kijai/comfyui-fluxtrainer issues and pull requests

#94 - it works great! thank you master!!!

Issue - State: open - Opened by AGXAI 3 days ago

#94 - it works great! thank you master!!!

Issue - State: open - Opened by AGXAI 3 days ago

#93 - InitFluxLoRATraining Couldn't build proto file into descriptor pool: duplicate file name sentencepiece_model.proto

Issue - State: open - Opened by wazitang 5 days ago

#93 - InitFluxLoRATraining Couldn't build proto file into descriptor pool: duplicate file name sentencepiece_model.proto

Issue - State: open - Opened by wazitang 5 days ago

#92 - Hello! I hope you will support the Chinese translation, which will especially help your plugin to carry forward! Although your plugin is already pretty good.

Issue - State: open - Opened by ab1314625 13 days ago

#92 - Hello! I hope you will support the Chinese translation, which will especially help your plugin to carry forward! Although your plugin is already pretty good.

Issue - State: open - Opened by ab1314625 13 days ago

#91 - torch.cuda.OutOfMemoryError

Issue - State: open - Opened by mneyanmels 14 days ago - 5 comments

#91 - torch.cuda.OutOfMemoryError

Issue - State: open - Opened by mneyanmels 14 days ago - 5 comments

#90 - Cyrillic alphabet in the cache causes this error

Issue - State: open - Opened by Liteyn 16 days ago

#90 - Cyrillic alphabet in the cache causes this error

Issue - State: open - Opened by Liteyn 16 days ago

#89 - An error about InitFlux LoRATraining

Issue - State: open - Opened by linseasea 17 days ago - 2 comments

#89 - An error about InitFlux LoRATraining

Issue - State: open - Opened by linseasea 17 days ago - 2 comments

#88 - Getting error after training and saving the model when running in Linux environment

Issue - State: open - Opened by CCJetWing 21 days ago - 1 comment

#88 - Getting error after training and saving the model when running in Linux environment

Issue - State: open - Opened by CCJetWing 21 days ago - 1 comment

#87 - Regarding SD35 support

Issue - State: open - Opened by liangzoey 21 days ago - 1 comment

#87 - Regarding SD35 support

Issue - State: open - Opened by liangzoey 21 days ago - 1 comment

#86 - Unable to load vocabulary from file. Please check that the provided vocabulary is accessible and not corrupted.

Issue - State: closed - Opened by Piranese 24 days ago - 1 comment

#86 - Unable to load vocabulary from file. Please check that the provided vocabulary is accessible and not corrupted.

Issue - State: closed - Opened by Piranese 24 days ago - 1 comment

#85 - Segmentation fault，Core dumped

Issue - State: open - Opened by ggggxm 26 days ago - 1 comment

#85 - Segmentation fault，Core dumped

Issue - State: open - Opened by ggggxm 26 days ago - 1 comment

#84 - any example of flux DB training?

Issue - State: closed - Opened by whmc76 30 days ago

#84 - any example of flux DB training?

Issue - State: closed - Opened by whmc76 30 days ago

#83 - STOP training when activating tab with another Comfyui (asyncio???)

Issue - State: open - Opened by Maelstrom2014 about 1 month ago

#83 - STOP training when activating tab with another Comfyui (asyncio???)

Issue - State: open - Opened by Maelstrom2014 about 1 month ago

#82 - torch.cuda.OutOfMemoryError

Issue - State: closed - Opened by vallestutz about 1 month ago - 3 comments

#82 - torch.cuda.OutOfMemoryError

Issue - State: closed - Opened by vallestutz about 1 month ago - 3 comments

#81 - the training is slower too much than use aitoolkit in all the same settings

Issue - State: closed - Opened by whmc76 about 1 month ago - 5 comments

#81 - the training is slower too much than use aitoolkit in all the same settings

Issue - State: closed - Opened by whmc76 about 1 month ago - 5 comments

#80 - InitFluxLoRATraining: "Cannot copy out of meta tensor; no data!"

Issue - State: open - Opened by dkhold about 1 month ago - 4 comments

#80 - InitFluxLoRATraining: "Cannot copy out of meta tensor; no data!"

