uzlm/alloma-1B-Instruct
TEXT GENERATIONConcurrency Cost:1Model Size:1BQuant:BF16Ctx Length:32kPublished:Sep 3, 2025License:llama3.2Architecture:Transformer0.0K Warm

The uzlm/alloma-1B-Instruct is a 1 billion parameter Llama-series model developed by Examy.me and Teamwork.uz, specifically optimized for the Uzbek language. It features a customized tokenizer that significantly improves inference speed and effective context length for Uzbek text, achieving 1.7 tokens per Uzbek word compared to ~3.5 in original Llama models. This model is designed for efficient deployment on resource-constrained environments, requiring only 2 GB of VRAM with quantization, making it suitable for small GPUs, edge devices, and mobile applications.

Loading preview...