update readme
Browse files
README.md
CHANGED
|
@@ -40,6 +40,11 @@ The model conversion was made with [onnx-convert](https://github.com/nixiesearch
|
|
| 40 |
python convert.sh --model_id intfloat/e5-small-v2 --quantize QInt8 --optimize 2
|
| 41 |
```
|
| 42 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 43 |
## License
|
| 44 |
|
| 45 |
Apache 2.0
|
|
|
|
| 40 |
python convert.sh --model_id intfloat/e5-small-v2 --quantize QInt8 --optimize 2
|
| 41 |
```
|
| 42 |
|
| 43 |
+
There are two versions of model available:
|
| 44 |
+
|
| 45 |
+
* `model.onnx` - Float32 version, with optimize=2
|
| 46 |
+
* `model_opt2_QInt8.onnx` - QInt8 quantized version, with optimize=2
|
| 47 |
+
|
| 48 |
## License
|
| 49 |
|
| 50 |
Apache 2.0
|