79cff23691bdf7bc142a280818fc76b4

This model is a fine-tuned version of deepseek-ai/DeepSeek-R1-Distill-Qwen-7B on the nyu-mll/glue [cola] dataset. It achieves the following results on the evaluation set:

• Loss: 2.8845
• Data Size: 1.0
• Epoch Runtime: 238.7417
• Accuracy: 0.7432
• F1 Macro: 0.7063
• Rouge1: 0.7422
• Rouge2: 0.0
• Rougel: 0.7441
• Rougelsum: 0.7432

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	Rouge1	Rouge2	Rougel	Rougelsum
No log	0	0	5.7324	0	3.5682	0.6855	0.4097	0.6865	0.0	0.6855	0.6855
No log	1	267	20.8798	0.0078	6.2864	0.6768	0.4552	0.6777	0.0	0.6768	0.6768
No log	2	534	12.1900	0.0156	15.2472	0.3330	0.2803	0.3330	0.0	0.3320	0.3330
No log	3	801	3.3454	0.0312	27.0312	0.5098	0.5096	0.5098	0.0	0.5088	0.5088
No log	4	1068	2.3817	0.0625	42.5132	0.7266	0.6462	0.7256	0.0	0.7256	0.7266
0.2699	5	1335	2.3495	0.125	61.6060	0.6963	0.4477	0.6973	0.0	0.6963	0.6963
2.2706	6	1602	2.3196	0.25	90.7402	0.7041	0.6139	0.7031	0.0	0.7031	0.7041
2.0848	7	1869	2.1867	0.5	138.8149	0.7441	0.6571	0.7446	0.0	0.7451	0.7441
1.7785	8.0	2136	2.5318	1.0	245.0176	0.7686	0.6972	0.7686	0.0	0.7686	0.7676
1.0437	9.0	2403	3.5520	1.0	241.3882	0.7852	0.7214	0.7852	0.0	0.7861	0.7852
0.9005	10.0	2670	3.2318	1.0	231.7098	0.7480	0.6530	0.7485	0.0	0.7480	0.7480
0.5771	11.0	2937	2.8845	1.0	238.7417	0.7432	0.7063	0.7422	0.0	0.7441	0.7432

Framework versions

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