Acereason Nemotron 14B GGUF
AceReason-Nemotron-14B是一個通過強化學習訓練的數學和代碼推理模型,在多個數學和代碼推理基準測試中表現出色。
下載量 326
發布時間 : 6/14/2025
模型概述
該模型專注於數學和代碼推理任務,通過強化學習訓練,在數學和編程問題解決方面具有優異性能。
模型特點
強化學習訓練
完全通過強化學習(RL)訓練的數學和代碼推理模型
系統性研究
通過廣泛的消融實驗系統地研究了RL訓練過程
性能提升
在數學和代碼推理基準測試中取得了出色的成績
分階段訓練
先在僅含數學的提示上進行RL訓練,然後在僅含代碼的提示上進行RL訓練
模型能力
數學問題解答
代碼生成
逐步推理
複雜問題解決
使用案例
教育
數學競賽題解答
解決高級數學競賽題目,如AIME競賽題
在AIME 2024和2025競賽題上表現優異
編程
代碼問題解決
根據問題描述生成Python代碼解決方案
在LiveCodeBench基準測試中表現良好
🚀 QuantFactory/AceReason-Nemotron-14B-GGUF
這是使用llama.cpp創建的nvidia/AceReason-Nemotron-14B的量化版本,可用於文本生成。
🚀 快速開始
本項目是基於llama.cpp對原始模型進行量化處理得到的版本,可用於文本生成任務。下面將為你介紹模型的相關信息、使用方法等內容。
✨ 主要特性
- 強化學習訓練:AceReason-Nemotron-14B是一個完全通過強化學習(RL)訓練的數學和代碼推理模型,從DeepSeek-R1-Distilled-Qwen-14B開始訓練,在多個數學和代碼推理基準測試中取得了出色的成績。
- 系統性研究:通過廣泛的消融實驗系統地研究了RL訓練過程,並提出了一種簡單而有效的方法:先在僅含數學的提示上進行RL訓練,然後在僅含代碼的提示上進行RL訓練。
- 性能提升:研究發現,僅含數學的RL不僅顯著提高了強蒸餾模型在數學基準測試上的性能,還提升了代碼推理任務的性能;擴展的僅含代碼的RL進一步提高了代碼基準測試的性能,同時對數學結果的影響極小。
📦 安裝指南
文檔未提及具體安裝步驟,故跳過。
💻 使用示例
基礎用法
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = 'nvidia/AceReason-Nemotron-14B'
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype="auto", device_map="auto")
prompt = "Jen enters a lottery by picking $4$ distinct numbers from $S=\\{1,2,3,\\cdots,9,10\\}.$ $4$ numbers are randomly chosen from $S.$ She wins a prize if at least two of her numbers were $2$ of the randomly chosen numbers, and wins the grand prize if all four of her numbers were the randomly chosen numbers. The probability of her winning the grand prize given that she won a prize is $\\tfrac{m}{n}$ where $m$ and $n$ are relatively prime positive integers. Find $m+n$."
messages = [{"role": "user", "content": prompt}]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to("cuda")
generated_ids = model.generate(
**model_inputs,
max_new_tokens=32768,
temperature=0.6,
top_p=0.95
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
高級用法
# 以下是針對不同類型問題的使用建議代碼
# 1. 數學問題建議:不包含系統提示,將所有指令直接放在用戶提示中,並使用特定格式要求答案
# 2. 代碼問題建議:根據是否有起始代碼,構建不同的提示信息
# 數學問題指令
math_instruction = "Please reason step by step, and put your final answer within \\boxed{}."
# 代碼問題指令
question = "" # code question
starter_code = "" # starter code function header
code_instruction_nostartercode = """Write Python code to solve the problem. Please place the solution code in the following format:\n```python\n# Your solution code here\n```"""
code_instruction_hasstartercode = """Please place the solution code in the following format:\n```python\n# Your solution code here\n```"""
if starter_code != "":
question += "\n\n" + "Solve the problem starting with the provided function header.\n\nFunction header:\n" + "```\n" + starter_code + "\n```"
question += "\n\n" + code_instruction_hasstartercode
else:
question += "\n\n" + code_instruction_nostartercode
final_prompt = "<|User|>" + question + "<|Assistant|><think>\n"
# 推理引擎配置
# 我們用於評估的推理引擎是 **vLLM==0.7.3**,使用 top-p=0.95,temperature=0.6,max_tokens=32768
📚 詳細文檔
最新消息
- 2025年6月11日:我們在AceReason評估中分享了評估工具包,包括:
- 運行推理和評分的腳本
- LiveCodeBench (avg@8):每個月(2023/5 - 2025/5)的模型預測文件和分數
- AIME24/25 (avg@64):模型預測文件和分數
- 2025年6月2日:我們很高興在AceReason-Math上分享我們的數學強化學習訓練數據集。
評估結果
我們在AIME 2024、AIME 2025、LiveCodeBench v5(2024/08/01 - 2025/02/01)和LiveCodeBench v6(2025/02/01 - 2025/05/01)上,將我們的模型與Qwen2.5和Llama3.1模型家族中可比大小的競爭推理模型進行了評估。更多評估結果可在我們的技術報告中找到。
模型 | AIME 2024 (avg@64) |
AIME 2025 (avg@64) |
LCB v5 (avg@8) |
LCB v6 (avg@8) |
---|---|---|---|---|
QwQ-32B | 79.5 | 65.8 | 63.4 | - |
DeepSeek-R1-671B | 79.8 | 70.0 | 65.9 | - |
Llama-Nemotron-Ultra-253B | 80.8 | 72.5 | 66.3 | - |
o3-mini (medium) | 79.6 | 76.7 | 67.4 | - |
Light-R1-14B | 74 | 60.2 | 57.9 | 51.5 |
DeepCoder-14B (32K Inference) | 71 | 56.1 | 57.9 | 50.4 |
OpenMath-Nemotron-14B | 76.3 | 63.0 | - | - |
OpenCodeReasoning-Nemotron-14B | - | - | 59.4 | 54.1 |
Llama-Nemotron-Super-49B-v1 | 67.5 | 60.0 | 45.5 | - |
DeepSeek-R1-Distilled-Qwen-14B | 69.7 | 50.2 | 53.1 | 47.9 |
DeepSeek-R1-Distilled-Qwen-32B | 72.6 | 54.9 | 57.2 | - |
AceReason-Nemotron-7B 🤖 | 69.0 | 53.6 | 51.8 | 44.1 |
AceReason-Nemotron-14B 🤖 | 78.6 | 67.4 | 61.1 | 54.9 |
評估工具包
請查看https://huggingface.co/nvidia/AceReason-Nemotron-14B/blob/main/README_EVALUATION.md 中的評估代碼、腳本和緩存的預測文件。
聯繫方式
- Yang Chen (yachen@nvidia.com)
- Zhuolin Yang (zhuoliny@nvidia.com)
- Zihan Liu (zihanl@nvidia.com)
- Chankyu Lee (chankyul@nvidia.com)
- Wei Ping (wping@nvidia.com)
🔧 技術細節
文檔未提供具體技術實現細節,故跳過。
📄 許可證
你對該模型的使用受NVIDIA開放模型許可證的約束。
📚 引用
@article{chen2025acereason,
title={AceReason-Nemotron: Advancing Math and Code Reasoning through Reinforcement Learning},
author={Chen, Yang and Yang, Zhuolin and Liu, Zihan and Lee, Chankyu and Xu, Peng and Shoeybi, Mohammad and Catanzaro, Bryan and Ping, Wei},
journal={arXiv preprint arXiv:2505.16400},
year={2025}
}
💡 使用建議
- 不包含系統提示,將所有指令直接放在用戶提示中。
- 對於數學問題,建議使用以下指令:請逐步推理,並將最終答案放在\boxed{}內。
- 對於代碼問題,根據是否有起始代碼,構建不同的提示信息。
- 我們用於評估的推理引擎是 vLLM==0.7.3,使用 top-p=0.95,temperature=0.6,max_tokens=32768。
Phi 2 GGUF
其他
Phi-2是微軟開發的一個小型但強大的語言模型,具有27億參數,專注於高效推理和高質量文本生成。
大型語言模型 支持多種語言
P
TheBloke
41.5M
205
Roberta Large
MIT
基於掩碼語言建模目標預訓練的大型英語語言模型,採用改進的BERT訓練方法
大型語言模型 英語
R
FacebookAI
19.4M
212
Distilbert Base Uncased
Apache-2.0
DistilBERT是BERT基礎模型的蒸餾版本,在保持相近性能的同時更輕量高效,適用於序列分類、標記分類等自然語言處理任務。
大型語言模型 英語
D
distilbert
11.1M
669
Llama 3.1 8B Instruct GGUF
Meta Llama 3.1 8B Instruct 是一個多語言大語言模型,針對多語言對話用例進行了優化,在常見的行業基準測試中表現優異。
大型語言模型 英語
L
modularai
9.7M
4
Xlm Roberta Base
MIT
XLM-RoBERTa是基於100種語言的2.5TB過濾CommonCrawl數據預訓練的多語言模型,採用掩碼語言建模目標進行訓練。
大型語言模型 支持多種語言
X
FacebookAI
9.6M
664
Roberta Base
MIT
基於Transformer架構的英語預訓練模型,通過掩碼語言建模目標在海量文本上訓練,支持文本特徵提取和下游任務微調
大型語言模型 英語
R
FacebookAI
9.3M
488
Opt 125m
其他
OPT是由Meta AI發佈的開放預訓練Transformer語言模型套件,參數量從1.25億到1750億,旨在對標GPT-3系列性能,同時促進大規模語言模型的開放研究。
大型語言模型 英語
O
facebook
6.3M
198
1
基於transformers庫的預訓練模型,適用於多種NLP任務
大型語言模型
Transformers

