xzuyn

Adds [CAME: Confidence-guided Adaptive Memory Efficient Optimization](https://arxiv.org/abs/2307.02047). Requires you to `pip install pytorch-came`. Github: https://github.com/yangluo7/CAME

`Hardware Requirement` in the readme lacks critical info

1

All that's provided is the training method, bits, and model sizes. Nothing is said about things such as context length, batch size, optimizations, etc. The examples could be at batch...

[Bug]: Artifacts on left and bottom sides of images with SD3.5M LoRAs

13

### What happened? I'm seeing artifacts with LoRAs I trained on SD3.5M, but only when generated outside of OneTrainer. The sample images generated during training look perfectly fine, but the...

Qwen3 Does Not Work With `fix_untrained_tokens`

2

### Please check that this issue hasn't been reported before. - [x] I searched previous [Bug Reports](https://github.com/axolotl-ai-cloud/axolotl/labels/bug) didn't find any similar reports. ### Expected Behavior I'd expect the tokens to...