Comments (2)
Hi there, instance level mixed precision with jmp is not built into Haiku, however you can implement this in a few ways. I've added them in this colab notebook: https://colab.research.google.com/gist/tomhennigan/874de0420c55f7bd062f24f7ec6b0e51/instance-level-half-precision.ipynb
Approximately the options are:
- Create a subclass and use that for instances you want in a different precision.
class LowPrecisionLinear(hk.Linear):
pass
half_policy = jmp.get_policy('compute=half')
hk.mixed_precision.set_policy(LowPrecisionLinear, half_policy)
def f(x):
net = hk.Sequential([
hk.Linear(300), jnp.tan,
hk.Linear(100), jnp.tan,
LowPrecisionLinear(10),
])
return net(x)
- Create a wrapper function that applies a specific policy before calling the call method.
def wrap_with_policy(mod: hk.Module, policy: jmp.Policy):
cls = type(mod)
@functools.wraps(mod.__call__)
def wrapper(*args, **kwargs):
old_policy = hk.mixed_precision.get_policy(hk.Linear)
hk.mixed_precision.set_policy(cls, policy)
try:
return mod(*args, **kwargs)
finally:
if old_policy is not None:
hk.mixed_precision.set_policy(cls, old_policy)
else:
hk.mixed_precision.clear_policy(cls)
return wrapper
half_policy = jmp.get_policy('compute=half')
def f(x):
net = hk.Sequential([
hk.Linear(300), jnp.tan,
hk.Linear(100), jnp.tan,
wrap_with_policy(hk.Linear(10), half_policy),
])
return net(x)
from dm-haiku.
Thank you for the suggested options. This is of great help to me.
from dm-haiku.
Related Issues (20)
- Suggestion: alias `Transformed`(WithState) apply to __call__ HOT 2
- Is there a way to load parameters from Flax model? HOT 2
- Support model examples HOT 7
- Change to jax.interpreters.xla for JAX==0.4.14 HOT 3
- Warning: hk.LayerNorm when used in transformer decoder causes violation of autoregressive property HOT 1
- Reservoir Computing with Haiku
- Efficiency difference in using jax.lax.fori_loop vs looping over identical layers? HOT 2
- Please publish requirements.txt fix to pip
- How to use `apply` with additional parameters? HOT 1
- hk.Conv2DTranspose takes FOREVER to initialize and compile HOT 1
- 0.4.16 timeline HOT 2
- How to export haiku network parameters into Pytorch network?
- Modules got silently "reused" with `hk.vmap` HOT 2
- Wrong gradients in a Haiku network
- Direct Feedback Alignment
- Issue with wheels including docs and examples folder
- `haiku.experimental.flax` is not part of newest pip release HOT 1
- Train multiple hk.nets.MLP with one optimizer HOT 2
- TypeError: 'type' object is not subscriptable HOT 4
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from dm-haiku.