# High-precision text generation

A 32B-parameter quantized large language model based on ArliAI/QwQ-32B-ArliAI-RpR-v4, quantized with llama.cpp at various precisions, suitable for text generation tasks.

Valkyrie-49B-v1 is a 49B-parameter large language model based on llama.cpp, offering multiple quantization versions suitable for different hardware configurations.

Autoregressive-32B is a Multiverse-32B baseline model built based on autoregressive modeling, providing strong support for text generation tasks.

This is a 12B-parameter large language model, processed with llama.cpp's imatrix quantization, offering multiple quantized versions to accommodate different hardware requirements.

INTELLECT 2 is a large language model launched by PrimeIntellect, supporting a context length of 40960 tokens, trained using the QwQ architecture and GRPO reinforcement learning framework.

This is a project that uses the mergekit tool to merge multiple pre-trained language models with a 12B parameter scale. The aim is to integrate the advantages of different models and improve language processing capabilities.

A large-scale language model fine-tuned specifically for Japanese novel writing support

Quantized version of OpenThinker2-32B, using llama.cpp for imatrix quantization, supports multiple quantization types, suitable for text generation tasks.

LoyalMaid-12B is a large language model with a 12B parameter scale, obtained by fusing multiple pre-trained language models using the mergekit tool.

A Japanese reasoning model developed based on Qwen2.5-32B-Instruct, integrating the ChatVector of QwQ-32B and optimizing Japanese reasoning performance.

Fused from multiple 12B-parameter large language models, equipped with text generation and dialogue capabilities

A 24B-parameter instruction-tuned model based on the Mistral architecture, suitable for text generation tasks

This is a GGUF format model converted from mlabonne/gemma-3-27b-it-abliterated, suitable for the llama.cpp framework.

Cydonia-24B-v2.1 is a 24B parameter large language model, processed with llama.cpp's imatrix quantization, offering multiple quantized versions to suit different hardware requirements.

A merged model based on multiple Qwen series 32B parameter models, enhanced with Model Stock method for improved multilingual processing capabilities

Lamarckvergence-14B is a pre-trained language model merged via mergekit, combining Lamarck-14B-v0.7 and Qwenvergence-14B-v12-Prose-DS. It ranks first among models with fewer than 15B parameters on the Open LLM Leaderboard.

Llamacpp imatrix quantized version of Nera_Noctis-12B, based on Nitral-AI/Nera_Noctis-12B model, supporting English text generation tasks.

FP8 quantized version of Meta-Llama-3.1-70B-Instruct, suitable for multilingual commercial and research purposes, especially ideal for assistant-like chat scenarios.

Bielik-7B-Instruct-v0.1 is a Polish large language model fine-tuned for instructions, based on Bielik-7B-v0.1, developed by the SpeakLeash team in collaboration with ACK Cyfronet AGH, specializing in Polish language understanding and processing tasks.

Quantized version of Mistral-Evolved-11b-v0.1, using llama.cpp for quantization, offering multiple quantization options to suit different needs.

A 7B-parameter large language model based on the Mistral architecture, trained with DPO (Direct Preference Optimization), demonstrating excellent performance in multiple benchmarks

QuartetAnemoi-70B-t0.0001 is a 70B-parameter large language model that merges multiple excellent models through a custom NearSwap algorithm, excelling in storytelling while avoiding clichés.

Daring Maid-20B-V1.1 is an upgraded version based on DaringMaid-20B, with the main update being the replacement of Noromaid-13b from v0.1.1 to v0.3, along with a slight increase in Noromaid's weight to ensure better compatibility.

Daring Maid-20B is a text generation model based on the fusion of multiple excellent models, aiming to create a smarter and more instruction-following Noromaid model.