Added docker files for dev and prod.

Added functionality to not proceed with llm summarization, if no menu items are detected in input image.

Signed-off-by: Amith Adiraju <[email protected]>

.gitignore

@@ -0,0 +1,3 @@
+misc.txt
+test_cas.py
+test_train_llm.py

Dockerfile.dev

@@ -0,0 +1,22 @@
+# 
+FROM python:3.9-slim
+
+# This is directory inside container
+WORKDIR /cont_transmiim
+
+# Copy requirements file
+COPY ./requirements.txt /cont_transmiim/requirements.txt
+
+# 
+RUN pip3 install --no-cache-dir --upgrade -r /cont_transmiim/requirements.txt
+
+# Copy remaining parts of code to container
+COPY ./src /cont_transmiim/src
+
+# Expose this specific port to web traffic
+EXPOSE 8501
+
+# Telling docker on how to test if a container is still working
+HEALTHCHECK CMD curl --fail http://localhost:8501/_stcore/health
+
+ENTRYPOINT [ "streamlit", "run", "src/main.py", "--server.port=8501", "--server.address=0.0.0.0" ]

Dockerfile.prod

@@ -0,0 +1,29 @@
+# 
+FROM python:3.9-slim
+
+# This is directory inside container
+WORKDIR /cont_transmiim
+
+# Install Git on Container to clone our repo
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    curl \
+    software-properties-common \
+    git \
+    && rm -rf /var/lib/apt/lists/*
+
+# Cloning code into container from a public repository
+# cont_transmiim -> src, Docker, requirements.txt
+RUN git clone -b feat_streamlit_gs https://github.com/amithadiraju1694/Transmiim.git .
+
+# 
+RUN pip3 install -r requirements.txt
+
+# Expose this specific port to web traffic
+EXPOSE 8501
+
+# Telling docker on how to test if a container is still working
+HEALTHCHECK CMD curl --fail http://localhost:8501/_stcore/health
+
+
+ENTRYPOINT [ "streamlit", "run", "src/main.py", "--server.port=8501", "--server.address=0.0.0.0" ]

app.py

@@ -42,30 +42,37 @@ if uploaded_file is not None:
         # Call the extract_filter_img function
         filtered_text = extract_filter_img(image, text_extractor)
         en_filter = time.perf_counter()
+
+        num_items_detected = len(filtered_text)
+        if num_items_detected == 0:
+            st.write("We couldn't detect any menu items ( indian for now ) from your image, please try a different image.")
         
-        msg2 = st.empty()
-        msg2.write("All pre-processing done, transcribing your menu items now ....")
-        st_trans_llm = time.perf_counter()
-        translated_text_dict = transcribe_menu_model(menu_texts=filtered_text,
-                                                     text_tokenizer=item_tokenizer,
-                                                     text_summarizer=item_summarizer
-                                                    )
-        
-        msg3 = st.empty()                                            
-        msg3.write("Done transcribing ... ")
-        en_trans_llm = time.perf_counter()
+        elif num_items_detected > 0:
+            st.write(f"Detected {num_items_detected} menu items ( indian ) from your input image ... ")
+            
+            msg2 = st.empty()
+            msg2.write("All pre-processing done, transcribing your menu items now ....")
+            st_trans_llm = time.perf_counter()
+            translated_text_dict = transcribe_menu_model(menu_texts=filtered_text,
+                                                        text_tokenizer=item_tokenizer,
+                                                        text_summarizer=item_summarizer
+                                                        )
+            
+            msg3 = st.empty()                                            
+            msg3.write("Done transcribing ... ")
+            en_trans_llm = time.perf_counter()
 
-        msg1.empty(); msg2.empty(); msg3.empty()
-        st.success("Image processed successfully! " )
+            msg1.empty(); msg2.empty(); msg3.empty()
+            st.success("Image processed successfully! " )
 
-        if DEBUG_MODE:
-            filter_time_sec = en_filter - st_filter
-            llm_time_sec = en_trans_llm - st_trans_llm
-            total_time_sec = filter_time_sec + llm_time_sec
+            if DEBUG_MODE:
+                filter_time_sec = en_filter - st_filter
+                llm_time_sec = en_trans_llm - st_trans_llm
+                total_time_sec = filter_time_sec + llm_time_sec
+                
+                st.write("Time took to extract and filter text {}".format(filter_time_sec))
+                st.write("Time took to summarize by LLM {}".format(llm_time_sec))
+                st.write('Overall time taken in seconds: {}'.format(total_time_sec))
             
-            st.write("Time took to extract and filter text {}".format(filter_time_sec))
-            st.write("Time took to summarize by LLM {}".format(llm_time_sec))
-            st.write('Overall time taken in seconds: {}'.format(total_time_sec))
-        
-        st.table(translated_text_dict)
+            st.table(translated_text_dict)