Update README.md
Browse files
README.md
CHANGED
|
@@ -95,8 +95,6 @@ I have added a linear quantization node before the `sentence_embedding` output s
|
|
| 95 |
|
| 96 |
This is compatible with the [qdrant](https://github.com/qdrant/qdrant) uint8 datatype for collections.
|
| 97 |
|
| 98 |
-
No benchmarks, but it in my limited testing it's exactly equivalent to the FP32 output of the uint8 quantized ONNX model.
|
| 99 |
-
|
| 100 |
# Quantization method
|
| 101 |
|
| 102 |
Linear quantization for the scale -.3 to 0.3, which is what sentence_embedding is normalized to.
|
|
|
|
| 95 |
|
| 96 |
This is compatible with the [qdrant](https://github.com/qdrant/qdrant) uint8 datatype for collections.
|
| 97 |
|
|
|
|
|
|
|
| 98 |
# Quantization method
|
| 99 |
|
| 100 |
Linear quantization for the scale -.3 to 0.3, which is what sentence_embedding is normalized to.
|