Twitter-domain vocabulary
According to the model card, the model extends an Indonesian BERT model with additive domain-specific vocabulary for Twitter.
Open Source Model Profile · indolem
indobertweet-base-uncased
indobertweet-base-uncased is a BERT-family fill-mask model from indolem. According to the model card, it is a large-scale pretrained model for Indonesian Twitter with domain-specific vocabulary.
- Publisher
- indolem
- Task
- fill-mask
- Model type
- bert
- License
- apache-2.0
- Library
- transformers
- Publication status
- Approved for indexing
Model overview
indobertweet-base-uncased is published by indolem as a fill-mask model for Indonesian Twitter. The captured configuration identifies BertForMaskedLM with model type bert. According to the model card, it extends an Indonesian BERT model with additive domain-specific vocabulary and is presented in an EMNLP 2021 paper.
Recorded capabilities
Documented preprocessing
According to the model card, usage lowercases words, converts mentions and URLs to @USER and HTTPURL, and translates emoticons with the emoji package.
Reported seven-dataset results
According to the model card table, the vocabulary-adapted variant averages 86.5 across sentiment, emotion, hate-speech, and NER sets, above the listed IndoBERT baselines.
Use cases in the source record
- Indonesian Twitter analysis such as sentiment, emotion, hate-speech, and NER workflows covered by the card's reported seven-dataset table.
- Fill-mask modeling with the documented preprocessing of lowercasing, @USER and HTTPURL replacement, and emoji translation.
Limitations and unknowns
- Reported benchmark figures and baseline comparisons are publisher claims and were not independently verified.
- No parameter count was extracted from this record.
- No context-window value was extracted from this record.
- Provider state is historical snapshot data, not independently refreshed current availability.
Source and provenance
Source: indolem/indobertweet-base-uncased
Captured: Unknown. Processed: 2026-09-07T19:34:47.096437+00:00.
IndoBERTweet 🐦 1. Paper Fajri Koto, Jey Han Lau, and Timothy Baldwin. IndoBERTweet: A Pretrained Language Model for Indonesian Twitter with Effective Domain-Specific Vocabulary Initialization . In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing ( EMNLP 2021 ), Dominican Republic (virtual). 2. About IndoBERTweet is the first large-scale pretrained model for Indonesian Twitter that is trained by extending a monolingually trained Indonesian BERT model with additive domain-specific vocabulary. In this paper, we show that initializing domain-specific vocabulary with average-pooling of BERT subword…
F001F002F003F004F005F006F007F008F009F010F011F012F013F014F015F016