metadata

language:
  - en
license: other
tags:
  - chat
license_name: tongyi-qianwen
license_link: https://huggingface.co/Qwen/Qwen2.5-72B-Instruct/blob/main/LICENSE
pipeline_tag: text-generation
library_name: transformers

Dracarys2-72B-Instruct

Introduction

We introduce the latest in the Smaug series, the Dracarys family of finetunes targeting coding performance improvements across a variety of base models.

This variant is a finetune of Qwen2.5-72B-Instruct

Compared to Qwen2.5-72B-Instruct, Dracarys has better LiveCodeBench scores (see evaluation results below).

Model Description

Developed by: Abacus.AI
License: https://huggingface.co/Qwen/Qwen2.5-72B-Instruct/blob/main/LICENSE
Finetuned from model: Qwen2.5-72B-Instruct.

How to use

The prompt format is unchanged from Qwen2.5-72B-Instruct (see evaluations for prompt details for LCB)

Use with transformers

See the snippet below for usage with Transformers:

import transformers
import torch

model_id = "abacusai/Dracarys2-72B-Instruct"

pipeline = transformers.pipeline(
    "text-generation",
    model=model_id,
    model_kwargs={"torch_dtype": torch.bfloat16},
    device_map="auto",
)

messages = [
    {"role": "system", "content": "You are data science coding assistant that generates Python code using Pandas and Numpy."},
    {"role": "user", "content": "Write code to select rows from the dataframe `df` having the maximum `temp` for each `city`"},
]

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

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

outputs = pipeline(
    prompt,
    max_new_tokens=256,
    eos_token_id=terminators,
    do_sample=True,
    temperature=0.6,
    top_p=0.9,
)
print(outputs[0]["generated_text"][len(prompt):])

Evaluation Results

LiveCodeBench

Model	Code Generation	Code Execution (COT)	Test Output Prediction
Dracarys2-72B-Instruct	53.80	89.12	59.61
Qwen2.5-72B-Instruct	53.03	88.72	46.28

Breakdown of LiveCodeBench CodeGeneration

Model	Easy	Medium	Hard
Dracarys2-72B-Instruct	88.79	50.28	9.47
Qwen2.5-72B-Instruct	86.99	49.59	9.99

Breakdown of LiveCodeBench TestOutputPrediction

Model	Easy	Medium	Hard
Dracarys2-72B-Instruct	79.25	53.76	37.63
Qwen2.5-72B-Instruct	68.43	39.46	22.22