einstein-model-v1

The context data used for training is einstein_dataset.json, thanks to https://www.kaggle.com/abdelrahmanm2003.

Run this instructions in a new Notebook to run the model:

# Install the llama-cpp-python package
!pip install llama-cpp-python

from llama_cpp import Llama

# Load the pretrained Einstein model
llm = Llama.from_pretrained(
    repo_id="m1zhab/einstein-model-v1",
    filename="model-unsloth-Q4_K_M.gguf",
)

# Function to generate chat completions
def chat_complete(prompt):
    resp = llm.create_chat_completion(
        messages=[
            {
                "role": "system",
                "content": (
                    "You're a model acting as Albert Einstein, with a knowledge cutoff "
                    "of April, 1955. When asked about inventions and scientific developments "
                    "past this date, you're supposed to answer that you don't know."
                ),
            },
            {"role": "user", "content": prompt},
        ]
    )
    return resp["choices"][0]["message"]["content"]

# Example usage
prompt = "Where were you born, tell me about your childhood."
msg = chat_complete(prompt)

print(msg)

Uploaded model

• Developed by: m1zhab
• License: apache-2.0
• Finetuned from model : unsloth/llama-3.2-3b-instruct-unsloth-bnb-4bit

This llama model was trained 2x faster with Unsloth and Huggingface's TRL library.

<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>