Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / XLabs-AI/x-flux issues and pull requests
#114 - Timesteps different from the original BFL's implementation?
Issue -
State: open - Opened by dkenw 2 days ago
#113 - Lora + Controlnet
Issue -
State: open - Opened by fefespn 3 days ago
#112 - Number of controlnet blocks
Issue -
State: open - Opened by bjtuln 3 days ago
- 1 comment
#111 - Questions Regarding Training Costs, Dtype Error on H100, and ControlNet Loss Behavior
Issue -
State: open - Opened by zyyyz 4 days ago
#110 - Adds parameters for inference on two gpus
Pull Request -
State: open - Opened by raulc0399 7 days ago
#108 - (out of memory)OOM for traing controlnet on flux-dev-fp8 on A100 40G
Issue -
State: open - Opened by kongzijian 10 days ago
- 6 comments
#107 - Error when train controlnet using multi GPU
Issue -
State: open - Opened by 1goodone 11 days ago
#106 - Requesting feature for latent caching.
Issue -
State: open - Opened by aravindhv10 12 days ago
- 1 comment
#105 - How to use multi gpu to run main.py with ip adapter?
Issue -
State: open - Opened by Vova-B 12 days ago
#104 - Fixed batch issue
Pull Request -
State: open - Opened by aravindhv10 12 days ago
#103 - Cannot access gated repo for url https://huggingface.co/black-forest-labs/FLUX.1-dev/resolve/main/ae.safetensors.
Issue -
State: open - Opened by szhao1968 15 days ago
- 1 comment
#102 - Broadcasting issue when using training batch size > 1
Issue -
State: open - Opened by aravindhv10 16 days ago
- 2 comments
#101 - 【FAIL!】Local dir is used to import model weights【FLUX.1,IP-Adapter】
Issue -
State: open - Opened by XHB-ZMM 16 days ago
- 1 comment
#100 - 【FAIL!】Local dir is used to import model weights【FLUX.1,IP-Adapter】
Issue -
State: closed - Opened by XHB-ZMM 16 days ago
#99 - VRAM requirement to load ControlNet for inference?
Issue -
State: open - Opened by yuchen1984 17 days ago
#98 - How to use my lora for inpainting
Issue -
State: open - Opened by thesby 17 days ago
#97 - Questions about training?
Issue -
State: open - Opened by Laidawang 18 days ago
#96 - Mac M2. XlabsSampler Torch not compiled with CUDA enabled
Issue -
State: open - Opened by zhoudeheng 18 days ago
#95 - Chinese version of README
Pull Request -
State: open - Opened by ParzHe 18 days ago
#94 - Training Failed, SIGTERM exitcode: -9
Issue -
State: open - Opened by JimikoSK 19 days ago
#93 - What about new concept: flux carrier?
Issue -
State: closed - Opened by eletroswing 20 days ago
#92 - Inpaint
Issue -
State: closed - Opened by eletroswing 20 days ago
#91 - One place of code redundancy
Issue -
State: closed - Opened by wendashi 21 days ago
- 1 comment
#90 - 怎样把三个CN合起来到xlabs的采样器
Issue -
State: open - Opened by maiqunshan 21 days ago
#89 - Validation logs for controlnet training
Issue -
State: open - Opened by nemoooooooooo 22 days ago
- 1 comment
#88 - how to use multi controlnet?
Issue -
State: open - Opened by wly-ai-bj 22 days ago
#87 - controlnet depth
Issue -
State: open - Opened by malichen-cv 23 days ago
#86 - Out of memory on A100 when training controlnet using the given training script.
Issue -
State: open - Opened by oozz-sj 24 days ago
- 3 comments
#85 - Can you provide the training code for IP-Adapter?
Issue -
State: open - Opened by Amo5 24 days ago
- 1 comment
#84 - num_processes > 1, train_flux_lora_deepspeed failed
Issue -
State: open - Opened by woshinwyx 24 days ago
- 1 comment
#83 - How to use IP-Adapter + ControlNet
Issue -
State: closed - Opened by jaebbb 24 days ago
- 1 comment
#82 - How many images and the corresponding iterations are needed for training lora?
Issue -
State: open - Opened by datong-new 25 days ago
- 1 comment
#81 - Any Tutorial on how to run this repo locally with flux-schnell?
