Major update. Support for 15 LLMs, World Flora Online taxonomy validation, geolocation, 2 OCR methods, significant UI changes, stability improvements, consistent JSON parsing

app.py

@@ -94,15 +94,16 @@ def content_input_images_hf():
 
 
 
-def create_download_button(zip_filepath):
-    with open(zip_filepath, 'rb') as f:
-        bytes_io = BytesIO(f.read())
-    st.download_button(
-        label=f"Download Results for{st.session_state['processing_add_on']}",type='primary',
-        data=bytes_io,
-        file_name=os.path.basename(zip_filepath),
-        mime='application/zip'
-    )
+def create_download_button(zip_filepath, col):
+    with col:
+        with open(zip_filepath, 'rb') as f:
+            bytes_io = BytesIO(f.read())
+        st.download_button(
+            label=f"Download Results for{st.session_state['processing_add_on']}",type='primary',
+            data=bytes_io,
+            file_name=os.path.basename(zip_filepath),
+            mime='application/zip'
+        )
 
 
 def delete_directory(dir_path):
@@ -373,7 +374,7 @@ class JSONReport:
         # Display the map if map_data is not empty
         if not map_data.empty:
             with self.json_map:
-                st.map(map_data, zoom=4, size='size', color='color')
+                st.map(map_data, zoom=4, size='size', color='color', use_container_width=True)
 
     def set_text(self, text_main=None, text_middle=None, text_right=None):
         if text_main:
@@ -1287,7 +1288,7 @@ def build_LLM_prompt_config():
                 if new_filename:
                     if check_unique_mapping_assignments():
                         if check_prompt_yaml_filename(new_filename):
-                            save_prompt_yaml(new_filename)
+                            save_prompt_yaml(new_filename, col_left_save)
                         else:
                             st.error("File name can only contain letters, numbers, underscores, and dashes. Cannot contain spaces.")
                     else:
@@ -1587,7 +1588,7 @@ def content_header():
 
                 path_custom_prompts = os.path.join(st.session_state.dir_home,'custom_prompts',st.session_state.config['leafmachine']['project']['prompt_version'])
                 # Call the machine function.
-                st.session_state['formatted_json'], st.session_state['formatted_json_WFO'], st.session_state['formatted_json_GEO'], total_cost, n_failed_OCR, n_failed_LLM_calls = voucher_vision(None,
+                st.session_state['formatted_json'], st.session_state['formatted_json_WFO'], st.session_state['formatted_json_GEO'], total_cost, n_failed_OCR, n_failed_LLM_calls, st.session_state['zip_filepath'] = voucher_vision(None,
                                                                                                                 st.session_state.dir_home, 
                                                                                                                 path_custom_prompts, 
                                                                                                                 None, 
@@ -1609,7 +1610,8 @@ def content_header():
                 else:
                     st.info(f":money_with_wings: This run cost :heavy_dollar_sign:{total_cost:.4f}")
                 st.balloons()
-
+            if st.session_state['zip_filepath']:
+                create_download_button(st.session_state['zip_filepath'], col_run_1)
         else:
             st.button("Start Processing", type='primary', disabled=True)
             st.error(":heavy_exclamation_mark: Required API keys not set. Please visit the 'API Keys' tab and set the Google Vision OCR API key and at least one LLM key.")
@@ -1617,7 +1619,8 @@ def content_header():
         if st.session_state['formatted_json']:
             if st.session_state['hold_output']:
                 st.session_state['json_report'].set_JSON(st.session_state['formatted_json'], st.session_state['formatted_json_WFO'], st.session_state['formatted_json_GEO'])
-   
+                if st.session_state['zip_filepath']:
+                    create_download_button(st.session_state['zip_filepath'], col_run_1)
     
  
     
@@ -2361,7 +2364,8 @@ if 'settings_filename' not in st.session_state:
     st.session_state['settings_filename'] = None
 if 'loaded_settings_filename' not in st.session_state:
     st.session_state['loaded_settings_filename'] = None
-
+if 'zip_filepath' not in st.session_state:
+    st.session_state['zip_filepath'] = None
 
 # Initialize session_state variables if they don't exist
 if 'prompt_info' not in st.session_state:

vouchervision/API_validation.py

@@ -1,4 +1,4 @@
-import os, io, openai, vertexai, json
+import os, io, openai, vertexai, json, tempfile
 import google.generativeai as genai
 from mistralai.client import MistralClient
 from mistralai.models.chat_completion import ChatMessage
@@ -171,12 +171,24 @@ class APIvalidation:
         except Exception as e:  # Replace with a more specific exception if possible
             return False
         
+    def get_google_credentials(self):
+        # Convert JSON key from string to a dictionary
+        service_account_json_str = os.getenv('google_service_account_json')
+
+        with tempfile.NamedTemporaryFile(mode="w+", delete=False,suffix=".json") as temp:
+            temp.write(service_account_json_str)
+            temp_filename = temp.name
+        
+        return temp_filename
+
+
     def check_google_vertex_genai_api_key(self):
         results = {"palm2": False, "gemini": False}
         if self.cfg_private:
             try:
                 # Assuming genai and vertexai are clients for Google services
-                os.environ["GOOGLE_API_KEY"] = self.cfg_private['google_palm']['google_palm_api']
+                # os.environ["GOOGLE_API_KEY"] = self.cfg_private['google_palm']['google_palm_api']
+                os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = self.get_google_credentials()
                 # genai.configure(api_key=self.cfg_private['google_palm']['google_palm_api'])
                 vertexai.init(project= self.cfg_private['google_palm']['project_id'], location=self.cfg_private['google_palm']['location'])
 
