Run 4. Outer Step 26. Inner Step 0.
Browse files- config.json +46 -46
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
config.json
CHANGED
|
@@ -6,7 +6,7 @@
|
|
| 6 |
"1": "SUCCESS",
|
| 7 |
"10": "SUCCESS",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
-
"101": "
|
| 10 |
"102": "NON_PARTICIPATING",
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
"104": "NON_PARTICIPATING",
|
|
@@ -16,8 +16,8 @@
|
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
"109": "NON_PARTICIPATING",
|
| 18 |
"11": "NON_PARTICIPATING",
|
| 19 |
-
"110": "
|
| 20 |
-
"111": "
|
| 21 |
"112": "NON_PARTICIPATING",
|
| 22 |
"113": "SUCCESS",
|
| 23 |
"114": "NON_PARTICIPATING",
|
|
@@ -27,8 +27,8 @@
|
|
| 27 |
"118": "NON_PARTICIPATING",
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
"12": "SUCCESS",
|
| 30 |
-
"120": "
|
| 31 |
-
"121": "
|
| 32 |
"122": "SUCCESS",
|
| 33 |
"123": "NON_PARTICIPATING",
|
| 34 |
"124": "NON_PARTICIPATING",
|
|
@@ -44,9 +44,9 @@
|
|
| 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": "NON_PARTICIPATING",
|
| 52 |
"140": "NON_PARTICIPATING",
|
|
@@ -60,18 +60,18 @@
|
|
| 60 |
"148": "SUCCESS",
|
| 61 |
"149": "NON_PARTICIPATING",
|
| 62 |
"15": "SUCCESS",
|
| 63 |
-
"150": "
|
| 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": "
|
| 75 |
"161": "NON_PARTICIPATING",
|
| 76 |
"162": "NON_PARTICIPATING",
|
| 77 |
"163": "NON_PARTICIPATING",
|
|
@@ -82,15 +82,15 @@
|
|
| 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": "SUCCESS",
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
"177": "SUCCESS",
|
| 93 |
-
"178": "
|
| 94 |
"179": "NON_PARTICIPATING",
|
| 95 |
"18": "SUCCESS",
|
| 96 |
"180": "NON_PARTICIPATING",
|
|
@@ -99,8 +99,8 @@
|
|
| 99 |
"183": "NON_PARTICIPATING",
|
| 100 |
"184": "NON_PARTICIPATING",
|
| 101 |
"185": "NON_PARTICIPATING",
|
| 102 |
-
"186": "
|
| 103 |
-
"187": "
|
| 104 |
"188": "NON_PARTICIPATING",
|
| 105 |
"189": "NON_PARTICIPATING",
|
| 106 |
"19": "NON_PARTICIPATING",
|
|
@@ -109,34 +109,34 @@
|
|
| 109 |
"192": "NON_PARTICIPATING",
|
| 110 |
"193": "NON_PARTICIPATING",
|
| 111 |
"194": "NON_PARTICIPATING",
|
| 112 |
-
"195": "
|
| 113 |
-
"196": "
|
| 114 |
"197": "NON_PARTICIPATING",
|
| 115 |
"198": "NON_PARTICIPATING",
|
| 116 |
"199": "NON_PARTICIPATING",
|
| 117 |
"2": "SUCCESS",
|
| 118 |
"20": "SUCCESS",
|
| 119 |
-
"200": "
|
| 120 |
-
"201": "
|
| 121 |
-
"202": "
|
| 122 |
"203": "SUCCESS",
|
| 123 |
"204": "NON_PARTICIPATING",
|
| 124 |
"205": "NON_PARTICIPATING",
|
| 125 |
"206": "NON_PARTICIPATING",
|
| 126 |
