Run 4. Outer Step 11. Inner Step 0.
Browse files- config.json +75 -75
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
config.json
CHANGED
|
@@ -1,33 +1,33 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "
|
| 3 |
"activation_function": "gelu_new",
|
| 4 |
"all_reduce_scores": {
|
| 5 |
"0": "NON_PARTICIPATING",
|
| 6 |
-
"1": "
|
| 7 |
"10": "SUCCESS",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
-
"101": "
|
| 10 |
-
"102": "
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
-
"104": "
|
| 13 |
"105": "NON_PARTICIPATING",
|
| 14 |
"106": "NON_PARTICIPATING",
|
| 15 |
-
"107": "
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
-
"109": "
|
| 18 |
"11": "SUCCESS",
|
| 19 |
"110": "SUCCESS",
|
| 20 |
"111": "NON_PARTICIPATING",
|
| 21 |
"112": "NON_PARTICIPATING",
|
| 22 |
"113": "NON_PARTICIPATING",
|
| 23 |
-
"114": "
|
| 24 |
"115": "NON_PARTICIPATING",
|
| 25 |
"116": "NON_PARTICIPATING",
|
| 26 |
-
"117": "
|
| 27 |
"118": "NON_PARTICIPATING",
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
-
"12": "
|
| 30 |
-
"120": "
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
"122": "SUCCESS",
|
| 33 |
"123": "NON_PARTICIPATING",
|
|
@@ -40,7 +40,7 @@
|
|
| 40 |
"13": "NON_PARTICIPATING",
|
| 41 |
"130": "NON_PARTICIPATING",
|
| 42 |
"131": "NON_PARTICIPATING",
|
| 43 |
-
"132": "
|
| 44 |
"133": "NON_PARTICIPATING",
|
| 45 |
"134": "SUCCESS",
|
| 46 |
"135": "SUCCESS",
|
|
@@ -54,40 +54,40 @@
|
|
| 54 |
"142": "SUCCESS",
|
| 55 |
"143": "NON_PARTICIPATING",
|
| 56 |
"144": "NON_PARTICIPATING",
|
| 57 |
-
"145": "
|
| 58 |
"146": "NON_PARTICIPATING",
|
| 59 |
-
"147": "
|
| 60 |
-
"148": "
|
| 61 |
-
"149": "
|
| 62 |
-
"15": "
|
| 63 |
"150": "NON_PARTICIPATING",
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
-
"153": "
|
| 67 |
-
"154": "
|
| 68 |
"155": "NON_PARTICIPATING",
|
| 69 |
"156": "SUCCESS",
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
"159": "NON_PARTICIPATING",
|
| 73 |
"16": "NON_PARTICIPATING",
|
| 74 |
-
"160": "
|
| 75 |
"161": "NON_PARTICIPATING",
|
| 76 |
"162": "NON_PARTICIPATING",
|
| 77 |
"163": "NON_PARTICIPATING",
|
| 78 |
"164": "NON_PARTICIPATING",
|
| 79 |
"165": "NON_PARTICIPATING",
|
| 80 |
"166": "NON_PARTICIPATING",
|
| 81 |
-
"167": "
|
| 82 |
"168": "NON_PARTICIPATING",
|
| 83 |
"169": "NON_PARTICIPATING",
|
| 84 |
"17": "SUCCESS",
|
| 85 |
-
"170": "
|
| 86 |
"171": "NON_PARTICIPATING",
|
| 87 |
-
"172": "
|
| 88 |
"173": "NON_PARTICIPATING",
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
-
"175": "
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
"177": "SUCCESS",
|
| 93 |
"178": "SUCCESS",
|
|
@@ -97,7 +97,7 @@
|
|
| 97 |
"181": "NON_PARTICIPATING",
|
| 98 |
"182": "NON_PARTICIPATING",
|
| 99 |
"183": "NON_PARTICIPATING",
|
| 100 |
-
"184": "
|
| 101 |
"185": "NON_PARTICIPATING",
|
| 102 |
"186": "SUCCESS",
|
| 103 |
"187": "NON_PARTICIPATING",
|
|
