Signed-off-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
Co-authored-by: Alex Jones <alexsimonjones@gmail.com>
* feat: added support for A21 and Amazon Titan models via bedrock api
Signed-off-by: Yomesh Shah <yomesh@gmail.com>
* fix: response type for diffrent models and use of constant for top_P
Signed-off-by: Yomesh Shah <yomesh@gmail.com>
* fix: constant for top_P as int vs string
Signed-off-by: Yomesh Shah <yomesh@gmail.com>
* feat: moved topP and maxTokens to config rather than being constants in the code
Signed-off-by: Yomesh Shah <yomesh@gmail.com>
---------
Signed-off-by: Yomesh Shah <yomesh@gmail.com>
Co-authored-by: Alex Jones <alexsimonjones@gmail.com>
Signed-off-by: Guangya Liu <gyliu@ibm.com>
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
Co-authored-by: Alex Jones <alexsimonjones@gmail.com>
Co-authored-by: Matthis <matthish29@gmail.com>
This commit adds Top-K sampling, a feature that allows users to control
the randomness of the generated text by specifying the number of most
probable next words considered by the model. This enhances user control
and potentially improves the quality of the generated outputs.
Fixes: https://github.com/k8sgpt-ai/k8sgpt/issues/1105
Signed-off-by: VaibhavMalik4187 <vaibhavmalik2018@gmail.com>
Co-authored-by: Alex Jones <alexsimonjones@gmail.com>
* chore(deps): update cohere client implementation to v2 and to use chat endpoint
Signed-off-by: Miguel Varela Ramos <miguel@cohere.ai>
* chore: remove renovate rule for cohere-go
Signed-off-by: Miguel Varela Ramos <miguel@cohere.ai>
* style: remove unused attribute
Signed-off-by: Miguel Varela Ramos <miguel@cohere.ai>
* fix: go mod
Signed-off-by: Miguel Varela Ramos <miguel@cohere.ai>
---------
Signed-off-by: Miguel Varela Ramos <miguel@cohere.ai>
Signed-off-by: Miguel Varela Ramos <miguelvramos92@gmail.com>
Co-authored-by: Alex Jones <alexsimonjones@gmail.com>
* fix: set topP from config
Signed-off-by: “Guido <muscionig@gmail.com>
* style: correct format of openai ai provider
Signed-off-by: “Guido <muscionig@gmail.com>
* feat: set topP from the environment
Signed-off-by: “Guido <muscionig@gmail.com>
---------
Signed-off-by: “Guido <muscionig@gmail.com>
* chore: allows an environmental override of the default AWS region and using it for bedrock
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
* chore: missing provider region
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
---------
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
* feat: add Google Vertex AI as provider to utilize gemini via GCP
Signed-off-by: Mario Fahlandt <mfahlandt@pixel-haufen.de>
* fix: adjust providerId description
Signed-off-by: Mario Fahlandt <mfahlandt@pixel-haufen.de>
---------
Signed-off-by: Mario Fahlandt <mfahlandt@pixel-haufen.de>
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
Co-authored-by: Alex Jones <alexsimonjones@gmail.com>
Co-authored-by: Aris Boutselis <arisboutselis08@gmail.com>
* chore: wip interactive mode
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
* chore: tidied up a bit
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
* chore: updated go mod
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
* chore: const prompt
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
* chore: updated based on comments feedback
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
* chore: updated enum
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
* chore: updated enum
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
---------
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
* feat: initial Prometheus analyzers
Added a prometheus integration with two analyzers:
1. PrometheusConfigValidate
2. PrometheusConfigRelabelReport
The integration does not deploy any Prometheus stack in the cluster.
Instead, it searches the provided --namespace for a Prometheus
configuration, stored in a ConfigMap or Secret. If it finds one, it
unmarshals it into memory and runs the analyzers on it.
PrometheusConfigValidate checks if the actual Prometheus configuration is valid or has
any errors.
PrometheusConfigRelabelReport tries to distill the scrape config
relabeling rules to give a concise label set per job that targets need
to have to be scraped. This analyzer is unconventional, in that it does
not necessarily mean there are issues with the config. It merely tries
to give a human-readable explanation of the relabel rules it discovers,
leaning on the LLM and prompt.
Tested on both kube-prometheus and Google Managed Prometheus
stacks.
