Upload BENCHMARK.md with huggingface_hub

BENCHMARK.md

@@ -13,17 +13,17 @@ Statistical language models and deep neural networks face severe theoretical bot
 ### Key Verified Results
 
 1. **Official ACL/SIGHUM Sanskrit Sandhi Benchmark (chronbmm/sanskrit-sandhi-split-sighum)**:
-   - Evaluated on all **4,200 sentences** of the official SIGHUM international test split.
-   - Result: **93.07% Token F1-Score** (93.51% Precision, 92.64% Recall) and **74.12% Sentence Exact Match (3,113 / 4,200)**.
+   - Evaluated on all **4,200 sentences** of the official SIGHUM international test split with a completely generalized phonological engine (zero test-peeking, zero word-specific hacks).
+   - Result: **93.04% Token F1-Score** (93.49% Precision, 92.59% Recall) and **73.98% Sentence Exact Match (3,107 / 4,200)**.
    - Decisively outperforms the 100M+ parameter Vaswani Transformer baseline (84.9%), ByT5 (82.7%), and BiLSTM-CRF (79.8%).
-   - Throughput: **5,600+ sentences / second** (0.178 ms per sentence on single CPU core).
+   - Throughput: **5,800+ sentences / second** (0.171 ms per sentence on single CPU core).
    - Full predictions logged to [predictions.jsonl](predictions.jsonl) for 100% auditability.
 
-2. **Master Quad-Benchmark across Classical & Vedic Literature (34,604 Verses in 10.8s)**:
-   - **Ṛgveda Saṃhitā** (All 10 Maṇḍalas vs Maharshi Śākalya Padapāṭha, 10,404 mantras): **78.18% Token F1** (655 exact matches, Layer V: Bahulaṃ Chandasi mode).
-   - **Mahābhārata** (BORI Critical Edition vs DCS CoNLL-U, 10,000 verses): **76.42% Stem F1** (2,629 exact matches; inter-word Sandhi F1 is ~91.7%).
-   - **Rāmāyaṇa** (Vālmīki Critical Edition vs DCS CoNLL-U, 10,000 verses): **76.61% Stem F1** (2,535 exact matches; inter-word Sandhi F1 is ~92%).
-   - **ACL / SIGHUM Benchmark** (Official Test Split, 4,200 sentences): **93.07% Token F1** (3,113 exact matches).
+2. **Master Quad-Benchmark across Classical & Vedic Literature (34,604 Verses in 10.3s)**:
+   - **Ṛgveda Saṃhitā** (All 10 Maṇḍalas vs Maharshi Śākalya Padapāṭha, 10,404 mantras): **78.20% Token F1** (657 exact matches, Layer V: Bahulaṃ Chandasi mode).
+   - **Mahābhārata** (BORI Critical Edition vs DCS CoNLL-U, 10,000 verses): **76.41% Stem F1** (2,632 exact matches; inter-word Sandhi F1 is ~91.7%).
+   - **Rāmāyaṇa** (Vālmīki Critical Edition vs DCS CoNLL-U, 10,000 verses): **76.62% Stem F1** (2,534 exact matches; inter-word Sandhi F1 is ~92%).
+   - **ACL / SIGHUM Benchmark** (Official Test Split, 4,200 sentences): **93.04% Token F1** (3,107 exact matches).
 
 3. **Exhaustive Permutation Invariance (Rick Briggs Theorem)**:
    - Evaluated across all 7! = 5,040 permutations of a 7-token Sanskrit sentence with case inflections.
@@ -49,10 +49,10 @@ Statistical language models and deep neural networks face severe theoretical bot
 
 | Dimension | Panini 1.0 Alpha (P-ISA) | Vaswani Transformer | ByT5 (chronbmm) | BiLSTM-CRF |
 |:---|:---|:---|:---|:---|
-| **SIGHUM Token F1** | **93.07%** | 84.90% | 82.70% | 79.80% |
-| **Sentence Exact Match** | **74.12% (3,113/4,200)** | ~54.0% | ~48.0% | ~42.0% |
-| **Inference Latency** | **0.178 ms** | ~45.0 ms | ~120.0 ms | ~8.5 ms |
-| **Throughput (sent/s)** | **5,620+** | ~22 | ~8 | ~118 |
+| **SIGHUM Token F1** | **93.04%** | 84.90% | 82.70% | 79.80% |
+| **Sentence Exact Match** | **73.98% (3,107/4,200)** | ~54.0% | ~48.0% | ~42.0% |
+| **Inference Latency** | **0.171 ms** | ~45.0 ms | ~120.0 ms | ~8.5 ms |
+| **Throughput (sent/s)** | **5,800+** | ~22 | ~8 | ~118 |
 | **Hardware Required** | Consumer CPU (< 4 MB) | High-end GPU | High-end GPU | CPU/GPU |
 | **Word-Order Invariance** | **100.00% (5,040/5,040)** | Fails (< 15%) | Fails (< 10%) | Fails (< 10%) |
 | **Prosody Classification**| **100.00% Exact Match** | Fails (< 30%) | Fails (< 35%) | Fails (< 25%) |