Pinned Loading

1. ft_lab ft_lab Public

   A reproducible research toolkit for systematically evaluating Full Fine-tuning, LoRA and QLoRA methods using the TinyLlama-1.1B model.

   Python 5 1

2. lingwa_mvp lingwa_mvp Public

   Built a multilingual voice-translation MVP using Flutter, Whisper, OpenAI, and TTS with modular controllers and async, cross-platform pipeline.

   C++

3. llm_infra_lab llm_infra_lab Public

   LLM infra with JAX pmap, vLLM-style inference mock, FSDP loop, and K8s/Terraform scaffolds.

   Python 3

4. nvda_stack_agent nvda_stack_agent Public

   Built a government-grade AI agent integrating NVIDIA NIM for fast inference and NeMo Guardrails for strict safety, with FastAPI orchestration and offline fallback.

   Python

5. rlhf_lab rlhf_lab Public

   Unified framework comparing SFT, PPO, DPO, IPO with shared pipelines and metrics for small-GPU RLHF.

   Python

6. reasoning_lab reasoning_lab Public

   Colab-friendly benchmark comparing CoT, SC, RAG×CoT, ReAct, and FSM with unified evaluation and configs.

   Python