608 : layers.69.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 626 :   29360128 : layers.69.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 627 :   29360128 : layers.69.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 628 :   29360128 : layers.69.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 629 :       8192 : layers.69.attention_norm.weight     : [8192]          : torch.bfloat16
 630 :       8192 : layers.69.ffn_norm.weight           : [8192]          : torch.bfloat16
 631 :    8388608 : layers.70.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 632 :    1048576 : layers.70.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 633 :    1048576 : layers.70.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 634 :    8388608 : layers.70.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 635 :   29360128 : layers.70.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 636 :   29360128 : layers.70.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 637 :   29360128 : layers.70.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 638 :       8192 : layers.70.attention_norm.weight     : [8192]          : torch.bfloat16
 639 :       8192 : layers.70.ffn_norm.weight           : [8192]          : torch.bfloat16
 640 :    8388608 : layers.71.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 641 :    1048576 : layers.71.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 642 :    1048576 : layers.71.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 643 :    8388608 : layers.71.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 644 :   29360128 : layers.71.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 645 :   29360128 : layers.71.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 646 :   29360128 : layers.71.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 647 :       8192 : layers.71.attention_norm.weight     : [8192]          : torch.bfloat16
 648 :       8192 : layers.71.ffn_norm.weight           : [8192]          : torch.bfloat16
 649 :    8388608 : layers.72.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 650 :    1048576 : layers.72.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 651 :    1048576 : layers.72.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 652 :    8388608 : layers.72.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 653 :   29360128 : layers.72.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 654 :   29360128 : layers.72.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 655 :   29360128 : layers.72.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 656 :       8192 : layers.72.attention_norm.weight     : [8192]          : torch.bfloat16
 657 :       8192 : layers.72.ffn_norm.weight           : [8192]          : torch.bfloat16
 658 :    8388608 : layers.73.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 659 :    1048576 : layers.73.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 660 :    1048576 : layers.73.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 661 :    8388608 : layers.73.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 662 :   29360128 : layers.73.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 663 :   29360128 : layers.73.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 664 :   29360128 : layers.73.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 665 :       8192 : layers.73.attention_norm.weight     : [8192]          : torch.bfloat16
 666 :       8192 : layers.73.ffn_norm.weight           : [8192]          : torch.bfloat16
 667 :    8388608 : layers.74.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 668 :    1048576 : layers.74.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 669 :    1048576 : layers.74.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 670 :    8388608 : layers.74.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 671 :   29360128 : layers.74.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 672 :   29360128 : layers.74.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 673 :   29360128 : layers.74.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 674 :       8192 : layers.74.attention_norm.weight     : [8192]          : torch.bfloat16
 675 :       8192 : layers.74.ffn_norm.weight           : [8192]          : torch.bfloat16
 676 :    8388608 : layers.75.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 677 :    1048576 : layers.75.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 678 :    1048576 : layers.75.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 679 :    8388608 : layers.75.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 680 :   29360128 : layers.75.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 681 :   29360128 : layers.75.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 682 :   29360128 : layers.75.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 683 :       8192 : layers.75.attention_norm.weight     : [8192]          : torch.bfloat16
 684 :       8192 : layers.75.ffn_norm.weight           : [8192]          : torch.bfloat16
 685 :    8388608 : layers.76.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 686 :    1048576 : layers.76.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 687 :    1048576 : layers.76.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 688 :    8388608 : layers.76.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 689 :   29360128 : layers.76.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 690 :   29360128 : layers.76.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 691 :   29360128 : layers.76.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 692 :       8192 : layers.76.attention_norm.weight     : [8192]          : torch.bfloat16
 693 :       8192 : layers.76.ffn_norm.weight           : [8192]          : torch.bfloat16
 694 :    8388608 : layers.77.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 695 :    1048576 : layers.77.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 696 :    1048576 : layers.77.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 697 :    8388608 : layers.77.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 698 :   29360128 : layers.77.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 699 :   29360128 : layers.77.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 700 :   29360128 : layers.77.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 701 :       8192 : layers.77.attention_norm.weight     : [8192]          : torch.bfloat16
 702 :       8192 : layers.77.ffn_norm.weight           : [8192]          : torch.bfloat16
 703 :    8388608 : layers.78.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 704 :    1048576 : layers.78.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 705 :    1048576 : layers.78.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 706 :    8388608 : layers.78.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 707 :   29360128 : layers.78.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 708 :   29360128 : layers.78.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 709 :   29360128 : layers.78.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 710 :       8192 : layers.78.attention_norm.weight     : [8192]          : torch.bfloat16
 711 :       8192 : layers.78.ffn_norm.weight           : [8192]          : torch.bfloat16
 712 :    8388608 : layers.79.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 713 :    1048576 : layers.79.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 714 :    1048576 : layers.79.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 715 :    8388608 : layers.79.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 716 :   29360128 : layers.79.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 717 :   29360128 : layers.79.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 718 :   29360128 : layers.79.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 719 :       8192 : layers.79.attention_norm.weight     : [8192]          : torch.bfloat16
 720 :       8192 : layers.79.ffn_norm.weight           : [8192]          : torch.bfloat16
 721 :       8192 : norm.weight                         : [8192]          : torch.bfloat16
 722 :  131334144 : output.weight                       : [16032, 8192]   : torch.bfloat16

-----------------------------------------------------------------------------
17640734720 params in total.
35281469440 bytes in total.
12.15 sec, 6.27 sec, 2768.56 MB/s
-----------------------------------------------------------------------------

[3/8]: Loading original/consolidated.02.pth
   0 :  131334144 : tok_embeddings.weight               : [16032, 8192]   : torch.bfloat16
   1 :    8388608 : layers.0.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
   2 :    1048576 : layers.0.attention.wk.weight        : [128, 8192]     : torch.bfloat16
   3 :    1048576 : layers.0.attention.wv.weight        : [128, 8192]     : torch.bfloat16
   4 :    8388608 : layers.0.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
   5 :   29360128 : layers.0.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
   6 :   29360128 : layers.0.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
   7 :   29360128 : layers.0.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
   8 :       8192 : layers.0.attention_norm.weight      : [8192]          : torch.bfloat16
   9 :       8192 : layers.0.ffn_norm.weight            : [8192]          : torch.bfloat16
  10 :    8388608 : layers.1.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  11 :    1048576 : layers.1.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  12 :    1048576 : layers.1.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  13 :    8388608 : layers.1.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  14 :   29360128 : layers.1.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  15 :   29360128 : layers.1.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  16 :   29360128 : layers.1.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  17 :       8192 : layers.1.attention_norm.weight      : [8192]          : torch.bfloat16
  18 :       8192 : layers.1.ffn_norm.weight            : [8192]          : torch.bfloat16
  19 :    8388608 : layers.2.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  20 :    1048576 : layers.2.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  21 :    1048576 : layers.2.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  22 :    8388608 : layers.2.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  23 :   29360128 : layers.2.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  24 :   29360128 : layers.2.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  25 :   29360128 : layers.2.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  26 :       8192 : layers.2.attention_norm.weight      : [8192]          : torch.bfloat16
  27 :       8192 : layers.2.ffn_norm.weight            : [8192]          : torch.bfloat16
  28 :    8388608 : layers.3.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  29 :    1048576 : layers.3.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  30 :    1048576 : layers.3.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  31 :    8388608 : layers.3.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  32 :   29360128 : layers.3.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  33 :   29360128 : layers.3.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  34 :   29360128 : layers.3.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  35 :       8192 : layers.3.attention_norm.weight      : [8192]          : torch.bfloat16
  36 :       8192 : layers.3.ffn_norm.weight            : [8192]          : torch.bfloat16
  37 :    8388608 : layers.4.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  38 :    1048576 : layers.4.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  39 :    1048576 : layers.4.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  40 :    8388608 : layers.4.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  41 :   29360128 : layers.4.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  42 :   29360128 : layers.4.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  43 :   29360128 : layers.4.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  44 :       8192 : layers.4.attention_norm.weight      : [8192]          : torch.bfloat16
  45 :       8192 : layers.4.ffn_norm.weight            : [8192]          : torch.bfloat16
  46 :    8388608 : layers.5.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  47 :    1048576 : layers.5.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  48 :    1048576 : layers.5.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  49 :    8388608 : layers.5.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  50 :   29360128 : layers.5.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  51 :   29360128 : layers.5.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  52 :   29360128 : layers.5.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  53 :       8192 : layers.5.attention_norm.weight      : [8192]          : torch.bfloat16
  54 :       8192 : layers.5.ffn_norm.weight            : [8192]          : torch.bfloat16
  55 :    8388608 : layers.6.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  56 :    1048576 : layers.6.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  57 :    1048576 : layers.6.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  58 :    8388608 : layers.6.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  59 :   29360128 : layers.6.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  60 :   29360128 : layers.6.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  61 :   29360128 : layers.6.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  62 :       8192 : layers.6.attention_norm.weight      : [8192]          : torch.bfloat16
  63 :       8192 : layers.6.ffn_norm.weight            : [8192]          : torch.bfloat16
  64 :    8388608 : layers.7.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  65 :    1048576 : layers.7.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  66 :    1048576 : layers.7.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  67 :    8388608 : layers.7.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  68 :   29360128 : layers.7.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  69 :   29360128 : layers.7.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  70 :   29360128 : layers.7.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  71 :       8192 : layers.7.attention_norm.weight      : [8192]          : torch.bfloat16
  72 :       8192 : layers.7.ffn_norm.weight            : [8192]          : torch.bfloat16
  73 :    8388608 : layers.8.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  74 :    1048576 : layers.8.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  75 :    1048576 : layers.8.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  76 :    8388608 : layers.8.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  77 :   29360128 : layers.8.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  78 :   29360128 : layers.8.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  79 :   29360128 : layers.8.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  80 :       8192 : layers.8.attention_norm.weight      : [8192]          : torch.bfloat16
  81 :       8192 : layers.8.ffn_norm.weight            : [8192]          : torch.bfloat16
  82 :    8388608 : layers.9.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  83 :    1048576 : layers.9.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  84 :    1048576 : layers.9.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  85 :    8388608 : layers.9.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  86 :   29360128 : layers.9.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  87 :   29360128 : layers.9.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  88 :   29360128 : layers.9.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  89 :       8192 : layers.9.attention_norm.weight      : [8192]          : torch.bfloat16
  90 :       8192 : layers.9.ffn_norm.weight            : [8192]          : torch.bfloat16
  91 :    8388608 : layers.10.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
  92 :    1048576 : layers.10.attention.wk.weight       : [128, 8192]     : torch.bfloat16
  93 :    1048576 : layers.10.attention.wv.weight       : [128, 8192]     : torch.bfloat16
  94 :    8388608 : layers.10.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
  95 :   29360128 : layers.10.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
  96 :   29360128 : layers.10.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
  97 :   29360128 : layers.10.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
  98 :       8192 : layers.10.attention_norm.weight     : [8192]          : torch.bfloat16
  99 :       8192 : layers.10.ffn_norm.weight           : [8192]          : torch.bfloat16
 100 :    8388608 : layers.11.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 101 :    1048576 : layers.11.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 102 :    1048576 : layers.11.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 103 :    8388608 : layers.11.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 104 :   29360128 : layers.11.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 105 :   29360128 : layers.11.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 106 :   29360128 : layers.11.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 107 :       8192 : layers.11.attention_norm.weight     : [8192]          : torch.bfloat16
 108 :       8192 : layers.11.ffn_norm.weight           : [8192]          : torch.bfloat16
 109 :    8388608 : layers.12.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 110 :    1048576 : layers.12.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 111 :    1048576 : layers.12.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 112 :    8388608 : layers.12.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 113 :   29360128 : layers.12.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 114 :   29360128 : layers.12.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 115 :   29360128 : layers.12.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 116 :       8192 : layers.12.attention_norm.weight     : [8192]          : torch.bfloat16
 117 :       8192 : layers.12.ffn_norm.weight           : [8192]          : torch.bfloat16
 118 :    8388608 : layers.13.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 119 :    1048576 : layers.13.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 120 :    1048576 : layers.13.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 121 :    8388608 : layers.13.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 122 :   29360128 : layers.13.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 123 :   29360128 : layers.13.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 124 :   29360128 : layers.13.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 125 :       8192 : layers.13.attention_norm.weight     : [8192]          : torch.bfloat16
 126 :       8192 : layers.13.ffn_norm.weight           : [8192]          : torch.bfloat16
 127 :    8388608 : layers.14.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 128 :    1048576 : layers.14.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 129 :    1048576 : layers.14.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 130 :    8388608 : layers.14.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 131 :   29360128 : layers.14.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 132 :   29360128 : layers.14.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 133 :   29360128 : layers.14.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 134 :       8192 : layers.14.attention_norm.weight     : [8192]          : torch.bfloat16
 135 :       8192 : layers.14.ffn_norm.weight           : [8192]          : torch.bfloat16
 136 :    8388608 : layers.15.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 137 :    1048576 : layers.15.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 138 :    1048576 : layers.15.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 139 :    8388608 : layers.15.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 140 :   29360128 : layers.15.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 141 :   29360128 : layers.15.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 142 :   29360128 : layers.15.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 143 :       8192 : layers.15.attention_norm.weight     : [8192]          : torch.bfloat16
 144 :       8192 : layers.15.ffn_norm.weight           : [8192]          : torch.bfloat16
 145 :    8388608 : layers.16.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 146 :    1048576 : layers.16.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 147 :    1048576 : layers.16.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 148 :    8388608 : layers.16.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 149 :   29360128 : layers.16.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 150 :   29360128 : layers.16.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 151 :   29360128 : layers.16.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 152 :       8192 : layers.16.attention_norm.weight     : [8192]          : torch.bfloat16
 153 :       8192 : layers.16.ffn_norm.weight           : [8192]          : torch.bfloat16
 154 :    8388608 : layers.17.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 155 :    1048576 : layers.17.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 156 :    1048576 : layers.17.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 157 :    8388608 : layers.17.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 158 :   29360128 : layers.17.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 159 :   29360128 : layers.17.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 160 :   29360128 : layers.17.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 161 :       8192 : layers.17.attention_norm.weight     : [8192]          : torch.bfloat16
 162 :       8192 : layers.17.ffn_norm.weight           : [8192]          : torch.bfloat16
 163 :    8388608 : layers.18.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 164 :    1048576 : layers.18.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 165 :    1048576 : layers.18.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 166 :    8388608 : layers.18.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 167 :   29360128 : layers.18.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 168 :   29360128 : layers.18.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 169 :   29360128 : layers.18.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 170 :       8192 : layers.18.attention_norm.weight     : [8192]          : torch.bfloat16
 171 :       8192 : layers.18.ffn_norm.weight           : [8192]          : torch.bfloat16
 172 :    8388608 : layers.19.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 173 :    1048576 : layers.19.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 174 :    1048576 : layers.19.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 175 :    8388608 : layers.19.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 176 :   29360128 : layers.19.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 177 :   29360128 : layers.19.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 178 :   29360128 : layers.19.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 179 :       8192 : layers.19.attention_norm.weight     : [8192]          : torch.bfloat16
 180 :       8192 : layers.19.ffn_norm.weight           : [8192]          : torch.bfloat16
 181 :    8388608 : layers.20.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 182 :    1048576 : layers.20.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 183 :    1048576 : layers.20.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 184 :    8388608 : layers.20.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 185 :   29360128 : layers.20.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 186 :   29360128 : layers.20.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 187 :   29360128 : layers.20.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 188 :       8192 : layers.20.attention_norm.weight     : [8192]          : torch.bfloat16
 189 :       8192 : layers.20.ffn_norm.weight           : [8192]          : torch.bfloat16
 190 :    8388608 : layers.21.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 191 :    1048576 : layers.21.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 192 :    1048576 : layers.21.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 193 :    8388608 : layers.21.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 194 :   29360128 : layers.21.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 195 :   29360128 : layers.21.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 196 :   29360128 : layers.21.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 197 :       8192 : layers.21.attention_norm.weight     : [8192]          : torch.bfloat16
 198 :       8192 : layers.21.ffn_norm.weight           : [8192]          : torch.bfloat16
 199 :    8388608 : layers.22.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 200 :    1048576 : layers.22.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 201 :    1048576 : layers.22.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 202 :    8388608 : layers.22.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 203 :   29360128 : layers.22.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 204 :   29360128 : layers.22.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 205 :   29360128 : layers.22.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 206 :       8192 : layers.22.attention_norm.weight     : [8192]          : torch.bfloat16
 207 :       8192 : layers.22.ffn_norm.weight           : [8192]          : torch.bfloat16
 208 :    8388608 : layers.23.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 209 :    1048576 : layers.23.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 210 :    1048576 : layers.23.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 211 :    8388608 : layers.23.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 212 :   29360128 : layers.23.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 213 :   29360128 : layers.23.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 214 :   29360128 : layers.23.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 215 :       8192 : layers.23.attention_norm.weight     : [8192]          : torch.bfloat16
 216 :       8192 : layers.23.ffn_norm.weight           : [8192]          : torch.bfloat16
 217 :    8388608 : layers.24.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 218 :    1048576 : layers.24.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 219 :    1048576 : layers.24.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 220 :    8388608 : layers.24.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 221 :   29360128 : layers.24.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 222 :   29360128 : layers.24.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 223 :   29360128 : layers.24.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 224 :       8192 : layers.24.attention_norm.weight     : [8192]          : torch.bfloat16
 225 :       8192 : layers.24.ffn_norm.weight           : [8192]          : torch.bfloat16
 226 :    8388608 : layers.25.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 227 :    1048576 : layers.25.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 228 :    1048576 : layers.25.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 229 :    8388608 : layers.25.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 230 :   29360128 : layers.25.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 231 :   29360128 : layers.25.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 232 :   29360128 : layers.25.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 233 :       8192 : layers.25.attention_norm.weight     : [8192]          : torch.bfloat16
 234 :       8192 : layers.25.ffn_norm.weight           : [8192]          : torch.bfloat16
 235 :    8388608 : layers.26.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 236 :    1048576 : layers.26.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 237 :    1048576 : layers.26.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 238 :    8388608 : layers.26.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 239 :   29360128 : layers.26.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 240 :   29360128 : layers.26.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 241 :   29360128 : layers.26.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 242 :       8192 : layers.26.attention_norm.weight     : [8192]          : torch.bfloat16
 243 :       8192 : layers.26.ffn_norm.weight           : [8192]          : torch.bfloat16
 244 :    8388608 : layers.27.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 245 :    1048576 : layers.27.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 246 :    1048576 : layers.27.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 247 :    8388608 : layers.27.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 248 :   29360128 : layers.27.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 249 :   29360128 : layers.27.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 250 :   29360128 : layers.27.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 251 :       8192 : layers.27.attention_norm.weight     : [8192]          : torch.bfloat16
 252 :       8192 : layers.27.ffn_norm.weight           : [8192]          : torch.bfloat16
 253 :    8388608 : layers.28.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 254 :    1048576 : layers.28.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 255 :    1048576 : layers.28.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 256 :    8388608 : layers.28.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 257 :   29360128 : layers.28.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 258 :   29360128 : layers.28.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 259 :   29360128 : layers.28.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 260 :       8192 : layers.28.attention_norm.weight     : [8192]          : torch.bfloat16
 261 :       8192 : layers.28.ffn_norm.weight           : [8192]          : torch.bfloat16
 262 :    8388608 : layers.29.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 263 :    1048576 : layers.29.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 264 :    1048576 : layers.29.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 265 :    8388608 : layers.29.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 266 :   29360128 : layers.29.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 267 :   29360128 : layers.29.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 268 :   29360128 : layers.29.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 269 :       8192 : layers.29.attention_norm.weight     : [8192]          : torch.bfloat16
 270 :       8192 : layers.29.ffn_norm.weight           : [8192]          : torch.bfloat16
 271 :    8388608 : layers.30.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 272 :    1048576 : layers.30.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 273 :    1048576 : layers.30.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 274 :    8388608 : layers.30.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 275 :   29360128 : layers.30.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 276 :   29360128 : layers.30.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 277 :   29360128 : layers.30.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 278 :       8192 : layers.30.attention_norm.weight     : [8192]          : torch.bfloat16
 279 :       8192 : layers.30.ffn_norm.weight           : [8192]          : torch.bfloat16
 280 :    8388608 : layers.31.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 281 :    1048576 : layers.31.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 282 :    1048576 : layers.31.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 283 :    8388608 : layers.31.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 284 :   29360128 : layers.31.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 285 :   29360128 : layers.31.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 286 :   29360128 : layers.31.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 287 :       8192 : layers.31.attention_norm.weight     : [8192]          : torch.bfloat16
 288 :       8192 : layers.31.ffn_norm.weight           : [8192]          : torch.bfloat16
 289 :    8388608 : layers.32.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 290 :    1048576 : layers.32.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 291 :    1048576 : layers.32.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 292 :    8388608 : layers.32.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 293 :   29360128 : layers.32.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 294 :   29360128 : layers.32.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 295 :   29360128 : layers.32.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 296 :       8192 : layers.32.attention_norm.weight     : [8192]          : torch.bfloat16
 297 :       8192 : layers.32.ffn_norm.weight           : [8192]          : torch.bfloat16
 298 :    8388608 : layers.33.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 299 :    1048576 : layers.33.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 300 :    1048576 : layers.33.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 301 :    8388608 : layers.33.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 302 :   29360128 : layers.33.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 303 :   29360128 : layers.33.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 304 :   29360128 : layers.33.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 305 :       8192 : layers.33.attention_norm.weight     : [8192]          : torch.bfloat16
 306 :       8192 : layers.33.ffn_norm.weight           : [8192]          : torch.bfloat16
 307 :    8388608 : layers.34.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 308 :    1048576 : layers.34.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 309 :    1048576 : layers.34.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 310 :    8388608 : layers.34.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 311 :   29360128 : layers.34.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 312 :   29360128 : layers.34.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 313 :   29360128 : layers.34.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 314 :       8192 : layers.34.attention_norm.weight     : [8192]          : torch.bfloat16
 315 :       8192 : layers.34.ffn_norm.weight           : [8192]          : torch.bfloat16
 316 :    8388608 : layers.35.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 317 :    1048576 : layers.35.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 318 :    1048576 : layers.35.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 319 :    8388608 : layers.35.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 320 :   29360128 : layers.35.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 321 :   29360128 : layers.35.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 322 :   29360128 : layers.35.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 323 :       8192 : layers.35.attention_norm.weight     : [8192]          : torch.bfloat16
 324 :       8192 : layers.35.ffn_norm.weight           : [8192]          : torch.bfloat16
 325 :    8388608 : layers.36.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 326 :    1048576 : layers.36.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 327 :    1048576 : layers.36.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 328 :    8388608 : layers.36.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 329 :   29360128 : layers.36.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 330 :   29360128 : layers.36.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 331 :   29360128 : layers.36.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 332 :       8192 : layers.36.attention_norm.weight     : [8192]          : torch.bfloat16
 333 :       8192 : layers.36.ffn_norm.weight           : [8192]          : torch.bfloat16
 334 :    8388608 : layers.37.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 335 :    1048576 : layers.37.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 336 :    1048576 : layers.37.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 337 :    8388608 : layers.37.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 338 :   29360128 : layers.37.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 339 :   29360128 : layers.37.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 340 :   29360128 : layers.37.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 341 :       8192 : layers.37.attention_norm.weight     : [8192]          : torch.bfloat16
 342 :       8192 : layers.37.ffn_norm.weight           : [8192]          : torch.bfloat16
 343 :    8388608 : layers.38.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 344 :    1048576 : layers.38.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 345 :    1048576 : layers.38.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 346 :    8388608 : layers.38.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 347 :   29360128 : layers.38.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 348 :   29360128 : layers.38.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 349 :   29360128 : layers.38.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 350 :       8192 : layers.38.attention_norm.weight     : [8192]          : torch.bfloat16
 351 :       8192 : layers.38.ffn_norm.weight           : [8192]          : torch.bfloat16
 352 :    8388608 : layers.39.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 353 :    1048576 : layers.39.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 354 :    1048576 : layers.39.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 355 :    8388608 : layers.39.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 356 :   29360128 : layers.39.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 357 :   29360128 : layers.39.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 358 :   29360128 : layers.39.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 359 :       8192 : layers.39.attention_norm.weight     : [8192]          : torch.bfloat16
 360 :       8192 : layers.39.ffn_norm.weight           : [8192]          : torch.bfloat16
 361 :    8388608 : layers.40.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 362 :    1048576 : layers.40.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 363 :    1048576 : layers.40.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 364 :    8388608 : layers.40.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 365 :   29360128 : layers.40.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 366 :   29360128 : layers.40.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 367 :   29360128 : layers.40.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 368 :       8192 : layers.40.attention_norm.weight     : [8192]          : torch.bfloat16
 369 :       8192 : layers.40.ffn_norm.weight           : [8192]          : torch.bfloat16
 370 :    8388608 : layers.41.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 371 :    1048576 : layers.41.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 372 :    1048576 : layers.41.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 373 :    8388608 : layers.41.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 374 :   29360128 : layers.41.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 375 :   29360128 : layers.41.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 376 :   29360128 : layers.41.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 377 :       8192 : layers.41.attention_norm.weight     : [8192]          : torch.bfloat16
 378 :       8192 : layers.41.ffn_norm.weight           : [8192]          : torch.bfloat16
 379 :    8388608 : layers.42.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 380 :    1048576 : layers.42.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 381 :    1048576 : layers.42.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 382 :    8388608 : layers.42.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 383 :   29360128 : layers.42.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 384 :   29360128 : layers.42.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 385 :   29360128 : layers.42.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 386 :       8192 : layers.42.attention_norm.weight     : [8192]          : torch.bfloat16
 387 :       8192 : layers.42.ffn_norm.weight           : [8192]          : torch.bfloat16
 388 :    8388608 : layers.43.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 389 :    1048576 : layers.43.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 390 :    1048576 : layers.43.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 391 :    8388608 : layers.43.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 392 :   29360128 : layers.43.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 393 :   29360128 : layers.43.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 394 :   29360128 : layers.43.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 395 :       8192 : layers.43.attention_norm.weight     : [8192]          : torch.bfloat16
 396 :       8192 : layers.43.ffn_norm.weight           : [8192]          : torch.bfloat16
 397 :    8388608 : layers.44.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 398 :    1048576 : layers.44.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 399 :    1048576 : layers.44.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 400 :    8388608 : layers.44.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 401 :   29360128 : layers.44.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 402 :   29360128 : layers.44.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 403 :   29360128 : layers.44.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 404 :       8192 : layers.44.attention_norm.weight     : [8192]          : torch.bfloat16
 405 :       8192 : layers.44.ffn_norm.weight           : [8192]          : torch.bfloat16
 406 :    8388608 : layers.45.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 407 :    1048576 : layers.45.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 408 :    1048576 : layers.45.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 409 :    8388608 : layers.45.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 410 :   29360128 : layers.45.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 411 :   29360128 : layers.45.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 412 :   29360128 : layers.45.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 413 :       8192 : layers.45.attention_norm.weight     : [8192]          : torch.bfloat16
 414 :       8192 : layers.45.ffn_norm.weight           : [8192]          : torch.bfloat16
