mannix

llama3.1-8b-abliterated

Ablitered v3 llama-3.1 8b with uncensored prompt

65.9K Pulls 38 Tags Updated 2 years ago

gemma2-9b-simpo

Fine-tuned google/gemma-2-9b-it on princeton-nlp/gemma2-ultrafeedback-armorm with the SimPO objective.

11.5K Pulls 24 Tags Updated 2 years ago

llamax3-8b-alpaca

LLaMAX is a multilingual language model, developed through continued pre-training on Llama3, and supports over 100 languages

10.8K Pulls 16 Tags Updated 2 years ago

llama3-uncensored

llama3-8b with uncensored GuruBot prompt

10.3K Pulls 1 Tag Updated 2 years ago

llama3.1-8b-lexi

This is an uncensored version of Llama 3.1 8B Instruct with an uncensored prompt.

tools

10.3K Pulls 45 Tags Updated 2 years ago

deepseek-coder-v2-lite-instruct

An open-source Mixture-of-Experts code language model that achieves performance comparable to GPT4-Turbo in code-specific tasks.

8,131 Pulls 23 Tags Updated 2 years ago

defog-llama3-sqlcoder-8b

A capable language model for text to SQL generation for Postgres, Redshift and Snowflake that is on-par with the most capable generalist frontier models.

7,996 Pulls 16 Tags Updated 2 years ago

llama3-8b-ablitered-v3

Ablitered v3 llama-3 8b with uncensored prompt

5,134 Pulls 25 Tags Updated 2 years ago

qwen3.6-27b-a3b-coder

Pruned to 184e version of 35b-a3b with LCB and MultiPL-E HE targeting for Coding

vision tools thinking

3,972 Pulls 47 Tags Updated 1 month ago

hermes-3-llama-3.1-8b

Hermes 3 Llama-3.1 8b Model by NousResearch

tools

3,302 Pulls 22 Tags Updated 2 years ago

dolphin-2.9-llama3-8b

The uncensored Dolphin model based on Mistral that excels at coding tasks. Updated to version 2.9 with llama 3.

2,813 Pulls 2 Tags Updated 2 years ago

jan-nano

Jan-Nano is a compact 4-billion parameter language model specifically designed and trained for deep research tasks.

tools thinking

2,140 Pulls 4 Tags Updated 1 year ago

qwq-32b-abilterated

QwQ is an experimental research model focused on advancing AI reasoning capabilities. Abliterated with uncensored prompt, i-matrix quants.

tools

2,067 Pulls 17 Tags Updated 1 year ago

llava-phi3

A new small LLaVA model fine-tuned from Phi 3 Mini [I-Quants]

vision

2,065 Pulls 4 Tags Updated 2 years ago

qwen2.5-coder

The latest series of Code-Specific Qwen models, with significant improvements in code generation, code reasoning, and code fixing. I-Quant models.

tools

2,027 Pulls 47 Tags Updated 1 year ago

phi3-mini-4k

Phi-3 Mini is a lightweight 3B state-of-the-art open models by Microsoft. Updated in July 2024.

1,686 Pulls 19 Tags Updated 2 years ago

qwen3.6-27b-a3b-coderx

Code-targeted 184-of-256 expert cut of Qwen3.6-35B-A3B, LCB + MultiPL-E/HumanEval targeted, with MTP self-speculative decoding and vision tags. LCB v6 72.73 vs 61.04 base

vision tools thinking

1,643 Pulls 39 Tags Updated 2 months ago

dolphin-2.9.2-qwen2-72b

This model is based on Qwen2-72b, Dolphin-2.9.2 has a variety of instruction, conversational, and coding skills. It also has initial agentic abilities and supports function calling. Dolphin is uncensored.

1,553 Pulls 20 Tags Updated 2 years ago

gemma4-98e-v7-coder

An even more improved version of Gemma-4 98e coder variant, the best 20b coder

vision tools thinking

1,489 Pulls 27 Tags Updated 3 months ago

Bonsai-2-27B

Ternary Bonsai 2 27B (Qwen 3.8 27B) for xOllama

vision tools thinking

1,481 Pulls 13 Tags Updated 3 weeks ago

gemma4-98e-v6-coder

The best 20b coding model just got better! Beats the bigger 26b brother in Python and code reasoning

vision tools thinking

1,451 Pulls 62 Tags Updated 3 months ago

gemma2-9b-sppo-iter3

This model was developed using Self-Play Preference Optimization at iteration 3, based on the google/gemma-2-9b-it architecture as starting point.

