creae Dockerfile, update autotab_app relative paths in param.py, decreseasing requirements.txt

Dockerfile

@@ -11,8 +11,11 @@ COPY README.md /README.md
 COPY Procfile /Procfile
 COPY requirements.txt /requirements.txt
 COPY setup.py /setup.py
+COPY api/api_app.py /api/api_app.py
 
 RUN pip install --upgrade pip 
 RUN pip install -r requirements.txt
+RUN apt-get update -y && apt-get install -y --no-install-recommends build-essential gcc \
+                                        libsndfile1 
 
-CMD uvicorn api.simple:app --host 0.0.0.0 --port $PORT
+CMD uvicorn api.api_app:app --host 0.0.0.0 --port $PORT

autotab/param.py

@@ -1,7 +1,12 @@
 ###### ENVIRONMENT VARIABLES FOR AUTOTAB #########
+import os
 
 GCP = True
 BUCKET_NAME = "wagon-data-737-sadriwala"
 DATA_BUCKET_FOLDER = "data/spec_repr"
-LOCAL_DATA = "/mnt/d/data-science/le-wagon/autotab/data/spec_repr/"
-LOCAL_MODEL = "/mnt/d/data-science/le-wagon/autotab/h5-model/full_val0_75acc_weights.h5"
+LOCAL_DATA = os.path.abspath("data/spec_repr") + "/"
+LOCAL_MODEL = os.path.abspath("h5-model/full_val0_75acc_weights.h5")
+
+if __name__== '__main__':
+    print(LOCAL_DATA)
+    print(LOCAL_MODEL)

requirements.txt

@@ -10,128 +10,128 @@ streamlit
 # data science
 numpy
 pandas
-scikit-learn
+# scikit-learn
 tensorflow
 
-# tests/linter #optional
-black
-coverage
-flake8
-pytest
-yapf
+# # tests/linter #optional
+# black
+# coverage
+# flake8
+# pytest
+# yapf
 
-# API #optional
-gcsfs
-mlflow
-s3fs
+# # API #optional
+# gcsfs
+# mlflow
+# s3fs
 
-# utilities #optional
-six>=1.14
-joblib
-memoized-property
-termcolor
+# # utilities #optional
+# six>=1.14
+# joblib
+# memoized-property
+# termcolor
 
 #guitar-set #primary
 librosa
 jams
 sox
 
-#installed as minimal packages by Le Wagon #optional
-anyio
-argon2-cffi
-astroid
-attrs
-Babel
-backcall
-bleach
-certifi
-cffi
-charset-normalizer
-cycler
-debugpy
-decorator
-defusedxml
-entrypoints
-fonttools
-idna
-importlib-resources
-iniconfig
-ipdb
-ipykernel
-ipython
-ipython-genutils
-isort
-jedi
-Jinja2
-joblib
-json5
-jsonschema
-jupyter-client
-jupyter-core
-jupyter-server
-jupyterlab
-jupyterlab-pygments
-jupyterlab-server
-kiwisolver
-lazy-object-proxy
-MarkupSafe
-matplotlib
-matplotlib-inline
-mccabe
-mistune
-nbclassic
-nbclient
-nbconvert
-nbformat
-nest-asyncio
-notebook
-numpy
-packaging
-pandas
-pandocfilters
-parso
-pexpect
-pickleshare
-Pillow
-pip
-platformdirs
-pluggy
-prometheus-client
-prompt-toolkit
-ptyprocess
-py
-pycparser
-Pygments
-pylint
-pyparsing
-pyrsistent
-pytest
-python-dateutil
-pytz
-pyzmq
-requests
-scikit-learn
-scipy
-seaborn
-Send2Trash
-setuptools
-setuptools-scm
-six
-sniffio
-terminado
-testpath
-threadpoolctl
-toml
-tomli
-tornado
-traitlets
-typing_extensions
-urllib3
-wcwidth
-webencodings
-websocket-client
-wrapt
-zipp
+# #installed as minimal packages by Le Wagon #optional
+# anyio
+# argon2-cffi
+# astroid
+# attrs
+# Babel
+# backcall
+# bleach
+# certifi
+# cffi
+# charset-normalizer
+# cycler
+# debugpy
+# decorator
+# defusedxml
+# entrypoints
+# fonttools
+# idna
+# importlib-resources
+# iniconfig
+# ipdb
+# ipykernel
+# ipython
+# ipython-genutils
+# isort
+# jedi
+# Jinja2
+# joblib
+# json5
+# jsonschema
+# jupyter-client
+# jupyter-core
+# jupyter-server
+# jupyterlab
+# jupyterlab-pygments
+# jupyterlab-server
+# kiwisolver
+# lazy-object-proxy
+# MarkupSafe
+# matplotlib
+# matplotlib-inline
+# mccabe
+# mistune
+# nbclassic
+# nbclient
+# nbconvert
+# nbformat
+# nest-asyncio
+# notebook
+# numpy
+# packaging
+# pandas
+# pandocfilters
+# parso
+# pexpect
+# pickleshare
+# Pillow
+# pip
+# platformdirs
+# pluggy
+# prometheus-client
+# prompt-toolkit
+# ptyprocess
+# py
+# pycparser
+# Pygments
+# pylint
+# pyparsing
+# pyrsistent
+# pytest
+# python-dateutil
+# pytz
+# pyzmq
+# requests
+# scikit-learn
+# scipy
+# seaborn
+# Send2Trash
+# setuptools
+# setuptools-scm
+# six
+# sniffio
+# terminado
+# testpath
+# threadpoolctl
+# toml
+# tomli
+# tornado
+# traitlets
+# typing_extensions
+# urllib3
+# wcwidth
+# webencodings
+# websocket-client
+# wrapt
+# zipp
 
