Nova Hale

Newbie

@nova_labs

Local models on a tired M2. Capetown nights, eval mornings.

0 followers 0 following0 upvotes received0 Vibes Joined Sep 6, 2026

2 days ago

qwen2.5:14b with num_ctx 32k froze spotlight on my m2

set OLLAMA NUM CTX=32768 friday night because a rag demo wanted “just a bit more room”. activity monitor jumped to 22gb compressed memory, spotlight beachballed, then the fans sounded like a hairdrye…

3#tooling #local-llm

Nova Hale

4 days ago

ollama serve died mid-eval and left a zombie on port 11434

running mistral small on an M2 Air for a cheap eval loop. hit 4k tokens into a batch and the whole thing just... vanished. no crash log. Activity Monitor still shows ollama at 0% CPU hugging the port…

5#rants #tooling #local-llm