Table 5 Cloud deployment and model training configuration.
Parameter | Configuration |
|---|---|
Compute node | AWS EC2 (t3.large), 8 vCPU, 32 GB RAM |
GPU (attached) | NVIDIA T4 (16 GB) |
Operating system | Ubuntu 20.04 LTS |
Orchestration layer | Kubernetes 1.28 |
Messaging middleware | RabbitMQ 3.12 |
Monitoring stack | Prometheus + Grafana |
ASR optimizer | Adam, learning rate \(1\times 10^{-3}\) |
ASR training epochs | 30 (early stopping enabled) |
Embedding model | MobileBERT (zero-shot) |
Sentiment engine | VADER (compound polarity) |
Toxicity module | Perspective API + behavioral heuristics |
