Правила форума  |  Как оформлять свои темы и комментарии  |  Как делать крутые скриншоты  |  ВАЖНО

Вы должны войти, чтобы создавать сообщения и темы.

после твикера сдохли все нейронки [Решено]

Георгий@time2be
10 Сообщений
Pro
#1 · 11.08.2025, 01:54

откат не помог. фордж, олдфото, фокус - все отказываются работать...

не пойму где вообще проблема - что переустановить(((((((((

фордж пишет это, несмотря на то что я venv переустановил:

Спойлер

To create a public link, set `share=True` in `launch()`.
Startup time: 27.0s (prepare environment: 10.3s, launcher: 0.5s, import torch: 6.2s, initialize shared: 0.2s, other imports: 0.3s, load scripts: 3.5s, create ui: 3.6s, gradio launch: 2.4s).
Environment vars changed: {'stream': False, 'inference_memory': 1024.0, 'pin_shared_memory': False}
[GPU Setting] You will use 93.75% GPU memory (15355.00 MB) to load weights, and use 6.25% GPU memory (1024.00 MB) to do matrix computation.
Loading Model: {'checkpoint_info': {'filename': 'D:\\_Models\\Flux\\flux1KontextDev_v10Fp8E4m3fn.safetensors', 'hash': '753d0aee'}, 'additional_modules': ['C:\\_Forge\\webui\\models\\text_encoder\\clip_l.safetensors', 'C:\\_Forge\\webui\\models\\VAE\\ae.safetensors', 'C:\\_Forge\\webui\\models\\text_encoder\\t5xxl_fp8_e4m3fn.safetensors'], 'unet_storage_dtype': None}
[Unload] Trying to free all memory for cuda:0 with 0 models keep loaded ... Done.
StateDict Keys: {'transformer': 780, 'vae': 244, 'text_encoder': 196, 'text_encoder_2': 220, 'ignore': 0}
Using Detected T5 Data Type: torch.float8_e4m3fn
Using Detected UNet Type: torch.float8_e4m3fn
Working with z of shape (1, 16, 32, 32) = 16384 dimensions.
K-Model Created: {'storage_dtype': torch.float8_e4m3fn, 'computation_dtype': torch.bfloat16}
Model loaded in 0.8s (unload existing model: 0.2s, forge model load: 0.6s).
Skipping unconditional conditioning when CFG = 1. Negative Prompts are ignored.
[Unload] Trying to free 7723.54 MB for cuda:0 with 0 models keep loaded ... Done.
[Memory Management] Target: JointTextEncoder, Free GPU: 15208.15 MB, Model Require: 5153.49 MB, Previously Loaded: 0.00 MB, Inference Require: 1024.00 MB, Remaining: 9030.66 MB, All loaded to GPU.
Moving model(s) has taken 2.69 seconds
C:\_Forge\webui\venv\lib\site-packages\transformers\models\clip\modeling_clip.py:539: UserWarning: 1Torch was not compiled with flash attention. (Triggered internally at ..\aten\src\ATen\native\transformers\cuda\sdp_utils.cpp:455.)
attn_output = torch.nn.functional.scaled_dot_product_attention(
Exception in thread MemMon:
Traceback (most recent call last):
File "C:\Users\0072\AppData\Local\Programs\Python\Python310\lib\threading.py", line 1016, in _bootstrap_inner
self.run()
File "C:\_Forge\webui\modules\memmon.py", line 53, in run
free, total = self.cuda_mem_get_info()
File "C:\_Forge\webui\modules\memmon.py", line 34, in cuda_mem_get_info
return torch.cuda.mem_get_info(index)
File "C:\_Forge\webui\venv\lib\site-packages\torch\cuda\memory.py", line 655, in mem_get_info
return torch.cuda.cudart().cudaMemGetInfo(device)
RuntimeError: CUDA error: an illegal memory access was encountered
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

