Is your feature request related to a problem? Please describe.
When restarting Ollama Web UI, any Open AI API key added in settings is cleared. This means each time the an API key needs to be re-added to get access to ChatGPT models.
Describe the solution you'd like
The option for persistent storage of the API key.
Describe alternatives you've considered
I haven't found an alternative, I've just re-entered the key each time the Docker container is restarted.
Additional context
I'm running Ollama Web UI in a Docker container on Windows (WSL2).
Originally created by @Grepta on GitHub (Jan 23, 2024).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/546
**Is your feature request related to a problem? Please describe.**
When restarting Ollama Web UI, any Open AI API key added in settings is cleared. This means each time the an API key needs to be re-added to get access to ChatGPT models.
**Describe the solution you'd like**
The option for persistent storage of the API key.
**Describe alternatives you've considered**
I haven't found an alternative, I've just re-entered the key each time the Docker container is restarted.
**Additional context**
I'm running Ollama Web UI in a Docker container on Windows (WSL2).
You can set a container environment variable by adding -e OPENAI_API_KEY=your_api_key to your docker run command, or adding another line to the environment section of the ollama-webui service in your Docker Compose file, depending how you launched it.
<!-- gh-comment-id:1905168779 -->
@justinh-rahb commented on GitHub (Jan 23, 2024):
You can set a container environment variable by adding `-e OPENAI_API_KEY=your_api_key` to your `docker run` command, or adding another line to the environment section of the ollama-webui service in your Docker Compose file, depending how you launched it.
It may be worth considering moving the OpenAI implementation / API key to the user level as this bears a cost. That way the server "owner" wouldn't have to worry about users and their usage on the paid API.
<!-- gh-comment-id:1909187465 -->
@mtompkins commented on GitHub (Jan 25, 2024):
It may be worth considering moving the OpenAI implementation / API key to the user level as this bears a cost. That way the server "owner" wouldn't have to worry about users and their usage on the paid API.
It may be worth considering moving the OpenAI implementation / API key to the user level as this bears a cost. That way the server "owner" wouldn't have to worry about users and their usage on the paid API.
@mtompkins This was the previous implementation, it was moved from client-side to backend in #381
<!-- gh-comment-id:1909188692 -->
@justinh-rahb commented on GitHub (Jan 25, 2024):
> It may be worth considering moving the OpenAI implementation / API key to the user level as this bears a cost. That way the server "owner" wouldn't have to worry about users and their usage on the paid API.
@mtompkins This was the previous implementation, it was moved from client-side to backend in #381
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Originally created by @Grepta on GitHub (Jan 23, 2024).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/546
Is your feature request related to a problem? Please describe.
When restarting Ollama Web UI, any Open AI API key added in settings is cleared. This means each time the an API key needs to be re-added to get access to ChatGPT models.
Describe the solution you'd like
The option for persistent storage of the API key.
Describe alternatives you've considered
I haven't found an alternative, I've just re-entered the key each time the Docker container is restarted.
Additional context
I'm running Ollama Web UI in a Docker container on Windows (WSL2).
@justinh-rahb commented on GitHub (Jan 23, 2024):
You can set a container environment variable by adding
-e OPENAI_API_KEY=your_api_keyto yourdocker runcommand, or adding another line to the environment section of the ollama-webui service in your Docker Compose file, depending how you launched it.@mtompkins commented on GitHub (Jan 25, 2024):
It may be worth considering moving the OpenAI implementation / API key to the user level as this bears a cost. That way the server "owner" wouldn't have to worry about users and their usage on the paid API.
@justinh-rahb commented on GitHub (Jan 25, 2024):
@mtompkins This was the previous implementation, it was moved from client-side to backend in #381