Skip to content

10 Complete Workflow (Kaggle)

Source: notebooks/10-complete-workflow-llamatelemetry-v0-1-1-e1.ipynb

Notebook focus

This page is a cell-by-cell walkthrough of the notebook, explaining the intent of each step and showing the exact code executed.

Cell-by-cell walkthrough

Cell 1 (Markdown)

10 Complete Workflow (Kaggle)

End-to-end pipeline: OTLP environment setup, server preset launch, instrumented client, and inference — all in one notebook.

What you will learn: - Load Grafana OTLP credentials from Kaggle Secrets - Start llama-server from a preset - Create an instrumented client with OpenTelemetry - Run inference with full observability

Requirements: Kaggle T4 x2. Optional: Grafana Cloud OTLP secrets.

Cell 2 (Markdown)

1) Install

Cell 3 (Code)

Summary: Installs required dependencies and runtime tools.

!pip -q install git+https://github.com/llamatelemetry/llamatelemetry.git@v0.1.1

Cell 4 (Markdown)

2) Load OTLP environment from Kaggle Secrets

Reads GRAFANA_OTLP_ENDPOINT and GRAFANA_OTLP_HEADERS (or OTLP_ENDPOINT / OTLP_TOKEN) from Kaggle Secrets and sets them as environment variables.

Cell 5 (Code)

Summary: Imports core libraries: llamatelemetry.

from llamatelemetry.kaggle.pipeline import (
    KagglePipelineConfig,
    load_grafana_otlp_env_from_kaggle,
    start_server_from_preset,
    setup_otel_and_client,
)
from llamatelemetry.kaggle import ServerPreset

otlp_env = load_grafana_otlp_env_from_kaggle()
print(f"OTLP configured: {bool(otlp_env)}")

Cell 6 (Markdown)

3) Start llama-server from a preset

Cell 7 (Code)

Summary: Works with GGUF models, quantization, or metadata.

model_path = "/kaggle/input/your-model/model.gguf"

manager = start_server_from_preset(model_path, ServerPreset.KAGGLE_DUAL_T4)
print(f"Server started: {manager.check_server_health()}")

Cell 8 (Markdown)

4) Setup OpenTelemetry + instrumented client

setup_otel_and_client() returns a dict with: - client — InstrumentedLlamaCppClient - tracer — OpenTelemetry tracer - meter — OpenTelemetry meter

Cell 9 (Code)

Summary: Sets up Graphistry for graph visualization or analytics.

cfg = KagglePipelineConfig(
    enable_graphistry=False,
    enable_llama_metrics=True,
)

ctx = setup_otel_and_client("http://127.0.0.1:8080", cfg)
client = ctx["client"]
print(f"Pipeline context keys: {list(ctx.keys())}")

Cell 10 (Markdown)

5) Run inference with telemetry

Cell 11 (Code)

Summary: Works with GGUF models, quantization, or metadata.

resp = client.chat_completions({
    "model": "local-gguf",
    "messages": [{"role": "user", "content": "Hello from the complete workflow!"}],
    "max_tokens": 64,
})
print(resp.choices[0].message.content)

Cell 12 (Markdown)

6) Check server metrics

Cell 13 (Code)

Summary: Imports core libraries: json.

import json

health = manager.get_health()
print(json.dumps(health, indent=2, default=str))

Cell 14 (Markdown)

7) Cleanup

Cell 15 (Code)

Summary: Cleans up or shuts down running resources.

manager.stop_server()
print("Done.")