Run 4. Outer Step 55. Inner Step 0.
Browse files- config.json +42 -42
- 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",
|
|
@@ -13,7 +13,7 @@
|
|
| 13 |
"105": "NON_PARTICIPATING",
|
| 14 |
"106": "NON_PARTICIPATING",
|
| 15 |
"107": "NON_PARTICIPATING",
|
| 16 |
-
"108": "
|
| 17 |
"109": "NON_PARTICIPATING",
|
| 18 |
"11": "SUCCESS",
|
| 19 |
"110": "NON_PARTICIPATING",
|
|
@@ -28,17 +28,17 @@
|
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
"12": "SUCCESS",
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 31 |
-
"121": "
|
| 32 |
"122": "NON_PARTICIPATING",
|
| 33 |
"123": "SUCCESS",
|
| 34 |
"124": "NON_PARTICIPATING",
|
| 35 |
"125": "NON_PARTICIPATING",
|
| 36 |
-
"126": "
|
| 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",
|
|
@@ -48,15 +48,15 @@
|
|
| 48 |
"137": "NON_PARTICIPATING",
|
| 49 |
"138": "NON_PARTICIPATING",
|
| 50 |
"139": "NON_PARTICIPATING",
|
| 51 |
-
"14": "
|
| 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": "NON_PARTICIPATING",
|
| 61 |
"149": "NON_PARTICIPATING",
|
| 62 |
"15": "SUCCESS",
|
|
@@ -64,7 +64,7 @@
|
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
"153": "SUCCESS",
|
| 67 |
-
"154": "
|
| 68 |
"155": "NON_PARTICIPATING",
|
| 69 |
"156": "NON_PARTICIPATING",
|
| 70 |
"157": "NON_PARTICIPATING",
|
|
@@ -79,26 +79,26 @@
|
|
| 79 |
"165": "NON_PARTICIPATING",
|
| 80 |
"166": "NON_PARTICIPATING",
|
| 81 |
"167": "NON_PARTICIPATING",
|
| 82 |
-
"168": "
|
| 83 |
"169": "NON_PARTICIPATING",
|
| 84 |
"17": "SUCCESS",
|
| 85 |
"170": "NON_PARTICIPATING",
|
| 86 |
-
"171": "
|
| 87 |
"172": "NON_PARTICIPATING",
|
| 88 |
-
"173": "
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
"175": "NON_PARTICIPATING",
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
-
"177": "
|
| 93 |
"178": "NON_PARTICIPATING",
|
| 94 |
"179": "NON_PARTICIPATING",
|
| 95 |
"18": "SUCCESS",
|
| 96 |
-
"180": "
|
| 97 |
"181": "NON_PARTICIPATING",
|
| 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",
|
|
@@ -118,16 +118,16 @@
|
|
| 118 |
"20": "SUCCESS",
|
| 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": "
|
| 129 |
"21": "SUCCESS",
|
| 130 |
-
"210": "
|
| 131 |
"211": "NON_PARTICIPATING",
|
| 132 |
"212": "NON_PARTICIPATING",
|
| 133 |
"213": "NON_PARTICIPATING",
|
|
@@ -139,7 +139,7 @@
|
|
| 139 |
"219": "NON_PARTICIPATING",
|
| 140 |
"22": "SUCCESS",
|
| 141 |
"220": "SUCCESS",
|
| 142 |
-
"221": "
|
| 143 |
"222": "NON_PARTICIPATING",
|
| 144 |
"223": "NON_PARTICIPATING",
|
| 145 |
"224": "NON_PARTICIPATING",
|
|
@@ -149,34 +149,34 @@
|
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
"229": "SUCCESS",
|
| 151 |
"23": "SUCCESS",
|
| 152 |
-
"230": "
|
| 153 |
-
"231": "
|
| 154 |
"232": "NON_PARTICIPATING",
|
| 155 |
"233": "NON_PARTICIPATING",
|
| 156 |
"234": "NON_PARTICIPATING",
|
| 157 |
"235": "NON_PARTICIPATING",
|
| 158 |
-
"236": "
|
| 159 |