 415 :    8388608 : layers.46.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 416 :    1048576 : layers.46.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 417 :    1048576 : layers.46.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 418 :    8388608 : layers.46.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 419 :   29360128 : layers.46.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 420 :   29360128 : layers.46.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 421 :   29360128 : layers.46.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 422 :       8192 : layers.46.attention_norm.weight     : [8192]          : torch.bfloat16
 423 :       8192 : layers.46.ffn_norm.weight           : [8192]          : torch.bfloat16
 424 :    8388608 : layers.47.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 425 :    1048576 : layers.47.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 426 :    1048576 : layers.47.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 427 :    8388608 : layers.47.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 428 :   29360128 : layers.47.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 429 :   29360128 : layers.47.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 430 :   29360128 : layers.47.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 431 :       8192 : layers.47.attention_norm.weight     : [8192]          : torch.bfloat16
 432 :       8192 : layers.47.ffn_norm.weight           : [8192]          : torch.bfloat16
 433 :    8388608 : layers.48.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 434 :    1048576 : layers.48.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 435 :    1048576 : layers.48.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 436 :    8388608 : layers.48.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 437 :   29360128 : layers.48.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 438 :   29360128 : layers.48.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 439 :   29360128 : layers.48.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 440 :       8192 : layers.48.attention_norm.weight     : [8192]          : torch.bfloat16
 441 :       8192 : layers.48.ffn_norm.weight           : [8192]          : torch.bfloat16
 442 :    8388608 : layers.49.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 443 :    1048576 : layers.49.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 444 :    1048576 : layers.49.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 445 :    8388608 : layers.49.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 446 :   29360128 : layers.49.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 447 :   29360128 : layers.49.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 448 :   29360128 : layers.49.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 449 :       8192 : layers.49.attention_norm.weight     : [8192]          : torch.bfloat16
 450 :       8192 : layers.49.ffn_norm.weight           : [8192]          : torch.bfloat16
 451 :    8388608 : layers.50.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 452 :    1048576 : layers.50.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 453 :    1048576 : layers.50.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 454 :    8388608 : layers.50.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 455 :   29360128 : layers.50.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 456 :   29360128 : layers.50.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 457 :   29360128 : layers.50.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 458 :       8192 : layers.50.attention_norm.weight     : [8192]          : torch.bfloat16
 459 :       8192 : layers.50.ffn_norm.weight           : [8192]          : torch.bfloat16
 460 :    8388608 : layers.51.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 461 :    1048576 : layers.51.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 462 :    1048576 : layers.51.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 463 :    8388608 : layers.51.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 464 :   29360128 : layers.51.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 465 :   29360128 : layers.51.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 466 :   29360128 : layers.51.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 467 :       8192 : layers.51.attention_norm.weight     : [8192]          : torch.bfloat16
 468 :       8192 : layers.51.ffn_norm.weight           : [8192]          : torch.bfloat16
 469 :    8388608 : layers.52.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 470 :    1048576 : layers.52.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 471 :    1048576 : layers.52.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 472 :    8388608 : layers.52.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 473 :   29360128 : layers.52.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 474 :   29360128 : layers.52.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 475 :   29360128 : layers.52.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 476 :       8192 : layers.52.attention_norm.weight     : [8192]          : torch.bfloat16
 477 :       8192 : layers.52.ffn_norm.weight           : [8192]          : torch.bfloat16
 478 :    8388608 : layers.53.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 479 :    1048576 : layers.53.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 480 :    1048576 : layers.53.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 481 :    8388608 : layers.53.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 482 :   29360128 : layers.53.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 483 :   29360128 : layers.53.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 484 :   29360128 : layers.53.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 485 :       8192 : layers.53.attention_norm.weight     : [8192]          : torch.bfloat16
 486 :       8192 : layers.53.ffn_norm.weight           : [8192]          : torch.bfloat16
 487 :    8388608 : layers.54.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 488 :    1048576 : layers.54.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 489 :    1048576 : layers.54.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 490 :    8388608 : layers.54.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 491 :   29360128 : layers.54.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 492 :   29360128 : layers.54.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 493 :   29360128 : layers.54.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 494 :       8192 : layers.54.attention_norm.weight     : [8192]          : torch.bfloat16
 495 :       8192 : layers.54.ffn_norm.weight           : [8192]          : torch.bfloat16
 496 :    8388608 : layers.55.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 497 :    1048576 : layers.55.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 498 :    1048576 : layers.55.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 499 :    8388608 : layers.55.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 500 :   29360128 : layers.55.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 501 :   29360128 : layers.55.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 502 :   29360128 : layers.55.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 503 :       8192 : layers.55.attention_norm.weight     : [8192]          : torch.bfloat16
 504 :       8192 : layers.55.ffn_norm.weight           : [8192]          : torch.bfloat16
 505 :    8388608 : layers.56.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 506 :    1048576 : layers.56.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 507 :    1048576 : layers.56.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 508 :    8388608 : layers.56.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 509 :   29360128 : layers.56.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 510 :   29360128 : layers.56.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 511 :   29360128 : layers.56.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 512 :       8192 : layers.56.attention_norm.weight     : [8192]          : torch.bfloat16
 513 :       8192 : layers.56.ffn_norm.weight           : [8192]          : torch.bfloat16
 514 :    8388608 : layers.57.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 515 :    1048576 : layers.57.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 516 :    1048576 : layers.57.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 517 :    8388608 : layers.57.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 518 :   29360128 : layers.57.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 519 :   29360128 : layers.57.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 520 :   29360128 : layers.57.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 521 :       8192 : layers.57.attention_norm.weight     : [8192]          : torch.bfloat16
 522 :       8192 : layers.57.ffn_norm.weight           : [8192]          : torch.bfloat16
 523 :    8388608 : layers.58.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 524 :    1048576 : layers.58.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 525 :    1048576 : layers.58.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 526 :    8388608 : layers.58.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 527 :   29360128 : layers.58.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 528 :   29360128 : layers.58.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 529 :   29360128 : layers.58.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 530 :       8192 : layers.58.attention_norm.weight     : [8192]          : torch.bfloat16
 531 :       8192 : layers.58.ffn_norm.weight           : [8192]          : torch.bfloat16
 532 :    8388608 : layers.59.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 533 :    1048576 : layers.59.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 534 :    1048576 : layers.59.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 535 :    8388608 : layers.59.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 536 :   29360128 : layers.59.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 537 :   29360128 : layers.59.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 538 :   29360128 : layers.59.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 539 :       8192 : layers.59.attention_norm.weight     : [8192]          : torch.bfloat16
 540 :       8192 : layers.59.ffn_norm.weight           : [8192]          : torch.bfloat16
 541 :    8388608 : layers.60.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 542 :    1048576 : layers.60.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 543 :    1048576 : layers.60.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 544 :    8388608 : layers.60.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 545 :   29360128 : layers.60.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 546 :   29360128 : layers.60.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 547 :   29360128 : layers.60.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 548 :       8192 : layers.60.attention_norm.weight     : [8192]          : torch.bfloat16
 549 :       8192 : layers.60.ffn_norm.weight           : [8192]          : torch.bfloat16
 550 :    8388608 : layers.61.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 551 :    1048576 : layers.61.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 552 :    1048576 : layers.61.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 553 :    8388608 : layers.61.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 554 :   29360128 : layers.61.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 555 :   29360128 : layers.61.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 556 :   29360128 : layers.61.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 557 :       8192 : layers.61.attention_norm.weight     : [8192]          : torch.bfloat16
 558 :       8192 : layers.61.ffn_norm.weight           : [8192]          : torch.bfloat16
 559 :    8388608 : layers.62.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 560 :    1048576 : layers.62.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 561 :    1048576 : layers.62.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 562 :    8388608 : layers.62.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 563 :   29360128 : layers.62.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 564 :   29360128 : layers.62.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 565 :   29360128 : layers.62.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 566 :       8192 : layers.62.attention_norm.weight     : [8192]          : torch.bfloat16
 567 :       8192 : layers.62.ffn_norm.weight           : [8192]          : torch.bfloat16
 568 :    8388608 : layers.63.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 569 :    1048576 : layers.63.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 570 :    1048576 : layers.63.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 571 :    8388608 : layers.63.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 572 :   29360128 : layers.63.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 573 :   29360128 : layers.63.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 574 :   29360128 : layers.63.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 575 :       8192 : layers.63.attention_norm.weight     : [8192]          : torch.bfloat16
 576 :       8192 : layers.63.ffn_norm.weight           : [8192]          : torch.bfloat16
 577 :    8388608 : layers.64.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 578 :    1048576 : layers.64.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 579 :    1048576 : layers.64.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 580 :    8388608 : layers.64.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 581 :   29360128 : layers.64.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 582 :   29360128 : layers.64.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 583 :   29360128 : layers.64.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 584 :       8192 : layers.64.attention_norm.weight     : [8192]          : torch.bfloat16
 585 :       8192 : layers.64.ffn_norm.weight           : [8192]          : torch.bfloat16
 586 :    8388608 : layers.65.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 587 :    1048576 : layers.65.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 588 :    1048576 : layers.65.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 589 :    8388608 : layers.65.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 590 :   29360128 : layers.65.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 591 :   29360128 : layers.65.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 592 :   29360128 : layers.65.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 593 :       8192 : layers.65.attention_norm.weight     : [8192]          : torch.bfloat16
 594 :       8192 : layers.65.ffn_norm.weight           : [8192]          : torch.bfloat16
 595 :    8388608 : layers.66.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 596 :    1048576 : layers.66.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 597 :    1048576 : layers.66.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 598 :    8388608 : layers.66.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 599 :   29360128 : layers.66.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 600 :   29360128 : layers.66.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 601 :   29360128 : layers.66.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 602 :       8192 : layers.66.attention_norm.weight     : [8192]          : torch.bfloat16
 603 :       8192 : layers.66.ffn_norm.weight           : [8192]          : torch.bfloat16
 604 :    8388608 : layers.67.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 605 :    1048576 : layers.67.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 606 :    1048576 : layers.67.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 607 :    8388608 : layers.67.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 608 :   29360128 : layers.67.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 609 :   29360128 : layers.67.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 610 :   29360128 : layers.67.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 611 :       8192 : layers.67.attention_norm.weight     : [8192]          : torch.bfloat16
 612 :       8192 : layers.67.ffn_norm.weight           : [8192]          : torch.bfloat16
 613 :    8388608 : layers.68.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 614 :    1048576 : layers.68.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 615 :    1048576 : layers.68.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 616 :    8388608 : layers.68.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 617 :   29360128 : layers.68.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 618 :   29360128 : layers.68.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 619 :   29360128 : layers.68.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 620 :       8192 : layers.68.attention_norm.weight     : [8192]          : torch.bfloat16
 621 :       8192 : layers.68.ffn_norm.weight           : [8192]          : torch.bfloat16
 622 :    8388608 : layers.69.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 623 :    1048576 : layers.69.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 624 :    1048576 : layers.69.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 625 :    8388608 : layers.69.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 626 :   29360128 : layers.69.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 627 :   29360128 : layers.69.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 628 :   29360128 : layers.69.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 629 :       8192 : layers.69.attention_norm.weight     : [8192]          : torch.bfloat16
 630 :       8192 : layers.69.ffn_norm.weight           : [8192]          : torch.bfloat16
 631 :    8388608 : layers.70.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 632 :    1048576 : layers.70.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 633 :    1048576 : layers.70.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 634 :    8388608 : layers.70.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 635 :   29360128 : layers.70.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 636 :   29360128 : layers.70.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 637 :   29360128 : layers.70.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 638 :       8192 : layers.70.attention_norm.weight     : [8192]          : torch.bfloat16
 639 :       8192 : layers.70.ffn_norm.weight           : [8192]          : torch.bfloat16
 640 :    8388608 : layers.71.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 641 :    1048576 : layers.71.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 642 :    1048576 : layers.71.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 643 :    8388608 : layers.71.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 644 :   29360128 : layers.71.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 645 :   29360128 : layers.71.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 646 :   29360128 : layers.71.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 647 :       8192 : layers.71.attention_norm.weight     : [8192]          : torch.bfloat16
 648 :       8192 : layers.71.ffn_norm.weight           : [8192]          : torch.bfloat16
 649 :    8388608 : layers.72.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 650 :    1048576 : layers.72.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 651 :    1048576 : layers.72.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 652 :    8388608 : layers.72.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 653 :   29360128 : layers.72.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 654 :   29360128 : layers.72.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 655 :   29360128 : layers.72.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 656 :       8192 : layers.72.attention_norm.weight     : [8192]          : torch.bfloat16
 657 :       8192 : layers.72.ffn_norm.weight           : [8192]          : torch.bfloat16
 658 :    8388608 : layers.73.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 659 :    1048576 : layers.73.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 660 :    1048576 : layers.73.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 661 :    8388608 : layers.73.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 662 :   29360128 : layers.73.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 663 :   29360128 : layers.73.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 664 :   29360128 : layers.73.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 665 :       8192 : layers.73.attention_norm.weight     : [8192]          : torch.bfloat16
 666 :       8192 : layers.73.ffn_norm.weight           : [8192]          : torch.bfloat16
 667 :    8388608 : layers.74.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 668 :    1048576 : layers.74.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 669 :    1048576 : layers.74.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 670 :    8388608 : layers.74.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 671 :   29360128 : layers.74.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 672 :   29360128 : layers.74.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 673 :   29360128 : layers.74.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 674 :       8192 : layers.74.attention_norm.weight     : [8192]          : torch.bfloat16
 675 :       8192 : layers.74.ffn_norm.weight           : [8192]          : torch.bfloat16
 676 :    8388608 : layers.75.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 677 :    1048576 : layers.75.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 678 :    1048576 : layers.75.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 679 :    8388608 : layers.75.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 680 :   29360128 : layers.75.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 681 :   29360128 : layers.75.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 682 :   29360128 : layers.75.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 683 :       8192 : layers.75.attention_norm.weight     : [8192]          : torch.bfloat16
 684 :       8192 : layers.75.ffn_norm.weight           : [8192]          : torch.bfloat16
 685 :    8388608 : layers.76.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 686 :    1048576 : layers.76.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 687 :    1048576 : layers.76.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 688 :    8388608 : layers.76.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 689 :   29360128 : layers.76.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 690 :   29360128 : layers.76.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 691 :   29360128 : layers.76.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 692 :       8192 : layers.76.attention_norm.weight     : [8192]          : torch.bfloat16
 693 :       8192 : layers.76.ffn_norm.weight           : [8192]          : torch.bfloat16
 694 :    8388608 : layers.77.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 695 :    1048576 : layers.77.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 696 :    1048576 : layers.77.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 697 :    8388608 : layers.77.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 698 :   29360128 : layers.77.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 699 :   29360128 : layers.77.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 700 :   29360128 : layers.77.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 701 :       8192 : layers.77.attention_norm.weight     : [8192]          : torch.bfloat16
 702 :       8192 : layers.77.ffn_norm.weight           : [8192]          : torch.bfloat16
 703 :    8388608 : layers.78.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 704 :    1048576 : layers.78.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 705 :    1048576 : layers.78.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 706 :    8388608 : layers.78.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 707 :   29360128 : layers.78.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 708 :   29360128 : layers.78.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 709 :   29360128 : layers.78.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 710 :       8192 : layers.78.attention_norm.weight     : [8192]          : torch.bfloat16
 711 :       8192 : layers.78.ffn_norm.weight           : [8192]          : torch.bfloat16
 712 :    8388608 : layers.79.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 713 :    1048576 : layers.79.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 714 :    1048576 : layers.79.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 715 :    8388608 : layers.79.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 716 :   29360128 : layers.79.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 717 :   29360128 : layers.79.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 718 :   29360128 : layers.79.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 719 :       8192 : layers.79.attention_norm.weight     : [8192]          : torch.bfloat16
 720 :       8192 : layers.79.ffn_norm.weight           : [8192]          : torch.bfloat16
 721 :       8192 : norm.weight                         : [8192]          : torch.bfloat16
 722 :  131334144 : output.weight                       : [16032, 8192]   : torch.bfloat16

-----------------------------------------------------------------------------
26461102080 params in total.
52922204160 bytes in total.
18.55 sec, 6.39 sec, 2721.17 MB/s
-----------------------------------------------------------------------------

[4/8]: Loading original/consolidated.03.pth
   0 :  131334144 : tok_embeddings.weight               : [16032, 8192]   : torch.bfloat16
   1 :    8388608 : layers.0.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
   2 :    1048576 : layers.0.attention.wk.weight        : [128, 8192]     : torch.bfloat16
   3 :    1048576 : layers.0.attention.wv.weight        : [128, 8192]     : torch.bfloat16
   4 :    8388608 : layers.0.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
   5 :   29360128 : layers.0.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
   6 :   29360128 : layers.0.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
   7 :   29360128 : layers.0.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
   8 :       8192 : layers.0.attention_norm.weight      : [8192]          : torch.bfloat16
   9 :       8192 : layers.0.ffn_norm.weight            : [8192]          : torch.bfloat16
  10 :    8388608 : layers.1.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  11 :    1048576 : layers.1.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  12 :    1048576 : layers.1.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  13 :    8388608 : layers.1.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  14 :   29360128 : layers.1.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  15 :   29360128 : layers.1.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  16 :   29360128 : layers.1.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  17 :       8192 : layers.1.attention_norm.weight      : [8192]          : torch.bfloat16
  18 :       8192 : layers.1.ffn_norm.weight            : [8192]          : torch.bfloat16
  19 :    8388608 : layers.2.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  20 :    1048576 : layers.2.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  21 :    1048576 : layers.2.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  22 :    8388608 : layers.2.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  23 :   29360128 : layers.2.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  24 :   29360128 : layers.2.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  25 :   29360128 : layers.2.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  26 :       8192 : layers.2.attention_norm.weight      : [8192]          : torch.bfloat16
  27 :       8192 : layers.2.ffn_norm.weight            : [8192]          : torch.bfloat16
  28 :    8388608 : layers.3.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  29 :    1048576 : layers.3.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  30 :    1048576 : layers.3.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  31 :    8388608 : layers.3.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  32 :   29360128 : layers.3.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  33 :   29360128 : layers.3.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  34 :   29360128 : layers.3.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  35 :       8192 : layers.3.attention_norm.weight      : [8192]          : torch.bfloat16
  36 :       8192 : layers.3.ffn_norm.weight            : [8192]          : torch.bfloat16
  37 :    8388608 : layers.4.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  38 :    1048576 : layers.4.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  39 :    1048576 : layers.4.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  40 :    8388608 : layers.4.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  41 :   29360128 : layers.4.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  42 :   29360128 : layers.4.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  43 :   29360128 : layers.4.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  44 :       8192 : layers.4.attention_norm.weight      : [8192]          : torch.bfloat16
  45 :       8192 : layers.4.ffn_norm.weight            : [8192]          : torch.bfloat16
  46 :    8388608 : layers.5.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  47 :    1048576 : layers.5.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  48 :    1048576 : layers.5.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  49 :    8388608 : layers.5.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  50 :   29360128 : layers.5.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  51 :   29360128 : layers.5.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  52 :   29360128 : layers.5.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  53 :       8192 : layers.5.attention_norm.weight      : [8192]          : torch.bfloat16
  54 :       8192 : layers.5.ffn_norm.weight            : [8192]          : torch.bfloat16
  55 :    8388608 : layers.6.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  56 :    1048576 : layers.6.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  57 :    1048576 : layers.6.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  58 :    8388608 : layers.6.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  59 :   29360128 : layers.6.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  60 :   29360128 : layers.6.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  61 :   29360128 : layers.6.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  62 :       8192 : layers.6.attention_norm.weight      : [8192]          : torch.bfloat16
  63 :       8192 : layers.6.ffn_norm.weight            : [8192]          : torch.bfloat16
  64 :    8388608 : layers.7.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  65 :    1048576 : layers.7.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  66 :    1048576 : layers.7.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  67 :    8388608 : layers.7.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  68 :   29360128 : layers.7.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  69 :   29360128 : layers.7.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  70 :   29360128 : layers.7.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  71 :       8192 : layers.7.attention_norm.weight      : [8192]          : torch.bfloat16
  72 :       8192 : layers.7.ffn_norm.weight            : [8192]          : torch.bfloat16
  73 :    8388608 : layers.8.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  74 :    1048576 : layers.8.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  75 :    1048576 : layers.8.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  76 :    8388608 : layers.8.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  77 :   29360128 : layers.8.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  78 :   29360128 : layers.8.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  79 :   29360128 : layers.8.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  80 :       8192 : layers.8.attention_norm.weight      : [8192]          : torch.bfloat16
  81 :       8192 : layers.8.ffn_norm.weight            : [8192]          : torch.bfloat16
  82 :    8388608 : layers.9.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  83 :    1048576 : layers.9.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  84 :    1048576 : layers.9.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  85 :    8388608 : layers.9.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  86 :   29360128 : layers.9.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  87 :   29360128 : layers.9.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  88 :   29360128 : layers.9.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  89 :       8192 : layers.9.attention_norm.weight      : [8192]          : torch.bfloat16
  90 :       8192 : layers.9.ffn_norm.weight            : [8192]          : torch.bfloat16
  91 :    8388608 : layers.10.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
  92 :    1048576 : layers.10.attention.wk.weight       : [128, 8192]     : torch.bfloat16
  93 :    1048576 : layers.10.attention.wv.weight       : [128, 8192]     : torch.bfloat16
  94 :    8388608 : layers.10.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
  95 :   29360128 : layers.10.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
  96 :   29360128 : layers.10.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
  97 :   29360128 : layers.10.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
  98 :       8192 : layers.10.attention_norm.weight     : [8192]          : torch.bfloat16
  99 :       8192 : layers.10.ffn_norm.weight           : [8192]          : torch.bfloat16
 100 :    8388608 : layers.11.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 101 :    1048576 : layers.11.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 102 :    1048576 : layers.11.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 103 :    8388608 : layers.11.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 104 :   29360128 : layers.11.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 105 :   29360128 : layers.11.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 106 :   29360128 : layers.11.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 107 :       8192 : layers.11.attention_norm.weight     : [8192]          : torch.bfloat16
 108 :       8192 : layers.11.ffn_norm.weight           : [8192]          : torch.bfloat16
 109 :    8388608 : layers.12.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 110 :    1048576 : layers.12.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 111 :    1048576 : layers.12.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 112 :    8388608 : layers.12.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 113 :   29360128 : layers.12.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 114 :   29360128 : layers.12.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 115 :   29360128 : layers.12.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 116 :       8192 : layers.12.attention_norm.weight     : [8192]          : torch.bfloat16
 117 :       8192 : layers.12.ffn_norm.weight           : [8192]          : torch.bfloat16
 118 :    8388608 : layers.13.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 119 :    1048576 : layers.13.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 120 :    1048576 : layers.13.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 121 :    8388608 : layers.13.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 122 :   29360128 : layers.13.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 123 :   29360128 : layers.13.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 124 :   29360128 : layers.13.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 125 :       8192 : layers.13.attention_norm.weight     : [8192]          : torch.bfloat16
 126 :       8192 : layers.13.ffn_norm.weight           : [8192]          : torch.bfloat16
 127 :    8388608 : layers.14.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 128 :    1048576 : layers.14.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 129 :    1048576 : layers.14.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 130 :    8388608 : layers.14.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 131 :   29360128 : layers.14.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 132 :   29360128 : layers.14.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 133 :   29360128 : layers.14.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 134 :       8192 : layers.14.attention_norm.weight     : [8192]          : torch.bfloat16
 135 :       8192 : layers.14.ffn_norm.weight           : [8192]          : torch.bfloat16
 136 :    8388608 : layers.15.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 137 :    1048576 : layers.15.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 138 :    1048576 : layers.15.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 139 :    8388608 : layers.15.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 140 :   29360128 : layers.15.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 141 :   29360128 : layers.15.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 142 :   29360128 : layers.15.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 143 :       8192 : layers.15.attention_norm.weight     : [8192]          : torch.bfloat16
 144 :       8192 : layers.15.ffn_norm.weight           : [8192]          : torch.bfloat16
 145 :    8388608 : layers.16.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 146 :    1048576 : layers.16.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 147 :    1048576 : layers.16.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 148 :    8388608 : layers.16.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 149 :   29360128 : layers.16.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 150 :   29360128 : layers.16.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 151 :   29360128 : layers.16.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 152 :       8192 : layers.16.attention_norm.weight     : [8192]          : torch.bfloat16
 153 :       8192 : layers.16.ffn_norm.weight           : [8192]          : torch.bfloat16
 154 :    8388608 : layers.17.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 155 :    1048576 : layers.17.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 156 :    1048576 : layers.17.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 157 :    8388608 : layers.17.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 158 :   29360128 : layers.17.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 159 :   29360128 : layers.17.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 160 :   29360128 : layers.17.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 161 :       8192 : layers.17.attention_norm.weight     : [8192]          : torch.bfloat16
 162 :       8192 : layers.17.ffn_norm.weight           : [8192]          : torch.bfloat16
 163 :    8388608 : layers.18.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 164 :    1048576 : layers.18.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 165 :    1048576 : layers.18.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 166 :    8388608 : layers.18.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 167 :   29360128 : layers.18.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 168 :   29360128 : layers.18.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 169 :   29360128 : layers.18.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 170 :       8192 : layers.18.attention_norm.weight     : [8192]          : torch.bfloat16
 171 :       8192 : layers.18.ffn_norm.weight           : [8192]          : torch.bfloat16
 172 :    8388608 : layers.19.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 173 :    1048576 : layers.19.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 174 :    1048576 : layers.19.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 175 :    8388608 : layers.19.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 176 :   29360128 : layers.19.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 177 :   29360128 : layers.19.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 178 :   29360128 : layers.19.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 179 :       8192 : layers.19.attention_norm.weight     : [8192]          : torch.bfloat16
 180 :       8192 : layers.19.ffn_norm.weight           : [8192]          : torch.bfloat16
 181 :    8388608 : layers.20.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 182 :    1048576 : layers.20.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 183 :    1048576 : layers.20.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 184 :    8388608 : layers.20.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 185 :   29360128 : layers.20.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 186 :   29360128 : layers.20.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 187 :   29360128 : layers.20.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 188 :       8192 : layers.20.attention_norm.weight     : [8192]          : torch.bfloat16
 189 :       8192 : layers.20.ffn_norm.weight           : [8192]          : torch.bfloat16
 190 :    8388608 : layers.21.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 191 :    1048576 : layers.21.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 192 :    1048576 : layers.21.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 193 :    8388608 : layers.21.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 194 :   29360128 : layers.21.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 195 :   29360128 : layers.21.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 196 :   29360128 : layers.21.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 197 :       8192 : layers.21.attention_norm.weight     : [8192]          : torch.bfloat16
 198 :       8192 : layers.21.ffn_norm.weight           : [8192]          : torch.bfloat16
 199 :    8388608 : layers.22.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 200 :    1048576 : layers.22.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 201 :    1048576 : layers.22.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 202 :    8388608 : layers.22.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 203 :   29360128 : layers.22.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 204 :   29360128 : layers.22.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 205 :   29360128 : layers.22.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 206 :       8192 : layers.22.attention_norm.weight     : [8192]          : torch.bfloat16
