bddcdd7f73257a564691f579c1239a9f

This model is a fine-tuned version of Qwen/Qwen2.5-1.5B on the nyu-mll/glue [stsb] dataset. It achieves the following results on the evaluation set:

• Loss: 3.0761
• Data Size: 1.0
• Epoch Runtime: 44.0479
• Mse: 0.7692
• Mae: 0.6911
• R2: 0.6559

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	91.4851	0	3.9236	22.8725	4.2807	-9.2317
No log	1	179	61.2800	0.0078	4.2263	15.3208	3.3462	-5.8536
No log	2	358	31.5547	0.0156	5.9445	7.8890	2.3658	-2.5290
No log	3	537	4.2059	0.0312	8.7044	1.0519	0.8157	0.5294
No log	4	716	3.0331	0.0625	11.7948	0.7585	0.6920	0.6607
No log	5	895	4.6288	0.125	15.6273	1.1578	0.8708	0.4821
0.8543	6	1074	4.4606	0.25	20.0331	1.1157	0.8022	0.5009
3.0884	7	1253	3.2438	0.5	30.7546	0.8110	0.6936	0.6372
2.4561	8.0	1432	2.2554	1.0	51.5671	0.5640	0.5975	0.7477
12.4135	9.0	1611	2.4206	1.0	43.9917	0.6052	0.6002	0.7293
3.9307	10.0	1790	2.7132	1.0	46.4324	0.6783	0.6446	0.6966
2.1976	11.0	1969	3.1616	1.0	48.3478	0.7906	0.6778	0.6464
2.2256	12.0	2148	3.0761	1.0	44.0479	0.7692	0.6911	0.6559

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.3.0
• Tokenizers 0.22.1