Add EasyAnimateV5.1 text-to-video, image-to-video, control-to-video generation model #10626

bubbliiiing · 2025-01-22T08:49:52Z

What does this PR do?

This PR converts the EasyAnimateV5.1 model into a diffuser-supported inference model, including three complete pipelines and corresponding modules.

Before submitting

This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
Did you read the contributor guideline?
Did you read our philosophy doc (important for complex PRs)?
Was this discussed/approved via a GitHub issue or the forum? Please add a link to it if that's the case.
Did you make sure to update the documentation with your changes? Here are the
documentation guidelines, and
here are tips on formatting docstrings.
Did you write any new necessary tests?

Who can review?

Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.

@a-r-r-o-w

a-r-r-o-w

Thank you for the PR @bubbliiiing! This is in great shape and already mostly in the implementation style used in diffusers 🤗

I've left some comments from a quick look through the PR. Happy to help make any of the required changes to help bring the PR to completion

.gitignore

src/diffusers/models/attention_processor.py

a-r-r-o-w · 2025-01-22T14:51:10Z

src/diffusers/models/attention_processor.py

+        encoder_hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        image_rotary_emb: Optional[torch.Tensor] = None,
+        attn2: Attention = None,


This seems similar to Flux/SD3/HunyuanVideo's Joint-attention processors that concatenate the visual and text tokens. Let's do it the same way as done here:

diffusers/src/diffusers/models/transformers/transformer_hunyuan_video.py

Line 367 in 8d6f6d6

self.attn = Attention(

diffusers/src/diffusers/models/transformers/transformer_hunyuan_video.py

Line 41 in 8d6f6d6

class HunyuanVideoAttnProcessor2_0:

Does this mean using add_q, add_k, add_v instead of using attn2?

There are two things we could do here:

Either convert the state dict of the original-format models (that you currently have on the HuggingFace Hub) and update them to diffusers-format (which would make attn2.to_q -> add_q_proj, attn2.to_k -> add_k_proj, attn2.to_v -> add_v_proj

Create a custom attention class similar to Attention and MochiAttention in which you are free to use layer naming of your choice (so basically keeping the same to_q, to_k and to_v.

The first approach is more closely aligned with diffusers code style but would require you to update multiple checkpoints -- but we are transitioning to a single file modeling format, so if you choose to go with second approach for convenience, that works for us as well. Essentially, irrespective of the design you choose, we need to make sure:

When the forward of a layer is called, it only takes tensors as input and produces tensors as output.

Taking intermediate layers as input to forward, or making calls to other layers out-of-order randomly, is not supported by our design style of different current/upcoming features

cc @DN6 here in case you have thoughts about the single file format and model-specific Attention classes

I moved attn2 to the processor's init; does this meet the requirements?

src/diffusers/models/autoencoders/autoencoder_kl_magvit.py

a-r-r-o-w · 2025-01-22T14:59:35Z

src/diffusers/models/autoencoders/autoencoder_kl_magvit.py

+        for name, module in self.named_children():
+            _set_3dgroupnorm_for_submodule(name, module)
+
+    def single_forward(self, x: torch.Tensor) -> torch.Tensor:


This is very different from diffusers-style implementation of encoder/decoder. Could we follow the style as done in:

diffusers/src/diffusers/models/autoencoders/autoencoder_dc.py

Line 202 in 8d6f6d6

class Encoder(nn.Module):

diffusers/src/diffusers/models/autoencoders/autoencoder_kl_mochi.py

Line 691 in 8d6f6d6

class AutoencoderKLMochi(ModelMixin, ConfigMixin):

diffusers/src/diffusers/models/autoencoders/autoencoder_kl_hunyuan_video.py

Line 457 in 8d6f6d6

class HunyuanVideoEncoder3D(nn.Module):

Sorry, does this mean that I cannot use functions like set_padding_one_frame?

Do I need to use this conv_cache in autoencoder_kl_mochi?

