metadata
dataset_info:
features:
- name: file_name
dtype: string
- name: uni
dtype: string
- name: wylie
dtype: string
- name: url
dtype: string
- name: dept
dtype: string
- name: grade
dtype: int64
- name: char_len
dtype: int64
- name: audio_len
dtype: float64
- name: original_id
dtype: string
- name: strata
dtype: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 1983504
num_examples: 3071
- name: validation
num_bytes: 340118
num_examples: 543
download_size: 798221
dataset_size: 2323622
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
Dataset Information
- Total audio duration: 5.6 hours
- Number of audio segments: 3071
- Content type: Teaching material of Garchen Rinpche
- Speaker age group: 70-90 years
Dataset Statistics
Configuration: default
Split: train
Total Rows: 3,071
dept
- Type: categorical
- Data Type:
object - Unique Values: 1
Value Distribution:
| Value | Count | Percentage |
|---|---|---|
STT_GR |
3,071 | 100.00% |
grade
- Type: numerical
- Data Type:
int64 - Sum: 3,071.00
- Average: 1.00
char_len
- Type: numerical
- Data Type:
int64 - Sum: 322,271.00
- Average: 104.94
audio_len
- Type: numerical
- Data Type:
float64 - Sum: 20,215.33
- Average: 6.58
strata
- Type: categorical
- Data Type:
object - Unique Values: 7
Value Distribution:
| Value | Count | Percentage |
|---|---|---|
70-80__long__Teaching |
1,636 | 53.27% |
70-80__medium__Teaching |
773 | 25.17% |
70-80__long__Prayer |
252 | 8.21% |
70-80__long__Q&A |
218 | 7.10% |
70-80__medium__Prayer |
94 | 3.06% |
70-80__medium__Q&A |
54 | 1.76% |
70-80__short__Teaching |
44 | 1.43% |
__index_level_0__
- Type: numerical
- Data Type:
int64 - Sum: 5,562,812.00
- Average: 1,811.40
Split: validation
Total Rows: 543
dept
- Type: categorical
- Data Type:
object - Unique Values: 1
Value Distribution:
| Value | Count | Percentage |
|---|---|---|
STT_GR |
543 | 100.00% |
grade
- Type: numerical
- Data Type:
int64 - Sum: 543.00
- Average: 1.00
char_len
- Type: numerical
- Data Type:
int64 - Sum: 54,420.00
- Average: 100.22
audio_len
- Type: numerical
- Data Type:
float64 - Sum: 3,438.28
- Average: 6.33
strata
- Type: categorical
- Data Type:
object - Unique Values: 7
Value Distribution:
| Value | Count | Percentage |
|---|---|---|
70-80__long__Teaching |
289 | 53.22% |
70-80__medium__Teaching |
136 | 25.05% |
70-80__long__Prayer |
44 | 8.10% |
70-80__long__Q&A |
39 | 7.18% |
70-80__medium__Prayer |
17 | 3.13% |
70-80__medium__Q&A |
10 | 1.84% |
70-80__short__Teaching |
8 | 1.47% |
__index_level_0__
- Type: numerical
- Data Type:
int64 - Sum: 979,093.00
- Average: 1,803.12