learning_rate: 5e-06
train_batch_size: 32
eval_batch_size: 64
seed: 1024
gradient_accumulation_steps: 8
total_train_batch_size: 256
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.05
num_epochs: 20.0

Training Loss	Epoch	Step	Validation Loss
5.0473	0.4248	1000	5.0233
4.5125	0.8495	2000	4.5246
4.0498	1.2740	3000	4.0511
3.7344	1.6987	4000	3.7227
3.4596	2.1232	5000	3.4661
3.3253	2.5479	6000	3.3277
3.243	2.9727	7000	3.2375
3.1534	3.3972	8000	3.1808
3.1165	3.8219	9000	3.1339
3.0558	4.2464	10000	3.1016
3.0382	4.6711	11000	3.0752
2.9669	5.0956	12000	3.0545
2.9732	5.5203	13000	3.0399
2.9642	5.9451	14000	3.0161
2.9233	6.3695	15000	3.0080
2.9238	6.7943	16000	2.9950
2.8728	7.2188	17000	2.9866
2.8796	7.6435	18000	2.9766
2.8314	8.0680	19000	2.9725
2.8479	8.4927	20000	2.9632
2.8474	8.9175	21000	2.9558
2.8131	9.3419	22000	2.9520
2.8169	9.7667	23000	2.9436
2.7759	10.1911	24000	2.9473
2.7885	10.6159	25000	2.9414
2.7448	11.0404	26000	2.9390
2.7642	11.4651	27000	2.9355
2.7631	11.8899	28000	2.9273
2.7327	12.3143	29000	2.9303
2.743	12.7391	30000	2.9269
2.711	13.1635	31000	2.9293
2.7193	13.5883	32000	2.9227
2.7137	14.0127	33000	2.9248
2.7011	14.4375	34000	2.9193
2.7038	14.8623	35000	2.9168
2.6771	15.2867	36000	2.9184
2.6873	15.7115	37000	2.9154
2.6619	16.1359	38000	2.9190
2.6648	16.5607	39000	2.9151
2.6718	16.9855	40000	2.9122
2.6562	17.4099	41000	2.9157
2.6538	17.8347	42000	2.9118
2.6366	18.2591	43000	2.9139
2.6392	18.6839	44000	2.9134
2.6311	19.1083	45000	2.9134
2.6278	19.5331	46000	2.9131
2.631	19.9578	47000	2.9125

kanishka/opt-babylm1_no_preemption_laugh-lemmatized_seed-1024_5e-6

opt-babylm1_no_preemption_laugh-lemmatized_seed-1024_5e-6