Crashing after faceswap on Colab Pro

[sitzbrau]

Read Troubleshoot

[x] I confirm that I have read the Troubleshoot guide before making this issue.

Describe the problem
Fooocus is running without problems on Google Colab Pro, V100 GPU, but when i enable the faceswap application display error and stop

edit: solution
https://github.com/lllyasviel/Fooocus/issues/2327#issuecomment-1963649657

Full Console Log
Paste the full console log here. You will make our job easier if you give a full log.

[Parameters] Adaptive CFG = 7
[Parameters] Sharpness = 2
[Parameters] ADM Scale = 1.5 : 0.8 : 0.3
[Parameters] CFG = 4.0
[Parameters] Seed = 7248250527830904799
[Fooocus] Downloading control models ...
[Fooocus] Loading control models ...
extra clip vision: ['vision_model.embeddings.position_ids']
[Parameters] Sampler = dpmpp_2m_sde_gpu - karras
[Parameters] Steps = 30 - 15
[Fooocus] Initializing ...
[Fooocus] Loading models ...
Refiner unloaded.
model_type EPS
UNet ADM Dimension 2816
Using pytorch attention in VAE
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
Using pytorch attention in VAE
extra {'cond_stage_model.clip_l.text_projection', 'cond_stage_model.clip_l.logit_scale', 'cond_stage_model.clip_g.transformer.text_model.embeddings.position_ids'}
Base model loaded: /content/drive/MyDrive/Fooocus/models/checkpoints/poltergeistXL_v10.safetensors
Request to load LoRAs [['sd_xl_offset_example-lora_1.0.safetensors', 0.1], ['None', 1.0], ['None', 1.0], ['None', 1.0], ['None', 1.0]] for model [/content/drive/MyDrive/Fooocus/models/checkpoints/poltergeistXL_v10.safetensors].
Loaded LoRA [/content/drive/MyDrive/Fooocus/models/loras/sd_xl_offset_example-lora_1.0.safetensors] for UNet [/content/drive/MyDrive/Fooocus/models/checkpoints/poltergeistXL_v10.safetensors] with 788 keys at weight 0.1.
Requested to load SDXLClipModel
Loading 1 new model
^C

[mashb1t]

Please check the resource allocation in the Colab playbook, upper right, in the graphs and send a screenshot showing the numbers before and after it crashed. Thanks!

[sitzbrau]

Please check the resource allocation in the Colab playbook, upper right, in the graphs and send a screenshot showing the numbers before and after it crashed. Thanks!

https://imgur.com/a/x8NSicO

App started successful. Use the app with http://127.0.0.1:7865/ or 127.0.0.1:7865 or https://254f4ddb8853f0cf79.gradio.live/
[Parameters] Adaptive CFG = 7
[Parameters] Sharpness = 2
[Parameters] ADM Scale = 1.5 : 0.8 : 0.3
[Parameters] CFG = 4.0
[Parameters] Seed = 880228361787244384
[Fooocus] Downloading control models ...
[Fooocus] Loading control models ...
extra clip vision: ['vision_model.embeddings.position_ids']
[Parameters] Sampler = dpmpp_2m_sde_gpu - karras
[Parameters] Steps = 60 - 30
[Fooocus] Initializing ...
[Fooocus] Loading models ...
Refiner unloaded.
model_type EPS
UNet ADM Dimension 2816
Using pytorch attention in VAE
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
Using pytorch attention in VAE
extra {'cond_stage_model.clip_l.text_projection', 'cond_stage_model.clip_g.transformer.text_model.embeddings.position_ids', 'cond_stage_model.clip_l.logit_scale'}
Base model loaded: /content/drive/MyDrive/Fooocus/models/checkpoints/poltergeistXL_v12.safetensors
Request to load LoRAs [['sd_xl_offset_example-lora_1.0.safetensors', 0.1], ['None', 1.0], ['None', 1.0], ['None', 1.0], ['None', 1.0]] for model [/content/drive/MyDrive/Fooocus/models/checkpoints/poltergeistXL_v12.safetensors].
Loaded LoRA [/content/drive/MyDrive/Fooocus/models/loras/sd_xl_offset_example-lora_1.0.safetensors] for UNet [/content/drive/MyDrive/Fooocus/models/checkpoints/poltergeistXL_v12.safetensors] with 788 keys at weight 0.1.
Requested to load SDXLClipModel
Loading 1 new model
^C

[mashb1t]

@sitzbrau please check if applying these parameters solves the issues for you: https://github.com/lllyasviel/Fooocus/pull/1710/files
#1710

--share --attention-split --disable-offload-from-vram --always-high-vram

[sitzbrau]

@sitzbrau please check if applying these parameters solves the issues for you: https://github.com/lllyasviel/Fooocus/pull/1710/files

--share --attention-split --disable-offload-from-vram --always-high-vram

now it's working, thanks