got prompt
Using pytorch attention in VAE
Using pytorch attention in VAE
VAE load device: cuda:0, offload device: cpu, dtype: torch.bfloat16
SELECTED: input1
[MultiGPU Core Patching] text_encoder_device_patched returning device: cuda:0 (current_text_encoder_device=cuda:0)
CLIP/text encoder model load device: cuda:0, offload device: cpu, current: cpu, dtype: torch.float16
Requested to load QwenImageTEModel_
loaded partially 9827.8 9826.07177734375 0
0 models unloaded.
loaded partially 9826.071777439118 9826.07177734375 0
[MultiGPU Core Patching] Successfully patched ModelPatcher.partially_load
gguf qtypes: F32 (1087), BF16 (6), Q8_0 (840)
model weight dtype torch.bfloat16, manual cast: None
model_type FLUX
[MultiGPU DisTorch V2] Full allocation string: #cuda:0;16.0;cpu
SELECTED: input1
Requested to load QwenImage
===============================================
DisTorch2 Model Virtual VRAM Analysis
===============================================
Object Role Original(GB) Total(GB) Virt(GB)
-----------------------------------------------
cuda:0 recip 12.00GB 28.00GB +16.00GB
cpu donor 15.93GB 0.00GB -15.93GB
-----------------------------------------------
model model 20.27GB 4.27GB -16.00GB
[MultiGPU DisTorch V2] Model size (20.27GB) is larger than 90% of available VRAM on: cuda:0 (10.80GB).
[MultiGPU DisTorch V2] To prevent an OOM error, set 'virtual_vram_gb' to at least 9.47.
==================================================
[MultiGPU DisTorch V2] Final Allocation String:
cuda:0,0.3556;cpu,1.0000
==================================================
DisTorch2 Model Device Allocations
==================================================
Device VRAM GB Dev % Model GB Dist %
--------------------------------------------------
cuda:0 12.00 35.6% 4.27 21.1%
cpu 15.93 100.0% 15.93 78.9%
--------------------------------------------------
DisTorch2 Model Layer Distribution
--------------------------------------------------
Layer Type Layers Memory (MB) % Total
--------------------------------------------------
Linear 846 20861.41 100.0%
RMSNorm 241 0.07 0.0%
LayerNorm 241 0.00 0.0%
--------------------------------------------------
DisTorch2 Model Final Device/Layer Assignments
--------------------------------------------------
Device Layers Memory (MB) % Total
--------------------------------------------------
cuda:0 (<0.01%) 485 2.34 0.0%
cuda:0 182 4546.11 21.8%
cpu 661 16313.03 78.2%
--------------------------------------------------
[MultiGPU DisTorch V2] DisTorch loading completed.
[MultiGPU DisTorch V2] Total memory: 20861.48MB
100%|███████████████████████████████████████████████████████████████████████████████| 30/30 [1:15:56<00:00, 151.87s/it]
大佬,真没招了,一张图出了1个小时15分钟,是哪里设置不对还是配置不够啊,16G内存,RTX3060,12G显存