Instructions to use humanlong/improving-self-evolution-mbpp with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use humanlong/improving-self-evolution-mbpp with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("humanlong/improving-self-evolution-mbpp", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Download round5_comparisons.json from humanlong/improving-self-evolution-mbpp: direct link, hf CLI and curl.
- Browser
- Download file 764 Bytes
-
https://huggingface.co/humanlong/improving-self-evolution-mbpp/resolve/main/round5_comparisons.json
- Command line
-
hf download hf://humanlong/improving-self-evolution-mbpp/round5_comparisons.json
-
curl -L -o round5_comparisons.json https://huggingface.co/humanlong/improving-self-evolution-mbpp/resolve/main/round5_comparisons.json
764 Bytes
| [ | |
| { | |
| "candidate": "spectral_soft/round_5/evaluation", | |
| "reference": "base/evaluation", | |
| "pass1_delta": 0.0225, | |
| "pass1_ci95": [ | |
| 0.012, | |
| 0.0335 | |
| ], | |
| "noninferior_at_1pct": true, | |
| "coverage4_delta": -0.3152965855006672, | |
| "coverage4_ci95": [ | |
| -0.39445073313153484, | |
| -0.2395721128340615 | |
| ], | |
| "paired_eligible_tasks": 245 | |
| }, | |
| { | |
| "candidate": "spectral_soft/round_5/evaluation", | |
| "reference": "spd_hard/round_5/evaluation", | |
| "pass1_delta": -0.01175, | |
| "pass1_ci95": [ | |
| -0.019, | |
| -0.00475 | |
| ], | |
| "noninferior_at_1pct": false, | |
| "coverage4_delta": 0.2929389999111115, | |
| "coverage4_ci95": [ | |
| 0.23322470344098753, | |
| 0.3531874808062307 | |
| ], | |
| "paired_eligible_tasks": 251 | |
| } | |
| ] | |