Running on an RTX 2070 graphics card causes an error, incompatible with bf16 models
got prompt
WARNING: [Errno 2] No such file or directory: 'E:\Comfyui_test\ComfyUI_cu128_15_0\ComfyUI\input\clipspace/clipspace-painted-masked-1772420453967.png'
Using pytorch attention in VAE
Using pytorch attention in VAE
VAE load device: cuda:0, offload device: cpu, dtype: torch.float16
😺dzNodes: LayerStyle -> MaskGrow Processed 1 mask(s).
Found quantization metadata version 1
Using MixedPrecisionOps for text encoder
CLIP/text encoder model load device: cuda:0, offload device: cpu, current: cpu, dtype: torch.float16
samples.shape torch.Size([1, 3, 1024, 1024])
sample_masks.shape torch.Size([1, 3, 1024, 1024])
noise_mask.shape torch.Size([1, 1024, 1024])
Requested to load WanVAE
loaded completely; 3542.00 MB usable, 242.03 MB loaded, full load: True
Requested to load QwenImageTEModel_
FETCH ComfyRegistry Data: 10/133
loaded partially; 5704.80 MB usable, 5250.55 MB loaded, 2659.73 MB offloaded, 453.25 MB buffer reserved, lowvram patches: 0
VRAM < 15GiB, enabling CPU offload
WARNING, you accessed scaled_fp8 from the model config object which doesn't exist. Please fix your code.
model weight dtype torch.bfloat16, manual cast: torch.float32
model_type FLUX
Requested to load NunchakuQwenImage
0%| | 0/4 [00:00<?, ?it/s]Assertion failed: isTypeMatch(_in_feats.dtype()), file C:\Users\nunchaku\Desktop\actions-runner_work\nunchaku\nunchaku\src\kernels\awq\gemv_awq.cu, line 254