📁📄 Welcome to My Resume Classifier Repository
This work presents a resume-classifier ML model which can categorize resumes from a input directory containing pdf files of resumes. I used "Blurr" to train the model and the script.py file can be run directly from the commandline terminal to get the output.
Data was collected from the link: https://www.kaggle.com/datasets/snehaanbhawal/resume-dataset
In a tottal, there are 3447 data in the file named as Resume.csv.
I used Fastai and Blurr with "distilroberta-base" model as it it easier to use and time effiecient.
Finetuned a distilrobera-base model from HuggingFace Transformers using Fastai and Blurr. The model training notebook can be viewed from the "notebook" folder of this repository. The model achieved an accuracy of approximately 96%.
As the size of the dataset was small in this project, it was splitted into 2 sets: train datset and validation dataset. I used the validation dataset for the test dataset in this model.
The F1 scores of this model are:
F1 Score (Micro) = 0.5473145780051151
F1 Score (Macro) = 0.37185516078339154
ROC curve and AUC score:
The trained model has a memory of 300+MB. I compressed this model using ONNX quantization and brought it under 80MB.
For model deplyoment, please use the script.py file. This Python script is executable from the command line as follows:
python script.py "path/to/dir".The script takes a directory containing the resumes to be categorized as input. Using the trained model, the script categorizes each resume.
• For each resume, it moves the resume to the respective category folder.
• The script also creates and writes a CSV file named as categorized_resumes.csv containing two columns: filename and category.