feat: use smaller Qwen3 non-thinking model
Browse files- Dockerfile +1 -1
Dockerfile
CHANGED
|
@@ -25,7 +25,7 @@ RUN uv pip install -r requirements.txt
|
|
| 25 |
COPY --chown=user scripts/* /app/scripts/
|
| 26 |
RUN chmod +x scripts/docker-deploy.sh
|
| 27 |
|
| 28 |
-
ENV NML_AI_CHAT_MODEL="huggingface:Qwen/Qwen3-
|
| 29 |
ENV NML_AI_VS_CONFIG="/app/scripts/vector-stores.json"
|
| 30 |
ENV RUNNING_IN_DOCKER=1
|
| 31 |
|
|
|
|
| 25 |
COPY --chown=user scripts/* /app/scripts/
|
| 26 |
RUN chmod +x scripts/docker-deploy.sh
|
| 27 |
|
| 28 |
+
ENV NML_AI_CHAT_MODEL="huggingface:Qwen/Qwen3-4B-Instruct-2507:cheapest"
|
| 29 |
ENV NML_AI_VS_CONFIG="/app/scripts/vector-stores.json"
|
| 30 |
ENV RUNNING_IN_DOCKER=1
|
| 31 |
|