Run 4. Outer Step 53. Inner Step 0.
Browse files- config.json +43 -43
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
config.json
CHANGED
|
@@ -14,12 +14,12 @@
|
|
| 14 |
"106": "NON_PARTICIPATING",
|
| 15 |
"107": "SUCCESS",
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
-
"109": "
|
| 18 |
"11": "SUCCESS",
|
| 19 |
"110": "NON_PARTICIPATING",
|
| 20 |
-
"111": "
|
| 21 |
"112": "NON_PARTICIPATING",
|
| 22 |
-
"113": "
|
| 23 |
"114": "NON_PARTICIPATING",
|
| 24 |
"115": "NON_PARTICIPATING",
|
| 25 |
"116": "NON_PARTICIPATING",
|
|
@@ -31,22 +31,22 @@
|
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
"122": "NON_PARTICIPATING",
|
| 33 |
"123": "NON_PARTICIPATING",
|
| 34 |
-
"124": "
|
| 35 |
"125": "NON_PARTICIPATING",
|
| 36 |
"126": "SUCCESS",
|
| 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": "
|
| 48 |
"137": "NON_PARTICIPATING",
|
| 49 |
-
"138": "
|
| 50 |
"139": "NON_PARTICIPATING",
|
| 51 |
"14": "SUCCESS",
|
| 52 |
"140": "NON_PARTICIPATING",
|
|
@@ -57,37 +57,37 @@
|
|
| 57 |
"145": "NON_PARTICIPATING",
|
| 58 |
"146": "NON_PARTICIPATING",
|
| 59 |
"147": "NON_PARTICIPATING",
|
| 60 |
-
"148": "
|
| 61 |
"149": "NON_PARTICIPATING",
|
| 62 |
-
"15": "
|
| 63 |
"150": "NON_PARTICIPATING",
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
"153": "NON_PARTICIPATING",
|
| 67 |
-
"154": "
|
| 68 |
"155": "NON_PARTICIPATING",
|
| 69 |
"156": "NON_PARTICIPATING",
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
-
"159": "
|
| 73 |
"16": "SUCCESS",
|
| 74 |
"160": "NON_PARTICIPATING",
|
| 75 |
"161": "NON_PARTICIPATING",
|
| 76 |
-
"162": "
|
| 77 |
"163": "NON_PARTICIPATING",
|
| 78 |
-
"164": "
|
| 79 |
"165": "SUCCESS",
|
| 80 |
"166": "NON_PARTICIPATING",
|
| 81 |
"167": "NON_PARTICIPATING",
|
| 82 |
"168": "NON_PARTICIPATING",
|
| 83 |
-
"169": "
|
| 84 |
"17": "SUCCESS",
|
| 85 |
"170": "NON_PARTICIPATING",
|
| 86 |
"171": "NON_PARTICIPATING",
|
| 87 |
"172": "NON_PARTICIPATING",
|
| 88 |
"173": "NON_PARTICIPATING",
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
-
"175": "
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
"177": "NON_PARTICIPATING",
|
| 93 |
"178": "NON_PARTICIPATING",
|
|
@@ -98,14 +98,14 @@
|
|
| 98 |
"182": "NON_PARTICIPATING",
|
| 99 |
"183": "NON_PARTICIPATING",
|
| 100 |
"184": "NON_PARTICIPATING",
|
| 101 |
-
"185": "
|
| 102 |
"186": "NON_PARTICIPATING",
|
| 103 |
-
"187": "
|
| 104 |
"188": "NON_PARTICIPATING",
|
| 105 |
"189": "NON_PARTICIPATING",
|
| 106 |
"19": "SUCCESS",
|
| 107 |
"190": "NON_PARTICIPATING",
|
| 108 |
-
"191": "
|
| 109 |
"192": "NON_PARTICIPATING",
|
| 110 |
"193": "NON_PARTICIPATING",
|
| 111 |
"194": "NON_PARTICIPATING",
|
|
@@ -132,7 +132,7 @@
|
|
| 132 |
"212": "NON_PARTICIPATING",
|
| 133 |
"213": "NON_PARTICIPATING",
|
| 134 |
"214": "NON_PARTICIPATING",
|
| 135 |
-
"215": "
|
| 136 |
"216": "NON_PARTICIPATING",
|
| 137 |
"217": "NON_PARTICIPATING",
|
| 138 |
"218": "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",
|
|
@@ -149,29 +149,29 @@
|
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
"229": "SUCCESS",
|
