I've been struggling with this error a day now. I'm trying to run Kubeflow pipeline for model fine tuning using VertexAI on colab. This is the code that I'm using
# Get the current date for model naming
date = datetime.datetime.now().strftime("%Y-%m-%d")
MODEL_NAME = f"qa-model-{date}"
# Pipeline arguments
pipeline_arguments = {
"model_display_name": MODEL_NAME,
"location": REGION,
"project": PROJECT_ID,
"large_model_reference": "gemini-1.0-pro-002", # Model reference for question answering
"train_steps": TRAINING_STEPS,
"dataset_uri": TRAINING_DATA_URI,
"evaluation_interval": EVALUATION_INTERVAL,
"evaluation_data_uri": EVALUATION_DATA_URI,
}
# Define the pipeline template path for model tuning
template_path = "https://us-kfp.pkg.dev/ml-pipeline/large-language-model-pipelines/tune-large-model/v2.0.0"
# Create a pipeline job for fine-tuning the model
job = aiplatform.PipelineJob(
template_path=template_path,
display_name="fine-tune-gemini-qa",
parameter_values=pipeline_arguments,
location=REGION,
pipeline_root="gs://zeinab-llmops-vertexai/pipeline_root", # GCS path for storing pipeline artifacts
enable_caching=True
)
# Submit the pipeline job
job.submit()
Obviously I'm trying to use Gemini, but I get the following error which complains about using Bison and suggests me to use Gemini!!
_InactiveRpcError: <_InactiveRpcError of RPC that terminated with:
status = StatusCode.FAILED_PRECONDITION
details = "Bison model tuning is deprecated. Please migrate to Gemini tuning."
debug_error_string = "UNKNOWN:Error received from peer ipv4:172.217.203.95:443 {created_time:"2025-02-28T06:30:14.731451386+00:00", grpc_status:9, grpc_message:"Bison model tuning is deprecated. Please migrate to Gemini tuning."}"
>
The above exception was the direct cause of the following exception:
FailedPrecondition Traceback (most recent call last)
/usr/local/lib/python3.11/dist-packages/google/api_core/grpc_helpers.py in error_remapped_callable(*args, **kwargs)
76 return callable_(*args, **kwargs)
77 except grpc.RpcError as exc:
---> 78 raise exceptions.from_grpc_error(exc) from exc
79
80 return error_remapped_callable
FailedPrecondition: 400 Bison model tuning is deprecated. Please migrate to Gemini tuning.
I have a paid GCP account with enough credit, and I have Vertex AI user role on the project. I would appreciate any help!
You can find how to tune Gemini here: https://cloud.google.com/vertex-ai/generative-ai/docs/models/gemini-use-supervised-tuning
The SDK syntax you are using is indeed from deprecated bison