Filter by type:
Filter by tags:
moondream2-20250414
Moondream is a small vision language model designed to run efficiently everywhere.
Repository: localaiLicense: apache-2.0
smolvlm-256m-instruct
SmolVLM-256M is the smallest multimodal model in the world. It accepts arbitrary sequences of image and text inputs to produce text outputs. It's designed for efficiency. SmolVLM can answer questions about images, describe visual content, or transcribe text. Its lightweight architecture makes it suitable for on-device applications while maintaining strong performance on multimodal tasks. It can run inference on one image with under 1GB of GPU RAM.
Repository: localaiLicense: apache-2.0
smolvlm-500m-instruct
SmolVLM-500M is a tiny multimodal model, member of the SmolVLM family. It accepts arbitrary sequences of image and text inputs to produce text outputs. It's designed for efficiency. SmolVLM can answer questions about images, describe visual content, or transcribe text. Its lightweight architecture makes it suitable for on-device applications while maintaining strong performance on multimodal tasks. It can run inference on one image with 1.23GB of GPU RAM.
Repository: localaiLicense: apache-2.0
smolvlm-instruct
SmolVLM is a compact open multimodal model that accepts arbitrary sequences of image and text inputs to produce text outputs. Designed for efficiency, SmolVLM can answer questions about images, describe visual content, create stories grounded on multiple images, or function as a pure language model without visual inputs. Its lightweight architecture makes it suitable for on-device applications while maintaining strong performance on multimodal tasks.
Repository: localaiLicense: apache-2.0
smolvlm2-2.2b-instruct
SmolVLM2-2.2B is a lightweight multimodal model designed to analyze video content. The model processes videos, images, and text inputs to generate text outputs - whether answering questions about media files, comparing visual content, or transcribing text from images. Despite its compact size, requiring only 5.2GB of GPU RAM for video inference, it delivers robust performance on complex multimodal tasks. This efficiency makes it particularly well-suited for on-device applications where computational resources may be limited.
Repository: localaiLicense: apache-2.0
smolvlm2-500m-video-instruct
SmolVLM2-500M-Video is a lightweight multimodal model designed to analyze video content.
The model processes videos, images, and text inputs to generate text outputs - whether answering questions about media files, comparing visual content, or transcribing text from images. Despite its compact size, requiring only 1.8GB of GPU RAM for video inference, it delivers robust performance on complex multimodal tasks.
This efficiency makes it particularly well-suited for on-device applications where computational resources may be limited.
Repository: localaiLicense: apache-2.0
smolvlm2-256m-video-instruct
SmolVLM2-256M-Video is a lightweight multimodal model designed to analyze video content. The model processes videos, images, and text inputs to generate text outputs - whether answering questions about media files, comparing visual content, or transcribing text from images. Despite its compact size, requiring only 1.38GB of GPU RAM for video inference. This efficiency makes it particularly well-suited for on-device applications that require specific domain fine-tuning and computational resources may be limited.
Repository: localaiLicense: apache-2.0
gemma-3-27b-it
Google/gemma-3-27b-it is an open-source, state-of-the-art vision-language model built from the same research and technology used to create the Gemini models. It is multimodal, handling text and image input and generating text output, with open weights for both pre-trained variants and instruction-tuned variants. Gemma 3 models have a large, 128K context window, multilingual support in over 140 languages, and are available in more sizes than previous versions. They are well-suited for a variety of text generation and image understanding tasks, including question answering, summarization, and reasoning. Their relatively small size makes it possible to deploy them in environments with limited resources such as laptops, desktops or your own cloud infrastructure, democratizing access to state of the art AI models and helping foster innovation for everyone.
Repository: localaiLicense: gemma
sicariussicariistuff_x-ray_alpha
This is a pre-alpha proof-of-concept of a real fully uncensored vision model.
Why do I say "real"? The few vision models we got (qwen, llama 3.2) were "censored," and their fine-tunes were made only to the text portion of the model, as training a vision model is a serious pain.
