Skip to content
/ cache-dit Public
forked from vipshop/cache-dit

An Unified Cache Acceleration Toolbox for DiTs: Qwen-Image, Qwen-Image-Edit, FLUX.1, Wan2.1, etc.♥️

License

View license
4 stars 7 forks Branches Tags Activity
Star
Notifications You must be signed in to change notification settings

xlite-dev/cache-dit

BranchesTags
 
 

Repository files navigation

🤗 CacheDiT: An Unified and Training-free Cache Acceleration
Toolbox for Diffusion Transformers

🔥Unified Cache APIs | DBCache | Hybrid TaylorSeer | Hybrid Cache CFG🔥

♥️ Cache Acceleration with One-line Code ~ ♥️

🔥News

📖Contents

⚙️Installation

You can install the stable release of cache-dit from PyPI:

pip3 install -U cache-dit

Or you can install the latest develop version from GitHub:

pip3 install git+https://github.com/vipshop/cache-dit.git

🔥Supported Models

Currently, cache-dit library supports almost Any Diffusion Transformers (with Transformer Blocks that match the specific Input and Output patterns). Please check 🎉Unified Cache APIs for more details. Here are just some of the tested models listed:

🎉Unified Cache APIs

Currently, for any Diffusion models with Transformer Blocks that match the specific Input/Output patterns, we can use the Unified Cache APIs from cache-dit, namely, the cache_dit.enable_cache(...) API. The supported patterns are listed as follows:

(IN: hidden_states, encoder_hidden_states, ...) -> (OUT: hidden_states, encoder_hidden_states)  
(IN: hidden_states, encoder_hidden_states, ...) -> (OUT: encoder_hidden_states, hidden_states)  
(IN: hidden_states, encoder_hidden_states, ...) -> (OUT: hidden_states)
(IN: hidden_states, ...) -> (OUT: hidden_states) # TODO, DiT, Lumina2, etc.

After the cache_dit.enable_cache(...) API is called, you just need to call the pipe as normal. The pipe param can be any Diffusion Pipeline. Please refer to Qwen-Image as an example. The Unified Cache APIs are currently in the experimental phase; please stay tuned for updates.

import cache_dit
from diffusers import DiffusionPipeline 

# can be any diffusion pipeline
pipe = DiffusionPipeline.from_pretrained("Qwen/Qwen-Image")

# one line code with default cache options.
cache_dit.enable_cache(pipe) 

# or, enable cache with custom settings.
cache_dit.enable_cache(
    pipe, transformer=pipe.transformer,
    blocks=pipe.transformer.transformer_blocks,
    return_hidden_states_first=False,
    **cache_dit.default_options(),
)

# just call the pipe as normal.
output = pipe(...)

# then, summary the cache stats.
stats = cache_dit.summary(pipe)

After finishing each inference of pipe(...), you can call the cache_dit.summary(...) API on pipe to get the details of the cache stats for the current inference (markdown table format). You can set details param as True to show more details of cache stats.

⚡️Cache Steps and Residual Diffs Statistics: QwenImagePipeline

| Cache Steps | Diffs P00 | Diffs P25 | Diffs P50 | Diffs P75 | Diffs P95 |
|-------------|-----------|-----------|-----------|-----------|-----------|
| 23          | 0.04      | 0.082     | 0.115     | 0.152     | 0.245     |
...

⚡️DBCache: Dual Block Cache

DBCache: Dual Block Caching for Diffusion Transformers. We have enhanced FBCache into a more general and customizable cache algorithm, namely DBCache, enabling it to achieve fully UNet-style cache acceleration for DiT models. Different configurations of compute blocks (F8B12, etc.) can be customized in DBCache. Moreover, it can be entirely training-free. DBCache can strike a perfect balance between performance and precision!

DBCache, L20x1 , Steps: 28, "A cat holding a sign that says hello world with complex background"

Baseline(L20x1) F1B0 (0.08) F1B0 (0.20) F8B8 (0.15) F12B12 (0.20) F16B16 (0.20)
24.85s 15.59s 8.58s 15.41s 15.11s 17.74s
Baseline(L20x1) F1B0 (0.08) F8B8 (0.12) F8B12 (0.12) F8B16 (0.20) F8B20 (0.20)
27.85s 6.04s 5.88s 5.77s 6.01s 6.20s

DBCache, L20x4 , Steps: 20, case to show the texture recovery ability of DBCache

These case studies demonstrate that even with relatively high thresholds (such as 0.12, 0.15, 0.2, etc.) under the DBCache F12B12 or F8B16 configuration, the detailed texture of the kitten's fur, colored cloth, and the clarity of text can still be preserved. This suggests that users can leverage DBCache to effectively balance performance and precision in their workflows!

