Run 4. Outer Step 64. Inner Step 0.
Browse files- config.json +47 -47
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
config.json
CHANGED
|
@@ -6,39 +6,39 @@
|
|
| 6 |
"1": "SUCCESS",
|
| 7 |
"10": "NON_PARTICIPATING",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
-
"101": "
|
| 10 |
"102": "NON_PARTICIPATING",
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
-
"104": "
|
| 13 |
"105": "NON_PARTICIPATING",
|
| 14 |
-
"106": "
|
| 15 |
-
"107": "
|
| 16 |
-
"108": "
|
| 17 |
"109": "NON_PARTICIPATING",
|
| 18 |
"11": "SUCCESS",
|
| 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": "
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
"12": "SUCCESS",
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
-
"122": "
|
| 33 |
"123": "NON_PARTICIPATING",
|
| 34 |
-
"124": "
|
| 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": "
|
| 42 |
"131": "NON_PARTICIPATING",
|
| 43 |
"132": "NON_PARTICIPATING",
|
| 44 |
"133": "SUCCESS",
|
|
@@ -56,7 +56,7 @@
|
|
| 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",
|
|
@@ -87,24 +87,24 @@
|
|
| 87 |
"172": "NON_PARTICIPATING",
|
| 88 |
"173": "NON_PARTICIPATING",
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
-
"175": "
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
-
"177": "
|
| 93 |
-
"178": "
|
| 94 |
"179": "NON_PARTICIPATING",
|
| 95 |
"18": "SUCCESS",
|
| 96 |
"180": "NON_PARTICIPATING",
|
| 97 |
-
"181": "
|
| 98 |
-
"182": "
|
| 99 |
"183": "NON_PARTICIPATING",
|
| 100 |
"184": "NON_PARTICIPATING",
|
| 101 |
"185": "NON_PARTICIPATING",
|
| 102 |
-
"186": "
|
| 103 |
"187": "NON_PARTICIPATING",
|
| 104 |
"188": "NON_PARTICIPATING",
|
| 105 |
"189": "NON_PARTICIPATING",
|
| 106 |
"19": "SUCCESS",
|
| 107 |
-
"190": "
|
| 108 |
"191": "NON_PARTICIPATING",
|
| 109 |
"192": "NON_PARTICIPATING",
|
| 110 |
"193": "NON_PARTICIPATING",
|
|
@@ -113,10 +113,10 @@
|
|
| 113 |
"196": "NON_PARTICIPATING",
|
| 114 |
"197": "NON_PARTICIPATING",
|
| 115 |
"198": "NON_PARTICIPATING",
|
| 116 |
-
"199": "
|
| 117 |
"2": "SUCCESS",
|
| 118 |
"20": "SUCCESS",
|
| 119 |
-
"200": "
|
| 120 |
"201": "NON_PARTICIPATING",
|
| 121 |
"202": "NON_PARTICIPATING",
|
| 122 |
"203": "NON_PARTICIPATING",
|
|
@@ -149,27 +149,27 @@
|
|
| 149 |
"228": "SUCCESS",
|
| 150 |
"229": "NON_PARTICIPATING",
|
| 151 |
"23": "SUCCESS",
|
| 152 |
-
"230": "
|
| 153 |
"231": "NON_PARTICIPATING",
|
| 154 |
-
"232": "
|
| 155 |
"233": "NON_PARTICIPATING",
|
| 156 |
"234": "NON_PARTICIPATING",
|
| 157 |
"235": "NON_PARTICIPATING",
|
| 158 |
-
"236": "
|
| 159 |
"237": "NON_PARTICIPATING",
|
| 160 |
"238": "NON_PARTICIPATING",
|
| 161 |
-
"239": "
|
| 162 |
"24": "SUCCESS",
|
| 163 |
"240": "NON_PARTICIPATING",
|
| 164 |
-
"241": "
|
| 165 |
"242": "NON_PARTICIPATING",
|
| 166 |
"243": "NON_PARTICIPATING",
|
| 167 |
"244": "NON_PARTICIPATING",
|
| 168 |
"245": "NON_PARTICIPATING",
|
| 169 |
-
"246": "
|
| 170 |
"247": "NON_PARTICIPATING",
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
-
"249": "
|
| 173 |
"25": "SUCCESS",
|
| 174 |
"250": "NON_PARTICIPATING",
|
| 175 |
"251": "NON_PARTICIPATING",
|
|
@@ -181,12 +181,12 @@
|
|
| 181 |
"27": "SUCCESS",
|
| 182 |
"28": "SUCCESS",
|
| 183 |
"29": "SUCCESS",
|
| 184 |
-
"3": "
|
| 185 |
"30": "NON_PARTICIPATING",
|
| 186 |
