AttributeError: 'ImageEncoder' object has no attribute 'img_size'

[BenDes21]

Any solution ?

[Celtmant]

same problem

[Celtmant]

Any solution ?

In general, the problem is in Sam 2, I switched to the Sam 1 loader and everything works on the model from 2 gig,

[neverbiasu]

Could u show me your fully traceback?

[SaintLandline]

!!! Exception during processing!!! 'ImageEncoder' object has no attribute 'img_size'
Traceback (most recent call last):
  File "C:\ComfyUI_windows_portable\ComfyUI\execution.py", line 152, in recursive_execute
    output_data, output_ui = get_output_data(obj, input_data_all)
                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\ComfyUI_windows_portable\ComfyUI\execution.py", line 82, in get_output_data
    return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)
                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\ComfyUI_windows_portable\ComfyUI\execution.py", line 75, in map_node_over_list
    results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\ComfyUI_windows_portable\ComfyUI\custom_nodes\comfyui_segment_anything\node.py", line 325, in main
    (images, masks) = sam_segment(
                      ^^^^^^^^^^^^
  File "C:\ComfyUI_windows_portable\ComfyUI\custom_nodes\comfyui_segment_anything\node.py", line 240, in sam_segment
    predictor = SamPredictorHQ(sam_model, sam_is_hq)
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\ComfyUI_windows_portable\ComfyUI\custom_nodes\comfyui_segment_anything\sam_hq\predictor.py", line 22, in __init__
    super().__init__(sam_model=sam_model)
  File "C:\ComfyUI_windows_portable\python_embeded\Lib\site-packages\segment_anything\predictor.py", line 31, in __init__
    self.transform = ResizeLongestSide(sam_model.image_encoder.img_size)
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\ComfyUI_windows_portable\python_embeded\Lib\site-packages\torch\nn\modules\module.py", line 1709, in __getattr__
    raise AttributeError(f"'{type(self).__name__}' object has no attribute '{name}'")
AttributeError: 'ImageEncoder' object has no attribute 'img_size'

[coollofty]

same problem, who solved?

[Maelstrom2014]

same problem

[neverbiasu]

U r using comfyui_segment_anything node, I'll change our nodes name to distinguish it from the former.

[coollofty]

U r using comfyui_segment_anything node, I'll change our nodes name to distinguish it from the former.

It's not a problem with the name. I changed the name in the code and confirmed that the wrong node was not used, but the error still persists

NODE_CLASS_MAPPINGS = {
    'SAM2ModelLoader (sam2)': SAM2ModelLoader,
    'GroundingDinoModelLoader (sam2)': GroundingDinoModelLoader,
    'GroundingDinoSAMSegment (sam2)': GroundingDinoSAMSegment,
    'InvertMask (sam2)': InvertMask,
    "IsMaskEmpty (sam2)": IsMaskEmptyNode,
}

NODE_DISPLAY_NAME_MAPPINGS = {
    'SAM2ModelLoader': 'SAM2 Model Loader',
    'GroundingDinoModelLoader': 'Grounding Dino Model Loader',
    'GroundingDinoSAMSegment': 'Grounding Dino SAM Segment',
    'InvertMask': 'Invert Mask',
    "IsMaskEmpty": "Is Mask Empty",
}

Error during runtime:

