Picture

Amar Saini @_Epoching_

Deep Learning Research California, USA Joined October 2015
  • Tweets
    912
  • Followers
    377
  • Following
    294
  • Likes
    3K
Amar Saini retweeted
rasbt Profile Picture

Sebastian Raschka @rasbt

a month ago

Couldn't resist. Here's a pure PyTorch from-scratch re-implementation of Gemma 3 270M in a Jupyter Notebook (uses about 1.49 GB RAM): github.com/rasbt/LLMs-fro…

rasbt Profile Picture

Sebastian Raschka @rasbt

2 months ago

Couldn't resist. Here's a pure PyTorch from-scratch re-implementation of Gemma 3 270M in a Jupyter Notebook (uses about 1.49 GB RAM): github.com/rasbt/LLMs-fro… https://t.co/9vF9U29pWh

Amar Saini retweeted
luisgnet Profile Picture

Luis @luisgnet

a month ago

so I built a tiktok for doomscrolling arxiv papers: elixiv.org/foryou

Amar Saini retweeted
jeremyphoward Profile Picture

Jeremy Howard @jeremyphoward

2 months ago

"🤯Imagine this: - Build a web app & interact with it on the same screen as your code - Use live vars from your REPL directly in AI - Turn any function into AI tool instantly A live environment fusing the best ideas from Literate Programming, Smalltalk, & Jupyter."

HamelHusain Profile Picture

Hamel Husain @HamelHusain

2 months ago

"🤯Imagine this: - Build a web app & interact with it on the same screen as your code - Use live vars from your REPL directly in AI - Turn any function into AI tool instantly A live environment fusing the best ideas from Literate Programming, Smalltalk, & Jupyter."

Amar Saini retweeted
bartoldson Profile Picture

Brian Bartoldson @bartoldson

6 months ago

🚀 We fixed a major LLM post-training bottleneck! Our new method (TBA) combines trajectory balance with asynchronous training to speed up LLM RL 5-50x while improving results+scalability. For example, using VinePPO's GSM8K setup, we obtain +1.2% accuracy and 50x faster RL.

Amar Saini retweeted
tom_doerr Profile Picture

Tom Dörr @tom_doerr

8 months ago

GPU-accelerated NumPy/SciPy library

Amar Saini retweeted
itsPaulAi Profile Picture

Paul Couvert @itsPaulAi

9 months ago

Hugging Face released a free course on agents You can learn how to create: - Code agents that solve problem with code - Retrieval agents that supply grounded context - Custom functional agents that do whatever you need Definitely the best I've seen so far. Link below

Amar Saini retweeted
janusch_patas Profile Picture

MrNeRF @janusch_patas

10 months ago

MV-DUSt3R+: Single-Stage Scene Reconstruction from Sparse Views In 2 Seconds Contributions: • We present MV-DUSt3R, a novel feed-forward network for pose-free scene reconstruction from sparse multi-view input. It not only runs 48 ∼ 78× faster than DUSt3R for 4 ∼ 24 views but…

Amar Saini retweeted
luisgnet Profile Picture

Luis @luisgnet

9 months ago

PH Launch #3 🫡 LFG 🚀 producthunt.com/posts/elixiv

Amar Saini retweeted
SakanaAILabs Profile Picture

Sakana AI @SakanaAILabs

9 months ago

Introducing ASAL: Automating the Search for Artificial Life with Foundation Models sakana.ai/asal/ Artificial Life (ALife) research holds key insights that can transform and accelerate progress in AI. By speeding up ALife discovery with AI, we accelerate our…

Amar Saini retweeted
jeremyphoward Profile Picture

Jeremy Howard @jeremyphoward

9 months ago

I'll get straight to the point. We trained 2 new models. Like BERT, but modern. ModernBERT. Not some hypey GenAI thing, but a proper workhorse model, for retrieval, classification, etc. Real practical stuff. It's much faster, more accurate, longer context, and more useful. 🧵

Amar Saini retweeted
RickyTQChen Profile Picture

Ricky T. Q. Chen @RickyTQChen

10 months ago

Want to learn continuous & discrete Flow Matching? We've just released: 📙 A guide covering Flow Matching basics & advanced methods arxiv.org/abs/2412.06264. 💻 An open source codebase with image & text examples github.com/facebookresear…. 🗣️ A Flow Matching tutorial #NeurIPS2024.

Amar Saini retweeted
SakanaAILabs Profile Picture

Sakana AI @SakanaAILabs

10 months ago

Introducing An Evolved Universal Transformer Memory sakana.ai/namm Neural Attention Memory Models (NAMMs) are a new kind of neural memory system for Transformers that not only boost their performance and efficiency but are also transferable to other foundation models,…

Amar Saini retweeted
AIatMeta Profile Picture

AI at Meta @AIatMeta

10 months ago

As we continue to explore new post-training techniques, today we're releasing Llama 3.3 — a new open source model that delivers leading performance and quality across text-based use cases such as synthetic data generation at a fraction of the inference cost.

