init
Browse files- README.md +0 -42
- subset_test/train-00000-of-00001.parquet +0 -3
- test_s2s.sh +0 -17
README.md
CHANGED
@@ -1,42 +0,0 @@
|
|
1 |
-
---
|
2 |
-
dataset_info:
|
3 |
-
config_name: subset_test
|
4 |
-
features:
|
5 |
-
- name: jaA.audio
|
6 |
-
dtype: audio
|
7 |
-
- name: enA.audio
|
8 |
-
dtype: audio
|
9 |
-
- name: line_no
|
10 |
-
dtype: int64
|
11 |
-
- name: enA.id
|
12 |
-
dtype: string
|
13 |
-
- name: enA.url
|
14 |
-
dtype: string
|
15 |
-
- name: enA.duration_start
|
16 |
-
dtype: int64
|
17 |
-
- name: enA.duration_end
|
18 |
-
dtype: int64
|
19 |
-
- name: enA.laser_score
|
20 |
-
dtype: float64
|
21 |
-
- name: jaA.id
|
22 |
-
dtype: string
|
23 |
-
- name: jaA.url
|
24 |
-
dtype: string
|
25 |
-
- name: jaA.duration_start
|
26 |
-
dtype: int64
|
27 |
-
- name: jaA.duration_end
|
28 |
-
dtype: int64
|
29 |
-
- name: jaA.laser_score
|
30 |
-
dtype: float64
|
31 |
-
splits:
|
32 |
-
- name: train
|
33 |
-
num_bytes: 1530085.0
|
34 |
-
num_examples: 8
|
35 |
-
download_size: 1506623
|
36 |
-
dataset_size: 1530085.0
|
37 |
-
configs:
|
38 |
-
- config_name: subset_test
|
39 |
-
data_files:
|
40 |
-
- split: train
|
41 |
-
path: subset_test/train-*
|
42 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
subset_test/train-00000-of-00001.parquet
DELETED
@@ -1,3 +0,0 @@
|
|
1 |
-
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:7379569d017aa9190e7db157eb329b306ea2fd4ab511c85e8a8f904f97ea93cb
|
3 |
-
size 1506623
|
|
|
|
|
|
|
|
test_s2s.sh
DELETED
@@ -1,17 +0,0 @@
|
|
1 |
-
# audio
|
2 |
-
export DATASET_ID=test
|
3 |
-
export DIRECTION="enA-jaA"
|
4 |
-
export LINE_NO_START=0
|
5 |
-
export LINE_NO_END=10
|
6 |
-
export HF_DATASET="experiment-process-seamless-align"
|
7 |
-
python fetch_dataset_s2s.py
|
8 |
-
# tokenize
|
9 |
-
export DATASET_ID=test_tokenized
|
10 |
-
export DIRECTION="enA-jaA"
|
11 |
-
export HF_DATASET="experiment-process-seamless-align"
|
12 |
-
python tokenize_dataset_s2s.py
|
13 |
-
# speaker embedding
|
14 |
-
export DATASET_ID=test
|
15 |
-
export DIRECTION="enA-jaA"
|
16 |
-
export HF_DATASET="experiment-process-seamless-align"
|
17 |
-
python attach_speaker_embedding_s2s.py
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|