ed874697366c57dc95940cc46b49b7ba

This model is a fine-tuned version of openai-community/gpt2 on the google/boolq dataset. It achieves the following results on the evaluation set:

• Loss: 1.0356
• Data Size: 1.0
• Epoch Runtime: 23.2010
• Accuracy: 0.7261
• F1 Macro: 0.6981
• Rouge1: 0.7270
• Rouge2: 0.0
• Rougel: 0.7255
• Rougelsum: 0.7258

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	8.9091	0	3.2784	0.3787	0.2747	0.3787	0.0	0.3793	0.3790
No log	1	294	3.4782	0.0078	3.7885	0.6008	0.4249	0.6008	0.0	0.6011	0.6005
No log	2	588	0.9471	0.0156	3.7762	0.5729	0.4995	0.5732	0.0	0.5729	0.5723
No log	3	882	0.7552	0.0312	4.2588	0.5432	0.5169	0.5435	0.0	0.5435	0.5426
0.0481	4	1176	0.6790	0.0625	5.0083	0.6032	0.5255	0.6029	0.0	0.6031	0.6036
0.0563	5	1470	0.6748	0.125	6.3832	0.6281	0.4350	0.6284	0.0	0.6278	0.6284
0.0975	6	1764	0.6239	0.25	9.3238	0.6612	0.5954	0.6618	0.0	0.6612	0.6615
0.6123	7	2058	0.6561	0.5	14.1194	0.6078	0.6077	0.6081	0.0	0.6085	0.6078
0.5517	8.0	2352	0.5673	1.0	24.4125	0.7111	0.6883	0.7117	0.0	0.7108	0.7111
0.4635	9.0	2646	0.6108	1.0	23.2478	0.7374	0.7118	0.7377	0.0	0.7374	0.7371
0.3124	10.0	2940	0.7318	1.0	23.6046	0.7328	0.7067	0.7328	0.0	0.7322	0.7328
0.2118	11.0	3234	0.9414	1.0	23.7510	0.7374	0.7124	0.7377	0.0	0.7374	0.7371
0.1472	12.0	3528	1.0356	1.0	23.2010	0.7261	0.6981	0.7270	0.0	0.7255	0.7258

Framework versions

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