Run 4. Outer Step 30. Inner Step 0.
Browse files- config.json +61 -64
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
config.json
CHANGED
|
@@ -6,45 +6,45 @@
|
|
| 6 |
"1": "SUCCESS",
|
| 7 |
"10": "NON_PARTICIPATING",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
-
"101": "
|
| 10 |
"102": "NON_PARTICIPATING",
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
"104": "NON_PARTICIPATING",
|
| 13 |
"105": "NON_PARTICIPATING",
|
| 14 |
"106": "SUCCESS",
|
| 15 |
-
"107": "
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
"109": "NON_PARTICIPATING",
|
| 18 |
"11": "NON_PARTICIPATING",
|
| 19 |
-
"110": "
|
| 20 |
"111": "NON_PARTICIPATING",
|
| 21 |
"112": "NON_PARTICIPATING",
|
| 22 |
-
"113": "
|
| 23 |
"114": "NON_PARTICIPATING",
|
| 24 |
"115": "NON_PARTICIPATING",
|
| 25 |
"116": "NON_PARTICIPATING",
|
| 26 |
"117": "NON_PARTICIPATING",
|
| 27 |
"118": "NON_PARTICIPATING",
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
-
"12": "
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 31 |
"121": "SUCCESS",
|
| 32 |
-
"122": "
|
| 33 |
-
"123": "
|
| 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": "
|
| 41 |
"130": "NON_PARTICIPATING",
|
| 42 |
"131": "NON_PARTICIPATING",
|
| 43 |
-
"132": "
|
| 44 |
"133": "NON_PARTICIPATING",
|
| 45 |
"134": "NON_PARTICIPATING",
|
| 46 |
-
"135": "
|
| 47 |
-
"136": "
|
| 48 |
"137": "NON_PARTICIPATING",
|
| 49 |
"138": "NON_PARTICIPATING",
|
| 50 |
"139": "NON_PARTICIPATING",
|
|
@@ -52,26 +52,26 @@
|
|
| 52 |
"140": "NON_PARTICIPATING",
|
| 53 |
"141": "NON_PARTICIPATING",
|
| 54 |
"142": "NON_PARTICIPATING",
|
| 55 |
-
"143": "
|
| 56 |
"144": "NON_PARTICIPATING",
|
| 57 |
-
"145": "
|
| 58 |
"146": "NON_PARTICIPATING",
|
| 59 |
"147": "NON_PARTICIPATING",
|
| 60 |
-
"148": "
|
| 61 |
-
"149": "
|
| 62 |
"15": "SUCCESS",
|
| 63 |
-
"150": "
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
"153": "NON_PARTICIPATING",
|
| 67 |
"154": "SUCCESS",
|
| 68 |
"155": "NON_PARTICIPATING",
|
| 69 |
-
"156": "
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
-
"158": "
|
| 72 |
-
"159": "
|
| 73 |
"16": "SUCCESS",
|
| 74 |
-
"160": "
|
| 75 |
"161": "NON_PARTICIPATING",
|
| 76 |
"162": "NON_PARTICIPATING",
|
| 77 |
"163": "NON_PARTICIPATING",
|
|
@@ -79,15 +79,15 @@
|
|
| 79 |
"165": "NON_PARTICIPATING",
|
| 80 |
"166": "NON_PARTICIPATING",
|
| 81 |
"167": "NON_PARTICIPATING",
|
| 82 |
-
"168": "
|
| 83 |
-
"169": "
|
| 84 |
"17": "SUCCESS",
|
| 85 |
"170": "NON_PARTICIPATING",
|
| 86 |
-
"171": "
|
| 87 |
-
"172": "
|
| 88 |
"173": "NON_PARTICIPATING",
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
-
"175": "
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
"177": "NON_PARTICIPATING",
|
| 93 |
"178": "NON_PARTICIPATING",
|
|
@@ -98,7 +98,7 @@
|
|
| 98 |
"182": "NON_PARTICIPATING",
|
| 99 |
"183": "NON_PARTICIPATING",
|
| 100 |
"184": "SUCCESS",
|
| 101 |
-
"185": "
|
