lucataco/animate-diff-sdxl-lcm | Run with an API on Replicate

lucataco / animate-diff-sdxl-lcm

Animate Your Personalized Text-to-Image Diffusion Models with SDXL and LCM (Updated 1 year, 7 months ago)

Cold

Public
332 runs
A100 (80GB)
Paper
License

Iterate in playground

Run with an API

Playground API Examples README Versions

Input

motion_module

string

Select a Motion Model (currently only one available)

Default: "mm_sdxl_v10_beta"

checkpoint

string

Select a model checkpoint

Default: "dynavision"

use_checkpoint

boolean

Default: false

aspect

string

Aspect ratio

Default: "1:1"

video_length

integer

(minimum: 16)

Video length

Default: 16

prompt

string

Shift + Return to add a new line

A panda standing on a surfboard in the ocean in sunset, 4k, high resolution. Realistic, Cinematic, high resolutionA panda standing on a surfboard in the ocean in sunset, 4k, high resolution. Realistic, Cinematic, high resolution

Input prompt

Default: "A panda standing on a surfboard in the ocean in sunset, 4k, high resolution. Realistic, Cinematic, high resolution"

n_prompt

string

Shift + Return to add a new line

Negative prompt

Default: ""

steps

integer

(minimum: 1, maximum: 100)

Number of inference steps

Default: 6

guidance_scale

number

(minimum: 1, maximum: 10)

guidance scale

Default: 1

seed

integer

(minimum: 0, maximum: 2147483647)

Seed (0 = random, maximum: 2147483647)

mp4

boolean

Returns .mp4 if true or .gif if false

Default: true

Run this model in Node.js with one line of code:

npx create-replicate --model=lucataco/animate-diff-sdxl-lcm

or set up a project from scratch

Install Replicate’s Node.js client library:

npm install replicate

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Import and set up the client:

import Replicate from "replicate";
import fs from "node:fs";

const replicate = new Replicate({
  auth: process.env.REPLICATE_API_TOKEN,
});

Run lucataco/animate-diff-sdxl-lcm using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "lucataco/animate-diff-sdxl-lcm:64fe561273282266ba4b51fd84e89603412c088882d1f1c325d875c09b3816af",
  {
    input: {
      mp4: true,
      seed: 0,
      steps: 6,
      aspect: "1:1",
      prompt: "A panda standing on a surfboard in the ocean in sunset, 4k, high resolution. Realistic, Cinematic, high resolution",
      n_prompt: "",
      checkpoint: "dynavision",
      video_length: 16,
      motion_module: "mm_sdxl_v10_beta",
      guidance_scale: 1,
      use_checkpoint: false
    }
  }
);

// To access the file URL:
console.log(output.url()); //=> "http://example.com"

// To write the file to disk:
fs.writeFile("my-image.png", output);

To learn more, take a look at the guide on getting started with Node.js.

Install Replicate’s Python client library:

pip install replicate

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Import the client:

import replicate

Run lucataco/animate-diff-sdxl-lcm using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "lucataco/animate-diff-sdxl-lcm:64fe561273282266ba4b51fd84e89603412c088882d1f1c325d875c09b3816af",
    input={
        "mp4": True,
        "seed": 0,
        "steps": 6,
        "aspect": "1:1",
        "prompt": "A panda standing on a surfboard in the ocean in sunset, 4k, high resolution. Realistic, Cinematic, high resolution",
        "n_prompt": "",
        "checkpoint": "dynavision",
        "video_length": 16,
        "motion_module": "mm_sdxl_v10_beta",
        "guidance_scale": 1,
        "use_checkpoint": False
    }
)
print(output)

To learn more, take a look at the guide on getting started with Python.

