Run 4. Outer Step 60. Inner Step 0.
Browse files- config.json +46 -47
- inner_optimizer.pt +1 -1
- model.safetensors +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",
|
|
@@ -11,13 +11,13 @@
|
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
"104": "NON_PARTICIPATING",
|
| 13 |
"105": "NON_PARTICIPATING",
|
| 14 |
-
"106": "
|
| 15 |
-
"107": "
|
| 16 |
-
"108": "
|
| 17 |
"109": "NON_PARTICIPATING",
|
| 18 |
"11": "SUCCESS",
|
| 19 |
"110": "NON_PARTICIPATING",
|
| 20 |
-
"111": "
|
| 21 |
"112": "NON_PARTICIPATING",
|
| 22 |
"113": "NON_PARTICIPATING",
|
| 23 |
"114": "NON_PARTICIPATING",
|
|
@@ -30,15 +30,15 @@
|
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 31 |
"121": "SUCCESS",
|
| 32 |
"122": "NON_PARTICIPATING",
|
| 33 |
-
"123": "
|
| 34 |
-
"124": "
|
| 35 |
"125": "NON_PARTICIPATING",
|
| 36 |
"126": "NON_PARTICIPATING",
|
| 37 |
"127": "NON_PARTICIPATING",
|
| 38 |
"128": "NON_PARTICIPATING",
|
| 39 |
"129": "NON_PARTICIPATING",
|
| 40 |
"13": "SUCCESS",
|
| 41 |
-
"130": "
|
| 42 |
"131": "NON_PARTICIPATING",
|
| 43 |
"132": "NON_PARTICIPATING",
|
| 44 |
"133": "NON_PARTICIPATING",
|
|
@@ -52,18 +52,18 @@
|
|
| 52 |
"140": "NON_PARTICIPATING",
|
| 53 |
"141": "NON_PARTICIPATING",
|
| 54 |
"142": "NON_PARTICIPATING",
|
| 55 |
-
"143": "
|
| 56 |
"144": "NON_PARTICIPATING",
|
| 57 |
"145": "NON_PARTICIPATING",
|
| 58 |
"146": "NON_PARTICIPATING",
|
| 59 |
-
"147": "
|
| 60 |
-
"148": "
|
| 61 |
"149": "NON_PARTICIPATING",
|
| 62 |
"15": "SUCCESS",
|
| 63 |
"150": "NON_PARTICIPATING",
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
-
"153": "
|
| 67 |
"154": "SUCCESS",
|
| 68 |
"155": "NON_PARTICIPATING",
|
| 69 |
"156": "NON_PARTICIPATING",
|
|
@@ -89,16 +89,16 @@
|
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
"175": "NON_PARTICIPATING",
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
-
"177": "
|
| 93 |
-
"178": "
|
| 94 |
"179": "NON_PARTICIPATING",
|
| 95 |
"18": "SUCCESS",
|
| 96 |
"180": "NON_PARTICIPATING",
|
| 97 |
-
"181": "
|
| 98 |
"182": "NON_PARTICIPATING",
|
| 99 |
"183": "NON_PARTICIPATING",
|
| 100 |
"184": "NON_PARTICIPATING",
|
| 101 |
-
"185": "
|
| 102 |
"186": "NON_PARTICIPATING",
|
| 103 |
"187": "NON_PARTICIPATING",
|
| 104 |
"188": "NON_PARTICIPATING",
|
|
@@ -113,7 +113,7 @@
|
|
| 113 |
"196": "NON_PARTICIPATING",
|
| 114 |
"197": "NON_PARTICIPATING",
|
| 115 |
"198": "NON_PARTICIPATING",
|
| 116 |
-
"199": "
|
| 117 |
"2": "SUCCESS",
|
| 118 |
"20": "SUCCESS",
|
| 119 |
"200": "NON_PARTICIPATING",
|
|
@@ -125,9 +125,9 @@
|
|
| 125 |
"206": "NON_PARTICIPATING",
|
| 126 |
"207": "NON_PARTICIPATING",
|
| 127 |
"208": "NON_PARTICIPATING",
|
| 128 |
-
"209": "
|
| 129 |
"21": "SUCCESS",
|
| 130 |
-
"210": "
|
| 131 |
"211": "NON_PARTICIPATING",
|
| 132 |
"212": "NON_PARTICIPATING",
|
| 133 |
"213": "NON_PARTICIPATING",
|
|
@@ -135,10 +135,10 @@