"207": "NON_PARTICIPATING",
|
| 127 |
-
"208": "
|
| 128 |
"209": "NON_PARTICIPATING",
|
| 129 |
"21": "SUCCESS",
|
| 130 |
-
"210": "
|
| 131 |
-
"211": "
|
| 132 |
"212": "NON_PARTICIPATING",
|
| 133 |
"213": "SUCCESS",
|
| 134 |
"214": "NON_PARTICIPATING",
|
| 135 |
"215": "NON_PARTICIPATING",
|
| 136 |
"216": "NON_PARTICIPATING",
|
| 137 |
"217": "NON_PARTICIPATING",
|
| 138 |
-
"218": "
|
| 139 |
-
"219": "
|
| 140 |
"22": "SUCCESS",
|
| 141 |
"220": "NON_PARTICIPATING",
|
| 142 |
"221": "NON_PARTICIPATING",
|
|
@@ -147,8 +147,8 @@
|
|
| 147 |
"226": "NON_PARTICIPATING",
|
| 148 |
"227": "NON_PARTICIPATING",
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
-
"229": "
|
| 151 |
-
"23": "
|
| 152 |
"230": "NON_PARTICIPATING",
|
| 153 |
"231": "NON_PARTICIPATING",
|
| 154 |
"232": "NON_PARTICIPATING",
|
|
@@ -161,26 +161,26 @@
|
|
| 161 |
"239": "NON_PARTICIPATING",
|
| 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": "NON_PARTICIPATING",
|
| 170 |
-
"247": "
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
-
"249": "
|
| 173 |
"25": "SUCCESS",
|
| 174 |
-
"250": "
|
| 175 |
"251": "NON_PARTICIPATING",
|
| 176 |
-
"252": "
|
| 177 |
"253": "NON_PARTICIPATING",
|
| 178 |
"254": "SUCCESS",
|
| 179 |
-
"255": "
|
| 180 |
"26": "SUCCESS",
|
| 181 |
"27": "SUCCESS",
|
| 182 |
"28": "SUCCESS",
|
| 183 |
-
"29": "
|
| 184 |
"3": "SUCCESS",
|
| 185 |
"30": "NON_PARTICIPATING",
|
| 186 |
"31": "SUCCESS",
|
|
@@ -206,7 +206,7 @@
|
|
| 206 |
"5": "NON_PARTICIPATING",
|
| 207 |
"50": "SUCCESS",
|
| 208 |
"51": "SUCCESS",
|
| 209 |
-
"52": "
|
| 210 |
"53": "NON_PARTICIPATING",
|
| 211 |
"54": "SUCCESS",
|
| 212 |
"55": "NON_PARTICIPATING",
|
|
@@ -228,7 +228,7 @@
|
|
| 228 |
"7": "SUCCESS",
|
| 229 |
"70": "NON_PARTICIPATING",
|
| 230 |
"71": "NON_PARTICIPATING",
|
| 231 |
-
"72": "
|
| 232 |
"73": "NON_PARTICIPATING",
|
| 233 |
"74": "SUCCESS",
|
| 234 |
"75": "SUCCESS",
|
|
@@ -238,7 +238,7 @@
|
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "SUCCESS",
|
| 240 |
"80": "SUCCESS",
|
| 241 |
-
"81": "
|
| 242 |
"82": "SUCCESS",
|
| 243 |
"83": "SUCCESS",
|
| 244 |
"84": "SUCCESS",
|
|
@@ -254,7 +254,7 @@
|
|
| 254 |
"93": "NON_PARTICIPATING",
|
| 255 |
"94": "SUCCESS",
|
| 256 |
"95": "SUCCESS",
|
| 257 |
-
"96": "
|
| 258 |
"97": "NON_PARTICIPATING",
|
| 259 |
"98": "SUCCESS",
|
| 260 |
"99": "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": "SUCCESS",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
+
"101": "NON_PARTICIPATING",
|
| 10 |
"102": "NON_PARTICIPATING",
