apunkt/dograh
1
0
Fork 0
mirror of https://github.com/dograh-hq/dograh.git synced 2026-06-22 08:38:13 +02:00
Code Issues Projects Releases Packages Wiki Activity

fix: fix interruption handling for Gemini Live

Browse source 
1. Fixes #236
2. Fix run_inference for variable extraction for Gemini Live
This commit is contained in:
Abhishek Kumar 2026-04-15 19:29:07 +05:30
parent 14e6f29f2f
commit e31b38122e
12 changed files with 48 additions and 15 deletions
Download patch file Download diff file Expand all files Collapse all files

4
docs/configurations/inference-providers.mdx
View file

@ -73,7 +73,7 @@ For example, if you only want to change the voice for a specific agent:
You can also switch an individual agent to use a **Realtime** provider (such as Gemini Live) even if the global configuration uses standard LLM + TTS + STT. Toggle the **Realtime** switch in the Model Overrides tab, then configure the realtime provider, model, and voice.
<Note>
When an agent uses a Realtime provider, it replaces the separate LLM, TTS, and STT services with a single speech-to-speech model. The individual LLM/TTS/STT override tabs are hidden in this mode.
When an agent uses a Realtime provider, it replaces the separate TTS and STT services with a single speech-to-speech model. An **LLM** is still required alongside the Realtime model — it's used for out-of-band tasks like variable extraction and QA analysis, which the realtime service does not handle. Context compaction is not applicable in Realtime mode and is ignored if enabled.
</Note>
## Gemini 3.1 Live
@ -119,5 +119,5 @@ To use Gemini 3.1 Live with Dograh, you need a Google Gemini API key. Follow the
6. Select the language (currently `en` is supported).
<Note>
When using a Realtime provider like Gemini Live, you do not need to configure separate LLM, TTS, and STT services — the realtime model handles all three.
When using a Realtime provider like Gemini Live, you do not need to configure separate TTS and STT services — the realtime model handles speech in and out. However, you **must** still configure an **LLM** under the LLM tab: it powers variable extraction and QA analysis, which the realtime service does not perform.
</Note>