yihuizhou's picture
Update README.md
90d6de8 verified
metadata
license: apache-2.0
base_model:
  - intfloat/multilingual-e5-small

quantization and input maxlength

coreML: using linear quantize nbit=8

input max = 128

mlpackages for swift

note

i tried turn it into float`6, but it changed too much for its prediction. quantization using linear nbit=8, it works almost like the original.