tocmo0nlord/axolotl
1
0
Fork 0
Code Issues Pull Requests Actions 3 Packages Projects Releases Wiki Activity

Built site for gh-pages

Browse Source
This commit is contained in:
Quarto GHA Workflow Runner
2025-02-18 15:07:29 +00:00
parent 46dcae42f1
commit c6be6e42bd
4 changed files with 42 additions and 42 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
docs/lora_optims.html
View File

@@ -415,7 +415,7 @@ pre > code.sourceCode > span > a:first-child::before { text-decoration: underlin
<ul>
<li>One or more NVIDIA or AMD GPUs (in order to use the Triton kernels)
<ul>
<li>AMD can be used with experimental Triton support by setting the environment variable <code>TORCH_ROCM_AOTRITON_ENABLE_EXPERIMENTAL=1</code></li>
<li>Note: Set <code>TORCH_ROCM_AOTRITON_ENABLE_EXPERIMENTAL=1</code> to enable <a href="https://github.com/ROCm/aotriton/issues/16#issuecomment-2346675491">memory-efficient attention on AMD GPUs</a></li>
</ul></li>
<li>Targeted LoRA adapters cannot use Dropout
<ul>