Qwen2.5-1.5B

Qwen2.5-1.5B features 1.5 billion parameters, delivering significantly improved performance over the 0.5B version while maintaining relatively low computational requirements. It retains the same 32K-token context window and 8K-token generation capability, enabling support for a wide range of tasks. This preserves the key advantage of lightweight models—the balance between performance and efficiency. The model demonstrates strong capabilities in logical reasoning, contextual understanding, and coherent text generation across multiple languages (including Russian), while remaining lightweight enough for deployment on computationally constrained, power-sensitive devices.

Qwen2.5-1.5B stands out for its fast response times, versatility, and ability to operate in resource-limited environments while maintaining high-quality responses and instruction comprehension. The model is particularly well-suited for enterprise chatbots, smart assistants, on-premise support systems, as well as educational and research projects where data privacy and autonomous operation are critical.

Announce Date: 19.09.2024
Parameters: 1.5B
Context: 32K
Attention Type: Full Attention
VRAM requirements: 1.6 GB using 4 bits quantization
Developer: Alibaba
Transformers Version: 4.43.1
License: Apache 2.0

Use our pre-built public endpoints to test inference and explore Qwen2.5-1.5B capabilities.

Rent your own physically dedicated instance with hourly or long-term monthly billing.

We recommend deploying private instances in the following scenarios:

• maximize endpoint performance,
• enable full context for long sequences,
• ensure top-tier security for data processing in an isolated, dedicated environment,
• use custom weights, such as fine-tuned models or LoRA adapters.

Contact our dedicated neural networks support team at nn@immers.cloud or send your request to the sales department at sale@immers.cloud.