Building on HF

Prithiv Sakthi PRO

prithivMLmods

👁 hugging-science

👁 Image
clem's profile picture 👁 Image
ai-lazzman's profile picture 👁 Image
p32k's profile picture

https://linktr.ee/prithivsakthi

AI & ML interests

computer vision, nlp, multimodality - HuggingFace Fellow ML 🤗

Recent Activity

updated a Space about 3 hours ago

prithivMLmods/Flux.2-Klein-Small-Decoder

updated a Space about 3 hours ago

prithivMLmods/flux-klein-kv-edit-consistency-fast

liked a model about 17 hours ago

prithivMLmods/gemma-4-12B-it-heretic_decensored

View all activity

Organizations

prithivMLmods 's collections 201

StyleCaption

Collection of Multimodal Models for Captioning -> Experimental

gemma 4 gguf

Collection of gemma 4/4[MoE] GGUF

MTP Qwen 3.5/3.6 Stable

Collection of Qwen 3.5/3.6 MTP Featuring GGUF

Stable Heretic / Multimodal Models

Revised Weights

Qwen → GLM-5.1 (Distilled)

Collection of GLM-5.1 math reasoning-distilled, abliteration variants of Qwen3.5/3.6

CapQwen3.6 - Caption Qwen3.6 Distilled (exp)

Collection of Qwen3.6 Captioning Expert System Models (+abliteration )

Garnet OCR

Collection of Garnet OCR Models

Qwen3.6-27B Compressions

Collection of various compressed versions of the Qwen3.6-27B native multimodal model.

Gemma 4 Compression(s) - MoE

Collection of Various Compression Schemes for Gemma 4 +MoE Calibration

Gemma 4 Compression(s)

Collection of Various Compression Schemes for Gemma 4

Qwen3.5 abliterated v2-MAX [BF16]

Version 2 in the Unredacted series, with much lower refusal limits, is available in different compression schemes.

VGA [* Polaris Series]

capable of accurately locating and understanding 'any' object | State: Experimental, Category: Object Detection, High-depth analysis in visual tasks

Qwen3.5 Caption [`Gliese Series ]

Expert Image Captioning System

Qwen Image Edit [Layout Bbox]

Collection of Region Bounding LoRAs

Multimodal FP8 Dynamic Collection

Collection of Compressed Multimodal VLM Editions on the Hub [ vLLM / LLM Compressor ]

Qwen Image Edit - AIOs / Compression

Collection of FP8 compression techniques, Diffusers–Transformers converted weights, and more for QIE.

Unredacted MAX - VL

multi-stage trained continual "abliteration" approach for the qwen3-vl series models

Multimodal Doc Models – Iterations up to [Feb 2026]

Iterations of my doc ocr models timeline of continual training built on top of qwen-vl models.

Qwen Image Edit (exps)

adapter LoRA developed for Qwen’s Qwen-Image-Edit-2511 image-to-image model

Physical Long-Horizon Reasoning

SFT, RL

Dynamic markdowns

Document-AI, HTML, OCR

Qwen3 VisionCaption

abliterated & [safe and controlled caption generation]

GGUF Captioning VLMs

All-in-One Quants

Kontext CAM Angles

Multiple Adapters for Heterogeneous Angles

Qwen VL Captions

Qwen3VL, Qwen2.5VL

General / Caption Abliteration

Qwen2.5-VL and Qwen3-VL

MM: VLM-Parsing

List of Models for Multimodal VLM Parsing

tool calling

agentic reasoning, web search orchestration

LZO-1 (Lossless Zoom Operator)

I2I Optimization

Radiation Qwen/Llama (exp)

Polished Tokens, Improved Multilingual Support, and Controlled Abliteration.

VL caption — < Sep 15 ’25

Multimodal models

Gliese-Post1.0

post1.0 ft. camel-doc-ocr-062825

DeepCaption attr.

Vision Language Attribution

Image Generation Apps [Collection]

Flux.1 (Krea, Dev, Schnell), Qwen-Image, Nano-Banana, Qwen-Image-Edit

VL Abliterated-Caption

qwen2.5-vl : 3b, 7b caption

Qwen3-4B-2507-GGUF

max - f32

Vision-Language for Reasoning (VLr) - Lumian2

Document Retrieval, Content Extraction, and Analysis Recognition.

