huggingface/diffusers

Pull Requests Commits

enable group_offload cases and quanto cases on XPU (#11405)

yao-matrix committed 1 year ago

Verified 9ce89e2e

[tests] add tests to check for graph breaks, recompilation, cuda syncs in pipelines during torch.compile() (#11085)

sayakpaul committed 1 year ago

Verified aa5f5d41

[train_dreambooth_lora.py] Set LANCZOS as default interpolation mode for resizing (#11421)

merterbak committed 1 year ago

Verified bd96a084

Modular Diffusers Guiders (#11311)

a-r-r-o-w committed 1 year ago

Verified b863bdd6

Fix typos in strings and comments (#11407)

co63oc committed 1 year ago

Verified f00a9957

[BUG] fixed WAN docstring (#11226)

ishan-modi committed 1 year ago

Verified e8312e7c

Fix Flux IP adapter argument in the pipeline example (#11402)

AeroDEmi committed 1 year ago

Verified 79868345

[HiDream LoRA] optimizations + small updates (#11381)

linoytsaban committed 1 year ago

Verified edd78804

Kolors additional pipelines, community contrib (#11372)

Teriks committed 1 year ago

Verified b4be4228

[Feature] Added Xlab Controlnet support (#11249)

ishan-modi committed 1 year ago

Verified a4f9c3cb

[train_dreambooth_flux] Add LANCZOS as the default interpolation mode for image resizing (#11395)

ishandutta0098 committed 1 year ago

Verified 4b60f4b6

Fix group offloading with block_level and use_stream=True (#11375)

a-r-r-o-w committed 1 year ago

Verified 6cef71de

Update README_hidream.md (#11386)

AMEERAZAM08 committed 1 year ago

Verified 026507c0

[HiDream] move deprecation to 0.35.0 (#11384)

yiyixuxu committed 1 year ago

Verified 448c72a2

Update modeling imports (#11129)

a-r-r-o-w committed 1 year ago

Verified f108ad88

[LoRA] add LoRA support to HiDream and fine-tuning script (#11281)

linoytsaban committed 1 year ago

Verified e30d3bf5

Add stochastic sampling to FlowMatchEulerDiscreteScheduler (#11369)

apolinario committed 1 year ago

Verified 6ab62c74

[Refactor] Minor Improvement for import utils (#11161)

ishan-modi committed 1 year ago

Verified f59df3bb

Support different-length pos/neg prompts for FLUX.1-schnell variants like Chroma (#11120)

josephrocca committed 1 year ago

Verified a00c73a5

[cogview4][feat] Support attention mechanism with variable-length support and batch packing (#11349)

OleehyO committed 1 year ago

Verified 0434db9a

Add Serialized Type Name kwarg in Model Output (#10502)

anzr299 committed 1 year ago

Verified aff574fb

[BUG] fixes in kadinsky pipeline (#11080)

ishan-modi committed 1 year ago

Verified 79ea8eb2

Fix Wan I2V prepare_latents dtype (#11371)

a-r-r-o-w committed 1 year ago

Verified e7f3a737

fix issue that training flux controlnet was unstable and validation r… (#11373)

PromeAIpro committed 1 year ago

Verified 7a4a126d

[train_dreambooth_lora_sdxl.py] Fix the LR Schedulers when num_train_epochs is passed in a distributed training env (#11240)

kghamilton89 committed 1 year ago

Verified 0dec414d

[Flux LoRAs] fix lr scheduler bug in distributed scenarios (#11242)

linoytsaban committed 1 year ago

Verified 44eeba07

[Wan2.1-FLF2V] update conversion script (#11365)

yiyixuxu committed 1 year ago

Verified 5873377a

update output for Hidream transformer (#11366)

yiyixuxu committed 1 year ago

Verified 5a2e0f71

Fix: `StableDiffusionXLControlNetAdapterInpaintPipeline` incorrectly inherited `StableDiffusionLoraLoaderMixin` (#11357)

Kazuki-Yoda committed 1 year ago

Verified ef47726e

support Wan-FLF2V (#11353)

yiyixuxu committed 1 year ago

Verified 0021bfa1

Newer Older