The newly introduced low_cpu_mem_usage argument was not propagated to the add_adapter method of all PeftModel task types. This is now fixed and tests were added.
Tests were only added for LoRA to avoid running many relatively slow and redundant tests and because low_cpu_mem_usage is not implemented for prompt learning models.
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.
The newly introduced
low_cpu_mem_usage
argument was not propagated to theadd_adapter
method of allPeftModel
task types. This is now fixed and tests were added.Tests were only added for LoRA to avoid running many relatively slow and redundant tests and because
low_cpu_mem_usage
is not implemented for prompt learning models.