Run Local DeepSeek-R1-Distill-Qwen32B on Ollama, Error

[jazzlee008]

Which version of the app are you using?

VSCode 1.96.4, Roo-Code 3.3.4

Which API Provider are you using?

Ollama

Which Model are you using?

DeepSeek-R1-Distill-Qwen32B download from huggingface

What happened?

The local LLM can answer the first question and showed API Streaming Failed after that.

Error messages:

API Streaming Failed
Cannot read properties of undefined (reading 'type')

[TASK RESUMPTION] This task was interrupted just now. It may or may not be complete, so please reassess the task context. Be aware that the project state may have changed since then. The current working directory is now 'e:/mycoding/python/AutoPG'. If the task has not been completed, retry the last step before interruption and proceed with completing the task.

Note: If you previously attempted a tool use that the user did not provide a result for, you should assume the tool use was not successful and assess whether you should retry. If the last tool was a browser_action, the browser has been closed and you must launch a new browser if needed.

IMPORTANT: If the last tool use was a write_to_file that was interrupted, the file was reverted back to its original state before the interrupted edit, and you do NOT need to re-read the file as you already have its up-to-date contents.

<environment_details>

VSCode Visible Files

bubble_sort.py

VSCode Open Tabs

bubble_sort.py

Current Time

1/28/2025, 1:46:19 PM (Asia/Shanghai, UTC+8:00)

Current Mode

code

Current Working Directory (e:/mycoding/python/AutoPG) Files

bubble_sort.py
</environment_details>

Steps to reproduce

1. Setup LLM configuration
2. Ask Roo to build up bubble sort code
3. LLM show the code in chat window
4. Boom, red error showed there
5. At same situation, Google Gemini was running well

Relevant API REQUEST output

[TASK RESUMPTION] This task was interrupted just now. It may or may not be complete, so please reassess the task context. Be aware that the project state may have changed since then. The current working directory is now 'e:/mycoding/python/AutoPG'. If the task has not been completed, retry the last step before interruption and proceed with completing the task.

Note: If you previously attempted a tool use that the user did not provide a result for, you should assume the tool use was not successful and assess whether you should retry. If the last tool was a browser_action, the browser has been closed and you must launch a new browser if needed.

IMPORTANT: If the last tool use was a write_to_file that was interrupted, the file was reverted back to its original state before the interrupted edit, and you do NOT need to re-read the file as you already have its up-to-date contents.

<environment_details>
# VSCode Visible Files
bubble_sort.py

# VSCode Open Tabs
bubble_sort.py

# Current Time
1/28/2025, 1:46:19 PM (Asia/Shanghai, UTC+8:00)

# Current Mode
code

# Current Working Directory (e:/mycoding/python/AutoPG) Files
bubble_sort.py
</environment_details>

Additional context

No response

[jsfaint]

I got same issue with deepseek online version.
Roo Code version is 3.3.4

[Szpadel]

this should be fixed with #606

[markamccorkle]

Not unique to this exact model, but I get a continued Roo is having trouble... while using any Ollama models including this one. Are we saying 3.3.4 should work with Ollama with local R1 models and actually create and edit files?

[jazzlee008]

Updated roo-code version, error changed to "Unexpected API Response: The language model did not provide any assistant messages. This may indicate an issue with the API or the model's output."

Version
3.3.11
Last Updated
2025-02-05, 21:09:29

[nissa-seru]

Closing as dupe since this will be addressed by action plan for #612