Views
No views yet
1models:
2- model: google/gemma-2-2b
3- model: google/gemma-2-2b-it
4 parameters:
5 density:
6 - filter: model.layers.0.self_attn.q_proj
7 value: 0.00539
8 - filter: model.layers.1.self_attn.q_proj
9 value: 0.3256
10 - filter: model.layers.2.self_attn.q_proj
11 value: 0.03843
12 - filter: model.layers.3.self_attn.q_proj
13 value: 0.14886
14 - filter: model.layers.4.self_attn.q_proj
15 value: 0.41323
16 - filter: model.layers.5.self_attn.q_proj
17 value: 0.33564
18 - filter: model.layers.6.self_attn.q_proj
19 value: 0.03716
20 - filter: model.layers.7.self_attn.q_proj
21 value: 0.1084
22 - filter: model.layers.8.self_attn.q_proj
23 value: 0.16658
24 - filter: model.layers.9.self_attn.q_proj
25 value: 0.19716
26 - filter: model.layers.10.self_attn.q_proj
27 value: 0.10008
28 - filter: model.layers.11.self_attn.q_proj
29 value: 0.37198
30 - filter: model.layers.12.self_attn.q_proj
31 value: 0.3622
32 - filter: model.layers.13.self_attn.q_proj
33 value: 0.26806
34 - filter: model.layers.14.self_attn.q_proj
35 value: 0.73192
36 - filter: model.layers.15.self_attn.q_proj
37 value: 0.14731
38 - filter: model.layers.16.self_attn.q_proj
39 value: 0.46923
40 - filter: model.layers.17.self_attn.q_proj
41 value: 0.63063
42 - filter: model.layers.18.self_attn.q_proj
43 value: 0.65673
44 - filter: model.layers.19.self_attn.q_proj
45 value: 0.66558
46 - filter: model.layers.20.self_attn.q_proj
47 value: 0.63758
48 - filter: model.layers.21.self_attn.q_proj
49 value: 0.64362
50 - filter: model.layers.22.self_attn.q_proj
51 value: 0.40341
52 - filter: model.layers.23.self_attn.q_proj
53 value: 0.12982
54 - filter: model.layers.24.self_attn.q_proj
55 value: 0.04552
56 - filter: model.layers.25.self_attn.q_proj
57 value: 0.03919
58 - filter: model.layers.0.self_attn.k_proj
59 value: 0.00592
60 - filter: model.layers.1.self_attn.k_proj
61 value: 0.18684
62 - filter: model.layers.2.self_attn.k_proj
63 value: 0.02603
64 - filter: model.layers.3.self_attn.k_proj
65 value: 0.07283
66 - filter: model.layers.4.self_attn.k_proj
67 value: 0.27067
68 - filter: model.layers.5.self_attn.k_proj
69 value: 0.18189
70 - filter: model.layers.6.self_attn.k_proj
71 value: 0.10695
72 - filter: model.layers.7.self_attn.k_proj
73 value: 0.17169
74 - filter: model.layers.8.self_attn.k_proj
75 value: 0.08753
76 - filter: model.layers.9.self_attn.k_proj
77 value: 0.12807
78 - filter: model.layers.10.self_attn.k_proj
79 value: 0.07783
80 - filter: model.layers.11.self_attn.k_proj
81 value: 0.41518
82 - filter: model.layers.12.self_attn.k_proj
83 value: 0.42307
84 - filter: model.layers.13.self_attn.k_proj
85 value: 0.2896
86 - filter: model.layers.14.self_attn.k_proj
87 value: 0.52953
88 - filter: model.layers.15.self_attn.k_proj
89 value: 0.25035
90 - filter: model.layers.16.self_attn.k_proj
91 value: 0.57679
92 - filter: model.layers.17.self_attn.k_proj
93 value: 0.52768
94 - filter: model.layers.18.self_attn.k_proj
95 value: 0.666
96 - filter: model.layers.19.self_attn.k_proj
97 value: 0.67796
98 - filter: model.layers.20.self_attn.k_proj
99 value: 0.58825
