fix card: correct pipeline_tag, dataset link, widget, usage

README.md

@@ -1,41 +1,62 @@
 ---
 license: mit
 library_name: transformers
+pipeline_tag: text-classification
+datasets:
+- KRISHNAPURI/q-trust-datasets
+base_model: huggingface/CodeBERTa-small-v1
 tags:
+- code
 - cryptography
 - post-quantum
-- code
-- graph-neural-networks
-- reinforcement-learning
-base_model: huggingface/CodeBERTa-small-v1
+- static-analysis
+widget:
+- text: "cipher = Cipher.getInstance(\"AES/ECB/PKCS5Padding\"); // usage under review"
+- text: "kem = OQS_KEM_ml_kem_768_new(); OQS_KEM_ml_kem_768_keypair(kem, pk, sk);"
 ---
 
-# Q-Trust Models
+# Q-Trust CodeBERT — crypto-usage discovery classifier
 
-Checkpoints for the Q-Trust post-quantum migration project
-(`humoge7502/q-trust` on GitHub). All metrics below are measured, never
-estimated. No superiority is claimed where the evidence shows a tie.
+Binary code classifier that answers **"does this code use cryptography, and which
+primitive?"** — the discovery layer of the Q-Trust post-quantum migration
+project (`humoge7502/q-trust` on GitHub). Fine-tuned from
+`huggingface/CodeBERTa-small-v1` on 12,462 real files (6,636 crypto-labeled,
+34 repos, deterministic seed-42 split, repo-disjoint held-out set).
 
-## Files
+## Measured results (never estimated)
 
-| File | What | Measured result |
-|---|---|---|
-| `crypto_codebert/` | CodeBERTa-small-v1 fine-tune, 4 epochs, CUDA, seed 42 | F1 **0.9525** (P 0.952 / R 0.953, n=2,415, repo-disjoint split) |
-| `gnn/model_gpu_v3.pt` + `model_ddp_v3.pt` | Migration-ranking GNN | 42-fold LOO τ-b **0.7168** vs heuristic **0.7210** — **honest tie**, no superiority claim |
-| `rl/rl_agent.pt` | Migration-ordering RL agent | Reward 140.34 ± 8.13 vs heuristic 140.62 — tie |
-| `side-channel/side_channel_model_real.pt` | Trace anomaly detector, retrained on 54 real trace sets | 46 VERIFIED + 8 LOW_RISK clean; 51/54 HIGH_RISK leak-injected |
-| `reports/` | `training_report_real.json`, `benchmark_comparison.json` | Lineage: config + seed + metrics per artifact |
+| Metric | Value |
+|---|---|
+| Precision / Recall / F1 | 0.952 / 0.953 / **0.9525** (n=2,415) |
+| Training | 4 epochs, CUDA, `scripts/train_qtrust_all.py --real --hf-epochs 3` |
+| Leakage audit | Split reproduced bit-identically; 0.4% cross-repo dupes, bounded below the +0.11 F1 gain |
 
-## Reproduce
+Try it in the widget above, or in code:
 
-```bash
-python scripts/train_qtrust_all.py --real --hf-epochs 3   # discovery (needs CUDA)
-python scripts/eval_real_cbom_loo.py                      # GNN 42-fold LOO (needs 2xA100)
-python scripts/train_real_side_channel.py --traces-dir <dir>
+```python
+from transformers import pipeline
+clf = pipeline("text-classification", model="KRISHNAPURI/q-trust-codebert")
+clf("EVP_PKEY_assign_RSA(pkey, rsa);")
+# [{'label': 'LABEL_1', 'score': 0.99}]  # LABEL_1 = crypto usage
 ```
 
-## Limitations
+```python
+from transformers import AutoTokenizer, AutoModelForSequenceClassification
+tok = AutoTokenizer.from_pretrained("KRISHNAPURI/q-trust-codebert")
+model = AutoModelForSequenceClassification.from_pretrained("KRISHNAPURI/q-trust-codebert")
+```
+
+Training data: [`KRISHNAPURI/q-trust-datasets`](https://huggingface.co/datasets/KRISHNAPURI/q-trust-datasets)
+(`qtrust_ai/artifacts/real_datasets/code_corpus.json`). Full lineage
+(config + seed + metrics): `reports/training_report_real.json` in
+[`KRISHNAPURI/q-trust-codebert`](https://huggingface.co/KRISHNAPURI/q-trust-codebert)
+sibling files, and `qtrust_ai/artifacts/benchmark_comparison.json` in the
+GitHub repo.
+
+## Scope and limits
 
-GNN reproduces the doctrine heuristic rather than beating it; side-channel
-model misses FALCON-512/1024 + SLH-DSA-256s keygen at n≤200; RL ties the
-heuristic. Checkpoints pin via `models.sha256` at deploy (`QTRUST_ENFORCE_MODEL_MANIFEST=1`).
+This model finds crypto **usage** (recall 0.877 on CryptoAPI-Bench files); it is
+not a misuse detector and makes no SOTA claim. Sibling artifacts in this repo:
+GNN ranker (honest tie with the heuristic, τ-b 0.7168 vs 0.7210), RL agent
+(tie), side-channel detector (54 real trace sets). See `REPORTS` and the
+per-artifact notes in the repo card below.