GraphRAG with llama-index by rbrugaro · Pull Request #793 · opea-project/GenAIComps

rbrugaro · 2024-10-15T21:59:29Z

Description

GraphRAG components using llama-index:

comps/dataprep/neo4j/llama-index/extract_graph_neo4j.py
comps/retrievers/neo4l/llama-index/retriever_community_answers_neo4j.py

later i will add a GenAIExample that will combine these 2 with the DocSum to generate the final answers from the community summaries

Signed-off-by: rbrugaro <rita.brugarolas.brufau@intel.com>

for more information, see https://pre-commit.ci

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

for more information, see https://pre-commit.ci

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

…into graphRAG_LI

for more information, see https://pre-commit.ci

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

ichbinblau

Hi, @rbrugaro Thanks for enabling Graphrag with llama-index. I found some inconsistent places in the commits during the test. PTAL.

comps/retrievers/neo4j/llama_index/README.md

comps/dataprep/neo4j/llama_index/README.md

rbrugaro · 2024-10-23T03:54:57Z

@ichbinblau Thanks for catching this!, PR still in draft since i need to do some clean up and will take care of the READMEs

…triever Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

for more information, see https://pre-commit.ci

codecov · 2024-10-23T21:11:29Z

Codecov Report

Attention: Patch coverage is 9.52381% with 38 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
comps/cores/mega/gateway.py	7.31%	38 Missing ⚠️

Files with missing lines	Coverage Δ
comps/cores/mega/constants.py	`98.41% <100.00%> (+0.02%)`	⬆️
comps/cores/mega/gateway.py	`30.24% <7.31%> (-2.07%)`	⬇️

... and 1 file with indirect coverage changes

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

for more information, see https://pre-commit.ci

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

tests/retrievers/test_retrievers_neo4j_llama_index_on_intel_hpu.sh

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

…into graphRAG_LI

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

ashahba

Thanks @rbrugaro for this PR!
This looks like a great PR and my comments and change requests are mostly BKMs for code consistency.

comps/cores/mega/gateway.py

comps/dataprep/neo4j/llama_index/README.md

tests/dataprep/test_dataprep_neo4j_llama_index_on_intel_hpu.sh

tests/retrievers/test_retrievers_neo4j_llama_index_on_intel_hpu.sh

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

ashahba

LGTM!

* graphRAG dataprep llama-index validated w openai endpoints Signed-off-by: rbrugaro <rita.brugarolas.brufau@intel.com> * llama-index graphRAG retrieval validated with openai models Signed-off-by: rbrugaro <rita.brugarolas.brufau@intel.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * validated code usng TGI/TEI instead of openai Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * compose.yaml for dataprep validated with neo4j, TGI/TEI, openai Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * graphRAG retriever validated and full compose.yaml Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * minor fix Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * add graphRAGGateway WIP Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * graphragGateway working in E2E Example Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * fix schedule in orchestrator to support ChatCompletionRequest input Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * change default to TGI instead of openAI and add test code for neo4jretriever Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * test code for dataprep-neo4j microservice Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * improved READMES Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * update docker build path for tests Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * minor fix Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * fix typo in container name Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * resolve image name conflict for hub publishing Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * add tgi validation to miicroservice tests Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * rename test script to trigger cicd correctly w hpu Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * rename test containers for cicd Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * update HF_TOKEN in TGI/TEI test scripts Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * swaped larger model so the graph isnt empty in ci test Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * set 4 hpu for 70B model in ci test Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * add extra time for large model loading cicd Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * fix tgi gaudi shard args Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * switch to chat cause chat template is needed Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * enable logs in test Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * use locally downloaded model in CI machine Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * use local model path and reduce wait time Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * clear ports before ci run Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * fix cache model access Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * fix cache model access Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * incrased wait time for tgi shards ready Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * wait until tgi connected Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * switch back to small model for testing Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * minor readability fixes Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> * README fixes Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> --------- Signed-off-by: rbrugaro <rita.brugarolas.brufau@intel.com> Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