Unhandled exception caught in c10/util/AbortHandler.h
00007FFD5CAB805400007FFD5CA9C190 torch_python.dll!THPGenerator_initDefaultGenerator [<unknown file> @ <unknown line number>]
00007FFDEE9C1F9800007FFDEE9C1F80 ucrtbase.dll!terminate [<unknown file> @ <unknown line number>]
00007FFDD8931AAB00007FFDD8931150 VCRUNTIME140_1.dll!_NLG_Return2 [<unknown file> @ <unknown line number>]
00007FFDD893231700007FFDD8931150 VCRUNTIME140_1.dll!_NLG_Return2 [<unknown file> @ <unknown line number>]
00007FFDD893411900007FFDD8934070 VCRUNTIME140_1.dll!_CxxFrameHandler4 [<unknown file> @ <unknown line number>]
00007FFDF11928BF00007FFDF11927A0 ntdll.dll!_chkstk [<unknown file> @ <unknown line number>]
00007FFDF114255400007FFDF11420D0 ntdll.dll!RtlRaiseException [<unknown file> @ <unknown line number>]
00007FFDF11422A700007FFDF11420D0 ntdll.dll!RtlRaiseException [<unknown file> @ <unknown line number>]
00007FFDEEE0533900007FFDEEE052D0 KERNELBASE.dll!RaiseException [<unknown file> @ <unknown line number>]
00007FFDDAD466C000007FFDDAD46630 VCRUNTIME140.dll!CxxThrowException [<unknown file> @ <unknown line number>]
00007FFD7475319E00007FFD74753130 c10.dll!c10::detail::torchCheckFail [<unknown file> @ <unknown line number>]
00007FFD8E30F93F00007FFD8E30F640 c10_cuda.dll!c10::cuda::c10_cuda_check_implementation [<unknown file> @ <unknown line number>]
00007FFD8E31524200007FFD8E310BA0 c10_cuda.dll!c10::cuda::warn_or_error_on_sync [<unknown file> @ <unknown line number>]
00007FFD8E3150CF00007FFD8E310BA0 c10_cuda.dll!c10::cuda::warn_or_error_on_sync [<unknown file> @ <unknown line number>]
00007FFD7471A31C00007FFD7471A2F0 c10.dll!c10::StorageImpl::~StorageImpl [<unknown file> @ <unknown line number>]
00007FFC6534962500007FFC65347E40 torch_cpu.dll!at::DynamicLibrary::sym [<unknown file> @ <unknown line number>]
00007FFD7470EFC200007FFD7470EF50 c10.dll!c10::ConstantSymNodeImpl<bool>::~ConstantSymNodeImpl<bool> [<unknown file> @ <unknown line number>]
00007FFD7473D43500007FFD7473D3C0 c10.dll!c10::TensorImpl::~TensorImpl [<unknown file> @ <unknown line number>]
00007FFC6534966500007FFC65347E40 torch_cpu.dll!at::DynamicLibrary::sym [<unknown file> @ <unknown line number>]
00007FFC652EE71D00007FFC652EE690 torch_cpu.dll!at::TensorBase::reset [<unknown file> @ <unknown line number>]
00007FFDD89310EA <unknown symbol address> VCRUNTIME140_1.dll!<unknown symbol> [<unknown file> @ <unknown line number>]
00007FFDD8932CB000007FFDD8931150 VCRUNTIME140_1.dll!_NLG_Return2 [<unknown file> @ <unknown line number>]
00007FFDD8933B0D00007FFDD8931150 VCRUNTIME140_1.dll!_NLG_Return2 [<unknown file> @ <unknown line number>]
00007FFDD893224400007FFDD8931150 VCRUNTIME140_1.dll!_NLG_Return2 [<unknown file> @ <unknown line number>]
00007FFDD893411900007FFDD8934070 VCRUNTIME140_1.dll!_CxxFrameHandler4 [<unknown file> @ <unknown line number>]
00007FFC6AC7343C00007FFC69108050 torch_cpu.dll!torch::ModuleDef::unsafe_arena_set_allocated_torchscript_arena [<unknown file> @ <unknown line number>]
00007FFDF119293F00007FFDF11927A0 ntdll.dll!_chkstk [<unknown file> @ <unknown line number>]
00007FFDF112093900007FFDF1120600 ntdll.dll!RtlUnwindEx [<unknown file> @ <unknown line number>]
00007FFDD8933DAE00007FFDD8931150 VCRUNTIME140_1.dll!_NLG_Return2 [<unknown file> @ <unknown line number>]
00007FFDD893163500007FFDD8931150 VCRUNTIME140_1.dll!_NLG_Return2 [<unknown file> @ <unknown line number>]
00007FFDD89319D500007FFDD8931150 VCRUNTIME140_1.dll!_NLG_Return2 [<unknown file> @ <unknown line number>]
00007FFDD893231700007FFDD8931150 VCRUNTIME140_1.dll!_NLG_Return2 [<unknown file> @ <unknown line number>]
