Search results

[tracker] Open add-model PRs

installationInstallation and setup issuesInstallation and setup issues

platformPlatform-specific (Windows/macOS)Platform-specific (Windows/macOS)

scope: attentionAttention backends (VSA, STA, Flash, etc.)Attention backends (VSA, STA, Flash, etc.)

scope: distributedSP, FSDP, USP, multi-nodeSP, FSDP, USP, multi-node

scope: docsDocumentationDocumentation

scope: inferenceInference pipeline, serving, CLIInference pipeline, serving, CLI

scope: modelModel architecture (DiTs, encoders, VAEs)Model architecture (DiTs, encoders, VAEs)

scope: trainingTraining pipeline, methods, configsTraining pipeline, methods, configs

Status: Open.

#1407 In hao-ai-lab/FastVideo;

· H1yori233 opened on May 27, 2026

[bug]: HunyuanVideo-1.5 generation crashes at text_encoding.py:221 — 'list' object has no attribute 'strip'

installationInstallation and setup issuesInstallation and setup issues

performancePerformance and memory issuesPerformance and memory issues

scope: attentionAttention backends (VSA, STA, Flash, etc.)Attention backends (VSA, STA, Flash, etc.)

scope: dataData preprocessing, datasetsData preprocessing, datasets

scope: distributedSP, FSDP, USP, multi-nodeSP, FSDP, USP, multi-node

scope: docsDocumentationDocumentation

scope: inferenceInference pipeline, serving, CLIInference pipeline, serving, CLI

scope: modelModel architecture (DiTs, encoders, VAEs)Model architecture (DiTs, encoders, VAEs)

scope: trainingTraining pipeline, methods, configsTraining pipeline, methods, configs

Status: Open.

#1387 In hao-ai-lab/FastVideo;

· Mister-Raggs opened on May 23, 2026

RMSNorm.forward_native doesn't unwrap DTensor weight under FSDP CPU offload

bugSomething isn't workingSomething isn't working

help wantedExtra attention is neededExtra attention is needed

installationInstallation and setup issuesInstallation and setup issues

scope: attentionAttention backends (VSA, STA, Flash, etc.)Attention backends (VSA, STA, Flash, etc.)

scope: distributedSP, FSDP, USP, multi-nodeSP, FSDP, USP, multi-node

scope: modelModel architecture (DiTs, encoders, VAEs)Model architecture (DiTs, encoders, VAEs)

scope: trainingTraining pipeline, methods, configsTraining pipeline, methods, configs

Status: Open.

#1379 In hao-ai-lab/FastVideo;

· SolitaryThinker opened on May 21, 2026

[Bug] Downstream perf CI stage metric key mapping bug

bugSomething isn't workingSomething isn't working

scope: attentionAttention backends (VSA, STA, Flash, etc.)Attention backends (VSA, STA, Flash, etc.)

scope: inferenceInference pipeline, serving, CLIInference pipeline, serving, CLI

Status: Open.

Bug

#1377 In hao-ai-lab/FastVideo;

· Satyam-53 opened on May 21, 2026

[RFC]: Robust performance regression tracking enhancement

enhancementNew feature or requestNew feature or request

scope: infraCI, tests, Docker, buildCI, tests, Docker, build

type: ciCI/CD infrastructureCI/CD infrastructure

Status: Open.

Task

#1374 In hao-ai-lab/FastVideo;

· Satyam-53 opened on May 20, 2026

[Bug] Is the MoE expert routing in _sample_timesteps intended to be flow-shift-aware?

installationInstallation and setup issuesInstallation and setup issues

scope: attentionAttention backends (VSA, STA, Flash, etc.)Attention backends (VSA, STA, Flash, etc.)

scope: distributedSP, FSDP, USP, multi-nodeSP, FSDP, USP, multi-node

scope: inferenceInference pipeline, serving, CLIInference pipeline, serving, CLI

scope: modelModel architecture (DiTs, encoders, VAEs)Model architecture (DiTs, encoders, VAEs)

scope: trainingTraining pipeline, methods, configsTraining pipeline, methods, configs

Status: Open.

#1370 In hao-ai-lab/FastVideo;

· Xiaoyuan-Isaac-Wang opened on May 18, 2026

[Feature] distribute fastvideo kernels through the kernels project

installationInstallation and setup issuesInstallation and setup issues

scope: attentionAttention backends (VSA, STA, Flash, etc.)Attention backends (VSA, STA, Flash, etc.)

scope: docsDocumentationDocumentation

scope: kernelCUDA kernels, fastvideo-kernelCUDA kernels, fastvideo-kernel

Status: Open.

#1318 In hao-ai-lab/FastVideo;

· sayakpaul opened on May 11, 2026

[Bug] Several models hardcode supported attention backends. Causing confusion

installationInstallation and setup issuesInstallation and setup issues

scope: attentionAttention backends (VSA, STA, Flash, etc.)Attention backends (VSA, STA, Flash, etc.)

scope: docsDocumentationDocumentation

scope: inferenceInference pipeline, serving, CLIInference pipeline, serving, CLI

scope: modelModel architecture (DiTs, encoders, VAEs)Model architecture (DiTs, encoders, VAEs)

Status: Open.

#1254 In hao-ai-lab/FastVideo;

· watbulb opened on Apr 25, 2026

[Bug] Failed to distill Wan2.1-Fun-1.3B-InP

installationInstallation and setup issuesInstallation and setup issues

scope: attentionAttention backends (VSA, STA, Flash, etc.)Attention backends (VSA, STA, Flash, etc.)

scope: dataData preprocessing, datasetsData preprocessing, datasets

scope: docsDocumentationDocumentation

scope: inferenceInference pipeline, serving, CLIInference pipeline, serving, CLI

scope: modelModel architecture (DiTs, encoders, VAEs)Model architecture (DiTs, encoders, VAEs)

scope: trainingTraining pipeline, methods, configsTraining pipeline, methods, configs

Status: Open.

#1229 In hao-ai-lab/FastVideo;

· InuyashaLee opened on Apr 13, 2026

[Feature] ltx2.3 support

Status: Open.

#1171 In hao-ai-lab/FastVideo;

· zwukong opened on Mar 14, 2026

[RFC]: FastWan2.2-TI2V-5B-FullAttn-Diffusers output doesn't follow input image

Status: Open.

#1153 In hao-ai-lab/FastVideo;

· nanwanj opened on Mar 4, 2026

[Bug] Model FastWan-2.1-T2V-1.3B-Diffusers fails during inference

Status: Open.

#1146 In hao-ai-lab/FastVideo;

· dsynkd opened on Mar 1, 2026