Vision-Language for Reasoning (VLr)

Thinking / No-Thinking

rVL and Captioning Models

Image and video frame understanding, contextual analysis, and natural language reasoning, Abliterated Captioning / Uncensored Image Captioning,

Poseidon Reasoning

(LLM) research, benchmarking, and STEM-focused

Open-Omega: A Fusion of Math, Science, and Coding

reason/scientific thought mixture

Corvus-OCR-Caption-Mix

variable-dimensional pairs dataset

Re-think SmolLM

trl / @exp

SigLIP2 070225, 070125

new exps

Doc VL

drex [doc ], virex [ video (image ++) exp ]

Tron Experimentals

Reasoning and Non-Reasoning

GCIRS Reasoning Qwen

RL~Reward Signal

MoT Experimental Reasoning Traces R1

Disaster Classification Approach Exp

siglip2 base

Galactic (Sombrero) Opus R1

Code, Math, Moderately Abliterated

Open classifiers

image detection

Pet It

There are two means of refuge from the miseries of life: music and cats

OpenSDI Diffusion-Generated Image Classification

spotting diffusion-generated images in realistic, open-world scenarios

Qwen_3 GGUF

Upto Full precision (float32) version

QWQ_Qwen3 R1 LCoT

CoT, LCoT - QwQ / R1 Math

SigLIP2 05102025

Moderation, Balance, Classifiers

Core & DocScope OCR Models

DocVQA, RealWorldQA, MTVQA

Qwen3 Moderate Behavioral Flexibility

moderately abliterated and improved context awareness and moderate behavioral flexibility variant of Qwen3.

General Optimization Problems [ RPO ] for Reasoning !

Robust Policy Optimization (RPO)

Edge Device LLM - Advanced RL v3

advanced reinforcement learning (RL)

SigLIP2 Content Filters - Models v.2

Moderation, Balance, Classifiers

Edge LLM - Advanced RL

advanced reinforcement learning (RL)

Galactic LLM Exp4

Math, Reasoning

Fashion Product Full Net Experimentals

Usage, Season, BaseColour, ArticleType, SubCategory, MasterCategory, Gender

SigLIP2 Content Filters - Models v.1

Moderation, Balance, Classifiers

Multi-Source Experimentals

Domain Specific Classification Models : SigLIP2

Clean / Radio : Mono Voice

Names are randomly assigned to the respective TTS model based on gender, and the generated voice will have a radio and clean essence. [ Llama ]

Raptor X

Raptor X: Advanced UI Coder, Reasoning and General-Purpose Coder, and Math LLM

Galactic Qwen Exp 2

Math, Code, General-Purpose : Multipurpose

Image Classification Exp 032025

vit, siglip

LLM's >=40 Avg [14B]

OpenLLM Leaderboard

Galactic Qwen 14B's Test

Experimentals

14B Perceived Pattern

Experimentals

Sombrero QwQ Elite

Coding, Mathematical Problem Solving

Midranger Experimentals

14b - 32b experimental ft

Opus + Elite Optimized

general purpose 14b llms

Siglip2 Custom

SiglipForImageClassification

Deepfake Classification 022025

based on recent dataset

Purposeful Conversational Models 🚅

providing assistance

Deepfake Quality Assess Datasets

Classification of Good and Bad Quality Deepfakes { Datasets }

Deepfake Quality Assessment

Controllable Classification of Good and Bad Quality Deepfakes

Bellatrix Tiny 🤏🏻

1.5B, 1B, 0.5B param's

Reasoning Saturations🫙

7B, 14B [ Saturation ]

Vision Infer Custom

VLM - Multimodality

Phi-4 Custom FT

phi 4 custom ft [ qwq, lcot, math ]

PyThagorean 🐍

Solve Math with Python

LwQ -Llama 🦙

Llama with Questions for reasoning / long chain of thought

ChemQwen🧪

ChemQwen Based on InChI

GWQ

Open Eval < 500, long chain of thought and keyword based answering, reasoning

Deepthink and Reasoning

Best for Deepthink and Reasoning

Custom Build Models

Text Generation Cook with it's GGUF

GPT-Generated Unified Format (GGUF)

ease of reading

LoRA Space Collections

Flux & Sdxl

FLUX LoRA DLC2

🔥

584

100+ Impressive LoRA's For Flux.1

Stable Spaces

Only Stable

Flux LoRA Collections

Flux THE LoRA

gemma 4 qat

+ quantization-aware training (qat) ggufs for the gemma 4 family of models

MTP Qwen 3.5/3.6 [MoE] Stable

Collection of Qwen 3.5/3.6 MoE | MTP Featuring GGUF

QIE/FRIE1.1 — Test LoRAs [May2026]

