haixuantao
0026331df6
Minor release fix
1 year ago
haixuantao
35e72cbc15
Add latest news section and improve readme initial section
1 year ago
haixuantao
27c1047c96
Bump dora version
1 year ago
Haixuan Xavier Tao
074908cca6
Add qwenvl2 5 ( #767 )
1 year ago
haixuantao
4d83d10756
Remove past_key_value blocked by transformers cache error https://github.com/huggingface/transformers/issues/34678
1 year ago
haixuantao
071de278ef
Add caching of historic key values
1 year ago
haixuantao
57fd132998
fix CI ignore node-hub
1 year ago
haixuantao
f659b1dad5
fix pytest
1 year ago
haixuantao
0f2361ba2f
Minor refactoring
1 year ago
haixuanTao
f1e7c0eb8a
Removing flash_attn as it is too hard to auto-install
1 year ago
haixuanTao
1f6d109caa
Fixing linux installation of flash-attn
1 year ago
haixuantao
f801931caf
Ignore node-hub folder
1 year ago
haixuantao
76cbed9bef
Set python to version 3.10 within venv
1 year ago
haixuantao
ebcda89675
fix flash attn install
1 year ago
haixuantao
a247999ef4
Add readme within llama factory
1 year ago
haixuantao
627b3f7d8d
Using default pypi within llama-cpp-python
1 year ago
Haixuan Xavier Tao
de4ff66b89
Bump upload artifact version ( #769 )
This is a small PR to fix current failing CI/CD issues.
1 year ago
haixuantao
2e0c05f251
Add flash_attn dependency to qwenvl 2.5
1 year ago
haixuantao
5fdb6fcd09
Add mlx_lm dependencies
1 year ago
haixuantao
4de0d46ea6
Add index pypi for llama-cpp-python windows and linux platform
1 year ago
haixuantao
85d06a6e11
Bump upload artifact version
1 year ago
haixuantao
aa32c2483f
Adding speech to speech example within vlm
1 year ago
haixuantao
afa8dacfee
Fix qwen2.5
1 year ago
haixuantao
938449d86b
Adding activation word configuration
1 year ago
haixuantao
fa15fdd4d1
Enable multi torch version
1 year ago
haixuantao
f67fa63772
Fix qwen errors
1 year ago
haixuantao
8358e3a0f9
fix outetts pyproject issue with metal flag for llama-cpp-python
1 year ago
haixuantao
9bf2da2a35
Merge remote-tracking branch 'origin/add-qwen' into add-qwenvl2-5
1 year ago
Haixuan Xavier Tao
88892863f9
Reduce silence duration ( #768 )
This is a very minor PR to reduce default silence in VAD.
1 year ago
haixuantao
7da62f3a88
Adding text to speech into qwenvl demo
1 year ago
haixuantao
beb97c5349
Add additional bad sentences
1 year ago
haixuantao
1981cbba4c
Fix torch for qwenvl2 for avoiding errors
1 year ago
haixuantao
21e8f98b12
Minor pyaudio stream fix
1 year ago
haixuantao
65135134c7
Add dora-qwen2-5-vl
1 year ago
Haixuan Xavier Tao
a9391b8af8
Use mlx whisper instead of lightning-whisper ( #766 )
This is a small refactoring of whisper mlx for macos.
1 year ago
Haixuan Xavier Tao
b58c534bfb
Add a uv flag to make it possible to automatically replace `pip` with `uv pip` and prepend run command with `uv run` ( #765 )
This uv flag makes it possible to run dora within a `uv venv` envirement
instead of a venv.
This has the advantage, to:
- not depend on a shell to run command within the virtual env.
- be faster than normal venv
- be easier to maintain a clear separation of environment.
## Example
https://github.com/user-attachments/assets/fbb58afb-76df-4447-85db-87806367846b
## Usage
```bash
uv venv
dora build dataflow.yml --uv
dora run dataflow.yml --uv
```
1 year ago
haixuantao
ac0e6a31c6
Adding qwen llm model
1 year ago
haixuantao
5be3cbbb9c
Merge remote-tracking branch 'origin/use-mlx-whisper' into add-uv-flag
1 year ago
haixuantao
28af216c8f
Add uv flag to enable build and run through uv run and uv pip
1 year ago
haixuantao
053f3a1876
Reduce silence duration
1 year ago
haixuantao
3883874a31
Use mlx whisper
1 year ago
Haixuan Xavier Tao
196bc7f3f8
chore: use workspace edition ( #764 )
1 year ago
yjhmelody
ca614753a1
chore: use workspace edition
1 year ago
Haixuan Xavier Tao
69e5cd7558
Print python stdout without buffer even for script ( #761 )
Adding ** PYTHONUNBUFFERED=1** makes python script print out to stdout
without buffering.
Previously this was fixed with a python flag but that was not called
when python was used as a script.
This resolve the previous issue.
1 year ago
Haixuan Xavier Tao
18d26d137c
Add ios lidar ( #762 )
This PR add the possibility to work with IOS lidar as well as texture
rgb-d image within rerun
<img width="1800" alt="Screenshot 2025-01-22 at 09 20 50"
src="https://github.com/user-attachments/assets/97ec6efe-db79-437b-9b4b-6f254da38333 "
/>
1 year ago
haixuantao
91e291d4aa
Fix Typo
1 year ago
haixuantao
6cc835d290
Group depth camera into a single folder
1 year ago
haixuantao
cbe73522fe
Adding opencv dependency
1 year ago
haixuantao
97de48fc70
Remove / within depth naming
1 year ago
haixuantao
745a12943e
Add note on device app installation
1 year ago