error 11 / LMStudio / MBP
#1
by
Vicnent
- opened
HI
with the BF16 model :
Error loading model.
(Exit code: 11). Please check settings and try loading the model again.
2025/11/24 -
All up to date :
LMStudio : Version 0.3.32 (0.3.32),
Metal LLama.cpp 1.58.0
LM Studio MLX 0.32.0
MBP M3 MAX / 128 GB RAM
Lmstudio says the model is given for 262 144 tokens (but even it is 25k, it doesn't load)
You could try setting a context window size of either 32K or 64K, as although the official Qwen3-4B-Instruct model officially supports up to 256K, its training data is limited to sequences of at most 32K tokens across multiple turns. This may cause the framework to attempt allocating excessively large memory, resulting in errors or crashes.