GitHub · Where software is built

Experimental ROCM & PYTORCH builds for ALL GFX103X (rdna2)
#435 · patientx opened on Feb 7, 2026
48
Native pytorch build for AMD on Windows [Official amd 7000-9000 support & two choices for 6000 series]
#170 · alexsarmiento opened on Jun 6, 2025
108
Running Comfyui-Zluda with Ryzen 2200G-2400G-3200G-3400G apus
#77 · patientx opened on Feb 15, 2025

Labels Milestones New issue

Text encorder not working.

#437

· dkspwndj opened

on Feb 11, 2026

Experimental ROCM & PYTORCH builds for ALL GFX103X (rdna2)

#435

· patientx opened

on Feb 7, 2026

Found comfy_kitchen backend triton: {'available': True, 'disabled': True,

#434

· IdiotSavante opened

on Feb 7, 2026

TextEncodeQwenImageEditPlus_lrzjason: GET was unable to find an engine to execute this computation

#432

· sudonode opened

on Feb 5, 2026

semi-official ROCM & PYTORCH install with GFX1030 (6800's and 6900's) and GFX1032 (6600's)

#431

· patientx opened

on Feb 4, 2026

[Bug] FaceDetailer / VAE Decode crash on RX 6700 XT (gfx1031) - "GET was unable to find an engine"

#430

· Kaichi34 opened

on Feb 2, 2026

With ACEStep I get this error : VAEDecodeAudio 'function' object has no attribute 'func'

#424

· Deadlyforce opened

on Jan 27, 2026

[Windows] ComfyUI-ZLUDA fails to start - compilation error with struct timeval visibility on ROCm 6.4

#411

· dwojewod opened

on Jan 7, 2026

Slow generate time with 9070xt

#406

· GuraGuraNoMi opened

on Dec 30, 2025

Triton Error "error: 'stdio.h' file not found" (Python 3.12.7 - RX 5700XT)

#405

· Lorizzo-Mirco opened

on Dec 30, 2025

VAEEncode: GET was unable to find an engine to execute this computation

#404

· konkoltonov opened

on Dec 30, 2025

No CUDA GPUS are available

#400

· Gfdcvn opened

on Dec 26, 2025