DialoGPT-large-faqs-block-size-128-bs-16-lr-2e-5

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: 2.7873

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	40	3.3953
No log	2.0	80	2.7368
No log	3.0	120	2.4963
No log	4.0	160	2.4083
No log	5.0	200	2.3677
No log	6.0	240	2.3529
No log	7.0	280	2.3669
No log	8.0	320	2.4104
No log	9.0	360	2.4576
No log	10.0	400	2.5224
No log	11.0	440	2.5940
No log	12.0	480	2.6281
1.7771	13.0	520	2.6656
1.7771	14.0	560	2.6991
1.7771	15.0	600	2.7157
1.7771	16.0	640	2.7565
1.7771	17.0	680	2.7790
1.7771	18.0	720	2.7847
1.7771	19.0	760	2.7866
1.7771	20.0	800	2.7873

Framework versions

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