A monolingual automatic speech recognition model for
Telugu
, fine-tuned from
openai/whisper-large-v3
. Part of
BuzzASR
,
a suite of 102 language-specialized ASR models (Findings of EMNLP 2026).
This model uses
full fine-tuning (native per-language tokenizer replacement + text multitask fine-tuning)
.
Results (normalized CER / WER, %)
Test set
CER
WER
Whisper-large-v3 (zero-shot) CER
FLEURS
13.53
47.06
82.65
Common Voice 25
1.22
8.21
103.03
Combined
13.03
45.49
74.67
~5.7x CER reduction over Whisper zero-shot on the combined test set.
telugu huggingface.co is an AI model on huggingface.co that provides telugu's model effect (), which can be used instantly with this BuzzASR telugu model. huggingface.co supports a free trial of the telugu model, and also provides paid use of the telugu. Support call telugu model through api, including Node.js, Python, http.
telugu huggingface.co is an online trial and call api platform, which integrates telugu's modeling effects, including api services, and provides a free online trial of telugu, you can try telugu online for free by clicking the link below.
telugu is an open source model from GitHub that offers a free installation service, and any user can find telugu on GitHub to install. At the same time, huggingface.co provides the effect of telugu install, users can directly use telugu installed effect in huggingface.co for debugging and trial. It also supports api for free installation.