From bc5a7c0111c9e8a8a0ef590ff11593fe81287dd7 Mon Sep 17 00:00:00 2001
From: John Walz <john@validmind.ai>
Date: Tue, 11 Feb 2025 12:41:30 -0500
Subject: [PATCH 1/4] fix: context length issues

---
 validmind/ai/test_descriptions.py | 29 ++++++++++++++++++++++++++++-
 1 file changed, 28 insertions(+), 1 deletion(-)

diff --git a/validmind/ai/test_descriptions.py b/validmind/ai/test_descriptions.py
index b52bc3820..fefa978fd 100644
--- a/validmind/ai/test_descriptions.py
+++ b/validmind/ai/test_descriptions.py
@@ -7,6 +7,8 @@
 from concurrent.futures import ThreadPoolExecutor
 from typing import List, Optional, Union
 
+import tiktoken
+
 from ..client_config import client_config
 from ..logging import get_logger
 from ..utils import NumpyEncoder, md_to_html, test_id_to_name
@@ -35,6 +37,25 @@ def _get_llm_global_context():
     return context if context_enabled and context else None
 
 
+def _truncate_summary(summary: str, max_tokens: int = 100_000):
+    if len(summary) < max_tokens:
+        # since string itself is less than max_tokens, definitely small enough
+        return summary
+
+    # TODO: better context length handling
+    encoding = tiktoken.encoding_for_model("gpt-4o")
+    summary_tokens = encoding.encode(summary)
+
+    if len(summary_tokens) > max_tokens:
+        summary = (
+            encoding.decode(summary_tokens[:max_tokens])
+            + "...[truncated]"
+            + encoding.decode(summary_tokens[-100:])
+        )
+
+    return summary
+
+
 def generate_description(
     test_id: str,
     test_description: str,
@@ -107,7 +128,13 @@ def wrapped():
                 title=title,
             )
         except Exception as e:
-            logger.error(f"Failed to generate description: {e}")
+            if "maximum context length" in str(e):
+                logger.warning(
+                    f"Test result {test_id} is too large to generate a description"
+                )
+            else:
+                logger.warning(f"Failed to generate description for {test_id}: {e}")
+            logger.warning(f"Using default description for {test_id}")
 
             return test_description
 

From e140d7f60933399bf0f60109b35a0f88fc83acc4 Mon Sep 17 00:00:00 2001
From: John Walz <john@validmind.ai>
Date: Tue, 11 Feb 2025 12:42:10 -0500
Subject: [PATCH 2/4] 2.8.10

---
 pyproject.toml           | 2 +-
 validmind/__version__.py | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/pyproject.toml b/pyproject.toml
index c42db1a9e..3a5b25455 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -10,7 +10,7 @@ description = "ValidMind Library"
 license = "Commercial License"
 name = "validmind"
 readme = "README.pypi.md"
-version = "2.8.9"
+version = "2.8.10"
 
 [tool.poetry.dependencies]
 python = ">=3.8.1,<3.12"
diff --git a/validmind/__version__.py b/validmind/__version__.py
index e71de367e..71322b8ca 100644
--- a/validmind/__version__.py
+++ b/validmind/__version__.py
@@ -1 +1 @@
-__version__ = "2.8.9"
+__version__ = "2.8.10"

From c4d65dd5c39e045cfcdd85a160a1b10872eab925 Mon Sep 17 00:00:00 2001
From: John Walz <john@validmind.ai>
Date: Tue, 11 Feb 2025 13:00:22 -0500
Subject: [PATCH 3/4] chore: adding warning when truncating test result for llm

---
 validmind/ai/test_descriptions.py | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/validmind/ai/test_descriptions.py b/validmind/ai/test_descriptions.py
index fefa978fd..10ff33fe7 100644
--- a/validmind/ai/test_descriptions.py
+++ b/validmind/ai/test_descriptions.py
@@ -37,7 +37,7 @@ def _get_llm_global_context():
     return context if context_enabled and context else None
 
 
-def _truncate_summary(summary: str, max_tokens: int = 100_000):
+def _truncate_summary(summary: str, test_id: str, max_tokens: int = 100_000):
     if len(summary) < max_tokens:
         # since string itself is less than max_tokens, definitely small enough
         return summary
@@ -47,6 +47,10 @@ def _truncate_summary(summary: str, max_tokens: int = 100_000):
     summary_tokens = encoding.encode(summary)
 
     if len(summary_tokens) > max_tokens:
+        logger.warning(
+            "Truncating test result due to context length restrictions..."
+            f"Generated description for {test_id} may be innacurate"
+        )
         summary = (
             encoding.decode(summary_tokens[:max_tokens])
             + "...[truncated]"
@@ -100,7 +104,7 @@ def generate_description(
             "test_name": test_name,
             "test_description": test_description,
             "title": title,
-            "summary": summary,
+            "summary": _truncate_summary(summary, test_id),
             "figures": [
                 figure._get_b64_url() for figure in ([] if tables else figures)
             ],

From 634d6b3036ef9badd930a7118a49a3dd6d5892bb Mon Sep 17 00:00:00 2001
From: John Walz <john@validmind.ai>
Date: Tue, 11 Feb 2025 13:04:04 -0500
Subject: [PATCH 4/4] chore: updating warning message

---
 validmind/ai/test_descriptions.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/validmind/ai/test_descriptions.py b/validmind/ai/test_descriptions.py
index 10ff33fe7..48747627a 100644
--- a/validmind/ai/test_descriptions.py
+++ b/validmind/ai/test_descriptions.py
@@ -48,8 +48,8 @@ def _truncate_summary(summary: str, test_id: str, max_tokens: int = 100_000):
 
     if len(summary_tokens) > max_tokens:
         logger.warning(
-            "Truncating test result due to context length restrictions..."
-            f"Generated description for {test_id} may be innacurate"
+            f"Truncating {test_id} due to context length restrictions..."
+            " Generated description may be innacurate"
         )
         summary = (
             encoding.decode(summary_tokens[:max_tokens])