Run 4. Outer Step 20. Inner Step 0.
Browse files- config.json +34 -37
- inner_optimizer.pt +1 -1
- model.safetensors +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": "SUCCESS",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
"101": "SUCCESS",
|
|
@@ -14,10 +14,10 @@
|
|
| 14 |
"106": "SUCCESS",
|
| 15 |
"107": "SUCCESS",
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
-
"109": "
|
| 18 |
"11": "NON_PARTICIPATING",
|
| 19 |
"110": "SUCCESS",
|
| 20 |
-
"111": "
|
| 21 |
"112": "NON_PARTICIPATING",
|
| 22 |
"113": "NON_PARTICIPATING",
|
| 23 |
"114": "SUCCESS",
|
|
@@ -27,14 +27,14 @@
|
|
| 27 |
"118": "SUCCESS",
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
"12": "SUCCESS",
|
| 30 |
-
"120": "
|
| 31 |
-
"121": "
|
| 32 |
"122": "SUCCESS",
|
| 33 |
"123": "SUCCESS",
|
| 34 |
"124": "NON_PARTICIPATING",
|
| 35 |
"125": "NON_PARTICIPATING",
|
| 36 |
"126": "NON_PARTICIPATING",
|
| 37 |
-
"127": "
|
| 38 |
"128": "NON_PARTICIPATING",
|
| 39 |
"129": "NON_PARTICIPATING",
|
| 40 |
"13": "NON_PARTICIPATING",
|
|
@@ -43,7 +43,7 @@
|
|
| 43 |
"132": "SUCCESS",
|
| 44 |
"133": "NON_PARTICIPATING",
|
| 45 |
"134": "NON_PARTICIPATING",
|
| 46 |
-
"135": "
|
| 47 |
"136": "NON_PARTICIPATING",
|
| 48 |
"137": "NON_PARTICIPATING",
|
| 49 |
"138": "NON_PARTICIPATING",
|
|
@@ -56,20 +56,20 @@
|
|
| 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": "
|
| 68 |
"155": "NON_PARTICIPATING",
|
| 69 |
"156": "SUCCESS",
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
-
"159": "
|
| 73 |
"16": "SUCCESS",
|
| 74 |
"160": "SUCCESS",
|
| 75 |
"161": "NON_PARTICIPATING",
|
|
@@ -79,7 +79,7 @@
|
|
| 79 |
"165": "NON_PARTICIPATING",
|
| 80 |
"166": "NON_PARTICIPATING",
|
| 81 |
"167": "SUCCESS",
|
| 82 |
-
"168": "
|
| 83 |
"169": "NON_PARTICIPATING",
|
| 84 |
"17": "SUCCESS",
|
| 85 |
"170": "SUCCESS",
|
|
@@ -100,13 +100,13 @@
|
|
| 100 |
"184": "SUCCESS",
|
| 101 |
"185": "NON_PARTICIPATING",
|
| 102 |
"186": "SUCCESS",
|
| 103 |
-
"187": "
|
| 104 |
"188": "NON_PARTICIPATING",
|
| 105 |
"189": "NON_PARTICIPATING",
|
| 106 |
"19": "NON_PARTICIPATING",
|
| 107 |
"190": "NON_PARTICIPATING",
|
| 108 |
-
"191": "
|
| 109 |
-
"192": "
|
| 110 |
"193": "SUCCESS",
|
| 111 |
"194": "NON_PARTICIPATING",
|
| 112 |
"195": "SUCCESS",
|
|
@@ -136,10 +136,10 @@
|
|
| 136 |
"216": "NON_PARTICIPATING",
|
| 137 |
"217": "NON_PARTICIPATING",
|
| 138 |
"218": "SUCCESS",
|
| 139 |
-
"219": "
|
| 140 |
"22": "SUCCESS",
|
| 141 |
"220": "NON_PARTICIPATING",
|
| 142 |
-
"221": "
|
| 143 |
"222": "NON_PARTICIPATING",
|
| 144 |
"223": "SUCCESS",
|
| 145 |
"224": "NON_PARTICIPATING",
|
|
@@ -150,7 +150,7 @@
|
|
| 150 |
"229": "NON_PARTICIPATING",
|
| 151 |
"23": "SUCCESS",
|
| 152 |
"230": "NON_PARTICIPATING",
|
| 153 |
-
"231": "
|
| 154 |
"232": "NON_PARTICIPATING",
|
| 155 |
"233": "NON_PARTICIPATING",
|
| 156 |
"234": "NON_PARTICIPATING",
|
|
@@ -161,7 +161,7 @@
|
|
| 161 |
"239": "NON_PARTICIPATING",
|
| 162 |
"24": "SUCCESS",
