Run 4. Outer Step 62. Inner Step 2.
Browse files- config.json +44 -44
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
config.json
CHANGED
|
@@ -3,37 +3,37 @@
|
|
| 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": "
|
| 10 |
-
"102": "
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
-
"104": "
|
| 13 |
"105": "SUCCESS",
|
| 14 |
-
"106": "
|
| 15 |
"107": "NON_PARTICIPATING",
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 17 |
"109": "NON_PARTICIPATING",
|
| 18 |
"11": "SUCCESS",
|
| 19 |
-
"110": "
|
| 20 |
-
"111": "
|
| 21 |
"112": "NON_PARTICIPATING",
|
| 22 |
"113": "NON_PARTICIPATING",
|
| 23 |
"114": "NON_PARTICIPATING",
|
| 24 |
"115": "NON_PARTICIPATING",
|
| 25 |
"116": "NON_PARTICIPATING",
|
| 26 |
"117": "NON_PARTICIPATING",
|
| 27 |
-
"118": "
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
"12": "SUCCESS",
|
| 30 |
-
"120": "
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
-
"122": "
|
| 33 |
-
"123": "
|
| 34 |
"124": "NON_PARTICIPATING",
|
| 35 |
-
"125": "
|
| 36 |
-
"126": "
|
| 37 |
"127": "NON_PARTICIPATING",
|
| 38 |
"128": "NON_PARTICIPATING",
|
| 39 |
"129": "NON_PARTICIPATING",
|
|
@@ -42,31 +42,31 @@
|
|
| 42 |
"131": "NON_PARTICIPATING",
|
| 43 |
"132": "NON_PARTICIPATING",
|
| 44 |
"133": "NON_PARTICIPATING",
|
| 45 |
-
"134": "
|
| 46 |
"135": "NON_PARTICIPATING",
|
| 47 |
"136": "NON_PARTICIPATING",
|
| 48 |
-
"137": "
|
| 49 |
-
"138": "
|
| 50 |
"139": "SUCCESS",
|
| 51 |
-
"14": "
|
| 52 |
"140": "NON_PARTICIPATING",
|
| 53 |
-
"141": "
|
| 54 |
"142": "NON_PARTICIPATING",
|
| 55 |
-
"143": "
|
| 56 |
"144": "NON_PARTICIPATING",
|
| 57 |
"145": "NON_PARTICIPATING",
|
| 58 |
"146": "NON_PARTICIPATING",
|
| 59 |
"147": "NON_PARTICIPATING",
|
| 60 |
"148": "NON_PARTICIPATING",
|
| 61 |
"149": "NON_PARTICIPATING",
|
| 62 |
-
"15": "
|
| 63 |
-
"150": "
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
"153": "NON_PARTICIPATING",
|
| 67 |
"154": "NON_PARTICIPATING",
|
| 68 |
"155": "NON_PARTICIPATING",
|
| 69 |
-
"156": "
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
"159": "NON_PARTICIPATING",
|
|
@@ -83,11 +83,11 @@
|
|
| 83 |
"169": "NON_PARTICIPATING",
|
| 84 |
"17": "SUCCESS",
|
| 85 |
"170": "NON_PARTICIPATING",
|
| 86 |
-
"171": "
|
| 87 |
"172": "NON_PARTICIPATING",
|
| 88 |
"173": "NON_PARTICIPATING",
|
| 89 |
"174": "NON_PARTICIPATING",
|
| 90 |
-
"175": "
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
"177": "SUCCESS",
|
| 93 |
"178": "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",
|
|
@@ -121,10 +121,10 @@
|
|
| 121 |
"202": "SUCCESS",
|
| 122 |
"203": "NON_PARTICIPATING",
|
