Text Generation Models

Model · Text generation

Mangaba-Nano-1M

Mangaba AI

Um modelo de linguagem de 1,4 milhão de parâmetros, treinado do zero em português do Brasil pela Mangaba AI. Relatório técnico completo (dados, arquitetura, Muon, ablações, controles e destilação): https://huggingface.co/spaces/mangaba-ai/mangaba-nano-artigo Não há modelo base: os pesos começaram aleatórios e o tokenizer também foi treinado do zero sobre o mesmo corpus. O treino rodou nas GPUs NVIDIA L40S da própria Mangaba, em datacenter no Brasil. É um modelo de pesquisa e de borda: cabe em 5,6 MB, roda em qualquer CPU e serve como base para estudar escalas minúsculas, fazer ajuste fino em tarefas estreitas (classificação, autocompletar, extração simples) ou rodar em microcontroladores e…

Model · Text generation

MANN-GPT

Wuff Mann

因果 MANN · 全局记忆 v2 · 动态思考网络（DTN） MANN-GPT v6 is a memory-augmented autoregressive language-model architecture built around causal MANN layers, a formally qualified Global Memory v2 path, and a Dynamic Thinking Network (DTN). This repository is the frozen architecture and code archive for the first formally qualified bilingual continuation line. MANN-GPT v6 是一种记忆增强型自回归语言模型架构，其核心由 因果 MANN 层（causal MANN layers）、已完成正式资格验证的 Global Memory v2 全局记忆路径，以及 动态思考网络（Dynamic Thinking Network, DTN） 构成。本仓库保存首条完成正式资格验证的中英双语继续训练线路所对应的冻结架构与代码归档。 The canonical Phase-1 checkpoint passed value qualification, Global-Memory causal-identity retention, exact Q2 preservation, fixed-D3 retention, strict autoregressive…

Model · Text generation

MarinSkyRL-Open-MOPD-SmolLM3-3B-step-32

Open Athena

This is a MarinSkyRL-native Open-MOPD student after 32 optimizer steps. It starts from the authors' mixed-domain SFT checkpoint. Student responses were scored by the authors' math, code, and instruction-following RL teachers, routed by domain. The objective uses the student's selected top-16 token IDs and a clipped policy surrogate. This is an early checkpoint, not the authors' step-200 final model. The checkpoint is an unquantized, six-file Hugging Face export of the durable MarinSkyRL globalstep32 FSDP2 checkpoint. The policy export was used for the independent step-32 evaluation. The export's model.safetensors SHA-256 is bb7326640142069bc2e1fba5f54f15e0cccb1ff861f34f318b372eaab7abaf4b.…

Model · Text generation

mario-html-4B-GGUF

Wesjos Max

A visual-UI language model fine-tuned from Qwen3.5-4B, quantized to GGUF for llama.cpp. It answers in a small semantic HTML component system — bento grids, metric cards, steppers, callouts — wrapped in one root, instead of plain Markdown. Reasoning-native, and no system prompt required. One root element; everything else is an optional child: ai-view-header is not in the source repo's theme.css — it renders with default header styling. Fine, but define your own if you want tighter control. Apache License 2.0.

Model · Text generation

math-slm-qwen2.5-0.5b-v4

Mandavi Singh

A domain-specific small language model for step-by-step math problem solving, built by team03 (SLM Learners) for the Pramana SLM++ Bootcamp Round 2 submission. For an OpenAI-compatible endpoint, serve with servehf.py (stdlib + transformers only, no Ollama needed). Precision note: training ran in bf16 compute (QLoRA 4-bit NF4 base), but the merged checkpoint uploaded here is float16 (the merge step reloads the base in fp16). - Public Hugging Face datasets pulled via pulldata.py; licenses verified through the HF API on 2026-09-05 and recorded in datamanifest.md. - Held-out eval set built with buildheldouteval.py from raw ExamBench rows never used in training, with a final overlap check that…

Standard BF16 Hugging Face model, exported from training step 100. Weights, configuration, tokenizer, and the thinking chat template are at the repository root. Source checkpoint: hi-todayis-jh/maxrl-no-eos-qwen3-1.7b-compression-bs32-n16-32k-146103-step100 at commit 57eb43d677a67c0b80f242386b66f514b8838843. Load this repository directly with AutoModelForCausalLM.frompretrained or vLLM. Use enablethinking=True in the tokenizer chat template for the existing thinking evaluations. The export manifest records source provenance, file checksums, tensor validation, and a successful Transformers CPU load and forward pass.

