Qwen2.5-LLM: Extending the boundary of LLMs
- Source
- qwenlm.github.io
- Author
- Qwen Team
- Date
Context
Qwen's Qwen2.5 language-model family gives developers more size choices, but the variants do not share identical deployment terms. The release's model table spans seven dense models from 0.5B to 72B parameters, including new 3B, 14B, and 32B options.
The smaller variants, at 0.5B, 1.5B, and 3B, list 32,000- windows, while models from 7B upward list 128,000. Licensing also differs: the announcement places the 3B model under the Qwen Research License and the 72B model under the Qwen License, with the others under Apache 2.0. Choosing a size therefore also means checking how much input it accepts and which license applies.
- token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
- context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
- LLM — A large language model — the neural network behind tools like Claude and ChatGPT, trained on huge amounts of text to predict what comes next.
“Our research indicates a significant interest among users in models within the 10-30B range for production use, as well as 3B models for mobile applications.”
Checking sign-in…
Loading comments…