From 985af0aca41919237f0a16f7d27e7176bccb582c Mon Sep 17 00:00:00 2001 From: Kashif Rasul Date: Tue, 14 Feb 2023 10:43:59 +0100 Subject: [PATCH 1/5] initial docs about KarrasDiffusionSchedulers --- docs/source/en/api/schedulers/overview.mdx | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) diff --git a/docs/source/en/api/schedulers/overview.mdx b/docs/source/en/api/schedulers/overview.mdx index d27fbe10c528..106838a21a06 100644 --- a/docs/source/en/api/schedulers/overview.mdx +++ b/docs/source/en/api/schedulers/overview.mdx @@ -83,4 +83,24 @@ The class [`SchedulerOutput`] contains the outputs from any schedulers `step(... ### KarrasDiffusionSchedulers +`KarrasDiffusionSchedulers` encompasses the main generalization of schedulers in Diffusers. The schedulers in this class are distinguished at a hight level by their noise sampling strategy; the type of network and scaling; and finally the training strategy or how the loss is weighed. + +The different schedulers, depending on the type of ODE solver, fall into the above taxonomy and provides a good abstraction for the design of the different schedulers. The main schedulers in this class are give by the following `Enum`: + +```python +class KarrasDiffusionSchedulers(Enum): + DDIMScheduler = 1 + DDPMScheduler = 2 + PNDMScheduler = 3 + LMSDiscreteScheduler = 4 + EulerDiscreteScheduler = 5 + HeunDiscreteScheduler = 6 + EulerAncestralDiscreteScheduler = 7 + DPMSolverMultistepScheduler = 8 + DPMSolverSinglestepScheduler = 9 + KDPM2DiscreteScheduler = 10 + KDPM2AncestralDiscreteScheduler = 11 + DEISMultistepScheduler = 12 +``` + [[autodoc]] schedulers.scheduling_utils.KarrasDiffusionSchedulers From b0418ec76e8fae5340cc829e571ca33f1fca63b3 Mon Sep 17 00:00:00 2001 From: Kashif Rasul Date: Tue, 14 Feb 2023 10:47:01 +0100 Subject: [PATCH 2/5] typo --- docs/source/en/api/schedulers/overview.mdx | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/source/en/api/schedulers/overview.mdx b/docs/source/en/api/schedulers/overview.mdx index 106838a21a06..23643cb2bb6c 100644 --- a/docs/source/en/api/schedulers/overview.mdx +++ b/docs/source/en/api/schedulers/overview.mdx @@ -83,9 +83,9 @@ The class [`SchedulerOutput`] contains the outputs from any schedulers `step(... ### KarrasDiffusionSchedulers -`KarrasDiffusionSchedulers` encompasses the main generalization of schedulers in Diffusers. The schedulers in this class are distinguished at a hight level by their noise sampling strategy; the type of network and scaling; and finally the training strategy or how the loss is weighed. +`KarrasDiffusionSchedulers` encompasses the main generalization of schedulers in Diffusers. The schedulers in this class are distinguished, at a high level, by their noise sampling strategy; the type of network and scaling; and finally the training strategy or how the loss is weighed. -The different schedulers, depending on the type of ODE solver, fall into the above taxonomy and provides a good abstraction for the design of the different schedulers. The main schedulers in this class are give by the following `Enum`: +The different schedulers, depending on the type of ODE solver, fall into the above taxonomy and provides a good abstraction for the design of the main schedulers provided by Diffusers. The schedulers in this class are give by the following `Enum`: ```python class KarrasDiffusionSchedulers(Enum): From c736d24bb7fa4cc2b1614c9b8cbcd7daaf9a8f27 Mon Sep 17 00:00:00 2001 From: Kashif Rasul Date: Tue, 14 Feb 2023 10:49:42 +0100 Subject: [PATCH 3/5] grammer --- docs/source/en/api/schedulers/overview.mdx | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/source/en/api/schedulers/overview.mdx b/docs/source/en/api/schedulers/overview.mdx index 23643cb2bb6c..2730ac8a666f 100644 --- a/docs/source/en/api/schedulers/overview.mdx +++ b/docs/source/en/api/schedulers/overview.mdx @@ -85,7 +85,7 @@ The class [`SchedulerOutput`] contains the outputs from any schedulers `step(... `KarrasDiffusionSchedulers` encompasses the main generalization of schedulers in Diffusers. The schedulers in this class are