xiulinyang/GPT2_BABYLM10MCR_50000

10MCR_50000_41

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 4.1225
• Accuracy: 0.3627

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0006
• train_batch_size: 32
• eval_batch_size: 32
• seed: 41
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 2000
• num_epochs: 20.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.5345	1.0	1387	5.5310	0.2657
1.3433	2.0	2774	5.0589	0.2901
1.1935	3.0	4161	4.7845	0.2995
1.1513	4.0	5548	4.6243	0.3058
1.1217	5.0	6935	4.5012	0.3140
1.0756	6.0	8322	4.3986	0.3246
1.0563	7.0	9709	4.3107	0.3362
1.0251	8.0	11096	4.2487	0.3448
1.0088	9.0	12483	4.2077	0.3508
1.0028	10.0	13870	4.1780	0.3543
0.9847	11.0	15257	4.1589	0.3566
0.9775	12.0	16644	4.1493	0.3579
0.9668	13.0	18031	4.1424	0.3592
0.9576	14.0	19418	4.1357	0.3601
0.9503	15.0	20805	4.1290	0.3609
0.9387	16.0	22192	4.1266	0.3613
0.9302	17.0	23579	4.1250	0.3617
0.926	18.0	24966	4.1233	0.3620
0.9138	19.0	26353	4.1249	0.3624
0.9072	20.0	27740	4.1225	0.3627

Framework versions

• Transformers 4.51.3
• Pytorch 2.5.1+cu121
• Datasets 3.6.0
• Tokenizers 0.21.1