Run 4. Outer Step 98. Inner Step 2.
Browse files- config.json +31 -31
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
config.json
CHANGED
|
@@ -6,7 +6,7 @@
|
|
| 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",
|
|
@@ -25,8 +25,8 @@
|
|
| 25 |
"116": "NON_PARTICIPATING",
|
| 26 |
"117": "NON_PARTICIPATING",
|
| 27 |
"118": "NON_PARTICIPATING",
|
| 28 |
-
"119": "
|
| 29 |
-
"12": "
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
"122": "NON_PARTICIPATING",
|
|
@@ -37,16 +37,16 @@
|
|
| 37 |
"127": "NON_PARTICIPATING",
|
| 38 |
"128": "NON_PARTICIPATING",
|
| 39 |
"129": "NON_PARTICIPATING",
|
| 40 |
-
"13": "
|
| 41 |
"130": "NON_PARTICIPATING",
|
| 42 |
"131": "NON_PARTICIPATING",
|
| 43 |
"132": "NON_PARTICIPATING",
|
| 44 |
"133": "NON_PARTICIPATING",
|
| 45 |
-
"134": "
|
| 46 |
"135": "NON_PARTICIPATING",
|
| 47 |
"136": "NON_PARTICIPATING",
|
| 48 |
"137": "NON_PARTICIPATING",
|
| 49 |
-
"138": "
|
| 50 |
"139": "NON_PARTICIPATING",
|
| 51 |
"14": "SUCCESS",
|
| 52 |
"140": "NON_PARTICIPATING",
|
|
@@ -62,7 +62,7 @@
|
|
| 62 |
"15": "NON_PARTICIPATING",
|
| 63 |
"150": "NON_PARTICIPATING",
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
-
"152": "
|
| 66 |
"153": "NON_PARTICIPATING",
|
| 67 |
"154": "NON_PARTICIPATING",
|
| 68 |
"155": "NON_PARTICIPATING",
|
|
@@ -78,20 +78,20 @@
|
|
| 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": "NON_PARTICIPATING",
|
| 86 |
"171": "NON_PARTICIPATING",
|
| 87 |
"172": "NON_PARTICIPATING",
|
| 88 |
-
"173": "
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
"175": "NON_PARTICIPATING",
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
"177": "NON_PARTICIPATING",
|
| 93 |
"178": "NON_PARTICIPATING",
|
| 94 |
-
"179": "
|
| 95 |
"18": "SUCCESS",
|
| 96 |
"180": "NON_PARTICIPATING",
|
| 97 |
"181": "NON_PARTICIPATING",
|
|
@@ -116,18 +116,18 @@
|
|
| 116 |
"199": "NON_PARTICIPATING",
|
| 117 |
"2": "SUCCESS",
|
| 118 |
"20": "NON_PARTICIPATING",
|
| 119 |
-
"200": "
|
| 120 |
"201": "NON_PARTICIPATING",
|
| 121 |
"202": "NON_PARTICIPATING",
|
| 122 |
"203": "NON_PARTICIPATING",
|
| 123 |
"204": "NON_PARTICIPATING",
|
| 124 |
-
"205": "
|
| 125 |
"206": "NON_PARTICIPATING",
|
| 126 |
"207": "NON_PARTICIPATING",
|
| 127 |
"208": "NON_PARTICIPATING",
|
| 128 |
"209": "NON_PARTICIPATING",
|
| 129 |
"21": "SUCCESS",
|
| 130 |
-
"210": "
|
| 131 |
"211": "NON_PARTICIPATING",
|
| 132 |
"212": "NON_PARTICIPATING",
|
| 133 |
"213": "NON_PARTICIPATING",
|
|
@@ -141,7 +141,7 @@
|
|
| 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",
|
|
@@ -151,15 +151,15 @@
|
|
| 151 |
"23": "SUCCESS",
|
| 152 |
"230": "NON_PARTICIPATING",
|
| 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": "NON_PARTICIPATING",
|
| 162 |
-
"24": "
|
| 163 |
"240": "NON_PARTICIPATING",
|
| 164 |
"241": "NON_PARTICIPATING",
|
| 165 |
"242": "NON_PARTICIPATING",
|
|
@@ -168,7 +168,7 @@
|
|
| 168 |
"245": "NON_PARTICIPATING",
|
| 169 |
"246": "NON_PARTICIPATING",
|
| 170 |
"247": "NON_PARTICIPATING",
|
| 171 |
-
"248": "
|
| 172 |
"249": "NON_PARTICIPATING",
|
| 173 |
"25": "SUCCESS",
|
| 174 |
