title: lisa + gradio + fastapi + CUDA emoji: ⚡ colorFrom: red colorTo: purple sdk: gradio sdk_version: 5.9.1 app_file: app.py pinned: true
LISA (Reasoning Segmentation via Large Language Model) on cuda, now with huggingface ZeroGPU support!
Exec jupyter on the remote server with port forwarding on localhost
- checkout repo, install venv with jupyter
- port forwarding in localhost wiht private key:
ssh -i /path/to/private_key name@endpoint.com -L 8889:localhost:8889 -N -f - start the jupyter-lab server
- connect to page in localhost
Commands to work on remote virtual machines (e.g. SaturnCloud) after clone and git lfs install
cd ~/workspace/lisa-on-cuda/
rm -rf lisa_venv
python3 -m venv lisa_venv
ln -s lisa_venv/ venv
source venv/bin/activate
pip --version
which python
python -m pip install pip wheel --upgrade
python -m pip install pytest pytest-cov jupyterlab
python -m pip install -r requirements.txt
nohup jupyter-lab &
tail -F nohup.out
Jupyterlab Howto
To run the test.ipynb notebook you should already:
- cloned project https://huggingface.co/spaces/aletrn/lisa-on-cuda with active git lfs
- created and activated a virtualenv
- installed jupyterlab dependencies from requirements_jupyter.txt
- installed dependencies from requirements.txt
Hardware requirements for local usage
- an nvidia gpu with 10 or 12GB of memory (a T4 should suffice)
- at least 16GB of system ram
Hardware requirements on huggingface ZeroGPU
Right now (July 2024) huggingface let use ZeroGPU Nvidia A100 GPUs.
See LISA for details on the original project. Note that the authors don't keep the project updated anymore.
Dependencies and HuggingFace demos with Gradio SDK
HuggingFace demos based on Gradio SDK (you need that to use ZeroGPU hardware) needs updated requirements.txt. You can keep your requirements.txt in sync with the dependencies installed in the project using this python command (from samgis-core):
python -m samgis_core.utilities.update_requirements_txt --req_no_version_path requirements_no_versions.txt --req_output_path requirements.txt
About the parameters:
- input argument
--req_no_version_pathis a file with the dependencies package list without version declared - output argument
--req_output_pathis the output requirements.txt
This command simply freeze the installed packages and filter it using the dependencies package list from the input argument.
If you need to modify the requirements_no_versions.txt file, avoid inserting
python(this is required only bypoetry)gradio,gradio-client(installed directly by HuggingFace, selecting the version in header section of the README.md file)spaces(installed directly by HuggingFace, version selected connected by the gradio version)
In case of doubt check the HuggingFace container log for the correct spaces package version.
1 maintainer
Release files for lisa-on-cuda 1.4.14
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| lisa_on_cuda-1.4.14.tar.gz | 95.3 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| lisa_on_cuda-1.4.14-py3-none-any.whl | Python 3 | none | any | Details |
Total release size:214.7 kB
Release files / lisa_on_cuda-1.4.14.tar.gz
| Download URL | lisa_on_cuda-1.4.14.tar.gz |
|---|---|
| Size | 95.3 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
f3c81203a64f99e112910524b6d31e79c9ed7c464f992d0ac05fdfbf96a7bc2e
|
|
BLAKE2b-256 checksum How to use checksums |
dda998fb14e0306b0afd388f620b23ed123697117e1c23f0aceb8895de1bbf3e
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
poetry/2.1.1 CPython/3.12.9 Linux/5.15.167.4-microsoft-standard-WSL2
|
Release files / lisa_on_cuda-1.4.14-py3-none-any.whl
| Download URL | lisa_on_cuda-1.4.14-py3-none-any.whl |
|---|---|
| Size | 119.5 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
ca1aa90662139b79ac7f2150cd95eee6866435cd72884b21c48c0ba117e58e08
|
|
BLAKE2b-256 checksum How to use checksums |
d53d815d96b13d3b4f0cfd2b111ff66e90d14b8f5c565bc6348fbade994cc3e9
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
poetry/2.1.1 CPython/3.12.9 Linux/5.15.167.4-microsoft-standard-WSL2
|
