CoolFace
Modelpublic

philschmid/MiniLMv2-L12-H384-emotion

sourceHugging Faceupdated 5y agoView on Hugging Face
0likes16downloads
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. -->

MiniLMv2-L12-H384-emotion

This model is a fine-tuned version of nreimers/MiniLMv2-L12-H384-distilled-from-RoBERTa-Large on the emotion dataset. It achieves the following results on the evaluation set:

  • —Loss: 0.2069
  • —Accuracy: 0.925

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: 3e-05
  • —trainbatchsize: 16
  • —evalbatchsize: 32
  • —seed: 42
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —lrschedulerwarmup_steps: 500
  • —num_epochs: 8
  • —mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation LossAccuracy
0.87451.010000.66730.81
0.34662.020000.28160.918
0.22013.030000.23670.9215
0.17614.040000.20690.925
0.14355.050000.20890.922
0.14546.060000.21680.923
0.10417.070000.20810.924
0.09538.080000.21330.9245

Framework versions

  • —Transformers 4.12.3
  • —Pytorch 1.9.1
  • —Datasets 1.15.1
  • —Tokenizers 0.10.3