Improve performance of gneissweb_classification: issue1017

[issei-ibm]

Why are these changes needed?

This PR addresses the issue of throughput of gneissweb_classification transform.

Related issue number (if any).

#1017
#1034

[touma-I]

@issei-ibm Could you please address the issue identified [here] (https://github.com/IBM/data-prep-kit/actions/runs/13233406276/job/36935004453?pr=1029)

==================================== ERRORS ====================================
_ ERROR collecting language/gneissweb_classification/test/test_gneissweb_classification.py _
../venv/lib/python3.10/site-packages/huggingface_hub/utils/_http.py:406: in hf_raise_for_status
    response.raise_for_status()
../venv/lib/python3.10/site-packages/requests/models.py:1024: in raise_for_status
    raise HTTPError(http_error_msg, response=self)
E   requests.exceptions.HTTPError: 401 Client Error: Unauthorized for url: https://huggingface.co/facebook/fasttext-language-identification/resolve/main/model.bin

The above exception was the direct cause of the following exception:

[issei-ibm]

@touma-I Could you look at my comment here?