0ec236299b317d76c0f94de06fc85471

This model is a fine-tuned version of meta-llama/Llama-3.2-1B-Instruct on the contemmcm/cls_mmlu dataset. It achieves the following results on the evaluation set:

• Loss: 5.8746
• Data Size: 1.0
• Epoch Runtime: 78.5633
• Accuracy: 0.2779
• F1 Macro: 0.2241

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	9.9364	0	2.9379	0.2420	0.1876
No log	1	438	11.6798	0.0078	3.5674	0.2453	0.0998
No log	2	876	6.0949	0.0156	5.5874	0.2374	0.1504
No log	3	1314	5.8527	0.0312	8.3921	0.2487	0.0996
No log	4	1752	5.9142	0.0625	11.6323	0.2354	0.1503
0.3881	5	2190	6.0489	0.125	17.1439	0.2620	0.1625
0.7521	6	2628	5.5870	0.25	26.2689	0.2666	0.1350
5.6417	7	3066	5.6066	0.5	44.1598	0.2460	0.0999
5.6381	8.0	3504	5.6027	1.0	81.3182	0.2527	0.1008
5.5548	9.0	3942	5.7473	1.0	80.5339	0.2533	0.1011
5.1021	10.0	4380	5.8746	1.0	78.5633	0.2779	0.2241

Framework versions

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