07aac8d827bc5d497e9e92a875f357fe

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

• Loss: 0.6961
• Data Size: 1.0
• Epoch Runtime: 59.5930
• Accuracy: 0.5625
• F1 Macro: 0.36
• Rouge1: 0.5625
• Rouge2: 0.0
• Rougel: 0.5625
• Rougelsum: 0.5625

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	1.0667	0	0.8538	0.5625	0.3905	0.5625	0.0	0.5625	0.5625
No log	1	19	12.4922	0.0078	1.5088	0.4375	0.3043	0.4375	0.0	0.4375	0.4375
No log	2	38	5.4522	0.0156	9.2289	0.4375	0.3043	0.4375	0.0	0.4375	0.4375
No log	3	57	1.6213	0.0312	23.9743	0.5625	0.36	0.5625	0.0	0.5625	0.5625
No log	4	76	1.1822	0.0625	38.1882	0.5625	0.36	0.5625	0.0	0.5625	0.5625
No log	5	95	0.7029	0.125	46.5441	0.4375	0.3043	0.4375	0.0	0.4375	0.4375
0.2618	6	114	0.8439	0.25	51.6454	0.4375	0.3043	0.4375	0.0	0.4375	0.4375
0.2618	7	133	0.6961	0.5	66.8749	0.4688	0.4603	0.4688	0.0	0.4688	0.4688
0.6014	8.0	152	0.6901	1.0	25.3694	0.5625	0.36	0.5625	0.0	0.5625	0.5625
0.6014	9.0	171	0.7054	1.0	27.2986	0.5625	0.36	0.5625	0.0	0.5625	0.5625
0.6014	10.0	190	0.7043	1.0	38.4476	0.5469	0.3535	0.5469	0.0	0.5469	0.5469
0.7266	11.0	209	0.8239	1.0	49.8182	0.4375	0.3043	0.4375	0.0	0.4375	0.4375
0.7266	12.0	228	0.6961	1.0	59.5930	0.5625	0.36	0.5625	0.0	0.5625	0.5625

Framework versions

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