In the model implementations, we usually only try to keep (atleast in the latest model integrations):

Submodel initializations

Forward method

So, unless a helper function like set_padding_one_frame is used in multiple locations, I would suggest directly substituting its code in the forward implementation. If a helper function is required, let's make it a private function by prefixing the function name with an underscore

The conv_cache saves a few computations when running the VAE encode/decode process from repeated frames that are used as padding. As such, it is not required to implement it if it is not needed for framewise encoding and decoding.

src/diffusers/models/autoencoders/autoencoder_kl_magvit.py

src/diffusers/models/downsampling.py

src/diffusers/models/normalization.py

bubbliiiing · 2025-01-23T03:55:33Z

Sorry for not standardizing some parts; I will make the necessary modifications. Also, I would like to ask if I need to add test files in tests/pipelines and add documentation in docs/source/en/api/pipelines?

a-r-r-o-w · 2025-01-23T21:25:43Z

Yes, we will need a test for all three pipelines as well as model tests in tests/models, as well as the documentation pages. Ofcourse, I will try to help you with everything :)

Also, congratulations on the release! I tried out the original repository example and the model is very good! 🎉

a-r-r-o-w · 2025-02-06T01:50:16Z

Thank you so much for addressing the reviews @bubbliiiing! The PR is almost ready to merge IMO. There are just a few more small changes to make to align to our model implementation design. Is it okay if I quickly push some last changes to this branch directly?

bubbliiiing · 2025-02-06T16:25:26Z

Thank you so much for addressing the reviews @bubbliiiing! The PR is almost ready to merge IMO. There are just a few more small changes to make to align to our model implementation design. Is it okay if I quickly push some last changes to this branch directly?

Of course, thank you very much for your help.

nitinmukesh · 2025-02-10T17:53:21Z

For some reason this isn't working anymore, it was working earlier. :(

import torch
from diffusers import BitsAndBytesConfig as DiffusersBitsAndBytesConfig, EasyAnimateTransformer3DModel, EasyAnimatePipeline
from diffusers.utils import export_to_video
from transformers import AutoModel
dtype = torch.bfloat16
quant_config = DiffusersBitsAndBytesConfig(load_in_4bit=True, bnb_4bit_quant_type="nf4", bnb_4bit_compute_dtype=dtype)
text_encoder_4bit = AutoModel.from_pretrained(
    "alibaba-pai/EasyAnimateV5.1-12b-zh",
    subfolder="text_encoder",
    quantization_config=quant_config,
    torch_dtype=torch.bfloat16,
)
transformer_4bit = EasyAnimateTransformer3DModel.from_pretrained(
    "alibaba-pai/EasyAnimateV5.1-12b-zh",
    subfolder="transformer",
    quantization_config=quant_config,
    torch_dtype=dtype,
)
pipeline = EasyAnimatePipeline.from_pretrained(
    "alibaba-pai/EasyAnimateV5.1-12b-zh",
    text_encoder=text_encoder_4bit,
    transformer=transformer_4bit,
    torch_dtype=dtype,
)
pipeline.enable_model_cpu_offload()
pipeline.vae.enable_slicing()
pipeline.vae.enable_tiling()
prompt = "A cat walks on the grass, realistic style."
negative_prompt = "bad detailed"
video = pipeline(
    prompt=prompt, 
    negative_prompt=negative_prompt, 
    num_frames=81, 
    num_inference_steps=40,
    width=512,
    height=320
).frames[0]
export_to_video(video, "cat2.mp4", fps=8)

It just crashes

(venv) C:\aiOWN\lumina2>python EasyAnimate5.1.py
`low_cpu_mem_usage` was None, now default to True since model is quantized.
Downloading shards: 100%|███████████████████████████████████████████████████| 5/5 [00:00<?, ?it/s]
`Qwen2VLRotaryEmbedding` can now be fully parameterized by passing the model config through the `config` argument. All other arguments will be removed in v4.46
Loading checkpoint shards: 100%|████████████████████████████████████| 5/5 [00:19<00:00,  3.90s/it]
The config attributes {'add_ref_latent_in_control_model': True, 'clip_channels': None, 'enable_clip_in_inpaint': False, 'ref_channels': None, 'swa_layers': None} were passed to EasyAnimateTransformer3DModel, but are not expected and will be ignored. Please verify your config.json configuration file.

