metadata
license: mit
language:
- en
This is the quantized (INT8) ONNX variant of the bge-large-en-v1.5 model for embeddings created with DeepSparse Optimum for ONNX export/inference pipeline and Neural Magic's Sparsify for One-Shot quantization.
Model achieves 100% accuracy recovery on the STSB validation dataset vs. dense ONNX variant.
Other sparse and quantized bge ONNX models:
zeroshot/bge-large-en-v1.5-sparse
zeroshot/bge-base-en-v1.5-sparse
zeroshot/bge-base-en-v1.5-quant