Collection of Qwen Image Editing LoRAs

MiniCPM-V-4.6/5.0 — abliterated/uncensored

Collection of Uncensored Variants of MiniCPM-V-4.6 Multimodal Models

Qwen → DS V4 Flash (Distilled)

Collection of DeepSeek V4 reasoning-distilled, abliteration variants of Qwen3.5/3.6

Q3.5-Opus-DA

Collection of Opus-Reasoning-Distilled abliteration variants of Qwen3.5

Auto-Aggressive (Uncensored Qwen 3.6 Native Multimodal)

Collection of Uncensored Variants of Qwen3.6 Native Multimodal Models

Qwen3.6-35B-A3B Compressions

Collection of various compressed versions of the Qwen3.6-35B-A3B native multimodal model.

Gemma 4 Uncensored [MAX] + Compression(s) - [`β ]

Collection of Uncensored Variants of Gemma 4 Native Multimodal Models

Gemma-4 F32 GGUF

Collection of Gemma-4 Quants

Feb 15–Mar 25 '26

Multimodal Collections Sublist

Hall of Multimodal Apps

Audio, Image, Image-to-Text

Qwen3.5 Unredacted MAX

“abliteration” models – experimental. / no mtp

Qwen3.5 GGUF

Collection of Lightweight Qwen3.5 GGUF Models

Qwen 2.5 VL Unredacted MAX

“abliteration” models – experimental.

Unredacted MAX - VL [FP8]

[FP8_DYNAMIC] multi-stage trained continual "abliteration" approach for the qwen3-vl series models

Unredacted MAX - VL [GGUF]

multi-stage trained continual "abliteration" approach for the qwen3-vl series models (GGUF)

QIE Jan 23, 26

adapter LoRA developed for Qwen’s Qwen-Image-Edit-2511 image-to-image model

Qwen Image Edit (Object-Manipulator)

Add or remove the specified objects, flexible for both single-image and multi-image modes.

Computer Use Agent [SFT] – December 15, 2025

Localization, Tool Calling, and Navigation

Gliese VLMs Collection

Collection of Gliese series OCR models

MetaCLIP2 Image Classification Experiments

Domain-Specific Downstream Tasks

GGUF OCR VLMs

All-in-One Quants

Gliese OCR Post-x.0

Qwen2.5VL OCR (Model Tree)

Qwen3-VL Abliteration

Qwen3-VL Abliterated Model Collection [ Version 1.0 ]

VLM 2.0 - Oct 08'25

Multilingual upstream VLMs - Caption, OCR, Markdown, HTML Render

Gliese Edge

Multimodal VLM and LLM : 'Gliese' Collection of Models

Kepler Qwen (27/09)

Abliterated Reasoning & polished token probabilities, enhancing balanced multilingual generation

UI-Code Web.f

Polished and Controlled Gen

i2i - Kontext (exp)

Photo Restore, Polaroid, Inter-style Mix and more...

Caption 3o — filtered

filtered for an sfw captioning set of image pairs.

Qwen3 with DeepSeek v3.1 Traces

random event simulations, logical problem analysis, and structured reasoning tasks

Mixed-Model Traces

experimental

Qwen-Image-Exp-LoRA

Illustration Design, Style Intermix

Multimodal VLMs - Aug'25

Reasoning, Document Extraction, Table Reconstruction

VLMer Experimental 0818 / 0817

vlm extended reasoning, captioning

R1 Traces × Qwen3 Experiment

Flux.1-Krea-Merged-Schnell / Dev

(FLUX.1-schnell + FLUX.1-Krea-dev) | (Flux.1-Dev + Flux.1-Krea-Dev)

Flux.1-Krea-Merged-Dev

(Flux.1-Dev + Flux.1-Krea-Dev)