rbrugaro and others added 14 commits October 15, 2024 13:53

graphRAG dataprep llama-index validated w openai endpoints

b0a95a3

Signed-off-by: rbrugaro <rita.brugarolas.brufau@intel.com>

llama-index graphRAG retrieval validated with openai models

4e1db1f

Signed-off-by: rbrugaro <rita.brugarolas.brufau@intel.com>

[pre-commit.ci] auto fixes from pre-commit.com hooks

ba1cffb

for more information, see https://pre-commit.ci

[pre-commit.ci] auto fixes from pre-commit.com hooks

dc97611

for more information, see https://pre-commit.ci

validated code usng TGI/TEI instead of openai

d4a6459

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

compose.yaml for dataprep validated with neo4j, TGI/TEI, openai

012a503

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

fix merge conflict

a0668b7

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

[pre-commit.ci] auto fixes from pre-commit.com hooks

f345bb6

for more information, see https://pre-commit.ci

[pre-commit.ci] auto fixes from pre-commit.com hooks

0cf55a8

for more information, see https://pre-commit.ci

graphRAG retriever validated and full compose.yaml

1ceb659

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

minor fix

8e034cd

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

Merge branch 'graphRAG_LI' of https://github.com/rbrugaro/GenAIComps …

3ca5298

…into graphRAG_LI

[pre-commit.ci] auto fixes from pre-commit.com hooks

9eee5cb

for more information, see https://pre-commit.ci

add graphRAGGateway WIP

109a04e

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

rbrugaro mentioned this pull request Oct 22, 2024

Graph rag opea-project/GenAIExamples#1007

Merged

rbrugaro added 2 commits October 22, 2024 19:10

graphragGateway working in E2E Example

7e2c2ea

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

fix schedule in orchestrator to support ChatCompletionRequest input

7e5d810

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

ichbinblau reviewed Oct 23, 2024

View reviewed changes

rbrugaro and others added 5 commits October 23, 2024 06:23

change default to TGI instead of openAI and add test code for neo4jre…

7b0e51a

…triever Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

test code for dataprep-neo4j microservice

85f0efe

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

improved READMES

9d103a7

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

Merge remote-tracking branch 'upstream/main' into graphRAG_LI

08b107c

[pre-commit.ci] auto fixes from pre-commit.com hooks

7a8fe54

for more information, see https://pre-commit.ci

rbrugaro marked this pull request as ready for review October 23, 2024 21:19

rbrugaro requested a review from lvliang-intel as a code owner October 23, 2024 21:19

rbrugaro and others added 3 commits October 24, 2024 03:40

update docker build path for tests

8ddbabd

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

[pre-commit.ci] auto fixes from pre-commit.com hooks

9feeaae

for more information, see https://pre-commit.ci

minor fix

a1c7238

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

chensuyue reviewed Oct 28, 2024

View reviewed changes

tests/retrievers/test_retrievers_neo4j_llama_index_on_intel_hpu.sh Outdated Show resolved Hide resolved

rbrugaro added 2 commits October 28, 2024 17:32

update HF_TOKEN in TGI/TEI test scripts

8fcc940

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

swaped larger model so the graph isnt empty in ci test

18f1ad7

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

rbrugaro mentioned this pull request Oct 28, 2024

GraphRAG QA Example opea-project/GenAIExamples#1036

Closed

4 tasks

rbrugaro added 13 commits October 28, 2024 21:47

set 4 hpu for 70B model in ci test

be2b3e5

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

add extra time for large model loading cicd

d0ae033

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

fix tgi gaudi shard args

5791644

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

switch to chat cause chat template is needed

2501099

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

enable logs in test

e4c839e

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

Merge branch 'main' into graphRAG_LI

d752dd5

use locally downloaded model in CI machine

057dde2

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

Merge branch 'graphRAG_LI' of https://github.com/rbrugaro/GenAIComps …

a0eb001

…into graphRAG_LI

use local model path and reduce wait time

847e132

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

clear ports before ci run

fee0735

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

fix cache model access

7acdfad

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

fix cache model access

f9ac6ba

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

incrased wait time for tgi shards ready

09d011f

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

ashahba requested changes Oct 29, 2024

View reviewed changes

rbrugaro added 5 commits October 29, 2024 21:28

wait until tgi connected

cd0e4f9

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

switch back to small model for testing

eae835b

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

minor readability fixes

7b09de6

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

Merge branch 'main' into graphRAG_LI

ebbc556

README fixes

2a33510

Signed-off-by: Rita Brugarolas <rita.brugarolas.brufau@intel.com>

rbrugaro requested a review from ashahba October 30, 2024 03:34

XuhuiRen approved these changes Oct 30, 2024

View reviewed changes

ashahba approved these changes Oct 30, 2024

View reviewed changes

ashahba requested review from lvliang-intel and removed request for lvliang-intel October 30, 2024 05:43

ashahba merged commit 19330ea into opea-project:main Oct 30, 2024

Conversation

rbrugaro commented Oct 15, 2024

Description

Uh oh!

ichbinblau left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

rbrugaro commented Oct 23, 2024

Uh oh!

codecov bot commented Oct 23, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Uh oh!

ashahba left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

ashahba left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

codecov bot commented Oct 23, 2024 •

edited

Loading