[tx] Add cuDNN flash attention #1

	name: SkyRL-tx-GPU

	on:
	push:
	branches: [ main ]
	paths:
	- 'skyrl-tx/**'
	- '.github/workflows/gpu_skyrl_tx.yaml'
	pull_request:
	paths:
	- 'skyrl-tx/**'
	- '.github/workflows/gpu_skyrl_tx.yaml'
	workflow_dispatch:

	permissions:
	checks: write
	contents: read

	concurrency:
	group: skyrl-tx-gpu-${{ github.workflow }}-${{ github.ref }}
	cancel-in-progress: true

	jobs:
	skyrl_tx_gpu_tests:
	runs-on: ubuntu-latest
	defaults:
	run:
	shell: bash
	working-directory: ./skyrl-tx

	steps:
	- uses: actions/checkout@v4
	- name: Set up Python
	uses: actions/setup-python@v5
	with:
	python-version: '3.12'
	cache: 'pip'
	- name: Install the latest version of uv
	uses: astral-sh/setup-uv@v6
	with:
	activate-environment: true
	- name: Install dependencies
	run: uv pip install anyscale==0.24.79 typer==0.9.0
	- name: GPU tests
	env:
	ANYSCALE_CLI_TOKEN: ${{ secrets.ANYSCALE_CLI_TOKEN }}
	ANYSCALE_HOST: https://console.anyscale.com
	run: \|
	anyscale job submit -f ci/anyscale_gpu_ci.yaml --timeout 10000
	anyscale job wait --cloud sky-anyscale-aws-us-east-1 --name skyrl-tx-gpu-ci --timeout 10000

Provide feedback