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Run lucataco/animate-diff-sdxl-lcm using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

curl -s -X POST \
  -H "Authorization: Bearer $REPLICATE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -H "Prefer: wait" \
  -d $'{
    "version": "lucataco/animate-diff-sdxl-lcm:64fe561273282266ba4b51fd84e89603412c088882d1f1c325d875c09b3816af",
    "input": {
      "mp4": true,
      "seed": 0,
      "steps": 6,
      "aspect": "1:1",
      "prompt": "A panda standing on a surfboard in the ocean in sunset, 4k, high resolution. Realistic, Cinematic, high resolution",
      "n_prompt": "",
      "checkpoint": "dynavision",
      "video_length": 16,
      "motion_module": "mm_sdxl_v10_beta",
      "guidance_scale": 1,
      "use_checkpoint": false
    }
  }' \
  https://api.replicate.com/v1/predictions

To learn more, take a look at Replicate’s HTTP API reference docs.

Output

{
  "completed_at": "2023-11-18T06:25:43.139643Z",
  "created_at": "2023-11-18T06:18:13.775148Z",
  "data_removed": false,
  "error": null,
  "id": "yivsjvlbjp3n4x3tkdjyvinflm",
  "input": {
    "mp4": true,
    "seed": 0,
    "steps": 6,
    "aspect": "1:1",
    "prompt": "A panda standing on a surfboard in the ocean in sunset, 4k, high resolution. Realistic, Cinematic, high resolution",
    "n_prompt": "",
    "checkpoint": "dynavision",
    "video_length": 16,
    "motion_module": "mm_sdxl_v10_beta",
    "guidance_scale": 1,
    "use_checkpoint": false
  },
  "logs": "loaded temporal unet's pretrained weights from /AnimateDiff/models/StableDiffusion/unet ...\nThe config attributes {'dropout': 0.0, 'reverse_transformer_layers_per_block': None} were passed to UNet3DConditionModel, but are not expected and will be ignored. Please verify your config.json configuration file.\n### missing keys: 420;\n### unexpected keys: 0;\n### Temporal Module Parameters: 236.7792 M\nLoading motion module from /AnimateDiff/models/Motion_Module/mm_sdxl_v10_beta.ckpt...\nUsing seed: 0\nsampling: A panda standing on a surfboard in the ocean in sunset, 4k, high resolution. Realistic, Cinematic, high resolution ...\n  0%|          | 0/6 [00:00<?, ?it/s]\n 17%|█▋        | 1/6 [00:03<00:18,  3.74s/it]\n 33%|███▎      | 2/6 [00:05<00:09,  2.45s/it]\n 50%|█████     | 3/6 [00:06<00:06,  2.04s/it]\n 67%|██████▋   | 4/6 [00:08<00:03,  1.85s/it]\n 83%|████████▎ | 5/6 [00:09<00:01,  1.74s/it]\n100%|██████████| 6/6 [00:11<00:00,  1.68s/it]\n100%|██████████| 6/6 [00:11<00:00,  1.92s/it]\nffmpeg version 4.4.2-0ubuntu0.22.04.1 Copyright (c) 2000-2021 the FFmpeg developers\nbuilt