Starred repositories
verl: Volcano Engine Reinforcement Learning for LLMs
Implement a ChatGPT-like LLM in PyTorch from scratch, step by step
Implementing DeepSeek R1's GRPO algorithm from scratch
DeepSiteForOpenAI 是一个灵活的开发工具，它将 DeepSite 的强大功能与 OpenAI 接口无缝集成，支持自定义接口，使用openai 风格的接口为开发者提供了一个高效、智能的编程环境。这个工具允许用户通过自然语言描述来生成代码，实现"氛围编程"（Vibe coding）体验，让编程变得更加直观和高效。
mirror of https://huggingface.co/spaces/enzostvs/deepsite
A high-throughput and memory-efficient inference and serving engine for LLMs
Official inference framework for 1-bit LLMs
Tool to download models from Huggingface Hub and convert them to GGML/GGUF for llama.cpp
Get up and running with Llama 3.3, DeepSeek-R1, Phi-4, Gemma 3, Mistral Small 3.1 and other large language models.
Is ChatGPT Good at Search? LLMs as Re-Ranking Agent [EMNLP 2023 Outstanding Paper Award]
A blazing fast inference solution for text embeddings models
🌐 Make websites accessible for AI agents. Automate tasks online with ease.
No fortress, purely open ground. OpenManus is Coming.
Ray is an AI compute engine. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.
Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)
Flax is a neural network library for JAX that is designed for flexibility.
HIP: C++ Heterogeneous-Compute Interface for Portability
Play ChatGPT and other LLM with Xiaomi AI Speaker
idootop / mi-service-lite
Forked from inu1255/mi-serviceNode.js client for XiaoMi Cloud Service
PyTorch original implementation of Cross-lingual Language Model Pretraining.
The official repo of Qwen (通义千问) chat & pretrained large language model proposed by Alibaba Cloud.