Run 4. Outer Step 23. Inner Step 0. Peers 88.
Browse files- config.json +43 -43
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
- outer_optimizer.pt +1 -1
config.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "
|
| 3 |
"activation_function": "gelu_new",
|
| 4 |
"all_reduce_scores": {
|
| 5 |
"0": "NON_PARTICIPATING",
|
|
@@ -11,10 +11,10 @@
|
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
"104": "NON_PARTICIPATING",
|
| 13 |
"105": "NON_PARTICIPATING",
|
| 14 |
-
"106": "
|
| 15 |
"107": "SUCCESS",
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
-
"109": "
|
| 18 |
"11": "NON_PARTICIPATING",
|
| 19 |
"110": "SUCCESS",
|
| 20 |
"111": "NON_PARTICIPATING",
|
|
@@ -24,26 +24,26 @@
|
|
| 24 |
"115": "NON_PARTICIPATING",
|
| 25 |
"116": "NON_PARTICIPATING",
|
| 26 |
"117": "NON_PARTICIPATING",
|
| 27 |
-
"118": "
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
"12": "SUCCESS",
|
| 30 |
-
"120": "
|
| 31 |
"121": "SUCCESS",
|
| 32 |
"122": "NON_PARTICIPATING",
|
| 33 |
"123": "SUCCESS",
|
| 34 |
"124": "NON_PARTICIPATING",
|
| 35 |
"125": "NON_PARTICIPATING",
|
| 36 |
"126": "NON_PARTICIPATING",
|
| 37 |
-
"127": "
|
| 38 |
"128": "NON_PARTICIPATING",
|
| 39 |
-
"129": "
|
| 40 |
"13": "NON_PARTICIPATING",
|
| 41 |
"130": "NON_PARTICIPATING",
|
| 42 |
"131": "NON_PARTICIPATING",
|
| 43 |
-
"132": "
|
| 44 |
"133": "NON_PARTICIPATING",
|
| 45 |
"134": "NON_PARTICIPATING",
|
| 46 |
-
"135": "
|
| 47 |
"136": "NON_PARTICIPATING",
|
| 48 |
"137": "NON_PARTICIPATING",
|
| 49 |
"138": "NON_PARTICIPATING",
|
|
@@ -51,25 +51,25 @@
|
|
| 51 |
"14": "NON_PARTICIPATING",
|
| 52 |
"140": "NON_PARTICIPATING",
|
| 53 |
"141": "NON_PARTICIPATING",
|
| 54 |
-
"142": "
|
| 55 |
-
"143": "
|
| 56 |
"144": "NON_PARTICIPATING",
|
| 57 |
"145": "NON_PARTICIPATING",
|
| 58 |
"146": "NON_PARTICIPATING",
|
| 59 |
-
"147": "
|
| 60 |
-
"148": "
|
| 61 |
-
"149": "
|
| 62 |
"15": "SUCCESS",
|
| 63 |
-
"150": "
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
"153": "NON_PARTICIPATING",
|
| 67 |
"154": "SUCCESS",
|
| 68 |
-
"155": "
|
| 69 |
-
"156": "
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
-
"159": "
|
| 73 |
"16": "SUCCESS",
|
| 74 |
"160": "SUCCESS",
|
| 75 |
"161": "NON_PARTICIPATING",
|
|
@@ -78,16 +78,16 @@
|
|
| 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": "SUCCESS",
|
| 86 |
"171": "NON_PARTICIPATING",
|
| 87 |
-
"172": "
|
| 88 |
"173": "NON_PARTICIPATING",
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
-
"175": "
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
"177": "NON_PARTICIPATING",
|
| 93 |
"178": "NON_PARTICIPATING",
|
|
@@ -98,9 +98,9 @@
|
|
| 98 |
"182": "NON_PARTICIPATING",
|
| 99 |
"183": "NON_PARTICIPATING",
|
| 100 |
"184": "SUCCESS",
