How to get the original code by saved codebook and index

Nobody has claimed this yet.

Hi, I am trying to get quantized by saved codebook and index, but self.codebooks will always get 0. How can I do it?
```
import torch
from vector_quantize_pytorch import ResidualVQ

residual_vq = ResidualVQ(
dim = 256,
codebook_size = 256,
num_quantizers = 2,
kmeans_init = True,
kmeans_iters = 10
)

x = torch.randn(1, 1024, 256)

residual_vq.eval()

quantized, indices, commit_loss = residual_vq(x)

codebook = residual_vq.codebooks # get the codebook

torch.save(dict(codebook = codebook, indices = indices), 'codebook.pt')
codebook = torch.load('codebook.pt')['codebook']
indices = torch.load('codebook.pt')['indices']

residual_vq_reinit = ResidualVQ(
dim = 256,
codebook_size = 256,
num_quantizers = 2,
kmeans_init = True,
kmeans_iters = 10
)
residual_vq_reinit.codebook = codebook

quantized_out = residual_vq_reinit.get_codes_from_indices(indices)

assert torch.all(quantized == quantized_out.sum(dim = 0))
```

Start by running the provided reproduction with ResidualVQ and inspect the codebooks attribute alongside get_codes_from_indices. Trace how the saved codebook is stored and how the reinitialized instance reads its codebook, then verify that reconstruction from the saved indices matches the original quantized output.

Written by the indexing model from the issue text.