Model · Text generation

Meet8-0.6B

M7

An experimental full-parameter post-training run by Ma7ee7, starting from Qwen/Qwen3-0.6B-Base. - Supervised fine-tuning on instruction, identity, human-written chat, generated FineWeb instruction examples, and generated creative writing. - Paired direct-answer and /think warmup using short reasoning examples. - Full-parameter GRPO on math, science, and logic tasks. - Approximately 3,500 total RL optimizer steps. Use the saved tokenizer chat template. The following system message substantially improves identity and reasoning-mode compliance: You are Meet8-0.6B, developed by Ma7ee7. If the user's message begins with /think, give a brief explanation inside..., then give your final response…

Model · Text generation

menasaat-qwen3-4b-oman-qlora-v2

Jaifar Alshizawi

Second-generation LoRA adapter fine-tuning Qwen/Qwen3-4B (Apache 2.0) on a combined bilingual corpus: the Oman Wikipedia instruction set plus 1,095 tourism-grounded pairs from menasaat/menasaat-oman-tourism-places. مُحوِّل LoRA يُخصِّص Qwen3-4B على مجموعة موسّعة تجمع ويكيبيديا العُمانية وبيانات الأماكن السياحية. Tourism knowledge: opening hours, locations (wilaya/governorate), best visiting seasons, and visitor tips for 159 places across all 11 governorates — plus Arabic location questions (أين تقع...؟) answered from the dataset. Tourism answers are grounded in the compiled dataset; verify time-sensitive details (hours, seasons) before high-stakes use. General-knowledge capability matches…

Model · Text generation

menasaat-qwen3-4b-oman-qlora-v2

Menasaat | Virtual Platforms LLC

Second-generation LoRA adapter fine-tuning Qwen/Qwen3-4B (Apache 2.0) on a combined bilingual corpus: the Oman Wikipedia instruction set plus 1,095 tourism-grounded pairs from menasaat/menasaat-oman-tourism-places. مُحوِّل LoRA يُخصِّص Qwen3-4B على مجموعة موسّعة تجمع ويكيبيديا العُمانية وبيانات الأماكن السياحية. Tourism knowledge: opening hours, locations (wilaya/governorate), best visiting seasons, and visitor tips for 159 places across all 11 governorates — plus Arabic location questions (أين تقع...؟) answered from the dataset. Tourism answers are grounded in the compiled dataset; verify time-sensitive details (hours, seasons) before high-stakes use. General-knowledge capability matches…

Model · Text generation

MiMo-V2.6-Flash-RL

Xiaomi MiMo

Scaling Reinforcement Learning Toward Self-Improvement MiMo-V2.6-Flash-RL is the efficiency-balanced checkpoint of the MiMo-V2.6 series. The series is built to scale reinforcement learning toward self-improvement — scaling RL compute, environment diversity, and grader compute together, so the model keeps expanding its capability frontier through exploration and feedback. Key features include: - Native Omnimodal + Long Horizon: Text, image, video, and audio in one model; 1M tokens for long repositories, tool traces, and multi-session agent runs. - Multi-Prefix Multi-Teacher On-Policy Distillation (MOPD2): After mixed RL, MOPD2 combines autonomous student rollouts with prefix-conditioned…

Model · Text generation

MiMo-V2.6-Flash-RL

Zachary Howard

Scaling Reinforcement Learning Toward Self-Improvement MiMo-V2.6-Flash-RL is the efficiency-balanced checkpoint of the MiMo-V2.6 series. The series is built to scale reinforcement learning toward self-improvement — scaling RL compute, environment diversity, and grader compute together, so the model keeps expanding its capability frontier through exploration and feedback. Key features include: - Native Omnimodal + Long Horizon: Text, image, video, and audio in one model; 1M tokens for long repositories, tool traces, and multi-session agent runs. - Multi-Prefix Multi-Teacher On-Policy Distillation (MOPD2): After mixed RL, MOPD2 combines autonomous student rollouts with prefix-conditioned…