bubbliiiing · 2025-02-11T03:36:25Z

由于某种原因，它不再起作用了，之前它是起作用的。:(

import torch
from diffusers import BitsAndBytesConfig as DiffusersBitsAndBytesConfig, EasyAnimateTransformer3DModel, EasyAnimatePipeline
from diffusers.utils import export_to_video
from transformers import AutoModel
dtype = torch.bfloat16
quant_config = DiffusersBitsAndBytesConfig(load_in_4bit=True, bnb_4bit_quant_type="nf4", bnb_4bit_compute_dtype=dtype)
text_encoder_4bit = AutoModel.from_pretrained(
    "alibaba-pai/EasyAnimateV5.1-12b-zh",
    subfolder="text_encoder",
    quantization_config=quant_config,
    torch_dtype=torch.bfloat16,
)
transformer_4bit = EasyAnimateTransformer3DModel.from_pretrained(
    "alibaba-pai/EasyAnimateV5.1-12b-zh",
    subfolder="transformer",
    quantization_config=quant_config,
    torch_dtype=dtype,
)
pipeline = EasyAnimatePipeline.from_pretrained(
    "alibaba-pai/EasyAnimateV5.1-12b-zh",
    text_encoder=text_encoder_4bit,
    transformer=transformer_4bit,
    torch_dtype=dtype,
)
pipeline.enable_model_cpu_offload()
pipeline.vae.enable_slicing()
pipeline.vae.enable_tiling()
prompt = "A cat walks on the grass, realistic style."
negative_prompt = "bad detailed"
video = pipeline(
    prompt=prompt, 
    negative_prompt=negative_prompt, 
    num_frames=81, 
    num_inference_steps=40,
    width=512,
    height=320
).frames[0]
export_to_video(video, "cat2.mp4", fps=8)

它只是崩溃了

(venv) C:\aiOWN\lumina2>python EasyAnimate5.1.py
`low_cpu_mem_usage` was None, now default to True since model is quantized.
Downloading shards: 100%|███████████████████████████████████████████████████| 5/5 [00:00<?, ?it/s]
`Qwen2VLRotaryEmbedding` can now be fully parameterized by passing the model config through the `config` argument. All other arguments will be removed in v4.46
Loading checkpoint shards: 100%|████████████████████████████████████| 5/5 [00:19<00:00,  3.90s/it]
The config attributes {'add_ref_latent_in_control_model': True, 'clip_channels': None, 'enable_clip_in_inpaint': False, 'ref_channels': None, 'swa_layers': None} were passed to EasyAnimateTransformer3DModel, but are not expected and will be ignored. Please verify your config.json configuration file.

Thank you for your feedback, I'll give it a try.

bubbliiiing · 2025-02-11T06:06:25Z

For some reason this isn't working anymore, it was working earlier. :(

import torch
from diffusers import BitsAndBytesConfig as DiffusersBitsAndBytesConfig, EasyAnimateTransformer3DModel, EasyAnimatePipeline
from diffusers.utils import export_to_video
from transformers import AutoModel
dtype = torch.bfloat16
quant_config = DiffusersBitsAndBytesConfig(load_in_4bit=True, bnb_4bit_quant_type="nf4", bnb_4bit_compute_dtype=dtype)
text_encoder_4bit = AutoModel.from_pretrained(
    "alibaba-pai/EasyAnimateV5.1-12b-zh",
    subfolder="text_encoder",
    quantization_config=quant_config,
    torch_dtype=torch.bfloat16,
)
transformer_4bit = EasyAnimateTransformer3DModel.from_pretrained(
    "alibaba-pai/EasyAnimateV5.1-12b-zh",
    subfolder="transformer",
    quantization_config=quant_config,
    torch_dtype=dtype,
)
pipeline = EasyAnimatePipeline.from_pretrained(
    "alibaba-pai/EasyAnimateV5.1-12b-zh",
    text_encoder=text_encoder_4bit,
    transformer=transformer_4bit,
    torch_dtype=dtype,
)
pipeline.enable_model_cpu_offload()
pipeline.vae.enable_slicing()
pipeline.vae.enable_tiling()
prompt = "A cat walks on the grass, realistic style."
negative_prompt = "bad detailed"
video = pipeline(
    prompt=prompt, 
    negative_prompt=negative_prompt, 
    num_frames=81, 
    num_inference_steps=40,
    width=512,
    height=320
).frames[0]
export_to_video(video, "cat2.mp4", fps=8)

It just crashes

(venv) C:\aiOWN\lumina2>python EasyAnimate5.1.py
`low_cpu_mem_usage` was None, now default to True since model is quantized.
Downloading shards: 100%|███████████████████████████████████████████████████| 5/5 [00:00<?, ?it/s]
`Qwen2VLRotaryEmbedding` can now be fully parameterized by passing the model config through the `config` argument. All other arguments will be removed in v4.46
Loading checkpoint shards: 100%|████████████████████████████████████| 5/5 [00:19<00:00,  3.90s/it]
The config attributes {'add_ref_latent_in_control_model': True, 'clip_channels': None, 'enable_clip_in_inpaint': False, 'ref_channels': None, 'swa_layers': None} were passed to EasyAnimateTransformer3DModel, but are not expected and will be ignored. Please verify your config.json configuration file.

