Run 6. Outer Step 82. Inner Step 5.
Browse files- config.json +48 -48
- gradients.pt +1 -1
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
config.json
CHANGED
|
@@ -23,17 +23,17 @@
|
|
| 23 |
"115": "NON_PARTICIPATING",
|
| 24 |
"116": "NOT_ALIVE",
|
| 25 |
"117": "NOT_ALIVE",
|
| 26 |
-
"118": "
|
| 27 |
"119": "NON_PARTICIPATING",
|
| 28 |
"12": "NOT_ALIVE",
|
| 29 |
"120": "NOT_ALIVE",
|
| 30 |
"121": "NOT_ALIVE",
|
| 31 |
-
"122": "
|
| 32 |
"123": "NON_PARTICIPATING",
|
| 33 |
"124": "NON_PARTICIPATING",
|
| 34 |
"125": "NOT_ALIVE",
|
| 35 |
"126": "NOT_ALIVE",
|
| 36 |
-
"127": "
|
| 37 |
"128": "NON_PARTICIPATING",
|
| 38 |
"129": "NOT_ALIVE",
|
| 39 |
"13": "NOT_ALIVE",
|
|
@@ -43,40 +43,40 @@
|
|
| 43 |
"133": "NON_PARTICIPATING",
|
| 44 |
"134": "NOT_ALIVE",
|
| 45 |
"135": "NON_PARTICIPATING",
|
| 46 |
-
"136": "
|
| 47 |
"137": "NON_PARTICIPATING",
|
| 48 |
"138": "NON_PARTICIPATING",
|
| 49 |
"139": "NON_PARTICIPATING",
|
| 50 |
"14": "NOT_ALIVE",
|
| 51 |
-
"140": "
|
| 52 |
"141": "NON_PARTICIPATING",
|
| 53 |
-
"142": "
|
| 54 |
"143": "NOT_ALIVE",
|
| 55 |
"144": "SUCCESS",
|
| 56 |
"145": "NOT_ALIVE",
|
| 57 |
"146": "NON_PARTICIPATING",
|
| 58 |
"147": "NOT_ALIVE",
|
| 59 |
-
"148": "
|
| 60 |
-
"149": "
|
| 61 |
-
"15": "
|
| 62 |
"150": "NOT_ALIVE",
|
| 63 |
"151": "NON_PARTICIPATING",
|
| 64 |
"152": "NOT_ALIVE",
|
| 65 |
"153": "NOT_ALIVE",
|
| 66 |
-
"154": "
|
| 67 |
-
"155": "
|
| 68 |
"156": "NOT_ALIVE",
|
| 69 |
"157": "NOT_ALIVE",
|
| 70 |
"158": "NON_PARTICIPATING",
|
| 71 |
-
"159": "
|
| 72 |
"16": "NOT_ALIVE",
|
| 73 |
"160": "NON_PARTICIPATING",
|
| 74 |
"161": "NON_PARTICIPATING",
|
| 75 |
"162": "NON_PARTICIPATING",
|
| 76 |
-
"163": "
|
| 77 |
-
"164": "
|
| 78 |
"165": "NON_PARTICIPATING",
|
| 79 |
-
"166": "
|
| 80 |
"167": "NOT_ALIVE",
|
| 81 |
"168": "NON_PARTICIPATING",
|
| 82 |
"169": "NON_PARTICIPATING",
|
|
@@ -92,13 +92,13 @@
|
|
| 92 |
"178": "NON_PARTICIPATING",
|
| 93 |
"179": "NON_PARTICIPATING",
|
| 94 |
"18": "NOT_ALIVE",
|
| 95 |
-
"180": "
|
| 96 |
"181": "NOT_ALIVE",
|
| 97 |
-
"182": "
|
| 98 |
-
"183": "
|
| 99 |
"184": "NON_PARTICIPATING",
|
| 100 |
"185": "NON_PARTICIPATING",
|
| 101 |
-
"186": "
|
| 102 |
"187": "NOT_ALIVE",
|
| 103 |
"188": "NON_PARTICIPATING",
|
| 104 |
"189": "NON_PARTICIPATING",
|
|
@@ -106,7 +106,7 @@
|
