# Dolphin 2.9.2 Qwen2 72B

> 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.

Provider: Cognitive Computations · Type: LLM · State: Ready

model path: `accounts/fireworks/models/dolphin-2-9-2-qwen2-72b`

[Deploy on demand](https://app.fireworks.ai/dashboard/deployments/create?baseModel=accounts/fireworks/models/dolphin-2-9-2-qwen2-72b)

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.

## Dolphin 2.9.2 Qwen2 72B API Features

- **On-demand Deployment**: On-demand deployments allow you to use Dolphin 2.9.2 Qwen2 72B on dedicated GPUs with Fireworks' high-performance serving stack with high reliability and no rate limits. · [Docs](https://docs.fireworks.ai/guides/ondemand-deployments)

## Metadata

- **State**: Ready
- **Created on**: 7 JUN, 2024
- **Kind**: Base model
- **Provider**: Cognitive Computations
- **Huggingface**: [cognitivecomputations/dolphin-2.9.2-qwen2-72b](https://huggingface.co/cognitivecomputations/dolphin-2.9.2-qwen2-72b)

## Specifications

- **Calibrated**: No
- **Mixture-of-Experts**: No
- **Parameters**: 72.7B

## Supporting Functionality

- **Fine-tuning**: Not Supported
- **Serverless**: Not Supported
- **Context Length**: 131k tokens
- **Function Calling**: Not Supported
- **Embeddings**: Not Supported
- **Rerankers**: Not Supported
- **Supporting image input**: Not Supported

## Dolphin 2.9.2 Qwen2 72B FAQs

### What is Dolphin 2.9.2 Qwen2 72B and who developed it?

Dolphin 2.9.2 Qwen2 72B is a fine-tuned version of Qwen2-72B, developed by Cognitive Computations and hosted on Fireworks. The model was curated by Eric Hartford, Lucas Atkins, and Fernando Fernandes. It is designed for instruction-following, conversational tasks, and early agentic behaviors.

### What applications and use cases does Dolphin 2.9.2 Qwen2 72B excel at?

The model is optimized for:

- Conversational AI
- Code generation and reasoning
- Agentic systems
- Enterprise RAG workflows
- Search and multimedia 

It supports initial function-calling logic, though tool integration is not built-in.

### What is the maximum context length for Dolphin 2.9.2 Qwen2 72B?

The model supports a context length of 131,072 tokens via rope scaling (YaRN), inherited from its Qwen2-72B base.

### What is the usable context window for Dolphin 2.9.2 Qwen2 72B?

The full 131K token window is available in Fireworks on-demand deployments.

### What are known failure modes of Dolphin 2.9.2 Qwen2 72B?

- Uncensored behavior: The model is intentionally unaligned and can comply with unethical prompts; users are warned to implement safety filters before use
- No vision support, embeddings, or reranker capabilities
- Function calling is rudimentary; lacks full schema/tool use integration
- Fine-tuned on 8K token sequences, which may limit long-range coherence despite large context window

### How many parameters does Dolphin 2.9.2 Qwen2 72B have?

The model has 72.7 billion parameters.

### Is fine-tuning supported for Dolphin 2.9.2 Qwen2 72B?

Yes. Fireworks supports LoRA-based fine-tuning for this model, available via the platform’s RFT offering.

### What rate limits apply on the shared endpoint?

- Serverless: Not supported
- On-demand: Available with no rate limits via dedicated infrastructure

### What license governs commercial use of Dolphin 2.9.2 Qwen2 72B?

The model is licensed under the Tongyi Qianwen license, inherited from its base (Qwen2-72B).
