gpt2-p10k

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

• Loss: 0.0241

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: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.2	200	0.0558
No log	0.4	400	0.1944
0.2826	0.6	600	0.3970
0.2826	0.8	800	0.6245
0.8928	1.0	1000	2.0545
0.8928	1.2	1200	0.3789
0.8928	1.4	1400	0.4120
0.5735	1.6	1600	0.9738
0.5735	1.8	1800	1.4284
3.2584	2.0	2000	3.8628
3.2584	2.2	2200	0.6803
3.2584	2.4	2400	0.4168
1.1454	2.6	2600	0.0628
1.1454	2.8	2800	0.0353
0.0693	3.0	3000	0.0301
0.0693	3.2	3200	0.0294
0.0693	3.4	3400	0.0284
0.0299	3.6	3600	0.0279
0.0299	3.8	3800	0.0274
0.0287	4.0	4000	0.0274
0.0287	4.2	4200	0.0271
0.0287	4.4	4400	0.0260
0.0274	4.6	4600	0.0260
0.0274	4.8	4800	0.0261
0.0267	5.0	5000	0.0257
0.0267	5.2	5200	0.0255
0.0267	5.4	5400	0.0255
0.0263	5.6	5600	0.0254
0.0263	5.8	5800	0.0250
0.0259	6.0	6000	0.0250
0.0259	6.2	6200	0.0252
0.0259	6.4	6400	0.0253
0.0256	6.6	6600	0.0250
0.0256	6.8	6800	0.0247
0.0253	7.0	7000	0.0256
0.0253	7.2	7200	0.0247
0.0253	7.4	7400	0.0245
0.0251	7.6	7600	0.0245
0.0251	7.8	7800	0.0245
0.0251	8.0	8000	0.0246
0.0251	8.2	8200	0.0244
0.0251	8.4	8400	0.0246
0.0252	8.6	8600	0.0243
0.0252	8.8	8800	0.0242
0.0244	9.0	9000	0.0242
0.0244	9.2	9200	0.0242
0.0244	9.4	9400	0.0242
0.0247	9.6	9600	0.0242
0.0247	9.8	9800	0.0241
0.0245	10.0	10000	0.0241

Framework versions

• Transformers 4.40.2
• Pytorch 2.3.0
• Datasets 2.19.1
• Tokenizers 0.19.1