Run 4. Outer Step 58. Inner Step 0. Peers 91.
Browse files- config.json +37 -37
- 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",
|
|
@@ -12,11 +12,11 @@
|
|
| 12 |
"104": "NON_PARTICIPATING",
|
| 13 |
"105": "NON_PARTICIPATING",
|
| 14 |
"106": "NON_PARTICIPATING",
|
| 15 |
-
"107": "
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
"109": "NON_PARTICIPATING",
|
| 18 |
"11": "SUCCESS",
|
| 19 |
-
"110": "
|
| 20 |
"111": "SUCCESS",
|
| 21 |
"112": "NON_PARTICIPATING",
|
| 22 |
"113": "NON_PARTICIPATING",
|
|
@@ -30,12 +30,12 @@
|
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
"122": "NON_PARTICIPATING",
|
| 33 |
-
"123": "
|
| 34 |
-
"124": "
|
| 35 |
"125": "NON_PARTICIPATING",
|
| 36 |
-
"126": "
|
| 37 |
"127": "NON_PARTICIPATING",
|
| 38 |
-
"128": "
|
| 39 |
"129": "NON_PARTICIPATING",
|
| 40 |
"13": "SUCCESS",
|
| 41 |
"130": "SUCCESS",
|
|
@@ -46,7 +46,7 @@
|
|
| 46 |
"135": "NON_PARTICIPATING",
|
| 47 |
"136": "SUCCESS",
|
| 48 |
"137": "NON_PARTICIPATING",
|
| 49 |
-
"138": "
|
| 50 |
"139": "NON_PARTICIPATING",
|
| 51 |
"14": "SUCCESS",
|
| 52 |
"140": "NON_PARTICIPATING",
|
|
@@ -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",
|
|
@@ -75,8 +75,8 @@
|
|
| 75 |
"161": "NON_PARTICIPATING",
|
| 76 |
"162": "NON_PARTICIPATING",
|
| 77 |
"163": "NON_PARTICIPATING",
|
| 78 |
-
"164": "
|
| 79 |
-
"165": "
|
| 80 |
"166": "NON_PARTICIPATING",
|
| 81 |
"167": "NON_PARTICIPATING",
|
| 82 |
"168": "NON_PARTICIPATING",
|
|
@@ -87,20 +87,20 @@
|
|
| 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",
|
| 94 |
"179": "NON_PARTICIPATING",
|
| 95 |
-
"18": "
|
| 96 |
"180": "NON_PARTICIPATING",
|
| 97 |
-
"181": "
|
| 98 |
"182": "NON_PARTICIPATING",
|
| 99 |
"183": "NON_PARTICIPATING",
|
| 100 |
"184": "NON_PARTICIPATING",
|
| 101 |
"185": "NON_PARTICIPATING",
|
| 102 |
"186": "NON_PARTICIPATING",
|
| 103 |
-
"187": "
|
| 104 |
"188": "NON_PARTICIPATING",
|
| 105 |
"189": "NON_PARTICIPATING",
|
| 106 |
"19": "SUCCESS",
|
|
@@ -115,41 +115,41 @@
|
|
| 115 |
"198": "NON_PARTICIPATING",
|
| 116 |
"199": "NON_PARTICIPATING",
|
| 117 |
"2": "SUCCESS",
|
| 118 |
-
"20": "
|
| 119 |
"200": "NON_PARTICIPATING",
|
| 120 |
"201": "NON_PARTICIPATING",
|
| 121 |
-
"202": "
|
| 122 |
"203": "NON_PARTICIPATING",
|
| 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": "NON_PARTICIPATING",
|
| 131 |
"211": "NON_PARTICIPATING",
|
| 132 |
"212": "NON_PARTICIPATING",
|
| 133 |
"213": "NON_PARTICIPATING",
|
| 134 |
-
"214": "
|
| 135 |
-
"215": "
|
| 136 |
"216": "NON_PARTICIPATING",
|
| 137 |
"217": "NON_PARTICIPATING",
|
| 138 |
"218": "NON_PARTICIPATING",
