thudm / cogvideox-i2v

Image-to-Video Diffusion Models with An Expert Transformer

Cold

Public
910 runs
L40S
GitHub
Weights
Paper
License

Run with an API

Playground API Examples README Versions

Input

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

npx create-replicate --model=thudm/cogvideox-i2v

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";

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

Run thudm/cogvideox-i2v using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "thudm/cogvideox-i2v:82caaa79520e03b3963c975e38dd68ac8a6b18a8a39c19fd8dc03b4ed4b91c58",
  {
    input: {
      image: "https://replicate.delivery/pbxt/Lf97CMO0Sz0sZ0IuQarZRT8TbcMz4pCurtiLSKWDBPSTMb1S/input.jpg",
      prompt: "Starry sky slowly rotating.",
      num_frames: 49,
      guidance_scale: 6,
      num_inference_steps: 50
    }
  }
);

// 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 thudm/cogvideox-i2v using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "thudm/cogvideox-i2v:82caaa79520e03b3963c975e38dd68ac8a6b18a8a39c19fd8dc03b4ed4b91c58",
    input={
        "image": "https://replicate.delivery/pbxt/Lf97CMO0Sz0sZ0IuQarZRT8TbcMz4pCurtiLSKWDBPSTMb1S/input.jpg",
        "prompt": "Starry sky slowly rotating.",
        "num_frames": 49,
        "guidance_scale": 6,
        "num_inference_steps": 50
    }
)
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 thudm/cogvideox-i2v 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": "thudm/cogvideox-i2v:82caaa79520e03b3963c975e38dd68ac8a6b18a8a39c19fd8dc03b4ed4b91c58",
    "input": {
      "image": "https://replicate.delivery/pbxt/Lf97CMO0Sz0sZ0IuQarZRT8TbcMz4pCurtiLSKWDBPSTMb1S/input.jpg",
      "prompt": "Starry sky slowly rotating.",
      "num_frames": 49,
      "guidance_scale": 6,
      "num_inference_steps": 50
    }
  }' \
  https://api.replicate.com/v1/predictions