The only actually trained and uncensored vision model I am aware of is ToriiGate; the rest of the vision models are just the stock vision + a fine-tuned LLM.
Repository: localaiLicense: gemma
planetoid_27b_v.2
This is a merge of pre-trained gemma3 language models
Goal of this merge was to create good uncensored gemma 3 model good for assistant and roleplay, with uncensored vision.
First, vision: i dont know is it normal, but it slightly hallucinate (maybe q3 is too low?), but lack any refusals and otherwise work fine. I used default gemma 3 27b mmproj.
Second, text: it is slow on my hardware, slower than 24b mistral, speed close to 32b QWQ. Model is smart even on q3, responses are adequate in length and are interesting to read. Model is quite attentive to context, tested up to 8k - no problems or degradation spotted. (beware of your typos, it will copy yours mistakes) Creative capabilities are good too, model will create good plot for you, if you let it. Model follows instructions fine, it is really good in "adventure" type of cards. Russian is supported, is not too great, maybe on higher quants is better. Refusals was not encountered.
However, i find this model not unbiased enough. It is close to neutrality, but i want it more "dark". Positivity highly depends on prompts. With good enough cards model can do wonders.
Tested on Q3_K_L, t 1.04.
Repository: localaiLicense: gemma
l3.3-prikol-70b-v0.2
A merge of some Llama 3.3 models because um uh yeah
Went extra schizo on the recipe, hoping for an extra fun result, and... Well, I guess it's an overall improvement over the previous revision. It's a tiny bit smarter, has even more distinct swipes and nice dialogues, but for some reason it's damn sloppy.
I've published the second step of this merge as a separate model, and I'd say the results are more interesting, but not as usable as this one. https://huggingface.co/Nohobby/AbominationSnowPig
Prompt format: Llama3 OR Llama3 Context and ChatML Instruct. It actually works a bit better this way
Repository: localaiLicense: llama3.3
steyrcannon-0.2-qwen2.5-72b
SteyrCannon-0.2 is an updated revision from the original SteyrCannon. This uses EVA-Qwen2.5-72B-v0.2. Nothing else has changed.This model was merged using the TIES merge method using EVA-UNIT-01/EVA-Qwen2.5-72B-v0.2 as a base.
The following models were included in the merge:
anthracite-org/magnum-v4-72b
EVA-UNIT-01/EVA-Qwen2.5-72B-v0.2
Repository: localaiLicense: apache-2.0
qwen_qwen2.5-vl-7b-instruct
In the past five months since Qwen2-VL’s release, numerous developers have built new models on the Qwen2-VL vision-language models, providing us with valuable feedback. During this period, we focused on building more useful vision-language models. Today, we are excited to introduce the latest addition to the Qwen family: Qwen2.5-VL.
Key Enhancements:
Understand things visually: Qwen2.5-VL is not only proficient in recognizing common objects such as flowers, birds, fish, and insects, but it is highly capable of analyzing texts, charts, icons, graphics, and layouts within images.
Being agentic: Qwen2.5-VL directly plays as a visual agent that can reason and dynamically direct tools, which is capable of computer use and phone use.
Understanding long videos and capturing events: Qwen2.5-VL can comprehend videos of over 1 hour, and this time it has a new ability of cpaturing event by pinpointing the relevant video segments.
Capable of visual localization in different formats: Qwen2.5-VL can accurately localize objects in an image by generating bounding boxes or points, and it can provide stable JSON outputs for coordinates and attributes.
Generating structured outputs: for data like scans of invoices, forms, tables, etc. Qwen2.5-VL supports structured outputs of their contents, benefiting usages in finance, commerce, etc.
Model Architecture Updates:
Dynamic Resolution and Frame Rate Training for Video Understanding:
We extend dynamic resolution to the temporal dimension by adopting dynamic FPS sampling, enabling the model to comprehend videos at various sampling rates. Accordingly, we update mRoPE in the time dimension with IDs and absolute time alignment, enabling the model to learn temporal sequence and speed, and ultimately acquire the ability to pinpoint specific moments.