00007FFDD893411900007FFDD8934070 VCRUNTIME140_1.dll!_CxxFrameHandler4 [<unknown file> @ <unknown line number>]
00007FFD5D13054000007FFD5CF89090 torch_python.dll!c10d::PythonCommHook::runHook [<unknown file> @ <unknown line number>]
00007FFDF11928BF00007FFDF11927A0 ntdll.dll!_chkstk [<unknown file> @ <unknown line number>]
00007FFDF114255400007FFDF11420D0 ntdll.dll!RtlRaiseException [<unknown file> @ <unknown line number>]
00007FFDF11422A700007FFDF11420D0 ntdll.dll!RtlRaiseException [<unknown file> @ <unknown line number>]
00007FFDEEE0533900007FFDEEE052D0 KERNELBASE.dll!RaiseException [<unknown file> @ <unknown line number>]
00007FFDDAD466C000007FFDDAD46630 VCRUNTIME140.dll!CxxThrowException [<unknown file> @ <unknown line number>]
00007FFD7475319E00007FFD74753130 c10.dll!c10::detail::torchCheckFail [<unknown file> @ <unknown line number>]
00007FFD8E30F93F00007FFD8E30F640 c10_cuda.dll!c10::cuda::c10_cuda_check_implementation [<unknown file> @ <unknown line number>]
00007FFC6D74FC9700007FFC6D7118E0 torch_cuda.dll!at::native::slow_conv2d_forward_out_cuda [<unknown file> @ <unknown line number>]
00007FFC6563A5DA00007FFC65639740 torch_cpu.dll!at::native::copy_ignoring_overlaps [<unknown file> @ <unknown line number>]
00007FFC656396AD00007FFC656395A0 torch_cpu.dll!at::native::copy_ [<unknown file> @ <unknown line number>]
00007FFC6607666B00007FFC66076500 torch_cpu.dll!at::_ops::copy_::call [<unknown file> @ <unknown line number>]
00007FFC6584D35E00007FFC6584CE90 torch_cpu.dll!at::native::_to_copy [<unknown file> @ <unknown line number>]
00007FFC6648A55600007FFC664889D0 torch_cpu.dll!at::compositeexplicitautograd::view_copy_symint_outf [<unknown file> @ <unknown line number>]
00007FFC6645D06C00007FFC6641D610 torch_cpu.dll!at::compositeexplicitautograd::bucketize_outf [<unknown file> @ <unknown line number>]
00007FFC65C166F400007FFC65B70CD0 torch_cpu.dll!at::TensorMaker::make_tensor [<unknown file> @ <unknown line number>]
00007FFC65CDAE4C00007FFC65CDAD70 torch_cpu.dll!at::_ops::_to_copy::redispatch [<unknown file> @ <unknown line number>]
00007FFC6621FDF200007FFC6620E7D0 torch_cpu.dll!at::_ops::view_as_real::redispatch [<unknown file> @ <unknown line number>]
00007FFC66227FCC00007FFC6620E7D0 torch_cpu.dll!at::_ops::view_as_real::redispatch [<unknown file> @ <unknown line number>]
00007FFC65C4BD5400007FFC65C4BB40 torch_cpu.dll!at::_ops::_to_copy::call [<unknown file> @ <unknown line number>]
00007FFC658560FE00007FFC65855BB0 torch_cpu.dll!at::native::to_dense_backward [<unknown file> @ <unknown line number>]
00007FFC6585586C00007FFC65855770 torch_cpu.dll!at::native::to [<unknown file> @ <unknown line number>]
00007FFC66621EDD00007FFC6661B0B0 torch_cpu.dll!at::compositeimplicitautograd::where [<unknown file> @ <unknown line number>]
00007FFC6660312200007FFC665C8970 torch_cpu.dll!at::compositeimplicitautograd::broadcast_to_symint [<unknown file> @ <unknown line number>]
00007FFC65E0F6CC00007FFC65E0F500 torch_cpu.dll!at::_ops::to_device::call [<unknown file> @ <unknown line number>]
00007FFC652F174300007FFC652F1710 torch_cpu.dll!at::Tensor::to [<unknown file> @ <unknown line number>]
00007FFD5C7C1DCF00007FFD5C6E1D40 torch_python.dll!THPPointer<_object>::get [<unknown file> @ <unknown line number>]
00007FFD5C7AF5F700007FFD5C6E1D40 torch_python.dll!THPPointer<_object>::get [<unknown file> @ <unknown line number>]
00007FFD7592000000007FFD7591F8E0 python310.dll!PyMem_RawMalloc [<unknown file> @ <unknown line number>]
00007FFD75927B2000007FFD759271D0 python310.dll!PyEval_EvalFrameDefault [<unknown file> @ <unknown line number>]
00007FFD759280ED00007FFD759271D0 python310.dll!PyEval_EvalFrameDefault [<unknown file> @ <unknown line number>]

