ViT

数据集目录格式
└─imageNet-1k
   ├─train                # 训练数据集
   └─val                  # 评估数据集

# pretrain
python run_mindformer.py --config ./configs/vit/run_vit_base_p16_224_100ep.yaml --run_mode train

# evaluate
python run_mindformer.py --config ./configs/vit/run_vit_base_p16_224_100ep.yaml --run_mode eval --dataset_dir [DATASET_PATH]

# predict
python run_mindformer.py --config ./configs/vit/run_vit_base_p16_224_100ep.yaml --run_mode predict --predict_data [PATH_TO_IMAGE]

from mindformers import ViTForImageClassification, ViTConfig

ViTForImageClassification.show_support_list()
# 输出：
# - support list of ViTForImageClassification is:
# -    ['vit_base_p16']
# - -------------------------------------

# 模型标志加载模型
model = ViTForImageClassification.from_pretrained("vit_base_p16")

#模型配置加载模型
config = ViTConfig.from_pretrained("vit_base_p16")
# {'patch_size': 16, 'in_chans': 3, 'embed_dim': 768, 'depth': 12, 'num_heads': 12, 'mlp_ratio': 4,
# ..., 'batch_size': 32, 'image_size': 224, 'num_classes': 1000}
model = ViTForImageClassification(config)

from mindformers.trainer import Trainer
from mindformers.tools.image_tools import load_image

# 初始化任务
vit_trainer = Trainer(
    task='image_classification',
    model='vit_base_p16',
    train_dataset="imageNet-1k/train",
    eval_dataset="imageNet-1k/val")
img = load_image("https://ascend-repo-modelzoo.obs.cn-east-2.myhuaweicloud.com/XFormer_for_mindspore/clip/sunflower.png")

# 方式1：使用现有的预训练权重进行finetune， 并使用finetune获得的权重进行eval和推理
vit_trainer.train(resume_or_finetune_from_checkpoint="mae_vit_base_p16", do_finetune=True)
vit_trainer.evaluate(eval_checkpoint=True)
predict_result = vit_trainer.predict(predict_checkpoint=True, input_data=img, top_k=3)
print(predict_result)

# 方式2: 重头开始训练，并使用训练好的权重进行eval和推理
vit_trainer.train()
vit_trainer.evaluate(eval_checkpoint=True)
predict_result = vit_trainer.predict(predict_checkpoint=True, input_data=img, top_k=3)
print(predict_result)

# 方式3： 从obs下载训练好的权重并进行eval和推理
vit_trainer.evaluate()
predict_result = vit_trainer.predict(input_data=img, top_k=3)
print(predict_result)

from mindformers.pipeline import pipeline
from mindformers.tools.image_tools import load_image


pipeline_task = pipeline("image_classification", model='vit_base_p16')
img = load_image("https://ascend-repo-modelzoo.obs.cn-east-2.myhuaweicloud.com/XFormer_for_mindspore/clip/sunflower.png")
pipeline_result = pipeline_task(img, top_k=3)

python mindformers/models/vit/convert_weight.py --torch_path "PATH OF ViT-Base.pth" --mindspore_path "SAVE PATH OF vit_base_p16.ckpt"