Correct the inaccurate docstring for int4_row_quantize()

Summary:
X-link: facebookresearch/FBGEMM#1930

When  I was reading the code in `quantize.py`, the docstring in int4_row_quantize() seems inaccurate.

The `int4_row_quantize()` function returns `wq` with shape `[N, K]`, not `[N, K // 2]`:

```
# Line 101: Concatenate chunks back to original K dimension 
out = torch.cat(out, dim=-1)  

# Line 104: Convert to int8 dtype 
out = out.to(dtype=torch.int8)`
```

So `wq` is `[N, K]` stored as `int8` elements, where each `int8` element contains a single int4 value. And after it, the actual packing could be done via:

https://www.internalfb.com/code/fbsource/[c12bfdd174f7897f4615f98b630f2ac612c471fa]/fbcode/deeplearning/fbgemm/fbgemm_gpu/experimental/gen_ai/gen_ai/quantize.py?lines=18

Differential Revision:
D82919480

Privacy Context Container: 151967047006994

Author: geng-meta

fbgemm_gpu/experimental/gen_ai/gen_ai/quantize.py

@@ -81,7 +81,7 @@ def int4_row_quantize(
         x (Tensor): [N, K] Higher precision weight tensor to quantize.
         group_size (int): Number of elements to calculate group scale for.
     Returns:
-        wq (Tensor): [N, K // 2] Quantized int4 tensor stored in int8 elements.
+        wq (Tensor): [N, K] Quantized int4 tensor stored in int8 elements.
         group_scale (Tensor): [K / group_size, N] FP32 Scale per group.
     """
     n_bit = 4  # Number of target bits.