Debug an ML Script and Verify Implementation
Budget: $30 – $250 USD
I need help debugging an ML Script. Once the script is debugged, I need to verify the script is implemented correctly.
Debug error:
Run training file with python train.py. You will see concatenation error. Fix error.
Verify Implementation:
In record_data.py, only record full screen.
In super_data_loader.py, inside multi input (start from line 229) and under the return is 2 inputs: the video frames and the key strokes along with output y. Line 237 is the same but with images. At 252, you read in individual frames. You can now take subsection from frame and make it frame_speedometer and another subsection frame_minimap. Now in line 270, you return 3 variables/images. No need to save data separately. You may need to do function calling where one function calls to get all data. You can maybe do something in 232 section where you have array of inputs have 2 more inputs. Read one image and manipulate it in ram
In models.py, make input 3 and 4. In data_loader has to have same sequence as model.py. Add 2 more lines between 232 and 233. First one will be mini map and other will be speedometer. Then go to model.py, make keystrokes input 4 and make input 2 mini map and input 3 be speedometer. Sequence of inputs has to be maintained
Debug error:
Run training file with python train.py. You will see concatenation error. Fix error.
Verify Implementation:
In record_data.py, only record full screen.
In super_data_loader.py, inside multi input (start from line 229) and under the return is 2 inputs: the video frames and the key strokes along with output y. Line 237 is the same but with images. At 252, you read in individual frames. You can now take subsection from frame and make it frame_speedometer and another subsection frame_minimap. Now in line 270, you return 3 variables/images. No need to save data separately. You may need to do function calling where one function calls to get all data. You can maybe do something in 232 section where you have array of inputs have 2 more inputs. Read one image and manipulate it in ram
In models.py, make input 3 and 4. In data_loader has to have same sequence as model.py. Add 2 more lines between 232 and 233. First one will be mini map and other will be speedometer. Then go to model.py, make keystrokes input 4 and make input 2 mini map and input 3 be speedometer. Sequence of inputs has to be maintained