src/speech_decoding/models/v14_converged_v3/objective.py

Selected training excerpt · lines 500–502 · historical width comment corrected to the released width · source SHA-256 5718458b537d

500        pred_in = scatter_visible(501            self.enc_to_pred(z), pack.idx, grid.total, self.mask_token502        )  # (B, total, 192)