Press any key to continue . . .

Голосуйте - палец вниз.0Голосуйте - палец вверх.0
Георгий@time2be
10 СообщенийАвтор темы
Pro
#2 · 11.08.2025, 02:04

а вот фокус что говорит:

Спойлер

To create a public link, set `share=True` in `launch()`.
extra {'cond_stage_model.clip_l.logit_scale', 'cond_stage_model.clip_l.text_projection'}
left over keys: dict_keys(['cond_stage_model.clip_l.transformer.text_model.embeddings.position_ids'])
Base model loaded: D:\_Models\SDXL\realisticStockPhoto_v20.safetensors
VAE loaded: None
Request to load LoRAs [('SDXL_FILM_PHOTOGRAPHY_STYLE_V1.safetensors', 0.25)] for model [D:\_Models\SDXL\realisticStockPhoto_v20.safetensors].
IMPORTANT: You are using gradio version 3.41.2, however version 4.44.1 is available, please upgrade.
--------
IMPORTANT: You are using gradio version 3.41.2, however version 4.44.1 is available, please upgrade.
--------
Loaded LoRA [D:\_Lora\SDXL\SDXL_FILM_PHOTOGRAPHY_STYLE_V1.safetensors] for UNet [D:\_Models\SDXL\realisticStockPhoto_v20.safetensors] with 722 keys at weight 0.25.
Loaded LoRA [D:\_Lora\SDXL\SDXL_FILM_PHOTOGRAPHY_STYLE_V1.safetensors] for CLIP [D:\_Models\SDXL\realisticStockPhoto_v20.safetensors] with 264 keys at weight 0.25.
Fooocus V2 Expansion: Vocab with 642 words.
Fooocus Expansion engine loaded for cuda:0, use_fp16 = True.
Requested to load SDXLClipModel
Requested to load GPT2LMHeadModel
Loading 2 new models
Exception in thread Thread-2 (worker):
Traceback (most recent call last):
File "threading.py", line 1016, in _bootstrap_inner
File "threading.py", line 953, in run
File "D:\_FE\Fooocus\modules\async_worker.py", line 293, in worker
import modules.default_pipeline as pipeline
File "D:\_FE\Fooocus\modules\default_pipeline.py", line 270, in <module>
refresh_everything(
File "D:\_FE\python_embeded\lib\site-packages\torch\utils\_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
File "D:\_FE\python_embeded\lib\site-packages\torch\utils\_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
File "D:\_FE\Fooocus\modules\default_pipeline.py", line 265, in refresh_everything
prepare_text_encoder(async_call=True)
File "D:\_FE\python_embeded\lib\site-packages\torch\utils\_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
File "D:\_FE\python_embeded\lib\site-packages\torch\utils\_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
File "D:\_FE\Fooocus\modules\default_pipeline.py", line 228, in prepare_text_encoder
ldm_patched.modules.model_management.load_models_gpu([final_clip.patcher, final_expansion.patcher])
File "D:\_FE\Fooocus\modules\patch.py", line 447, in patched_load_models_gpu
y = ldm_patched.modules.model_management.load_models_gpu_origin(*args, **kwargs)
File "D:\_FE\Fooocus\ldm_patched\modules\model_management.py", line 437, in load_models_gpu
cur_loaded_model = loaded_model.model_load(lowvram_model_memory)
File "D:\_FE\Fooocus\ldm_patched\modules\model_management.py", line 304, in model_load
raise e
File "D:\_FE\Fooocus\ldm_patched\modules\model_management.py", line 300, in model_load
self.real_model = self.model.patch_model(device_to=patch_model_to) #TODO: do something with loras and offloading to CPU
File "D:\_FE\Fooocus\ldm_patched\modules\model_patcher.py", line 202, in patch_model
out_weight = self.calculate_weight(self.patches[key], temp_weight, key).to(weight.dtype)
File "D:\_FE\Fooocus\modules\patch.py", line 78, in calculate_weight_patched
mat1 = ldm_patched.modules.model_management.cast_to_device(v[0], weight.device, torch.float32)
File "D:\_FE\Fooocus\ldm_patched\modules\model_management.py", line 617, in cast_to_device
return tensor.to(device, non_blocking=non_blocking).to(dtype, non_blocking=non_blocking)
RuntimeError: CUDA error: an illegal memory access was encountered
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