Model · Text generation

MiMo-V2.6-Pro-RL

Xiaomi MiMo

Scaling Reinforcement Learning Toward Self-Improvement MiMo-V2.6-Pro-RL is the flagship checkpoint of the MiMo-V2.6 series. The series is built to scale reinforcement learning toward self-improvement — scaling RL compute, environment diversity, and grader compute together, so the model keeps expanding its capability frontier through exploration and feedback. Key features include: - Native Omnimodal + Long Horizon: Text, image, video, and audio in one model; 1M tokens for long repositories, tool traces, and multi-session agent runs. - Multi-Prefix Multi-Teacher On-Policy Distillation (MOPD2): After mixed RL, MOPD2 combines autonomous student rollouts with prefix-conditioned single-turn…

Model · Text generation

MiMo-V2.6-Pro-RL

Zachary Howard

Scaling Reinforcement Learning Toward Self-Improvement MiMo-V2.6-Pro-RL is the flagship checkpoint of the MiMo-V2.6 series. The series is built to scale reinforcement learning toward self-improvement — scaling RL compute, environment diversity, and grader compute together, so the model keeps expanding its capability frontier through exploration and feedback. Key features include: - Native Omnimodal + Long Horizon: Text, image, video, and audio in one model; 1M tokens for long repositories, tool traces, and multi-session agent runs. - Multi-Prefix Multi-Teacher On-Policy Distillation (MOPD2): After mixed RL, MOPD2 combines autonomous student rollouts with prefix-conditioned single-turn…

This repository contains a text-only Mini-K3-1H v2 pretraining checkpoint from a controlled 20-architecture comparison. The family retains Kimi-K3's KDA and Gated MLA operators, block Attention Residuals, Stable LatentMoE, SiTU activations, output gates, and Quantile Balancing at approximately one billion logical parameters. The exact architecture for this repository is listed below; some ablations deliberately replace the baseline KDA/MLA ratio, decay granularity, convolution length, or positional encoding. - Hidden width / attention heads / KDA head width: 1024 / 12 / 128 - Vocabulary / BOS / generation EOS / PAD: 163840 / 163584 / 163586 / 163839 control state retained in FP32 where…

This repository contains a text-only Mini-K3-1H v2 pretraining checkpoint from a controlled 20-architecture comparison. The family retains Kimi-K3's KDA and Gated MLA operators, block Attention Residuals, Stable LatentMoE, SiTU activations, output gates, and Quantile Balancing at approximately one billion logical parameters. The exact architecture for this repository is listed below; some ablations deliberately replace the baseline KDA/MLA ratio, decay granularity, convolution length, or positional encoding. - Hidden width / attention heads / KDA head width: 1024 / 12 / 128 - Vocabulary / BOS / generation EOS / PAD: 163840 / 163584 / 163586 / 163839 control state retained in FP32 where…

This repository contains a text-only Mini-K3-1H checkpoint from the controlled Attention Residuals follow-up study. The comparison includes standard PreNorm, three block granularities, and full sublayer-level AttnRes. The original Block-4 baseline is the separately published nkkbr/Mini-K3-1H-v2; this model's only architectural change is the depth-wise residual mechanism. - Hidden width / attention heads / KDA head width: 1024 / 12 / 128 - Residual mode / decoder layers per block: block / 1 - Vocabulary / BOS / generation EOS / PAD: 163840 / 163584 / 163586 / 163839 control state retained in FP32 where defined by the implementation Every numbered checkpoint is an immutable Git tag. main…

This repository contains a text-only Mini-K3-1H v2 pretraining checkpoint from a controlled 20-architecture comparison. The family retains Kimi-K3's KDA and Gated MLA operators, block Attention Residuals, Stable LatentMoE, SiTU activations, output gates, and Quantile Balancing at approximately one billion logical parameters. The exact architecture for this repository is listed below; some ablations deliberately replace the baseline KDA/MLA ratio, decay granularity, convolution length, or positional encoding. - Hidden width / attention heads / KDA head width: 1024 / 12 / 128 - Vocabulary / BOS / generation EOS / PAD: 163840 / 163584 / 163586 / 163839 control state retained in FP32 where…

