19.03.2024 aktualisiert


100 % verfügbar
Data Analyst / Scientist
Berlin, Deutschland
Berlin +50 km
M.Sc Data ScienceSkills
APIsKünstliche IntelligenzData AnalysisArchitekturBugfixingClusteranalyseDatenbankenDatensammlungWeb ScrapingData MiningDatenmodellData-ProfilingDatenbank DesignInformationsextraktionPython
- ML/AI
- bug fixing
- Data Science
- Machine Learning
- REST API
- database design
- database
- data analysis
- R
- data modelling
- information extraction
- Team Collaboration
- data engineering
- communication
- data profiling
- Research & Development
- deep-learning
- R Studio
- Decision Tree
- Programming
- Logistic Regression
- Time series
- Predictive modelling
- Cluster analysis
- Text mining
- Data Mining
- Linear regression
- scikit-learn
- Python
- Pandas
- PyTorch
- SQL
- Statistics
- LLMs Prompt Architecture
- LLMs Prompt Design
- LLMs Prompt Engineering
- Numpy
- Hugging Face
- MatPlotLib
- IBM SPSS Statistics
- AWS Glue DataBrew
- AWS Athena
- Rapid Minner
- Jupyter Notebook
Erfahrung
Berufserfahrung0 - 5 Jahre
Freelancer seit2024
Sprachen
DeutschGrundkenntnisseEnglischverhandlungssicher
Projekthistorie
The focus of this thesis is to evaluate the chain-of-thought (CoT) prompting architecture of Large Language Models (LLMs) using one or
more datasets. I used Iterative Chain-of-thought (Iter CoT) for evaluation and also used different LLMs apart from Chat GPT or GPT
models since they were used to analyze the prompt architecture in the original research paper of Iterative Chain-of-thought (Iter CoT). I
utilized the two Stanford Alpaca's low-rank adaptation (LoRA) which are 13B and 30B. Furthermore, I employed two 13B LLMs that
have been fine-tuned from the Meta's LLaMa2 model. In this thesis, I employed pre-trained LLMs and fine-tuned the prompt to gain
better performance from the LLM using the Iter CoT architecture, and I then analyzed the output performance and also performed error
analysis. Evaluation showed that iterative instructions improved accuracy of answer. Alpaca LoRA 13b improved the correctness
accuracy from 44.77% to 92.53% through three revisions. Additionally, this approach reduced the failure rate in answering questions.
The alpaca LoRA 13b had the lowest failure rate of 7.46%. Furthermore, the Iter-CoT method outperformed the manual CoT method,
and the alpaca LoRA 13b and 30b outperformed the ChatGPT model.
more datasets. I used Iterative Chain-of-thought (Iter CoT) for evaluation and also used different LLMs apart from Chat GPT or GPT
models since they were used to analyze the prompt architecture in the original research paper of Iterative Chain-of-thought (Iter CoT). I
utilized the two Stanford Alpaca's low-rank adaptation (LoRA) which are 13B and 30B. Furthermore, I employed two 13B LLMs that
have been fine-tuned from the Meta's LLaMa2 model. In this thesis, I employed pre-trained LLMs and fine-tuned the prompt to gain
better performance from the LLM using the Iter CoT architecture, and I then analyzed the output performance and also performed error
analysis. Evaluation showed that iterative instructions improved accuracy of answer. Alpaca LoRA 13b improved the correctness
accuracy from 44.77% to 92.53% through three revisions. Additionally, this approach reduced the failure rate in answering questions.
The alpaca LoRA 13b had the lowest failure rate of 7.46%. Furthermore, the Iter-CoT method outperformed the manual CoT method,
and the alpaca LoRA 13b and 30b outperformed the ChatGPT model.
This application could be used in the online store or chatbot. In the chatbot, the customer can give a picture of a cloth. The chatbot can
quickly search that cloth in the store and display the exact same or some similar clothes to the customer. Here similarity means the
similar patterns and textures in the clothes. It is a more time saving and effective way of cloth search in the online store using the
chatbots. This also helps customers to make quick and correct decisions through visualisation of exact or similar clothes which have
different colours and textures.
quickly search that cloth in the store and display the exact same or some similar clothes to the customer. Here similarity means the
similar patterns and textures in the clothes. It is a more time saving and effective way of cloth search in the online store using the
chatbots. This also helps customers to make quick and correct decisions through visualisation of exact or similar clothes which have
different colours and textures.
During my data science masters courses, I participated in a datathon in Berlin in 2022. There I worked on datasets of 1 district in Berlin
and later I included this dataset with data of another district in my project of one data science course. I prepared a dataset from those 2datasets of 2 Berlin districts to design parking spots. I illustrated some important findings which are helpful for the district city council
officials.
and later I included this dataset with data of another district in my project of one data science course. I prepared a dataset from those 2datasets of 2 Berlin districts to design parking spots. I illustrated some important findings which are helpful for the district city council
officials.