autograd-ts 0.1.1 → 0.1.2
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +20 -21
- package/package.json +1 -1
package/README.md
CHANGED
|
@@ -1,5 +1,4 @@
|
|
|
1
1
|
# autograd-ts
|
|
2
|
-
|
|
3
2
|
A tiny Typescript library that implements **reverse-mode automatic differentiation** (autograd) and neural network training.
|
|
4
3
|
|
|
5
4
|
The goal of the project is exploring the fundamental mechanism that allows neural networks to learn.
|
|
@@ -10,6 +9,25 @@ Producing an output is straightforward. However, learning requires determining h
|
|
|
10
9
|
|
|
11
10
|
This library explores the mechanism that makes that possible.
|
|
12
11
|
|
|
12
|
+
## Usage
|
|
13
|
+
|
|
14
|
+
Install:
|
|
15
|
+
|
|
16
|
+
```bash
|
|
17
|
+
npm install autograd-ts
|
|
18
|
+
```
|
|
19
|
+
|
|
20
|
+
## Import
|
|
21
|
+
```ts
|
|
22
|
+
import {
|
|
23
|
+
Value,
|
|
24
|
+
val,
|
|
25
|
+
MLP,
|
|
26
|
+
mse,
|
|
27
|
+
sgd,
|
|
28
|
+
} from 'autograd-ts';
|
|
29
|
+
```
|
|
30
|
+
|
|
13
31
|
---
|
|
14
32
|
|
|
15
33
|
# Understanding Learning
|
|
@@ -225,28 +243,9 @@ This feedback loop is the mechanism that allows neural networks to learn.
|
|
|
225
243
|
---
|
|
226
244
|
|
|
227
245
|
# Project Scope
|
|
228
|
-
|
|
229
|
-
## Usage
|
|
230
|
-
|
|
231
|
-
Install:
|
|
232
|
-
|
|
233
|
-
```bash
|
|
234
|
-
npm install autograd-ts
|
|
235
|
-
```
|
|
236
|
-
|
|
237
|
-
## Import
|
|
238
|
-
```ts
|
|
239
|
-
import {
|
|
240
|
-
Value,
|
|
241
|
-
val,
|
|
242
|
-
MLP,
|
|
243
|
-
mse,
|
|
244
|
-
sgd,
|
|
245
|
-
} from 'autograd-ts';
|
|
246
|
-
```
|
|
246
|
+
Version 0.1.0 focuses on a minimal learning system.
|
|
247
247
|
|
|
248
248
|
## Core Autograd Engine
|
|
249
|
-
Version 0.1.0 focuses on a minimal learning system.
|
|
250
249
|
|
|
251
250
|
* Value
|
|
252
251
|
* add
|