3MPER0RR/TinyLlama-3MPER0RR-abliterated

TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:1.1BQuant:BF16Context Size:2kPublished:Aug 30, 2026License:mitArchitecture:Transformer Open Weights Featherless Exclusive Cold

TinyLlama-3MPER0RR-abliterated by 3MPER0RR is an experimental 1.1 billion parameter language model with a 2048-token context length. This model has undergone an experimental modification process, making it a unique variant of the lightweight TinyLlama architecture. It is intended for research and experimentation into the effects of such modifications on language model behavior.

Loading preview...

Overview

3MPER0RR/TinyLlama-3MPER0RR-abliterated is an experimental 1.1 billion parameter language model based on the TinyLlama architecture, featuring a 2048-token context length. This model has been subjected to an "abliteration" process, an experimental modification designed to alter its core characteristics. The project aims to explore the effects of these modifications on lightweight language models.

Key Characteristics

  • Experimental Modification: The model has undergone a unique, experimental modification process.
  • Lightweight Architecture: Built upon the TinyLlama base, making it suitable for resource-constrained environments.
  • Research Focus: Primarily intended for research and experimentation into modified language model behaviors.
  • Attribution: Original model authors and licenses are fully attributed to their respective creators.

Intended Use

This model is specifically designed for:

  • Research: Investigating the impact of experimental modifications on LLMs.
  • Experimentation: Testing hypotheses related to model architecture and behavior post-modification.
  • Exploration: Understanding the altered characteristics of lightweight language models.

Finetunes
1 models