distinguished, at a high level, by their noise sampling strategy; the type of network and scaling; and finally the training strategy or how the loss is weighed. -The different schedulers, depending on the type of ODE solver, fall into the above taxonomy and provides a good abstraction for the design of the main schedulers provided by Diffusers. The schedulers in this class are give by the following `Enum`: +The different schedulers, depending on the type of ODE solver, fall into the above taxonomy and provide a good abstraction for the design of the main schedulers implemented in Diffusers. The schedulers in this class are give by the following `Enum`: ```python class KarrasDiffusionSchedulers(Enum): From 47b8e16477a7a574fcafe351f7903498f56e2423 Mon Sep 17 00:00:00 2001 From: Kashif Rasul Date: Tue, 14 Feb 2023 11:29:14 +0100 Subject: [PATCH 4/5] Update docs/source/en/api/schedulers/overview.mdx Co-authored-by: Pedro Cuenca --- docs/source/en/api/schedulers/overview.mdx | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/source/en/api/schedulers/overview.mdx b/docs/source/en/api/schedulers/overview.mdx index 2730ac8a666f..c5fab520a1ea 100644 --- a/docs/source/en/api/schedulers/overview.mdx +++ b/docs/source/en/api/schedulers/overview.mdx @@ -85,7 +85,7 @@ The class [`SchedulerOutput`] contains the outputs from any schedulers `step(... `KarrasDiffusionSchedulers` encompasses the main generalization of schedulers in Diffusers. The schedulers in this class are distinguished, at a high level, by their noise sampling strategy; the type of network and scaling; and finally the training strategy or how the loss is weighed. -The different schedulers, depending on the type of ODE solver, fall into the above taxonomy and provide a good abstraction for the design of the main schedulers implemented in Diffusers. The schedulers in this class are give by the following `Enum`: +The different schedulers, depending on the type of ODE solver, fall into the above taxonomy and provide a good abstraction for the design of the main schedulers implemented in Diffusers. The schedulers in this class are given by the following `Enum`: ```python class KarrasDiffusionSchedulers(Enum): From d7f566fd7a79deb371527617dae5906ccacc3633 Mon Sep 17 00:00:00 2001 From: Kashif Rasul Date: Wed, 15 Feb 2023 09:12:02 +0100 Subject: [PATCH 5/5] do not list the schedulers explicitly --- docs/source/en/api/schedulers/overview.mdx | 18 +----------------- 1 file changed, 1 insertion(+), 17 deletions(-) diff --git a/docs/source/en/api/schedulers/overview.mdx b/docs/source/en/api/schedulers/overview.mdx index c5fab520a1ea..35c26a721964 100644 --- a/docs/source/en/api/schedulers/overview.mdx +++ b/docs/source/en/api/schedulers/overview.mdx @@ -85,22 +85,6 @@ The class [`SchedulerOutput`] contains the outputs from any schedulers `step(... `KarrasDiffusionSchedulers` encompasses the main generalization of schedulers in Diffusers. The schedulers in this class are distinguished, at a high level, by their noise sampling strategy; the type of network and scaling; and finally the training strategy or how the loss is weighed. -The different schedulers, depending on the type of ODE solver, fall into the above taxonomy and provide a good abstraction for the design of the main schedulers implemented in Diffusers. The schedulers in this class are given by the following `Enum`: - -```python -class KarrasDiffusionSchedulers(Enum): - DDIMScheduler = 1 - DDPMScheduler = 2 - PNDMScheduler = 3 - LMSDiscreteScheduler = 4 - EulerDiscreteScheduler = 5 - HeunDiscreteScheduler = 6 - EulerAncestralDiscreteScheduler = 7 - DPMSolverMultistepScheduler = 8 - DPMSolverSinglestepScheduler = 9 - KDPM2DiscreteScheduler = 10 - KDPM2AncestralDiscreteScheduler = 11 - DEISMultistepScheduler = 12 -``` +The different schedulers, depending on the type of ODE solver, fall into the above taxonomy and provide a good abstraction for the design of the main schedulers implemented in Diffusers. The schedulers in this class are given below: [[autodoc]] schedulers.scheduling_utils.KarrasDiffusionSchedulers