It seems to be working fine in my environment.

nitinmukesh · 2025-02-12T06:35:28Z

@bubbliiiing Sorry for the trouble.
It's working now after reinstalling the PR.

nitinmukesh · 2025-02-12T10:24:35Z

20250212_154949_370380_easyanimate51_bnb.mp4

A little white rabbit with glasses was sitting on a chair in a cafe reading a newspaper. There was a cup of hot coffee on the table.

Nice output. Please add support of pipeline.enable_sequential_cpu_offload().

a-r-r-o-w · 2025-02-12T10:39:35Z

@nitinmukesh enable_sequential_cpu_offload() should work with any diffusers model out of the box. Are you facing any particular error? Apologies about iterating slow on this PR as there are many integrations happening in parallel, but will try to wrap it up soon to fix any sequential offloading problem you might be facing

bubbliiiing · 2025-02-12T10:42:10Z

enable_sequential_cpu_offload

It seems that qwen2vl does not support enable_sequential_cpu_offload and will prompt the following error message:

nitinmukesh · 2025-02-12T10:48:22Z

Exactly this is the error I'm getting.
Currently with int4 and enable_model_cpu_offload it is working fine for 10 GB VRAM.
With sequential it should work for as low as 6 GB VRAM. However if it is not supported for qwen2vl we can simply ignore adding enable_sequential_cpu_offload.

a-r-r-o-w · 2025-02-12T10:50:11Z

cc @SunMarc here for sequential cpu offload related issues in Qwen2VL

nitinmukesh · 2025-02-12T10:50:51Z

@nitinmukesh enable_sequential_cpu_offload() should work with any diffusers model out of the box. Are you facing any particular error? Apologies about iterating slow on this PR as there are many integrations happening in parallel, but will try to wrap it up soon to fix any sequential offloading problem you might be facing

No need to aplogize, you guys are doing very good. I'm not in hurry ()maybe excited to see all the development happening) just trying all the features that are being added and posting my feedback. I want to see this library grow as it's very easy to use.

nitinmukesh · 2025-02-12T10:56:22Z

@a-r-r-o-w

Here is the code to reproduce the error

[pipeline.enable_model_cpu_offload() works fine]

import torch
from diffusers import BitsAndBytesConfig as DiffusersBitsAndBytesConfig, EasyAnimateTransformer3DModel, EasyAnimatePipeline
from diffusers.utils import export_to_video
from transformers import AutoModel
dtype = torch.bfloat16
quant_config = DiffusersBitsAndBytesConfig(load_in_4bit=True, bnb_4bit_quant_type="nf4", bnb_4bit_compute_dtype=dtype)
text_encoder_4bit = AutoModel.from_pretrained(
    "alibaba-pai/EasyAnimateV5.1-12b-zh",
    subfolder="text_encoder",
    quantization_config=quant_config,
    torch_dtype=torch.bfloat16,
)
transformer_4bit = EasyAnimateTransformer3DModel.from_pretrained(
    "alibaba-pai/EasyAnimateV5.1-12b-zh",
    subfolder="transformer",
    quantization_config=quant_config,
    torch_dtype=dtype,
)
pipeline = EasyAnimatePipeline.from_pretrained(
    "alibaba-pai/EasyAnimateV5.1-12b-zh",
    text_encoder=text_encoder_4bit,
    transformer=transformer_4bit,
    torch_dtype=dtype,
)
pipeline.enable_sequential_cpu_offload()
pipeline.vae.enable_slicing()
pipeline.vae.enable_tiling()
prompt = "A cat walks on the grass, realistic style."
negative_prompt = "bad detailed"
video = pipeline(
    prompt=prompt, 
    negative_prompt=negative_prompt, 
    num_frames=81, 
    num_inference_steps=40,
    width=512,
    height=320
).frames[0]
export_to_video(video, "cat2.mp4", fps=8)

and log

Log 1: #10626 (comment)