|
|
| 135 |
"215": "NON_PARTICIPATING",
|
| 136 |
"216": "NON_PARTICIPATING",
|
| 137 |
"217": "NON_PARTICIPATING",
|
| 138 |
-
"218": "
|
| 139 |
-
"219": "
|
| 140 |
"22": "SUCCESS",
|
| 141 |
-
"220": "
|
| 142 |
"221": "NON_PARTICIPATING",
|
| 143 |
"222": "NON_PARTICIPATING",
|
| 144 |
"223": "NON_PARTICIPATING",
|
|
@@ -147,27 +147,27 @@
|
|
| 147 |
"226": "NON_PARTICIPATING",
|
| 148 |
"227": "NON_PARTICIPATING",
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
-
"229": "
|
| 151 |
"23": "SUCCESS",
|
| 152 |
-
"230": "
|
| 153 |
"231": "NON_PARTICIPATING",
|
| 154 |
-
"232": "
|
| 155 |
"233": "NON_PARTICIPATING",
|
| 156 |
"234": "NON_PARTICIPATING",
|
| 157 |
"235": "NON_PARTICIPATING",
|
| 158 |
"236": "NON_PARTICIPATING",
|
| 159 |
"237": "NON_PARTICIPATING",
|
| 160 |
-
"238": "
|
| 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": "NON_PARTICIPATING",
|
| 173 |
"25": "SUCCESS",
|
|
@@ -176,18 +176,18 @@
|
|
| 176 |
"252": "NON_PARTICIPATING",
|
| 177 |
"253": "NON_PARTICIPATING",
|
| 178 |
"254": "NON_PARTICIPATING",
|
| 179 |
-
"255": "
|
| 180 |
"26": "SUCCESS",
|
| 181 |
"27": "SUCCESS",
|
| 182 |
"28": "SUCCESS",
|
| 183 |
"29": "SUCCESS",
|
| 184 |
"3": "SUCCESS",
|
| 185 |
"30": "NON_PARTICIPATING",
|
| 186 |
-
"31": "
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "SUCCESS",
|
| 190 |
-
"35": "
|
| 191 |
"36": "SUCCESS",
|
| 192 |
"37": "SUCCESS",
|
| 193 |
"38": "SUCCESS",
|
|
@@ -209,8 +209,8 @@
|
|
| 209 |
"52": "NON_PARTICIPATING",
|
| 210 |
"53": "SUCCESS",
|
| 211 |
"54": "SUCCESS",
|
| 212 |
-
"55": "
|
| 213 |
-
"56": "
|
| 214 |
"57": "SUCCESS",
|
| 215 |
"58": "NON_PARTICIPATING",
|
| 216 |
"59": "SUCCESS",
|
|
@@ -218,9 +218,9 @@
|
|
| 218 |
"60": "SUCCESS",
|
| 219 |
"61": "NON_PARTICIPATING",
|
| 220 |
"62": "SUCCESS",
|
| 221 |
-
"63": "
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
-
"65": "
|
| 224 |
"66": "SUCCESS",
|
| 225 |
"67": "SUCCESS",
|
| 226 |
"68": "SUCCESS",
|
|
@@ -233,8 +233,8 @@
|
|
| 233 |
"74": "SUCCESS",
|
| 234 |
"75": "SUCCESS",
|
| 235 |
"76": "SUCCESS",
|
| 236 |
-
"77": "
|
| 237 |
-
"78": "
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "SUCCESS",
|
| 240 |
"80": "SUCCESS",
|
|
@@ -244,9 +244,9 @@
|
|
| 244 |
"84": "SUCCESS",
|
| 245 |
"85": "NON_PARTICIPATING",
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
-
"87": "
|
| 248 |
"88": "SUCCESS",
|
| 249 |
-
"89": "
|
| 250 |
"9": "SUCCESS",
|
| 251 |
"90": "SUCCESS",
|
| 252 |
"91": "SUCCESS",
|
|
@@ -268,21 +268,20 @@
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
-
|
| 272 |
-
|
| 273 |
-
|
| 274 |
-
|
| 275 |
-
|
| 276 |
-
5607178
|
| 277 |
],
|
| 278 |
"block_size": 1024,
|
| 279 |
"bos_token_id": 50256,
|
| 280 |
"embd_pdrop": 0.1,
|
| 281 |
"eos_token_id": 50256,
|
| 282 |
"initializer_range": 0.02,
|
| 283 |
-
"inner_step":
|
| 284 |
