Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / PixArt-alpha/PixArt-sigma issues and pull requests

#100 - LoRA training from local models

Issue - State: open - Opened by GavChap 4 months ago - 2 comments

#99 - How can PixArt-sigma generate higher resolution images beyond training-resolution?

Issue - State: closed - Opened by llstela 4 months ago - 2 comments
Labels: Answered

#98 - how to give a lora scale in inference

Issue - State: open - Opened by xiaowen0110 4 months ago - 1 comment

#98 - how to give a lora scale in inference

Issue - State: open - Opened by xiaowen0110 4 months ago - 1 comment

#97 - Make save_model_steps optional in train.py

Pull Request - State: closed - Opened by congruency 4 months ago - 1 comment

#97 - Make save_model_steps optional in train.py

Pull Request - State: closed - Opened by congruency 4 months ago - 1 comment

#96 - Add merge_transformers script

Pull Request - State: closed - Opened by frutiemax92 4 months ago

#96 - Add merge_transformers script

Pull Request - State: closed - Opened by frutiemax92 4 months ago

#95 - Fix convert diffusers to pixart

Pull Request - State: closed - Opened by frutiemax92 4 months ago

#95 - Fix convert diffusers to pixart

Pull Request - State: closed - Opened by frutiemax92 4 months ago

#94 - Add convert_diffusers_to_pixart script

Pull Request - State: closed - Opened by frutiemax92 4 months ago

#94 - Add convert_diffusers_to_pixart script

Pull Request - State: closed - Opened by frutiemax92 4 months ago

#93 - Train DMD !

Issue - State: closed - Opened by ApolloRay 4 months ago - 1 comment
Labels: Answered

#93 - Train DMD !

Issue - State: closed - Opened by ApolloRay 4 months ago - 1 comment
Labels: Answered

#92 - Inference is quite slow on A10

Issue - State: open - Opened by souvikqb 4 months ago - 11 comments

#92 - Inference is quite slow on A10

Issue - State: open - Opened by souvikqb 4 months ago - 11 comments

#91 - add_noise training details

Issue - State: open - Opened by kelisiya 4 months ago - 4 comments
Labels: Answered

#91 - add_noise training details

Issue - State: open - Opened by kelisiya 4 months ago - 4 comments
Labels: Answered

#90 - how to train sigma lora?

Issue - State: open - Opened by YLongJin 4 months ago - 4 comments

#90 - how to train sigma lora?

Issue - State: open - Opened by YLongJin 4 months ago - 4 comments

#89 - VRAM memory requirements for lora / finetuning?

Issue - State: open - Opened by GavChap 4 months ago - 4 comments
Labels: Answered

#89 - VRAM memory requirements for lora / finetuning?

Issue - State: open - Opened by GavChap 4 months ago - 4 comments
Labels: Answered

#88 - Details about instability when training with diffusers

Issue - State: open - Opened by samedii 4 months ago - 4 comments
Labels: help wanted

#88 - Details about instability when training with diffusers

Issue - State: open - Opened by samedii 4 months ago - 4 comments
Labels: help wanted

#87 - Are there any plans to release the code for ControlNetPipeline?

Issue - State: open - Opened by MingfuYAN 4 months ago - 1 comment
Labels: Answered

#86 - Inpainting

Issue - State: open - Opened by berryxue 4 months ago - 3 comments
Labels: Answered

#85 - [bug report] multi-aspect ratio training

Issue - State: open - Opened by linnanwang 5 months ago - 3 comments
Labels: Answered

#85 - [bug report] multi-aspect ratio training

Issue - State: open - Opened by linnanwang 5 months ago - 3 comments
Labels: Answered

#84 - Train text encoder

Pull Request - State: open - Opened by raulc0399 5 months ago - 2 comments

#84 - Train text encoder

Pull Request - State: open - Opened by raulc0399 5 months ago - 2 comments

#83 - Training pixart sigma with any aspect ratio?

Issue - State: open - Opened by NguyenNhoTrung 5 months ago - 3 comments
Labels: Answered

#83 - Training pixart sigma with any aspect ratio?

Issue - State: open - Opened by NguyenNhoTrung 5 months ago - 3 comments
Labels: Answered

#82 - Comfy UI Quickstart docker and workflows

Pull Request - State: closed - Opened by congruency 5 months ago - 3 comments

#82 - Comfy UI Quickstart docker and workflows

Pull Request - State: closed - Opened by congruency 5 months ago - 3 comments

#81 - What are the extra 4 channels from the denoiser?

