* init

* fix: remove package

* #3: project organization

* #3-feat: build chat api

* #3 rename func

* Remove unused dependencies from go.sum

* #3: Add .env to .gitignore

---------

Co-authored-by: Max <mkrueger190@gmail.com>

* #12 Inited the Gateway, ServerManager components + CLI

* #12 Inited ServerManager

* #12 Used hertz as a HTTP server

* #12 Added todo

* #13 Added the health endpoint

* #3: Update dependencies, fix package names, add validators

* #3: Remove unused fields in ProviderConfig struct

* #3: Update OpenAI provider configuration validation

* #3: Refactor OpenAI provider config structure

* #3: Update go.mod and go.sum files

* #3: add comments

* #3: Refactor package name in OpenAI API and chat files

* #3: build OpenAI config

* #3: Build API Request

* #3: Refactor param checking in BuildAPIRequest function

* #3: build request based on client parameters

---------

Co-authored-by: Max <mkrueger190@gmail.com>

* #15: Update dependencies in go.mod and add input validation using validator package

* #15: Fix validation error in BuildAPIRequest function

* #15: Refactor OpenAI provider configuration structure

* #15: TODO: Import provider configs to the config list

* #15: Import OpenAI provider configuration and update BuildAPIRequest function

* #15: comments

---------

Co-authored-by: Max <mkrueger190@gmail.com>

* #24: Refactor OpenAI provider config struct

* #24: Update OpenAiProviderConfig Messages field validation

---------

Co-authored-by: Max <mkrueger190@gmail.com>

* #8 Setting up CI checks via GH Actions

* #8 Fixing lint errors in the existing codebase

* #8 Fixed remaining lint errors

* #8 Installed tools via makefile, merged tests with the lint yaml

* #8 Disabled troublesome checks

* #8 fixing the minor version for gosec

* #8 Run gosec directly in the job

* #14 Inited Glide logging based on Zap

* #14 introducing a new telemetry struct

* #14: Introduced a new telemetry struct to hold all observability components

* #14: Added the router struct

* #14: Configured hertz logger to obey the general logging config

* #14 Added more static checks

* #35: Updated readme

* #35: Added more sections

* #35: minor adjustments

* #35 Optimized the logo

* #35: Added more useful info

Centered the logo

made badge color more readable

* #11: Added a new config struct

* #11: Added API config and connected it to the general config

* #11 Added config provider & loading a config file

* #11 Added directive expander

* #11 Added a secret field

* #11 Linting

* #11 Added default configs & replaced mapstruct back to YAML & simplified expander setup

* #11 finished directive expander

* #11 linting

* #11 Added a test for a single $ char

* #11 linting

* #11 fixed typo

* #29: openai client

* #29: add sample

* #29: Refactor OpenAI provider configuration

* #29: Refactor OpenAiClient struct and methods

* #29: Refactor OpenAiClient struct and methods

* #29: refactor

* 29: chat converted to Hertz

* #29: Update OpenAI provider configuration and chat request validation

* #29: clean up

* #29: Remove unused code and refactor parseStreamingChatResponse function

* #29: Update dependencies in go.mod and go.sum files

* #29: build client init

* #29: CreatChatRequest

* #29: Fix import and log package names

* #29: add run method

* #29: tests passing

* #29: tests passing

* #29: add todo

* #29: go mod tody

* #29: gofmt

* #29: gofumpt

* #29: lint

* #29: lint

* #29: lint

* #29: fix Implicit memory aliasing

* #29: lint

* #29: update request from defaultParams

* #29: lint

* #29: lint

* #29: emove unused variable in OpenAiClient constructor

* #29: Update client

* #29: Refactor OpenAI chat functionality

* #29: Fix OpenAI client error handling and validation

* #29: Fix OpenAI client error handling and validation

* #29: Set the appropriate base URL in OpenAiClient constructor

* #29: Refactor OpenAI provider configuration

* #29: comments

* #29: Refactor OpenAI client and related functions

* #29: chores

* #29: chores

* #29: comment

* #29: create a yaml for provider global configs

* #29: comments

* #29: comments

* #29: update http client

* #29: Add file path validation and error handling in openaiclient.go

* #29: move common helpers to types.go

* #29: chores

* #29: Refactor OpenAI client and chat request creation

* #29: Remove unused dependencies and update dependencies

* #29: Update Unified Data structure

* #29: remove provider

* #29: clean up

* #29: clean up

* #29: lint

* #29: remove hertz comment

* #29: pass unified data to chat method

* #29: init logging

* #29: init logging

* #29: init logging

* #29: lint

---------

Co-authored-by: Max <mkrueger190@gmail.com>

* #39: Report test coverage to CodeCov

* #39: Fixing linting

* #39: linting

* #39 Adding the coverage badge to the readme

* Fail on no config file

* Use the dev config in the make run command

* Covered a few exceptional cases

* Linting

* Linting

- Cleaned up and restructured provider/openai codebase. Separated unified schemas to the api package.
- Added and exposed the language chat API with the unified request/response schemas. Updated the bruno collection with this request
- Added an example of client tests
- Connected Glide API with the underlying model provider (OpenAI client is hardcoded for now)
- Implemented default value setting for nested nillable config items
- Implemented provider setting validation on the model item level