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
"104": "NON_PARTICIPATING",
|
|
|
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
"109": "NON_PARTICIPATING",
|
| 18 |
"11": "NON_PARTICIPATING",
|
| 19 |
+
"110": "SUCCESS",
|
| 20 |
+
"111": "NON_PARTICIPATING",
|
| 21 |
"112": "NON_PARTICIPATING",
|
| 22 |
"113": "SUCCESS",
|
| 23 |
"114": "NON_PARTICIPATING",
|
|
|
|
| 27 |
"118": "NON_PARTICIPATING",
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
"12": "SUCCESS",
|
| 30 |
+
"120": "NON_PARTICIPATING",
|
| 31 |
+
"121": "NON_PARTICIPATING",
|
| 32 |
"122": "SUCCESS",
|
| 33 |
"123": "NON_PARTICIPATING",
|
| 34 |
"124": "NON_PARTICIPATING",
|
|
|
|
| 44 |
"133": "NON_PARTICIPATING",
|
| 45 |
"134": "NON_PARTICIPATING",
|
| 46 |
"135": "NON_PARTICIPATING",
|
| 47 |
+
"136": "SUCCESS",
|
| 48 |
"137": "NON_PARTICIPATING",
|
| 49 |
+
"138": "NON_PARTICIPATING",
|
| 50 |
"139": "NON_PARTICIPATING",
|
| 51 |
"14": "NON_PARTICIPATING",
|
| 52 |
"140": "NON_PARTICIPATING",
|
|
|
|
| 60 |
"148": "SUCCESS",
|
| 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": "SUCCESS",
|
| 74 |
+
"160": "SUCCESS",
|
| 75 |
"161": "NON_PARTICIPATING",
|
| 76 |
"162": "NON_PARTICIPATING",
|
| 77 |
"163": "NON_PARTICIPATING",
|
|
|
|
| 82 |
"168": "NON_PARTICIPATING",
|
| 83 |
"169": "NON_PARTICIPATING",
|
| 84 |
"17": "SUCCESS",
|
| 85 |
+
"170": "NON_PARTICIPATING",
|
| 86 |
"171": "NON_PARTICIPATING",
|
| 87 |
+
"172": "SUCCESS",
|
| 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",
|
|
|
|
| 99 |
"183": "NON_PARTICIPATING",
|
| 100 |
"184": "NON_PARTICIPATING",
|
| 101 |
"185": "NON_PARTICIPATING",
|
| 102 |
+
"186": "SUCCESS",
|
| 103 |
+
"187": "NON_PARTICIPATING",
|
| 104 |
"188": "NON_PARTICIPATING",
|
| 105 |
"189": "NON_PARTICIPATING",
|
| 106 |
"19": "NON_PARTICIPATING",
|
|
|
|
| 109 |
"192": "NON_PARTICIPATING",
|
| 110 |
"193": "NON_PARTICIPATING",
|
| 111 |
"194": "NON_PARTICIPATING",
|
| 112 |
+
"195": "NON_PARTICIPATING",
|
| 113 |
+
"196": "SUCCESS",
|
| 114 |
"197": "NON_PARTICIPATING",
|
| 115 |
"198": "NON_PARTICIPATING",
|
| 116 |
"199": "NON_PARTICIPATING",
|
| 117 |
"2": "SUCCESS",
|
| 118 |
"20": "SUCCESS",
|
| 119 |
+
"200": "SUCCESS",
|
| 120 |
+
"201": "SUCCESS",
|
| 121 |
+
"202": "SUCCESS",
|
| 122 |
"203": "SUCCESS",
|
| 123 |
"204": "NON_PARTICIPATING",
|
| 124 |
"205": "NON_PARTICIPATING",
|
| 125 |
"206": "NON_PARTICIPATING",
|
| 126 |
"207": "NON_PARTICIPATING",
|
| 127 |
+
"208": "NON_PARTICIPATING",
|
| 128 |
"209": "NON_PARTICIPATING",
|
| 129 |
"21": "SUCCESS",
|
| 130 |
+
"210": "SUCCESS",
