{"id":46244346,"url":"https://github.com/mobilint/mblt-model-zoo","last_synced_at":"2026-05-08T06:21:48.170Z","repository":{"id":285334710,"uuid":"957755378","full_name":"mobilint/mblt-model-zoo","owner":"mobilint","description":"Mobilint Model Zoo Project","archived":false,"fork":false,"pushed_at":"2026-04-22T01:06:35.000Z","size":5681,"stargazers_count":20,"open_issues_count":0,"forks_count":0,"subscribers_count":2,"default_branch":"master","last_synced_at":"2026-04-22T03:03:41.436Z","etag":null,"topics":["ai-accelerators","computer-vision","edge-ai","mobilint","quantization","quantized-neural-networks","transformer"],"latest_commit_sha":null,"homepage":"https://www.mobilint.com/","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"bsd-3-clause","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/mobilint.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":"AGENTS.md","dco":null,"cla":null}},"created_at":"2025-03-31T04:23:31.000Z","updated_at":"2026-04-22T01:05:32.000Z","dependencies_parsed_at":null,"dependency_job_id":"2b2cc1db-867e-4d83-9f83-fbcd607190b7","html_url":"https://github.com/mobilint/mblt-model-zoo","commit_stats":null,"previous_names":["mobilint/mblt-models"],"tags_count":22,"template":false,"template_full_name":null,"purl":"pkg:github/mobilint/mblt-model-zoo","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mobilint%2Fmblt-model-zoo","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mobilint%2Fmblt-model-zoo/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mobilint%2Fmblt-model-zoo/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mobilint%2Fmblt-model-zoo/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/mobilint","download_url":"https://codeload.github.com/mobilint/mblt-model-zoo/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mobilint%2Fmblt-model-zoo/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":32323234,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-26T23:26:28.701Z","status":"online","status_checked_at":"2026-04-27T02:00:06.769Z","response_time":128,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["ai-accelerators","computer-vision","edge-ai","mobilint","quantization","quantized-neural-networks","transformer"],"created_at":"2026-03-03T21:01:58.862Z","updated_at":"2026-04-27T05:06:54.200Z","avatar_url":"https://github.com/mobilint.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Mobilint Model Zoo\n\n\u003cdiv align=\"center\"\u003e\n\u003cp\u003e\n\u003ca href=\"https://www.mobilint.com/\" target=\"_blank\"\u003e\n\u003cimg src=\"https://raw.githubusercontent.com/mobilint/mblt-model-zoo/master/assets/Mobilint_Logo_Primary.png\" alt=\"Mobilint Logo\" width=\"60%\"\u003e\n\u003c/a\u003e\n\u003c/p\u003e\n\u003c/div\u003e\n\n**mblt-model-zoo** is a curated collection of AI models optimized by [Mobilint](https://www.mobilint.com/)’s Neural Processing Units (NPUs).\n\nDesigned to help developers accelerate deployment, Mobilint's Model Zoo offers access to public, pre-trained, and pre-quantized models for vision, language, and multimodal tasks. Along with performance results, we provide pre- and post-processing tools to help developers evaluate, fine-tune, and integrate the models with ease.\n\n## Installation\n\n[![PyPI - Version](https://img.shields.io/pypi/v/mblt-model-zoo?logo=pypi\u0026logoColor=white)](https://pypi.org/project/mblt-model-zoo/)\n[![PyPI Downloads](https://static.pepy.tech/badge/mblt-model-zoo?period=total\u0026units=INTERNATIONAL_SYSTEM\u0026left_color=BLACK\u0026right_color=GREEN\u0026left_text=downloads)](https://clickpy.clickhouse.com/dashboard/mblt-model-zoo)\n[![PyPI - Python Version](https://img.shields.io/pypi/pyversions/mblt-model-zoo?logo=python\u0026logoColor=gold)](https://pypi.org/project/mblt-model-zoo/)\n\n- Prepare environment equipped with Mobilint's NPU. In case you are not a Mobilint customer, please contact [us](mailto:tech-support@mobilint.com).\n- Install **mblt-model-zoo** using pip:\n\n```bash\npip install mblt-model-zoo\n```\n\n- If you want to install the latest version from the source, clone the repository and install it:\n\n```bash\ngit clone https://github.com/mobilint/mblt-model-zoo.git\ncd mblt-model-zoo\npip install -e .\n```\n\n## Quick Start Guide\n\n### Initializing Quantized Model Class\n\n**mblt-model-zoo** provides a quantized model with associated pre- and post-processing tools. The following code snippet shows how to use the pre-trained model for inference.\n\n```python\nfrom mblt_model_zoo.vision import ResNet50\n\n# Load the pre-trained model. \n# Automatically download the model if not found in the local cache.\nresnet50 = ResNet50() \n\n# Load the model trained with a different recipe\n# Currently, the default is \"DEFAULT\", or \"IMAGENET1K_V1\".\nresnet50 = ResNet50(model_type = \"IMAGENET1K_V2\")\n\n# Download the model to local directory and load it\nresnet50 = ResNet50(local_path = \"path/to/local/\") # the file will be downloaded to \"path/to/local/model.mxq\"\n\n# Load the model from a local path or download as filename and file path you want\nresnet50 = ResNet50(local_path = \"path/to/local/model.mxq\")\n\n# Set inference mode for better performance\n# ARIES supports \"single\", \"multi\", \"global4\", and \"global8\" inference mode. Default is \"global8\"\nresnet50 = ResNet50(infer_mode = \"global8\")\n\n# (Beta) If you are holding a model compiled for REGULUS, enable inference on the REGULUS device.