| Parameter | Value |
|---|---|
| Algorithm | Byte-Level BPE |
| Language | ['arb_Arab', 'fas_Arab'] |
| Target Vocab Size | 16,000 |
| Final Vocab Size | 16,960 |
| Pre-tokenizer | custom:arb_Arab |
| Number handling | ltr_3digit |
| Contraction handling | True |
| Normalizer | NONE |
| Special Tokens | <s>, </s>, <pad>, <unk> |
| Training Shards | 4 |
1from transformers import AutoTokenizer
2
3tokenizer = AutoTokenizer.from_pretrained("flexitok/bpe_script_Arab_16000")
4tokens = tokenizer.encode("Hello, world!")tokenizer.json — Full HuggingFace tokenizervocab.json — Vocabulary mappingmerges.txt — BPE merge rules| Text | Tokens | Token IDs |
|---|---|---|
Hello, world! 12345 This is a test. こんにちは | H, ell, o, ,, Ġ, w, orld, !, Ġ, 123, 45, Ġ, Th, is, Ġ, is, Ġ, a, Ġ, t | 42, 5027, 81, 14, 223, 89, 12762, 3, 223, 16853, 5208, 223, 5728, 1147, 223, 1147, 223, 67, 223, 86 |