-
I am trying to finetune kinetics400- vivit_base_factorised_encoder and facing this error: **AttributeError: module 'optax' has no attribute 'momentum' ,** of course tried to change it with adam, sgd …
-
Trying to call `distributed_shampoo` (as the version in the JAX folder doesn't support Optax) without setting any parameter other than `learning_rate` and `block_size` results in this error:
```py
…
-
It'd be great to add a flavor of the CCSA algorithm. The quadratic approximate works just as well as MMA in my experience, but is much easier to implement. A true functional (ie stateless) implementat…
-
> jaxlib.xla_extension.XlaRuntimeError: UNIMPLEMENTED: No registered implementation for custom call to te_scaled_upper_triang_masked_softmax_forward for platform CUDA
```python
from transformer_en…
-
I am trying to use Jax in my Julia codebase for something that Zygote cannot do well (meta-learning). Someone recommended PythonCall as a solution to some issues I was having with PyCall.
So far, P…
-
Hi, I trying for run to saycan code.
but, I met some errors
my setting
OS: Ubuntu 22.04
GPU: RTX3090, nvidia-driver 535, Cuda: 12.2, cuDNN: 8.9.5
chex 0.1.8
optax …
-
**Describe the bug**
Hey Kris, love your framework! Working with a custom environment, and your discrete action unit test works perfect locally. Don't spend much time investigating this yet, just cre…
-
It says "Additional, Lion still requires momentum tracking in bfloat16, which can be expensive for training giant models. One potential solution is to factorize the momentum to save memory.", how to …
-
Is there any interest in adding the [Adan optimizer](https://twitter.com/davisblalock/status/1561976182567870465?t=AHwO3of5ivzgE0dW06hZvA&s=35) to optax? If so I can do it
-
Clear output executed at 7:53 PM (0 minutes ago) executed in 13.212s
/usr/local/lib/python3.7/dist-packages/flax/optim/base.py:52: DeprecationWarning: Use `optax` instead of `flax.optim`. Refer to…