100 - filter: model.layers.21.self_attn.k_proj
101 value: 0.51447
102 - filter: model.layers.22.self_attn.k_proj
103 value: 0.33317
104 - filter: model.layers.23.self_attn.k_proj
105 value: 0.05987
106 - filter: model.layers.24.self_attn.k_proj
107 value: 0.02903
108 - filter: model.layers.25.self_attn.k_proj
109 value: 0.08715
110 - filter: model.layers.0.self_attn.v_proj
111 value: 0.03025
112 - filter: model.layers.1.self_attn.v_proj
113 value: 0.14137
114 - filter: model.layers.2.self_attn.v_proj
115 value: 0.00286
116 - filter: model.layers.3.self_attn.v_proj
117 value: 0.09155
118 - filter: model.layers.4.self_attn.v_proj
119 value: 0.25172
120 - filter: model.layers.5.self_attn.v_proj
121 value: 0.2062
122 - filter: model.layers.6.self_attn.v_proj
123 value: 0.06811
124 - filter: model.layers.7.self_attn.v_proj
125 value: 0.01334
126 - filter: model.layers.8.self_attn.v_proj
127 value: 0.18251
128 - filter: model.layers.9.self_attn.v_proj
129 value: 0.10415
130 - filter: model.layers.10.self_attn.v_proj
131 value: 0.04
132 - filter: model.layers.11.self_attn.v_proj
133 value: 0.22664
134 - filter: model.layers.12.self_attn.v_proj
135 value: 0.16337
136 - filter: model.layers.13.self_attn.v_proj
137 value: 0.09347
138 - filter: model.layers.14.self_attn.v_proj
139 value: 0.52549
140 - filter: model.layers.15.self_attn.v_proj
141 value: 0.02152
142 - filter: model.layers.16.self_attn.v_proj
143 value: 0.25325
144 - filter: model.layers.17.self_attn.v_proj
145 value: 0.43041
146 - filter: model.layers.18.self_attn.v_proj
147 value: 0.49113
148 - filter: model.layers.19.self_attn.v_proj
149 value: 0.32307
150 - filter: model.layers.20.self_attn.v_proj
151 value: 0.41922
152 - filter: model.layers.21.self_attn.v_proj
153 value: 0.32166
154 - filter: model.layers.22.self_attn.v_proj
155 value: 0.34057
156 - filter: model.layers.23.self_attn.v_proj
157 value: 0.24319
158 - filter: model.layers.24.self_attn.v_proj
159 value: 0.07956
160 - filter: model.layers.25.self_attn.v_proj
161 value: 0.13849
162 - filter: model.layers.0.self_attn.o_proj
163 value: 0.03265
164 - filter: model.layers.1.self_attn.o_proj
165 value: 0.22645
166 - filter: model.layers.2.self_attn.o_proj
167 value: 0.06134
168 - filter: model.layers.3.self_attn.o_proj
169 value: 0.16228
170 - filter: model.layers.4.self_attn.o_proj
171 value: 0.07924
172 - filter: model.layers.5.self_attn.o_proj
173 value: 0.1259
174 - filter: model.layers.6.self_attn.o_proj
175 value: 0.09982
176 - filter: model.layers.7.self_attn.o_proj
177 value: 0.02826
178 - filter: model.layers.8.self_attn.o_proj
179 value: 0.03906
180 - filter: model.layers.9.self_attn.o_proj
181 value: 1.0
182 - filter: model.layers.10.self_attn.o_proj
183 value: 0.51227
184 - filter: model.layers.11.self_attn.o_proj
185 value: 0.34671
186 - filter: model.layers.12.self_attn.o_proj
187 value: 0.70159
188 - filter: model.layers.13.self_attn.o_proj
189 value: 0.4424
190 - filter: model.layers.14.self_attn.o_proj
191 value: 0.74354
192 - filter: model.layers.15.self_attn.o_proj
193 value: 0.16266
194 - filter: model.layers.16.self_attn.o_proj
