cat-llm 0.0.23__tar.gz → 0.0.25__tar.gz

{cat_llm-0.0.23 → cat_llm-0.0.25}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cat-llm
-Version: 0.0.23
+Version: 0.0.25
 Summary: A tool for categorizing text data and images using LLMs and vision models
 Project-URL: Documentation, https://github.com/chrissoria/cat-llm#readme
 Project-URL: Issues, https://github.com/chrissoria/cat-llm/issues

{cat_llm-0.0.23 → cat_llm-0.0.25}/src/catllm/CERAD_functions.py

@@ -1,6 +1,20 @@
 # a function specifically for CERAD Constructional Praxis test
 # specifically for pictures of drawings of shapes like circles, diamonds, rectangles, and cubes
-# SPDX-FileCopyrightText: 2025-present Christopher Soria <
+
+"""
+Areas for improvement:
+1. Prompt refinement: adjusting the prompt so that it produces a more accurate score.
+2. Image preprocessing: adjusting the images so that they are easier to be analyzed by the models.
+3. Model selection: using a different model that is better suited for image analysis.
+4. Model Ensembling: using multiple models and combining their scores to produce a more accurate score.
+5. Prompt ensembling: using multiple prompts and combining their scores to produce a more accurate score.
+6. Post-processing: adjusting the way scores are calculated after the model has output its assessment.
+7. Efficiency: optimizing the code to run faster, cheaper, and more efficiently.
+8. Drawn-format versatility: making the function more versatile to handle different scenarios, such as shapes drawn on tablets.
+9. Image input flexibility: allowing the function to accept images in various formats, such as URLs or raw image data.
+10. Test variety: expanding or adding functions to handle score more tests relevant for cogntive assesment, such as the MMSE.
+11. Error handling: improving error handling to better manage unexpected inputs or model failures.
+"""
 def cerad_drawn_score(
     shape, 
     image_input,
@@ -91,7 +105,7 @@ def cerad_drawn_score(
     link1 = []
     extracted_jsons = []
 
-    for i, img_path in enumerate(tqdm(image_files, desc="Categorising images"), start=0):
+    for i, img_path in enumerate(tqdm(image_files, desc="Scoring images"), start=0):
     # Check validity first
         if img_path is None or not os.path.exists(img_path):
             link1.append("Skipped NaN input or invalid path")
@@ -118,7 +132,7 @@ def cerad_drawn_score(
                     f"You are an image-tagging assistant trained in the CERAD Constructional Praxis test.\n"
                     f"Task ► Examine the attached image and decide, **for each category below**, "
                     f"whether it is PRESENT (1) or NOT PRESENT (0).\n\n"
-                    f"Image is expected to show within it a drawing of a {shape}.\n\n"
+                    f"{reference_text}"
                     f"Categories:\n{categories_str}\n\n"
                     f"Output format ► Respond with **only** a JSON object whose keys are the "
                     f"quoted category numbers ('1', '2', …) and whose values are 1 or 0. "
@@ -132,7 +146,6 @@ def cerad_drawn_score(
                 "image_url": {"url": encoded_image, "detail": "high"},
             },
         ]
-        print(prompt)
         if model_source == "OpenAI":
             from openai import OpenAI
             client = OpenAI(api_key=api_key)

{cat_llm-0.0.23 → cat_llm-0.0.25}/src/catllm/__about__.py

@@ -1,7 +1,7 @@
 # SPDX-FileCopyrightText: 2025-present Christopher Soria <chrissoria@berkeley.edu>
 #
 # SPDX-License-Identifier: MIT
-__version__ = "0.0.23"
+__version__ = "0.0.25"
 __author__ = "Chris Soria"
 __email__ = "chrissoria@berkeley.edu"
 __title__ = "cat-llm"