|
| 106 |
"190": "NOT_ALIVE",
|
| 107 |
"191": "SUCCESS",
|
| 108 |
"192": "NON_PARTICIPATING",
|
| 109 |
-
"193": "
|
| 110 |
"194": "NON_PARTICIPATING",
|
| 111 |
"195": "NON_PARTICIPATING",
|
| 112 |
"196": "NOT_ALIVE",
|
|
@@ -120,7 +120,7 @@
|
|
| 120 |
"202": "NOT_ALIVE",
|
| 121 |
"203": "NON_PARTICIPATING",
|
| 122 |
"204": "NON_PARTICIPATING",
|
| 123 |
-
"205": "
|
| 124 |
"206": "NON_PARTICIPATING",
|
| 125 |
"207": "NOT_ALIVE",
|
| 126 |
"208": "NOT_ALIVE",
|
|
@@ -133,22 +133,22 @@
|
|
| 133 |
"214": "NON_PARTICIPATING",
|
| 134 |
"215": "NON_PARTICIPATING",
|
| 135 |
"216": "NON_PARTICIPATING",
|
| 136 |
-
"217": "
|
| 137 |
"218": "NON_PARTICIPATING",
|
| 138 |
-
"219": "
|
| 139 |
"22": "NOT_ALIVE",
|
| 140 |
"220": "NOT_ALIVE",
|
| 141 |
"221": "NOT_ALIVE",
|
| 142 |
"222": "NOT_ALIVE",
|
| 143 |
"223": "NOT_ALIVE",
|
| 144 |
"224": "NON_PARTICIPATING",
|
| 145 |
-
"225": "
|
| 146 |
-
"226": "
|
| 147 |
"227": "NOT_ALIVE",
|
| 148 |
"228": "NON_PARTICIPATING",
|
| 149 |
"229": "NOT_ALIVE",
|
| 150 |
"23": "NOT_ALIVE",
|
| 151 |
-
"230": "
|
| 152 |
"231": "NOT_ALIVE",
|
| 153 |
"232": "NOT_ALIVE",
|
| 154 |
"233": "NOT_ALIVE",
|
|
@@ -161,11 +161,11 @@
|
|
| 161 |
"24": "NOT_ALIVE",
|
| 162 |
"240": "NOT_ALIVE",
|
| 163 |
"241": "NON_PARTICIPATING",
|
| 164 |
-
"242": "
|
| 165 |
"243": "NOT_ALIVE",
|
| 166 |
"244": "NON_PARTICIPATING",
|
| 167 |
"245": "NOT_ALIVE",
|
| 168 |
-
"246": "
|
| 169 |
"247": "NON_PARTICIPATING",
|
| 170 |
"248": "NOT_ALIVE",
|
| 171 |
"249": "NOT_ALIVE",
|
|
@@ -173,13 +173,13 @@
|
|
| 173 |
"250": "NON_PARTICIPATING",
|
| 174 |
"251": "NOT_ALIVE",
|
| 175 |
"252": "NOT_ALIVE",
|
| 176 |
-
"253": "
|
| 177 |
"254": "NON_PARTICIPATING",
|
| 178 |
-
"255": "
|
| 179 |
"26": "NON_PARTICIPATING",
|
| 180 |
"27": "NON_PARTICIPATING",
|
| 181 |
-
"28": "
|
| 182 |
-
"29": "
|
| 183 |
"3": "NOT_ALIVE",
|
| 184 |
"30": "NOT_ALIVE",
|
| 185 |
"31": "NOT_ALIVE",
|
|
@@ -187,16 +187,16 @@
|
|
| 187 |
"33": "NON_PARTICIPATING",
|
| 188 |
"34": "NON_PARTICIPATING",
|
| 189 |
"35": "NON_PARTICIPATING",
|
| 190 |
-
"36": "
|
| 191 |
"37": "NON_PARTICIPATING",
|
| 192 |
"38": "NOT_ALIVE",
|
| 193 |
"39": "NOT_ALIVE",
|
| 194 |
"4": "NOT_ALIVE",
|
| 195 |
"40": "NON_PARTICIPATING",
|
| 196 |
"41": "NON_PARTICIPATING",
|
| 197 |
-
"42": "
|
| 198 |