@@ -106,23 +106,23 @@
|
|
| 106 |
"19": "NON_PARTICIPATING",
|
| 107 |
"190": "NON_PARTICIPATING",
|
| 108 |
"191": "NON_PARTICIPATING",
|
| 109 |
-
"192": "
|
| 110 |
-
"193": "
|
| 111 |
-
"194": "
|
| 112 |
"195": "SUCCESS",
|
| 113 |
"196": "NON_PARTICIPATING",
|
| 114 |
"197": "NON_PARTICIPATING",
|
| 115 |
"198": "NON_PARTICIPATING",
|
| 116 |
"199": "NON_PARTICIPATING",
|
| 117 |
-
"2": "
|
| 118 |
"20": "SUCCESS",
|
| 119 |
"200": "SUCCESS",
|
| 120 |
-
"201": "
|
| 121 |
-
"202": "
|
| 122 |
-
"203": "
|
| 123 |
"204": "NON_PARTICIPATING",
|
| 124 |
"205": "NON_PARTICIPATING",
|
| 125 |
-
"206": "
|
| 126 |
"207": "NON_PARTICIPATING",
|
| 127 |
"208": "NON_PARTICIPATING",
|
| 128 |
"209": "NON_PARTICIPATING",
|
|
@@ -133,22 +133,22 @@
|
|
| 133 |
"213": "NON_PARTICIPATING",
|
| 134 |
"214": "NON_PARTICIPATING",
|
| 135 |
"215": "NON_PARTICIPATING",
|
| 136 |
-
"216": "
|
| 137 |
"217": "NON_PARTICIPATING",
|
| 138 |
-
"218": "
|
| 139 |
"219": "NON_PARTICIPATING",
|
| 140 |
-
"22": "
|
| 141 |
"220": "NON_PARTICIPATING",
|
| 142 |
"221": "NON_PARTICIPATING",
|
| 143 |
-
"222": "
|
| 144 |
-
"223": "
|
| 145 |
"224": "NON_PARTICIPATING",
|
| 146 |
"225": "SUCCESS",
|
| 147 |
-
"226": "
|
| 148 |
"227": "NON_PARTICIPATING",
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
"229": "NON_PARTICIPATING",
|
| 151 |
-
"23": "
|
| 152 |
"230": "NON_PARTICIPATING",
|
| 153 |
"231": "NON_PARTICIPATING",
|
| 154 |
"232": "NON_PARTICIPATING",
|
|
@@ -161,7 +161,7 @@
|
|
| 161 |
"239": "NON_PARTICIPATING",
|
| 162 |
"24": "NON_PARTICIPATING",
|
| 163 |
"240": "NON_PARTICIPATING",
|
| 164 |
-
"241": "
|
| 165 |
"242": "NON_PARTICIPATING",
|
| 166 |
"243": "NON_PARTICIPATING",
|
| 167 |
"244": "NON_PARTICIPATING",
|
|
@@ -169,9 +169,9 @@
|
|
| 169 |
"246": "NON_PARTICIPATING",
|
| 170 |
"247": "SUCCESS",
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
-
"249": "
|
| 173 |
"25": "SUCCESS",
|
| 174 |
-
"250": "
|
| 175 |
"251": "NON_PARTICIPATING",
|
| 176 |
"252": "NON_PARTICIPATING",
|
| 177 |
"253": "NON_PARTICIPATING",
|
|
@@ -189,75 +189,75 @@
|
|
| 189 |
"34": "SUCCESS",
|
| 190 |
"35": "NON_PARTICIPATING",
|
| 191 |
"36": "NON_PARTICIPATING",
|
| 192 |
-
"37": "
|
| 193 |
"38": "NON_PARTICIPATING",
|
| 194 |
"39": "SUCCESS",
|
| 195 |
"4": "SUCCESS",
|
| 196 |
-
"40": "
|
| 197 |
-
"41": "
|
| 198 |
"42": "NON_PARTICIPATING",
|
| 199 |
-
"43": "
|
| 200 |
"44": "SUCCESS",
|
| 201 |
-
"45": "
|
| 202 |
"46": "NON_PARTICIPATING",
|
| 203 |
"47": "NON_PARTICIPATING",
|
| 204 |
"48": "NON_PARTICIPATING",
|
| 205 |
"49": "SUCCESS",
|
| 206 |
"5": "NON_PARTICIPATING",
|
| 207 |
-
"50": "
|
| 208 |
"51": "SUCCESS",
|
| 209 |
"52": "NON_PARTICIPATING",
|
| 210 |
-
"53": "
|
| 211 |
"54": "NON_PARTICIPATING",
|
| 212 |
"55": "SUCCESS",
|
| 213 |
-
"56": "
|
| 214 |
-
"57": "
|
| 215 |
"58": "SUCCESS",
|
| 216 |
"59": "NON_PARTICIPATING",
|
| 217 |
-
"6": "
