pere commited on
Commit
141aff0
1 Parent(s): 0f09adb

Update export_models.sh

Browse files
Files changed (1) hide show
  1. export_models.sh +60 -0
export_models.sh ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ #!/bin/bash
2
+ pip install "optimum[exporters]>=1.14.1" tensorflow
3
+ git lfs track *.onnx*
4
+
5
+ python << END
6
+ from transformers import WhisperForConditionalGeneration, TFWhisperForConditionalGeneration, WhisperTokenizerFast
7
+ import shutil
8
+
9
+ # Backup generation_config.json
10
+ shutil.copyfile('./generation_config.json', './generation_config_backup.json')
11
+
12
+ print("Saving model to PyTorch...", end=" ")
13
+ model = WhisperForConditionalGeneration.from_pretrained("./", from_flax=True)
14
+ model.save_pretrained("./", safe_serialization=True)
15
+ model.save_pretrained("./", safe_serialization=False)
16
+ print("Done.")
17
+
18
+ print("Saving model to TensorFlow...", end=" ")
19
+ tf_model = TFWhisperForConditionalGeneration.from_pretrained("./")
20
+ tf_model.save_pretrained("./")
21
+ print("Done.")
22
+
23
+ # Restore the backup of generation_config.json
24
+ shutil.move('./generation_config_backup.json', './generation_config.json')
25
+
26
+ print("Saving model to ONNX...", end=" ")
27
+ from optimum.onnxruntime import ORTModelForSpeechSeq2Seq
28
+ ort_model = ORTModelForSpeechSeq2Seq.from_pretrained("./", export=True)
29
+ ort_model.save_pretrained("./onnx")
30
+ print("Done")
31
+
32
+ print("Saving model to CTranslate...", end=" ")
33
+ ct2-transformers-converter --model . --output_dir ct2
34
+ cp ct2/model.bin .
35
+ cp ct2/vocabulary.json .
36
+ cp config.json config_hf.json
37
+ jq -s '.[0] * .[1]' ct2/config.json config_hf.json > config.json
38
+ print("Done")
39
+
40
+
41
+ END
42
+
43
+ echo "Saving model to GGML (whisper.cpp)..."
44
+ wget -O convert-h5-to-ggml.py "https://raw.githubusercontent.com/NbAiLab/nb-whisper/main/convert-h5-to-ggml.py"
45
+ mkdir -p whisper/assets
46
+ wget -O whisper/assets/mel_filters.npz "https://github.com/openai/whisper/raw/c5d42560760a05584c1c79546a098287e5a771eb/whisper/assets/mel_filters.npz"
47
+ python ./convert-h5-to-ggml.py ./ ./ ./
48
+ rm ./convert-h5-to-ggml.py
49
+ rm -rf ./whisper
50
+ echo "Done"
51
+
52
+ echo "Quantizing GGML model..."
53
+ git clone --depth 1 https://github.com/ggerganov/whisper.cpp --branch v1.5.1
54
+ cd whisper.cpp/
55
+ make -j 32
56
+ make quantize -j 32
57
+ ./quantize ../ggml-model.bin ../ggml-model-q5_0.bin q5_0
58
+ cd ..
59
+ rm -rf whisper.cpp
60
+ echo "Done"