Model card

README.md

@@ -0,0 +1,53 @@
+---
+license: mit
+base_model: microsoft/unixcoder-base
+library_name: transformers
+pipeline_tag: text-classification
+datasets:
+- basakdemirok/AIGCodeSet
+tags:
+- code
+- python
+- ai-generated-code-detection
+---
+
+# CODer: UniXcoder checkpoints
+
+Fine-tuned [UniXcoder](https://huggingface.co/microsoft/unixcoder-base) classifiers that detect LLM-written Python code, trained on [AIGCodeSet](https://huggingface.co/datasets/basakdemirok/AIGCodeSet). They are the models behind *Detecting LLM-Written Code on AIGCodeSet: Tuned Code Encoders versus N-gram Baselines*. The code, the predictions and the paper are at [github.com/mparts/CODer](https://github.com/mparts/CODer).
+
+## Layout
+
+24 checkpoints: six seeds (`seed3` to `seed8`) × four task variants. The paper's seed ensemble averages the six seeds' probabilities.
+
+| Folder | Task | Labels |
+|---|---|---|
+| `seed<N>/binary` | human vs LLM, all data | 0 human, 1 ai |
+| `seed<N>/binary_generate` | human vs LLM, without LLM-edited code | 0 human, 1 ai |
+| `seed<N>/multiclass` | human or which LLM, all data | 0 Human, 1 GEMINI, 2 LLAMA, 3 CODESTRAL |
+| `seed<N>/multiclass_generate` | human or which LLM, without LLM-edited code | 0 Human, 1 GEMINI, 2 LLAMA, 3 CODESTRAL |
+
+The configs keep the default label names (`LABEL_0`, `LABEL_1`, ...); the table gives their meaning.
+
+## Usage
+
+~~~python
+from transformers import AutoModelForSequenceClassification, AutoTokenizer
+
+repo, variant = "mparts/CODer-unixcoder", "seed3/binary"
+tok = AutoTokenizer.from_pretrained(repo, subfolder=variant)
+model = AutoModelForSequenceClassification.from_pretrained(repo, subfolder=variant)
+~~~
+
+The models were trained on whitespace-normalised code (line endings unified, tabs expanded to four spaces, trailing spaces and outer blank lines removed: `normalise()` in `src/code_origin/data/preprocess.py`), truncated to 256 tokens. Normalise inputs the same way.
+
+## Training
+
+Fine-tuned on the train split of the repository's problem-grouped split of AIGCodeSet, in which no programming problem is shared between train, dev and test: 4 epochs, learning rate 2e-5, batch size 16, max length 256, weight decay 0.01, warmup ratio 0.1. The saved weights are from the epoch with the best dev macro-F1.
+
+## Limitations
+
+Trained on Python solutions to 317 CodeNet problems, against three LLMs (CodeLlama 34B, Codestral 22B, Gemini 1.5 Flash). It is a research artefact for this setting, not a general-purpose detector of AI-written code.
+
+## License
+
+The weights are released under the MIT license, like the code. The training data, AIGCodeSet, is under CDLA-Permissive-2.0.