DBCache provides configurable parameters for custom optimization, enabling a balanced trade-off between performance and precision:

  • Fn: Specifies that DBCache uses the first n Transformer blocks to fit the information at time step t, enabling the calculation of a more stable L1 diff and delivering more accurate information to subsequent blocks.
  • Bn: Further fuses approximate information in the last n Transformer blocks to enhance prediction accuracy. These blocks act as an auto-scaler for approximate hidden states that use residual cache.

  • warmup_steps: (default: 0) DBCache does not apply the caching strategy when the number of running steps is less than or equal to this value, ensuring the model sufficiently learns basic features during warmup.
  • max_cached_steps: (default: -1) DBCache disables the caching strategy when the previous cached steps exceed this value to prevent precision degradation.
  • residual_diff_threshold: The value of residual diff threshold, a higher value leads to faster performance at the cost of lower precision.

For a good balance between performance and precision, DBCache is configured by default with F8B0, 8 warmup steps, and unlimited cached steps.

import cache_dit
from diffusers import FluxPipeline

pipe = FluxPipeline.from_pretrained(
    "black-forest-labs/FLUX.1-dev",
    torch_dtype=torch.bfloat16,
).to("cuda")

# Default options, F8B0, good balance between performance and precision
cache_options = cache_dit.default_options()

# Custom options, F8B8, higher precision
cache_options = {
    "cache_type": cache_dit.DBCache,
    "warmup_steps": 8,
    "max_cached_steps": -1, # -1 means no limit
    "Fn_compute_blocks": 8, # Fn, F8, etc.
    "Bn_compute_blocks": 8, # Bn, B8, etc.
    "residual_diff_threshold": 0.12,
}

cache_dit.enable_cache(pipe, **cache_options)

Moreover, users configuring higher Bn values (e.g., F8B16) while aiming to maintain good performance can specify Bn_compute_blocks_ids to work with Bn. DBCache will only compute the specified blocks, with the remaining estimated using the previous step's residual cache.

# Custom options, F8B16, higher precision with good performance.
cache_options = {
    # 0, 2, 4, ..., 14, 15, etc. [0,16)
    "Bn_compute_blocks_ids": cache_dit.block_range(0, 16, 2),
    # If the L1 difference is below this threshold, skip Bn blocks 
    # not in `Bn_compute_blocks_ids`(1, 3,..., etc), Otherwise, 
    # compute these blocks.
    "non_compute_blocks_diff_threshold": 0.08,
}

🔥Hybrid TaylorSeer

We have supported the TaylorSeers: From Reusing to Forecasting: Accelerating Diffusion Models with TaylorSeers algorithm to further improve the precision of DBCache in cases where the cached steps are large, namely, Hybrid TaylorSeer + DBCache. At timesteps with significant intervals, the feature similarity in diffusion models decreases substantially, significantly harming the generation quality.

$$ \mathcal{F}_{\text {pred }, m}\left(x_{t-k}^l\right)=\mathcal{F}\left(x_t^l\right)+\sum_{i=1}^m \frac{\Delta^i \mathcal{F}\left(x_t^l\right)}{i!\cdot N^i}(-k)^i $$

TaylorSeer employs a differential method to approximate the higher-order derivatives of features and predict features in future timesteps with Taylor series expansion. The TaylorSeer implemented in cache-dit supports both hidden states and residual cache types. That is $\mathcal{F}_{\text {pred }, m}\left(x_{t-k}^l\right)$ can be a residual cache or a hidden-state cache.

cache_options = {
    # TaylorSeer options
    "enable_taylorseer": True,
    "enable_encoder_taylorseer": True,
    # Taylorseer cache type cache be hidden_states or residual.
    "taylorseer_cache_type": "residual",
    # Higher values of n_derivatives will lead to longer 
    # computation time but may improve precision significantly.
    "taylorseer_kwargs": {
        "n_derivatives": 2, # default is 2.
    },
    "warmup_steps": 3, # prefer: >= n_derivatives + 1
    "residual_diff_threshold": 0.12,
}

Important

Please note that if you have used TaylorSeer as the calibrator for approximate hidden states, the Bn param of DBCache can be set to 0. In essence, DBCache's Bn is also act as a calibrator, so you can choose either Bn > 0 or TaylorSeer. We recommend using the configuration scheme of TaylorSeer + DBCache FnB0.