Streamlined and Efficient Vision Encoder
We enhance both training and inference speeds by strategically implementing window attention into the ViT. The ViT architecture is further optimized with SwiGLU and RMSNorm, aligning it with the structure of the Qwen2.5 LLM.
Repository: localaiLicense: apache-2.0
qwen_qwen2.5-vl-72b-instruct
In the past five months since Qwen2-VL’s release, numerous developers have built new models on the Qwen2-VL vision-language models, providing us with valuable feedback. During this period, we focused on building more useful vision-language models. Today, we are excited to introduce the latest addition to the Qwen family: Qwen2.5-VL.
Key Enhancements:
Understand things visually: Qwen2.5-VL is not only proficient in recognizing common objects such as flowers, birds, fish, and insects, but it is highly capable of analyzing texts, charts, icons, graphics, and layouts within images.
Being agentic: Qwen2.5-VL directly plays as a visual agent that can reason and dynamically direct tools, which is capable of computer use and phone use.
Understanding long videos and capturing events: Qwen2.5-VL can comprehend videos of over 1 hour, and this time it has a new ability of cpaturing event by pinpointing the relevant video segments.
Capable of visual localization in different formats: Qwen2.5-VL can accurately localize objects in an image by generating bounding boxes or points, and it can provide stable JSON outputs for coordinates and attributes.
Generating structured outputs: for data like scans of invoices, forms, tables, etc. Qwen2.5-VL supports structured outputs of their contents, benefiting usages in finance, commerce, etc.
Model Architecture Updates:
Dynamic Resolution and Frame Rate Training for Video Understanding:
We extend dynamic resolution to the temporal dimension by adopting dynamic FPS sampling, enabling the model to comprehend videos at various sampling rates. Accordingly, we update mRoPE in the time dimension with IDs and absolute time alignment, enabling the model to learn temporal sequence and speed, and ultimately acquire the ability to pinpoint specific moments.
Streamlined and Efficient Vision Encoder
We enhance both training and inference speeds by strategically implementing window attention into the ViT. The ViT architecture is further optimized with SwiGLU and RMSNorm, aligning it with the structure of the Qwen2.5 LLM.
Repository: localaiLicense: apache-2.0
facebook_kernelllm
We introduce KernelLLM, a large language model based on Llama 3.1 Instruct, which has been trained specifically for the task of authoring GPU kernels using Triton. KernelLLM translates PyTorch modules into Triton kernels and was evaluated on KernelBench-Triton (see here). KernelLLM aims to democratize GPU programming by making kernel development more accessible and efficient.
KernelLLM's vision is to meet the growing demand for high-performance GPU kernels by automating the generation of efficient Triton implementations. As workloads grow larger and more diverse accelerator architectures emerge, the need for tailored kernel solutions has increased significantly. Although a number of works exist, most of them are limited to test-time optimization, while others tune on solutions traced of KernelBench problems itself, thereby limiting the informativeness of the results towards out-of-distribution generalization. To the best of our knowledge KernelLLM is the first LLM finetuned on external (torch, triton) pairs, and we hope that making our model available can accelerate progress towards intelligent kernel authoring systems.
KernelLLM Workflow for Triton Kernel Generation: Our approach uses KernelLLM to translate PyTorch code (green) into Triton kernel candidates. Input and output components are marked in bold. The generations are validated against unit tests, which run kernels with random inputs of known shapes. This workflow allows us to evaluate multiple generations (pass@k) by increasing the number of kernel candidate generations. The best kernel implementation is selected and returned (green output).
The model was trained on approximately 25,000 paired examples of PyTorch modules and their equivalent Triton kernel implementations, and additional synthetically generated samples. Our approach combines filtered code from TheStack [Kocetkov et al. 2022] and synthetic examples generated through torch.compile() and additional prompting techniques. The filtered and compiled dataset is [KernelBook]](https://huggingface.co/datasets/GPUMODE/KernelBook).
