distilgpt2-finetuned-custom-mail

This model is a fine-tuned version of distilgpt2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 3.1905

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: 2e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss
No log	1.0	7	3.5915
No log	2.0	14	3.4986
No log	3.0	21	3.4418
No log	4.0	28	3.3970
No log	5.0	35	3.3569
No log	6.0	42	3.3207
No log	7.0	49	3.2972
No log	8.0	56	3.2806
No log	9.0	63	3.2620
No log	10.0	70	3.2451
No log	11.0	77	3.2302
No log	12.0	84	3.2177
No log	13.0	91	3.2083
No log	14.0	98	3.2024
No log	15.0	105	3.1984
No log	16.0	112	3.1962
No log	17.0	119	3.1938
No log	18.0	126	3.1920
No log	19.0	133	3.1913
No log	20.0	140	3.1905

Framework versions

• Transformers 4.23.1
• Pytorch 1.12.1+cu113
• Datasets 2.6.1
• Tokenizers 0.13.1