"inner_steps": 0,
|
| 285 |
-
"last_allreduce_block":
|
| 286 |
"layer_norm_epsilon": 1e-05,
|
| 287 |
"model_type": "gpt_optimized",
|
| 288 |
"n_embd": 1280,
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "distributed/optimized-gpt2-1b",
|
| 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": "SUCCESS",
|
| 17 |
"109": "NON_PARTICIPATING",
|
| 18 |
"11": "SUCCESS",
|
| 19 |
"110": "NON_PARTICIPATING",
|
| 20 |
+
"111": "SUCCESS",
|
| 21 |
"112": "NON_PARTICIPATING",
|
| 22 |
"113": "NON_PARTICIPATING",
|
| 23 |
"114": "NON_PARTICIPATING",
|
|
|
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 31 |
"121": "SUCCESS",
|
| 32 |
"122": "NON_PARTICIPATING",
|
| 33 |
+
"123": "NON_PARTICIPATING",
|
| 34 |
+
"124": "SUCCESS",
|
| 35 |
"125": "NON_PARTICIPATING",
|
| 36 |
"126": "NON_PARTICIPATING",
|
| 37 |
"127": "NON_PARTICIPATING",
|
| 38 |
"128": "NON_PARTICIPATING",
|
| 39 |
"129": "NON_PARTICIPATING",
|
| 40 |
"13": "SUCCESS",
|
| 41 |
+
"130": "NON_PARTICIPATING",
|
| 42 |
"131": "NON_PARTICIPATING",
|
| 43 |
"132": "NON_PARTICIPATING",
|
| 44 |
"133": "NON_PARTICIPATING",
|
|
|
|
| 52 |
"140": "NON_PARTICIPATING",
|
| 53 |
"141": "NON_PARTICIPATING",
|
| 54 |
"142": "NON_PARTICIPATING",
|
| 55 |
+
"143": "NON_PARTICIPATING",
|
| 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": "NON_PARTICIPATING",
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
+
"153": "NON_PARTICIPATING",
|
| 67 |
"154": "SUCCESS",
|
| 68 |
"155": "NON_PARTICIPATING",
|
| 69 |
"156": "NON_PARTICIPATING",
|
|
|
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
"175": "NON_PARTICIPATING",
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
+
"177": "NON_PARTICIPATING",
|
| 93 |
+
"178": "SUCCESS",
|
| 94 |
"179": "NON_PARTICIPATING",
|
| 95 |
"18": "SUCCESS",
|
| 96 |
"180": "NON_PARTICIPATING",
|
| 97 |
+
"181": "SUCCESS",
|
| 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",
|
|
|
|
| 113 |
"196": "NON_PARTICIPATING",
|
| 114 |
"197": "NON_PARTICIPATING",
|
| 115 |
"198": "NON_PARTICIPATING",
|
| 116 |
+
"199": "SUCCESS",
|
| 117 |
"2": "SUCCESS",
|
| 118 |
"20": "SUCCESS",
|
| 119 |
"200": "NON_PARTICIPATING",
|
|
|
|
| 125 |
"206": "NON_PARTICIPATING",
|
| 126 |
"207": "NON_PARTICIPATING",
|
| 127 |
"208": "NON_PARTICIPATING",
|
| 128 |
+
"209": "SUCCESS",
|
| 129 |
"21": "SUCCESS",
|
| 130 |
+
"210": "SUCCESS",
|
| 131 |
"211": "NON_PARTICIPATING",
|
| 132 |
"212": "NON_PARTICIPATING",
|
| 133 |
"213": "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",
|
| 143 |
"222": "NON_PARTICIPATING",
|
| 144 |
"223": "NON_PARTICIPATING",
|
|
|
|
| 147 |
"226": "NON_PARTICIPATING",
|
| 148 |
"227": "NON_PARTICIPATING",
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
+
"229": "NON_PARTICIPATING",
|
| 151 |
"23": "SUCCESS",
|
| 152 |
+
"230": "NON_PARTICIPATING",
|
| 153 |
"231": "NON_PARTICIPATING",
