ollama - 💡(How to fix) Fix is your article about MLX support fake? [4 comments, 3 participants]

Error: pull model manifest: 400: {"error":"Repository is not GGUF or is not compatible with llama.cpp"} Error: pull model manifest: 400: {"error":"Repository is not GGUF or is not compatible with llama.cpp"}

TL;DR

The issue is likely due to incompatibility between the Ollama version and the model repositories, and trying a different model or updating Ollama might resolve the issue.

Guidance

• The error message suggests that the repository is not compatible with llama.cpp, indicating a potential version or compatibility issue.
• Try running the model with a different repository or version to see if the issue persists.
• The mention of Ollama 0.19 in the blog post and the user's version being 0.20.5 might imply a version mismatch, so checking the compatibility of the model with the current Ollama version is necessary.
• Verify that the model repositories used (hf.co/NexVeridian/gemma-4-31B-it-8bit and hf.co/nvidia/Gemma-4-31B-IT-NVFP4) are compatible with the current Ollama version.

Notes

The issue lacks specific details about the OS, GPU, CPU, and Ollama version, which could be relevant for troubleshooting. Additionally, the empty log output does not provide further insights into the issue.

Recommendation

Apply workaround: Try a different model or repository to see if the issue is specific to the current models being used, as this can help determine if the problem is with the models or the Ollama setup.