meetween/SpeechLMM-v2.0-L-7B

Different architecture vs Qwen/Qwen2.5-Omni-7B. Different architecture: only 1104 of 2448/3057 tensors match by name and shape.

On the Hub · Downloads on Model Pulse