videomae-base-finetuned-ucf101-subset

from transformers import VideoMAEFeatureExtractor, VideoMAEForVideoClassification
import torch

model = VideoMAEForVideoClassification.from_pretrained('sahild23d23e23re23r/videomae-base-finetuned-ucf101-subset')
feature_extractor = VideoMAEFeatureExtractor.from_pretrained('sahild23d23e23re23r/videomae-base-finetuned-ucf101-subset')

# Suponiendo que tienes el video en forma de fotogramas
frames = [...] 
inputs = feature_extractor(frames, return_tensors="pt")

with torch.no_grad():
    logits = model(**inputs).logits

predicted_label = logits.argmax(-1).item()
print('Etiqueta predicha:', predicted_label)

Parámetros de entrenamiento

• learning_rate: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam con betas=(0.9,0.999) y epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 28

Versiones del framework

• Transformers 4.34.0
• Pytorch 2.1.0
• Datasets 2.12.0
• Tokenizers 0.14.1