| 102 |
"186": "NON_PARTICIPATING",
|
| 103 |
"187": "NON_PARTICIPATING",
|
| 104 |
"188": "NON_PARTICIPATING",
|
|
@@ -107,10 +107,10 @@
|
|
| 107 |
"190": "NON_PARTICIPATING",
|
| 108 |
"191": "NON_PARTICIPATING",
|
| 109 |
"192": "NON_PARTICIPATING",
|
| 110 |
-
"193": "
|
| 111 |
"194": "NON_PARTICIPATING",
|
| 112 |
"195": "NON_PARTICIPATING",
|
| 113 |
-
"196": "
|
| 114 |
"197": "NON_PARTICIPATING",
|
| 115 |
"198": "NON_PARTICIPATING",
|
| 116 |
"199": "NON_PARTICIPATING",
|
|
@@ -118,36 +118,36 @@
|
|
| 118 |
"20": "SUCCESS",
|
| 119 |
"200": "NON_PARTICIPATING",
|
| 120 |
"201": "NON_PARTICIPATING",
|
| 121 |
-
"202": "
|
| 122 |
-
"203": "
|
| 123 |
"204": "NON_PARTICIPATING",
|
| 124 |
"205": "NON_PARTICIPATING",
|
| 125 |
"206": "NON_PARTICIPATING",
|
| 126 |
"207": "NON_PARTICIPATING",
|
| 127 |
-
"208": "
|
| 128 |
"209": "NON_PARTICIPATING",
|
| 129 |
"21": "SUCCESS",
|
| 130 |
-
"210": "
|
| 131 |
"211": "SUCCESS",
|
| 132 |
"212": "NON_PARTICIPATING",
|
| 133 |
-
"213": "
|
| 134 |
-
"214": "
|
| 135 |
"215": "NON_PARTICIPATING",
|
| 136 |
"216": "NON_PARTICIPATING",
|
| 137 |
"217": "NON_PARTICIPATING",
|
| 138 |
"218": "NON_PARTICIPATING",
|
| 139 |
-
"219": "
|
| 140 |
"22": "SUCCESS",
|
| 141 |
"220": "NON_PARTICIPATING",
|
| 142 |
"221": "NON_PARTICIPATING",
|
| 143 |
"222": "NON_PARTICIPATING",
|
| 144 |
-
"223": "
|
| 145 |
"224": "NON_PARTICIPATING",
|
| 146 |
-
"225": "
|
| 147 |
"226": "NON_PARTICIPATING",
|
| 148 |
"227": "NON_PARTICIPATING",
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
-
"229": "
|
| 151 |
"23": "NON_PARTICIPATING",
|
| 152 |
"230": "NON_PARTICIPATING",
|
| 153 |
"231": "NON_PARTICIPATING",
|
|
@@ -157,7 +157,7 @@
|
|
| 157 |
"235": "NON_PARTICIPATING",
|
| 158 |
"236": "NON_PARTICIPATING",
|
| 159 |
"237": "NON_PARTICIPATING",
|
| 160 |
-
"238": "
|
| 161 |
"239": "NON_PARTICIPATING",
|
| 162 |
"24": "SUCCESS",
|
| 163 |
"240": "NON_PARTICIPATING",
|
|
@@ -173,11 +173,11 @@
|
|
| 173 |
"25": "SUCCESS",
|
| 174 |
"250": "SUCCESS",
|
| 175 |
"251": "NON_PARTICIPATING",
|
| 176 |
-
"252": "
|
| 177 |
"253": "NON_PARTICIPATING",
|
| 178 |
"254": "NON_PARTICIPATING",
|
| 179 |
"255": "SUCCESS",
|
| 180 |
-
"26": "
|
| 181 |
"27": "SUCCESS",
|
| 182 |
"28": "SUCCESS",
|
| 183 |
"29": "SUCCESS",
|
|
@@ -187,14 +187,14 @@
|
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "NON_PARTICIPATING",
|
| 190 |
-
"35": "
|
| 191 |
"36": "NON_PARTICIPATING",
|
| 192 |
"37": "SUCCESS",
|
| 193 |
-
"38": "
|
| 194 |
-
"39": "
|
| 195 |
"4": "NON_PARTICIPATING",
|
| 196 |
"40": "SUCCESS",
|
| 197 |
-
"41": "
|
| 198 |
"42": "NON_PARTICIPATING",
|
| 199 |
"43": "SUCCESS",
|
| 200 |
"44": "SUCCESS",
|
|
@@ -208,14 +208,14 @@
|
|
| 208 |
"51": "SUCCESS",
|
| 209 |
"52": "SUCCESS",
|
| 210 |
"53": "NON_PARTICIPATING",
|
| 211 |
-
"54": "
|
| 212 |
"55": "SUCCESS",
|
| 213 |
"56": "SUCCESS",
|
| 214 |
"57": "SUCCESS",
|
