Mistral-NeMo-Minitron-8B-Chat

This is an instruction-tuned version of nvidia/Mistral-NeMo-Minitron-8B-Base that has underwent supervised fine-tuning with 32k instruction-response pairs from the teknium/OpenHermes-2.5 dataset.

How to use

Chat Format

Given the nature of the training data, the Mistral-NeMo-Minitron-8B chat model is best suited for prompts using the chat format as follows. You can provide the prompt as a question with a generic template as follows:

<|im_start|>system
You are a helpful assistant.<|im_end|>
<|im_start|>user
Question?<|im_end|>
<|im_start|>assistant

For example:

<|im_start|>system
You are a helpful assistant.<|im_end|>
<|im_start|>user
How to explain Internet for a medieval knight?<|im_end|>
<|im_start|>assistant

where the model generates the text after <|im_start|>assistant .

Sample inference code

This code snippets show how to get quickly started with running the model on a GPU:

import torch 
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline 

torch.random.manual_seed(0) 

model_id = "rasyosef/Mistral-NeMo-Minitron-8B-Chat"
model = AutoModelForCausalLM.from_pretrained( 
    model_id,  
    device_map="auto",  
    torch_dtype=torch.bfloat16 
) 

tokenizer = AutoTokenizer.from_pretrained(model_id) 

messages = [ 
    {"role": "system", "content": "You are a helpful AI assistant."}, 
    {"role": "user", "content": "Can you provide ways to eat combinations of bananas and dragonfruits?"}, 
    {"role": "assistant", "content": "Sure! Here are some ways to eat bananas and dragonfruits together: 1. Banana and dragonfruit smoothie: Blend bananas and dragonfruits together with some milk and honey. 2. Banana and dragonfruit salad: Mix sliced bananas and dragonfruits together with some lemon juice and honey."}, 
    {"role": "user", "content": "What about solving an 2x + 3 = 7 equation?"}, 
] 

pipe = pipeline( 
    "text-generation", 
    model=model, 
    tokenizer=tokenizer, 
) 

generation_args = { 
    "max_new_tokens": 256, 
    "return_full_text": False, 
    "temperature": 0.0, 
    "do_sample": False, 
} 

output = pipe(messages, **generation_args) 
print(output[0]['generated_text'])  

Note: If you want to use flash attention, call AutoModelForCausalLM.from_pretrained() with attn_implementation="flash_attention_2"

Benchmarks

These benchmarks were run using EleutherAI's lm-evaluation-harness

• IFEval (Instruction Following Evaluation): IFEval is a fairly interesting dataset that tests the capability of models to clearly follow explicit instructions, such as “include keyword x” or “use format y”. The models are tested on their ability to strictly follow formatting instructions rather than the actual contents generated, allowing strict and rigorous metrics to be used.
  • Score: 45.83

Demo

Here's a colab notebook with a chat interface, you can use this to interact with the chat model.

https://huggingface.co./rasyosef/Mistral-NeMo-Minitron-8B-Chat/blob/main/Mistral_NeMo_Minitron_8B_chatbot.ipynb

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	17.22
IFEval (0-Shot)	44.52
BBH (3-Shot)	26.04
MATH Lvl 5 (4-Shot)	0.76
GPQA (0-shot)	3.47
MuSR (0-shot)	12.94
MMLU-PRO (5-shot)	15.60