 207 :       8192 : layers.22.ffn_norm.weight           : [8192]          : torch.bfloat16
 208 :    8388608 : layers.23.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 209 :    1048576 : layers.23.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 210 :    1048576 : layers.23.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 211 :    8388608 : layers.23.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 212 :   29360128 : layers.23.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 213 :   29360128 : layers.23.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 214 :   29360128 : layers.23.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 215 :       8192 : layers.23.attention_norm.weight     : [8192]          : torch.bfloat16
 216 :       8192 : layers.23.ffn_norm.weight           : [8192]          : torch.bfloat16
 217 :    8388608 : layers.24.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 218 :    1048576 : layers.24.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 219 :    1048576 : layers.24.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 220 :    8388608 : layers.24.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 221 :   29360128 : layers.24.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 222 :   29360128 : layers.24.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 223 :   29360128 : layers.24.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 224 :       8192 : layers.24.attention_norm.weight     : [8192]          : torch.bfloat16
 225 :       8192 : layers.24.ffn_norm.weight           : [8192]          : torch.bfloat16
 226 :    8388608 : layers.25.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 227 :    1048576 : layers.25.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 228 :    1048576 : layers.25.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 229 :    8388608 : layers.25.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 230 :   29360128 : layers.25.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 231 :   29360128 : layers.25.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 232 :   29360128 : layers.25.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 233 :       8192 : layers.25.attention_norm.weight     : [8192]          : torch.bfloat16
 234 :       8192 : layers.25.ffn_norm.weight           : [8192]          : torch.bfloat16
 235 :    8388608 : layers.26.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 236 :    1048576 : layers.26.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 237 :    1048576 : layers.26.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 238 :    8388608 : layers.26.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 239 :   29360128 : layers.26.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 240 :   29360128 : layers.26.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 241 :   29360128 : layers.26.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 242 :       8192 : layers.26.attention_norm.weight     : [8192]          : torch.bfloat16
 243 :       8192 : layers.26.ffn_norm.weight           : [8192]          : torch.bfloat16
 244 :    8388608 : layers.27.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 245 :    1048576 : layers.27.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 246 :    1048576 : layers.27.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 247 :    8388608 : layers.27.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 248 :   29360128 : layers.27.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 249 :   29360128 : layers.27.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 250 :   29360128 : layers.27.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 251 :       8192 : layers.27.attention_norm.weight     : [8192]          : torch.bfloat16
 252 :       8192 : layers.27.ffn_norm.weight           : [8192]          : torch.bfloat16
 253 :    8388608 : layers.28.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 254 :    1048576 : layers.28.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 255 :    1048576 : layers.28.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 256 :    8388608 : layers.28.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 257 :   29360128 : layers.28.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 258 :   29360128 : layers.28.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 259 :   29360128 : layers.28.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 260 :       8192 : layers.28.attention_norm.weight     : [8192]          : torch.bfloat16
 261 :       8192 : layers.28.ffn_norm.weight           : [8192]          : torch.bfloat16
 262 :    8388608 : layers.29.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 263 :    1048576 : layers.29.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 264 :    1048576 : layers.29.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 265 :    8388608 : layers.29.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 266 :   29360128 : layers.29.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 267 :   29360128 : layers.29.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 268 :   29360128 : layers.29.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 269 :       8192 : layers.29.attention_norm.weight     : [8192]          : torch.bfloat16
 270 :       8192 : layers.29.ffn_norm.weight           : [8192]          : torch.bfloat16
 271 :    8388608 : layers.30.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 272 :    1048576 : layers.30.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 273 :    1048576 : layers.30.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 274 :    8388608 : layers.30.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 275 :   29360128 : layers.30.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 276 :   29360128 : layers.30.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 277 :   29360128 : layers.30.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 278 :       8192 : layers.30.attention_norm.weight     : [8192]          : torch.bfloat16
 279 :       8192 : layers.30.ffn_norm.weight           : [8192]          : torch.bfloat16
 280 :    8388608 : layers.31.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 281 :    1048576 : layers.31.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 282 :    1048576 : layers.31.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 283 :    8388608 : layers.31.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 284 :   29360128 : layers.31.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 285 :   29360128 : layers.31.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 286 :   29360128 : layers.31.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 287 :       8192 : layers.31.attention_norm.weight     : [8192]          : torch.bfloat16
 288 :       8192 : layers.31.ffn_norm.weight           : [8192]          : torch.bfloat16
 289 :    8388608 : layers.32.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 290 :    1048576 : layers.32.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 291 :    1048576 : layers.32.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 292 :    8388608 : layers.32.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 293 :   29360128 : layers.32.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 294 :   29360128 : layers.32.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 295 :   29360128 : layers.32.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 296 :       8192 : layers.32.attention_norm.weight     : [8192]          : torch.bfloat16
 297 :       8192 : layers.32.ffn_norm.weight           : [8192]          : torch.bfloat16
 298 :    8388608 : layers.33.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 299 :    1048576 : layers.33.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 300 :    1048576 : layers.33.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 301 :    8388608 : layers.33.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 302 :   29360128 : layers.33.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 303 :   29360128 : layers.33.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 304 :   29360128 : layers.33.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 305 :       8192 : layers.33.attention_norm.weight     : [8192]          : torch.bfloat16
 306 :       8192 : layers.33.ffn_norm.weight           : [8192]          : torch.bfloat16
 307 :    8388608 : layers.34.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 308 :    1048576 : layers.34.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 309 :    1048576 : layers.34.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 310 :    8388608 : layers.34.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 311 :   29360128 : layers.34.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 312 :   29360128 : layers.34.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 313 :   29360128 : layers.34.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 314 :       8192 : layers.34.attention_norm.weight     : [8192]          : torch.bfloat16
 315 :       8192 : layers.34.ffn_norm.weight           : [8192]          : torch.bfloat16
 316 :    8388608 : layers.35.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 317 :    1048576 : layers.35.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 318 :    1048576 : layers.35.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 319 :    8388608 : layers.35.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 320 :   29360128 : layers.35.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 321 :   29360128 : layers.35.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 322 :   29360128 : layers.35.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 323 :       8192 : layers.35.attention_norm.weight     : [8192]          : torch.bfloat16
 324 :       8192 : layers.35.ffn_norm.weight           : [8192]          : torch.bfloat16
 325 :    8388608 : layers.36.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 326 :    1048576 : layers.36.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 327 :    1048576 : layers.36.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 328 :    8388608 : layers.36.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 329 :   29360128 : layers.36.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 330 :   29360128 : layers.36.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 331 :   29360128 : layers.36.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 332 :       8192 : layers.36.attention_norm.weight     : [8192]          : torch.bfloat16
 333 :       8192 : layers.36.ffn_norm.weight           : [8192]          : torch.bfloat16
 334 :    8388608 : layers.37.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 335 :    1048576 : layers.37.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 336 :    1048576 : layers.37.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 337 :    8388608 : layers.37.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 338 :   29360128 : layers.37.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 339 :   29360128 : layers.37.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 340 :   29360128 : layers.37.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 341 :       8192 : layers.37.attention_norm.weight     : [8192]          : torch.bfloat16
 342 :       8192 : layers.37.ffn_norm.weight           : [8192]          : torch.bfloat16
 343 :    8388608 : layers.38.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 344 :    1048576 : layers.38.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 345 :    1048576 : layers.38.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 346 :    8388608 : layers.38.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 347 :   29360128 : layers.38.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 348 :   29360128 : layers.38.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 349 :   29360128 : layers.38.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 350 :       8192 : layers.38.attention_norm.weight     : [8192]          : torch.bfloat16
 351 :       8192 : layers.38.ffn_norm.weight           : [8192]          : torch.bfloat16
 352 :    8388608 : layers.39.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 353 :    1048576 : layers.39.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 354 :    1048576 : layers.39.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 355 :    8388608 : layers.39.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 356 :   29360128 : layers.39.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 357 :   29360128 : layers.39.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 358 :   29360128 : layers.39.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 359 :       8192 : layers.39.attention_norm.weight     : [8192]          : torch.bfloat16
 360 :       8192 : layers.39.ffn_norm.weight           : [8192]          : torch.bfloat16
 361 :    8388608 : layers.40.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 362 :    1048576 : layers.40.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 363 :    1048576 : layers.40.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 364 :    8388608 : layers.40.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 365 :   29360128 : layers.40.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 366 :   29360128 : layers.40.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 367 :   29360128 : layers.40.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 368 :       8192 : layers.40.attention_norm.weight     : [8192]          : torch.bfloat16
 369 :       8192 : layers.40.ffn_norm.weight           : [8192]          : torch.bfloat16
 370 :    8388608 : layers.41.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 371 :    1048576 : layers.41.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 372 :    1048576 : layers.41.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 373 :    8388608 : layers.41.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 374 :   29360128 : layers.41.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 375 :   29360128 : layers.41.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 376 :   29360128 : layers.41.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 377 :       8192 : layers.41.attention_norm.weight     : [8192]          : torch.bfloat16
 378 :       8192 : layers.41.ffn_norm.weight           : [8192]          : torch.bfloat16
 379 :    8388608 : layers.42.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 380 :    1048576 : layers.42.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 381 :    1048576 : layers.42.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 382 :    8388608 : layers.42.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 383 :   29360128 : layers.42.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 384 :   29360128 : layers.42.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 385 :   29360128 : layers.42.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 386 :       8192 : layers.42.attention_norm.weight     : [8192]          : torch.bfloat16
 387 :       8192 : layers.42.ffn_norm.weight           : [8192]          : torch.bfloat16
 388 :    8388608 : layers.43.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 389 :    1048576 : layers.43.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 390 :    1048576 : layers.43.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 391 :    8388608 : layers.43.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 392 :   29360128 : layers.43.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 393 :   29360128 : layers.43.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 394 :   29360128 : layers.43.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 395 :       8192 : layers.43.attention_norm.weight     : [8192]          : torch.bfloat16
 396 :       8192 : layers.43.ffn_norm.weight           : [8192]          : torch.bfloat16
 397 :    8388608 : layers.44.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 398 :    1048576 : layers.44.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 399 :    1048576 : layers.44.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 400 :    8388608 : layers.44.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 401 :   29360128 : layers.44.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 402 :   29360128 : layers.44.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 403 :   29360128 : layers.44.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 404 :       8192 : layers.44.attention_norm.weight     : [8192]          : torch.bfloat16
 405 :       8192 : layers.44.ffn_norm.weight           : [8192]          : torch.bfloat16
 406 :    8388608 : layers.45.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 407 :    1048576 : layers.45.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 408 :    1048576 : layers.45.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 409 :    8388608 : layers.45.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 410 :   29360128 : layers.45.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 411 :   29360128 : layers.45.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 412 :   29360128 : layers.45.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 413 :       8192 : layers.45.attention_norm.weight     : [8192]          : torch.bfloat16
 414 :       8192 : layers.45.ffn_norm.weight           : [8192]          : torch.bfloat16
 415 :    8388608 : layers.46.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 416 :    1048576 : layers.46.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 417 :    1048576 : layers.46.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 418 :    8388608 : layers.46.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 419 :   29360128 : layers.46.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 420 :   29360128 : layers.46.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 421 :   29360128 : layers.46.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 422 :       8192 : layers.46.attention_norm.weight     : [8192]          : torch.bfloat16
 423 :       8192 : layers.46.ffn_norm.weight           : [8192]          : torch.bfloat16
 424 :    8388608 : layers.47.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 425 :    1048576 : layers.47.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 426 :    1048576 : layers.47.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 427 :    8388608 : layers.47.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 428 :   29360128 : layers.47.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 429 :   29360128 : layers.47.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 430 :   29360128 : layers.47.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 431 :       8192 : layers.47.attention_norm.weight     : [8192]          : torch.bfloat16
 432 :       8192 : layers.47.ffn_norm.weight           : [8192]          : torch.bfloat16
 433 :    8388608 : layers.48.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 434 :    1048576 : layers.48.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 435 :    1048576 : layers.48.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 436 :    8388608 : layers.48.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 437 :   29360128 : layers.48.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 438 :   29360128 : layers.48.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 439 :   29360128 : layers.48.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 440 :       8192 : layers.48.attention_norm.weight     : [8192]          : torch.bfloat16
 441 :       8192 : layers.48.ffn_norm.weight           : [8192]          : torch.bfloat16
 442 :    8388608 : layers.49.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 443 :    1048576 : layers.49.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 444 :    1048576 : layers.49.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 445 :    8388608 : layers.49.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 446 :   29360128 : layers.49.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 447 :   29360128 : layers.49.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 448 :   29360128 : layers.49.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 449 :       8192 : layers.49.attention_norm.weight     : [8192]          : torch.bfloat16
 450 :       8192 : layers.49.ffn_norm.weight           : [8192]          : torch.bfloat16
 451 :    8388608 : layers.50.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 452 :    1048576 : layers.50.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 453 :    1048576 : layers.50.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 454 :    8388608 : layers.50.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 455 :   29360128 : layers.50.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 456 :   29360128 : layers.50.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 457 :   29360128 : layers.50.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 458 :       8192 : layers.50.attention_norm.weight     : [8192]          : torch.bfloat16
 459 :       8192 : layers.50.ffn_norm.weight           : [8192]          : torch.bfloat16
 460 :    8388608 : layers.51.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 461 :    1048576 : layers.51.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 462 :    1048576 : layers.51.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 463 :    8388608 : layers.51.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 464 :   29360128 : layers.51.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 465 :   29360128 : layers.51.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 466 :   29360128 : layers.51.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 467 :       8192 : layers.51.attention_norm.weight     : [8192]          : torch.bfloat16
 468 :       8192 : layers.51.ffn_norm.weight           : [8192]          : torch.bfloat16
 469 :    8388608 : layers.52.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 470 :    1048576 : layers.52.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 471 :    1048576 : layers.52.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 472 :    8388608 : layers.52.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 473 :   29360128 : layers.52.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 474 :   29360128 : layers.52.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 475 :   29360128 : layers.52.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 476 :       8192 : layers.52.attention_norm.weight     : [8192]          : torch.bfloat16
 477 :       8192 : layers.52.ffn_norm.weight           : [8192]          : torch.bfloat16
 478 :    8388608 : layers.53.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 479 :    1048576 : layers.53.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 480 :    1048576 : layers.53.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 481 :    8388608 : layers.53.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 482 :   29360128 : layers.53.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 483 :   29360128 : layers.53.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 484 :   29360128 : layers.53.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 485 :       8192 : layers.53.attention_norm.weight     : [8192]          : torch.bfloat16
 486 :       8192 : layers.53.ffn_norm.weight           : [8192]          : torch.bfloat16
 487 :    8388608 : layers.54.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 488 :    1048576 : layers.54.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 489 :    1048576 : layers.54.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 490 :    8388608 : layers.54.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 491 :   29360128 : layers.54.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 492 :   29360128 : layers.54.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 493 :   29360128 : layers.54.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 494 :       8192 : layers.54.attention_norm.weight     : [8192]          : torch.bfloat16
 495 :       8192 : layers.54.ffn_norm.weight           : [8192]          : torch.bfloat16
 496 :    8388608 : layers.55.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 497 :    1048576 : layers.55.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 498 :    1048576 : layers.55.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 499 :    8388608 : layers.55.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 500 :   29360128 : layers.55.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 501 :   29360128 : layers.55.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 502 :   29360128 : layers.55.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 503 :       8192 : layers.55.attention_norm.weight     : [8192]          : torch.bfloat16
 504 :       8192 : layers.55.ffn_norm.weight           : [8192]          : torch.bfloat16
 505 :    8388608 : layers.56.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 506 :    1048576 : layers.56.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 507 :    1048576 : layers.56.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 508 :    8388608 : layers.56.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 509 :   29360128 : layers.56.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 510 :   29360128 : layers.56.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 511 :   29360128 : layers.56.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 512 :       8192 : layers.56.attention_norm.weight     : [8192]          : torch.bfloat16
 513 :       8192 : layers.56.ffn_norm.weight           : [8192]          : torch.bfloat16
 514 :    8388608 : layers.57.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 515 :    1048576 : layers.57.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 516 :    1048576 : layers.57.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 517 :    8388608 : layers.57.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 518 :   29360128 : layers.57.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 519 :   29360128 : layers.57.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 520 :   29360128 : layers.57.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 521 :       8192 : layers.57.attention_norm.weight     : [8192]          : torch.bfloat16
 522 :       8192 : layers.57.ffn_norm.weight           : [8192]          : torch.bfloat16
 523 :    8388608 : layers.58.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 524 :    1048576 : layers.58.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 525 :    1048576 : layers.58.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 526 :    8388608 : layers.58.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 527 :   29360128 : layers.58.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 528 :   29360128 : layers.58.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 529 :   29360128 : layers.58.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 530 :       8192 : layers.58.attention_norm.weight     : [8192]          : torch.bfloat16
 531 :       8192 : layers.58.ffn_norm.weight           : [8192]          : torch.bfloat16
 532 :    8388608 : layers.59.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 533 :    1048576 : layers.59.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 534 :    1048576 : layers.59.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 535 :    8388608 : layers.59.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 536 :   29360128 : layers.59.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 537 :   29360128 : layers.59.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 538 :   29360128 : layers.59.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 539 :       8192 : layers.59.attention_norm.weight     : [8192]          : torch.bfloat16
 540 :       8192 : layers.59.ffn_norm.weight           : [8192]          : torch.bfloat16
 541 :    8388608 : layers.60.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 542 :    1048576 : layers.60.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 543 :    1048576 : layers.60.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 544 :    8388608 : layers.60.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 545 :   29360128 : layers.60.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 546 :   29360128 : layers.60.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 547 :   29360128 : layers.60.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 548 :       8192 : layers.60.attention_norm.weight     : [8192]          : torch.bfloat16
 549 :       8192 : layers.60.ffn_norm.weight           : [8192]          : torch.bfloat16
 550 :    8388608 : layers.61.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 551 :    1048576 : layers.61.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 552 :    1048576 : layers.61.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 553 :    8388608 : layers.61.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 554 :   29360128 : layers.61.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 555 :   29360128 : layers.61.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 556 :   29360128 : layers.61.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 557 :       8192 : layers.61.attention_norm.weight     : [8192]          : torch.bfloat16
 558 :       8192 : layers.61.ffn_norm.weight           : [8192]          : torch.bfloat16
 559 :    8388608 : layers.62.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 560 :    1048576 : layers.62.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 561 :    1048576 : layers.62.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 562 :    8388608 : layers.62.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 563 :   29360128 : layers.62.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 564 :   29360128 : layers.62.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 565 :   29360128 : layers.62.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 566 :       8192 : layers.62.attention_norm.weight     : [8192]          : torch.bfloat16
 567 :       8192 : layers.62.ffn_norm.weight           : [8192]          : torch.bfloat16
 568 :    8388608 : layers.63.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 569 :    1048576 : layers.63.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 570 :    1048576 : layers.63.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 571 :    8388608 : layers.63.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 572 :   29360128 : layers.63.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 573 :   29360128 : layers.63.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 574 :   29360128 : layers.63.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 575 :       8192 : layers.63.attention_norm.weight     : [8192]          : torch.bfloat16
 576 :       8192 : layers.63.ffn_norm.weight           : [8192]          : torch.bfloat16
 577 :    8388608 : layers.64.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 578 :    1048576 : layers.64.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 579 :    1048576 : layers.64.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 580 :    8388608 : layers.64.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 581 :   29360128 : layers.64.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 582 :   29360128 : layers.64.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 583 :   29360128 : layers.64.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 584 :       8192 : layers.64.attention_norm.weight     : [8192]          : torch.bfloat16
 585 :       8192 : layers.64.ffn_norm.weight           : [8192]          : torch.bfloat16
 586 :    8388608 : layers.65.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 587 :    1048576 : layers.65.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 588 :    1048576 : layers.65.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 589 :    8388608 : layers.65.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 590 :   29360128 : layers.65.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 591 :   29360128 : layers.65.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 592 :   29360128 : layers.65.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 593 :       8192 : layers.65.attention_norm.weight     : [8192]          : torch.bfloat16
 594 :       8192 : layers.65.ffn_norm.weight           : [8192]          : torch.bfloat16
 595 :    8388608 : layers.66.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 596 :    1048576 : layers.66.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 597 :    1048576 : layers.66.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 598 :    8388608 : layers.66.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 599 :   29360128 : layers.66.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 600 :   29360128 : layers.66.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 601 :   29360128 : layers.66.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 602 :       8192 : layers.66.attention_norm.weight     : [8192]          : torch.bfloat16
 603 :       8192 : layers.66.ffn_norm.weight           : [8192]          : torch.bfloat16
 604 :    8388608 : layers.67.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 605 :    1048576 : layers.67.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 606 :    1048576 : layers.67.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 607 :    8388608 : layers.67.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 608 :   29360128 : layers.67.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 609 :   29360128 : layers.67.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 610 :   29360128 : layers.67.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 611 :       8192 : layers.67.attention_norm.weight     : [8192]          : torch.bfloat16
 612 :       8192 : layers.67.ffn_norm.weight           : [8192]          : torch.bfloat16
 613 :    8388608 : layers.68.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 614 :    1048576 : layers.68.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 615 :    1048576 : layers.68.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 616 :    8388608 : layers.68.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 617 :   29360128 : layers.68.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 618 :   29360128 : layers.68.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 619 :   29360128 : layers.68.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 620 :       8192 : layers.68.attention_norm.weight     : [8192]          : torch.bfloat16
 621 :       8192 : layers.68.ffn_norm.weight           : [8192]          : torch.bfloat16
 622 :    8388608 : layers.69.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 623 :    1048576 : layers.69.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 624 :    1048576 : layers.69.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 625 :    8388608 : layers.69.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 626 :   29360128 : layers.69.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 627 :   29360128 : layers.69.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 628 :   29360128 : layers.69.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 629 :       8192 : layers.69.attention_norm.weight     : [8192]          : torch.bfloat16
 630 :       8192 : layers.69.ffn_norm.weight           : [8192]          : torch.bfloat16
 631 :    8388608 : layers.70.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 632 :    1048576 : layers.70.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 633 :    1048576 : layers.70.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 634 :    8388608 : layers.70.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 635 :   29360128 : layers.70.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 636 :   29360128 : layers.70.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 637 :   29360128 : layers.70.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 638 :       8192 : layers.70.attention_norm.weight     : [8192]          : torch.bfloat16
 639 :       8192 : layers.70.ffn_norm.weight           : [8192]          : torch.bfloat16
 640 :    8388608 : layers.71.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 641 :    1048576 : layers.71.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 642 :    1048576 : layers.71.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 643 :    8388608 : layers.71.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 644 :   29360128 : layers.71.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 645 :   29360128 : layers.71.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 646 :   29360128 : layers.71.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 647 :       8192 : layers.71.attention_norm.weight     : [8192]          : torch.bfloat16
 648 :       8192 : layers.71.ffn_norm.weight           : [8192]          : torch.bfloat16
 649 :    8388608 : layers.72.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 650 :    1048576 : layers.72.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 651 :    1048576 : layers.72.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 652 :    8388608 : layers.72.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 653 :   29360128 : layers.72.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 654 :   29360128 : layers.72.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 655 :   29360128 : layers.72.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 656 :       8192 : layers.72.attention_norm.weight     : [8192]          : torch.bfloat16
 657 :       8192 : layers.72.ffn_norm.weight           : [8192]          : torch.bfloat16
 658 :    8388608 : layers.73.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 659 :    1048576 : layers.73.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 660 :    1048576 : layers.73.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 661 :    8388608 : layers.73.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 662 :   29360128 : layers.73.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 663 :   29360128 : layers.73.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 664 :   29360128 : layers.73.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 665 :       8192 : layers.73.attention_norm.weight     : [8192]          : torch.bfloat16
 666 :       8192 : layers.73.ffn_norm.weight           : [8192]          : torch.bfloat16
 667 :    8388608 : layers.74.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 668 :    1048576 : layers.74.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 669 :    1048576 : layers.74.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 670 :    8388608 : layers.74.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 671 :   29360128 : layers.74.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 672 :   29360128 : layers.74.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 673 :   29360128 : layers.74.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 674 :       8192 : layers.74.attention_norm.weight     : [8192]          : torch.bfloat16
 675 :       8192 : layers.74.ffn_norm.weight           : [8192]          : torch.bfloat16
 676 :    8388608 : layers.75.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 677 :    1048576 : layers.75.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 678 :    1048576 : layers.75.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 679 :    8388608 : layers.75.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 680 :   29360128 : layers.75.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 681 :   29360128 : layers.75.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 682 :   29360128 : layers.75.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 683 :       8192 : layers.75.attention_norm.weight     : [8192]          : torch.bfloat16
 684 :       8192 : layers.75.ffn_norm.weight           : [8192]          : torch.bfloat16
 685 :    8388608 : layers.76.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 686 :    1048576 : layers.76.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 687 :    1048576 : layers.76.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 688 :    8388608 : layers.76.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 689 :   29360128 : layers.76.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 690 :   29360128 : layers.76.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 691 :   29360128 : layers.76.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 692 :       8192 : layers.76.attention_norm.weight     : [8192]          : torch.bfloat16
 693 :       8192 : layers.76.ffn_norm.weight           : [8192]          : torch.bfloat16
 694 :    8388608 : layers.77.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 695 :    1048576 : layers.77.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 696 :    1048576 : layers.77.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 697 :    8388608 : layers.77.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 698 :   29360128 : layers.77.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 699 :   29360128 : layers.77.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 700 :   29360128 : layers.77.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 701 :       8192 : layers.77.attention_norm.weight     : [8192]          : torch.bfloat16
 702 :       8192 : layers.77.ffn_norm.weight           : [8192]          : torch.bfloat16
 703 :    8388608 : layers.78.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 704 :    1048576 : layers.78.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 705 :    1048576 : layers.78.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 706 :    8388608 : layers.78.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 707 :   29360128 : layers.78.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 708 :   29360128 : layers.78.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 709 :   29360128 : layers.78.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 710 :       8192 : layers.78.attention_norm.weight     : [8192]          : torch.bfloat16
 711 :       8192 : layers.78.ffn_norm.weight           : [8192]          : torch.bfloat16
 712 :    8388608 : layers.79.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 713 :    1048576 : layers.79.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 714 :    1048576 : layers.79.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 715 :    8388608 : layers.79.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 716 :   29360128 : layers.79.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 717 :   29360128 : layers.79.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 718 :   29360128 : layers.79.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 719 :       8192 : layers.79.attention_norm.weight     : [8192]          : torch.bfloat16
 720 :       8192 : layers.79.ffn_norm.weight           : [8192]          : torch.bfloat16
 721 :       8192 : norm.weight                         : [8192]          : torch.bfloat16
 722 :  131334144 : output.weight                       : [16032, 8192]   : torch.bfloat16

-----------------------------------------------------------------------------
35281469440 params in total.
70562938880 bytes in total.
