nlp projects

lacaille mot 4b supreme2 is a high-efficiency, multi-domain model fine-tuned on qwen3-4b using the mixture of thoughts (mot) dataset enhanced with code, math, science expert clusters.

sombrero r1 14b elite13 is a fine-tuned variant of the deepseek-r1-distill-qwen-14b model, enhanced through rl to serve as a high-performance reasoning assistant model.

galactic qwen 14b exp2 model is optimized for general-purpose reasoning and answering, excelling in contextual understanding, and multi-step problem-solving.

draconis qwen3 math 4b preview is fine-tuned on the qwen3-4b architecture, optimized for excellence in mathematical reasoning, logical problem solving, and structured content generation.

sombrero opus 14b elite5 model is optimized for general-purpose reasoning and answering, excelling in contextual understanding, and multi-step problem-solving.

equuleus opus 14b exp model is optimized for general-purpose reasoning and answering, excelling in contextual understanding, and multi-step problem-solving.

sombrero opus 14b elite6 model is optimized for general-purpose reasoning and answering, excelling in contextual understanding, and multi-step problem-solving.

gauss opus 14b r999 model is optimized for advanced problem-solving. fine-tuned with specialized datasets in mathematics, physics, and formal logic, it delivers structured outputs.

dinobot opus 14b exp [abliterated] is a high-performance, general-purpose language model based on the qwen 2.5 14b architecture. it is designed to excel in reasoning, explanation.