Finegrain/refiners
1
0
Fork 0
mirror of https://github.com/finegrain-ai/refiners.git synced 2024-11-21 21:58:47 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Add a note that we mitigate non-zero SNR in DDIM.

Browse source
This commit is contained in:
Pierre Chapuis 2024-02-26 10:47:56 +01:00
parent 87f798778f
commit 91d1b46aa9
1 changed files with 2 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

3
src/refiners/foundationals/latent_diffusion/solvers/solver.py
View file

@ -33,7 +33,8 @@ class TimestepSpacing(str, Enum):
Attributes: Attributes:
LINSPACE: Sample N steps with linear interpolation, return a floating-point tensor. LINSPACE: Sample N steps with linear interpolation, return a floating-point tensor.
LINSPACE_ROUNDED: Same as LINSPACE but return an integer tensor with rounded timesteps. LINSPACE_ROUNDED: Same as LINSPACE but return an integer tensor with rounded timesteps.
LEADING: Sample N+1 steps, do not include the last timestep (i.e. bad - non-zero SNR). Used in DDIM. LEADING: Sample N+1 steps, do not include the last timestep (i.e. bad - non-zero SNR).
Used in DDIM, with a mitigation for that issue.
TRAILING: Sample N+1 steps, do not include the first timestep. TRAILING: Sample N+1 steps, do not include the first timestep.
CUSTOM: Use custom timespacing in solver (override `_generate_timesteps`, see DPM). CUSTOM: Use custom timespacing in solver (override `_generate_timesteps`, see DPM).
""" """