nelsonjchen/op-replay-clipper:f97b40e7 – Run with an API on Replicate

Version

You're looking at a specific version of this model. Jump to the model overview.

nelsonjchen /op-replay-clipper:f97b40e7

Input

renderType

string

UI renders with UI. Forward, Wide, and Driver process the raw, segmented, and low-compatibility HEVC video files into a portable H.264 MP4 file, are fast transcodes, and are great for quick previews. 360 requires viewing/uploading the video file in VLC or YouTube to pan around in a 🌐 sphere. Forward Upon Wide roughly overlays Forward video on Wide video. 360 Forward Upon Wide is 360 with Forward Upon Wide as the forward video.

Default: "ui"

route

string

Shift + Return to add a new line

https://connect.comma.ai/a2a0ccea32023010/1690488131496/1690488151496https://connect.comma.ai/a2a0ccea32023010/1690488131496/1690488151496

🔗 comma connect URL (e.g. https://connect.comma.ai/fe18f736cb0d7813/1698620773416/1698620855707, this is the preferred input method and includes dongle ID with start/end times.) OR #️⃣ route ID (e.g. a2a0ccea32023010|2023-07-27--13-01-19. Note that any segment ID "--" appended to the end will be ignored as"startSecond" is used instead, but route id portion of input will still accepted) (⚠️ "Public Access" must be enabled or a valid JWT Token must be provided. All required files for render type in Comma Connect must be uploaded from device. Please see the Quick Usage section of the README on GitHub at https://github.com/nelsonjchen/op-replay-clipper#quick-usage for instructions on generating an appropiate comma connect URL.)

Default: "https://connect.comma.ai/a2a0ccea32023010/1690488131496/1690488151496"

startSeconds

integer

(minimum: 0)

Start time in seconds for #️⃣ Route ID route input only. (🔗 comma connect URL already has the start time embedded in it and this input will be ignored in favor of that)

Default: 50

lengthSeconds

integer

(minimum: 5, maximum: 300)

Length of clip in seconds #️⃣ Route ID route input only. (🔗 comma connect URL already has the length time indirectly embedded in it from the embedded end time and this input will be ignored in favor of that. The minimum and maximum length will still be enforced)

Default: 20

smearAmount

integer

(minimum: 5, maximum: 40)

(UI Render only) Smear amount (Let the video start this time before beginning recording, useful for making sure the radar triangle (△), if not present, is present to be rendered at the start if necessary)

Default: 5

speedhackRatio

number

(minimum: 0.1, maximum: 7)

(UI Render only) Speedhack ratio (Higher ratio renders faster but renders may be more unstable and have artifacts) (Suggestion: 0.1-0.3 for jitter-free, 1-3 for fast renders)

Default: 1

metric

boolean

(UI Render only) Render in metric units (km/h)

Default: false

forwardUponWideH

number

(minimum: 1.9, maximum: 2.3)

(Forward Upon Wide Renders only) H-position of the forward video overlay on wide. Different devices can have different offsets from differing user mounting or factory calibration.

Default: 2.2

fileSize

integer

(minimum: 10, maximum: 100)

Rough size of clip output in MB.

Default: 25

jwtToken

string

Shift + Return to add a new line

Optional JWT Token from https://jwt.comma.ai for non-"Public access" routes. ⚠️ DO NOT SHARE THIS TOKEN WITH ANYONE as https://jwt.comma.ai generates JWT tokens valid for 1 year and they are irrevocable. Please use the safer, optionally temporary, more granular, and revocable "Public Access" toggle option on comma connect if possible. For more info, please see https://github.com/nelsonjchen/op-replay-clipper#jwt-token-input .

Default: ""

notes

string

Shift + Return to add a new line

Notes Text field. Doesn't affect output. For your own reference.

Default: ""

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

npx create-replicate --model=nelsonjchen/op-replay-clipper

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 nelsonjchen/op-replay-clipper using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "nelsonjchen/op-replay-clipper:f97b40e7ccea3fac7d49a7e41600dd6ad199a524c28ed7a1f78ac7b3f78e4129",
  {
    input: {
      notes: "",
      route: "https://connect.comma.ai/a2a0ccea32023010/1690488131496/1690488151496",
      metric: false,
      fileSize: 25,
      jwtToken: "",
      renderType: "ui",
      smearAmount: 5,
      startSeconds: 50,
      lengthSeconds: 20,
      speedhackRatio: 1,
      forwardUponWideH: 2.2
    }
  }
);
console.log(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 nelsonjchen/op-replay-clipper using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "nelsonjchen/op-replay-clipper:f97b40e7ccea3fac7d49a7e41600dd6ad199a524c28ed7a1f78ac7b3f78e4129",
    input={
        "notes": "",
        "route": "https://connect.comma.ai/a2a0ccea32023010/1690488131496/1690488151496",
        "metric": False,
        "fileSize": 25,
        "jwtToken": "",
        "renderType": "ui",
        "smearAmount": 5,
        "startSeconds": 50,
        "lengthSeconds": 20,
        "speedhackRatio": 1,
        "forwardUponWideH": 2.2
    }
)
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 nelsonjchen/op-replay-clipper 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": "f97b40e7ccea3fac7d49a7e41600dd6ad199a524c28ed7a1f78ac7b3f78e4129",
    "input": {
      "notes": "",
      "route": "https://connect.comma.ai/a2a0ccea32023010/1690488131496/1690488151496",
      "metric": false,
      "fileSize": 25,
      "jwtToken": "",
      "renderType": "ui",
      "smearAmount": 5,
      "startSeconds": 50,
      "lengthSeconds": 20,
      "speedhackRatio": 1,
      "forwardUponWideH": 2.2
    }
  }' \
  https://api.replicate.com/v1/predictions

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

You can run this model locally using Cog. First, install Cog:

brew install cog

If you don’t have Homebrew, there are other installation options available.

Run this to download the model and run it in your local environment:

cog predict r8.im/nelsonjchen/op-replay-clipper@sha256:f97b40e7ccea3fac7d49a7e41600dd6ad199a524c28ed7a1f78ac7b3f78e4129 \
  -i 'notes=""' \
  -i 'route="https://connect.comma.ai/a2a0ccea32023010/1690488131496/1690488151496"' \
  -i 'metric=false' \
  -i 'fileSize=25' \
  -i 'jwtToken=""' \
  -i 'renderType="ui"' \
  -i 'smearAmount=5' \
  -i 'startSeconds=50' \
  -i 'lengthSeconds=20' \
  -i 'speedhackRatio=1' \
  -i 'forwardUponWideH=2.2'

To learn more, take a look at the Cog documentation.

Run this to download the model and run it in your local environment:

docker run -d -p 5000:5000 --gpus=all r8.im/nelsonjchen/op-replay-clipper@sha256:f97b40e7ccea3fac7d49a7e41600dd6ad199a524c28ed7a1f78ac7b3f78e4129
curl -s -X POST \
  -H "Content-Type: application/json" \
  -d $'{
    "input": {
      "notes": "",
      "route": "https://connect.comma.ai/a2a0ccea32023010/1690488131496/1690488151496",
      "metric": false,
      "fileSize": 25,
      "jwtToken": "",
      "renderType": "ui",
      "smearAmount": 5,
      "startSeconds": 50,
      "lengthSeconds": 20,
      "speedhackRatio": 1,
      "forwardUponWideH": 2.2
    }
  }' \
  http://localhost:5000/predictions

To learn more, take a look at the Cog documentation.

Output

No output yet! Press "Submit" to start a prediction.