comfy-trt-test

comfy-trt-test copied to clipboard

ModuleNotFoundError: No module named 'comfy_trt'

9

Hey there! So I've tried to install it all on portable comfyui with python_embeded python.exe, and when I try to interract with convert_unet.py, it always keeps saying ModuleNotFoundError: No module...

OliviaOliveiira

Wanted To Share This Workflow Video Showcasing Your Awesome Custom Node

9

https://www.youtube.com/watch?v=kklIU2RqVOk

boricuapab

Getting a whole bunch of errors when running on a 2070 in WSL

1

python3 convert_unet.py --ckpt_path ../../models/checkpoints/sd_xl_base_1.0_0.9vae.safetensors Total VRAM 8192 MB, total RAM 48178 MB xformers version: 0.0.23.post1 Set vram state to: NORMAL_VRAM Device: cuda:0 NVIDIA GeForce RTX 2070 : native VAE dtype:...

GavChap

Dimension Error with SDXL models with batch size max > 1

5

Thank you for your work on this project This error occurs during image generation using a converted model. It does not occur with SD 1.5 models but has been observed...

Tliks

Cannot Build the Unet Model -Help

33

`python convert_unet.py --ckpt_path "H:\ComfyUI\models\checkpoints\SDXL\turbovisionxlSuperFastXL_tvxlV431vae.safetensors" --batch_max 4 --height_min 512 --width_min 512 --height_max 1024 --width_max 1024 --token_count_max 300` warnings.warn(warning) ERROR:root:Exporting to ONNX failed. ONNX export failed on an operator with unrecognized namespace...

Piscabo

Converted Turbo XL Successfully but was getting black images

3

I think something changed with comfy that is causing the turbo xl tensor rt model to render our black images. Im on this comfy commit ``` commit b0aab1e4ea3dfefe09c4f07de0e5237558097e22 (HEAD ->...

boricuapab

Speedup not so much, only 24%

8

First , I built it on A 10G by ```bash docker pull nvcr.io/nvidia/pytorch:23.11-py3 docker run --gpus all -it -p 80:8188 \ -v /root/acceleration:/acceleration -v ${PWD}:/app -w /app nvcr.io/nvidia/pytorch:23.11-py3 pip install...

onefish51