Not really sure why they have a HuggingFace presence.

xfalcox 4 days ago | parent | next [-]

Having a public tokenizer is quite useful, specially for embeddings. It allows you to do the chunking locally without going to the internet.

	▲	jgalt212 4 days ago \| parent [-]
		and you don't have to re-embed everything if the provider sunsets a model.

To inform us that these models are based on Qwen 2.5. :D