search

SciPhi-AI / R2R

The Elasticsearch for RAG. Build, scale, and deploy state of the art Retrieval-Augmented Generation applications
https://r2r-docs.sciphi.ai/
MIT License
3.28k stars 244 forks source link

ValueError: Missing required keys in 'kg' config: batch_size #1028

Closed sunsssyc closed 2 weeks ago

sunsssyc commented 2 weeks ago

When using the guide to configure R2R for automatic knowledge graph construction during input file ingestion, the constructed graph can optionally be utilized in downstream R2R RAG processes. Knowledge graphs are generated locally using the newly released Triplex model.

r2r docker-down to bring down existing R2R Docker, if running.

r2r serve --docker --config-name=local_llm_neo4j_kg

For cloud providers, use config-name=neo4j_kg

the container r2r-r2r-1 can not start

the logs

self.loaded_app = import_from_string(self.app) 2024-09-02 22:00:36 File "/usr/local/lib/python3.10/site-packages/uvicorn/importer.py", line 21, in import_from_string 2024-09-02 22:00:36 module = importlib.import_module(module_str) 2024-09-02 22:00:36 File "/usr/local/lib/python3.10/importlib/init.py", line 126, in import_module 2024-09-02 22:00:36 return _bootstrap._gcd_import(name[level:], package, level) 2024-09-02 22:00:36 File "", line 1050, in _gcd_import 2024-09-02 22:00:36 File "", line 1027, in _find_and_load 2024-09-02 22:00:36 File "", line 1006, in _find_and_load_unlocked 2024-09-02 22:00:36 File "", line 688, in _load_unlocked 2024-09-02 22:00:36 File "", line 883, in exec_module 2024-09-02 22:00:36 File "", line 241, in _call_with_frames_removed 2024-09-02 22:00:36 File "/app/core/main/app_entry.py", line 68, in 2024-09-02 22:00:36 app = r2r_app( 2024-09-02 22:00:36 File "/app/core/main/app_entry.py", line 33, in r2r_app 2024-09-02 22:00:36 config = R2RConfig.from_toml(config_path) 2024-09-02 22:00:36 File "/app/core/main/assembly/config.py", line 144, in from_toml 2024-09-02 22:00:36 return cls(config_data, base_path=Path(config_path).parent) 2024-09-02 22:00:36 File "/app/core/main/assembly/config.py", line 88, in init 2024-09-02 22:00:36 self._validate_config_section(default_config, section, keys) 2024-09-02 22:00:36 File "/app/core/main/assembly/config.py", line 129, in _validate_config_section 2024-09-02 22:00:36 raise ValueError( 2024-09-02 22:00:36 ValueError: Missing required keys in 'kg' config: batch_size

I executed the Docker deployment on a Mac.

shreyaspimpalgaonkar commented 2 weeks ago

hey, could you share the command that you ran?

sunsssyc commented 2 weeks ago

r2r serve --docker --config-name=local_llm_neo4j_kg

shreyaspimpalgaonkar commented 2 weeks ago

Thanks for flagging! You can do a quick fix by adding batch_size to the config. I'll push this fix out in the next release.

The config is stored in R2R/py/core/configs/local_llm_neo4j_kg.toml

[kg]
provider = "neo4j"
batch_size = 256
kg_extraction_prompt = "graphrag_triplet_extraction_zero_shot"

It should work now!

sunsssyc commented 2 weeks ago

Thank you so much