Merge branch 'dev_multi_zh-hans' of https://github.com/JinZr/icefall into dev_multi_zh-hans

This commit is contained in:
zr_jin 2023-09-02 17:00:56 +08:00
commit 28a361795f
3 changed files with 6 additions and 12 deletions

View File

@ -43,7 +43,7 @@ torch.set_num_threads(1)
torch.set_num_interop_threads(1)
def compute_fbank_thchs30(num_mel_bins: int = 80, speed_perturb: bool = False):
def compute_fbank_magicdata(num_mel_bins: int = 80, speed_perturb: bool = False):
src_dir = Path("data/manifests/magicdata")
output_dir = Path("data/fbank")
num_jobs = min(30, os.cpu_count())
@ -78,11 +78,9 @@ def compute_fbank_thchs30(num_mel_bins: int = 80, speed_perturb: bool = False):
recordings=m["recordings"],
supervisions=m["supervisions"],
)
if "train" in partition:
if "train" in partition and speed_perturb:
cut_set = (
(cut_set + cut_set.perturb_speed(0.9) + cut_set.perturb_speed(1.1))
if speed_perturb
else cut_set
)
cut_set = cut_set.compute_and_store_features(
extractor=extractor,

View File

@ -43,7 +43,7 @@ torch.set_num_threads(1)
torch.set_num_interop_threads(1)
def compute_fbank_thchs30(num_mel_bins: int = 80, speed_perturb: bool = False):
def compute_fbank_primewords(num_mel_bins: int = 80, speed_perturb: bool = False):
src_dir = Path("data/manifests/primewords")
output_dir = Path("data/fbank")
num_jobs = min(15, os.cpu_count())
@ -78,11 +78,9 @@ def compute_fbank_thchs30(num_mel_bins: int = 80, speed_perturb: bool = False):
recordings=m["recordings"],
supervisions=m["supervisions"],
)
if "train" in partition:
if "train" in partition and speed_perturb:
cut_set = (
(cut_set + cut_set.perturb_speed(0.9) + cut_set.perturb_speed(1.1))
if speed_perturb
else cut_set
)
cut_set = cut_set.compute_and_store_features(
extractor=extractor,

View File

@ -43,7 +43,7 @@ torch.set_num_threads(1)
torch.set_num_interop_threads(1)
def compute_fbank_thchs30(num_mel_bins: int = 80, speed_perturb: bool = False):
def compute_fbank_stcmds(num_mel_bins: int = 80, speed_perturb: bool = False):
src_dir = Path("data/manifests/stcmds")
output_dir = Path("data/fbank")
num_jobs = min(15, os.cpu_count())
@ -78,11 +78,9 @@ def compute_fbank_thchs30(num_mel_bins: int = 80, speed_perturb: bool = False):
recordings=m["recordings"],
supervisions=m["supervisions"],
)
if "train" in partition:
if "train" in partition and speed_perturb:
cut_set = (
(cut_set + cut_set.perturb_speed(0.9) + cut_set.perturb_speed(1.1))
if speed_perturb
else cut_set
)
cut_set = cut_set.compute_and_store_features(
extractor=extractor,