Issue -
State: open - Opened by Hassan-jr 25 days ago
#80 - Adapters not working in huggingface
Issue -
State: open - Opened by Xuzzo 26 days ago
#79 - how to use with IP-Adapter-FaceID
Issue -
State: open - Opened by eletroswing 26 days ago
#78 - How to use controlnet for background generation
Issue -
State: open - Opened by appleyang123 27 days ago
#77 - TypeError: DoubleStreamBlockLoraProcessor.forward() missing 1 required positional argument: 'pe'
Issue -
State: closed - Opened by MidnightRambo 29 days ago
- 2 comments
#76 - train_flux_lora_deepspeed.py Error : The size of tensor a (2) must match the size of tensor b (64)
Issue -
State: open - Opened by AfterHAL 29 days ago
- 2 comments
#75 - NotImplementedError: Cannot copy out of meta tensor; no data! Please use torch.nn.Module.to_empty() instead of torch.nn.Module.to() when moving module from meta to a different device.
Issue -
State: closed - Opened by starinskycc 29 days ago
- 2 comments
#74 - Sample code to use ip adapter .
Issue -
State: open - Opened by Tanmaypatil123 30 days ago
- 1 comment
#73 - Lora training's output_dir parameter is problematic
Issue -
State: open - Opened by qqtpi 30 days ago
#72 - how to select Cuda device to train lora ? (train_flux_lora_deepspeed.py)
Issue -
State: open - Opened by AfterHAL 30 days ago
- 2 comments
#71 - I have this....(⊙o⊙)…can not use XlabsSampler...
Issue -
State: closed - Opened by henger5566 30 days ago
- 6 comments
#70 - Quantization every time?
Issue -
State: open - Opened by SoftologyPro about 1 month ago
- 1 comment
#69 - Hi, for model convert?
Issue -
State: open - Opened by peki12345 about 1 month ago
- 2 comments
#68 - Full body image to character generation/ full body character generation
Issue -
State: open - Opened by khawar-islam about 1 month ago
#67 - fixed same seed on each image when using num_images_per_prompt
Pull Request -
State: closed - Opened by charlesmccarthy about 1 month ago
- 1 comment
#66 - RuntimeError: Given groups=1, weight of size [128, 3, 3, 3], expected input[1, 4, 512, 512] to have 3 channels, but got 4 channels instead
Issue -
State: closed - Opened by rbur0425 about 1 month ago
- 2 comments
#65 - Error occurred when executing ApplyFluxIPAdapter:
Issue -
State: open - Opened by plugcrypt about 1 month ago
- 3 comments
#64 - safetensors_rust.SafetensorError: Error while deserializing header: HeaderTooLarge
Issue -
State: closed - Opened by LilyXiehn about 1 month ago
- 1 comment
#63 - RuntimeError: mat1 and mat2 must have the same dtype, but got Half and BFloat16
Issue -
State: open - Opened by chenzhongzheng about 1 month ago
#62 - how to load and use the trained checkpoint with `dit.bin`
Issue -
State: open - Opened by youngwanLEE about 1 month ago
- 1 comment
#61 - Questions about flux-lora-collection
Issue -
State: open - Opened by dydxdt about 1 month ago
#60 - OSError: No such device (os error 19)
Issue -
State: open - Opened by chenzhongzheng about 1 month ago
- 4 comments
#59 - Any plan about inpainting controlnet?
Issue -
State: open - Opened by yinian-lw about 1 month ago
- 3 comments
#58 - How to prepare data for controlnet training
Issue -
State: closed - Opened by ymlab about 1 month ago
- 3 comments
#57 - ⭐ Add dev_req.txt for train
Pull Request -
State: closed - Opened by kadirnar about 1 month ago
- 1 comment
#56 - can we train this model in 1024 pix?
Issue -
State: open - Opened by heart-du about 1 month ago
- 3 comments
#55 - SIGKILL when running train_flux_lora_deepspeed.py
Issue -
State: open - Opened by k8tems about 1 month ago
- 8 comments
#54 - Is it possible to use this controlnet network for inpaint tasks and what needs to be changed?
Issue -
State: open - Opened by lgs00 about 1 month ago
- 2 comments
#53 - How to use zero3 to train the model?
