DiarizationLM-Gemma-4-v1
Google DeepMind🇺🇸 USFrontier· Oct 4, 2026
TextVisionopen weights
DiarizationLM is a Large Language Model framework designed to post-process, correct, and optimize automatic speech recognition (ASR) and speaker diarization outputs.
From the model card
- License
- apache-2.0
- Modalities
- text, image → text
1 variant
| E4B | E4B | HF |
Sources
HuggingFacePaperProvenance
- Detected
- Oct 5, 2026
- First source
- huggingface
- Description
- From the model card
- Editorially reviewed
- —