I’m no expert, but:
problems with creating a blog on Astro have already been discussed here (not working).
and problems with using Ollama and qwen2.5 have been discussed here (hallucinates).
you can use search function to find a topics about those.
Failed to fetch is an issue with your model response.
It looks like you Ollama running and bolt.diy detects it, so can you just confirm it returns a response using that model from the command line?
but solution is not given there
Does Ollama return a response from the cli?
its a longshot but you could try getting a different identical qwen2.5 model. depending on who quantized it will have different output results.
It almost looks like the api is not working.
yeah…
because there is no absolute solution.
for example qwen used to work perfectly for me, but now it doesn’t work at all
there is Debug Info in the settings. try to upload it here, maybe someone can help.