"250": "NON_PARTICIPATING",
|
|
@@ -178,8 +178,8 @@
|
|
| 178 |
"254": "NON_PARTICIPATING",
|
| 179 |
"255": "NON_PARTICIPATING",
|
| 180 |
"26": "NON_PARTICIPATING",
|
| 181 |
-
"27": "
|
| 182 |
-
"28": "
|
| 183 |
"29": "SUCCESS",
|
| 184 |
"3": "SUCCESS",
|
| 185 |
"30": "NON_PARTICIPATING",
|
|
@@ -187,7 +187,7 @@
|
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "NON_PARTICIPATING",
|
| 190 |
-
"35": "
|
| 191 |
"36": "NON_PARTICIPATING",
|
| 192 |
"37": "SUCCESS",
|
| 193 |
"38": "NON_PARTICIPATING",
|
|
@@ -205,7 +205,7 @@
|
|
| 205 |
"49": "NON_PARTICIPATING",
|
| 206 |
"5": "NON_PARTICIPATING",
|
| 207 |
"50": "SUCCESS",
|
| 208 |
-
"51": "
|
| 209 |
"52": "NON_PARTICIPATING",
|
| 210 |
"53": "SUCCESS",
|
| 211 |
"54": "SUCCESS",
|
|
@@ -215,10 +215,10 @@
|
|
| 215 |
"58": "NON_PARTICIPATING",
|
| 216 |
"59": "NON_PARTICIPATING",
|
| 217 |
"6": "SUCCESS",
|
| 218 |
-
"60": "
|
| 219 |
"61": "SUCCESS",
|
| 220 |
"62": "SUCCESS",
|
| 221 |
-
"63": "
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
"65": "NON_PARTICIPATING",
|
| 224 |
"66": "NON_PARTICIPATING",
|
|
@@ -229,7 +229,7 @@
|
|
| 229 |
"70": "NON_PARTICIPATING",
|
| 230 |
"71": "NON_PARTICIPATING",
|
| 231 |
"72": "NON_PARTICIPATING",
|
| 232 |
-
"73": "
|
| 233 |
"74": "NON_PARTICIPATING",
|
| 234 |
"75": "NON_PARTICIPATING",
|
| 235 |
"76": "SUCCESS",
|
|
@@ -245,11 +245,11 @@
|
|
| 245 |
"85": "NON_PARTICIPATING",
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
"87": "SUCCESS",
|
| 248 |
-
"88": "
|
| 249 |
"89": "SUCCESS",
|
| 250 |
"9": "NON_PARTICIPATING",
|
| 251 |
"90": "SUCCESS",
|
| 252 |
-
"91": "
|
| 253 |
"92": "NON_PARTICIPATING",
|
| 254 |
"93": "SUCCESS",
|
| 255 |
"94": "SUCCESS",
|
|
@@ -268,17 +268,17 @@
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
-
|
| 272 |
-
|
| 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":
|
| 280 |
"inner_steps": 0,
|
| 281 |
-
"last_allreduce_block":
|
| 282 |
"layer_norm_epsilon": 1e-05,
|
| 283 |
"model_type": "gpt_optimized",
|
| 284 |
"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",
|
|
|
|
| 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",
|
|
|
|
| 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": "SUCCESS",
|
| 50 |
"139": "NON_PARTICIPATING",
|
| 51 |
"14": "SUCCESS",
|
| 52 |
"140": "NON_PARTICIPATING",
|
|
|
|
| 62 |
"15": "NON_PARTICIPATING",
|
| 63 |
"150": "NON_PARTICIPATING",
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
+
"152": "NON_PARTICIPATING",
|
| 66 |
"153": "NON_PARTICIPATING",
|
| 67 |
"154": "NON_PARTICIPATING",
|
| 68 |
"155": "NON_PARTICIPATING",
|
|
|
|
| 78 |
"164": "NON_PARTICIPATING",
|
| 79 |
"165": "NON_PARTICIPATING",
|
| 80 |
"166": "NON_PARTICIPATING",
|
| 81 |
+
"167": "SUCCESS",
|
| 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": "SUCCESS",
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
"175": "NON_PARTICIPATING",
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
"177": "NON_PARTICIPATING",
|
| 93 |
"178": "NON_PARTICIPATING",
|
| 94 |
+
"179": "SUCCESS",
|
| 95 |
"18": "SUCCESS",
|
| 96 |
"180": "NON_PARTICIPATING",
|
| 97 |