with gcc 11 (Ubuntu 11.2.0-19ubuntu1)\nconfiguration: --prefix=/usr --extra-version=0ubuntu0.22.04.1 --toolchain=hardened --libdir=/usr/lib/x86_64-linux-gnu --incdir=/usr/include/x86_64-linux-gnu --arch=amd64 --enable-gpl --disable-stripping --enable-gnutls --enable-ladspa --enable-libaom --enable-libass --enable-libbluray --enable-libbs2b --enable-libcaca --enable-libcdio --enable-libcodec2 --enable-libdav1d --enable-libflite --enable-libfontconfig --enable-libfreetype --enable-libfribidi --enable-libgme --enable-libgsm --enable-libjack --enable-libmp3lame --enable-libmysofa --enable-libopenjpeg --enable-libopenmpt --enable-libopus --enable-libpulse --enable-librabbitmq --enable-librubberband --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libsrt --enable-libssh --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvorbis --enable-libvpx --enable-libwebp --enable-libx265 --enable-libxml2 --enable-libxvid --enable-libzimg --enable-libzmq --enable-libzvbi --enable-lv2 --enable-omx --enable-openal --enable-opencl --enable-opengl --enable-sdl2 --enable-pocketsphinx --enable-librsvg --enable-libmfx --enable-libdc1394 --enable-libdrm --enable-libiec61883 --enable-chromaprint --enable-frei0r --enable-libx264 --enable-shared\nlibavutil      56. 70.100 / 56. 70.100\nlibavcodec     58.134.100 / 58.134.100\nlibavformat    58. 76.100 / 58. 76.100\nlibavdevice    58. 13.100 / 58. 13.100\nlibavfilter     7.110.100 /  7.110.100\nlibswscale      5.  9.100 /  5.  9.100\nlibswresample   3.  9.100 /  3.  9.100\nlibpostproc    55.  9.100 / 55.  9.100\nInput #0, image2, from '/tmp/tmpp1m3aene/*.png':\nDuration: 00:00:00.64, start: 0.000000, bitrate: N/A\nStream #0:0: Video: png, rgb24(pc), 1024x1024, 25 fps, 25 tbr, 25 tbn, 25 tbc\nStream mapping:\nStream #0:0 -> #0:0 (png (native) -> h264 (libx264))\nPress [q] to stop, [?] for help\n[libx264 @ 0x5590b3e4e5c0] using cpu capabilities: MMX2 SSE2Fast SSSE3 SSE4.2 AVX FMA3 BMI2 AVX2\n[libx264 @ 0x5590b3e4e5c0] profile High, level 3.2, 4:2:0, 8-bit\n[libx264 @ 0x5590b3e4e5c0] 264 - core 163 r3060 5db6aa6 - H.264/MPEG-4 AVC codec - Copyleft 2003-2021 - http://www.videolan.org/x264.html - options: cabac=1 ref=3 deblock=1:0:0 analyse=0x3:0x113 me=hex subme=7 psy=1 psy_rd=1.00:0.00 mixed_ref=1 me_range=16 chroma_me=1 trellis=1 8x8dct=1 cqm=0 deadzone=21,11 fast_pskip=1 chroma_qp_offset=-2 threads=15 lookahead_threads=2 sliced_threads=0 nr=0 decimate=1 interlaced=0 bluray_compat=0 constrained_intra=0 bframes=3 b_pyramid=2 b_adapt=1 b_bias=0 direct=1 weightb=1 open_gop=0 weightp=2 keyint=250 keyint_min=25 scenecut=40 intra_refresh=0 rc=cqp