The checkpoint contains the FTA inference modules under encoder.*, ctc_head.*, blank_gate.*, and blank_prior.*. It is an icefall checkpoint, not a Transformers AutoModel.from_pretrained() package.

Loading

python

import torch

checkpoint = torch.load(
    "pretrained-epoch40-avg10.pt",
    map_location="cpu",
    weights_only=False,
)
state_dict = checkpoint["model"]

Use the model definition and decoding pipeline in conformer_ctc2/decode_vi_blank_v2.py from the accompanying icefall codebase.

Notes

FTA-specific training components require the matching model implementation. For reproducible decoding, use the same token inventory, SentencePiece model, and decoding configuration as the original experiment.

NYCU-MLLab/learning_frame_token_responsibilities_via_optimal_transport

FTA Conformer CTC — LibriSpeech 100h

Checkpoint

Loading

Notes

NYCU-MLLab/
learning_frame_token_responsibilities_via_optimal_transport