MCPcopy Create free account
hub / github.com/zai-org/CodeGeeX / process_sample

Method process_sample

codegeex/data/processor.py:128–141  ·  view source on GitHub ↗

Process a sample.

(self, sample: LabelSample)

Source from the content-addressed store, hash-verified

126 "labels": label
127 }
128 def process_sample(self, sample: LabelSample) -> Iterable[Dict[str, List[int]]]:
129 """
130 Process a sample.
131 """
132 prompt_tokens = self._tokenize(sample.prompt)
133 label = sample.label
134
135
136 if len(prompt_tokens) > self._max_seq_len:
137 if self._discard_overlong:
138 return
139 prompt_tokens=prompt_tokens[-self._max_seq_len:]
140
141 yield self.pad_seq(prompt_tokens, label, extra=sample.extra)
142
143 def process_sample_strict(self, sample: LabelSample) -> List[Dict[str, List[int]]]:
144 """

Callers 1

process_sample_strictMethod · 0.95

Calls 1

pad_seqMethod · 0.95

Tested by

no test coverage detected