030e60e4949fdf26930d716dbcd8639b

This model is a fine-tuned version of Qwen/Qwen2.5-3B on the nyu-mll/glue [cola] dataset. It achieves the following results on the evaluation set:

• Loss: 2.6070
• Data Size: 1.0
• Epoch Runtime: 91.6652
• Accuracy: 0.6836
• F1 Macro: 0.4340
• Rouge1: 0.6836
• Rouge2: 0.0
• Rougel: 0.6846
• Rougelsum: 0.6836

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	5.3980	0	3.9840	0.6826	0.4201	0.6836	0.0	0.6826	0.6826
No log	1	267	13.1002	0.0078	5.2651	0.3115	0.2375	0.3105	0.0	0.3115	0.3115
No log	2	534	3.8134	0.0156	8.4708	0.4180	0.4038	0.4180	0.0	0.4180	0.4175
No log	3	801	4.6969	0.0312	14.3924	0.6885	0.4078	0.6895	0.0	0.6885	0.6885
No log	4	1068	2.3961	0.0625	22.0729	0.7295	0.6258	0.7295	0.0	0.7285	0.7285
0.2189	5	1335	3.2111	0.125	30.6897	0.6885	0.4078	0.6895	0.0	0.6885	0.6885
3.2688	6	1602	2.3983	0.25	37.4425	0.6914	0.4320	0.6924	0.0	0.6919	0.6914
2.6049	7	1869	2.7742	0.5	55.2769	0.5361	0.5304	0.5352	0.0	0.5352	0.5366
2.4899	8.0	2136	2.6070	1.0	91.6652	0.6836	0.4340	0.6836	0.0	0.6846	0.6836

Framework versions

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