Issue - State: open - Opened by dkhold about 1 month ago - 4 comments

#79 - [Feature Request] Node to connect to tfboard, wandb, etc

Issue - State: open - Opened by clyce about 1 month ago - 3 comments

#79 - [Feature Request] Node to connect to tfboard, wandb, etc

Issue - State: open - Opened by clyce about 1 month ago - 3 comments

#78 - General question about the TrainDatasetAdd Node

Issue - State: closed - Opened by BlackVortexAI about 2 months ago - 2 comments

#78 - General question about the TrainDatasetAdd Node

Issue - State: closed - Opened by BlackVortexAI about 2 months ago - 2 comments

#77 - what model should i use

Issue - State: open - Opened by Chengym2023 about 2 months ago - 3 comments

#77 - what model should i use

Issue - State: open - Opened by Chengym2023 about 2 months ago - 3 comments

#76 - Flux.1.dev.fp8 CKPT trainning the avr_loss keep 'nan'

Issue - State: open - Opened by Lecho303 about 2 months ago - 8 comments

#76 - Flux.1.dev.fp8 CKPT trainning the avr_loss keep 'nan'

Issue - State: open - Opened by Lecho303 about 2 months ago - 8 comments

#75 - what could i do?

Issue - State: open - Opened by Chengym2023 about 2 months ago

#75 - what could i do?

Issue - State: open - Opened by Chengym2023 about 2 months ago

#74 - Make FluxTrainEnd an output node

Pull Request - State: closed - Opened by chrisgoringe about 2 months ago - 1 comment

#74 - Make FluxTrainEnd an output node

Pull Request - State: closed - Opened by chrisgoringe about 2 months ago - 1 comment

#73 - I still don't understand why increasing the batch size makes the training slower

Issue - State: open - Opened by PGCRT about 2 months ago - 2 comments

#73 - I still don't understand why increasing the batch size makes the training slower

Issue - State: open - Opened by PGCRT about 2 months ago - 2 comments

#72 - question about image caption file types

Issue - State: closed - Opened by enkie358 about 2 months ago - 2 comments

#72 - question about image caption file types

Issue - State: closed - Opened by enkie358 about 2 months ago - 2 comments

#71 - Regularization Issue with Multiple Training Datasets

Issue - State: open - Opened by zhristophe about 2 months ago - 2 comments

#71 - Regularization Issue with Multiple Training Datasets

Issue - State: open - Opened by zhristophe about 2 months ago - 2 comments

#70 - Is there a way to train with multi GPUs

Issue - State: open - Opened by RaySteve312 about 2 months ago - 3 comments

#70 - Is there a way to train with multi GPUs

Issue - State: open - Opened by RaySteve312 about 2 months ago - 3 comments

#69 - Can you explain the "String Constant Multiline" Node in your workflow?

Issue - State: closed - Opened by q5sys about 2 months ago - 2 comments

#69 - Can you explain the "String Constant Multiline" Node in your workflow?

Issue - State: closed - Opened by q5sys about 2 months ago - 2 comments

#68 - ImportError: numpy.core.multiarray failed to import

Issue - State: open - Opened by zark119 about 2 months ago

#68 - ImportError: numpy.core.multiarray failed to import

Issue - State: open - Opened by zark119 about 2 months ago

#67 - how to train multiple or all attn blocks

Issue - State: closed - Opened by whmc76 about 2 months ago - 5 comments

#67 - how to train multiple or all attn blocks

Issue - State: closed - Opened by whmc76 about 2 months ago - 5 comments

#66 - Adds TrainDatasetAddMultiReso

Pull Request - State: open - Opened by wanderor 2 months ago

#66 - Adds TrainDatasetAddMultiReso

Pull Request - State: open - Opened by wanderor 2 months ago

#65 - extract flux lora report "File does not contain tensor double_blocks.0.img_attn.proj.weight" error

Issue - State: open - Opened by deepfree2023 2 months ago - 5 comments

#65 - extract flux lora report "File does not contain tensor double_blocks.0.img_attn.proj.weight" error

Issue - State: open - Opened by deepfree2023 2 months ago - 5 comments

#64 - HOW to resume training?

Issue - State: closed - Opened by Maelstrom2014 2 months ago - 1 comment

#64 - HOW to resume training?

