Upload folder using huggingface_hub

graph.py

@@ -92,8 +92,8 @@ class SheamiState(TypedDict):
     process_desc: str
     units_processed: int
     units_total: int
-    overall_units_processed : int
-    overall_units_total : int
+    overall_units_processed: int
+    overall_units_total: int
 
 
 import re
@@ -163,7 +163,7 @@ def fn_init_node(state: SheamiState):
     state["units_total"] = 0
     state["process_desc"] = ""
     state["overall_units_processed"] = 0
-    state["overall_units_total"] = 6 # 6 steps totally
+    state["overall_units_total"] = 6  # 6 steps totally
 
     return state
 
@@ -174,8 +174,19 @@ def reset_process_desc(state: SheamiState, process_desc: str):
     state["units_total"] = 0
     return state
 
+
 async def fn_increment_index_node(state: SheamiState):
     state["current_index"] += 1
+    total_reports = len(state["uploaded_reports"])
+    try:
+        report_file_name = state["uploaded_reports"][
+            state["current_index"]
+        ].report_file_name
+        state["process_desc"] = (
+            f"Standardizing {state["current_index"]+1} of {total_reports} reports - {report_file_name} ..."
+        )
+    except:
+        pass
     return state
 
 
@@ -222,11 +233,8 @@ def fn_is_report_available_to_process(state: SheamiState) -> str:
         return "done"
 
 
-async def fn_testname_standardizer_node(state: SheamiState):
-    logger.info("%s| Standardizing Test Names: started", state["thread_id"])
-    state["messages"].append("Standardizing Test Names: started")
-    # collect unique names
-    unique_names = list(
+def get_unique_test_names(state: SheamiState):
+    return list(
         {
             result.test_name
             for report in state["standardized_reports"]
@@ -234,6 +242,13 @@ async def fn_testname_standardizer_node(state: SheamiState):
         }
     )
 
+
+async def fn_testname_standardizer_node(state: SheamiState):
+    logger.info("%s| Standardizing Test Names: started", state["thread_id"])
+    state["messages"].append("Standardizing Test Names: started")
+    # collect unique names
+    unique_names = get_unique_test_names(state)
+
     # run through LLM
     response = testname_standardizer_chain.invoke({"test_names": unique_names})
     raw_text = response.content
@@ -251,6 +266,7 @@ async def fn_testname_standardizer_node(state: SheamiState):
         for result in report.lab_results:
             result.test_name = normalization_map.get(result.test_name, result.test_name)
     logger.info("%s| Standardizing Test Names: finished", state["thread_id"])
+    state["messages"].append(f"Processed {len(unique_names)} tests")
     state["messages"].append("Standardizing Test Names: finished")
     return state