Log 2:

(venv) C:\aiOWN\diffuser_webui>python easyanimate_bnb.py
`low_cpu_mem_usage` was None, now default to True since model is quantized.
Downloading shards: 100%|███████████████████████████████████████████████████| 5/5 [00:00<?, ?it/s]
`Qwen2VLRotaryEmbedding` can now be fully parameterized by passing the model config through the `config` argument. All other arguments will be removed in v4.46
Loading checkpoint shards: 100%|████████████████████████████████████| 5/5 [00:34<00:00,  6.83s/it]
The config attributes {'add_ref_latent_in_control_model': True, 'clip_channels': None, 'enable_clip_in_inpaint': False, 'ref_channels': None, 'swa_layers': None} were passed to EasyAnimateTransformer3DModel, but are not expected and will be ignored. Please verify your config.json configuration file.
Expected types for text_encoder: ['Qwen2VLForConditionalGeneration', 'BertModel'], got Qwen2VLModel.
Loading pipeline components...:   0%|                                       | 0/5 [00:00<?, ?it/s]The config attributes {'force_upcast': True, 'mid_block_use_attention': False, 'sample_size': 256, 'slice_mag_vae': False, 'slice_compression_vae': False, 'cache_compression_vae': False, 'cache_mag_vae': True, 'use_tiling': False, 'norm_type': None, 'use_tiling_encoder': False, 'use_tiling_decoder': False, 'mid_block_attention_type': 'spatial'} were passed to AutoencoderKLMagvit, but are not expected and will be ignored. Please verify your config.json configuration file.
Loading pipeline components...: 100%|███████████████████████████████| 5/5 [00:01<00:00,  4.86it/s]
Traceback (most recent call last):
  File "C:\aiOWN\diffuser_webui\easyanimate_bnb.py", line 30, in <module>
    video = pipeline(
  File "C:\aiOWN\diffuser_webui\venv\lib\site-packages\torch\utils\_contextlib.py", line 116, in decorate_context
    return func(*args, **kwargs)
  File "C:\aiOWN\diffuser_webui\venv\lib\site-packages\diffusers\pipelines\easyanimate\pipeline_easyanimate.py", line 803, in __call__
    ) = self.encode_prompt(
  File "C:\aiOWN\diffuser_webui\venv\lib\site-packages\diffusers\pipelines\easyanimate\pipeline_easyanimate.py", line 383, in encode_prompt
    prompt_embeds = text_encoder(
  File "C:\aiOWN\diffuser_webui\venv\lib\site-packages\torch\nn\modules\module.py", line 1736, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "C:\aiOWN\diffuser_webui\venv\lib\site-packages\torch\nn\modules\module.py", line 1747, in _call_impl
    return forward_call(*args, **kwargs)
  File "C:\aiOWN\diffuser_webui\venv\lib\site-packages\accelerate\hooks.py", line 176, in new_forward
    output = module._old_forward(*args, **kwargs)
  File "C:\aiOWN\diffuser_webui\venv\lib\site-packages\transformers\models\qwen2_vl\modeling_qwen2_vl.py", line 1082, in forward
    inputs_embeds = self.embed_tokens(input_ids)
  File "C:\aiOWN\diffuser_webui\venv\lib\site-packages\torch\nn\modules\module.py", line 1736, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "C:\aiOWN\diffuser_webui\venv\lib\site-packages\torch\nn\modules\module.py", line 1747, in _call_impl
    return forward_call(*args, **kwargs)
  File "C:\aiOWN\diffuser_webui\venv\lib\site-packages\accelerate\hooks.py", line 171, in new_forward
    args, kwargs = module._hf_hook.pre_forward(module, *args, **kwargs)
  File "C:\aiOWN\diffuser_webui\venv\lib\site-packages\accelerate\hooks.py", line 370, in pre_forward
    return send_to_device(args, self.execution_device), send_to_device(
  File "C:\aiOWN\diffuser_webui\venv\lib\site-packages\accelerate\utils\operations.py", line 174, in send_to_device
    return honor_type(
  File "C:\aiOWN\diffuser_webui\venv\lib\site-packages\accelerate\utils\operations.py", line 81, in honor_type
    return type(obj)(generator)
  File "C:\aiOWN\diffuser_webui\venv\lib\site-packages\accelerate\utils\operations.py", line 175, in <genexpr>
    tensor, (send_to_device(t, device, non_blocking=non_blocking, skip_keys=skip_keys) for t in tensor)
  File "C:\aiOWN\diffuser_webui\venv\lib\site-packages\accelerate\utils\operations.py", line 155, in send_to_device
    return tensor.to(device, non_blocking=non_blocking)
NotImplementedError: Cannot copy out of meta tensor; no data!

