a0268468b1e0561fd1b6cc905629b669

This model is a fine-tuned version of FacebookAI/roberta-large on the contemmcm/cls_mmlu dataset. It achieves the following results on the evaluation set:

• Loss: 1.3967
• Data Size: 1.0
• Epoch Runtime: 69.1588
• 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	1.4608	0	2.6725	0.2453	0.0985
No log	1	438	1.3949	0.0078	3.5603	0.2493	0.1542
No log	2	876	1.4053	0.0156	4.4231	0.2453	0.0985
No log	3	1314	1.4903	0.0312	5.9256	0.2487	0.0996
No log	4	1752	1.3935	0.0625	8.3380	0.2527	0.1008
0.0789	5	2190	1.4071	0.125	12.6449	0.2527	0.1008
0.186	6	2628	1.3925	0.25	21.2752	0.2487	0.0996
1.4052	7	3066	1.3882	0.5	38.2134	0.2533	0.1011
1.4034	8.0	3504	1.3966	1.0	69.6853	0.2527	0.1008
1.4028	9.0	3942	1.4049	1.0	69.2272	0.2533	0.1011
1.3995	10.0	4380	1.3904	1.0	68.9935	0.2527	0.1008
1.3942	11.0	4818	1.3967	1.0	69.1588	0.2527	0.1008

Framework versions

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