24.61 sec, 6.07 sec, 2734.13 MB/s
-----------------------------------------------------------------------------

[5/8]: Loading original/consolidated.04.pth
   0 :  131334144 : tok_embeddings.weight               : [16032, 8192]   : torch.bfloat16
   1 :    8388608 : layers.0.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
   2 :    1048576 : layers.0.attention.wk.weight        : [128, 8192]     : torch.bfloat16
   3 :    1048576 : layers.0.attention.wv.weight        : [128, 8192]     : torch.bfloat16
   4 :    8388608 : layers.0.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
   5 :   29360128 : layers.0.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
   6 :   29360128 : layers.0.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
   7 :   29360128 : layers.0.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
   8 :       8192 : layers.0.attention_norm.weight      : [8192]          : torch.bfloat16
   9 :       8192 : layers.0.ffn_norm.weight            : [8192]          : torch.bfloat16
  10 :    8388608 : layers.1.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  11 :    1048576 : layers.1.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  12 :    1048576 : layers.1.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  13 :    8388608 : layers.1.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  14 :   29360128 : layers.1.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  15 :   29360128 : layers.1.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  16 :   29360128 : layers.1.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  17 :       8192 : layers.1.attention_norm.weight      : [8192]          : torch.bfloat16
  18 :       8192 : layers.1.ffn_norm.weight            : [8192]          : torch.bfloat16
  19 :    8388608 : layers.2.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  20 :    1048576 : layers.2.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  21 :    1048576 : layers.2.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  22 :    8388608 : layers.2.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  23 :   29360128 : layers.2.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  24 :   29360128 : layers.2.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  25 :   29360128 : layers.2.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  26 :       8192 : layers.2.attention_norm.weight      : [8192]          : torch.bfloat16
  27 :       8192 : layers.2.ffn_norm.weight            : [8192]          : torch.bfloat16
  28 :    8388608 : layers.3.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  29 :    1048576 : layers.3.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  30 :    1048576 : layers.3.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  31 :    8388608 : layers.3.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  32 :   29360128 : layers.3.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  33 :   29360128 : layers.3.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  34 :   29360128 : layers.3.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  35 :       8192 : layers.3.attention_norm.weight      : [8192]          : torch.bfloat16
  36 :       8192 : layers.3.ffn_norm.weight            : [8192]          : torch.bfloat16
  37 :    8388608 : layers.4.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  38 :    1048576 : layers.4.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  39 :    1048576 : layers.4.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  40 :    8388608 : layers.4.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  41 :   29360128 : layers.4.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  42 :   29360128 : layers.4.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  43 :   29360128 : layers.4.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  44 :       8192 : layers.4.attention_norm.weight      : [8192]          : torch.bfloat16
  45 :       8192 : layers.4.ffn_norm.weight            : [8192]          : torch.bfloat16
  46 :    8388608 : layers.5.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  47 :    1048576 : layers.5.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  48 :    1048576 : layers.5.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  49 :    8388608 : layers.5.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  50 :   29360128 : layers.5.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  51 :   29360128 : layers.5.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  52 :   29360128 : layers.5.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  53 :       8192 : layers.5.attention_norm.weight      : [8192]          : torch.bfloat16
  54 :       8192 : layers.5.ffn_norm.weight            : [8192]          : torch.bfloat16
  55 :    8388608 : layers.6.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  56 :    1048576 : layers.6.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  57 :    1048576 : layers.6.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  58 :    8388608 : layers.6.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  59 :   29360128 : layers.6.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  60 :   29360128 : layers.6.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  61 :   29360128 : layers.6.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  62 :       8192 : layers.6.attention_norm.weight      : [8192]          : torch.bfloat16
  63 :       8192 : layers.6.ffn_norm.weight            : [8192]          : torch.bfloat16
  64 :    8388608 : layers.7.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  65 :    1048576 : layers.7.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  66 :    1048576 : layers.7.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  67 :    8388608 : layers.7.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  68 :   29360128 : layers.7.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  69 :   29360128 : layers.7.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  70 :   29360128 : layers.7.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  71 :       8192 : layers.7.attention_norm.weight      : [8192]          : torch.bfloat16
  72 :       8192 : layers.7.ffn_norm.weight            : [8192]          : torch.bfloat16
  73 :    8388608 : layers.8.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  74 :    1048576 : layers.8.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  75 :    1048576 : layers.8.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  76 :    8388608 : layers.8.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  77 :   29360128 : layers.8.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  78 :   29360128 : layers.8.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  79 :   29360128 : layers.8.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  80 :       8192 : layers.8.attention_norm.weight      : [8192]          : torch.bfloat16
  81 :       8192 : layers.8.ffn_norm.weight            : [8192]          : torch.bfloat16
  82 :    8388608 : layers.9.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  83 :    1048576 : layers.9.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  84 :    1048576 : layers.9.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  85 :    8388608 : layers.9.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  86 :   29360128 : layers.9.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  87 :   29360128 : layers.9.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  88 :   29360128 : layers.9.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  89 :       8192 : layers.9.attention_norm.weight      : [8192]          : torch.bfloat16
  90 :       8192 : layers.9.ffn_norm.weight            : [8192]          : torch.bfloat16
  91 :    8388608 : layers.10.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
  92 :    1048576 : layers.10.attention.wk.weight       : [128, 8192]     : torch.bfloat16
  93 :    1048576 : layers.10.attention.wv.weight       : [128, 8192]     : torch.bfloat16
  94 :    8388608 : layers.10.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
  95 :   29360128 : layers.10.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
  96 :   29360128 : layers.10.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
  97 :   29360128 : layers.10.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
  98 :       8192 : layers.10.attention_norm.weight     : [8192]          : torch.bfloat16
  99 :       8192 : layers.10.ffn_norm.weight           : [8192]          : torch.bfloat16
 100 :    8388608 : layers.11.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 101 :    1048576 : layers.11.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 102 :    1048576 : layers.11.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 103 :    8388608 : layers.11.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 104 :   29360128 : layers.11.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 105 :   29360128 : layers.11.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 106 :   29360128 : layers.11.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 107 :       8192 : layers.11.attention_norm.weight     : [8192]          : torch.bfloat16
 108 :       8192 : layers.11.ffn_norm.weight           : [8192]          : torch.bfloat16
 109 :    8388608 : layers.12.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 110 :    1048576 : layers.12.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 111 :    1048576 : layers.12.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 112 :    8388608 : layers.12.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 113 :   29360128 : layers.12.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 114 :   29360128 : layers.12.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 115 :   29360128 : layers.12.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 116 :       8192 : layers.12.attention_norm.weight     : [8192]          : torch.bfloat16
 117 :       8192 : layers.12.ffn_norm.weight           : [8192]          : torch.bfloat16
 118 :    8388608 : layers.13.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 119 :    1048576 : layers.13.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 120 :    1048576 : layers.13.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 121 :    8388608 : layers.13.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 122 :   29360128 : layers.13.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 123 :   29360128 : layers.13.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 124 :   29360128 : layers.13.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 125 :       8192 : layers.13.attention_norm.weight     : [8192]          : torch.bfloat16
 126 :       8192 : layers.13.ffn_norm.weight           : [8192]          : torch.bfloat16
 127 :    8388608 : layers.14.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 128 :    1048576 : layers.14.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 129 :    1048576 : layers.14.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 130 :    8388608 : layers.14.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 131 :   29360128 : layers.14.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 132 :   29360128 : layers.14.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 133 :   29360128 : layers.14.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 134 :       8192 : layers.14.attention_norm.weight     : [8192]          : torch.bfloat16
 135 :       8192 : layers.14.ffn_norm.weight           : [8192]          : torch.bfloat16
 136 :    8388608 : layers.15.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 137 :    1048576 : layers.15.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 138 :    1048576 : layers.15.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 139 :    8388608 : layers.15.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 140 :   29360128 : layers.15.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 141 :   29360128 : layers.15.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 142 :   29360128 : layers.15.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 143 :       8192 : layers.15.attention_norm.weight     : [8192]          : torch.bfloat16
 144 :       8192 : layers.15.ffn_norm.weight           : [8192]          : torch.bfloat16
 145 :    8388608 : layers.16.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 146 :    1048576 : layers.16.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 147 :    1048576 : layers.16.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 148 :    8388608 : layers.16.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 149 :   29360128 : layers.16.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 150 :   29360128 : layers.16.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 151 :   29360128 : layers.16.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 152 :       8192 : layers.16.attention_norm.weight     : [8192]          : torch.bfloat16
 153 :       8192 : layers.16.ffn_norm.weight           : [8192]          : torch.bfloat16
 154 :    8388608 : layers.17.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 155 :    1048576 : layers.17.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 156 :    1048576 : layers.17.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 157 :    8388608 : layers.17.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 158 :   29360128 : layers.17.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 159 :   29360128 : layers.17.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 160 :   29360128 : layers.17.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 161 :       8192 : layers.17.attention_norm.weight     : [8192]          : torch.bfloat16
 162 :       8192 : layers.17.ffn_norm.weight           : [8192]          : torch.bfloat16
 163 :    8388608 : layers.18.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 164 :    1048576 : layers.18.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 165 :    1048576 : layers.18.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 166 :    8388608 : layers.18.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 167 :   29360128 : layers.18.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 168 :   29360128 : layers.18.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 169 :   29360128 : layers.18.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 170 :       8192 : layers.18.attention_norm.weight     : [8192]          : torch.bfloat16
 171 :       8192 : layers.18.ffn_norm.weight           : [8192]          : torch.bfloat16
 172 :    8388608 : layers.19.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 173 :    1048576 : layers.19.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 174 :    1048576 : layers.19.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 175 :    8388608 : layers.19.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 176 :   29360128 : layers.19.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 177 :   29360128 : layers.19.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 178 :   29360128 : layers.19.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 179 :       8192 : layers.19.attention_norm.weight     : [8192]          : torch.bfloat16
 180 :       8192 : layers.19.ffn_norm.weight           : [8192]          : torch.bfloat16
 181 :    8388608 : layers.20.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 182 :    1048576 : layers.20.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 183 :    1048576 : layers.20.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 184 :    8388608 : layers.20.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 185 :   29360128 : layers.20.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 186 :   29360128 : layers.20.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 187 :   29360128 : layers.20.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 188 :       8192 : layers.20.attention_norm.weight     : [8192]          : torch.bfloat16
 189 :       8192 : layers.20.ffn_norm.weight           : [8192]          : torch.bfloat16
 190 :    8388608 : layers.21.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 191 :    1048576 : layers.21.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 192 :    1048576 : layers.21.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 193 :    8388608 : layers.21.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 194 :   29360128 : layers.21.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 195 :   29360128 : layers.21.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 196 :   29360128 : layers.21.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 197 :       8192 : layers.21.attention_norm.weight     : [8192]          : torch.bfloat16
 198 :       8192 : layers.21.ffn_norm.weight           : [8192]          : torch.bfloat16
 199 :    8388608 : layers.22.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 200 :    1048576 : layers.22.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 201 :    1048576 : layers.22.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 202 :    8388608 : layers.22.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 203 :   29360128 : layers.22.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 204 :   29360128 : layers.22.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 205 :   29360128 : layers.22.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 206 :       8192 : layers.22.attention_norm.weight     : [8192]          : torch.bfloat16
 207 :       8192 : layers.22.ffn_norm.weight           : [8192]          : torch.bfloat16
 208 :    8388608 : layers.23.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 209 :    1048576 : layers.23.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 210 :    1048576 : layers.23.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 211 :    8388608 : layers.23.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 212 :   29360128 : layers.23.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 213 :   29360128 : layers.23.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 214 :   29360128 : layers.23.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 215 :       8192 : layers.23.attention_norm.weight     : [8192]          : torch.bfloat16
 216 :       8192 : layers.23.ffn_norm.weight           : [8192]          : torch.bfloat16
 217 :    8388608 : layers.24.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 218 :    1048576 : layers.24.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 219 :    1048576 : layers.24.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 220 :    8388608 : layers.24.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 221 :   29360128 : layers.24.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 222 :   29360128 : layers.24.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 223 :   29360128 : layers.24.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 224 :       8192 : layers.24.attention_norm.weight     : [8192]          : torch.bfloat16
 225 :       8192 : layers.24.ffn_norm.weight           : [8192]          : torch.bfloat16
 226 :    8388608 : layers.25.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 227 :    1048576 : layers.25.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 228 :    1048576 : layers.25.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 229 :    8388608 : layers.25.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 230 :   29360128 : layers.25.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 231 :   29360128 : layers.25.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 232 :   29360128 : layers.25.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 233 :       8192 : layers.25.attention_norm.weight     : [8192]          : torch.bfloat16
 234 :       8192 : layers.25.ffn_norm.weight           : [8192]          : torch.bfloat16
 235 :    8388608 : layers.26.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 236 :    1048576 : layers.26.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 237 :    1048576 : layers.26.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 238 :    8388608 : layers.26.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 239 :   29360128 : layers.26.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 240 :   29360128 : layers.26.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 241 :   29360128 : layers.26.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 242 :       8192 : layers.26.attention_norm.weight     : [8192]          : torch.bfloat16
 243 :       8192 : layers.26.ffn_norm.weight           : [8192]          : torch.bfloat16
 244 :    8388608 : layers.27.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 245 :    1048576 : layers.27.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 246 :    1048576 : layers.27.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 247 :    8388608 : layers.27.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 248 :   29360128 : layers.27.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 249 :   29360128 : layers.27.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 250 :   29360128 : layers.27.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 251 :       8192 : layers.27.attention_norm.weight     : [8192]          : torch.bfloat16
 252 :       8192 : layers.27.ffn_norm.weight           : [8192]          : torch.bfloat16
 253 :    8388608 : layers.28.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 254 :    1048576 : layers.28.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 255 :    1048576 : layers.28.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 256 :    8388608 : layers.28.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 257 :   29360128 : layers.28.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 258 :   29360128 : layers.28.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 259 :   29360128 : layers.28.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 260 :       8192 : layers.28.attention_norm.weight     : [8192]          : torch.bfloat16
 261 :       8192 : layers.28.ffn_norm.weight           : [8192]          : torch.bfloat16
 262 :    8388608 : layers.29.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 263 :    1048576 : layers.29.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 264 :    1048576 : layers.29.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 265 :    8388608 : layers.29.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 266 :   29360128 : layers.29.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 267 :   29360128 : layers.29.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 268 :   29360128 : layers.29.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 269 :       8192 : layers.29.attention_norm.weight     : [8192]          : torch.bfloat16
 270 :       8192 : layers.29.ffn_norm.weight           : [8192]          : torch.bfloat16
 271 :    8388608 : layers.30.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 272 :    1048576 : layers.30.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 273 :    1048576 : layers.30.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 274 :    8388608 : layers.30.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 275 :   29360128 : layers.30.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 276 :   29360128 : layers.30.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 277 :   29360128 : layers.30.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 278 :       8192 : layers.30.attention_norm.weight     : [8192]          : torch.bfloat16
 279 :       8192 : layers.30.ffn_norm.weight           : [8192]          : torch.bfloat16
 280 :    8388608 : layers.31.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 281 :    1048576 : layers.31.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 282 :    1048576 : layers.31.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 283 :    8388608 : layers.31.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 284 :   29360128 : layers.31.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 285 :   29360128 : layers.31.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 286 :   29360128 : layers.31.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 287 :       8192 : layers.31.attention_norm.weight     : [8192]          : torch.bfloat16
 288 :       8192 : layers.31.ffn_norm.weight           : [8192]          : torch.bfloat16
 289 :    8388608 : layers.32.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 290 :    1048576 : layers.32.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 291 :    1048576 : layers.32.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 292 :    8388608 : layers.32.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 293 :   29360128 : layers.32.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 294 :   29360128 : layers.32.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 295 :   29360128 : layers.32.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 296 :       8192 : layers.32.attention_norm.weight     : [8192]          : torch.bfloat16
 297 :       8192 : layers.32.ffn_norm.weight           : [8192]          : torch.bfloat16
 298 :    8388608 : layers.33.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 299 :    1048576 : layers.33.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 300 :    1048576 : layers.33.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 301 :    8388608 : layers.33.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 302 :   29360128 : layers.33.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 303 :   29360128 : layers.33.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 304 :   29360128 : layers.33.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 305 :       8192 : layers.33.attention_norm.weight     : [8192]          : torch.bfloat16
 306 :       8192 : layers.33.ffn_norm.weight           : [8192]          : torch.bfloat16
 307 :    8388608 : layers.34.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 308 :    1048576 : layers.34.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 309 :    1048576 : layers.34.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 310 :    8388608 : layers.34.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 311 :   29360128 : layers.34.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 312 :   29360128 : layers.34.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 313 :   29360128 : layers.34.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 314 :       8192 : layers.34.attention_norm.weight     : [8192]          : torch.bfloat16
 315 :       8192 : layers.34.ffn_norm.weight           : [8192]          : torch.bfloat16
 316 :    8388608 : layers.35.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 317 :    1048576 : layers.35.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 318 :    1048576 : layers.35.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 319 :    8388608 : layers.35.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 320 :   29360128 : layers.35.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 321 :   29360128 : layers.35.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 322 :   29360128 : layers.35.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 323 :       8192 : layers.35.attention_norm.weight     : [8192]          : torch.bfloat16
 324 :       8192 : layers.35.ffn_norm.weight           : [8192]          : torch.bfloat16
 325 :    8388608 : layers.36.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 326 :    1048576 : layers.36.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 327 :    1048576 : layers.36.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 328 :    8388608 : layers.36.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 329 :   29360128 : layers.36.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 330 :   29360128 : layers.36.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 331 :   29360128 : layers.36.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 332 :       8192 : layers.36.attention_norm.weight     : [8192]          : torch.bfloat16
 333 :       8192 : layers.36.ffn_norm.weight           : [8192]          : torch.bfloat16
 334 :    8388608 : layers.37.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 335 :    1048576 : layers.37.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 336 :    1048576 : layers.37.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 337 :    8388608 : layers.37.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 338 :   29360128 : layers.37.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 339 :   29360128 : layers.37.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 340 :   29360128 : layers.37.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 341 :       8192 : layers.37.attention_norm.weight     : [8192]          : torch.bfloat16
 342 :       8192 : layers.37.ffn_norm.weight           : [8192]          : torch.bfloat16
 343 :    8388608 : layers.38.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 344 :    1048576 : layers.38.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 345 :    1048576 : layers.38.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 346 :    8388608 : layers.38.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 347 :   29360128 : layers.38.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 348 :   29360128 : layers.38.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 349 :   29360128 : layers.38.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 350 :       8192 : layers.38.attention_norm.weight     : [8192]          : torch.bfloat16
 351 :       8192 : layers.38.ffn_norm.weight           : [8192]          : torch.bfloat16
 352 :    8388608 : layers.39.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 353 :    1048576 : layers.39.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 354 :    1048576 : layers.39.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 355 :    8388608 : layers.39.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 356 :   29360128 : layers.39.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 357 :   29360128 : layers.39.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 358 :   29360128 : layers.39.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 359 :       8192 : layers.39.attention_norm.weight     : [8192]          : torch.bfloat16
 360 :       8192 : layers.39.ffn_norm.weight           : [8192]          : torch.bfloat16
 361 :    8388608 : layers.40.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 362 :    1048576 : layers.40.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 363 :    1048576 : layers.40.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 364 :    8388608 : layers.40.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 365 :   29360128 : layers.40.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 366 :   29360128 : layers.40.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 367 :   29360128 : layers.40.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 368 :       8192 : layers.40.attention_norm.weight     : [8192]          : torch.bfloat16
 369 :       8192 : layers.40.ffn_norm.weight           : [8192]          : torch.bfloat16
 370 :    8388608 : layers.41.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 371 :    1048576 : layers.41.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 372 :    1048576 : layers.41.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 373 :    8388608 : layers.41.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 374 :   29360128 : layers.41.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 375 :   29360128 : layers.41.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 376 :   29360128 : layers.41.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 377 :       8192 : layers.41.attention_norm.weight     : [8192]          : torch.bfloat16
 378 :       8192 : layers.41.ffn_norm.weight           : [8192]          : torch.bfloat16
 379 :    8388608 : layers.42.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 380 :    1048576 : layers.42.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 381 :    1048576 : layers.42.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 382 :    8388608 : layers.42.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 383 :   29360128 : layers.42.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 384 :   29360128 : layers.42.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 385 :   29360128 : layers.42.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 386 :       8192 : layers.42.attention_norm.weight     : [8192]          : torch.bfloat16
 387 :       8192 : layers.42.ffn_norm.weight           : [8192]          : torch.bfloat16
 388 :    8388608 : layers.43.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 389 :    1048576 : layers.43.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 390 :    1048576 : layers.43.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 391 :    8388608 : layers.43.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 392 :   29360128 : layers.43.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 393 :   29360128 : layers.43.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 394 :   29360128 : layers.43.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 395 :       8192 : layers.43.attention_norm.weight     : [8192]          : torch.bfloat16
 396 :       8192 : layers.43.ffn_norm.weight           : [8192]          : torch.bfloat16
 397 :    8388608 : layers.44.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 398 :    1048576 : layers.44.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 399 :    1048576 : layers.44.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 400 :    8388608 : layers.44.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 401 :   29360128 : layers.44.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 402 :   29360128 : layers.44.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 403 :   29360128 : layers.44.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 404 :       8192 : layers.44.attention_norm.weight     : [8192]          : torch.bfloat16
 405 :       8192 : layers.44.ffn_norm.weight           : [8192]          : torch.bfloat16
 406 :    8388608 : layers.45.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 407 :    1048576 : layers.45.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 408 :    1048576 : layers.45.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 409 :    8388608 : layers.45.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 410 :   29360128 : layers.45.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 411 :   29360128 : layers.45.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 412 :   29360128 : layers.45.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 413 :       8192 : layers.45.attention_norm.weight     : [8192]          : torch.bfloat16
 414 :       8192 : layers.45.ffn_norm.weight           : [8192]          : torch.bfloat16
 415 :    8388608 : layers.46.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 416 :    1048576 : layers.46.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 417 :    1048576 : layers.46.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 418 :    8388608 : layers.46.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 419 :   29360128 : layers.46.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 420 :   29360128 : layers.46.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 421 :   29360128 : layers.46.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 422 :       8192 : layers.46.attention_norm.weight     : [8192]          : torch.bfloat16
 423 :       8192 : layers.46.ffn_norm.weight           : [8192]          : torch.bfloat16
 424 :    8388608 : layers.47.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 425 :    1048576 : layers.47.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 426 :    1048576 : layers.47.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 427 :    8388608 : layers.47.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 428 :   29360128 : layers.47.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 429 :   29360128 : layers.47.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 430 :   29360128 : layers.47.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 431 :       8192 : layers.47.attention_norm.weight     : [8192]          : torch.bfloat16
 432 :       8192 : layers.47.ffn_norm.weight           : [8192]          : torch.bfloat16
 433 :    8388608 : layers.48.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 434 :    1048576 : layers.48.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 435 :    1048576 : layers.48.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 436 :    8388608 : layers.48.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 437 :   29360128 : layers.48.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 438 :   29360128 : layers.48.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 439 :   29360128 : layers.48.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 440 :       8192 : layers.48.attention_norm.weight     : [8192]          : torch.bfloat16
 441 :       8192 : layers.48.ffn_norm.weight           : [8192]          : torch.bfloat16
 442 :    8388608 : layers.49.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 443 :    1048576 : layers.49.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 444 :    1048576 : layers.49.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 445 :    8388608 : layers.49.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 446 :   29360128 : layers.49.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 447 :   29360128 : layers.49.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 448 :   29360128 : layers.49.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 449 :       8192 : layers.49.attention_norm.weight     : [8192]          : torch.bfloat16
 450 :       8192 : layers.49.ffn_norm.weight           : [8192]          : torch.bfloat16
 451 :    8388608 : layers.50.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 452 :    1048576 : layers.50.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 453 :    1048576 : layers.50.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 454 :    8388608 : layers.50.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 455 :   29360128 : layers.50.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 456 :   29360128 : layers.50.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 457 :   29360128 : layers.50.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 458 :       8192 : layers.50.attention_norm.weight     : [8192]          : torch.bfloat16
 459 :       8192 : layers.50.ffn_norm.weight           : [8192]          : torch.bfloat16
 460 :    8388608 : layers.51.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 461 :    1048576 : layers.51.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 462 :    1048576 : layers.51.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 463 :    8388608 : layers.51.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 464 :   29360128 : layers.51.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 465 :   29360128 : layers.51.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 466 :   29360128 : layers.51.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 467 :       8192 : layers.51.attention_norm.weight     : [8192]          : torch.bfloat16
 468 :       8192 : layers.51.ffn_norm.weight           : [8192]          : torch.bfloat16
 469 :    8388608 : layers.52.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 470 :    1048576 : layers.52.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 471 :    1048576 : layers.52.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 472 :    8388608 : layers.52.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 473 :   29360128 : layers.52.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 474 :   29360128 : layers.52.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 475 :   29360128 : layers.52.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 476 :       8192 : layers.52.attention_norm.weight     : [8192]          : torch.bfloat16
 477 :       8192 : layers.52.ffn_norm.weight           : [8192]          : torch.bfloat16
 478 :    8388608 : layers.53.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 479 :    1048576 : layers.53.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 480 :    1048576 : layers.53.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 481 :    8388608 : layers.53.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 482 :   29360128 : layers.53.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 483 :   29360128 : layers.53.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 484 :   29360128 : layers.53.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 485 :       8192 : layers.53.attention_norm.weight     : [8192]          : torch.bfloat16
 486 :       8192 : layers.53.ffn_norm.weight           : [8192]          : torch.bfloat16
 487 :    8388608 : layers.54.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 488 :    1048576 : layers.54.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 489 :    1048576 : layers.54.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 490 :    8388608 : layers.54.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 491 :   29360128 : layers.54.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 492 :   29360128 : layers.54.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 493 :   29360128 : layers.54.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 494 :       8192 : layers.54.attention_norm.weight     : [8192]          : torch.bfloat16
 495 :       8192 : layers.54.ffn_norm.weight           : [8192]          : torch.bfloat16
 496 :    8388608 : layers.55.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 497 :    1048576 : layers.55.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 498 :    1048576 : layers.55.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 499 :    8388608 : layers.55.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 500 :   29360128 : layers.55.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 501 :   29360128 : layers.55.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 502 :   29360128 : layers.55.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 503 :       8192 : layers.55.attention_norm.weight     : [8192]          : torch.bfloat16
 504 :       8192 : layers.55.ffn_norm.weight           : [8192]          : torch.bfloat16
 505 :    8388608 : layers.56.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 506 :    1048576 : layers.56.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 507 :    1048576 : layers.56.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 508 :    8388608 : layers.56.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 509 :   29360128 : layers.56.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 510 :   29360128 : layers.56.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 511 :   29360128 : layers.56.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 512 :       8192 : layers.56.attention_norm.weight     : [8192]          : torch.bfloat16
 513 :       8192 : layers.56.ffn_norm.weight           : [8192]          : torch.bfloat16
 514 :    8388608 : layers.57.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 515 :    1048576 : layers.57.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 516 :    1048576 : layers.57.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 517 :    8388608 : layers.57.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 518 :   29360128 : layers.57.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 519 :   29360128 : layers.57.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 520 :   29360128 : layers.57.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 521 :       8192 : layers.57.attention_norm.weight     : [8192]          : torch.bfloat16
 522 :       8192 : layers.57.ffn_norm.weight           : [8192]          : torch.bfloat16
 523 :    8388608 : layers.58.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 524 :    1048576 : layers.58.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 525 :    1048576 : layers.58.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 526 :    8388608 : layers.58.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 527 :   29360128 : layers.58.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 528 :   29360128 : layers.58.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 529 :   29360128 : layers.58.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 530 :       8192 : layers.58.attention_norm.weight     : [8192]          : torch.bfloat16
 531 :       8192 : layers.58.ffn_norm.weight           : [8192]          : torch.bfloat16
 532 :    8388608 : layers.59.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 533 :    1048576 : layers.59.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 534 :    1048576 : layers.59.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 535 :    8388608 : layers.59.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 536 :   29360128 : layers.59.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 537 :   29360128 : layers.59.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 538 :   29360128 : layers.59.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 539 :       8192 : layers.59.attention_norm.weight     : [8192]          : torch.bfloat16
 540 :       8192 : layers.59.ffn_norm.weight           : [8192]          : torch.bfloat16
 541 :    8388608 : layers.60.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 542 :    1048576 : layers.60.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 543 :    1048576 : layers.60.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 544 :    8388608 : layers.60.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 545 :   29360128 : layers.60.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 546 :   29360128 : layers.60.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 547 :   29360128 : layers.60.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 548 :       8192 : layers.60.attention_norm.weight     : [8192]          : torch.bfloat16
 549 :       8192 : layers.60.ffn_norm.weight           : [8192]          : torch.bfloat16
 550 :    8388608 : layers.61.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 551 :    1048576 : layers.61.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 552 :    1048576 : layers.61.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 553 :    8388608 : layers.61.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 554 :   29360128 : layers.61.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 555 :   29360128 : layers.61.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 556 :   29360128 : layers.61.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 557 :       8192 : layers.61.attention_norm.weight     : [8192]          : torch.bfloat16
 558 :       8192 : layers.61.ffn_norm.weight           : [8192]          : torch.bfloat16
 559 :    8388608 : layers.62.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 560 :    1048576 : layers.62.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 561 :    1048576 : layers.62.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 562 :    8388608 : layers.62.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 563 :   29360128 : layers.62.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 564 :   29360128 : layers.62.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 565 :   29360128 : layers.62.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 566 :       8192 : layers.62.attention_norm.weight     : [8192]          : torch.bfloat16
 567 :       8192 : layers.62.ffn_norm.weight           : [8192]          : torch.bfloat16
 568 :    8388608 : layers.63.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 569 :    1048576 : layers.63.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 570 :    1048576 : layers.63.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 571 :    8388608 : layers.63.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 572 :   29360128 : layers.63.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 573 :   29360128 : layers.63.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 574 :   29360128 : layers.63.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 575 :       8192 : layers.63.attention_norm.weight     : [8192]          : torch.bfloat16
 576 :       8192 : layers.63.ffn_norm.weight           : [8192]          : torch.bfloat16
 577 :    8388608 : layers.64.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 578 :    1048576 : layers.64.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 579 :    1048576 : layers.64.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 580 :    8388608 : layers.64.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 581 :   29360128 : layers.64.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 582 :   29360128 : layers.64.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 583 :   29360128 : layers.64.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 584 :       8192 : layers.64.attention_norm.weight     : [8192]          : torch.bfloat16
 585 :       8192 : layers.64.ffn_norm.weight           : [8192]          : torch.bfloat16
 586 :    8388608 : layers.65.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 587 :    1048576 : layers.65.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 588 :    1048576 : layers.65.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 589 :    8388608 : layers.65.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 590 :   29360128 : layers.65.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 591 :   29360128 : layers.65.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 592 :   29360128 : layers.65.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 593 :       8192 : layers.65.attention_norm.weight     : [8192]          : torch.bfloat16
 594 :       8192 : layers.65.ffn_norm.weight           : [8192]          : torch.bfloat16
 595 :    8388608 : layers.66.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 596 :    1048576 : layers.66.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 597 :    1048576 : layers.66.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 598 :    8388608 : layers.66.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 599 :   29360128 : layers.66.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 600 :   29360128 : layers.66.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 601 :   29360128 : layers.66.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 602 :       8192 : layers.66.attention_norm.weight     : [8192]          : torch.bfloat16
 603 :       8192 : layers.66.ffn_norm.weight           : [8192]          : torch.bfloat16
 604 :    8388608 : layers.67.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 605 :    1048576 : layers.67.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 606 :    1048576 : layers.67.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 607 :    8388608 : layers.67.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 608 :   29360128 : layers.67.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 609 :   29360128 : layers.67.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 610 :   29360128 : layers.67.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 611 :       8192 : layers.67.attention_norm.weight     : [8192]          : torch.bfloat16
 612 :       8192 : layers.67.ffn_norm.weight           : [8192]          : torch.bfloat16
 613 :    8388608 : layers.68.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 614 :    1048576 : layers.68.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 615 :    1048576 : layers.68.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 616 :    8388608 : layers.68.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 617 :   29360128 : layers.68.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 618 :   29360128 : layers.68.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 619 :   29360128 : layers.68.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 620 :       8192 : layers.68.attention_norm.weight     : [8192]          : torch.bfloat16
 621 :       8192 : layers.68.ffn_norm.weight           : [8192]          : torch.bfloat16
 622 :    8388608 : layers.69.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 623 :    1048576 : layers.69.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 624 :    1048576 : layers.69.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 625 :    8388608 : layers.69.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 626 :   29360128 : layers.69.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 627 :   29360128 : layers.69.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 628 :   29360128 : layers.69.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 629 :       8192 : layers.69.attention_norm.weight     : [8192]          : torch.bfloat16
 630 :       8192 : layers.69.ffn_norm.weight           : [8192]          : torch.bfloat16
 631 :    8388608 : layers.70.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 632 :    1048576 : layers.70.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 633 :    1048576 : layers.70.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 634 :    8388608 : layers.70.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 635 :   29360128 : layers.70.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 636 :   29360128 : layers.70.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 637 :   29360128 : layers.70.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 638 :       8192 : layers.70.attention_norm.weight     : [8192]          : torch.bfloat16
 639 :       8192 : layers.70.ffn_norm.weight           : [8192]          : torch.bfloat16
 640 :    8388608 : layers.71.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 641 :    1048576 : layers.71.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 642 :    1048576 : layers.71.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 643 :    8388608 : layers.71.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 644 :   29360128 : layers.71.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 645 :   29360128 : layers.71.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 646 :   29360128 : layers.71.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 647 :       8192 : layers.71.attention_norm.weight     : [8192]          : torch.bfloat16
 648 :       8192 : layers.71.ffn_norm.weight           : [8192]          : torch.bfloat16
 649 :    8388608 : layers.72.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 650 :    1048576 : layers.72.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 651 :    1048576 : layers.72.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 652 :    8388608 : layers.72.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 653 :   29360128 : layers.72.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 654 :   29360128 : layers.72.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 655 :   29360128 : layers.72.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 656 :       8192 : layers.72.attention_norm.weight     : [8192]          : torch.bfloat16
 657 :       8192 : layers.72.ffn_norm.weight           : [8192]          : torch.bfloat16
 658 :    8388608 : layers.73.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 659 :    1048576 : layers.73.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 660 :    1048576 : layers.73.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 661 :    8388608 : layers.73.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 662 :   29360128 : layers.73.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 663 :   29360128 : layers.73.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 664 :   29360128 : layers.73.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 665 :       8192 : layers.73.attention_norm.weight     : [8192]          : torch.bfloat16
 666 :       8192 : layers.73.ffn_norm.weight           : [8192]          : torch.bfloat16
 667 :    8388608 : layers.74.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 668 :    1048576 : layers.74.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 669 :    1048576 : layers.74.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 670 :    8388608 : layers.74.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 671 :   29360128 : layers.74.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 672 :   29360128 : layers.74.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 673 :   29360128 : layers.74.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 674 :       8192 : layers.74.attention_norm.weight     : [8192]          : torch.bfloat16
 675 :       8192 : layers.74.ffn_norm.weight           : [8192]          : torch.bfloat16
 676 :    8388608 : layers.75.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 677 :    1048576 : layers.75.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 678 :    1048576 : layers.75.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 679 :    8388608 : layers.75.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 680 :   29360128 : layers.75.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 681 :   29360128 : layers.75.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 682 :   29360128 : layers.75.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 683 :       8192 : layers.75.attention_norm.weight     : [8192]          : torch.bfloat16
 684 :       8192 : layers.75.ffn_norm.weight           : [8192]          : torch.bfloat16
 685 :    8388608 : layers.76.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 686 :    1048576 : layers.76.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 687 :    1048576 : layers.76.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 688 :    8388608 : layers.76.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 689 :   29360128 : layers.76.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 690 :   29360128 : layers.76.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 691 :   29360128 : layers.76.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 692 :       8192 : layers.76.attention_norm.weight     : [8192]          : torch.bfloat16
 693 :       8192 : layers.76.ffn_norm.weight           : [8192]          : torch.bfloat16
 694 :    8388608 : layers.77.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 695 :    1048576 : layers.77.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 696 :    1048576 : layers.77.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 697 :    8388608 : layers.77.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 698 :   29360128 : layers.77.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 699 :   29360128 : layers.77.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 700 :   29360128 : layers.77.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 701 :       8192 : layers.77.attention_norm.weight     : [8192]          : torch.bfloat16
 702 :       8192 : layers.77.ffn_norm.weight           : [8192]          : torch.bfloat16
 703 :    8388608 : layers.78.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 704 :    1048576 : layers.78.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 705 :    1048576 : layers.78.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 706 :    8388608 : layers.78.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 707 :   29360128 : layers.78.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 708 :   29360128 : layers.78.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 709 :   29360128 : layers.78.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 710 :       8192 : layers.78.attention_norm.weight     : [8192]          : torch.bfloat16
 711 :       8192 : layers.78.ffn_norm.weight           : [8192]          : torch.bfloat16
 712 :    8388608 : layers.79.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 713 :    1048576 : layers.79.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 714 :    1048576 : layers.79.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 715 :    8388608 : layers.79.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 716 :   29360128 : layers.79.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 717 :   29360128 : layers.79.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 718 :   29360128 : layers.79.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 719 :       8192 : layers.79.attention_norm.weight     : [8192]          : torch.bfloat16
 720 :       8192 : layers.79.ffn_norm.weight           : [8192]          : torch.bfloat16
 721 :       8192 : norm.weight                         : [8192]          : torch.bfloat16
 722 :  131334144 : output.weight                       : [16032, 8192]   : torch.bfloat16

-----------------------------------------------------------------------------
44101836800 params in total.
