support FluxDisableGuidance

PulidFluxHook.py

@@ -70,7 +70,7 @@ def __call__(self, input_args, extra_options):
         img = out['img']
         temp_img = img
         for i, callback in enumerate(self.callback):
-            if sigma <= self.kwargs[i]["sigma_start"] and sigma >= self.kwargs[i]["sigma_end"]:
+            if self.kwargs[i]["sigma_start"] >= sigma >= self.kwargs[i]["sigma_end"]:
                 img = img + callback(temp_img,
                                      pulid_model=self.kwargs[i]['pulid_model'],
                                      ca_idx=self.kwargs[i]['ca_idx'],
@@ -118,7 +118,7 @@ def __call__(self, input_args, extra_options):
         real_img, txt = img[:, txt.shape[1]:, ...], img[:, :txt.shape[1], ...]
         temp_img = real_img
         for i, callback in enumerate(self.callback):
-            if sigma <= self.kwargs[i]["sigma_start"] and sigma >= self.kwargs[i]["sigma_end"]:
+            if self.kwargs[i]["sigma_start"] >= sigma >= self.kwargs[i]["sigma_end"]:
                 real_img = real_img + callback(temp_img,
                                                pulid_model=self.kwargs[i]['pulid_model'],
                                                ca_idx=self.kwargs[i]['ca_idx'],
@@ -159,9 +159,8 @@ def pulid_forward_orig(
     img = self.img_in(img)
     vec = self.time_in(timestep_embedding(timesteps, 256).to(img.dtype))
     if self.params.guidance_embed:
-        if guidance is None:
-            raise ValueError("Didn't get guidance strength for guidance distilled model.")
-        vec = vec + self.guidance_in(timestep_embedding(guidance, 256).to(img.dtype))
+        if guidance is not None:
+            vec = vec + self.guidance_in(timestep_embedding(guidance, 256).to(img.dtype))
 
     vec = vec + self.vector_in(y)
     txt = self.txt_in(txt)
@@ -171,9 +170,6 @@ def pulid_forward_orig(
 
     blocks_replace = patches_replace.get("dit", {})
 
-    # pulid_temp_attrs = transformer_options.get(PatchKeys.pulid_patch_key_attrs, {})
-    # pulid_temp_attrs['timesteps'] = timesteps
-
     for i, block in enumerate(self.double_blocks):
         # 0 -> 18
         if ("double_block", i) in blocks_replace:
@@ -212,8 +208,6 @@ def block_wrap(args):
                 if add is not None:
                     img += add
 
-    # pulid_temp_attrs['double_blocks_txt'] = txt
-
     img = torch.cat((txt, img), 1)
 
     for i, block in enumerate(self.single_blocks):

README.md

@@ -3,6 +3,7 @@
 - Solved [ComfyUI-PuLID-Flux](https://github.com/balazik/ComfyUI-PuLID-Flux) model pollution problem.
 - Supported use with `TeaCache` (Need use with [ComfyUI_Patches_ll](https://github.com/lldacing/ComfyUI_Patches_ll)).
 - Supported use with [Comfy-WaveSpeed](https://github.com/chengzeyi/Comfy-WaveSpeed), supported by `Comfy-WaveSpeed` in [commit-36ba3c8](https://github.com/chengzeyi/Comfy-WaveSpeed/commit/36ba3c8b74735d4521828507a4bf323df1a9a9d0).
+- Supported simple use with `First Block Cache` (Can use with [ComfyUI_Patches_ll](https://github.com/lldacing/ComfyUI_Patches_ll)).
 
 Must uninstall or disable `ComfyUI-PuLID-Flux` and other PuLID-Flux nodes before install this plugin. Due to certain reasons, I used the same node's name `ApplyPulidFlux`.
 

README_CN.md

@@ -3,6 +3,7 @@
 - 解决插件 [ComfyUI-PuLID-Flux](https://github.com/balazik/ComfyUI-PuLID-Flux) 存在的模型污染问题。
 - 支持使用[TeaCache](https://github.com/ali-vilab/TeaCache)加速（`TeaCache`加速需要配合[ComfyUI_Patches_ll](https://github.com/lldacing/ComfyUI_Patches_ll)使用）。
 - 支持使用[Comfy-WaveSpeed](https://github.com/chengzeyi/Comfy-WaveSpeed)加速, Comfy-WaveSpeed在[提交记录-36ba3c8](https://github.com/chengzeyi/Comfy-WaveSpeed/commit/36ba3c8b74735d4521828507a4bf323df1a9a9d0)中提供支持。
+- 支持使用简单的`First Block Cache`加速（可以配合[ComfyUI_Patches_ll](https://github.com/lldacing/ComfyUI_Patches_ll)使用）。
 
 在安装此插件之前，必须卸载或禁用`ComfyUI-PuLID-Flux`和其他PuLID Flux节点, 因为由于某些原因，我使用了同样的节点名`ApplyPulidFlux`
 
@@ -36,7 +37,7 @@
   - 使用 [TeaCache](https://github.com/ali-vilab/TeaCache)加速, 必须加在[`FluxForwardOverrider` and `ApplyTeaCachePatch`](https://github.com/lldacing/ComfyUI_Patches_ll)之前.
   - 使用 [Comfy-WaveSpeed](https://github.com/chengzeyi/Comfy-WaveSpeed)加速, 必须加在[`ApplyFBCacheOnModel`](https://github.com/lldacing/ComfyUI_Patches_ll)之前.
 - FixPulidFluxPatch (已弃用)
-  - 如果想使用 [TeaCache](https://github.com/ali-vilab/TeaCache)加速, 必须~~加在 `ApplyPulidFlux` 节点之后, 并~~在后面连接节点 [`FluxForwardOverrider` and `ApplyTeaCachePatch`](https://github.com/lldacing/ComfyUI_Patches_ll).
+  - 如果想使用 [TeaCache](https://github.com/ali-vilab/TeaCache)加速, 必须加在 `ApplyPulidFlux` 节点之后, 并在后面连接节点 [`FluxForwardOverrider` and `ApplyTeaCachePatch`](https://github.com/lldacing/ComfyUI_Patches_ll).
 
 ## 感谢
 

pyproject.toml

@@ -1,7 +1,7 @@
 [project]
 name = "comfyui_pulid_flux_ll"
 description = "The implementation for PuLID-Flux, support use with TeaCache and WaveSpeed, no model pollution."
-version = "1.0.7"
+version = "1.0.8"
 license = {file = "LICENSE"}
 dependencies = ['facexlib', 'insightface', 'onnxruntime', 'onnxruntime-gpu; sys_platform != "darwin" and platform_machine == "x86_64"', 'ftfy', 'timm']