|
| 101 |
-
"185": "
|
| 102 |
"186": "SUCCESS",
|
| 103 |
-
"187": "
|
| 104 |
"188": "NON_PARTICIPATING",
|
| 105 |
"189": "NON_PARTICIPATING",
|
| 106 |
"19": "NON_PARTICIPATING",
|
|
@@ -116,10 +116,10 @@
|
|
| 116 |
"199": "NON_PARTICIPATING",
|
| 117 |
"2": "SUCCESS",
|
| 118 |
"20": "SUCCESS",
|
| 119 |
-
"200": "
|
| 120 |
"201": "NON_PARTICIPATING",
|
| 121 |
"202": "NON_PARTICIPATING",
|
| 122 |
-
"203": "
|
| 123 |
"204": "NON_PARTICIPATING",
|
| 124 |
"205": "NON_PARTICIPATING",
|
| 125 |
"206": "NON_PARTICIPATING",
|
|
@@ -130,25 +130,25 @@
|
|
| 130 |
"210": "NON_PARTICIPATING",
|
| 131 |
"211": "NON_PARTICIPATING",
|
| 132 |
"212": "NON_PARTICIPATING",
|
| 133 |
-
"213": "
|
| 134 |
"214": "SUCCESS",
|
| 135 |
-
"215": "
|
| 136 |
"216": "NON_PARTICIPATING",
|
| 137 |
"217": "NON_PARTICIPATING",
|
| 138 |
-
"218": "
|
| 139 |
"219": "SUCCESS",
|
| 140 |
"22": "SUCCESS",
|
| 141 |
"220": "NON_PARTICIPATING",
|
| 142 |
-
"221": "
|
| 143 |
"222": "NON_PARTICIPATING",
|
| 144 |
-
"223": "
|
| 145 |
"224": "NON_PARTICIPATING",
|
| 146 |
"225": "SUCCESS",
|
| 147 |
-
"226": "
|
| 148 |
"227": "NON_PARTICIPATING",
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
"229": "NON_PARTICIPATING",
|
| 151 |
-
"23": "
|
| 152 |
"230": "NON_PARTICIPATING",
|
| 153 |
"231": "NON_PARTICIPATING",
|
| 154 |
"232": "NON_PARTICIPATING",
|
|
@@ -159,9 +159,9 @@
|
|
| 159 |
"237": "NON_PARTICIPATING",
|
| 160 |
"238": "NON_PARTICIPATING",
|
| 161 |
"239": "NON_PARTICIPATING",
|
| 162 |
-
"24": "
|
| 163 |
"240": "NON_PARTICIPATING",
|
| 164 |
-
"241": "
|
| 165 |
"242": "NON_PARTICIPATING",
|
| 166 |
"243": "NON_PARTICIPATING",
|
| 167 |
"244": "SUCCESS",
|
|
@@ -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",
|
|
@@ -184,9 +184,9 @@
|
|
| 184 |
"3": "SUCCESS",
|
| 185 |
"30": "NON_PARTICIPATING",
|
| 186 |
"31": "NON_PARTICIPATING",
|
| 187 |
-
"32": "
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
-
"34": "
|
| 190 |
"35": "NON_PARTICIPATING",
|
| 191 |
"36": "NON_PARTICIPATING",
|
| 192 |
"37": "SUCCESS",
|
|
@@ -209,16 +209,16 @@
|
|
| 209 |
"52": "NON_PARTICIPATING",
|
| 210 |
"53": "NON_PARTICIPATING",
|
| 211 |
"54": "SUCCESS",
|
| 212 |
-
"55": "
|
| 213 |
"56": "SUCCESS",
|
| 214 |
"57": "SUCCESS",
|
| 215 |
"58": "SUCCESS",
|
| 216 |
"59": "NON_PARTICIPATING",
|
| 217 |
-
"6": "
|
| 218 |
"60": "SUCCESS",
|
| 219 |
-
"61": "
|
| 220 |
"62": "SUCCESS",
|
| 221 |
-
"63": "
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
"65": "SUCCESS",
|
| 224 |
"66": "NON_PARTICIPATING",
|
|
@@ -242,7 +242,7 @@
|
|
| 242 |
"82": "SUCCESS",
|
| 243 |
"83": "NON_PARTICIPATING",
|
| 244 |
"84": "SUCCESS",
|
| 245 |
-
"85": "
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
"87": "SUCCESS",
