Fangjun Kuang 63563d16d3
Fix setting joiner dim (#2027)
Fixes incorrect computation of encoder_dim when encoder_dim is a comma-separated list of integers by ensuring numeric (not lexicographic) max is used.

Fixes #2018

- Replace int(max(params.encoder_dim.split(","))) (lexicographic max on strings) with max(_to_int_tuple(params.encoder_dim)) (numeric max).
- Apply the fix consistently across all affected training scripts.
2025-09-19 09:42:41 +08:00
..
2022-12-13 16:13:26 +08:00
2025-09-19 09:42:41 +08:00
2023-10-25 00:03:33 +08:00
2022-03-18 11:39:06 +08:00

Introduction

This recipe includes some different ASR models trained with TedLium3.

Transducers

There are various folders containing the name transducer in this folder. The following table lists the differences among them.

Encoder Decoder Comment
transducer_stateless Conformer Embedding + Conv1d
pruned_transducer_stateless Conformer Embedding + Conv1d Using k2 pruned RNN-T loss

The decoder in transducer_stateless is modified from the paper Rnn-Transducer with Stateless Prediction Network. We place an additional Conv1d layer right after the input embedding layer.