@andrew-n-carr.bsky.social
co-founder leading science at Cartwheel
AI writer for TLDR AI Newsletter
co-founder Arcade
Past - Codegen at OpenAI, Brain at GoogleAI, world ranked Tetris player
AI writer for TLDR AI Newsletter
co-founder Arcade
Past - Codegen at OpenAI, Brain at GoogleAI, world ranked Tetris player
I thought this was an interesting graphic
February 16, 2025 at 10:37 PM
I thought this was an interesting graphic
They did it all without Jira... Amazing
December 26, 2024 at 2:55 AM
They did it all without Jira... Amazing
Genmo has released LoRA training capabilities for their generative video model Mochi
github.com/genmoai/moch...
Trains quickly on a single 80GB GPU.
github.com/genmoai/moch...
Trains quickly on a single 80GB GPU.
November 27, 2024 at 12:24 AM
Genmo has released LoRA training capabilities for their generative video model Mochi
github.com/genmoai/moch...
Trains quickly on a single 80GB GPU.
github.com/genmoai/moch...
Trains quickly on a single 80GB GPU.
Cool new paper from NVIDIA about a hybrid state space + attention model that performs extremely well as a small model. Their 1.5B model even out performs Llama 3.2 3B
arxiv: arxiv.org/abs/2411.13676
arxiv: arxiv.org/abs/2411.13676
November 22, 2024 at 8:27 PM
Cool new paper from NVIDIA about a hybrid state space + attention model that performs extremely well as a small model. Their 1.5B model even out performs Llama 3.2 3B
arxiv: arxiv.org/abs/2411.13676
arxiv: arxiv.org/abs/2411.13676
Inference Scaling Laws of DeepSeek-R1-Lite-Preview
Longer Reasoning, Better Performance. DeepSeek-R1-Lite-Preview shows steady score improvements on AIME as thought length increases.
Longer Reasoning, Better Performance. DeepSeek-R1-Lite-Preview shows steady score improvements on AIME as thought length increases.
November 20, 2024 at 3:36 PM
Inference Scaling Laws of DeepSeek-R1-Lite-Preview
Longer Reasoning, Better Performance. DeepSeek-R1-Lite-Preview shows steady score improvements on AIME as thought length increases.
Longer Reasoning, Better Performance. DeepSeek-R1-Lite-Preview shows steady score improvements on AIME as thought length increases.
Impressive Results of DeepSeek-R1-Lite-Preview Across Benchmarks!
November 20, 2024 at 3:36 PM
Impressive Results of DeepSeek-R1-Lite-Preview Across Benchmarks!
Fun probability fact, the likelihood that two randomly drawn numbers are coprime is 61%!
November 20, 2024 at 1:56 AM
Fun probability fact, the likelihood that two randomly drawn numbers are coprime is 61%!
I have nothing to say. Just enjoy this validation loss curve for a moment
November 19, 2024 at 11:09 PM
I have nothing to say. Just enjoy this validation loss curve for a moment