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
5.2617	0.4205	1000	5.2241
4.7816	0.8410	2000	4.8340
4.2486	1.2611	3000	4.2311
3.8748	1.6816	4000	3.8693
3.5628	2.1018	5000	3.5683
3.4175	2.5223	6000	3.4130
3.3097	2.9428	7000	3.3074
3.2238	3.3629	8000	3.2413
3.1736	3.7834	9000	3.1924
3.1098	4.2035	10000	3.1558
3.0895	4.6240	11000	3.1208
3.0196	5.0442	12000	3.0997
3.0244	5.4647	13000	3.0777
3.0078	5.8852	14000	3.0602
2.9691	6.3053	15000	3.0481
2.9641	6.7258	16000	3.0326
2.9124	7.1459	17000	3.0257
2.9188	7.5664	18000	3.0127
2.9162	7.9869	19000	3.0019
2.8825	8.4070	20000	2.9992
2.8819	8.8275	21000	2.9857
2.8417	9.2477	22000	2.9876
2.8478	9.6682	23000	2.9768
2.8072	10.0883	24000	2.9783
2.8234	10.5088	25000	2.9709
2.8313	10.9293	26000	2.9641
2.7982	11.3494	27000	2.9645
2.8016	11.7699	28000	2.9595
2.7671	12.1901	29000	2.9607
2.7771	12.6106	30000	2.9509
2.7526	13.0307	31000	2.9547
2.7524	13.4512	32000	2.9507
2.7619	13.8717	33000	2.9432
2.7276	14.2918	34000	2.9489
2.7422	14.7123	35000	2.9424
2.7166	15.1325	36000	2.9442
2.7133	15.5530	37000	2.9409
2.7254	15.9735	38000	2.9372
2.7016	16.3936	39000	2.9404
2.7052	16.8141	40000	2.9368
2.6786	17.2342	41000	2.9388
2.6922	17.6547	42000	2.9364
2.6688	18.0748	43000	2.9377
2.6673	18.4953	44000	2.9361
2.6762	18.9158	45000	2.9343
2.6647	19.3360	46000	2.9356
2.6605	19.7565	47000	2.9350

kanishka/opt-babylm1_more_preemption_laugh_seed-1024_5e-6

opt-babylm1_more_preemption_laugh_seed-1024_5e-6