Multimodal VLMs - Until July'25

Multimodal VLMs for Domain-Specific Tasks: OCR, Reasoning, and Captioning

Explora x Poseidon Reasoning

Open-Omega : Exp

qwen2.5, 3

Captioning / OCR / DocTable

Mixed Model Operations

07/11 ~ Visual Understanding

Deep Caption VL - Trained with variable-dimensional pairs

Behemoth Tiny Thought [exp vl]

Chain-of-Thought Reasoning, Detailed Image Captioning

Camel Doc MM

enhances document comprehension capabilities

VisionScope OCR Experimentals

Based on Qwen2.5 VL, Qwen2 VL

Blitzar Experimentals

coder

MoT Supreme

Mixture-of-Thoughts

Open Voice Classification

audio classifiers

Open Classification 2805

(direct, reactive, retrieval-based responses), (reasoning, planning, deeper analysis)

Qwen_3 Experimental-2

trajectory

Explicit Content Filters

vit, siglip2

OpenG

Math, Code, Reasoning – Edge Device LLMs

Multiclass Image Classification 05142025

classification net.

Qwen_3 Experimental

Cot, LCoT - Reasoning

QWQ_Qwen3 Fine-Thinking

CoT, LCoT - QwQ Math

Content Filters SigLIP2/ViT

Moderation, Balance, Classifiers

SigLIP2 Content Filters 052025 Patch 1

Moderation, Balance, Classifiers

SigLIP2 Content Filters 042025 Final

Moderation, Balance, Classifiers

Multilabel Image Classification Datasets

classify one or more scene categories

Edge supervised fine-tuning (SFT)

standard language modeling objective

Edge LLM - Advanced RL v2

advanced reinforcement learning (RL)

Mini Galactic Qwen X RL - Edge Device Models

Coder, Math. [ Reinforcement learning ]

Y.2 Galactic Qwen

Y.2 for Coding and Math

Galactic LLM Exp 3

Math, Code, General-Purpose

SigLIP2 Content Filters - Datasets

Moderation, Balance, Classifiers

Messy OCR's - VL 2B

Qwen2 VL

SigLIP2 042025

Age, Gender, Race and More ..

DomainNet 0324

Multi-Source Domain Adaptation : Infograph, Quickdraw, Real, Sketch, Fashion

Galactic Qwen Exp

Math, Code, General-Purpose

Abliterated Edge Device LLM's

0.5B - 1B - 3B

Math Sage

Mathematical Models, <=58 Avg

Full-entry fine-tuning of SigLIP2

Test finetune

Multimodal Implementations

Comprehensive Demo of Multimodal VLMs on the Hub

Sombrero SM

StreamlinedMemory

r999

general purpose r999 multilingual LLMs

Stable Coders

Coding Logic and Creativity !!

Viper-Hybrid-Coder

reasoning+coding

Image Annotation & Captioning

Trained based on Flux-generated images and their enhanced prompt captioning and annotations in a simple JSON format.

Coder Models Equilibrium 🧑🏻‍💻

Coder Models

Reasoning is all you need 🌠

The garage for 14B and 7B reasoning models, listed based on benchmarks.

Optimus Reasoning

14b, 7b reasoning models

Reasoning Exp Domain Models 🧠

Based on Qwen's 2 vL

Opus Elite 💭

Model contention for gradual increments in reasoning-specific model evaluations.

Express 🚅

Express Tiny LLM's

QwQ Reasoning 🐤

conversational, instruct, 14b Convos < 350 in OpenLLM Evals

Text Generation, Vision Language, Image Generation

Triangulum Series

Reasoning, Long Chain of Thought, Keyword based problem solve

Vision Language Models

Grounding, chat

Stranger Zone Collections [ Org ]

Artificial general intelligence

SD 3.5 Large LoRA

large, medium, turbo lora

Collection Zero & Demo ( Recently Updated )

Image Gen - Text -to-Image

SDXL Dev Models

Are just in initial stage sdxl LoRA

StyleCaption

Collection of Multimodal Models for Captioning -> Experimental

gemma 4 qat

+ quantization-aware training (qat) ggufs for the gemma 4 family of models

gemma 4 gguf

Collection of gemma 4/4[MoE] GGUF

MTP Qwen 3.5/3.6 [MoE] Stable

Collection of Qwen 3.5/3.6 MoE | MTP Featuring GGUF

MTP Qwen 3.5/3.6 Stable

Collection of Qwen 3.5/3.6 MTP Featuring GGUF

QIE/FRIE1.1 — Test LoRAs [May2026]