|
| 139 |
"219": "NON_PARTICIPATING",
|
| 140 |
"22": "SUCCESS",
|
| 141 |
-
"220": "
|
| 142 |
"221": "NON_PARTICIPATING",
|
| 143 |
"222": "NON_PARTICIPATING",
|
| 144 |
-
"223": "
|
| 145 |
"224": "NON_PARTICIPATING",
|
| 146 |
-
"225": "
|
| 147 |
"226": "NON_PARTICIPATING",
|
| 148 |
"227": "NON_PARTICIPATING",
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
"229": "SUCCESS",
|
| 151 |
"23": "SUCCESS",
|
| 152 |
-
"230": "
|
| 153 |
"231": "NON_PARTICIPATING",
|
| 154 |
"232": "NON_PARTICIPATING",
|
| 155 |
"233": "NON_PARTICIPATING",
|
|
@@ -160,10 +160,10 @@
|
|
| 160 |
"238": "NON_PARTICIPATING",
|
| 161 |
"239": "NON_PARTICIPATING",
|
| 162 |
"24": "SUCCESS",
|
| 163 |
-
"240": "
|
| 164 |
-
"241": "
|
| 165 |
"242": "NON_PARTICIPATING",
|
| 166 |
-
"243": "
|
| 167 |
"244": "NON_PARTICIPATING",
|
| 168 |
"245": "NON_PARTICIPATING",
|
| 169 |
"246": "NON_PARTICIPATING",
|
|
@@ -171,7 +171,7 @@
|
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
"249": "SUCCESS",
|
| 173 |
"25": "SUCCESS",
|
| 174 |
-
"250": "
|
| 175 |
"251": "NON_PARTICIPATING",
|
| 176 |
"252": "NON_PARTICIPATING",
|
| 177 |
"253": "NON_PARTICIPATING",
|
|
@@ -183,7 +183,7 @@
|
|
| 183 |
"29": "SUCCESS",
|
| 184 |
"3": "SUCCESS",
|
| 185 |
"30": "SUCCESS",
|
| 186 |
-
"31": "
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "SUCCESS",
|
|
@@ -192,23 +192,23 @@
|
|
| 192 |
"37": "SUCCESS",
|
| 193 |
"38": "SUCCESS",
|
| 194 |
"39": "SUCCESS",
|
| 195 |
-
"4": "
|
| 196 |
"40": "SUCCESS",
|
| 197 |
"41": "SUCCESS",
|
| 198 |
"42": "NON_PARTICIPATING",
|
| 199 |
"43": "SUCCESS",
|
| 200 |
-
"44": "
|
| 201 |
"45": "SUCCESS",
|
| 202 |
"46": "NON_PARTICIPATING",
|
| 203 |
"47": "NON_PARTICIPATING",
|
| 204 |
"48": "NON_PARTICIPATING",
|
| 205 |
"49": "SUCCESS",
|
| 206 |
-
"5": "
|
| 207 |
"50": "SUCCESS",
|
| 208 |
"51": "SUCCESS",
|
| 209 |
"52": "SUCCESS",
|
| 210 |
"53": "SUCCESS",
|
| 211 |
-
"54": "
|
| 212 |
"55": "SUCCESS",
|
| 213 |
"56": "SUCCESS",
|
| 214 |
"57": "SUCCESS",
|
|
@@ -220,7 +220,7 @@
|
|
| 220 |
"62": "SUCCESS",
|
| 221 |
"63": "SUCCESS",
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
-
"65": "
|
| 224 |
"66": "SUCCESS",
|
| 225 |
"67": "SUCCESS",
|
| 226 |
"68": "SUCCESS",
|
|
@@ -230,8 +230,8 @@
|
|
| 230 |
"71": "NON_PARTICIPATING",
|
| 231 |
"72": "SUCCESS",
|
| 232 |
"73": "NON_PARTICIPATING",
|
| 233 |
-
"74": "
|
| 234 |
-
"75": "
|
| 235 |
"76": "SUCCESS",
|
| 236 |
"77": "SUCCESS",
|
| 237 |
"78": "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": "FAIL",
|
| 7 |
"10": "NON_PARTICIPATING",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
