Run 4. Outer Step 95. Inner Step 1.
Browse files- config.json +30 -30
- 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": "NON_PARTICIPATING",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
"101": "NON_PARTICIPATING",
|
|
@@ -30,7 +30,7 @@
|
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
"122": "NON_PARTICIPATING",
|
| 33 |
-
"123": "
|
| 34 |
"124": "NON_PARTICIPATING",
|
| 35 |
"125": "NON_PARTICIPATING",
|
| 36 |
"126": "NON_PARTICIPATING",
|
|
@@ -38,7 +38,7 @@
|
|
| 38 |
"128": "NON_PARTICIPATING",
|
| 39 |
"129": "NON_PARTICIPATING",
|
| 40 |
"13": "NON_PARTICIPATING",
|
| 41 |
-
"130": "
|
| 42 |
"131": "NON_PARTICIPATING",
|
| 43 |
"132": "NON_PARTICIPATING",
|
| 44 |
"133": "NON_PARTICIPATING",
|
|
@@ -70,7 +70,7 @@
|
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
"159": "NON_PARTICIPATING",
|
| 73 |
-
"16": "
|
| 74 |
"160": "NON_PARTICIPATING",
|
| 75 |
"161": "NON_PARTICIPATING",
|
| 76 |
"162": "NON_PARTICIPATING",
|
|
@@ -81,7 +81,7 @@
|
|
| 81 |
"167": "NON_PARTICIPATING",
|
| 82 |
"168": "NON_PARTICIPATING",
|
| 83 |
"169": "NON_PARTICIPATING",
|
| 84 |
-
"17": "
|
| 85 |
"170": "NON_PARTICIPATING",
|
| 86 |
"171": "NON_PARTICIPATING",
|
| 87 |
"172": "NON_PARTICIPATING",
|
|
@@ -92,8 +92,8 @@
|
|
| 92 |
"177": "NON_PARTICIPATING",
|
| 93 |
"178": "NON_PARTICIPATING",
|
| 94 |
"179": "NON_PARTICIPATING",
|
| 95 |
-
"18": "
|
| 96 |
-
"180": "
|
| 97 |
"181": "NON_PARTICIPATING",
|
| 98 |
"182": "NON_PARTICIPATING",
|
| 99 |
"183": "NON_PARTICIPATING",
|
|
@@ -112,9 +112,9 @@
|
|
| 112 |
"195": "NON_PARTICIPATING",
|
| 113 |
"196": "NON_PARTICIPATING",
|
| 114 |
"197": "NON_PARTICIPATING",
|
| 115 |
-
"198": "
|
| 116 |
"199": "NON_PARTICIPATING",
|
| 117 |
-
"2": "
|
| 118 |
"20": "NON_PARTICIPATING",
|
| 119 |
"200": "NON_PARTICIPATING",
|
| 120 |
"201": "NON_PARTICIPATING",
|
|
@@ -126,7 +126,7 @@
|
|
| 126 |
"207": "NON_PARTICIPATING",
|
| 127 |
"208": "NON_PARTICIPATING",
|
| 128 |
"209": "NON_PARTICIPATING",
|
| 129 |
-
"21": "
|
| 130 |
"210": "NON_PARTICIPATING",
|
| 131 |
"211": "NON_PARTICIPATING",
|
| 132 |
"212": "NON_PARTICIPATING",
|
|
@@ -137,11 +137,11 @@
|
|
| 137 |
"217": "NON_PARTICIPATING",
|
| 138 |
"218": "NON_PARTICIPATING",
|
| 139 |
"219": "NON_PARTICIPATING",
|
| 140 |
-
"22": "
|
| 141 |
"220": "NON_PARTICIPATING",
|
| 142 |
"221": "NON_PARTICIPATING",
|
| 143 |
"222": "NON_PARTICIPATING",
|
| 144 |
-
"223": "
|
| 145 |
"224": "NON_PARTICIPATING",
|
| 146 |
"225": "NON_PARTICIPATING",
|
| 147 |
"226": "NON_PARTICIPATING",
|
|
