CLIP ViT-B/32 (LAION) โ€” BonfyreFPQ v12

Native .fpq v12 quantized weights for laion/CLIP-ViT-B-32-laion2B-s34B-b79K.

Model Info

Property Value
Architecture Multimodal (Vision-Language)
Parameters 151M
Original Size 577 MB (safetensors)
FPQ v12 Size 164 MB
Compression 3.5ร—
Avg Cosine 0.999876
Worst Cosine 0.998477

Format

BonfyreFPQ v12 native format: rANS entropy-coded E8 lattice coordinates with 6-bit packed tiles and FP16 scales.

Decode to safetensors

bonfyre-fpqx decode clip-vit-b32-laion-v12-fpq.fpq output.safetensors

Source

Converted from laion/CLIP-ViT-B-32-laion2B-s34B-b79K using BonfyreFPQ v12.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support