Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Jul 27, 2025

• Add accelerator API to RPC distributed examples: ddp_rpc, parameter_server, rnn (#1371)

  Show description for e9a4e75
  jafraustro

  authored
  e9a4e75

  Copy full SHA for e9a4e75

Commits on Jul 21, 2025

• mark fsdp1 as deprecated and point to fsdp2 (#1374)

  Show description for 99f5c4e
  weifengpy

  authored
  99f5c4e

  Copy full SHA for 99f5c4e

Commits on Jul 16, 2025

• fsdp_tp_example fsdp1-> fsdp2

  

  wwwjn

  authored and

  soumith

  committed
  2174f26

  Copy full SHA for 2174f26

• remove manual n_heads change

  

  wwwjn

  authored and

  soumith

  committed
  8c1555d

  Copy full SHA for 8c1555d

Commits on Jul 14, 2025

• Bump torch from 1.9.0 to 2.2.0 in /distributed/rpc/batch

  Show description for 58370dc
  

  dependabot[bot]

  authored and

  soumith

  committed
  58370dc

  Copy full SHA for 58370dc

• Fixed "--help" output in README.md

  Show description for bd1254b
  

  Anthony Smith

  authored and

  soumith

  committed
  bd1254b

  Copy full SHA for bd1254b

• Enhance README and examples for Tensor Parallelism

  Show description for 86ed9e4
  

  jafraustro

  authored and

  soumith

  committed
  86ed9e4

  Copy full SHA for 86ed9e4

• Add torch.accelerator API to mingGPT example (#34)

  Show description for 698a89e
  

  gera-aldama

  authored and

  soumith

  committed
  698a89e

  Copy full SHA for 698a89e

• add min_gpu verification similart tu tensor_parallel_example.py example

  Show description for f84bcb3
  

  jafraustro

  authored and

  soumith

  committed
  f84bcb3

  Copy full SHA for f84bcb3

• Update requirements.txt to specify torch version >=2.7

  Show description for eab402d
  

  jafraustro

  authored and

  soumith

  committed
  eab402d

  Copy full SHA for eab402d

• Refactor DDP example to use Accelerator API

  Show description for 3302012
  

  jafraustro

  authored and

  soumith

  committed
  3302012

  Copy full SHA for 3302012

• Switch to torchrun for distributed launches

  Show description for d86d205
  

  jafraustro

  authored and

  soumith

  committed
  d86d205

  Copy full SHA for d86d205

Commits on Jul 10, 2025

• Add `accelerator` API to GCN example.

  Show description for d47f0f3
  

  jafraustro

  authored and

  soumith

  committed
  d47f0f3

  Copy full SHA for d47f0f3

• Update .github/workflows/main_distributed.yaml

  Show description for c04a5a1
  

  2 peopleauthored and

  soumith

  committed
  c04a5a1

  Copy full SHA for c04a5a1

• Update .github/workflows/main_distributed.yaml

  Show description for d384911
  

  2 peopleauthored and

  soumith

  committed
  d384911

  Copy full SHA for d384911

• Update .github/workflows/main_distributed.yaml

  Show description for 2640946
  

  2 peopleauthored and

  soumith

  committed
  2640946

  Copy full SHA for 2640946

• Updating Python to 3.10, fsdp_tp_example.py to accelerator

  

  githubsgi

  authored and

  soumith

  committed
  997c123

  Copy full SHA for 997c123

• Updating comments in multiple files.

  

  githubsgi

  authored and

  soumith

  committed
  3fc7853

  Copy full SHA for 3fc7853

• Removing pre's from requirement.txt, ...

  

  githubsgi

  authored and

  soumith

  committed
  bac5d0f

  Copy full SHA for bac5d0f

• Bumping up Python from 3.8 to 3.9 in wf yaml.

  

  githubsgi

  authored and

  soumith

  committed
  a5b6297

  Copy full SHA for a5b6297

• Bumping up Pyhon version to 3.9 .

  

  githubsgi

  authored and

  soumith

  committed
  fe39e70

  Copy full SHA for fe39e70

• Updating CI torch version.

  

  githubsgi

  authored and

  soumith

  committed
  56cc1a6

  Copy full SHA for 56cc1a6

• Moving CommDebugMode code to a seperate PR.

  

  githubsgi

  authored and

  soumith

  committed
  3055e3e

  Copy full SHA for 3055e3e

• cuda to accelerator

  

  githubsgi

  authored and

  soumith

  committed
  f8c79f1

  Copy full SHA for f8c79f1

• Cuda to accelerator, +CommDebugMode

  

  githubsgi

  authored and

  soumith

  committed
  6bfe651

  Copy full SHA for 6bfe651

• Update main.py tutorial, set persistent_workers, make GPU execution much faster

  Show description for d0b7e37
  

  maxwbuckley

  authored and

  soumith

  committed
  d0b7e37

  Copy full SHA for d0b7e37

Commits on Jul 7, 2025

• Fixed use_accel not defined issue (#1363)

  BLOrange-AMD

  authored
  4206858

  Copy full SHA for 4206858

Commits on Jun 30, 2025

• imagenet: fix typo addressing args.gpu (#1361)

  dvrogozh

  authored
  16554e5

  Copy full SHA for 16554e5

Commits on Jun 25, 2025

• Revert "Add Differentiable Physics: Mass-Spring System example" (#1360)

  Show description for 8d408d2
  msaroufim

  authored
  8d408d2

  Copy full SHA for 8d408d2

• Add Differentiable Physics: Mass-Spring System example (#1359)

  Show description for 7c35995
  

  abhitorch81

  and

  AbhiLegend

  authored
  7c35995

  Copy full SHA for 7c35995

Commits on Jun 24, 2025

• Add accelerate API support for Super Resolution example (#1358)

  Show description for 6f61614
  framoncg

  authored
  6f61614

  Copy full SHA for 6f61614

Commits on Jun 17, 2025

• Change run_python_examples.sh to run word_language_model/generate.py after each call to word_language_model/main.py

  

  framoncg

  authored and

  soumith

  committed
  2944a9d

  Copy full SHA for 2944a9d

• Update word_language_model/generate.py to remove duplicates, use abc order

  Show description for 6595d7b
  

  2 peopleauthored and

  soumith

  committed
  6595d7b

  Copy full SHA for 6595d7b

• Fix torch>= 2.6 generate.py compatibility in Word_language_example

  

  framoncg

  authored and

  soumith

  committed
  a86bda0

  Copy full SHA for a86bda0

• Removing unneeded print

  Show description for 37986af
  

  eromomon

  authored and

  soumith

  committed
  37986af

  Copy full SHA for 37986af

Pagination

Previous

Next