V7 HULK — val_loss=0.03017

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+training_curves.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,28 @@
+---
+license: mit
+tags: [chess, pytorch, resnet, se-resnet, position-evaluation, policy-head, arabic]
+language: [ar, en]
+---
+# ♟️ Chess Position Evaluator — V7 HULK
+
+> **Ahmed Darwish** | [@engdarwish](https://huggingface.co/engdarwish)
+
+## Architecture
+SE-ResNet-20 with Dual Head (Value + Policy)
+
+| Metric | Value |
+|--------|-------|
+| Parameters | 47,450,753 |
+| Best Val Loss | 0.030174 |
+| Architecture | SE-ResNet-20 |
+| Input | 18×8×8 board tensor |
+| Value Output | [-1, 1] (tanh) |
+| Policy Output | 20,480 move logits |
+
+## Training Data
+- **Kaggle Chess Evaluations** — 12.9M positions (Stockfish evals)
+- **Lichess 2023-10** — 2M moves from ELO 2000+ games
+
+## Contact
+📧 eahmeddarwish@gmail.com
+🌐 [GitHub](https://github.com/engrdarwish)

checkpoint.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22832a8ba0c708b4949e2f5e234ec3a1553df172bb1bb2df9887131adb56175f
+size 569751541

config.json

@@ -0,0 +1,13 @@
+{
+  "version": "V7_HULK",
+  "architecture": "SE-ResNet-20 Dual Head",
+  "in_channels": 18,
+  "num_filters": 256,
+  "num_res_blocks": 20,
+  "policy_size": 20480,
+  "best_val_loss": 0.030174075044405294,
+  "author": "Ahmed Darwish",
+  "email": "eahmeddarwish@gmail.com",
+  "huggingface": "https://huggingface.co/engdarwish",
+  "github": "https://github.com/engrdarwish"
+}

model_weights.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e05367310c42d705cb6b385878264bccbcb79508894bbbc2f8b7d06cd524bc24
+size 189994667