How to download dataset from huggingface using actions?

Hello, friends.

I created an action to download a CSV file, but for some reason I get a message about large size every time (“response_data”: “ResponseTooLargeError”). Although the file size is only a few hundred kilobytes and this should not be a problem. Perhaps huggingface is blocked? Or what is the problem?

OpenAPI 3.0 specification:

openapi: 3.0.0
info:
  title: Hugging Face Cryptocurrency Dataset API
  description: API for downloading cryptocurrency OHLCV datasets from Hugging Face.
  version: 1.0.0
servers:
  - url: https://huggingface.co
    description: Hugging Face API Server
paths:
  /datasets/arthurneuron/cryptocurrency-futures-ohlcv-dataset-1m/resolve/main/{symbol}USDT.csv:
    get:
      operationId: downloadCryptoData
      summary: Downloads OHLCV data for a specified cryptocurrency token.
      parameters:
        - name: symbol
          in: path
          required: true
          description: Symbol of the cryptocurrency token.
          schema:
            type: string
        - name: download
          in: query
          required: true
          description: Flag to trigger the download of the dataset.
          schema:
            type: boolean
            default: true
      responses:
        '200':
          description: A CSV file containing OHLCV data for the specified cryptocurrency token.
          content:
            text/csv:
              schema:
                type: string
                format: binary
        '404':
          description: Dataset not found

Wow, I’ve been waiting for 3 months to get this post approved. :slight_smile:
Now I know that GPTs considers CSV files as raw text and rewrites them in chat code.