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

Output

{
  "completed_at": "2024-09-21T16:31:24.720449Z",
  "created_at": "2024-09-21T16:22:09.382000Z",
  "data_removed": false,
  "error": null,
  "id": "qb7hex84wsrgj0cj2nj95qk4y4",
  "input": {
    "image": "https://replicate.delivery/pbxt/Lf97CMO0Sz0sZ0IuQarZRT8TbcMz4pCurtiLSKWDBPSTMb1S/input.jpg",
    "prompt": "Starry sky slowly rotating.",
    "num_frames": 49,
    "guidance_scale": 6,
    "num_inference_steps": 50
  },
  "logs": "Using seed: 10643\n  0%|          | 0/50 [00:00<?, ?it/s]\n  2%|▏         | 1/50 [00:12<10:16, 12.59s/it]\n  4%|▍         | 2/50 [00:20<07:41,  9.62s/it]\n  6%|▌         | 3/50 [00:27<06:47,  8.66s/it]\n  8%|▊         | 4/50 [00:35<06:18,  8.22s/it]\n 10%|█         | 5/50 [00:42<05:59,  7.99s/it]\n 12%|█▏        | 6/50 [00:50<05:45,  7.85s/it]\n 14%|█▍        | 7/50 [00:57<05:34,  7.77s/it]\n 16%|█▌        | 8/50 [01:05<05:24,  7.72s/it]\n 18%|█▊        | 9/50 [01:13<05:14,  7.68s/it]\n 20%|██        | 10/50 [01:20<05:06,  7.65s/it]\n 22%|██▏       | 11/50 [01:28<04:57,  7.64s/it]\n 24%|██▍       | 12/50 [01:35<04:49,  7.63s/it]\n 26%|██▌       | 13/50 [01:43<04:42,  7.62s/it]\n 28%|██▊       | 14/50 [01:51<04:34,  7.62s/it]\n 30%|███       | 15/50 [01:58<04:26,  7.62s/it]\n 32%|███▏      | 16/50 [02:06<04:19,  7.63s/it]\n 34%|███▍      | 17/50 [02:14<04:11,  7.63s/it]\n 36%|███▌      | 18/50 [02:21<04:04,  7.63s/it]\n 38%|███▊      | 19/50 [02:29<03:56,  7.63s/it]\n 40%|████      | 20/50 [02:37<03:49,  7.63s/it]\n 42%|████▏     | 21/50 [02:44<03:41,  7.63s/it]\n 44%|████▍     | 22/50 [02:52<03:33,  7.64s/it]\n 46%|████▌     | 23/50 [02:59<03:26,  7.64s/it]\n 48%|████▊     | 24/50 [03:07<03:18,  7.64s/it]\n 50%|█████     | 25/50 [03:15<03:10,  7.64s/it]\n 52%|█████▏    | 26/50 [03:22<03:03,  7.64s/it]\n 54%|█████▍    | 27/50 [03:30<02:55,  7.64s/it]\n 56%|█████▌    | 28/50 [03:38<02:48,  7.64s/it]\n 58%|█████▊    | 29/50 [03:45<02:40,  7.64s/it]\n 60%|██████    | 30/50 [03:53<02:32,  7.64s/it]\n 62%|██████▏   | 31/50 [04:01<02:25,  7.65s/it]\n 64%|██████▍   | 32/50 [04:08<02:17,  7.65s/it]\n 66%|██████▌   | 33/50 [04:16<02:10,  7.66s/it]\n 68%|██████▊   | 34/50 [04:24<02:02,  7.66s/it]\n 70%|███████   | 35/50 [04:31<01:54,  7.66s/it]\n 72%|███████▏  | 36/50 [04:39<01:47,  7.66s/it]\n 74%|███████▍  | 37/50 [04:47<01:39,  7.66s/it]\n 76%|███████▌  | 38/50 [04:54<01:31,  7.66s/it]\n 78%|███████▊  | 39/50 [05:02<01:24,  7.66s/it]\n 80%|████████  | 40/50 [05:10<01:16,  7.66s/it]\n 82%|████████▏ | 41/50 [05:17<01:08,  7.66s/it]\n 84%|████████▍ | 42/50 [05:25<01:01,  7.66s/it]\n 86%|████████▌ | 43/50 [05:32<00:53,  7.66s/it]\n 88%|████████▊ | 44/50 [05:40<00:45,  7.66s/it]\n 90%|█████████ | 45/50 [05:48<00:38,  7.66s/it]\n 92%|█████████▏| 46/50 [05:55<00:30,  7.66s/it]\n 94%|█████████▍| 47/50 [06:03<00:22,  7.66s/it]\n 96%|█████████▌| 48/50 [06:11<00:15,  7.66s/it]\n 98%|█████████▊| 49/50 [06:18<00:07,  7.66s/it]\n100%|██████████| 50/50 [06:26<00:00,  7.66s/it]\n100%|██████████| 50/50 [06:26<00:00,  7.73s/it]",
  "metrics": {
    "predict_time": 406.217533968,
    "total_time": 555.338449
  },
  "output": "https://replicate.delivery/pbxt/ru4kvN7GrOIlNVQGin3bYZsKYsacFmIqZty4tfEusqmtNlvJA/out.mp4",
  "started_at": "2024-09-21T16:24:38.502915Z",
  "status": "succeeded",
  "urls": {
    "get": "https://api.replicate.com/v1/predictions/qb7hex84wsrgj0cj2nj95qk4y4",
    "cancel": "https://api.replicate.com/v1/predictions/qb7hex84wsrgj0cj2nj95qk4y4/cancel"
  },
  "version": "82caaa79520e03b3963c975e38dd68ac8a6b18a8a39c19fd8dc03b4ed4b91c58"
}

