gatrag-context-classifier 1.0.0__tar.gz

gatrag_context_classifier-1.0.0/PKG-INFO

@@ -0,0 +1,78 @@
+Metadata-Version: 2.5
+Name: gatrag-context-classifier
+Version: 1.0.0
+Summary: NLI cross-encoder gate that scores whether retrieved RAG context is sufficient to answer a query.
+Project-URL: Homepage, https://github.com/YOUR_USER/gatrag-context-classifier
+Author-email: Gattupalli Saketh <18saketh@gmail.com>
+License-Expression: MIT
+Keywords: classifier,context,cross-encoder,nli,rag
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: Programming Language :: Python :: 3
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Python: >=3.9
+Requires-Dist: numpy>=1.24
+Provides-Extra: model
+Requires-Dist: sentence-transformers>=3.0; extra == 'model'
+Requires-Dist: torch>=2.0; extra == 'model'
+Description-Content-Type: text/markdown
+
+GATRAG
+
+Guarded Abstention / Answering for Retrieval-Augmented Generation (RAG)
+
+GATRAG is a lightweight context-sufficiency gate for RAG pipelines. It sits between the retriever and the LLM/generator, evaluates whether retrieved passages contain enough information to answer a user's query, filters weak passages, and provides an explicit generate / abstain decision.
+
+Instead of always sending retrieved documents to an LLM, GATRAG uses an NLI cross-encoder to estimate whether the retrieved context actually supports the query.
+
+Core idea: retrieve → validate context → generate only when the context is sufficient.
+
+This can help RAG applications reduce unnecessary generation, expose an inspectable confidence signal, and avoid confidently answering when retrieval produced weak or irrelevant context.
+
+Features
+
+NLI-based context sufficiency classification.
+
+Per-document relevance/sufficiency scoring.
+
+Document filtering and ranking.
+
+Optional whole-context ("packed") scoring.
+
+Explicit SUFFICIENT, INSUFFICIENT, and UNCERTAIN results.
+
+Simple boolean generation gate through validate_context_simple().
+
+Batch validation through batch_validate().
+
+Supports plain strings, SimpleDocument, and document-like objects exposing page_content and optional metadata.
+
+Configurable score thresholds, uncertainty margins, model, device, batching, context length, and ranking limits.
+
+Optional conservative fallback when model inference fails.
+
+Compatible with CPU and CUDA environments.
+
+Small public API with typed configuration and inspectable results.
+
+Installation
+
+Basic installation
+
+pip install gatrag-context-classifier
+
+Model-enabled installation
+
+The default classifier loads an NLI model through sentence-transformers. Install the model extra:
+
+pip install "gatrag[model]"
+
+Or with uv:
+
+uv add "gatrag[model]"
+
+The default model is:
+
+cross-encoder/nli-deberta-v3-small
+
+The model is loaded automatically when ContextSufficientClassifier is created without a custom model.

gatrag_context_classifier-1.0.0/README.md

@@ -0,0 +1,59 @@
+GATRAG
+
+Guarded Abstention / Answering for Retrieval-Augmented Generation (RAG)
+
+GATRAG is a lightweight context-sufficiency gate for RAG pipelines. It sits between the retriever and the LLM/generator, evaluates whether retrieved passages contain enough information to answer a user's query, filters weak passages, and provides an explicit generate / abstain decision.
+
+Instead of always sending retrieved documents to an LLM, GATRAG uses an NLI cross-encoder to estimate whether the retrieved context actually supports the query.
+
+Core idea: retrieve → validate context → generate only when the context is sufficient.
+
+This can help RAG applications reduce unnecessary generation, expose an inspectable confidence signal, and avoid confidently answering when retrieval produced weak or irrelevant context.
+
+Features
+
+NLI-based context sufficiency classification.
+
+Per-document relevance/sufficiency scoring.
+
+Document filtering and ranking.
+
+Optional whole-context ("packed") scoring.
+
+Explicit SUFFICIENT, INSUFFICIENT, and UNCERTAIN results.
+
+Simple boolean generation gate through validate_context_simple().
+
+Batch validation through batch_validate().
+
+Supports plain strings, SimpleDocument, and document-like objects exposing page_content and optional metadata.
+
+Configurable score thresholds, uncertainty margins, model, device, batching, context length, and ranking limits.
+
+Optional conservative fallback when model inference fails.
+
+Compatible with CPU and CUDA environments.
+
+Small public API with typed configuration and inspectable results.
+
+Installation
+
+Basic installation
+
+pip install gatrag-context-classifier
+
+Model-enabled installation
+
+The default classifier loads an NLI model through sentence-transformers. Install the model extra:
+
+pip install "gatrag[model]"
+
+Or with uv:
+
+uv add "gatrag[model]"
+
+The default model is:
+
+cross-encoder/nli-deberta-v3-small
+
+The model is loaded automatically when ContextSufficientClassifier is created without a custom model.

gatrag_context_classifier-1.0.0/pyproject.toml

@@ -0,0 +1,49 @@
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[project]
+name = "gatrag-context-classifier"
+version = "1.0.0"
+description = "NLI cross-encoder gate that scores whether retrieved RAG context is sufficient to answer a query."
+readme = "README.md"
+requires-python = ">=3.9"
+license = "MIT"
+authors = [
+  { name = "Gattupalli Saketh", email = "18saketh@gmail.com" },
+]
+keywords = ["rag", "nli", "cross-encoder", "context", "classifier"]
+classifiers = [
+  "Development Status :: 4 - Beta",
+  "Intended Audience :: Developers",
+  "Programming Language :: Python :: 3",
+  "Topic :: Scientific/Engineering :: Artificial Intelligence",
+]
+dependencies = [
+  "numpy>=1.24",
+]
+
+[project.optional-dependencies]
+model = [
+  "sentence-transformers>=3.0",
+  "torch>=2.0",
+]
+
+[project.urls]
+Homepage = "https://github.com/YOUR_USER/gatrag-context-classifier"
+
+[tool.hatch.build.targets.wheel]
+packages = ["src/gatrag"]
+
+[tool.hatch.build.targets.sdist]
+include = [
+  "src/gatrag",
+  "README.md",
+  "pyproject.toml",
+]
+
+[tool.uv]
+package = true
+
+
+