mradermacher/VLM-R1-Qwen2.5VL-3B-OVD-0321-GGUF
Same weights, converted (fewer layers) vs omlab/VLM-R1-Qwen2.5VL-3B-OVD-0321. mradermacher/VLM-R1-Qwen2.5VL-3B-OVD-0321-GGUF matches part of omlab/VLM-R1-Qwen2.5VL-3B-OVD-0321 (434 of its 824 tensors; the rest of omlab/VLM-R1-Qwen2.5VL-3B-OVD-0321 has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/F32).
Tensors changed: Embeddings 0/1, Norms 0/73, MLP 0/108, Attention 0/252