GGUF quantizations of https://huggingface.co/lodestones/Chroma

Update (July 10, 2025): Requantized and re-uploaded v42 because Chroma's author re-uploaded new weights for this version.

Note: Q3 and below quants degrade in quality so badly that I skipped creating them. More versions will be added later, stay tuned.

Quality test (v41)

Sampler name: Euler

Scheduler: beta

Steps: 30


Q4_0

Q4_0

Q4_1

Q4_1

Q4_K_M

Q4_K_M

Q4_K_S

Q4_K_S

Q5_0

Q5_0

Q5_1

Q5_1

Q5_K_M

Q5_K_M

Q5_K_S

Q5_K_S

Q6_K

Q6_K

Q8_0

Q8_0

Downloads last month
1,482
GGUF
Model size
8.9B params
Architecture
flux
Hardware compatibility
Log In to view the estimation

4-bit

5-bit

6-bit

8-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support