Instructions to use jeiku/Writing_Mistral with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use jeiku/Writing_Mistral with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("models/jeiku_Rosa_v1_7B") model = PeftModel.from_pretrained(base_model, "jeiku/Writing_Mistral") - Notebooks
- Google Colab
- Kaggle
Download training_log.json from jeiku/Writing_Mistral: direct link, hf CLI and curl.
- Browser
- Download file 432 Bytes
-
https://huggingface.co/jeiku/Writing_Mistral/resolve/main/training_log.json
- Command line
-
hf download hf://jeiku/Writing_Mistral/training_log.json
-
curl -L -o training_log.json https://huggingface.co/jeiku/Writing_Mistral/resolve/main/training_log.json
432 Bytes
| { | |
| "base_model_name": "jeiku_Rosa_v1_7B", | |
| "base_model_class": "MistralForCausalLM", | |
| "base_loaded_in_4bit": false, | |
| "base_loaded_in_8bit": false, | |
| "projections": "q, v", | |
| "loss": 2.093, | |
| "learning_rate": 0.0, | |
| "epoch": 9.8, | |
| "current_steps": 454, | |
| "train_runtime": 250.1064, | |
| "train_samples_per_second": 23.15, | |
| "train_steps_per_second": 0.06, | |
| "total_flos": 4.166767776419021e+16, | |
| "train_loss": 2.4269789377848308 | |
| } |