|
| 218 |
-
"60": "
|
| 219 |
"61": "NON_PARTICIPATING",
|
| 220 |
-
"62": "
|
| 221 |
-
"63": "
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
"65": "NON_PARTICIPATING",
|
| 224 |
"66": "NON_PARTICIPATING",
|
| 225 |
-
"67": "
|
| 226 |
"68": "NON_PARTICIPATING",
|
| 227 |
"69": "NON_PARTICIPATING",
|
| 228 |
"7": "NON_PARTICIPATING",
|
| 229 |
-
"70": "
|
| 230 |
"71": "NON_PARTICIPATING",
|
| 231 |
"72": "NON_PARTICIPATING",
|
| 232 |
"73": "NON_PARTICIPATING",
|
| 233 |
"74": "NON_PARTICIPATING",
|
| 234 |
-
"75": "
|
| 235 |
"76": "NON_PARTICIPATING",
|
| 236 |
"77": "NON_PARTICIPATING",
|
| 237 |
"78": "SUCCESS",
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "NON_PARTICIPATING",
|
| 240 |
-
"80": "
|
| 241 |
"81": "NON_PARTICIPATING",
|
| 242 |
-
"82": "
|
| 243 |
"83": "NON_PARTICIPATING",
|
| 244 |
-
"84": "
|
| 245 |
"85": "SUCCESS",
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
-
"87": "
|
| 248 |
"88": "SUCCESS",
|
| 249 |
-
"89": "
|
| 250 |
"9": "NON_PARTICIPATING",
|
| 251 |
"90": "NON_PARTICIPATING",
|
| 252 |
"91": "NON_PARTICIPATING",
|
| 253 |
-
"92": "
|
| 254 |
"93": "NON_PARTICIPATING",
|
| 255 |
"94": "NON_PARTICIPATING",
|
| 256 |
-
"95": "
|
| 257 |
-
"96": "
|
| 258 |
"97": "NON_PARTICIPATING",
|
| 259 |
"98": "NON_PARTICIPATING",
|
| 260 |
-
"99": "
|
| 261 |
},
|
| 262 |
"architectures": [
|
| 263 |
"GPTOptim"
|
|
@@ -268,20 +268,20 @@
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
-
|
| 272 |
-
|
| 273 |
-
|
| 274 |
-
|
| 275 |
-
|
| 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":
|
| 283 |
"inner_steps": 0,
|
| 284 |
-
"last_allreduce_block":
|
| 285 |
"layer_norm_epsilon": 1e-05,
|
| 286 |
"model_type": "gpt_optimized",
|
| 287 |
"n_embd": 1280,
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "distributed/optimized-gpt2-1b",
|
| 3 |
"activation_function": "gelu_new",
|
| 4 |
"all_reduce_scores": {
|
| 5 |
"0": "NON_PARTICIPATING",
|
| 6 |
+
"1": "NON_PARTICIPATING",
|
| 7 |
"10": "SUCCESS",
|
| 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": "NON_PARTICIPATING",
|
| 18 |
"11": "SUCCESS",
|
| 19 |
"110": "SUCCESS",
|
| 20 |
"111": "NON_PARTICIPATING",
|
| 21 |
"112": "NON_PARTICIPATING",
|
| 22 |
"113": "NON_PARTICIPATING",
|
| 23 |
+
"114": "SUCCESS",
|
| 24 |
"115": "NON_PARTICIPATING",
|
| 25 |
"116": "NON_PARTICIPATING",
|
| 26 |
+
"117": "NON_PARTICIPATING",
|
| 27 |
"118": "NON_PARTICIPATING",
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
+
"12": "NON_PARTICIPATING",
|
| 30 |
+
"120": "SUCCESS",
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
"122": "SUCCESS",
|
| 33 |
"123": "NON_PARTICIPATING",
|
|
|
|
| 40 |
"13": "NON_PARTICIPATING",
|
| 41 |
"130": "NON_PARTICIPATING",
|
| 42 |
"131": "NON_PARTICIPATING",
|
| 43 |
+
"132": "NON_PARTICIPATING",
|
| 44 |
"133": "NON_PARTICIPATING",
|
| 45 |
"134": "SUCCESS",
|
| 46 |
"135": "SUCCESS",
|
|
|
|
| 54 |
