Dolly-v2-7b
| Property | Value |
|---|---|
| Base Model | Pythia-6.9b |
| License | MIT |
| Training Data | databricks-dolly-15k |
| Language | English |
What is dolly-v2-7b?
Dolly-v2-7b is an instruction-following large language model developed by Databricks, designed to be commercially viable and openly accessible. Built upon EleutherAI's Pythia-6.9b architecture, this model has been fine-tuned on a carefully curated dataset of approximately 15,000 instruction-response pairs generated by Databricks employees.
Implementation Details
The model leverages the transformers library and can be deployed using PyTorch. It supports both basic instruction following and more complex tasks across multiple domains. Notable technical aspects include support for bfloat16 precision to optimize memory usage and compatibility with popular frameworks like LangChain.
- Supports multiple instruction types including brainstorming, classification, and QA
- Implements efficient inference through device mapping and acceleration
- Provides flexible deployment options with custom instruction pipeline
Core Capabilities
- Instruction following across 7 major capability domains
- Information extraction and summarization
- Closed and open-ended question answering
- Text generation and classification tasks
- Context-aware responses with reference passage support
Frequently Asked Questions
Q: What makes this model unique?
Dolly-v2-7b stands out for being one of the first commercially viable, instruction-tuned language models released under a permissive license. While not state-of-the-art, it demonstrates surprisingly high-quality instruction-following capabilities not present in its base model.
Q: What are the recommended use cases?
The model is best suited for instruction-based tasks such as content generation, question answering, and summarization. However, users should be aware of its limitations with complex programming tasks, mathematical operations, and factual accuracy.
