Errors on chunking and cordinates

Token shape: 512: This means the model should be able to handle this without exceeding the token limit, which is good.

Type of coordinates[0]: <class 'tuple'>: This shows that the coordinates are tuples, which means we have to use them as indices in a different way.

Value of coordinates[0]: (0, 8): This likely indicates the row and column indices, respectively.

An error occurred: 'str' object has no attribute 'values': This means that chunk.iloc[coordinates[0]].values is a string and doesn't have the attribute "values." We need to fix this line.

An error occurred: iloc cannot enlarge its target object: This could indicate an indexing error and needs further investigation.

Let's start by addressing these issues.

The coordinates being a tuple suggest that they should be used as (row, column) indices directly.
The 'str' object has no attribute 'values' error suggests that we might need to remove .values from chunk.iloc[coordinates[0]].values.

app.py

@@ -30,13 +30,16 @@ def ask_llm_chunk(chunk, questions):
         for coordinates in predicted_answer_coordinates:
             st.write(f"Type of coordinates[0]: {type(coordinates[0])}")  # Debugging line
             st.write(f"Value of coordinates[0]: {coordinates[0]}")  # Debugging line
+            
             if len(coordinates) == 1:
-                answers.append(chunk.iloc[coordinates[0]].values)
+                row, col = coordinates[0]
+                answers.append(chunk.iloc[row, col])
             else:
                 cell_values = []
                 for coordinate in coordinates:
-                    cell_values.append(chunk.iloc[coordinate].values)
-                answers.append(", ".join(cell_values))
+                    row, col = coordinate
+                    cell_values.append(chunk.iloc[row, col])
+                answers.append(", ".join(map(str, cell_values)))
         return answers
     except Exception as e:
         st.write(f"An error occurred: {e}")
@@ -44,6 +47,7 @@ def ask_llm_chunk(chunk, questions):
 
 
 
+
 MAX_ROWS_PER_CHUNK = 200
 
 def summarize_map_reduce(data, questions):