e1cb0ddaecaf2fdc8575527837db502c

This model is a fine-tuned version of google-bert/bert-large-uncased-whole-word-masking-finetuned-squad on the contemmcm/cls_mmlu dataset. It achieves the following results on the evaluation set:

• Loss: 1.3914
• Data Size: 1.0
• Epoch Runtime: 44.6866
• Accuracy: 0.2527
• F1 Macro: 0.1008

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Accuracy	F1 Macro
No log	0	0	1.4805	0	1.7962	0.2453	0.0985
No log	1	438	1.5063	0.0078	2.4461	0.2527	0.1008
No log	2	876	1.4117	0.0156	3.0836	0.2527	0.1008
No log	3	1314	1.4564	0.0312	4.4745	0.2487	0.0996
No log	4	1752	1.3931	0.0625	6.3609	0.2527	0.1008
0.0795	5	2190	1.4010	0.125	9.2850	0.2453	0.0985
0.1868	6	2628	1.3984	0.25	14.8389	0.2487	0.0996
1.4087	7	3066	1.3915	0.5	24.5142	0.2487	0.0996
1.4009	8.0	3504	1.3900	1.0	46.0448	0.2487	0.0996
1.4033	9.0	3942	1.3948	1.0	45.9171	0.2533	0.1011
1.4005	10.0	4380	1.3908	1.0	45.5259	0.2527	0.1008
1.4111	11.0	4818	1.3912	1.0	45.4028	0.2533	0.1011
1.4059	12.0	5256	1.3914	1.0	44.6866	0.2527	0.1008

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.3.0
• Tokenizers 0.22.1