alt Hacker NewsI can't Gemma4 to actually finish a turn properly, it's always ending abruptly or making malformed tool calls. It's probably something I've misconfigured in oMLX or Opencode.
Replies
acrispino • yesterday at 2:25 AM
possibly a problem with the chat template
https://huggingface.co/google/gemma-4-31B-it/discussions/118
clusterhacks • last Monday at 8:26 PM
Huh. Same problem, and I run with llama.cpp. In my case, Gemma4-31B (4-bit quant though) will just stop sometimes.
Same problem with Gemma 4 + oMLX + OpenCode. The thinking and tool calling seems to be parsed fine in other clients such as Open WebUI. This really shouldn’t even matter because the client isn’t responsible for parsing the output, but it’s happening anyway.