PyTorch(@PyTorch) 's Twitter Profileg
PyTorch

@PyTorch

Tensors and neural networks in Python with strong hardware acceleration. PyTorch is an open source project at the Linux Foundation. #PyTorchFoundation

ID:776585502606721024

linkhttp://pytorch.org calendar_today16-09-2016 00:56:26

2,0K Tweets

375,5K Followers

77 Following

Follow People
PyTorch(@PyTorch) 's Twitter Profile Photo

Part 2️⃣: Understanding GPU Memory 🤔

In our latest post, we will use the Memory Snapshot to visualize a GPU memory leak caused by reference cycles, and then locate and remove them in our code using the Reference Cycle Detector.

Read more: hubs.la/Q02drBVw0

Part 2️⃣: Understanding GPU Memory 🤔 In our latest post, we will use the Memory Snapshot to visualize a GPU memory leak caused by reference cycles, and then locate and remove them in our code using the Reference Cycle Detector. Read more: hubs.la/Q02drBVw0
account_circle
PyTorch(@PyTorch) 's Twitter Profile Photo

Hear how Meta reports experiences in using PyTorch 2 to accelerate production models, resulting in up to ~25% training-time reduction. 👀🔥👇

hubs.la/Q02djDdc0

account_circle
PyTorch(@PyTorch) 's Twitter Profile Photo

Latest blog post: A more generalized model transformation solution 🔥

More performant and user-friendly, bringing performance improvements to both model training and inference without manual efforts.

Check it out 👉 hubs.la/Q02d9Ffm0

Latest blog post: A more generalized model transformation solution 🔥 More performant and user-friendly, bringing performance improvements to both model training and inference without manual efforts. Check it out 👉 hubs.la/Q02d9Ffm0
account_circle
PyTorch(@PyTorch) 's Twitter Profile Photo

Understanding GPU Memory 1: Visualizing All Allocations over Time 👀

In part 1 of this series, we show how to use Memory Snapshot, the Memory Profiler, and the Reference Cycle Detector to debug out of memory errors and improve memory usage.

Read more: hubs.la/Q02d6VZQ0

Understanding GPU Memory 1: Visualizing All Allocations over Time 👀 In part 1 of this series, we show how to use Memory Snapshot, the Memory Profiler, and the Reference Cycle Detector to debug out of memory errors and improve memory usage. Read more: hubs.la/Q02d6VZQ0
account_circle
PyTorch(@PyTorch) 's Twitter Profile Photo

Cool PyTorch use case 🔥

Hear from Intel on how they use PyTorch for seismic imaging and collecting dinosaur bones.

👉👉👉 hubs.la/Q02cQnsB0

Cool PyTorch use case 🔥 Hear from Intel on how they use PyTorch for seismic imaging and collecting dinosaur bones. 👉👉👉 hubs.la/Q02cQnsB0
account_circle
PyTorch(@PyTorch) 's Twitter Profile Photo

Introducing PyPose, one of our latest ecosystem tools 🔥

PyPose is a PyTorch-based robotics-oriented library that provides a set of tools and algorithms for connecting deep learning with physics-based optimization.

Learn more 👇
hubs.ly/Q02cbc7y0

account_circle
PyTorch(@PyTorch) 's Twitter Profile Photo

Welcome to the newest member of the PyTorch Foundation, Snowflake ❄️

We're excited to see what they bring to our community through data management and deep learning technologies. 🤝

Read more: hubs.la/Q02b_VHL0

Welcome to the newest member of the PyTorch Foundation, Snowflake ❄️ We're excited to see what they bring to our community through data management and deep learning technologies. 🤝 Read more: hubs.la/Q02b_VHL0
account_circle
PyTorch(@PyTorch) 's Twitter Profile Photo

10x performance on LLaMa 7B, all in native PyTorch. No C++ needed.

