Run 4. Outer Step 38. Inner Step 0.
Browse files- config.json +59 -60
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
config.json
CHANGED
|
@@ -7,44 +7,44 @@
|
|
| 7 |
"10": "NON_PARTICIPATING",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
"101": "NON_PARTICIPATING",
|
| 10 |
-
"102": "
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
"104": "NON_PARTICIPATING",
|
| 13 |
"105": "NON_PARTICIPATING",
|
| 14 |
"106": "NON_PARTICIPATING",
|
| 15 |
-
"107": "
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
"109": "SUCCESS",
|
| 18 |
"11": "NON_PARTICIPATING",
|
| 19 |
"110": "NON_PARTICIPATING",
|
| 20 |
"111": "NON_PARTICIPATING",
|
| 21 |
"112": "NON_PARTICIPATING",
|
| 22 |
-
"113": "
|
| 23 |
"114": "NON_PARTICIPATING",
|
| 24 |
"115": "NON_PARTICIPATING",
|
| 25 |
"116": "NON_PARTICIPATING",
|
| 26 |
-
"117": "
|
| 27 |
"118": "NON_PARTICIPATING",
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
"12": "SUCCESS",
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
"122": "NON_PARTICIPATING",
|
| 33 |
-
"123": "
|
| 34 |
"124": "NON_PARTICIPATING",
|
| 35 |
"125": "NON_PARTICIPATING",
|
| 36 |
"126": "NON_PARTICIPATING",
|
| 37 |
-
"127": "
|
| 38 |
"128": "NON_PARTICIPATING",
|
| 39 |
"129": "NON_PARTICIPATING",
|
| 40 |
"13": "SUCCESS",
|
| 41 |
-
"130": "
|
| 42 |
"131": "NON_PARTICIPATING",
|
| 43 |
"132": "NON_PARTICIPATING",
|
| 44 |
"133": "NON_PARTICIPATING",
|
| 45 |
"134": "NON_PARTICIPATING",
|
| 46 |
"135": "NON_PARTICIPATING",
|
| 47 |
-
"136": "
|
| 48 |
"137": "NON_PARTICIPATING",
|
| 49 |
"138": "NON_PARTICIPATING",
|
| 50 |
"139": "NON_PARTICIPATING",
|
|
@@ -57,17 +57,17 @@
|
|
| 57 |
"145": "NON_PARTICIPATING",
|
| 58 |
"146": "NON_PARTICIPATING",
|
| 59 |
"147": "NON_PARTICIPATING",
|
| 60 |
-
"148": "
|
| 61 |
"149": "NON_PARTICIPATING",
|
| 62 |
-
"15": "
|
| 63 |
-
"150": "
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
-
"153": "
|
| 67 |
"154": "NON_PARTICIPATING",
|
| 68 |
"155": "NON_PARTICIPATING",
|
| 69 |
-
"156": "
|
| 70 |
-
"157": "
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
"159": "NON_PARTICIPATING",
|
| 73 |
"16": "SUCCESS",
|
|
@@ -76,20 +76,20 @@
|
|
| 76 |
"162": "NON_PARTICIPATING",
|
| 77 |
"163": "NON_PARTICIPATING",
|
| 78 |
"164": "NON_PARTICIPATING",
|
| 79 |
-
"165": "
|
| 80 |
"166": "NON_PARTICIPATING",
|
| 81 |
"167": "NON_PARTICIPATING",
|
| 82 |
"168": "NON_PARTICIPATING",
|
| 83 |
-
"169": "
|
| 84 |
"17": "SUCCESS",
|
| 85 |
"170": "NON_PARTICIPATING",
|
| 86 |
-
"171": "
|
| 87 |
"172": "NON_PARTICIPATING",
|
| 88 |
"173": "NON_PARTICIPATING",
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
