07389fe75bfb1ccb0e881cbcad2f348e

This model is a fine-tuned version of facebook/opt-6.7b on the nyu-mll/glue [qqp] dataset. It achieves the following results on the evaluation set:

• Loss: 0.5064
• Data Size: 1.0
• Epoch Runtime: 4564.0172
• Accuracy: 0.7824
• F1 Macro: 0.7638
• Rouge1: 0.7824
• Rouge2: 0.0
• Rougel: 0.7826
• Rougelsum: 0.7824

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	0.7932	0	101.2218	0.5212	0.4880	0.5212	0.0	0.5211	0.5210
1.1125	1	11370	0.8546	0.0078	136.0722	0.6320	0.3873	0.6319	0.0	0.6320	0.6318
0.697	2	22740	0.6579	0.0156	175.5793	0.6320	0.3872	0.6318	0.0	0.6319	0.6317
0.6291	3	34110	0.6202	0.0312	252.1490	0.6773	0.5611	0.6771	0.0	0.6772	0.6771
0.5387	4	45480	0.5599	0.0625	393.0325	0.7282	0.6797	0.7280	0.0	0.7282	0.7281
0.5332	5	56850	0.5391	0.125	669.3164	0.7267	0.7137	0.7266	0.0	0.7267	0.7266
0.4894	6	68220	0.4992	0.25	1223.3237	0.7546	0.7313	0.7545	0.0	0.7546	0.7544
0.478	7	79590	0.5050	0.5	2343.5491	0.7498	0.7413	0.7498	0.0	0.7497	0.7497
0.4237	8.0	90960	0.4668	1.0	4572.9409	0.7814	0.7546	0.7813	0.0	0.7815	0.7812
0.4143	9.0	102330	0.4593	1.0	4561.9115	0.7790	0.7660	0.7789	0.0	0.7791	0.7789
0.3938	10.0	113700	0.4585	1.0	4568.8218	0.7886	0.7684	0.7885	0.0	0.7887	0.7884
0.3827	11.0	125070	0.4821	1.0	4570.3422	0.7796	0.7681	0.7796	0.0	0.7796	0.7795
0.3455	12.0	136440	0.4752	1.0	4567.8479	0.7846	0.7678	0.7846	0.0	0.7846	0.7845
0.3309	13.0	147810	0.4775	1.0	4568.5633	0.7857	0.7655	0.7855	0.0	0.7857	0.7856
0.3095	14.0	159180	0.5064	1.0	4564.0172	0.7824	0.7638	0.7824	0.0	0.7826	0.7824

Framework versions

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