1,149 Pulls 24 Tags Updated 2 years ago

omnimerge-v4-mtp

Qwen/Qwen3.6-27B + 3 Qwen3.6 fine-tunes with MLP-passthrough surgery - MTP quants

vision tools thinking

1,144 Pulls 39 Tags Updated 1 month ago

smallthinker-abliterated

A new small reasoning model fine-tuned from the Qwen 2.5 3B Instruct model. I-Quants models, abliterated with uncensored prompt.

1,095 Pulls 15 Tags Updated 1 year ago

omnimerge-v4

Qwen/Qwen3.6-27B + 3 Qwen3.6 fine-tunes with MLP-passthrough surgery

vision tools thinking

1,066 Pulls 53 Tags Updated 1 month ago

flux2-klein

4B diffusion model · Qwen3-4B text encoder · FLUX.2 VAE · 1024×1024 in four steps

4b

1,064 Pulls 1 Tag Updated 1 month ago

whisper

whisper.cpp large-v3 turbo · 99 languages · translation to English

989 Pulls 1 Tag Updated 1 month ago

gemma2-9b

Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models

910 Pulls 24 Tags Updated 2 years ago

omnimerge-v6

v4's merge rebased on Qwen/Qwen3.8-27B (+ 3 Qwen3.6 fine-tunes) — +6.5pp LiveCodeBench, MTP + vision

vision tools thinking

895 Pulls 39 Tags Updated 1 month ago

Bonsai-2-27B-CRACK

Ternary CRACK · Abliterated · No guardrails · PQ2_0 ternary 2.13 bpw · 7.2 GB

vision tools thinking

878 Pulls 7 Tags Updated 2 weeks ago

gemma4-98e-v7-coderx

Gemma-4 98e coder max variant, top notch coding skills at the expense of science knowledge

vision tools thinking

854 Pulls 27 Tags Updated 3 months ago

dolphin-mixtral

Uncensored, 8x7b and 8x22b fine-tuned models based on the Mixtral mixture of experts models that excels at coding tasks. Created by Eric Hartford.

837 Pulls 2 Tags Updated 2 years ago

JackOD-9B-Coder

Four-way omnimerge_v2 merge over Qwen3.5-9B (DeltaCoder 0.55 / Qwopus3.5-9B-Coder 0.30 / Ornith-1.5-9B 0.15), tuned for multi-turn agentic coding — it finishes tasks instead of looping. 9B dense, MTP head

tools thinking

803 Pulls 28 Tags Updated 1 week ago

ornith-1.5-27b-a3b-coderx

REAP-stability-floor 184-expert prune of Ornith-1.5-35B-A3B (~27B, A3B active, top-8) — the recommended arm of the pair. Beats the 256-expert teacher on LiveCodeBench v6 (+10.4pp)

vision tools thinking

735 Pulls 39 Tags Updated 1 month ago

qwen2-math-7b

Qwen2-Math is a series of specialized math language models built upon the Qwen2 LLMs

700 Pulls 21 Tags Updated 2 years ago

deepseek-v2-lite-instruct

A strong, economical, and efficient Mixture-of-Experts language model.

671 Pulls 8 Tags Updated 2 years ago

llama3.1-storm

Llama-3.1-Storm-8B outperforms both Llama-3.1-8B-Instruct and Hermes-3-Llama-3.1-8B!

tools

644 Pulls 13 Tags Updated 2 years ago

kokoro

Kokoro 82M · audio.cpp engine · 54 voices · mp3 at 24 kHz

82m

635 Pulls 1 Tag Updated 4 weeks ago

llama3-groq-tool-8b

A series of models from Groq that represent a significant advancement in open-source AI capabilities for tool use/function calling.

tools

623 Pulls 18 Tags Updated 2 years ago

qwq-32b

QwQ is an experimental research model focused on advancing AI reasoning capabilities. i-matrix quantizations.

tools

619 Pulls 23 Tags Updated 1 year ago

internlm2.5-20b

InternLM2.5 has open-sourced a 20 billion parameter base model and a chat model tailored for practical scenarios.

596 Pulls 14 Tags Updated 2 years ago

hermes-3-llama-3.1-70b

Hermes 3 Llama-3.1 70b Model by NousResearch

tools

586 Pulls 22 Tags Updated 2 years ago

llama3.1-70b

New state-of-the-art model from Meta available in 8B, 70B and 405B sizes

tools

584 Pulls 34 Tags Updated 2 years ago

llama3-12b

Meta-Llama-3-12B-Instruct is a depth upscaling merge of llama3-8b from M. Labonne

566 Pulls 22 Tags Updated 2 years ago

gemma4-98e-v5-coder

Pruned to 98 experts gemma-4 a4b 26b v5-coder. Best 20b coder model overall

tools thinking

521 Pulls 31 Tags Updated 5 months ago

llama-3.3

New state of the art 70B model. Llama 3.3 70B offers similar performance compared to Llama 3.1 405B model. I-Quants models.

tools

452 Pulls 9 Tags Updated 1 year ago

z-image

Diffusion model · Qwen3-4B text encoder · 1024×1024 in eight steps

434 Pulls 1 Tag Updated 1 month ago

smaug-llama3-8b

This model was built using the Smaug recipe for improving performance on real world multi-turn conversations applied to meta-llama/Meta-Llama-3-8B-Instruct.

