From 0efc87352ce7d3903dbdedbfa5d01ca5e1cb19e7 Mon Sep 17 00:00:00 2001
From: Shi Xian <40013335+R1ckShi@users.noreply.github.com>
Date: 星期四, 05 十二月 2024 15:15:38 +0800
Subject: [PATCH] Merge pull request #2267 from modelscope/dev_sx2
---
examples/aishell/e_branchformer/conf/e_branchformer_12e_6d_2048_256.yaml | 12 +++++++-----
1 files changed, 7 insertions(+), 5 deletions(-)
diff --git a/examples/aishell/e_branchformer/conf/e_branchformer_12e_6d_2048_256.yaml b/examples/aishell/e_branchformer/conf/e_branchformer_12e_6d_2048_256.yaml
index dea580e..6438ae1 100644
--- a/examples/aishell/e_branchformer/conf/e_branchformer_12e_6d_2048_256.yaml
+++ b/examples/aishell/e_branchformer/conf/e_branchformer_12e_6d_2048_256.yaml
@@ -94,18 +94,20 @@
dataset: AudioDataset
dataset_conf:
index_ds: IndexDSJsonl
- batch_sampler: DynamicBatchLocalShuffleSampler
- batch_type: example # example or length
- batch_size: 1 # if batch_type is example, batch_size is the numbers of samples; if length, batch_size is source_token_len+target_token_len;
+ batch_sampler: EspnetStyleBatchSampler
+ batch_type: length # example or length
+ batch_size: 10000 # if batch_type is example, batch_size is the numbers of samples; if length, batch_size is source_token_len+target_token_len;
max_token_length: 2048 # filter samples if source_token_len+target_token_len > max_token_length,
- buffer_size: 500
+ buffer_size: 1024
shuffle: True
num_workers: 4
+ preprocessor_speech: SpeechPreprocessSpeedPerturb
+ preprocessor_speech_conf:
+ speed_perturb: [0.9, 1.0, 1.1]
tokenizer: CharTokenizer
tokenizer_conf:
unk_symbol: <unk>
- split_with_space: true
ctc_conf:
--
Gitblit v1.9.1