* 22: Installed & exposed API documentation

* #22 Left a note about API docs in the readme.md

* #22 Check if API docs were updated in CI

* #22: linting

* #22: install swag CLI

* #51: Build routers & models based on provided config

* #51 Fixed circular import issue

* #51: Checked model config before ending initialization

* #51: updated the dev config to setup the openAI model

* 51: Fixing vuln issues

* #51 logged more context

* #51: Collected routers as slice

* #51: Added the new lang routers API

* #51: Hide api key filed completely

Updated some of the description and spelling

* #54: Refactor UnifiedChatResponse struct and add ProviderResponse and TokenCount structs

* #54: Unified response created and tested - passing

* #54: lint

* #54: lint

---------

Co-authored-by: Max <mkrueger190@gmail.com>

* #54: Refactor UnifiedChatResponse struct and add ProviderResponse and TokenCount structs

* #54: Unified response created and tested - passing

* #54: lint

* #5: init

* #5: build cohere chat client

* #5: lint

* #5: Cohere provider working

* #5: lint

* #5: remove api key lol

* #5: Update OpenAI references to Cohere

* #5: Refactor cohere client test and update response ID

* #5: lint

* #5: lint

---------

Co-authored-by: Max <mkrueger190@gmail.com>

* #61: Create OpenAI Chat Response Schema

* #61: Test unmarshling

* #61: unmarshal openai response

* #61: Update CohereChatCompletion struct and related types

* #61: Update Cohere chat response mapping

* #61: lint

* #61: lint

---------

Co-authored-by: Max <mkrueger190@gmail.com>

* #4: add azure openai client

* #4: update testdata

* #4: lint

* #4: add tests

* #4: fix var name

* #4: add tests

* #4: tests

* #4: fix model response

---------

Co-authored-by: Max <mkrueger190@gmail.com>

typo

* #60: Add Anthropic provider

* Update chat provider response schema, replace responseID

* #60: Fix field naming conventions in language.go and chat.go

* #60: reduce complexity of toModel func and lint

* #60: add test

* #60: lint

---------

Co-authored-by: Max <mkrueger190@gmail.com>

* #9 Inited goreleaser

* #9: Inited the goreleaser

* #9 Added a GH Action to release binaries on tags

* #9: Setup release of binaries & announcement

* #9: fixed syntax errors

* #9 Commented out some lines that could cause noise & made the release workflow be dependant on the lint & vuln workflows

* #9: allow to run linting & vuln checks on tags

* #9 fixing dependencies between workflows

* #9 Tried to check the workflow name

* #9 Trying a different approach to reuse the other workflows in the release workflow

* #9 Disabled linux arm builds

* #9 disabled build for linux 386

* #9 removed s390x arch as it fails

* #9 removed the ppc64le arch because of failures

* #9: Removed riscv64 arch

* #9: Tried to fit a big number into uint32 on 32bit systems

* #9: fixed the cmd

* #9: fixed linting

* #9: Removed 386 as a arch target

* #9: ignored win arm64 pair

* #9: removed openbsd because of netpoll issues

* #9 Removed win from the OS target

* #9: try to force it release things

* #9: Fixed the token assignment

* #9: Added images and make commands to build them

* #9: Specified the linux as the target OS to fix issues with binary

* #9: Publishing the latest tags

* #9: completed the release workflows for images

* #9: disabled login into ghcr temporarily

* #9 commented more steps to speed up testing

* #9 exposing more info & change the working dir

* #9 using working directory to check out the right path

* #9: trying to fix the build time to keep the same in bins and images

* #9: Fixing the empty date

* #9 trying to fix the build time

* #9 fixing the ref

* #9 Trying to push images to GHCR

* #9 removed unneeded deps

* #9 Uncommented announcements on new releases

* #9: Passing Discord secrets to goreleaser

* #9: Added the basic homebrew tap setup

* #9 Made sure releases happen on the main branch only

* #9: Trying to use the deployment key to update the brew tap

- Made sure it's not possible to override the required fields by nil values accidentally by misformatting the config file
- used github.com/go-playground/validator/v10 for config validation
- created custom validation errors
- Made sure router and model IDs are unique
- Made sure there is at least one lang router configured
- Made sure that each router has at least one model configured
- Warn users in case there is only one model in the router's pool

* #41: Exposed the basic HTTP server configs

* #41: Added the link to hertz configs

* 📝 Updated the installation methods

* 📝 Using the demo repo as a preferred quickest way to get started

* #81: Update request schema so a model can be specified

* #81: Add model field to JSON payload

* #81: refactor getPayload function

* #81: lint

* #81: Refactor Chat function to support message override, add back pointer

* #81: add docs

---------

Co-authored-by: Max <mkrueger190@gmail.com>

* 📝 #104 Added the marketecture

* #104 Added the discord chat badge

- Normalize response latency by response token count 
- renamed TokenCount to TokenUsage

* 🧹 Actualized links everywhere