Issue - State: closed - Opened by Maelstrom2014 2 months ago - 1 comment

#63 - Failed to use trained lora to generate image, the error is “r2.weight addmm_cuda not implemented for float8”

Issue - State: closed - Opened by Bellzs 2 months ago - 1 comment

#63 - Failed to use trained lora to generate image, the error is “r2.weight addmm_cuda not implemented for float8”

Issue - State: closed - Opened by Bellzs 2 months ago - 1 comment

#62 - ERROR: 'utf-8' codec can't decode byte 0x92 in position 717: invalid start byte

Issue - State: open - Opened by kuzman123 2 months ago - 2 comments

#62 - ERROR: 'utf-8' codec can't decode byte 0x92 in position 717: invalid start byte

Issue - State: open - Opened by kuzman123 2 months ago - 2 comments

#61 - Fails on restart

Issue - State: closed - Opened by chrisgoringe 2 months ago - 2 comments

#61 - Fails on restart

Issue - State: closed - Opened by chrisgoringe 2 months ago - 2 comments

#60 - TRAINING savings on minimal LOSS for better LORA

Issue - State: open - Opened by Maelstrom2014 2 months ago - 4 comments

#60 - TRAINING savings on minimal LOSS for better LORA

Issue - State: open - Opened by Maelstrom2014 2 months ago - 4 comments

#59 - ERROR: OOM on [FluxTrainValidate], but training w|o problem

Issue - State: open - Opened by Maelstrom2014 2 months ago - 2 comments

#59 - ERROR: OOM on [FluxTrainValidate], but training w|o problem

Issue - State: open - Opened by Maelstrom2014 2 months ago - 2 comments

#58 - Failed to validate prompt for output 808

Issue - State: open - Opened by RaySteve312 2 months ago - 1 comment

#58 - Failed to validate prompt for output 808

Issue - State: open - Opened by RaySteve312 2 months ago - 1 comment

#57 - InitFluxLoRATraining——Reserved escape sequence used (line 9 column 1 char 167)

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

#57 - InitFluxLoRATraining——Reserved escape sequence used (line 9 column 1 char 167)

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

#56 - Visualize NODE, normalize_y not normalizing :)

Issue - State: open - Opened by Maelstrom2014 2 months ago - 1 comment

#56 - Visualize NODE, normalize_y not normalizing :)

Issue - State: open - Opened by Maelstrom2014 2 months ago - 1 comment

#55 - ERROR on start of training. AttributeError: 'LoRANetwork' object has no attribute 'prepare_optimizer_params'

Issue - State: closed - Opened by Maelstrom2014 2 months ago - 1 comment

#55 - ERROR on start of training. AttributeError: 'LoRANetwork' object has no attribute 'prepare_optimizer_params'

Issue - State: closed - Opened by Maelstrom2014 2 months ago - 1 comment

#54 - 更新后提示InitFluxLoRATraining Reserved escape sequence used (line 9 column 1 char 167)

Issue - State: open - Opened by zangchenyang 2 months ago - 1 comment

#54 - 更新后提示InitFluxLoRATraining Reserved escape sequence used (line 9 column 1 char 167)

Issue - State: open - Opened by zangchenyang 2 months ago - 1 comment

#53 - The issue of facial contamination in LoRA

Issue - State: open - Opened by chaorenai 2 months ago - 3 comments

#53 - The issue of facial contamination in LoRA

Issue - State: open - Opened by chaorenai 2 months ago - 3 comments

#52 - How to sequentially train multiple LoRA models using scheduled tasks?

Issue - State: open - Opened by oooing 2 months ago - 1 comment

#52 - How to sequentially train multiple LoRA models using scheduled tasks?

Issue - State: open - Opened by oooing 2 months ago - 1 comment

#51 - (Resolved) 6x Slower than FluxGYM?

Issue - State: closed - Opened by PGCRT 2 months ago - 19 comments

#51 - (Resolved) 6x Slower than FluxGYM?

Issue - State: closed - Opened by PGCRT 2 months ago - 19 comments

#50 - [Feature request]: Support empty captions

Issue - State: open - Opened by MNeMoNiCuZ 2 months ago - 2 comments

#50 - [Feature request]: Support empty captions

Issue - State: open - Opened by MNeMoNiCuZ 2 months ago - 2 comments

#49 - Error occurred when executing InitFluxLoRATraining - '' is not a directory

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

#49 - Error occurred when executing InitFluxLoRATraining - '' is not a directory

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

#48 - [Feature Request] Allow to specify exact epochs

Issue - State: closed - Opened by Apacchi88 2 months ago - 2 comments

#48 - [Feature Request] Allow to specify exact epochs

Issue - State: closed - Opened by Apacchi88 2 months ago - 2 comments

#47 - New Error After Update: module 'torch' has no attribute 'float8_e4m3fnuz'

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

#47 - New Error After Update: module 'torch' has no attribute 'float8_e4m3fnuz'

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

#46 - An error occurred while running to the InitFlux LoRATraining node，Can't load tokenizer for 'openai/clip-vit-large-patch14

Issue - State: open - Opened by tongdaofang 2 months ago - 1 comment

GitHub / kijai/comfyui-fluxtrainer issues and pull requests