195 value: 0.71896
196 - filter: model.layers.17.self_attn.o_proj
197 value: 0.53179
198 - filter: model.layers.18.self_attn.o_proj
199 value: 0.18472
200 - filter: model.layers.19.self_attn.o_proj
201 value: 0.09507
202 - filter: model.layers.20.self_attn.o_proj
203 value: 0.13472
204 - filter: model.layers.21.self_attn.o_proj
205 value: 0.23246
206 - filter: model.layers.22.self_attn.o_proj
207 value: 0.10599
208 - filter: model.layers.23.self_attn.o_proj
209 value: 0.00282
210 - filter: model.layers.24.self_attn.o_proj
211 value: 0.09864
212 - filter: model.layers.25.self_attn.o_proj
213 value: 0.00961
214 - filter: model.layers.0.mlp.gate_proj
215 value: 0.4928
216 - filter: model.layers.1.mlp.gate_proj
217 value: 0.08775
218 - filter: model.layers.2.mlp.gate_proj
219 value: 0.0002
220 - filter: model.layers.3.mlp.gate_proj
221 value: 0.12885
222 - filter: model.layers.4.mlp.gate_proj
223 value: 0.50973
224 - filter: model.layers.5.mlp.gate_proj
225 value: 0.29085
226 - filter: model.layers.6.mlp.gate_proj
227 value: 0.06577
228 - filter: model.layers.7.mlp.gate_proj
229 value: 0.11425
230 - filter: model.layers.8.mlp.gate_proj
231 value: 0.12968
232 - filter: model.layers.9.mlp.gate_proj
233 value: 0.19222
234 - filter: model.layers.10.mlp.gate_proj
235 value: 0.17268
236 - filter: model.layers.11.mlp.gate_proj
237 value: 0.10512
238 - filter: model.layers.12.mlp.gate_proj
239 value: 0.02651
240 - filter: model.layers.13.mlp.gate_proj
241 value: 0.04687
242 - filter: model.layers.14.mlp.gate_proj
243 value: 0.14716
244 - filter: model.layers.15.mlp.gate_proj
245 value: 0.03147
246 - filter: model.layers.16.mlp.gate_proj
247 value: 0.05726
248 - filter: model.layers.17.mlp.gate_proj
249 value: 0.04511
250 - filter: model.layers.18.mlp.gate_proj
251 value: 0.25063
252 - filter: model.layers.19.mlp.gate_proj
253 value: 0.26688
254 - filter: model.layers.20.mlp.gate_proj
255 value: 0.37716
256 - filter: model.layers.21.mlp.gate_proj
257 value: 0.3862
258 - filter: model.layers.22.mlp.gate_proj
259 value: 0.27378
260 - filter: model.layers.23.mlp.gate_proj
261 value: 0.08641
262 - filter: model.layers.24.mlp.gate_proj
263 value: 0.37212
264 - filter: model.layers.25.mlp.gate_proj
265 value: 0.31151
266 - filter: model.layers.0.mlp.up_proj
267 value: 0.24228
268 - filter: model.layers.1.mlp.up_proj
269 value: 0.06887
270 - filter: model.layers.2.mlp.up_proj
271 value: 0.20625
272 - filter: model.layers.3.mlp.up_proj
273 value: 0.27546
274 - filter: model.layers.4.mlp.up_proj
275 value: 0.59112
276 - filter: model.layers.5.mlp.up_proj
277 value: 0.41203
278 - filter: model.layers.6.mlp.up_proj
279 value: 0.07411
280 - filter: model.layers.7.mlp.up_proj
281 value: 0.05424
282 - filter: model.layers.8.mlp.up_proj
283 value: 0.23792
284 - filter: model.layers.9.mlp.up_proj
285 value: 0.1677
286 - filter: model.layers.10.mlp.up_proj
287 value: 0.17054
288 - filter: model.layers.11.mlp.up_proj
289 value: 0.18762
290 - filter: model.layers.12.mlp.up_proj
291 value: 0.08044
292 - filter: model.layers.13.mlp.up_proj
293 value: 0.0021
294 - filter: model.layers.14.mlp.up_proj