405 Pulls 22 Tags Updated 2 years ago

replete-coder-llama3-8b

Replete-Coder-llama3-8b is a general purpose model that is specially trained in coding in over 100 coding languages.

402 Pulls 10 Tags Updated 2 years ago

wan2.2

Wan2.2 TI2V 5B (GGUF) · umt5-xxl text encoder · Wan2.2 VAE · 640×352, 33 frames at 24 fps

390 Pulls 2 Tags Updated 4 weeks ago

gemma2-2b

Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models

343 Pulls 11 Tags Updated 2 years ago

mixtral_7bx2_moe

A high-quality Mixture of Experts (MoE) model with open weights by Mistral AI.

343 Pulls 11 Tags Updated 2 years ago

llama3.1-8b

New state-of-the-art model from Meta available in 8B, 70B and 405B sizes.

tools

336 Pulls 41 Tags Updated 2 years ago

ornith-1.5-27b-a3b-coder

Code-targeted 184-expert prune of Ornith-1.5-35B-A3B (256e→184e, ~27B, A3B active, top-8). Beats the 256-expert teacher on LiveCodeBench v6 (+6.5pp) and HumanEval+ (+2.4pp) at ~28% fewer experts. Native MTP head + vision tower

vision tools thinking

323 Pulls 39 Tags Updated 1 month ago

smaug-llama3-70b

This model was built using a new Smaug recipe for improving performance on real world multi-turn conversations applied to Meta-Llama-3-70B-Instruct

310 Pulls 9 Tags Updated 2 years ago

qwen2-7b

Qwen2 is the new series of Qwen large language models

299 Pulls 4 Tags Updated 2 years ago

llama3-sppo-iter3

Meta Llama-3-8b with Self-Play Preference Optimization for Language Model Alignment at iteration 3

281 Pulls 21 Tags Updated 2 years ago

smaug-llama3-70b-32k

This model was built using a new Smaug recipe for improving performance on real world multi-turn conversations applied to Meta-Llama-3-70B-Instruct, 32k context length

269 Pulls 4 Tags Updated 2 years ago

jackod-9b-coder-dflash

Four-way omnimerge_v2 merge over Qwen3.5-9B (DeltaCoder 0.55 / Qwopus3.5-9B-Coder 0.30 / Ornith-1.5-9B 0.15), tuned for multi-turn agentic coding — it finishes tasks instead of looping. 9B dense, DFlash drafter

tools thinking

255 Pulls 28 Tags Updated 1 week ago

gemma4-98e-v4

Pruned to 98 experts gemma-4 a4b 26b v4

tools thinking

251 Pulls 30 Tags Updated 5 months ago

replete-coder-merged-8b

Replete-Coder-Merged-8b is a general purpose model that is specially trained in coding in over 100 coding languages

247 Pulls 21 Tags Updated 2 years ago

gemma4-31b-he1

Pruned/masked heads version of gemma4-31b

tools thinking

239 Pulls 26 Tags Updated 5 months ago

replete-adapted-llama3-8b

Replete-Adapted-llama3-8b is a general purpose model that is specially trained in coding in over 100 coding languages.

230 Pulls 17 Tags Updated 2 years ago

llama3-gradient

This model extends LLama-3 70B's context length from 8k to over 1m tokens. [I-Quants]

223 Pulls 4 Tags Updated 2 years ago

smaug-qwen2-72b

The latest in the Smaug series - a finetune of Qwen2-72B-Instruct

208 Pulls 21 Tags Updated 2 years ago

outetts

OuteTTS 0.3 500M · WavTokenizer vocoder

204 Pulls 1 Tag Updated 1 month ago

wan2.1

Wan2.1 T2V 1.3B · umt5-xxl text encoder · Wan2.1 VAE · 640×352, 33 frames at 16 fps

201 Pulls 2 Tags Updated 4 weeks ago

starling-lm-10.7b

Starling-LM-10.7B-beta, an open large language model (LLM) trained by Reinforcement Learning from AI Feedback (RLAIF)

165 Pulls 13 Tags Updated 2 years ago

qwen2-math-1.5b

Qwen2-Math is a series of specialized math language models built upon the Qwen2 LLMs

160 Pulls 7 Tags Updated 2 years ago

eurus-2-7b-prime

Eurus-2-7B-PRIME is trained using PRIME (Process Reinforcement through IMplicit rEward) method, an open-source solution for online reinforcement learning (RL) with process rewards, to advance reasoning abilities of language models.

