whisper.cpp/models
st-gr eb23f4ef16
openvino : fix convert-whisper-to-openvino.py (#1890)
Fix issue: Conversion from Whisper to OpenVino failed #1870

convert-whisper-to-openvino.py stopped working with OpenVINO version 2023.0.0-10926-b4452d56304-releases/2023/0 .

Error was: TypeError: load(): incompatible function arguments. The following argument types are supported:
    1. (self: openvino._pyopenvino.FrontEnd, path: object) -> ov::frontend::InputModel

Tested successfully with a large-v3 conversion.

Co-authored-by: Stefan Grundmann <grundmanns@sandiego.gov>
2024-02-22 15:11:35 +02:00
..
.gitignore Initial release 2022-09-25 22:13:49 +03:00
convert-h5-to-coreml.py whisper : make large version explicit + fix data size units (#1493) 2023-11-15 19:42:25 +02:00
convert-h5-to-ggml.py minor : improve C++ and Python style (#768) 2023-04-29 10:06:25 +03:00
convert-pt-to-ggml.py whisper : add support for large v3 (#1444) 2023-11-07 15:30:18 +02:00
convert-whisper-to-coreml.py coreml : fix ANE optimized encoder (#1716) 2024-01-04 16:28:30 +02:00
convert-whisper-to-openvino.py openvino : fix convert-whisper-to-openvino.py (#1890) 2024-02-22 15:11:35 +02:00
download-coreml-model.sh models : make all scripts to be POSIX Compliant (#1725) 2024-01-12 14:11:04 +02:00
download-ggml-model.cmd whisper : make large version explicit + fix data size units (#1493) 2023-11-15 19:42:25 +02:00
download-ggml-model.sh docs : make model options / model install methods clearer (#1806) 2024-01-26 17:39:54 +02:00
for-tests-ggml-base.bin Adding dummy models for testing purposes 2022-10-08 11:43:42 +03:00
for-tests-ggml-base.en.bin Adding dummy models for testing purposes 2022-10-08 11:43:42 +03:00
for-tests-ggml-large.bin Adding dummy models for testing purposes 2022-10-08 11:43:42 +03:00
for-tests-ggml-medium.bin Adding dummy models for testing purposes 2022-10-08 11:43:42 +03:00
for-tests-ggml-medium.en.bin Adding dummy models for testing purposes 2022-10-08 11:43:42 +03:00
for-tests-ggml-small.bin Adding dummy models for testing purposes 2022-10-08 11:43:42 +03:00
for-tests-ggml-small.en.bin Adding dummy models for testing purposes 2022-10-08 11:43:42 +03:00
for-tests-ggml-tiny.bin Adding dummy models for testing purposes 2022-10-08 11:43:42 +03:00
for-tests-ggml-tiny.en.bin Adding dummy models for testing purposes 2022-10-08 11:43:42 +03:00
generate-coreml-interface.sh models : make all scripts to be POSIX Compliant (#1725) 2024-01-12 14:11:04 +02:00
generate-coreml-model.sh models : make all scripts to be POSIX Compliant (#1725) 2024-01-12 14:11:04 +02:00
ggml_to_pt.py models : add ggml_to_pt script (#1042) 2023-06-25 15:29:54 +03:00
README.md docs : make model options / model install methods clearer (#1806) 2024-01-26 17:39:54 +02:00
requirements-coreml.txt models : add update py requirements 2024-02-13 11:51:32 +02:00
requirements-openvino.txt models : add update py requirements 2024-02-13 11:51:32 +02:00

Whisper model files in custom ggml format

The original Whisper PyTorch models provided by OpenAI are converted to custom ggml format in order to be able to load them in C/C++. Conversion is performed using the convert-pt-to-ggml.py script.

There are three ways to obtain ggml models:

1. Use download-ggml-model.sh to download pre-converted models

Example download:

$ ./download-ggml-model.sh base.en
Downloading ggml model base.en ...
models/ggml-base.en.bin          100%[=============================================>] 141.11M  5.41MB/s    in 22s
Done! Model 'base.en' saved in 'models/ggml-base.en.bin'
You can now use it like this:

  $ ./main -m models/ggml-base.en.bin -f samples/jfk.wav

2. Manually download pre-converted models

ggml models are available from the following locations:

3. Convert with convert-pt-to-ggml.py

Download one of the models provided by OpenAI and generate the ggml files using the convert-pt-to-ggml.py script.