Traceback (most recent call last): File "F:\ComfyUI\ComfyUI\execution.py", line 323, in execute output_data, output_ui, has_subgraph = get_output_data(obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\ComfyUI\execution.py", line 198, in get_output_data return_values = _map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\ComfyUI\execution.py", line 169, in _map_node_over_list process_inputs(input_dict, i) File "F:\ComfyUI\ComfyUI\execution.py", line 158, in process_inputs results.append(getattr(obj, func)(**inputs)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\ComfyUI\custom_nodes\ComfyUI-SAM2\node.py", line 311, in main (images, masks) = sam_segment( ^^^^^^^^^^^^ File "F:\ComfyUI\ComfyUI\custom_nodes\ComfyUI-SAM2\node.py", line 224, in sam_segment predictor = SAM2ImagePredictor(sam_model) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\ComfyUI\custom_nodes\ComfyUI-SAM2\sam2\sam2_image_predictor.py", line 41, in init self._transforms = SAM2Transforms( ^^^^^^^^^^^^^^^ File "F:\ComfyUI\ComfyUI\custom_nodes\ComfyUI-SAM2\sam2\utils\transforms.py", line 28, in init self.transforms = torch.jit.script( ^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\python_embeded\Lib\site-packages\torch\jit_script.py", line 1432, in script return _script_impl( ^^^^^^^^^^^^^ File "F:\ComfyUI\python_embeded\Lib\site-packages\torch\jit_script.py", line 1146, in _script_impl return torch.jit._recursive.create_script_module( ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\python_embeded\Lib\site-packages\torch\jit_recursive.py", line 556, in create_script_module concrete_type = get_module_concrete_type(nn_module, share_types) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\python_embeded\Lib\site-packages\torch\jit_recursive.py", line 505, in get_module_concrete_type concrete_type = concrete_type_store.get_or_create_concrete_type(nn_module) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\python_embeded\Lib\site-packages\torch\jit_recursive.py", line 437, in get_or_create_concrete_type concrete_type_builder = infer_concrete_type_builder(nn_module) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\python_embeded\Lib\site-packages\torch\jit_recursive.py", line 286, in infer_concrete_type_builder sub_concrete_type = get_module_concrete_type(item, share_types) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\python_embeded\Lib\site-packages\torch\jit_recursive.py", line 505, in get_module_concrete_type concrete_type = concrete_type_store.get_or_create_concrete_type(nn_module) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\python_embeded\Lib\site-packages\torch\jit_recursive.py", line 437, in get_or_create_concrete_type concrete_type_builder = infer_concrete_type_builder(nn_module) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\python_embeded\Lib\site-packages\torch\jit_recursive.py", line 397, in infer_concrete_type_builder attr_type, inferred = infer_type(name, value) ^^^^^^^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\python_embeded\Lib\site-packages\torch\jit_recursive.py", line 236, in infer_type attr_type = torch._C._jit_try_infer_type(item) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\python_embeded\Lib\site-packages\torch\jit\annotations.py", line 492, in try_ann_to_type scripted_class = torch.jit._script._recursive_compile_class(ann, loc) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\python_embeded\Lib\site-packages\torch\jit_script.py", line 1602, in _recursive_compile_class rcb = _jit_internal.createResolutionCallbackForClassMethods(obj) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\python_embeded\Lib\site-packages\torch_jit_internal.py", line 473, in createResolutionCallbackForClassMethods captures.update(get_type_hint_captures(fn)) ^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\ComfyUI\python_embeded\Lib\site-packages\torch_jit_internal.py", line 387, in get_type_hint_captures raise OSError( OSError: Failed to get source for <function Enum._generate_nextvalue at 0x00000128B33F9080> using inspect.getsource

[drake7707]

Having the same problem:

2024-09-24 13:16:45.202028 #28 [GroundingDinoModelLoader (segment anything)]: 4.87s
!!! Exception during processing !!! 'ImageEncoder' object has no attribute 'img_size'
Traceback (most recent call last):
  File "/comfyui/execution.py", line 317, in execute
    output_data, output_ui, has_subgraph = get_output_data(obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb)
                                           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/comfyui/execution.py", line 192, in get_output_data
    return_values = _map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb)
                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/comfyui/execution.py", line 169, in _map_node_over_list
    process_inputs(input_dict, i)
  File "/comfyui/execution.py", line 158, in process_inputs
    results.append(getattr(obj, func)(**inputs))
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/comfyui/custom_nodes/comfyui_segment_anything/node.py", line 325, in main
    (images, masks) = sam_segment(
                      ^^^^^^^^^^^^
  File "/comfyui/custom_nodes/comfyui_segment_anything/node.py", line 240, in sam_segment
    predictor = SamPredictorHQ(sam_model, sam_is_hq)
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/comfyui/custom_nodes/comfyui_segment_anything/sam_hq/predictor.py", line 22, in __init__
    super().__init__(sam_model=sam_model)
  File "/opt/conda/lib/python3.11/site-packages/segment_anything/predictor.py", line 31, in __init__
    self.transform = ResizeLongestSide(sam_model.image_encoder.img_size)
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/conda/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1729, in __getattr__
    raise AttributeError(f"'{type(self).__name__}' object has no attribute '{name}'")
AttributeError: 'ImageEncoder' object has no attribute 'img_size'

Diving deeper into the code it's because SAM2 uses sam2.modeling.backbones.image_encoder whereas the SAM1 uses ImageEncoderViTHQ and TinyViT which both have an img_size property, which is apparently set to 1024 by default

[drake7707]

Oh ok I understand what's going on. I have to uninstall the segment anything custom nodes so it's a naming issue. Once the nodes use all the nodes from this repo it works.

[coollofty]

Oh ok I understand what's going on. I have to uninstall the segment anything custom nodes so it's a naming issue. Once the nodes use all the nodes from this repo it works.

No, it's not you understand..... Look at this workflow, it uses all the nodes from this repo as mentioned above, and I have changed the names of all the nodes, so the nodes won't be used incorrectly, but the errors same

[drake7707]

I had both ComfyUI-segment-anything-2 and comfyui_segment_anything in my custom nodes. Both of them have naming collisions with this repo. When I moved them out of my custom_nodes and restarted everything worked