Mark James Adams

Results 5 comments of Mark James Adams

With [Homebrew](https://brew.sh) ```sh brew install [email protected] pip3 install torch torchvision pip3 install setuptools_rust pip3 install -U git+https://github.com/huggingface/diffusers.git pip3 install transformers scipy ftfy ``` Then start `python3` and follow the instructions...

Looking at the code and the equations referenced, [`sample_euler` in k_diffusion](https://github.com/crowsonkb/k-diffusion/blob/4567328deec0f90a549dde53c1858cf7ed0140dd/k_diffusion/sampling.py#L51) is the same or very similar to [`KarrasVeScheduler` in diffusers](https://github.com/huggingface/diffusers/blob/5164c9faa9d40f55fcc4ff1940f1171a5d43b9e0/src/diffusers/schedulers/scheduling_karras_ve.py#L25). So that would be a good starting point for...

**MacBook Pro 14-inch, 2023; Apple M2 Pro, 8-P-Core, 4-E-core, 19-GPU-core; 32GB Memory** Model: stable-diffusion-2-base Guidance Scale: 7.5 Step count: 25 GPU: 11.0s, 11.1s, 11.0s ANE: 10.6s, 10.8s, 10.7s, GPU+ANE: 10.5s,...

Maybe also this parameter could have a different name since the value it accepts is interpreted as `-log(P)` and would also depend on the value of the `logp` argument.

I think the idea is to use `gwas2vcf` to create a VCF for each separate sumstats file, then merge them together using a tool like [`bcftools merge`](https://samtools.github.io/bcftools/bcftools.html#merge).