I am trying to build a Marketing Mix Model using the MCMC implementation from Jax and Black Jax. However when running the model, the sample acceptance probability stays zero, and so there are no ...

I'm trying to use JAX for implementing point cloud processing. However, I found that training becomes extremely slow due to my implementation of the following index_points_3d operation, which performs ...

I was trying to make a model testing series I ran into some horrible compilation behavior. The model runs about 5x faster on my CPU (Xeon W-2225, 4 core) than on my RTX 3060 (12GB) OR on my T1000 (4GB)...

I'm currently using Flax for neural network implementations. My model takes two inputs: x and θ. It first processes x through an LSTM, then concatenates the LSTM's output with θ — or more precisely, ...

We have been fetching the signing certificate from the Azure Key Vault and adding it to the local Windows store, which allows for the signing of macros in the .xla application through Excel. var ...

I turned on XLA when running TensorFLow, and in order to further optimize the fused kernels, I added export XLA_FLAGS="--xla_dump_to=/tmp/xla_dump", and got the dumped IRs, including lmhlo....

I am trying to compile the tensorflow serving project and I would like to have llvm/mlir compiled as a shared objects. The project is tensorflow serving -> tensorflow -> xla and compiles to a ...

Trying to understand the semantics of https://github.com/openxla/stablehlo/blob/main/docs/spec.md#scatter. Many of attributes do not have any explanation or definition. Can someone please explain how ...

I'm looking for a tool to calculate the FLOPs when given the computational graph of XLA-HLO. Is someone know some HLO cost models or analytical models for print the FLOPs of operator node for ...

In short, I have the problem of getting the following error when running a keras_cv/retina_net based object-detection model: "No registered 'RaggedTensorToTensor' OpKernel for XLA_GPU_JIT devices ...

When I run my python script, I had the messages below: WARNING: All log messages before absl::InitializeLog() is called are written to STDERR I0000 00:00:1701341037.989729 1542352 device_compiler.h:...

TL;DR Is (a and b) equivalent to tf.logical_and(a, b) in terms of optimization and performance? (a and b are tensorflow tensors) Details: I use python with tensorflow. My first priority is to make the ...

XLA can be enabled using model = tf.function(model, jit_compile=True). Some model types are faster that way, some are slower. So far, so good. But why can model = tf.function(model, jit_compile=None) ...

How can I pass user-defined variables to xlam files? I have written a macro and saved on an xlam file. I reference the xlam in an xlsm file and I call the macro from the xlsm. I want to pass a user-...

I'm fairly new to this and have little to no experience. I had a notebook running PyTorch that I wanted to run a Google Cloud TPU VM. Machine specs: - Ubuntu - TPU v2-8 - pt-2.0 I should have 8 cores....