Issue - State: closed - Opened by samedii 5 months ago - 5 comments

#80 - Fix convert_diffusers_to_pipeline.py

Pull Request - State: closed - Opened by frutiemax92 5 months ago

#80 - Fix convert_diffusers_to_pipeline.py

Pull Request - State: closed - Opened by frutiemax92 5 months ago

#79 - where can I find config.json?

Issue - State: closed - Opened by 1sheep1 5 months ago - 2 comments
Labels: Answered

#79 - where can I find config.json?

Issue - State: closed - Opened by 1sheep1 5 months ago - 2 comments
Labels: Answered

#78 - train dataset

Issue - State: closed - Opened by dangweili 5 months ago - 1 comment
Labels: Answered

#77 - In some scenarios, there are poor performance issues in 2K checkpoint.

Issue - State: closed - Opened by kelisiya 5 months ago - 1 comment
Labels: Answered

#76 - Diffusers model loading Issues

Issue - State: closed - Opened by nbardy 5 months ago - 2 comments
Labels: Answered

#75 - Modify trainer to source validation prompts from config

Pull Request - State: closed - Opened by congruency 5 months ago

#75 - Modify trainer to source validation prompts from config

Pull Request - State: closed - Opened by congruency 5 months ago

#73 - Img2Img pipeline for pixart

Issue - State: open - Opened by frutiemax92 5 months ago - 3 comments
Labels: Answered

#72 - Fine-tuning script based on diffusers

Issue - State: closed - Opened by youngwanLEE 5 months ago - 3 comments
Labels: Answered

#72 - Fine-tuning script based on diffusers

Issue - State: closed - Opened by youngwanLEE 5 months ago - 3 comments
Labels: Answered

#71 - interface.py中168行拼写bug：imrage_size

Issue - State: closed - Opened by liangwq 5 months ago - 1 comment
Labels: Fixed

#71 - interface.py中168行拼写bug：imrage_size

Issue - State: closed - Opened by liangwq 5 months ago - 1 comment
Labels: Fixed

#70 - PixArt-Sigma-XL-2-512-MS do not have multi-aspect/resolution embedding?

Issue - State: closed - Opened by LinB203 5 months ago - 14 comments

#69 - How to img2img?

Issue - State: open - Opened by berryxue 5 months ago - 1 comment
Labels: Answered

#68 - Is lora only for alpha?

Issue - State: closed - Opened by berryxue 5 months ago - 1 comment
Labels: Answered

#67 - Max_length in lora training

Issue - State: closed - Opened by berryxue 5 months ago - 1 comment
Labels: Answered

#67 - Max_length in lora training

Issue - State: closed - Opened by berryxue 5 months ago - 1 comment
Labels: Answered

#66 - Store model on CPU during convert_pixart_to_diffusers.py process

Pull Request - State: closed - Opened by congruency 5 months ago

#66 - Store model on CPU during convert_pixart_to_diffusers.py process

Pull Request - State: closed - Opened by congruency 5 months ago

#65 - Question about Samplers

Issue - State: closed - Opened by YanzuoLu 5 months ago - 2 comments
Labels: Answered

#65 - Question about Samplers

Issue - State: closed - Opened by YanzuoLu 5 months ago - 2 comments
Labels: Answered

#64 - Support lora training and inference

Pull Request - State: closed - Opened by lawrence-cj 5 months ago

#63 - ValueError: Invalid sample size issue when testing 2K checkpoint

Issue - State: closed - Opened by MahdiMohseni0033 5 months ago - 4 comments
Labels: Answered

#62 - Add inference_pipeline.py

Pull Request - State: closed - Opened by frutiemax92 5 months ago

#61 - When will authors release the Lora training scripts?

