Text Generation Models

Model · Text generation

Winery-Qwen3.8-Flash-Next-Cuvee-GGUF

Winery

One blend of Qwen3.8-Flash-Next (125B MoE, 512 experts, 10 active) in three bottles, sized so it runs on a gaming PC with Winery Strata. That's a fork of Strata with a wine-red glass UI, where Cuvée is the first model in the installer. - Short thinking from Swift 1.5 (UkisAI) - Refusals poured out by Huihui's abliteration Visit the Cuvée Space for the expert-cellar map and a bottle picker for your RAM. All three keep the dense weights in Q6K/Q80 and the 28.8 GB PLE table in IQ4NL (Strata reads it from the SSD). They were quantised with Unsloth's imatrix for Flash-Next. Slim drops half the experts, the same move as ISTA-DASLab's Coder. The difference is that it keeps the 256 experts routed…

Wrench is a pruned, task-specific developer-tool SLM derived from Qwen3.6-35B-A3B. It contains 3,881,244,016 parameters and stays below the 4.25B parameter ceiling. This is a public experimental artifact. It is downloadable and reproducible, but it is not a claim that the final 4M retrieval-quality or MiniMax-parity gates have passed. The canonical distribution format is Hugging Face Safetensors. The package embeds the tokenizer hook, deterministic mechanical lookup runtime, long-context overlay, hash-bound package metadata, and the FreeToken launcher. It is intended to feel like one model directory, not a separately installed harness. The bundled launcher requires a compatible FreeToken…

YANchor-4B is a general-purpose recurrent model that preserves crucial memory as ANchors for retrieval during subsequent reasoning. Its multidimensional memory mechanism enables effective long-horizon reasoning with O(N) generation time and O(1) history-state memory. For example, on challenging math problems, YANchor achieves 82.93% mean pass@1 on AIME 2024–2026 and 63.64% on HMMT, substantially outperforming linear-time, constant-state counterparts, including 7B–14B models. YANchor delivers several-fold higher batched long-generation throughput than Transformer and hybrid baselines on H100. Evaluations across dozens of benchmarks demonstrate its superiority in general-purpose capabilities.…

Model · Text generation

yejin-korean-8b-v3-cpt6000-distill-qwen36-lora-lr2e4-nothink-v2

Hajunho

Korean instruction-following language model. Built from Yejin 8B v3 (pretrained from scratch by Yejin: cooler8/yejin-korean-8b-v3-base): continued pretraining for 6,000 steps on public Korean data, Yejin chat vector from our own SFT model, then LoRA (rank 64, lr 2e-4). The LoRA data mixes our phase-1 format data (45,013 samples) with 39,290 short Korean solutions written by Qwen/Qwen3.6-27B (Apache-2.0) for questions from our own puzzle generator, KMMLU train, AI Hub school exams and AI Hub reading comprehension; only solutions that reached the known answer were kept (sequence-level distillation; no teacher weights or logits). No benchmark test items were used, and the teacher solutions…

This model is a fine-tuned version of Qwen/Qwen2.5-3B-Instruct. It has been trained using TRL. This model was trained with SFT. - PEFT 0.21.0

English · 中文 ds4 is a native C/CUDA inference engine plus an offline toolchain that runs DeepSeek V4.1 Flash on a single NVIDIA DGX Spark (GB10, 128 GB unified memory). Everything described here is our own work: the vector-quantization format, the solvers that produce the sidecar and the post-training file, the CUDA kernels, and the rulers we judge all of it with. The model architecture is DeepSeek's and is not re-explained here — read the official release for that. How to read this page. It unfolds in five steps; stop wherever you have what you need. 1. The numbers — what runs, how fast, how close to the original. 2. Three convictions — why the system is shaped this way. 3. The…

This is the model card of a transformers model that has been pushed on the Hub. Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. Use the code below to get started with the model. Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).

weighted/imatrix quants of https://huggingface.co/ZenithLLM/zen-ingress-1-3b-dpo-merged For a convenient overview and download list, visit our model page for this model. static quants are available at https://huggingface.co/mradermacher/zen-ingress-1-3b-dpo-merged-GGUF If you are unsure how to use GGUF files, refer to one of TheBloke's READMEs for more details, including on how to concatenate multi-part files. (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants) Here is a handy graph by ikawrakow comparing some lower-quality quant And here are Artefact2's thoughts on the matter…

Model · Text generation

Zenyx-MCEN-Delta-Mamba

Anamitra Sarkar

A ~565M-parameter language model that interleaves Mamba-3 (MIMO) and Gated DeltaNet-2 (GDN-2) sequence mixers with two MCEN stages (a reversible fixed-point "reasoning" core with a delta-rule memory), trained from scratch on the Zenyx-V3 pretraining mixture with the DeepSeek tokenizer. This repository holds training checkpoints, logs and diagnostics of an ongoing run; it is not a finished release. Data mixture — 121 sources in the frozen manifest (data/manifestv2.json, created 2026-10-10 06:04:26, hash 9454e840c2b9f032…); grouped shares: The comparison models are the same author's: Zenyx-v2-base (85M unique parameters, 32 effective layers, 32k vocabulary) and Zenyx-V3-Base (1.58B-parameter…