"101": "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",
|
| 18 |
"11": "SUCCESS",
|
| 19 |
+
"110": "SUCCESS",
|
| 20 |
"111": "SUCCESS",
|
| 21 |
"112": "NON_PARTICIPATING",
|
| 22 |
"113": "NON_PARTICIPATING",
|
|
|
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
"122": "NON_PARTICIPATING",
|
| 33 |
+
"123": "SUCCESS",
|
| 34 |
+
"124": "NON_PARTICIPATING",
|
| 35 |
"125": "NON_PARTICIPATING",
|
| 36 |
+
"126": "NON_PARTICIPATING",
|
| 37 |
"127": "NON_PARTICIPATING",
|
| 38 |
+
"128": "SUCCESS",
|
| 39 |
"129": "NON_PARTICIPATING",
|
| 40 |
"13": "SUCCESS",
|
| 41 |
"130": "SUCCESS",
|
|
|
|
| 46 |
"135": "NON_PARTICIPATING",
|
| 47 |
"136": "SUCCESS",
|
| 48 |
"137": "NON_PARTICIPATING",
|
| 49 |
+
"138": "SUCCESS",
|
| 50 |
"139": "NON_PARTICIPATING",
|
| 51 |
"14": "SUCCESS",
|
| 52 |
"140": "NON_PARTICIPATING",
|
|
|
|
| 56 |
"144": "NON_PARTICIPATING",
|
| 57 |
"145": "NON_PARTICIPATING",
|
| 58 |
"146": "NON_PARTICIPATING",
|
| 59 |
+
"147": "SUCCESS",
|
| 60 |
"148": "NON_PARTICIPATING",
|
| 61 |
"149": "NON_PARTICIPATING",
|
| 62 |
"15": "SUCCESS",
|
|
|
|
| 75 |
"161": "NON_PARTICIPATING",
|
| 76 |
"162": "NON_PARTICIPATING",
|
| 77 |
"163": "NON_PARTICIPATING",
|
| 78 |
+
"164": "SUCCESS",
|
| 79 |
+
"165": "NON_PARTICIPATING",
|
| 80 |
"166": "NON_PARTICIPATING",
|
| 81 |
"167": "NON_PARTICIPATING",
|
| 82 |
"168": "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",
|
| 94 |
"179": "NON_PARTICIPATING",
|
| 95 |
+
"18": "NON_PARTICIPATING",
|
| 96 |
"180": "NON_PARTICIPATING",
|
| 97 |
+
"181": "NON_PARTICIPATING",
|
| 98 |
"182": "NON_PARTICIPATING",
|
| 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",
|
|
|
|
| 115 |
"198": "NON_PARTICIPATING",
|
| 116 |
"199": "NON_PARTICIPATING",
|
| 117 |
"2": "SUCCESS",
|
| 118 |
+
"20": "FAIL",
|
| 119 |
"200": "NON_PARTICIPATING",
|
| 120 |
"201": "NON_PARTICIPATING",
|
| 121 |
+
"202": "SUCCESS",
|
| 122 |
"203": "NON_PARTICIPATING",
|
| 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": "NON_PARTICIPATING",
|
| 131 |
"211": "NON_PARTICIPATING",
|
| 132 |
"212": "NON_PARTICIPATING",
|
| 133 |
"213": "NON_PARTICIPATING",
|
| 134 |
+
"214": "SUCCESS",
|
| 135 |
+
"215": "NON_PARTICIPATING",
|
| 136 |
"216": "NON_PARTICIPATING",
|
| 137 |
"217": "NON_PARTICIPATING",
|
| 138 |
"218": "NON_PARTICIPATING",
|
| 139 |
"219": "NON_PARTICIPATING",
|
| 140 |
"22": "SUCCESS",
|
| 141 |
+
"220": "SUCCESS",
|
| 142 |
"221": "NON_PARTICIPATING",
|
| 143 |
"222": "NON_PARTICIPATING",
|
| 144 |
+
"223": "NON_PARTICIPATING",
|
| 145 |
"224": "NON_PARTICIPATING",
|
| 146 |
+