@@ -166,7 +166,7 @@
|
|
| 166 |
"243": "NON_PARTICIPATING",
|
| 167 |
"244": "NON_PARTICIPATING",
|
| 168 |
"245": "NON_PARTICIPATING",
|
| 169 |
-
"246": "
|
| 170 |
"247": "NON_PARTICIPATING",
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
"249": "NON_PARTICIPATING",
|
|
@@ -180,7 +180,7 @@
|
|
| 180 |
"26": "NON_PARTICIPATING",
|
| 181 |
"27": "NON_PARTICIPATING",
|
| 182 |
"28": "NON_PARTICIPATING",
|
| 183 |
-
"29": "
|
| 184 |
"3": "SUCCESS",
|
| 185 |
"30": "NON_PARTICIPATING",
|
| 186 |
"31": "NON_PARTICIPATING",
|
|
@@ -198,8 +198,8 @@
|
|
| 198 |
"42": "NON_PARTICIPATING",
|
| 199 |
"43": "NON_PARTICIPATING",
|
| 200 |
"44": "NON_PARTICIPATING",
|
| 201 |
-
"45": "
|
| 202 |
-
"46": "
|
| 203 |
"47": "NON_PARTICIPATING",
|
| 204 |
"48": "NON_PARTICIPATING",
|
| 205 |
"49": "NON_PARTICIPATING",
|
|
@@ -208,7 +208,7 @@
|
|
| 208 |
"51": "NON_PARTICIPATING",
|
| 209 |
"52": "NON_PARTICIPATING",
|
| 210 |
"53": "SUCCESS",
|
| 211 |
-
"54": "
|
| 212 |
"55": "NON_PARTICIPATING",
|
| 213 |
"56": "NON_PARTICIPATING",
|
| 214 |
"57": "SUCCESS",
|
|
@@ -217,15 +217,15 @@
|
|
| 217 |
"6": "SUCCESS",
|
| 218 |
"60": "NON_PARTICIPATING",
|
| 219 |
"61": "NON_PARTICIPATING",
|
| 220 |
-
"62": "
|
| 221 |
"63": "NON_PARTICIPATING",
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
"65": "NON_PARTICIPATING",
|
| 224 |
"66": "NON_PARTICIPATING",
|
| 225 |
-
"67": "
|
| 226 |
"68": "SUCCESS",
|
| 227 |
"69": "SUCCESS",
|
| 228 |
-
"7": "
|
| 229 |
"70": "NON_PARTICIPATING",
|
| 230 |
"71": "NON_PARTICIPATING",
|
| 231 |
"72": "NON_PARTICIPATING",
|
|
@@ -241,22 +241,22 @@
|
|
| 241 |
"81": "SUCCESS",
|
| 242 |
"82": "NON_PARTICIPATING",
|
| 243 |
"83": "NON_PARTICIPATING",
|
| 244 |
-
"84": "
|
| 245 |
"85": "NON_PARTICIPATING",
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
-
"87": "
|
| 248 |
"88": "NON_PARTICIPATING",
|
| 249 |
-
"89": "
|
| 250 |
"9": "NON_PARTICIPATING",
|
| 251 |
-
"90": "
|
| 252 |
"91": "NON_PARTICIPATING",
|
| 253 |
"92": "NON_PARTICIPATING",
|
| 254 |
"93": "SUCCESS",
|
| 255 |
-
"94": "
|
| 256 |
"95": "NON_PARTICIPATING",
|
| 257 |
"96": "NON_PARTICIPATING",
|
| 258 |
"97": "NON_PARTICIPATING",
|
| 259 |
-
"98": "
|
| 260 |
"99": "NON_PARTICIPATING"
|
| 261 |
},
|
| 262 |
"architectures": [
|
|
@@ -268,17 +268,17 @@
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
-
|
| 272 |
-
|
| 273 |
],
|
| 274 |
"block_size": 1024,
|
| 275 |
"bos_token_id": 50256,
|
| 276 |
"embd_pdrop": 0.1,
|
| 277 |
"eos_token_id": 50256,
|
| 278 |
"initializer_range": 0.02,
|
| 279 |
-
