1e151ca42cd81c945e904fbd09cf6cb9

This model is a fine-tuned version of distilbert/distilgpt2 on the nyu-mll/glue [cola] dataset. It achieves the following results on the evaluation set:

• Loss: 0.7609
• Data Size: 1.0
• Epoch Runtime: 11.7662
• Accuracy: 0.7451
• F1 Macro: 0.6632
• Rouge1: 0.7451
• Rouge2: 0.0
• Rougel: 0.7461
• Rougelsum: 0.7441

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	6.5225	0	1.3553	0.3115	0.2375	0.3105	0.0	0.3115	0.3115
No log	1	267	2.9999	0.0078	1.7172	0.3105	0.2370	0.3096	0.0	0.3096	0.3105
No log	2	534	1.1866	0.0156	1.5736	0.6650	0.4447	0.6660	0.0	0.6650	0.6641
No log	3	801	0.8160	0.0312	2.0617	0.6055	0.4979	0.6055	0.0	0.6064	0.6055
No log	4	1068	0.6201	0.0625	2.1709	0.6855	0.4067	0.6865	0.0	0.6855	0.6855
0.0593	5	1335	0.6281	0.125	2.7571	0.6855	0.4067	0.6865	0.0	0.6855	0.6855
0.6148	6	1602	0.6450	0.25	4.0643	0.6885	0.4078	0.6895	0.0	0.6885	0.6885
0.5935	7	1869	0.6544	0.5	6.6404	0.6855	0.4067	0.6865	0.0	0.6855	0.6855
0.5543	8.0	2136	0.5893	1.0	11.6649	0.7012	0.4866	0.7021	0.0	0.7012	0.7012
0.4819	9.0	2403	0.6055	1.0	11.5366	0.7266	0.5821	0.7266	0.0	0.7266	0.7256
0.4271	10.0	2670	0.6059	1.0	11.4565	0.7197	0.6145	0.7197	0.0	0.7192	0.7197
0.3136	11.0	2937	0.6672	1.0	11.6719	0.7354	0.6503	0.7354	0.0	0.7354	0.7354
0.2923	12.0	3204	0.7609	1.0	11.7662	0.7451	0.6632	0.7451	0.0	0.7461	0.7441

Framework versions

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