 uvicorn
 fastapi

streamlit/autotab_app.py

@@ -5,7 +5,6 @@ import pandas as pd
 import numpy as np 
 from autotab.TabDataReprGen import TabDataReprGen
 from PIL import Image
-import base64
 
 st.set_page_config(
     page_title="AutoTab tab generator",
@@ -43,29 +42,34 @@ CSS = """
 """
 st.write(f'<style>{CSS}</style>', unsafe_allow_html=True)
 
-# st.title('Model')
-
-model = tp.load_model_and_weights()
-model.load_weights('./h5-model/full_val0_75acc_weights.h5')
-
-# st.write(model.weights[0].shape)
 
 st.set_option('deprecation.showfileUploaderEncoding', False)
 
 uploaded_file = st.file_uploader("choose a music file:", type="wav")
 
+processed_file = None
+
 mode = st.radio('Choose Mode of Tab production:', ('squeezed notes', 'changed notes'))
 st.write('<style>div.row-widget.stRadio > div{flex-direction:row;}</style>', unsafe_allow_html=True)
 
+model = tp.load_model_and_weights()
+model.load_weights('./h5-model/full_val0_75acc_weights.h5')
+genrep = TabDataReprGen()
 
+if uploaded_file is not None:
+    x_new = genrep.load_rep_from_raw_file(uploaded_file)
+    y_pred = model.predict(x_new)
+    processed_file= uploaded_file
+    
 if uploaded_file is not None and mode == 'squeezed notes': 
     st.title("""
             The predicted squeezed Tabs: 
     """)
-    genrep = TabDataReprGen()
-    x_new = genrep.load_rep_from_raw_file(uploaded_file)
-    # st.write(x_new.shape)
-    y_pred = model.predict(x_new)
+    if uploaded_file != processed_file:        
+        x_new = genrep.load_rep_from_raw_file(uploaded_file)
+        # st.write(x_new.shape)
+        y_pred = model.predict(x_new)
+        processed_file = uploaded_file
     
     expanded_tab = tp.make_full_tab(y_pred, len(y_pred))
     display_tab = tp.make_squeezed_tab(expanded_tab)
@@ -75,10 +79,11 @@ if uploaded_file is not None and mode == 'changed notes':
     st.title("""
             The predicted changed notes Tabs: 
     """)
-    genrep = TabDataReprGen()
-    x_new = genrep.load_rep_from_raw_file(uploaded_file)
-    # st.write(x_new.shape)
-    y_pred = model.predict(x_new)
+    if uploaded_file != processed_file:        
+        x_new = genrep.load_rep_from_raw_file(uploaded_file)
+        # st.write(x_new.shape)
+        y_pred = model.predict(x_new)
+        processed_file = uploaded_file
     
     expanded_tab = tp.make_full_tab(y_pred, len(y_pred))
     display_tab = tp.make_dynamic_tab(expanded_tab)