94e7b9cf98bb10a5aad19f56290e99c5

This model is a fine-tuned version of facebook/opt-2.7b on the nyu-mll/glue [qnli] dataset. It achieves the following results on the evaluation set:

• Loss: 1.8727
• Data Size: 1.0
• Epoch Runtime: 735.2675
• Accuracy: 0.7061
• F1 Macro: 0.7015
• Rouge1: 0.7061
• Rouge2: 0.0
• Rougel: 0.7064
• Rougelsum: 0.7063

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	0.8133	0	10.9501	0.5085	0.4906	0.5083	0.0	0.5081	0.5090
No log	1	3273	0.7121	0.0078	16.7026	0.5059	0.3363	0.5055	0.0	0.5059	0.5057
0.0149	2	6546	0.7137	0.0156	27.4726	0.5066	0.3454	0.5062	0.0	0.5064	0.5064
0.8136	3	9819	0.7306	0.0312	40.9555	0.5057	0.3359	0.5053	0.0	0.5057	0.5056
0.6957	4	13092	0.7404	0.0625	64.0052	0.4943	0.3308	0.4947	0.0	0.4943	0.4944
0.6963	5	16365	0.6568	0.125	106.3617	0.6140	0.6095	0.6138	0.0	0.6142	0.6140
0.6729	6	19638	0.6738	0.25	197.0527	0.5866	0.5596	0.5864	0.0	0.5866	0.5864
0.636	7	22911	0.6730	0.5	375.2546	0.5939	0.5696	0.5943	0.0	0.5944	0.5938
0.543	8.0	26184	0.5849	1.0	730.0642	0.7224	0.7139	0.7226	0.0	0.7228	0.7222
0.4262	9.0	29457	0.5755	1.0	733.6856	0.7392	0.7374	0.7392	0.0	0.7397	0.7392
0.2872	10.0	32730	0.6829	1.0	735.3039	0.7215	0.7202	0.7215	0.0	0.7219	0.7213
0.1723	11.0	36003	0.9874	1.0	737.1306	0.7261	0.7253	0.7259	0.0	0.7265	0.7259
0.0753	12.0	39276	1.3682	1.0	733.7647	0.7114	0.7102	0.7112	0.0	0.7116	0.7115
0.074	13.0	42549	1.8727	1.0	735.2675	0.7061	0.7015	0.7061	0.0	0.7064	0.7063

Framework versions

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