1
unslothai
6.2M
1
Llama 3.1 8B Instruct
Llama 3.1是Meta推出的多語言大語言模型系列,包含8B、70B和405B參數規模,支持8種語言和代碼生成,優化了多語言對話場景。
大型語言模型
Transformers 支持多種語言

L
meta-llama
5.7M
3,898
T5 Base
Apache-2.0
T5基礎版是由Google開發的文本到文本轉換Transformer模型,參數規模2.2億,支持多語言NLP任務。
大型語言模型 支持多種語言
T
google-t5
5.4M
702
精選推薦AI模型
Llama 3 Typhoon V1.5x 8b Instruct
專為泰語設計的80億參數指令模型,性能媲美GPT-3.5-turbo,優化了應用場景、檢索增強生成、受限生成和推理任務
大型語言模型
Transformers 支持多種語言

L
scb10x
3,269
16
Cadet Tiny
Openrail
Cadet-Tiny是一個基於SODA數據集訓練的超小型對話模型,專為邊緣設備推理設計,體積僅為Cosmo-3B模型的2%左右。
對話系統
Transformers 英語

C
ToddGoldfarb
2,691
6
Roberta Base Chinese Extractive Qa
基於RoBERTa架構的中文抽取式問答模型,適用於從給定文本中提取答案的任務。
問答系統 中文
R
uer
2,694
98