a-r-r-o-w · 2025-02-24T12:05:48Z

@bubbliiiing I've verified that the -diffusers suffixed checkpoints works with the T2V pipeline. I've refactored the VAE to mostly diffusers-expected format. Only the pipelines remain now, specifically the following changes:

Removing einops dependency
Creating a specific RoPE layer similar to how we do in HunyuanVideo and other latest model integrations
Removing calls to Image.open. The users will always have to pass a PIL.Image.Image, np.array or torch.Tensor themselves
Some other minimal cleanup

I can take these up when I find some time this week, but if you or someone from your team has the time to look into this, it would be very helpful too!

output.mp4

bubbliiiing · 2025-02-25T02:06:25Z

Sure, I'll do my best to make the necessary fixes. If there are any parts I haven't covered, please let me know and I'll be happy to help with the modifications. Thank you!

bubbliiiing · 2025-02-25T06:02:05Z

I have fixed some issues. Please take another look when you have time.

a-r-r-o-w

Thanks @bubbliiiing, the latest changes are great! @yiyixuxu Could you give this a look?

…easyanimate

src/diffusers/pipelines/easyanimate/pipeline_easyanimate.py

src/diffusers/pipelines/easyanimate/pipeline_easyanimate_control.py

…ol.py

HuggingFaceDocBuilderDev · 2025-02-27T02:40:51Z

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

yiyixuxu

awesome! thank you

yiyixuxu · 2025-02-28T23:25:00Z

we have some tests are still failing though,

bubbliiiing added 2 commits January 22, 2025 08:26

Update EasyAnimate V5.1

5609fc2

Merge branch 'main' of https://github.com/bubbliiiing/diffusers

4978561

a-r-r-o-w reviewed Jan 22, 2025

View reviewed changes

yiyixuxu added the roadmap Add to current release roadmap label Jan 22, 2025

bubbliiiing added 3 commits February 4, 2025 15:12

Add docs && add tests && Fix comments problems in transformer3d and vae

0b01118

delete comments and remove useless import

914f460

delete process

19fcc7d

Update EXAMPLE_DOC_STRING

a7821be

nitinmukesh mentioned this pull request Feb 12, 2025

Update 7b && Support low vram inference aigc-apps/EasyAnimate#196

Merged

a-r-r-o-w added 6 commits February 14, 2025 16:48

Merge branch 'main' into easyanimate

a6a5509

rename transformer file

6c0d81d

make fix-copies

414bf8f

make style

98602d8

refactor pt. 1

02f8c26

update toctree.yml

d5b3db9

yiyixuxu added the close-to-merge label Feb 24, 2025

bubbliiiing added 2 commits February 25, 2025 06:00

Fix problem in comments

528d97e

Merge branch 'main' of github.com:bubbliiiing/diffusers

2e1b4f5

a-r-r-o-w approved these changes Feb 27, 2025

View reviewed changes

a-r-r-o-w requested a review from yiyixuxu February 27, 2025 01:20

a-r-r-o-w added 3 commits February 27, 2025 02:23

Merge branch 'main' into easyanimate

9e8a249

Merge branch 'main' of https://github.com/bubbliiiing/diffusers into …

96ccfb5

…easyanimate

refactor tiling; remove einops dependency

58edc80

a-r-r-o-w reviewed Feb 27, 2025

View reviewed changes

src/diffusers/pipelines/easyanimate/pipeline_easyanimate.py Outdated Show resolved Hide resolved

src/diffusers/pipelines/easyanimate/pipeline_easyanimate_control.py Show resolved Hide resolved

src/diffusers/pipelines/easyanimate/pipeline_easyanimate_control.py Outdated Show resolved Hide resolved