|
| 154 |
+
"232": "SUCCESS",
|
| 155 |
"233": "NON_PARTICIPATING",
|
| 156 |
"234": "NON_PARTICIPATING",
|
| 157 |
"235": "NON_PARTICIPATING",
|
| 158 |
"236": "NON_PARTICIPATING",
|
| 159 |
"237": "NON_PARTICIPATING",
|
| 160 |
+
"238": "NON_PARTICIPATING",
|
| 161 |
"239": "NON_PARTICIPATING",
|
| 162 |
"24": "SUCCESS",
|
| 163 |
"240": "NON_PARTICIPATING",
|
| 164 |
+
"241": "SUCCESS",
|
| 165 |
"242": "NON_PARTICIPATING",
|
| 166 |
"243": "NON_PARTICIPATING",
|
| 167 |
"244": "NON_PARTICIPATING",
|
| 168 |
"245": "NON_PARTICIPATING",
|
| 169 |
"246": "NON_PARTICIPATING",
|
| 170 |
+
"247": "NON_PARTICIPATING",
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
"249": "NON_PARTICIPATING",
|
| 173 |
"25": "SUCCESS",
|
|
|
|
| 176 |
"252": "NON_PARTICIPATING",
|
| 177 |
"253": "NON_PARTICIPATING",
|
| 178 |
"254": "NON_PARTICIPATING",
|
| 179 |
+
"255": "NON_PARTICIPATING",
|
| 180 |
"26": "SUCCESS",
|
| 181 |
"27": "SUCCESS",
|
| 182 |
"28": "SUCCESS",
|
| 183 |
"29": "SUCCESS",
|
| 184 |
"3": "SUCCESS",
|
| 185 |
"30": "NON_PARTICIPATING",
|
| 186 |
+
"31": "NON_PARTICIPATING",
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "SUCCESS",
|
| 190 |
+
"35": "SUCCESS",
|
| 191 |
"36": "SUCCESS",
|
| 192 |
"37": "SUCCESS",
|
| 193 |
"38": "SUCCESS",
|
|
|
|
| 209 |
"52": "NON_PARTICIPATING",
|
| 210 |
"53": "SUCCESS",
|
| 211 |
"54": "SUCCESS",
|
| 212 |
+
"55": "NON_PARTICIPATING",
|
| 213 |
+
"56": "SUCCESS",
|
| 214 |
"57": "SUCCESS",
|
| 215 |
"58": "NON_PARTICIPATING",
|
| 216 |
"59": "SUCCESS",
|
|
|
|
| 218 |
"60": "SUCCESS",
|
| 219 |
"61": "NON_PARTICIPATING",
|
| 220 |
"62": "SUCCESS",
|
| 221 |
+
"63": "SUCCESS",
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
+
"65": "SUCCESS",
|
| 224 |
"66": "SUCCESS",
|
| 225 |
"67": "SUCCESS",
|
| 226 |
"68": "SUCCESS",
|
|
|
|
| 233 |
"74": "SUCCESS",
|
| 234 |
"75": "SUCCESS",
|
| 235 |
"76": "SUCCESS",
|
| 236 |
+
"77": "NON_PARTICIPATING",
|
| 237 |
+
"78": "SUCCESS",
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "SUCCESS",
|
| 240 |
"80": "SUCCESS",
|
|
|
|
| 244 |
"84": "SUCCESS",
|
| 245 |
"85": "NON_PARTICIPATING",
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
+
"87": "FAIL",
|
| 248 |
"88": "SUCCESS",
|
| 249 |
+
"89": "NON_PARTICIPATING",
|
| 250 |
"9": "SUCCESS",
|
| 251 |
"90": "SUCCESS",
|
| 252 |
"91": "SUCCESS",
|
|
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
+
5607323,
|
| 272 |
+
5607327,
|
| 273 |
+
5607330,
|
| 274 |
+
5607334,
|
| 275 |
+
5607338
|
|
|
|
| 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": 5607244,
|
| 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 sha256:4a6f2ca753fdcf14b353d1294109f5b02a02498ee88b2e47c5dd122fa12bba4e
|
| 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 sha256:71ebdf993501cb16fb54b328887bc4a06d87255e7dbf734e70458d19fd8f6c13
|
| 3 |
size 4040701744
|