From b37e7b15e02fe8ac42c6591650dc2bfe86b1e8ac Mon Sep 17 00:00:00 2001
From: 游雁 <zhifu.gzf@alibaba-inc.com>
Date: 星期三, 22 五月 2024 15:11:08 +0800
Subject: [PATCH] wenetspeech
---
funasr/bin/compute_audio_cmvn.py | 10 ++++++----
1 files changed, 6 insertions(+), 4 deletions(-)
diff --git a/funasr/bin/compute_audio_cmvn.py b/funasr/bin/compute_audio_cmvn.py
index 91bb2ac..697da94 100644
--- a/funasr/bin/compute_audio_cmvn.py
+++ b/funasr/bin/compute_audio_cmvn.py
@@ -52,7 +52,7 @@
frontend=frontend,
tokenizer=None,
is_training=False,
- **kwargs.get("dataset_conf")
+ **kwargs.get("dataset_conf"),
)
# dataloader
@@ -68,11 +68,13 @@
dataset_train, collate_fn=dataset_train.collator, **batch_sampler_train
)
- iter_stop = int(kwargs.get("scale", 1.0) * len(dataloader_train))
-
total_frames = 0
for batch_idx, batch in enumerate(dataloader_train):
- if batch_idx >= iter_stop:
+ iter_stop = int(kwargs.get("scale", -1.0) * len(dataloader_train))
+ log_step = iter_stop // 100
+ if batch_idx % log_step == 0:
+ print(f"prcessed: {batch_idx}/{iter_stop}")
+ if batch_idx >= iter_stop and iter_stop > 0.0:
break
fbank = batch["speech"].numpy()[0, :, :]
--
Gitblit v1.9.1