zuzett/lfm2.5-vl-450m-gguf overview

{
  "metadata": {},
  "card_data": {
    "library_name": "transformers",
    "license": "other",
    "license_name": "lfm1.0",
    "license_link": "LICENSE",
    "language": [
      "en",
      "ja",
      "ko",
      "fr",
      "es",
      "de",
      "ar",
      "zh",
      "pt"
    ],
    "pipeline_tag": "image-text-to-text",
    "tags": [
      "liquid",
      "lfm2",
      "lfm2-vl",
      "edge",
      "lfm2.5-vl",
      "lfm2.5"
    ],
    "base_model": "LiquidAI/LFM2.5-350M",
    "frontmatter": {
      "library_name": "transformers",
      "license": "other",
      "license_name": "lfm1.0",
      "license_link": "LICENSE",
      "language": [
        "en",
        "ja",
        "ko",
        "fr",
        "es",
        "de",
        "ar",
        "zh",
        "pt"
      ],
      "pipeline_tag": "image-text-to-text",
      "tags": [
        "liquid",
        "lfm2",
        "lfm2-vl",
        "edge",
        "lfm2.5-vl",
        "lfm2.5"
      ],
      "base_model": "LiquidAI/LFM2.5-350M"
    },
    "hero_image_url": "https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/2b08LKpev0DNEk6DlnWkY.png",
    "summary": "LFM2.5‑VL-450M is Liquid AI's refreshed version of the first vision-language model, LFM2-VL-450M, built on an updated backbone LFM2.5-350M and tuned for stronger real-world performance. Find more about LFM2.5 family of models in our blog post. * **Enhanced instruction following** on vision and language tasks. * **Improved multilingual vision understanding** in Arabic, Chinese, French, German, Japanese, Korean, Portuguese and Spanish. * **Bounding box prediction and object detection** for grounded visual understanding. * **Function calling support** for text-only input. 🎥⚡️ You can try LFM2.5-VL-450M running locally in your browser with our real-time video stream captioning WebGPU demo 🎥⚡️ Alternatively, try the API model on the Playground.",
    "quick_links": [],
    "benchmark_table_html": "",
    "readme_markdown": "---\nlibrary_name: transformers\nlicense: other\nlicense_name: lfm1.0\nlicense_link: LICENSE\nlanguage:\n- en\n- ja\n- ko\n- fr\n- es\n- de\n- ar\n- zh\n- pt\npipeline_tag: image-text-to-text\ntags:\n- liquid\n- lfm2\n- lfm2-vl\n- edge\n- lfm2.5-vl\n- lfm2.5\nbase_model: LiquidAI/LFM2.5-350M\n---\n\nUsing <a href=\"https://github.com/ggerganov/llama.cpp/\">llama.cpp</a> release <a href=\"https://github.com/ggerganov/llama.cpp/releases/tag/b8763\">b8763</a> for quantization.\nComputing over 400 chunks, n_ctx=1024\n\n<center>\n<div style=\"text-align: center;\">\n  <img \n    src=\"https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/2b08LKpev0DNEk6DlnWkY.png\" \n    alt=\"Liquid AI\"\n    style=\"width: 100%; max-width: 100%; height: auto; display: inline-block; margin-bottom: 0.5em; margin-top: 0.5em;\"\n  />\n</div>\n<div style=\"display: flex; justify-content: center; gap: 0.5em;\">\n<a href=\"https://playground.liquid.ai/chat?model=lfm2.5-vl-450m\"><strong>Try LFM</strong></a> • <a href=\"https://docs.liquid.ai/lfm/getting-started/welcome\"><strong>Docs</strong></a> • <a href=\"https://leap.liquid.ai/\"><strong>LEAP</strong></a> • <a href=\"https://discord.com/invite/liquid-ai\"><strong>Discord</strong></a>\n</div>\n</center>\n\n<br>\n\n# LFM2.5‑VL-450M\n\nLFM2.5‑VL-450M is [Liquid AI](https://www.liquid.ai/)'s refreshed version of the first vision-language model, [LFM2-VL-450M](https://huggingface.co/LiquidAI/LFM2-VL-450M), built on an updated backbone [LFM2.5-350M](https://huggingface.co/LiquidAI/LFM2.5-350M) and tuned for stronger real-world performance. Find more about LFM2.5 family of models in our [blog post](http://www.liquid.ai/blog/lfm2-5-vl-450m).