- Fixed document-rag workspace problem
- OpenAI text-completion processor now puts 'not-set' in the token
if no token is set (new OpenAI library requires it to be set to
something.
- Update tests
Use max_completion_tokens for OpenAI and Azure OpenAI providers:
The OpenAI API deprecated max_tokens in favor of
max_completion_tokens for chat completions. Newer models
(gpt-4o, o1, o3) reject the old parameter with a 400 error.
AZURE_API_VERSION env var now overrides the default API version:
(falls back to 2024-12-01-preview).
Update tests to test for expected structures
* Tidy up duplicate tech specs in doc directory
* Streaming LLM text-completion service tech spec.
* text-completion and prompt interfaces
* streaming change applied to all LLMs, so far tested with VertexAI
* Skip Pinecone unit tests, upstream module issue is affecting things, tests are passing again
* Added agent streaming, not working and has broken tests
* Remove some 'unnecessary' parameters from OpenAI invocation. The OpenAI
API is getting complicated with the API and SDK changing on OpenAI's end,
but this not getting mapped through to other services which are 'compatible'
with OpenAI.
* Update OpenAI test for this change
* Trying running tests with Python 3.13
* Implement KG extraction agent (kg-extract-agent)
* Using ReAct framework (agent-manager-react)
* ReAct manager had an issue when emitting JSON, which conflicts which ReAct manager's own JSON messages, so refactored ReAct manager to use traditional ReAct messages, non-JSON structure.
* Minor refactor to take the prompt template client out of prompt-template so it can be more readily used by other modules. kg-extract-agent uses this framework.
- Changed GoogleAIStudio LLM code to match latest documentation
- Very minor tweak to vertexai LLM code - just matching what's in SDK docs
no actual change to implementation.
- Tweaked VertexAI container build to speed up in dev
- Comments in LLM code to mention which docs it was built from. Google
SDKs are confusing ATM.
- Keeps processing in different flows separate so that data can go to different stores / collections etc.
- Potentially supports different processing flows
- Tidies the processing API with common base-classes for e.g. LLMs, and automatic configuration of 'clients' to use the right queue names in a flow
- prompt-template takes config from the config-svc, dynamically reloads
as new config appears.
- agent-react takes config from config-svc, dynamically reloads
- Fixed lack of data in config queue, needed to take the Earliest, not the
Latest values.
- Changed text-completion and knowledge-query tool to both use 'query'
as the argument.
- Prompt and agent no longer have command line args to supply
configuration.
Reconfigure so that AZURE_TOKEN, AZURE_MODEL and AZURE_ENDPOINT
can be used to set the token/model/endpoint parameters. This allows it to
be deployed in K8s and use secrets to set these environment variables
* - Refactored retry for rate limits into the base class
- ConsumerProducer is derived from Consumer to simplify code
- Added rate_limit_count metrics for rate limit events
* Add rate limit events to VertexAI and Google AI Studio
* Added Grafana rate limit dashboard
* Add rate limit handling to all LLMs
* Added single-target command-line config generator. Mainly using for
testing ATM.
* Slightly tweak the config decode so that components can over-ride the
'with' method which injects parameters.
* Deliberately break the prompt-generic template. Could do better, this
is temporary.
* Add 'prompt-overrides' component, injects new prompts.
* Removed prompt generic reference, not used
* prompt-generic is no longer supported
- Change templates to interpolate environment variables in docker compose
- Change templates to invoke secrets for environment variable credentials in K8s configuration
- Update LLMs to pull in credentials from environment variables if not specified
