Instructions to use Likich/open-coding-mistral7b-single_code-qlora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use Likich/open-coding-mistral7b-single_code-qlora with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2") model = PeftModel.from_pretrained(base_model, "Likich/open-coding-mistral7b-single_code-qlora") - Notebooks
- Google Colab
- Kaggle
File size: 423 Bytes
9bba433 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 | {
"n_test": 100,
"valid_json_rate": 1.0,
"nonempty_rate": 1.0,
"exact_set_match": 0.2,
"mean_set_f1": 0.2,
"average_codes": 1.0,
"model_name": "mistral7b",
"base_model": "mistralai/Mistral-7B-Instruct-v0.2",
"adapter": "artifacts/mistral7b",
"task_mode": "single_code",
"thresholds": {
"min_valid_json_rate": 0.9,
"min_nonempty_rate": 0.95,
"max_average_codes": 3.0
},
"passed": true
} |