本项目从零开始构建并优化了一个千万参数级别的大规模预训练语言模型，涵盖预训练、有监督微调（SFT）和R1推理蒸馏三个阶段。项目采用自定义Transformer架构（包括RMSNorm、分组注意力、多Query机制、SwiGLU激活和RoPE位置编码），实现高效的长文本处理和自回归生成。同时，开发了基于BBPE编码的分词器，针对中文进行了优化，构建了轻量级词表。预训练阶段通过混合精度训练、梯度累积和Cosine Annealing学习率调度，实现全流程训练；微调阶段则利用开源SFT数据集，通过特殊损失掩码优化指令遵循；R1推理蒸馏阶段采用Deepseek-R1数据及损失加权策略，赋予模型逐步推理（慢思考）的能力，显著提升复杂任务的回答质量。
☆23Mar 10, 2025Updated last year
Alternatives and similar repositories for WeakWater-30M
Users that are interested in WeakWater-30M are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- [KDD 2026] Official implementation of "FaST: Efficient and Effective Long-Horizon Forecasting for Large-Scale Spatial-Temporal Graphs via…☆20Jun 1, 2026Updated 3 months ago
- FastTrack4LLM 是一个为大模型学习者准备的大模型学习与实践框架，帮助他们轻松掌握大模型的核心原理与训练流程，让每个人都能真正理解大模型的内部机制。本项目不仅完整复现了 LLaMA、Qwen、DeepSeek 等主流开源大模型架构，还覆盖了大模型的全生命周期：To…☆31Nov 6, 2025Updated 10 months ago
- BUPT神经网络与深度学习课设☆10Dec 29, 2023Updated 2 years ago
- Load Tensorflow pb file using Bert/TextCNNs, an ensemble model using Java.☆10Aug 20, 2021Updated 5 years ago
- 最基本最小白的自然语言处理入门读物，基于deepseek-r1，涵盖了传统NLP和现代大模型☆30Jan 16, 2026Updated 8 months ago
- Virtual machines for every use case on DigitalOcean • AdGet dependable uptime with 99.99% SLA, simple security tools, and predictable monthly pricing with DigitalOcean's virtual machines, called Droplets.
- 本项目对Deepseek-R1-Distill-Qwen-7B进行心理咨询CoT数据的LoRA微调，以进一步提升Deepseek-R1-Distill-Qwen-7B在心理咨询领域的慢思考能力。☆12Mar 11, 2025Updated last year
- 一个基于transformers的自定义命名实体识别模型示例☆17Jul 4, 2021Updated 5 years ago
- This repository includes code and materials for the paper "Efficient PRM Training Data Synthesis via Formal Verification" (ACL 2026 Findi…☆20Apr 7, 2026Updated 5 months ago
- Paris multilayer transport network☆11Sep 13, 2021Updated 5 years ago
- ☆15Aug 31, 2025Updated last year
- 使用django对情感分析功能进行封装，里面包含使用情感词典和Bert模型进行情感分类，最后可以使用tensorFlow serving将模型部署在docker中运行。☆13Sep 23, 2019Updated 7 years ago
- Unofficial implementation of the Ask-LLM paper 'How to Train Data-Efficient LLMs', arXiv:2402.09668.☆12Jun 19, 2024Updated 2 years ago
- 这是一个可通过网页远程登录管理、可接入讯飞星火、ChatGPT等大语言模型的微信聊天机器人，使用微信网页版协议。☆16Feb 20, 2024Updated 2 years ago
- NewsApp包含客户端源码、服务端源码、数据库文件。 基于Miscrosoft人工智能项目ProjectOxford中的Recognition Emotion做的， 主要是基于用户的面部表情来推送不同类别的新闻。 Emotion API可以参考：https://www.p…☆10Mar 2, 2016Updated 10 years ago
- Managed hosting for WordPress and PHP on Cloudways • AdManaged hosting for WordPress, Magento, Laravel, or PHP apps, on multiple cloud providers. Deploy in minutes on Cloudways by DigitalOcean.
