From 184d30dd9cc607e800e88a331d270b6f23dba7d3 Mon Sep 17 00:00:00 2001 From: Hui Zhang Date: Tue, 7 Sep 2021 07:20:49 +0000 Subject: [PATCH] relase librispeech audio max len to 30 second --- deepspeech/io/dataloader.py | 2 +- examples/librispeech/s0/conf/deepspeech2.yaml | 2 +- examples/librispeech/s0/conf/deepspeech2_online.yaml | 2 +- examples/librispeech/s2/conf/transformer.yaml | 2 +- examples/tiny/s0/conf/deepspeech2.yaml | 2 +- examples/tiny/s1/conf/chunk_confermer.yaml | 2 +- 6 files changed, 6 insertions(+), 6 deletions(-) diff --git a/deepspeech/io/dataloader.py b/deepspeech/io/dataloader.py index a35a0bc0..310f5f58 100644 --- a/deepspeech/io/dataloader.py +++ b/deepspeech/io/dataloader.py @@ -44,7 +44,7 @@ def feat_dim_and_vocab_size(data_json: List[Dict[Text, Any]], def batch_collate(x): - """de-tuple. + """de-minibatch, since user compose batch. Args: x (List[Tuple]): [(utts, xs, ilens, ys, olens)] diff --git a/examples/librispeech/s0/conf/deepspeech2.yaml b/examples/librispeech/s0/conf/deepspeech2.yaml index 031d684d..2c31e66e 100644 --- a/examples/librispeech/s0/conf/deepspeech2.yaml +++ b/examples/librispeech/s0/conf/deepspeech2.yaml @@ -4,7 +4,7 @@ data: dev_manifest: data/manifest.dev-clean test_manifest: data/manifest.test-clean min_input_len: 0.0 - max_input_len: 27.0 # second + max_input_len: 30.0 # second min_output_len: 0.0 max_output_len: .inf min_output_input_ratio: 0.00 diff --git a/examples/librispeech/s0/conf/deepspeech2_online.yaml b/examples/librispeech/s0/conf/deepspeech2_online.yaml index 27f59f3f..87445c0b 100644 --- a/examples/librispeech/s0/conf/deepspeech2_online.yaml +++ b/examples/librispeech/s0/conf/deepspeech2_online.yaml @@ -4,7 +4,7 @@ data: dev_manifest: data/manifest.dev-clean test_manifest: data/manifest.test-clean min_input_len: 0.0 - max_input_len: 27.0 # second + max_input_len: 30.0 # second min_output_len: 0.0 max_output_len: .inf min_output_input_ratio: 0.00 diff --git a/examples/librispeech/s2/conf/transformer.yaml b/examples/librispeech/s2/conf/transformer.yaml index f7c27d1f..4c60913e 100644 --- a/examples/librispeech/s2/conf/transformer.yaml +++ b/examples/librispeech/s2/conf/transformer.yaml @@ -22,7 +22,7 @@ collator: batch_frames_out: 0 batch_frames_inout: 0 augmentation_config: conf/augmentation.json - num_workers: 2 + num_workers: 0 subsampling_factor: 1 num_encs: 1 diff --git a/examples/tiny/s0/conf/deepspeech2.yaml b/examples/tiny/s0/conf/deepspeech2.yaml index aeb33f58..c93217d3 100644 --- a/examples/tiny/s0/conf/deepspeech2.yaml +++ b/examples/tiny/s0/conf/deepspeech2.yaml @@ -4,7 +4,7 @@ data: dev_manifest: data/manifest.tiny test_manifest: data/manifest.tiny min_input_len: 0.0 - max_input_len: 27.0 + max_input_len: 30.0 min_output_len: 0.0 max_output_len: 400.0 min_output_input_ratio: 0.05 diff --git a/examples/tiny/s1/conf/chunk_confermer.yaml b/examples/tiny/s1/conf/chunk_confermer.yaml index 1b701aa2..96da3d9f 100644 --- a/examples/tiny/s1/conf/chunk_confermer.yaml +++ b/examples/tiny/s1/conf/chunk_confermer.yaml @@ -4,7 +4,7 @@ data: dev_manifest: data/manifest.tiny test_manifest: data/manifest.tiny min_input_len: 0.5 # second - max_input_len: 20.0 # second + max_input_len: 30.0 # second min_output_len: 0.0 # tokens max_output_len: 400.0 # tokens min_output_input_ratio: 0.05