Collection of Qwen Image Editing LoRAs

Stable Heretic / Multimodal Models

Revised Weights

MiniCPM-V-4.6/5.0 — abliterated/uncensored

Collection of Uncensored Variants of MiniCPM-V-4.6 Multimodal Models

Qwen → GLM-5.1 (Distilled)

Collection of GLM-5.1 math reasoning-distilled, abliteration variants of Qwen3.5/3.6

Qwen → DS V4 Flash (Distilled)

Collection of DeepSeek V4 reasoning-distilled, abliteration variants of Qwen3.5/3.6

CapQwen3.6 - Caption Qwen3.6 Distilled (exp)

Collection of Qwen3.6 Captioning Expert System Models (+abliteration )

Q3.5-Opus-DA

Collection of Opus-Reasoning-Distilled abliteration variants of Qwen3.5

Garnet OCR

Collection of Garnet OCR Models

Auto-Aggressive (Uncensored Qwen 3.6 Native Multimodal)

Collection of Uncensored Variants of Qwen3.6 Native Multimodal Models

Qwen3.6-27B Compressions

Collection of various compressed versions of the Qwen3.6-27B native multimodal model.

Qwen3.6-35B-A3B Compressions

Collection of various compressed versions of the Qwen3.6-35B-A3B native multimodal model.

Gemma 4 Compression(s) - MoE

Collection of Various Compression Schemes for Gemma 4 +MoE Calibration

Gemma 4 Uncensored [MAX] + Compression(s) - [`β ]

Collection of Uncensored Variants of Gemma 4 Native Multimodal Models

Gemma 4 Compression(s)

Collection of Various Compression Schemes for Gemma 4

Gemma-4 F32 GGUF

Collection of Gemma-4 Quants

Qwen3.5 abliterated v2-MAX [BF16]

Version 2 in the Unredacted series, with much lower refusal limits, is available in different compression schemes.

Feb 15–Mar 25 '26

Multimodal Collections Sublist

VGA [* Polaris Series]

capable of accurately locating and understanding 'any' object | State: Experimental, Category: Object Detection, High-depth analysis in visual tasks

Hall of Multimodal Apps

Audio, Image, Image-to-Text

Qwen3.5 Caption [`Gliese Series ]

Expert Image Captioning System

Qwen3.5 Unredacted MAX

“abliteration” models – experimental. / no mtp

Qwen Image Edit [Layout Bbox]

Collection of Region Bounding LoRAs

Qwen3.5 GGUF

Collection of Lightweight Qwen3.5 GGUF Models

Multimodal FP8 Dynamic Collection

Collection of Compressed Multimodal VLM Editions on the Hub [ vLLM / LLM Compressor ]

Qwen 2.5 VL Unredacted MAX

“abliteration” models – experimental.

Qwen Image Edit - AIOs / Compression

Collection of FP8 compression techniques, Diffusers–Transformers converted weights, and more for QIE.

Unredacted MAX - VL [FP8]

[FP8_DYNAMIC] multi-stage trained continual "abliteration" approach for the qwen3-vl series models

Unredacted MAX - VL

multi-stage trained continual "abliteration" approach for the qwen3-vl series models

Unredacted MAX - VL [GGUF]

multi-stage trained continual "abliteration" approach for the qwen3-vl series models (GGUF)

Multimodal Doc Models – Iterations up to [Feb 2026]

Iterations of my doc ocr models timeline of continual training built on top of qwen-vl models.

QIE Jan 23, 26

adapter LoRA developed for Qwen’s Qwen-Image-Edit-2511 image-to-image model

Qwen Image Edit (exps)

adapter LoRA developed for Qwen’s Qwen-Image-Edit-2511 image-to-image model

Qwen Image Edit (Object-Manipulator)

Add or remove the specified objects, flexible for both single-image and multi-image modes.