| 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": "NON_PARTICIPATING",
|
| 159 |
"237": "NON_PARTICIPATING",
|
| 160 |
-
"238": "
|
| 161 |
-
"239": "
|
| 162 |
"24": "SUCCESS",
|
| 163 |
-
"240": "
|
| 164 |
-
"241": "
|
| 165 |
"242": "NON_PARTICIPATING",
|
| 166 |
-
"243": "
|
| 167 |
"244": "NON_PARTICIPATING",
|
| 168 |
-
"245": "
|
| 169 |
"246": "NON_PARTICIPATING",
|
| 170 |
"247": "NON_PARTICIPATING",
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
-
"249": "
|
| 173 |
"25": "SUCCESS",
|
| 174 |
-
"250": "
|
| 175 |
"251": "NON_PARTICIPATING",
|
| 176 |
"252": "NON_PARTICIPATING",
|
| 177 |
"253": "NON_PARTICIPATING",
|
|
@@ -182,12 +182,12 @@
|
|
| 182 |
"28": "SUCCESS",
|
| 183 |
"29": "SUCCESS",
|
| 184 |
"3": "SUCCESS",
|
| 185 |
-
"30": "
|
| 186 |
"31": "NON_PARTICIPATING",
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "SUCCESS",
|
| 190 |
-
"35": "
|
| 191 |
"36": "SUCCESS",
|
| 192 |
"37": "SUCCESS",
|
| 193 |
"38": "SUCCESS",
|
|
@@ -200,7 +200,7 @@
|
|
| 200 |
"44": "SUCCESS",
|
| 201 |
"45": "SUCCESS",
|
| 202 |
"46": "NON_PARTICIPATING",
|
| 203 |
-
"47": "
|
| 204 |
"48": "NON_PARTICIPATING",
|
| 205 |
"49": "SUCCESS",
|
| 206 |
"5": "SUCCESS",
|
|
@@ -225,7 +225,7 @@
|
|
| 225 |
"67": "SUCCESS",
|
| 226 |
"68": "SUCCESS",
|
| 227 |
"69": "SUCCESS",
|
| 228 |
-
"7": "
|
| 229 |
"70": "NON_PARTICIPATING",
|
| 230 |
"71": "NON_PARTICIPATING",
|
| 231 |
"72": "SUCCESS",
|
|
@@ -247,8 +247,8 @@
|
|
| 247 |
"87": "SUCCESS",
|
| 248 |
"88": "SUCCESS",
|
| 249 |
"89": "SUCCESS",
|
| 250 |
-
"9": "
|
| 251 |
-
"90": "
|
| 252 |
"91": "SUCCESS",
|
| 253 |
"92": "SUCCESS",
|
| 254 |
"93": "NON_PARTICIPATING",
|
|
@@ -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,
|
|
|
|
| 14 |
"106": "NON_PARTICIPATING",
|
| 15 |
"107": "SUCCESS",
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
+
"109": "SUCCESS",
|
| 18 |
"11": "SUCCESS",
|
| 19 |
"110": "NON_PARTICIPATING",
|
| 20 |
+
"111": "NON_PARTICIPATING",
|
| 21 |
"112": "NON_PARTICIPATING",
|
| 22 |
+
"113": "SUCCESS",
|
| 23 |
"114": "NON_PARTICIPATING",
|
| 24 |
"115": "NON_PARTICIPATING",
|
| 25 |
"116": "NON_PARTICIPATING",
|
|
|
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
"122": "NON_PARTICIPATING",
|
| 33 |
"123": "NON_PARTICIPATING",
|
| 34 |
+
"124": "NON_PARTICIPATING",
|
| 35 |
"125": "NON_PARTICIPATING",
|
| 36 |
"126": "SUCCESS",
|
| 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",
|
|
|
|
| 57 |
"145": "NON_PARTICIPATING",
|
| 58 |
"146": "NON_PARTICIPATING",
|
| 59 |
"147": "NON_PARTICIPATING",
|
| 60 |
+
"148": "SUCCESS",
|
| 61 |
"149": "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",
|
| 69 |
"156": "NON_PARTICIPATING",
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
+
"159": "SUCCESS",
|
| 73 |
"16": "SUCCESS",
|
| 74 |
"160": "NON_PARTICIPATING",
|
| 75 |
"161": "NON_PARTICIPATING",
|
| 76 |
+
"162": "SUCCESS",
|
| 77 |
"163": "NON_PARTICIPATING",
|
| 78 |
+
"164": "SUCCESS",
|
| 79 |
"165": "SUCCESS",
|
| 80 |
"166": "NON_PARTICIPATING",
|
| 81 |
"167": "NON_PARTICIPATING",
|
| 82 |
"168": "NON_PARTICIPATING",
|
| 83 |
+
"169": "SUCCESS",
|
| 84 |
