a92c401f84d87638c978908ad9f47027

This model is a fine-tuned version of deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B on the nyu-mll/glue dataset. It achieves the following results on the evaluation set:

• Loss: 2.0579
• Data Size: 1.0
• Epoch Runtime: 45.5877
• Mse: 0.5147
• Mae: 0.5557
• R2: 0.7698

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	Mse	Mae	R2
No log	0	0	25.6358	0	3.8811	6.4103	2.0828	-1.8676
No log	1	179	23.8544	0.0078	4.1876	5.9642	1.9817	-1.6680
No log	2	358	18.1604	0.0156	6.0646	4.5407	1.7715	-1.0312
No log	3	537	6.9024	0.0312	9.5076	1.7262	1.1003	0.2278
No log	4	716	3.0629	0.0625	10.9463	0.7658	0.6853	0.6574
No log	5	895	4.1229	0.125	15.7994	1.0309	0.8144	0.5388
0.463	6	1074	2.3133	0.25	20.1048	0.5784	0.5883	0.7412
2.9086	7	1253	3.0553	0.5	30.3291	0.7641	0.7083	0.6582
2.1353	8.0	1432	2.7786	1.0	51.9329	0.6945	0.6545	0.6893
1.1545	9.0	1611	2.5051	1.0	43.8829	0.6263	0.6069	0.7198
0.946	10.0	1790	2.1844	1.0	46.3699	0.5463	0.5781	0.7556
0.7317	11.0	1969	2.1323	1.0	47.8290	0.5332	0.5693	0.7615
0.5233	12.0	2148	2.2708	1.0	44.1612	0.5679	0.6047	0.7460
0.4518	13.0	2327	1.9611	1.0	45.0843	0.4905	0.5554	0.7806
0.4331	14.0	2506	2.0977	1.0	45.1636	0.5246	0.5522	0.7653
0.3029	15.0	2685	1.8147	1.0	43.5439	0.4538	0.5191	0.7970
0.2731	16.0	2864	1.9673	1.0	42.8446	0.4920	0.5414	0.7799
0.2468	17.0	3043	1.8620	1.0	44.9792	0.4656	0.5201	0.7917
0.2374	18.0	3222	1.9823	1.0	43.4776	0.4957	0.5427	0.7783
0.2261	19.0	3401	2.0579	1.0	45.5877	0.5147	0.5557	0.7698

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.1