English | MiniCPM5-2B-Math is a math and proof model built on MiniCPM5: a dense 2.5B model based on MiniCPM5-2B and post-trained for long-form mathematical reasoning and proof writing. It runs on a single consumer GPU or a laptop and keeps the native 128K context. - Strong for its size. As a plain single call with up to 128K tokens of thinking, 92.3 on AIME 2025, 89.8 on AIME 2026 and 81.8 on HMMT February 2026 (avg@16), and 100 / 93.3 / 90.9 with a 16-sample vote. The official general-purpose MiniCPM5-2B scores 87.1 / 89.8 / 68.4 under the same protocol. - Writes checkable proofs. On IMO-ProofBench Basic, an average single call with up to 128K tokens writes a complete proof for 21 of the…

GGUF builds of MiniCPM5-2B-Math, a 2.5B math and proof specialist, for llama.cpp, Ollama and LM Studio. Pair it with the MiniCPM-Math Harness to run a small proof swarm on a laptop. Converted with llama.cpp b10964 (converthftogguf.py, tokenizer pre-type minicpm5) and quantized with llama-quantize. The files load on llama.cpp b9354 or newer, which added the MiniCPM5 tokenizer; use b9833 or newer, which added the MiniCPM5 chat parser that separates the reasoning from the answer. The recommended sampling settings (temperature 0.9, top-p 0.95, min-p 0, top-k off) are embedded in the GGUF metadata, so llama.cpp uses them by default. All results on the model card were measured with the BF16…

Model · Text generation

minimind-math-v3

Jiwubu

minimind-math-v3 是一个基于 MiniMind 项目从零训练的小型中文数学语言模型，权重已转换为 Qwen3 兼容格式，可直接使用 HuggingFace transformers 加载推理。 - 架构：Dense Transformer（Decoder-only），未使用 MoE - 数据：以数学/推理语料为主的预训练数据（pretraint2tmini.jsonl），并经过数学 SFT 训练 - Tokenizer：基于 MiniMind 自训 tokenizer 修改的 patched tokenizer，详见下方说明 本模型的预训练与 SFT 均基于 jiwubu/minimind 代码仓库完成。相对上游 jingyaogong/minimind 的主要修改是新增 math/ 目录，用于生成数学 SFT 用的加减乘除算术数据（genmathdataaddsub.py、genmathdatamuldiv.py 等），并提供配套评测脚本（evalmath.py）。 最简单的方式是用仓库自带的交互式验证脚本（chat template + 贪心解码 + 流式输出）： 或直接用 transformers： - 预训练 2 个 epoch，batch size 32，最大序列长度 340 - 优化器 AdamW，初始学习率 5e-4（带 warmup + 余弦衰减） - 混合精度：bfloat16 本模型使用基于 MiniMind 原版 tokenizer 修改的 patched tokenizer（Sequence[Digits, ByteLevel]）：…

Ministral-3-8B-Instruct-2512 post-trained for the Polish history matura (extended level) during the Warsaw Model Trainers alpha=64) on top of mistralai/Ministral-3-8B-Instruct-2512-BF16 (the bf16 twin of the FP8 -2512 release). closedem = exact match on closed items. Graded scores (deepseek-flash-high@think via benchmark/bench.py) are reported separately.

Ministral-3-8B-Instruct-2512 post-trained for the Polish history matura (extended level) during the Warsaw Model Trainers alpha=64) on top of mistralai/Ministral-3-8B-Instruct-2512-BF16 (the bf16 twin of the FP8 -2512 release). closedem = exact match on closed items. Graded scores (deepseek-flash-high@think via benchmark/bench.py) are reported separately.

