Instructions to use berkeley-nest/Starling-RM-7B-alpha with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use berkeley-nest/Starling-RM-7B-alpha with Transformers:
# Load model directly from transformers import AutoTokenizer, LLMForSequenceRegression tokenizer = AutoTokenizer.from_pretrained("berkeley-nest/Starling-RM-7B-alpha") model = LLMForSequenceRegression.from_pretrained("berkeley-nest/Starling-RM-7B-alpha") - Notebooks
- Google Colab
- Kaggle
Prompt Format
#4
by ekurtulus - opened
Is there a specific prompt format for this model?
GPT4 User: Your instruction goes here
<|end_of_turn|>GPT4 Assistant:
Oh my apologies for missing the question here. Actually our reward model has a different prompt format as language model. It's included in the example code: "[INST] Hello? [/INST] Hi, how can I help you?"