search

microsoft / graphrag

A modular graph-based Retrieval-Augmented Generation (RAG) system
https://microsoft.github.io/graphrag/
MIT License
16.82k stars 1.58k forks source link

[Bug]: Cant disable gleanings #613

Closed therealTT closed 1 month ago

therealTT commented 1 month ago

Describe the bug

I tried setting the max_gleanings to 0 in settings.yaml to prevent the gleaning step from happening but in the indexing-engine.log file max_gleanings is still 1 and I can see the gleaning response in cache folder.

Steps to reproduce

  1. Run python -m graphrag.index --init --root ./ragtest
  2. change max_gleanings value under entity extraction to 0 (I also tried with None, null. Both didn't work)
  3. Run python -m graphrag.index --root ./ragtest

Expected Behavior

In the indexing-engine.log file under entity_extraction max_gleaning should be 0.

GraphRAG Config Used

encoding_model: cl100k_base skip_workflows: [] llm: api_key: ${GRAPHRAG_API_KEY} type: openai_chat # or azure_openai_chat model: llama3-70b-8192 model_supports_json: true # recommended if this is available for your model.

max_tokens: 4000

request_timeout: 180.0

api_base: https://api.groq.com/openai/v1

api_version: 2024-02-15-preview

organization:

deployment_name:

tokens_per_minute: 5000 # set a leaky bucket throttle requests_per_minute: 30 # set a leaky bucket throttle max_retries: 10 max_retry_wait: 60.0

sleep_on_rate_limit_recommendation: true # whether to sleep when azure suggests wait-times

concurrent_requests: 25 # the number of parallel inflight requests that may be made

temperature: 0 # temperature for sampling

top_p: 1 # top-p sampling

n: 1 # Number of completions to generate

parallelization: stagger: 0.3

num_threads: 50 # the number of threads to use for parallel processing

async_mode: threaded # or asyncio

embeddings:

parallelization: override the global parallelization settings for embeddings

async_mode: threaded # or asyncio llm: api_key: ${AZURE_API_KEY} type: azure_openai_embedding # or openai_embedding model: text-embedding-3-small api_base: https://graphrag-dataeaze.openai.azure.com/ api_version: 2024-02-15-preview

organization: 

deployment_name: graphrag_test
# tokens_per_minute: 150_000 # set a leaky bucket throttle
# requests_per_minute: 10_000 # set a leaky bucket throttle
# max_retries: 10
# max_retry_wait: 10.0
# sleep_on_rate_limit_recommendation: true # whether to sleep when azure suggests wait-times
# concurrent_requests: 25 # the number of parallel inflight requests that may be made
# batch_size: 16 # the number of documents to send in a single request
# batch_max_tokens: 8191 # the maximum number of tokens to send in a single request
# target: required # or optional

chunks: size: 300 overlap: 100 group_by_columns: [id] # by default, we don't allow chunks to cross documents

input: type: file # or blob file_type: text # or csv base_dir: "input" file_encoding: utf-8 file_pattern: ".*\.txt$"

cache: type: file # or blob base_dir: "cache"

connection_string:

container_name:

storage: type: file # or blob base_dir: "output/${timestamp}/artifacts"

connection_string:

container_name:

reporting: type: file # or console, blob base_dir: "output/${timestamp}/reports"

connection_string:

container_name:

entity_extraction:

llm: override the global llm settings for this task

parallelization: override the global parallelization settings for this task

async_mode: override the global async_mode settings for this task

prompt: "prompts/entity_extraction.txt" entity_types: ["Know Your Customer (KYC)", "Credit Cards", "Payment", "Anti-Money Laundering (AML)", "Bank Rate", "Repo Rate", "Debt", "Cash Reserve Ratio", "Commercial Banks", "Wilful Defaulters", "Financial Institutions", "Act", "Law", "Regulation", "Credit Information Companies", "Funds", "Risk Management", "Cooperative Banks", "Dealers", "Stocks", "Monetary Policy", "Inflation Control", "Banking Ombudsman", "Non-Banking Financial Companies (NBFCs)", "Financial Stability", "Basel Norms", "Liquidity Adjustment Facility (LAF)", "Statutory Liquidity Ratio (SLR)", "Priority Sector Lending", "Financial Inclusion", "Digital Payments", "Unified Payments Interface (UPI)", "Real Time Gross Settlement (RTGS)", "National Electronic Funds Transfer (NEFT)", "Deposit Insurance", "Currency Management", "Treasury Bills", "Government Securities", "Microfinance", "Payment Banks"]

max_gleanings: 0

summarize_descriptions:

llm: override the global llm settings for this task

parallelization: override the global parallelization settings for this task

async_mode: override the global async_mode settings for this task

prompt: "prompts/summarize_descriptions.txt" max_length: 500

claim_extraction:

llm: override the global llm settings for this task

parallelization: override the global parallelization settings for this task

async_mode: override the global async_mode settings for this task

enabled: true

prompt: "prompts/claim_extraction.txt" description: "Any claims or facts that could be relevant to information discovery." max_gleanings: 0

community_reports:

llm: override the global llm settings for this task

parallelization: override the global parallelization settings for this task

async_mode: override the global async_mode settings for this task

prompt: "prompts/community_report.txt" max_length: 1000 max_input_length: 4000

cluster_graph: max_cluster_size: 10

embed_graph: enabled: false # if true, will generate node2vec embeddings for nodes

num_walks: 10

walk_length: 40

window_size: 2

iterations: 3

random_seed: 597832

umap: enabled: false # if true, will generate UMAP embeddings for nodes

snapshots: graphml: false raw_entities: false top_level_nodes: false

local_search:

text_unit_prop: 0.5

community_prop: 0.1

conversation_history_max_turns: 5

top_k_mapped_entities: 10

top_k_relationships: 10

llm_temperature: 0 # temperature for sampling

llm_top_p: 1 # top-p sampling

llm_n: 1 # Number of completions to generate

max_tokens: 12000

global_search:

llm_temperature: 0 # temperature for sampling

llm_top_p: 1 # top-p sampling

llm_n: 1 # Number of completions to generate

max_tokens: 12000

data_max_tokens: 12000

map_max_tokens: 1000

reduce_max_tokens: 2000

concurrency: 32

Logs and screenshots

image Image from indexing-engine.log

Additional Information

therealTT commented 1 month ago

UPDATE: Setting max gleanings to -1 in settings.yaml seems to have the intended effect.