Run 4. Outer Step 93. Inner Step 0. Peers 73.
Browse files- config.json +33 -33
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
- outer_optimizer.pt +1 -1
config.json
CHANGED
|
@@ -3,7 +3,7 @@
|
|
| 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",
|
|
@@ -11,7 +11,7 @@
|
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
"104": "NON_PARTICIPATING",
|
| 13 |
"105": "NON_PARTICIPATING",
|
| 14 |
-
"106": "
|
| 15 |
"107": "NON_PARTICIPATING",
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
"109": "NON_PARTICIPATING",
|
|
@@ -32,10 +32,10 @@
|
|
| 32 |
"122": "NON_PARTICIPATING",
|
| 33 |
"123": "NON_PARTICIPATING",
|
| 34 |
"124": "NON_PARTICIPATING",
|
| 35 |
-
"125": "
|
| 36 |
"126": "NON_PARTICIPATING",
|
| 37 |
"127": "NON_PARTICIPATING",
|
| 38 |
-
"128": "
|
| 39 |
"129": "NON_PARTICIPATING",
|
| 40 |
"13": "SUCCESS",
|
| 41 |
"130": "NON_PARTICIPATING",
|
|
@@ -62,7 +62,7 @@
|
|
| 62 |
"15": "SUCCESS",
|
| 63 |
"150": "NON_PARTICIPATING",
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
-
"152": "
|
| 66 |
"153": "NON_PARTICIPATING",
|
| 67 |
"154": "NON_PARTICIPATING",
|
| 68 |
"155": "NON_PARTICIPATING",
|
|
@@ -70,14 +70,14 @@
|
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
"159": "NON_PARTICIPATING",
|
| 73 |
-
"16": "
|
| 74 |
"160": "NON_PARTICIPATING",
|
| 75 |
"161": "SUCCESS",
|
| 76 |
"162": "SUCCESS",
|
| 77 |
"163": "NON_PARTICIPATING",
|
| 78 |
"164": "NON_PARTICIPATING",
|
| 79 |
"165": "NON_PARTICIPATING",
|
| 80 |
-
"166": "
|
| 81 |
"167": "NON_PARTICIPATING",
|
| 82 |
"168": "NON_PARTICIPATING",
|
| 83 |
"169": "NON_PARTICIPATING",
|
|
@@ -86,18 +86,18 @@
|
|
| 86 |
"171": "NON_PARTICIPATING",
|
| 87 |
"172": "NON_PARTICIPATING",
|
| 88 |
"173": "NON_PARTICIPATING",
|
| 89 |
-
"174": "
|
| 90 |
"175": "NON_PARTICIPATING",
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
-
"177": "
|
| 93 |
"178": "NON_PARTICIPATING",
|
| 94 |
"179": "NON_PARTICIPATING",
|
| 95 |
"18": "SUCCESS",
|
| 96 |
"180": "NON_PARTICIPATING",
|
| 97 |
"181": "NON_PARTICIPATING",
|
| 98 |
"182": "NON_PARTICIPATING",
|
| 99 |
-
"183": "
|
| 100 |
-
"184": "
|
| 101 |
"185": "NON_PARTICIPATING",
|
| 102 |
"186": "NON_PARTICIPATING",
|
| 103 |
"187": "NON_PARTICIPATING",
|
|
@@ -106,40 +106,40 @@
|
|
| 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",
|
| 113 |
"196": "NON_PARTICIPATING",
|
| 114 |
"197": "NON_PARTICIPATING",
|
| 115 |
"198": "NON_PARTICIPATING",
|
| 116 |
-
"199": "
|
| 117 |
"2": "SUCCESS",
|
| 118 |
"20": "SUCCESS",
|
| 119 |
"200": "NON_PARTICIPATING",
|
| 120 |
"201": "NON_PARTICIPATING",
|
| 121 |
"202": "NON_PARTICIPATING",
|
| 122 |
"203": "NON_PARTICIPATING",
|
| 123 |
-
"204": "
|
| 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",
|
| 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": "
|
| 143 |
"222": "NON_PARTICIPATING",
|
| 144 |
"223": "NON_PARTICIPATING",
|
| 145 |
"224": "NON_PARTICIPATING",
|
|
@@ -149,7 +149,7 @@
|
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
"229": "NON_PARTICIPATING",
|
| 151 |
"23": "NON_PARTICIPATING",
|
| 152 |
-
"230": "
|
| 153 |
"231": "NON_PARTICIPATING",
|
| 154 |
"232": "NON_PARTICIPATING",
|
| 155 |
"233": "NON_PARTICIPATING",
|
|
@@ -167,19 +167,19 @@
|
|
| 167 |
"244": "NON_PARTICIPATING",
|
| 168 |
"245": "NON_PARTICIPATING",
|
| 169 |
"246": "NON_PARTICIPATING",
|
| 170 |
-
"247": "
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
"249": "NON_PARTICIPATING",
|
| 173 |
"25": "SUCCESS",
|
| 174 |
-
"250": "
|
| 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": "
|
| 182 |
-
"28": "
|
| 183 |
"29": "SUCCESS",
|
| 184 |
"3": "NON_PARTICIPATING",
|
| 185 |
"30": "NON_PARTICIPATING",
|
|
@@ -199,17 +199,17 @@
|
|
| 199 |