88203673600 bytes in total.
30.09 sec, 5.48 sec, 2795.71 MB/s
-----------------------------------------------------------------------------

[6/8]: Loading original/consolidated.05.pth
   0 :  131334144 : tok_embeddings.weight               : [16032, 8192]   : torch.bfloat16
   1 :    8388608 : layers.0.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
   2 :    1048576 : layers.0.attention.wk.weight        : [128, 8192]     : torch.bfloat16
   3 :    1048576 : layers.0.attention.wv.weight        : [128, 8192]     : torch.bfloat16
   4 :    8388608 : layers.0.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
   5 :   29360128 : layers.0.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
   6 :   29360128 : layers.0.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
   7 :   29360128 : layers.0.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
   8 :       8192 : layers.0.attention_norm.weight      : [8192]          : torch.bfloat16
   9 :       8192 : layers.0.ffn_norm.weight            : [8192]          : torch.bfloat16
  10 :    8388608 : layers.1.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  11 :    1048576 : layers.1.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  12 :    1048576 : layers.1.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  13 :    8388608 : layers.1.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  14 :   29360128 : layers.1.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  15 :   29360128 : layers.1.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  16 :   29360128 : layers.1.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  17 :       8192 : layers.1.attention_norm.weight      : [8192]          : torch.bfloat16
  18 :       8192 : layers.1.ffn_norm.weight            : [8192]          : torch.bfloat16
  19 :    8388608 : layers.2.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  20 :    1048576 : layers.2.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  21 :    1048576 : layers.2.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  22 :    8388608 : layers.2.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  23 :   29360128 : layers.2.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  24 :   29360128 : layers.2.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  25 :   29360128 : layers.2.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  26 :       8192 : layers.2.attention_norm.weight      : [8192]          : torch.bfloat16
  27 :       8192 : layers.2.ffn_norm.weight            : [8192]          : torch.bfloat16
  28 :    8388608 : layers.3.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  29 :    1048576 : layers.3.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  30 :    1048576 : layers.3.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  31 :    8388608 : layers.3.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  32 :   29360128 : layers.3.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  33 :   29360128 : layers.3.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  34 :   29360128 : layers.3.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  35 :       8192 : layers.3.attention_norm.weight      : [8192]          : torch.bfloat16
  36 :       8192 : layers.3.ffn_norm.weight            : [8192]          : torch.bfloat16
  37 :    8388608 : layers.4.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  38 :    1048576 : layers.4.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  39 :    1048576 : layers.4.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  40 :    8388608 : layers.4.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  41 :   29360128 : layers.4.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  42 :   29360128 : layers.4.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  43 :   29360128 : layers.4.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  44 :       8192 : layers.4.attention_norm.weight      : [8192]          : torch.bfloat16
  45 :       8192 : layers.4.ffn_norm.weight            : [8192]          : torch.bfloat16
  46 :    8388608 : layers.5.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  47 :    1048576 : layers.5.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  48 :    1048576 : layers.5.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  49 :    8388608 : layers.5.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  50 :   29360128 : layers.5.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  51 :   29360128 : layers.5.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  52 :   29360128 : layers.5.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  53 :       8192 : layers.5.attention_norm.weight      : [8192]          : torch.bfloat16
  54 :       8192 : layers.5.ffn_norm.weight            : [8192]          : torch.bfloat16
  55 :    8388608 : layers.6.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  56 :    1048576 : layers.6.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  57 :    1048576 : layers.6.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  58 :    8388608 : layers.6.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  59 :   29360128 : layers.6.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  60 :   29360128 : layers.6.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  61 :   29360128 : layers.6.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  62 :       8192 : layers.6.attention_norm.weight      : [8192]          : torch.bfloat16
  63 :       8192 : layers.6.ffn_norm.weight            : [8192]          : torch.bfloat16
  64 :    8388608 : layers.7.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  65 :    1048576 : layers.7.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  66 :    1048576 : layers.7.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  67 :    8388608 : layers.7.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  68 :   29360128 : layers.7.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  69 :   29360128 : layers.7.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  70 :   29360128 : layers.7.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  71 :       8192 : layers.7.attention_norm.weight      : [8192]          : torch.bfloat16
  72 :       8192 : layers.7.ffn_norm.weight            : [8192]          : torch.bfloat16
  73 :    8388608 : layers.8.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  74 :    1048576 : layers.8.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  75 :    1048576 : layers.8.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  76 :    8388608 : layers.8.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  77 :   29360128 : layers.8.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  78 :   29360128 : layers.8.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  79 :   29360128 : layers.8.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  80 :       8192 : layers.8.attention_norm.weight      : [8192]          : torch.bfloat16
  81 :       8192 : layers.8.ffn_norm.weight            : [8192]          : torch.bfloat16
  82 :    8388608 : layers.9.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  83 :    1048576 : layers.9.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  84 :    1048576 : layers.9.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  85 :    8388608 : layers.9.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  86 :   29360128 : layers.9.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  87 :   29360128 : layers.9.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  88 :   29360128 : layers.9.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  89 :       8192 : layers.9.attention_norm.weight      : [8192]          : torch.bfloat16
  90 :       8192 : layers.9.ffn_norm.weight            : [8192]          : torch.bfloat16
  91 :    8388608 : layers.10.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
  92 :    1048576 : layers.10.attention.wk.weight       : [128, 8192]     : torch.bfloat16
  93 :    1048576 : layers.10.attention.wv.weight       : [128, 8192]     : torch.bfloat16
  94 :    8388608 : layers.10.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
  95 :   29360128 : layers.10.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
  96 :   29360128 : layers.10.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
  97 :   29360128 : layers.10.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
  98 :       8192 : layers.10.attention_norm.weight     : [8192]          : torch.bfloat16
  99 :       8192 : layers.10.ffn_norm.weight           : [8192]          : torch.bfloat16
 100 :    8388608 : layers.11.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 101 :    1048576 : layers.11.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 102 :    1048576 : layers.11.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 103 :    8388608 : layers.11.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 104 :   29360128 : layers.11.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 105 :   29360128 : layers.11.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 106 :   29360128 : layers.11.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 107 :       8192 : layers.11.attention_norm.weight     : [8192]          : torch.bfloat16
 108 :       8192 : layers.11.ffn_norm.weight           : [8192]          : torch.bfloat16
 109 :    8388608 : layers.12.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 110 :    1048576 : layers.12.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 111 :    1048576 : layers.12.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 112 :    8388608 : layers.12.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 113 :   29360128 : layers.12.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 114 :   29360128 : layers.12.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 115 :   29360128 : layers.12.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 116 :       8192 : layers.12.attention_norm.weight     : [8192]          : torch.bfloat16
 117 :       8192 : layers.12.ffn_norm.weight           : [8192]          : torch.bfloat16
 118 :    8388608 : layers.13.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 119 :    1048576 : layers.13.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 120 :    1048576 : layers.13.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 121 :    8388608 : layers.13.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 122 :   29360128 : layers.13.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 123 :   29360128 : layers.13.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 124 :   29360128 : layers.13.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 125 :       8192 : layers.13.attention_norm.weight     : [8192]          : torch.bfloat16
 126 :       8192 : layers.13.ffn_norm.weight           : [8192]          : torch.bfloat16
 127 :    8388608 : layers.14.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 128 :    1048576 : layers.14.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 129 :    1048576 : layers.14.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 130 :    8388608 : layers.14.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 131 :   29360128 : layers.14.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 132 :   29360128 : layers.14.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 133 :   29360128 : layers.14.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 134 :       8192 : layers.14.attention_norm.weight     : [8192]          : torch.bfloat16
 135 :       8192 : layers.14.ffn_norm.weight           : [8192]          : torch.bfloat16
 136 :    8388608 : layers.15.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 137 :    1048576 : layers.15.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 138 :    1048576 : layers.15.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 139 :    8388608 : layers.15.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 140 :   29360128 : layers.15.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 141 :   29360128 : layers.15.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 142 :   29360128 : layers.15.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 143 :       8192 : layers.15.attention_norm.weight     : [8192]          : torch.bfloat16
 144 :       8192 : layers.15.ffn_norm.weight           : [8192]          : torch.bfloat16
 145 :    8388608 : layers.16.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 146 :    1048576 : layers.16.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 147 :    1048576 : layers.16.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 148 :    8388608 : layers.16.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 149 :   29360128 : layers.16.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 150 :   29360128 : layers.16.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 151 :   29360128 : layers.16.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 152 :       8192 : layers.16.attention_norm.weight     : [8192]          : torch.bfloat16
 153 :       8192 : layers.16.ffn_norm.weight           : [8192]          : torch.bfloat16
 154 :    8388608 : layers.17.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 155 :    1048576 : layers.17.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 156 :    1048576 : layers.17.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 157 :    8388608 : layers.17.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 158 :   29360128 : layers.17.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 159 :   29360128 : layers.17.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 160 :   29360128 : layers.17.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 161 :       8192 : layers.17.attention_norm.weight     : [8192]          : torch.bfloat16
 162 :       8192 : layers.17.ffn_norm.weight           : [8192]          : torch.bfloat16
 163 :    8388608 : layers.18.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 164 :    1048576 : layers.18.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 165 :    1048576 : layers.18.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 166 :    8388608 : layers.18.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 167 :   29360128 : layers.18.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 168 :   29360128 : layers.18.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 169 :   29360128 : layers.18.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 170 :       8192 : layers.18.attention_norm.weight     : [8192]          : torch.bfloat16
 171 :       8192 : layers.18.ffn_norm.weight           : [8192]          : torch.bfloat16
 172 :    8388608 : layers.19.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 173 :    1048576 : layers.19.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 174 :    1048576 : layers.19.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 175 :    8388608 : layers.19.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 176 :   29360128 : layers.19.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 177 :   29360128 : layers.19.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 178 :   29360128 : layers.19.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 179 :       8192 : layers.19.attention_norm.weight     : [8192]          : torch.bfloat16
 180 :       8192 : layers.19.ffn_norm.weight           : [8192]          : torch.bfloat16
 181 :    8388608 : layers.20.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 182 :    1048576 : layers.20.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 183 :    1048576 : layers.20.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 184 :    8388608 : layers.20.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 185 :   29360128 : layers.20.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 186 :   29360128 : layers.20.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 187 :   29360128 : layers.20.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 188 :       8192 : layers.20.attention_norm.weight     : [8192]          : torch.bfloat16
 189 :       8192 : layers.20.ffn_norm.weight           : [8192]          : torch.bfloat16
 190 :    8388608 : layers.21.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 191 :    1048576 : layers.21.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 192 :    1048576 : layers.21.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 193 :    8388608 : layers.21.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 194 :   29360128 : layers.21.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 195 :   29360128 : layers.21.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 196 :   29360128 : layers.21.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 197 :       8192 : layers.21.attention_norm.weight     : [8192]          : torch.bfloat16
 198 :       8192 : layers.21.ffn_norm.weight           : [8192]          : torch.bfloat16
 199 :    8388608 : layers.22.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 200 :    1048576 : layers.22.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 201 :    1048576 : layers.22.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 202 :    8388608 : layers.22.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 203 :   29360128 : layers.22.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 204 :   29360128 : layers.22.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 205 :   29360128 : layers.22.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 206 :       8192 : layers.22.attention_norm.weight     : [8192]          : torch.bfloat16
 207 :       8192 : layers.22.ffn_norm.weight           : [8192]          : torch.bfloat16
 208 :    8388608 : layers.23.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 209 :    1048576 : layers.23.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 210 :    1048576 : layers.23.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 211 :    8388608 : layers.23.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 212 :   29360128 : layers.23.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 213 :   29360128 : layers.23.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 214 :   29360128 : layers.23.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 215 :       8192 : layers.23.attention_norm.weight     : [8192]          : torch.bfloat16
 216 :       8192 : layers.23.ffn_norm.weight           : [8192]          : torch.bfloat16
 217 :    8388608 : layers.24.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 218 :    1048576 : layers.24.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 219 :    1048576 : layers.24.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 220 :    8388608 : layers.24.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 221 :   29360128 : layers.24.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 222 :   29360128 : layers.24.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 223 :   29360128 : layers.24.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 224 :       8192 : layers.24.attention_norm.weight     : [8192]          : torch.bfloat16
 225 :       8192 : layers.24.ffn_norm.weight           : [8192]          : torch.bfloat16
 226 :    8388608 : layers.25.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 227 :    1048576 : layers.25.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 228 :    1048576 : layers.25.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 229 :    8388608 : layers.25.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 230 :   29360128 : layers.25.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 231 :   29360128 : layers.25.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 232 :   29360128 : layers.25.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 233 :       8192 : layers.25.attention_norm.weight     : [8192]          : torch.bfloat16
 234 :       8192 : layers.25.ffn_norm.weight           : [8192]          : torch.bfloat16
 235 :    8388608 : layers.26.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 236 :    1048576 : layers.26.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 237 :    1048576 : layers.26.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 238 :    8388608 : layers.26.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 239 :   29360128 : layers.26.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 240 :   29360128 : layers.26.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 241 :   29360128 : layers.26.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 242 :       8192 : layers.26.attention_norm.weight     : [8192]          : torch.bfloat16
 243 :       8192 : layers.26.ffn_norm.weight           : [8192]          : torch.bfloat16
 244 :    8388608 : layers.27.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 245 :    1048576 : layers.27.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 246 :    1048576 : layers.27.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 247 :    8388608 : layers.27.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 248 :   29360128 : layers.27.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 249 :   29360128 : layers.27.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 250 :   29360128 : layers.27.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 251 :       8192 : layers.27.attention_norm.weight     : [8192]          : torch.bfloat16
 252 :       8192 : layers.27.ffn_norm.weight           : [8192]          : torch.bfloat16
 253 :    8388608 : layers.28.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 254 :    1048576 : layers.28.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 255 :    1048576 : layers.28.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 256 :    8388608 : layers.28.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 257 :   29360128 : layers.28.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 258 :   29360128 : layers.28.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 259 :   29360128 : layers.28.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 260 :       8192 : layers.28.attention_norm.weight     : [8192]          : torch.bfloat16
 261 :       8192 : layers.28.ffn_norm.weight           : [8192]          : torch.bfloat16
 262 :    8388608 : layers.29.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 263 :    1048576 : layers.29.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 264 :    1048576 : layers.29.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 265 :    8388608 : layers.29.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 266 :   29360128 : layers.29.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 267 :   29360128 : layers.29.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 268 :   29360128 : layers.29.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 269 :       8192 : layers.29.attention_norm.weight     : [8192]          : torch.bfloat16
 270 :       8192 : layers.29.ffn_norm.weight           : [8192]          : torch.bfloat16
 271 :    8388608 : layers.30.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 272 :    1048576 : layers.30.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 273 :    1048576 : layers.30.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 274 :    8388608 : layers.30.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 275 :   29360128 : layers.30.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 276 :   29360128 : layers.30.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 277 :   29360128 : layers.30.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 278 :       8192 : layers.30.attention_norm.weight     : [8192]          : torch.bfloat16
 279 :       8192 : layers.30.ffn_norm.weight           : [8192]          : torch.bfloat16
 280 :    8388608 : layers.31.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 281 :    1048576 : layers.31.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 282 :    1048576 : layers.31.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 283 :    8388608 : layers.31.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 284 :   29360128 : layers.31.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 285 :   29360128 : layers.31.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 286 :   29360128 : layers.31.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 287 :       8192 : layers.31.attention_norm.weight     : [8192]          : torch.bfloat16
 288 :       8192 : layers.31.ffn_norm.weight           : [8192]          : torch.bfloat16
 289 :    8388608 : layers.32.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 290 :    1048576 : layers.32.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 291 :    1048576 : layers.32.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 292 :    8388608 : layers.32.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 293 :   29360128 : layers.32.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 294 :   29360128 : layers.32.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 295 :   29360128 : layers.32.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 296 :       8192 : layers.32.attention_norm.weight     : [8192]          : torch.bfloat16
 297 :       8192 : layers.32.ffn_norm.weight           : [8192]          : torch.bfloat16
 298 :    8388608 : layers.33.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 299 :    1048576 : layers.33.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 300 :    1048576 : layers.33.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 301 :    8388608 : layers.33.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 302 :   29360128 : layers.33.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 303 :   29360128 : layers.33.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 304 :   29360128 : layers.33.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 305 :       8192 : layers.33.attention_norm.weight     : [8192]          : torch.bfloat16
 306 :       8192 : layers.33.ffn_norm.weight           : [8192]          : torch.bfloat16
 307 :    8388608 : layers.34.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 308 :    1048576 : layers.34.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 309 :    1048576 : layers.34.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 310 :    8388608 : layers.34.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 311 :   29360128 : layers.34.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 312 :   29360128 : layers.34.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 313 :   29360128 : layers.34.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 314 :       8192 : layers.34.attention_norm.weight     : [8192]          : torch.bfloat16
 315 :       8192 : layers.34.ffn_norm.weight           : [8192]          : torch.bfloat16
 316 :    8388608 : layers.35.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 317 :    1048576 : layers.35.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 318 :    1048576 : layers.35.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 319 :    8388608 : layers.35.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 320 :   29360128 : layers.35.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 321 :   29360128 : layers.35.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 322 :   29360128 : layers.35.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 323 :       8192 : layers.35.attention_norm.weight     : [8192]          : torch.bfloat16
 324 :       8192 : layers.35.ffn_norm.weight           : [8192]          : torch.bfloat16
 325 :    8388608 : layers.36.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 326 :    1048576 : layers.36.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 327 :    1048576 : layers.36.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 328 :    8388608 : layers.36.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 329 :   29360128 : layers.36.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 330 :   29360128 : layers.36.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 331 :   29360128 : layers.36.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 332 :       8192 : layers.36.attention_norm.weight     : [8192]          : torch.bfloat16
 333 :       8192 : layers.36.ffn_norm.weight           : [8192]          : torch.bfloat16
 334 :    8388608 : layers.37.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 335 :    1048576 : layers.37.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 336 :    1048576 : layers.37.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 337 :    8388608 : layers.37.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 338 :   29360128 : layers.37.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 339 :   29360128 : layers.37.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 340 :   29360128 : layers.37.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 341 :       8192 : layers.37.attention_norm.weight     : [8192]          : torch.bfloat16
 342 :       8192 : layers.37.ffn_norm.weight           : [8192]          : torch.bfloat16
 343 :    8388608 : layers.38.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 344 :    1048576 : layers.38.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 345 :    1048576 : layers.38.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 346 :    8388608 : layers.38.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 347 :   29360128 : layers.38.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 348 :   29360128 : layers.38.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 349 :   29360128 : layers.38.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 350 :       8192 : layers.38.attention_norm.weight     : [8192]          : torch.bfloat16
 351 :       8192 : layers.38.ffn_norm.weight           : [8192]          : torch.bfloat16
 352 :    8388608 : layers.39.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 353 :    1048576 : layers.39.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 354 :    1048576 : layers.39.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 355 :    8388608 : layers.39.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 356 :   29360128 : layers.39.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 357 :   29360128 : layers.39.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 358 :   29360128 : layers.39.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 359 :       8192 : layers.39.attention_norm.weight     : [8192]          : torch.bfloat16
 360 :       8192 : layers.39.ffn_norm.weight           : [8192]          : torch.bfloat16
 361 :    8388608 : layers.40.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 362 :    1048576 : layers.40.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 363 :    1048576 : layers.40.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 364 :    8388608 : layers.40.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 365 :   29360128 : layers.40.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 366 :   29360128 : layers.40.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 367 :   29360128 : layers.40.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 368 :       8192 : layers.40.attention_norm.weight     : [8192]          : torch.bfloat16
 369 :       8192 : layers.40.ffn_norm.weight           : [8192]          : torch.bfloat16
 370 :    8388608 : layers.41.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 371 :    1048576 : layers.41.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 372 :    1048576 : layers.41.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 373 :    8388608 : layers.41.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 374 :   29360128 : layers.41.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 375 :   29360128 : layers.41.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 376 :   29360128 : layers.41.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 377 :       8192 : layers.41.attention_norm.weight     : [8192]          : torch.bfloat16
 378 :       8192 : layers.41.ffn_norm.weight           : [8192]          : torch.bfloat16
 379 :    8388608 : layers.42.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 380 :    1048576 : layers.42.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 381 :    1048576 : layers.42.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 382 :    8388608 : layers.42.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 383 :   29360128 : layers.42.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 384 :   29360128 : layers.42.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 385 :   29360128 : layers.42.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 386 :       8192 : layers.42.attention_norm.weight     : [8192]          : torch.bfloat16
 387 :       8192 : layers.42.ffn_norm.weight           : [8192]          : torch.bfloat16
 388 :    8388608 : layers.43.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 389 :    1048576 : layers.43.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 390 :    1048576 : layers.43.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 391 :    8388608 : layers.43.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 392 :   29360128 : layers.43.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 393 :   29360128 : layers.43.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 394 :   29360128 : layers.43.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 395 :       8192 : layers.43.attention_norm.weight     : [8192]          : torch.bfloat16
 396 :       8192 : layers.43.ffn_norm.weight           : [8192]          : torch.bfloat16
 397 :    8388608 : layers.44.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 398 :    1048576 : layers.44.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 399 :    1048576 : layers.44.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 400 :    8388608 : layers.44.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 401 :   29360128 : layers.44.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 402 :   29360128 : layers.44.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 403 :   29360128 : layers.44.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 404 :       8192 : layers.44.attention_norm.weight     : [8192]          : torch.bfloat16
 405 :       8192 : layers.44.ffn_norm.weight           : [8192]          : torch.bfloat16
 406 :    8388608 : layers.45.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 407 :    1048576 : layers.45.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 408 :    1048576 : layers.45.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 409 :    8388608 : layers.45.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 410 :   29360128 : layers.45.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 411 :   29360128 : layers.45.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 412 :   29360128 : layers.45.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 413 :       8192 : layers.45.attention_norm.weight     : [8192]          : torch.bfloat16
 414 :       8192 : layers.45.ffn_norm.weight           : [8192]          : torch.bfloat16
 415 :    8388608 : layers.46.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 416 :    1048576 : layers.46.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 417 :    1048576 : layers.46.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 418 :    8388608 : layers.46.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 419 :   29360128 : layers.46.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 420 :   29360128 : layers.46.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 421 :   29360128 : layers.46.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 422 :       8192 : layers.46.attention_norm.weight     : [8192]          : torch.bfloat16
 423 :       8192 : layers.46.ffn_norm.weight           : [8192]          : torch.bfloat16
 424 :    8388608 : layers.47.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 425 :    1048576 : layers.47.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 426 :    1048576 : layers.47.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 427 :    8388608 : layers.47.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 428 :   29360128 : layers.47.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 429 :   29360128 : layers.47.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 430 :   29360128 : layers.47.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 431 :       8192 : layers.47.attention_norm.weight     : [8192]          : torch.bfloat16
 432 :       8192 : layers.47.ffn_norm.weight           : [8192]          : torch.bfloat16
 433 :    8388608 : layers.48.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 434 :    1048576 : layers.48.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 435 :    1048576 : layers.48.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 436 :    8388608 : layers.48.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 437 :   29360128 : layers.48.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 438 :   29360128 : layers.48.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 439 :   29360128 : layers.48.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 440 :       8192 : layers.48.attention_norm.weight     : [8192]          : torch.bfloat16
 441 :       8192 : layers.48.ffn_norm.weight           : [8192]          : torch.bfloat16
 442 :    8388608 : layers.49.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 443 :    1048576 : layers.49.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 444 :    1048576 : layers.49.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 445 :    8388608 : layers.49.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 446 :   29360128 : layers.49.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 447 :   29360128 : layers.49.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 448 :   29360128 : layers.49.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 449 :       8192 : layers.49.attention_norm.weight     : [8192]          : torch.bfloat16
 450 :       8192 : layers.49.ffn_norm.weight           : [8192]          : torch.bfloat16
 451 :    8388608 : layers.50.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 452 :    1048576 : layers.50.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 453 :    1048576 : layers.50.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 454 :    8388608 : layers.50.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 455 :   29360128 : layers.50.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 456 :   29360128 : layers.50.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 457 :   29360128 : layers.50.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 458 :       8192 : layers.50.attention_norm.weight     : [8192]          : torch.bfloat16
 459 :       8192 : layers.50.ffn_norm.weight           : [8192]          : torch.bfloat16
 460 :    8388608 : layers.51.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 461 :    1048576 : layers.51.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 462 :    1048576 : layers.51.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 463 :    8388608 : layers.51.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 464 :   29360128 : layers.51.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 465 :   29360128 : layers.51.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 466 :   29360128 : layers.51.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 467 :       8192 : layers.51.attention_norm.weight     : [8192]          : torch.bfloat16
 468 :       8192 : layers.51.ffn_norm.weight           : [8192]          : torch.bfloat16
 469 :    8388608 : layers.52.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 470 :    1048576 : layers.52.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 471 :    1048576 : layers.52.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 472 :    8388608 : layers.52.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 473 :   29360128 : layers.52.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 474 :   29360128 : layers.52.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 475 :   29360128 : layers.52.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 476 :       8192 : layers.52.attention_norm.weight     : [8192]          : torch.bfloat16
 477 :       8192 : layers.52.ffn_norm.weight           : [8192]          : torch.bfloat16
 478 :    8388608 : layers.53.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 479 :    1048576 : layers.53.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 480 :    1048576 : layers.53.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 481 :    8388608 : layers.53.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 482 :   29360128 : layers.53.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 483 :   29360128 : layers.53.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 484 :   29360128 : layers.53.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 485 :       8192 : layers.53.attention_norm.weight     : [8192]          : torch.bfloat16
 486 :       8192 : layers.53.ffn_norm.weight           : [8192]          : torch.bfloat16
 487 :    8388608 : layers.54.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 488 :    1048576 : layers.54.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 489 :    1048576 : layers.54.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 490 :    8388608 : layers.54.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 491 :   29360128 : layers.54.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 492 :   29360128 : layers.54.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 493 :   29360128 : layers.54.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 494 :       8192 : layers.54.attention_norm.weight     : [8192]          : torch.bfloat16
 495 :       8192 : layers.54.ffn_norm.weight           : [8192]          : torch.bfloat16
 496 :    8388608 : layers.55.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 497 :    1048576 : layers.55.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 498 :    1048576 : layers.55.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 499 :    8388608 : layers.55.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 500 :   29360128 : layers.55.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 501 :   29360128 : layers.55.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 502 :   29360128 : layers.55.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 503 :       8192 : layers.55.attention_norm.weight     : [8192]          : torch.bfloat16
 504 :       8192 : layers.55.ffn_norm.weight           : [8192]          : torch.bfloat16
 505 :    8388608 : layers.56.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 506 :    1048576 : layers.56.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 507 :    1048576 : layers.56.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 508 :    8388608 : layers.56.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 509 :   29360128 : layers.56.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 510 :   29360128 : layers.56.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 511 :   29360128 : layers.56.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 512 :       8192 : layers.56.attention_norm.weight     : [8192]          : torch.bfloat16
 513 :       8192 : layers.56.ffn_norm.weight           : [8192]          : torch.bfloat16
 514 :    8388608 : layers.57.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 515 :    1048576 : layers.57.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 516 :    1048576 : layers.57.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 517 :    8388608 : layers.57.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 518 :   29360128 : layers.57.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 519 :   29360128 : layers.57.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 520 :   29360128 : layers.57.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 521 :       8192 : layers.57.attention_norm.weight     : [8192]          : torch.bfloat16
 522 :       8192 : layers.57.ffn_norm.weight           : [8192]          : torch.bfloat16
 523 :    8388608 : layers.58.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 524 :    1048576 : layers.58.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 525 :    1048576 : layers.58.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 526 :    8388608 : layers.58.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 527 :   29360128 : layers.58.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 528 :   29360128 : layers.58.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 529 :   29360128 : layers.58.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 530 :       8192 : layers.58.attention_norm.weight     : [8192]          : torch.bfloat16
 531 :       8192 : layers.58.ffn_norm.weight           : [8192]          : torch.bfloat16
 532 :    8388608 : layers.59.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 533 :    1048576 : layers.59.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 534 :    1048576 : layers.59.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 535 :    8388608 : layers.59.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 536 :   29360128 : layers.59.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 537 :   29360128 : layers.59.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 538 :   29360128 : layers.59.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 539 :       8192 : layers.59.attention_norm.weight     : [8192]          : torch.bfloat16
 540 :       8192 : layers.59.ffn_norm.weight           : [8192]          : torch.bfloat16
