Module: RubyLLM::Protocols::Gemini::FileTranscription

Included in:
Interactions::Transcription, RubyLLM::Providers::VertexAI::Transcription
Defined in:
lib/ruby_llm/protocols/gemini/file_transcription.rb

Overview

:nodoc: all

Instance Method Summary collapse

Instance Method Details

#transcribe(audio_file, model:, language:, format: nil, speaker_names: nil, speaker_references: nil, provider_options: {}, prompt: nil, temperature: nil) ⇒ Object 



7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
 
# File 'lib/ruby_llm/protocols/gemini/file_transcription.rb', line 7

def transcribe(audio_file, model:, language:, format: nil, speaker_names: nil,
               speaker_references: nil, provider_options: {}, prompt: nil, temperature: nil)
  raise_transcription_streaming_unsupported if block_given?
  validate_transcription_request(format:, speaker_references:, temperature:)
  attachments = Attachment.wrap(audio_file, config: @config)
  unless attachments.one? && attachments.first.audio?
    raise ArgumentError, 'Dedicated transcription requires exactly one audio file'
  end

  track_usage(:transcription) do
    payload = render_transcription_payload(attachments.first, model:, language:, speaker_names:,
                                                              provider_options:, prompt:)
    response = @connection.post(transcription_url(model), payload, usage: @usage_tracker)
    parse_transcription_response(response, model:)
  end
end

#validate_transcription_request(format:, speaker_references:, temperature:) ⇒ Object

Raises:

  • (ArgumentError) 


24
25
26
27
28
 
# File 'lib/ruby_llm/protocols/gemini/file_transcription.rb', line 24

def validate_transcription_request(format:, speaker_references:, temperature:)
  return unless format || speaker_references || temperature

  raise ArgumentError, 'Dedicated transcription does not accept format, speaker references, or temperature'
end