Amar Saini retweeted
jeremyphoward Profile Picture

Jeremy Howard @jeremyphoward

10 months ago

Wow there is a *lot* of llms.txt activity happening now. Like… a lot lot… directory.llmstxt.cloud

Amar Saini retweeted
rasbt Profile Picture

Sebastian Raschka @rasbt

12 months ago

The Llama 3.2 1B and 3B models are my favorite LLMs -- small but very capable. If you want to understand how the architectures look like under the hood, I implemented them from scratch (one of the best ways to learn): github.com/rasbt/LLMs-fro…

Amar Saini retweeted
_akhaliq Profile Picture

AK @_akhaliq

a year ago

CCTV_YOLO Fast Real-time Object Detection with High-Res Output This project demonstrates real-time object detection using the YOLOv5n6 model with low-resolution inference for high-speed processing, while drawing the results on high-resolution frames. The object detection…

Amar Saini retweeted
danielhanchen Profile Picture

Daniel Han @danielhanchen

a year ago

My analysis of Llama 3.2: 1. New 1B and 3B text only LLMs 9 trillion tokens 2. New 11B and 90B vision multimodal models 3. 128K context length 4. 1B and 3B used some distillation from 8B and 70B 5. VLM 6 billion img, text pairs 6. CLIP MLP GeLU + cross attention Long analysis:…

Amar Saini retweeted
AIatMeta Profile Picture

AI at Meta @AIatMeta

a year ago

📣 Introducing Llama 3.2: Lightweight models for edge devices, vision models and more! What’s new? • Llama 3.2 1B & 3B models deliver state-of-the-art capabilities for their class for several on-device use cases — with support for @Arm, @MediaTek & @Qualcomm on day one. •…

Amar Saini retweeted
ClementDelangue Profile Picture

clem 🤗 @ClementDelangue

a year ago

Once again, an AI system is not "thinking", it's "processing", "running predictions",... just like Google or computers do. Giving the false impression that technology systems are human is just cheap snake oil and marketing to fool you into thinking it's more clever than it is.

Amar Saini retweeted
_arohan_ Profile Picture

rohan anil @_arohan_

a year ago

Insane timeline.

RealJosephus Profile Picture

Joseph @RealJosephus

a year ago

Insane timeline.

Trends for United States

Nicki
187 B posts
Bengals
50,6 B posts
Jets
86,4 B posts
Zac Taylor
7.084 posts
Barney B
14,9 B posts
Tyreek
43 B posts
Druski
8.626 posts
Kevin Hart
4.665 posts
Jake Browning
4.839 posts
Broncos
40,3 B posts
Burrow
12 B posts
Garrett Wilson
6.891 posts
Cocaine Barbie
23,2 B posts
Dolphins
57,1 B posts
Bo Nix
6.867 posts
#WWERaw
48,7 B posts
Justin Fields
11,3 B posts
Aaron Glenn
3.759 posts
Dobbins
7.862 posts
ABCDEFG
4.886 posts

You might like

VFX artist & AI Educator. Exploring new tools from the worlds of AI, digital art, film, and games in a fun and project-oriented way. Tutorials on YouTube!
Mickmumpitz
@mickmumpitz
13K Followers
Interpretability/Finetuning @AnthropicAI Previously: Staff ML Engineer @stripe, Wrote BMLPA by @OReillyMedia, Head of AI at @InsightFellows, ML @Zipcar
Emmanuel Ameisen
@mlpowered
10K Followers
ML Research | Multimodality, Self-improvement, LRMs | Keras3 collaborator, contributes to JAX ecosystem, TF-addons maintainer | @GoogleDevExpert in ML-JAX | OSS
Aakash Kumar Nain
@A_K_Nain
12K Followers
3D computer vision and ML for AR and robotics, @Google and TU Munich
Federico Tombari
@fedassa
2K Followers
AI researcher & engineer~ sharing practical tips on NLP & CV. trying to stay up to date on 🔥 papers. working on #FashionAI 💃💅 🤖❤ language learning🐍🛼🍫☕🤑
Ceyda Cinarel
@ceyda_cinarel
2K Followers
We're rekindling the dream of personal computing by making reliable software creation accessible to all. Join us: https://t.co/9UF0rR6YUr
Imbue
@imbue_ai
7K Followers
Prof @WisconsinCS. Making AI reliable for the open world.
Sharon Y. Li
@SharonYixuanLi
11K Followers
Founder - https://t.co/rmc2uywaIp Youtube - https://t.co/7UTfrP1GJt
Vishnu - Jarvislabs.a...
@vishnuvig
2K Followers
@huggingface engineer. I'm the reason your LLM frontend has a jinja2cpp dependency. Sometimes yells about housing and trans rights instead of working He/him
Matthew Carrigan
@carrigmat
14K Followers
Building 3D embodiment for AI agents so that they can see, understand & interact. 🏂 Platform: https://t.co/X0fdkUyZwQ 🤖 API: https://t.co/mKPn7UES2W
Meshcapade
@meshcapade
7K Followers
Farcasting back to Hyperion
Wayde Gilliam
@waydegilliam
2K Followers
Teaching Pixels New Tricks | Research @runwayml
Elad Richardson
@EladRichardson
2K Followers
3D Machine Learning @HuggingFace, IndividualKex on TikTok/YT, PhD Brown
dylan
@dylan_ebert_
9K Followers
Applied AI @weights_biases | ex-Facebook Safety | https://t.co/a7i7G5dkLG | 🇮🇪 | Came for the bants, stayed for the rants
Morgan McGuire
@morgymcg
3K Followers
VP & Chief AI Scientist@Reliance Jio, Volunteer@iSPIRT. Visiting Faculty @ ISB, 2x founder, Prev: Google Research, Fashiate, Yahoo Labs. Views personal!
Gaurav Aggarwal
@fooobar
12K Followers