"175": "NON_PARTICIPATING",
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
-
"177": "
|
| 93 |
"178": "NON_PARTICIPATING",
|
| 94 |
"179": "NON_PARTICIPATING",
|
| 95 |
"18": "SUCCESS",
|
|
@@ -100,16 +100,16 @@
|
|
| 100 |
"184": "NON_PARTICIPATING",
|
| 101 |
"185": "NON_PARTICIPATING",
|
| 102 |
"186": "NON_PARTICIPATING",
|
| 103 |
-
"187": "
|
| 104 |
"188": "NON_PARTICIPATING",
|
| 105 |
-
"189": "
|
| 106 |
"19": "NON_PARTICIPATING",
|
| 107 |
-
"190": "
|
| 108 |
"191": "NON_PARTICIPATING",
|
| 109 |
-
"192": "
|
| 110 |
"193": "NON_PARTICIPATING",
|
| 111 |
"194": "NON_PARTICIPATING",
|
| 112 |
-
"195": "
|
| 113 |
"196": "NON_PARTICIPATING",
|
| 114 |
"197": "NON_PARTICIPATING",
|
| 115 |
"198": "NON_PARTICIPATING",
|
|
@@ -118,17 +118,17 @@
|
|
| 118 |
"20": "SUCCESS",
|
| 119 |
"200": "NON_PARTICIPATING",
|
| 120 |
"201": "NON_PARTICIPATING",
|
| 121 |
-
"202": "
|
| 122 |
-
"203": "
|
| 123 |
"204": "NON_PARTICIPATING",
|
| 124 |
-
"205": "
|
| 125 |
"206": "NON_PARTICIPATING",
|
| 126 |
"207": "NON_PARTICIPATING",
|
| 127 |
"208": "NON_PARTICIPATING",
|
| 128 |
-
"209": "
|
| 129 |
"21": "SUCCESS",
|
| 130 |
"210": "NON_PARTICIPATING",
|
| 131 |
-
"211": "
|
| 132 |
"212": "NON_PARTICIPATING",
|
| 133 |
"213": "NON_PARTICIPATING",
|
| 134 |
"214": "NON_PARTICIPATING",
|
|
@@ -141,20 +141,20 @@
|
|
| 141 |
"220": "NON_PARTICIPATING",
|
| 142 |
"221": "NON_PARTICIPATING",
|
| 143 |
"222": "NON_PARTICIPATING",
|
| 144 |
-
"223": "
|
| 145 |
"224": "NON_PARTICIPATING",
|
| 146 |
"225": "NON_PARTICIPATING",
|
| 147 |
"226": "NON_PARTICIPATING",
|
| 148 |
"227": "NON_PARTICIPATING",
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
-
"229": "
|
| 151 |
-
"23": "
|
| 152 |
"230": "NON_PARTICIPATING",
|
| 153 |
"231": "NON_PARTICIPATING",
|
| 154 |
"232": "NON_PARTICIPATING",
|
| 155 |
"233": "NON_PARTICIPATING",
|
| 156 |
"234": "NON_PARTICIPATING",
|
| 157 |
-
"235": "
|
| 158 |
"236": "NON_PARTICIPATING",
|
| 159 |
"237": "NON_PARTICIPATING",
|
| 160 |
"238": "NON_PARTICIPATING",
|
|
@@ -163,9 +163,9 @@
|
|
| 163 |
"240": "NON_PARTICIPATING",
|
| 164 |
"241": "NON_PARTICIPATING",
|
| 165 |
"242": "SUCCESS",
|
| 166 |
-
"243": "
|
| 167 |
-
"244": "
|
| 168 |
-
"245": "
|
| 169 |
"246": "NON_PARTICIPATING",
|
| 170 |
"247": "NON_PARTICIPATING",
|
| 171 |
"248": "NON_PARTICIPATING",
|
|
@@ -173,45 +173,45 @@
|
|
| 173 |
"25": "SUCCESS",
|
| 174 |
"250": "NON_PARTICIPATING",
|
| 175 |
"251": "NON_PARTICIPATING",
|
| 176 |
-
"252": "
|
| 177 |
"253": "NON_PARTICIPATING",
|
| 178 |
"254": "NON_PARTICIPATING",
|
| 179 |
-
"255": "
|
| 180 |
"26": "SUCCESS",
|
| 181 |
