LanguageModelSession.GenerationError.exceededContextWindowSize(_:)

Declaration

case exceededContextWindowSize(LanguageModelSession.GenerationError.Context)

Mentioned in

• Analyzing the runtime performance of your Foundation Models app
• Generating content and performing tasks with Foundation Models

Discussion

This error occurs when you use the available tokens for the context window of 4,096 tokens. The token count includes instructions, prompts, and outputs for a session instance. A single token corresponds to approximately three to four characters in languages like English, Spanish, or German, and one token per character in languages like Japanese, Chinese, and Korean.

Start a new session when you exceed the content window size, and try again using a shorter prompt or shorter output length.

For more information on managing the context window size, see TN3193: Managing the on-device foundation model’s context window.

See Also

Generation errors

• LanguageModelSession.GenerationError.assetsUnavailable(_:)
• LanguageModelSession.GenerationError.decodingFailure(_:)
• LanguageModelSession.GenerationError.guardrailViolation(_:)
• LanguageModelSession.GenerationError.rateLimited(_:)
• LanguageModelSession.GenerationError.refusal(_:_:)
• LanguageModelSession.GenerationError.concurrentRequests(_:)
• LanguageModelSession.GenerationError.unsupportedGuide(_:)
• LanguageModelSession.GenerationError.unsupportedLanguageOrLocale(_:)
• LanguageModelSession.GenerationError.Context
• LanguageModelSession.GenerationError.Refusal