004247259dc20293acf19cda27906bdb

This model is a fine-tuned version of Qwen/Qwen2.5-1.5B on the contemmcm/cls_mmlu dataset. It achieves the following results on the evaluation set:

• Loss: 15.6912
• Data Size: 1.0
• Epoch Runtime: 113.1814
• Accuracy: 0.2852
• F1 Macro: 0.2848

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	17.9240	0	3.9888	0.2553	0.1438
No log	1	438	8.5223	0.0078	4.9615	0.2593	0.2356
No log	2	876	7.5222	0.0156	7.2820	0.2493	0.1951
No log	3	1314	6.4074	0.0312	11.6446	0.2434	0.1847
No log	4	1752	5.8273	0.0625	16.6596	0.25	0.1608
0.3679	5	2190	5.8369	0.125	24.2477	0.2527	0.1008
0.7661	6	2628	5.6940	0.25	39.5744	0.2606	0.1491
5.5894	7	3066	5.6361	0.5	63.5454	0.2699	0.2192
5.5147	8.0	3504	5.5449	1.0	117.0560	0.2892	0.2531
4.7651	9.0	3942	6.3449	1.0	111.6204	0.2939	0.2460
3.1101	10.0	4380	7.5634	1.0	111.7050	0.2945	0.2879
2.1903	11.0	4818	9.0287	1.0	112.0035	0.2899	0.2844
1.2029	12.0	5256	15.6912	1.0	113.1814	0.2852	0.2848

Framework versions

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