"27": "SUCCESS",
|
| 182 |
"28": "SUCCESS",
|
| 183 |
"29": "SUCCESS",
|
| 184 |
"3": "SUCCESS",
|
| 185 |
-
"30": "
|
| 186 |
-
"31": "
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
-
"34": "
|
| 190 |
"35": "SUCCESS",
|
| 191 |
-
"36": "
|
| 192 |
"37": "SUCCESS",
|
| 193 |
"38": "SUCCESS",
|
| 194 |
"39": "SUCCESS",
|
| 195 |
"4": "SUCCESS",
|
| 196 |
"40": "SUCCESS",
|
| 197 |
"41": "SUCCESS",
|
| 198 |
-
"42": "
|
| 199 |
"43": "SUCCESS",
|
| 200 |
"44": "SUCCESS",
|
| 201 |
"45": "SUCCESS",
|
| 202 |
"46": "NON_PARTICIPATING",
|
| 203 |
-
"47": "
|
| 204 |
"48": "NON_PARTICIPATING",
|
| 205 |
"49": "SUCCESS",
|
| 206 |
-
"5": "
|
| 207 |
"50": "SUCCESS",
|
| 208 |
"51": "SUCCESS",
|
| 209 |
"52": "SUCCESS",
|
| 210 |
-
"53": "
|
| 211 |
"54": "SUCCESS",
|
| 212 |
"55": "SUCCESS",
|
| 213 |
"56": "SUCCESS",
|
| 214 |
-
"57": "
|
| 215 |
"58": "NON_PARTICIPATING",
|
| 216 |
"59": "SUCCESS",
|
| 217 |
"6": "NON_PARTICIPATING",
|
|
@@ -223,18 +223,18 @@
|
|
| 223 |
"65": "SUCCESS",
|
| 224 |
"66": "SUCCESS",
|
| 225 |
"67": "SUCCESS",
|
| 226 |
-
"68": "
|
| 227 |
-
"69": "
|
| 228 |
"7": "SUCCESS",
|
| 229 |
"70": "NON_PARTICIPATING",
|
| 230 |
"71": "NON_PARTICIPATING",
|
| 231 |
"72": "SUCCESS",
|
| 232 |
"73": "NON_PARTICIPATING",
|
| 233 |
-
"74": "
|
| 234 |
"75": "SUCCESS",
|
| 235 |
"76": "SUCCESS",
|
| 236 |
"77": "SUCCESS",
|
| 237 |
-
"78": "
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "SUCCESS",
|
| 240 |
"80": "SUCCESS",
|
|
@@ -242,8 +242,8 @@
|
|
| 242 |
"82": "SUCCESS",
|
| 243 |
"83": "NON_PARTICIPATING",
|
| 244 |
"84": "SUCCESS",
|
| 245 |
-
"85": "
|
| 246 |
-
"86": "
|
| 247 |
"87": "SUCCESS",
|
| 248 |
"88": "SUCCESS",
|
| 249 |
"89": "SUCCESS",
|
|
@@ -268,21 +268,20 @@
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
-
|
| 272 |
-
|
| 273 |
-
|
| 274 |
-
|
| 275 |
-
|
| 276 |
-
5552069
|
| 277 |
],
|
| 278 |
"block_size": 1024,
|
| 279 |
"bos_token_id": 50256,
|
| 280 |
"embd_pdrop": 0.1,
|
| 281 |
"eos_token_id": 50256,
|
| 282 |
"initializer_range": 0.02,
|
| 283 |
-
"inner_step":
|
| 284 |
"inner_steps": 0,
|
| 285 |
-
"last_allreduce_block":
|
| 286 |
"layer_norm_epsilon": 1e-05,
|
| 287 |
"model_type": "gpt_optimized",
|
| 288 |
"n_embd": 1280,
|
|
|
|
| 7 |
"10": "NON_PARTICIPATING",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
"101": "NON_PARTICIPATING",
|
| 10 |
+
"102": "SUCCESS",
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
"104": "NON_PARTICIPATING",
|
| 13 |
"105": "NON_PARTICIPATING",
|
| 14 |
"106": "NON_PARTICIPATING",
|
| 15 |
+
"107": "NON_PARTICIPATING",
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
"109": "SUCCESS",
