Dataset Viewer
Auto-converted to Parquet Duplicate
source_id
stringclasses
10 values
source_name
stringclasses
10 values
repository_url
stringclasses
10 values
commit_hash
stringclasses
1 value
relative_path
stringlengths
7
105
file_type
stringclasses
7 values
primary_category
stringclasses
11 values
secondary_categories
listlengths
0
3
license
stringclasses
2 values
content
stringlengths
80
468k
char_count
int64
80
468k
word_count
int64
7
41.8k
sha256
stringlengths
64
64
quality_scores
dict
doc_id
stringlengths
14
14
lightrag
LightRAG
https://github.com/HKUDS/LightRAG
HEAD
AGENTS.md
.md
J_agentic_rag_and_graph_rag
[ "A_architecture_and_orchestration", "C_indexing_and_vector_databases", "B_embedding_and_chunking" ]
MIT
# Repository Guidelines ## Project Overview LightRAG is a Retrieval-Augmented Generation (RAG) framework that uses graph-based knowledge representation for enhanced information retrieval. The system extracts entities and relationships from documents, builds a knowledge graph, and uses multiple retrieval modes (`local...
34,756
4,425
558604e3fe05d70fe0eafbfa2695d20c2cad3753c5ae6227d07c1e6be75004eb
{ "is_clean": true, "avg_line_length": 86.2, "code_ratio": 0 }
cpt_doc_000001
lightrag
LightRAG
https://github.com/HKUDS/LightRAG
HEAD
README-id.md
.md
J_agentic_rag_and_graph_rag
[ "K_repository_level_code_understanding", "D_dense_sparse_hybrid_retrieval", "F_rag_evaluation_and_metrics" ]
MIT
<div align="center"> <div style="margin: 20px 0;"> <img src="./assets/logo.png" width="120" height="120" alt="LightRAG Logo" style="border-radius: 20px; box-shadow: 0 8px 32px rgba(0, 217, 255, 0.3);"> </div> # 🚀 LightRAG: Retrieval-Augmented Generation yang Sederhana dan Cepat <div align="center"> <a href="https...
58,804
6,241
a3aff8ea5180966baa63c4832e5f2efc07906ac4e3ebf111c239ea9e73c0e3d0
{ "is_clean": true, "avg_line_length": 83.5, "code_ratio": 0 }
cpt_doc_000002
lightrag
LightRAG
https://github.com/HKUDS/LightRAG
HEAD
README-ja.md
.md
J_agentic_rag_and_graph_rag
[ "K_repository_level_code_understanding" ]
MIT
<div align="center"> <div style="margin: 20px 0;"> <img src="./assets/logo.png" width="120" height="120" alt="LightRAG Logo" style="border-radius: 20px; box-shadow: 0 8px 32px rgba(0, 217, 255, 0.3);"> </div> # 🚀 LightRAG: シンプルかつ高速な検索拡張生成(RAG) <div align="center"> <a href="https://www.star-history.com/hkuds/light...
40,396
2,373
db49f5ef9a3f8c6bb14d74aa830c8ff9cf478f45ceb0aa00afca051886c7706c
{ "is_clean": true, "avg_line_length": 57.4, "code_ratio": 0 }
cpt_doc_000003
lightrag
LightRAG
https://github.com/HKUDS/LightRAG
HEAD
README-zh.md
.md
J_agentic_rag_and_graph_rag
[ "K_repository_level_code_understanding" ]
MIT
"<div align=\"center\">\n\n<div style=\"margin: 20px 0;\">\n <img src=\"./assets/logo.png\" width=\(...TRUNCATED)
34,496
2,345
bc7c62504c15a0ccf4033f846385a72227d56f35d867fe6e55ed87f3b41534b2
{ "is_clean": true, "avg_line_length": 49, "code_ratio": 0 }
cpt_doc_000004
lightrag
LightRAG
https://github.com/HKUDS/LightRAG
HEAD
README.md
.md
K_repository_level_code_understanding
[ "J_agentic_rag_and_graph_rag", "D_dense_sparse_hybrid_retrieval", "F_rag_evaluation_and_metrics" ]
MIT
"<div align=\"center\">\n\n<div style=\"margin: 20px 0;\">\n <img src=\"./assets/logo.png\" width=\(...TRUNCATED)
56,633
6,460
