The models that end in dates are specific models that don’t change. The ones that don’t end in dates are pointers to the latest recommended version and do change.
LLMS
-
Fluency Does Not Equal Intelligence in Large Language Models
By
–
Good essay on the fallacy of assuming that LLMs are intelligent because they are fluent.
-
LoRAX v0.2: Sparse SGMV and Tensor Parallel Inference
By
–
Announcing LoRAX v0.2 Sparse SGMV: vectorize LoRA and base model requests in same batch Tensor Parallel SGMV: multi-GPU, multi-LoRA vectorized inference ExLlama v2 kernels for faster GPT-Q (thanks Florian Zimmermeister!)
…and more! https://
pbase.ai/3T2Nemt -
EPFL’s Llama-2 Medical Fine-Tuned Model Released
By
–
A version of Llama-2 fine-tuned for medicine from EPFL.
-
LLMs Mathematical Capabilities: A Potential Breakthrough?
By
–
Could one say that is a breakthrough, because usually LLMs aren’t able to do maths, right?
-
AI Transforms Music Creation with Machine Learning and LLMs
By
–
Transforming the future of music creation https://
bit.ly/3SYIDBZ
#AI #MachineLearning #DeepLearning #LLMs #DataScience -
Amazon Trains Next-Gen Titan LLM with NVIDIA NeMo
By
–
Explore how @amazon leveraged the NVIDIA NeMo framework, GPUs, and EFA from @awscloud to train its next-generation LLM, giving some of the largest Amazon Titan foundation models customers a faster, more accessible solution for #generativeAI. #AWSreinvent
-
Fine-tuning and derivatives of core AI models
By
–
What the situation with derivatives of core models, like fine tunes for example?
-
Guest lecture on intuitions for understanding large language models
By
–
It was an honor to give a guest lecture yesterday at Stanford’s CS330 class, "Deep Multi-Task and Meta-Learning"! I discussed a few very simple intuitions for how I personally think about large language models. Slides: https://
docs.google.com/presentation/d
/1hQUd3pF8_2Gr2Obc89LKjmHL0DlH-uof9M0yFVd3FA4/edit?usp=sharing
… Here are the six intuitions: (1) -
GPU Memory Limits for Large Model Training on A100
By
–
No I’ve gone higher, 2k would OOM error on an 80gb A100