f57b822af04303de56ae9ce426c2f4ac

This model is a fine-tuned version of facebook/opt-125m on the nyu-mll/glue [sst2] dataset. It achieves the following results on the evaluation set:

• Loss: 0.4166
• Data Size: 1.0
• Epoch Runtime: 149.4917
• Accuracy: 0.8819
• F1 Macro: 0.8817
• Rouge1: 0.8819
• Rouge2: 0.0
• Rougel: 0.8819
• Rougelsum: 0.8831

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	0.8020	0	1.4856	0.5116	0.3714	0.5116	0.0	0.5116	0.5104
No log	1	2104	0.4183	0.0078	3.2572	0.8299	0.8298	0.8310	0.0	0.8299	0.8299
No log	2	4208	0.5029	0.0156	3.7173	0.7743	0.7635	0.7731	0.0	0.7743	0.7755
0.009	3	6312	0.4130	0.0312	6.1791	0.8356	0.8349	0.8356	0.0	0.8356	0.8356
0.2967	4	8416	0.3165	0.0625	10.7564	0.8530	0.8522	0.8530	0.0	0.8530	0.8530
0.2934	5	10520	0.2755	0.125	19.7290	0.8877	0.8875	0.8877	0.0	0.8877	0.8877
0.2299	6	12624	0.2739	0.25	37.5943	0.8819	0.8818	0.8819	0.0	0.8819	0.8819
0.1874	7	14728	0.2382	0.5	73.2081	0.9086	0.9085	0.9086	0.0	0.9086	0.9086
0.1542	8.0	16832	0.3261	1.0	144.6199	0.8738	0.8738	0.8738	0.0	0.875	0.8738
0.1257	9.0	18936	0.3737	1.0	144.2787	0.875	0.8747	0.875	0.0	0.875	0.8738
0.1144	10.0	21040	0.3436	1.0	146.7333	0.8854	0.8854	0.8854	0.0	0.8854	0.8843
0.0864	11.0	23144	0.4166	1.0	149.4917	0.8819	0.8817	0.8819	0.0	0.8819	0.8831

Framework versions

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