Explore tweets tagged as #Pytorch
@PyTorch
PyTorch
16 hours
Startup founders, this is your moment. The #PyTorch Startup Showcase is Oct 21. Pitch live, get noticed, & grow your #AI vision. Apply by Sept 14. 👉 https://t.co/9lzewZw2VW
Tweet media one
0
5
9
@PyTorch
PyTorch
6 days
FlashAttention in 3D? Our latest blog explores the #kernel design of 2-Simplicial #Attention, modeling the algorithm with a hardware aligned design and rewriting the entire kernel in TLX (Triton Low Level Extensions). 🔗 https://t.co/7Enro9MHTI #PyTorch #OpenSourceAI
Tweet media one
4
35
192
@Hesamation
ℏεsam
7 days
gentle reminder that if you start learning AI today from scratch for 2 hrs/day, you begin 2026 knowing: > python, jupyter, numpy (2 weeks) > calculus, linear algebra (4 weeks) > DL basics w/ pytorch (4 weeks) > practice basic projects (1 week) > transformers and LLMs (4 weeks)
Tweet media one
26
157
2K
@buildwithyash
Yash
4 days
Day 19 of python by doing - started building weather dashboard using flask to learn api - learned matliplot - wrote the model script for the project - will prepare ppt for sih tomorrow #PyTorch #flask #Python #100DayChallenge #ai #ArtificialIntelligence
Tweet media one
Tweet media two
1
0
8
@AnushElangovan
Anush Elangovan
2 days
@LodestoneE621 We usually recommend users to use our publicly released rocm-pytorch training docker for best stability and performance of training with ROCm. We have tested on our upcoming release candidate of PyTorch Training Docker v25.8, which uses ROCm 6.4.3 release with PyTorch 2.8, and
Tweet media one
0
0
6
@yuji_fujima
YUJI
20 hours
高性能VLAモデル(ロボの動きを出すAIモデル)の最新バージョン「π0.5」がオープンソースに。Pytorchもサポート。 推論に必要な最低スペックは、 GPUメモリ:8GB以上(ex RTX 4090)
Tweet media one
@JieWang_ZJUI
Jie Wang
2 days
pi-05 has a huge improvement in architecture, integrating more auxiliary information into training. And the performance is more impressive: it can even pick up a tiny piece of orange peel! Thanks @physical_int for open-source amazing models, stay tuned for more cool demos!
0
6
29
@_MattJiang_
Yuxuan Jiang
1 day
Excited to be presenting at #PyTorchConference 2025 (Oct 22–23, SF)! 🎉 I’ll share TrainCheck — detecting silent bugs in deep learning training via invariant inference & proactive checks across model/optimizer/data states. Looking forward to connecting with the PyTorch
Tweet media one
0
3
9
@victor_explore
Victor
5 hours
🤯This Deep Learning Playlist has one of the best animations. Link in comment 33 videos covering everything from backpropagation to coding your own transformer from scratch in PyTorch.
Tweet media one
1
6
45
@PyTorch
PyTorch
1 day
Who else is heading to San Francisco for #PyTorchCon? 🎉 See the leaders, builders, & innovators already signed up, then add your name to the list! October 22-23 is your chance to be in the room where #AI moves forward. 👀 https://t.co/s2lf2xPG4I 🎟️ https://t.co/nSs7KIH1vC
0
7
29
@svpino
Santiago
8 days
Here are my 3 favorite Machine Learning books. These books will teach you the following: 1. The Machine Learning workflow 2. ML algorithms and techniques 3. Deep Learning Tools (Keras, PyTorch, and Scikit-Learn) If you are going to buy one book, make it one of these.
Tweet media one
10
41
292
@DatawithPooja
Pooja Pawar, PhD
10 days
Top Tools Every Data Science Professional Must Know Python | Pandas | NumPy | Matplotlib | Seaborn | PyTorch | Scikit-learn | Dask | Keras Tableau, SQL, Excel & more! Which one is your favorite? #DataScience #Python #AI #MachineLearning #BigData #Analytics #SQL
Tweet media one
0
2
13
@DeepSpeedAI
DeepSpeed
2 days
Step into the future of AI at #PyTorchCon 2025, Oct 22–23 in San Francisco 🔥 Join the DeepSpeed keynote and technical talks. Register: https://t.co/6iogY2eetT + Oct 21 co-located events: Measuring Intelligence, Open Agent & AI Infra Summits / Startup Showcase & PyTorch Training
0
2
6
@tm23twt
ƬⲘ ⚔️
9 days
just coding transformer architecture in pytorch around 6-7 classes - positional encoding, multihead attention, position wise feed forward, layer norm, encoder decoder layer, transformer embeddings & some more idk if it will be done by today, will share the code later :)
Tweet media one
Tweet media two
6
14
220
@imraj4_u
techgeek
6 days
You can hate meta for its privacy policies but you can’t deny that Facebook engineers have made remarkable contribution to programming. React changed the web app development and PyTorch made AI coding way faster. Both React and PyTorch are developed by @Meta
Tweet media one
1
1
2