MCPcopy Create free account
hub / github.com/NVIDIA/TensorRT / question_features

Function question_features

demo/BERT/inference_c.py:118–120  ·  view source on GitHub ↗
(tokens, question)

Source from the content-addressed store, hash-verified

116 max_seq_length = args.sequence_length
117
118 def question_features(tokens, question):
119 # Extract features from the paragraph and question
120 return dp.convert_example_to_features(tokens, question, tokenizer, max_seq_length, doc_stride, args.max_query_length)
121
122 # The first context created will use the 0th profile. A new context must be created
123 # for each additional profile needed. Here, we only use batch size 1, thus we only need the first profile.

Callers 1

inference_c.pyFile · 0.70

Calls

no outgoing calls

Tested by

no test coverage detected