CAMeLBERT-mix-sentiment — Apple Core AI, Arabic sentiment

A Core AI (.aimodel) build of CAMeL-Lab/bert-base-arabic-camelbert-mix-sentiment for on-device Arabic sentiment on Apple Silicon.

Upstream ships pytorch_model.bin, flax_model.msgpack and tf_model.h5 — and no MLX, no ONNX, no Core AI. 100,274 downloads, and this is the Core AI build.

Outputs

One graph, two outputs. Keeping both costs nothing at this size and means neither the trained classifier nor the reusable vectors are thrown away.

Read this before reading the outputs

Select outputs by shape, not by name. The runtime names tensors after the op that produced them. On this graph that is div_1 for the vectors and addmm_73 for the logits — the names in the wrapper's dict do not survive into the .aimodel. A one-line size == 3 / size == 768 pick is the robust way, and it is what the measurement script does.

Verified equivalence

Compared against the original PyTorch model on six fixed inputs at the exported shape:

Worst across all six: 2.38e-06, and 6/6 labels agree with PyTorch. That residual is relative ~1e-6 on logits of magnitude ~3 — float32 rounding accumulated through the classifier layer, not a difference in the model.

One colloquial case, and what it does not show

الطلب اتأخر أسبوعين وما حد رد عليا. ("the order was two weeks late and nobody answered me") is classified neutral by this model. PyTorch classifies it neutral too and the conversion reproduces that exactly — so this is the model's own behaviour on a dialect complaint, and it carries through unchanged.

One sentence, not a survey: dialect sentiment corpora were not run, and no claim is made about how often this happens.

Use

import asyncio, numpy as np
from coreai.runtime import AIModel, NDArray, SpecializationOptions
from transformers import AutoTokenizer

LABELS = ["positive", "negative", "neutral"]
SEQUENCE_LENGTH = 512          # fixed shape: pad to it

async def analyse(text: str):
    tok = AutoTokenizer.from_pretrained(".")
    model = await AIModel.load("camelbert-arabic-sentiment.aimodel", SpecializationOptions.default())
    fn = model.load_function("main")
    enc = tok([text], padding="max_length", truncation=True, max_length=SEQUENCE_LENGTH,
              return_tensors="np")
    out = await fn({"input_ids": NDArray(enc["input_ids"].astype("int32")),
                    "attention_mask": NDArray(enc["attention_mask"].astype("int32"))})
    flat = {k: np.asarray(v.numpy()).reshape(-1) for k, v in out.items()}
    logits = next(v for v in flat.values() if v.size == 3)
    vectors = next(v for v in flat.values() if v.size == 768)
    return LABELS[int(np.argmax(logits))], logits, vectors

print(asyncio.run(analyse("الخدمة كانت ممتازة وأنصح به بشدة."))[0])   # positive

What is in here

Limits

• Fixed at 512 tokens; pad rather than passing a shorter sequence.
• Six inputs measured. Dialect variation, long documents and the 512-token boundary were not.
• No speed or memory benchmark. Equivalence only.
• The classification is 3-way and single-label; no confidence calibration was measured.

Licence and attribution

Apache-2.0, as declared by upstream. A derivative work of CAMeL-Lab/bert-base-arabic-camelbert-mix-sentiment; credit for the model is theirs.