Physical Long-Horizon Reasoning

SFT, RL

Computer Use Agent [SFT] – December 15, 2025

Localization, Tool Calling, and Navigation

Dynamic markdowns

Document-AI, HTML, OCR

Gliese VLMs Collection

Collection of Gliese series OCR models

Qwen3 VisionCaption

abliterated & [safe and controlled caption generation]

MetaCLIP2 Image Classification Experiments

Domain-Specific Downstream Tasks

Multiple Adapters for Heterogeneous Angles

Gliese OCR Post-x.0

Qwen2.5VL OCR (Model Tree)

Qwen VL Captions

Qwen3VL, Qwen2.5VL

Qwen3-VL Abliteration

Qwen3-VL Abliterated Model Collection [ Version 1.0 ]

General / Caption Abliteration

Qwen2.5-VL and Qwen3-VL

VLM 2.0 - Oct 08'25

Multilingual upstream VLMs - Caption, OCR, Markdown, HTML Render

MM: VLM-Parsing

List of Models for Multimodal VLM Parsing

Gliese Edge

Multimodal VLM and LLM : 'Gliese' Collection of Models

tool calling

agentic reasoning, web search orchestration

Kepler Qwen (27/09)

Abliterated Reasoning & polished token probabilities, enhancing balanced multilingual generation

LZO-1 (Lossless Zoom Operator)

I2I Optimization

UI-Code Web.f

Polished and Controlled Gen

Radiation Qwen/Llama (exp)

Polished Tokens, Improved Multilingual Support, and Controlled Abliteration.

i2i - Kontext (exp)

Photo Restore, Polaroid, Inter-style Mix and more...

VL caption — < Sep 15 ’25

Multimodal models

Caption 3o — filtered

filtered for an sfw captioning set of image pairs.

Gliese-Post1.0

post1.0 ft. camel-doc-ocr-062825

Qwen3 with DeepSeek v3.1 Traces

random event simulations, logical problem analysis, and structured reasoning tasks

DeepCaption attr.

Vision Language Attribution

Illustration Design, Style Intermix

MoD Experimental !

Mixture of Domains

Multimodal VLMs - Aug'25

Reasoning, Document Extraction, Table Reconstruction

Image Generation Apps [Collection]

Flux.1 (Krea, Dev, Schnell), Qwen-Image, Nano-Banana, Qwen-Image-Edit

VLMer Experimental 0818 / 0817

vlm extended reasoning, captioning

VL Abliterated-Caption

qwen2.5-vl : 3b, 7b caption

R1 Traces × Qwen3 Experiment

Vision-Language for Reasoning (VLr) - Lumian2

Flux.1-Krea-Merged-Schnell / Dev

(FLUX.1-schnell + FLUX.1-Krea-dev) | (Flux.1-Dev + Flux.1-Krea-Dev)

Camel-Doc-OCR-080125

Document Retrieval, Content Extraction, and Analysis Recognition.

Flux.1-Krea-Merged-Dev

(Flux.1-Dev + Flux.1-Krea-Dev)

Vision-Language for Reasoning (VLr)

Thinking / No-Thinking

Multimodal VLMs - Until July'25

Multimodal VLMs for Domain-Specific Tasks: OCR, Reasoning, and Captioning

rVL and Captioning Models

Image and video frame understanding, contextual analysis, and natural language reasoning, Abliterated Captioning / Uncensored Image Captioning,

Explora x Poseidon Reasoning

Poseidon Reasoning

(LLM) research, benchmarking, and STEM-focused

Open-Omega : Exp

qwen2.5, 3

Open-Omega: A Fusion of Math, Science, and Coding

reason/scientific thought mixture

Captioning / OCR / DocTable

Mixed Model Operations

Corvus-OCR-Caption-Mix

variable-dimensional pairs dataset

07/11 ~ Visual Understanding

Deep Caption VL - Trained with variable-dimensional pairs

Re-think SmolLM

trl / @exp

Behemoth Tiny Thought [exp vl]

Chain-of-Thought Reasoning, Detailed Image Captioning

SigLIP2 070225, 070125

new exps

Camel Doc MM

enhances document comprehension capabilities

Doc VL

drex [doc ], virex [ video (image ++) exp ]

VisionScope OCR Experimentals

Based on Qwen2.5 VL, Qwen2 VL

Tron Experimentals

Reasoning and Non-Reasoning

Blitzar Experimentals

MoT Experimental Reasoning Traces R1

Mixture-of-Thoughts

Open Voice Classification

audio classifiers

The Prime X

math, code

Open Classification 2805

Disaster Classification Approach Exp

siglip2 base

Cortex Dual ~ DiMind

(direct, reactive, retrieval-based responses), (reasoning, planning, deeper analysis)

Galactic (Sombrero) Opus R1

Code, Math, Moderately Abliterated

Qwen_3 Experimental-2

trajectory

Open classifiers

image detection

Explicit Content Filters

vit, siglip2

Pet It

There are two means of refuge from the miseries of life: music and cats

OpenG

Math, Code, Reasoning – Edge Device LLMs

OpenSDI Diffusion-Generated Image Classification

spotting diffusion-generated images in realistic, open-world scenarios

Multiclass Image Classification 05142025

classification net.

