7f73adcdf5baaee0456d6b8f705ded8b

This model is a fine-tuned version of deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B on the contemmcm/amazon_reviews_2013 [cell-phone] dataset. It achieves the following results on the evaluation set:

• Loss: 7.0451
• Data Size: 1.0
• Epoch Runtime: 585.6772
• Accuracy: 0.6867
• F1 Macro: 0.6125

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	10.8692	0	39.3696	0.1461	0.1118
No log	1	1973	4.4333	0.0078	43.4493	0.5401	0.4237
0.111	2	3946	4.2262	0.0156	49.5090	0.5826	0.4864
3.7109	3	5919	3.6245	0.0312	60.0230	0.6268	0.5091
3.3662	4	7892	3.1600	0.0625	76.7956	0.6725	0.5881
3.1069	5	9865	3.2843	0.125	110.0763	0.6639	0.5282
3.077	6	11838	3.0563	0.25	177.6305	0.6845	0.5349
2.8524	7	13811	2.9521	0.5	314.8432	0.6867	0.6313
2.3612	8.0	15784	2.9180	1.0	591.4517	0.7001	0.6362
1.0246	9.0	17757	4.3063	1.0	585.3902	0.6966	0.6276
0.579	10.0	19730	5.7404	1.0	585.8918	0.6604	0.6139
0.563	11.0	21703	6.3530	1.0	588.3526	0.6838	0.6067
0.3482	12.0	23676	7.0451	1.0	585.6772	0.6867	0.6125

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.2.0
• Tokenizers 0.22.1