Chess Challenge submission by Sunxt25

README.md

@@ -7,20 +7,20 @@ tags:
 license: mit
 ---
 
-# chess-Sunxt25
+# chess-sunxt25
 
 Chess model submitted to the LLM Course Chess Challenge.
 
 ## Submission Info
 
 - **Submitted by**: [Sunxt25](https://huggingface.co/Sunxt25)
-- **Parameters**: 878,336
+- **Parameters**: 921,480
 - **Organization**: LLM-course
 
 ## Model Details
 
 - **Architecture**: Chess Transformer (GPT-style)
-- **Vocab size**: 144
-- **Embedding dim**: 128
-- **Layers**: 5
-- **Heads**: 4
+- **Vocab size**: 149
+- **Embedding dim**: 120
+- **Layers**: 6
+- **Heads**: 8

config.json

@@ -2,9 +2,6 @@
   "architectures": [
     "ChessForCausalLM"
   ],
-  "auto_map": {
-    "AutoTokenizer": "chess_tokenizer_custom.ChessTokenizer"
-  },
   "bos_token_id": 1,
   "dropout": 0.1,
   "dtype": "float32",
@@ -12,12 +9,12 @@
   "layer_norm_epsilon": 1e-05,
   "model_type": "chess_transformer",
   "n_ctx": 256,
-  "n_embd": 128,
-  "n_head": 4,
-  "n_inner": 384,
-  "n_layer": 5,
+  "n_embd": 120,
+  "n_head": 8,
+  "n_inner": 360,
+  "n_layer": 6,
   "pad_token_id": 0,
   "tie_weights": true,
   "transformers_version": "4.57.5",
-  "vocab_size": 144
-}
+  "vocab_size": 149
+}

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f2b168eb150974e5eff71ecdc4c7dc06a708917e31fc17001362a4550da9da07
-size 3518768
+oid sha256:010cb99e29355984a756f59f47dd9bda7c4e3945f0f6600d54eb6bff6ec75ef1
+size 3692360

tokenizer_config.json

@@ -1,12 +1,37 @@
 {
   "added_tokens_decoder": {
-    "0": { "content": "[PAD]", "special": true },
-    "1": { "content": "[BOS]", "special": true },
-    "2": { "content": "[EOS]", "special": true },
-    "3": { "content": "[UNK]", "special": true }
-  },
-  "auto_map": {
-    "AutoTokenizer": "chess_tokenizer_custom.ChessTokenizer"
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "[BOS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "[EOS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
   },
   "bos_token": "[BOS]",
   "clean_up_tokenization_spaces": false,
@@ -16,4 +41,4 @@
   "pad_token": "[PAD]",
   "tokenizer_class": "ChessTokenizer",
   "unk_token": "[UNK]"
-}
+}

vocab.json

@@ -142,5 +142,10 @@
   "e8_t": 140,
   "f8_t": 141,
   "g8_t": 142,
-  "h8_t": 143
+  "h8_t": 143,
+  "(x)": 144,
+  "(+)": 145,
+  "(+*)": 146,
+  "(o)": 147,
+  "(O)": 148
 }