Add ONNX weights for browser inference

3ddb732 verified 3 months ago

549 Bytes

---
library_name: transformers.js
pipeline_tag: text-classification
---
# CV match model for browser inference

This is the ONNX export of the Sri Lankan job-resume Cross-Encoder. It is intended to run locally in a browser using Transformers.js, so the free static Hugging Face Space does not need a paid CPU runtime.

The model accepts a job description as the first text and a candidate profile as the second text. Apply a sigmoid to the single output logit and compare the resulting probability with the decision threshold `0.4399277865886688`.

	---
	library_name: transformers.js
	pipeline_tag: text-classification
	---

	# CV match model for browser inference

	This is the ONNX export of the Sri Lankan job-resume Cross-Encoder. It is intended to run locally in a browser using Transformers.js, so the free static Hugging Face Space does not need a paid CPU runtime.

	The model accepts a job description as the first text and a candidate profile as the second text. Apply a sigmoid to the single output logit and compare the resulting probability with the decision threshold `0.4399277865886688`.