Output of randn_tensor changes based on dtype, making seeds non-reproducible if the model precision is changed

### Describe the bug

Using `diffusers.utils.torch_utils.randn_tensor` to create noise will create different random noise depending on the tensor dtype even if a generator with a manual seed is passed, meaning seeds are not reproducible between dtypes.

For example, the [prepare_latents](https://github.com/huggingface/diffusers/blob/ccc8321651ebb879f70e563274b2d03c84c18f2f/src/diffusers/pipelines/aura_flow/pipeline_aura_flow.py#L572) call in the auraflow pipeline passes the dtype [directly to the randn_tensor](https://github.com/huggingface/diffusers/blob/ccc8321651ebb879f70e563274b2d03c84c18f2f/src/diffusers/pipelines/aura_flow/pipeline_aura_flow.py#L379) function. Running this model in BF16, FP16 and FP32 will result in completely different images.

Unsure if this is intended behavior. Generating the seed in FP32 then converting to the target dtype would mean better portability between environments. Alternatively, letting the user specify the behavior by manually passing a dtype themselves would also fix this.

### Reproduction

Minimal example that creates a tensor that is equivalent to a 1024x1024 image in SDXL latent space. Interestingly enough the output of this function **does match** for very small tensors.

```py
import torch
import diffusers
from diffusers.utils.torch_utils import randn_tensor

test_shape = (1, 4, 128, 128)
test_dtypes = [torch.float32, torch.float16, torch.bfloat16]

for dtype in test_dtypes:
generator = torch.Generator(device="cpu").manual_seed(22)
noise = diffusers.utils.torch_utils.randn_tensor(test_shape, generator=generator, device=torch.device("cpu"), dtype=dtype)
print(f"Random {dtype} tensor: {noise.flatten()[:8]}") # print first 8 elements for testing
```

For image model test, the default auraflow pipeline was used, with the model loaded as a gguf file. Changing the dtype in the pipeline arg + the two args for the transformer causes the output to be different. Editing the pipeline to have the noise generation always happen in FP32 makes the outputs match.

### Logs

```shell
Output of the above example, consistent between multiple systems:

python noise_test.py
Random torch.float32 tensor: tensor([ 0.3920, 0.0734, -0.0045, -0.0535, -0.0589, 0.6002, 2.0421, 1.3273])
Random torch.float16 tensor: tensor([ 0.2698, -0.3406, 0.1014, 0.0960, -0.6147, 0.6489, 0.0311, -0.5171],
dtype=torch.float16)
Random torch.bfloat16 tensor: tensor([-0.5195, -1.1797, 0.4219, 1.2891, 0.6602, -0.9141, 2.3906, 2.2188],
dtype=torch.bfloat16)
```

### System Info

Tested on latest diffusers installed from git on windows with torch `2.1.1+cu121`.
Verified on diffusers `0.32.2` with torch `2.7.0.dev20250302+cu126`

### Who can help?

@yiyixuxu @sayakpaul

Beginne in src/diffusers/utils/torch_utils.py bei randn_tensor und untersuche anschließend den verknüpften prepare_latents-Pfad in src/diffusers/pipelines/aura_flow/pipeline_aura_flow.py. Führe die bereitgestellte Reproduktion mit festgelegtem Seed für float32, float16 und bfloat16 aus und prüfe vorhandene Utility-Tests, falls vorhanden. Erledigt ist die Aufgabe, wenn das erwartete Verhalten mit festgelegtem Seed entschieden, durch einen Regressionstest dokumentiert und über die relevanten Dtypes hinweg konsistent bleibt.

Vom Indexierungsmodell aus dem Issue-Text verfasst.