Kennedy Selvadurai, PhDinLevel Up Codingscapy: A Powerful and Versatile Python Library for Advance Network Packet ManipulationsAnd a cameo by tshark, a command-line packet filter extraordinaire12h ago12h ago
Kennedy Selvadurai, PhDinAI AdvancesImpact of a Knowledge Graph in RAG Powered By CassandraWould LlamaIndex’s ease of knowledge graph generation be panacea for LLM hallucination?Jun 27Jun 27
Kennedy Selvadurai, PhDinAI AdvancesAutomating Instruction Generation off any Document for LLM Fine-TuningUsing Mistral 7B with a novel approach to gain comprehensive coverage of documents for effective instruction generation.Apr 191Apr 191
Kennedy Selvadurai, PhDinAI AdvancesUnveiling the Power of Advanced Retrievers through Visual InsightsIs context from non-redundant nearest neighbors sufficient for LLM to generate accurate responses?Mar 181Mar 181
Kennedy Selvadurai, PhDinAI AdvancesComprehensive Evaluation of Gemma 2B in a RAG setup with Basic and Advanced RetrieversA multi-dimensional evaluation of Gemma vs TinyLlama for resource constrained systemsMar 82Mar 82
Kennedy Selvadurai, PhDinAI AdvancesVisualizing FAISS Vector Space to Understand its Influence on RAG PerformanceVisualizing embeddings using renumics-spotlight reveals useful insights into RAG generation behavior.Feb 264Feb 264
Kennedy Selvadurai, PhDinTowards Data ScienceQuantized Mistral 7B vs TinyLlama for Resource-Constrained SystemsPerformance comparison between these models for accuracy and response time in a RAG question-answering setup.Feb 174Feb 174
Kennedy Selvadurai, PhDEfficient Handling of Ad Hoc JSON Data with Embedded Time SeriesIf you work with different JSON data regularly, workflows here may give hours in a day back.Nov 10, 20231Nov 10, 20231
Kennedy Selvadurai, PhDQuerying Internal Documents using Mistral 7B with Context from an Ensemble RetrieverCould the lost quality of a quantized Mistral 7B model be redeemed with an ensemble retriever?Oct 27, 20233Oct 27, 20233
Kennedy Selvadurai, PhDinDev GeniusObjectively Testing Language Models on Question-Answering PerformanceA more informed way to evaluate current open-source LLMs for use within your corporate walls.Sep 7, 2023Sep 7, 2023