From cff59114459d58984d106973482e969ada7b3eb0 Mon Sep 17 00:00:00 2001 From: Tony Wang <78333580+tonywang10101@users.noreply.github.com> Date: Sun, 22 Oct 2023 08:29:41 +0800 Subject: [PATCH] chore(model): move the model_hub_* files from vdp to model (#374) Because - move the model_hub_* files from vdp to model This commit - remove model hub files here --- model-hub/model_hub_cpu.json | 81 ----------------------------------- model-hub/model_hub_gpu.json | 82 ------------------------------------ 2 files changed, 163 deletions(-) delete mode 100644 model-hub/model_hub_cpu.json delete mode 100644 model-hub/model_hub_gpu.json diff --git a/model-hub/model_hub_cpu.json b/model-hub/model_hub_cpu.json deleted file mode 100644 index f6f48857..00000000 --- a/model-hub/model_hub_cpu.json +++ /dev/null @@ -1,81 +0,0 @@ -[ - { - "id": "mobilenetv2", - "description": "An efficient image classification model, pretrained on ImageNet dataset which contains images from 1,000 classes.", - "task": "TASK_CLASSIFICATION", - "model_definition": "model-definitions/github", - "configuration": { - "repository": "instill-ai/model-mobilenetv2-dvc", - "tag": "v1.0-cpu" - } - }, - { - "id": "vit-base-patch16-224", - "description": "Vision Transformer (ViT) model pretrained on ImageNet-21k and fine-tuned on ImageNet 2012, which contains images from 1,000 classes, at resolution 224x224.", - "task": "TASK_CLASSIFICATION", - "model_definition": "model-definitions/huggingface", - "configuration": { - "repo_id": "google/vit-base-patch16-224" - } - }, - { - "id": "yolov7", - "description": "YOLOv7 is a state-of-the-art real-time object detector pretrained on MS COCO dataset with 80 object classes.", - "task": "TASK_DETECTION", - "model_definition": "model-definitions/github", - "configuration": { - "repository": "instill-ai/model-yolov7-dvc", - "tag": "v1.0-cpu" - } - }, - { - "id": "yolov7-pose", - "description": "a keypoint detector, extended on the basis of YOLOv7, to detect keypoints in the human body. The model is pretrained on MS COCO dataset with 17 keypoints.", - "task": "TASK_KEYPOINT", - "model_definition": "model-definitions/github", - "configuration": { - "repository": "instill-ai/model-yolov7-pose-dvc", - "tag": "v1.0-cpu" - } - }, - { - "id": "ocr-psnet-easyocr", - "description": "An OCR model that combines the PSNet model to localise bounding boxes that contain texts and the EasyOCR model to recognise texts in the detected bounding boxes.", - "task": "TASK_OCR", - "model_definition": "model-definitions/github", - "configuration": { - "repository": "instill-ai/model-ocr-dvc", - "tag": "v1.0-cpu" - } - }, - { - "id": "mask-rcnn", - "description": "Mask R-CNN is a state-of-the-art instance segmentation model, pretrained on MS COCO dataset with 80 object classes.", - "task": "TASK_INSTANCE_SEGMENTATION", - "model_definition": "model-definitions/github", - "configuration": { - "repository": "instill-ai/model-instance-segmentation-dvc", - "tag": "v1.0-cpu" - } - }, - { - "id": "lraspp", - "description": "A semantic segmentation model based on MobileNetV3 from the OpenMMLab semantic segmentation toolbox and benchmark.", - "task": "TASK_SEMANTIC_SEGMENTATION", - "model_definition": "model-definitions/github", - "configuration": { - "repository": "instill-ai/model-semantic-segmentation-dvc", - "tag": "v1.0-cpu" - } - }, - { - "id": "stable-diffusion-1-5-fp32-txt2img", - "description": "Stable Diffusion v2 generates high quality images based on text prompts.", - "task": "TASK_TEXT_TO_IMAGE", - "model_definition": "model-definitions/github", - "configuration": { - "repository": "instill-ai/model-diffusion-dvc", - "tag": "v1.5-cpu" - } - } -] diff --git a/model-hub/model_hub_gpu.json b/model-hub/model_hub_gpu.json deleted file mode 100644 index 4a047764..00000000 --- a/model-hub/model_hub_gpu.json +++ /dev/null @@ -1,82 +0,0 @@ -[ - { - "id": "mobilenetv2", - "description": "An efficient image classification model, pretrained on ImageNet dataset which contains images from 1,000 classes.", - "task": "TASK_CLASSIFICATION", - "model_definition": "model-definitions/github", - "configuration": { - "repository": "instill-ai/model-mobilenetv2-dvc", - "tag": "v1.0-gpu" - } - }, - { - "id": "yolov7", - "description": "YOLOv7 is a state-of-the-art real-time object detector pretrained on MS COCO dataset with 80 object classes.", - "task": "TASK_DETECTION", - "model_definition": "model-definitions/github", - "configuration": { - "repository": "instill-ai/model-yolov7-dvc", - "tag": "v1.0-gpu" - } - }, - { - "id": "yolov7-pose", - "description": "a keypoint detector, extended on the basis of YOLOv7, to detect keypoints in the human body. The model is pretrained on MS COCO dataset with 17 keypoints.", - "task": "TASK_KEYPOINT", - "model_definition": "model-definitions/github", - "configuration": { - "repository": "instill-ai/model-yolov7-pose-dvc", - "tag": "v1.0-gpu" - } - }, - { - "id": "mask-rcnn", - "description": "Mask R-CNN is a state-of-the-art instance segmentation model, pretrained on MS COCO dataset with 80 object classes.", - "task": "TASK_INSTANCE_SEGMENTATION", - "model_definition": "model-definitions/github", - "configuration": { - "repository": "instill-ai/model-instance-segmentation-dvc", - "tag": "v1.0-gpu" - } - }, - { - "id": "stomata-mask-rcnn", - "description": "The instance segmentation model based on Mask R-CNN from Detectron2, fine-tuned on the 'stomata200-mix' dataset with 1 object class 'stomata'.", - "task": "TASK_INSTANCE_SEGMENTATION", - "model_definition": "model-definitions/github", - "configuration": { - "repository": "instill-ai/model-stomata-instance-segmentation-dvc", - "tag": "v2.0-gpu" - } - }, - { - "id": "lraspp", - "description": "A semantic segmentation model based on MobileNetV3 from the OpenMMLab semantic segmentation toolbox and benchmark.", - "task": "TASK_SEMANTIC_SEGMENTATION", - "model_definition": "model-definitions/github", - "configuration": { - "repository": "instill-ai/model-semantic-segmentation-dvc", - "tag": "v1.0-gpu" - } - }, - { - "id": "stable-diffusion-1-5-fp16-txt2img", - "description": "Stable Diffusion v2 generates high quality images based on text prompts.", - "task": "TASK_TEXT_TO_IMAGE", - "model_definition": "model-definitions/github", - "configuration": { - "repository": "instill-ai/model-diffusion-dvc", - "tag": "v1.5-fp16-gpu0" - } - }, - { - "id": "gpt-2", - "description": "GPT-2, from OpenAI, is trained to generate text based on your prompts.", - "task": "TASK_TEXT_GENERATION", - "model_definition": "model-definitions/github", - "configuration": { - "repository": "instill-ai/model-gpt2-megatron-dvc", - "tag": "fp32-345m-1-gpu" - } - } -]