\n\n* **Enhanced instruction following** on vision and language tasks.\n* **Improved multilingual vision understanding** in Arabic, Chinese, French, German, Japanese, Korean, Portuguese and Spanish.\n* **Bounding box prediction and object detection** for grounded visual understanding.\n* **Function calling support** for text-only input.\n  \n🎥⚡️ You can try LFM2.5-VL-450M running locally in your browser with our real-time video stream captioning [WebGPU demo](https://huggingface.co/spaces/LiquidAI/LFM2.5-VL-450M-WebGPU) 🎥⚡️ \n\nAlternatively, try the API model on the [Playground](https://playground.liquid.ai/chat?model=lfm2.5-vl-450m).\n\n## 📄 Model details\n\nLFM2.5-VL-450M is a general-purpose vision-language model with the following features:\n\n- **LM Backbone**: LFM2.5-350M\n- **Vision encoder**: SigLIP2 NaFlex shape‑optimized 86M\n- **Context length**: 32,768 tokens\n- **Vocabulary size**: 65,536\n- **Languages**: English, Arabic, Chinese, French, German, Japanese, Korean, Portuguese, and Spanish\n- **Native resolution processing**: handles images up to 512*512 pixels without upscaling and preserves non-standard aspect ratios without distortion\n- **Tiling strategy**: splits large images into non-overlapping 512×512 patches and includes thumbnail encoding for global context\n- **Inference-time flexibility**: user-tunable maximum image tokens and tile count for speed/quality tradeoff without retraining\n- **Generation parameters**: \n  - text: `temperature=0.1`, `min_p=0.15`, `repetition_penalty=1.05`\n  - vision: `min_image_tokens=32` `max_image_tokens=256`, `do_image_splitting=True`\n\n| Model | Description |\n|-------|-------------|\n| [**LFM2.5-VL-450M**](https://huggingface.co/LiquidAI/LFM2.5-VL-450M) | Original model checkpoint in native format. Best for fine-tuning or inference with Transformers and vLLM. |\n| [LFM2.5-VL-450M-GGUF](https://huggingface.co/LiquidAI/LFM2.5-VL-450M-GGUF) | Quantized format for llama.cpp and compatible tools. Optimized for CPU inference and local deployment with reduced memory usage. |\n| [LFM2.5-VL-450M-ONNX](https://huggingface.co/LiquidAI/LFM2.5-VL-450M-ONNX) | ONNX Runtime format for cross-platform deployment. Enables hardware-accelerated inference across diverse environments (cloud, edge, mobile). |\n| [LFM2.5-VL-450M-MLX-8bit](https://huggingface.co/LiquidAI/LFM2.5-VL-450M-MLX-8bit) | MLX format for Apple Silicon. Optimized for fast on-device inference on Mac with [mlx-vlm](https://github.com/Blaizzy/mlx-vlm). Also available in [4bit](https://huggingface.co/LiquidAI/LFM2.5-VL-450M-MLX-4bit), [5bit](https://huggingface.co/LiquidAI/LFM2.5-VL-450M-MLX-5bit), [6bit](https://huggingface.co/LiquidAI/LFM2.5-VL-450M-MLX-6bit), and [bf16](https://huggingface.co/LiquidAI/LFM2.5-VL-450M-MLX-bf16). |\n\nWe recommend using it for general vision-language workloads, captioning and object detection. It’s not well-suited for knowledge-intensive tasks or fine-grained OCR.\n\n### Chat Template\n\nLFM2.5-VL uses a ChatML-like format. See the [Chat Template documentation](https://docs.liquid.ai/lfm/key-concepts/chat-template#vision-models) for details.