CoolFace
Modelpublic

NasimB/switchboard-log-rarity-seed

sourceHugging Facemitupdated 3y agoView on Hugging Face
0likes14downloads
Model Card

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

switchboard-log-rarity-seed

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

  • Loss: 4.1008

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: 0.0005
  • trainbatchsize: 64
  • evalbatchsize: 64
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lrschedulertype: cosine
  • lrschedulerwarmup_steps: 1000
  • num_epochs: 6
  • mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation Loss
6.35820.295005.3483
5.03370.5810004.9322
4.70730.8715004.6918
4.44391.1720004.5506
4.2941.4625004.4317
4.1871.7530004.3272
4.08152.0435004.2480
3.88912.3340004.2093
3.85682.6245004.1546
3.83192.9250004.0999
3.63923.2155004.0964
3.59193.560004.0644
3.56143.7965004.0333
3.47524.0870004.0305
3.31144.3775004.0258
3.30714.6680004.0137
3.29114.9685003.9998
3.15785.2590004.0124
3.13065.5495004.0113
3.12285.83100004.0107

Framework versions

  • Transformers 4.26.1
  • Pytorch 1.11.0+cu113
  • Datasets 2.13.0
  • Tokenizers 0.13.3