Windows PC ML Environment Setup
Budget: $1,500 – $3,000 USD
I'm seeking a professional to assist with setting up my Windows PC for machine learning tasks.
Key Responsibilities:
- Remote access to my PC to build an environment supporting ML tools, predominantly TensorFlow and PyTorch.
- Configuration of an image database and relevant file structure.
- Coding of a Python script to build upon an existing crack detection model, in the training of a custom model. See video tut: https://www.youtube.com/watch?v=vFGxM2KLs10
- Integration of an existing .json file that contains annotations in COCO 1.0.
- Testing of the script on unannotated imagery and generating output.
Requirements:
- Proficient in Python programming with extensive experience in machine learning tools.
- Familiarity with the Jupyter Notebook programming environment.
- Previous experience in setting up and configuring Windows PC for machine learning tasks is an advantage.
My PC: Windows 11 Pro, x64-based PC, 13th Gen Intel Core i9-13900K, 3000Mhz, 24 Cores, 32 Logical Processors, 192 GB RAM, RTX 4090 gpu
Regarding the subject matter, the imagery comes from uniformly lit, planar geometry, namely bird eye views to concrete bridge decks. The 7952x5304 rasters resolve down to .03mm features, the defect of interest being hairline cracks. We have as a starting point annotations using a poly line tool, the short term need being simply to highlight any cracks, independent of width or length. A longer term project may evolve upon reaching this first milestone, automating feature extraction, classification can wait.
Key Responsibilities:
- Remote access to my PC to build an environment supporting ML tools, predominantly TensorFlow and PyTorch.
- Configuration of an image database and relevant file structure.
- Coding of a Python script to build upon an existing crack detection model, in the training of a custom model. See video tut: https://www.youtube.com/watch?v=vFGxM2KLs10
- Integration of an existing .json file that contains annotations in COCO 1.0.
- Testing of the script on unannotated imagery and generating output.
Requirements:
- Proficient in Python programming with extensive experience in machine learning tools.
- Familiarity with the Jupyter Notebook programming environment.
- Previous experience in setting up and configuring Windows PC for machine learning tasks is an advantage.
My PC: Windows 11 Pro, x64-based PC, 13th Gen Intel Core i9-13900K, 3000Mhz, 24 Cores, 32 Logical Processors, 192 GB RAM, RTX 4090 gpu
Regarding the subject matter, the imagery comes from uniformly lit, planar geometry, namely bird eye views to concrete bridge decks. The 7952x5304 rasters resolve down to .03mm features, the defect of interest being hairline cracks. We have as a starting point annotations using a poly line tool, the short term need being simply to highlight any cracks, independent of width or length. A longer term project may evolve upon reaching this first milestone, automating feature extraction, classification can wait.