|
| 18 |
"11": "NON_PARTICIPATING",
|
| 19 |
"110": "NON_PARTICIPATING",
|
| 20 |
"111": "NON_PARTICIPATING",
|
| 21 |
"112": "NON_PARTICIPATING",
|
| 22 |
+
"113": "NON_PARTICIPATING",
|
| 23 |
"114": "NON_PARTICIPATING",
|
| 24 |
"115": "NON_PARTICIPATING",
|
| 25 |
"116": "NON_PARTICIPATING",
|
| 26 |
+
"117": "SUCCESS",
|
| 27 |
"118": "NON_PARTICIPATING",
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
"12": "SUCCESS",
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
"122": "NON_PARTICIPATING",
|
| 33 |
+
"123": "SUCCESS",
|
| 34 |
"124": "NON_PARTICIPATING",
|
| 35 |
"125": "NON_PARTICIPATING",
|
| 36 |
"126": "NON_PARTICIPATING",
|
| 37 |
+
"127": "NON_PARTICIPATING",
|
| 38 |
"128": "NON_PARTICIPATING",
|
| 39 |
"129": "NON_PARTICIPATING",
|
| 40 |
"13": "SUCCESS",
|
| 41 |
+
"130": "NON_PARTICIPATING",
|
| 42 |
"131": "NON_PARTICIPATING",
|
| 43 |
"132": "NON_PARTICIPATING",
|
| 44 |
"133": "NON_PARTICIPATING",
|
| 45 |
"134": "NON_PARTICIPATING",
|
| 46 |
"135": "NON_PARTICIPATING",
|
| 47 |
+
"136": "SUCCESS",
|
| 48 |
"137": "NON_PARTICIPATING",
|
| 49 |
"138": "NON_PARTICIPATING",
|
| 50 |
"139": "NON_PARTICIPATING",
|
|
|
|
| 57 |
"145": "NON_PARTICIPATING",
|
| 58 |
"146": "NON_PARTICIPATING",
|
| 59 |
"147": "NON_PARTICIPATING",
|
| 60 |
+
"148": "NON_PARTICIPATING",
|
| 61 |
"149": "NON_PARTICIPATING",
|
| 62 |
+
"15": "NON_PARTICIPATING",
|
| 63 |
+
"150": "NON_PARTICIPATING",
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
+
"153": "SUCCESS",
|
| 67 |
"154": "NON_PARTICIPATING",
|
| 68 |
"155": "NON_PARTICIPATING",
|
| 69 |
+
"156": "SUCCESS",
|
| 70 |
+
"157": "SUCCESS",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
"159": "NON_PARTICIPATING",
|
| 73 |
"16": "SUCCESS",
|
|
|
|
| 76 |
"162": "NON_PARTICIPATING",
|
| 77 |
"163": "NON_PARTICIPATING",
|
| 78 |
"164": "NON_PARTICIPATING",
|
| 79 |
+
"165": "NON_PARTICIPATING",
|
| 80 |
"166": "NON_PARTICIPATING",
|
| 81 |
"167": "NON_PARTICIPATING",
|
| 82 |
"168": "NON_PARTICIPATING",
|
| 83 |
+
"169": "NON_PARTICIPATING",
|
| 84 |
"17": "SUCCESS",
|
| 85 |
"170": "NON_PARTICIPATING",
|
| 86 |
+
"171": "NON_PARTICIPATING",
|
| 87 |
"172": "NON_PARTICIPATING",
|
| 88 |
"173": "NON_PARTICIPATING",
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
"175": "NON_PARTICIPATING",
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
+
"177": "NON_PARTICIPATING",
|
| 93 |
"178": "NON_PARTICIPATING",
|
| 94 |
"179": "NON_PARTICIPATING",
|
| 95 |
"18": "SUCCESS",
|
|
|
|
| 100 |
"184": "NON_PARTICIPATING",
|
| 101 |
"185": "NON_PARTICIPATING",
|
| 102 |
"186": "NON_PARTICIPATING",
|
| 103 |
+
