learning_rate: 5e-06
train_batch_size: 32
eval_batch_size: 64
seed: 211
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
5.4198	0.4303	1000	5.4380
4.7349	0.8605	2000	4.7845
4.241	1.2904	3000	4.3278
3.8765	1.7207	4000	3.9904
3.5718	2.1506	5000	3.7079
3.4203	2.5809	6000	3.5557
3.3169	3.0108	7000	3.4696
3.2454	3.4410	8000	3.4020
3.1936	3.8713	9000	3.3507
3.1276	4.3012	10000	3.3197
3.1097	4.7315	11000	3.2871
3.0463	5.1614	12000	3.2759
3.0413	5.5916	13000	3.2444
3.0013	6.0215	14000	3.2374
2.9838	6.4518	15000	3.2207
2.982	6.8821	16000	3.2050
2.9424	7.3119	17000	3.2055
2.9415	7.7422	18000	3.1893
2.9018	8.1721	19000	3.1835
2.906	8.6024	20000	3.1788
2.8656	9.0323	21000	3.1725
2.8663	9.4625	22000	3.1650
2.8795	9.8928	23000	3.1558
2.839	10.3227	24000	3.1620
2.8437	10.7530	25000	3.1506
2.8036	11.1829	26000	3.1522
2.8195	11.6131	27000	3.1481
2.7785	12.0430	28000	3.1492
2.7926	12.4733	29000	3.1411
2.798	12.9036	30000	3.1386
2.7676	13.3335	31000	3.1407
2.7774	13.7637	32000	3.1352
2.7472	14.1936	33000	3.1385
2.7536	14.6239	34000	3.1361
2.7202	15.0538	35000	3.1401
2.7365	15.4841	36000	3.1343
2.741	15.9143	37000	3.1285
2.7152	16.3442	38000	3.1361
2.7167	16.7745	39000	3.1320
2.7042	17.2044	40000	3.1348
2.7047	17.6346	41000	3.1340
2.6828	18.0645	42000	3.1343
2.688	18.4948	43000	3.1316
2.6892	18.9251	44000	3.1308
2.6748	19.3550	45000	3.1335
2.6753	19.7852	46000	3.1332

kanishka/opt-babylm1-ntb-ntx_seed-211_5e-6

opt-babylm1-ntb-ntx_seed-211_5e-6