This modelcard is a fine tuned version of Bert-base-uncased for text classification given candidates profile description on the resume.
Model Details
- Language(s) (NLP): English
- License: [More Information Needed]
- Finetuned from model [bert-base-uncased]:
Bias, Risks, and Limitations
Risks inherited from base model apply
How to Get Started with the Model
Training Details
Training Procedure
Preprocessing [optional]
[More Information Needed]
Training Hyperparameters
- Training regime: [More Information Needed]
Evaluation
Step Training Loss Validation Loss Accuracy F1 Precision Recall 50 3.191700 3.144314 0.045576 0.005990 0.043515 0.042929 100 2.940400 2.529969 0.313673 0.198400 0.326865 0.263533 150 2.289300 1.829516 0.638070 0.522103 0.562795 0.570456 200 1.582400 1.251463 0.784182 0.678805 0.699407 0.707561 250 1.120000 1.036136 0.773458 0.668376 0.700246 0.693412 300 0.886400 0.898808 0.812332 0.733046 0.757457 0.741408 350 0.799100 0.825964 0.824397 0.745086 0.766041 0.752816 400 0.848900 0.734574 0.836461 0.773499 0.812788 0.768352 450 0.699000 0.745959 0.833780 0.759005 0.780415 0.763101 500 0.510300 0.710035 0.844504 0.771996 0.794520 0.775719 550 0.514200 0.679396 0.840483 0.768999 0.785576 0.773328 600 0.516700 0.684477 0.843164 0.770532 0.785427 0.774498 650 0.483200 0.686520 0.840483 0.768139 0.780840 0.772857
Metrics
eval_loss eval_Accuracy eval_F1 eval_Precision eval_Recall
train 0.479060 0.897583 0.839754 0.911787 0.840669 val 0.710035 0.844504 0.771996 0.794520 0.775719
Hardware
colab T4