Example conversion, assuming the original PyTorch files have been downloaded into ~/.cache/whisper. Change ~/path/to/repo/whisper/ to the location for your copy of the Whisper source:

mkdir models/whisper-medium
python models/convert-pt-to-ggml.py ~/.cache/whisper/medium.pt ~/path/to/repo/whisper/ ./models/whisper-medium
mv ./models/whisper-medium/ggml-model.bin models/ggml-medium.bin
rmdir models/whisper-medium

Available models

Model Disk SHA
tiny 75 MiB bd577a113a864445d4c299885e0cb97d4ba92b5f
tiny.en 75 MiB c78c86eb1a8faa21b369bcd33207cc90d64ae9df
base 142 MiB 465707469ff3a37a2b9b8d8f89f2f99de7299dac
base.en 142 MiB 137c40403d78fd54d454da0f9bd998f78703390c
small 466 MiB 55356645c2b361a969dfd0ef2c5a50d530afd8d5
small.en 466 MiB db8a495a91d927739e50b3fc1cc4c6b8f6c2d022
small.en-tdrz 465 MiB b6c6e7e89af1a35c08e6de56b66ca6a02a2fdfa1
medium 1.5 GiB fd9727b6e1217c2f614f9b698455c4ffd82463b4
medium.en 1.5 GiB 8c30f0e44ce9560643ebd10bbe50cd20eafd3723
large-v1 2.9 GiB b1caaf735c4cc1429223d5a74f0f4d0b9b59a299
large-v2 2.9 GiB 0f4c8e34f21cf1a914c59d8b3ce882345ad349d6
large-v2-q5_0 1.1 GiB 00e39f2196344e901b3a2bd5814807a769bd1630
large-v3 2.9 GiB ad82bf6a9043ceed055076d0fd39f5f186ff8062
large-v3-q5_0 1.1 GiB e6e2ed78495d403bef4b7cff42ef4aaadcfea8de

Models are multilingual unless the model name includes .en. Models ending in -q5_0 are quantized. Models ending in -tdrz support local diarization (marking of speaker turns) using tinydiarize. More information about models is available upstream (openai/whisper). The list above is a subset of the models supported by the download-ggml-model.sh script, but many more are available at https://huggingface.co/ggerganov/whisper.cpp/tree/main and elsewhere.

Model files for testing purposes

The model files prefixed with for-tests- are empty (i.e. do not contain any weights) and are used by the CI for testing purposes. They are directly included in this repository for convenience and the Github Actions CI uses them to run various sanitizer tests.

Fine-tuned models

There are community efforts for creating fine-tuned Whisper models using extra training data. For example, this blog post describes a method for fine-tuning using Hugging Face (HF) Transformer implementation of Whisper. The produced models are in slightly different format compared to the original OpenAI format. To read the HF models you can use the convert-h5-to-ggml.py script like this:

git clone https://github.com/openai/whisper
git clone https://github.com/ggerganov/whisper.cpp

# clone HF fine-tuned model (this is just an example)
git clone https://huggingface.co/openai/whisper-medium

# convert the model to ggml
python3 ./whisper.cpp/models/convert-h5-to-ggml.py ./whisper-medium/ ./whisper .

Distilled models

Initial support for https://huggingface.co/distil-whisper is available.

Currently, the chunk-based transcription strategy is not implemented, so there can be sub-optimal quality when using the distilled models with whisper.cpp.

# clone OpenAI whisper and whisper.cpp
git clone https://github.com/openai/whisper
git clone https://github.com/ggerganov/whisper.cpp

# get the models
cd whisper.cpp/models
git clone https://huggingface.co/distil-whisper/distil-medium.en
git clone https://huggingface.co/distil-whisper/distil-large-v2

# convert to ggml
python3 ./convert-h5-to-ggml.py ./distil-medium.en/ ../../whisper .
mv ggml-model.bin ggml-medium.en-distil.bin

python3 ./convert-h5-to-ggml.py ./distil-large-v2/ ../../whisper .
mv ggml-model.bin ggml-large-v2-distil.bin