 541 :    8388608 : layers.60.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 542 :    1048576 : layers.60.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 543 :    1048576 : layers.60.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 544 :    8388608 : layers.60.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 545 :   29360128 : layers.60.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 546 :   29360128 : layers.60.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 547 :   29360128 : layers.60.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 548 :       8192 : layers.60.attention_norm.weight     : [8192]          : torch.bfloat16
 549 :       8192 : layers.60.ffn_norm.weight           : [8192]          : torch.bfloat16
 550 :    8388608 : layers.61.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 551 :    1048576 : layers.61.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 552 :    1048576 : layers.61.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 553 :    8388608 : layers.61.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 554 :   29360128 : layers.61.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 555 :   29360128 : layers.61.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 556 :   29360128 : layers.61.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 557 :       8192 : layers.61.attention_norm.weight     : [8192]          : torch.bfloat16
 558 :       8192 : layers.61.ffn_norm.weight           : [8192]          : torch.bfloat16
 559 :    8388608 : layers.62.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 560 :    1048576 : layers.62.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 561 :    1048576 : layers.62.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 562 :    8388608 : layers.62.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 563 :   29360128 : layers.62.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 564 :   29360128 : layers.62.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 565 :   29360128 : layers.62.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 566 :       8192 : layers.62.attention_norm.weight     : [8192]          : torch.bfloat16
 567 :       8192 : layers.62.ffn_norm.weight           : [8192]          : torch.bfloat16
 568 :    8388608 : layers.63.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 569 :    1048576 : layers.63.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 570 :    1048576 : layers.63.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 571 :    8388608 : layers.63.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 572 :   29360128 : layers.63.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 573 :   29360128 : layers.63.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 574 :   29360128 : layers.63.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 575 :       8192 : layers.63.attention_norm.weight     : [8192]          : torch.bfloat16
 576 :       8192 : layers.63.ffn_norm.weight           : [8192]          : torch.bfloat16
 577 :    8388608 : layers.64.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 578 :    1048576 : layers.64.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 579 :    1048576 : layers.64.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 580 :    8388608 : layers.64.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 581 :   29360128 : layers.64.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 582 :   29360128 : layers.64.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 583 :   29360128 : layers.64.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 584 :       8192 : layers.64.attention_norm.weight     : [8192]          : torch.bfloat16
 585 :       8192 : layers.64.ffn_norm.weight           : [8192]          : torch.bfloat16
 586 :    8388608 : layers.65.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 587 :    1048576 : layers.65.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 588 :    1048576 : layers.65.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 589 :    8388608 : layers.65.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 590 :   29360128 : layers.65.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 591 :   29360128 : layers.65.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 592 :   29360128 : layers.65.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 593 :       8192 : layers.65.attention_norm.weight     : [8192]          : torch.bfloat16
 594 :       8192 : layers.65.ffn_norm.weight           : [8192]          : torch.bfloat16
 595 :    8388608 : layers.66.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 596 :    1048576 : layers.66.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 597 :    1048576 : layers.66.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 598 :    8388608 : layers.66.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 599 :   29360128 : layers.66.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 600 :   29360128 : layers.66.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 601 :   29360128 : layers.66.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 602 :       8192 : layers.66.attention_norm.weight     : [8192]          : torch.bfloat16
 603 :       8192 : layers.66.ffn_norm.weight           : [8192]          : torch.bfloat16
 604 :    8388608 : layers.67.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 605 :    1048576 : layers.67.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 606 :    1048576 : layers.67.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 607 :    8388608 : layers.67.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 608 :   29360128 : layers.67.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 609 :   29360128 : layers.67.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 610 :   29360128 : layers.67.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 611 :       8192 : layers.67.attention_norm.weight     : [8192]          : torch.bfloat16
 612 :       8192 : layers.67.ffn_norm.weight           : [8192]          : torch.bfloat16
 613 :    8388608 : layers.68.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 614 :    1048576 : layers.68.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 615 :    1048576 : layers.68.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 616 :    8388608 : layers.68.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 617 :   29360128 : layers.68.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 618 :   29360128 : layers.68.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 619 :   29360128 : layers.68.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 620 :       8192 : layers.68.attention_norm.weight     : [8192]          : torch.bfloat16
 621 :       8192 : layers.68.ffn_norm.weight           : [8192]          : torch.bfloat16
 622 :    8388608 : layers.69.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 623 :    1048576 : layers.69.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 624 :    1048576 : layers.69.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 625 :    8388608 : layers.69.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 626 :   29360128 : layers.69.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 627 :   29360128 : layers.69.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 628 :   29360128 : layers.69.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 629 :       8192 : layers.69.attention_norm.weight     : [8192]          : torch.bfloat16
 630 :       8192 : layers.69.ffn_norm.weight           : [8192]          : torch.bfloat16
 631 :    8388608 : layers.70.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 632 :    1048576 : layers.70.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 633 :    1048576 : layers.70.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 634 :    8388608 : layers.70.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 635 :   29360128 : layers.70.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 636 :   29360128 : layers.70.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 637 :   29360128 : layers.70.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 638 :       8192 : layers.70.attention_norm.weight     : [8192]          : torch.bfloat16
 639 :       8192 : layers.70.ffn_norm.weight           : [8192]          : torch.bfloat16
 640 :    8388608 : layers.71.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 641 :    1048576 : layers.71.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 642 :    1048576 : layers.71.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 643 :    8388608 : layers.71.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 644 :   29360128 : layers.71.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 645 :   29360128 : layers.71.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 646 :   29360128 : layers.71.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 647 :       8192 : layers.71.attention_norm.weight     : [8192]          : torch.bfloat16
 648 :       8192 : layers.71.ffn_norm.weight           : [8192]          : torch.bfloat16
 649 :    8388608 : layers.72.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 650 :    1048576 : layers.72.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 651 :    1048576 : layers.72.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 652 :    8388608 : layers.72.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 653 :   29360128 : layers.72.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 654 :   29360128 : layers.72.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 655 :   29360128 : layers.72.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 656 :       8192 : layers.72.attention_norm.weight     : [8192]          : torch.bfloat16
 657 :       8192 : layers.72.ffn_norm.weight           : [8192]          : torch.bfloat16
 658 :    8388608 : layers.73.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 659 :    1048576 : layers.73.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 660 :    1048576 : layers.73.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 661 :    8388608 : layers.73.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 662 :   29360128 : layers.73.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 663 :   29360128 : layers.73.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 664 :   29360128 : layers.73.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 665 :       8192 : layers.73.attention_norm.weight     : [8192]          : torch.bfloat16
 666 :       8192 : layers.73.ffn_norm.weight           : [8192]          : torch.bfloat16
 667 :    8388608 : layers.74.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 668 :    1048576 : layers.74.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 669 :    1048576 : layers.74.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 670 :    8388608 : layers.74.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 671 :   29360128 : layers.74.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 672 :   29360128 : layers.74.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 673 :   29360128 : layers.74.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 674 :       8192 : layers.74.attention_norm.weight     : [8192]          : torch.bfloat16
 675 :       8192 : layers.74.ffn_norm.weight           : [8192]          : torch.bfloat16
 676 :    8388608 : layers.75.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 677 :    1048576 : layers.75.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 678 :    1048576 : layers.75.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 679 :    8388608 : layers.75.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 680 :   29360128 : layers.75.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 681 :   29360128 : layers.75.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 682 :   29360128 : layers.75.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 683 :       8192 : layers.75.attention_norm.weight     : [8192]          : torch.bfloat16
 684 :       8192 : layers.75.ffn_norm.weight           : [8192]          : torch.bfloat16
 685 :    8388608 : layers.76.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 686 :    1048576 : layers.76.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 687 :    1048576 : layers.76.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 688 :    8388608 : layers.76.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 689 :   29360128 : layers.76.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 690 :   29360128 : layers.76.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 691 :   29360128 : layers.76.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 692 :       8192 : layers.76.attention_norm.weight     : [8192]          : torch.bfloat16
 693 :       8192 : layers.76.ffn_norm.weight           : [8192]          : torch.bfloat16
 694 :    8388608 : layers.77.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 695 :    1048576 : layers.77.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 696 :    1048576 : layers.77.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 697 :    8388608 : layers.77.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 698 :   29360128 : layers.77.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 699 :   29360128 : layers.77.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 700 :   29360128 : layers.77.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 701 :       8192 : layers.77.attention_norm.weight     : [8192]          : torch.bfloat16
 702 :       8192 : layers.77.ffn_norm.weight           : [8192]          : torch.bfloat16
 703 :    8388608 : layers.78.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 704 :    1048576 : layers.78.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 705 :    1048576 : layers.78.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 706 :    8388608 : layers.78.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 707 :   29360128 : layers.78.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 708 :   29360128 : layers.78.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 709 :   29360128 : layers.78.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 710 :       8192 : layers.78.attention_norm.weight     : [8192]          : torch.bfloat16
 711 :       8192 : layers.78.ffn_norm.weight           : [8192]          : torch.bfloat16
 712 :    8388608 : layers.79.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 713 :    1048576 : layers.79.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 714 :    1048576 : layers.79.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 715 :    8388608 : layers.79.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 716 :   29360128 : layers.79.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 717 :   29360128 : layers.79.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 718 :   29360128 : layers.79.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 719 :       8192 : layers.79.attention_norm.weight     : [8192]          : torch.bfloat16
 720 :       8192 : layers.79.ffn_norm.weight           : [8192]          : torch.bfloat16
 721 :       8192 : norm.weight                         : [8192]          : torch.bfloat16
 722 :  131334144 : output.weight                       : [16032, 8192]   : torch.bfloat16

-----------------------------------------------------------------------------
52922204160 params in total.
105844408320 bytes in total.
35.50 sec, 5.41 sec, 2843.42 MB/s
-----------------------------------------------------------------------------

[7/8]: Loading original/consolidated.06.pth
   0 :  131334144 : tok_embeddings.weight               : [16032, 8192]   : torch.bfloat16
   1 :    8388608 : layers.0.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
   2 :    1048576 : layers.0.attention.wk.weight        : [128, 8192]     : torch.bfloat16
   3 :    1048576 : layers.0.attention.wv.weight        : [128, 8192]     : torch.bfloat16
   4 :    8388608 : layers.0.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
   5 :   29360128 : layers.0.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
   6 :   29360128 : layers.0.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
   7 :   29360128 : layers.0.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
   8 :       8192 : layers.0.attention_norm.weight      : [8192]          : torch.bfloat16
   9 :       8192 : layers.0.ffn_norm.weight            : [8192]          : torch.bfloat16
  10 :    8388608 : layers.1.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  11 :    1048576 : layers.1.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  12 :    1048576 : layers.1.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  13 :    8388608 : layers.1.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  14 :   29360128 : layers.1.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  15 :   29360128 : layers.1.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  16 :   29360128 : layers.1.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  17 :       8192 : layers.1.attention_norm.weight      : [8192]          : torch.bfloat16
  18 :       8192 : layers.1.ffn_norm.weight            : [8192]          : torch.bfloat16
  19 :    8388608 : layers.2.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  20 :    1048576 : layers.2.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  21 :    1048576 : layers.2.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  22 :    8388608 : layers.2.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  23 :   29360128 : layers.2.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  24 :   29360128 : layers.2.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  25 :   29360128 : layers.2.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  26 :       8192 : layers.2.attention_norm.weight      : [8192]          : torch.bfloat16
  27 :       8192 : layers.2.ffn_norm.weight            : [8192]          : torch.bfloat16
  28 :    8388608 : layers.3.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  29 :    1048576 : layers.3.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  30 :    1048576 : layers.3.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  31 :    8388608 : layers.3.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  32 :   29360128 : layers.3.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  33 :   29360128 : layers.3.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  34 :   29360128 : layers.3.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  35 :       8192 : layers.3.attention_norm.weight      : [8192]          : torch.bfloat16
  36 :       8192 : layers.3.ffn_norm.weight            : [8192]          : torch.bfloat16
  37 :    8388608 : layers.4.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  38 :    1048576 : layers.4.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  39 :    1048576 : layers.4.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  40 :    8388608 : layers.4.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  41 :   29360128 : layers.4.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  42 :   29360128 : layers.4.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  43 :   29360128 : layers.4.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  44 :       8192 : layers.4.attention_norm.weight      : [8192]          : torch.bfloat16
  45 :       8192 : layers.4.ffn_norm.weight            : [8192]          : torch.bfloat16
  46 :    8388608 : layers.5.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  47 :    1048576 : layers.5.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  48 :    1048576 : layers.5.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  49 :    8388608 : layers.5.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  50 :   29360128 : layers.5.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  51 :   29360128 : layers.5.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  52 :   29360128 : layers.5.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  53 :       8192 : layers.5.attention_norm.weight      : [8192]          : torch.bfloat16
  54 :       8192 : layers.5.ffn_norm.weight            : [8192]          : torch.bfloat16
  55 :    8388608 : layers.6.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  56 :    1048576 : layers.6.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  57 :    1048576 : layers.6.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  58 :    8388608 : layers.6.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  59 :   29360128 : layers.6.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  60 :   29360128 : layers.6.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  61 :   29360128 : layers.6.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  62 :       8192 : layers.6.attention_norm.weight      : [8192]          : torch.bfloat16
  63 :       8192 : layers.6.ffn_norm.weight            : [8192]          : torch.bfloat16
  64 :    8388608 : layers.7.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  65 :    1048576 : layers.7.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  66 :    1048576 : layers.7.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  67 :    8388608 : layers.7.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  68 :   29360128 : layers.7.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  69 :   29360128 : layers.7.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  70 :   29360128 : layers.7.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  71 :       8192 : layers.7.attention_norm.weight      : [8192]          : torch.bfloat16
  72 :       8192 : layers.7.ffn_norm.weight            : [8192]          : torch.bfloat16
  73 :    8388608 : layers.8.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  74 :    1048576 : layers.8.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  75 :    1048576 : layers.8.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  76 :    8388608 : layers.8.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  77 :   29360128 : layers.8.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  78 :   29360128 : layers.8.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  79 :   29360128 : layers.8.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  80 :       8192 : layers.8.attention_norm.weight      : [8192]          : torch.bfloat16
  81 :       8192 : layers.8.ffn_norm.weight            : [8192]          : torch.bfloat16
  82 :    8388608 : layers.9.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  83 :    1048576 : layers.9.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  84 :    1048576 : layers.9.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  85 :    8388608 : layers.9.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  86 :   29360128 : layers.9.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  87 :   29360128 : layers.9.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  88 :   29360128 : layers.9.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  89 :       8192 : layers.9.attention_norm.weight      : [8192]          : torch.bfloat16
  90 :       8192 : layers.9.ffn_norm.weight            : [8192]          : torch.bfloat16
  91 :    8388608 : layers.10.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
  92 :    1048576 : layers.10.attention.wk.weight       : [128, 8192]     : torch.bfloat16
  93 :    1048576 : layers.10.attention.wv.weight       : [128, 8192]     : torch.bfloat16
  94 :    8388608 : layers.10.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
  95 :   29360128 : layers.10.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
  96 :   29360128 : layers.10.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
  97 :   29360128 : layers.10.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
  98 :       8192 : layers.10.attention_norm.weight     : [8192]          : torch.bfloat16
  99 :       8192 : layers.10.ffn_norm.weight           : [8192]          : torch.bfloat16
 100 :    8388608 : layers.11.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 101 :    1048576 : layers.11.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 102 :    1048576 : layers.11.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 103 :    8388608 : layers.11.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 104 :   29360128 : layers.11.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 105 :   29360128 : layers.11.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 106 :   29360128 : layers.11.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 107 :       8192 : layers.11.attention_norm.weight     : [8192]          : torch.bfloat16
 108 :       8192 : layers.11.ffn_norm.weight           : [8192]          : torch.bfloat16
 109 :    8388608 : layers.12.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 110 :    1048576 : layers.12.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 111 :    1048576 : layers.12.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 112 :    8388608 : layers.12.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 113 :   29360128 : layers.12.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 114 :   29360128 : layers.12.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 115 :   29360128 : layers.12.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 116 :       8192 : layers.12.attention_norm.weight     : [8192]          : torch.bfloat16
 117 :       8192 : layers.12.ffn_norm.weight           : [8192]          : torch.bfloat16
 118 :    8388608 : layers.13.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 119 :    1048576 : layers.13.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 120 :    1048576 : layers.13.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 121 :    8388608 : layers.13.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 122 :   29360128 : layers.13.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 123 :   29360128 : layers.13.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 124 :   29360128 : layers.13.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 125 :       8192 : layers.13.attention_norm.weight     : [8192]          : torch.bfloat16
 126 :       8192 : layers.13.ffn_norm.weight           : [8192]          : torch.bfloat16
 127 :    8388608 : layers.14.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 128 :    1048576 : layers.14.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 129 :    1048576 : layers.14.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 130 :    8388608 : layers.14.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 131 :   29360128 : layers.14.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 132 :   29360128 : layers.14.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 133 :   29360128 : layers.14.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 134 :       8192 : layers.14.attention_norm.weight     : [8192]          : torch.bfloat16
 135 :       8192 : layers.14.ffn_norm.weight           : [8192]          : torch.bfloat16
 136 :    8388608 : layers.15.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 137 :    1048576 : layers.15.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 138 :    1048576 : layers.15.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 139 :    8388608 : layers.15.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 140 :   29360128 : layers.15.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 141 :   29360128 : layers.15.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 142 :   29360128 : layers.15.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 143 :       8192 : layers.15.attention_norm.weight     : [8192]          : torch.bfloat16
 144 :       8192 : layers.15.ffn_norm.weight           : [8192]          : torch.bfloat16
 145 :    8388608 : layers.16.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 146 :    1048576 : layers.16.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 147 :    1048576 : layers.16.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 148 :    8388608 : layers.16.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 149 :   29360128 : layers.16.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 150 :   29360128 : layers.16.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 151 :   29360128 : layers.16.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 152 :       8192 : layers.16.attention_norm.weight     : [8192]          : torch.bfloat16
 153 :       8192 : layers.16.ffn_norm.weight           : [8192]          : torch.bfloat16
 154 :    8388608 : layers.17.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 155 :    1048576 : layers.17.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 156 :    1048576 : layers.17.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 157 :    8388608 : layers.17.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 158 :   29360128 : layers.17.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 159 :   29360128 : layers.17.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 160 :   29360128 : layers.17.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 161 :       8192 : layers.17.attention_norm.weight     : [8192]          : torch.bfloat16
 162 :       8192 : layers.17.ffn_norm.weight           : [8192]          : torch.bfloat16
 163 :    8388608 : layers.18.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 164 :    1048576 : layers.18.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 165 :    1048576 : layers.18.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 166 :    8388608 : layers.18.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 167 :   29360128 : layers.18.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 168 :   29360128 : layers.18.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 169 :   29360128 : layers.18.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 170 :       8192 : layers.18.attention_norm.weight     : [8192]          : torch.bfloat16
 171 :       8192 : layers.18.ffn_norm.weight           : [8192]          : torch.bfloat16
 172 :    8388608 : layers.19.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 173 :    1048576 : layers.19.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 174 :    1048576 : layers.19.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 175 :    8388608 : layers.19.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 176 :   29360128 : layers.19.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 177 :   29360128 : layers.19.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 178 :   29360128 : layers.19.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 179 :       8192 : layers.19.attention_norm.weight     : [8192]          : torch.bfloat16
 180 :       8192 : layers.19.ffn_norm.weight           : [8192]          : torch.bfloat16
 181 :    8388608 : layers.20.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 182 :    1048576 : layers.20.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 183 :    1048576 : layers.20.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 184 :    8388608 : layers.20.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 185 :   29360128 : layers.20.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 186 :   29360128 : layers.20.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 187 :   29360128 : layers.20.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 188 :       8192 : layers.20.attention_norm.weight     : [8192]          : torch.bfloat16
 189 :       8192 : layers.20.ffn_norm.weight           : [8192]          : torch.bfloat16
 190 :    8388608 : layers.21.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 191 :    1048576 : layers.21.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 192 :    1048576 : layers.21.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 193 :    8388608 : layers.21.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 194 :   29360128 : layers.21.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 195 :   29360128 : layers.21.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 196 :   29360128 : layers.21.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 197 :       8192 : layers.21.attention_norm.weight     : [8192]          : torch.bfloat16
 198 :       8192 : layers.21.ffn_norm.weight           : [8192]          : torch.bfloat16
 199 :    8388608 : layers.22.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 200 :    1048576 : layers.22.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 201 :    1048576 : layers.22.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 202 :    8388608 : layers.22.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 203 :   29360128 : layers.22.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 204 :   29360128 : layers.22.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 205 :   29360128 : layers.22.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 206 :       8192 : layers.22.attention_norm.weight     : [8192]          : torch.bfloat16
 207 :       8192 : layers.22.ffn_norm.weight           : [8192]          : torch.bfloat16
 208 :    8388608 : layers.23.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 209 :    1048576 : layers.23.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 210 :    1048576 : layers.23.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 211 :    8388608 : layers.23.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 212 :   29360128 : layers.23.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 213 :   29360128 : layers.23.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 214 :   29360128 : layers.23.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 215 :       8192 : layers.23.attention_norm.weight     : [8192]          : torch.bfloat16
 216 :       8192 : layers.23.ffn_norm.weight           : [8192]          : torch.bfloat16
 217 :    8388608 : layers.24.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 218 :    1048576 : layers.24.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 219 :    1048576 : layers.24.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 220 :    8388608 : layers.24.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 221 :   29360128 : layers.24.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 222 :   29360128 : layers.24.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 223 :   29360128 : layers.24.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 224 :       8192 : layers.24.attention_norm.weight     : [8192]          : torch.bfloat16
 225 :       8192 : layers.24.ffn_norm.weight           : [8192]          : torch.bfloat16
 226 :    8388608 : layers.25.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 227 :    1048576 : layers.25.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 228 :    1048576 : layers.25.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 229 :    8388608 : layers.25.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 230 :   29360128 : layers.25.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 231 :   29360128 : layers.25.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 232 :   29360128 : layers.25.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 233 :       8192 : layers.25.attention_norm.weight     : [8192]          : torch.bfloat16
 234 :       8192 : layers.25.ffn_norm.weight           : [8192]          : torch.bfloat16
 235 :    8388608 : layers.26.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 236 :    1048576 : layers.26.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 237 :    1048576 : layers.26.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 238 :    8388608 : layers.26.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 239 :   29360128 : layers.26.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 240 :   29360128 : layers.26.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 241 :   29360128 : layers.26.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 242 :       8192 : layers.26.attention_norm.weight     : [8192]          : torch.bfloat16
 243 :       8192 : layers.26.ffn_norm.weight           : [8192]          : torch.bfloat16
 244 :    8388608 : layers.27.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 245 :    1048576 : layers.27.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 246 :    1048576 : layers.27.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 247 :    8388608 : layers.27.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 248 :   29360128 : layers.27.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 249 :   29360128 : layers.27.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 250 :   29360128 : layers.27.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 251 :       8192 : layers.27.attention_norm.weight     : [8192]          : torch.bfloat16
 252 :       8192 : layers.27.ffn_norm.weight           : [8192]          : torch.bfloat16
 253 :    8388608 : layers.28.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 254 :    1048576 : layers.28.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 255 :    1048576 : layers.28.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 256 :    8388608 : layers.28.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 257 :   29360128 : layers.28.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 258 :   29360128 : layers.28.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 259 :   29360128 : layers.28.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 260 :       8192 : layers.28.attention_norm.weight     : [8192]          : torch.bfloat16
 261 :       8192 : layers.28.ffn_norm.weight           : [8192]          : torch.bfloat16
 262 :    8388608 : layers.29.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 263 :    1048576 : layers.29.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 264 :    1048576 : layers.29.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 265 :    8388608 : layers.29.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 266 :   29360128 : layers.29.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 267 :   29360128 : layers.29.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 268 :   29360128 : layers.29.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 269 :       8192 : layers.29.attention_norm.weight     : [8192]          : torch.bfloat16
 270 :       8192 : layers.29.ffn_norm.weight           : [8192]          : torch.bfloat16
 271 :    8388608 : layers.30.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 272 :    1048576 : layers.30.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 273 :    1048576 : layers.30.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 274 :    8388608 : layers.30.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 275 :   29360128 : layers.30.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 276 :   29360128 : layers.30.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 277 :   29360128 : layers.30.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 278 :       8192 : layers.30.attention_norm.weight     : [8192]          : torch.bfloat16
 279 :       8192 : layers.30.ffn_norm.weight           : [8192]          : torch.bfloat16
 280 :    8388608 : layers.31.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 281 :    1048576 : layers.31.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 282 :    1048576 : layers.31.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 283 :    8388608 : layers.31.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 284 :   29360128 : layers.31.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 285 :   29360128 : layers.31.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 286 :   29360128 : layers.31.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 287 :       8192 : layers.31.attention_norm.weight     : [8192]          : torch.bfloat16
 288 :       8192 : layers.31.ffn_norm.weight           : [8192]          : torch.bfloat16
 289 :    8388608 : layers.32.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 290 :    1048576 : layers.32.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 291 :    1048576 : layers.32.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 292 :    8388608 : layers.32.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 293 :   29360128 : layers.32.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 294 :   29360128 : layers.32.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 295 :   29360128 : layers.32.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 296 :       8192 : layers.32.attention_norm.weight     : [8192]          : torch.bfloat16
 297 :       8192 : layers.32.ffn_norm.weight           : [8192]          : torch.bfloat16
 298 :    8388608 : layers.33.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 299 :    1048576 : layers.33.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 300 :    1048576 : layers.33.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 301 :    8388608 : layers.33.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 302 :   29360128 : layers.33.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 303 :   29360128 : layers.33.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 304 :   29360128 : layers.33.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 305 :       8192 : layers.33.attention_norm.weight     : [8192]          : torch.bfloat16
 306 :       8192 : layers.33.ffn_norm.weight           : [8192]          : torch.bfloat16
 307 :    8388608 : layers.34.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 308 :    1048576 : layers.34.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 309 :    1048576 : layers.34.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 310 :    8388608 : layers.34.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 311 :   29360128 : layers.34.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 312 :   29360128 : layers.34.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 313 :   29360128 : layers.34.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 314 :       8192 : layers.34.attention_norm.weight     : [8192]          : torch.bfloat16
 315 :       8192 : layers.34.ffn_norm.weight           : [8192]          : torch.bfloat16
 316 :    8388608 : layers.35.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 317 :    1048576 : layers.35.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 318 :    1048576 : layers.35.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 319 :    8388608 : layers.35.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 320 :   29360128 : layers.35.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 321 :   29360128 : layers.35.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 322 :   29360128 : layers.35.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 323 :       8192 : layers.35.attention_norm.weight     : [8192]          : torch.bfloat16
 324 :       8192 : layers.35.ffn_norm.weight           : [8192]          : torch.bfloat16
 325 :    8388608 : layers.36.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 326 :    1048576 : layers.36.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 327 :    1048576 : layers.36.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 328 :    8388608 : layers.36.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 329 :   29360128 : layers.36.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 330 :   29360128 : layers.36.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 331 :   29360128 : layers.36.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 332 :       8192 : layers.36.attention_norm.weight     : [8192]          : torch.bfloat16
 333 :       8192 : layers.36.ffn_norm.weight           : [8192]          : torch.bfloat16
 334 :    8388608 : layers.37.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 335 :    1048576 : layers.37.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 336 :    1048576 : layers.37.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 337 :    8388608 : layers.37.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 338 :   29360128 : layers.37.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 339 :   29360128 : layers.37.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 340 :   29360128 : layers.37.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 341 :       8192 : layers.37.attention_norm.weight     : [8192]          : torch.bfloat16
 342 :       8192 : layers.37.ffn_norm.weight           : [8192]          : torch.bfloat16
 343 :    8388608 : layers.38.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 344 :    1048576 : layers.38.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 345 :    1048576 : layers.38.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 346 :    8388608 : layers.38.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 347 :   29360128 : layers.38.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 348 :   29360128 : layers.38.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 349 :   29360128 : layers.38.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 350 :       8192 : layers.38.attention_norm.weight     : [8192]          : torch.bfloat16
 351 :       8192 : layers.38.ffn_norm.weight           : [8192]          : torch.bfloat16
 352 :    8388608 : layers.39.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 353 :    1048576 : layers.39.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 354 :    1048576 : layers.39.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 355 :    8388608 : layers.39.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 356 :   29360128 : layers.39.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 357 :   29360128 : layers.39.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 358 :   29360128 : layers.39.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 359 :       8192 : layers.39.attention_norm.weight     : [8192]          : torch.bfloat16
 360 :       8192 : layers.39.ffn_norm.weight           : [8192]          : torch.bfloat16