|
| 248 |
"88": "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,
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "Joooorrit/19_hot",
|
| 3 |
"activation_function": "gelu_new",
|
| 4 |
"all_reduce_scores": {
|
| 5 |
"0": "NON_PARTICIPATING",
|
|
|
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
"104": "NON_PARTICIPATING",
|
| 13 |
"105": "NON_PARTICIPATING",
|
| 14 |
+
"106": "SUCCESS",
|
| 15 |
"107": "SUCCESS",
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
+
"109": "SUCCESS",
|
| 18 |
"11": "NON_PARTICIPATING",
|
| 19 |
"110": "SUCCESS",
|
| 20 |
"111": "NON_PARTICIPATING",
|
|
|
|
| 24 |
"115": "NON_PARTICIPATING",
|
| 25 |
"116": "NON_PARTICIPATING",
|
| 26 |
"117": "NON_PARTICIPATING",
|
| 27 |
+
"118": "SUCCESS",
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
"12": "SUCCESS",
|
| 30 |
+
"120": "SUCCESS",
|
| 31 |
"121": "SUCCESS",
|
| 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": "NON_PARTICIPATING",
|
| 39 |
+
"129": "SUCCESS",
|
| 40 |
"13": "NON_PARTICIPATING",
|
| 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": "NON_PARTICIPATING",
|
|
|
|
| 51 |
"14": "NON_PARTICIPATING",
|
| 52 |
"140": "NON_PARTICIPATING",
|
| 53 |
"141": "NON_PARTICIPATING",
|
| 54 |
+
"142": "NON_PARTICIPATING",
|
| 55 |
+
"143": "SUCCESS",
|
| 56 |
"144": "NON_PARTICIPATING",
|
| 57 |
"145": "NON_PARTICIPATING",
|
| 58 |
"146": "NON_PARTICIPATING",
|
| 59 |
+
"147": "NON_PARTICIPATING",
|
| 60 |
+
"148": "SUCCESS",
|
| 61 |
+
"149": "NON_PARTICIPATING",
|
| 62 |
"15": "SUCCESS",
|
| 63 |
+
"150": "SUCCESS",
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
"153": "NON_PARTICIPATING",
|
| 67 |
"154": "SUCCESS",
|
| 68 |
+
"155": "NON_PARTICIPATING",
|
| 69 |
+
"156": "SUCCESS",
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
+
"159": "SUCCESS",
|
| 73 |
"16": "SUCCESS",
|
| 74 |
"160": "SUCCESS",
|
| 75 |
"161": "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",
|
| 84 |
"17": "SUCCESS",
|
| 85 |
"170": "SUCCESS",
|
| 86 |
"171": "NON_PARTICIPATING",
|
| 87 |
+
"172": "NON_PARTICIPATING",
|
| 88 |
"173": "NON_PARTICIPATING",
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
+
"175": "NON_PARTICIPATING",
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
"177": "NON_PARTICIPATING",
|
| 93 |
"178": "NON_PARTICIPATING",
|
|
|
|
| 98 |
"182": "NON_PARTICIPATING",
|
| 99 |
"183": "NON_PARTICIPATING",
|
| 100 |
"184": "SUCCESS",
|
| 101 |
+
"185": "SUCCESS",
|
| 102 |
"186": "SUCCESS",
|
| 103 |
+
"187": "NON_PARTICIPATING",
|
| 104 |
"188": "NON_PARTICIPATING",
|
| 105 |
"189": "NON_PARTICIPATING",
|
| 106 |
"19": "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": "NON_PARTICIPATING",
|
| 125 |
"206": "NON_PARTICIPATING",
|
|
|
|
| 130 |
