Skip to content

Tags: gmagogsfm/torchtune

Tags

v0.5.0-rc1

Toggle v0.5.0-rc1's commit message

Verified

This commit was created on GitHub.com and signed with GitHub’s verified signature.
[Fix] Llama 3.2 Vision decoder_trainable flag fixed (pytorch#2150)

v0.4.0

Toggle v0.4.0's commit message
Update Qwen2.5 configs (pytorch#1999)

v0.4.0-rc4

Toggle v0.4.0-rc4's commit message
Update Qwen2.5 configs (pytorch#1999)

v0.4.0-rc3

Toggle v0.4.0-rc3's commit message
Pass quantization_kwargs to CLIP builders (pytorch#1994)

v0.4.0-rc2

Toggle v0.4.0-rc2's commit message
Refactor Recipe State Dict Code (pytorch#1964)

v0.4.0-rc1

Toggle v0.4.0-rc1's commit message

Verified

This commit was created on GitHub.com and signed with GitHub’s verified signature.
Fix lora single device fine tune checkpoint saving & nan loss when us…

…e_dora=True (pytorch#1909)

v0.3.1

Toggle v0.3.1's commit message
Fix CLIP pos embedding interpolation to work on DTensors (pytorch#1739)

v0.3.1-rc3

Toggle v0.3.1-rc3's commit message
Fix CLIP pos embedding interpolation to work on DTensors (pytorch#1739)

v0.3.1-rc2

Toggle v0.3.1-rc2's commit message
Bump version 0.3.1 (pytorch#1720)

v0.3.1-rc1

Toggle v0.3.1-rc1's commit message

Verified

This commit was created on GitHub.com and signed with GitHub’s verified signature.
MPS support (pytorch#1706)