mbtree=0 qp=17 ip_ratio=1.40 pb_ratio=1.30 aq=0\nOutput #0, mp4, to '/tmp/tmpp1m3aene/out.mp4':\nMetadata:\nencoder         : Lavf58.76.100\nStream #0:0: Video: h264 (avc1 / 0x31637661), yuv420p(tv, progressive), 1024x1024, q=2-31, 25 fps, 12800 tbn\nMetadata:\nencoder         : Lavc58.134.100 libx264\nSide data:\ncpb: bitrate max/min/avg: 0/0/0 buffer size: 0 vbv_delay: N/A\nframe=    1 fps=0.0 q=0.0 size=       0kB time=00:00:00.00 bitrate=N/A speed=   0x\n[mp4 @ 0x5590b3e49d40] Starting second pass: moving the moov atom to the beginning of the file\nframe=   16 fps=0.0 q=-1.0 Lsize=    1058kB time=00:00:00.52 bitrate=16666.7kbits/s speed=1.16x\nvideo:1057kB audio:0kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 0.097927%\n[libx264 @ 0x5590b3e4e5c0] frame I:1     Avg QP:14.00  size: 90593\n[libx264 @ 0x5590b3e4e5c0] frame P:8     Avg QP:17.00  size: 70904\n[libx264 @ 0x5590b3e4e5c0] frame B:7     Avg QP:18.86  size: 60567\n[libx264 @ 0x5590b3e4e5c0] consecutive B-frames: 25.0% 50.0%  0.0% 25.0%\n[libx264 @ 0x5590b3e4e5c0] mb I  I16..4: 22.6% 60.6% 16.8%\n[libx264 @ 0x5590b3e4e5c0] mb P  I16..4: 19.8% 41.7%  8.0%  P16..4: 20.0%  6.8%  2.9%  0.0%  0.0%    skip: 0.8%\n[libx264 @ 0x5590b3e4e5c0] mb B  I16..4:  6.6% 27.6%  6.6%  B16..8: 25.7% 12.8%  2.7%  direct:12.2%  skip: 5.9%  L0:44.9% L1:43.1% BI:12.0%\n[libx264 @ 0x5590b3e4e5c0] 8x8 transform intra:62.4% inter:89.2%\n[libx264 @ 0x5590b3e4e5c0] coded y,uvDC,uvAC intra: 69.9% 92.2% 70.9% inter: 41.1% 85.5% 13.7%\n[libx264 @ 0x5590b3e4e5c0] i16 v,h,dc,p:  3% 21%  2% 74%\n[libx264 @ 0x5590b3e4e5c0] i8 v,h,dc,ddl,ddr,vr,hd,vl,hu:  4% 74%  6%  2%  2%  2%  5%  2%  3%\n[libx264 @ 0x5590b3e4e5c0] i4 v,h,dc,ddl,ddr,vr,hd,vl,hu:  8% 61%  4%  2%  4%  4% 10%  3%  4%\n[libx264 @ 0x5590b3e4e5c0] i8c dc,h,v,p: 19% 61%  6% 14%\n[libx264 @ 0x5590b3e4e5c0] Weighted P-Frames: Y:25.0% UV:25.0%\n[libx264 @ 0x5590b3e4e5c0] ref P L0: 58.6%  5.0% 20.5% 11.6%  4.4%\n[libx264 @ 0x5590b3e4e5c0] ref B L0: 78.5% 18.3%  3.2%\n[libx264 @ 0x5590b3e4e5c0] ref B L1: 97.2%  2.8%\n[libx264 @ 0x5590b3e4e5c0] kb/s:13522.41",
  "metrics": {
    "predict_time": 59.312788,
    "total_time": 449.364495
  },
  "output": "https://replicate.delivery/pbxt/jZ1zUMfvHVSEJqopozeT2t9JKZfvWKyf7glocSG0f3cycFMPC/out.mp4",
  "started_at": "2023-11-18T06:24:43.826855Z",
  "status": "succeeded",
  "urls": {
    "get": "https://api.replicate.com/v1/predictions/yivsjvlbjp3n4x3tkdjyvinflm",
    "cancel": "https://api.replicate.com/v1/predictions/yivsjvlbjp3n4x3tkdjyvinflm/cancel"
  },
  "version": "64fe561273282266ba4b51fd84e89603412c088882d1f1c325d875c09b3816af"
}

