diff --git a/models/cv/object_detection/rtmdet/igie/README.md b/models/cv/object_detection/rtmdet/igie/README.md index cdc20956440176842d94528fccc1309a79729cad..9f7c7f94fd4b725e73e29d520f10ba9c397e10fa 100644 --- a/models/cv/object_detection/rtmdet/igie/README.md +++ b/models/cv/object_detection/rtmdet/igie/README.md @@ -21,15 +21,22 @@ Dataset: - to download the labels dataset. - to download the validation dataset. - to download the train dataset. + - to download the annotations dataset. ```bash unzip -q -d ./ coco2017labels.zip unzip -q -d ./coco/images/ train2017.zip unzip -q -d ./coco/images/ val2017.zip +unzip -q -d ./coco annotations_trainval2017.zip coco ├── annotations -│   └── instances_val2017.json +│ └── instances_train2017.json +│ └── instances_val2017.json +│ └── captions_train2017.json +│ └── captions_val2017.json +│ └── person_keypoints_train2017.json +│ └── person_keypoints_val2017.json ├── images │   ├── train2017 │   └── val2017 diff --git a/models/cv/pose_estimation/hrnetpose/igie/README.md b/models/cv/pose_estimation/hrnetpose/igie/README.md index b3e860b28ce41e43cbea0c9c44b912a1ab3c92a1..565af0e0df80e27194f05b9e53002568663c64af 100644 --- a/models/cv/pose_estimation/hrnetpose/igie/README.md +++ b/models/cv/pose_estimation/hrnetpose/igie/README.md @@ -21,15 +21,22 @@ Dataset: - to download the labels dataset. - to download the validation dataset. - to download the train dataset. + - to download the annotations dataset. ```bash unzip -q -d ./ coco2017labels.zip unzip -q -d ./coco/images/ train2017.zip unzip -q -d ./coco/images/ val2017.zip +unzip -q -d ./coco annotations_trainval2017.zip coco ├── annotations -│   └── instances_val2017.json +│ └── instances_train2017.json +│ └── instances_val2017.json +│ └── captions_train2017.json +│ └── captions_val2017.json +│ └── person_keypoints_train2017.json +│ └── person_keypoints_val2017.json ├── images │   ├── train2017 │   └── val2017 diff --git a/models/cv/pose_estimation/lightweight_openpose/ixrt/README.md b/models/cv/pose_estimation/lightweight_openpose/ixrt/README.md index f6b5c621f124a449fe8f5f8c0bcc8530cbb7b0d2..8ff1df64183540b59455495874e7ecea806b0423 100644 --- a/models/cv/pose_estimation/lightweight_openpose/ixrt/README.md +++ b/models/cv/pose_estimation/lightweight_openpose/ixrt/README.md @@ -19,8 +19,42 @@ inference (no flip or any post-processing done). ### Prepare Resources -- dataset: -- checkpoints: +Pretrained model: + +Dataset: + - to download the labels dataset. + - to download the validation dataset. + - to download the train dataset. + - to download the annotations dataset. + +```bash +unzip -q -d ./ coco2017labels.zip +unzip -q -d ./coco/images/ train2017.zip +unzip -q -d ./coco/images/ val2017.zip +unzip -q -d ./coco annotations_trainval2017.zip + +coco +├── annotations +│ └── instances_train2017.json +│ └── instances_val2017.json +│ └── captions_train2017.json +│ └── captions_val2017.json +│ └── person_keypoints_train2017.json +│ └── person_keypoints_val2017.json +├── images +│   ├── train2017 +│   └── val2017 +├── labels +│   ├── train2017 +│   └── val2017 +├── LICENSE +├── README.txt +├── test-dev2017.txt +├── train2017.cache +├── train2017.txt +├── val2017.cache +└── val2017.txt +``` ### Install Dependencies @@ -50,7 +84,7 @@ onnxsim ./lightweight-human-pose-estimation.pytorch/human-pose-estimation.onnx . ## Model Inference ```bash -export DATASETS_DIR=/Path/to/coco_pose/ +export DATASETS_DIR=/Path/to/coco/ export CHECKPOINTS_DIR=/Path/to/checkpoints/ ``` diff --git a/models/cv/pose_estimation/rtmpose/igie/README.md b/models/cv/pose_estimation/rtmpose/igie/README.md index