"43": "NON_PARTICIPATING",
|
| 199 |
-
"44": "
|
| 200 |
"45": "NON_PARTICIPATING",
|
| 201 |
"46": "NON_PARTICIPATING",
|
| 202 |
"47": "NOT_ALIVE",
|
|
@@ -208,11 +208,11 @@
|
|
| 208 |
"52": "NOT_ALIVE",
|
| 209 |
"53": "NOT_ALIVE",
|
| 210 |
"54": "NOT_ALIVE",
|
| 211 |
-
"55": "
|
| 212 |
"56": "NON_PARTICIPATING",
|
| 213 |
"57": "NON_PARTICIPATING",
|
| 214 |
"58": "NOT_ALIVE",
|
| 215 |
-
"59": "
|
| 216 |
"6": "NOT_ALIVE",
|
| 217 |
"60": "NON_PARTICIPATING",
|
| 218 |
"61": "NOT_ALIVE",
|
|
@@ -232,17 +232,17 @@
|
|
| 232 |
"74": "NOT_ALIVE",
|
| 233 |
"75": "NON_PARTICIPATING",
|
| 234 |
"76": "NOT_ALIVE",
|
| 235 |
-
"77": "
|
| 236 |
-
"78": "
|
| 237 |
"79": "NOT_ALIVE",
|
| 238 |
"8": "NOT_ALIVE",
|
| 239 |
"80": "NOT_ALIVE",
|
| 240 |
-
"81": "
|
| 241 |
"82": "NON_PARTICIPATING",
|
| 242 |
-
"83": "
|
| 243 |
"84": "NON_PARTICIPATING",
|
| 244 |
"85": "NOT_ALIVE",
|
| 245 |
-
"86": "
|
| 246 |
"87": "NOT_ALIVE",
|
| 247 |
"88": "NOT_ALIVE",
|
| 248 |
"89": "NOT_ALIVE",
|
|
@@ -251,12 +251,12 @@
|
|
| 251 |
"91": "NOT_ALIVE",
|
| 252 |
"92": "NON_PARTICIPATING",
|
| 253 |
"93": "NOT_ALIVE",
|
| 254 |
-
"94": "
|
| 255 |
"95": "NON_PARTICIPATING",
|
| 256 |
"96": "NON_PARTICIPATING",
|
| 257 |
"97": "NOT_ALIVE",
|
| 258 |
"98": "NON_PARTICIPATING",
|
| 259 |
-
"99": "
|
| 260 |
},
|
| 261 |
"architectures": [
|
| 262 |
"LlamaForCausalLM"
|
|
@@ -264,17 +264,17 @@
|
|
| 264 |
"attention_bias": false,
|
| 265 |
"attention_dropout": 0.0,
|
| 266 |
"block_list": [
|
| 267 |
-
|
| 268 |
-
|
| 269 |
],
|
| 270 |
"bos_token_id": 1,
|
| 271 |
"eos_token_id": 2,
|
| 272 |
"hidden_act": "silu",
|
| 273 |
"hidden_size": 2048,
|
| 274 |
"initializer_range": 0.02,
|
| 275 |
-
"inner_step":
|
| 276 |
"intermediate_size": 5632,
|
| 277 |
-
"last_allreduce_block":
|
| 278 |
"max_position_embeddings": 2048,
|
| 279 |
"mlp_bias": false,
|
| 280 |
"model_type": "llama",
|
|
|
|
| 23 |
"115": "NON_PARTICIPATING",
|
| 24 |
"116": "NOT_ALIVE",
|
| 25 |
"117": "NOT_ALIVE",
|
| 26 |
+
"118": "NON_PARTICIPATING",
|
| 27 |
"119": "NON_PARTICIPATING",
|
| 28 |
"12": "NOT_ALIVE",
|
| 29 |
"120": "NOT_ALIVE",
|
| 30 |
"121": "NOT_ALIVE",
|
| 31 |
+
"122": "NON_PARTICIPATING",
|
| 32 |
"123": "NON_PARTICIPATING",
|
| 33 |
"124": "NON_PARTICIPATING",
|
| 34 |
"125": "NOT_ALIVE",
|
| 35 |
"126": "NOT_ALIVE",
|
| 36 |
