Rope updates #800

	name: GPU Builds and Tests

	on:
	create:
	workflow_dispatch:
	push:
	branches:
	- master
	pull_request:
	branches:
	- master
	- llama3

	jobs:
	build-and-test-llama3:
	runs-on: ubicloud-gpu-standard-1-latest
	env:
	HF_TOKEN: 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
	steps:
	- name: Checkout code
	uses: actions/checkout@v4
	- run: echo "::add-mask::$HF_TOKEN"

	- name: Install OpenMP
	run: sudo apt-get update && sudo apt-get install -y libomp-dev

	- name: Install dependencies
	run: pip install -r requirements.txt

	- name: Run preprocessing
	run: python dev/data/tinyshakespeare.py --model_desc llama-3

	- name: Train model
	run: python train_llama3.py --write_tensors 1 --dtype float32 --offload 1

	- name: Build FP32 precision
	run: PRECISION=FP32 make test_llama3cu

	- name: Run default
	run: ./test_llama3cu

	- name: Build BF16 precision
	run: PRECISION=BF16 make train_llama3cu test_llama3cu

	- name: Run blocking
	run: CUDA_LAUNCH_BLOCKING=1 ./test_llama3cu

	- name: Run default
	run: ./test_llama3cu

Provide feedback