Add the epoch-14 DrugCLIP checkpoint scored in PR 94

README.md

@@ -0,0 +1,32 @@
+---
+license: other
+tags:
+  - drug-discovery
+  - drugclip
+---
+
+# DrugCLIP epoch 14
+
+Selected fine-tune checkpoint from the paper-scale DrugCLIP run. Held-out scores below are the corrected rescore in [this report](https://github.com/openssi-dev/ai4sci-tasks/pull/94#issuecomment-6174348916).
+
+`model.pt` SHA-256: `bb108dcec5608b04eee546a0ed57248ab5f956cb61799a8c5611ab9c8da7c772`
+
+Equal-target mean EF at 1% is **16.108**.
+
+| Benchmark | Targets | ROC-AUC | BEDROC (80.5) | EF 1% |
+| --- | ---: | ---: | ---: | ---: |
+| DUD-E | 102 | 0.769 | 0.295 | 18.056 |
+| LIT-PCBA | 15 | 0.577 | 0.038 | 2.861 |
+
+Random ranking is EF 1.
+
+The checkpoint is fine-tune epoch 14, selected by development BEDROC 0.4288. Fine-tuning stopped at epoch 34 after patience 20. The model has 95,305,418 parameters. Pocket atoms were scored with element tokens, a 511-atom test crop, and the author pooling (DUD-E median/MAD, then max; LIT-PCBA max).
+
+Exact benchmark means from that rescore:
+
+| Benchmark | ROC-AUC | BEDROC (80.5) | EF 1% |
+| --- | ---: | ---: | ---: |
+| DUD-E | 0.7686053801137084 | 0.29458757122655027 | 18.055988235981932 |
+| LIT-PCBA | 0.5774894830416493 | 0.0381821961785978 | 2.861166091212881 |
+
+Overall equal-target mean EF at 1% is 16.107934114857695.

model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb108dcec5608b04eee546a0ed57248ab5f956cb61799a8c5611ab9c8da7c772
+size 381378659

model_config.json

@@ -0,0 +1,13 @@
+{
+  "attention_heads": 64,
+  "dropout": 0.1,
+  "embedding_dim": 128,
+  "encoder_layers": 15,
+  "encoder_width": 512,
+  "ffn_width": 2048,
+  "gradient_checkpointing": true,
+  "max_molecule_atoms": 256,
+  "max_pocket_atoms": 256,
+  "molecule_vocab_size": 31,
+  "pocket_vocab_size": 10
+}

model_manifest.json

@@ -0,0 +1,11 @@
+{
+  "checkpoint": {
+    "path": "model.pt",
+    "sha256": "bb108dcec5608b04eee546a0ed57248ab5f956cb61799a8c5611ab9c8da7c772"
+  },
+  "method": "paper-dual-unimol",
+  "model_config": "model_config.json",
+  "parameter_count": 95305418,
+  "schema": "target-based-drug-screening-model/v2",
+  "training_summary": "training_summary.json"
+}