Run 4. Outer Step 93. Inner Step 2.
Browse files- config.json +64 -80
- inner_optimizer.pt +2 -2
- model.safetensors +1 -1
config.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "
|
| 3 |
"activation_function": "gelu_new",
|
| 4 |
"all_reduce_scores": {
|
| 5 |
"0": "NON_PARTICIPATING",
|
| 6 |
-
"1": "
|
| 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",
|
|
@@ -31,21 +31,21 @@
|
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
"122": "NON_PARTICIPATING",
|
| 33 |
"123": "NON_PARTICIPATING",
|
| 34 |
-
"124": "
|
| 35 |
-
"125": "
|
| 36 |
"126": "NON_PARTICIPATING",
|
| 37 |
"127": "NON_PARTICIPATING",
|
| 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": "NON_PARTICIPATING",
|
| 48 |
-
"137": "
|
| 49 |
"138": "NON_PARTICIPATING",
|
| 50 |
"139": "NON_PARTICIPATING",
|
| 51 |
"14": "SUCCESS",
|
|
@@ -56,23 +56,23 @@
|
|
| 56 |
"144": "NON_PARTICIPATING",
|
| 57 |
"145": "NON_PARTICIPATING",
|
| 58 |
"146": "NON_PARTICIPATING",
|
| 59 |
-
"147": "
|
| 60 |
"148": "NON_PARTICIPATING",
|
| 61 |
"149": "NON_PARTICIPATING",
|
| 62 |
"15": "SUCCESS",
|
| 63 |
-
"150": "
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
-
"153": "
|
| 67 |
"154": "NON_PARTICIPATING",
|
| 68 |
"155": "NON_PARTICIPATING",
|
| 69 |
"156": "NON_PARTICIPATING",
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
"159": "NON_PARTICIPATING",
|
| 73 |
-
"16": "
|
| 74 |
"160": "NON_PARTICIPATING",
|
| 75 |
-
"161": "
|
| 76 |
"162": "SUCCESS",
|
| 77 |
"163": "NON_PARTICIPATING",
|
| 78 |
"164": "NON_PARTICIPATING",
|
|
@@ -86,7 +86,7 @@
|
|
| 86 |
"171": "NON_PARTICIPATING",
|
| 87 |
"172": "NON_PARTICIPATING",
|
| 88 |
"173": "NON_PARTICIPATING",
|
| 89 |
-
"174": "
|
| 90 |
"175": "NON_PARTICIPATING",
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
"177": "NON_PARTICIPATING",
|
|
@@ -96,17 +96,17 @@
|
|
| 96 |
"180": "NON_PARTICIPATING",
|
| 97 |
"181": "NON_PARTICIPATING",
|
| 98 |
"182": "NON_PARTICIPATING",
|
| 99 |
-
"183": "
|
| 100 |
"184": "NON_PARTICIPATING",
|
| 101 |
-
"185": "
|
| 102 |
-
"186": "
|
| 103 |
"187": "NON_PARTICIPATING",
|
| 104 |
"188": "NON_PARTICIPATING",
|
| 105 |
"189": "NON_PARTICIPATING",
|
| 106 |
"19": "SUCCESS",
|
| 107 |
"190": "NON_PARTICIPATING",
|
| 108 |
"191": "NON_PARTICIPATING",
|
| 109 |
-
"192": "
|
| 110 |
"193": "NON_PARTICIPATING",
|
| 111 |
"194": "NON_PARTICIPATING",
|
| 112 |
"195": "NON_PARTICIPATING",
|
|
@@ -119,12 +119,12 @@
|
|
| 119 |
"200": "NON_PARTICIPATING",
|
| 120 |
"201": "NON_PARTICIPATING",
|
| 121 |
"202": "NON_PARTICIPATING",
|
| 122 |
-
"203": "
|
| 123 |
-
"204": "
|
| 124 |
"205": "SUCCESS",
|
| 125 |
"206": "NON_PARTICIPATING",
|
| 126 |
"207": "NON_PARTICIPATING",
|
| 127 |
-
"208": "
|
| 128 |
"209": "NON_PARTICIPATING",
|
| 129 |
