Code

Serialize Config from Model#7

Code

Merged

bfineran merged 14 commits intomainneuralmagic/compressed-tensors:mainfrom serialize_configneuralmagic/compressed-tensors:serialize_configCopy head branch name to clipboardApr 16, 2024

+234-19

Commits

Commits on Apr 12, 2024

• Apply quantization config implementation

  

  Benjamin

  committed
  9aae8e8

  Copy full SHA for 9aae8e8

• add TODO

  

  Benjamin

  committed
  8465015

  Copy full SHA for 8465015

• integrate full lifecycle support, QuantizationStatus updates, add tinyllama test

  

  Benjamin

  committed
  24e04b6

  Copy full SHA for 24e04b6

• fix comment

  

  Benjamin

  committed
  b5a07c4

  Copy full SHA for b5a07c4

• initial implementation

  

  Sara Adkins

  committed
  7142a71

  Copy full SHA for 7142a71

• add unit test

  

  Sara Adkins

  committed
  23e9ae8

  Copy full SHA for 23e9ae8

Commits on Apr 15, 2024

• Merge branch 'main' into serialize_config

  

  Sara Adkins

  committed
  dd77890

  Copy full SHA for dd77890

• cleanup is_quantized

  

  Sara Adkins

  committed
  b9c9530

  Copy full SHA for b9c9530

Commits on Apr 16, 2024

• clean up targets and ignore lists

  

  Sara Adkins

  committed
  845bfb9

  Copy full SHA for 845bfb9

• global compression ratio and docstrings

  

  Sara Adkins

  committed
  1a7984c

  Copy full SHA for 1a7984c

• make sure scale/zp on correct device

  

  Sara Adkins

  committed
  faa93c9

  Copy full SHA for faa93c9

• helper for model quantization

  

  Sara Adkins

  committed
  caeab7d

  Copy full SHA for caeab7d

• Merge branch 'fix_device_mismatch' into serialize_config

  

  Sara Adkins

  committed
  e7e6f43

  Copy full SHA for e7e6f43

• Merge branch 'main' into serialize_config

  bfineran

  authored
  ec2ef84

  Copy full SHA for ec2ef84