* 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
* Changed the JSONNET for prompt-template so that prompt templates are
simpler. This tidies default-prompts.jsonnet so it's easier to see
where the prompts are.
* Updated prompt override templates to match new structure
- 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
* Separate Prom metrics, different processors as different jobs
* Create producers before consumers, may streamline startup.
* Bump version
* Add Pulsar init command, will replace pulsar-admin invocations.
* Integrate tg-init-pulsar with YAMLs
* Update YAMLs
* Add metering-rag for RAG token metrics
* Make metering executable
* Add metering-rag to Prom metrics
* Metricate the right queue
* Update templates to include metering-rag
- Fix Docker Compose relative directory problem
- Added CONFIGDIR environment variable for docker compose
- Correct Pulsar URL defaults for graph utils to be localhost again
Added templates which produce K8s resources. With the provided GCP wrapper, it works on GCP K8s cluster. This isn't stable enough for other folks to use so will need more piloting before it can be documented and released.
* Added prompt-template, allows definiton, relationships and kg query
to be specified in config / command-line.
* Bump version & add prompt-templates to YAMLs
* Apply to graph rag flow
* Break out different templates
* - Added docker-compose-storage.yaml, just the infrastructure bits
- Tidied storage invocation
* Util, sits on chunker output and reports histogram of chunk sizes
* Added a rate limit exception
* Reduce request/response timeouts because looks like there are major issues
* Add rate limit exception catch to all consumers
* Version to 0.6.3