Model · Text generation

Zenyx-V3-Base

Anamitra Sarkar

Zenyx V3 is an efficient 1.5B-parameter Mixture-of-Experts (MoE) foundation model built for low-latency inference and high throughput. It is written from scratch in JAX/Flax and trained on TPU v5e-8. Latest checkpoint: step 111,200 (core tests re-run, next section) · the full benchmark tables further down were measured at step 107,200 · 81.0B tokens seen at step 107,200 Same protocol as the step-107,200 tables below (Kaggle T4, fp16, 0-shot, full evaluation sets, log-likelihood ranking, identical code and items). Only these tests were re-run; every other table in this card is still step 107,200. All differences in the first four rows are within about one to one-and-a-half standard errors of…

This model is a fine-tuned version of francesca9805/ppt-wc-zipf-newlex-zho-before-100mb-packed-bfdisoseed455. It has been trained using TRL. This model was trained with SFT.

This model is a fine-tuned version of francesca9805/zho-hans-10mb-ppt-Dp-100mb-packed-bfdisoseed10. It has been trained using TRL. This model was trained with SFT.

This model is a fine-tuned version of francesca9805/zho-hans-10mb-ppt-Dp-10mb-packed-bfdisoseed3407. It has been trained using TRL. This model was trained with SFT.

This is the model card of a transformers model that has been pushed on the Hub. Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. Use the code below to get started with the model. Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).

Model · Text generation

zoey

Dakuwon Moody

Status: experimental test completed; continuous background mode is disabled. The test did not change the serving checkpoint, app, or model weight files in this repository. This update documents the experiment only. A separate controller completed two successive coach → learner cycles without chat input, then stopped at its configured limit. It supplied a planning request and a follow-up, retaining the actual first coach answer as conversation history. The second lesson drafted with the checkpoint learned from the first. The student used the existing Map5/SEE memory and linear reranker, plus the frozen FP32 GPT2-medium chooser. The external coach was Qwen/Qwen2.5-32B-Instruct, revision…

Model · Text generation

Zrald-AI-model-quant-qwen-3.8-27b

Gerald Bustilla

High-efficiency, hardware-tested GGUF releases of Qwen3.8-27B (27 Billion Parameters, Dense Architecture). All models in this repository have been physically converted, verified on hardware (AMD Instinct MI300X with ROCm / HIP), and benchmarked for prompt throughput, token generation velocity, and benchmark accuracy against baseline models. This repository provides three specialized model tiers: - zraldv1-ac (Accuracy-Priority Tier): 17.08 GiB (18.3 GB). Near-lossless retention (99.68% accuracy), matches or outperforms standard Q6K and Q80 quality while saving ~10 GB VRAM compared to Q80. - zraldv1-ba (Balanced Sweet-Spot): 14.46 GiB (15.5 GB). Optimal balance (99.12% accuracy), fits…

Model · Text generation

Zrald-AI-qwen-3.8-27b-zraldv1-ac

Gerald Bustilla

Official high-efficiency GGUF release of Qwen3.8-27B optimized for Accuracy Priority Tier. This repository contains zraldv1-ac.gguf (17.08 GiB / 18.34 GB), physically benchmarked on AMD Instinct MI300X hardware. For cross-comparison tables against standard Q80, Q6K, Q5K, Q4K, and Q2K models, visit the master repository: - Base model by the Qwen Team (Alibaba) under Apache 2.0. - Runtime by Georgi Gerganov and the llama.cpp community.

Model · Text generation

Zrald-AI-qwen-3.8-27b-zraldv1-ba

Gerald Bustilla

Official high-efficiency GGUF release of Qwen3.8-27B optimized for Balanced Sweet-Spot Tier. This repository contains zraldv1-ba.gguf (14.46 GiB / 15.52 GB), physically benchmarked on AMD Instinct MI300X hardware. For cross-comparison tables against standard Q80, Q6K, Q5K, Q4K, and Q2K models, visit the master repository: - Base model by the Qwen Team (Alibaba) under Apache 2.0. - Runtime by Georgi Gerganov and the llama.cpp community.

Model · Text generation

Zrald-AI-qwen-3.8-27b-zraldv1-cs

Gerald Bustilla

Official high-efficiency GGUF release of Qwen3.8-27B optimized for Compressed Size Tier. This repository contains zraldv1-cs.gguf (10.18 GiB / 10.93 GB), physically benchmarked on AMD Instinct MI300X hardware. For cross-comparison tables against standard Q80, Q6K, Q5K, Q4K, and Q2K models, visit the master repository: - Base model by the Qwen Team (Alibaba) under Apache 2.0. - Runtime by Georgi Gerganov and the llama.cpp community.