+
"127": "NON_PARTICIPATING",
|
| 37 |
"128": "NON_PARTICIPATING",
|
| 38 |
"129": "NOT_ALIVE",
|
| 39 |
"13": "NOT_ALIVE",
|
|
|
|
| 43 |
"133": "NON_PARTICIPATING",
|
| 44 |
"134": "NOT_ALIVE",
|
| 45 |
"135": "NON_PARTICIPATING",
|
| 46 |
+
"136": "SUCCESS",
|
| 47 |
"137": "NON_PARTICIPATING",
|
| 48 |
"138": "NON_PARTICIPATING",
|
| 49 |
"139": "NON_PARTICIPATING",
|
| 50 |
"14": "NOT_ALIVE",
|
| 51 |
+
"140": "NON_PARTICIPATING",
|
| 52 |
"141": "NON_PARTICIPATING",
|
| 53 |
+
"142": "NON_PARTICIPATING",
|
| 54 |
"143": "NOT_ALIVE",
|
| 55 |
"144": "SUCCESS",
|
| 56 |
"145": "NOT_ALIVE",
|
| 57 |
"146": "NON_PARTICIPATING",
|
| 58 |
"147": "NOT_ALIVE",
|
| 59 |
+
"148": "NON_PARTICIPATING",
|
| 60 |
+
"149": "NOT_ALIVE",
|
| 61 |
+
"15": "SUCCESS",
|
| 62 |
"150": "NOT_ALIVE",
|
| 63 |
"151": "NON_PARTICIPATING",
|
| 64 |
"152": "NOT_ALIVE",
|
| 65 |
"153": "NOT_ALIVE",
|
| 66 |
+
"154": "NON_PARTICIPATING",
|
| 67 |
+
"155": "NON_PARTICIPATING",
|
| 68 |
"156": "NOT_ALIVE",
|
| 69 |
"157": "NOT_ALIVE",
|
| 70 |
"158": "NON_PARTICIPATING",
|
| 71 |
+
"159": "NON_PARTICIPATING",
|
| 72 |
"16": "NOT_ALIVE",
|
| 73 |
"160": "NON_PARTICIPATING",
|
| 74 |
"161": "NON_PARTICIPATING",
|
| 75 |
"162": "NON_PARTICIPATING",
|
| 76 |
+
"163": "NOT_ALIVE",
|
| 77 |
+
"164": "NON_PARTICIPATING",
|
| 78 |
"165": "NON_PARTICIPATING",
|
| 79 |
+
"166": "SUCCESS",
|
| 80 |
"167": "NOT_ALIVE",
|
| 81 |
"168": "NON_PARTICIPATING",
|
| 82 |
"169": "NON_PARTICIPATING",
|
|
|
|
| 92 |
"178": "NON_PARTICIPATING",
|
| 93 |
"179": "NON_PARTICIPATING",
|
| 94 |
"18": "NOT_ALIVE",
|
| 95 |
+
"180": "NOT_ALIVE",
|
| 96 |
"181": "NOT_ALIVE",
|
| 97 |
+
"182": "NON_PARTICIPATING",
|
| 98 |
+
"183": "NOT_ALIVE",
|
| 99 |
"184": "NON_PARTICIPATING",
|
| 100 |
"185": "NON_PARTICIPATING",
|
| 101 |
+
"186": "NON_PARTICIPATING",
|
| 102 |
"187": "NOT_ALIVE",
|
| 103 |
"188": "NON_PARTICIPATING",
|
| 104 |
"189": "NON_PARTICIPATING",
|
|
|
|
| 106 |
"190": "NOT_ALIVE",
|
| 107 |
"191": "SUCCESS",
|
| 108 |
"192": "NON_PARTICIPATING",
|
| 109 |
+
"193": "SUCCESS",
|
| 110 |
"194": "NON_PARTICIPATING",
|
| 111 |
"195": "NON_PARTICIPATING",
|
| 112 |
"196": "NOT_ALIVE",
|
|
|
|
| 120 |
"202": "NOT_ALIVE",
|
| 121 |
"203": "NON_PARTICIPATING",
|
| 122 |
"204": "NON_PARTICIPATING",
|
| 123 |
+
"205": "SUCCESS",
|
| 124 |
