fix: correct CLIP Vision model file extensions to .bin

Changed CLIP Vision model destination filenames from .safetensors to .bin
to match actual PyTorch model format. The files are ZIP/pickle format with
'PK' magic bytes, not safetensors format, which was causing "header too large"
deserialization errors in ComfyUI IPAdapterUnifiedLoader.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

comfyui_models.yaml

@@ -219,7 +219,7 @@ model_categories:
       notes: Text-image understanding model
       files:
         - source: "pytorch_model.bin"
-          dest: "CLIP-ViT-H-14-laion2B-s32B-b79K.safetensors"
+          dest: "CLIP-ViT-H-14-laion2B-s32B-b79K.bin"
 
     - repo_id: laion/CLIP-ViT-bigG-14-laion2B-39B-b160k
       description: CLIP G - For SDXL IP-Adapter
@@ -232,7 +232,7 @@ model_categories:
       notes: Larger CLIP model for SDXL
       files:
         - source: "open_clip_pytorch_model.bin"
-          dest: "CLIP-ViT-bigG-14-laion2B-39B-b160k.safetensors"
+          dest: "CLIP-ViT-bigG-14-laion2B-39B-b160k.bin"
 
     - repo_id: google/siglip-so400m-patch14-384
       description: SigLIP - For FLUX models