@@ -209,7 +221,7 @@ class APIvalidation:
             try:
                 # Assuming genai and vertexai are clients for Google services
                 # os.environ["GOOGLE_API_KEY"] = os.getenv('PALM_API_KEY')
-                genai.configure(api_key=os.getenv('PALM_API_KEY'))
+                genai.configure(api_key=os.getenv('GOOGLE_API_KEY'))
                 vertexai.init(project= os.getenv('GOOGLE_PROJECT_ID'), location=os.getenv('GOOGLE_LOCATION'))
 
                 try:

vouchervision/LLM_local_MistralAI.py

@@ -12,6 +12,7 @@ from vouchervision.utils_taxonomy_WFO import validate_taxonomy_WFO
 from vouchervision.utils_geolocate_HERE import validate_coordinates_here
 
 '''
+Local Pipielines:
 https://python.langchain.com/docs/integrations/llms/huggingface_pipelines
 '''
 

vouchervision/VoucherVision_Config_Builder.py

@@ -41,7 +41,7 @@ def build_VV_config(loaded_cfg=None):
 
         LLM_version_user = 'Azure GPT 3.5 Instruct' #'Azure GPT 4 Turbo 1106-preview'
         prompt_version = 'version_5.yaml' # from ["Version 1", "Version 1 No Domain Knowledge", "Version 2"]
-        use_LeafMachine2_collage_images = True # Use LeafMachine2 collage images
+        use_LeafMachine2_collage_images = False # Use LeafMachine2 collage images
         do_create_OCR_helper_image = True
 
         batch_size = 500

vouchervision/utils_VoucherVision.py

@@ -1,5 +1,5 @@
 import openai
-import os, json, glob, shutil, yaml, torch, logging
+import os, json, glob, shutil, yaml, torch, logging, tempfile
 import openpyxl
 from openpyxl import Workbook, load_workbook
 import vertexai
@@ -408,12 +408,17 @@ class VoucherVision():
         else:
             return False
         
-    
+    def get_google_credentials(self):
+        # Convert JSON key from string to a dictionary
+        service_account_json_str = os.getenv('google_service_account_json')
 
+        with tempfile.NamedTemporaryFile(mode="w+", delete=False,suffix=".json") as temp:
+            temp.write(service_account_json_str)
+            temp_filename = temp.name
         
+        return temp_filename   
     
     
-
     def set_API_keys(self):
         if self.is_hf:
             openai_api_key = os.getenv('OPENAI_API_KEY')
@@ -425,6 +430,7 @@ class VoucherVision():
             open_cage_api_key = os.getenv('open_cage_geocode')
             google_project_id = os.getenv('GOOGLE_PROJECT_ID')
             google_project_location = os.getenv('GOOGLE_LOCATION')
+            os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = self.get_google_credentials()
 
 
 

vouchervision/vouchervision_main.py

@@ -1,7 +1,7 @@
 '''
 VoucherVision - based on LeafMachine2 Processes
 '''
-import os, inspect, sys, logging, subprocess
+import os, inspect, sys, logging, subprocess, shutil
 from time import perf_counter
 currentdir = os.path.dirname(os.path.dirname(inspect.getfile(inspect.currentframe())))
 parentdir = os.path.dirname(currentdir)
@@ -14,7 +14,7 @@ from data_project import Project_Info
 from LM2_logger import start_logging
 from fetch_data import fetch_data
 from utils_VoucherVision import VoucherVision, space_saver
-
+from vouchervision.utils_hf import upload_to_drive
 
 def voucher_vision(cfg_file_path, dir_home, path_custom_prompts, cfg_test, progress_report, json_report, path_api_cost=None, test_ind = None, is_hf = True, is_real_run=False):
     t_overall = perf_counter()
@@ -78,7 +78,21 @@ def voucher_vision(cfg_file_path, dir_home, path_custom_prompts, cfg_test, progr
 
     Voucher_Vision.close_logger_handlers()
 
-    return last_JSON_response, final_WFO_record, final_GEO_record, total_cost, Voucher_Vision.n_failed_OCR, Voucher_Vision.n_failed_LLM_calls 
+    zip_filepath = None
+    if is_hf:
+        # Create Higging Face zip file
+        dir_to_zip = os.path.join(Dirs.dir_home, Dirs.run_name)  
+        zip_filename = Dirs.run_name  
+
+        # Creating a zip file
+        zip_filepath = make_zipfile(dir_to_zip, zip_filename)
+        upload_to_drive(zip_filepath, zip_filename)
+
+    return last_JSON_response, final_WFO_record, final_GEO_record, total_cost, Voucher_Vision.n_failed_OCR, Voucher_Vision.n_failed_LLM_calls, zip_filepath 
+
+def make_zipfile(source_dir, output_filename):
+    shutil.make_archive(output_filename, 'zip', source_dir)
+    return output_filename + '.zip'
 
 def voucher_vision_OCR_test(cfg_file_path, dir_home, cfg_test, path_to_crop):
     # get_n_overall = progress_report.get_n_overall()