Prep for v1.0.3 (#434)

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "ontogpt"
-version = "1.0.2"
+version = "1.0.3"
 description = "OntoGPT"
 authors = ["Chris Mungall <cjmungall@lbl.gov>", "J. Harry Caufield <jhc@lbl.gov>"]
 license = "BSD-3"

src/ontogpt/engines/generic_engine.py

@@ -10,7 +10,6 @@
 from pydantic import BaseModel
 
 from ontogpt.engines.knowledge_engine import KnowledgeEngine
-from ontogpt.prompts.qa import GENERIC_QA_PROMPT
 
 logger = logging.getLogger(__name__)
 
@@ -48,11 +47,10 @@ def run(
  question_collection: QuestionCollection,
  template_path: Union[str, Path] = "",
  ) -> Iterator[Question]:
- if template_path is None:
- template_path = GENERIC_QA_PROMPT
  if isinstance(template_path, Path):
  template_path = str(template_path)
  if isinstance(template_path, str):
+ # Note: The default is a string, so this will always be true
  # create a Jinja2 template object
  with open(template_path) as file:
  template_txt = file.read()

src/ontogpt/engines/halo_engine.py

@@ -257,7 +257,7 @@ def get_element_score(self, element: OntologyElement, tokens: Set[int]) -> float
  score = self.element_scores[element_name]
  else:
  score = 0
- for _, v in element.dict().items():
+ for _, v in element.model_dump().items():
  if v:
  score += 1
  self.element_scores[element_name] = score

src/ontogpt/engines/mapping_engine.py

@@ -1,4 +1,5 @@
 """Synonym engine."""
+
 import logging
 from copy import deepcopy
 from dataclasses import dataclass

src/ontogpt/engines/reasoner_engine.py

@@ -1,4 +1,5 @@
 """Reasoner engine."""
+
 import logging
 import re
 from dataclasses import dataclass
@@ -72,21 +73,21 @@ class ReasonerResultSet(BaseModel):
 
 @dataclass
 class ReasonerEngine(KnowledgeEngine):
- """Engine for performing reasoning using GPT.
+ """Engine for performing reasoning using an LLM.
 
  This engine takes as input an Ontology, and a query Task,
- and then translates this to a GPT prompt that asks GPT to
+ and then translates this to an LLM prompt that asks the LLM to
  perform the task over the ontology after reasoning over it.
 
  The Task is typically a query such as finding superclasses of
  a given class.
 
  This is intended primarily for investigation purposes. For practical
  scenarios, it is recommended to use a dedicated OWL reasoner. The goal
- of this engine is to evaluate the extent to which GPT can perform
+ of this engine is to evaluate the extent to which the LLM can perform
  reasoning-like tasks, including deduction and abduction (explanation).
 
- Due to token-length constraints on GPT models, it is usually necessary
+ Due to token-length constraints on some models, it is usually necessary
  to extract a submodule prior to reasoning. This can be done using the
  OntologyExtractor:
 

tests/unit/test_clients/test_pubmed_client_unit.py

@@ -1,4 +1,5 @@
 """PubMed client tests."""
+
 import unittest
 
 from ontogpt.clients import PubmedClient

tests/unit/test_converters/test_ontology_converter.py

@@ -1,4 +1,5 @@
 """Core tests."""
+
 import logging
 import unittest
 

tests/unit/test_ontex/test_extract.py

@@ -1,4 +1,5 @@
-"""Core tests."""
+"""Core tests for extraction."""
+
 import csv
 import logging
 import unittest