haoline 0.2.0

HaoLine (好线) - Universal Model Inspector. See what's really inside your models.

HaoLine (好线)

Universal Model Inspector — See what's really inside your models.

A graph-level analysis tool that inspects neural network architectures (ONNX, PyTorch, TensorFlow, TensorRT), computes static complexity metrics, surfaces structural risk signals, and auto-generates human-readable reports.

---

Installation

pip install haoline

# With LLM summaries (requires API key)
pip install haoline[llm]

# With all features (LLM, PDF export, GPU metrics)
pip install haoline[full]

---

Quick Start

Analyze a Model

# Basic analysis
haoline analyze model.onnx

# With hardware profile
haoline analyze model.onnx --hardware rtx4090

# Generate HTML report with charts
haoline analyze model.onnx --out-html report.html --with-plots

# Full analysis with AI summary
haoline analyze model.onnx --hardware auto --with-plots --llm-summary --out-html full_report.html

Compare Model Variants

# Compare FP32, FP16, INT8 quantizations
haoline compare \
  --models resnet_fp32.onnx resnet_fp16.onnx resnet_int8.onnx \
  --baseline-precision fp32 \
  --out-html comparison.html --with-charts

---

Features

• Parameter counts - Per node, per block, and globally
• FLOP estimates - Identify compute hotspots
• Memory analysis - Peak activation memory and VRAM requirements
• Risk signals - Detect problematic architecture patterns
• Hardware estimates - GPU utilization predictions for 30+ NVIDIA profiles
• Runtime profiling - Actual inference benchmarks with ONNX Runtime
• Visualizations - Operator histograms, parameter/FLOPs distribution charts
• LLM Summaries - AI-generated executive summaries
• Interactive graphs - D3.js neural network visualizations
• Shareable Reports - HTML, Markdown, PDF, and JSON output formats

---

Supported Formats

Format	Read	Write
ONNX	✅	-
PyTorch (.pt, .pth)	✅	-
TensorFlow SavedModel	✅	-
TensorRT Engine	🔜	-
SafeTensors	🔜	-

---

Python API

from haoline import ModelInspector

inspector = ModelInspector()
report = inspector.inspect("model.onnx")

# Get metrics
print(f"Parameters: {report.params.total:,}")
print(f"FLOPs: {report.flops.total:,}")
print(f"Peak Memory: {report.memory.peak_activation_bytes / 1e9:.2f} GB")

# Export
report.to_json("report.json")
report.to_markdown("model_card.md")
report.to_html("report.html")

---

LLM Setup (Optional)

HaoLine can generate AI-powered summaries of your model architectures. Install with LLM support and set the API key for your preferred provider:

pip install haoline[llm]

Provider	Environment Variable	Get API Key
OpenAI	OPENAI_API_KEY	platform.openai.com
Anthropic	ANTHROPIC_API_KEY	console.anthropic.com
Google Gemini	GOOGLE_API_KEY	aistudio.google.com
xAI Grok	XAI_API_KEY	console.x.ai

# Set your preferred provider's API key
export OPENAI_API_KEY="sk-..."           # OpenAI
export ANTHROPIC_API_KEY="sk-ant-..."    # Anthropic
export GOOGLE_API_KEY="..."              # Google Gemini
export XAI_API_KEY="xai-..."             # xAI Grok

# Windows PowerShell
$env:OPENAI_API_KEY="sk-..."

# Use in analysis (auto-detects available provider)
haoline model.onnx --llm-summary

---

Security Notice

⚠️ Loading untrusted models is inherently risky.

Like PyTorch's torch.load(), HaoLine uses pickle and numpy.load(allow_pickle=True) when loading JAX model parameters. These can execute arbitrary code if the model file is malicious.

Best practices:

• Only analyze models from trusted sources
• Run in a sandboxed environment (Docker, VM) when analyzing unknown models
• Review model provenance before loading

---

License

MIT License - See LICENSE for details.

---

Etymology

HaoLine (好线) combines:

• 好 (hǎo) = "good" in Chinese
• A personal touch from the author

"The good path through your neural network."