DialoGPT-large-faqs-block-size-256-bs-16-lr-1e-05

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

• Loss: 2.4550

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: 1e-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	20	5.0495
No log	2.0	40	4.2366
No log	3.0	60	3.6512
No log	4.0	80	3.3026
No log	5.0	100	3.0554
No log	6.0	120	2.8980
No log	7.0	140	2.7913
No log	8.0	160	2.7087
No log	9.0	180	2.6477
No log	10.0	200	2.5997
No log	11.0	220	2.5631
No log	12.0	240	2.5290
No log	13.0	260	2.5126
No log	14.0	280	2.4872
No log	15.0	300	2.4787
No log	16.0	320	2.4684
No log	17.0	340	2.4628
No log	18.0	360	2.4591
No log	19.0	380	2.4565
No log	20.0	400	2.4550

Framework versions

• Transformers 4.33.0.dev0
• Pytorch 2.0.1+cu118
• Datasets 2.14.4.dev0
• Tokenizers 0.13.3