AI21: Jamba Instruct
- 250K Context
- 0.5/M Input Tokens
- 0.7/M Output Tokens
- Ai21
- Text 2 text
- 02 Dec, 2024
The Jamba-Instruct model, introduced by AI21 Labs, is an instruction-tuned variant of their hybrid SSM-Transformer Jamba model, specifically optimized for enterprise applications.
- 256K Context Window: 它可以处理大量信息,相当于一本400页的小说,这对于涉及大文件的任务(如财务报告或法律文件)非常有利
- Safety and Accuracy: Jamba-Instruct 设计了增强的安全功能,以确保在企业环境中的安全部署,降低实施的风险和成本
Read their announcement to learn more.
Jamba has a knowledge cutoff of February 2024.