Class: Informers::QuestionAnsweringPipeline

Inherits:
Pipeline
  • Object
show all
Defined in:
lib/informers/pipelines.rb

Instance Method Summary collapse

Constructor Details

#initialize(**options) ⇒ QuestionAnsweringPipeline

Returns a new instance of QuestionAnsweringPipeline.



203
204
205
 
# File 'lib/informers/pipelines.rb', line 203

def initialize(**options)
  super(**options)
end

Instance Method Details

#call(question, context, top_k: 1) ⇒ Object 



207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
 
# File 'lib/informers/pipelines.rb', line 207

def call(question, context, top_k: 1)
  # Run tokenization
  inputs = @tokenizer.(question,
    text_pair: context,
    padding: true,
    truncation: true,
    return_offsets: true
  )

  output = @model.(inputs)

  to_return = []
  output.start_logits.length.times do |j|
    ids = inputs[:input_ids][j]
    sep_index = ids.index(@tokenizer.sep_token_id)
    offsets = inputs[:offsets][j]

    s1 = Utils.softmax(output.start_logits[j])
      .map.with_index
      .select { |x| x[1] > sep_index }
    e1 = Utils.softmax(output.end_logits[j])
      .map.with_index
      .select { |x| x[1] > sep_index }

    options = s1.product(e1)
      .select { |x| x[0][1] <= x[1][1] }
      .map { |x| [x[0][1], x[1][1], x[0][0] * x[1][0]] }
      .sort_by { |v| -v[2] }

    [options.length, top_k].min.times do |k|
      start, end_, score = options[k]

      answer_tokens = ids.slice(start, end_ + 1)

      answer = @tokenizer.decode(answer_tokens,
        skip_special_tokens: true
      )

      to_return << {
        answer:,
        score:,
        start: offsets[start][0],
        end: offsets[end_][1]
      }
    end
  end

  question.is_a?(Array) ? to_return : to_return[0]
end