* Add prefixes to image models to solve the problem of duplicate models
* Fix the issue that response_format is not set, causing the b64_json parameter to be ignored.
* fix zhipu embedding error when input is array but not string
* fix: only use the first one
---------
Co-authored-by: 蔡新疆 <cxj@icc.link>
Co-authored-by: JustSong <songquanpeng@foxmail.com>
* feat: support for the ollama vision model
`llava` model, pass test
* Update main.go
format code
* chore: remove useless log
---------
Co-authored-by: nongqiqin <nongqiqin@tipdm.com>
Co-authored-by: JustSong <songquanpeng@foxmail.com>
* Update model.go
* Support Gemini tool_calls.
* Fix gemini tool calls (also keep support functions).
* Fixed the problem of arguments not being stringified.
Fix panic: candidate.Content.Parts out of range
* feat: support aws bedrockruntime claude3
closes#622, closes#749, closes#1300
* fix: convert to aws claude model id
* fix: Update AWS adapter to handle stream completions and calculate usage metrics
Based on the file summaries provided, here are the important bullet points for the commit message:
- Add functionality to handle stream completion events from AWS in the relay/adaptor/aws/main.go file
- Marshall AWS response to OpenAI format and calculate usage metrics in the same file
- Implement a custom render function for streaming events in the same file
- Improve error handling for JSON unmarshalling and marshalling errors in the same file
* fix: Implement AWS handler with usage tracking and error handling
- Implemented streaming response handling for AWS handler
- Set response content type to text/event-stream
- Added error handling for failed marshaling/unmarshaling
- Updated return values to include `relaymodel.ErrorWithStatusCode` and `relaymodel.Usage`
- Improved error handling and response formatting for AWS adaptor
* fix: Refactor AWS Adapter for Improved Model Mapping and Error Handling
* Refactor AWS adapter to improve model management
- Replace hardcoded model list in `adapter.go` with a function to get models from `awsModelIDMap`
- Update `GetModelList` function to return model list directly
- Add `GetChannelName` function to get channel name from `Adaptor` object
* Improve error handling and code organization in main.go
- Replace switch statement with a map to map AWS model IDs to OpenAI model IDs
- Return an error if the model is not found in the map
- Use a single return statement instead of wrapping multiple return statements in the `awsModelID` function
- Add a new error message for when the model is not found in the map in the `Handler` function
* fix: bug fix
* chore: change variable name & package
* chore: change variable name
* perf: update config related code
---------
Co-authored-by: JustSong <songquanpeng@foxmail.com>
* Refactor AWS adapter to improve model management
- Replace hardcoded model list in `adapter.go` with a function to get models from `awsModelIDMap`
- Update `GetModelList` function to return model list directly
- Add `GetChannelName` function to get channel name from `Adaptor` object
* Improve error handling and code organization in main.go
- Replace switch statement with a map to map AWS model IDs to OpenAI model IDs
- Return an error if the model is not found in the map
- Use a single return statement instead of wrapping multiple return statements in the `awsModelID` function
- Add a new error message for when the model is not found in the map in the `Handler` function
- Implemented streaming response handling for AWS handler
- Set response content type to text/event-stream
- Added error handling for failed marshaling/unmarshaling
- Updated return values to include `relaymodel.ErrorWithStatusCode` and `relaymodel.Usage`
- Improved error handling and response formatting for AWS adaptor
Based on the file summaries provided, here are the important bullet points for the commit message:
- Add functionality to handle stream completion events from AWS in the relay/adaptor/aws/main.go file
- Marshall AWS response to OpenAI format and calculate usage metrics in the same file
- Implement a custom render function for streaming events in the same file
- Improve error handling for JSON unmarshalling and marshalling errors in the same file
- Updated context key for model name in `Handler` and `StreamHandler` functions of `main.go` for consistency and potential future usage
- Functionality remains unchanged in this commit
- Refactor constant definitions and organization
- Clean up package level variables and functions
- Introduce new `relaymode` and `apitype` packages for constant definitions
- Refactor and simplify code in several packages including `openai`, `relay/channel/baidu`, `relay/util`, `relay/controller`, `relay/channeltype`
- Add helper functions in `relay/channeltype` package to convert channel type constants to corresponding API type constants
- Remove deprecated functions such as `ResponseText2Usage` from `relay/channel/openai/helper.go`
- Modify code in `relay/util/validation.go` and related files to use new `validator.ValidateTextRequest` function
- Rename `util` package to `relaymode` and update related imports in several packages
* Fix ali ConvertRequest function to use baidu keyword
* Support Ali stable-diffusion-xl and wanx-v1 model
* Support Ali stable-diffusion-xl and wanx-v1 model
* Support Ali stable-diffusion-xl and wanx-v1 model
* chore: update ali constants and model ratio
---------
Co-authored-by: JustSong <songquanpeng@foxmail.com>
Co-authored-by: JustSong <39998050+songquanpeng@users.noreply.github.com>
* Updated relay/channel/gemini package to use gin-gonic/gin for routing
* Added timeouts, environment variable for proxy, and error handling for HTTP clients in relay/util/init.go
* Improved error handling, URL path cases, and channel selection logic in middleware/distributor.go
* Added Content-Type header, closed request bodies, and added context to requests in relay/channel/common.go
* Upgraded various dependencies in go.mod
* Modified the GetRequestURL method in relay/channel/gemini/adaptor.go to include a "key" parameter in the URL and set a default version of "v1beta"