"187": "NON_PARTICIPATING",
|
| 104 |
"188": "NON_PARTICIPATING",
|
| 105 |
+
"189": "NON_PARTICIPATING",
|
| 106 |
"19": "NON_PARTICIPATING",
|
| 107 |
+
"190": "NON_PARTICIPATING",
|
| 108 |
"191": "NON_PARTICIPATING",
|
| 109 |
+
"192": "NON_PARTICIPATING",
|
| 110 |
"193": "NON_PARTICIPATING",
|
| 111 |
"194": "NON_PARTICIPATING",
|
| 112 |
+
"195": "NON_PARTICIPATING",
|
| 113 |
"196": "NON_PARTICIPATING",
|
| 114 |
"197": "NON_PARTICIPATING",
|
| 115 |
"198": "NON_PARTICIPATING",
|
|
|
|
| 118 |
"20": "SUCCESS",
|
| 119 |
"200": "NON_PARTICIPATING",
|
| 120 |
"201": "NON_PARTICIPATING",
|
| 121 |
+
"202": "SUCCESS",
|
| 122 |
+
"203": "NON_PARTICIPATING",
|
| 123 |
"204": "NON_PARTICIPATING",
|
| 124 |
+
"205": "NON_PARTICIPATING",
|
| 125 |
"206": "NON_PARTICIPATING",
|
| 126 |
"207": "NON_PARTICIPATING",
|
| 127 |
"208": "NON_PARTICIPATING",
|
| 128 |
+
"209": "SUCCESS",
|
| 129 |
"21": "SUCCESS",
|
| 130 |
"210": "NON_PARTICIPATING",
|
| 131 |
+
"211": "NON_PARTICIPATING",
|
| 132 |
"212": "NON_PARTICIPATING",
|
| 133 |
"213": "NON_PARTICIPATING",
|
| 134 |
"214": "NON_PARTICIPATING",
|
|
|
|
| 141 |
"220": "NON_PARTICIPATING",
|
| 142 |
"221": "NON_PARTICIPATING",
|
| 143 |
"222": "NON_PARTICIPATING",
|
| 144 |
+
"223": "NON_PARTICIPATING",
|
| 145 |
"224": "NON_PARTICIPATING",
|
| 146 |
"225": "NON_PARTICIPATING",
|
| 147 |
"226": "NON_PARTICIPATING",
|
| 148 |
"227": "NON_PARTICIPATING",
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
+
"229": "NON_PARTICIPATING",
|
| 151 |
+
"23": "SUCCESS",
|
| 152 |
"230": "NON_PARTICIPATING",
|
| 153 |
"231": "NON_PARTICIPATING",
|
| 154 |
"232": "NON_PARTICIPATING",
|
| 155 |
"233": "NON_PARTICIPATING",
|
| 156 |
"234": "NON_PARTICIPATING",
|
| 157 |
+
"235": "SUCCESS",
|
| 158 |
"236": "NON_PARTICIPATING",
|
| 159 |
"237": "NON_PARTICIPATING",
|
| 160 |
"238": "NON_PARTICIPATING",
|
|
|
|
| 163 |
"240": "NON_PARTICIPATING",
|
| 164 |
"241": "NON_PARTICIPATING",
|
| 165 |
"242": "SUCCESS",
|
| 166 |
+
"243": "SUCCESS",
|
| 167 |
+
"244": "SUCCESS",
|
| 168 |
+
"245": "SUCCESS",
|
| 169 |
"246": "NON_PARTICIPATING",
|
| 170 |
"247": "NON_PARTICIPATING",
|
| 171 |
"248": "NON_PARTICIPATING",
|
|
|
|
| 173 |
"25": "SUCCESS",
|
| 174 |
"250": "NON_PARTICIPATING",
|
| 175 |
"251": "NON_PARTICIPATING",
|
| 176 |
+
"252": "NON_PARTICIPATING",
|
| 177 |
"253": "NON_PARTICIPATING",
|
| 178 |
"254": "NON_PARTICIPATING",
|
| 179 |
+
"255": "NON_PARTICIPATING",
|
| 180 |
"26": "SUCCESS",
|
| 181 |
"27": "SUCCESS",
|
| 182 |
"28": "SUCCESS",
|
| 183 |
"29": "SUCCESS",
|
| 184 |
"3": "SUCCESS",
|
| 185 |
+
"30": "SUCCESS",
|
| 186 |
+
"31": "NON_PARTICIPATING",
