Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Zipformer recipe for CommonVoice #1546

Merged
merged 44 commits into from
Apr 9, 2024
Merged

Conversation

JinZr
Copy link
Collaborator

@JinZr JinZr commented Mar 12, 2024

No description provided.

@JinZr JinZr marked this pull request as ready for review March 13, 2024 03:20
@JinZr JinZr changed the title [WIP] Zipformer recipe for CommonVoice Zipformer recipe for CommonVoice Mar 13, 2024
@JinZr
Copy link
Collaborator Author

JinZr commented Mar 15, 2024

this PR provides support for character-based languages (mozilla lang code: yue, zh-CN, zh-TW, zh-HK, more languages to be supported), integration of word segmentation for cantonese (yue and zh-HK) using pycantonese, as well as a zipformer recipe for commonvoice corpus.

Copy link
Collaborator

@csukuangfj csukuangfj left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks!

Left some minor comments.

egs/commonvoice/ASR/RESULTS.md Outdated Show resolved Hide resolved
egs/commonvoice/ASR/RESULTS.md Outdated Show resolved Hide resolved
parser.add_argument(
"--subset",
type=str,
default="train",
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Could you add a choices here? Otherwise, it is not clear what values are valid.

# limitations under the License.

"""
This script takes a text file "data/lang_char/text" as input, the file consist of
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
This script takes a text file "data/lang_char/text" as input, the file consist of
This script takes a text file "data/lang_char/text" as input; the file consists of

egs/commonvoice/ASR/local/word_segment_yue.py Show resolved Hide resolved
egs/commonvoice/ASR/local/word_segment_yue.py Outdated Show resolved Hide resolved
egs/commonvoice/ASR/local/word_segment_yue.py Outdated Show resolved Hide resolved
egs/commonvoice/ASR/local/word_segment_yue.py Outdated Show resolved Hide resolved
egs/commonvoice/ASR/prepare.sh Show resolved Hide resolved
egs/commonvoice/ASR/pruned_transducer_stateless7/train.py Outdated Show resolved Hide resolved
@JinZr JinZr merged commit f2e36ec into k2-fsa:master Apr 9, 2024
163 checks passed
@JinZr JinZr deleted the dev/cv-zipformer branch April 9, 2024 03:37
@wangjl1993
Copy link

There is a bug in CommonVoice(EN) recipe(prepare.sh: line 342 ):"text" is in "supervisions", so it should be extracted by using jq '.supervisions[].text'
image

@JinZr
Copy link
Collaborator Author

JinZr commented Oct 9, 2024 via email

@JinZr
Copy link
Collaborator Author

JinZr commented Oct 9, 2024 via email

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants