Suggestions
Kishan Bhoopalam
Founder @ Leadmatic
Kishan Bhoopalam is a professional with a diverse background in data science and entrepreneurship. He is currently a Member at RevGen GENIUS, based in New York City.2 Bhoopalam's career has included several notable positions and achievements:
Professional Experience
Current Roles::
Previous Positions::
Education and Skills
Bhoopalam holds an M.S. in Computer Science with a focus on Machine Learning and Artificial Intelligence.3 He is fluent in English, demonstrating native or bilingual proficiency.3
Online Presence
His LinkedIn profile can be found under the username "kishanb".3 This platform likely provides more detailed information about his professional experiences and connections.
Kishan Bhoopalam's diverse roles in data science and his entrepreneurial ventures, particularly with Leadmatic, suggest he has a strong background in applying advanced analytics and AI technologies in business contexts.
Highlights
🎉 I'm thrilled to announce tinyChat, a new, open-source Large Language Model (LLM) that's under 1% the size of GPT-3.5.
Why make a Large Language Model smaller? They’re supposed to be large after all. We see several reasons the open-source machine learning community can benefit from smaller models.
-
Small models can act as co-agents with larger models like ChatGPT with smaller repetitive tasks distributed to smaller models saving compute costs and creating efficiencies.
-
Smaller architectures create constraints and those constraints create opportunities for exploring more efficient model architectures and novel datasets.
-
We would like to see a future where small, high-performance models could expand access to generative AI and create accessibility. Smaller models could mean tasks are run locally on a device rather than in a centralized cloud infrastructure so the advancement of small efficient models is worth pursuing.
We created this model thanks in part to the rapid advancements in generative AI and open-source machine learning. Leveraging Google’s Flan-T5-Large as our base model, Databricks dolly-15k dataset, and Low-Rank Adaptation of Large Language Models (LoRA), we were able to create a model that improves on Flan-T5’s capabilities and perform tasks like summarization and creative text generation.
This is a v1 with many limitations but we plan on iterating and improving the model over time. Our hope is that future versions of tinyChat play a vital role in the AI ecosystem!
Check out our blog post for a detailed dive into tinyChat's creation, capabilities, and the road ahead: https://lnkd.in/gJKkZ_rn Github Repo: https://lnkd.in/g2pBmgBz Huggingface Model Hub: https://lnkd.in/gi_ej7N3
#tinyChat #AI #NLP #GenerativeAI #OpenSource #MachineLearning
