“sdp”的搜索结果

Install Triton + SageAttention to accelerate ComfyUI (the sageattn attention_mode and inductor torch.compile used by WanVideoWrapper / many video graphs). Windows-first (triton-windows + woct0rdho prebuilt SageAttention wheels matched to torch/CUDA/python into the RIGHT python), plus Linux (official triton + build) and Mac (N/A → sdpa/MPS). Also covers the SAFE sdpa / no-compile fallback so an example that assumes sageattn + torch.compile still runs when these aren't installed (video-extend TRAP 5). Use when a loader crashes with "No module named 'sageattention'" or reports triton unavailable, when asked to speed up Wan/video workflows, or when deciding whether to install acceleration vs. fall back.

中文简介正在准备中。当前显示原文说明。

Fornece treinamento distribuído de LLM nativo do PyTorch usando torchtitan com paralelismo 4D (FSDP2, TP, PP, CP). Use para fazer pré-treinamento de Llama 3.1, DeepSeek V3 ou modelos customizados em escala de 8 a 512+ GPUs com Float8, torch.compile e checkpoint distribuído.

中文简介正在准备中。当前显示原文说明。

Otimiza atenção em transformers com Flash Attention para ganho de 2-4x em velocidade e redução de 10-20x em memória. Use ao treinar/executar transformers com sequências longas (>512 tokens), ao encontrar problemas de memória GPU com atenção, ou quando precisa de inferência mais rápida. Suporta SDPA nativo do PyTorch, biblioteca flash-attn, H100 FP8 e sliding window attention.

中文简介正在准备中。当前显示原文说明。

API de treinamento distribuído mais simples. 4 linhas para adicionar suporte distribuído a qualquer script PyTorch. API unificada para DeepSpeed/FSDP/Megatron/DDP. Posicionamento automático de device, precisão mista (FP16/BF16/FP8). Config interativo, comando de launch único. Padrão do ecossistema HuggingFace.

中文简介正在准备中。当前显示原文说明。

Framework PyTorch de alto nível com classe Trainer, treinamento distribuído automático (DDP/FSDP/DeepSpeed), sistema de callbacks e boilerplate mínimo. Escala de laptop para supercomputador com o mesmo código. Use quando quiser loops de treinamento limpos com boas práticas integradas.

中文简介正在准备中。当前显示原文说明。

Deploy a Software-Defined Perimeter using the CSA v2.0 specification with Single Packet Authorization, mutual TLS, and SDP controller/gateway configuration to enforce zero trust network access.

中文简介正在准备中。当前显示原文说明。

Deploy a Software-Defined Perimeter using the CSA v2.0 specification with Single Packet Authorization, mutual TLS, and SDP controller/gateway configuration to enforce zero trust network access.

中文简介正在准备中。当前显示原文说明。

High-level PyTorch framework with Trainer class, automatic distributed training (DDP/FSDP/DeepSpeed), callbacks system, and minimal boilerplate. Scales from laptop to supercomputer with same code. Use when you want clean training loops with built-in best practices.

中文简介正在准备中。当前显示原文说明。

Simplest distributed training API. 4 lines to add distributed support to any PyTorch script. Unified API for DeepSpeed/FSDP/Megatron/DDP. Automatic device placement, mixed precision (FP16/BF16/FP8). Interactive config, single launch command. HuggingFace ecosystem standard.

中文简介正在准备中。当前显示原文说明。

Provides PyTorch-native distributed LLM pretraining using torchtitan with 4D parallelism (FSDP2, TP, PP, CP). Use when pretraining Llama 3.1, DeepSeek V3, or custom models at scale from 8 to 512+ GPUs with Float8, torch.compile, and distributed checkpointing.

中文简介正在准备中。当前显示原文说明。

High-level PyTorch framework with Trainer class, automatic distributed training (DDP/FSDP/DeepSpeed), callbacks system, and minimal boilerplate. Scales from laptop to supercomputer with same code. Use when you want clean training loops with built-in best practices.

中文简介正在准备中。当前显示原文说明。

Optimizes transformer attention with Flash Attention for 2-4x speedup and 10-20x memory reduction. Use when training/running transformers with long sequences (>512 tokens), encountering GPU memory issues with attention, or need faster inference. Supports PyTorch native SDPA, flash-attn library, H100 FP8, and sliding window attention.

中文简介正在准备中。当前显示原文说明。

Simplest distributed training API. 4 lines to add distributed support to any PyTorch script. Unified API for DeepSpeed/FSDP/Megatron/DDP. Automatic device placement, mixed precision (FP16/BF16/FP8). Interactive config, single launch command. HuggingFace ecosystem standard.

中文简介正在准备中。当前显示原文说明。

Deploy a Software-Defined Perimeter using the CSA v2.0 specification with Single Packet Authorization, mutual TLS, and SDP controller/gateway configuration to enforce zero trust network access.

中文简介正在准备中。当前显示原文说明。

为处理长输入的Transformer模型接入Flash Attention，加快训练与推理中的注意力计算并减少GPU显存占用，涵盖安装、性能测量和输出比较。

• 加快长输入模型的训练
• 减少长文本推理的显存占用
• 比较处理速度与模型输出

Guides LLM fine-tuning with Axolotl, from YAML configuration and dataset formats to distributed training across GPUs. Covers LoRA and QLoRA, methods that train on preferences, multimodal support, and troubleshooting using documentation references and code examples.

中文简介正在准备中。当前显示英语简介。

• Prepare YAML settings for LLM fine-tuning
• Explore LoRA, QLoRA, and preference training options
• Check dataset formats and input handling

Guides LLM training with verl using rewards to improve model responses. Covers data preparation, reward functions, GRPO and PPO configuration, distributed GPU training, and troubleshooting memory shortages or unstable training.

中文简介正在准备中。当前显示英语简介。

• Train math reasoning models with GRPO and answer-based rewards
• Configure PPO training with a separate value-estimation model
• Run large-model training across multiple GPU nodes

Practical guidance for training MoE VLMs in Megatron Bridge. Compares FSDP and 3D-parallel approaches, using rounded lessons from Qwen3-VL, Qwen3-Next, and other multimodal experiments.

中文简介正在准备中。当前显示原文说明。

Set up infrastructure for fine-tuning LLMs with QLoRA, LoRA, and full fine-tuning using Hugging Face TRL, Axolotl, and distributed training with DeepSpeed or FSDP. Covers dataset prep, training runs, and model export.

中文简介正在准备中。当前显示原文说明。

Use Hugging Face Accelerate for PyTorch training-loop migration, distributed launch/configuration, DeepSpeed/FSDP/TPU backend setup, big-model inference/offload, checkpointing, tracking, and troubleshooting.

中文简介正在准备中。当前显示原文说明。

Think and work like an expert Deep Learning Scientist. Use when a task calls for Deep Learning Scientist judgment. Reasons from CNN/Transformer inductive bias, Li et al. loss landscapes, grokking/mode connectivity, and Kaplan/Chinchilla scaling (~20 tokens/param); designs ResNet/ViT/DiT/MoE/FlashAttention stacks with FLOPs-matched ablations; trains AdamW+cosine/WSD via Megatron-FSDP/DeepSpeed; evaluates FID/MMLU-Pro/MMLU-CF with lm-eval decontamination and Pineau/NeurIPS reproducibility checklists.

中文简介正在准备中。当前显示原文说明。

Use when training models across multiple GPUs or nodes, handling large models that don't fit in memory, or optimizing training throughput - covers DDP, FSDP, DeepSpeed ZeRO, model/data parallelism, and gradient checkpointingUse when ", " mentioned.

中文简介正在准备中。当前显示原文说明。

Guides multi-GPU pre-training: DDP, FSDP2, ZeRO, tensor/pipeline/expert parallelism, fp8/Muon. Use when scaling a run, training MoE, or reproducing GPT-2 on rented GPUs.

中文简介正在准备中。当前显示原文说明。

对 msMemScope 显存数据做全链路处理：根据用户的显存问题诉求（OOM、显存泄漏、显存占比拆解、Step 间对比等），指导并帮助用户以最合适的 msMemScope 采集配置完成显存数据采集（支持命令行、Python 接口、mstx 打点三种采集方式，覆盖训练、推理 vLLM-Ascend、FSDP、强化学习 verl 等场景）；对已采集的 memscope_dump_*.csv/db 数据做系统性解读与问题诊断（整体显存曲线趋势分析、显存使用大头按模块/组件拆解分析、显存峰值点各模块与流程占比分析，以及对显存泄漏、OOM 等典型问题的根因定位与原因分析）。

中文简介正在准备中。当前显示原文说明。