 361 :    8388608 : layers.40.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 362 :    1048576 : layers.40.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 363 :    1048576 : layers.40.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 364 :    8388608 : layers.40.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 365 :   29360128 : layers.40.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 366 :   29360128 : layers.40.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 367 :   29360128 : layers.40.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 368 :       8192 : layers.40.attention_norm.weight     : [8192]          : torch.bfloat16
 369 :       8192 : layers.40.ffn_norm.weight           : [8192]          : torch.bfloat16
 370 :    8388608 : layers.41.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 371 :    1048576 : layers.41.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 372 :    1048576 : layers.41.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 373 :    8388608 : layers.41.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 374 :   29360128 : layers.41.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 375 :   29360128 : layers.41.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 376 :   29360128 : layers.41.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 377 :       8192 : layers.41.attention_norm.weight     : [8192]          : torch.bfloat16
 378 :       8192 : layers.41.ffn_norm.weight           : [8192]          : torch.bfloat16
 379 :    8388608 : layers.42.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 380 :    1048576 : layers.42.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 381 :    1048576 : layers.42.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 382 :    8388608 : layers.42.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 383 :   29360128 : layers.42.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 384 :   29360128 : layers.42.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 385 :   29360128 : layers.42.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 386 :       8192 : layers.42.attention_norm.weight     : [8192]          : torch.bfloat16
 387 :       8192 : layers.42.ffn_norm.weight           : [8192]          : torch.bfloat16
 388 :    8388608 : layers.43.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 389 :    1048576 : layers.43.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 390 :    1048576 : layers.43.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 391 :    8388608 : layers.43.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 392 :   29360128 : layers.43.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 393 :   29360128 : layers.43.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 394 :   29360128 : layers.43.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 395 :       8192 : layers.43.attention_norm.weight     : [8192]          : torch.bfloat16
 396 :       8192 : layers.43.ffn_norm.weight           : [8192]          : torch.bfloat16
 397 :    8388608 : layers.44.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 398 :    1048576 : layers.44.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 399 :    1048576 : layers.44.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 400 :    8388608 : layers.44.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 401 :   29360128 : layers.44.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 402 :   29360128 : layers.44.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 403 :   29360128 : layers.44.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 404 :       8192 : layers.44.attention_norm.weight     : [8192]          : torch.bfloat16
 405 :       8192 : layers.44.ffn_norm.weight           : [8192]          : torch.bfloat16
 406 :    8388608 : layers.45.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 407 :    1048576 : layers.45.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 408 :    1048576 : layers.45.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 409 :    8388608 : layers.45.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 410 :   29360128 : layers.45.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 411 :   29360128 : layers.45.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 412 :   29360128 : layers.45.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 413 :       8192 : layers.45.attention_norm.weight     : [8192]          : torch.bfloat16
 414 :       8192 : layers.45.ffn_norm.weight           : [8192]          : torch.bfloat16
 415 :    8388608 : layers.46.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 416 :    1048576 : layers.46.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 417 :    1048576 : layers.46.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 418 :    8388608 : layers.46.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 419 :   29360128 : layers.46.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 420 :   29360128 : layers.46.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 421 :   29360128 : layers.46.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 422 :       8192 : layers.46.attention_norm.weight     : [8192]          : torch.bfloat16
 423 :       8192 : layers.46.ffn_norm.weight           : [8192]          : torch.bfloat16
 424 :    8388608 : layers.47.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 425 :    1048576 : layers.47.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 426 :    1048576 : layers.47.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 427 :    8388608 : layers.47.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 428 :   29360128 : layers.47.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 429 :   29360128 : layers.47.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 430 :   29360128 : layers.47.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 431 :       8192 : layers.47.attention_norm.weight     : [8192]          : torch.bfloat16
 432 :       8192 : layers.47.ffn_norm.weight           : [8192]          : torch.bfloat16
 433 :    8388608 : layers.48.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 434 :    1048576 : layers.48.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 435 :    1048576 : layers.48.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 436 :    8388608 : layers.48.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 437 :   29360128 : layers.48.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 438 :   29360128 : layers.48.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 439 :   29360128 : layers.48.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 440 :       8192 : layers.48.attention_norm.weight     : [8192]          : torch.bfloat16
 441 :       8192 : layers.48.ffn_norm.weight           : [8192]          : torch.bfloat16
 442 :    8388608 : layers.49.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 443 :    1048576 : layers.49.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 444 :    1048576 : layers.49.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 445 :    8388608 : layers.49.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 446 :   29360128 : layers.49.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 447 :   29360128 : layers.49.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 448 :   29360128 : layers.49.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 449 :       8192 : layers.49.attention_norm.weight     : [8192]          : torch.bfloat16
 450 :       8192 : layers.49.ffn_norm.weight           : [8192]          : torch.bfloat16
 451 :    8388608 : layers.50.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 452 :    1048576 : layers.50.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 453 :    1048576 : layers.50.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 454 :    8388608 : layers.50.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 455 :   29360128 : layers.50.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 456 :   29360128 : layers.50.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 457 :   29360128 : layers.50.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 458 :       8192 : layers.50.attention_norm.weight     : [8192]          : torch.bfloat16
 459 :       8192 : layers.50.ffn_norm.weight           : [8192]          : torch.bfloat16
 460 :    8388608 : layers.51.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 461 :    1048576 : layers.51.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 462 :    1048576 : layers.51.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 463 :    8388608 : layers.51.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 464 :   29360128 : layers.51.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 465 :   29360128 : layers.51.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 466 :   29360128 : layers.51.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 467 :       8192 : layers.51.attention_norm.weight     : [8192]          : torch.bfloat16
 468 :       8192 : layers.51.ffn_norm.weight           : [8192]          : torch.bfloat16
 469 :    8388608 : layers.52.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 470 :    1048576 : layers.52.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 471 :    1048576 : layers.52.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 472 :    8388608 : layers.52.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 473 :   29360128 : layers.52.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 474 :   29360128 : layers.52.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 475 :   29360128 : layers.52.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 476 :       8192 : layers.52.attention_norm.weight     : [8192]          : torch.bfloat16
 477 :       8192 : layers.52.ffn_norm.weight           : [8192]          : torch.bfloat16
 478 :    8388608 : layers.53.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 479 :    1048576 : layers.53.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 480 :    1048576 : layers.53.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 481 :    8388608 : layers.53.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 482 :   29360128 : layers.53.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 483 :   29360128 : layers.53.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 484 :   29360128 : layers.53.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 485 :       8192 : layers.53.attention_norm.weight     : [8192]          : torch.bfloat16
 486 :       8192 : layers.53.ffn_norm.weight           : [8192]          : torch.bfloat16
 487 :    8388608 : layers.54.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 488 :    1048576 : layers.54.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 489 :    1048576 : layers.54.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 490 :    8388608 : layers.54.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 491 :   29360128 : layers.54.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 492 :   29360128 : layers.54.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 493 :   29360128 : layers.54.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 494 :       8192 : layers.54.attention_norm.weight     : [8192]          : torch.bfloat16
 495 :       8192 : layers.54.ffn_norm.weight           : [8192]          : torch.bfloat16
 496 :    8388608 : layers.55.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 497 :    1048576 : layers.55.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 498 :    1048576 : layers.55.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 499 :    8388608 : layers.55.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 500 :   29360128 : layers.55.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 501 :   29360128 : layers.55.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 502 :   29360128 : layers.55.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 503 :       8192 : layers.55.attention_norm.weight     : [8192]          : torch.bfloat16
 504 :       8192 : layers.55.ffn_norm.weight           : [8192]          : torch.bfloat16
 505 :    8388608 : layers.56.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 506 :    1048576 : layers.56.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 507 :    1048576 : layers.56.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 508 :    8388608 : layers.56.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 509 :   29360128 : layers.56.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 510 :   29360128 : layers.56.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 511 :   29360128 : layers.56.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 512 :       8192 : layers.56.attention_norm.weight     : [8192]          : torch.bfloat16
 513 :       8192 : layers.56.ffn_norm.weight           : [8192]          : torch.bfloat16
 514 :    8388608 : layers.57.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 515 :    1048576 : layers.57.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 516 :    1048576 : layers.57.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 517 :    8388608 : layers.57.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 518 :   29360128 : layers.57.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 519 :   29360128 : layers.57.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 520 :   29360128 : layers.57.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 521 :       8192 : layers.57.attention_norm.weight     : [8192]          : torch.bfloat16
 522 :       8192 : layers.57.ffn_norm.weight           : [8192]          : torch.bfloat16
 523 :    8388608 : layers.58.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 524 :    1048576 : layers.58.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 525 :    1048576 : layers.58.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 526 :    8388608 : layers.58.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 527 :   29360128 : layers.58.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 528 :   29360128 : layers.58.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 529 :   29360128 : layers.58.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 530 :       8192 : layers.58.attention_norm.weight     : [8192]          : torch.bfloat16
 531 :       8192 : layers.58.ffn_norm.weight           : [8192]          : torch.bfloat16
 532 :    8388608 : layers.59.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 533 :    1048576 : layers.59.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 534 :    1048576 : layers.59.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 535 :    8388608 : layers.59.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 536 :   29360128 : layers.59.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 537 :   29360128 : layers.59.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 538 :   29360128 : layers.59.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 539 :       8192 : layers.59.attention_norm.weight     : [8192]          : torch.bfloat16
 540 :       8192 : layers.59.ffn_norm.weight           : [8192]          : torch.bfloat16
 541 :    8388608 : layers.60.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 542 :    1048576 : layers.60.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 543 :    1048576 : layers.60.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 544 :    8388608 : layers.60.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 545 :   29360128 : layers.60.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 546 :   29360128 : layers.60.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 547 :   29360128 : layers.60.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 548 :       8192 : layers.60.attention_norm.weight     : [8192]          : torch.bfloat16
 549 :       8192 : layers.60.ffn_norm.weight           : [8192]          : torch.bfloat16
 550 :    8388608 : layers.61.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 551 :    1048576 : layers.61.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 552 :    1048576 : layers.61.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 553 :    8388608 : layers.61.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 554 :   29360128 : layers.61.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 555 :   29360128 : layers.61.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 556 :   29360128 : layers.61.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 557 :       8192 : layers.61.attention_norm.weight     : [8192]          : torch.bfloat16
 558 :       8192 : layers.61.ffn_norm.weight           : [8192]          : torch.bfloat16
 559 :    8388608 : layers.62.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 560 :    1048576 : layers.62.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 561 :    1048576 : layers.62.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 562 :    8388608 : layers.62.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 563 :   29360128 : layers.62.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 564 :   29360128 : layers.62.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 565 :   29360128 : layers.62.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 566 :       8192 : layers.62.attention_norm.weight     : [8192]          : torch.bfloat16
 567 :       8192 : layers.62.ffn_norm.weight           : [8192]          : torch.bfloat16
 568 :    8388608 : layers.63.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 569 :    1048576 : layers.63.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 570 :    1048576 : layers.63.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 571 :    8388608 : layers.63.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 572 :   29360128 : layers.63.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 573 :   29360128 : layers.63.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 574 :   29360128 : layers.63.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 575 :       8192 : layers.63.attention_norm.weight     : [8192]          : torch.bfloat16
 576 :       8192 : layers.63.ffn_norm.weight           : [8192]          : torch.bfloat16
 577 :    8388608 : layers.64.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 578 :    1048576 : layers.64.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 579 :    1048576 : layers.64.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 580 :    8388608 : layers.64.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 581 :   29360128 : layers.64.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 582 :   29360128 : layers.64.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 583 :   29360128 : layers.64.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 584 :       8192 : layers.64.attention_norm.weight     : [8192]          : torch.bfloat16
 585 :       8192 : layers.64.ffn_norm.weight           : [8192]          : torch.bfloat16
 586 :    8388608 : layers.65.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 587 :    1048576 : layers.65.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 588 :    1048576 : layers.65.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 589 :    8388608 : layers.65.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 590 :   29360128 : layers.65.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 591 :   29360128 : layers.65.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 592 :   29360128 : layers.65.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 593 :       8192 : layers.65.attention_norm.weight     : [8192]          : torch.bfloat16
 594 :       8192 : layers.65.ffn_norm.weight           : [8192]          : torch.bfloat16
 595 :    8388608 : layers.66.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 596 :    1048576 : layers.66.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 597 :    1048576 : layers.66.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 598 :    8388608 : layers.66.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 599 :   29360128 : layers.66.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 600 :   29360128 : layers.66.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 601 :   29360128 : layers.66.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 602 :       8192 : layers.66.attention_norm.weight     : [8192]          : torch.bfloat16
 603 :       8192 : layers.66.ffn_norm.weight           : [8192]          : torch.bfloat16
 604 :    8388608 : layers.67.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 605 :    1048576 : layers.67.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 606 :    1048576 : layers.67.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 607 :    8388608 : layers.67.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 608 :   29360128 : layers.67.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 609 :   29360128 : layers.67.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 610 :   29360128 : layers.67.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 611 :       8192 : layers.67.attention_norm.weight     : [8192]          : torch.bfloat16
 612 :       8192 : layers.67.ffn_norm.weight           : [8192]          : torch.bfloat16
 613 :    8388608 : layers.68.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 614 :    1048576 : layers.68.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 615 :    1048576 : layers.68.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 616 :    8388608 : layers.68.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 617 :   29360128 : layers.68.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 618 :   29360128 : layers.68.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 619 :   29360128 : layers.68.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 620 :       8192 : layers.68.attention_norm.weight     : [8192]          : torch.bfloat16
 621 :       8192 : layers.68.ffn_norm.weight           : [8192]          : torch.bfloat16
 622 :    8388608 : layers.69.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 623 :    1048576 : layers.69.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 624 :    1048576 : layers.69.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 625 :    8388608 : layers.69.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 626 :   29360128 : layers.69.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 627 :   29360128 : layers.69.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 628 :   29360128 : layers.69.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 629 :       8192 : layers.69.attention_norm.weight     : [8192]          : torch.bfloat16
 630 :       8192 : layers.69.ffn_norm.weight           : [8192]          : torch.bfloat16
 631 :    8388608 : layers.70.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 632 :    1048576 : layers.70.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 633 :    1048576 : layers.70.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 634 :    8388608 : layers.70.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 635 :   29360128 : layers.70.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 636 :   29360128 : layers.70.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 637 :   29360128 : layers.70.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 638 :       8192 : layers.70.attention_norm.weight     : [8192]          : torch.bfloat16
 639 :       8192 : layers.70.ffn_norm.weight           : [8192]          : torch.bfloat16
 640 :    8388608 : layers.71.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 641 :    1048576 : layers.71.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 642 :    1048576 : layers.71.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 643 :    8388608 : layers.71.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 644 :   29360128 : layers.71.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 645 :   29360128 : layers.71.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 646 :   29360128 : layers.71.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 647 :       8192 : layers.71.attention_norm.weight     : [8192]          : torch.bfloat16
 648 :       8192 : layers.71.ffn_norm.weight           : [8192]          : torch.bfloat16
 649 :    8388608 : layers.72.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 650 :    1048576 : layers.72.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 651 :    1048576 : layers.72.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 652 :    8388608 : layers.72.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 653 :   29360128 : layers.72.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 654 :   29360128 : layers.72.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 655 :   29360128 : layers.72.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 656 :       8192 : layers.72.attention_norm.weight     : [8192]          : torch.bfloat16
 657 :       8192 : layers.72.ffn_norm.weight           : [8192]          : torch.bfloat16
 658 :    8388608 : layers.73.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 659 :    1048576 : layers.73.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 660 :    1048576 : layers.73.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 661 :    8388608 : layers.73.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 662 :   29360128 : layers.73.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 663 :   29360128 : layers.73.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 664 :   29360128 : layers.73.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 665 :       8192 : layers.73.attention_norm.weight     : [8192]          : torch.bfloat16
 666 :       8192 : layers.73.ffn_norm.weight           : [8192]          : torch.bfloat16
 667 :    8388608 : layers.74.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 668 :    1048576 : layers.74.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 669 :    1048576 : layers.74.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 670 :    8388608 : layers.74.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 671 :   29360128 : layers.74.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 672 :   29360128 : layers.74.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 673 :   29360128 : layers.74.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 674 :       8192 : layers.74.attention_norm.weight     : [8192]          : torch.bfloat16
 675 :       8192 : layers.74.ffn_norm.weight           : [8192]          : torch.bfloat16
 676 :    8388608 : layers.75.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 677 :    1048576 : layers.75.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 678 :    1048576 : layers.75.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 679 :    8388608 : layers.75.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 680 :   29360128 : layers.75.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 681 :   29360128 : layers.75.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 682 :   29360128 : layers.75.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 683 :       8192 : layers.75.attention_norm.weight     : [8192]          : torch.bfloat16
 684 :       8192 : layers.75.ffn_norm.weight           : [8192]          : torch.bfloat16
 685 :    8388608 : layers.76.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 686 :    1048576 : layers.76.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 687 :    1048576 : layers.76.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 688 :    8388608 : layers.76.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 689 :   29360128 : layers.76.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 690 :   29360128 : layers.76.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 691 :   29360128 : layers.76.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 692 :       8192 : layers.76.attention_norm.weight     : [8192]          : torch.bfloat16
 693 :       8192 : layers.76.ffn_norm.weight           : [8192]          : torch.bfloat16
 694 :    8388608 : layers.77.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 695 :    1048576 : layers.77.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 696 :    1048576 : layers.77.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 697 :    8388608 : layers.77.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 698 :   29360128 : layers.77.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 699 :   29360128 : layers.77.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 700 :   29360128 : layers.77.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 701 :       8192 : layers.77.attention_norm.weight     : [8192]          : torch.bfloat16
 702 :       8192 : layers.77.ffn_norm.weight           : [8192]          : torch.bfloat16
 703 :    8388608 : layers.78.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 704 :    1048576 : layers.78.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 705 :    1048576 : layers.78.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 706 :    8388608 : layers.78.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 707 :   29360128 : layers.78.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 708 :   29360128 : layers.78.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 709 :   29360128 : layers.78.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 710 :       8192 : layers.78.attention_norm.weight     : [8192]          : torch.bfloat16
 711 :       8192 : layers.78.ffn_norm.weight           : [8192]          : torch.bfloat16
 712 :    8388608 : layers.79.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 713 :    1048576 : layers.79.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 714 :    1048576 : layers.79.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 715 :    8388608 : layers.79.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 716 :   29360128 : layers.79.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 717 :   29360128 : layers.79.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 718 :   29360128 : layers.79.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 719 :       8192 : layers.79.attention_norm.weight     : [8192]          : torch.bfloat16
 720 :       8192 : layers.79.ffn_norm.weight           : [8192]          : torch.bfloat16
 721 :       8192 : norm.weight                         : [8192]          : torch.bfloat16
 722 :  131334144 : output.weight                       : [16032, 8192]   : torch.bfloat16

-----------------------------------------------------------------------------
61742571520 params in total.
123485143040 bytes in total.
40.41 sec, 4.91 sec, 2913.90 MB/s
-----------------------------------------------------------------------------

[8/8]: Loading original/consolidated.07.pth
   0 :  131334144 : tok_embeddings.weight               : [16032, 8192]   : torch.bfloat16
   1 :    8388608 : layers.0.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
   2 :    1048576 : layers.0.attention.wk.weight        : [128, 8192]     : torch.bfloat16
   3 :    1048576 : layers.0.attention.wv.weight        : [128, 8192]     : torch.bfloat16
   4 :    8388608 : layers.0.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
   5 :   29360128 : layers.0.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
   6 :   29360128 : layers.0.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
   7 :   29360128 : layers.0.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
   8 :       8192 : layers.0.attention_norm.weight      : [8192]          : torch.bfloat16
   9 :       8192 : layers.0.ffn_norm.weight            : [8192]          : torch.bfloat16
  10 :    8388608 : layers.1.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  11 :    1048576 : layers.1.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  12 :    1048576 : layers.1.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  13 :    8388608 : layers.1.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  14 :   29360128 : layers.1.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  15 :   29360128 : layers.1.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  16 :   29360128 : layers.1.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  17 :       8192 : layers.1.attention_norm.weight      : [8192]          : torch.bfloat16
  18 :       8192 : layers.1.ffn_norm.weight            : [8192]          : torch.bfloat16
  19 :    8388608 : layers.2.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  20 :    1048576 : layers.2.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  21 :    1048576 : layers.2.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  22 :    8388608 : layers.2.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  23 :   29360128 : layers.2.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  24 :   29360128 : layers.2.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  25 :   29360128 : layers.2.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  26 :       8192 : layers.2.attention_norm.weight      : [8192]          : torch.bfloat16
  27 :       8192 : layers.2.ffn_norm.weight            : [8192]          : torch.bfloat16
  28 :    8388608 : layers.3.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  29 :    1048576 : layers.3.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  30 :    1048576 : layers.3.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  31 :    8388608 : layers.3.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  32 :   29360128 : layers.3.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  33 :   29360128 : layers.3.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  34 :   29360128 : layers.3.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  35 :       8192 : layers.3.attention_norm.weight      : [8192]          : torch.bfloat16
  36 :       8192 : layers.3.ffn_norm.weight            : [8192]          : torch.bfloat16
  37 :    8388608 : layers.4.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  38 :    1048576 : layers.4.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  39 :    1048576 : layers.4.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  40 :    8388608 : layers.4.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  41 :   29360128 : layers.4.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  42 :   29360128 : layers.4.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  43 :   29360128 : layers.4.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  44 :       8192 : layers.4.attention_norm.weight      : [8192]          : torch.bfloat16
  45 :       8192 : layers.4.ffn_norm.weight            : [8192]          : torch.bfloat16
  46 :    8388608 : layers.5.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  47 :    1048576 : layers.5.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  48 :    1048576 : layers.5.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  49 :    8388608 : layers.5.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  50 :   29360128 : layers.5.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  51 :   29360128 : layers.5.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  52 :   29360128 : layers.5.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  53 :       8192 : layers.5.attention_norm.weight      : [8192]          : torch.bfloat16
  54 :       8192 : layers.5.ffn_norm.weight            : [8192]          : torch.bfloat16
  55 :    8388608 : layers.6.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  56 :    1048576 : layers.6.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  57 :    1048576 : layers.6.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  58 :    8388608 : layers.6.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  59 :   29360128 : layers.6.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  60 :   29360128 : layers.6.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  61 :   29360128 : layers.6.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  62 :       8192 : layers.6.attention_norm.weight      : [8192]          : torch.bfloat16
  63 :       8192 : layers.6.ffn_norm.weight            : [8192]          : torch.bfloat16
  64 :    8388608 : layers.7.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  65 :    1048576 : layers.7.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  66 :    1048576 : layers.7.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  67 :    8388608 : layers.7.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  68 :   29360128 : layers.7.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  69 :   29360128 : layers.7.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  70 :   29360128 : layers.7.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  71 :       8192 : layers.7.attention_norm.weight      : [8192]          : torch.bfloat16
  72 :       8192 : layers.7.ffn_norm.weight            : [8192]          : torch.bfloat16
  73 :    8388608 : layers.8.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  74 :    1048576 : layers.8.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  75 :    1048576 : layers.8.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  76 :    8388608 : layers.8.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  77 :   29360128 : layers.8.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  78 :   29360128 : layers.8.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  79 :   29360128 : layers.8.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  80 :       8192 : layers.8.attention_norm.weight      : [8192]          : torch.bfloat16
  81 :       8192 : layers.8.ffn_norm.weight            : [8192]          : torch.bfloat16
  82 :    8388608 : layers.9.attention.wq.weight        : [1024, 8192]    : torch.bfloat16
  83 :    1048576 : layers.9.attention.wk.weight        : [128, 8192]     : torch.bfloat16
  84 :    1048576 : layers.9.attention.wv.weight        : [128, 8192]     : torch.bfloat16
  85 :    8388608 : layers.9.attention.wo.weight        : [8192, 1024]    : torch.bfloat16
  86 :   29360128 : layers.9.feed_forward.w1.weight     : [3584, 8192]    : torch.bfloat16
  87 :   29360128 : layers.9.feed_forward.w3.weight     : [3584, 8192]    : torch.bfloat16
  88 :   29360128 : layers.9.feed_forward.w2.weight     : [8192, 3584]    : torch.bfloat16
  89 :       8192 : layers.9.attention_norm.weight      : [8192]          : torch.bfloat16
  90 :       8192 : layers.9.ffn_norm.weight            : [8192]          : torch.bfloat16
  91 :    8388608 : layers.10.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
  92 :    1048576 : layers.10.attention.wk.weight       : [128, 8192]     : torch.bfloat16
  93 :    1048576 : layers.10.attention.wv.weight       : [128, 8192]     : torch.bfloat16
  94 :    8388608 : layers.10.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
  95 :   29360128 : layers.10.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
  96 :   29360128 : layers.10.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
  97 :   29360128 : layers.10.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
  98 :       8192 : layers.10.attention_norm.weight     : [8192]          : torch.bfloat16
  99 :       8192 : layers.10.ffn_norm.weight           : [8192]          : torch.bfloat16
 100 :    8388608 : layers.11.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 101 :    1048576 : layers.11.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 102 :    1048576 : layers.11.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 103 :    8388608 : layers.11.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 104 :   29360128 : layers.11.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 105 :   29360128 : layers.11.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 106 :   29360128 : layers.11.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 107 :       8192 : layers.11.attention_norm.weight     : [8192]          : torch.bfloat16
 108 :       8192 : layers.11.ffn_norm.weight           : [8192]          : torch.bfloat16
 109 :    8388608 : layers.12.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 110 :    1048576 : layers.12.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 111 :    1048576 : layers.12.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 112 :    8388608 : layers.12.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 113 :   29360128 : layers.12.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 114 :   29360128 : layers.12.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 115 :   29360128 : layers.12.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 116 :       8192 : layers.12.attention_norm.weight     : [8192]          : torch.bfloat16
 117 :       8192 : layers.12.ffn_norm.weight           : [8192]          : torch.bfloat16
 118 :    8388608 : layers.13.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 119 :    1048576 : layers.13.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 120 :    1048576 : layers.13.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 121 :    8388608 : layers.13.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 122 :   29360128 : layers.13.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 123 :   29360128 : layers.13.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 124 :   29360128 : layers.13.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 125 :       8192 : layers.13.attention_norm.weight     : [8192]          : torch.bfloat16
 126 :       8192 : layers.13.ffn_norm.weight           : [8192]          : torch.bfloat16
 127 :    8388608 : layers.14.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 128 :    1048576 : layers.14.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 129 :    1048576 : layers.14.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 130 :    8388608 : layers.14.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 131 :   29360128 : layers.14.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 132 :   29360128 : layers.14.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 133 :   29360128 : layers.14.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 134 :       8192 : layers.14.attention_norm.weight     : [8192]          : torch.bfloat16
 135 :       8192 : layers.14.ffn_norm.weight           : [8192]          : torch.bfloat16
 136 :    8388608 : layers.15.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 137 :    1048576 : layers.15.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 138 :    1048576 : layers.15.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 139 :    8388608 : layers.15.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 140 :   29360128 : layers.15.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 141 :   29360128 : layers.15.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 142 :   29360128 : layers.15.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 143 :       8192 : layers.15.attention_norm.weight     : [8192]          : torch.bfloat16
 144 :       8192 : layers.15.ffn_norm.weight           : [8192]          : torch.bfloat16
 145 :    8388608 : layers.16.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 146 :    1048576 : layers.16.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 147 :    1048576 : layers.16.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 148 :    8388608 : layers.16.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 149 :   29360128 : layers.16.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 150 :   29360128 : layers.16.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 151 :   29360128 : layers.16.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 152 :       8192 : layers.16.attention_norm.weight     : [8192]          : torch.bfloat16
 153 :       8192 : layers.16.ffn_norm.weight           : [8192]          : torch.bfloat16
 154 :    8388608 : layers.17.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 155 :    1048576 : layers.17.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 156 :    1048576 : layers.17.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 157 :    8388608 : layers.17.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 158 :   29360128 : layers.17.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 159 :   29360128 : layers.17.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 160 :   29360128 : layers.17.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 161 :       8192 : layers.17.attention_norm.weight     : [8192]          : torch.bfloat16
 162 :       8192 : layers.17.ffn_norm.weight           : [8192]          : torch.bfloat16
 163 :    8388608 : layers.18.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 164 :    1048576 : layers.18.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 165 :    1048576 : layers.18.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 166 :    8388608 : layers.18.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 167 :   29360128 : layers.18.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 168 :   29360128 : layers.18.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 169 :   29360128 : layers.18.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 170 :       8192 : layers.18.attention_norm.weight     : [8192]          : torch.bfloat16
 171 :       8192 : layers.18.ffn_norm.weight           : [8192]          : torch.bfloat16
 172 :    8388608 : layers.19.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 173 :    1048576 : layers.19.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 174 :    1048576 : layers.19.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 175 :    8388608 : layers.19.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 176 :   29360128 : layers.19.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 177 :   29360128 : layers.19.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 178 :   29360128 : layers.19.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 179 :       8192 : layers.19.attention_norm.weight     : [8192]          : torch.bfloat16
