MythoMax-L2-13B

Gryphe · August 2023

activeOpen Weightdecoder onlytext
Parameters13B
Context Window4K tokens

Why It Matters

Became the go-to model for creative writing communities, demonstrating that model merging could produce specialized capabilities that no single training run achieved.

Description

One of the most popular creative writing and roleplay AI models, based on LLaMA 2 13B. Created by merging multiple separately fine-tuned models — a technique that combines the strengths of different training approaches into a single model — optimized specifically for narrative storytelling, character consistency, and creative fiction.

Notable Milestones

  • Dominant model on SillyTavern and other roleplay platforms
  • Popularized the model merging technique in the open-source community

Key Innovations

Instruction Tuning
Instruction TuningFine-tuning a model on instruction-response pairs so it follows user commands more reliably.
Open Weight
Open WeightModel weights are publicly released but training data/code may not be. Enables fine-tuning but not full reproduction.

Family Tree

Built On

Lineage

LLaMALLaMA 2MythoMax-L2-13B