Haohuan Wang
Haohuan Wang
Hi, I am trying to add c++11 completion support to my emacs configuration, what I did is add a std=c++11 flag into ac-clang-flags by using "(setq ac-clang-flags (append '("-std=c++11") ac-clang-flags))"...
i am working on keras2->keras3 adoptions for some of the models i am working with and i started to maintain a set of backend agnostic ops that used in our...
i am moving some of my models to keras3 and using jax backend for training. i am on a slightly older tensorboard (2.12) so when it pulls in keras callback...
this is a follow up from https://github.com/keras-team/keras/pull/19885 discussion where i am trying to make torch / keras well played together on tracking parameters. the solution i ended up with: 1....
i am trying to serve a keras3 model with tf saved model but running into issue when trying to load variables from using `saved_model_cli` or tensorflow c++ apis. using examples...
### Description ```python import jax import jax.numpy as jnp print(f"JAX version: {jax.__version__}") print(f"Devices: {jax.devices()}") print() jax_gpu_array = jax.device_put(jnp.array([1.0, 2.0, 3.0, 4.0], dtype=jnp.float32), jax.devices("gpu")[0]) # Test copy=True (expected to fail with...
for users doesn't use summary writer from tensorflow or tf.dataset but still want to use clu (eg. periodic actions are really nice); we should allow user to drop tensorflow as...