"17": "SUCCESS",
|
| 85 |
"170": "NON_PARTICIPATING",
|
| 86 |
"171": "NON_PARTICIPATING",
|
| 87 |
"172": "NON_PARTICIPATING",
|
| 88 |
"173": "NON_PARTICIPATING",
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
+
"175": "SUCCESS",
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
"177": "NON_PARTICIPATING",
|
| 93 |
"178": "NON_PARTICIPATING",
|
|
|
|
| 98 |
"182": "NON_PARTICIPATING",
|
| 99 |
"183": "NON_PARTICIPATING",
|
| 100 |
"184": "NON_PARTICIPATING",
|
| 101 |
+
"185": "SUCCESS",
|
| 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": "SUCCESS",
|
| 109 |
"192": "NON_PARTICIPATING",
|
| 110 |
"193": "NON_PARTICIPATING",
|
| 111 |
"194": "NON_PARTICIPATING",
|
|
|
|
| 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",
|
|
|
|
| 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",
|
|
|
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
"229": "SUCCESS",
|
| 151 |
"23": "SUCCESS",
|
| 152 |
+
"230": "SUCCESS",
|
| 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": "SUCCESS",
|
| 161 |
+
"239": "SUCCESS",
|
| 162 |
"24": "SUCCESS",
|
| 163 |
+
"240": "NON_PARTICIPATING",
|
| 164 |
+
"241": "NON_PARTICIPATING",
|
| 165 |
"242": "NON_PARTICIPATING",
|
| 166 |
+
"243": "NON_PARTICIPATING",
|
| 167 |
"244": "NON_PARTICIPATING",
|
| 168 |
+
"245": "SUCCESS",
|
| 169 |
"246": "NON_PARTICIPATING",
|
| 170 |
"247": "NON_PARTICIPATING",
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
+
"249": "NON_PARTICIPATING",
|
| 173 |
"25": "SUCCESS",
|
| 174 |
+
"250": "SUCCESS",
|
| 175 |
"251": "NON_PARTICIPATING",
|
| 176 |
"252": "NON_PARTICIPATING",
|
| 177 |
"253": "NON_PARTICIPATING",
|
|
|
|
| 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": "SUCCESS",
|
| 190 |
+
"35": "NON_PARTICIPATING",
|
| 191 |
"36": "SUCCESS",
|
| 192 |
"37": "SUCCESS",
|
| 193 |
"38": "SUCCESS",
|
|
|
|
| 200 |
"44": "SUCCESS",
|
| 201 |
"45": "SUCCESS",
|
| 202 |
"46": "NON_PARTICIPATING",
|
| 203 |
+
"47": "SUCCESS",
|
| 204 |
"48": "NON_PARTICIPATING",
|
| 205 |
"49": "SUCCESS",
|
| 206 |
"5": "SUCCESS",
|
|
|
|
| 225 |
"67": "SUCCESS",
|
| 226 |
"68": "SUCCESS",
|
| 227 |
"69": "SUCCESS",
|
| 228 |
+
"7": "NON_PARTICIPATING",
|
| 229 |
"70": "NON_PARTICIPATING",
|
| 230 |
"71": "NON_PARTICIPATING",
|
| 231 |
"72": "SUCCESS",
|
|
|
|
| 247 |
"87": "SUCCESS",
|
| 248 |
"88": "SUCCESS",
|
| 249 |
"89": "SUCCESS",
|
| 250 |
+
"9": "NON_PARTICIPATING",
|
| 251 |
+
"90": "NON_PARTICIPATING",
|
| 252 |
"91": "SUCCESS",
|
| 253 |
"92": "SUCCESS",
|
| 254 |
"93": "NON_PARTICIPATING",
|
|
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
+
5587040,
|
| 272 |
+
5587068,
|
| 273 |
+
5587073,
|
| 274 |
+
5587077,
|
| 275 |
+
5587081
|
| 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": 5576204,
|
| 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 8081782503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7db7c3dedaa0442d85b6ddf661cbbd2d2f2fb1636fd3de18319394bc16a63fdf
|
| 3 |
size 8081782503
|
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 sha256:9ba35fe72ed12b54c4e91ff5e916450689614454f76f3fc50436b55c099335a8
|
| 3 |
size 4040701744
|