| 215 |
"58": "NON_PARTICIPATING",
|
| 216 |
"59": "NON_PARTICIPATING",
|
| 217 |
"6": "NON_PARTICIPATING",
|
| 218 |
-
"60": "
|
| 219 |
"61": "NON_PARTICIPATING",
|
| 220 |
"62": "SUCCESS",
|
| 221 |
"63": "SUCCESS",
|
|
@@ -223,7 +223,7 @@
|
|
| 223 |
"65": "SUCCESS",
|
| 224 |
"66": "SUCCESS",
|
| 225 |
"67": "SUCCESS",
|
| 226 |
-
"68": "
|
| 227 |
"69": "NON_PARTICIPATING",
|
| 228 |
"7": "SUCCESS",
|
| 229 |
"70": "NON_PARTICIPATING",
|
|
@@ -232,17 +232,17 @@
|
|
| 232 |
"73": "NON_PARTICIPATING",
|
| 233 |
"74": "SUCCESS",
|
| 234 |
"75": "SUCCESS",
|
| 235 |
-
"76": "
|
| 236 |
"77": "SUCCESS",
|
| 237 |
"78": "SUCCESS",
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "SUCCESS",
|
| 240 |
"80": "SUCCESS",
|
| 241 |
-
"81": "
|
| 242 |
"82": "SUCCESS",
|
| 243 |
-
"83": "
|
| 244 |
"84": "SUCCESS",
|
| 245 |
-
"85": "
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
"87": "SUCCESS",
|
| 248 |
"88": "SUCCESS",
|
|
@@ -253,11 +253,11 @@
|
|
| 253 |
"92": "NON_PARTICIPATING",
|
| 254 |
"93": "NON_PARTICIPATING",
|
| 255 |
"94": "SUCCESS",
|
| 256 |
-
"95": "
|
| 257 |
"96": "SUCCESS",
|
| 258 |
"97": "SUCCESS",
|
| 259 |
"98": "SUCCESS",
|
| 260 |
-
"99": "
|
| 261 |
},
|
| 262 |
"architectures": [
|
| 263 |
"GPTOptim"
|
|
@@ -268,23 +268,20 @@
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
-
|
| 272 |
-
|
| 273 |
-
|
| 274 |
-
|
| 275 |
-
|
| 276 |
-
5539814,
|
| 277 |
-
5539816,
|
| 278 |
-
5539818
|
| 279 |
],
|
| 280 |
"block_size": 1024,
|
| 281 |
"bos_token_id": 50256,
|
| 282 |
"embd_pdrop": 0.1,
|
| 283 |
"eos_token_id": 50256,
|
| 284 |
"initializer_range": 0.02,
|
| 285 |
-
"inner_step":
|
| 286 |
"inner_steps": 0,
|
| 287 |
-
"last_allreduce_block":
|
| 288 |
"layer_norm_epsilon": 1e-05,
|
| 289 |
"model_type": "gpt_optimized",
|
| 290 |
"n_embd": 1280,
|
|
|
|
| 6 |
"1": "SUCCESS",
|
| 7 |
"10": "NON_PARTICIPATING",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
+
"101": "SUCCESS",
|
| 10 |
"102": "NON_PARTICIPATING",
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
"104": "NON_PARTICIPATING",
|
| 13 |
"105": "NON_PARTICIPATING",
|
| 14 |
"106": "SUCCESS",
|
| 15 |
+
"107": "NON_PARTICIPATING",
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
"109": "NON_PARTICIPATING",
|
| 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": "NON_PARTICIPATING",
|
| 27 |
"118": "NON_PARTICIPATING",
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
+
"12": "FAIL",
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 31 |
"121": "SUCCESS",
|
| 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",
|
|
|
|
| 52 |
"140": "NON_PARTICIPATING",
|
| 53 |
"141": "NON_PARTICIPATING",
|
| 54 |
"142": "NON_PARTICIPATING",
|
| 55 |
+
"143": "NON_PARTICIPATING",
|
| 56 |
"144": "NON_PARTICIPATING",
|
| 57 |
+
"145": "SUCCESS",
|
| 58 |
"146": "NON_PARTICIPATING",
|
| 59 |
"147": "NON_PARTICIPATING",
|
| 60 |
+
"148": "SUCCESS",