Generated in

6 minutes 46 seconds

Tweak it ShareReport

Using seed: 10643
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:12<10:16, 12.59s/it]
  4%|▍         | 2/50 [00:20<07:41,  9.62s/it]
  6%|▌         | 3/50 [00:27<06:47,  8.66s/it]
  8%|▊         | 4/50 [00:35<06:18,  8.22s/it]
 10%|█         | 5/50 [00:42<05:59,  7.99s/it]
 12%|█▏        | 6/50 [00:50<05:45,  7.85s/it]
 14%|█▍        | 7/50 [00:57<05:34,  7.77s/it]
 16%|█▌        | 8/50 [01:05<05:24,  7.72s/it]
 18%|█▊        | 9/50 [01:13<05:14,  7.68s/it]
 20%|██        | 10/50 [01:20<05:06,  7.65s/it]
 22%|██▏       | 11/50 [01:28<04:57,  7.64s/it]
 24%|██▍       | 12/50 [01:35<04:49,  7.63s/it]
 26%|██▌       | 13/50 [01:43<04:42,  7.62s/it]
 28%|██▊       | 14/50 [01:51<04:34,  7.62s/it]
 30%|███       | 15/50 [01:58<04:26,  7.62s/it]
 32%|███▏      | 16/50 [02:06<04:19,  7.63s/it]
 34%|███▍      | 17/50 [02:14<04:11,  7.63s/it]
 36%|███▌      | 18/50 [02:21<04:04,  7.63s/it]
 38%|███▊      | 19/50 [02:29<03:56,  7.63s/it]
 40%|████      | 20/50 [02:37<03:49,  7.63s/it]
 42%|████▏     | 21/50 [02:44<03:41,  7.63s/it]
 44%|████▍     | 22/50 [02:52<03:33,  7.64s/it]
 46%|████▌     | 23/50 [02:59<03:26,  7.64s/it]
 48%|████▊     | 24/50 [03:07<03:18,  7.64s/it]
 50%|█████     | 25/50 [03:15<03:10,  7.64s/it]
 52%|█████▏    | 26/50 [03:22<03:03,  7.64s/it]
 54%|█████▍    | 27/50 [03:30<02:55,  7.64s/it]
 56%|█████▌    | 28/50 [03:38<02:48,  7.64s/it]
 58%|█████▊    | 29/50 [03:45<02:40,  7.64s/it]
 60%|██████    | 30/50 [03:53<02:32,  7.64s/it]
 62%|██████▏   | 31/50 [04:01<02:25,  7.65s/it]
 64%|██████▍   | 32/50 [04:08<02:17,  7.65s/it]
 66%|██████▌   | 33/50 [04:16<02:10,  7.66s/it]
 68%|██████▊   | 34/50 [04:24<02:02,  7.66s/it]
 70%|███████   | 35/50 [04:31<01:54,  7.66s/it]
 72%|███████▏  | 36/50 [04:39<01:47,  7.66s/it]
 74%|███████▍  | 37/50 [04:47<01:39,  7.66s/it]
 76%|███████▌  | 38/50 [04:54<01:31,  7.66s/it]
 78%|███████▊  | 39/50 [05:02<01:24,  7.66s/it]
 80%|████████  | 40/50 [05:10<01:16,  7.66s/it]
 82%|████████▏ | 41/50 [05:17<01:08,  7.66s/it]
 84%|████████▍ | 42/50 [05:25<01:01,  7.66s/it]
 86%|████████▌ | 43/50 [05:32<00:53,  7.66s/it]
 88%|████████▊ | 44/50 [05:40<00:45,  7.66s/it]
 90%|█████████ | 45/50 [05:48<00:38,  7.66s/it]
 92%|█████████▏| 46/50 [05:55<00:30,  7.66s/it]
 94%|█████████▍| 47/50 [06:03<00:22,  7.66s/it]
 96%|█████████▌| 48/50 [06:11<00:15,  7.66s/it]
 98%|█████████▊| 49/50 [06:18<00:07,  7.66s/it]
100%|██████████| 50/50 [06:26<00:00,  7.66s/it]
100%|██████████| 50/50 [06:26<00:00,  7.73s/it]

Examples

View more examples

Run time and cost

This model costs approximately $0.48 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 L40S GPU hardware. Predictions typically complete within 9 minutes.

Readme

CogVideoX

This is the image-to-video generation demo, for text-to-video go to https://replicate.com/chenxwh/cogvideox-t2v

CogVideoX is an open-source version of the video generation model originating from QingYing.

Citation

🌟 If you find our work helpful, please leave us a star and cite our paper.

@article{yang2024cogvideox,
  title={CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer},
  author={Yang, Zhuoyi and Teng, Jiayan and Zheng, Wendi and Ding, Ming and Huang, Shiyu and Xu, Jiazheng and Yang, Yuanming and Hong, Wenyi and Zhang, Xiaohan and Feng, Guanyu and others},
  journal={arXiv preprint arXiv:2408.06072},
  year={2024}
}

We welcome your contributions! You can click here for more information.

License Agreement

The code in this repository is released under the Apache 2.0 License.

The CogVideoX-5B model (Transformers module, include I2V and T2V) is released under the CogVideoX LICENSE.