Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / google/neural-tangents issues and pull requests

#88 - NaN occurs when backprop through the kernel including ReLu layer

Issue - State: closed - Opened by ZuowenWang0000 almost 4 years ago - 2 comments
Labels: bug

#87 - kernel function decorated with batch doesn't return the right shape

Issue - State: closed - Opened by qixuanf almost 4 years ago - 5 comments
Labels: bug

#86 - sparse support for stax.Aggregate function

Issue - State: open - Opened by jl626 almost 4 years ago - 4 comments
Labels: enhancement

#85 - RuntimeError: CUDA operation failed: cudaGetErrorString symbol not found

Issue - State: closed - Opened by kventinel almost 4 years ago - 4 comments
Labels: bug

#84 - Shared weights in Neural Nets

Issue - State: open - Opened by jiahai-feng about 4 years ago - 4 comments
Labels: enhancement

#83 - Question about the dimension of NTK

Issue - State: closed - Opened by chenwydj about 4 years ago - 2 comments
Labels: question

#82 - Option `do_stabilize=True` leads to TypeError in kernel execution

Issue - State: closed - Opened by White-Link about 4 years ago - 1 comment
Labels: bug

#81 - Calculating Marginal log-likelihood of NNGP

Issue - State: closed - Opened by TZeng20 about 4 years ago - 7 comments
Labels: question

#80 - How to mask NN layer by nt.stax.DotGeneral?

Issue - State: open - Opened by Kangfei about 4 years ago - 3 comments
Labels: question

#79 - Is there any way to use ``neural_tangents.predict`` library with ``stax.Aggregate`` function?

Issue - State: closed - Opened by jl626 about 4 years ago - 6 comments
Labels: bug

#73 - gradient of kernel_fn with autograd

Issue - State: closed - Opened by emited about 4 years ago - 4 comments
Labels: question

#72 - Difference between "gradient_descent_mse" and "gradient_descent_mse_ensemble"

Issue - State: closed - Opened by lionelmessi6410 about 4 years ago - 5 comments
Labels: question

#70 - when the dimension of ntk_train_train exceed 32767, gradient_descent_mse returns error result.

Issue - State: closed - Opened by ghost about 4 years ago - 4 comments
Labels: bug

#68 - How to calculate empirical NTK of model being used in a classification task

Issue - State: open - Opened by uditsaxena about 4 years ago - 10 comments
Labels: question

#65 - Question: Is it possible to implement recurrent neural networks?

Issue - State: open - Opened by alpha358 about 4 years ago - 3 comments
Labels: enhancement

#57 - Difference between the General `Kernel` and `AnalyticKernel`

Issue - State: closed - Opened by DarrenZhang01 over 4 years ago - 4 comments
Labels: question

#44 - Create pypi.yml

Pull Request - State: open - Opened by romanngg over 4 years ago

#30 - Memory and computation efficiency problems for empirical NTK kernel

Issue - State: open - Opened by maciejkorzepa over 4 years ago - 12 comments
Labels: enhancement

#29 - Memory and running time issues for CNN

Issue - State: closed - Opened by HornHehhf over 4 years ago - 11 comments
Labels: enhancement

#28 - Memory Constraint (Approximation Available?)

Issue - State: open - Opened by ginward over 4 years ago - 8 comments
Labels: enhancement

#23 - Support high-dimensional, variable-lengths sequences

Pull Request - State: closed - Opened by icml2020-attention almost 5 years ago - 2 comments

#21 - `utils.is_on_cpu` breaks inside of JIT.

Issue - State: open - Opened by sschoenholz almost 5 years ago - 2 comments
Labels: bug

#18 - very large memory footprint for a simple UNet

Issue - State: open - Opened by kayhan-batmanghelich almost 5 years ago - 7 comments
Labels: enhancement

#13 - Batchnorm

Pull Request - State: open - Opened by thegregyang almost 5 years ago - 4 comments