ji4chenli/t2v-turbo-v2 – Run with an API on Replicate

ji4chenli / t2v-turbo-v2

Enhancing Video Model Post-Training through Data, Reward, and Conditional Guidance Design

Cold

Public
598 runs
L40S
GitHub
Paper

Run with an API

Playground API Examples README Versions

Input

prompt

string

Shift + Return to add a new line

With the style of low-poly game art, A majestic, white horse gallops gracefully across a moonlit beachWith the style of low-poly game art, A majestic, white horse gallops gracefully across a moonlit beach

Input prompt

Default: "With the style of low-poly game art, A majestic, white horse gallops gracefully across a moonlit beach"

num_inference_steps

integer

(minimum: 4, maximum: 50)

Number of denoising steps

Default: 16

guidance_scale

number

(minimum: 2, maximum: 14)

Scale for classifier-free guidance

Default: 7.5

motion_gs

number

Set guidance for motion

Default: 0.05

percentage

number

(minimum: 0, maximum: 0.5)

Percentage of steps to apply motion guidance (v2 w/ MG only)

Default: 0.5

num_frames

integer

Number of Video Frames

Default: 16

fps

integer

FPS of the output video.

Default: 8

seed

integer

Random seed. Leave blank to randomize the seed

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

npx create-replicate --model=ji4chenli/t2v-turbo-v2

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 ji4chenli/t2v-turbo-v2 using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "ji4chenli/t2v-turbo-v2:cdd7ad235a237a2076808758531221bae1a84422f30becd4fc881c9940093ce5",
  {
    input: {
      fps: 8,
      prompt: "With the style of low-poly game art, A majestic, white horse gallops gracefully across a moonlit beach",
      motion_gs: 0.05,
      num_frames: 16,
      percentage: 0.5,
      guidance_scale: 7.5,
      num_inference_steps: 16
    }
  }
);

// 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 ji4chenli/t2v-turbo-v2 using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "ji4chenli/t2v-turbo-v2:cdd7ad235a237a2076808758531221bae1a84422f30becd4fc881c9940093ce5",
    input={
        "fps": 8,
        "prompt": "With the style of low-poly game art, A majestic, white horse gallops gracefully across a moonlit beach",
        "motion_gs": 0.05,
        "num_frames": 16,
        "percentage": 0.5,
        "guidance_scale": 7.5,
        "num_inference_steps": 16
    }
)
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 ji4chenli/t2v-turbo-v2 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": "cdd7ad235a237a2076808758531221bae1a84422f30becd4fc881c9940093ce5",
    "input": {
      "fps": 8,
      "prompt": "With the style of low-poly game art, A majestic, white horse gallops gracefully across a moonlit beach",
      "motion_gs": 0.05,
      "num_frames": 16,
      "percentage": 0.5,
      "guidance_scale": 7.5,
      "num_inference_steps": 16
    }
  }' \
  https://api.replicate.com/v1/predictions

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

Output

{
  "completed_at": "2024-10-14T11:29:36.010195Z",
  "created_at": "2024-10-14T11:27:34.894000Z",
  "data_removed": false,
  "error": null,
  "id": "00acswwqxsrgm0cjhb487eg5g0",
  "input": {
    "fps": 8,
    "prompt": "With the style of low-poly game art, A majestic, white horse gallops gracefully across a moonlit beach",
    "motion_gs": 0.05,
    "num_frames": 16,
    "percentage": 0.5,
    "guidance_scale": 7.5,
    "num_inference_steps": 16
  },
  "logs": "Using seed: 23534\n  0%|          | 0/16 [00:00<?, ?it/s]\n  6%|▋         | 1/16 [00:00<00:06,  2.31it/s]\n 12%|█▎        | 2/16 [00:00<00:03,  3.55it/s]\n 19%|█▉        | 3/16 [00:00<00:03,  3.27it/s]\n 25%|██▌       | 4/16 [00:01<00:03,  3.15it/s]\n 31%|███▏      | 5/16 [00:01<00:03,  3.09it/s]\n 38%|███▊      | 6/16 [00:01<00:03,  3.05it/s]\n 44%|████▍     | 7/16 [00:02<00:02,  3.03it/s]\n 50%|█████     | 8/16 [00:02<00:02,  3.02it/s]\n 56%|█████▋    | 9/16 [00:02<00:02,  3.01it/s]\n 62%|██████▎   | 10/16 [00:03<00:01,  3.01it/s]\n 69%|██████▉   | 11/16 [00:03<00:01,  3.00it/s]\n 75%|███████▌  | 12/16 [00:03<00:01,  3.00it/s]\n 81%|████████▏ | 13/16 [00:04<00:01,  3.00it/s]\n 88%|████████▊ | 14/16 [00:04<00:00,  3.00it/s]\n 94%|█████████▍| 15/16 [00:04<00:00,  3.00it/s]\n100%|██████████| 16/16 [00:05<00:00,  2.99it/s]\n100%|██████████| 16/16 [00:05<00:00,  3.02it/s]",
  "metrics": {
    "predict_time": 8.009425347,
    "total_time": 121.116195
  },
  "output": "https://replicate.delivery/pbxt/K0zaYeQK51W6BK8EtxDC5tes1uSslcChobSYX3e8MSr9UWNnA/out.mp4",
  "started_at": "2024-10-14T11:29:28.000770Z",
  "status": "succeeded",
  "urls": {
    "get": "https://api.replicate.com/v1/predictions/00acswwqxsrgm0cjhb487eg5g0",
    "cancel": "https://api.replicate.com/v1/predictions/00acswwqxsrgm0cjhb487eg5g0/cancel"
  },
  "version": "cdd7ad235a237a2076808758531221bae1a84422f30becd4fc881c9940093ce5"
}

Generated in

8.0 seconds

Tweak it ShareReport

Examples

View more examples

Run time and cost

This model costs approximately $0.025 to run on Replicate, or 40 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 26 seconds.

Readme

T2V-Turbo-v2: Enhancing Video Model Post-Training through Data, Reward, and Conditional Guidance Design

Project Page: https://t2v-turbo-v2.github.io/

T2V-Turbo-v2