CoolFace
Modelpublic

domenicrosati/deberta-v3-large-survey-fluency-rater-gpt4

sourceHugging Facemitupdated 3y agoView on Hugging Face
0likes6downloads
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. -->

deberta-v3-large-survey-fluency-rater-gpt4

This model is a fine-tuned version of microsoft/deberta-v3-large on the None dataset. It achieves the following results on the evaluation set:

  • —Loss: 0.9754
  • —Krippendorff: -0.0424
  • —Spearman: 0.1432
  • —Absolute Agreement: 0.6897
  • —Agreement Within One: 0.7676

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: 6e-06
  • —trainbatchsize: 16
  • —evalbatchsize: 16
  • —seed: 42
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —lrschedulerwarmup_steps: 1000
  • —num_epochs: 20
  • —mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation LossKrippendorffSpearmanAbsolute AgreementAgreement Within One
No log1.0501.9660-0.7952nan0.02781.0
No log2.01001.9587-0.7952nan0.02781.0
No log3.01501.9548-0.7155-0.32360.02780.9306
No log4.02002.0047-0.7155-0.32360.02780.9306
No log5.02502.2550-0.3248-0.39960.1250.7361
No log6.03002.4123-0.1335-0.10460.250.7361
No log7.03502.4828-0.1883-0.23240.20830.7361
No log8.04002.4175-0.3139-0.40860.1250.7361
No log9.04502.6742-0.1335-0.10460.250.7361
1.291710.05002.8155-0.2880nan0.29170.7222
1.291711.05503.0146-0.2880nan0.29170.7222
1.291712.06003.0548-0.1729-0.19930.22220.7361
1.291713.06503.3351-0.2880nan0.29170.7222
1.291714.07003.0936-0.1883-0.23240.20830.7361
1.291715.07503.3908-0.1335-0.10460.250.7361
1.291716.08002.3496-0.375-0.39160.08330.7778
1.291717.08503.6256-0.1729-0.19930.22220.7361
1.291718.09003.3170-0.2037-0.31210.15280.7361
1.291719.09503.9354-0.2369-0.32850.15280.7361
0.720620.010003.6945-0.1997-0.26130.19440.7361

Framework versions

  • —Transformers 4.26.0
  • —Pytorch 1.13.1
  • —Datasets 2.10.1
  • —Tokenizers 0.12.1