\n\n```\n<|startoftext|><|im_start|>system\nYou are a helpful multimodal assistant by Liquid AI.<|im_end|>\n<|im_start|>user\n<image>Describe this image.<|im_end|>\n<|im_start|>assistant\nThis image shows a Caenorhabditis elegans (C. elegans) nematode.<|im_end|>\n```\n\nYou can use [`processor.apply_chat_template()`](https://huggingface.co/docs/transformers/en/chat_templating_multimodal) to format your messages automatically.\n\n## 🏃 Inference\n\nYou can run LFM2.5-VL-450M with Hugging Face [`transformers`](https://github.com/huggingface/transformers) v5.1 or newer:\n\n```bash\npip install transformers pillow\n```\n\n```python\nfrom transformers import AutoProcessor, AutoModelForImageTextToText\nfrom transformers.image_utils import load_image\n\n# Load model and processor\nmodel_id = \"LiquidAI/LFM2.5-VL-450M\"\nmodel = AutoModelForImageTextToText.from_pretrained(\n    model_id,\n    device_map=\"auto\",\n    dtype=\"bfloat16\"\n)\nprocessor = AutoProcessor.from_pretrained(model_id)\n\n# Load image and create conversation\nurl = \"https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg\"\nimage = load_image(url)\nconversation = [\n    {\n        \"role\": \"user\",\n        \"content\": [\n            {\"type\": \"image\", \"image\": image},\n            {\"type\": \"text\", \"text\": \"What is in this image?\"},\n        ],\n    },\n]\n\n# Generate Answer\ninputs = processor.apply_chat_template(\n    conversation,\n    add_generation_prompt=True,\n    return_tensors=\"pt\",\n    return_dict=True,\n    tokenize=True,\n).to(model.device)\noutputs = model.generate(**inputs, max_new_tokens=64)\nprocessor.batch_decode(outputs, skip_special_tokens=True)[0]\n\n# This image captures the iconic Statue of Liberty standing majestically on Liberty Island in New York City. The statue, a symbol of freedom and democracy, is prominently featured in the foreground, its greenish-gray hue contrasting beautifully with the surrounding water.\n```\n\n### Visual grounding\n\nLFM2.5-VL-450M supports bounding box prediction:\n\n```python\nurl = \"https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg\"\nimage = load_image(url)\nquery = \"status\"\nprompt = f'Detect all instances of: {query}. Response must be a JSON array: [{\"label\": ..., \"bbox\": [x1, y1, x2, y2]}, ...]. Coordinates are normalized to [0,1].'\n\nconversation = [\n    {\n        \"role\": \"user\",\n        \"content\": [\n            {\"type\": \"image\", \"image\": image},\n            {\"type\": \"text\", \"text\": prompt},\n        ],\n    },\n]\n\n# Generate Answer\ninputs = processor.apply_chat_template(\n    conversation,\n    add_generation_prompt=True,\n    return_tensors=\"pt\",\n    return_dict=True,\n    tokenize=True,\n).to(model.device)\noutputs = model.generate(**inputs, max_new_tokens=64)\nprocessor.batch_decode(outputs, skip_special_tokens=True)[0]\n\n# [{\"label\": \"statue\", \"bbox\": [0.3, 0.25, 0.4, 0.65]}]\n```\n\n### Tool Use\n\nLFM2.5 supports function calling for text only input by applying the chat template with the tokenizer. See the [Tool Use documentation](https://docs.liquid.ai/lfm/key-concepts/tool-use) for the full guide.\n\n```python\ntools = [{\n    \"name\": \"get_weather\",\n    \"description\": \"Get current weather for a location\",\n    \"parameters\": {\n        \"type\": \"object\",\n        \"properties\": {\"location\": {\"type\": \"string\"}},\n        \"required\": [\"location\"]\n    }\n}]\n\nmessages = [{\"role\": \"user\", \"content\": \"What's the weather in Paris?