"225": "SUCCESS",
|
| 147 |
"226": "NON_PARTICIPATING",
|
| 148 |
"227": "NON_PARTICIPATING",
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
"229": "SUCCESS",
|
| 151 |
"23": "SUCCESS",
|
| 152 |
+
"230": "SUCCESS",
|
| 153 |
"231": "NON_PARTICIPATING",
|
| 154 |
"232": "NON_PARTICIPATING",
|
| 155 |
"233": "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",
|
| 166 |
+
"243": "NON_PARTICIPATING",
|
| 167 |
"244": "NON_PARTICIPATING",
|
| 168 |
"245": "NON_PARTICIPATING",
|
| 169 |
"246": "NON_PARTICIPATING",
|
|
|
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
"249": "SUCCESS",
|
| 173 |
"25": "SUCCESS",
|
| 174 |
+
"250": "SUCCESS",
|
| 175 |
"251": "NON_PARTICIPATING",
|
| 176 |
"252": "NON_PARTICIPATING",
|
| 177 |
"253": "NON_PARTICIPATING",
|
|
|
|
| 183 |
"29": "SUCCESS",
|
| 184 |
"3": "SUCCESS",
|
| 185 |
"30": "SUCCESS",
|
| 186 |
+
"31": "SUCCESS",
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "SUCCESS",
|
|
|
|
| 192 |
"37": "SUCCESS",
|
| 193 |
"38": "SUCCESS",
|
| 194 |
"39": "SUCCESS",
|
| 195 |
+
"4": "FAIL",
|
| 196 |
"40": "SUCCESS",
|
| 197 |
"41": "SUCCESS",
|
| 198 |
"42": "NON_PARTICIPATING",
|
| 199 |
"43": "SUCCESS",
|
| 200 |
+
"44": "FAIL",
|
| 201 |
"45": "SUCCESS",
|
| 202 |
"46": "NON_PARTICIPATING",
|
| 203 |
"47": "NON_PARTICIPATING",
|
| 204 |
"48": "NON_PARTICIPATING",
|
| 205 |
"49": "SUCCESS",
|
| 206 |
+
"5": "FAIL",
|
| 207 |
"50": "SUCCESS",
|
| 208 |
"51": "SUCCESS",
|
| 209 |
"52": "SUCCESS",
|
| 210 |
"53": "SUCCESS",
|
| 211 |
+
"54": "FAIL",
|
| 212 |
"55": "SUCCESS",
|
| 213 |
"56": "SUCCESS",
|
| 214 |
"57": "SUCCESS",
|
|
|
|
| 220 |
"62": "SUCCESS",
|
| 221 |
"63": "SUCCESS",
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
+
"65": "FAIL",
|
| 224 |
"66": "SUCCESS",
|
| 225 |
"67": "SUCCESS",
|
| 226 |
"68": "SUCCESS",
|
|
|
|
| 230 |
"71": "NON_PARTICIPATING",
|
| 231 |
"72": "SUCCESS",
|
| 232 |
"73": "NON_PARTICIPATING",
|
| 233 |
+
"74": "FAIL",
|
| 234 |
+
"75": "FAIL",
|
| 235 |
"76": "SUCCESS",
|
| 236 |
"77": "SUCCESS",
|
| 237 |
"78": "SUCCESS",
|
|
|
|
| 275 |
"initializer_range": 0.02,
|
| 276 |
"inner_step": 0,
|
| 277 |
"inner_steps": 0,
|
| 278 |
+
"last_allreduce_block": 5586129,
|
| 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:bf08af80e71d175751ce02fbd6d6c87042b51fc592ab8bdb47207174a85d6a34
|
| 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 sha256:126c9144bb4d1735c07dda5ed227396053890e84be9398dd4a7e5f7a1bf8f3cc
|
| 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:af44269bac8c19a41f811d90d7b57313bf123e923af5b886fcef77de1e968163
|
| 3 |
size 4040805354
|