bb474ca2df14838287e240f58e07af31

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: 2.9459
• Data Size: 1.0
• Epoch Runtime: 2556.2673
• Accuracy: 0.8046
• F1 Macro: 0.8041
• Rouge1: 0.8046
• Rouge2: 0.0
• Rougel: 0.8047
• Rougelsum: 0.8045

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	6.9209	0	19.3062	0.3417	0.2743	0.3416	0.0	0.3421	0.3419
3.9212	1	12271	2.3606	0.0078	38.8233	0.7662	0.7642	0.7663	0.0	0.7662	0.7664
2.2858	2	24542	2.1807	0.0156	60.8958	0.7860	0.7832	0.7858	0.0	0.7859	0.7862
2.1527	3	36813	2.1088	0.0312	101.1695	0.7949	0.7923	0.7949	0.0	0.7947	0.7949
2.0216	4	49084	1.9452	0.0625	181.3092	0.7998	0.7984	0.7996	0.0	0.7997	0.7998
1.8085	5	61355	1.9626	0.125	334.2134	0.8100	0.8081	0.8099	0.0	0.8100	0.8102
1.8206	6	73626	2.0119	0.25	654.9177	0.8023	0.8019	0.8023	0.0	0.8022	0.8021
1.4947	7	85897	1.9356	0.5	1288.6444	0.8111	0.8108	0.8111	0.0	0.8110	0.8112
1.299	8.0	98168	1.9056	1.0	2540.0394	0.8214	0.8213	0.8213	0.0	0.8213	0.8213
0.8721	9.0	110439	2.3002	1.0	2548.0075	0.8146	0.8140	0.8145	0.0	0.8146	0.8146
0.6403	10.0	122710	2.7267	1.0	2568.1349	0.8078	0.8070	0.8078	0.0	0.8077	0.8077
0.5967	11.0	134981	2.7967	1.0	2639.9235	0.8144	0.8137	0.8144	0.0	0.8144	0.8143
0.5792	12.0	147252	2.9459	1.0	2556.2673	0.8046	0.8041	0.8046	0.0	0.8047	0.8045

Framework versions

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