istupakov commited on
Commit
bb52c0c
·
1 Parent(s): 25f52b2

Upload folder using huggingface_hub

Browse files
README.md CHANGED
@@ -23,17 +23,25 @@ print(model.recognize("test.wav"))
23
 
24
  ## Code for models export
25
 
26
- Export Whisper to ONNX with `onnxruntime` ([whisper.convert_to_onnx](https://github.com/microsoft/onnxruntime/blob/main/onnxruntime/python/tools/transformers/models/whisper/README.md)).
27
 
28
- Download model and export with Beam Search and Forced Decoder Input Ids:
29
  ```shell
30
- python3 -m onnxruntime.transformers.models.whisper.convert_to_onnx -m openai/whisper-base --output whisper-onnx --use_external_data_format --use_forced_decoder_ids --optimize_onnx --precision fp32
31
  ```
32
 
33
- Save preprocessor and tokenizer configs
34
  ```py
35
- from transformers import WhisperProcessor
36
 
37
- processor = WhisperProcessor.from_pretrained("openai/whisper-base")
38
  processor.save_pretrained("whisper-onnx")
39
- ```
 
 
 
 
 
 
 
 
 
23
 
24
  ## Code for models export
25
 
26
+ Read onnxruntime [instruction](https://github.com/microsoft/onnxruntime/blob/main/onnxruntime/python/tools/transformers/models/whisper/README.md) for convert Whisper to ONNX.
27
 
28
+ Download model and export with *Beam Search* and *Forced Decoder Input Ids*:
29
  ```shell
30
+ python3 -m onnxruntime.transformers.models.whisper.convert_to_onnx -m openai/whisper-base --output ./whisper-onnx --use_forced_decoder_ids --optimize_onnx --precision fp32
31
  ```
32
 
33
+ Save tokenizer config
34
  ```py
35
+ from transformers import WhisperTokenizer
36
 
37
+ processor = WhisperTokenizer.from_pretrained("openai/whisper-base")
38
  processor.save_pretrained("whisper-onnx")
39
+ ```
40
+
41
+ Create `config.json`
42
+ ```json
43
+ {
44
+ "model_type": "whisper-ort",
45
+ "features_size": 80
46
+ }
47
+ ```
config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "model_type": "whisper-ort",
3
+ "features_size": 80
4
+ }
preprocessor_config.json DELETED
@@ -1,15 +0,0 @@
1
- {
2
- "chunk_length": 30,
3
- "dither": 0.0,
4
- "feature_extractor_type": "WhisperFeatureExtractor",
5
- "feature_size": 80,
6
- "hop_length": 160,
7
- "n_fft": 400,
8
- "n_samples": 480000,
9
- "nb_max_frames": 3000,
10
- "padding_side": "right",
11
- "padding_value": 0.0,
12
- "processor_class": "WhisperProcessor",
13
- "return_attention_mask": false,
14
- "sampling_rate": 16000
15
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
whisper-base_beamsearch.int8.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:537719570b1b859d5d805283f509bce621d8ebbba8189609d8fb12064a68530c
3
+ size 107166530
whisper-base_beamsearch.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:732180688d163f91283769140db2a63a47da21ca1c1379a6c686b4bd62f2c412
3
+ size 398028367