 180 :       8192 : layers.19.ffn_norm.weight           : [8192]          : torch.bfloat16
 181 :    8388608 : layers.20.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 182 :    1048576 : layers.20.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 183 :    1048576 : layers.20.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 184 :    8388608 : layers.20.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 185 :   29360128 : layers.20.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 186 :   29360128 : layers.20.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 187 :   29360128 : layers.20.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 188 :       8192 : layers.20.attention_norm.weight     : [8192]          : torch.bfloat16
 189 :       8192 : layers.20.ffn_norm.weight           : [8192]          : torch.bfloat16
 190 :    8388608 : layers.21.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 191 :    1048576 : layers.21.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 192 :    1048576 : layers.21.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 193 :    8388608 : layers.21.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 194 :   29360128 : layers.21.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 195 :   29360128 : layers.21.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 196 :   29360128 : layers.21.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 197 :       8192 : layers.21.attention_norm.weight     : [8192]          : torch.bfloat16
 198 :       8192 : layers.21.ffn_norm.weight           : [8192]          : torch.bfloat16
 199 :    8388608 : layers.22.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 200 :    1048576 : layers.22.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 201 :    1048576 : layers.22.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 202 :    8388608 : layers.22.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 203 :   29360128 : layers.22.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 204 :   29360128 : layers.22.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 205 :   29360128 : layers.22.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 206 :       8192 : layers.22.attention_norm.weight     : [8192]          : torch.bfloat16
 207 :       8192 : layers.22.ffn_norm.weight           : [8192]          : torch.bfloat16
 208 :    8388608 : layers.23.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 209 :    1048576 : layers.23.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 210 :    1048576 : layers.23.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 211 :    8388608 : layers.23.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 212 :   29360128 : layers.23.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 213 :   29360128 : layers.23.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 214 :   29360128 : layers.23.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 215 :       8192 : layers.23.attention_norm.weight     : [8192]          : torch.bfloat16
 216 :       8192 : layers.23.ffn_norm.weight           : [8192]          : torch.bfloat16
 217 :    8388608 : layers.24.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 218 :    1048576 : layers.24.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 219 :    1048576 : layers.24.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 220 :    8388608 : layers.24.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 221 :   29360128 : layers.24.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 222 :   29360128 : layers.24.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 223 :   29360128 : layers.24.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 224 :       8192 : layers.24.attention_norm.weight     : [8192]          : torch.bfloat16
 225 :       8192 : layers.24.ffn_norm.weight           : [8192]          : torch.bfloat16
 226 :    8388608 : layers.25.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 227 :    1048576 : layers.25.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 228 :    1048576 : layers.25.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 229 :    8388608 : layers.25.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 230 :   29360128 : layers.25.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 231 :   29360128 : layers.25.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 232 :   29360128 : layers.25.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 233 :       8192 : layers.25.attention_norm.weight     : [8192]          : torch.bfloat16
 234 :       8192 : layers.25.ffn_norm.weight           : [8192]          : torch.bfloat16
 235 :    8388608 : layers.26.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 236 :    1048576 : layers.26.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 237 :    1048576 : layers.26.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 238 :    8388608 : layers.26.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 239 :   29360128 : layers.26.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 240 :   29360128 : layers.26.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 241 :   29360128 : layers.26.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 242 :       8192 : layers.26.attention_norm.weight     : [8192]          : torch.bfloat16
 243 :       8192 : layers.26.ffn_norm.weight           : [8192]          : torch.bfloat16
 244 :    8388608 : layers.27.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 245 :    1048576 : layers.27.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 246 :    1048576 : layers.27.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 247 :    8388608 : layers.27.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 248 :   29360128 : layers.27.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 249 :   29360128 : layers.27.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 250 :   29360128 : layers.27.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 251 :       8192 : layers.27.attention_norm.weight     : [8192]          : torch.bfloat16
 252 :       8192 : layers.27.ffn_norm.weight           : [8192]          : torch.bfloat16
 253 :    8388608 : layers.28.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 254 :    1048576 : layers.28.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 255 :    1048576 : layers.28.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 256 :    8388608 : layers.28.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 257 :   29360128 : layers.28.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 258 :   29360128 : layers.28.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 259 :   29360128 : layers.28.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 260 :       8192 : layers.28.attention_norm.weight     : [8192]          : torch.bfloat16
 261 :       8192 : layers.28.ffn_norm.weight           : [8192]          : torch.bfloat16
 262 :    8388608 : layers.29.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 263 :    1048576 : layers.29.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 264 :    1048576 : layers.29.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 265 :    8388608 : layers.29.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 266 :   29360128 : layers.29.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 267 :   29360128 : layers.29.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 268 :   29360128 : layers.29.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 269 :       8192 : layers.29.attention_norm.weight     : [8192]          : torch.bfloat16
 270 :       8192 : layers.29.ffn_norm.weight           : [8192]          : torch.bfloat16
 271 :    8388608 : layers.30.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 272 :    1048576 : layers.30.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 273 :    1048576 : layers.30.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 274 :    8388608 : layers.30.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 275 :   29360128 : layers.30.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 276 :   29360128 : layers.30.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 277 :   29360128 : layers.30.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 278 :       8192 : layers.30.attention_norm.weight     : [8192]          : torch.bfloat16
 279 :       8192 : layers.30.ffn_norm.weight           : [8192]          : torch.bfloat16
 280 :    8388608 : layers.31.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 281 :    1048576 : layers.31.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 282 :    1048576 : layers.31.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 283 :    8388608 : layers.31.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 284 :   29360128 : layers.31.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 285 :   29360128 : layers.31.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 286 :   29360128 : layers.31.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 287 :       8192 : layers.31.attention_norm.weight     : [8192]          : torch.bfloat16
 288 :       8192 : layers.31.ffn_norm.weight           : [8192]          : torch.bfloat16
 289 :    8388608 : layers.32.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 290 :    1048576 : layers.32.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 291 :    1048576 : layers.32.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 292 :    8388608 : layers.32.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 293 :   29360128 : layers.32.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 294 :   29360128 : layers.32.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 295 :   29360128 : layers.32.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 296 :       8192 : layers.32.attention_norm.weight     : [8192]          : torch.bfloat16
 297 :       8192 : layers.32.ffn_norm.weight           : [8192]          : torch.bfloat16
 298 :    8388608 : layers.33.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 299 :    1048576 : layers.33.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 300 :    1048576 : layers.33.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 301 :    8388608 : layers.33.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 302 :   29360128 : layers.33.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 303 :   29360128 : layers.33.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 304 :   29360128 : layers.33.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 305 :       8192 : layers.33.attention_norm.weight     : [8192]          : torch.bfloat16
 306 :       8192 : layers.33.ffn_norm.weight           : [8192]          : torch.bfloat16
 307 :    8388608 : layers.34.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 308 :    1048576 : layers.34.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 309 :    1048576 : layers.34.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 310 :    8388608 : layers.34.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 311 :   29360128 : layers.34.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 312 :   29360128 : layers.34.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 313 :   29360128 : layers.34.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 314 :       8192 : layers.34.attention_norm.weight     : [8192]          : torch.bfloat16
 315 :       8192 : layers.34.ffn_norm.weight           : [8192]          : torch.bfloat16
 316 :    8388608 : layers.35.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 317 :    1048576 : layers.35.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 318 :    1048576 : layers.35.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 319 :    8388608 : layers.35.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 320 :   29360128 : layers.35.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 321 :   29360128 : layers.35.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 322 :   29360128 : layers.35.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 323 :       8192 : layers.35.attention_norm.weight     : [8192]          : torch.bfloat16
 324 :       8192 : layers.35.ffn_norm.weight           : [8192]          : torch.bfloat16
 325 :    8388608 : layers.36.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 326 :    1048576 : layers.36.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 327 :    1048576 : layers.36.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 328 :    8388608 : layers.36.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 329 :   29360128 : layers.36.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 330 :   29360128 : layers.36.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 331 :   29360128 : layers.36.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 332 :       8192 : layers.36.attention_norm.weight     : [8192]          : torch.bfloat16
 333 :       8192 : layers.36.ffn_norm.weight           : [8192]          : torch.bfloat16
 334 :    8388608 : layers.37.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 335 :    1048576 : layers.37.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 336 :    1048576 : layers.37.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 337 :    8388608 : layers.37.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 338 :   29360128 : layers.37.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 339 :   29360128 : layers.37.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 340 :   29360128 : layers.37.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 341 :       8192 : layers.37.attention_norm.weight     : [8192]          : torch.bfloat16
 342 :       8192 : layers.37.ffn_norm.weight           : [8192]          : torch.bfloat16
 343 :    8388608 : layers.38.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 344 :    1048576 : layers.38.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 345 :    1048576 : layers.38.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 346 :    8388608 : layers.38.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 347 :   29360128 : layers.38.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 348 :   29360128 : layers.38.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 349 :   29360128 : layers.38.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 350 :       8192 : layers.38.attention_norm.weight     : [8192]          : torch.bfloat16
 351 :       8192 : layers.38.ffn_norm.weight           : [8192]          : torch.bfloat16
 352 :    8388608 : layers.39.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 353 :    1048576 : layers.39.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 354 :    1048576 : layers.39.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 355 :    8388608 : layers.39.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 356 :   29360128 : layers.39.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 357 :   29360128 : layers.39.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 358 :   29360128 : layers.39.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 359 :       8192 : layers.39.attention_norm.weight     : [8192]          : torch.bfloat16
 360 :       8192 : layers.39.ffn_norm.weight           : [8192]          : torch.bfloat16
 361 :    8388608 : layers.40.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 362 :    1048576 : layers.40.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 363 :    1048576 : layers.40.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 364 :    8388608 : layers.40.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 365 :   29360128 : layers.40.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 366 :   29360128 : layers.40.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 367 :   29360128 : layers.40.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 368 :       8192 : layers.40.attention_norm.weight     : [8192]          : torch.bfloat16
 369 :       8192 : layers.40.ffn_norm.weight           : [8192]          : torch.bfloat16
 370 :    8388608 : layers.41.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 371 :    1048576 : layers.41.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 372 :    1048576 : layers.41.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 373 :    8388608 : layers.41.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 374 :   29360128 : layers.41.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 375 :   29360128 : layers.41.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 376 :   29360128 : layers.41.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 377 :       8192 : layers.41.attention_norm.weight     : [8192]          : torch.bfloat16
 378 :       8192 : layers.41.ffn_norm.weight           : [8192]          : torch.bfloat16
 379 :    8388608 : layers.42.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 380 :    1048576 : layers.42.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 381 :    1048576 : layers.42.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 382 :    8388608 : layers.42.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 383 :   29360128 : layers.42.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 384 :   29360128 : layers.42.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 385 :   29360128 : layers.42.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 386 :       8192 : layers.42.attention_norm.weight     : [8192]          : torch.bfloat16
 387 :       8192 : layers.42.ffn_norm.weight           : [8192]          : torch.bfloat16
 388 :    8388608 : layers.43.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 389 :    1048576 : layers.43.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 390 :    1048576 : layers.43.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 391 :    8388608 : layers.43.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 392 :   29360128 : layers.43.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 393 :   29360128 : layers.43.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 394 :   29360128 : layers.43.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 395 :       8192 : layers.43.attention_norm.weight     : [8192]          : torch.bfloat16
 396 :       8192 : layers.43.ffn_norm.weight           : [8192]          : torch.bfloat16
 397 :    8388608 : layers.44.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 398 :    1048576 : layers.44.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 399 :    1048576 : layers.44.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 400 :    8388608 : layers.44.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 401 :   29360128 : layers.44.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 402 :   29360128 : layers.44.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 403 :   29360128 : layers.44.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 404 :       8192 : layers.44.attention_norm.weight     : [8192]          : torch.bfloat16
 405 :       8192 : layers.44.ffn_norm.weight           : [8192]          : torch.bfloat16
 406 :    8388608 : layers.45.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 407 :    1048576 : layers.45.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 408 :    1048576 : layers.45.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 409 :    8388608 : layers.45.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 410 :   29360128 : layers.45.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 411 :   29360128 : layers.45.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 412 :   29360128 : layers.45.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 413 :       8192 : layers.45.attention_norm.weight     : [8192]          : torch.bfloat16
 414 :       8192 : layers.45.ffn_norm.weight           : [8192]          : torch.bfloat16
 415 :    8388608 : layers.46.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 416 :    1048576 : layers.46.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 417 :    1048576 : layers.46.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 418 :    8388608 : layers.46.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 419 :   29360128 : layers.46.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 420 :   29360128 : layers.46.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 421 :   29360128 : layers.46.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 422 :       8192 : layers.46.attention_norm.weight     : [8192]          : torch.bfloat16
 423 :       8192 : layers.46.ffn_norm.weight           : [8192]          : torch.bfloat16
 424 :    8388608 : layers.47.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 425 :    1048576 : layers.47.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 426 :    1048576 : layers.47.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 427 :    8388608 : layers.47.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 428 :   29360128 : layers.47.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 429 :   29360128 : layers.47.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 430 :   29360128 : layers.47.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 431 :       8192 : layers.47.attention_norm.weight     : [8192]          : torch.bfloat16
 432 :       8192 : layers.47.ffn_norm.weight           : [8192]          : torch.bfloat16
 433 :    8388608 : layers.48.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 434 :    1048576 : layers.48.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 435 :    1048576 : layers.48.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 436 :    8388608 : layers.48.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 437 :   29360128 : layers.48.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 438 :   29360128 : layers.48.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 439 :   29360128 : layers.48.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 440 :       8192 : layers.48.attention_norm.weight     : [8192]          : torch.bfloat16
 441 :       8192 : layers.48.ffn_norm.weight           : [8192]          : torch.bfloat16
 442 :    8388608 : layers.49.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 443 :    1048576 : layers.49.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 444 :    1048576 : layers.49.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 445 :    8388608 : layers.49.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 446 :   29360128 : layers.49.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 447 :   29360128 : layers.49.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 448 :   29360128 : layers.49.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 449 :       8192 : layers.49.attention_norm.weight     : [8192]          : torch.bfloat16
 450 :       8192 : layers.49.ffn_norm.weight           : [8192]          : torch.bfloat16
 451 :    8388608 : layers.50.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 452 :    1048576 : layers.50.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 453 :    1048576 : layers.50.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 454 :    8388608 : layers.50.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 455 :   29360128 : layers.50.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 456 :   29360128 : layers.50.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 457 :   29360128 : layers.50.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 458 :       8192 : layers.50.attention_norm.weight     : [8192]          : torch.bfloat16
 459 :       8192 : layers.50.ffn_norm.weight           : [8192]          : torch.bfloat16
 460 :    8388608 : layers.51.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 461 :    1048576 : layers.51.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 462 :    1048576 : layers.51.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 463 :    8388608 : layers.51.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 464 :   29360128 : layers.51.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 465 :   29360128 : layers.51.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 466 :   29360128 : layers.51.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 467 :       8192 : layers.51.attention_norm.weight     : [8192]          : torch.bfloat16
 468 :       8192 : layers.51.ffn_norm.weight           : [8192]          : torch.bfloat16
 469 :    8388608 : layers.52.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 470 :    1048576 : layers.52.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 471 :    1048576 : layers.52.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 472 :    8388608 : layers.52.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 473 :   29360128 : layers.52.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 474 :   29360128 : layers.52.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 475 :   29360128 : layers.52.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 476 :       8192 : layers.52.attention_norm.weight     : [8192]          : torch.bfloat16
 477 :       8192 : layers.52.ffn_norm.weight           : [8192]          : torch.bfloat16
 478 :    8388608 : layers.53.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 479 :    1048576 : layers.53.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 480 :    1048576 : layers.53.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 481 :    8388608 : layers.53.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 482 :   29360128 : layers.53.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 483 :   29360128 : layers.53.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 484 :   29360128 : layers.53.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 485 :       8192 : layers.53.attention_norm.weight     : [8192]          : torch.bfloat16
 486 :       8192 : layers.53.ffn_norm.weight           : [8192]          : torch.bfloat16
 487 :    8388608 : layers.54.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 488 :    1048576 : layers.54.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 489 :    1048576 : layers.54.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 490 :    8388608 : layers.54.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 491 :   29360128 : layers.54.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 492 :   29360128 : layers.54.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 493 :   29360128 : layers.54.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 494 :       8192 : layers.54.attention_norm.weight     : [8192]          : torch.bfloat16
 495 :       8192 : layers.54.ffn_norm.weight           : [8192]          : torch.bfloat16
 496 :    8388608 : layers.55.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 497 :    1048576 : layers.55.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 498 :    1048576 : layers.55.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 499 :    8388608 : layers.55.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 500 :   29360128 : layers.55.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 501 :   29360128 : layers.55.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 502 :   29360128 : layers.55.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 503 :       8192 : layers.55.attention_norm.weight     : [8192]          : torch.bfloat16
 504 :       8192 : layers.55.ffn_norm.weight           : [8192]          : torch.bfloat16
 505 :    8388608 : layers.56.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 506 :    1048576 : layers.56.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 507 :    1048576 : layers.56.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 508 :    8388608 : layers.56.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 509 :   29360128 : layers.56.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 510 :   29360128 : layers.56.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 511 :   29360128 : layers.56.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 512 :       8192 : layers.56.attention_norm.weight     : [8192]          : torch.bfloat16
 513 :       8192 : layers.56.ffn_norm.weight           : [8192]          : torch.bfloat16
 514 :    8388608 : layers.57.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 515 :    1048576 : layers.57.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 516 :    1048576 : layers.57.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 517 :    8388608 : layers.57.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 518 :   29360128 : layers.57.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 519 :   29360128 : layers.57.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 520 :   29360128 : layers.57.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 521 :       8192 : layers.57.attention_norm.weight     : [8192]          : torch.bfloat16
 522 :       8192 : layers.57.ffn_norm.weight           : [8192]          : torch.bfloat16
 523 :    8388608 : layers.58.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 524 :    1048576 : layers.58.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 525 :    1048576 : layers.58.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 526 :    8388608 : layers.58.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 527 :   29360128 : layers.58.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 528 :   29360128 : layers.58.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 529 :   29360128 : layers.58.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 530 :       8192 : layers.58.attention_norm.weight     : [8192]          : torch.bfloat16
 531 :       8192 : layers.58.ffn_norm.weight           : [8192]          : torch.bfloat16
 532 :    8388608 : layers.59.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 533 :    1048576 : layers.59.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 534 :    1048576 : layers.59.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 535 :    8388608 : layers.59.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 536 :   29360128 : layers.59.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 537 :   29360128 : layers.59.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 538 :   29360128 : layers.59.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 539 :       8192 : layers.59.attention_norm.weight     : [8192]          : torch.bfloat16
 540 :       8192 : layers.59.ffn_norm.weight           : [8192]          : torch.bfloat16
 541 :    8388608 : layers.60.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 542 :    1048576 : layers.60.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 543 :    1048576 : layers.60.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 544 :    8388608 : layers.60.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 545 :   29360128 : layers.60.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 546 :   29360128 : layers.60.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 547 :   29360128 : layers.60.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 548 :       8192 : layers.60.attention_norm.weight     : [8192]          : torch.bfloat16
 549 :       8192 : layers.60.ffn_norm.weight           : [8192]          : torch.bfloat16
 550 :    8388608 : layers.61.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 551 :    1048576 : layers.61.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 552 :    1048576 : layers.61.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 553 :    8388608 : layers.61.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 554 :   29360128 : layers.61.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 555 :   29360128 : layers.61.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 556 :   29360128 : layers.61.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 557 :       8192 : layers.61.attention_norm.weight     : [8192]          : torch.bfloat16
 558 :       8192 : layers.61.ffn_norm.weight           : [8192]          : torch.bfloat16
 559 :    8388608 : layers.62.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 560 :    1048576 : layers.62.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 561 :    1048576 : layers.62.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 562 :    8388608 : layers.62.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 563 :   29360128 : layers.62.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 564 :   29360128 : layers.62.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 565 :   29360128 : layers.62.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 566 :       8192 : layers.62.attention_norm.weight     : [8192]          : torch.bfloat16
 567 :       8192 : layers.62.ffn_norm.weight           : [8192]          : torch.bfloat16
 568 :    8388608 : layers.63.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 569 :    1048576 : layers.63.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 570 :    1048576 : layers.63.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 571 :    8388608 : layers.63.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 572 :   29360128 : layers.63.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 573 :   29360128 : layers.63.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 574 :   29360128 : layers.63.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 575 :       8192 : layers.63.attention_norm.weight     : [8192]          : torch.bfloat16
 576 :       8192 : layers.63.ffn_norm.weight           : [8192]          : torch.bfloat16
 577 :    8388608 : layers.64.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 578 :    1048576 : layers.64.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 579 :    1048576 : layers.64.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 580 :    8388608 : layers.64.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 581 :   29360128 : layers.64.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 582 :   29360128 : layers.64.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 583 :   29360128 : layers.64.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 584 :       8192 : layers.64.attention_norm.weight     : [8192]          : torch.bfloat16
 585 :       8192 : layers.64.ffn_norm.weight           : [8192]          : torch.bfloat16
 586 :    8388608 : layers.65.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 587 :    1048576 : layers.65.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 588 :    1048576 : layers.65.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 589 :    8388608 : layers.65.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 590 :   29360128 : layers.65.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 591 :   29360128 : layers.65.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 592 :   29360128 : layers.65.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 593 :       8192 : layers.65.attention_norm.weight     : [8192]          : torch.bfloat16
 594 :       8192 : layers.65.ffn_norm.weight           : [8192]          : torch.bfloat16
 595 :    8388608 : layers.66.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 596 :    1048576 : layers.66.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 597 :    1048576 : layers.66.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 598 :    8388608 : layers.66.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 599 :   29360128 : layers.66.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 600 :   29360128 : layers.66.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 601 :   29360128 : layers.66.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 602 :       8192 : layers.66.attention_norm.weight     : [8192]          : torch.bfloat16
 603 :       8192 : layers.66.ffn_norm.weight           : [8192]          : torch.bfloat16
 604 :    8388608 : layers.67.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 605 :    1048576 : layers.67.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 606 :    1048576 : layers.67.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 607 :    8388608 : layers.67.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 608 :   29360128 : layers.67.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 609 :   29360128 : layers.67.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 610 :   29360128 : layers.67.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 611 :       8192 : layers.67.attention_norm.weight     : [8192]          : torch.bfloat16
 612 :       8192 : layers.67.ffn_norm.weight           : [8192]          : torch.bfloat16
 613 :    8388608 : layers.68.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 614 :    1048576 : layers.68.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 615 :    1048576 : layers.68.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 616 :    8388608 : layers.68.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 617 :   29360128 : layers.68.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 618 :   29360128 : layers.68.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 619 :   29360128 : layers.68.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 620 :       8192 : layers.68.attention_norm.weight     : [8192]          : torch.bfloat16
 621 :       8192 : layers.68.ffn_norm.weight           : [8192]          : torch.bfloat16
 622 :    8388608 : layers.69.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 623 :    1048576 : layers.69.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 624 :    1048576 : layers.69.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 625 :    8388608 : layers.69.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 626 :   29360128 : layers.69.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 627 :   29360128 : layers.69.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 628 :   29360128 : layers.69.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 629 :       8192 : layers.69.attention_norm.weight     : [8192]          : torch.bfloat16
 630 :       8192 : layers.69.ffn_norm.weight           : [8192]          : torch.bfloat16
 631 :    8388608 : layers.70.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 632 :    1048576 : layers.70.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 633 :    1048576 : layers.70.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 634 :    8388608 : layers.70.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 635 :   29360128 : layers.70.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 636 :   29360128 : layers.70.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 637 :   29360128 : layers.70.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 638 :       8192 : layers.70.attention_norm.weight     : [8192]          : torch.bfloat16
 639 :       8192 : layers.70.ffn_norm.weight           : [8192]          : torch.bfloat16
 640 :    8388608 : layers.71.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 641 :    1048576 : layers.71.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 642 :    1048576 : layers.71.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 643 :    8388608 : layers.71.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 644 :   29360128 : layers.71.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 645 :   29360128 : layers.71.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 646 :   29360128 : layers.71.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 647 :       8192 : layers.71.attention_norm.weight     : [8192]          : torch.bfloat16
 648 :       8192 : layers.71.ffn_norm.weight           : [8192]          : torch.bfloat16
 649 :    8388608 : layers.72.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 650 :    1048576 : layers.72.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 651 :    1048576 : layers.72.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 652 :    8388608 : layers.72.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 653 :   29360128 : layers.72.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 654 :   29360128 : layers.72.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 655 :   29360128 : layers.72.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 656 :       8192 : layers.72.attention_norm.weight     : [8192]          : torch.bfloat16
 657 :       8192 : layers.72.ffn_norm.weight           : [8192]          : torch.bfloat16
 658 :    8388608 : layers.73.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 659 :    1048576 : layers.73.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 660 :    1048576 : layers.73.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 661 :    8388608 : layers.73.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 662 :   29360128 : layers.73.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 663 :   29360128 : layers.73.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 664 :   29360128 : layers.73.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 665 :       8192 : layers.73.attention_norm.weight     : [8192]          : torch.bfloat16
 666 :       8192 : layers.73.ffn_norm.weight           : [8192]          : torch.bfloat16
 667 :    8388608 : layers.74.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 668 :    1048576 : layers.74.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 669 :    1048576 : layers.74.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 670 :    8388608 : layers.74.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 671 :   29360128 : layers.74.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 672 :   29360128 : layers.74.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 673 :   29360128 : layers.74.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 674 :       8192 : layers.74.attention_norm.weight     : [8192]          : torch.bfloat16
 675 :       8192 : layers.74.ffn_norm.weight           : [8192]          : torch.bfloat16
 676 :    8388608 : layers.75.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 677 :    1048576 : layers.75.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 678 :    1048576 : layers.75.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 679 :    8388608 : layers.75.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 680 :   29360128 : layers.75.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 681 :   29360128 : layers.75.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 682 :   29360128 : layers.75.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 683 :       8192 : layers.75.attention_norm.weight     : [8192]          : torch.bfloat16
 684 :       8192 : layers.75.ffn_norm.weight           : [8192]          : torch.bfloat16
 685 :    8388608 : layers.76.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 686 :    1048576 : layers.76.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 687 :    1048576 : layers.76.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 688 :    8388608 : layers.76.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 689 :   29360128 : layers.76.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 690 :   29360128 : layers.76.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 691 :   29360128 : layers.76.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 692 :       8192 : layers.76.attention_norm.weight     : [8192]          : torch.bfloat16
 693 :       8192 : layers.76.ffn_norm.weight           : [8192]          : torch.bfloat16
 694 :    8388608 : layers.77.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 695 :    1048576 : layers.77.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 696 :    1048576 : layers.77.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 697 :    8388608 : layers.77.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 698 :   29360128 : layers.77.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 699 :   29360128 : layers.77.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 700 :   29360128 : layers.77.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 701 :       8192 : layers.77.attention_norm.weight     : [8192]          : torch.bfloat16
 702 :       8192 : layers.77.ffn_norm.weight           : [8192]          : torch.bfloat16
 703 :    8388608 : layers.78.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 704 :    1048576 : layers.78.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 705 :    1048576 : layers.78.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 706 :    8388608 : layers.78.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 707 :   29360128 : layers.78.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 708 :   29360128 : layers.78.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 709 :   29360128 : layers.78.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 710 :       8192 : layers.78.attention_norm.weight     : [8192]          : torch.bfloat16
 711 :       8192 : layers.78.ffn_norm.weight           : [8192]          : torch.bfloat16
 712 :    8388608 : layers.79.attention.wq.weight       : [1024, 8192]    : torch.bfloat16
 713 :    1048576 : layers.79.attention.wk.weight       : [128, 8192]     : torch.bfloat16
 714 :    1048576 : layers.79.attention.wv.weight       : [128, 8192]     : torch.bfloat16
 715 :    8388608 : layers.79.attention.wo.weight       : [8192, 1024]    : torch.bfloat16
 716 :   29360128 : layers.79.feed_forward.w1.weight    : [3584, 8192]    : torch.bfloat16
 717 :   29360128 : layers.79.feed_forward.w3.weight    : [3584, 8192]    : torch.bfloat16
 718 :   29360128 : layers.79.feed_forward.w2.weight    : [8192, 3584]    : torch.bfloat16
 719 :       8192 : layers.79.attention_norm.weight     : [8192]          : torch.bfloat16
 720 :       8192 : layers.79.ffn_norm.weight           : [8192]          : torch.bfloat16
 721 :       8192 : norm.weight                         : [8192]          : torch.bfloat16
 722 :  131334144 : output.weight                       : [16032, 8192]   : torch.bfloat16
Total number of parameters: 70562938880
70.5629 B
141125877760 Bytes
131.4337 GB