Akarshan Biswas
Akarshan Biswas
The new llama.cpp backend supports running multiple models at the same time, given enough VRAM is available(disabled by default).
Initial attempt: Problem: After tool calling, the response received is a different response Possible consideration: attempt using responses like API.
@urmauur Can push the image attachment file chooser to tauri rfd fix here.
Already rebased and tested yesterday. [Screencast_20251104_212314.webm](https://github.com/user-attachments/assets/afad3478-3c63-463b-b59d-ff8296ea0151)
The conflicts needs to be resolved!
This is already solved with common cpus builds which targets all CPUs
> A program crash will queue the user to investigate what's wrong. Degraded performance just might make the user think the software/model is of low quality. Exactly! I see no...
``` [2025-09-27][09:18:11][tauri_plugin_llamacpp::process][INFO] Sending SIGTERM to PID 172 [2025-09-27][09:18:11][tauri_plugin_llamacpp::commands][INFO] [llamacpp] srv operator(): operator(): cleaning up before exit... [2025-09-27][09:18:11][tauri_plugin_llamacpp::process][INFO] Process exited gracefully: exit status: 0 ``` This 3 lines indicate that the...
Unfortunately, I cannot reproduce this. The logs (with the exception of an error from hardware-plugin) LGTM so far except that the server process seems to be stopped manually. I need...
Need testing on AMD GPU. Logs clearly say that llama-server is being terminated gracefully (with no errors or whatsoever). Some third party programs or scripts might be doing this. I...