Chess Challenge submission by julesdamidaux

README.md

@@ -14,13 +14,13 @@ Chess model submitted to the LLM Course Chess Challenge.
 ## Submission Info
 
 - **Submitted by**: [julesdamidaux](https://huggingface.co/julesdamidaux)
-- **Parameters**: 1,856,256
+- **Parameters**: 909,824
 - **Organization**: LLM-course
 
 ## Model Details
 
 - **Architecture**: Chess Transformer (GPT-style)
 - **Vocab size**: 1682
-- **Embedding dim**: 192
+- **Embedding dim**: 128
 - **Layers**: 4
 - **Heads**: 4

config.json

@@ -13,9 +13,9 @@
   "layer_norm_epsilon": 1e-05,
   "model_type": "chess_transformer",
   "n_ctx": 256,
-  "n_embd": 192,
+  "n_embd": 128,
   "n_head": 4,
-  "n_inner": 576,
+  "n_inner": 384,
   "n_layer": 4,
   "pad_token_id": 0,
   "tie_weights": true,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8265a8f5c4f8afb12ff59d3eaa8e0e41132059351f77cef803435d229c804204
-size 7429448
+oid sha256:e1ca8c5f2dc0f12113ccf254c1c414a39391dd4e00b5dfc049295105cf9fcdac
+size 3643696