No Priors Ep. 80 | With Andrej Karpathy from OpenAI and Tesla https://
bit.ly/3N9Cy1l
#AI #MachineLearning #DeepLearning #LLMs #DataScience
LLMS
-
Andrej Karpathy on AI Innovation OpenAI Tesla
By
–
-
Llama 3.1 Improves Context Length and Launches Multimodal Vision
By
–
New Features Improved Context Length On Llama 3.1 8B & 70B, we improved context lengths. Up to 16K on 8B & 64k on 70B! Launching Multimodal We are releasing Llama 3.1 11B & 90B w/ Vision. These models can analyze images & respond to text prompts!
-
Paris emerges as Europe’s leading AI startup hub with FAIR
By
–
False.
– Paris is the most vibrant startup scene in Europe at the moment, more attractive to investors than London.
– FAIR-Paris is one of the 3 largest locations for FAIR, along with Menlo Park and New York.
– Llama-1 was produced in Paris. 11 of the 13 authors of the initial -
GPU Autoscaling for Fine-tuned SLMs: Webinar Recap
By
–
Missed our webinar? Let’s talk #GPU autoscaling for #finetuned #slms! Learn how to: Handle traffic surges effortlessly Optimize GPU costs Hit throughput SLAs https://
pbase.ai/3NOtCyw -
Simple recipe for model error analysis in machine learning
By
–
A simple recipe for model error analysis https://
bit.ly/3XN0SuM
#AI #MachineLearning #DeepLearning #LLMs #DataScience -
Fast AI Inference with SambaNova Cloud and Llama 3.2
By
–
Why is fast #inference so exciting?
— SambaNova (@SambaNovaAI) 29 octobre 2024
🎥 @aton2006, shares the value of fast #inference.
With SambaNova Cloud, #devs can start building with lightning-fast AI inference on @AIatMeta's Llama 3.2 💻
Start developing ⤵️https://t.co/zm6RCXY00nWhy is fast #inference so exciting? @aton2006
, shares the value of fast #inference. With SambaNova Cloud, #devs can start building with lightning-fast AI inference on @AIatMeta
's Llama 3.2 Start developing http://
cloud.sambanova.ai -
Meta Layer Skip Enables 200% Faster Transformer Inference
By
–
Meta presents Layer Skip – up-to 200% fast inference > Applies layer dropout: low rates for early layers, high rates for later layers
> Uses early exit loss with shared exit for all transformer layers Inference: > Increases early exit accuracy without auxiliary layers
> -
Claude Now Available as Model Option at Same Price
By
–
Yep same price, no catch. Claude is just now available as one of the model options
-
GPT Action Simulation: Frame Testing and Prompt Optimization
By
–
not actually play play — i had a bunch of screen-caps and forward-simulated what GPT would take as an action.
I tried sending one frame, 10 frames, 24 frames, etc. and asked it to take one of several actions — and played with the prompting to adjust the state space and action