159 Pulls 23 Tags Updated 1 year ago

nous-hermes2-solar-10.7b

The powerful Solar based model by Nous Research that excels at scientific discussion and coding tasks.

158 Pulls 7 Tags Updated 2 years ago

smallthinker

A new small reasoning model fine-tuned from the Qwen 2.5 3B Instruct model. I-Quants models.

142 Pulls 14 Tags Updated 1 year ago

qwen2-1.5b

Qwen2 is the new series of Qwen large language models

136 Pulls 4 Tags Updated 2 years ago

gemma4-98e

Pruned to 98 experts gemma-4 a4b 26b v3

tools thinking

125 Pulls 27 Tags Updated 5 months ago

wizardlm2

State of the art large language model from Microsoft AI with improved performance on complex chat, multilingual, reasoning and agent use cases. More quantizations.

115 Pulls 2 Tags Updated 2 years ago

discopop-zephyr-7b-gemma

A 7B parameter GPT-like model fine-tuned on a mix of publicly available, synthetic datasets using DiscoPOP

103 Pulls 5 Tags Updated 2 years ago

supertonic

Supertonic 3 · audio.cpp engine · 10 voices · mp3 at 44.1 kHz

102 Pulls 2 Tags Updated 4 weeks ago

phi3-mini-cpo-simpo

Phi-3-mini-4K-instruct with CPO-SimPO

101 Pulls 16 Tags Updated 2 years ago

kittentts

KittenTTS Mini 0.8 · audio.cpp engine · 8 English voices · mp3 at 24 kHz

85 Pulls 1 Tag Updated 4 weeks ago

whistle

Cactus Compute's Whistle · 7 languages · word timestamps · 17 MB · runs on the CPU (Metal on a Mac) · optional Silero speech detector · OpenAI /v1/audio/transcriptions

80 Pulls 2 Tags Updated 2 weeks ago

stable-audio

10 Pulls 2 Tags Updated 4 hours ago

Crazy Overclocker, Amateur Coder, AI Pragmatist, Non-sense Lover. xOllama & osync developer

llama3.1-8b-abliterated

gemma2-9b-simpo

llamax3-8b-alpaca

llama3-uncensored

llama3.1-8b-lexi

deepseek-coder-v2-lite-instruct

defog-llama3-sqlcoder-8b

llama3-8b-ablitered-v3

qwen3.6-27b-a3b-coder

hermes-3-llama-3.1-8b

dolphin-2.9-llama3-8b

jan-nano

qwq-32b-abilterated

llava-phi3

qwen2.5-coder

phi3-mini-4k

qwen3.6-27b-a3b-coderx

dolphin-2.9.2-qwen2-72b

gemma4-98e-v7-coder

Bonsai-2-27B

gemma4-98e-v6-coder

gemma2-9b-sppo-iter3

omnimerge-v4-mtp

smallthinker-abliterated

omnimerge-v4

flux2-klein

whisper

gemma2-9b

omnimerge-v6

Bonsai-2-27B-CRACK

gemma4-98e-v7-coderx

dolphin-mixtral

JackOD-9B-Coder

ornith-1.5-27b-a3b-coderx

qwen2-math-7b

deepseek-v2-lite-instruct

llama3.1-storm

kokoro

llama3-groq-tool-8b

qwq-32b

internlm2.5-20b

hermes-3-llama-3.1-70b

llama3.1-70b

llama3-12b

gemma4-98e-v5-coder

llama-3.3

z-image

smaug-llama3-8b

replete-coder-llama3-8b

wan2.2

gemma2-2b

mixtral_7bx2_moe

llama3.1-8b

ornith-1.5-27b-a3b-coder

smaug-llama3-70b

qwen2-7b

llama3-sppo-iter3

smaug-llama3-70b-32k

jackod-9b-coder-dflash

gemma4-98e-v4

replete-coder-merged-8b

gemma4-31b-he1

replete-adapted-llama3-8b

llama3-gradient

smaug-qwen2-72b

outetts

wan2.1

starling-lm-10.7b

qwen2-math-1.5b

eurus-2-7b-prime

nous-hermes2-solar-10.7b

smallthinker

qwen2-1.5b

gemma4-98e

wizardlm2

discopop-zephyr-7b-gemma

supertonic

phi3-mini-cpo-simpo

kittentts