社区博客与文章

Community Articles

使用 PyTorch 完全分片数据并行技术加速大模型训练

基于 Habana Gaudi 的 Transformers 入门

〜不要〜重复自己

在 🤗 Transformers 中使用约束波束搜索引导文本生成

大语言模型：新的摩尔定律？

机器学习即代码的时代已经到来

深入理解 BigBird 的块稀疏注意力

长程 transformer 模型

如何成功将 🤗 API 客户的 transformer 模型推理速度加快 100 倍

基于 Transformers 的编码器-解码器模型

Reformer 模型 - 突破语言建模的极限

如何生成文本：通过 Transformers 用不同的解码方法生成文本

NEW Articles from Team or Enterprise organizations will get promoted to the main section.

社区博客与文章

Tensor Parallelism (TP) in Transformers: 5 Minutes to Understand

TurkColBERT: A Benchmark of Dense and Late-Interaction Models for Turkish Information Retrieval

Norm-Preserving Biprojected Abliteration

Ellora: Enhancing LLMs with LoRA - Standardized Recipes for Capability Enhancement

**An Edge-First Generalized LLM LoRA Fine-Tuning Framework for Heterogeneous GPUs**

Uncensor any LLM with abliteration

Code a simple RAG from scratch

AI Energy Score v2: Refreshed Leaderboard, now with Reasoning 🧠

Building Jobly: Semantic Job Matching with RAG and Vector Embeddings

Making Model Tuning Accessible: This is what we built observing 100s of users tune models!

DeepFabric: Generate, Train and Evaluate with Datasets curated for Model Behavior Training.

Gemini-3 Benchmarkathon

Engineering Notes: Training a LoRA for Z-Image Turbo with the Ostris AI Toolkit

Mastering Tensor Dimensions in Transformers

Curating datasets directly on the Hub

Building and evaluating Multimodal Rerankers

KV Caching Explained: Optimizing Transformer Inference Efficiency

A Review on the Evolvement of Load Balancing Strategy in MoE LLMs: Pitfalls and Lessons

From GRPO to DAPO and GSPO: What, Why, and How

GSMA Open-Telco LLM Benchmarks 2.0: The first dedicated LLM Evaluation for Telecoms

使用 PyTorch 完全分片数据并行技术加速大模型训练

基于 Habana Gaudi 的 Transformers 入门

〜不要〜重复自己

在 🤗 Transformers 中使用约束波束搜索引导文本生成

大语言模型：新的摩尔定律？

机器学习即代码的时代已经到来

深入理解 BigBird 的块稀疏注意力

长程 transformer 模型

如何成功将 🤗 API 客户的 transformer 模型推理速度加快 100 倍

基于 Transformers 的编码器-解码器模型

Reformer 模型 - 突破语言建模的极限

如何生成文本：通过 Transformers 用不同的解码方法生成文本

Tensor Parallelism (TP) in Transformers: 5 Minutes to Understand

TurkColBERT: A Benchmark of Dense and Late-Interaction Models for Turkish Information Retrieval

Norm-Preserving Biprojected Abliteration

Ellora: Enhancing LLMs with LoRA - Standardized Recipes for Capability Enhancement

**An Edge-First Generalized LLM LoRA Fine-Tuning Framework for Heterogeneous GPUs**

Uncensor any LLM with abliteration

Code a simple RAG from scratch

AI Energy Score v2: Refreshed Leaderboard, now with Reasoning 🧠

Building Jobly: Semantic Job Matching with RAG and Vector Embeddings

Making Model Tuning Accessible: This is what we built observing 100s of users tune models!

DeepFabric: Generate, Train and Evaluate with Datasets curated for Model Behavior Training.

Gemini-3 Benchmarkathon

Engineering Notes: Training a LoRA for Z-Image Turbo with the Ostris AI Toolkit

Mastering Tensor Dimensions in Transformers

Curating datasets directly on the Hub

Building and evaluating Multimodal Rerankers

KV Caching Explained: Optimizing Transformer Inference Efficiency

A Review on the Evolvement of Load Balancing Strategy in MoE LLMs: Pitfalls and Lessons

From GRPO to DAPO and GSPO: What, Why, and How

GSMA Open-Telco LLM Benchmarks 2.0: The first dedicated LLM Evaluation for Telecoms

An Edge-First Generalized LLM LoRA Fine-Tuning Framework for Heterogeneous GPUs

An Edge-First Generalized LLM LoRA Fine-Tuning Framework for Heterogeneous GPUs