converting-pytorch-to-litert

Converts PyTorch models (e.g. ResNet, timm, HuggingFace transformers) directly to LiteRT (.tflite) flatbuffer format. Use when converting PyTorch models to TFLite, setting up export environments, or troubleshooting torch-to-litert conversion bugs. Don't use for ONNX exports or converting existing TensorFlow models.

Quality

67%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Failed to scan

The risk profile of this skill

Fix and improve this skill with Tessl

tessl review fix ./SKILL.md

Quality

Content

47%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The content establishes useful environment and dependency rules but is incomplete: it cuts off mid-code-block at the pip install step and never provides the actual conversion procedure or any validation. What exists is concise and well-organized, but the core actionable workflow is missing.

Suggestions

Add the actual PyTorch→LiteRT conversion code or command sequence after the dependencies section so the core task is executable, and close the unterminated ```bash code fence.

Reframe sections 1–4 as an ordered conversion workflow (prepare env → install deps → write/run conversion → verify output) with explicit validation checkpoints, e.g. inspect the generated .tflite with litert-cli or load it to confirm ops.

Trim the 'ATTENTION ALL AGENTS...based on recent project findings' preamble and fold the ai-edge-torch deprecation note into the litert-torch step to reduce padding.

Dimension	Reasoning	Score
Conciseness	The body is lean and rule-driven with little concept re-explanation, but the 'ATTENTION ALL AGENTS...based on recent project findings' preamble and slightly redundant deprecation framing add minor padding, placing it just below the fully lean 5.	4 / 5
Actionability	It gives concrete environment commands (python3.11, the pip install line) but the body ends at dependencies with no actual PyTorch→LiteRT conversion code or steps, leaving only high-level hints for the core task — matching the 'minimal concrete guidance' anchor.	2 / 5
Workflow Clarity	Sections 1–4 are a list of rules rather than a sequenced conversion workflow, there are no validation checkpoints for a fragile conversion, and the document is truncated mid-code-block, fitting the 'rough sequence, gaps, validation absent' anchor.	2 / 5
Progressive Disclosure	At 27 lines with organized numbered sections and no bundle files, it is well-structured for a simple skill, but the truncated/incomplete content and absence of any signaled deep references keep it below the clean 5.	4 / 5
	Total	12 / 20 Passed

Description

87%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, well-scoped description that clearly states capabilities, usage triggers, and out-of-scope cases. It is concise, third-person, and highly distinct from neighboring conversion skills.

Dimension	Reasoning	Score
Specificity	Names the domain and several concrete actions — 'Converts PyTorch models...directly to LiteRT (.tflite) flatbuffer format', 'setting up export environments', 'troubleshooting torch-to-litert conversion bugs' — with minor coverage gaps, fitting the 'several specific actions' anchor rather than the comprehensive 5.	4 / 5
Completeness	Explicitly answers both what ('Converts PyTorch models...directly to LiteRT (.tflite) flatbuffer format') and when ('Use when converting PyTorch models to TFLite, setting up export environments, or troubleshooting...') with concrete trigger phrases, matching the top anchor.	5 / 5
Trigger Term Quality	Includes natural user phrasing such as 'PyTorch models to TFLite', 'torch-to-litert conversion bugs', and the '.tflite' extension with LiteRT/TFLite synonyms, but a few common variations are missing, so it sits above the 3 anchor but below comprehensive 5.	4 / 5
Distinctiveness Conflict Risk	Occupies a clear niche (direct PyTorch→LiteRT) and adds an explicit negative boundary — 'Don't use for ONNX exports or converting existing TensorFlow models' — minimizing conflict with adjacent skills.	5 / 5
	Total	18 / 20 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: google-ai-edge/litert-torch
Path: SKILL.md
Commit: 70d7619

Reviewed: 4 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.