295 value: 0.26389
296 - filter: model.layers.15.mlp.up_proj
297 value: 0.06886
298 - filter: model.layers.16.mlp.up_proj
299 value: 0.21476
300 - filter: model.layers.17.mlp.up_proj
301 value: 0.21777
302 - filter: model.layers.18.mlp.up_proj
303 value: 0.32111
304 - filter: model.layers.19.mlp.up_proj
305 value: 0.2885
306 - filter: model.layers.20.mlp.up_proj
307 value: 0.40549
308 - filter: model.layers.21.mlp.up_proj
309 value: 0.42539
310 - filter: model.layers.22.mlp.up_proj
311 value: 0.31214
312 - filter: model.layers.23.mlp.up_proj
313 value: 0.02931
314 - filter: model.layers.24.mlp.up_proj
315 value: 0.3693
316 - filter: model.layers.25.mlp.up_proj
317 value: 0.345
318 - filter: model.layers.0.mlp.down_proj
319 value: 0.06756
320 - filter: model.layers.1.mlp.down_proj
321 value: 0.03746
322 - filter: model.layers.2.mlp.down_proj
323 value: 0.09104
324 - filter: model.layers.3.mlp.down_proj
325 value: 0.06643
326 - filter: model.layers.4.mlp.down_proj
327 value: 0.05003
328 - filter: model.layers.5.mlp.down_proj
329 value: 0.0406
330 - filter: model.layers.6.mlp.down_proj
331 value: 0.01609
332 - filter: model.layers.7.mlp.down_proj
333 value: 0.09629
334 - filter: model.layers.8.mlp.down_proj
335 value: 0.08912
336 - filter: model.layers.9.mlp.down_proj
337 value: 0.12874
338 - filter: model.layers.10.mlp.down_proj
339 value: 0.04635
340 - filter: model.layers.11.mlp.down_proj
341 value: 0.0099
342 - filter: model.layers.12.mlp.down_proj
343 value: 0.03487
344 - filter: model.layers.13.mlp.down_proj
345 value: 0.04977
346 - filter: model.layers.14.mlp.down_proj
347 value: 0.00393
348 - filter: model.layers.15.mlp.down_proj
349 value: 0.00748
350 - filter: model.layers.16.mlp.down_proj
351 value: 0.06696
352 - filter: model.layers.17.mlp.down_proj
353 value: 0.02067
354 - filter: model.layers.18.mlp.down_proj
355 value: 0.10311
356 - filter: model.layers.19.mlp.down_proj
357 value: 0.009
358 - filter: model.layers.20.mlp.down_proj
359 value: 0.0215
360 - filter: model.layers.21.mlp.down_proj
361 value: 0.04196
362 - filter: model.layers.22.mlp.down_proj
363 value: 0.06326
364 - filter: model.layers.23.mlp.down_proj
365 value: 0.21737
366 - filter: model.layers.24.mlp.down_proj
367 value: 0.19338
368 - filter: model.layers.25.mlp.down_proj
369 value: 0.04905
370 weight:
371 - value: 1
372merge_method: ties
373base_model: google/gemma-2-2b
374parameters:
375 normalize: true
376 int8_mask: true
377dtype: bfloat16
378tokenizer_source: union1!pip install -qU transformers accelerate
2
3from transformers import AutoTokenizer
4import transformers
5import torch
6
7model = "choprahetarth/gemma-instruct-merge-named_correctly"
8messages = [{"role": "user", "content": "What is a large language model?"}]
9
10tokenizer = AutoTokenizer.from_pretrained(model)
11prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
12pipeline = transformers.pipeline(
13 "text-generation",
14 model=model,
15 torch_dtype=torch.float16,
16 device_map="auto",
17)
18
19outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
20print(outputs[0]["generated_text"])