Highlights:

Skills:Pytorch, machine Learning, Deep learning, NLP, Large Language model, transformers, Huggingface, DPO, LLM fine-tuning, Prompt Engineering, docker, Django, fastapi, flask, git, aws, kafka, gpt, langchain, langgraph, mysql, mangodb, graphql, selenium, llama2

Data Scientist 09/2023 - current
Krutrim, Bangalore, India
Data preprocessing for pretraning LLM , data visualization and data analysis Working with Mosaic platform for pre-tranining LLM for indic langauges Train our custom tokenizer with different hindic langauge using sentencepiece and huggingface Model inferecing using TensorRT-llm ,Vllm ,huggingface Write Custome Inferencing framework just like Vlllm in fastapi for MPT model Architecture Processing commoncrawl 96 snapshots data using ansible in large scale implement attention mechanism like MHA,GQA,MQA implement inflight
--
Senior Software Engineer 09/2022 - 09/2023
Renewbuy, Gurgaon, India
Industry: Insurance
Fine-tune LLM model with distributed gpus using huggingface and deepspeed(Zeros implemention) Peft lora and full fine-tuning analysis use DPO and RLHF for model Alignment model evaluation using own scripts for MMLU,NLU and other tasks Working with docker and kafka
--
Python Developer 02/2021 - 09/2022
Unfied credit solution, Gurgaon, India
Industry: Finance
Implementing OCR to extract data from pdf files as well captcha breaking with image processing Data Scrapping with selenium and requests ,BeautifulSoup Building API in Django and Maintaining SQL servers Build Restapi with Fastapi and Mongodb from scratch Build ETL data pipeline and automate data entry works
--

Uttrakhand Technical University 08/2018 - 04/2022
Dehradun, , India
Degree: Bachelor's Degree
Major:Computer Science Engineering
NA

Daud Ibrahim Dewan