Ritesh Kumar Maurya
Machine Learning Engineer
Open to Work
- Target role
- Machine Learning Engineer
- Availability
- Actively looking
- Experience
- Beginner
- Working preference
- Remote (global)
PythonPyTorchMachine LearningDeep LearningGenerative AILLMsDiffusion ModelsFlow Matching
Qwen2.5 Tool-Calling LoRA
ViewPyTorch, Qwen2.5, LoRA, PEFT, TRL, LLM, SFT, Tool Calling
93.89% tool accuracy and 98.75% valid JSON on 400 held-out samples
nanoFM: Flow Matching with DiT
ViewPyTorch, Flow Matching, DiT, Generative AI, Transformers
End-to-end Flow Matching implementation with DiT, trained from scratch on 5-class image generation
nanoSBGM: Score-Based Generative Modeling
ViewPyTorch, Score Matching, SDE, DiT, Generative AI, ODE
Implemented VE, VP & sub-VP SDEs with both stochastic and probability-flow ODE sampling