Create README.md

model_name,model_type,parameters_million,training_dataset_size_million,accuracy_percent,f1_score,inference_latency_ms,memory_usage_mb,power_consumption_watt
VisionNet-S,Computer Vision,45,12,91.2,0.89,18,850,120
VisionNet-L,Computer Vision,120,50,94.5,0.93,32,1450,210
TextPro-B,Transformer NLP,110,300,92.8,0.91,25,1300,190
TextPro-L,Transformer NLP,340,800,95.6,0.95,48,2800,350
AudioSense-T,Audio Classification,60,40,88.9,0.87,20,920,140
AudioSense-X,Audio Classification,150,120,93.4,0.92,37,1600,260
MultiAI-Base,Multimodal,220,500,94.2,0.93,41,2400,310
MultiAI-XL,Multimodal,550,1200,96.8,0.96,75,4200,480
EdgeLite-N,Edge AI,15,5,85.3,0.82,12,400,60
EdgeLite-Pro,Edge AI,28,10,89.7,0.88,15,520,75

README.md

@@ -0,0 +1,42 @@
+---
+license: mit
+task_categories:
+- tabular-classification
+language:
+- en
+tags:
+- artificial-intelligence
+- machine-learning
+- model-benchmark
+- deep-learning
+size_categories:
+- n<1K
+---
+
+# AI Model Performance Benchmark Dataset
+
+## Description
+This dataset contains structured benchmark data of artificial intelligence models across multiple domains including computer vision, natural language processing, audio classification, multimodal systems, and edge AI.
+
+It provides model specifications such as parameter size, training dataset scale, accuracy, F1 score, inference latency, memory usage, and estimated power consumption.
+
+## Columns
+- model_name: Name of the AI model
+- model_type: Category of AI system
+- parameters_million: Number of model parameters (in millions)
+- training_dataset_size_million: Training dataset size (in millions of samples)
+- accuracy_percent: Model accuracy percentage
+- f1_score: F1 performance score
+- inference_latency_ms: Inference speed in milliseconds
+- memory_usage_mb: Runtime memory usage in MB
+- power_consumption_watt: Estimated power consumption
+
+## Purpose
+This dataset is designed for:
+- AI performance benchmarking
+- Machine learning research simulation
+- Educational purposes
+- Model efficiency comparison studies
+
+## Format
+CSV structured dataset