vllm-sr/Decision-2.0-Kai-0.6B
Different architecture vs Qwen/Qwen3-0.6B-Base. Different architecture: only 0 of 310/10 tensors match by name and shape.
Different architecture vs Qwen/Qwen3-0.6B-Base. Different architecture: only 0 of 310/10 tensors match by name and shape.