Hugging Face's logo

datafreak
/
laya-chess

Laya
Safetensors
chess
decision-model
Model card Files
xet
 Community
datafreak commited on
Commit
d33ff17
·
verified ·
1 Parent(s): 7a8de87

step2220: step 2220, 142080 examples, eval {'val_loss': 1.5918, 'val_wp_mae': 0.1285, 'top_move_acc': 0.16333333333333333, 'step': 2000}

Browse files
Files changed (3) hide show
  1. README.md +2 -2
  2. chess_meta.json +97 -3
  3. model.safetensors +1 -1
README.md CHANGED
@@ -9,7 +9,7 @@ tags: [chess, laya, decision-model]
9
  for each legal move, a Laya `score` question predicts the side to move's win chance over 10 levels.
10
  The engine plays the move with the highest expected win chance (no search).
11
 
12
- - checkpoint: `step1147` (step 1147, 73,408 training examples)
13
- - latest eval: `{'val_loss': 2.4097, 'val_wp_mae': 0.2865, 'top_move_acc': 0.06, 'step': 0}`
14
  - encoding / levels / prompt template: see `chess_meta.json`
15
  - code: github.com/devroopsaha744/LayaChess
 
9
  for each legal move, a Laya `score` question predicts the side to move's win chance over 10 levels.
10
  The engine plays the move with the highest expected win chance (no search).
11
 
12
+ - checkpoint: `step2220` (step 2220, 142,080 training examples)
13
+ - latest eval: `{'val_loss': 1.5918, 'val_wp_mae': 0.1285, 'top_move_acc': 0.16333333333333333, 'step': 2000}`
14
  - encoding / levels / prompt template: see `chess_meta.json`
15
  - code: github.com/devroopsaha744/LayaChess
chess_meta.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "tag": "step1147",
3
  "final": false,
4
- "step": 1147,
5
- "consumed": 73408,
6
  "n_levels": 10,
7
  "crit": [
8
  "0-10%",
@@ -138,6 +138,100 @@
138
  {
139
  "step": 1100,
140
  "loss": 1.6392
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
141
  }
142
  ]
143
  }
 
1
  {
2
+ "tag": "step2220",
3
  "final": false,
4
+ "step": 2220,
5
+ "consumed": 142080,
6
  "n_levels": 10,
7
  "crit": [
8
  "0-10%",
 
138
  {
139
  "step": 1100,
140
  "loss": 1.6392
141
+ },
142
+ {
143
+ "step": 1150,
144
+ "loss": 1.6377
145
+ },
146
+ {
147
+ "step": 1200,
148
+ "loss": 1.6143
149
+ },
150
+ {
151
+ "step": 1250,
152
+ "loss": 1.6564
153
+ },
154
+ {
155
+ "step": 1300,
156
+ "loss": 1.6321
157
+ },
158
+ {
159
+ "step": 1350,
160
+ "loss": 1.6073
161
+ },
162
+ {
163
+ "step": 1400,
164
+ "loss": 1.6365
165
+ },
166
+ {
167
+ "step": 1450,
168
+ "loss": 1.6115
169
+ },
170
+ {
171
+ "step": 1500,
172
+ "loss": 1.6233
173
+ },
174
+ {
175
+ "step": 1550,
176
+ "loss": 1.6357
177
+ },
178
+ {
179
+ "step": 1600,
180
+ "loss": 1.6131
181
+ },
182
+ {
183
+ "step": 1650,
184
+ "loss": 1.6356
185
+ },
186
+ {
187
+ "step": 1700,
188
+ "loss": 1.5871
189
+ },
190
+ {
191
+ "step": 1750,
192
+ "loss": 1.631
193
+ },
194
+ {
195
+ "step": 1800,
196
+ "loss": 1.6044
197
+ },
198
+ {
199
+ "step": 1850,
200
+ "loss": 1.606
201
+ },
202
+ {
203
+ "step": 1900,
204
+ "loss": 1.6274
205
+ },
206
+ {
207
+ "step": 1950,
208
+ "loss": 1.6102
209
+ },
210
+ {
211
+ "step": 2000,
212
+ "loss": 1.6095
213
+ },
214
+ {
215
+ "val_loss": 1.5918,
216
+ "val_wp_mae": 0.1285,
217
+ "top_move_acc": 0.16333333333333333,
218
+ "step": 2000
219
+ },
220
+ {
221
+ "step": 2050,
222
+ "loss": 1.6272
223
+ },
224
+ {
225
+ "step": 2100,
226
+ "loss": 1.6018
227
+ },
228
+ {
229
+ "step": 2150,
230
+ "loss": 1.6003
231
+ },
232
+ {
233
+ "step": 2200,
234
+ "loss": 1.5952
235
  }
236
  ]
237
  }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9a51fa2150286fa28962e5f07d3786bf6d6e4ad5af5ff83be66d3be2d6e218cf
3
  size 842609220
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:137650d99cc30af206e7d2055a45fa419b0d0cd629ff993319897a3e2af58466
3
  size 842609220