usestrix
Strix can not connect to ollama local model on macos
When trying to use Strix with ollama as the local model on macos, the agent starts but can not connect
To Reproduce Steps to reproduce the behavior:
- Install Strix on macOS 14.4.1.
- Set up ollama locally with the model available.
- Configure Strix environment variables: ◦ export STRIX_LLM="ollama/deepseek-r1:8b" ◦ export LLM_API_BASE="http://localhost:11434" or "http://localhost:11434/v1" ◦ export LLM_API_KEY=local
- Run the command: strix --target http://xxx.cn
Strix runs but return below errors: LLM Request Failed Details: litellm.APIConnectionError: OllamaException - │ │ │ │ litellm.Timeout: Connection timed out after 180.0 seconds.
And the local ollama run smoothly as attached
I am not sure whether the configuration of Strix environment variables error or the local ollama models error, Could anyone can check and advise how to resolve this issue? Thanks.
I face the same issue when I run get-oss:20b on my MacBook Pro M2. I can run the LLM in terminal fairly fast but Strix hits the timeout and gets error 500 in 127.0.0.1/api/generate call. When I use a lighter model such as llama3.1:8b, it seems like Strix moves forward but keeps repeating a general response about the procedure and not really a penetration test. With openAI API and Key, it works fine. Not sure what is exactly the bottleneck
I face the same issue when I run get-oss:20b on my MacBook Pro M2. I can run the LLM in terminal fairly fast but Strix hits the timeout and gets error 500 in 127.0.0.1/api/generate call. When I use a lighter model such as llama3.1:8b, it seems like Strix moves forward but keeps repeating a general response about the procedure and not really a penetration test. With openAI API and Key, it works fine. Not sure what is exactly the bottleneck
I have fixed the issue as I use gpt-oss:20b-cloud models instead but face another 429 rate limit issue, then I tried to use local ollama and gpt-oss:20b and faced another exception as Connection timed out after 180.0 seconds. I have no idea on how to set the timeout more than 180 seconds, I think strix need more steps to optimize the services currently.