| 123 |
"204": "NON_PARTICIPATING",
|
| 124 |
-
"205": "
|
| 125 |
"206": "NON_PARTICIPATING",
|
| 126 |
"207": "NON_PARTICIPATING",
|
| 127 |
-
"208": "
|
| 128 |
"209": "NON_PARTICIPATING",
|
| 129 |
"21": "SUCCESS",
|
| 130 |
"210": "NON_PARTICIPATING",
|
|
@@ -133,7 +133,7 @@
|
|
| 133 |
"213": "NON_PARTICIPATING",
|
| 134 |
"214": "NON_PARTICIPATING",
|
| 135 |
"215": "NON_PARTICIPATING",
|
| 136 |
-
"216": "
|
| 137 |
"217": "NON_PARTICIPATING",
|
| 138 |
"218": "NON_PARTICIPATING",
|
| 139 |
"219": "NON_PARTICIPATING",
|
|
@@ -151,7 +151,7 @@
|
|
| 151 |
"23": "SUCCESS",
|
| 152 |
"230": "SUCCESS",
|
| 153 |
"231": "NON_PARTICIPATING",
|
| 154 |
-
"232": "
|
| 155 |
"233": "SUCCESS",
|
| 156 |
"234": "NON_PARTICIPATING",
|
| 157 |
"235": "NON_PARTICIPATING",
|
|
@@ -163,10 +163,10 @@
|
|
| 163 |
"240": "NON_PARTICIPATING",
|
| 164 |
"241": "NON_PARTICIPATING",
|
| 165 |
"242": "NON_PARTICIPATING",
|
| 166 |
-
"243": "
|
| 167 |
-
"244": "
|
| 168 |
"245": "NON_PARTICIPATING",
|
| 169 |
-
"246": "
|
| 170 |
"247": "NON_PARTICIPATING",
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
"249": "NON_PARTICIPATING",
|
|
@@ -175,7 +175,7 @@
|
|
| 175 |
"251": "NON_PARTICIPATING",
|
| 176 |
"252": "NON_PARTICIPATING",
|
| 177 |
"253": "NON_PARTICIPATING",
|
| 178 |
-
"254": "
|
| 179 |
"255": "NON_PARTICIPATING",
|
| 180 |
"26": "SUCCESS",
|
| 181 |
"27": "SUCCESS",
|
|
@@ -188,11 +188,11 @@
|
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "NON_PARTICIPATING",
|
| 190 |
"35": "SUCCESS",
|
| 191 |
-
"36": "
|
| 192 |
"37": "SUCCESS",
|
| 193 |
"38": "SUCCESS",
|
| 194 |
"39": "SUCCESS",
|
| 195 |
-
"4": "
|
| 196 |
"40": "SUCCESS",
|
| 197 |
"41": "SUCCESS",
|
| 198 |
"42": "SUCCESS",
|
|
@@ -220,9 +220,9 @@
|
|
| 220 |
"62": "SUCCESS",
|
| 221 |
"63": "SUCCESS",
|
| 222 |
"64": "SUCCESS",
|
| 223 |
-
"65": "
|
| 224 |
"66": "NON_PARTICIPATING",
|
| 225 |
-
"67": "
|
| 226 |
"68": "SUCCESS",
|
| 227 |
"69": "SUCCESS",
|
| 228 |
"7": "SUCCESS",
|
|
@@ -237,8 +237,8 @@
|
|
| 237 |
"78": "SUCCESS",
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "SUCCESS",
|
| 240 |
-
"80": "
|
| 241 |
-
"81": "
|
| 242 |
"82": "SUCCESS",
|
| 243 |
"83": "SUCCESS",
|
| 244 |
"84": "SUCCESS",
|
|
@@ -249,7 +249,7 @@
|
|
| 249 |
"89": "SUCCESS",
|
| 250 |
"9": "SUCCESS",
|
| 251 |
"90": "SUCCESS",
|
| 252 |
-
"91": "
|
| 253 |
"92": "SUCCESS",
|
| 254 |
"93": "NON_PARTICIPATING",
|
| 255 |
"94": "SUCCESS",
|
|
@@ -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": "NON_PARTICIPATING",
|
| 7 |
"10": "NON_PARTICIPATING",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
+