"142": "SUCCESS",
|
| 55 |
"143": "NON_PARTICIPATING",
|
| 56 |
"144": "NON_PARTICIPATING",
|
| 57 |
+
"145": "SUCCESS",
|
| 58 |
"146": "NON_PARTICIPATING",
|
| 59 |
+
"147": "SUCCESS",
|
| 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": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
"159": "NON_PARTICIPATING",
|
| 73 |
"16": "NON_PARTICIPATING",
|
| 74 |
+
"160": "NON_PARTICIPATING",
|
| 75 |
"161": "NON_PARTICIPATING",
|
| 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": "SUCCESS",
|
| 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": "SUCCESS",
|
| 93 |
"178": "SUCCESS",
|
|
|
|
| 97 |
"181": "NON_PARTICIPATING",
|
| 98 |
"182": "NON_PARTICIPATING",
|
| 99 |
"183": "NON_PARTICIPATING",
|
| 100 |
+
"184": "SUCCESS",
|
| 101 |
"185": "NON_PARTICIPATING",
|
| 102 |
"186": "SUCCESS",
|
| 103 |
"187": "NON_PARTICIPATING",
|
|
|
|
| 106 |
"19": "NON_PARTICIPATING",
|
| 107 |
"190": "NON_PARTICIPATING",
|
| 108 |
"191": "NON_PARTICIPATING",
|
| 109 |
+
"192": "NON_PARTICIPATING",
|
| 110 |
+
"193": "SUCCESS",
|
| 111 |
+
"194": "SUCCESS",
|
| 112 |
"195": "SUCCESS",
|
| 113 |
"196": "NON_PARTICIPATING",
|
| 114 |
"197": "NON_PARTICIPATING",
|
| 115 |
"198": "NON_PARTICIPATING",
|
| 116 |
"199": "NON_PARTICIPATING",
|
| 117 |
+
"2": "NON_PARTICIPATING",
|
| 118 |
"20": "SUCCESS",
|
| 119 |
"200": "SUCCESS",
|
| 120 |
+
"201": "SUCCESS",
|
| 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": "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": "NON_PARTICIPATING",
|
| 141 |
"220": "NON_PARTICIPATING",
|
| 142 |
"221": "NON_PARTICIPATING",
|
| 143 |
+
"222": "NON_PARTICIPATING",
|
| 144 |
+
"223": "NON_PARTICIPATING",
|
| 145 |
"224": "NON_PARTICIPATING",
|
| 146 |
"225": "SUCCESS",
|
| 147 |
+
"226": "SUCCESS",
|
| 148 |
"227": "NON_PARTICIPATING",
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
"229": "NON_PARTICIPATING",
|
| 151 |
+
"23": "NON_PARTICIPATING",
|
| 152 |
"230": "NON_PARTICIPATING",
|
| 153 |
"231": "NON_PARTICIPATING",
|
| 154 |
"232": "NON_PARTICIPATING",
|
|
|
|
| 161 |
"239": "NON_PARTICIPATING",
|
| 162 |
"24": "NON_PARTICIPATING",
|
| 163 |
"240": "NON_PARTICIPATING",
|
| 164 |
+
"241": "SUCCESS",
|
| 165 |
"242": "NON_PARTICIPATING",
|
| 166 |
"243": "NON_PARTICIPATING",
|
| 167 |
"244": "NON_PARTICIPATING",
|
|
|
|
| 169 |
"246": "NON_PARTICIPATING",
|
| 170 |
"247": "SUCCESS",
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
+
"249": "NON_PARTICIPATING",
|
| 173 |
"25": "SUCCESS",
|
| 174 |
+
"250": "NON_PARTICIPATING",
|
| 175 |
"251": "NON_PARTICIPATING",
|
| 176 |
"252": "NON_PARTICIPATING",
|
| 177 |
"253": "NON_PARTICIPATING",
|
|
|
|
| 189 |
"34": "SUCCESS",
|
| 190 |
"35": "NON_PARTICIPATING",
|
| 191 |
"36": "NON_PARTICIPATING",
|
| 192 |
+
"37": "NON_PARTICIPATING",
|
| 193 |
"38": "NON_PARTICIPATING",
|
| 194 |
"39": "SUCCESS",
|
| 195 |
"4": "SUCCESS",
|
| 196 |
+