543105a425787a7cbb0e4d74526ee9f03edd0af24639a0efde3eddca5c9560bd
{ "is_clean": true, "avg_line_length": 80.4, "code_ratio": 0 }
cpt_doc_000005
lightrag
LightRAG
https://github.com/HKUDS/LightRAG
HEAD
SECURITY.md
.md
I_security_and_guardrails
[ "K_repository_level_code_understanding", "J_agentic_rag_and_graph_rag" ]
MIT
"# Reporting Security Issues\n\nThe LightRAG team and community take security bugs seriously. We app(...TRUNCATED)
971
137
1c19ad0b80b5811b3ed406e49a3a7fb13e06fe19a98a087828b9cea44709d7e3
{ "is_clean": true, "avg_line_length": 53.9, "code_ratio": 0 }
cpt_doc_000006
lightrag
LightRAG
https://github.com/HKUDS/LightRAG
HEAD
docker-compose-full.yml
.yml
C_indexing_and_vector_databases
["B_embedding_and_chunking","E_reranking_and_query_transformation","G_grounded_generation_and_faithf(...TRUNCATED)
MIT
"# Full Docker Compose Deployment Sample Generated by Setup Wizard: `make base` and `make storage`\n(...TRUNCATED)
8,762
710
c950ffb16f3190d4df74ab273013fd896c0939dca891255b9652742d709caf9b
{ "is_clean": true, "avg_line_length": 33.3, "code_ratio": 0 }
cpt_doc_000007
lightrag
LightRAG
https://github.com/HKUDS/LightRAG
HEAD
docker-compose.podman.yml
.yml
B_embedding_and_chunking
[ "E_reranking_and_query_transformation", "I_security_and_guardrails", "J_agentic_rag_and_graph_rag" ]
MIT
"# Podman-compatible compose file for LightRAG\n#\n# Usage:\n# podman-compose -f docker-compose.po(...TRUNCATED)
2,565
299
0aa8c25fff43119aec24c5c6d27f352922cd1368e8d3da55f23d089fcb915a39
{ "is_clean": true, "avg_line_length": 45.8, "code_ratio": 0 }
cpt_doc_000008
lightrag
LightRAG
https://github.com/HKUDS/LightRAG
HEAD
docker-compose.yml
.yml
I_security_and_guardrails
[ "J_agentic_rag_and_graph_rag" ]
MIT
"services:\n lightrag:\n image: ghcr.io/hkuds/lightrag:latest\n build:\n context: .\n (...TRUNCATED)
1,929
212
593c7f1eb59d641edb11ffc2684a6792533f309d789929c1efbb84b6e19a214c
{ "is_clean": true, "avg_line_length": 43.8, "code_ratio": 0 }
cpt_doc_000009
lightrag
LightRAG
https://github.com/HKUDS/LightRAG
HEAD
pyproject.toml
.toml
J_agentic_rag_and_graph_rag
["K_repository_level_code_understanding","A_architecture_and_orchestration","B_embedding_and_chunkin(...TRUNCATED)
MIT
"[build-system]\nrequires = [\"setuptools>=64\", \"wheel\"]\nbuild-backend = \"setuptools.build_meta(...TRUNCATED)
10,634
1,114
167cd94b8bb928e994e7777e1b3891cb4a65519662e9fc35eb1843b2b91d7c2e
{ "is_clean": true, "avg_line_length": 39.4, "code_ratio": 0 }
cpt_doc_000010
End of preview. Expand in Data Studio

RAG Domain-Adaptive Pretraining Corpus (8.19M Tokens)

This dataset contains the curated, deduplicated pretraining corpus of production Retrieval-Augmented Generation (RAG) frameworks used to domain-adapt Qwen2.5-Coder-3B into RAG-CPT-3B.

Source Repositories

Includes curated, provenance-verified source code, architecture documentation, and test suites from:

  • LlamaIndex (llama-index-core)
  • LangChain & LangGraph
  • Haystack (deepset-ai/haystack)
  • Qdrant Client SDK
  • FastEmbed
  • ChromaDB
  • Microsoft GraphRAG
  • Ragas (Evaluation)
  • AutoRAG

Data Splits

  • Train: data/train.jsonl (44 MB, ~7.4M tokens)
  • Validation: data/validation.jsonl (4.1 MB, ~0.8M tokens)
Downloads last month
62

Models trained or fine-tuned on gaurav0506/rag-pretraining-corpus-8m