vulkan: support sqrt operation

[ddwkim]

Make sure to read the contributing guidelines before submitting a PR

This is sqrt implementation in F32 for Vulkan.
I have found the implementation was missing while porting a diffusion model scheduler in GGML.
The following is the result of test-backend-ops.

./build/bin/test-backend-ops -o SQRT
ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
ggml_cuda_init: found 1 CUDA devices:
  Device 0: NVIDIA RTX A6000, compute capability 8.6, VMM: yes
ggml_vulkan: Found 1 Vulkan devices:
ggml_vulkan: 0 = NVIDIA RTX A6000 (NVIDIA) | uma: 0 | fp16: 1 | bf16: 0 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: KHR_coopmat
Testing 3 devices

Backend 1/3: CUDA0
  Device description: NVIDIA RTX A6000
  Device memory: 48636 MB (46852 MB free)

  SQRT(type=f16,ne=[10,3,3,2]): OK
  SQRT(type=f32,ne=[10,3,3,2]): OK
  10842/10842 tests passed
  Backend CUDA0: OK
Backend 2/3: Vulkan0
  Device description: NVIDIA RTX A6000
  Device memory: 49140 MB (49140 MB free)

  SQRT(type=f16,ne=[10,3,3,2]): not supported [Vulkan0] 
  SQRT(type=f32,ne=[10,3,3,2]): OK
  10842/10842 tests passed
  Backend Vulkan0: OK
Backend 3/3: CPU
  Skipping CPU backend
3/3 backends passed
OK

[0cc4m]

Thank you, nice work.