After the latest commit model giving weak results
#14
by
Hrushi11
- opened
After the latest commit I believe the model is giving weak results, Earlier the results were quite good for a prompt like this.
prmpt = """dreamlikeart, cute golden retriever, night theme, 4k photography"""
It is giving such results. (image is attached) Also the inference time is reduced, the model is giving quite fast results.
I think its because I uploaded f16 vae, it should always be f32. Try now
Hmm, no, it's running fine locally
No still the results are not good, the inference time for some reason has also reduced
Try now
Check with the hosted inference API
Fixed
not-someone
changed discussion status to
closed
Yes, Working, Thanks