"21": "SUCCESS",
|
| 130 |
"210": "NON_PARTICIPATING",
|
|
@@ -133,10 +133,10 @@
|
|
| 133 |
"213": "NON_PARTICIPATING",
|
| 134 |
"214": "NON_PARTICIPATING",
|
| 135 |
"215": "NON_PARTICIPATING",
|
| 136 |
-
"216": "
|
| 137 |
-
"217": "
|
| 138 |
"218": "NON_PARTICIPATING",
|
| 139 |
-
"219": "
|
| 140 |
"22": "SUCCESS",
|
| 141 |
"220": "NON_PARTICIPATING",
|
| 142 |
"221": "NON_PARTICIPATING",
|
|
@@ -147,10 +147,10 @@
|
|
| 147 |
"226": "NON_PARTICIPATING",
|
| 148 |
"227": "NON_PARTICIPATING",
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
-
"229": "
|
| 151 |
-
"23": "
|
| 152 |
"230": "NON_PARTICIPATING",
|
| 153 |
-
"231": "
|
| 154 |
"232": "NON_PARTICIPATING",
|
| 155 |
"233": "NON_PARTICIPATING",
|
| 156 |
"234": "NON_PARTICIPATING",
|
|
@@ -164,97 +164,97 @@
|
|
| 164 |
"241": "NON_PARTICIPATING",
|
| 165 |
"242": "NON_PARTICIPATING",
|
| 166 |
"243": "NON_PARTICIPATING",
|
| 167 |
-
"244": "
|
| 168 |
"245": "NON_PARTICIPATING",
|
| 169 |
"246": "NON_PARTICIPATING",
|
| 170 |
-
"247": "
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
"249": "NON_PARTICIPATING",
|
| 173 |
"25": "SUCCESS",
|
| 174 |
"250": "NON_PARTICIPATING",
|
| 175 |
"251": "NON_PARTICIPATING",
|
| 176 |
-
"252": "
|
| 177 |
"253": "NON_PARTICIPATING",
|
| 178 |
"254": "NON_PARTICIPATING",
|
| 179 |
"255": "NON_PARTICIPATING",
|
| 180 |
"26": "SUCCESS",
|
| 181 |
"27": "SUCCESS",
|
| 182 |
-
"28": "
|
| 183 |
"29": "SUCCESS",
|
| 184 |
-
"3": "
|
| 185 |
-
"30": "
|
| 186 |
"31": "SUCCESS",
|
| 187 |
-
"32": "
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "SUCCESS",
|
| 190 |
"35": "SUCCESS",
|
| 191 |
-
"36": "
|
| 192 |
-
"37": "
|
| 193 |
"38": "SUCCESS",
|
| 194 |
-
"39": "
|
| 195 |
"4": "SUCCESS",
|
| 196 |
"40": "SUCCESS",
|
| 197 |
"41": "SUCCESS",
|
| 198 |
-
"42": "
|
| 199 |
-
"43": "
|
| 200 |
"44": "SUCCESS",
|
| 201 |
"45": "SUCCESS",
|
| 202 |
"46": "NON_PARTICIPATING",
|
| 203 |
-
"47": "
|
| 204 |
"48": "NON_PARTICIPATING",
|
| 205 |
"49": "SUCCESS",
|
| 206 |
-
"5": "
|
| 207 |
-
"50": "
|
| 208 |
"51": "SUCCESS",
|
| 209 |
"52": "SUCCESS",
|
| 210 |
-
"53": "
|
| 211 |
"54": "SUCCESS",
|
| 212 |
-
"55": "
|
| 213 |
"56": "SUCCESS",
|
| 214 |
-
"57": "
|
| 215 |
"58": "NON_PARTICIPATING",
|
| 216 |
"59": "SUCCESS",
|
| 217 |
"6": "NON_PARTICIPATING",
|
| 218 |
"60": "SUCCESS",
|
| 219 |
-
"61": "
|
| 220 |
"62": "SUCCESS",
|
| 221 |
"63": "SUCCESS",
|
| 222 |
-
"64": "
|
| 223 |
"65": "SUCCESS",
|
| 224 |
-
"66": "
|
| 225 |
"67": "SUCCESS",
|
| 226 |
-
"68": "
|
| 227 |
-
"69": "
|
| 228 |
"7": "SUCCESS",
|
| 229 |
-
"70": "
|
| 230 |
-
"71": "
|
| 231 |
"72": "SUCCESS",
|
| 232 |
-
"73": "
|
| 233 |
"74": "SUCCESS",
|
| 234 |
"75": "SUCCESS",
|
| 235 |
-
"76": "
|
| 236 |
"77": "SUCCESS",
|
