a02745c90c6bd40f1208ccb5a6cc7fc4

This model is a fine-tuned version of Qwen/Qwen2.5-0.5B on the nyu-mll/glue [qnli] dataset. It achieves the following results on the evaluation set:

• Loss: 3.4280
• Data Size: 1.0
• Epoch Runtime: 545.0345
• Accuracy: 0.8020
• F1 Macro: 0.8019
• Rouge1: 0.8020
• Rouge2: 0.0
• Rougel: 0.8024
• Rougelsum: 0.8018

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	9.2139	0	10.2240	0.4952	0.4096	0.4952	0.0	0.4956	0.4954
No log	1	3273	5.5353	0.0078	14.5682	0.5108	0.3527	0.5103	0.0	0.5108	0.5106
0.0843	2	6546	1.8616	0.0156	19.2074	0.7965	0.7965	0.7967	0.0	0.7967	0.7961
2.3859	3	9819	2.2006	0.0312	28.7076	0.7574	0.7540	0.7579	0.0	0.7574	0.7572
2.149	4	13092	2.1578	0.0625	44.6483	0.7461	0.7375	0.7461	0.0	0.7467	0.7461
1.9333	5	16365	1.8505	0.125	78.9568	0.7831	0.7808	0.7829	0.0	0.7835	0.7831
1.8623	6	19638	1.7715	0.25	144.7326	0.7994	0.7969	0.7994	0.0	0.7996	0.7993
1.6272	7	22911	1.8130	0.5	278.9048	0.8037	0.8029	0.8039	0.0	0.8038	0.8035
1.5843	8.0	26184	2.0011	1.0	532.0536	0.7882	0.7849	0.7882	0.0	0.7882	0.7884
1.0375	9.0	29457	1.7503	1.0	540.9046	0.8303	0.8303	0.8303	0.0	0.8306	0.8303
0.7357	10.0	32730	2.2392	1.0	540.7577	0.8210	0.8206	0.8213	0.0	0.8211	0.8209
0.5658	11.0	36003	2.7259	1.0	535.2230	0.8092	0.8091	0.8090	0.0	0.8092	0.8094
0.5091	12.0	39276	2.8143	1.0	527.6234	0.7956	0.7952	0.7954	0.0	0.7957	0.7957
0.3756	13.0	42549	3.4280	1.0	545.0345	0.8020	0.8019	0.8020	0.0	0.8024	0.8018

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.2.0
• Tokenizers 0.22.1