We finetuned Llama3.1-8B-Instruct on the created dataset using supervised instruction tuning and measured its ability to generate correct Triton kernels and corresponding calling code on KernelBench-Triton, our newly created variant of KernelBench [Ouyang et al. 2025] targeting Triton kernel generation. The torch code was used with a prompt template containing a format example as instruction during both training and evaluation. The model was trained for 10 epochs with a batch size of 32 and a standard SFT recipe with hyperparameters selected by perplexity on a held-out subset of the training data. Training took circa 12 hours wall clock time on 16 GPUs (192 GPU hours), and we report the best checkpoint's validation results.
Repository: localaiLicense: llama3.1
ultravox-v0_5-llama-3_1-8b
Ultravox is a multimodal Speech LLM built around a pretrained Llama3.1-8B-Instruct and whisper-large-v3-turbo backbone.
See https://ultravox.ai for the GitHub repo and more information.
Ultravox is a multimodal model that can consume both speech and text as input (e.g., a text system prompt and voice user message). The input to the model is given as a text prompt with a special <|audio|> pseudo-token, and the model processor will replace this magic token with embeddings derived from the input audio. Using the merged embeddings as input, the model will then generate output text as usual.
In a future revision of Ultravox, we plan to expand the token vocabulary to support generation of semantic and acoustic audio tokens, which can then be fed to a vocoder to produce voice output. No preference tuning has been applied to this revision of the model.
Repository: localaiLicense: llama3.1
mistralai_mistral-small-3.1-24b-instruct-2503
Building upon Mistral Small 3 (2501), Mistral Small 3.1 (2503) adds state-of-the-art vision understanding and enhances long context capabilities up to 128k tokens without compromising text performance. With 24 billion parameters, this model achieves top-tier capabilities in both text and vision tasks.
This model is an instruction-finetuned version of: Mistral-Small-3.1-24B-Base-2503.
Mistral Small 3.1 can be deployed locally and is exceptionally "knowledge-dense," fitting within a single RTX 4090 or a 32GB RAM MacBook once quantized.
Repository: localaiLicense: apache-2.0
mistralai_devstral-small-2505
Devstral is an agentic LLM for software engineering tasks built under a collaboration between Mistral AI and All Hands AI 🙌. Devstral excels at using tools to explore codebases, editing multiple files and power software engineering agents. The model achieves remarkable performance on SWE-bench which positionates it as the #1 open source model on this benchmark.
It is finetuned from Mistral-Small-3.1, therefore it has a long context window of up to 128k tokens. As a coding agent, Devstral is text-only and before fine-tuning from Mistral-Small-3.1 the vision encoder was removed.
For enterprises requiring specialized capabilities (increased context, domain-specific knowledge, etc.), we will release commercial models beyond what Mistral AI contributes to the community.
Learn more about Devstral in our blog post.
Key Features:
Agentic coding: Devstral is designed to excel at agentic coding tasks, making it a great choice for software engineering agents.
lightweight: with its compact size of just 24 billion parameters, Devstral is light enough to run on a single RTX 4090 or a Mac with 32GB RAM, making it an appropriate model for local deployment and on-device use.
Apache 2.0 License: Open license allowing usage and modification for both commercial and non-commercial purposes.
Context Window: A 128k context window.
Tokenizer: Utilizes a Tekken tokenizer with a 131k vocabulary size.
Repository: localaiLicense: apache-2.0
moondream2
a tiny vision language model that kicks ass and runs anywhere
Repository: localaiLicense: apache-2.0
llava-1.6-vicuna
LLaVA represents a novel end-to-end trained large multimodal model that combines a vision encoder and Vicuna for general-purpose visual and language understanding, achieving impressive chat capabilities mimicking spirits of the multimodal GPT-4 and setting a new state-of-the-art accuracy on Science QA.
Repository: localaiLicense: apache-2.0
llava-1.6-mistral
LLaVA represents a novel end-to-end trained large multimodal model that combines a vision encoder and Vicuna for general-purpose visual and language understanding, achieving impressive chat capabilities mimicking spirits of the multimodal GPT-4 and setting a new state-of-the-art accuracy on Science QA.
Repository: localaiLicense: apache-2.0
Page 1