* Added io and net/http packages in relay/channel/gemini/adaptor.go and relay/channel/gemini/main.go
* Added a struct for GeminiStreamResp and modified response handling in relay/channel/gemini/main.go
* Imported packages for io and net/http added, gin-gonic/gin package added, and error handling improved in relay/channel/gemini/main.go
- Refactor error handling using `errors/v2` package for improved error management and formatting
- Implement changes in relay/controller/text.go, controller/channel-test.go, controller/relay.go, common/message/email.go, relay/controller/audio.go, controller/channel-billing.go, relay/controller/helper.go, relay/channel/common.go, relay/channel/minimax/main.go, and common/message/main.go
- Add package imports and update package paths for better code organization
- Replace hardcoded strings with error variables for consistency and readability
- Implement secure methods for generating unique Message-ID
- Improve SMTP handling and authentication in email.go
- Update response handling and error messages for better user experience
- Add constant checks and modify functions based on imageRequest parameters in helper.go
- Use `http.NewRequest` instead of `fmt.Fprint` for improved request handling in relay/channel/common.go
- Update error handling and formatting for better consistency across all files
- Updated error handling across multiple files with `Laisky/errors/v2` package
- Replaced hardcoded error messages with `Laisky/errors` in relay/channel/tencent/adaptor.go
- Added a function to check if a request should be retried in relay/controller/relay.go
- Removed unused imports and variables, and updated comments in various files
- Changed Redis cache handling in model/cache.go
- Refactored error handling in relay/channel/tencent/main.go and relay/channel/baidu/main.go
- Updated import paths and error handling in model/user.go, model/redemption.go, and controller/github.go
- Added import for tiktoken-go package in relay/channel/openai/token.go
- Added GetSign and ParseConfig functions in relay/channel/tencent/main.go
- Replaced specific error imports with a more general one in relay/channel/ali/adaptor.go
- Updated import comments and function calls in relay/channel/ali/adaptor.go
- Added checks and custom error messages in model/token.go
- Removed unused functions and variables in relay/channel/baidu/adaptor.go
- Imported "github.com/Laisky/errors/v2" package in controller/channel-billing.go
- Removed unused import packages in [relay/channel/tencent/adaptor.go](http://relay/channel/tencent/adaptor.go) and relay/channel/palm/adaptor.go
- Updated go.mod and go.sum files with new dependencies and versions
- Introduced idle connection timeout for HTTP client transport
- Implemented custom timeout for HTTP client when `RelayTimeout` is 0
- Set ImpatientHTTPClient timeout to 5 seconds
- Added new config variable `IdleTimeout` with default value of 30 seconds
- Utilized shared transport object for all HTTP clients in relay/util package
- Implemented error messages with request details and request parameter validation for image requests
- Added URL logic and Azure channel support for image generation requests
- Integrated quota management and updated quota usage logging for image generation requests
- Mapped original model names to mapped model names and unmarshalled image responses into a struct for further processing
- Modify `PostConsumeQuota` function in `common.go` to check if `totalQuota` is less than 0 and add error logging
- Update log content in `RecordConsumeLog` function of `common.go` to include model ratio and group ratio
- Change the method of calculating user used quota and request count in `UpdateUserUsedQuotaAndRequestCount` function of `common.go`
- Implement calculation of `completionRatio` and update `quota` calculation based on `completionRatio` in `postConsumeQuota` function of `helper.go`
- Implement error handling for `model.PostConsumeTokenQuota` and `model.CacheUpdateUserQuota` in `helper.go`
- Update `UserUsedQuota` and `ChannelUsedQuota` with new `quota` value in `helper.go`
- Move `scanner.Split` instantiation to a new function
- Introduce a new regular expression to extract data from the response
- Utilize regular expressions to pre-process the event stream
- Refactor content response
- Update channel adaptor to support `claude_model`
- Remove `null` stop reasons from content responses
- Add logging for error responses
- Change content start, ping, and message delta types to return true
- Set stop reason to end turn when response does not include a stop reason
- Set content response stop reason to null
- Add error handling for unmarshalling stream responses
- Rename `Completion` to `Text` in type definitions and `StopReason` to `Delta.StopReason`
- Count tokens in the `Delta.Text` field of the response instead of `Completion`
- Remove `Model` from the full text response
- Trim \r from incoming data
- Refactor Anthropic adapter to work with the new Anthropic API and model requests
- Remove the default value for `MaxTokensToSample`
- Set `MaxTokens` to 500 instead of 1000000
- Use `system` messages as the system prompt instead of the first message
- Rewrite model name for relay text
- Simplify logic and move logging statements
- Remove a check that filtered out models by model-mapping
- Remove abilities for the model mapping
- Lower numeric tolerance for test files
- Improved error handling in various modules for better stability and responsiveness.
- Optimized code in several files for improved efficiency and readability.
- Enhanced user experience by providing more detailed error responses in the controller.
- Strengthened security by ignoring sensitive files in `.gitignore`.
- Improved error handling in various modules for better stability and responsiveness.
- Optimized code in several files for improved efficiency and readability.
- Enhanced user experience by providing more detailed error responses in the controller.
- Strengthened security by ignoring sensitive files in `.gitignore`.
