Conversation
Co-authored-by: jtracey93 <41163455+jtracey93@users.noreply.github.com>
Co-authored-by: jtracey93 <41163455+jtracey93@users.noreply.github.com>
Co-authored-by: jtracey93 <41163455+jtracey93@users.noreply.github.com>
|
@jtracey93 I see you mention this fixes #483, but it seems to do something different than what is requested there? |
|
@cjpais Very fair. I must have associated this very loosely in my mind. However, it does add the ability to set the URL for a custom LLM via an environment variable, which I would find very useful as today using Windows Foundry Local or other services such as ollama, etc.. I'm having to start those manually and then update the settings via the UI each time I launch Handy or restart my PC and its painful |
|
yeah understood, I will also get this in, I just wanted to clarify |
|
hey @cjpais - sorry to chase but do you think we could get this in soon as would really speed up my workflow 👍 |
|
@jtracey93 i need more time, features are really not a priority at the moment. I am really trying to stabilize the app and adding new features adds new requests and bug pathways. If you need it now, I highly recommend using your own build. I can also kick off a build for this PR so everyone can use this if they need. |
🧪 Test Build ReadyBuild artifacts for PR #633 are available for testing. Download artifacts from workflow run Artifacts expire after 30 days. |
…LLM env URL - PR cjpais#477: Graceful error handling in audio recorder worker thread instead of panicking when no microphone config is available or stream fails to build - PR cjpais#747: Lazy stream close for bluetooth mic latency - keeps mic stream open for 30s after recording stops in OnDemand mode, eliminating BT activation delay - PR cjpais#633: Support PHONARA_CUSTOM_LLM_BASE_URL env var to override LLM base URL at runtime (useful for local LLM services on dynamic ports) - PR cjpais#872: Bump macOS minimum system version from 10.13 to 10.15 - PR cjpais#976: Add tests confirming long repeating word stutter collapse works Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Before Submitting This PR
Please confirm you have done the following:
Human Written Description
I use Windows foundry local on my machine to provide my local LLMs. However, every time the service starts, it will pick a non-deterministic port to start on, which therefore means I need a way to provide the latest base URL for the custom LLMs every time. This pull request adds support for adding an environment variable which will be used and checked every time the LLM is needed to be called, which will meet the needs of my use case and many others.
Related Issues/Discussions
Fixes #483
Testing
Screenshots/Videos (if applicable)
AI Assistance
If AI was used: