CARVIEW |
Select Language
HTTP/2 200
date: Wed, 30 Jul 2025 04:06:58 GMT
content-type: text/html; charset=utf-8
vary: X-PJAX, X-PJAX-Container, Turbo-Visit, Turbo-Frame, X-Requested-With,Accept-Encoding, Accept, X-Requested-With
etag: W/"0174818e5c005f83bc19063c0a054482"
cache-control: max-age=0, private, must-revalidate
strict-transport-security: max-age=31536000; includeSubdomains; preload
x-frame-options: deny
x-content-type-options: nosniff
x-xss-protection: 0
referrer-policy: no-referrer-when-downgrade
content-security-policy: default-src 'none'; base-uri 'self'; child-src github.githubassets.com github.com/assets-cdn/worker/ github.com/assets/ gist.github.com/assets-cdn/worker/; connect-src 'self' uploads.github.com www.githubstatus.com collector.github.com raw.githubusercontent.com api.github.com github-cloud.s3.amazonaws.com github-production-repository-file-5c1aeb.s3.amazonaws.com github-production-upload-manifest-file-7fdce7.s3.amazonaws.com github-production-user-asset-6210df.s3.amazonaws.com *.rel.tunnels.api.visualstudio.com wss://*.rel.tunnels.api.visualstudio.com objects-origin.githubusercontent.com copilot-proxy.githubusercontent.com proxy.individual.githubcopilot.com proxy.business.githubcopilot.com proxy.enterprise.githubcopilot.com *.actions.githubusercontent.com wss://*.actions.githubusercontent.com productionresultssa0.blob.core.windows.net/ productionresultssa1.blob.core.windows.net/ productionresultssa2.blob.core.windows.net/ productionresultssa3.blob.core.windows.net/ productionresultssa4.blob.core.windows.net/ productionresultssa5.blob.core.windows.net/ productionresultssa6.blob.core.windows.net/ productionresultssa7.blob.core.windows.net/ productionresultssa8.blob.core.windows.net/ productionresultssa9.blob.core.windows.net/ productionresultssa10.blob.core.windows.net/ productionresultssa11.blob.core.windows.net/ productionresultssa12.blob.core.windows.net/ productionresultssa13.blob.core.windows.net/ productionresultssa14.blob.core.windows.net/ productionresultssa15.blob.core.windows.net/ productionresultssa16.blob.core.windows.net/ productionresultssa17.blob.core.windows.net/ productionresultssa18.blob.core.windows.net/ productionresultssa19.blob.core.windows.net/ github-production-repository-image-32fea6.s3.amazonaws.com github-production-release-asset-2e65be.s3.amazonaws.com insights.github.com wss://alive.github.com api.githubcopilot.com api.individual.githubcopilot.com api.business.githubcopilot.com api.enterprise.githubcopilot.com; font-src github.githubassets.com; form-action 'self' github.com gist.github.com copilot-workspace.githubnext.com objects-origin.githubusercontent.com; frame-ancestors 'none'; frame-src viewscreen.githubusercontent.com notebooks.githubusercontent.com; img-src 'self' data: blob: github.githubassets.com media.githubusercontent.com camo.githubusercontent.com identicons.github.com avatars.githubusercontent.com private-avatars.githubusercontent.com github-cloud.s3.amazonaws.com objects.githubusercontent.com release-assets.githubusercontent.com secured-user-images.githubusercontent.com/ user-images.githubusercontent.com/ private-user-images.githubusercontent.com opengraph.githubassets.com copilotprodattachments.blob.core.windows.net/github-production-copilot-attachments/ github-production-user-asset-6210df.s3.amazonaws.com customer-stories-feed.github.com spotlights-feed.github.com objects-origin.githubusercontent.com *.githubusercontent.com; manifest-src 'self'; media-src github.com user-images.githubusercontent.com/ secured-user-images.githubusercontent.com/ private-user-images.githubusercontent.com github-production-user-asset-6210df.s3.amazonaws.com gist.github.com; script-src github.githubassets.com; style-src 'unsafe-inline' github.githubassets.com; upgrade-insecure-requests; worker-src github.githubassets.com github.com/assets-cdn/worker/ github.com/assets/ gist.github.com/assets-cdn/worker/
server: github.com
content-encoding: gzip
accept-ranges: bytes
set-cookie: _gh_sess=gm6qW1VCNrDeEoXSikDDmzVbvkoR8PvDNIt3%2Fpo%2BIuAwBIw%2F7y9tntqpPRyjOuNGK0g%2FkS7wJ%2FPBF6EntDpKphsh%2F%2Fz9h25nIv5sKf6xry7p93TY6FgKUdIxl0IPfeNjq0a8leV%2Bj0adA1ymxMZRKo19hDCIkt3yZrpR8VzysPpUYI7H0%2Fu5t3HbGssy7wGwDlc6O2YsX84VlWJkbYaEiG4reFc8ytv54zj8NAG8K4L8BrmTxS11eBI9O0GlEqmldoN8YMeA99ZwlyfDtHd%2FAA%3D%3D--AEFrMJqMguiaQsNK--TF3PjtDVm8NHIjuGhpNAxw%3D%3D; Path=/; HttpOnly; Secure; SameSite=Lax
set-cookie: _octo=GH1.1.1869010144.1753848417; Path=/; Domain=github.com; Expires=Thu, 30 Jul 2026 04:06:57 GMT; Secure; SameSite=Lax
set-cookie: logged_in=no; Path=/; Domain=github.com; Expires=Thu, 30 Jul 2026 04:06:57 GMT; HttpOnly; Secure; SameSite=Lax
x-github-request-id: A44E:A6962:11DA8B1:15C2A1E:68899A61
Releases · PaddlePaddle/PaddleFleetX · GitHub
10 Jan 14:20
Loading
02 Dec 04:22
Loading
29 Nov 08:32
Loading
Skip to content
Navigation Menu
{{ message }}
-
Notifications
You must be signed in to change notification settings - Fork 165
Releases: PaddlePaddle/PaddleFleetX
Releases · PaddlePaddle/PaddleFleetX
PaddleFleetX v2.4.1
362ffe3
This commit was created on GitHub.com and signed with GitHub’s verified signature.
The key has expired.
Compare
- 更新 imagen 相关配置信息
- 发布 PaddleFleetX 2.4 镜像
- 修复 GPT 量化相关 bug
- ViT 和 imagen 的 benchmark 监控修复
Assets 2
PaddleFleetX v2.4.0
2a41090
This commit was created on GitHub.com and signed with GitHub’s verified signature.
The key has expired.
Compare
一、环境部署
- 为提升开发部署用户体验,全面适配了 PaddlePaddle 2.4,并发布了预安装镜像。
二、动态图训练
- 支持gradient accumulation。(#824)
- 修复dataloader int32 overflow的问题。(#818)
- 开源了 MoCo V1、V2 在 Imagenet1K 上的预训练和 linprob 微调代码以及Checkpoint,并达对齐精度
三、自动并行
- 在 345M、1.3B、6.7B 规模上支持 GPT 预训练模型的自动并行分布式训练,还支持了自动混合精度、分组切片、重计算与梯度累计优化策略。(#757 #801)
- 为了支持大模型分布式推理,实现了 GPT 生成模型的自适应转换,包括组网重切分与参数自动转换功能。(#815)
四、推理部署
- 优化GPT生成模型组网逻辑,添加自定义融合算子,减少动转静产生的同步操作,提升推理性能(#946)。
五、性能
- 在345M、1.3B、6.7B与175B模型上支持TensorFuse功能、适配使用FusedLinear、支持selective recompute、支持fp16 embedding。(#620,#626,#634,#635,#752)
- 在6.7B模型上适配sharding stage 2 reduce overlap、适配sharding stage 2 broadcast overlap、适配sharding stage 2多流broadcast。(#799,#812,#833)
- 在175B模型上适配interleave pipeline、适配pipeline recompute interval、支持pipeline非均匀且分的组网方式、支持sequence parallel策略。(#860,#881,#884,#734,#746,#819,#846,#854,#861)
- 相对于同等模型规模的Megatron(DeepSpeed),345M GPT 八卡性能超越竞品 14.2%、1.3B GPT 八卡性能超越竞品5.6%、6.7B GPT 16卡性能超越竞品11.7%、175B GPT 128卡性能超越竞品 0.4%。
六、调试工具
七、模型
Assets 2
PaddleFleetX v2.4.0rc
13b4341
This commit was created on GitHub.com and signed with GitHub’s verified signature.
The key has expired.
Compare
1、环境部署
开发支持包括 Docker/PyPI 等多种二次开发和部署环境,提升使用易用性,可被其他套件或平台安装集成
2、动态图训练
- 开源GPT大模型分布式训练代码及345M模型参数
- 开源了 ViT-B/16 在 Imagenet1K 上的预训练代码以及Checkpoint,并达到谷歌官方ViT公布的精度
- 开源Imagen模型代码,实现 Imagen 397M、2B 文图生成算法以及 256x256、1024x1024 2个超分扩散模型组网、训练、评估和推理功能
3、自动并行
实现GPT『动转静+自动并行』大模型训练,支持常见并行策略、优化策略和两者的任意组合使用,其中并行策略包括数据并行、张量并行、流水线并行和混合并行,优化策略包括重计算、混合精度(1/2/3)、梯度累加、Sharding(1/2/3)
4、推理部署
- 支持动转静模型导出和InferenceEngine推理部署通用能力
- 支持GPT系列模型导出和推理部署
5、量化压缩
- 支持动态图量化训练功能
- GPT-345M模型经过INT8量化,在LAMBDA任务上精度无损。(Baseline Accuracy: 44.17%; INT8量化后 Accuracy: 44.38%)
6、性能
- 训练:GPT-345M模型下,八卡性能超越竞品Megatron-LM 14.2%。GPT-1.3B模型下,八卡性能超越竞品Megatron-LM 5.6%
- 推理:Imagen对齐了 T5-11B 文本推理模型,性能超越 PyTorch 20%。解决 Imagen 1024x1024 长序列超分扩散模型显存占用过大的问题,模型吞吐提升35%
7、调试工具
覆盖包括分布式等多种调试需求,兼容VisualDL可视化工具,提升二次开发体验
Assets 2
You can’t perform that action at this time.