f985cdb9b2571556b9b78b76ff5dd3e1

This model is a fine-tuned version of deepseek-ai/DeepSeek-R1-Distill-Qwen-7B on the google/boolq dataset. It achieves the following results on the evaluation set:

• Loss: 2.9734
• Data Size: 1.0
• Epoch Runtime: 733.8298
• Accuracy: 0.7739
• F1 Macro: 0.7346
• Rouge1: 0.7745
• Rouge2: 0.0
• Rougel: 0.7736
• Rougelsum: 0.7736

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	13.0664	0	20.7375	0.6170	0.3926	0.6170	0.0	0.6164	0.6167
No log	1	294	24.5370	0.0078	25.9441	0.6213	0.3847	0.6213	0.0	0.6210	0.6213
No log	2	588	2.6574	0.0156	36.3139	0.6167	0.4511	0.6163	0.0	0.6161	0.6166
No log	3	882	3.5147	0.0312	56.9167	0.3854	0.2944	0.3854	0.0	0.3857	0.3857
0.2744	4	1176	2.8565	0.0625	83.2449	0.6860	0.6839	0.6863	0.0	0.6857	0.6857
0.1958	5	1470	2.0035	0.125	124.5276	0.7687	0.7496	0.7687	0.0	0.7687	0.7690
0.3183	6	1764	3.1689	0.25	206.4032	0.5086	0.4867	0.5086	0.0	0.5083	0.5083
1.8021	7	2058	2.6540	0.5	385.1019	0.6676	0.6674	0.6673	0.0	0.6674	0.6676
1.3798	8.0	2352	2.5272	1.0	733.7345	0.7341	0.7323	0.7344	0.0	0.7350	0.7338
0.713	9.0	2646	2.9734	1.0	733.8298	0.7739	0.7346	0.7745	0.0	0.7736	0.7736

Framework versions

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