I think btlm is cool too — but stablelm seems to work well too afaict. Where on kaggle would CC BY-SA be a problem? Are there some comps that require that solutions must be usable in closed source software?
LLMS
-
AI Model Training on Academic Papers for Student Education
By
–
No, it sounds like you're wrong on the facts of what the model did. It was *literally* trained on academic papers, and *only* returned information already in those papers. The virology students are encouraged to directly find, read, and discuss papers such as these.
-
LoRAX: Serving 100s of Fine-tuned LLMs Efficiently
By
–
LoRAX speaks for your #GPUs! Great seeing our latest innovation for dynamically serving 100s of #finetuned LLMs at the cost of serving 1 #LLM featured on @KDnuggets
. -
Microsoft’s 12-Lesson Generative AI Course for Beginners
By
–
Microsoft Tech Community : Generative AI for Beginners – A 12-Lesson Course – Microsoft Community Hub @Microsoft v/ @lee_stott Ht @theomitsa #GenAI #chatgpt #llm #gpt4 #OenAI #microsoft #promptengineering #GenerativeAI #LLAMA #ai #gpt4 #chatgpt #stats #statistics
-
Pretraining Without Attention: Novel ML Architecture Approach
By
–
oh yeah! That was Pretraining without attention: https://
arxiv.org/abs/2212.10544 -
Tensortream CCO Discusses Open Source, LLMs, Hardware Future
By
–
Thank you to @theinformation and @steph_palazzolo for sitting down with #tenstorrent CCO @DavidBennett__ to talk open source, LLMs and the future of hardware.
-
Groq AI enables faster language user interface for application creation
By
–
#AI powered by @GroqInc will enable a "language user interface" that allows anyone to create AI applications faster than ever before. Stop by booth #1681 at #SC23 to learn how.
-
Building LLM Chatbots with Cohere Chat Endpoint
By
–
Building LLM chatbots can feel overwhelming, but it doesn’t have to be. The Chat endpoint provides a simple API to build LLM chatbots. In this LLM University chapter, learn how to build a chatbot with the Chat endpoint. https://
txt.cohere.com/chatbot-chat-e
ndpoint/
… -
Goodside: high logprobs not cached responses cause joke reuse
By
–
I don’t think this is evidence of cached responses; the logprobs may just be that high. GPT has had issues with joke reuse at least since RLHF (and likely before with FeedMe) — e.g. ChatGPT (GPT-3.5) was found to reuse the same 25 jokes for 90% responses: https://
arxiv.org/abs/2306.04563 -
Developer Streams 30-Day NLLB Model Replication Project
By
–
He is still constantly learning and sharing. Aleksa even recently did a 30-day coding live streaming working on replicating Meta's NLLB (No Language Left Behind approach)