Description
Describe the bug
I'm following the blog Getting started with InstructLab for generative AI model tuning by @cedricclyburn.
My main goal is to create new information in an LLM. I was first trying with "Dad Jokes" and when I ran ilab generate after filling out a file as Cedric showed, I got the error
INFO 2025-06-06 16:15:37,450 root:352: Converting taxonomy to samples
failed to generate data with exception: Generating dataset failed with the following error: Error: No new leaf nodes found in the taxonomy.
After tweaking a number of things, I tried to see if my knowledge had been added, but it did not appear to be. So, just as an experiment, I removed all my files and tried to generate again, but I received the same error.
The messaging is not very enlightening. I realized that Cedric's post is close to a year old, and many of the commands in it have changed. Is there a better "Getting Started" post on how to add knowledge somewhere that you would recommend?
To Reproduce
Steps to reproduce the behavior:
- Go to '...'
- Click on '....'
- Scroll down to '....'
- See error
Expected behavior
An updated LLM!
Screenshots
Device Info (please complete the following information):
ilab system info
Platform:
sys.version: 3.11.12 (main, May 13 2025, 00:00:00) [GCC 15.1.1 20250425 (Red Hat 15.1.1-1)]
sys.platform: linux
os.name: posix
platform.release: 6.14.9-300.fc42.x86_64
platform.machine: x86_64
platform.node: tsweeney-thinkpadp1gen4i.bos.csb
platform.python_version: 3.11.12
os-release.ID: fedora
os-release.VERSION_ID: 42
os-release.PRETTY_NAME: Fedora Linux 42 (Workstation Edition)
os-release.VARIANT: Workstation Edition
memory.total: 62.49 GB
memory.available: 23.79 GB
memory.used: 18.62 GB
InstructLab:
instructlab.version: 0.25.0
instructlab-dolomite.version: 0.2.0
instructlab-eval.version: 0.5.1
instructlab-quantize.version: 0.1.0
instructlab-schema.version: 0.4.2
instructlab-sdg.version: 0.7.3
instructlab-training.version: 0.8.1
Torch:
torch.version: 2.5.1+cu124
torch.backends.cpu.capability: AVX512
torch.version.cuda: 12.4
torch.version.hip: None
torch.cuda.available: False
torch.backends.cuda.is_built: True
torch.backends.mps.is_built: False
torch.backends.mps.is_available: False
llama_cpp_python:
llama_cpp_python.version: 0.3.6
llama_cpp_python.supports_gpu_offload: False
Additional context