Generated in

59.3 seconds

Tweak itReport View full prediction

loaded temporal unet's pretrained weights from /AnimateDiff/models/StableDiffusion/unet ...
The config attributes {'dropout': 0.0, 'reverse_transformer_layers_per_block': None} were passed to UNet3DConditionModel, but are not expected and will be ignored. Please verify your config.json configuration file.
### missing keys: 420;
### unexpected keys: 0;
### Temporal Module Parameters: 236.7792 M
Loading motion module from /AnimateDiff/models/Motion_Module/mm_sdxl_v10_beta.ckpt...
Using seed: 0
sampling: A panda standing on a surfboard in the ocean in sunset, 4k, high resolution. Realistic, Cinematic, high resolution ...
  0%|          | 0/6 [00:00<?, ?it/s]
 17%|█▋        | 1/6 [00:03<00:18,  3.74s/it]
 33%|███▎      | 2/6 [00:05<00:09,  2.45s/it]
 50%|█████     | 3/6 [00:06<00:06,  2.04s/it]
 67%|██████▋   | 4/6 [00:08<00:03,  1.85s/it]
 83%|████████▎ | 5/6 [00:09<00:01,  1.74s/it]
100%|██████████| 6/6 [00:11<00:00,  1.68s/it]
100%|██████████| 6/6 [00:11<00:00,  1.92s/it]
ffmpeg version 4.4.2-0ubuntu0.22.04.1 Copyright (c) 2000-2021 the FFmpeg developers
built with gcc 11 (Ubuntu 11.2.0-19ubuntu1)
configuration: --prefix=/usr --extra-version=0ubuntu0.22.04.1 --toolchain=hardened --libdir=/usr/lib/x86_64-linux-gnu --incdir=/usr/include/x86_64-linux-gnu --arch=amd64 --enable-gpl --disable-stripping --enable-gnutls --enable-ladspa --enable-libaom --enable-libass --enable-libbluray --enable-libbs2b --enable-libcaca --enable-libcdio --enable-libcodec2 --enable-libdav1d --enable-libflite --enable-libfontconfig --enable-libfreetype --enable-libfribidi --enable-libgme --enable-libgsm --enable-libjack --enable-libmp3lame --enable-libmysofa --enable-libopenjpeg --enable-libopenmpt --enable-libopus --enable-libpulse --enable-librabbitmq --enable-librubberband --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libsrt --enable-libssh --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvorbis --enable-libvpx --enable-libwebp --enable-libx265 --enable-libxml2 --enable-libxvid --enable-libzimg --enable-libzmq --enable-libzvbi --enable-lv2 --enable-omx --enable-openal --enable-opencl --enable-opengl --enable-sdl2 --enable-pocketsphinx --enable-librsvg --enable-libmfx --enable-libdc1394 --enable-libdrm --enable-libiec61883 --enable-chromaprint --enable-frei0r --enable-libx264 --enable-shared
libavutil      56. 70.100 / 56. 70.100
libavcodec     58.134.100 / 58.134.100
libavformat    58. 76.100 / 58. 76.100
libavdevice    58. 13.100 / 58. 13.100
libavfilter     7.110.100 /  7.110.100
libswscale      5.  9.100 /  5.  9.100
libswresample   3.  9.100 /  3.  9.100
libpostproc    55.  9.100 / 55.  9.100
Input #0, image2, from '/tmp/tmpp1m3aene/*.png':
Duration: 00:00:00.64, start: 0.000000, bitrate: N/A
Stream #0:0: Video: png, rgb24(pc), 1024x1024, 25 fps, 25 tbr, 25 tbn, 25 tbc
Stream mapping:
Stream #0:0 -> #0:0 (png (native) -> h264 (libx264))
Press [q] to stop, [?] for help
[libx264 @ 0x5590b3e4e5c0] using cpu capabilities: MMX2 SSE2Fast SSSE3 SSE4.2 AVX FMA3 BMI2 AVX2
[libx264 @ 0x5590b3e4e5c0] profile High, level 3.2, 4:2:0, 8-bit