"210": "NON_PARTICIPATING",
|
| 131 |
"211": "NON_PARTICIPATING",
|
| 132 |
"212": "NON_PARTICIPATING",
|
| 133 |
+
"213": "SUCCESS",
|
| 134 |
"214": "SUCCESS",
|
| 135 |
+
"215": "NON_PARTICIPATING",
|
| 136 |
"216": "NON_PARTICIPATING",
|
| 137 |
"217": "NON_PARTICIPATING",
|
| 138 |
+
"218": "SUCCESS",
|
| 139 |
"219": "SUCCESS",
|
| 140 |
"22": "SUCCESS",
|
| 141 |
"220": "NON_PARTICIPATING",
|
| 142 |
+
"221": "NON_PARTICIPATING",
|
| 143 |
"222": "NON_PARTICIPATING",
|
| 144 |
+
"223": "SUCCESS",
|
| 145 |
"224": "NON_PARTICIPATING",
|
| 146 |
"225": "SUCCESS",
|
| 147 |
+
"226": "NON_PARTICIPATING",
|
| 148 |
"227": "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",
|
|
|
|
| 159 |
"237": "NON_PARTICIPATING",
|
| 160 |
"238": "NON_PARTICIPATING",
|
| 161 |
"239": "NON_PARTICIPATING",
|
| 162 |
+
"24": "NON_PARTICIPATING",
|
| 163 |
"240": "NON_PARTICIPATING",
|
| 164 |
+
"241": "NON_PARTICIPATING",
|
| 165 |
"242": "NON_PARTICIPATING",
|
| 166 |
"243": "NON_PARTICIPATING",
|
| 167 |
"244": "SUCCESS",
|
|
|
|
| 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",
|
|
|
|
| 184 |
"3": "SUCCESS",
|
| 185 |
"30": "NON_PARTICIPATING",
|
| 186 |
"31": "NON_PARTICIPATING",
|
| 187 |
+
"32": "NON_PARTICIPATING",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
+
"34": "NON_PARTICIPATING",
|
| 190 |
"35": "NON_PARTICIPATING",
|
| 191 |
"36": "NON_PARTICIPATING",
|
| 192 |
"37": "SUCCESS",
|
|
|
|
| 209 |
"52": "NON_PARTICIPATING",
|
| 210 |
"53": "NON_PARTICIPATING",
|
| 211 |
"54": "SUCCESS",
|
| 212 |
+
"55": "NON_PARTICIPATING",
|
| 213 |
"56": "SUCCESS",
|
| 214 |
"57": "SUCCESS",
|
| 215 |
"58": "SUCCESS",
|
| 216 |
"59": "NON_PARTICIPATING",
|
| 217 |
+
"6": "NON_PARTICIPATING",
|
| 218 |
"60": "SUCCESS",
|
| 219 |
+
"61": "NON_PARTICIPATING",
|
| 220 |
"62": "SUCCESS",
|
| 221 |
+
"63": "NON_PARTICIPATING",
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
"65": "SUCCESS",
|
| 224 |
"66": "NON_PARTICIPATING",
|
|
|
|
| 242 |
"82": "SUCCESS",
|
| 243 |
"83": "NON_PARTICIPATING",
|
| 244 |
"84": "SUCCESS",
|
| 245 |
+
"85": "NON_PARTICIPATING",
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
"87": "SUCCESS",
|
| 248 |
"88": "SUCCESS",
|
|
|
|
| 275 |
"initializer_range": 0.02,
|
| 276 |
"inner_step": 0,
|
| 277 |
"inner_steps": 0,
|
| 278 |
+
"last_allreduce_block": 5529444,
|
| 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:ec9885a5120cb0143cc71f757289b9ab5498f78e1945943fa0fa9cb7e63c8a39
|
| 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 sha256:8a256b0c2c1e4f1446ba82b204e63602b0cf61e85dcc97337a7b23b9edc5e72f
|
| 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:bfe92d8527ba211f71417e85eeb70e5f4e99eda1730a0978aee9e33ba635bb68
|
| 3 |
size 4040805354
|