DBCache F1B0 + TaylorSeer, L20x1, Steps: 28,
"A cat holding a sign that says hello world with complex background"

Baseline(L20x1) F1B0 (0.12) +TaylorSeer F1B0 (0.15) +TaylorSeer +compile
24.85s 12.85s 12.86s 10.27s 10.28s 8.48s

⚡️Hybrid Cache CFG

cache-dit supports caching for CFG (classifier-free guidance). For models that fuse CFG and non-CFG into a single forward step, or models that do not include CFG (classifier-free guidance) in the forward step, please set do_separate_classifier_free_guidance param to False (default). Otherwise, set it to True. For examples:

cache_options = {
    # CFG: classifier free guidance or not
    # For model that fused CFG and non-CFG into single forward step,
    # should set do_separate_classifier_free_guidance as False.
    # For example, set it as True for Wan 2.1 and set it as False 
    # for FLUX.1, HunyuanVideo, CogVideoX, Mochi.
    "do_separate_classifier_free_guidance": True, # Wan 2.1, Qwen-Image
    # Compute cfg forward first or not, default False, namely, 
    # 0, 2, 4, ..., -> non-CFG step; 1, 3, 5, ... -> CFG step.
    "cfg_compute_first": False,
    # Compute spearate diff values for CFG and non-CFG step, 
    # default True. If False, we will use the computed diff from 
    # current non-CFG transformer step for current CFG step.
    "cfg_diff_compute_separate": True,
}

🔥Torch Compile

By the way, cache-dit is designed to work compatibly with torch.compile. You can easily use cache-dit with torch.compile to further achieve a better performance. For example:

cache_dit.enable_cache(
    pipe, **cache_dit.default_options()
)
# Compile the Transformer module
pipe.transformer = torch.compile(pipe.transformer)

However, users intending to use cache-dit for DiT with dynamic input shapes should consider increasing the recompile limit of torch._dynamo. Otherwise, the recompile_limit error may be triggered, causing the module to fall back to eager mode.

torch._dynamo.config.recompile_limit = 96  # default is 8
torch._dynamo.config.accumulated_recompile_limit = 2048  # default is 256

Please check bench.py for more details.

🛠Metrics CLI

You can utilize the APIs provided by cache-dit to quickly evaluate the accuracy losses caused by different cache configurations. For example:

from cache_dit.metrics import compute_psnr
from cache_dit.metrics import compute_video_psnr
from cache_dit.metrics import FrechetInceptionDistance  # FID

FID = FrechetInceptionDistance()
image_psnr, n = compute_psnr("true.png", "test.png") # Num: n
image_fid,  n = FID.compute_fid("true_dir", "test_dir")
video_psnr, n = compute_video_psnr("true.mp4", "test.mp4") # Frames: n

Please check test_metrics.py for more details. Or, you can use cache-dit-metrics-cli tool. For examples:

cache-dit-metrics-cli -h  # show usage
# all: PSNR, FID, SSIM, MSE, ..., etc.
cache-dit-metrics-cli all  -i1 true.png -i2 test.png  # image
cache-dit-metrics-cli all  -i1 true_dir -i2 test_dir  # image dir
cache-dit-metrics-cli all  -v1 true.mp4 -v2 test.mp4  # video
cache-dit-metrics-cli all  -v1 true_dir -v2 test_dir  # video dir
cache-dit-metrics-cli fid  -i1 true_dir -i2 test_dir  # FID
cache-dit-metrics-cli psnr -i1 true_dir -i2 test_dir  # PSNR

👋Contribute

How to contribute? Star ⭐️ this repo to support us or check CONTRIBUTE.md.

©️License

The cache-dit codebase is adapted from FBCache. Special thanks to their excellent work! We have followed the original License from FBCache, please check LICENSE for more details.

©️Citations

@misc{cache-dit@2025,
  title={cache-dit: An Unified and Training-free Cache Acceleration Toolbox for Diffusion Transformers},
  url={https://github.com/vipshop/cache-dit.git},
  note={Open-source software available at https://github.com/vipshop/cache-dit.git},
  author={vipshop.com},
  year={2025}
}

About

An Unified Cache Acceleration Toolbox for DiTs: Qwen-Image, Qwen-Image-Edit, FLUX.1, Wan2.1, etc.♥️

Resources

Readme

License

View license
Activity
Custom properties

Stars

4 stars

Watchers

0 watching

Forks

0 forks
Report repository

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%