a-r-r-o-w and others added 4 commits February 27, 2025 03:25

fix docs path

9f04fa1

make fix-copies

0451caf

Update src/diffusers/pipelines/easyanimate/pipeline_easyanimate_contr…

7f1b78d

…ol.py

update _toctree.yml

0059a35

yiyixuxu approved these changes Feb 28, 2025

View reviewed changes

yiyixuxu and others added 6 commits February 28, 2025 23:21

Merge branch 'main' into main

ca252f6

fix test

8b616d2

update

905cd50

update

5c7d8ab

update

b4e73ba

make fix-copies

e511864

shethaadit approved these changes Mar 3, 2025

View reviewed changes

DN6 and others added 3 commits March 3, 2025 12:44

Merge branch 'main' into main

fce2f9e

fix tests

856231c

Merge branch 'main' into main

f4c810a

a-r-r-o-w merged commit 5e3b7d2 into huggingface:main Mar 3, 2025
23 of 27 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add EasyAnimateV5.1 text-to-video, image-to-video, control-to-video generation model #10626

Add EasyAnimateV5.1 text-to-video, image-to-video, control-to-video generation model #10626

bubbliiiing commented Jan 22, 2025 •

edited

Loading

a-r-r-o-w left a comment

a-r-r-o-w Jan 22, 2025

bubbliiiing Jan 26, 2025

a-r-r-o-w Jan 27, 2025

bubbliiiing Feb 4, 2025

a-r-r-o-w Jan 22, 2025

bubbliiiing Jan 26, 2025

bubbliiiing Jan 26, 2025

a-r-r-o-w Jan 27, 2025

bubbliiiing commented Jan 23, 2025

a-r-r-o-w commented Jan 23, 2025

a-r-r-o-w commented Feb 6, 2025

bubbliiiing commented Feb 6, 2025

nitinmukesh commented Feb 10, 2025

bubbliiiing commented Feb 11, 2025

bubbliiiing commented Feb 11, 2025

nitinmukesh commented Feb 12, 2025

nitinmukesh commented Feb 12, 2025

a-r-r-o-w commented Feb 12, 2025

bubbliiiing commented Feb 12, 2025

nitinmukesh commented Feb 12, 2025

a-r-r-o-w commented Feb 12, 2025

nitinmukesh commented Feb 12, 2025 •

edited

Loading

nitinmukesh commented Feb 12, 2025

a-r-r-o-w commented Feb 24, 2025

bubbliiiing commented Feb 25, 2025

bubbliiiing commented Feb 25, 2025

a-r-r-o-w left a comment •

edited

Loading

HuggingFaceDocBuilderDev commented Feb 27, 2025

yiyixuxu left a comment

yiyixuxu commented Feb 28, 2025

Add EasyAnimateV5.1 text-to-video, image-to-video, control-to-video generation model #10626

Add EasyAnimateV5.1 text-to-video, image-to-video, control-to-video generation model #10626

Conversation

bubbliiiing commented Jan 22, 2025 • edited Loading

What does this PR do?

Before submitting

Who can review?

a-r-r-o-w left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

bubbliiiing commented Jan 23, 2025

a-r-r-o-w commented Jan 23, 2025

a-r-r-o-w commented Feb 6, 2025

bubbliiiing commented Feb 6, 2025

nitinmukesh commented Feb 10, 2025

bubbliiiing commented Feb 11, 2025

bubbliiiing commented Feb 11, 2025

nitinmukesh commented Feb 12, 2025

nitinmukesh commented Feb 12, 2025

a-r-r-o-w commented Feb 12, 2025

bubbliiiing commented Feb 12, 2025

nitinmukesh commented Feb 12, 2025

a-r-r-o-w commented Feb 12, 2025

nitinmukesh commented Feb 12, 2025 • edited Loading

nitinmukesh commented Feb 12, 2025

a-r-r-o-w commented Feb 24, 2025

bubbliiiing commented Feb 25, 2025

bubbliiiing commented Feb 25, 2025

a-r-r-o-w left a comment • edited Loading

Choose a reason for hiding this comment

HuggingFaceDocBuilderDev commented Feb 27, 2025

yiyixuxu left a comment

Choose a reason for hiding this comment

yiyixuxu commented Feb 28, 2025

bubbliiiing commented Jan 22, 2025 •

edited

Loading

nitinmukesh commented Feb 12, 2025 •

edited

Loading

a-r-r-o-w left a comment •

edited

Loading