TypeError: 'NoneType' object is not subscriptable - 🅛🅣🅧 Add Video IC-LoRA Guide
I'm trying your workflow, but i'm stuck for time_scale_factor = scale_factors[0] on TypeError: 'NoneType' object is not subscriptable in 🅛🅣🅧 Add Video IC-LoRA Guide
TypeError: 'NoneType' object is not subscriptable
File "/execution.py", line 534, in execute
output_data, output_ui, has_subgraph, has_pending_tasks = await get_output_data(prompt_id, unique_id, obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, v3_data=v3_data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/execution.py", line 334, in get_output_data
return_values = await _async_map_node_over_list(prompt_id, unique_id, obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, v3_data=v3_data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/execution.py", line 308, in _async_map_node_over_list
await process_inputs(input_dict, i)
File "/execution.py", line 296, in process_inputs
result = f(**inputs)
^^^^^^^^^^^
File "/comfy_api/internal/init.py", line 149, in wrapped_func
return method(locked_class, **inputs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/comfy_api/latest/_io.py", line 1826, in EXECUTE_NORMALIZED
to_return = cls.execute(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/custom_nodes/ComfyUI-LTXVideo/iclora.py", line 166, in execute
time_scale_factor = scale_factors[0]
~~~~~~~~~~~~~^^^
Try updating ComfyUI-LTXVideo nodes.
If it's not that try to run other workflows, if they work then take the IC lora and guide and add them to that working WF. Also need to know what Eros model version and what distilled lora used too.
If it's not that try to run other workflows, if they work then take the IC lora and guide and add them to that working WF. Also need to know what Eros model version and what distilled lora used too.
it was the v1 fp8 from 5 days ago, about loras i tried everything, 1.1 ceil36, ceil72 and also condasafe. In theory now i can only try the beefy bf16 and the new one fp8 mixed you uploaded 22 hours ago.
Update
I tried the BF16 and it seems work well, but i didn't test the audio, only video.
Only a question: inside the nodes "LTXV audio vae loader" and "LTXV audio text encoder loader" what kind of checkpoint I have to use? During the first run i used the common LTX2_audio_vae_bf16.safetensors and i got the error. During the second run i used your beefy 10Eros_v1_bf16.safetensors inside all 3 nodes (checkpoint, vae and text encoder together with gemma3), because you wrote "BF16 loads as a checkpoint with clip and VAEs". Am I doing right?
The old fp8 is problematic on some versions or setups I think. worked for me and some others but also had those issues. The new fp8 is better in my opinion anyways with better audio. That new one and bf16 are normal checkpoints that load like ltx checkpoints do:
