# Role-playing Optimization

A Llamacpp imatrix quantized version based on allura-org/Q3-30B-A3B-Designant, suitable for various quantization needs, supporting role-playing and conversational tasks.

This is a 70B-parameter large language model based on Llama 3.3, fine-tuned with SFT+DPO, specializing in character-driven dialogues and creative content generation

A model fine-tuned for role-playing scenarios based on Qwen3-8B, suitable for Japanese role-playing dialogue generation.

François-PE is a role-playing fine-tuned model based on the Mistral architecture, which addresses underfitting issues through the KTO method and improves coherence.

A dialogue model fine-tuned based on Gemma 27b, aiming to replicate the text quality of Claude 3, supporting ChatML format for conversational interactions.

32-bit floating-point upgraded large language model, significantly improving text generation quality and quantization performance through full-component FP32 remastering

MiniMaid-L2 is a role-play specialized model further optimized from MiniMaid-L1, achieving outstanding performance among 3B-scale models through knowledge distillation and training on a larger dataset.

A thought-oriented model fine-tuned based on Meta's Llama 3.3-70B-Instruct, focusing on enhancing creative writing and role-playing capabilities

A dialogue model fine-tuned and optimized based on microsoft/phi-3-mini-4k-instruct, specifically designed for role-playing scenarios, requiring specific prompt formats for optimal results

A multilingual ChatML model merged using the SCE method, supporting English and Japanese dialogue

Archaeo-12B is a 12B-parameter large language model specifically designed for role-playing and creative writing, obtained by merging the Rei-12B and Francois-Huali-12B models

A fusion model specifically designed for role-playing and creative writing, combining Rei-12B and Francois-Huali-12B via Slerp algorithm

A 70B-parameter large language model based on the Llama 3.3 architecture, integrating features from multiple excellent models, focusing on storytelling, scene description, and dialogue interaction capabilities.

Rei-12B is a large language model designed to reproduce the writing quality of the Claude 3 series models (especially Sonnet and Opus), fine-tuned based on Mistral-Nemo-Instruct, supporting long context and role-playing.

This is a 12B-parameter language model merged using the mergekit tool, specializing in creative writing and role-playing tasks with a unique narrative style.

Slush is a two-stage model trained with high LoRA dropout rate, focusing on enhancing creativity and role-playing capabilities

Nemotron-Mini-4B-Instruct is a response generation model developed by NVIDIA, optimized for role-playing, retrieval-augmented generation, and function calling. It is fine-tuned based on Minitron-4B-Base and supports a context length of 4096 tokens.

A large language model fine-tuned on the BeyondDialogue dataset, specifically designed for Chinese-English role-playing scenarios

Smart-Lemon-Cookie-7B is a pre-trained language model merged using mergekit, combining the strengths of multiple models, suitable for text generation and role-playing tasks.

Based on the Mistral 7B large language model, fine-tuned through pre-trained model merging on the Date-DPO-v2 dataset, excelling in role-playing, creative writing, and general intelligent tasks.

A 7B-parameter merged model with visual/multimodal capabilities

Nina-v2-7B is a 7B-parameter large language model merged via SLERP fusion technology from Nina-7B and Hermes-2-Pro-Mistral-7B, excelling in role consistency

A 70B-parameter large language model merged using the DARE linear fusion method, optimized for role-playing and story creation

A 7B-parameter large language model focused on role-playing, featuring 8192 context length with emphasis on creativity and reduced clichés

7-billion-parameter role-playing specialized model, incorporating features from high-quality models like Starling-LM, excels in vivid role-playing and precise instruction following

A 120-layer, 103-billion-parameter hybrid model self-merged from Midnight-Rose-70B-v2.0.3, specifically designed for role-playing and story creation

Estopia is a model focused on enhancing the quality of dialogue and prose generation under instruction formats, excelling in character cards, detail retention, and guided narratives.

This model is created by spherical linear interpolation (slerp) fusion between Mistral-7B instruction-tuned version and Synatra-7B role-playing version, combining both instruction understanding and role-playing capabilities

Astrea-RP-v1-3B is a 3B-parameter large language model that integrates multiple models through task arithmetic, focusing on text generation tasks.

A model merged from Noromaid-13b-v0.1.1 and Chronos-13b-v2, suitable for role-playing and storytelling.

Experimental stitched model, created by merging Thorn-13B and Norn Maid-13B, excels in role-playing scenarios

MythoMist 7B is a highly experimental merged model based on Mistral, utilizing innovative algorithms for active benchmarking to achieve user objectives, specifically targeting the reduction of certain negatively associated vocabulary frequencies.

Noromaid-20b-v0.1.1 is a large language model suitable for role-playing, emotional role-playing, and general scenarios, jointly developed by IkariDev and Undi.

Psyfighter2 is a fusion model created by the KoboldAI community, enhanced with medical data to improve fictional capabilities, but not suitable for actual medical use.

Nethena-20B is a 20B-parameter large language model developed by NeverSleep, suitable for role-playing, emotional interaction, and general purposes.

Athnete is a 13B-parameter large language model based on the Alpaca format, suitable for role-playing, emotional role-playing, and general purposes.

Athena v4 is an experimental large language model suitable for role-playing, emotional role-playing, and general scenarios. Uses Alpaca format prompt templates.

A versatile chat/instruction hybrid model based on Mistral-7B, excelling in single-turn instructions, multi-turn conversations, role-playing, and text adventure game generation

The AWQ quantized version of MythoMax L2 13B, which can effectively improve inference efficiency.

A fusion model of Pygmalion-2 13B and MythoMax 13B, specializing in role-playing/chat scenarios