91cbcc5f6ad7c31e90bf7c977c225302

This model is a fine-tuned version of facebook/opt-125m on the google/boolq dataset. It achieves the following results on the evaluation set:

• Loss: 1.3196
• Data Size: 1.0
• Epoch Runtime: 23.4913
• Accuracy: 0.6734
• F1 Macro: 0.6517
• Rouge1: 0.6740
• Rouge2: 0.0
• Rougel: 0.6731
• Rougelsum: 0.6728

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	0.7327	0	3.2602	0.6081	0.4083	0.6078	0.0	0.6078	0.6078
No log	1	294	0.7389	0.0078	3.6040	0.4473	0.4417	0.4473	0.0	0.4476	0.4473
No log	2	588	0.6716	0.0156	3.5576	0.6137	0.4317	0.6137	0.0	0.6131	0.6140
No log	3	882	0.6609	0.0312	4.1385	0.6198	0.3879	0.6198	0.0	0.6192	0.6195
0.0292	4	1176	0.6609	0.0625	4.7772	0.6204	0.3859	0.6204	0.0	0.6201	0.6203
0.0546	5	1470	0.6584	0.125	6.0451	0.6219	0.3850	0.6219	0.0	0.6216	0.6216
0.093	6	1764	0.6289	0.25	8.8599	0.6547	0.5885	0.6547	0.0	0.6547	0.6555
0.5788	7	2058	0.6185	0.5	13.7694	0.6596	0.6478	0.6593	0.0	0.6599	0.6599
0.4664	8.0	2352	0.6135	1.0	24.4934	0.6774	0.6672	0.6774	0.0	0.6771	0.6774
0.2417	9.0	2646	0.9035	1.0	23.2813	0.6927	0.6741	0.6930	0.0	0.6924	0.6927
0.1696	10.0	2940	1.5665	1.0	23.5260	0.6673	0.6585	0.6673	0.0	0.6673	0.6667
0.1116	11.0	3234	1.7915	1.0	23.3186	0.6900	0.6749	0.6900	0.0	0.6896	0.6900
0.0833	12.0	3528	1.3196	1.0	23.4913	0.6734	0.6517	0.6740	0.0	0.6731	0.6728

Framework versions

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