llama-3-12b-instruct

Llama3 from 8B to 12B

We created a model from other cool models to combine everything into one cool model.

Model Details

Model Description

• Developed by: @ehristoforu
• Model type: Text Generation (conversational)
• Language(s) (NLP): English, Russian
• Finetuned from model: meta-llama/Meta-Llama-3-8B-Instruct

How to Get Started with the Model

Use the code below to get started with the model.

from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

model_id = "ehristoforu/llama-3-12b-instruct"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype=torch.bfloat16,
    device_map="auto",
)

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

input_ids = tokenizer.apply_chat_template(
    messages,
    add_generation_prompt=True,
    return_tensors="pt"
).to(model.device)

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

outputs = model.generate(
    input_ids,
    max_new_tokens=256,
    eos_token_id=terminators,
    do_sample=True,
    temperature=0.6,
    top_p=0.9,
)
response = outputs[0][input_ids.shape[-1]:]
print(tokenizer.decode(response, skip_special_tokens=True))

About merge

Base model: Meta-Llama-3-8B-Instruct

Merge models:

• Muhammad2003/Llama3-8B-OpenHermes-DPO
• IlyaGusev/saiga_llama3_8b
• NousResearch/Meta-Llama-3-8B-Instruct
• abacusai/Llama-3-Smaug-8B
• vicgalle/Configurable-Llama-3-8B-v0.2
• cognitivecomputations/dolphin-2.9-llama3-8b
• NeuralNovel/Llama-3-NeuralPaca-8b

Merge datasets:

• mlabonne/chatml-OpenHermes2.5-dpo-binarized-alpha
• tatsu-lab/alpaca
• vicgalle/configurable-system-prompt-multitask
• IlyaGusev/ru_turbo_saiga
• IlyaGusev/ru_sharegpt_cleaned
• IlyaGusev/oasst1_ru_main_branch
• IlyaGusev/gpt_roleplay_realm
• lksy/ru_instruct_gpt4