| 237 |
-
"78": "
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "SUCCESS",
|
| 240 |
"80": "SUCCESS",
|
| 241 |
-
"81": "
|
| 242 |
"82": "SUCCESS",
|
| 243 |
"83": "NON_PARTICIPATING",
|
| 244 |
"84": "SUCCESS",
|
| 245 |
"85": "SUCCESS",
|
| 246 |
-
"86": "
|
| 247 |
"87": "SUCCESS",
|
| 248 |
-
"88": "
|
| 249 |
"89": "SUCCESS",
|
| 250 |
-
"9": "
|
| 251 |
"90": "SUCCESS",
|
| 252 |
"91": "SUCCESS",
|
| 253 |
"92": "SUCCESS",
|
| 254 |
"93": "NON_PARTICIPATING",
|
| 255 |
-
"94": "
|
| 256 |
"95": "SUCCESS",
|
| 257 |
-
"96": "
|
| 258 |
"97": "SUCCESS",
|
| 259 |
"98": "SUCCESS",
|
| 260 |
"99": "SUCCESS"
|
|
@@ -268,33 +268,17 @@
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
-
|
| 272 |
-
|
| 273 |
-
5622459,
|
| 274 |
-
5622463,
|
| 275 |
-
5622467,
|
| 276 |
-
5622472,
|
| 277 |
-
5622476,
|
| 278 |
-
5622480,
|
| 279 |
-
5622484,
|
| 280 |
-
5622488,
|
| 281 |
-
5622493,
|
| 282 |
-
5622497,
|
| 283 |
-
5622501,
|
| 284 |
-
5622505,
|
| 285 |
-
5622510,
|
| 286 |
-
5622514,
|
| 287 |
-
5622518,
|
| 288 |
-
5622522
|
| 289 |
],
|
| 290 |
"block_size": 1024,
|
| 291 |
"bos_token_id": 50256,
|
| 292 |
"embd_pdrop": 0.1,
|
| 293 |
"eos_token_id": 50256,
|
| 294 |
"initializer_range": 0.02,
|
| 295 |
-
"inner_step":
|
| 296 |
"inner_steps": 0,
|
| 297 |
-
"last_allreduce_block":
|
| 298 |
"layer_norm_epsilon": 1e-05,
|
| 299 |
"model_type": "gpt_optimized",
|
| 300 |
"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": "NON_PARTICIPATING",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
"101": "NON_PARTICIPATING",
|
| 10 |
+
"102": "NON_PARTICIPATING",
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
"104": "NON_PARTICIPATING",
|
| 13 |
"105": "NON_PARTICIPATING",
|
|
|
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
"122": "NON_PARTICIPATING",
|
| 33 |
"123": "NON_PARTICIPATING",
|
| 34 |
+
"124": "NON_PARTICIPATING",
|
| 35 |
+
"125": "SUCCESS",
|
| 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": "NON_PARTICIPATING",
|
| 48 |
+
"137": "NON_PARTICIPATING",
|
| 49 |
"138": "NON_PARTICIPATING",
|
| 50 |
"139": "NON_PARTICIPATING",
|
| 51 |
"14": "SUCCESS",
|
|
|
|
| 56 |
"144": "NON_PARTICIPATING",
|
| 57 |
"145": "NON_PARTICIPATING",
|
| 58 |
"146": "NON_PARTICIPATING",
|
| 59 |
+
"147": "NON_PARTICIPATING",
|
| 60 |
"148": "NON_PARTICIPATING",
|
| 61 |
"149": "NON_PARTICIPATING",
|
| 62 |
"15": "SUCCESS",
|
| 63 |
+
"150": "NON_PARTICIPATING",
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
+
"153": "NON_PARTICIPATING",
|
| 67 |
"154": "NON_PARTICIPATING",
|
| 68 |
"155": "NON_PARTICIPATING",
|
| 69 |
"156": "NON_PARTICIPATING",
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
"159": "NON_PARTICIPATING",
|
| 73 |
+
"16": "NON_PARTICIPATING",
|
| 74 |
