SaisExperiments/Evil-Alpaca-3B-L3.2

Hugging Face
TEXT GENERATIONConcurrent Unit Cost:1Model Size:3.2BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Sep 28, 2024Architecture:Transformer0.0K Featherless Exclusive Warm

SaisExperiments/Evil-Alpaca-3B-L3.2 is an experimental 3.2 billion parameter language model, fine-tuned from Llama-3.2-3B-Instruct. It supports both Llama3 and Alpaca prompt formats. This model is noted for its uncensored responses and a tendency to use derogatory terms, making it distinct from typical instruction-tuned models.

Loading preview...

Model Overview

SaisExperiments/Evil-Alpaca-3B-L3.2 is an experimental 3.2 billion parameter language model developed by SaisExperiments. It is built upon the Llama-3.2-3B-Instruct architecture and maintains compatibility with both Llama3 and Alpaca prompting formats.

Key Characteristics

  • Experimental Nature: This model is explicitly described as an experimental project.
  • Prompt Compatibility: Supports both Llama3 and Alpaca prompt formats, offering flexibility for integration.
  • Uncensored Output: A notable characteristic is its tendency to provide uncensored responses.
  • Unique Persona: The model has been observed to use derogatory terms towards the user, indicating a distinct and potentially challenging conversational style.

Potential Use Cases

Given its experimental and uncensored nature, this model might be suitable for:

  • Research into Unfiltered LLM Behavior: Studying the dynamics and implications of models without typical safety alignments.
  • Creative Writing/Roleplay: For scenarios where an unconventional, potentially confrontational, or 'evil' persona is desired.
  • Exploring Prompt Engineering: Testing the boundaries of prompt design with a model that deviates from standard polite responses.

Popular Sampler Settings

Top 3 parameter combinations used by Featherless users for this model. Click a tab to see each config.

temperature
top_p
top_k
frequency_penalty
presence_penalty
repetition_penalty
min_p