vit-mic

Slim model weights, executable usage and framework expectations in one repository.

• Contract manifest
• Transformers usage
• Transformers expectations
• Fixture provenance

Contract

vit model integration contract

Public Transformers usage validated on a slim fixture on an NVIDIA A10G, with default, eager and bfloat16 variants and fresh-process baseline recording. This prototype declares fixture-only targets; full checkpoint coverage is separate.

Regeneration evidence.

Slim model

Slim fixture: mic-vit-slim

A 0.2M-parameter fixture for the vit MIC, covering google/vit-base-patch16-224. It is a test artifact, not a usable model, tagged test-artifact to keep it out of candidate-model search.

• Fixture key 3fc8366aafe8. It covers every checkpoint whose processor files are byte-identical and whose config matches after shrinking.
• Same architecture, smaller config: ViTForImageClassification, shrunk by the generic rule in sweep.py (num_hidden_layers 2).
• Processor files copied byte for byte from google/vit-base-patch16-224: preprocessor_config.json.
• Overfit to the contract's training step (300 steps). Random weights failed the input-sensitivity gate: pixel_values:reverse moved the logits by only 5.9e-05 (threshold 0.001).
• Same tensor names as the real checkpoint. The real checkpoint has every parameter, and so does the fixture.
• Saved in float32, the dtype google/vit-base-patch16-224 loads in by default.
• Finite in float32, bfloat16, and float16 (forward, loss, gradients).

Built by build_fixtures.py; build.json records the pinned sources, gate results, and file hashes. This repository includes the slim weights, usage and framework expectations; rebuild the other files with python example/build_fixtures.py --fixture mic-vit-slim.