|
| 61 |
+
"149": "SUCCESS",
|
| 62 |
"15": "SUCCESS",
|
| 63 |
+
"150": "NON_PARTICIPATING",
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
"153": "NON_PARTICIPATING",
|
| 67 |
"154": "SUCCESS",
|
| 68 |
"155": "NON_PARTICIPATING",
|
| 69 |
+
"156": "SUCCESS",
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
+
"158": "NON_PARTICIPATING",
|
| 72 |
+
"159": "NON_PARTICIPATING",
|
| 73 |
"16": "SUCCESS",
|
| 74 |
+
"160": "NON_PARTICIPATING",
|
| 75 |
"161": "NON_PARTICIPATING",
|
| 76 |
"162": "NON_PARTICIPATING",
|
| 77 |
"163": "NON_PARTICIPATING",
|
|
|
|
| 79 |
"165": "NON_PARTICIPATING",
|
| 80 |
"166": "NON_PARTICIPATING",
|
| 81 |
"167": "NON_PARTICIPATING",
|
| 82 |
+
"168": "SUCCESS",
|
| 83 |
+
"169": "SUCCESS",
|
| 84 |
"17": "SUCCESS",
|
| 85 |
"170": "NON_PARTICIPATING",
|
| 86 |
+
"171": "NON_PARTICIPATING",
|
| 87 |
+
"172": "SUCCESS",
|
| 88 |
"173": "NON_PARTICIPATING",
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
+
"175": "NON_PARTICIPATING",
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
"177": "NON_PARTICIPATING",
|
| 93 |
"178": "NON_PARTICIPATING",
|
|
|
|
| 98 |
"182": "NON_PARTICIPATING",
|
| 99 |
"183": "NON_PARTICIPATING",
|
| 100 |
"184": "SUCCESS",
|
| 101 |
+
"185": "SUCCESS",
|
| 102 |
"186": "NON_PARTICIPATING",
|
| 103 |
"187": "NON_PARTICIPATING",
|
| 104 |
"188": "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",
|
| 116 |
"199": "NON_PARTICIPATING",
|
|
|
|
| 118 |
"20": "SUCCESS",
|
| 119 |
"200": "NON_PARTICIPATING",
|
| 120 |
"201": "NON_PARTICIPATING",
|
| 121 |
+
"202": "NON_PARTICIPATING",
|
| 122 |
+
"203": "SUCCESS",
|
| 123 |
"204": "NON_PARTICIPATING",
|
| 124 |
"205": "NON_PARTICIPATING",
|
| 125 |
"206": "NON_PARTICIPATING",
|
| 126 |
"207": "NON_PARTICIPATING",
|
| 127 |
+
"208": "NON_PARTICIPATING",
|
| 128 |
"209": "NON_PARTICIPATING",
|
| 129 |
"21": "SUCCESS",
|
| 130 |
+
"210": "SUCCESS",
|
| 131 |
"211": "SUCCESS",
|
| 132 |
"212": "NON_PARTICIPATING",
|
| 133 |
+
"213": "NON_PARTICIPATING",
|
| 134 |
+
"214": "NON_PARTICIPATING",
|
| 135 |
"215": "NON_PARTICIPATING",
|
| 136 |
"216": "NON_PARTICIPATING",
|
| 137 |
"217": "NON_PARTICIPATING",
|
| 138 |
"218": "NON_PARTICIPATING",
|
| 139 |
+
"219": "NON_PARTICIPATING",
|
| 140 |
"22": "SUCCESS",
|
| 141 |
"220": "NON_PARTICIPATING",
|
| 142 |
"221": "NON_PARTICIPATING",
|
| 143 |
"222": "NON_PARTICIPATING",
|
| 144 |
+
"223": "SUCCESS",
|
| 145 |
"224": "NON_PARTICIPATING",
|
| 146 |
+
"225": "NON_PARTICIPATING",
|
| 147 |
"226": "NON_PARTICIPATING",
|
| 148 |
"227": "NON_PARTICIPATING",
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
+
"229": "SUCCESS",
|
| 151 |
"23": "NON_PARTICIPATING",
|
| 152 |
"230": "NON_PARTICIPATING",
|
| 153 |
"231": "NON_PARTICIPATING",
|
|
|
|
| 157 |
"235": "NON_PARTICIPATING",
|
| 158 |
"236": "NON_PARTICIPATING",
|
| 159 |
"237": "NON_PARTICIPATING",
|
| 160 |
+
"238": "SUCCESS",