"31": "NON_PARTICIPATING",
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
-
"34": "
|
| 190 |
"35": "SUCCESS",
|
| 191 |
"36": "SUCCESS",
|
| 192 |
"37": "SUCCESS",
|
|
@@ -205,13 +205,13 @@
|
|
| 205 |
"49": "SUCCESS",
|
| 206 |
"5": "SUCCESS",
|
| 207 |
"50": "SUCCESS",
|
| 208 |
-
"51": "
|
| 209 |
"52": "NON_PARTICIPATING",
|
| 210 |
"53": "SUCCESS",
|
| 211 |
"54": "SUCCESS",
|
| 212 |
"55": "SUCCESS",
|
| 213 |
"56": "SUCCESS",
|
| 214 |
-
"57": "
|
| 215 |
"58": "NON_PARTICIPATING",
|
| 216 |
"59": "SUCCESS",
|
| 217 |
"6": "NON_PARTICIPATING",
|
|
@@ -219,35 +219,35 @@
|
|
| 219 |
"61": "NON_PARTICIPATING",
|
| 220 |
"62": "SUCCESS",
|
| 221 |
"63": "SUCCESS",
|
| 222 |
-
"64": "
|
| 223 |
"65": "SUCCESS",
|
| 224 |
"66": "SUCCESS",
|
| 225 |
"67": "SUCCESS",
|
| 226 |
"68": "SUCCESS",
|
| 227 |
"69": "SUCCESS",
|
| 228 |
"7": "SUCCESS",
|
| 229 |
-
"70": "
|
| 230 |
"71": "NON_PARTICIPATING",
|
| 231 |
"72": "SUCCESS",
|
| 232 |
"73": "NON_PARTICIPATING",
|
| 233 |
"74": "SUCCESS",
|
| 234 |
-
"75": "
|
| 235 |
-
"76": "
|
| 236 |
"77": "SUCCESS",
|
| 237 |
-
"78": "
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "SUCCESS",
|
| 240 |
"80": "SUCCESS",
|
| 241 |
"81": "SUCCESS",
|
| 242 |
"82": "SUCCESS",
|
| 243 |
-
"83": "
|
| 244 |
"84": "SUCCESS",
|
| 245 |
"85": "NON_PARTICIPATING",
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
-
"87": "
|
| 248 |
"88": "SUCCESS",
|
| 249 |
"89": "SUCCESS",
|
| 250 |
-
"9": "
|
| 251 |
"90": "SUCCESS",
|
| 252 |
"91": "SUCCESS",
|
| 253 |
"92": "SUCCESS",
|
|
@@ -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,
|
|
|
|
| 6 |
"1": "SUCCESS",
|
| 7 |
"10": "NON_PARTICIPATING",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
+
"101": "NON_PARTICIPATING",
|
| 10 |
"102": "NON_PARTICIPATING",
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
+
"104": "SUCCESS",
|
| 13 |
"105": "NON_PARTICIPATING",
|
| 14 |
+
"106": "SUCCESS",
|
| 15 |
+
"107": "NON_PARTICIPATING",
|
| 16 |
+
"108": "NON_PARTICIPATING",
|
| 17 |
"109": "NON_PARTICIPATING",
|
| 18 |
"11": "SUCCESS",
|
| 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": "SUCCESS",
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
+
"122": "NON_PARTICIPATING",
|
| 33 |
"123": "NON_PARTICIPATING",
|
| 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": "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",
|
|
|
|
| 87 |
"172": "NON_PARTICIPATING",
|
| 88 |
"173": "NON_PARTICIPATING",
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
+
"175": "SUCCESS",
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
+
"177": "SUCCESS",
|
| 93 |
+
"178": "NON_PARTICIPATING",
|
| 94 |
"179": "NON_PARTICIPATING",
|
| 95 |
"18": "SUCCESS",
|
| 96 |
"180": "NON_PARTICIPATING",
|
| 97 |
+
"181": "NON_PARTICIPATING",
|
| 98 |
+
"182": "SUCCESS",
|
| 99 |
"183": "NON_PARTICIPATING",
|
| 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": "NON_PARTICIPATING",
|
| 110 |
"193": "NON_PARTICIPATING",
|
|
|
|
| 113 |
"196": "NON_PARTICIPATING",
|
| 114 |
"197": "NON_PARTICIPATING",
|
| 115 |
"198": "NON_PARTICIPATING",
|
| 116 |
+
"199": "SUCCESS",
|
| 117 |
"2": "SUCCESS",
|
| 118 |
"20": "SUCCESS",
|
| 119 |
+
"200": "NON_PARTICIPATING",
|
| 120 |
"201": "NON_PARTICIPATING",
|
| 121 |
"202": "NON_PARTICIPATING",
