Can't really use an OpenAI compatible API server for ilab data generate since it adds the path to the model name

[ivanbaldo]

Bug description If we want to use the real OpenAI API server with for example the gpt-4o model, currently we can't, since it sends /instructlab/models/gpt-4o through the API instead. This doesn't happen with InstructLab 0.17.1, the bug is in the ilab wrapper of RHELAI.

To Reproduce

export HF_TOKEN=notneeded
ilab init --non-interactive
ilab download --repository ibm/granite-7b-base
mkdir taxonomy/knowledge/netlabs
cat >taxonomy/knowledge/netlabs/qna.yaml <<EOT
version: 2
task_description: 'Test con datos locales de Netlabs'
created_by: ibaldo@netlabs.com.uy
domain: netlabs
seed_examples:
  - question: When did the 2024 Oscars happen?
    answer: |
      The 2024 Oscars were held on March 10, 2024.
  - question: What film had the most Oscar nominations in 2024?
    answer: |
      Oppenheimer had 13 Oscar nominations.
  - question: Who presented the 2024 Oscar for Best Original Screenplay and Best Adapted Screenplay?
    answer: |
      Octavia Spencer presented the award for Best Original Screenplay and Best Adapted Screenplay at the 2024 Oscars.
  - question: Who hosted the 2024 Oscars?
    answer: |
      Jimmy Kimmel hosted the 96th Academy Awards ceremony.
  - question: At the 2024 Oscars, who were the nominees for best director and who won?
    answer: |
      The nominees for director at the 2024 Oscars was Christopher Nolan for Oppenheimer,
      Justine Triet for Anatomy of a Fall, Martin Scorsese for Killers of the Flower Moon,
      Yorgos Lanthimos for Poor Things, and Jonathan Glazer for The Zone of Interest.
      Christopher Nolan won best director for Oppenheimer.
  - question: Did Billie Eilish perform at the 2024 Oscars?
    answer: |
      Yes Billie Eilish performed "What Was I Made For?" from Barbie at the 2024 Oscars.
document:
  repo: https://github.com/juliadenham/oscars2024_knowledge.git
  commit: main
  patterns:
    - oscars2024_results.md
EOT
ilab taxonomy diff
ilab generate --endpoint-url https://api.openai.com:443/v1 --model gpt-4o --api-key secret

Unexpected behavior

ilab generate --endpoint-url https://api.openai.com:443/v1 --model gpt-4o --api-key secret

WARNING: You need at least 2 GPUs to load full precision models
ilab generate --endpoint-url https://api.openai.com:443/v1 --model gpt-4o --api-key secret --model-family mixtral --num-instructions 5000 --model /instructlab/models/gpt-4o
You are using an aliased command, this will be deprecated in a future release. Please consider using `ilab data generate` instead
Generating synthetic data using '/instructlab/models/gpt-4o' model, taxonomy:'taxonomy' against https://api.openai.com:443/v1 server
DEBUG 2024-07-03 21:08:29,384 utils.py:581: read_taxonomy Found new taxonomy files:
DEBUG 2024-07-03 21:08:29,384 utils.py:583: read_taxonomy * knowledge/netlabs/qna.yaml
DEBUG 2024-07-03 21:08:29,673 utils.py:214: get_documents Processing files...
DEBUG 2024-07-03 21:08:29,674 utils.py:539: read_taxonomy_file Content from git repo fetched
Cannot find prompt.txt. Using default prompt depending on model-family.
DEBUG 2024-07-03 21:08:29,675 generate_data.py:417: generate_data Loaded 6 human-written seed instructions from taxonomy
DEBUG 2024-07-03 21:08:29,678 generate_data.py:461: generate_data Generating to: generated/generated_gpt-4o_2024-07-03T21_08_29.json
  0%|                                                                                                                            | 0/5000 [00:00<?, ?it/s]
Synthesizing new instructions. If you aren't satisfied with the generated instructions, interrupt training (Ctrl-C) and try adjusting your YAML files. Adding more examples may help.
INFO 2024-07-03 21:08:29,685 generate_data.py:505: generate_data Selected taxonomy path knowledge->netlabs
Generating dataset failed with the following error: Model /instructlab/models/gpt-4o is not served by the server.
These are the served models
['whisper-1', 'tts-1', 'dall-e-2', 'tts-1-hd-1106', 'tts-1-hd', 'gpt-4-turbo-2024-04-09', 'gpt-4-turbo', 
'gpt-3.5-turbo-1106', 'dall-e-3', 'gpt-4-0125-preview', 'gpt-4-turbo-preview', 'text-embedding-3-small', 
'text-embedding-3-large', 'gpt-3.5-turbo-16k', 'gpt-4-1106-preview', 'babbage-002', 'gpt-4o-2024-05-13', 
'gpt-4', 'gpt-4-0613', 'gpt-3.5-turbo-0125', 'tts-1-1106', 'gpt-3.5-turbo', 'gpt-3.5-turbo-instruct', 
'gpt-3.5-turbo-instruct-0914', 'text-embedding-ada-002', 'davinci-002', 'gpt-4o']
  0%|                                                                                                                            | 0/5000 [00:00<?, ?it/s]

Device Info

ilab sysinfo
WARNING: You need at least 2 GPUs to load full precision models
/usr/local/lib64/python3.11/site-packages/torch/cuda/__init__.py:619: UserWarning: Can't initialize NVML
  warnings.warn("Can't initialize NVML")
instructlab.version: 0.17.1
sys.version: 3.11.7 (main, May 16 2024, 00:00:00) [GCC 11.4.1 20231218 (Red Hat 11.4.1-3)]
sys.platform: linux
os.name: posix
platform.release: 5.14.0-467.el9.x86_64
platform.machine: x86_64
os-release.ID: rhel
os-release.VERSION_ID: 9.4
os-release.PRETTY_NAME: Red Hat Enterprise Linux 9.4 (Plow)
torch.version: 2.3.1+cu121
torch.backends.cpu.capability: AVX2
torch.version.cuda: 12.1
torch.version.hip: None
torch.cuda.available: True
torch.backends.cuda.is_built: True
torch.backends.mps.is_built: False
torch.backends.mps.is_available: False
torch.cuda.bf16: True
torch.cuda.current: 0
torch.cuda.0.name: NVIDIA A10G
torch.cuda.0.free: 21.7
torch.cuda.0.total: 22.0
torch.cuda.0.capability: 8.6
llama_cpp_python.version: 0.2.75
llama_cpp_python.supports_gpu_offload: True