VibeThinker-3B-heretic
> [!TIP] > This model is reproducible! > > See the README in the reproduce directory for more...
Метка: reasoning
gemma-4-12B-coder-fable5-composer2.5-MTP-NVFP4
> TL;DR — A local Python-coding assistant that thinks before it codes. 8.25 GB, runs on one 16...
Glint-Trace
Glint-Trace is a QLoRA adapter that teaches a tiny language model (Qwen 3.5 0.8B Base) to think before it answers. The...
granite-4.1-8b-Claude-Opus-4.6-Thinking-MAX-gguf
Этот репозиторий содержит квантования формата GGUF модели DavidAU/granite-4.1-8b-Claude-Opus-4.6-Thinking-MAX. Квантование проводили локально с использованием llama.cpp (сборка b9556). Модальности:Генерация текста...
MiniCPM5-1B-Uncensored
A fully uncensored version of openbmb/MiniCPM5-1B, produced with a single training-free stage: single-direction abliteration (Arditi et al., 2024)....
Cosmos-T-80M
Cosmos-T-80M is the first model in the Cosmos-T series — small, from-scratch, decoder-only Transformers pretrained on chain-of-thought data...
SR2AM-v0.1-8B
We argue that efficient agentic reasoning benefits from decomposing deliberation into three interacting systems: reactive execution (System I)...
GRM-2.6-Opus-GGUF
Original Model: OrionLLM/GRM-2.6-Opus Architecture: Qwen3.6-27B License: Apache 2.0 MTP Support: Yes Prompt: Create a complete SVG loading animation...
Qwen-3-0.6B-Claude-4.7-Opus-Distilled
Что произойдет, если вы возьмете аргументацию Клода 4.7 Opus и поместите ее в Qwen 3 0.6B? Вы получите...
Qwen3.6-27B-Opus-CoT-S1-Hermes-S2-SFT
> Calling for independent benchmarks. I’ve only run internal smoke tests against this model — full eval numbers...