Instructions to use FinGPT/fingpt-forecaster_dow30_llama2-7b_lora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use FinGPT/fingpt-forecaster_dow30_llama2-7b_lora with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("base_models/Llama-2-7b-chat-hf") model = PeftModel.from_pretrained(base_model, "FinGPT/fingpt-forecaster_dow30_llama2-7b_lora") - Notebooks
- Google Colab
- Kaggle
Runtime Error: ImportError: cannot import name 'driver' from 'triton.runtime' in FinGPT / Hugging Face Model Deployment
Dear FinGPT and Hugging Face Support Team,
I hope this message finds you well.
I am reaching out to report a runtime issue I encountered while deploying or running a model using the FinGPT/Hugging Face LLM integration. The application consistently fails with the following error trace:
Details
Environment: Python 3.11.17
Libraries involved: peft, bitsandbytes, triton
Runtime: (Please specify if running on FinGPT cloud, Hugging Face Space, or local deployment)
Error Summary: The bitsandbytes library is unable to import the driver module from triton.runtime, causing the model initialization to fail.
Request
I would appreciate your guidance on how to resolve or patch this issue.
If this is a known compatibility or dependency mismatch, kindly share the recommended versions of bitsandbytes and triton that work with the current FinGPT/Hugging Face runtime.
Thank you for your time and support.
Warm regards,
Mangesh Jadhav
B.Tech β Artificial Intelligence & Machine Learning
G.H. Raisoni College of Engineering, Pune
π§ mangeshjadhav126@gmail.com
π± +91 7498958281
