impresso-project
/

ocr-quality-assessor-unigram-light

Token Classification

ocr_qa_assessment

quality-assessment

Model card Files Files and versions Community

emanuelaboros commited on 16 days ago

Commit

70b421f

·

1 Parent(s): 3b1b05f

moved to language-identifier

Files changed (1) hide show

modeling_ocrqa.py +1 -1

modeling_ocrqa.py CHANGED Viewed

@@ -96,7 +96,7 @@ class QAAssessmentModel(PreTrainedModel):
         print(self.ocrqa_assessors)
         self.lang_pipeline = pipeline("langident",
-                                    model="impresso-project/impresso-langident",
                                     trust_remote_code=True,
                                     device="cpu")

         print(self.ocrqa_assessors)
         self.lang_pipeline = pipeline("langident",
+                                    model="impresso-project/language-identifier",
                                     trust_remote_code=True,
                                     device="cpu")