Ministral-3-8B-Instruct-2512 post-trained for the Polish history matura (extended level) during the Warsaw Model Trainers alpha=64) on top of mistralai/Ministral-3-8B-Instruct-2512-BF16 (the bf16 twin of the FP8 -2512 release). closedem = exact match on closed items. Graded scores (deepseek-flash-high@think via benchmark/bench.py) are reported separately.

Model · Text generation

ministral-8b-tlink-sentence-only-autoregressive

Anirban Saha Anik

Full-parameter fine-tuning for TLINK temporal relation classification. mistralai/Ministral-8B-Instruct-2410 AnirbanSaha/tlink-classification-sentence-only The evaluation/ directory contains predictions, the classification report, confusion matrix, and a machine-readable evaluation summary.

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).

Using llama.cpp release b3825 for quantization. Original model: https://huggingface.co/ArliAI/Mistral-Small-22B-ArliAI-RPMax-v1.1 All quants made using imatrix option with dataset from here Run them in LM Studio Some of these quants (Q3KXL, Q4KL etc) are the standard quantization method with the embeddings and output weights quantized to Q80 instead of what they would normally default to. Some say that this improves the quality, others don't notice any difference. If you use these models PLEASE COMMENT with your findings. I would like feedback that these are actually used and useful so I don't keep uploading quants no one is using. Thanks! First, make sure you have hugginface-cli installed…

One frozen public checkpoint is loaded at a time on a 24 GB GPU. A hash router picks it. Optional chains pass text from one expert to the next. A bag of those chains votes on the final string. A separate 128-dimensional cache, shared only by the routers, can slide along the chain. This repository is an assembly of published methods. It does not introduce a new routing, memory, or decoding algorithm. The bibliography at the bottom is the full set of ideas used in the code. Upstream weight files keep their own licenses and are not redistributed here. A request is hashed into a bag of word ids. A small classifier scores 138 expert ids. Confidence under 0.45 selects the int4 14B coder. An image…

Model · Text generation

Mochi-1.5

Eli Grenn

A small, unprofessional fine-tune of Qwen 2.5, made to have a slightly different personality. Based on huihui-ai/Qwen2.5-0.5B-Instruct-abliterated-v3. Mochi prioritizes having multiple different variants of the base model for different purposes, always using 0.5B parameters. It is only made for fun, and has no actual use if you're looking for something powerful or coherent. - No system prompt required. The personality is now baked in by default. You can still pass a system prompt to adjust behavior, but the Mochi vibe works without one. - Cleaner dataset. 868 examples, all without system prompts, so the model doesn't depend on one to behave correctly. - Factual accuracy is degraded compared…

A 162M-parameter decoder trained twice from the same initialization on the same 2.62 billion FineWeb-Edu tokens, once along each of two different routes through NVIDIA, AMD and Apple GPUs. The two runs are bitwise identical at every one of the 5,000 optimizer steps: training state, summed gradients, all 64 per-shard losses and the learning rate. Their final checkpoint files have the same SHA-256. This repository distributes that one final model with the evidence from both routes. The model is GPT-3 Small-sized and trained with GPT-3 Small's hyperparameters, but its blocks are modern Llama-style blocks; see the comparison below. It is a research artifact for reproducible training, not a…

Model · Text generation

moondream3.1-9B-A2B-GGUF

Amaimediacom

Local hardware: Mechrevo Kuangshi GM7AG0M — RTX 3060 Laptop 6GB GDDR6, 64GB DDR5, i7-12700H (14C/20T, 4.7GHz), Windows 11, Samsung 990 Pro. Good for imatrix and 0.6–35B-class work in RAM. 9B+ and searches need rented H200/Blackwell, typically $100 per quant. Released as part of the NOESIS Professional Multilingual Dubbing Automation Platform (framework: Deterministic Hybrid Control Framework for Frozen Neural Operators — DHCF-FNO). GGUF-конвертация и кванты текстового бэкбона Moondream 3.1 — vision-language модели с MoE-архитектурой (9B всего / 2B активных параметров). Конвертация и квантование выполнены AMAImedia; vision-часть (mmproj) в репозиторий не входит. Все файлы…

