dolphin-2.9.1-yi-1.5-34b

Use cases

• Research into LLM safety training and RLHF effects via comparison with aligned models
• Red-teaming and security research requiring a model that does not refuse
• Evaluating alignment removal in large instruction-tuned models
• Use cases requiring explicit content generation where base model safety filters are a constraint

Pros

• 34B scale provides strong instruction following and reasoning depth
• Yi-1.5-34B backbone has broad multilingual coverage
• Community fine-tune with documented training data sources

Cons

• No safety filtering means the model will comply with harmful requests — requires responsible deployment
• Yi-1.5 base license must be reviewed for commercial use — not Apache 2.0
• Community fine-tune with no formal support or maintenance guarantees
• Accuracy on benign tasks may be reduced by uncensored training data mix
• Deployment requires careful access control to prevent misuse

FAQ

What is dolphin-2.9.1-yi-1.5-34b used for?

Research into LLM safety training and RLHF effects via comparison with aligned models. Red-teaming and security research requiring a model that does not refuse. Evaluating alignment removal in large instruction-tuned models. Use cases requiring explicit content generation where base model safety filters are a constraint.

Is dolphin-2.9.1-yi-1.5-34b free to use?

dolphin-2.9.1-yi-1.5-34b is an open-source model published on HuggingFace. License terms vary by model — check the model card for the specific license.

How do I run dolphin-2.9.1-yi-1.5-34b locally?

Most HuggingFace models can be loaded with transformers or the appropriate framework library. See the model card for framework-specific instructions and hardware requirements.

Tags