4
4
| [ ** Why Haiku?** ] ( #why-haiku )
5
5
| [ ** Quickstart** ] ( #quickstart )
6
6
| [ ** Installation** ] ( #installation )
7
- | [ ** Examples** ] ( https://github.com/deepmind/haiku/tree/master/examples/ )
7
+ | [ ** Examples** ] ( https://github.com/deepmind/dm- haiku/tree/master/examples/ )
8
8
| [ ** User manual** ] ( #user-manual )
9
9
| [ ** Documentation** ] ( https://dm-haiku.readthedocs.io/ )
10
10
| [ ** Citing Haiku** ] ( #citing-haiku )
@@ -16,10 +16,13 @@ Haiku is a simple neural network library for
16
16
authors of [ Sonnet] ( https://github.com/deepmind/sonnet ) , a neural network
17
17
library for [ TensorFlow] ( https://github.com/tensorflow/tensorflow ) .
18
18
19
+ ** Disambiguation:** if you are looking for Haiku the operating system then
20
+ please see https://haiku-os.org/ .
21
+
19
22
NOTE: Haiku is currently ** alpha** . A number of researchers have tested Haiku
20
23
for several months and have reproduced a number of experiments at scale. Please
21
24
feel free to use Haiku, but be sure to test any assumptions and to
22
- [ let us know] ( https://github.com/deepmind/haiku/issues ) if things don't look
25
+ [ let us know] ( https://github.com/deepmind/dm- haiku/issues ) if things don't look
23
26
right!
24
27
25
28
## Overview
@@ -180,9 +183,9 @@ entries in `params` and `grads`. The result has the same structure as the previo
180
183
` params ` and can again be used with ` apply ` .
181
184
182
185
For more, see our
183
- [ examples directory] ( https://github.com/deepmind/haiku/tree/master/examples/ ) .
186
+ [ examples directory] ( https://github.com/deepmind/dm- haiku/tree/master/examples/ ) .
184
187
The
185
- [ MNIST example] ( https://github.com/deepmind/haiku/tree/master/examples/mnist.py )
188
+ [ MNIST example] ( https://github.com/deepmind/dm- haiku/tree/master/examples/mnist.py )
186
189
is a good place to start.
187
190
188
191
## Installation
@@ -198,7 +201,7 @@ to install JAX with the relevant accelerator support.
198
201
Then, install Haiku using pip:
199
202
200
203
``` bash
201
- $ pip install git+https://github.com/deepmind/haiku
204
+ $ pip install git+https://github.com/deepmind/dm- haiku
202
205
```
203
206
204
207
## User manual
@@ -348,7 +351,7 @@ for _ in range(10):
348
351
```
349
352
350
353
For a more complete look at distributed Haiku training, take a look at our
351
- [ ResNet-50 on ImageNet example] ( https://github.com/deepmind/haiku/tree/master/examples/imagenet/ ) .
354
+ [ ResNet-50 on ImageNet example] ( https://github.com/deepmind/dm- haiku/tree/master/examples/imagenet/ ) .
352
355
353
356
## Citing Haiku
354
357
@@ -358,12 +361,12 @@ To cite this repository:
358
361
@software{haiku2020github,
359
362
author = {Tom Hennigan and Trevor Cai and Tamara Norman and Igor Babuschkin},
360
363
title = {{H}aiku: {S}onnet for {JAX}},
361
- url = {https://github.com/deepmind/haiku},
364
+ url = {https://github.com/deepmind/dm- haiku},
362
365
version = {0.0.1a0},
363
366
year = {2020},
364
367
}
365
368
```
366
369
367
370
In this bibtex entry, the version number is intended to be from
368
- [ haiku/__ init__ .py] ( https://github.com/deepmind/haiku/blob/master/haiku/__init__.py ) ,
371
+ [ ` haiku/__init__.py ` ] ( https://github.com/deepmind/dm- haiku/blob/master/haiku/__init__.py ) ,
369
372
and the year corresponds to the project's open-source release.
0 commit comments