Code

Fine-tuning support#248

Code

Draft

pharmapsychotic wants to merge 41 commits intomainStability-AI/stability-sdk:mainfrom PLATFORM-339Stability-AI/stability-sdk:PLATFORM-339Copy head branch name to clipboard

+3,227-36

Commits

Commits on May 22, 2023

• First pass at fine tuning support in SDK

  pharmapsychotic

  committed
  d4c7d05

Commits on May 23, 2023

• Fine-tuning updates:

  Show description for c221c62
  pharmapsychotic

  committed
  c221c62

Commits on May 26, 2023

• Finetune updates:

  Show description for d9a40b8
  pharmapsychotic

  committed
  d9a40b8

Commits on Jun 8, 2023

• Update to latest api-interfaces

  pharmapsychotic

  committed
  6c56a6d

Commits on Jun 15, 2023

• Update to latest api-interfaces

  pharmapsychotic

  committed
  ab4c022

Commits on Jun 16, 2023

• Project.project_type -> Project.type

  pharmapsychotic

  committed
  2b3178e

Commits on Jun 20, 2023

• add pydantic dependency

  brianfitzgerald

  committed
  6665684

Commits on Jun 26, 2023

• Renamed object_name to object_prompt and addition of extras field for passing through extra training arguments for development and testing.

  pharmapsychotic

  committed
  039a5c6

Commits on Jun 30, 2023

• Bf/platform 339 (#243)

  brianfitzgerald

  authored
  42457d2

• Merge branch main into PLATFORM-339

  pharmapsychotic

  committed
  9d1395a

• Update for multi-LoRA and using model names in prompt

  pharmapsychotic

  committed
  45ed1d2

• Fix zip handling and fold install/cell cells

  pharmapsychotic

  committed
  6b8397e

• Extra error handling and logging for image uploads

  pharmapsychotic

  committed
  8326654

Commits on Jul 5, 2023

• Add fine tune support to StabilityInference

  pharmapsychotic

  committed
  8171e66

Commits on Jul 6, 2023

• Improve training status display

  pharmapsychotic

  committed
  75188fb

Commits on Jul 11, 2023

• Fine-tuning updates:

  Show description for 33071cc
  pharmapsychotic

  committed
  33071cc

Commits on Jul 12, 2023

• Better prompt model substitution

  pharmapsychotic

  committed
  99ab453

Commits on Jul 14, 2023

• Clean up

  pharmapsychotic

  committed
  a2784e3

Commits on Jul 19, 2023

• Update notebook for SDXL 0.9 fine-tuning

  pharmapsychotic

  committed
  f011706

Commits on Jul 21, 2023

• Fixes to <model:weight> syntax ensure the :weight is stripped and not passed through in prompt.

  pharmapsychotic

  committed
  0a04175

• Merge branch main into PLATFORM-339

  pharmapsychotic

  committed
  fb1fa2f

Commits on Jul 24, 2023

• Separate max image counts for face/object/style

  pharmapsychotic

  committed
  c519f6a

Commits on Jul 25, 2023

• Update client.StabilityInference to also use parse_models_from_prompts and add support to parser for generation.Prompt objects.

  pharmapsychotic

  committed
  e905c2a

Commits on Aug 2, 2023

• Increase max image size

  pharmapsychotic

  committed
  c35a25a

• Add stable-diffusion-xl-1024-v1-0 to finetune notebook

  pharmapsychotic

  committed
  4c6dcc3

Commits on Aug 10, 2023

• Bugfix: regex for token replacement in prompt

  brianfitzgerald

  committed
  25d92f9

• fix another edge case, and add unit tests

  brianfitzgerald

  committed
  5b01ef3

• update notebook defaults

  brianfitzgerald

  committed
  4855603

Commits on Aug 14, 2023

• update image upload logic

  brianfitzgerald

  committed
  5d3b669

Commits on Aug 28, 2023

• add external test link

  brianfitzgerald

  committed
  00611e6

Commits on Sep 20, 2023

• feat: added fine-tuning REST API Colab

  todd-elvers

  committed
  374c3fa

• feat: added fine-tuning REST API Colab

  todd-elvers

  committed
  3bb15ae

Commits on Sep 22, 2023

• chore: updated REST API colab to refrence prod preview url

  todd-elvers

  committed
  81418cd

• fix: corrected an indentation error

  todd-elvers

  committed
  08bf3ac

• fix: corrected an indentation error

  todd-elvers

  committed
  4fb309a

• fix: corrected an indentation error

  todd-elvers

  committed
  d5ef8f2

Commits on Sep 28, 2023

• chore: added a Colab I made for Adil

  todd-elvers

  committed
  cde43e1

Commits on Oct 2, 2023

• chore: fixed a small bug in existing colab

  todd-elvers

  committed
  e8c9145

• chore: more corrections to REST colabs

  todd-elvers

  committed
  f0fffd4

Commits on Oct 5, 2023

• Use production endpoint for fine tuning notebook

  brianfitzgerald

  committed
  7208fa5

• feat: updated REST API colab to allow for using more than one fine-tune at a time

  todd-elvers

  committed
  61dd932