Ishan Trivedi
LLM Reliability · Evaluation · Fine-Tuning · Production AI
- 1
- Followers
AudioForge — Parameter-Efficient Audio Classification
Built and evaluated an audio-classification pipeline on FSD50K, covering 36K+ training clips across 200 multi-label sound categories.
I compared a CNN trained from scratch against a pretrained Audio Spectrogram Transformer adapted using LoRA.
The LoRA setup trained only 0.52% of the AST’s 86.8M parameters while improving mean Average Precision from 0.302 to 0.557 — an 84% improvement.
Beyond model training, I handled the GPU infrastructure end-to-end: VRAM sizing, AWS EC2 provisioning, quota management, spot-instance fallback, cost-controlled start/stop workflows, and debugging failed training runs with smoke tests before full experiments.
What I worked on
Audio Spectrogram Transformer fine-tuning
LoRA / parameter-efficient training
Multi-label audio classification
PyTorch training and evaluation pipelines
GPU memory and infrastructure optimization
Experiment debugging and validation
Stack: Python · PyTorch · Hugging Face · AWS EC2
0
31
Autonomous AI red-team platform. Point it at any HTTP AI agent — a 5-agent
LangGraph pipeline attacks it, finds vulnerabilities, proposes defenses, and
streams results live to a React dashboard.
0
95
Built a production-oriented AI platform that accelerates engineering incident investigations using Retrieval-Augmented Generation (RAG). The system ingests engineering documentation, logs, runbooks, and deployment data to generate citation-backed answers for faster root cause analysis.
0
117
Fine-tuned a Qwen2.5-7B language model using QLoRA to automatically analyze system logs and produce structured incident reports containing root cause, severity, confidence score, and recommended fixes.
PyTorch • Hugging Face • PEFT • QLoRA • AWS EC2 • Transformers
0
159