Upload README.md

README.md

@@ -31,7 +31,6 @@ This repository is mainly intended for:
 - initialization for downstream molecular property prediction
 - autoregressive molecular string modeling research
 
-**Repository naming note.** In the paper, the improved fragment-link representation is called **FragLink**. In the current Hugging Face repository layout, the corresponding folders appear as **FragSeqV2**. Both names are kept below so that readers can map the paper terminology to the released files.
 
 ## Model Sources
 
@@ -63,7 +62,7 @@ The current repository layout contains checkpoints grouped by representation and
 - 278M
 - 650M
 
-### FragSeqV2 *(FragLink in the paper)*
+### FragLink
 - 1M
 - 4M
 - 16M
@@ -107,7 +106,7 @@ The paper studies five string representations:
 - **DeepSMILES**
 - **SAFE**
 - **FragSeq**
-- **FragLink** *(stored as `FragSeqV2` in the current repo layout)*
+- **FragLink**
 
 ### Scaling Grid
 
@@ -158,21 +157,21 @@ Downstream transfer is evaluated on nine MoleculeNet benchmarks:
 
 | Task | Metric | Best Released Representation | Score |
 |---|---:|---|---:|
-| BACE | ROC-AUC ↑ | FragLink / FragSeqV2 | 89.7 |
+| BACE | ROC-AUC ↑ | FragLink | 89.7 |
 | HIV | ROC-AUC ↑ | SAFE* | 83.3 |
 | BBBP | ROC-AUC ↑ | DeepSMILES | 97.8 |
 | SIDER | ROC-AUC ↑ | FragSeq | 68.8 |
 | Tox21 | ROC-AUC ↑ | FragSeq | 83.7 |
 | ClinTox | ROC-AUC ↑ | SMILES / DeepSMILES | 99.8 |
 | ESOL | RMSE ↓ | DeepSMILES | 0.362 |
-| FreeSolv | RMSE ↓ | FragLink / FragSeqV2 | 1.095 |
-| Lipophilicity | RMSE ↓ | FragLink / FragSeqV2 | 0.593 |
+| FreeSolv | RMSE ↓ | FragLink | 1.095 |
+| Lipophilicity | RMSE ↓ | FragLink | 0.593 |
 
 \* The paper notes that SAFE reaches the highest HIV score, but also points out that SAFE only covers about 83% of the original HIV test set in that comparison. For full context, please check the paper.
 
 ### Task-Level Takeaways
 
-- **FragLink / FragSeqV2** is especially strong on **BACE** and the **biophysics regression tasks**.
+- **FragLink** is especially strong on **BACE** and the **biophysics regression tasks**.
 - **SMILES** and **DeepSMILES** are strong on **HIV**, **BBBP**, and **ClinTox**.
 - **FragSeq** is particularly competitive on **SIDER** and **Tox21**.
 - There is **no single best representation for every downstream task**.
@@ -220,7 +219,7 @@ Examples:
 - `SMILES 152M`
 - `DeepSMILES 85M`
 - `FragSeq 43M`
-- `FragSeqV2 152M`
+- `FragLink 152M`
 - `SAFE 278M`
 
 Then load the selected checkpoint with the official codebase and the matching configuration.