|
Download setup_instructions.md from seochan99/ecb-datasets: direct link, hf CLI and curl.
- Browser
- Download file 2.34 kB
-
https://huggingface.co/datasets/seochan99/ecb-datasets/resolve/main/setup_instructions.md
- Command line
-
hf download hf://datasets/seochan99/ecb-datasets/setup_instructions.md
-
curl -L -o setup_instructions.md https://huggingface.co/datasets/seochan99/ecb-datasets/resolve/main/setup_instructions.md
2.34 kB
| # Hugging Face Dataset Upload Guide | |
| ## 1. Hugging Face Token Setup | |
| ### Method 1: Create Token on Web | |
| 1. Visit https://huggingface.co/settings/tokens | |
| 2. Click "New token" | |
| 3. Enter token name (e.g., "dataset-upload") | |
| 4. Select "Write" permission | |
| 5. Copy the token | |
| ### Method 2: Set Environment Variable | |
| ```bash | |
| export HUGGINGFACE_HUB_TOKEN="your_token_here" | |
| ``` | |
| ### Method 3: Use Login Command | |
| ```bash | |
| python3 -c "from huggingface_hub import login; login()" | |
| # Enter the token when prompted | |
| ``` | |
| ## 2. Dataset Upload Methods | |
| ### Method A: Python Script (Recommended) | |
| ```bash | |
| cd /Users/chan/ECB/hugging_face/dataset | |
| python3 upload_to_huggingface.py seochan99 ecb-datasets | |
| ``` | |
| ### Method B: Git Usage | |
| ```bash | |
| cd /Users/chan/ECB/hugging_face/dataset | |
| ./upload_to_huggingface.sh seochan99 ecb-datasets | |
| ``` | |
| ### Method C: Web Interface | |
| 1. Go to https://huggingface.co/new-dataset to create new dataset | |
| 2. "Add file" → "Upload files" to upload all files | |
| 3. Click "Commit changes" | |
| ## 3. Post-Upload Verification | |
| 1. **Dataset URL Check**: `https://huggingface.co/datasets/seochan99/ecb-datasets` | |
| 2. **File Structure Check**: Verify all folders and files are uploaded correctly | |
| 3. **README.md Check**: Ensure dataset description is displayed correctly | |
| ## 4. Dataset Usage | |
| The uploaded dataset can be used as follows: | |
| ```python | |
| from datasets import load_dataset | |
| # Load specific dataset configurations | |
| add_attribution = load_dataset("seochan99/ecb-datasets", "add_attribution") | |
| country_swap = load_dataset("seochan99/ecb-datasets", "country_swap") | |
| multi_loop_edits = load_dataset("seochan99/ecb-datasets", "multi_loop_edits") | |
| ``` | |
| ## 5. Troubleshooting | |
| ### Token-related Errors | |
| - Verify token is correct | |
| - Ensure token has "Write" permission | |
| ### Upload Failures | |
| - Check internet connection | |
| - Verify file size limits (Hugging Face has 5GB limit per file) | |
| - Check permissions | |
| ### Git-related Errors | |
| - Ensure Git is installed | |
| - Verify Hugging Face account is correct | |
| ## 6. Dataset Information | |
| - **Dataset Name**: ECB Datasets: Cultural Bias Evaluation in Generative Image Models | |
| - **Related Paper**: "Exposing Blindspots: Cultural Bias Evaluation in Generative Image Models" | |
| - **ECB Meaning**: Evaluation Cultural Bias | |
| - **Models**: Flux, HiDream, NextStep, Qwen, SD35 | |
| - **Countries**: Korea, China, India, Kenya, Nigeria, United States |