|
| 122 |
"203": "NON_PARTICIPATING",
|
|
|
|
| 149 |
"228": "SUCCESS",
|
| 150 |
"229": "NON_PARTICIPATING",
|
| 151 |
"23": "SUCCESS",
|
| 152 |
+
"230": "NON_PARTICIPATING",
|
| 153 |
"231": "NON_PARTICIPATING",
|
| 154 |
+
"232": "SUCCESS",
|
| 155 |
"233": "NON_PARTICIPATING",
|
| 156 |
"234": "NON_PARTICIPATING",
|
| 157 |
"235": "NON_PARTICIPATING",
|
| 158 |
+
"236": "NON_PARTICIPATING",
|
| 159 |
"237": "NON_PARTICIPATING",
|
| 160 |
"238": "NON_PARTICIPATING",
|
| 161 |
+
"239": "SUCCESS",
|
| 162 |
"24": "SUCCESS",
|
| 163 |
"240": "NON_PARTICIPATING",
|
| 164 |
+
"241": "SUCCESS",
|
| 165 |
"242": "NON_PARTICIPATING",
|
| 166 |
"243": "NON_PARTICIPATING",
|
| 167 |
"244": "NON_PARTICIPATING",
|
| 168 |
"245": "NON_PARTICIPATING",
|
| 169 |
+
"246": "SUCCESS",
|
| 170 |
"247": "NON_PARTICIPATING",
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
+
"249": "SUCCESS",
|
| 173 |
"25": "SUCCESS",
|
| 174 |
"250": "NON_PARTICIPATING",
|
| 175 |
"251": "NON_PARTICIPATING",
|
|
|
|
| 181 |
"27": "SUCCESS",
|
| 182 |
"28": "SUCCESS",
|
| 183 |
"29": "SUCCESS",
|
| 184 |
+
"3": "SUCCESS",
|
| 185 |
"30": "NON_PARTICIPATING",
|
| 186 |
"31": "NON_PARTICIPATING",
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
+
"34": "NON_PARTICIPATING",
|
| 190 |
"35": "SUCCESS",
|
| 191 |
"36": "SUCCESS",
|
| 192 |
"37": "SUCCESS",
|
|
|
|
| 205 |
"49": "SUCCESS",
|
| 206 |
"5": "SUCCESS",
|
| 207 |
"50": "SUCCESS",
|
| 208 |
+
"51": "FAIL",
|
| 209 |
"52": "NON_PARTICIPATING",
|
| 210 |
"53": "SUCCESS",
|
| 211 |
"54": "SUCCESS",
|
| 212 |
"55": "SUCCESS",
|
| 213 |
"56": "SUCCESS",
|
| 214 |
+
"57": "NON_PARTICIPATING",
|
| 215 |
"58": "NON_PARTICIPATING",
|
| 216 |
"59": "SUCCESS",
|
| 217 |
"6": "NON_PARTICIPATING",
|
|
|
|
| 219 |
"61": "NON_PARTICIPATING",
|
| 220 |
"62": "SUCCESS",
|
| 221 |
"63": "SUCCESS",
|
| 222 |
+
"64": "SUCCESS",
|
| 223 |
"65": "SUCCESS",
|
| 224 |
"66": "SUCCESS",
|
| 225 |
"67": "SUCCESS",
|
| 226 |
"68": "SUCCESS",
|
| 227 |
"69": "SUCCESS",
|
| 228 |
"7": "SUCCESS",
|
| 229 |
+
"70": "SUCCESS",
|
| 230 |
"71": "NON_PARTICIPATING",
|
| 231 |
"72": "SUCCESS",
|
| 232 |
"73": "NON_PARTICIPATING",
|
| 233 |
"74": "SUCCESS",
|
| 234 |
+
"75": "SUCCESS",
|
| 235 |
+
"76": "NON_PARTICIPATING",
|
| 236 |
"77": "SUCCESS",
|
| 237 |
+
"78": "SUCCESS",
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "SUCCESS",
|
| 240 |
"80": "SUCCESS",
|
| 241 |
"81": "SUCCESS",
|
| 242 |
"82": "SUCCESS",
|
| 243 |
+
"83": "SUCCESS",
|
| 244 |
"84": "SUCCESS",
|
| 245 |
"85": "NON_PARTICIPATING",
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
+
"87": "SUCCESS",
|
| 248 |
"88": "SUCCESS",
|
| 249 |
"89": "SUCCESS",
|
| 250 |
+
"9": "SUCCESS",
|
| 251 |
"90": "SUCCESS",
|
| 252 |
"91": "SUCCESS",
|
| 253 |
"92": "SUCCESS",
|
|
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
+
5615530,
|
| 272 |
+
5615535,
|
| 273 |
+
5615539,
|
| 274 |
+
5615544,
|
| 275 |
+
5615548
|
| 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": 5613899,
|
| 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:ebc41618cd9a1ff2fc7df56831f0863ea0bb88d53a3c679cb1f6cc3f3999d75d
|
| 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 sha256:29f25e4e1f5ad183d22736c38a618157c09202e62e790401715d03e9b9392f5f
|
| 3 |
size 4040701744
|