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
+
"34": "SUCCESS",
|
| 190 |
"35": "SUCCESS",
|
| 191 |
+
"36": "SUCCESS",
|
| 192 |
"37": "SUCCESS",
|
| 193 |
"38": "SUCCESS",
|
| 194 |
"39": "SUCCESS",
|
| 195 |
"4": "SUCCESS",
|
| 196 |
"40": "SUCCESS",
|
| 197 |
"41": "SUCCESS",
|
| 198 |
+
"42": "SUCCESS",
|
| 199 |
"43": "SUCCESS",
|
| 200 |
"44": "SUCCESS",
|
| 201 |
"45": "SUCCESS",
|
| 202 |
"46": "NON_PARTICIPATING",
|
| 203 |
+
"47": "SUCCESS",
|
| 204 |
"48": "NON_PARTICIPATING",
|
| 205 |
"49": "SUCCESS",
|
| 206 |
+
"5": "SUCCESS",
|
| 207 |
"50": "SUCCESS",
|
| 208 |
"51": "SUCCESS",
|
| 209 |
"52": "SUCCESS",
|
| 210 |
+
"53": "SUCCESS",
|
| 211 |
"54": "SUCCESS",
|
| 212 |
"55": "SUCCESS",
|
| 213 |
"56": "SUCCESS",
|
| 214 |
+
"57": "SUCCESS",
|
| 215 |
"58": "NON_PARTICIPATING",
|
| 216 |
"59": "SUCCESS",
|
| 217 |
"6": "NON_PARTICIPATING",
|
|
|
|
| 223 |
"65": "SUCCESS",
|
| 224 |
"66": "SUCCESS",
|
| 225 |
"67": "SUCCESS",
|
| 226 |
+
"68": "SUCCESS",
|
| 227 |
+
"69": "SUCCESS",
|
| 228 |
"7": "SUCCESS",
|
| 229 |
"70": "NON_PARTICIPATING",
|
| 230 |
"71": "NON_PARTICIPATING",
|
| 231 |
"72": "SUCCESS",
|
| 232 |
"73": "NON_PARTICIPATING",
|
| 233 |
+
"74": "NON_PARTICIPATING",
|
| 234 |
"75": "SUCCESS",
|
| 235 |
"76": "SUCCESS",
|
| 236 |
"77": "SUCCESS",
|
| 237 |
+
"78": "SUCCESS",
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "SUCCESS",
|
| 240 |
"80": "SUCCESS",
|
|
|
|
| 242 |
"82": "SUCCESS",
|
| 243 |
"83": "NON_PARTICIPATING",
|
| 244 |
"84": "SUCCESS",
|
| 245 |
+
"85": "NON_PARTICIPATING",
|
| 246 |
+
"86": "SUCCESS",
|
| 247 |
"87": "SUCCESS",
|
| 248 |
"88": "SUCCESS",
|
| 249 |
"89": "SUCCESS",
|
|
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
+
5552948,
|
| 272 |
+
5552952,
|
| 273 |
+
5552956,
|
| 274 |
+
5552960,
|
| 275 |
+
5552964
|
|
|
|
| 276 |
],
|
| 277 |
"block_size": 1024,
|
| 278 |
"bos_token_id": 50256,
|
| 279 |
"embd_pdrop": 0.1,
|
| 280 |
"eos_token_id": 50256,
|
| 281 |
"initializer_range": 0.02,
|
| 282 |
+
"inner_step": 0,
|
| 283 |
"inner_steps": 0,
|
| 284 |
+
"last_allreduce_block": 5552497,
|
| 285 |
"layer_norm_epsilon": 1e-05,
|
| 286 |
"model_type": "gpt_optimized",
|
| 287 |
"n_embd": 1280,
|
inner_optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 8081782026
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a3d40bf492e22a7dd9d9f13e8e3fed6931d06b801425b5a8f1b336fe232a445b
|
| 3 |
size 8081782026
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4040701744
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:44f67b33b2043d37d1496c5a7dea981857bb909fab7a305a89346611b9d43458
|
| 3 |
size 4040701744
|