Hugging Face's logo

hotchpotch
/
vespa-onnx-intfloat-multilingual-e5-large

Feature Extraction
Transformers
ONNX
xlm-roberta
text-embeddings-inference
Model card Files
xet
Community
1
hotchpotch commited on
Commit
6e9ecc4
·
verified ·
1 Parent(s): 4c2dc50

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -4,7 +4,7 @@ license: mit
4
 
5
  Converted [intfloat/multilingual-e5-large](https://huggingface.co/intfloat/multilingual-e5-large) model in onnx int8 format for use with [Vespa Embedding](https://docs.vespa.ai/en/embedding.html).
6
 
7
- If your CPU supports the avx-512 VNNI instruction set, such as vespa cloud, `multilingual-e5-large-int8-avx512_vnni.onnx` will perform best.
8
 
9
  The model was quantized using the [optimum](https://github.com/huggingface/optimum) toolkit.
10
 
 
4
 
5
  Converted [intfloat/multilingual-e5-large](https://huggingface.co/intfloat/multilingual-e5-large) model in onnx int8 format for use with [Vespa Embedding](https://docs.vespa.ai/en/embedding.html).
6
 
7
+ If your CPU supports the avx-512 VNNI instruction set, `multilingual-e5-large-int8-avx512_vnni.onnx` will perform best.
8
 
9
  The model was quantized using the [optimum](https://github.com/huggingface/optimum) toolkit.
10