@@ -91,7 +91,7 @@ <h3>SpeechBrain Basics</h3>
91
91
< a class ="active " href ="# "> SpeechBrain Basics</ a >
92
92
</ div >
93
93
< ul class ="blog_meta list ">
94
- < li > < a href ="about.html "> Plantiga P.< i class ="lnr lnr-user "> </ i > </ a > </ li >
94
+ < li > < a href ="about.html "> Plantinga P.< i class ="lnr lnr-user "> </ i > </ a > </ li >
95
95
< li > < a href ="# "> Jan. 2021< i class ="lnr lnr-calendar-full "> </ i > </ a > </ li >
96
96
< li > < a href ="# "> Difficulty: easy< i class ="lnr lnr-cog "> </ i > </ a > </ li >
97
97
< li > < a href ="# "> Time: 10min< i class ="lnr lnr-hourglass "> </ i > </ a > </ li >
@@ -101,7 +101,7 @@ <h3>SpeechBrain Basics</h3>
101
101
< div class ="col-md-9 ">
102
102
< div class ="blog_post ">
103
103
< div class ="blog_details ">
104
- < h2 > The Brain Class</ h2 >
104
+ < h2 > Brain Class</ h2 >
105
105
< p > One key component of deep learning is iterating the dataset multiple times and performing parameter updates.
106
106
This process is sometimes called the "training loop" and there are usually many stages to this loop.
107
107
SpeechBrain provides a convenient framework for organizing the training loop, in the form of a class known as the "Brain" class,
@@ -119,7 +119,7 @@ <h2>The Brain Class</h2>
119
119
< a class ="active " href ="# "> SpeechBrain Basics</ a >
120
120
</ div >
121
121
< ul class ="blog_meta list ">
122
- < li > < a href ="about.html "> Plantiga P.< i class ="lnr lnr-user "> </ i > </ a > </ li >
122
+ < li > < a href ="about.html "> Plantinga P.< i class ="lnr lnr-user "> </ i > </ a > </ li >
123
123
< li > < a href ="# "> Jan. 2021< i class ="lnr lnr-calendar-full "> </ i > </ a > </ li >
124
124
< li > < a href ="# "> Difficulty: easy< i class ="lnr lnr-cog "> </ i > </ a > </ li >
125
125
< li > < a href ="# "> Time: 15min< i class ="lnr lnr-hourglass "> </ i > </ a > </ li >
@@ -143,6 +143,33 @@ <h2>HyperPyYAML</h2>
143
143
</ div >
144
144
</ div >
145
145
</ article >
146
+ < article class ="row blog_item ">
147
+ < div class ="col-md-3 ">
148
+ < div class ="blog_info text-right ">
149
+ < div class ="post_tag ">
150
+ < a class ="active " href ="# "> SpeechBrain Basics</ a >
151
+ </ div >
152
+ < ul class ="blog_meta list ">
153
+ < li > < a href ="about.html "> Cornell S. & Rouhe A.< i class ="lnr lnr-user "> </ i > </ a > </ li >
154
+ < li > < a href ="# "> Jan. 2021< i class ="lnr lnr-calendar-full "> </ i > </ a > </ li >
155
+ < li > < a href ="# "> Difficulty: medium< i class ="lnr lnr-cog "> </ i > </ a > </ li >
156
+ < li > < a href ="# "> Time: 20min< i class ="lnr lnr-hourglass "> </ i > </ a > </ li >
157
+ </ ul >
158
+ </ div >
159
+ </ div >
160
+ < div class ="col-md-9 ">
161
+ < div class ="blog_post ">
162
+ < div class ="blog_details ">
163
+ < h2 > Data Loading Pipeline</ h2 >
164
+ < p > Setting up an efficient data loading pipeline is often a tedious task which involves creating the examples,
165
+ defining your torch.utils.data.Dataset class as well as different data sampling and augmentations strategies.
166
+ In SpeechBrain we provide efficient abstractions to simplify this time-consuming process without sacrificing
167
+ flexibility. In fact our data pipeline is built around the Pytorch one.</ p >
168
+ < a href ="https://colab.research.google.com/drive/1NEboTfb2EIBrc0nUd9NKwcG2eqf-kv3d?usp=sharing " class ="blog_btn "> Open in Google Colab</ a >
169
+ </ div >
170
+ </ div >
171
+ </ div >
172
+ </ article >
146
173
147
174
< nav class ="blog-pagination justify-content-center d-flex ">
148
175
< ul class ="pagination ">
0 commit comments