"160": "NON_PARTICIPATING",
|
| 75 |
+
"161": "SUCCESS",
|
| 76 |
"162": "SUCCESS",
|
| 77 |
"163": "NON_PARTICIPATING",
|
| 78 |
"164": "NON_PARTICIPATING",
|
|
|
|
| 86 |
"171": "NON_PARTICIPATING",
|
| 87 |
"172": "NON_PARTICIPATING",
|
| 88 |
"173": "NON_PARTICIPATING",
|
| 89 |
+
"174": "SUCCESS",
|
| 90 |
"175": "NON_PARTICIPATING",
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
"177": "NON_PARTICIPATING",
|
|
|
|
| 96 |
"180": "NON_PARTICIPATING",
|
| 97 |
"181": "NON_PARTICIPATING",
|
| 98 |
"182": "NON_PARTICIPATING",
|
| 99 |
+
"183": "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": "SUCCESS",
|
| 107 |
"190": "NON_PARTICIPATING",
|
| 108 |
"191": "NON_PARTICIPATING",
|
| 109 |
+
"192": "SUCCESS",
|
| 110 |
"193": "NON_PARTICIPATING",
|
| 111 |
"194": "NON_PARTICIPATING",
|
| 112 |
"195": "NON_PARTICIPATING",
|
|
|
|
| 119 |
"200": "NON_PARTICIPATING",
|
| 120 |
"201": "NON_PARTICIPATING",
|
| 121 |
"202": "NON_PARTICIPATING",
|
| 122 |
+
"203": "NON_PARTICIPATING",
|
| 123 |
+
"204": "NON_PARTICIPATING",
|
| 124 |
"205": "SUCCESS",
|
| 125 |
"206": "NON_PARTICIPATING",
|
| 126 |
"207": "NON_PARTICIPATING",
|
| 127 |
+
"208": "NON_PARTICIPATING",
|
| 128 |
"209": "NON_PARTICIPATING",
|
| 129 |
"21": "SUCCESS",
|
| 130 |
"210": "NON_PARTICIPATING",
|
|
|
|
| 133 |
"213": "NON_PARTICIPATING",
|
| 134 |
"214": "NON_PARTICIPATING",
|
| 135 |
"215": "NON_PARTICIPATING",
|
| 136 |
+
"216": "SUCCESS",
|
| 137 |
+
"217": "SUCCESS",
|
| 138 |
"218": "NON_PARTICIPATING",
|
| 139 |
+
"219": "SUCCESS",
|
| 140 |
"22": "SUCCESS",
|
| 141 |
"220": "NON_PARTICIPATING",
|
| 142 |
"221": "NON_PARTICIPATING",
|
|
|
|
| 147 |
"226": "NON_PARTICIPATING",
|
| 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",
|
| 155 |
"233": "NON_PARTICIPATING",
|
| 156 |
"234": "NON_PARTICIPATING",
|
|
|
|
| 164 |
"241": "NON_PARTICIPATING",
|
| 165 |
"242": "NON_PARTICIPATING",
|
| 166 |
"243": "NON_PARTICIPATING",
|
| 167 |
+
"244": "NON_PARTICIPATING",
|
| 168 |
"245": "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",
|
| 178 |
"254": "NON_PARTICIPATING",
|
| 179 |
"255": "NON_PARTICIPATING",
|
| 180 |
"26": "SUCCESS",
|
| 181 |
"27": "SUCCESS",
|
| 182 |
+
"28": "NON_PARTICIPATING",
|
| 183 |
"29": "SUCCESS",
|
| 184 |
+
"3": "NON_PARTICIPATING",
|
| 185 |
+
"30": "NON_PARTICIPATING",
|
| 186 |
"31": "SUCCESS",
|
| 187 |
+
"32": "NON_PARTICIPATING",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "SUCCESS",
|
| 190 |
"35": "SUCCESS",
|
| 191 |
+
"36": "NON_PARTICIPATING",
|
| 192 |
+
"37": "NON_PARTICIPATING",
|
| 193 |
"38": "SUCCESS",
|
| 194 |
+
"39": "NON_PARTICIPATING",
|
| 195 |
"4": "SUCCESS",
|
| 196 |
"40": "SUCCESS",
|
| 197 |
"41": "SUCCESS",
|
| 198 |
+
"42": "NON_PARTICIPATING",
|
| 199 |
+
