gorilla-openfunctions-v0-sharded

Gorilla-openfunctions-v0 - Sharded

<img src="https://gorilla.cs.berkeley.edu/assets/img/blog_post_4_gorilla_open_function_calling.png" alt="Gorilla Open Functions" width="800" style="margin-left:'auto' margin-right:'auto' display:'block'"/>

🧩🧩🧩 Just a sharded version of gorilla-openfunctions-v0.

💻 Using this version, you can smoothly load the model on Colab and play with it!

From the original model card:

Gorilla OpenFunctions extends Large Language Model(LLM) Chat Completion feature to formulate executable APIs call given natural language instructions and API context.

📣 NEW: try this model for Information Extraction

🧪🦍 Needle in a Jungle - Information Extraction via LLMs

I did an experiment: a notebook that extracts information from a given URL (or text) based on a user-provided structure.

Stack: 🏗️ Haystack LLM framework + Gorilla OpenFunctions

• 📝 Post full of details
• 📓 Notebook

General Usage

This version of the model is meant primarily to run smoothly on Colab. I suggest loading the model with 8-bit quantization, so that you have some free GPU to perform inference.

However, it is perfectly fine to load the model in half-precision or with stronger quantization (4-bit).

! pip install transformers accelerate bitsandbytes

import json
import torch
from transformers import pipeline

def get_prompt(user_query: str, functions: list = []) -> str:
    """
    Generates a conversation prompt based on the user's query and a list of functions.

    Parameters:
    - user_query (str): The user's query.
    - functions (list): A list of functions to include in the prompt.

    Returns:
    - str: The formatted conversation prompt.
    """
    if len(functions) == 0:
        return f"USER: <<question>> {user_query}\nASSISTANT: "
    functions_string = json.dumps(functions)
    return f"USER: <<question>> {user_query} <<function>> {functions_string}\nASSISTANT: "

# Pipeline setup
pipe = pipeline(
    "text-generation",
    model="anakin87/gorilla-openfunctions-v0-sharded",
    device_map="auto",
    model_kwargs={"load_in_8bit":True, "torch_dtype":torch.float16},
    max_new_tokens=128,
    batch_size=16
)

# Example usage
query: str = "Call me an Uber ride type \"Plus\" in Berkeley at zipcode 94704 in 10 minutes"
functions = [
    {
        "name": "Uber Carpool",
        "api_name": "uber.ride",
        "description": "Find suitable ride for customers given the location, type of ride, and the amount of time the customer is willing to wait as parameters",
        "parameters":  [
            {"name": "loc", "description": "Location of the starting place of the Uber ride"},
            {"name": "type", "enum": ["plus", "comfort", "black"], "description": "Types of Uber ride user is ordering"},
            {"name": "time", "description": "The amount of time in minutes the customer is willing to wait"}
        ]
    }
]

# Generate prompt and obtain model output
prompt = get_prompt(query, functions=functions)
output = pipe(prompt)

print(output[0]['generated_text'].rpartition("ASSISTANT:")[-1].strip())
# uber.ride(loc="berkeley", type="plus", time=10)