"181": "NON_PARTICIPATING",
|
|
|
|
| 116 |
"199": "NON_PARTICIPATING",
|
| 117 |
"2": "SUCCESS",
|
| 118 |
"20": "NON_PARTICIPATING",
|
| 119 |
+
"200": "SUCCESS",
|
| 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",
|
| 131 |
"211": "NON_PARTICIPATING",
|
| 132 |
"212": "NON_PARTICIPATING",
|
| 133 |
"213": "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",
|
|
|
|
| 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": "SUCCESS",
|
| 159 |
"237": "NON_PARTICIPATING",
|
| 160 |
"238": "NON_PARTICIPATING",
|
| 161 |
"239": "NON_PARTICIPATING",
|
| 162 |
+
"24": "SUCCESS",
|
| 163 |
"240": "NON_PARTICIPATING",
|
| 164 |
"241": "NON_PARTICIPATING",
|
| 165 |
"242": "NON_PARTICIPATING",
|
|
|
|
| 168 |
"245": "NON_PARTICIPATING",
|
| 169 |
"246": "NON_PARTICIPATING",
|
| 170 |
"247": "NON_PARTICIPATING",
|
| 171 |
+
"248": "SUCCESS",
|
| 172 |
"249": "NON_PARTICIPATING",
|
| 173 |
"25": "SUCCESS",
|
| 174 |
"250": "NON_PARTICIPATING",
|
|
|
|
| 178 |
"254": "NON_PARTICIPATING",
|
| 179 |
"255": "NON_PARTICIPATING",
|
| 180 |
"26": "NON_PARTICIPATING",
|
| 181 |
+
"27": "SUCCESS",
|
| 182 |
+
"28": "SUCCESS",
|
| 183 |
"29": "SUCCESS",
|
| 184 |
"3": "SUCCESS",
|
| 185 |
"30": "NON_PARTICIPATING",
|
|
|
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "NON_PARTICIPATING",
|
| 190 |
+
"35": "SUCCESS",
|
| 191 |
"36": "NON_PARTICIPATING",
|
| 192 |
"37": "SUCCESS",
|
| 193 |
"38": "NON_PARTICIPATING",
|
|
|
|
| 205 |
"49": "NON_PARTICIPATING",
|
| 206 |
"5": "NON_PARTICIPATING",
|
| 207 |
"50": "SUCCESS",
|
| 208 |
+
"51": "SUCCESS",
|
| 209 |
"52": "NON_PARTICIPATING",
|
| 210 |
"53": "SUCCESS",
|
| 211 |
"54": "SUCCESS",
|
|
|
|
| 215 |
"58": "NON_PARTICIPATING",
|
| 216 |
"59": "NON_PARTICIPATING",
|
| 217 |
"6": "SUCCESS",
|
| 218 |
+
"60": "SUCCESS",
|
| 219 |
"61": "SUCCESS",
|
| 220 |
"62": "SUCCESS",
|
| 221 |
+
"63": "SUCCESS",
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
"65": "NON_PARTICIPATING",
|
| 224 |
"66": "NON_PARTICIPATING",
|
|
|
|
| 229 |
"70": "NON_PARTICIPATING",
|
| 230 |
"71": "NON_PARTICIPATING",
|
| 231 |
"72": "NON_PARTICIPATING",
|
| 232 |
+
"73": "SUCCESS",
|
| 233 |
"74": "NON_PARTICIPATING",
|
| 234 |
"75": "NON_PARTICIPATING",
|
| 235 |
"76": "SUCCESS",
|
|
|
|
| 245 |
"85": "NON_PARTICIPATING",
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
"87": "SUCCESS",
|
| 248 |
+
"88": "SUCCESS",
|
| 249 |
"89": "SUCCESS",
|
| 250 |
"9": "NON_PARTICIPATING",
|
| 251 |
"90": "SUCCESS",
|
| 252 |
+
"91": "SUCCESS",
|
| 253 |
"92": "NON_PARTICIPATING",
|
| 254 |
"93": "SUCCESS",
|
| 255 |
"94": "SUCCESS",
|
|
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
+
5693329,
|
| 272 |
+
5693364
|
| 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": 5692898,
|
| 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 8081782026
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4436b5503bbc0b6001659d2ffc5ace9b1a6341fc123a441a44554c5c8d2df2a
|
| 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 sha256:904d4c812f65da91a579b73ee3a9adf886eb76fd0ce404dfc1302c1906afe9cc
|
| 3 |
size 4040701744
|