180B total / ~78B active per request — giant knowledge, agile inference. - 64 domain experts, zero overlap — every training example maps to exactly one expert. Auditable by construction. - Fully auditable data pipeline — raw → tagged → tokenized corpora, code included, in Morpho-72B-MoE-Data. - Resume-first engineering — every expert checkpoint (weights + loss + step) is versioned in Morpho-72B-MoE-Checkpoints. Any box can pick up exactly where the last one died. - 128K context via RoPE YaRN for long documents, codebases, and books. Eight sectors — Finance, Law, Medicine, Tech, Science, Business, Arts, Humanities — each with 8 specialists. Full roster with per-expert data counts is logged…

Stage 5 of the Morpho-180B pipeline (display name; SFT backbone: - Per-expert DPO, same 64 experts, same one-example-one-expert mapping as SFT. - Real preference data only (HH-RLHF, HelpSteer2, UltraFeedback, Nectar), routed per expert by the v5 taxonomy. - DPO adapters initialize from the converged SFT adapters — never from scratch. Stage 5 IN PROGRESS — 17/31 experts DONE (resumed 3x, AUTOSKIP active, final leg running on 97GB GPU box). 24 eligible experts remain in the queue; 33 thin experts (<200 pairs) keep SFT-only per policy. Each row is verified from the expert's bestloss.txt + DONE.txt on this repo. Progress tracked on the model card.

Model · Text generation

motif-3-bh-galaxy

Hyunggi Chang

This package runs the Motif-3 LLM of Motif Technologies on one Tenstorrent Blackhole Galaxy (32 Blackhole chips). - 53 layers of Grouped Differential Latent Attention (GDLA). - A 4-stream mHC residual. - 384 routed experts and 1 shared expert, all PolyNorm MLPs, with top-8 routing. - 1 MTP (multi-token-prediction) layer. - default: for all traffic. - mtp (opt-in): greedy requests decode up to 1.9× faster with the model's own MTP layer. - Built on tt-metal / ttnn as a plain vLLM model behind the vllm-tt-plugin. - Mesh (4, 8): attention is 8-way tensor-parallel in 4 groups of 8 user lanes; the MoE is 32-way expert-parallel (12 experts per chip). - Decode runs as captured Metal traces. The MTP…

A merged (LoRA-finetuned) variant of Qwen/Qwen2.5-1.5B-Instruct, fine-tuned for Cisco-related assistant tasks. This repository contains full merged weights — not a PEFT adapter. Load it directly with transformers; no peft and no adapter merging step is required. The model is tagged endpointscompatible and can be deployed directly to for a dedicated, OpenAI-compatible /v1/chat/completions API. Direct use: [e.g. answering Cisco product, configuration, and troubleshooting questions.] Out of scope: This model is not production-grade for safety- or compliance-critical networking decisions. Verify all configuration and command output before applying it to real hardware. Do not use it as the sole…

Model · Text generation

NadevA23_Kronumos-Kairos-v2-GGUF

Bartowski

Using llama.cpp release b11259 for quantization. Don't know which to choose? Grab Q4KM (4.78GB) - usually a good mix of size and performance. Download instructions available here First, make sure you have the Hugging Face CLI installed: These quants run with llama.cpp - installable in one line via llama.app: llama-server includes a built-in chat web UI, served at http://localhost:8080 by default. These quants were made with llama.cpp release b11259 - if this model's architecture is newly supported, you'll need that release or newer to run them. They also work in: quants made using imatrix option, with a calibration corpus rendered through this model's own chat template. The corpus pairs…

Model · Text generation

nanites-smoky-1b-chat

Nanites

A Llama-3.2-1B-Instruct model fine-tuned to play the role of an old resident of the Great Smoky Mountains of the 1930s in conversation. You talk to it; the old-timer answers in period voice with regional vocabulary. This is the chat-tuned sibling of the CosyVoice2 gender-pooled speech voices (Earl / male at Nanite-Labs/nanites-smoky-speech-earl, Ethel / female at Nanite-Labs/nanites-smoky-speech-ethel). The chat model provides the language behind the TTS; pair them for a text-and-speech experience. Given a modern-English question about life in the Smokies (work, family, food, music, religion, hardships, the mountains themselves), the model answers in 1-3 sentences in the voice of an old…

