PerSegmentComparison
Reference against translation for every segment, worst ones highlighted.
The corpus level BLEU, CHRF and TER metrics condense a whole split into one number. This table is the split itself: it pairs each reference with its translation and a sentence level score, and highlights the lowest scoring rows so the segments dragging the aggregate down are the first thing seen. A concentration of low scores in long or short segments, or around a particular topic, is exactly what the average cannot say.
Parameters
- highlight_count : integer, default=
5 - How many of the lowest scoring segments to highlight.
Methods
compute(self, y_true, y_pred, class_names: Optional[List[str]] = None) -> List[DashAI.back.core.artifacts.Artifact]
PerSegmentComparisonBuild the per segment comparison table.
Parameters
- y_true : array_like
- Reference translations, one per row.
- y_pred : array_like
- The model's translations for the same rows.
- class_names : Optional[List[str]]
- Unused; always None for translation.
Returns
- List[Artifact]
- A single table, one row per segment plus the average row.
get_credential(self, name: str)
ConfigObjectResolve a registered credential component by name.
Parameters
- name : str
- Credential component class name (e.g. "HuggingFaceCredential").
Returns
- BaseCredential
- An instance of the requested credential component.
get_metadata(cls) -> Dict[str, Any]
BaseReportGet metadata values for the current report.
Returns
- Dict[str, Any]
- UI metadata, including whether the report needs a model that outputs class probabilities.
get_schema(cls) -> dict
ConfigObjectGenerates the component related Json Schema.
Returns
- dict
- Dictionary representing the Json Schema of the component.
validate_and_transform(self, raw_data: dict) -> dict
ConfigObjectIt takes the data given by the user to initialize the model and returns it with all the objects that the model needs to work.
Parameters
- raw_data : dict
- A dictionary with the data provided by the user to initialize the model.
Returns
- dict
- A validated dictionary with the necessary objects.