yuvraj108c/comfyui-depth-anything-tensorrt

ComfyUI Depth Anything (v1/v2) Tensorrt Custom Node (up to 14x faster), licensed under CC BY-NC-SA 4.0

comfyui-depth-anything-tensorrt's Issues

I have no problem ,I just want to thank you

it's very helpful node,thanks a lot

Trouble installing requirements

After a fresh install of the module from the ComfyUI manager, I followed the instructions and ran pip install -r requirements.txt
However, this is the error I received when running pip

Looking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com
Collecting cuda-python (from -r requirements.txt (line 1))
  Downloading cuda_python-12.4.0-cp310-cp310-win_amd64.whl.metadata (12 kB)
Collecting tensorrt (from -r requirements.txt (line 2))
  Downloading tensorrt-8.6.1.post1.tar.gz (18 kB)
  Preparing metadata (setup.py) ... error
  error: subprocess-exited-with-error

  × python setup.py egg_info did not run successfully.
  │ exit code: 1
  ╰─> [16 lines of output]
      Traceback (most recent call last):
        File "<string>", line 2, in <module>
        File "<pip-setuptools-caller>", line 34, in <module>
        File "C:\Users\Adrian\AppData\Local\Temp\pip-install-r7pwndnd\tensorrt_14eb96209b3642c6b050758595e14a19\setup.py", line 103, in <module>
          if disable_internal_pip or nvidia_pip_index_url in parent_command_line() or nvidia_pip_index_url in pip_config_list():
        File "C:\Users\Adrian\AppData\Local\Temp\pip-install-r7pwndnd\tensorrt_14eb96209b3642c6b050758595e14a19\setup.py", line 96, in parent_command_line
          return subprocess.check_output(["ps", "-p", str(pid), "-o", "command", "--no-headers"]).decode()
        File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.10_3.10.3056.0_x64__qbz5n2kfra8p0\lib\subprocess.py", line 421, in check_output
          return run(*popenargs, stdout=PIPE, timeout=timeout, check=True,
        File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.10_3.10.3056.0_x64__qbz5n2kfra8p0\lib\subprocess.py", line 503, in run
          with Popen(*popenargs, **kwargs) as process:
        File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.10_3.10.3056.0_x64__qbz5n2kfra8p0\lib\subprocess.py", line 971, in __init__
          self._execute_child(args, executable, preexec_fn, close_fds,
        File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.10_3.10.3056.0_x64__qbz5n2kfra8p0\lib\subprocess.py", line 1456, in _execute_child
          hp, ht, pid, tid = _winapi.CreateProcess(executable, args,
      FileNotFoundError: [WinError 2] The system cannot find the file specified
      [end of output]

  note: This error originates from a subprocess, and is likely not a problem with pip.
error: metadata-generation-failed

This is on Windows 11. Am I missing another dependency?

fix single image inference

Serialization assertion plan

Error: [TRT] [E] 1: [runtime.cpp::parsePlan::314] Error Code 1: Serialization (Serialization assertion plan->header.magicTag == rt::kPLAN_MAGIC_TAG failed.)
File "/root/ComfyUI/custom_nodes/ComfyUI-Depth-Anything-Tensorrt/init.py", line 38, in main

System Info: Ubuntu 20.04 LTS, Cuda 12.1, Tensorrt 8.6.1, Python 3.10, RTX3090 GPU

I used your code and I also tried to re-export the onnx model, but still this error, is it because of the cuda version?

Depth Anything V2

Thoughts on supporting this?

Model Reloading

Every subsequent run reloads the model into memory. For streaming purposes, it adds lots of latency. Ideally the model should be cached in memory.