\nresnet50 = ResNet50(product = \"regulus\")\n\n# In summary, the model can be loaded with the following arguments. \n# You may customize those arguments to work with Mobilint's NPU.\nresnet50 = ResNet50(\n    local_path = None,\n    model_type = \"DEFAULT\",\n    infer_mode = \"global8\",\n    product = \"aries\",\n)\n\n```\n\n### Working with Quantized Model\n\nWith the image given as a path, PIL image, numpy array, or torch tensor, you can perform inference with the quantized model. The following code snippet shows how to use the quantized model for inference:\n\n```python\nimage_path = \"path/to/image.jpg\"\n\ninput_img = resnet50.preprocess(image_path) # Preprocess the input image\noutput = resnet50(input_img) # Perform inference with the quantized model\nresult = resnet50.postprocess(output) # Postprocess the output\n\nresult.plot(\n    source_path=image_path,\n    save_path=\"path/to/save/result.jpg\",\n)\n```\n\n### Listing Available Models\n\n**mblt-model-zoo** offers a function to list all available models. You can use the following code snippet to list the models for a specific task (e.g., image classification, object detection, etc.):\n\n```python\nfrom mblt_model_zoo.vision import list_models\nfrom pprint import pprint\n\navailable_models = list_models()\npprint(available_models)\n```\n\n## Model List\n\nWe provide the models that are quantized with our advanced quantization techniques. A list of available vision models is [here](mblt_model_zoo/vision/README.md).\n\n## Optional Extras\n\nWhen working with tasks other than vision, extra dependencies may be required. Those options can be installed via `pip install mblt-model-zoo[NAME]` or `pip install -e .[NAME]`.\n\nCurrently, these optional functions are only available on environment equipped with Mobilint's [ARIES](https://www.mobilint.com/aries).\n\n|Name|Use|Details|\n|-------|------|------|\n|transformers|For using HuggingFace transformers related models|[README.md](mblt_model_zoo/hf_transformers/README.md)|\n|MeloTTS|For using MeloTTS models|[README.md](mblt_model_zoo/MeloTTS/README.md)|\n\nFor the `transformers` extra, the repository also includes:\n- functional test instructions in [tests/transformers/TEST.md](tests/transformers/TEST.md)\n- benchmark script usage in [benchmark/transformers/README.md](benchmark/transformers/README.md)\n\n\u003e Note: The `MeloTTS` extra includes `unidic`, which requires an additional dictionary download step. Python packaging (PEP 517/518) does not support running arbitrary post-install commands automatically, so run `mblt-unidic-download` (or `python -m unidic download`) after installing the extra when needed.\n\n## Verbose Option\n\nBy default, model initialization stays quiet. To print the model file size and MD5 hash whenever an MXQ model loads, set the environment variable `MBLT_MODEL_ZOO_VERBOSE` to a truthy value before running your script:\n\n```bash\nexport MBLT_MODEL_ZOO_VERBOSE=true  # accepted values: true/1/yes/on (case-insensitive)\npython your_script.py\n```\n\n### Example Verbose Output\n\n```bash\nModel Initialized\nModel Size: 216.94 MB\nModel Hash: 23c262c43b4c1c453dd0326e249480a0\nDevice Number: 0\nCore Mode: single\nTarget Cores: [CoreId(cluster=Cluster.Cluster0, core=Core.Core0)]\nModel Variant 0\n        Input Shape: [(1, 200, 96), (1, 200, 96), (2, 200, 200)]\n        Output Shape: [(1, 102400, 1)]\nModel Variant 1\n        Input Shape: [(1, 300, 96), (1, 300, 96), (2, 300, 300)]\n        Output Shape: [(1, 153600, 1)]\nModel Variant 2\n        Input Shape: [(1, 400, 96), (1, 400, 96), (2, 400, 400)]\n        Output Shape: [(1, 204800, 1)]\nModel Variant 3\n        Input Shape: [(1, 500, 96), (1, 500, 96), (2, 500, 500)]\n        Output Shape: [(1, 256000, 1)]\nModel Variant 4\n        Input Shape: [(1, 600, 96), (1, 600, 96), (2, 600, 600)]\n        Output Shape: [(1, 307200, 1)]\nModel Variant 5\n        Input Shape: [(1, 900, 96), (1, 900, 96), (2, 900, 900)]\n        Output Shape: [(1, 460800, 1)]\n```\n\nUnset or set the variable to any other value to suppress these messages.\n\n## License\n\nThe Mobilint Model Zoo is released under BSD 3-Clause License. Please see the [LICENSE](https://github.com/mobilint/mblt-model-zoo/blob/master/LICENSE) file for more details.\n\nAdditionally, the license for each model provided in this package follows the terms specified in the source link provided with it.\n\n## Support \u0026 Issues\n\nIf you encounter any problems with this package, please feel free to contact [us](mailto:tech-support@mobilint.com).\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmobilint%2Fmblt-model-zoo","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmobilint%2Fmblt-model-zoo","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmobilint%2Fmblt-model-zoo/lists"}