Issue -
State: open - Opened by lgs00 about 1 month ago
- 1 comment
#52 - NoneType' object is not subscriptable
Issue -
State: open - Opened by kakachiex2 about 1 month ago
- 1 comment
#51 - How to generate several images at one time by main.py(lora)
Issue -
State: closed - Opened by arrowonstr about 1 month ago
- 2 comments
#50 - Support for pose controlnet
Issue -
State: open - Opened by joekendal about 1 month ago
- 1 comment
#49 - png为4个channel时报错
Issue -
State: closed - Opened by xqiprogramming about 1 month ago
- 3 comments
#48 - Confused about the checkpoint output files
Issue -
State: closed - Opened by arrowonstr about 1 month ago
- 3 comments
#47 - huggingface mirror
Issue -
State: open - Opened by jloveric about 1 month ago
#46 - How to train Flux LoRa with a local model dir?
Issue -
State: closed - Opened by arrowonstr about 1 month ago
- 3 comments
#45 - Inconsistency in Variable Naming: `x_0` Used as Gaussian Prior Instead of `x_1`
Issue -
State: closed - Opened by DarkMnDragon about 1 month ago
#44 - Error of running `train_flux_lora_deepspeed.py`
Issue -
State: closed - Opened by HelloWarcraft about 1 month ago
- 8 comments
#43 - Whe does my Training loss didn't go down when I train lora of flux?
Issue -
State: open - Opened by vincezh2000 about 1 month ago
- 6 comments
#42 - How to convert dit.bin from train_flux_deepseed script?
Issue -
State: open - Opened by tercumantanumut about 1 month ago
- 3 comments
#41 - How to convert Lora to safetensors
Issue -
State: closed - Opened by CCRss about 1 month ago
- 2 comments
#40 - Diffusers Support Request!
Issue -
State: open - Opened by frankjiang about 1 month ago
- 2 comments
#39 - Batch size > 1 got runtime error
Issue -
State: open - Opened by tristanwqy about 1 month ago
- 1 comment
#38 - How much vram needed for controlnet training?
Issue -
State: open - Opened by universewill about 1 month ago
- 11 comments
#37 - Questions about the controlnet
Issue -
State: open - Opened by roibenita about 1 month ago
#36 - Support for Inpainting and Depth ControlNet
Issue -
State: open - Opened by varadrane17 about 1 month ago
- 6 comments
#35 - Update main.py
Pull Request -
State: closed - Opened by iamwangyabin about 1 month ago
#34 - reason of T5 download !?
Issue -
State: open - Opened by al-swaiti about 1 month ago
- 1 comment
#33 - Is this just guessing?
Issue -
State: open - Opened by julioasotodv about 1 month ago
- 2 comments
#32 - Why does the loss look like this?
Issue -
State: open - Opened by Podidiving about 1 month ago
- 1 comment
#31 - RealismLoRA training data size?
Issue -
State: open - Opened by martintomov about 1 month ago
#30 - ControlNet for Flux Schnell possibly?
Issue -
State: closed - Opened by martintomov about 1 month ago
- 1 comment
#29 - Question about LoRA file identification and usage
Issue -
State: open - Opened by WarriorMama777 about 1 month ago
- 1 comment
#28 - Realistm training details
Issue -
State: open - Opened by amanikiruga about 1 month ago
#27 - not using autocast?
Issue -
State: closed - Opened by bghira about 1 month ago
- 3 comments
#26 - Is it possible to change the target location?
Issue -
State: open - Opened by jprian about 1 month ago
- 1 comment
#25 - Create all-in-one XFluxPipeline class
Pull Request -
State: closed - Opened by Anghellia about 1 month ago
#23 - Flux Lora in ComfyUI
Issue -
State: open - Opened by ffdown about 1 month ago
- 1 comment
#22 - Fixed grammatical errors
Pull Request -
State: closed - Opened by NormanMises about 1 month ago
- 1 comment
#19 - Add Replicate demo and API
Pull Request -
State: closed - Opened by lucataco about 1 month ago
- 1 comment
#18 - update requirements because it requires this too
Pull Request -
State: closed - Opened by neonsecret about 1 month ago
- 1 comment
#17 - chore: update util.py
Pull Request -
State: closed - Opened by eltociear about 1 month ago
#16 - Question About Training Loss
Issue -
State: closed - Opened by Hzj199 about 1 month ago
- 1 comment
#15 - Training recipe for flux-schnell?
Issue -
State: open - Opened by fffyanggg about 1 month ago
- 4 comments
#12 - A100 80GB lora training out of memory
Issue -
State: open - Opened by leonary about 1 month ago
- 12 comments
#5 - ComfyUI sample workflows
Issue -
State: open - Opened by kemaleksioglu about 1 month ago
- 3 comments
#2 - Is there a plan to support faceid
Issue -
State: open - Opened by libai-lab about 2 months ago
- 7 comments