Barzin Lotfabadi's picture

1

Barzin Lotfabadi

blotfaba

·

https://www.sanctus.ai

AI & ML interests

AI architecture design, foundational & theoretical AI research.

Recent Activity

upvoted a paper about 1 month ago

Less is More: Recursive Reasoning with Tiny Networks

reacted to burtenshaw's post with ❤️ 6 months ago

NEW UNIT in the Hugging Face Reasoning course. We dive deep into the algorithm behind DeepSeek R1 with an advanced and hands-on guide to interpreting GRPO. 🔗 https://huggingface.co/reasoning-course This unit is super useful if you’re tuning models with reinforcement learning. It will help with: - interpreting loss and reward progression during training runs - selecting effective parameters for training - reviewing and defining effective reward functions This unit also works up smoothly toward the existing practical exercises form @mlabonne and Unsloth. 📣 Shout out to @ShirinYamani who wrote the unit. Follow for more great content.

reacted to bartowski's post with 👍 12 months ago

Decided to try to check how many weights in a 70b F32 model would be squashed when converted to F16 (spoiler, it's shockingly few) The reason for this comparison is that it should represent the same percentage of squishing as bf16 to fp16 Had claude make me a script, using the new Reflection-70B, and these are the results: Total weights: 70553706496 Fully representable: 70530215524 Squashed: 23490972 Percentage squashed: 0.03% 0.03%!!!! A couple things to note, this uses a roundtrip of F32 -> F16 -> F32 and then torch.isclose to account for rounding errors that come up by the very nature of extremely accurate numbers, but it uses VERY small tolerances (rtol=1e-5, atol=1e-8) This is also examining EVERY weight that was stored at F32, and for most layers I was somewhere between 0% and 0.03% of weights being squashed, no major outliers. Overall, I feel even safer converting to F16 for llama.cpp, the extremely small number of weights that fall outside the range are likely so small that they don't actually play a role in the final output of the model at inference anyways.

View all activity

Organizations

blotfaba 's datasets

None public yet