Ascend Python Packages

nndeploy

一款简单易用和高性能的AI部署框架 | An Easy-to-Use and High-Performance AI Deployment Framework

uc-manager-cuda-cu129

Persist and reuse KV Cache to speedup your LLM.

uc-manager-cann910-a3

Persist and reuse KV Cache to speedup your LLM.

uc-manager-cann910-a2

Persist and reuse KV Cache to speedup your LLM.

uc-manager-cuda-cu130

Persist and reuse KV Cache to speedup your LLM.

uc-manager-cann901-a2

Persist and reuse KV Cache to speedup your LLM.

uc-manager-cann901-a3

Persist and reuse KV Cache to speedup your LLM.

torch-rechub

A Lighting Pytorch Framework for Recommendation Models, Easy-to-use and Easy-to-extend.

pocketllm

Inference engine for large language models on consumer GPUs, with deep optimization for older hardware

ascend-nputop

An interactive Ascend-NPU process viewer

uc-manager-cann910-a5

Persist and reuse KV Cache to speedup your LLM.

llm-cal

LLM inference hardware calculator — architecture-aware (MLA/NSA/MoE), engine-aware (vLLM/SGLang), honest-labeled. Reads real safetensors bytes; supports 53 GPUs (NVIDIA / AMD / Huawei Ascend / 沐曦 / 昆仑芯 / 壁仞 / 寒武纪 / 海光).

uc-manager-cann901-a5

Persist and reuse KV Cache to speedup your LLM.

activation-hard

cubestudio开源云原生一站式机器学习/深度学习/大模型AI平台/MaaS/mlops/人工智能平台/训推平台，算法全链路流程，多租户，算力租赁平台，token中转，拖拉拽任务流pipeline编排，多机多卡分布式训练，超参搜索，推理服务，VGPU虚拟化，云边端协同，边缘计算，自动化标注平台，deepseek等大模型sft微调/奖励模型/强化学习训练，vllm/ollama/mindie大模型多机推理，私有知识库llmops智能体，AI模型市场，支持国产异构算力调度,昇腾/寒武纪/海光/摩尔/沐曦等，支持ib/roce/RDMA，信创支持