|
| 131 |
+
"211": "NON_PARTICIPATING",
|
| 132 |
"212": "NON_PARTICIPATING",
|
| 133 |
"213": "SUCCESS",
|
| 134 |
"214": "NON_PARTICIPATING",
|
| 135 |
"215": "NON_PARTICIPATING",
|
| 136 |
"216": "NON_PARTICIPATING",
|
| 137 |
"217": "NON_PARTICIPATING",
|
| 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": "SUCCESS",
|
| 151 |
+
"23": "SUCCESS",
|
| 152 |
"230": "NON_PARTICIPATING",
|
| 153 |
"231": "NON_PARTICIPATING",
|
| 154 |
"232": "NON_PARTICIPATING",
|
|
|
|
| 161 |
"239": "NON_PARTICIPATING",
|
| 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": "NON_PARTICIPATING",
|
| 169 |
"246": "NON_PARTICIPATING",
|
| 170 |
+
"247": "SUCCESS",
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
+
"249": "SUCCESS",
|
| 173 |
"25": "SUCCESS",
|
| 174 |
+
"250": "SUCCESS",
|
| 175 |
"251": "NON_PARTICIPATING",
|
| 176 |
+
"252": "SUCCESS",
|
| 177 |
"253": "NON_PARTICIPATING",
|
| 178 |
"254": "SUCCESS",
|
| 179 |
+
"255": "SUCCESS",
|
| 180 |
"26": "SUCCESS",
|
| 181 |
"27": "SUCCESS",
|
| 182 |
"28": "SUCCESS",
|
| 183 |
+
"29": "NON_PARTICIPATING",
|
| 184 |
"3": "SUCCESS",
|
| 185 |
"30": "NON_PARTICIPATING",
|
| 186 |
"31": "SUCCESS",
|
|
|
|
| 206 |
"5": "NON_PARTICIPATING",
|
| 207 |
"50": "SUCCESS",
|
| 208 |
"51": "SUCCESS",
|
| 209 |
+
"52": "SUCCESS",
|
| 210 |
"53": "NON_PARTICIPATING",
|
| 211 |
"54": "SUCCESS",
|
| 212 |
"55": "NON_PARTICIPATING",
|
|
|
|
| 228 |
"7": "SUCCESS",
|
| 229 |
"70": "NON_PARTICIPATING",
|
| 230 |
"71": "NON_PARTICIPATING",
|
| 231 |
+
"72": "SUCCESS",
|
| 232 |
"73": "NON_PARTICIPATING",
|
| 233 |
"74": "SUCCESS",
|
| 234 |
"75": "SUCCESS",
|
|
|
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "SUCCESS",
|
| 240 |
"80": "SUCCESS",
|
| 241 |
+
"81": "SUCCESS",
|
| 242 |
"82": "SUCCESS",
|
| 243 |
"83": "SUCCESS",
|
| 244 |
"84": "SUCCESS",
|
|
|
|
| 254 |
"93": "NON_PARTICIPATING",
|
| 255 |
"94": "SUCCESS",
|
| 256 |
"95": "SUCCESS",
|
| 257 |
+
"96": "NON_PARTICIPATING",
|
| 258 |
"97": "NON_PARTICIPATING",
|
| 259 |
"98": "SUCCESS",
|
| 260 |
"99": "SUCCESS"
|
|
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
+
5534744,
|
| 272 |
+
5534747,
|
| 273 |
+
5534749,
|
| 274 |
+
5534752,
|
| 275 |
+
5534755
|
| 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": 5534554,
|
| 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:fe2b4dad6f1244c2d31b72e5c24c37b523cf30130c59d8d5e890c47906370605
|
| 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:e8dcc3d8a04ef583a65c7a4f910329f2f35d2780dd2e3534a0b723fef03eb6b1
|
| 3 |
size 4040701744
|