f836ba7a8b6b63b23b2d7dfe73e6a93b

This model is a fine-tuned version of facebook/opt-1.3b on the nyu-mll/glue [qqp] dataset. It achieves the following results on the evaluation set:

• Loss: 0.4847
• Data Size: 0.125
• Epoch Runtime: 270.3211
• Accuracy: 0.7635
• F1 Macro: 0.7472
• Rouge1: 0.7634
• Rouge2: 0.0
• Rougel: 0.7636
• Rougelsum: 0.7635

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	0.9487	0	55.2289	0.6244	0.3878	0.6243	0.0	0.6244	0.6242
0.7834	1	11370	0.4724	0.0078	68.1750	0.7671	0.7496	0.7672	0.0	0.7671	0.7670
0.5312	2	22740	0.5306	0.0156	83.1091	0.7342	0.6696	0.7343	0.0	0.7343	0.7342
0.5258	3	34110	0.5066	0.0312	111.1156	0.7470	0.7075	0.7470	0.0	0.7470	0.7470
0.4594	4	45480	0.5209	0.0625	162.7971	0.7303	0.7282	0.7303	0.0	0.7304	0.7302
0.5087	5	56850	0.4847	0.125	270.3211	0.7635	0.7472	0.7634	0.0	0.7636	0.7635

Framework versions

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