Commit
·
e143fda
1
Parent(s):
7bc7521
Create README.md
Browse files
README.md
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
## Typesense Public Embedding Models
|
2 |
+
We store our current supported embedding models in this repo and you can also convert your own models to ONNX format and create a PR to add it to our supported models list.
|
3 |
+
|
4 |
+
### Convert a model to ONNX format
|
5 |
+
|
6 |
+
#### Converting a Hugging Face Transformers Model
|
7 |
+
You can follow instructions from [this link](https://huggingface.co/docs/transformers/serialization#export-to-onnx) to convert any model from Hugging Face to ONNX format using ```optimum-cli```.
|
8 |
+
#### Converting a PyTorch Model
|
9 |
+
You can use ```torch.onnx``` [APIs](https://pytorch.org/docs/stable/onnx.html) to convert PyTorch models to ONNX.
|
10 |
+
#### Converting a Tensorflow Model
|
11 |
+
You can use ```tf2onnx``` [tool](https://onnxruntime.ai/docs/tutorials/tf-get-started.html#getting-started-converting-tensorflow-to-onnx) to convert Tensorflow models to ONNX.
|
12 |
+
|
13 |
+
### Creating model config
|
14 |
+
Before creating a PR with your ONNX model, you should store model file, vocab file and model config file under a folder with model name. Your model config must be named as ```config.json``` and should contain those keys:
|
15 |
+
| Key | Description | Optional |
|
16 |
+
|-----|-------------|----------|
|
17 |
+
|model_md5| MD5 checksum of model file as string| No |
|
18 |
+
|vocab_md5| MD5 checksum of vocab file as string| No |
|
19 |
+
|model_type| Model type (currently only ```bert``` and ```xlm_roberta``` supported)| No |
|
20 |
+
|vocab_file_name| File name of vocab file| No |
|
21 |
+
|indexing_prefix| Prefix to be added before embedding documents| Yes |
|
22 |
+
|query_prefix| Prefix to be added before embedding queries | Yes |
|