|
| 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",
|
|
@@ -208,16 +208,16 @@
|
|
| 208 |
"51": "SUCCESS",
|
| 209 |
"52": "NON_PARTICIPATING",
|
| 210 |
"53": "NON_PARTICIPATING",
|
| 211 |
-
"54": "
|
| 212 |
"55": "SUCCESS",
|
| 213 |
"56": "SUCCESS",
|
| 214 |
"57": "SUCCESS",
|
| 215 |
"58": "NON_PARTICIPATING",
|
| 216 |
"59": "NON_PARTICIPATING",
|
| 217 |
"6": "NON_PARTICIPATING",
|
| 218 |
-
"60": "
|
| 219 |
"61": "NON_PARTICIPATING",
|
| 220 |
-
"62": "
|
| 221 |
"63": "SUCCESS",
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
"65": "NON_PARTICIPATING",
|
|
@@ -252,11 +252,11 @@
|
|
| 252 |
"91": "SUCCESS",
|
| 253 |
"92": "NON_PARTICIPATING",
|
| 254 |
"93": "NON_PARTICIPATING",
|
| 255 |
-
"94": "
|
| 256 |
"95": "SUCCESS",
|
| 257 |
"96": "SUCCESS",
|
| 258 |
"97": "NON_PARTICIPATING",
|
| 259 |
-
"98": "
|
| 260 |
"99": "SUCCESS"
|
| 261 |
},
|
| 262 |
"architectures": [
|
|
@@ -268,23 +268,20 @@
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
-
|
| 272 |
-
|
| 273 |
-
|
| 274 |
-
|
| 275 |
-
|
| 276 |
-
5523935,
|
| 277 |
-
5523939,
|
| 278 |
-
5523943
|
| 279 |
],
|
| 280 |
"block_size": 1024,
|
| 281 |
"bos_token_id": 50256,
|
| 282 |
"embd_pdrop": 0.1,
|
| 283 |
"eos_token_id": 50256,
|
| 284 |
"initializer_range": 0.02,
|
| 285 |
-
"inner_step":
|
| 286 |
"inner_steps": 0,
|
| 287 |
-
"last_allreduce_block":
|
| 288 |
"layer_norm_epsilon": 1e-05,
|
| 289 |
"model_type": "gpt_optimized",
|
| 290 |
"n_embd": 1280,
|
|
|
|
| 3 |
"activation_function": "gelu_new",
|
| 4 |
"all_reduce_scores": {
|
| 5 |
"0": "NON_PARTICIPATING",
|
| 6 |
+
"1": "SUCCESS",
|
| 7 |
"10": "SUCCESS",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
"101": "SUCCESS",
|
|
|
|
| 14 |
"106": "SUCCESS",
|
| 15 |
"107": "SUCCESS",
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
+
"109": "NON_PARTICIPATING",
|
| 18 |
"11": "NON_PARTICIPATING",
|
| 19 |
"110": "SUCCESS",
|
| 20 |
+
"111": "SUCCESS",
|
| 21 |
"112": "NON_PARTICIPATING",
|
| 22 |
"113": "NON_PARTICIPATING",
|
| 23 |
"114": "SUCCESS",
|
|
|
|
| 27 |
"118": "SUCCESS",
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
"12": "SUCCESS",
|
| 30 |
+
"120": "NON_PARTICIPATING",
|
| 31 |
+
"121": "SUCCESS",
|
| 32 |
"122": "SUCCESS",
|
| 33 |
"123": "SUCCESS",
|
| 34 |
"124": "NON_PARTICIPATING",
|
| 35 |
"125": "NON_PARTICIPATING",
|
| 36 |
"126": "NON_PARTICIPATING",
|
| 37 |
+
"127": "SUCCESS",
|
| 38 |
"128": "NON_PARTICIPATING",
|
| 39 |
"129": "NON_PARTICIPATING",
|
| 40 |
"13": "NON_PARTICIPATING",
|
|
|
|
| 43 |
"132": "SUCCESS",
|
| 44 |
"133": "NON_PARTICIPATING",
|
| 45 |
"134": "NON_PARTICIPATING",
|
| 46 |
+
"135": "SUCCESS",
|
| 47 |
"136": "NON_PARTICIPATING",
|
| 48 |
"137": "NON_PARTICIPATING",
|
| 49 |
"138": "NON_PARTICIPATING",
|
|
|
|
| 56 |
"144": "NON_PARTICIPATING",
|
| 57 |
"145": "NON_PARTICIPATING",
|
| 58 |
"146": "NON_PARTICIPATING",
|
| 59 |
+
"147": "NON_PARTICIPATING",
|
| 60 |
+
"148": "NON_PARTICIPATING",
|
