-
### Summary
Keras Core (https://keras.io/keras_core/announcement/) looks like it might provide a convenient unified way to access the relevant tensor operations for TensorFlow, PyTorch and JAX via a …
-
I noticed earlier today that Haiku has SpectralNormalization -- very cool!
I'm interested in implementing an improved version, which does a much better job estimating the norm for convolutional lay…
-
### Problem Description:
I was going through the `issue_list` : [ Add creation functions to Jax frontend #8431 ]() and found out that some of the function present there to implement are actually cla…
-
I'm trying to differentiate the MJX step function via the autograd function `jax.grad()` in JAX, like:
```
def step(vel, pos):
mjx_data = mjx.make_data(mjx_model)
mjx_data = mjx_data.replace(q…
-
Woudn't it make sense to add something to wrap all this to make deserialization work? Currently it's lossy and you need to basically either define a init_param_from_params function or store the init p…
-
Hi,
I've finished the featuring step and got the output feature.pkl for the multiple prediction. Then I got these error messages:
(parafold) [linx@localhost ParallelFold-main]$ I0803 08:18:26.62…
-
Hi devs,
I'm trying to `vmap` over an root finding procedure involving vector-valued parameters. I gather this is currently not possible with `jaxopt.ScipyRootFinding` because this is a wrapper aro…
-
Currently this gives a bad error:
```python
import jax
import jax.numpy as jnp
from flax import linen as nn
class Einselm(nn.Module):
@nn.compact
def __call__(self, x):
self.sow('int…
-
### Description
I am getting following error if i load model in `predict ` model. it works perfectly in `eval` mode.
```
ValueError: Incompatible shapes for matmul arguments: (8, 1, 64) and (2…
-
### Description
I would like to export trax trained model as tf object, to serve it in tensorflow serving
...
### Environment information
```
Google Colab
$ pip freeze | grep trax
trax=…