Dataset viewer displays wrong statists

### Describe the bug

In [my dataset](https://huggingface.co/datasets/speedcell4/opus-unigram2), there is a column called `lang2`, and there are 94 different classes in total, but the viewer says there are 83 values only. This issue only arises in the `train` split. The total number of values is also 94 in the `test` and `dev` columns, viewer tells the correct number of them.

### Steps to reproduce the bug

```python3
from datasets import load_dataset

ds = load_dataset('speedcell4/opus-unigram2').unique('lang2')
for key, lang2 in ds.items():
print(key, len(lang2))
```

This script returns the following and tells that the `train` split has 94 values in the `lang2` column.

```
train 94
dev 94
test 94
zero 5
```

### Expected behavior

94 in the reviewer.

### Environment info

Collecting environment information...
PyTorch version: 2.4.1+cu121
Is debug build: False
CUDA used to build PyTorch: 12.1
ROCM used to build PyTorch: N/A

OS: CentOS Linux release 8.2.2004 (Core) (x86_64)
GCC version: (GCC) 8.3.1 20191121 (Red Hat 8.3.1-5)
Clang version: Could not collect
CMake version: version 3.11.4
Libc version: glibc-2.28

Python version: 3.9.20 (main, Oct 3 2024, 07:27:41) [GCC 11.2.0] (64-bit runtime)
Python platform: Linux-4.18.0-193.28.1.el8_2.x86_64-x86_64-with-glibc2.28
Is CUDA available: True
CUDA runtime version: 12.2.140
CUDA_MODULE_LOADING set to: LAZY
GPU models and configuration:
GPU 0: NVIDIA A100-SXM4-40GB
GPU 1: NVIDIA A100-SXM4-40GB
GPU 2: NVIDIA A100-SXM4-40GB
GPU 3: NVIDIA A100-SXM4-40GB
GPU 4: NVIDIA A100-SXM4-40GB
GPU 5: NVIDIA A100-SXM4-40GB
GPU 6: NVIDIA A100-SXM4-40GB
GPU 7: NVIDIA A100-SXM4-40GB

Nvidia driver version: 525.85.05
cuDNN version: Could not collect
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: True

CPU:
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Byte Order: Little Endian
CPU(s): 64
On-line CPU(s) list: 0-63
Thread(s) per core: 1
Core(s) per socket: 32
Socket(s): 2
NUMA node(s): 4
Vendor ID: AuthenticAMD
CPU family: 23
Model: 49
Model name: AMD EPYC 7542 32-Core Processor
Stepping: 0
CPU MHz: 3389.114
BogoMIPS: 5789.40
Virtualization: AMD-V
L1d cache: 32K
L1i cache: 32K
L2 cache: 512K
L3 cache: 16384K
NUMA node0 CPU(s): 0-15
NUMA node1 CPU(s): 16-31
NUMA node2 CPU(s): 32-47
NUMA node3 CPU(s): 48-63
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate ssbd mba ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irperf xsaveerptr wbnoinvd arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif umip rdpid overflow_recov succor smca

Versions of relevant libraries:
[pip3] numpy==1.26.4
[pip3] torch==2.4.1+cu121
[pip3] torchaudio==2.4.1+cu121
[pip3] torchdevice==0.1.1
[pip3] torchglyph==0.3.2
[pip3] torchmetrics==1.5.0
[pip3] torchrua==0.5.1
[pip3] torchvision==0.19.1+cu121
[pip3] triton==3.0.0
[pip3] datasets==3.0.1
[conda] numpy 1.26.4 pypi_0 pypi
[conda] torch 2.4.1+cu121 pypi_0 pypi
[conda] torchaudio 2.4.1+cu121 pypi_0 pypi
[conda] torchdevice 0.1.1 pypi_0 pypi
[conda] torchglyph 0.3.2 pypi_0 pypi
[conda] torchmetrics 1.5.0 pypi_0 pypi
[conda] torchrua 0.5.1 pypi_0 pypi
[conda] torchvision 0.19.1+cu121 pypi_0 pypi
[conda] triton 3.0.0 pypi_0 pypi

Start by running the Python reproduction with datasets.load_dataset and compare unique('lang2') for the train, dev, and test splits. Trace how the dataset viewer computes train-column statistics; done means the viewer reports 94 values for train, matching the other splits and the reproduction.

Written by the indexing model from the issue text.