\"}]\n\n# Apply chat template with tools\ninputs = processor.tokenizer.apply_chat_template(\n    messages,\n    tools=tools,\n    add_generation_prompt=True,\n    return_tensors=\"pt\",\n    return_dict=True,\n)\ninput_ids = inputs[\"input_ids\"].to(model.device)\noutputs = model.generate(input_ids, max_new_tokens=256)\nresponse = processor.tokenizer.decode(outputs[0, input_ids.shape[1]:], skip_special_tokens=False)\n\n# <|tool_call_start|>[get_weather(location=\"Paris\")]<|tool_call_end|>I am retrieving the current weather for Paris.<|im_end|>\n```\n\n| Name | Description | Docs | Notebook |\n|------|-------------|------|----------|\n| [Transformers](https://github.com/huggingface/transformers) | Simple inference with direct access to model internals. | <a href=\"https://docs.liquid.ai/lfm/inference/transformers#vision-models\">Link</a>| <a href=\"https://colab.research.google.com/drive/1WVQpf4XrHgHFkP0FnlZfx2nK8PugvQNZ?usp=sharing\"><img src=\"https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/vlOyMEjwHa_b_LXysEu2E.png\" width=\"110\" alt=\"Colab link\"></a> |\n| [vLLM](https://github.com/vllm-project/vllm) | High-throughput production deployments with GPU. | <a href=\"https://docs.liquid.ai/deployment/gpu-inference/vllm#vision-models\">Link</a> | <a href=\"https://colab.research.google.com/drive/1sUfQlqAvuAVB4bZ6akYVQPGmHtTDUNpF?usp=sharing\"><img src=\"https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/vlOyMEjwHa_b_LXysEu2E.png\" width=\"110\" alt=\"Colab link\"></a> |\n| [SGLang](https://github.com/vllm-project/vllm) | High-throughput production deployments with GPU. | <a href=\"https://docs.liquid.ai/deployment/gpu-inference/sglang#vision-models\">Link</a> | <a href=\"https://colab.research.google.com/drive/1qJlAFag223yFOZGzuMIkYUFhybM9ao5g?usp=sharing\"><img src=\"https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/vlOyMEjwHa_b_LXysEu2E.png\" width=\"110\" alt=\"Colab link\"></a> |\n| [llama.cpp](https://github.com/ggml-org/llama.cpp) | Cross-platform inference with CPU offloading. | <a href=\"https://docs.liquid.ai/lfm/inference/llama-cpp#vision-models\">Link</a> | <a href=\"https://colab.research.google.com/drive/1q2PjE6O_AahakRlkTNJGYL32MsdUcj7b?usp=sharing\"><img src=\"https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/vlOyMEjwHa_b_LXysEu2E.png\" width=\"110\" alt=\"Colab link\"></a> |\n\n## 🔧 Fine-tuning\n\nWe recommend fine-tuning LFM2.5-VL-450M model on your use cases to maximize performance.\n\n| Notebook  | Description                                                          | Link |\n|-----------|----------------------------------------------------------------------|------|\n| SFT (Unsloth) | Supervised Fine-Tuning with LoRA using Unsloth. | <a href=\"https://colab.research.google.com/drive/1FaR2HSe91YDe88TG97-JVxMygl-rL6vB?usp=sharing\"><img src=\"https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/vlOyMEjwHa_b_LXysEu2E.png\" width=\"110\" alt=\"Colab link\"></a> |\n| SFT (TRL) | Supervised Fine-Tuning with LoRA using TRL. | <a href=\"https://colab.research.google.com/drive/10530_jt_Joa5zH2wgYlyXosypq1R7PIz?usp=sharing\"><img src=\"https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/vlOyMEjwHa_b_LXysEu2E.png\" width=\"110\" alt=\"Colab link\"></a> |\n\n\n## 📊 Performance\n\nLFM2.5-VL-450M improves over LFM2-VL-450M across both vision and language benchmarks, while also adding two new capabilities: bounding box prediction on RefCOCO-M and function calling support measured by BFCLv4.