Edit model card

Introduction

Nxcode-CQ-7B-orpo is an Monolithic Preference Optimization without Reference Model fine-tune of Qwen/CodeQwen1.5-7B on 100k samples of high-quality ranking data.

Evalplus

EvalPlus pass@1
HumanEval 86.6
HumanEval+ 83.5
MBPP(v0.2.0) 82.3
MBPP+(v0.2.0) 70.4

We use a simple template to generate the solution for evalplus:

"Complete the following Python function:\n{prompt}"

Evalplus Leaderboard

Models HumanEval HumanEval+
GPT-4-Turbo (April 2024) 90.2 86.6
GPT-4 (May 2023) 88.4 81.17
GPT-4-Turbo (Nov 2023) 85.4 79.3
CodeQwen1.5-7B-Chat 83.5 78.7
claude-3-opus (Mar 2024) 82.9 76.8
DeepSeek-Coder-33B-instruct 81.1 75.0
WizardCoder-33B-V1.1 79.9 73.2
OpenCodeInterpreter-DS-33B 79.3 73.8
speechless-codellama-34B-v2.0 77.4 72
GPT-3.5-Turbo (Nov 2023) 76.8 70.7
Llama3-70B-instruct 76.2 70.7

Bigcode Leaderboard

Bigcode Leaderboard

09/05/2024

Top 1 average score.

Top 2 winrate.

image/png

Quickstart

Here provides a code snippet with apply_chat_template to show you how to load the tokenizer and model and how to generate contents. You should upgrade the transformers if you receive an error when loading the tokenizer

from transformers import AutoModelForCausalLM, AutoTokenizer
device = "cuda" # the device to load the model onto

model = AutoModelForCausalLM.from_pretrained(
    "NTQAI/Nxcode-CQ-7B-orpo",
    torch_dtype="auto",
    device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained("NTQAI/Nxcode-CQ-7B-orpo")

prompt = """Complete the following Python function:
from typing import List


def has_close_elements(numbers: List[float], threshold: float) -> bool:
    """ Check if in given list of numbers, are any two numbers closer to each other than
    given threshold.
    >>> has_close_elements([1.0, 2.0, 3.0], 0.5)
    False
    >>> has_close_elements([1.0, 2.8, 3.0, 4.0, 5.0, 2.0], 0.3)
    True
    """
"""
messages = [
    {"role": "user", "content": prompt}
]

inputs = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt").to(model.device)
outputs = model.generate(inputs, max_new_tokens=512, do_sample=False, top_k=50, top_p=0.95, num_return_sequences=1, eos_token_id=tokenizer.eos_token_id)
res = tokenizer.decode(outputs[0][len(inputs[0]):], skip_special_tokens=True)

Contact information

For personal communication related to this project, please contact Nha Nguyen Van (nha.nguyen@ntq-solution.com.vn).

Downloads last month
21,672
Safetensors
Model size
7.25B params
Tensor type
BF16
Β·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for NTQAI/Nxcode-CQ-7B-orpo

Adapters
2 models
Merges
1 model
Quantizations
11 models

Spaces using NTQAI/Nxcode-CQ-7B-orpo 4

Collection including NTQAI/Nxcode-CQ-7B-orpo