"40": "NON_PARTICIPATING",
|
| 197 |
+
"41": "NON_PARTICIPATING",
|
| 198 |
"42": "NON_PARTICIPATING",
|
| 199 |
+
"43": "NON_PARTICIPATING",
|
| 200 |
"44": "SUCCESS",
|
| 201 |
+
"45": "NON_PARTICIPATING",
|
| 202 |
"46": "NON_PARTICIPATING",
|
| 203 |
"47": "NON_PARTICIPATING",
|
| 204 |
"48": "NON_PARTICIPATING",
|
| 205 |
"49": "SUCCESS",
|
| 206 |
"5": "NON_PARTICIPATING",
|
| 207 |
+
"50": "NON_PARTICIPATING",
|
| 208 |
"51": "SUCCESS",
|
| 209 |
"52": "NON_PARTICIPATING",
|
| 210 |
+
"53": "NON_PARTICIPATING",
|
| 211 |
"54": "NON_PARTICIPATING",
|
| 212 |
"55": "SUCCESS",
|
| 213 |
+
"56": "NON_PARTICIPATING",
|
| 214 |
+
"57": "SUCCESS",
|
| 215 |
"58": "SUCCESS",
|
| 216 |
"59": "NON_PARTICIPATING",
|
| 217 |
+
"6": "NON_PARTICIPATING",
|
| 218 |
+
"60": "NON_PARTICIPATING",
|
| 219 |
"61": "NON_PARTICIPATING",
|
| 220 |
+
"62": "NON_PARTICIPATING",
|
| 221 |
+
"63": "NON_PARTICIPATING",
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
"65": "NON_PARTICIPATING",
|
| 224 |
"66": "NON_PARTICIPATING",
|
| 225 |
+
"67": "NON_PARTICIPATING",
|
| 226 |
"68": "NON_PARTICIPATING",
|
| 227 |
"69": "NON_PARTICIPATING",
|
| 228 |
"7": "NON_PARTICIPATING",
|
| 229 |
+
"70": "NON_PARTICIPATING",
|
| 230 |
"71": "NON_PARTICIPATING",
|
| 231 |
"72": "NON_PARTICIPATING",
|
| 232 |
"73": "NON_PARTICIPATING",
|
| 233 |
"74": "NON_PARTICIPATING",
|
| 234 |
+
"75": "NON_PARTICIPATING",
|
| 235 |
"76": "NON_PARTICIPATING",
|
| 236 |
"77": "NON_PARTICIPATING",
|
| 237 |
"78": "SUCCESS",
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "NON_PARTICIPATING",
|
| 240 |
+
"80": "NON_PARTICIPATING",
|
| 241 |
"81": "NON_PARTICIPATING",
|
| 242 |
+
"82": "NON_PARTICIPATING",
|
| 243 |
"83": "NON_PARTICIPATING",
|
| 244 |
+
"84": "NON_PARTICIPATING",
|
| 245 |
"85": "SUCCESS",
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
+
"87": "NON_PARTICIPATING",
|
| 248 |
"88": "SUCCESS",
|
| 249 |
+
"89": "NON_PARTICIPATING",
|
| 250 |
"9": "NON_PARTICIPATING",
|
| 251 |
"90": "NON_PARTICIPATING",
|
| 252 |
"91": "NON_PARTICIPATING",
|
| 253 |
+
"92": "SUCCESS",
|
| 254 |
"93": "NON_PARTICIPATING",
|
| 255 |
"94": "NON_PARTICIPATING",
|
| 256 |
+
"95": "NON_PARTICIPATING",
|
| 257 |
+
"96": "NON_PARTICIPATING",
|
| 258 |
"97": "NON_PARTICIPATING",
|
| 259 |
"98": "NON_PARTICIPATING",
|
| 260 |
+
"99": "NON_PARTICIPATING"
|
| 261 |
},
|
| 262 |
"architectures": [
|
| 263 |
"GPTOptim"
|
|
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
+
5507088,
|
| 272 |
+
5507102,
|
| 273 |
+
5507106,
|
| 274 |
+
5507109,
|
| 275 |
+
5507113
|
| 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": 5506953,
|
| 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 sha256:0d35779c9f5dbd0c6ddb93abd8dfd0286b96d68d9c25cc7d0fe47f3b1c2bae58
|
| 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:ec321ec87ff24fbee5e185d8508c22a34151ace9e6888abf32c75a2e877d9e03
|
| 3 |
size 4040701744
|