Model · Text generation

nanites-smoky-1b-chat-gguf

Nanites

Q4KM quant of the merged nanites-smoky-1b-chat model — a Llama-3.2-1B-Instruct fine-tune that plays an old resident of the Great Smoky Mountains of the 1930s. For Ollama / LM Studio / llama.cpp users who don't want the base-model + adapter merge step. Adapter (train output) lives at Nanite-Labs/nanites-smoky-1b-chat; this repo holds the merged + quantized GGUF for direct use. Alpaca-style (### Instruction: / ### Input: / ### Response:). Example prompt

Model · Text generation

nanites-whaler-1b-chat

Nanites

A Llama-3.2-1B-Instruct model fine-tuned to play the role of a 19th-century American whaling first mate in conversation. You talk to it; the mate answers in period voice. This is the chat-tuned sibling of Nanite-Labs/nanites-whaler-1b, which is a text completer that continues 19th-century whaling logbook entries. The chat version is a different animal — don't mix them up. Talk to it like a sailor would talk to the mate on deck. It answers in the voice of a 19th-c. ship's mate, brief and direct, using "try-works", "by D.R.", "the Captain", "no speak of it", etc. Multi-turn is supported — the mate sees your prior Q&A in the context window and can refer to earlier exchanges. period voice) + 8…

Model · Text generation

nanites-whaler-1b-chat-gguf

Nanites

Q4KM quant of the merged nanites-whaler-1b-chat model — a Llama-3.2-1B-Instruct fine-tune that plays a 19th-century American whaling first mate. For Ollama / LM Studio / llama.cpp users who don't want the base-model + adapter merge step. Adapter (train output) lives at Nanite-Labs/nanites-whaler-1b-chat; this repo holds the merged + quantized GGUF for direct use. Alpaca-style (### Instruction: / ### Input: / ### Response:). Example prompt

A tiny pirate-themed GPT trained from scratch on a piratized version of TinyStories, then SFT-tuned. Built as a learning project — closer to nanoGPT than to a production LM. - model.safetensors — model weights. - config.json — architecture config (load into training.config.Config). - piratebpe.json — tokenizer (load with tokenizers.Tokenizer.fromfile). - trainingmetadata.json — full training config + metrics snapshot. - banner.png — the banner above. This model is not a transformers model — it uses the custom GPT class from this repo. - Trained on a small synthetic corpus (TinyStories, piratized). Vocabulary, grammar, and world knowledge are extremely narrow. - Short context window (256…

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).

This repo contains some APEX quantizations of Naphula/Goetia-26B-A4B-v1.7, created using bartowski's imatrix found here. As explained by its creators, APEX is "a novel MoE-aware mixed-precision quantization technique for llama.cpp". I suggest reading their information on the how and why of the method. Quantized using llama.cpp #1692f9e50 For roleplaying, I personally strongly recommend nohurry's SillyTavern presets for Gemma 4 models. My only contribution is compute. This is neither my merge, nor my quantization methodology. Have fun (within the constraints of the license). Google - the base model. Naphula - the Mergelord. Everyone whose finetunes were included in the merge! mudler - for…

Model · Text generation

ncm-checkpoints

HazyResearch

Released tokenizer and generator pairs for Next-Concept Modeling. Each folder holds a config.yaml with the architecture and sampling settings and the.pt checkpoints it names.

Model · Text generation

nebium-small

Nabin Oli

Nebium-Small is a 117-million-parameter causal Transformer trained for autoregressive next-chess-move prediction over Lichess UCI move sequences. - Rotary Position Embeddings (RoPE) on attention query and key projections ($\theta = 10000$) - SwiGLU feed-forward transformation - RMSNorm pre-normalization - Causal mask with padding token masking - Byte-Pair Encoding (BPE) tokenizer trained on UCI move plies $$L(N, D) = 1.69 + \frac{406.4}{N^{0.34}} + \frac{410.7}{D^{0.28}}$$ MIT License.

Model · Text generation

nebium-small-gguf

Nabin Oli

Quantized and FP16 GGUF format binaries for Nebium-Small (117M parameters). Designed for low-latency CPU and GPU execution with llama.cpp and Ollama. MIT License.