Qwen_3 GGUF

Upto Full precision (float32) version

Qwen_3 Experimental

Cot, LCoT - Reasoning

QWQ_Qwen3 R1 LCoT

CoT, LCoT - QwQ / R1 Math

QWQ_Qwen3 Fine-Thinking

CoT, LCoT - QwQ Math

SigLIP2 05102025

Moderation, Balance, Classifiers

Content Filters SigLIP2/ViT

Moderation, Balance, Classifiers

Core & DocScope OCR Models

DocVQA, RealWorldQA, MTVQA

SigLIP2 Content Filters 052025 Patch 1

Moderation, Balance, Classifiers

Qwen3 Moderate Behavioral Flexibility

moderately abliterated and improved context awareness and moderate behavioral flexibility variant of Qwen3.

SigLIP2 Content Filters 042025 Final

Moderation, Balance, Classifiers

General Optimization Problems [ RPO ] for Reasoning !

Robust Policy Optimization (RPO)

Multilabel Image Classification Datasets

classify one or more scene categories

Edge Device LLM - Advanced RL v3

advanced reinforcement learning (RL)

Edge supervised fine-tuning (SFT)

standard language modeling objective

SigLIP2 Content Filters - Models v.2

Moderation, Balance, Classifiers

Edge LLM - Advanced RL v2

advanced reinforcement learning (RL)

Edge LLM - Advanced RL

advanced reinforcement learning (RL)

Mini Galactic Qwen X RL - Edge Device Models

Coder, Math. [ Reinforcement learning ]

Galactic LLM Exp4

Math, Reasoning

Y.2 Galactic Qwen

Y.2 for Coding and Math

Fashion Product Full Net Experimentals

Usage, Season, BaseColour, ArticleType, SubCategory, MasterCategory, Gender

Galactic LLM Exp 3

Math, Code, General-Purpose

SigLIP2 Content Filters - Models v.1

Moderation, Balance, Classifiers

SigLIP2 Content Filters - Datasets

Moderation, Balance, Classifiers

Multi-Source Experimentals

Domain Specific Classification Models : SigLIP2

Messy OCR's - VL 2B

Qwen2 VL

Clean / Radio : Mono Voice

Names are randomly assigned to the respective TTS model based on gender, and the generated voice will have a radio and clean essence. [ Llama ]

SigLIP2 042025

Age, Gender, Race and More ..

Raptor X

Raptor X: Advanced UI Coder, Reasoning and General-Purpose Coder, and Math LLM

DomainNet 0324

Multi-Source Domain Adaptation : Infograph, Quickdraw, Real, Sketch, Fashion

Galactic Qwen Exp 2

Math, Code, General-Purpose : Multipurpose

Galactic Qwen Exp

Math, Code, General-Purpose

Image Classification Exp 032025

vit, siglip

Abliterated Edge Device LLM's

Mathematical Models, <=58 Avg

Galactic Qwen 14B's Test

Experimentals

Full-entry fine-tuning of SigLIP2

Test finetune

14B Perceived Pattern

Experimentals

Multimodal Implementations

Comprehensive Demo of Multimodal VLMs on the Hub

Sombrero QwQ Elite

Coding, Mathematical Problem Solving

Sombrero SM

StreamlinedMemory

Midranger Experimentals

14b - 32b experimental ft

r999

general purpose r999 multilingual LLMs

Opus + Elite Optimized

general purpose 14b llms

Stable Coders

Coding Logic and Creativity !!

Siglip2 Custom

SiglipForImageClassification

Viper-Hybrid-Coder

reasoning+coding

Deepfake Classification 022025

based on recent dataset

Image Annotation & Captioning

Trained based on Flux-generated images and their enhanced prompt captioning and annotations in a simple JSON format.