"237": "NON_PARTICIPATING",
|
| 160 |
"238": "SUCCESS",
|
| 161 |
"239": "NON_PARTICIPATING",
|
| 162 |
"24": "SUCCESS",
|
| 163 |
"240": "NON_PARTICIPATING",
|
| 164 |
"241": "NON_PARTICIPATING",
|
| 165 |
-
"242": "
|
| 166 |
"243": "NON_PARTICIPATING",
|
| 167 |
"244": "NON_PARTICIPATING",
|
| 168 |
"245": "NON_PARTICIPATING",
|
| 169 |
"246": "NON_PARTICIPATING",
|
| 170 |
-
"247": "
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
-
"249": "
|
| 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": "
|
| 180 |
"26": "SUCCESS",
|
| 181 |
"27": "SUCCESS",
|
| 182 |
"28": "SUCCESS",
|
|
@@ -187,7 +187,7 @@
|
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "SUCCESS",
|
| 190 |
-
"35": "
|
| 191 |
"36": "SUCCESS",
|
| 192 |
"37": "SUCCESS",
|
| 193 |
"38": "SUCCESS",
|
|
@@ -195,7 +195,7 @@
|
|
| 195 |
"4": "SUCCESS",
|
| 196 |
"40": "SUCCESS",
|
| 197 |
"41": "SUCCESS",
|
| 198 |
-
"42": "
|
| 199 |
"43": "SUCCESS",
|
| 200 |
"44": "SUCCESS",
|
| 201 |
"45": "SUCCESS",
|
|
@@ -203,14 +203,14 @@
|
|
| 203 |
"47": "SUCCESS",
|
| 204 |
"48": "NON_PARTICIPATING",
|
| 205 |
"49": "SUCCESS",
|
| 206 |
-
"5": "
|
| 207 |
"50": "SUCCESS",
|
| 208 |
"51": "SUCCESS",
|
| 209 |
-
"52": "
|
| 210 |
"53": "SUCCESS",
|
| 211 |
"54": "SUCCESS",
|
| 212 |
"55": "SUCCESS",
|
| 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",
|
|
@@ -234,7 +234,7 @@
|
|
| 234 |
"75": "SUCCESS",
|
| 235 |
"76": "SUCCESS",
|
| 236 |
"77": "SUCCESS",
|
| 237 |
-
"78": "
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "SUCCESS",
|
| 240 |
"80": "SUCCESS",
|
|
@@ -268,20 +268,20 @@
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
-
|
| 272 |
-
|
| 273 |
-
|
| 274 |
-
|
| 275 |
-
|
| 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":
|
| 283 |
"inner_steps": 0,
|
| 284 |
-
"last_allreduce_block":
|
| 285 |
"layer_norm_epsilon": 1e-05,
|
| 286 |
"model_type": "gpt_optimized",
|
| 287 |
"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",
|
|
|
|
| 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": "NON_PARTICIPATING",
|
|
|
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
"12": "SUCCESS",
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 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": "NON_PARTICIPATING",
|
| 40 |
"13": "SUCCESS",
|
| 41 |
+
"130": "SUCCESS",
|
| 42 |
"131": "NON_PARTICIPATING",
|
| 43 |
"132": "NON_PARTICIPATING",
|
| 44 |
"133": "NON_PARTICIPATING",
|
|
|
|
| 48 |
"137": "NON_PARTICIPATING",
|
| 49 |
"138": "NON_PARTICIPATING",
|
| 50 |
"139": "NON_PARTICIPATING",
|
| 51 |
+
"14": "SUCCESS",
|
| 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": "SUCCESS",
|
| 60 |
"148": "NON_PARTICIPATING",
|
| 61 |
"149": "NON_PARTICIPATING",
|
| 62 |
"15": "SUCCESS",
|
|
|
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
"153": "SUCCESS",
|
| 67 |
+
"154": "SUCCESS",
|
| 68 |
"155": "NON_PARTICIPATING",
|
| 69 |
