control-lora-canny-rank256 controlnet-canny-sdxl-1.0 t2i-adapter_xl_canny 预处理器 canny invert (from white bg & black line) 黑底白线的稿子需要使用这个反转 Depth深度 简要说明 根据输入的图像,猜测深度信息适用于对场景的描绘还原尤其是富有多层次的空间图形白近黑远 原图 深度图 模型/预处理器 模型...
huggingface.co/stabilityai/control-lora/tree/main/control-LoRAs-rank128https://huggingface.co/stabilityai/control-lora/tree/main/control-LoRAs-rank256 stabilityai/control-lora at main 文件名: control-lora-canny-rank128.safetensors control-lora-depth-rank128.safetensors control-lora-canny-rank256....
"https://huggingface.co/stabilityai/control-lora/resolve/main/control-LoRAs-rank256/control-lora-canny-rank256.safetensors", "https://huggingface.co/stabilityai/control-lora/resolve/main/control-LoRAs-rank256/control-lora-depth-rank256.safetensors", ...
self.device = device if rank > (new_rank := min(self.out_features, self.in_features)): self.rank = new_rank self.lora_A = nn.Linear( in_features=in_features, out_features=self.rank, bias=False, dtype=dtype, device=device, ) self.lora_B = nn.Linear( in_features=self.rank, out...
The file sizes of these Control-LoRA are pretty reasonable: about 400 MB and 800 MB. A control weight of around 0.75 seems to be the sweet spot. The 128 and 256-rank LoRA perform very similarly. You can use the 128 variant if you want to conserve space. ...
depth): "sai_xl_depth_256lora.safetensors", resource_id(ResourceKind.controlnet, SDVersion.sd15, ControlMode.normal): None, resource_id(ResourceKind.controlnet, SDVersion.sd15, ControlMode.pose): "control_lora_rank128_v11p_sd15_openpose_fp16.safetensors", resource_id(ResourceKind.control...
ComfyUI Recommended Parameters sampler steps:30 CFG:7.0 sampler_name:dpmpp_2m_sde scheduler:karras denoise:0.93 controlnet_strength:1.0 stargt_percent:0.0 end_percent:0.9 Checkpoints mistoLine_rank256.safetensors : General usage version, for ComfyUI and AUTOMATIC1111-WebUI. mistoLine_fp16.safe...
mistoLine_rank256 control_v11p_sd15s2_lineart_anime control_v11p_sd15_lineart Normal and Depth Estimators Preprocessor Nodesd-webui-controlnet/otherControlNet/T2I-Adapter MiDaS Depth Map(normal) depthcontrol_v11f1p_sd15_depth control_depth ...
rank_zero_deprecation( Launching Web UI with arguments: ONNX: selected=CUDAExecutionProvider, available=['AzureExecutionProvider', 'CPUExecutionProvider'] Tag Autocomplete: Could not locate model-keyword extension, Lora trigger word completion will be limited to those added through the extra networks ...
zero.GatheredParameters(param, modifier_rank=None) for deepspeed support. it's starting to feel like we need an autocast manager wrapper in train utils 👍 1 examples/instruct_pix2pix/train_instruct_pix2pix_sdxl.py Outdated Show resolved examples/instruct_pix2pix/train_instruct_pix2pix_sd...