Голосуйте - палец вниз.0Голосуйте - палец вверх.0
Mikhail@satirov
13 439 Сообщений
Высшая репутация
Pro
#3 · 11.08.2025, 03:40

Георгий, Привет.
Тут твикером и не пахнет даже, первая ошибка ругается на PyTorch, CUDA и драйвера видеокарты, не может получить доступ к памяти, вторая ошибка что-то с моделями, пишет SDXLClipModel и GPT2LMHeadModel.

Голосуйте - палец вниз.0Голосуйте - палец вверх.0
«Любой дурак может знать. Дело в том, чтобы понять.» — Альберт Эйнштейн
Profile photo ofAndrey
Андрей@andrey
4 110 Сообщений
Админ
Pro
#4 · 11.08.2025, 08:28

Добрый день.

И что значит "после твикера"?
Вы посмотрели на твикер? Вы потрогали его? Вы обругали твикер? 
Что конкретно произошло? Причинно-следственная связь.

Голосуйте - палец вниз.0Голосуйте - палец вверх.0
Голова дана человеку не для того, чтобы кушать, а для того, чтобы думать.
Георгий@time2be
10 СообщенийАвтор темы
Pro
#5 · 11.08.2025, 11:12
Цитата: Mikhail от 11.08.2025, 03:40

Георгий, Привет.
Тут твикером и не пахнет даже, первая ошибка ругается на PyTorch, CUDA и драйвера видеокарты, не может получить доступ к памяти, вторая ошибка что-то с моделями, пишет SDXLClipModel и GPT2LMHeadModel.

до того как сделаол твики, почти все, все работало - а потом упс - вот и пытаюсь понять что не так произошло

Голосуйте - палец вниз.0Голосуйте - палец вверх.0
Георгий@time2be
10 СообщенийАвтор темы
Pro
#6 · 11.08.2025, 11:15
Цитата: Андрей от 11.08.2025, 08:28

Добрый день.

И что значит "после твикера"?
Вы посмотрели на твикер? Вы потрогали его? Вы обругали твикер? 
Что конкретно произошло? Причинно-следственная связь.

Добрый день! Я его купил, поставил, сделал пости все твики. увы, интерфейс не позволяет мне понять какие сделаны, какие нет... потом сделал "откакт" то есть отменил изменения... но тоже не помогло. вот не понимаю что делать

Голосуйте - палец вниз.0Голосуйте - палец вверх.0
Георгий@time2be
10 СообщенийАвтор темы
Pro
#7 · 11.08.2025, 13:09

кажется как то повредился .net - переставил и все ожило. снести тему?

Голосуйте - палец вниз.0Голосуйте - палец вверх.0