"206": "NON_PARTICIPATING",
|
| 125 |
"207": "NOT_ALIVE",
|
| 126 |
"208": "NOT_ALIVE",
|
|
|
|
| 133 |
"214": "NON_PARTICIPATING",
|
| 134 |
"215": "NON_PARTICIPATING",
|
| 135 |
"216": "NON_PARTICIPATING",
|
| 136 |
+
"217": "NON_PARTICIPATING",
|
| 137 |
"218": "NON_PARTICIPATING",
|
| 138 |
+
"219": "NON_PARTICIPATING",
|
| 139 |
"22": "NOT_ALIVE",
|
| 140 |
"220": "NOT_ALIVE",
|
| 141 |
"221": "NOT_ALIVE",
|
| 142 |
"222": "NOT_ALIVE",
|
| 143 |
"223": "NOT_ALIVE",
|
| 144 |
"224": "NON_PARTICIPATING",
|
| 145 |
+
"225": "NON_PARTICIPATING",
|
| 146 |
+
"226": "NOT_ALIVE",
|
| 147 |
"227": "NOT_ALIVE",
|
| 148 |
"228": "NON_PARTICIPATING",
|
| 149 |
"229": "NOT_ALIVE",
|
| 150 |
"23": "NOT_ALIVE",
|
| 151 |
+
"230": "NON_PARTICIPATING",
|
| 152 |
"231": "NOT_ALIVE",
|
| 153 |
"232": "NOT_ALIVE",
|
| 154 |
"233": "NOT_ALIVE",
|
|
|
|
| 161 |
"24": "NOT_ALIVE",
|
| 162 |
"240": "NOT_ALIVE",
|
| 163 |
"241": "NON_PARTICIPATING",
|
| 164 |
+
"242": "NON_PARTICIPATING",
|
| 165 |
"243": "NOT_ALIVE",
|
| 166 |
"244": "NON_PARTICIPATING",
|
| 167 |
"245": "NOT_ALIVE",
|
| 168 |
+
"246": "SUCCESS",
|
| 169 |
"247": "NON_PARTICIPATING",
|
| 170 |
"248": "NOT_ALIVE",
|
| 171 |
"249": "NOT_ALIVE",
|
|
|
|
| 173 |
"250": "NON_PARTICIPATING",
|
| 174 |
"251": "NOT_ALIVE",
|
| 175 |
"252": "NOT_ALIVE",
|
| 176 |
+
"253": "NOT_ALIVE",
|
| 177 |
"254": "NON_PARTICIPATING",
|
| 178 |
+
"255": "NOT_ALIVE",
|
| 179 |
"26": "NON_PARTICIPATING",
|
| 180 |
"27": "NON_PARTICIPATING",
|
| 181 |
+
"28": "NON_PARTICIPATING",
|
| 182 |
+
"29": "NOT_ALIVE",
|
| 183 |
"3": "NOT_ALIVE",
|
| 184 |
"30": "NOT_ALIVE",
|
| 185 |
"31": "NOT_ALIVE",
|
|
|
|
| 187 |
"33": "NON_PARTICIPATING",
|
| 188 |
"34": "NON_PARTICIPATING",
|
| 189 |
"35": "NON_PARTICIPATING",
|
| 190 |
+
"36": "NOT_ALIVE",
|
| 191 |
"37": "NON_PARTICIPATING",
|
| 192 |
"38": "NOT_ALIVE",
|
| 193 |
"39": "NOT_ALIVE",
|
| 194 |
"4": "NOT_ALIVE",
|
| 195 |
"40": "NON_PARTICIPATING",
|
| 196 |
"41": "NON_PARTICIPATING",
|
| 197 |
+
"42": "NOT_ALIVE",
|
| 198 |
"43": "NON_PARTICIPATING",
|
| 199 |
+
"44": "NON_PARTICIPATING",
|
| 200 |
"45": "NON_PARTICIPATING",
|
| 201 |
"46": "NON_PARTICIPATING",
|
| 202 |
"47": "NOT_ALIVE",
|
|
|
|
| 208 |
"52": "NOT_ALIVE",
|
| 209 |
"53": "NOT_ALIVE",
|
| 210 |
"54": "NOT_ALIVE",
|
| 211 |
+
"55": "NOT_ALIVE",
|