"43": "NON_PARTICIPATING",
|
| 200 |
"44": "SUCCESS",
|
| 201 |
"45": "SUCCESS",
|
| 202 |
"46": "NON_PARTICIPATING",
|
| 203 |
+
"47": "SUCCESS",
|
| 204 |
"48": "NON_PARTICIPATING",
|
| 205 |
"49": "SUCCESS",
|
| 206 |
+
"5": "NON_PARTICIPATING",
|
| 207 |
+
"50": "NON_PARTICIPATING",
|
| 208 |
"51": "SUCCESS",
|
| 209 |
"52": "SUCCESS",
|
| 210 |
+
"53": "NON_PARTICIPATING",
|
| 211 |
"54": "SUCCESS",
|
| 212 |
+
"55": "NON_PARTICIPATING",
|
| 213 |
"56": "SUCCESS",
|
| 214 |
+
"57": "NON_PARTICIPATING",
|
| 215 |
"58": "NON_PARTICIPATING",
|
| 216 |
"59": "SUCCESS",
|
| 217 |
"6": "NON_PARTICIPATING",
|
| 218 |
"60": "SUCCESS",
|
| 219 |
+
"61": "SUCCESS",
|
| 220 |
"62": "SUCCESS",
|
| 221 |
"63": "SUCCESS",
|
| 222 |
+
"64": "SUCCESS",
|
| 223 |
"65": "SUCCESS",
|
| 224 |
+
"66": "NON_PARTICIPATING",
|
| 225 |
"67": "SUCCESS",
|
| 226 |
+
"68": "NON_PARTICIPATING",
|
| 227 |
+
"69": "NON_PARTICIPATING",
|
| 228 |
"7": "SUCCESS",
|
| 229 |
+
"70": "SUCCESS",
|
| 230 |
+
"71": "SUCCESS",
|
| 231 |
"72": "SUCCESS",
|
| 232 |
+
"73": "SUCCESS",
|
| 233 |
"74": "SUCCESS",
|
| 234 |
"75": "SUCCESS",
|
| 235 |
+
"76": "NON_PARTICIPATING",
|
| 236 |
"77": "SUCCESS",
|
| 237 |
+
"78": "NON_PARTICIPATING",
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "SUCCESS",
|
| 240 |
"80": "SUCCESS",
|
| 241 |
+
"81": "NON_PARTICIPATING",
|
| 242 |
"82": "SUCCESS",
|
| 243 |
"83": "NON_PARTICIPATING",
|
| 244 |
"84": "SUCCESS",
|
| 245 |
"85": "SUCCESS",
|
| 246 |
+
"86": "NON_PARTICIPATING",
|
| 247 |
"87": "SUCCESS",
|
| 248 |
+
"88": "NON_PARTICIPATING",
|
| 249 |
"89": "SUCCESS",
|
| 250 |
+
"9": "SUCCESS",
|
| 251 |
"90": "SUCCESS",
|
| 252 |
"91": "SUCCESS",
|
| 253 |
"92": "SUCCESS",
|
| 254 |
"93": "NON_PARTICIPATING",
|
| 255 |
+
"94": "NON_PARTICIPATING",
|
| 256 |
"95": "SUCCESS",
|
| 257 |
+
"96": "NON_PARTICIPATING",
|
| 258 |
"97": "SUCCESS",
|
| 259 |
"98": "SUCCESS",
|
| 260 |
"99": "SUCCESS"
|
|
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
+
5687103,
|
| 272 |
+
5687134
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 273 |
],
|
| 274 |
"block_size": 1024,
|
| 275 |
"bos_token_id": 50256,
|
| 276 |
"embd_pdrop": 0.1,
|
| 277 |
"eos_token_id": 50256,
|
| 278 |
"initializer_range": 0.02,
|
| 279 |
+
"inner_step": 2,
|
| 280 |
"inner_steps": 0,
|
| 281 |
+
"last_allreduce_block": 5686307,
|
| 282 |
"layer_norm_epsilon": 1e-05,
|
| 283 |
"model_type": "gpt_optimized",
|
| 284 |
"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
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:357a9dfeb07d3336137ade9f60b73966438a06e629abc696027d3f824a8c97a6
|
| 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 sha256:6777ba28e6d2780055e535cc0d1453e7c0cd61cd8abcb338cbb798c440d27153
|
| 3 |
size 4040701744
|