Switch pytorch_inf from float32 to auto to fix OOM on 16GB machines#10

Merged

jewilder merged 1 commit into

microsoft:mainfrom

philnach:fix/pytorch-inf-bfloat16-memory

Apr 2, 2026

Commits on Apr 2, 2026

Fix pytorch_inf model loading: use torch_dtype=auto to match official example and fix OOM on 16GB machines
philnach
committed