Harsha

I recommend using `accelerate`. For more information, please refer to https://huggingface.co/docs/accelerate/usage_guides/memory or https://towardsdatascience.com/a-batch-too-large-finding-the-batch-size-that-fits-on-gpus-aef70902a9f1. I personally prefer the former as I find it to be much cleaner.

Unfortunately, `accelerate` only supports pytorch. Probably, will have to wait until tensorflow is supported.

Unfortunately, tensorflow doesn't have decorators/functions to auto-scale batch size like how lightning/accelerate for pytorch does. However, [here's](https://github.com/neuronets/nobrainer_training_scripts/blob/f0e5b9f41b9db1b07697487fb79d0c0d243cfb30/1.2.0/test_train.py#L213-L248) a naive example of accomplishing this. @satra let me know your thoughts about...

Hello! Any chance this PR will be merged to enable running the ec2 self-hosted runner as non-root?

> @machulav @hvgazula we've been using this branch for the past eight months at the Autoware Foundation without an issue, but I'd prefer it to get merged instead of using...

Thanks, @esteve. It worked.

You can simply comment out that line (and other related lines) and continue.

Hello, did you see this https://github.com/google-research/scenic/issues/547?

@sargun-nagpal Did you notice `*=` (Multiply AND) next to `neg_cost_loss` as well as `pos_cost_loss`?

Hello! Sorry for being unclear earlier. In fact, you derived the answer yourself 😉 . All you need to tell yourself is- In the equation from the article, `t` is...