Class: Cohere::Transcribe::Runtime::WordPipeline::PairBudgetReload

Inherits:
Object
  • Object
show all
Defined in:
lib/cohere/transcribe/runtime/word_pipeline.rb

Overview

Reloads alignment PCM with one worker and at most one lookahead. The next file may overlap MMS computation only when the current/next PCM pair fits the configured budget; otherwise the current reference is dropped before the next decode starts.

Constant Summary collapse

SAMPLE_BYTES = 
4

Instance Method Summary collapse

Constructor Details

#initialize(works, memory_byte_limit:, reload:, started: nil) ⇒ PairBudgetReload

Returns a new instance of PairBudgetReload.

Raises:

  • (ArgumentError) 


230
231
232
233
234
235
236
237
238
239
 
# File 'lib/cohere/transcribe/runtime/word_pipeline.rb', line 230

def initialize(works, memory_byte_limit:, reload:, started: nil)
  @works = works.to_a.freeze
  @memory_byte_limit = Integer(memory_byte_limit)
  raise ArgumentError, "memory_byte_limit must be positive" unless @memory_byte_limit.positive?
  raise ArgumentError, "reload must respond to call" unless reload.respond_to?(:call)
  raise ArgumentError, "started must respond to call" if started && !started.respond_to?(:call)

  @reload = reload
  @started = started
end

Instance Method Details

#each ⇒ Object 



241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
 
# File 'lib/cohere/transcribe/runtime/word_pipeline.rb', line 241

def each
  return enum_for(__method__) unless block_given?
  return if @works.empty?

  pending = submit(@works.first, 0)
  @started&.call
  @works.each_with_index do |work, index|
    current = pending
    pending = nil
    reload_result = current.value
    current = nil # rubocop:disable Lint/UselessAssignment -- release Thread#value before a serial decode

    next_work = @works[index + 1]
    pending = submit(next_work, index + 1) if next_work && pair_fits?(work, next_work)
    yield [work, reload_result].freeze
    reload_result = nil # rubocop:disable Lint/UselessAssignment -- release PCM before a serial decode
    unless pending || next_work.nil?
      GC.start(full_mark: false, immediate_mark: true, immediate_sweep: true)
      pending = submit(next_work, index + 1)
    end
  end
  completed = true
ensure
  cancel(pending) if defined?(pending) && pending
  cancel(current) if defined?(current) && current
  GC.start(full_mark: false, immediate_mark: true, immediate_sweep: true) if defined?(completed) && completed
end