e984c0a18b4bb0b3129fc01967cc11e8

This model is a fine-tuned version of deepseek-ai/DeepSeek-R1-Distill-Qwen-7B on the nyu-mll/glue [sst2] dataset. It achieves the following results on the evaluation set:

• Loss: 3.6548
• Data Size: 1.0
• Epoch Runtime: 1682.1848
• Accuracy: 0.8484
• F1 Macro: 0.8479
• Rouge1: 0.8484
• Rouge2: 0.0
• Rougel: 0.8484
• Rougelsum: 0.8484

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	4.4033	0	3.2051	0.4676	0.4285	0.4676	0.0	0.4676	0.4664
No log	1	2104	3.5203	0.0078	16.0487	0.4861	0.3291	0.4861	0.0	0.4861	0.4873
No log	2	4208	4.0656	0.0156	32.5313	0.5093	0.3374	0.5093	0.0	0.5093	0.5081
0.0762	3	6312	1.8766	0.0312	68.3174	0.7743	0.7743	0.7743	0.0	0.7743	0.7731
1.6694	4	8416	1.5764	0.0625	118.3939	0.8380	0.8372	0.8380	0.0	0.8380	0.8380
1.3589	5	10520	1.4074	0.125	230.0990	0.8414	0.8406	0.8414	0.0	0.8414	0.8414
1.0252	6	12624	1.3402	0.25	437.4193	0.8704	0.8701	0.8704	0.0	0.8704	0.8704
0.8041	7	14728	1.2264	0.5	856.1002	0.875	0.8750	0.875	0.0	0.875	0.875
0.6276	8.0	16832	1.5403	1.0	1702.3318	0.8542	0.8532	0.8530	0.0	0.8542	0.8542
0.5738	9.0	18936	2.3254	1.0	1693.5311	0.8461	0.8457	0.8449	0.0	0.8461	0.8461
0.7041	10.0	21040	2.1962	1.0	1689.3121	0.8461	0.8458	0.8461	0.0	0.8461	0.8461
0.3462	11.0	23144	3.6548	1.0	1682.1848	0.8484	0.8479	0.8484	0.0	0.8484	0.8484

Framework versions

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