Rewrite prediction.py - lazy load one model at a time

Fix predict signature

Lazy load one model at a time - fixes segfault on HF free tier

Fix relative path issues for scores.json and params.yaml

Remove last artifacts/training reference

Fix artifacts path and inputImage path for HF

Fix all checkpoints paths to /tmp/checkpoints

Fix permission error - use /tmp/checkpoints

Add model download from HF Hub on startup

AWS updates and finish

CI/CD updated

biomedclip.py upload

app.py runs

prediction.py added

dvc repro testing

DenseNet121.keras modelre-run

ResNet50V2 model training

model runs

running 2nd model

new model run with 150 epochs and learning rate of 0.0003. Model checkpointing, early stopping and reduce learning rate on plateau callbacks have been added to the model training component. Model evaluation component has been updated to log validation loss and accuracy to mlflow and save it in scores.json file.

minor update

model_comparison

evaluation_updates

update evaluation

updates main.py

updates

model run successfully

updated model training

updating config and logs

pipeline updates

retest code progress

Pipeline design and code updates

utils added and updated

common.py update

testing activity log setup

requirements added (all required packages installed in environment)

folder structure added