|
| 161 |
"239": "NON_PARTICIPATING",
|
| 162 |
"24": "SUCCESS",
|
| 163 |
"240": "NON_PARTICIPATING",
|
|
|
|
| 173 |
"25": "SUCCESS",
|
| 174 |
"250": "SUCCESS",
|
| 175 |
"251": "NON_PARTICIPATING",
|
| 176 |
+
"252": "NON_PARTICIPATING",
|
| 177 |
"253": "NON_PARTICIPATING",
|
| 178 |
"254": "NON_PARTICIPATING",
|
| 179 |
"255": "SUCCESS",
|
| 180 |
+
"26": "NON_PARTICIPATING",
|
| 181 |
"27": "SUCCESS",
|
| 182 |
"28": "SUCCESS",
|
| 183 |
"29": "SUCCESS",
|
|
|
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "NON_PARTICIPATING",
|
| 190 |
+
"35": "NON_PARTICIPATING",
|
| 191 |
"36": "NON_PARTICIPATING",
|
| 192 |
"37": "SUCCESS",
|
| 193 |
+
"38": "SUCCESS",
|
| 194 |
+
"39": "SUCCESS",
|
| 195 |
"4": "NON_PARTICIPATING",
|
| 196 |
"40": "SUCCESS",
|
| 197 |
+
"41": "SUCCESS",
|
| 198 |
"42": "NON_PARTICIPATING",
|
| 199 |
"43": "SUCCESS",
|
| 200 |
"44": "SUCCESS",
|
|
|
|
| 208 |
"51": "SUCCESS",
|
| 209 |
"52": "SUCCESS",
|
| 210 |
"53": "NON_PARTICIPATING",
|
| 211 |
+
"54": "NON_PARTICIPATING",
|
| 212 |
"55": "SUCCESS",
|
| 213 |
"56": "SUCCESS",
|
| 214 |
"57": "SUCCESS",
|
| 215 |
"58": "NON_PARTICIPATING",
|
| 216 |
"59": "NON_PARTICIPATING",
|
| 217 |
"6": "NON_PARTICIPATING",
|
| 218 |
+
"60": "FAIL",
|
| 219 |
"61": "NON_PARTICIPATING",
|
| 220 |
"62": "SUCCESS",
|
| 221 |
"63": "SUCCESS",
|
|
|
|
| 223 |
"65": "SUCCESS",
|
| 224 |
"66": "SUCCESS",
|
| 225 |
"67": "SUCCESS",
|
| 226 |
+
"68": "SUCCESS",
|
| 227 |
"69": "NON_PARTICIPATING",
|
| 228 |
"7": "SUCCESS",
|
| 229 |
"70": "NON_PARTICIPATING",
|
|
|
|
| 232 |
"73": "NON_PARTICIPATING",
|
| 233 |
"74": "SUCCESS",
|
| 234 |
"75": "SUCCESS",
|
| 235 |
+
"76": "SUCCESS",
|
| 236 |
"77": "SUCCESS",
|
| 237 |
"78": "SUCCESS",
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "SUCCESS",
|
| 240 |
"80": "SUCCESS",
|
| 241 |
+
"81": "SUCCESS",
|
| 242 |
"82": "SUCCESS",
|
| 243 |
+
"83": "NON_PARTICIPATING",
|
| 244 |
"84": "SUCCESS",
|
| 245 |
+
"85": "NON_PARTICIPATING",
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
"87": "SUCCESS",
|
| 248 |
"88": "SUCCESS",
|
|
|
|
| 253 |
"92": "NON_PARTICIPATING",
|
| 254 |
"93": "NON_PARTICIPATING",
|
| 255 |
"94": "SUCCESS",
|
| 256 |
+
"95": "SUCCESS",
|
| 257 |
"96": "SUCCESS",
|
| 258 |
"97": "SUCCESS",
|
| 259 |
"98": "SUCCESS",
|
| 260 |
+
"99": "SUCCESS"
|
| 261 |
},
|
| 262 |
"architectures": [
|
| 263 |
"GPTOptim"
|
|
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
+
5539865,
|
| 272 |
+
5539867,
|
| 273 |
+
5539869,
|
| 274 |
+
5539872,
|
| 275 |
+
5539874
|
|
|
|
|
|
|
|
|
|
| 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": 5539810,
|
| 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:dfb85cbf0cd52fa2e85cc7781441c87c6133c934083252001d41f9d837a364f6
|
| 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 sha256:130bff7c6b3e2aa7641fb0c9596e4ecde06a35c863a3c2c10f634717c82ebef8
|
| 3 |
size 4040701744
|