Edit model card
Llama-3 DPO Logo

Llama-3-8B-Instruct-v0.8

This model was developed based on MaziyarPanahi/Llama-3-8B-Instruct-v0.4 model.

⚑ Quantized GGUF

All GGUF models are available here: MaziyarPanahi/Llama-3-8B-Instruct-v0.8-GGUF

πŸ† Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric Value
Avg. 73.20
AI2 Reasoning Challenge (25-Shot) 71.67
HellaSwag (10-Shot) 87.77
MMLU (5-Shot) 68.30
TruthfulQA (0-shot) 63.90
Winogrande (5-shot) 79.08
GSM8k (5-shot) 68.46

MaziyarPanahi/Llama-3-8B-Instruct-v0.8 is the 5th best-performing 8B model on the Open LLM Leaderboard. (03/06/2024).

image/png

Prompt Template

This model uses ChatML prompt template:

<|begin_of_text|><|start_header_id|>system<|end_header_id|>

{system_prompt}<|eot_id|><|start_header_id|>user<|end_header_id|>

{prompt}<|eot_id|><|start_header_id|>assistant<|end_header_id|>

How to use

You can use this model by using MaziyarPanahi/Llama-3-8B-Instruct-v0.8 as the model name in Hugging Face's transformers library.

from transformers import AutoModelForCausalLM, AutoTokenizer, TextStreamer
from transformers import pipeline
import torch

model_id = "MaziyarPanahi/Llama-3-8B-Instruct-v0.8"

model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype=torch.bfloat16,
    device_map="auto",
    trust_remote_code=True,
    # attn_implementation="flash_attention_2"
)

tokenizer = AutoTokenizer.from_pretrained(
    model_id,
    trust_remote_code=True
)

streamer = TextStreamer(tokenizer)

pipeline = pipeline(
    "text-generation",
    model=model,
    tokenizer=tokenizer,
    model_kwargs={"torch_dtype": torch.bfloat16},
    streamer=streamer
)

# Then you can use the pipeline to generate text.

messages = [
    {"role": "system", "content": "You are a pirate chatbot who always responds in pirate speak!"},
    {"role": "user", "content": "Who are you?"},
]

prompt = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)

terminators = [
    tokenizer.eos_token_id,    
    tokenizer.convert_tokens_to_ids("<|eot_id|>")
]

outputs = pipeline(
    prompt,
    max_new_tokens=512,
    eos_token_id=terminators,
    do_sample=True,
    temperature=0.6,
    top_p=0.95,
)
print(outputs[0]["generated_text"][len(prompt):])
Downloads last month
1,463
Safetensors
Model size
8.03B params
Tensor type
BF16
Β·
Inference API
Input a message to start chatting with MaziyarPanahi/Llama-3-8B-Instruct-v0.8.
Inference API (serverless) has been turned off for this model.

Finetuned from

Collection including MaziyarPanahi/Llama-3-8B-Instruct-v0.8

Evaluation results