"inner_step":
|
| 280 |
"inner_steps": 0,
|
| 281 |
-
"last_allreduce_block":
|
| 282 |
"layer_norm_epsilon": 1e-05,
|
| 283 |
"model_type": "gpt_optimized",
|
| 284 |
"n_embd": 1280,
|
|
|
|
| 3 |
"activation_function": "gelu_new",
|
| 4 |
"all_reduce_scores": {
|
| 5 |
"0": "NON_PARTICIPATING",
|
| 6 |
+
"1": "SUCCESS",
|
| 7 |
"10": "NON_PARTICIPATING",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
"101": "NON_PARTICIPATING",
|
|
|
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
"122": "NON_PARTICIPATING",
|
| 33 |
+
"123": "SUCCESS",
|
| 34 |
"124": "NON_PARTICIPATING",
|
| 35 |
"125": "NON_PARTICIPATING",
|
| 36 |
"126": "NON_PARTICIPATING",
|
|
|
|
| 38 |
"128": "NON_PARTICIPATING",
|
| 39 |
"129": "NON_PARTICIPATING",
|
| 40 |
"13": "NON_PARTICIPATING",
|
| 41 |
+
"130": "NON_PARTICIPATING",
|
| 42 |
"131": "NON_PARTICIPATING",
|
| 43 |
"132": "NON_PARTICIPATING",
|
| 44 |
"133": "NON_PARTICIPATING",
|
|
|
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
"159": "NON_PARTICIPATING",
|
| 73 |
+
"16": "SUCCESS",
|
| 74 |
"160": "NON_PARTICIPATING",
|
| 75 |
"161": "NON_PARTICIPATING",
|
| 76 |
"162": "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",
|
|
|
|
| 92 |
"177": "NON_PARTICIPATING",
|
| 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",
|
|
|
|
| 112 |
"195": "NON_PARTICIPATING",
|
| 113 |
"196": "NON_PARTICIPATING",
|
| 114 |
"197": "NON_PARTICIPATING",
|
| 115 |
+
"198": "NON_PARTICIPATING",
|
| 116 |
"199": "NON_PARTICIPATING",
|
| 117 |
+
"2": "SUCCESS",
|
| 118 |
"20": "NON_PARTICIPATING",
|
| 119 |
"200": "NON_PARTICIPATING",
|
| 120 |
"201": "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",
|
|
|
|
| 137 |
"217": "NON_PARTICIPATING",
|
| 138 |
"218": "NON_PARTICIPATING",
|
| 139 |
"219": "NON_PARTICIPATING",
|
| 140 |
+
"22": "SUCCESS",
|
| 141 |
"220": "NON_PARTICIPATING",
|
| 142 |
"221": "NON_PARTICIPATING",
|
| 143 |
"222": "NON_PARTICIPATING",
|
| 144 |
+
"223": "NON_PARTICIPATING",
|
| 145 |
"224": "NON_PARTICIPATING",
|
| 146 |
"225": "NON_PARTICIPATING",
|
| 147 |
"226": "NON_PARTICIPATING",
|
|
|
|
| 166 |
"243": "NON_PARTICIPATING",
|
| 167 |
"244": "NON_PARTICIPATING",
|
| 168 |
"245": "NON_PARTICIPATING",
|
| 169 |
+
"246": "SUCCESS",
|
| 170 |
"247": "NON_PARTICIPATING",
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
"249": "NON_PARTICIPATING",
|
|
|
|
| 180 |
"26": "NON_PARTICIPATING",
|
| 181 |
"27": "NON_PARTICIPATING",
|
| 182 |
"28": "NON_PARTICIPATING",
|
| 183 |
+
"29": "SUCCESS",
|
| 184 |
"3": "SUCCESS",
|
| 185 |