\n\n### Vision benchmarks\n\n| Model              | MMStar | RealWorldQA | MMBench (dev en) | MMMU (val) | POPE | MMVet | BLINK | InfoVQA (val) | OCRBench | MM-IFEval | MMMB | CountBench | RefCOCO-M |\n|--------------------|--------|-------------|------------------|------------|------|-------|-------|---------------|----------|------------|------|------------|-----------|\n| **LFM2.5-VL-450M** | 43.00  | 58.43       | 60.91            | 32.67      | 86.93| 41.10 | 43.92 | 43.02         | 684      | 45.00      | 68.09| 73.31      | 81.28     |\n| LFM2-VL-450M       | 40.87  | 52.03       | 56.27            | 34.44      | 83.79| 33.85 | 42.61 | 44.56         | 657      | 33.09      | 54.29| 47.64      | -         |\n| SmolVLM2-500M      | 38.20  | 49.90       | 52.32            | 34.10      | 82.67| 29.90 | 40.70 | 24.64         | 609      | 11.27      | 46.79| 61.81      | -         |\n\nAll vision benchmark scores are obtained using [VLMEvalKit](https://github.com/open-compass/VLMEvalKit). Multilingual scores are based on the average of benchmarks translated by GPT-4.1-mini from English to Arabic, Chinese, French, German, Japanese, Korean, Portuguese, and Spanish.\n\n### Language benchmarks\n\n| Model              | GPQA | MMLU Pro | IFEval | Multi-IF | BFCLv4 |\n|--------------------|------|----------|--------|----------|--------|\n| **LFM2.5-VL-450M** | 25.66| 19.32    | 61.16  | 34.63    | 21.08  |\n| LFM2-VL-450M       | 23.13| 17.22    | 51.75  | 26.21    | -      |\n| SmolVLM2-500M      | 23.84| 13.57    | 30.14  | 6.82     | -      |\n\n## 📬 Contact\n\n- Got questions or want to connect? [Join our Discord community](https://discord.com/invite/liquid-ai)\n- If you are interested in custom solutions with edge deployment, please contact [our sales team](https://www.liquid.ai/contact).\n\n## Citation\n\n```\n@article{liquidai2025lfm2,\n title={LFM2 Technical Report},\n author={Liquid AI},\n journal={arXiv preprint arXiv:2511.23404},\n year={2025}\n}\n```",
    "related_quantizations": []
  },
  "tags": [
    "transformers",
    "gguf",
    "liquid",
    "lfm2",
    "lfm2-vl",
    "edge",
    "lfm2.5-vl",
    "lfm2.5",
    "image-text-to-text",
    "en",
    "ja",
    "ko",
    "fr",
    "es",
    "de",
    "ar",
    "zh",
    "pt",
    "arxiv:2511.23404",
    "base_model:LiquidAI/LFM2.5-350M",
    "base_model:quantized:LiquidAI/LFM2.5-350M",
    "license:other",
    "endpoints_compatible",
    "region:us",
    "conversational"
  ],
  "likes": 0,
  "downloads": 1000,
  "gated": false,
  "private": false,
  "last_modified": "2026-04-15T08:31:41.000Z",
  "created_at": "2026-04-15T08:03:33.000Z",
  "pipeline_tag": "image-text-to-text",
  "library_name": "transformers"
}

{
  "_id": "69df465507f4e6a31f836310",
  "id": "ZuzeTt/LFM2.5-VL-450M-GGUF",
  "modelId": "ZuzeTt/LFM2.5-VL-450M-GGUF",
  "sha": "7028ccdbf6fc03e20e91d99aa7c6cae38ff93a4d",
  "createdAt": "2026-04-15T08:03:33.000Z",
  "lastModified": "2026-04-15T08:31:41.000Z",
  "author": "ZuzeTt",
  "downloads": 1000,
  "likes": 0,
  "gated": false,
  "private": false,
  "pipeline_tag": "image-text-to-text",
  "library_name": "transformers",
  "siblings_count": 32
}