Code

Refactored LLM Integration #388

Code

Merged

copernico merged 37 commits intoSAP:mainSAP/project-kb:mainfrom lauraschauer:lascha-llm-servicelauraschauer/project-kb:lascha-llm-serviceCopy head branch name to clipboardJul 4, 2024

+716-475

Commits

Commits on Jun 24, 2024

• adjusts tests

  lauraschauer

  committed
  df82247

• basic refactoring mvp

  lauraschauer

  committed
  520e67d

• I can run the code, but the singleton pattern implemented with metaclasses still doesn't work

  lauraschauer

  committed
  da1d558

• singleton pattern works

  lauraschauer

  committed
  6b2b982

• move models into their own files

  lauraschauer

  committed
  3efb1a6

• updates tests

  lauraschauer

  committed
  f306298

• corrects information about missing commits to be logged as info instead of error

  lauraschauer

  committed
  a7cf5fd

• moves singleton metaclass into utils

  lauraschauer

  committed
  1cb3673

• adds tests

  lauraschauer

  committed
  15057d8

• changes variable name d to model_definition for clarity

  lauraschauer

  committed
  d55735d

• moves ai core sk filepath to config.yaml

  lauraschauer

  committed
  0a6d1c0

• adds StrOutputParser to model chain and prints returned URL to console

  lauraschauer

  committed
  8d105db

• streamlines error handling

  lauraschauer

  committed
  92a5401

• updates README with explanation of llm_service parameters in config.yaml

  Show description for 5601ef0
  lauraschauer

  committed
  5601ef0

• removes internal URL

  lauraschauer

  committed
  95b3917

• corrects mistake in singleton logger value

  lauraschauer

  committed
  beab86e

• adds singleton tests

  lauraschauer

  committed
  a672e7d

• refactors: LLMService is created in main, when there is access to the config.llm_service. When LLMService is needed again, just access the instance using LLMService() without having to supply the c…

  Show description for 8e9efc5
  lauraschauer

  committed
  8e9efc5

• refactors: simplifies model_instantiation, gets rid of two layer inheritance (no more SAPLLM)

  lauraschauer

  committed
  bb8701b

• updates tests

  lauraschauer

  committed
  291935c

• adds correct strings to return from _llm_type()

  lauraschauer

  committed
  1409b01

• corrects docstring

  lauraschauer

  committed
  1216710

• renames file

  lauraschauer

  committed
  72e0f30

• final touch-up

  lauraschauer

  committed
  8afa9c5

• fixes bug where a LLMService could not be created. Check logs. error was raised even though the use of llms was set to false

  Show description for cc85b0b
  lauraschauer

  committed
  cc85b0b

• fixes tests

  Show description for e45a136
  lauraschauer

  committed
  e45a136

• removes extra index urls

  lauraschauer

  committed
  be27723

• more env variables mocking for tests

  lauraschauer

  committed
  5a14377

• small modification to prompt

  lauraschauer

  committed
  32ef6fb

• removes pip options

  lauraschauer

  committed
  1d215b5

• adds more OpenAI models for third party providers

  lauraschauer

  committed
  f451ab4

• removes unused mock class

  lauraschauer

  committed
  988c3dd

• makes sure that --repository flag overrides config.yaml file

  lauraschauer

  committed
  7b497f5

• conceptional change: when llm_service is set in config.yaml, prospector assumes that LLM support is wanted. To not use LLM support, comment out the llm_service block in config.yaml

  lauraschauer

  committed
  0f8dd3d

Commits on Jul 2, 2024

• adds more fine-grained request error raising

  lauraschauer

  committed
  53638d3

Commits on Jul 4, 2024

• adds a check using regex to remove delimiters if they are returned by the LLM

  lauraschauer

  committed
  38060fb

• updates tests to take new version tags into account

  lauraschauer

  committed
  485cf83