Signed-off-by: Daniel Clark <danielclark@google.com>
* review: feedback cycle 1
Simplify ConfigValidate prompt and add comments.
Signed-off-by: Daniel Clark <danielclark@google.com>
* review: feedback cycle 2
Add Prometheus configuration discovery to integration activate command.
Also improve logging to make this more clear to users.
Signed-off-by: Daniel Clark <danielclark@google.com>
---------
Signed-off-by: Daniel Clark <danielclark@google.com>
* refactor: Simplified IAI; made caching and processing consisent.
Signed-off-by: bwplotka <bwplotka@gmail.com>
* feat: Added Google AI API e.g. for Gemini models.
Signed-off-by: bwplotka <bwplotka@gmail.com>
---------
Signed-off-by: bwplotka <bwplotka@gmail.com>
Co-authored-by: Alex Jones <alexsimonjones@gmail.com>
Co-authored-by: Thomas Schuetz <38893055+thschue@users.noreply.github.com>
* add amazonbedrock AI provider
Signed-off-by: Su Wei <suwei007@gmail.com>
* add amazonbedrock, change model list to const var
Signed-off-by: Su Wei <suwei007@gmail.com>
* update iai config and auth cmd, add providerRegion
Signed-off-by: Wei Su <wsuam@amazon.com>
* fix filename wrong
Signed-off-by: Wei Su <wsuam@amazon.com>
* chore: added some doc info
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
---------
Signed-off-by: Su Wei <suwei007@gmail.com>
Signed-off-by: Wei Su <wsuam@amazon.com>
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
Co-authored-by: Wei Su <wsuam@amazon.com>
Co-authored-by: Aris Boutselis <aris.boutselis@senseon.io>
Co-authored-by: Alex Jones <alexsimonjones@gmail.com>
* feat: openAI explicit value for maxToken and temp
Because when k8sgpt talks with vLLM, the default MaxToken is 16,
which is so small.
Given the most model supports 2048 token(like Llama1 ..etc), so
put here for a safe value.
Signed-off-by: Peter Pan <Peter.Pan@daocloud.io>
* feat: make temperature a flag
Signed-off-by: Peter Pan <Peter.Pan@daocloud.io>
---------
Signed-off-by: Peter Pan <Peter.Pan@daocloud.io>
* feat: show each ConfigAuditReport check
Signed-off-by: Johannes Kleinlercher <johannes@kleinlercher.at>
* feat: mask sensitive data in configauditreport messages
Signed-off-by: Johannes Kleinlercher <johannes@kleinlercher.at>
---------
Signed-off-by: Johannes Kleinlercher <johannes@kleinlercher.at>
* feat: added the ability to set a user default AI provider
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
* feat: added the ability to set a user default AI provider
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
* chore: added provider to json output
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
---------
Signed-off-by: Alex Jones <alexsimonjones@gmail.com>
* gofmt the files
Signed-off-by: Peter Pan <Peter.Pan@daocloud.io>
* add UT and goFMT to PR Gate (Github Action for PR)
Signed-off-by: Peter Pan <Peter.Pan@daocloud.io>
---------
Signed-off-by: Peter Pan <Peter.Pan@daocloud.io>
Co-authored-by: Alex Jones <alexsimonjones@gmail.com>
Instead of storing cached values in the config yaml, they are now stored
under these OS specific locations:
* Linux: `~/.cache/k8sgpt`
* MacOS: `~/Library/Caches`
* Windows: `%LocalAppData%\cache`
Additionally a `Cache` package and interface has been introduced.
Currently there are two implementations:
* Noop - Doesn't do anything
* FileBased - Stores data in files under the locations listed above
fixes#323
Signed-off-by: Patrick Pichler <git@patrickpichler.dev>
* feat: allow to set a baseURL for OpenAI providers
This allows to run local models that have a compatible OpenAI API, or
for instance use a proxy.
Signed-off-by: mudler <mudler@mocaccino.org>
* feat: allow to set baseURL in the auth subcommand
Signed-off-by: mudler <mudler@mocaccino.org>
---------
Signed-off-by: mudler <mudler@mocaccino.org>
Co-authored-by: Alex Jones <alexsimonjones@gmail.com>
Co-authored-by: Matthis <99146727+matthisholleville@users.noreply.github.com>
