This tokenizer contains 32,023 tokens, so models that use it should round the vocab size up to a multiple of 64/128 (ie 32064 or 32128) for CUDA reasons.
Compared to Original GPT-NeoX
The 'base tokenizer' in the below is the tokenizer from
EleutherAI/pythia-1b
Total tokens in base tokenizer: 50277
Total tokens in retrained tokenizer: 32023
Number of common tokens: 28219
Tokens unique to base tokenizer: 22058
Tokens unique to retrained tokenizer: 3804
Example common tokens:
['ounter', 'ĠRaymond', 'ĠIP', 'Ġcontroversy', '=', 'Ġsituations', 'Ġclimbed', 'Ġtrac', 'XY', 'Ġhave']
bpe-tokenizer-32k-smolNeoX huggingface.co is an AI model on huggingface.co that provides bpe-tokenizer-32k-smolNeoX's model effect (), which can be used instantly with this BEE-spoke-data bpe-tokenizer-32k-smolNeoX model. huggingface.co supports a free trial of the bpe-tokenizer-32k-smolNeoX model, and also provides paid use of the bpe-tokenizer-32k-smolNeoX. Support call bpe-tokenizer-32k-smolNeoX model through api, including Node.js, Python, http.
bpe-tokenizer-32k-smolNeoX huggingface.co is an online trial and call api platform, which integrates bpe-tokenizer-32k-smolNeoX's modeling effects, including api services, and provides a free online trial of bpe-tokenizer-32k-smolNeoX, you can try bpe-tokenizer-32k-smolNeoX online for free by clicking the link below.
BEE-spoke-data bpe-tokenizer-32k-smolNeoX online free url in huggingface.co:
bpe-tokenizer-32k-smolNeoX is an open source model from GitHub that offers a free installation service, and any user can find bpe-tokenizer-32k-smolNeoX on GitHub to install. At the same time, huggingface.co provides the effect of bpe-tokenizer-32k-smolNeoX install, users can directly use bpe-tokenizer-32k-smolNeoX installed effect in huggingface.co for debugging and trial. It also supports api for free installation.
bpe-tokenizer-32k-smolNeoX install url in huggingface.co: