[AFQ] Optimize tensor_flatten for runtime #4552

Job	Run time
test (CPU 2.4, linux.4xlarge, torch==2.4.0 --index-url https://download.pytorch.org/whl/cpu, cpu) / linux-job	8m 40s
test (CPU 2.3, linux.4xlarge, torch==2.3.0 --index-url https://download.pytorch.org/whl/cpu, cpu) / linux-job	7m 21s
test (CPU 2.2.2, linux.4xlarge, torch==2.2.2 --index-url https://download.pytorch.org/whl/cpu "nu... / linux-job	5m 18s
test (CPU Nightly, linux.4xlarge, --pre torch --index-url https://download.pytorch.org/whl/nightl... / linux-job	9m 26s
test (CUDA 2.2.2, linux.g5.12xlarge.nvidia.gpu, torch==2.2.2 "numpy<2" , cuda, 12.1) / linux-job	19m 28s
test (CUDA 2.3, linux.g5.12xlarge.nvidia.gpu, torch==2.3.0, cuda, 12.1) / linux-job	27m 22s
test (CUDA 2.4, linux.g5.12xlarge.nvidia.gpu, torch==2.4.0, cuda, 12.1) / linux-job	35m 58s
test (CUDA Nightly, linux.g5.12xlarge.nvidia.gpu, --pre torch --index-url https://download.pytorc... / linux-job	1h 0m 0s
	2h 53m 33s

Provide feedback