Comments (2)
We may want to prefer this syntax (in examples and docs) instead of the current keyword arguments in the constructor.
For example:
class MyNetwork(nn.Container):
def __init__(self):
super(nn.Container, self).__init__()
self.l1 = nn.Linear(5, 10)
self.l2 = nn.Linear(10, 20)
instead of:
class MyNetwork(nn.Container):
def __init__(self):
super(nn.Container, self).__init__(
l1=nn.Linear(5, 10),
l2=nn.Linear(10, 20),
)
One advantage is that the order of module iteration would match the order of assignment, which is not true for the constructor syntax (except in the upcoming Python 3.6)
from pytorch.
@colesbury good point. We should probably do the same for parameters
from pytorch.
Related Issues (20)
- Matmul with broadcast makes copy because of reshape to bmm, and it is slow
- Race in ProcessGroupNCCL shutdown HOT 2
- DISABLED test_mark_non_differentiable (__main__.TestAutogradWithCompiledAutograd) HOT 3
- Dataloader codeowner
- DISABLED test_creation_with_zeros_cuda_float8_e5m2 (__main__.TestFloat8DtypeCUDA) HOT 1
- DISABLED test_mark_non_differentiable_none (__main__.TestAutogradWithCompiledAutograd) HOT 2
- DISABLED test_aot_module_simplified (__main__.TestAOTModuleSimplified) HOT 2
- Stable Diffusion Model Error: torch._dynamo.exc.InternalTorchDynamoError: raw
- Fused Linear and Cross-Entropy Loss `torch.nn.functional.linear_cross_entropy` HOT 4
- HOP dispatch isn't faithful
- PyTorch Memory Management in GPU-to-CPU Transfers issue
- DISABLED test_multi_backward (__main__.TestAutogradWithCompiledAutograd) HOT 2
- torch.compile does not work since 2.2.1 on MacOS for some models
- inductor::_reinterpret_tensor() Expected a value of type 'List[int]' for argument 'size' but instead found type 'tuple'
- Verbose log: [__aot_joint_graph] could not reconstruct view by re-applying a ViewMeta sequence. HOT 2
- compiling profiler with ExecutionTraceObserver breaks HOT 1
- [export] Llama3 export with dynamic shapes fails with constraint violations HOT 5
- [feature request] allow torch.compile calls with compiler options, but without modifying global dynamo options
- Option to hard fail for cudagraphs and/or more cudagraph logging HOT 1
- torch.export of ResNet with dynamic height fails due to constraint violation HOT 2
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 pytorch.