User reports different Qwen3.5 35B-A3B results on Tinker versus Alibaba Cloud
A user asks whether others have used image inputs with Qwen3.5 35B-A3B on Tinker.
TLDR
A user reports seeing “very different sampling results” from Qwen3.5 35B-A3B on Tinker compared with the same model on Alibaba Cloud. They ask whether anyone else has used image inputs with the model on Tinker.
Combined views
—
1 Source, first seen ago