f13212fecbcef90d4f9bdeb7c7a669d31300b7c2..5cadf5c93e40479a20d9ccd9134fc4e05e5f1543 100644 --- a/models/cv/pose_estimation/rtmpose/igie/README.md +++ b/models/cv/pose_estimation/rtmpose/igie/README.md @@ -21,15 +21,22 @@ Dataset: - to download the labels dataset. - to download the validation dataset. - to download the train dataset. + - to download the annotations dataset. ```bash unzip -q -d ./ coco2017labels.zip unzip -q -d ./coco/images/ train2017.zip unzip -q -d ./coco/images/ val2017.zip +unzip -q -d ./coco annotations_trainval2017.zip coco ├── annotations -│   └── instances_val2017.json +│ └── instances_train2017.json +│ └── instances_val2017.json +│ └── captions_train2017.json +│ └── captions_val2017.json +│ └── person_keypoints_train2017.json +│ └── person_keypoints_val2017.json ├── images │   ├── train2017 │   └── val2017 diff --git a/models/cv/pose_estimation/rtmpose/ixrt/README.md b/models/cv/pose_estimation/rtmpose/ixrt/README.md index 80a5df9b4746749148bbd0214d3b4e1ed4adbef2..3e66c52ec4d91e11a786a1d4cfbc38671b74b2e2 100644 --- a/models/cv/pose_estimation/rtmpose/ixrt/README.md +++ b/models/cv/pose_estimation/rtmpose/ixrt/README.md @@ -21,15 +21,22 @@ Dataset: - to download the labels dataset. - to download the validation dataset. - to download the train dataset. + - to download the annotations dataset. ```bash unzip -q -d ./ coco2017labels.zip unzip -q -d ./coco/images/ train2017.zip unzip -q -d ./coco/images/ val2017.zip +unzip -q -d ./coco annotations_trainval2017.zip coco ├── annotations -│   └── instances_val2017.json +│ └── instances_train2017.json +│ └── instances_val2017.json +│ └── captions_train2017.json +│ └── captions_val2017.json +│ └── person_keypoints_train2017.json +│ └── person_keypoints_val2017.json ├── images │   ├── train2017 │   └── val2017 diff --git a/models/multimodal/vision_language_model/xlmroberta/vllm/offline_inference_embedding.py b/models/multimodal/vision_language_model/xlmroberta/vllm/offline_inference_embedding.py index 88974ed0930e9d3998de456ea491f8453f58d2f2..7ed8f7f2d38529d2fd131428f6451aeafaf1a731 100644 --- a/models/multimodal/vision_language_model/xlmroberta/vllm/offline_inference_embedding.py +++ b/models/multimodal/vision_language_model/xlmroberta/vllm/offline_inference_embedding.py @@ -68,4 +68,5 @@ if __name__ == "__main__": outputs = llm.encode(prompts) # Print the outputs. for output in outputs: - print(output.outputs.embedding) # list of hidden_size floats \ No newline at end of file + print(output.outputs.embedding) # list of hidden_size floats + print("Offline inference is successful!") \ No newline at end of file diff --git a/tests/model_info.json b/tests/model_info.json index 350f81d0e3a8fd8069f09d2e028055052aa29a4f..aa8dbd39e22c4907b1262a2cc744017a40ca788c 100644 --- a/tests/model_info.json +++ b/tests/model_info.json @@ -7213,8 +7213,8 @@ "demoType": "" }, { - "display_name": "LLaVA Next Base", - "model_name": "llava_next_base", + "display_name": "E5-V", + "model_name": "e5-v", "framework": "vllm", "release_version": "25.09", "release_sdk": "4.3.0", @@ -7226,8 +7226,8 @@ "mdims": "", "dataset": "", "license": "", - "model_path": "models/multimodal/vision_language_model/llava_next_base/vllm", - "readme_file": "models/multimodal/vision_language_model/llava_next_base/vllm/README.md", + "model_path": "models/multimodal/vision_language_model/e5-v/vllm", + "readme_file": "models/multimodal/vision_language_model/e5-v/vllm/README.md", "bitbucket_repo": "", "bitbucket_branch": "", "bitbucket_path": "", diff --git a/tests/run_vllm.py b/tests/run_vllm.py index 39964353ea4e2cf069f94a8b3355eaf5f5406f94..6aadb8dc313496f9197865adf05f28b8e2ebccea 100644 --- a/tests/run_vllm.py +++ b/tests/run_vllm.py @@ -144,13 +144,13 @@ def run_nlp_testcase(model): cd ../{model['model_path']} python3 offline_inference.py --model ./llama2-7b --max-tokens 256 -tp 1 --temperature 0.0 """ - elif model_name == "llama3-70b": - script = f""" - set -x - cd ../{model['model_path']} - export CUDA_VISIBLE_DEVICES=0,1,2,3 - python3 offline_inference.py --model ./llama3-70b --max-tokens 256 -tp 4 --temperature 0.0 - """ + # elif model_name == "llama3-70b": # need 8gpus to run + # script = f""" + # set -x + # cd ../{model['model_path']} + # export CUDA_VISIBLE_DEVICES=0,1,2,3 + # python3 offline_inference.py --model ./llama3-70b --max-tokens 256 -tp 4 --temperature 0.0 + # """ elif model_name == "qwen-7b": script = f""" set -x @@ -177,13 +177,13 @@ def run_nlp_testcase(model): export CUDA_VISIBLE_DEVICES=0,1,2,3 python3 offline_inference.py --model ./qwen1.5-32b --max-tokens 256 -tp 4 --temperature 0.0 """ - elif model_name == "qwen1.5-72b": - script = f""" - set -x - cd ../{model['model_path']} - export CUDA_VISIBLE_DEVICES=0,1 - python3 offline_inference.py --model ./qwen1.5-72b --max-tokens 256 -tp 2 --temperature 0.0 --max-model-len 3096 - """ + # elif model_name == "qwen1.5-72b": # need 8gpus to run + # script = f""" + # set -x + # cd ../{model['model_path']} + # export CUDA_VISIBLE_DEVICES=0,1 + # python3 offline_inference.py --model ./qwen1.5-72b --max-tokens 256 -tp 2 --temperature 0.0 --max-model-len 3096 + # """ elif model_name == "qwen2-7b": script = f""" set -x @@ -191,13 +191,13 @@ def run_nlp_testcase(model): export CUDA_VISIBLE_DEVICES=0 python3 offline_inference.py --model ./qwen2-7b --max-tokens 256 -tp 1 --temperature 0.0 """ - elif model_name == "qwen2-72b": - script = f""" - set -x - cd ../{model['model_path']} - export CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 - python3 offline_inference.py --model ./qwen2-72b --max-tokens 256 -tp 8 --temperature 0.0 --gpu-memory-utilization 0.98 --max-model-len 32768 - """ + # elif model_name == "qwen2-72b": # need 8gpus to run + # script = f""" + # set -x + # cd ../{model['model_path']} + # export CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 + # python3 offline_inference.py --model ./qwen2-72b --max-tokens 256 -tp 8 --temperature 0.0 --gpu-memory-utilization 0.98 --max-model-len 32768 + # """ elif model_name == "stablelm": script = f""" set -x @@ -215,13 +215,13 @@ def run_nlp_testcase(model): cd ../{model['model_path']} python3 offline_inference.py --model ./{model_name} --max-tokens 256 -tp {tp} --temperature 0.0 --max-model-len 3096 """ - elif model_name == "aria": - script = f""" - set -x - cd ../{model['model_path']} - export VLLM_ASSETS_CACHE=../vllm/ - python3 offline_inference_vision_language.py --model ./{model_name} --max-tokens 256 -tp 4 --trust-remote-code --temperature 0.0 --dtype bfloat16 --tokenizer-mode slow - """ + # elif model_name == "aria": # need 8gpus to run + # script = f""" + # set -x + # cd ../{model['model_path']} + # export VLLM_ASSETS_CACHE=../vllm/ + # python3 offline_inference_vision_language.py --model ./{model_name} --max-tokens 256 -tp 4 --trust-remote-code --temperature 0.0 --dtype bfloat16 --tokenizer-mode slow + # """ elif model_name == "chameleon_7b" or model_name == "fuyu_8b": script = f""" set -x @@ -314,7 +314,7 @@ def run_nlp_testcase(model): export ENABLE_FLASH_ATTENTION_WITH_HEAD_DIM_PADDING=1 python3 offline_inference_vision_language.py --model ./{model_name} -tp 4 --trust-remote-code --temperature 0.0 --max-token 256 """ - elif model_name == "llava_next_base": + elif model_name == "e5-v": script = f""" set -x cd ../{model['model_path']}