| 61 |
+
"149": "SUCCESS",
|
| 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": "SUCCESS",
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
+
"159": "SUCCESS",
|
| 73 |
"16": "SUCCESS",
|
| 74 |
"160": "SUCCESS",
|
| 75 |
"161": "NON_PARTICIPATING",
|
|
|
|
| 79 |
"165": "NON_PARTICIPATING",
|
| 80 |
"166": "NON_PARTICIPATING",
|
| 81 |
"167": "SUCCESS",
|
| 82 |
+
"168": "NON_PARTICIPATING",
|
| 83 |
"169": "NON_PARTICIPATING",
|
| 84 |
"17": "SUCCESS",
|
| 85 |
"170": "SUCCESS",
|
|
|
|
| 100 |
"184": "SUCCESS",
|
| 101 |
"185": "NON_PARTICIPATING",
|
| 102 |
"186": "SUCCESS",
|
| 103 |
+
"187": "NON_PARTICIPATING",
|
| 104 |
"188": "NON_PARTICIPATING",
|
| 105 |
"189": "NON_PARTICIPATING",
|
| 106 |
"19": "NON_PARTICIPATING",
|
| 107 |
"190": "NON_PARTICIPATING",
|
| 108 |
+
"191": "NON_PARTICIPATING",
|
| 109 |
+
"192": "NON_PARTICIPATING",
|
| 110 |
"193": "SUCCESS",
|
| 111 |
"194": "NON_PARTICIPATING",
|
| 112 |
"195": "SUCCESS",
|
|
|
|
| 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",
|
|
|
|
| 150 |
"229": "NON_PARTICIPATING",
|
| 151 |
"23": "SUCCESS",
|
| 152 |
"230": "NON_PARTICIPATING",
|
| 153 |
+
"231": "NON_PARTICIPATING",
|
| 154 |
"232": "NON_PARTICIPATING",
|
| 155 |
"233": "NON_PARTICIPATING",
|
| 156 |
"234": "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": "SUCCESS",
|
|
|
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
"249": "SUCCESS",
|
| 173 |
"25": "SUCCESS",
|
| 174 |
+
"250": "NON_PARTICIPATING",
|
| 175 |
"251": "NON_PARTICIPATING",
|
| 176 |
"252": "NON_PARTICIPATING",
|
| 177 |
"253": "NON_PARTICIPATING",
|
|
|
|
| 208 |
"51": "SUCCESS",
|
| 209 |
"52": "NON_PARTICIPATING",
|
| 210 |
"53": "NON_PARTICIPATING",
|
| 211 |
+
"54": "SUCCESS",
|
| 212 |
"55": "SUCCESS",
|
| 213 |
"56": "SUCCESS",
|
| 214 |
"57": "SUCCESS",
|
| 215 |
"58": "NON_PARTICIPATING",
|
| 216 |
"59": "NON_PARTICIPATING",
|
| 217 |
"6": "NON_PARTICIPATING",
|
| 218 |
+
"60": "SUCCESS",
|
| 219 |
"61": "NON_PARTICIPATING",
|
| 220 |
+
"62": "SUCCESS",
|
| 221 |
"63": "SUCCESS",
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
"65": "NON_PARTICIPATING",
|
|
|
|
| 252 |
"91": "SUCCESS",
|
| 253 |
"92": "NON_PARTICIPATING",
|
| 254 |
"93": "NON_PARTICIPATING",
|
| 255 |
+
"94": "SUCCESS",
|
| 256 |
"95": "SUCCESS",
|
| 257 |
"96": "SUCCESS",
|
| 258 |
"97": "NON_PARTICIPATING",
|
| 259 |
+
"98": "SUCCESS",
|
| 260 |
"99": "SUCCESS"
|
| 261 |
},
|
| 262 |
"architectures": [
|
|
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
+
5524039,
|
| 272 |
+
5524042,
|
| 273 |
+
5524046,
|
| 274 |
+
5524049,
|
| 275 |
+
5524053
|
|
|
|
|
|
|
|
|
|
| 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": 5523943,
|
| 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:e94fa6ac5551a36f60d15da29d09afdb270881cc81e1d5d124c11fbb9f83f935
|
| 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 sha256:22d2815d35e5139d8f03b0869a4072819f099cc8f8fcf4e0be28d232526efd42
|
| 3 |
size 4040701744
|