Model · Text generation

Nemotron-3.5-Bolt-74B-A9B-CRACK-24GB-GGUF

Dealign.ai

NemotronH hybrid Mamba2 + attention + MoE · 74B total / 9B active · VL (image + video) · 22 GB on disk This is NVIDIA's own Nemotron-3.5-Bolt-74B-A9B with its refusal training removed and quantized to NVIDIA's own 24GB mixed-precision layout. The multimodal vision tower is kept byte-identical to the BF16 source; it is shipped with NVIDIA's own mmproj file and the matching LiLiCorr speculative drafter. llama.cpp upstream b76b13379 or newer. Older forks cannot load Nemotron-H puzzle checkpoints. Core text + vision: With speculative decoding via LiLiCorr drafter: Thinking is a binary enablethinking chat-template kwarg (no effort levels): Default sampling: temperature=0 greedy for the cleanest…

Model · Text generation

NEO

NEO

NEO = Neural Expert Orchestrator. Release: v1 (September 29, 2026). Model card only; this repository does not yet contain model weights. NEO v1 is the HYPERneocloud serving name for DeepSeek-R1-Distill-Qwen-14B. The currently verified serving configuration uses upstream revision 1df8507178afcc1bef68cd8c393f61a886323761. It is not presented as newly trained or independently owned base weights. No NEO fine-tuned adapter is verified as loaded in this endpoint. - https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B - Exact revision was verified using the model-hub revision API. - Upstream card identifies the license as MIT. Preserve the original license and attribution in any…

游戏 UGC 评论「语义单元抽取」模型。基于 Qwen/Qwen3-4B，LoRA（r16 / alpha32 / dropout0.05）训练后合并进底座，fp16 权重，可直接用 transformers 加载，无 peft 依赖。 - 任务：从一条评论文本中抽取语义单元（mention / span / occurrence），只做抽取，不做分类 - 硬件：实测峰值显存 8.82 GB，T4/T10（16GB，Turing）可直接部署，attention 走 sdpa，无需 FlashAttention2 生成参数已固化在 generationconfig.json：贪心解码（dosample=False）+ maxnewtokens=2048，请勿改动，否则与评测结果不一致。 ok=False 表示输出 JSON 解析失败或存在非法 span（mention 不在正文 / occurrence 越界），此时 semanticunits 为已通过校验的子集；线上建议对 ok=False 记录走兜底（重试或降级）。 - JSON 成功率 98.99%，非法 span 0 - SDK 与评测输出平价验证：20/20 DEV 样本单元集合完全一致 1. dtype 必须 fp16：Turing 架构无 bf16 支持；Ampere+（A10/A100/4090）可改用 bf16 3. 时延：A100 mean 4.8s / p50 3.8s 每条；Turing 预计 8~15s，随单元数线性增长 4. 输入 text 建议 ≤ 1024 token（训练 maxseq…

Model · Text generation

nirca-mini

nirca - non-idempotent rolling context AI/Agent

Nirca Mini is a small chat model trained from scratch, with ternary weights: its linear layers' weights are -1, 0 or +1, with a scale and a bias for each block of 128 weights. It reads and writes bytes, so it needs no tokenizer. It - Trained from scratch in ternary, with quantization-aware training from almost the very start; the ternary weights were learned in training, not quantized after. Nirca was inspired by mini-AGI, and its architecture is Try it in your browser (WebGPU): https://tg-techie-agents.github.io/nirca-mini-webgpu/mini/ This card is written by program from the checkpoint's own record, at step 122,203 (4,592,602,367 tokens seen). a mixture of experts (MLX's 2-bit affine…

This model is a fine-tuned version of francesca9805/ppt-wc-zipf-newlex-nld-before-100mb-packed-bfdisoseed455. 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).

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).

โมเดล Thai GPT-2 ที่ Fine-tune จาก flax-community/gpt2-base-thai ด้วย THAITEXTS ข้อสอบปลายภาค 10301374 โดย 6704101346 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

nlp-lab5-8-gpt2

Kittiwong Meejan

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).