From a33852fd7a2e6a38732f48080c43d6aad52bd011 Mon Sep 17 00:00:00 2001 From: pkufool Date: Fri, 20 Aug 2021 09:53:55 +0800 Subject: [PATCH] Add RESULTS.md --- egs/librispeech/ASR/RESULTS.md | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) create mode 100644 egs/librispeech/ASR/RESULTS.md diff --git a/egs/librispeech/ASR/RESULTS.md b/egs/librispeech/ASR/RESULTS.md new file mode 100644 index 000000000..159147a3e --- /dev/null +++ b/egs/librispeech/ASR/RESULTS.md @@ -0,0 +1,23 @@ +## Results + +### LibriSpeech BPE training results (Conformer-CTC) +#### 2021-08-19 +(Wei Kang): Result of https://github.com/k2-fsa/icefall/pull/13 + +TensorBoard log is available at https://tensorboard.dev/experiment/GnRzq8WWQW62dK4bklXBTg/#scalars + +Pretrained model is available at https://huggingface.co/pkufool/conformer_ctc + +The best decoding results (WER) are listed below, we got this results by averaging models from epoch 15 to 34, and using `attention-decoder` decoder with num_paths equals to 100. + +||test-clean|test-other| +|--|--|--| +|WER| 2.57% | 5.94% | + +To get more unique paths, we scaled the lattice.scores with 0.5 (see https://github.com/k2-fsa/icefall/pull/10#discussion_r690951662 for more details), we searched the lm_score_scale and attention_score_scale for best results, the scales that produced the WER above are also listed below. + +||lm_scale|attention_scale| +|--|--|--| +|test-clean|1.3|1.2| +|test-other|1.2|1.1| +