4cf557b80c2abbcebf5cb6742ac3111a

This model is a fine-tuned version of Qwen/Qwen2.5-7B on the nyu-mll/glue [stsb] dataset. It achieves the following results on the evaluation set:

• Loss: 9.0028
• Data Size: 1.0
• Epoch Runtime: 167.2591
• Mse: 2.2516
• Mae: 1.2798
• R2: -0.0072

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	Mse	Mae	R2
No log	0	0	162.1640	0	4.9325	40.5406	5.3933	-17.1352
No log	1	179	216.4438	0.0078	6.0286	54.1147	5.4474	-23.2074
No log	2	358	78.2590	0.0156	15.2698	19.5655	3.7806	-7.7524
No log	3	537	71.3912	0.0312	26.6903	17.8493	3.8704	-6.9846
No log	4	716	20.5629	0.0625	36.6249	5.1418	1.8678	-1.3001
No log	5	895	14.8689	0.125	52.8272	3.7177	1.5604	-0.6631
16.3429	6	1074	8.3563	0.25	75.7176	2.0895	1.1038	0.0653
5.983	7	1253	4.2240	0.5	126.6140	1.0560	0.8367	0.5276
4.5779	8.0	1432	3.3550	1.0	195.2600	0.8391	0.7164	0.6246
6.1749	9.0	1611	7.1648	1.0	153.6322	1.7914	1.0613	0.1987
12.2715	10.0	1790	10.9483	1.0	170.9618	2.7376	1.3450	-0.2246
7.0884	11.0	1969	8.7058	1.0	157.1733	2.1772	1.2580	0.0260
9.5085	12.0	2148	9.0028	1.0	167.2591	2.2516	1.2798	-0.0072

Framework versions

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