Dolphin 2.9.2 Qwen2 72B is a fine-tuned version of the Qwen2 72B Large Language Model with a variety of instruction, conversational, and coding skills. Also supports function calling.
Fine-tuningDocs | Dolphin 2.9.2 Qwen2 72B can be customized with your data to improve responses. Fireworks uses LoRA to efficiently train and deploy your personalized model |
On-demand DeploymentDocs | On-demand deployments give you dedicated GPUs for Dolphin 2.9.2 Qwen2 72B using Fireworks' reliable, high-performance system with no rate limits. |
Dolphin 2.9.2 Qwen2 72B is a fine-tuned version of Qwen2-72B, developed by Cognitive Computations and hosted on Fireworks AI. The model was curated by Eric Hartford, Lucas Atkins, and Fernando Fernandes. It is designed for instruction-following, conversational tasks, and early agentic behaviors.
The model is optimized for:
It supports initial function-calling logic, though tool integration is not built-in.
The model supports a context length of 131,072 tokens via rope scaling (YaRN), inherited from its Qwen2-72B base.
The full 131K token window is available in Fireworks on-demand deployments.
The model has 72.7 billion parameters.
Yes. Fireworks supports LoRA-based fine-tuning for this model, available via the platform’s RFT offering.
The model is licensed under the Tongyi Qianwen license, inherited from its base (Qwen2-72B).
