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
5.2572	0.4212	1000	5.2225
4.6962	0.8424	2000	4.6983
4.1718	1.2632	3000	4.1679
3.8095	1.6844	4000	3.7988
3.5144	2.1053	5000	3.5161
3.3677	2.5265	6000	3.3689
3.2817	2.9477	7000	3.2779
3.2014	3.3685	8000	3.2222
3.1614	3.7897	9000	3.1730
3.0876	4.2106	10000	3.1397
3.0667	4.6318	11000	3.1104
3.0027	5.0526	12000	3.0845
3.0047	5.4738	13000	3.0713
2.9899	5.8950	14000	3.0509
2.9579	6.3159	15000	3.0431
2.9515	6.7371	16000	3.0237
2.8994	7.1579	17000	3.0170
2.9024	7.5791	18000	3.0056
2.9142	8.0	19000	2.9946
2.8702	8.4212	20000	2.9937
2.8774	8.8424	21000	2.9857
2.8328	9.2632	22000	2.9791
2.8397	9.6844	23000	2.9779
2.7992	10.1053	24000	2.9718
2.8177	10.5265	25000	2.9685
2.8145	10.9477	26000	2.9572
2.7861	11.3685	27000	2.9603
2.7929	11.7897	28000	2.9530
2.757	12.2106	29000	2.9557
2.7726	12.6318	30000	2.9507
2.7289	13.0526	31000	2.9496
2.7372	13.4738	32000	2.9480
2.757	13.8950	33000	2.9419
2.7225	14.3159	34000	2.9429
2.735	14.7371	35000	2.9393
2.6988	15.1579	36000	2.9430
2.7095	15.5791	37000	2.9390
2.716	16.0	38000	2.9335
2.6842	16.4212	39000	2.9375
2.6933	16.8424	40000	2.9343
2.6728	17.2632	41000	2.9361
2.6781	17.6844	42000	2.9328
2.6624	18.1053	43000	2.9358
2.6655	18.5265	44000	2.9335
2.6638	18.9477	45000	2.9318
2.6558	19.3685	46000	2.9330
2.6551	19.7897	47000	2.9326

kanishka/opt-babylm1_no_non_preemption_laugh-max_seed-1024_5e-6

opt-babylm1_no_non_preemption_laugh-max_seed-1024_5e-6