"30": "NON_PARTICIPATING",
|
| 186 |
"31": "NON_PARTICIPATING",
|
|
|
|
| 198 |
"42": "NON_PARTICIPATING",
|
| 199 |
"43": "NON_PARTICIPATING",
|
| 200 |
"44": "NON_PARTICIPATING",
|
| 201 |
+
"45": "SUCCESS",
|
| 202 |
+
"46": "SUCCESS",
|
| 203 |
"47": "NON_PARTICIPATING",
|
| 204 |
"48": "NON_PARTICIPATING",
|
| 205 |
"49": "NON_PARTICIPATING",
|
|
|
|
| 208 |
"51": "NON_PARTICIPATING",
|
| 209 |
"52": "NON_PARTICIPATING",
|
| 210 |
"53": "SUCCESS",
|
| 211 |
+
"54": "SUCCESS",
|
| 212 |
"55": "NON_PARTICIPATING",
|
| 213 |
"56": "NON_PARTICIPATING",
|
| 214 |
"57": "SUCCESS",
|
|
|
|
| 217 |
"6": "SUCCESS",
|
| 218 |
"60": "NON_PARTICIPATING",
|
| 219 |
"61": "NON_PARTICIPATING",
|
| 220 |
+
"62": "SUCCESS",
|
| 221 |
"63": "NON_PARTICIPATING",
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
"65": "NON_PARTICIPATING",
|
| 224 |
"66": "NON_PARTICIPATING",
|
| 225 |
+
"67": "SUCCESS",
|
| 226 |
"68": "SUCCESS",
|
| 227 |
"69": "SUCCESS",
|
| 228 |
+
"7": "SUCCESS",
|
| 229 |
"70": "NON_PARTICIPATING",
|
| 230 |
"71": "NON_PARTICIPATING",
|
| 231 |
"72": "NON_PARTICIPATING",
|
|
|
|
| 241 |
"81": "SUCCESS",
|
| 242 |
"82": "NON_PARTICIPATING",
|
| 243 |
"83": "NON_PARTICIPATING",
|
| 244 |
+
"84": "SUCCESS",
|
| 245 |
"85": "NON_PARTICIPATING",
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
+
"87": "SUCCESS",
|
| 248 |
"88": "NON_PARTICIPATING",
|
| 249 |
+
"89": "SUCCESS",
|
| 250 |
"9": "NON_PARTICIPATING",
|
| 251 |
+
"90": "SUCCESS",
|
| 252 |
"91": "NON_PARTICIPATING",
|
| 253 |
"92": "NON_PARTICIPATING",
|
| 254 |
"93": "SUCCESS",
|
| 255 |
+
"94": "SUCCESS",
|
| 256 |
"95": "NON_PARTICIPATING",
|
| 257 |
"96": "NON_PARTICIPATING",
|
| 258 |
"97": "NON_PARTICIPATING",
|
| 259 |
+
"98": "SUCCESS",
|
| 260 |
"99": "NON_PARTICIPATING"
|
| 261 |
},
|
| 262 |
"architectures": [
|
|
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
+
5689300,
|
| 272 |
+
5689316
|
| 273 |
],
|
| 274 |
"block_size": 1024,
|
| 275 |
"bos_token_id": 50256,
|
| 276 |
"embd_pdrop": 0.1,
|
| 277 |
"eos_token_id": 50256,
|
| 278 |
"initializer_range": 0.02,
|
| 279 |
+
"inner_step": 1,
|
| 280 |
"inner_steps": 0,
|
| 281 |
+
"last_allreduce_block": 5689165,
|
| 282 |
"layer_norm_epsilon": 1e-05,
|
| 283 |
"model_type": "gpt_optimized",
|
| 284 |
"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:fc1bee61a55b7e010e093f36bbd9b521f4d1a0682e11aa93c77212241cc88e11
|
| 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 sha256:8098df6efa5001898c80bf2c70cb010d19bb8c2d0d64aa8c82113ea5b77806b4
|
| 3 |
size 4040701744
|