"156": "NON_PARTICIPATING",
|
| 70 |
"157": "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": "NON_PARTICIPATING",
|
| 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": "SUCCESS",
|
| 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": "NON_PARTICIPATING",
|
| 100 |
"184": "NON_PARTICIPATING",
|
| 101 |
+
"185": "SUCCESS",
|
| 102 |
"186": "NON_PARTICIPATING",
|
| 103 |
"187": "NON_PARTICIPATING",
|
| 104 |
"188": "NON_PARTICIPATING",
|
|
|
|
| 118 |
"20": "SUCCESS",
|
| 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",
|
|
|
|
| 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",
|
|
|
|
| 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",
|
| 156 |
"234": "NON_PARTICIPATING",
|
| 157 |
"235": "NON_PARTICIPATING",
|
| 158 |
+
"236": "NON_PARTICIPATING",
|
| 159 |
"237": "NON_PARTICIPATING",
|
| 160 |
"238": "SUCCESS",
|
| 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",
|
| 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": "SUCCESS",
|
| 180 |
"26": "SUCCESS",
|
| 181 |
"27": "SUCCESS",
|
| 182 |
"28": "SUCCESS",
|
|
|
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "SUCCESS",
|
| 190 |
+
"35": "NON_PARTICIPATING",
|
| 191 |
"36": "SUCCESS",
|
| 192 |
"37": "SUCCESS",
|
| 193 |
"38": "SUCCESS",
|
|
|
|
| 195 |
"4": "SUCCESS",
|
| 196 |
"40": "SUCCESS",
|
| 197 |
"41": "SUCCESS",
|
| 198 |
+
"42": "SUCCESS",
|
| 199 |
"43": "SUCCESS",
|
| 200 |
"44": "SUCCESS",
|
| 201 |
"45": "SUCCESS",
|
|
|
|
| 203 |
"47": "SUCCESS",
|
| 204 |
"48": "NON_PARTICIPATING",
|
| 205 |
"49": "SUCCESS",
|
| 206 |
+
"5": "SUCCESS",
|
| 207 |
"50": "SUCCESS",
|
| 208 |
"51": "SUCCESS",
|
| 209 |
+
"52": "NON_PARTICIPATING",
|
| 210 |
"53": "SUCCESS",
|
| 211 |
"54": "SUCCESS",
|
| 212 |
"55": "SUCCESS",
|
| 213 |
+
"56": "FAIL",
|
| 214 |
"57": "SUCCESS",
|
| 215 |
"58": "NON_PARTICIPATING",
|
| 216 |
"59": "SUCCESS",
|
|
|
|
| 218 |
"60": "SUCCESS",
|
| 219 |
"61": "NON_PARTICIPATING",
|
| 220 |
"62": "SUCCESS",
|
| 221 |
+
"63": "FAIL",
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
+
"65": "NON_PARTICIPATING",
|
| 224 |
"66": "SUCCESS",
|
| 225 |
"67": "SUCCESS",
|
| 226 |
"68": "SUCCESS",
|
|
|
|
| 234 |
"75": "SUCCESS",
|
| 235 |
"76": "SUCCESS",
|
| 236 |
"77": "SUCCESS",
|
| 237 |
+
"78": "NON_PARTICIPATING",
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "SUCCESS",
|
| 240 |
"80": "SUCCESS",
|
|
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
+
5598080,
|
| 272 |
+
5598141,
|
| 273 |
+
5598145,
|
| 274 |
+
5598149,
|
| 275 |
+
5598153
|
| 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": 5598085,
|
| 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 sha256:947d97cde1f2589a1b11de936d83c3f55f3059a2447fb09709e9b6312092d6c2
|
| 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 sha256:075efad1f420165ff4bebedb7a530e2a586494b40fd90fec010614eb24a0d69e
|
| 3 |
size 4040701744
|