"43": "NON_PARTICIPATING",
|
| 200 |
"44": "SUCCESS",
|
| 201 |
"45": "SUCCESS",
|
| 202 |
-
"46": "
|
| 203 |
"47": "SUCCESS",
|
| 204 |
"48": "NON_PARTICIPATING",
|
| 205 |
"49": "SUCCESS",
|
| 206 |
-
"5": "
|
| 207 |
"50": "NON_PARTICIPATING",
|
| 208 |
"51": "SUCCESS",
|
| 209 |
"52": "SUCCESS",
|
| 210 |
"53": "NON_PARTICIPATING",
|
| 211 |
"54": "SUCCESS",
|
| 212 |
-
"55": "
|
| 213 |
"56": "SUCCESS",
|
| 214 |
"57": "NON_PARTICIPATING",
|
| 215 |
"58": "NON_PARTICIPATING",
|
|
@@ -240,19 +240,19 @@
|
|
| 240 |
"80": "SUCCESS",
|
| 241 |
"81": "NON_PARTICIPATING",
|
| 242 |
"82": "SUCCESS",
|
| 243 |
-
"83": "
|
| 244 |
"84": "SUCCESS",
|
| 245 |
"85": "SUCCESS",
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
"87": "SUCCESS",
|
| 248 |
-
"88": "
|
| 249 |
-
"89": "
|
| 250 |
"9": "SUCCESS",
|
| 251 |
"90": "SUCCESS",
|
| 252 |
"91": "SUCCESS",
|
| 253 |
"92": "SUCCESS",
|
| 254 |
"93": "NON_PARTICIPATING",
|
| 255 |
-
"94": "
|
| 256 |
"95": "SUCCESS",
|
| 257 |
"96": "NON_PARTICIPATING",
|
| 258 |
"97": "SUCCESS",
|
|
@@ -275,7 +275,7 @@
|
|
| 275 |
"initializer_range": 0.02,
|
| 276 |
"inner_step": 0,
|
| 277 |
"inner_steps": 0,
|
| 278 |
-
"last_allreduce_block":
|
| 279 |
"layer_norm_epsilon": 1e-05,
|
| 280 |
"model_type": "gpt_optimized",
|
| 281 |
"n_embd": 1280,
|
|
|
|
| 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",
|
|
|
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
"104": "NON_PARTICIPATING",
|
| 13 |
"105": "NON_PARTICIPATING",
|
| 14 |
+
"106": "NON_PARTICIPATING",
|
| 15 |
"107": "NON_PARTICIPATING",
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
"109": "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",
|
|
|
|
| 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",
|
|
|
|
| 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",
|
| 79 |
"165": "NON_PARTICIPATING",
|
| 80 |
+
"166": "NON_PARTICIPATING",
|
| 81 |
"167": "NON_PARTICIPATING",
|
| 82 |
"168": "NON_PARTICIPATING",
|
| 83 |
"169": "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",
|
| 93 |
"178": "NON_PARTICIPATING",
|
| 94 |
"179": "NON_PARTICIPATING",
|
| 95 |
"18": "SUCCESS",
|
| 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",
|
|
|
|
| 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",
|
| 113 |
"196": "NON_PARTICIPATING",
|
| 114 |
"197": "NON_PARTICIPATING",
|
| 115 |
"198": "NON_PARTICIPATING",
|
| 116 |
+
"199": "NON_PARTICIPATING",
|
| 117 |
"2": "SUCCESS",
|
| 118 |
"20": "SUCCESS",
|
| 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",
|
| 131 |
"211": "NON_PARTICIPATING",
|
| 132 |
"212": "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",
|
| 143 |
"222": "NON_PARTICIPATING",
|
| 144 |
"223": "NON_PARTICIPATING",
|
| 145 |
"224": "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",
|
|
|
|
| 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",
|
|
|
|
| 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",
|
|
|
|
| 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",
|
|
|
|
| 275 |
"initializer_range": 0.02,
|
| 276 |
"inner_step": 0,
|
| 277 |
"inner_steps": 0,
|
| 278 |
+
"last_allreduce_block": 5686307,
|
| 279 |
"layer_norm_epsilon": 1e-05,
|
| 280 |
"model_type": "gpt_optimized",
|
| 281 |
"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 sha256:5db4cf8bb32a236edb96631d30f4c06400a5758cfc84f0126493f936465ca302
|
| 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 sha256:00e0a1bd9f9498444a0f26aac576cae9be850466c373c25c9500b5cdefeae033
|
| 3 |
size 4040701744
|
outer_optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4040805354
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0abff9c8cb59b8fc85b3e71e24b403679e28b836909985b186a385777fda4ab2
|
| 3 |
size 4040805354
|