[libx264 @ 0x5590b3e4e5c0] 264 - core 163 r3060 5db6aa6 - H.264/MPEG-4 AVC codec - Copyleft 2003-2021 - http://www.videolan.org/x264.html - options: cabac=1 ref=3 deblock=1:0:0 analyse=0x3:0x113 me=hex subme=7 psy=1 psy_rd=1.00:0.00 mixed_ref=1 me_range=16 chroma_me=1 trellis=1 8x8dct=1 cqm=0 deadzone=21,11 fast_pskip=1 chroma_qp_offset=-2 threads=15 lookahead_threads=2 sliced_threads=0 nr=0 decimate=1 interlaced=0 bluray_compat=0 constrained_intra=0 bframes=3 b_pyramid=2 b_adapt=1 b_bias=0 direct=1 weightb=1 open_gop=0 weightp=2 keyint=250 keyint_min=25 scenecut=40 intra_refresh=0 rc=cqp mbtree=0 qp=17 ip_ratio=1.40 pb_ratio=1.30 aq=0
Output #0, mp4, to '/tmp/tmpp1m3aene/out.mp4':
Metadata:
encoder         : Lavf58.76.100
Stream #0:0: Video: h264 (avc1 / 0x31637661), yuv420p(tv, progressive), 1024x1024, q=2-31, 25 fps, 12800 tbn
Metadata:
encoder         : Lavc58.134.100 libx264
Side data:
cpb: bitrate max/min/avg: 0/0/0 buffer size: 0 vbv_delay: N/A
frame=    1 fps=0.0 q=0.0 size=       0kB time=00:00:00.00 bitrate=N/A speed=   0x
[mp4 @ 0x5590b3e49d40] Starting second pass: moving the moov atom to the beginning of the file
frame=   16 fps=0.0 q=-1.0 Lsize=    1058kB time=00:00:00.52 bitrate=16666.7kbits/s speed=1.16x
video:1057kB audio:0kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 0.097927%
[libx264 @ 0x5590b3e4e5c0] frame I:1     Avg QP:14.00  size: 90593
[libx264 @ 0x5590b3e4e5c0] frame P:8     Avg QP:17.00  size: 70904
[libx264 @ 0x5590b3e4e5c0] frame B:7     Avg QP:18.86  size: 60567
[libx264 @ 0x5590b3e4e5c0] consecutive B-frames: 25.0% 50.0%  0.0% 25.0%
[libx264 @ 0x5590b3e4e5c0] mb I  I16..4: 22.6% 60.6% 16.8%
[libx264 @ 0x5590b3e4e5c0] mb P  I16..4: 19.8% 41.7%  8.0%  P16..4: 20.0%  6.8%  2.9%  0.0%  0.0%    skip: 0.8%
[libx264 @ 0x5590b3e4e5c0] mb B  I16..4:  6.6% 27.6%  6.6%  B16..8: 25.7% 12.8%  2.7%  direct:12.2%  skip: 5.9%  L0:44.9% L1:43.1% BI:12.0%
[libx264 @ 0x5590b3e4e5c0] 8x8 transform intra:62.4% inter:89.2%
[libx264 @ 0x5590b3e4e5c0] coded y,uvDC,uvAC intra: 69.9% 92.2% 70.9% inter: 41.1% 85.5% 13.7%
[libx264 @ 0x5590b3e4e5c0] i16 v,h,dc,p:  3% 21%  2% 74%
[libx264 @ 0x5590b3e4e5c0] i8 v,h,dc,ddl,ddr,vr,hd,vl,hu:  4% 74%  6%  2%  2%  2%  5%  2%  3%
[libx264 @ 0x5590b3e4e5c0] i4 v,h,dc,ddl,ddr,vr,hd,vl,hu:  8% 61%  4%  2%  4%  4% 10%  3%  4%
[libx264 @ 0x5590b3e4e5c0] i8c dc,h,v,p: 19% 61%  6% 14%
[libx264 @ 0x5590b3e4e5c0] Weighted P-Frames: Y:25.0% UV:25.0%
[libx264 @ 0x5590b3e4e5c0] ref P L0: 58.6%  5.0% 20.5% 11.6%  4.4%
[libx264 @ 0x5590b3e4e5c0] ref B L0: 78.5% 18.3%  3.2%
[libx264 @ 0x5590b3e4e5c0] ref B L1: 97.2%  2.8%
[libx264 @ 0x5590b3e4e5c0] kb/s:13522.41

Examples

View more examples

Run time and cost

This model costs approximately $0.40 to run on Replicate, or 2 runs per $1, but this varies depending on your inputs. It is also open source and you can run it on your own computer with Docker.

This model runs on Nvidia A100 (80GB) GPU hardware. Predictions typically complete within 5 minutes. The predict time for this model varies significantly based on the inputs.

Readme

https://github.com/guoyww/AnimateDiff/tree/sdxl/

A beta-version of motion module for SDXL Now you can generate high-resolution videos on SDXL with/without personalized models. Checkpoint with better quality would be available soon. Stay tuned.