You can deploy this model and start earning money today!

gurcaneker/
bge-reranker-v2-m3-onnx

Name: bge-reranker-v2-m3-onnx
Availability: InStock

BAAI/bge-reranker-v2-m3

quantized

apache-2.0

Views

No views yet

bge-reranker-v2-m3-onnx

This repository is a one-time ONNX conversion of BAAI/bge-reranker-v2-m3, produced with Hugging Face's official optimum export tool from the original safetensors weights.

The original model is licensed under Apache License 2.0 (see the source model card). This derivative (format-converted, ONNX) distribution is provided under the same Apache-2.0 license, as permitted by the original license terms. No weights were modified beyond the ONNX export/graph optimization process.

Produced for use with fastembed's TextCrossEncoder.add_custom_model(), since the upstream BAAI repository does not ship an onnx/model.onnx artifact required by fastembed.

Source

Base model: BAAI/bge-reranker-v2-m3
Conversion tool: Hugging Face optimum
Conversion script: scripts/convert_reranker_onnx.py in professional-rag (private project repo)

Files

model.onnx — ONNX graph
model.onnx_data — external ONNX weights (>2GB, referenced by model.onnx)
config.json, tokenizer.json, tokenizer_config.json, sentencepiece.bpe.model, special_tokens_map.json — tokenizer/model config

gurcaneker/bge-reranker-v2-m3-onnx

bge-reranker-v2-m3-onnx

Source

Files

gurcaneker/
bge-reranker-v2-m3-onnx