RichardErkhov/ValiantLabs_-_Llama3-70B-Fireplace-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Llama3-70B-Fireplace - GGUF

• —Model creator: https://huggingface.co/ValiantLabs/
• —Original model: https://huggingface.co/ValiantLabs/Llama3-70B-Fireplace/

Original model description: --- language:

• —en pipeline_tag: text-generation tags:
• —fireplace
• —valiant
• —valiant-labs
• —llama
• —llama-3
• —llama-3-instruct
• —llama-3-instruct-70b
• —70b
• —function-calling
• —conversational
• —chat
• —instruct model_type: llama license: llama3 ---

Fireplace is a function-calling model for Llama 3 70b Instruct.

• —combines function-calling abilities with a high-performance, versatile chat model
• —function-calling utilizing the Llama 3 Instruct format

This version of Fireplace, like our previous Fireplace-13b and Fireplace-34b models, focuses on combining chat-instruct and function-calling only.

We're working now on Fireplace 2 for Llama 3, which will include function calling as one of several enhanced technical skills.

Version

This is the 2024-05-09 release of Fireplace for Llama 3 70b.

We're excited to bring additional releases for Fireplace and other models in our Build Tools lineup to Llama 3 soon!

Prompting Guide

Fireplace uses the Llama 3 Instruct prompt format:

<|beginoftext|><|startheaderid|>system<|endheaderid|>{{ systemprompt }}<|eotid|><|startheaderid|>user<|endheaderid|>{{ usermsg1 }}<|eotid|><|startheaderid|>assistant<|endheaderid|>{{ modelanswer1 }}<|eotid|>

Example input for function calling:

<|beginoftext|><|startheaderid|>system<|endheaderid|>\n\n You are Fireplace, an expert code assistant with access to the following functions. Use them if required - { "name": "calculatetip", "description": "Calculate the tip amount for a bill", "parameters": { "type": "object", "properties": { "billamount": { "type": "number", "description": "The total amount of the bill" }, "tippercentage": { "type": "number", "description": "The percentage of tip to be given" } }, "required": [ "billamount", "tippercentage" ] } } { "name": "checkwebsiteavailability", "description": "Check the availability of a website", "parameters": { "type": "object", "properties": { "url": { "type": "string", "description": "The URL of the website" } }, "required": [ "url" ] } } <|eotid|><|startheaderid|>user<|endheaderid|>\n\nHi, I need help with calculating a tip. My bill is $100 and I want to leave a 30% tip. <|eotid|><|startheaderid|>assistant<|endheader_id|>

For assistant handling of function responses, deliver them in a new user message:

<|startheaderid|>user<|endheaderid|>\n\n FUNCTION RESPONSE: {"status": "success", "message": "Email has been sent successfully"} <|eot_id|>

WARNING: text-generation-webui

When using Llama 3 Instruct models (including Fireplace) with text-generation-webui note that a current bug in webui can result in incorrect reading of the model's ending tokens, causing unfinished outputs and incorrect structure.

For a temporary workaround if you encounter this issue, edit Fireplace's tokenizer_config file as indicated:

from "eostoken": "<|endof_text|>",

to "eostoken": "<|eotid|>",

The Model

Fireplace is built on top of Llama 3 70b Instruct, the highest performance open-source model currently available.

This version of Fireplace uses the glaiveai/glaive-function-calling-v2 dataset converted to Llama 3 Instruct format.

Fireplace is created by Valiant Labs.

Check out our HuggingFace page for Shining Valiant 2 and our other models!

Follow us on X for updates on our models!

We care about open source. For everyone to use.

We encourage others to finetune further from our models.