Check out our second blog post in a series on Accelerating Generative AI using Native PyTorch. 🔥

hubs.la/Q02bx5Bn0

account_circle
PyTorch(@PyTorch) 's Twitter Profile Photo

Learn more about the Intel + PyTorch 2.1 contributions, including:

TorchInductor-CPU optimizations
CPU dynamic shape inference path
C++ wrapper
Flash-attention-based scaled dot product algorithm for CPU
PyTorch 2 export post-training quantization

🔥👇
hubs.la/Q02bqhNt0

account_circle
Linux Foundation Training & Certification(@LF_Training) 's Twitter Profile Photo

🐧 Cyber Monday is here 🐧! 65% off bundles, power bundles & IT professional programs + 50% off individual courses & certs + 10% off THRIVE-ONE annual subscriptions. Enroll today: hubs.la/Q029Nt_S0

🐧 Cyber Monday is here 🐧! 65% off bundles, power bundles & IT professional programs + 50% off individual courses & certs + 10% off THRIVE-ONE annual subscriptions. Enroll today: hubs.la/Q029Nt_S0 #Linux #CloudNative #ITCertifications #Training
account_circle
PyTorch(@PyTorch) 's Twitter Profile Photo

It's Cyber Monday, which means all LF trainings are up to 65% off!

This includes both of our PyTorch trainings 🎉 Save now: hubs.la/Q029M_zm0

It's Cyber Monday, which means all LF trainings are up to 65% off! This includes both of our PyTorch trainings 🎉 Save now: hubs.la/Q029M_zm0
account_circle
PyTorch(@PyTorch) 's Twitter Profile Photo

As we head into this Thanksgiving weekend, we want to express our gratitude to you, our community, for making PyTorch what it is today.

For those who celebrate, we wish you a safe and happy holiday with all of your favorite PieTorch (personally, we’re fans of cherry 🥧).

As we head into this Thanksgiving weekend, we want to express our gratitude to you, our community, for making PyTorch what it is today. For those who celebrate, we wish you a safe and happy holiday with all of your favorite PieTorch (personally, we’re fans of cherry 🥧).
account_circle
PyTorch(@PyTorch) 's Twitter Profile Photo

New blog series: Accelerating Generative AI using native PyTorch. 🔥

In this post we talk through new PyTorch performance features from the conference and how they can be used to produce an 8x faster, entirely PyTorch implementation of Segment Anything.

hubs.la/Q0298GN00

New blog series: Accelerating Generative AI using native PyTorch. 🔥 In this post we talk through new PyTorch performance features from the conference and how they can be used to produce an 8x faster, entirely PyTorch implementation of Segment Anything. hubs.la/Q0298GN00
account_circle
PyTorch(@PyTorch) 's Twitter Profile Photo

🎉 Fall PyTorch Docathon complete 🎉

Dedication, expertise, and tireless efforts of our open-source contributors have helped us improve PyTorch documentation. 💜

Congratulations to our winners!! Results here: hubs.la/Q0298s8H0

🎉 Fall PyTorch Docathon complete 🎉 Dedication, expertise, and tireless efforts of our open-source contributors have helped us improve PyTorch documentation. 💜 Congratulations to our winners!! Results here: hubs.la/Q0298s8H0
account_circle
PyTorch(@PyTorch) 's Twitter Profile Photo

Training deep learning models on such substantial volumes of data requires many techniques, like activation checkpointing.

Hear how Theator uses automatic differentiation and the new non-reentrant implementation of activation checkpointing. 👇

hubs.la/Q028fS0z0

account_circle
PyTorch(@PyTorch) 's Twitter Profile Photo

Improve the inference latencies of the Llama 2 family of models using PyTorch native optimizations 🔥

See how teams from IBM and Meta demonstrate ultra low latencies for 70B model inference in our latest blog post: hubs.la/Q0283P3L0

Improve the inference latencies of the Llama 2 family of models using PyTorch native optimizations 🔥 See how teams from IBM and Meta demonstrate ultra low latencies for 70B model inference in our latest blog post: hubs.la/Q0283P3L0
account_circle