gpt2_for_whole_train_result_1_2bce

This model is a fine-tuned version of gpt2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0290
• Accuracy: 0.995
• F1: 0.9953

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: 0.0001
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 64
• total_train_batch_size: 4096
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1000
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
1.7296	6.8817	50	0.7674	0.6645	0.7317
0.2968	13.7634	100	0.1240	0.9565	0.9605
0.0564	20.6452	150	0.0491	0.986	0.9869
0.0206	27.5269	200	0.0351	0.993	0.9934
0.0081	34.4086	250	0.0321	0.994	0.9944
0.0041	41.2903	300	0.0370	0.993	0.9935
0.002	48.1720	350	0.0334	0.995	0.9953
0.0017	55.0538	400	0.0339	0.9955	0.9958
0.001	61.9355	450	0.0326	0.995	0.9953
0.0005	68.8172	500	0.0371	0.996	0.9963
0.0005	75.6989	550	0.0333	0.996	0.9962
0.0007	82.5806	600	0.0344	0.9945	0.9948
0.0004	89.4624	650	0.0512	0.9925	0.9930
0.0024	96.3441	700	0.0290	0.995	0.9953

Framework versions

• Transformers 4.40.0
• Pytorch 2.4.1+cu121
• Datasets 3.1.0
• Tokenizers 0.19.1