Issue - State: closed - Opened by berryxue 5 months ago - 2 comments
Labels: Answered

#60 - Add convert_diffusers_to_pipeline script

Pull Request - State: closed - Opened by frutiemax92 5 months ago - 1 comment

#59 - Some problems with graio loading training weights

Issue - State: closed - Opened by kelisiya 5 months ago - 2 comments

#58 - Fix error for features using different image exts

Pull Request - State: closed - Opened by frutiemax92 5 months ago

#57 - [Script] Images folder convert script to data_info.json

Pull Request - State: closed - Opened by frutiemax92 5 months ago - 3 comments

#56 - VAE for the DMD model

Issue - State: closed - Opened by EternalEvan 5 months ago - 2 comments
Labels: Answered

#55 - Is it possible to use Controlnet, IPAdapter, and Lora

Issue - State: open - Opened by libai-lab 5 months ago - 7 comments
Labels: Answered

#54 - train speed seems too slow

Issue - State: closed - Opened by Hazarch 5 months ago - 1 comment
Labels: Answered

#53 - There is a gap with stable diffusion

Issue - State: closed - Opened by conansherry 5 months ago - 5 comments
Labels: Answered

#52 - 参考图

Issue - State: closed - Opened by linghu-cell 5 months ago - 3 comments
Labels: Answered

#51 - Fix error for single gpu

Pull Request - State: closed - Opened by frutiemax92 5 months ago

#50 - Question about VAE adaption stage

Issue - State: closed - Opened by piddnad 5 months ago - 7 comments
Labels: Answered

#49 - How to use the DMD checkpoint?

Issue - State: closed - Opened by EternalEvan 5 months ago - 5 comments
Labels: Answered

#48 - 高分辨率的图就是直接 resize 放大的？

Issue - State: closed - Opened by wodsoe 5 months ago - 1 comment
Labels: Answered

#47 - Cannot load a Lora safetensors file

Issue - State: closed - Opened by ukaprch 5 months ago - 5 comments

#46 - 256x256 PixArt-Sigma model, only get a FID-30k nearly ~49 on coco2014 val , is that normal?

Issue - State: open - Opened by Gus-Guo 5 months ago - 9 comments
Labels: Answered

#45 - Resume training confusion

Issue - State: closed - Opened by mcrisafu 5 months ago - 2 comments

#44 - must have sharegpt4v prompt in json ?

Issue - State: closed - Opened by kelisiya 5 months ago

#43 - How much graphics memory is required to run the above test?

Issue - State: closed - Opened by shams2023 5 months ago - 6 comments
Labels: Answered

#42 - CUDA out of memory

Issue - State: closed - Opened by linghu-cell 5 months ago - 9 comments
Labels: Answered

#41 - Create tmp folder in the output folder to avoid errors

Pull Request - State: closed - Opened by frutiemax92 5 months ago

#40 - Fix git clone step for t5 pretrained model

Pull Request - State: closed - Opened by frutiemax92 5 months ago - 1 comment

#39 - running the dmd demo raise error

Issue - State: closed - Opened by RyanHuangNLP 5 months ago - 2 comments
Labels: Answered

#38 - Would you like to provide more details of dmd training?

Issue - State: closed - Opened by RyanHuangNLP 5 months ago - 1 comment
Labels: Answered

#37 - why pixart-sigma-256x256 ckpt has no 'pos_embed', while other ckpts has?

Issue - State: closed - Opened by Gus-Guo 5 months ago - 3 comments
Labels: Answered

#36 - Small error in train script

Issue - State: closed - Opened by attashe 5 months ago - 2 comments
Labels: Fixed, Answered

#35 - Question about KV compression in diffusers

Issue - State: closed - Opened by lijing1996 5 months ago - 1 comment

#34 - why the checkpoints is biger

Issue - State: open - Opened by chongxian 5 months ago - 8 comments

#33 - Support HuggingFace mirror in download

Pull Request - State: closed - Opened by 7sunarni 5 months ago - 1 comment

#31 - Add Replicate demo and API

Pull Request - State: open - Opened by chenxwh 5 months ago

#30 - how do you pre-procecss 512-size images?

Issue - State: closed - Opened by EternalEvan 5 months ago - 2 comments

#29 - Unexpected results with `diffusers`

Issue - State: closed - Opened by chenxwh 5 months ago - 8 comments
Labels: Fixed

#27 - About pixart_sigma + controlnet

Issue - State: closed - Opened by EugeneCh1a 5 months ago - 1 comment
Labels: Answered

#26 - DMD load problems

Issue - State: closed - Opened by chuck-ma 5 months ago - 2 comments
Labels: Fixed

#25 - Text-to-Image Alignment Performance of the pixart-sigma Model

Issue - State: open - Opened by xiexiaoshinick 5 months ago - 4 comments
Labels: good first issue

#24 - Inquiry About Hardware Requirements and Training Details for PIXART-Σ

Issue - State: closed - Opened by sohowj 5 months ago - 2 comments