Purposeful Conversational Models 🚅

providing assistance

Coder Models Equilibrium 🧑🏻‍💻

Coder Models

Deepfake Quality Assess Datasets

Classification of Good and Bad Quality Deepfakes { Datasets }

Reasoning is all you need 🌠

The garage for 14B and 7B reasoning models, listed based on benchmarks.

Deepfake Quality Assessment

Controllable Classification of Good and Bad Quality Deepfakes

Optimus Reasoning

14b, 7b reasoning models

Bellatrix Tiny 🤏🏻

1.5B, 1B, 0.5B param's

Reasoning Exp Domain Models 🧠

Based on Qwen's 2 vL

Reasoning Saturations🫙

7B, 14B [ Saturation ]

Opus Elite 💭

Model contention for gradual increments in reasoning-specific model evaluations.

phi 4 custom ft [ qwq, lcot, math ]

QwQ Reasoning 🐤

conversational, instruct, 14b Convos < 350 in OpenLLM Evals

PyThagorean 🐍

Solve Math with Python

Calcium-Opus 💬

Opus for chat

LwQ -Llama 🦙

Llama with Questions for reasoning / long chain of thought

Omni Reasoner📒

Multimodality

ChemQwen🧪

ChemQwen Based on InChI

Blaze.1 🔥

Text Generation, Vision Language, Image Generation

GWQ

Open Eval < 500, long chain of thought and keyword based answering, reasoning

Triangulum Series

Reasoning, Long Chain of Thought, Keyword based problem solve

Deepthink and Reasoning

Best for Deepthink and Reasoning

Vision Language Models

Grounding, chat

Custom Build Models

Text Generation Cook with it's GGUF

Stranger Zone Collections [ Org ]

Artificial general intelligence

GPT-Generated Unified Format (GGUF)

ease of reading

SD 3.5 Large LoRA

large, medium, turbo lora

LoRA Space Collections

Flux & Sdxl

FLUX LoRA DLC2

🔥

584

100+ Impressive LoRA's For Flux.1

Collection Zero & Demo ( Recently Updated )

Image Gen - Text -to-Image

Stable Spaces

Only Stable

SDXL Dev Models

Are just in initial stage sdxl LoRA

Flux LoRA Collections

Flux THE LoRA

URL: https://huggingface.co/prithivMLmods/collections

⇱ prithivMLmods (Prithiv Sakthi)

Prithiv Sakthi PRO

AI & ML interests

Recent Activity

Organizations

prithivMLmods 's collections 201

prithivMLmods/vga-qwen3.5-moe-super-gguf

prithivMLmods/visual-grounding-anything

Wan2.2 I2v Fast

Flux.2 Klein Small Decoder

Qwen-Image-Edit-2511-LoRAs-Fast

FireRed Image Edit 1.0 Fast

Multimodal VLM Thinking

FLUX LoRA DLC2

Florence-2 Models

VisionScope-R2

Qwen3 VL HF Demo

Multimodal OCR2

Multimodal Edge Node

GLM OCR Demo

Super OCRs Demo

Multimodal OCR2

Multimodal OCR

VisionScope-R2

FLUX REALISM

Multimodal VLM Thinking

core OCR

Qwen 3.5 HF Demo

Multimodal OCR

Cheers HF Demo

Multimodal OCR

Multimodal VLM Thinking

FLUX LoRA DLC

FLUX LoRA DLC2

FLUX REALISM

prithivMLmods/vga-qwen3.5-moe-super-gguf

prithivMLmods/visual-grounding-anything

Qwen3 VL HF Demo

Multimodal OCR2

Multimodal Edge Node

GLM OCR Demo

Wan2.2 I2v Fast

Flux.2 Klein Small Decoder

Qwen-Image-Edit-2511-LoRAs-Fast

FireRed Image Edit 1.0 Fast

Super OCRs Demo

Multimodal OCR2

Multimodal OCR

VisionScope-R2

FLUX REALISM

Multimodal VLM Thinking

Multimodal VLM Thinking

core OCR

Qwen 3.5 HF Demo

Multimodal OCR

Cheers HF Demo

Multimodal OCR

Multimodal VLM Thinking

FLUX LoRA DLC2

FLUX LoRA DLC

FLUX LoRA DLC2

FLUX REALISM

Florence-2 Models

VisionScope-R2