| 212 |
"56": "NON_PARTICIPATING",
|
| 213 |
"57": "NON_PARTICIPATING",
|
| 214 |
"58": "NOT_ALIVE",
|
| 215 |
+
"59": "NOT_ALIVE",
|
| 216 |
"6": "NOT_ALIVE",
|
| 217 |
"60": "NON_PARTICIPATING",
|
| 218 |
"61": "NOT_ALIVE",
|
|
|
|
| 232 |
"74": "NOT_ALIVE",
|
| 233 |
"75": "NON_PARTICIPATING",
|
| 234 |
"76": "NOT_ALIVE",
|
| 235 |
+
"77": "NON_PARTICIPATING",
|
| 236 |
+
"78": "NON_PARTICIPATING",
|
| 237 |
"79": "NOT_ALIVE",
|
| 238 |
"8": "NOT_ALIVE",
|
| 239 |
"80": "NOT_ALIVE",
|
| 240 |
+
"81": "NOT_ALIVE",
|
| 241 |
"82": "NON_PARTICIPATING",
|
| 242 |
+
"83": "NON_PARTICIPATING",
|
| 243 |
"84": "NON_PARTICIPATING",
|
| 244 |
"85": "NOT_ALIVE",
|
| 245 |
+
"86": "NOT_ALIVE",
|
| 246 |
"87": "NOT_ALIVE",
|
| 247 |
"88": "NOT_ALIVE",
|
| 248 |
"89": "NOT_ALIVE",
|
|
|
|
| 251 |
"91": "NOT_ALIVE",
|
| 252 |
"92": "NON_PARTICIPATING",
|
| 253 |
"93": "NOT_ALIVE",
|
| 254 |
+
"94": "NOT_ALIVE",
|
| 255 |
"95": "NON_PARTICIPATING",
|
| 256 |
"96": "NON_PARTICIPATING",
|
| 257 |
"97": "NOT_ALIVE",
|
| 258 |
"98": "NON_PARTICIPATING",
|
| 259 |
+
"99": "NON_PARTICIPATING"
|
| 260 |
},
|
| 261 |
"architectures": [
|
| 262 |
"LlamaForCausalLM"
|
|
|
|
| 264 |
"attention_bias": false,
|
| 265 |
"attention_dropout": 0.0,
|
| 266 |
"block_list": [
|
| 267 |
+
6186902,
|
| 268 |
+
6186917
|
| 269 |
],
|
| 270 |
"bos_token_id": 1,
|
| 271 |
"eos_token_id": 2,
|
| 272 |
"hidden_act": "silu",
|
| 273 |
"hidden_size": 2048,
|
| 274 |
"initializer_range": 0.02,
|
| 275 |
+
"inner_step": 4,
|
| 276 |
"intermediate_size": 5632,
|
| 277 |
+
"last_allreduce_block": 6186771,
|
| 278 |
"max_position_embeddings": 2048,
|
| 279 |
"mlp_bias": false,
|
| 280 |
"model_type": "llama",
|
gradients.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 88389626
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1775b6450771a6e7c0ead691fcbad67930a06c72c4f44cd66fdde217e0cb44cf
|
| 3 |
size 88389626
|
inner_optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 8800564076
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6e0c165892f1f2dcbac53062cb877bf58ea819c690f6dd936d2ff744feb5e821
|
| 3 |
size 8800564076
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4400216536
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea82d902288e9a68f562a21b85fa3c41b2e1dbb4ce8d768eba25b0d35f27e0e8
|
| 3 |
size 4400216536
|