UniRec-0.1B — Core ML
Core ML conversion of
topdu/unirec-0.1b,
a unified recogniser for printed text, mathematical formulas and tables. Packaged
for on-device use by
MathOCR, a macOS app
that turns PDF pages without a text layer into Markdown with LaTeX.
Nothing here is a new model. The weights are the original authors'; this
repository only changes the format.
Contents
| File | Size | Input | Output |
|---|
UniRecEncoder.mlpackage | 82 MB | pixel_values [1,3,H,W] | cross_k, cross_v [6,1,6,S,128] |
UniRecDecoder.mlpackage | 191 MB | one token + cross-attention K/V | logits [1,56371] |
Both float16, minimum_deployment_target = macOS 15.
The split follows the maintainers' own ONNX export: the encoder also computes
the decoder's cross-attention key and value projections, which depend only on
the image and so are evaluated once per region rather than once per token.
The decoder is stateful — its self-attention KV cache lives on the Core ML
side rather than being passed in and out each step.
Requirements and limits
Run with compute units restricted to CPU and GPU.
1let configuration = MLModelConfiguration()
2configuration.computeUnits = .cpuAndGPU
This is not a performance preference. With the Neural Engine enabled, the
encoder is executed incorrectly: at a 192×640 input its output deviates from the
reference by 56%, against 0.19% on CPU+GPU. FocalSVTR's focal modulation
uses depthwise convolutions with kernels up to 15×15, which appears to be the
cause. The failure is silent — the model returns plausible but wrong text rather
than an error.
Other limits, fixed at conversion time:
- 512 encoder positions (
cross_len marks how many are real; pad the rest).
That is a region of roughly 960×544 px. Larger regions must be split.
- 512 generated tokens — the self-attention cache size.
Preprocessing
Identical to the original model:
- RGB.
- Fit inside 960×1408 preserving aspect ratio. Images already smaller are not
enlarged.
- Round both sides down to a multiple of 64, minimum 64.
- Bicubic resample.
- Scale to
[0,1], then (x - 0.5) / 0.5.
NCHW, float32.
Decoding
Greedy. Start from bos = 0, stop at eos = 2. The position index follows
M2M100's convention: position = pad_token_id + 1 + step, i.e. 2 + step.
Token IDs map to strings through
unirec_tokenizer_mapping.json in
topdu/unirec_0_1b_onnx.
Formulas are emitted as
\( … \) and
\[ … \]; tables as HTML.
Verification
Checked against the maintainers' ONNX export on six rendered crops — an inline
formula, three display equations, a nested radical and a multi-line paragraph.
All six produce token-for-token identical greedy sequences.
On an M1: encoder 29–217 ms per region, decoder ~10 ms per token.
The conversion and verification scripts are in the MathOCR repository under
tools/unirec_coreml/.
Licence and attribution
Apache-2.0, inherited from the original model.
- Original model and weights: topdu/unirec-0.1b
- Source implementation: Topdu/OpenOCR
Changes made in this redistribution, as Apache-2.0 §4(b) requires:
- Exported from PyTorch to Core ML (
.mlpackage, ML Program), float16.
- The decoding step was reimplemented with a fixed-size KV cache updated by a
masked write, because HuggingFace's
Cache class does not trace to a static
graph. Same weights, same arithmetic — verified against the original.
- The encoder additionally returns the decoder's cross-attention K/V
projections, matching the upstream ONNX export's split.