- A Real-Time Network Intrusion Detection System based on Transformer-LSTM model (Major Project)☆16May 27, 2024Updated 2 years ago
- 使用Sentencepiece对中文语料进行分词☆13Nov 30, 2023Updated 2 years ago
- An android app which converts text/voice input to American Sign Language(ASL)☆13Sep 8, 2016Updated 10 years ago
- EasyTTS是一个便捷的工具，旨在方便地使用第三方API服务来调用OpenAI的文本转语音（TTS）功能。 EasyTTS允许用户输入文本，并选择不同的模型、音色、格式来生成音频文件。☆10Nov 26, 2023Updated 2 years ago
- BUPT智能计算系统☆22Jan 2, 2024Updated 2 years ago
- 「城语」APP基于A级景区、历史古迹、文物保护单位等基础数据，利用先进的大模型能力实现智能化的Citywalk 路线规划，包括设计一条路线、生成路线攻略、生成景点的推荐理由等三大核心功能；利用大模型减少了人工编辑和推荐的工作量，并可以根据游客的需求进行个性化定制，提升了游客…☆19Feb 20, 2024Updated 2 years ago
- ☆19Sep 14, 2026Updated last week
- 同济大学计科机器学习大作业☆10Mar 22, 2025Updated last year
- Chinese Word Segmentation task based on BERT and implemented in Pytorch☆14Aug 14, 2020Updated 6 years ago
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- 基于 BPE 实现的中文分词。优化：预处理，并行计算，多字词，多词表☆14May 14, 2022Updated 4 years ago
- 基于MFCC特征构建单核GMM的0-9独立词语音识别，MFCC，GMM，sklearn，Isolated word recognition。☆10Nov 18, 2020Updated 5 years ago
- Classification and aggregation of russian news articles. University coursework.☆18Jan 21, 2019Updated 7 years ago
- Identify the type of news based on headlines and short descriptions☆16Mar 25, 2019Updated 7 years ago
- 本项目利用医学领域的 CoT 数据对 Deepseek-R1-Distill-Qwen-7B 进行微调，通过 QLoRA 量化和 Unsloth 加速训练，显著提升模型在复杂医学推理任务中的慢思考能力。知识蒸馏技术使轻量级模型获得大模型的推理优势，实现高效、准确且具有解释性…☆48Mar 10, 2025Updated last year
- ☆15Oct 19, 2025Updated 11 months ago
- 基于CNN的新浪新闻文本分类☆11Jul 22, 2019Updated 7 years ago
- [ICCV'25] Sparfels: Fast Reconstruction from Sparse Unposed Imagery☆16Jan 18, 2026Updated 8 months ago
- ADNI resting-state fMRI analysis scripts☆14Jul 30, 2015Updated 11 years ago
- 1-Click AI Models by DigitalOcean Gradient • AdDeploy popular AI models on DigitalOcean Gradient GPU virtual machines with just a single click. Zero configuration with optimized deployments.
- A Tight-fisted Optimizer (Tiger), implemented in PyTorch.☆12Jun 26, 2024Updated 2 years ago
- Testing the performance of CNN and BERT embeddings on GLUE tasks☆15Mar 24, 2023Updated 3 years ago
- ☆15Jul 17, 2025Updated last year
- Partitioning a large Embedding LUT into multiple devices. (for Pytorch)☆16Apr 30, 2017Updated 9 years ago
- Personalized PageRank (PPR) on GraphLab PowerGraph☆15Jan 3, 2017Updated 9 years ago
- 基于大模型ChatGLM，微调方式为LORA，集SFT、RM、PPO算法为一体项目☆14Jun 20, 2023Updated 3 years ago
- 监控哔哩哔哩直播间数据，实时保存至数据库，并在内置网页上查看精致的可视化统计图表。☆12Jan 4, 2022Updated 4 years ago