List your company

Tokenization on Tamil Wikipedia freq >= 3 (Full Evaluation)

1,699,970Total TokensSarvam-1 (68K Indic BPE)
864,4601,081,371.251,298,282.51,515,193.75Mar 6, 2026

Evaluation results

MethodLinks
Sarvam-1 (68K Indic BPE)Vocab=68,0962026.031,699,9703.5223.6
SentencePiece BPE 16KVocab=16,0002026.031,375,7782.85—
VerChol (morphological)Vocab=12,9912026.03914,6191.8933.5
VerChol (morphological)Vocab=32,9912026.03896,5951.8634.8