"101": "NON_PARTICIPATING",
|
| 10 |
+
"102": "NON_PARTICIPATING",
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
+
"104": "NON_PARTICIPATING",
|
| 13 |
"105": "SUCCESS",
|
| 14 |
+
"106": "SUCCESS",
|
| 15 |
"107": "NON_PARTICIPATING",
|
| 16 |
"108": "NON_PARTICIPATING",
|
| 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",
|
| 24 |
"115": "NON_PARTICIPATING",
|
| 25 |
"116": "NON_PARTICIPATING",
|
| 26 |
"117": "NON_PARTICIPATING",
|
| 27 |
+
"118": "NON_PARTICIPATING",
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
"12": "SUCCESS",
|
| 30 |
+
"120": "SUCCESS",
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
+
"122": "NON_PARTICIPATING",
|
| 33 |
+
"123": "NON_PARTICIPATING",
|
| 34 |
"124": "NON_PARTICIPATING",
|
| 35 |
+
"125": "NON_PARTICIPATING",
|
| 36 |
+
"126": "NON_PARTICIPATING",
|
| 37 |
"127": "NON_PARTICIPATING",
|
| 38 |
"128": "NON_PARTICIPATING",
|
| 39 |
"129": "NON_PARTICIPATING",
|
|
|
|
| 42 |
"131": "NON_PARTICIPATING",
|
| 43 |
"132": "NON_PARTICIPATING",
|
| 44 |
"133": "NON_PARTICIPATING",
|
| 45 |
+
"134": "NON_PARTICIPATING",
|
| 46 |
"135": "NON_PARTICIPATING",
|
| 47 |
"136": "NON_PARTICIPATING",
|
| 48 |
+
"137": "NON_PARTICIPATING",
|
| 49 |
+
"138": "NON_PARTICIPATING",
|
| 50 |
"139": "SUCCESS",
|
| 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": "NON_PARTICIPATING",
|
| 60 |
"148": "NON_PARTICIPATING",
|
| 61 |
"149": "NON_PARTICIPATING",
|
| 62 |
+
"15": "SUCCESS",
|
| 63 |
+
"150": "SUCCESS",
|
| 64 |
"151": "NON_PARTICIPATING",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
"153": "NON_PARTICIPATING",
|
| 67 |
"154": "NON_PARTICIPATING",
|
| 68 |
"155": "NON_PARTICIPATING",
|
| 69 |
+
"156": "SUCCESS",
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
"159": "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": "SUCCESS",
|
| 91 |
"176": "NON_PARTICIPATING",
|
| 92 |
"177": "SUCCESS",
|
| 93 |
"178": "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",
|
|
|
|
| 121 |
"202": "SUCCESS",
|
| 122 |
"203": "NON_PARTICIPATING",
|
| 123 |
"204": "NON_PARTICIPATING",
|
| 124 |
+
"205": "SUCCESS",
|
| 125 |
"206": "NON_PARTICIPATING",
|
| 126 |
"207": "NON_PARTICIPATING",
|
| 127 |
+
"208": "SUCCESS",
|
| 128 |
"209": "NON_PARTICIPATING",
|
| 129 |
"21": "SUCCESS",
|
| 130 |
"210": "NON_PARTICIPATING",
|
|
|
|
| 133 |
"213": "NON_PARTICIPATING",
|
| 134 |
"214": "NON_PARTICIPATING",
|
| 135 |
"215": "NON_PARTICIPATING",
|
| 136 |
+
"216": "NON_PARTICIPATING",
|
| 137 |
"217": "NON_PARTICIPATING",
|
| 138 |
"218": "NON_PARTICIPATING",
|
| 139 |
"219": "NON_PARTICIPATING",
|
|
|
|
| 151 |
"23": "SUCCESS",
|
| 152 |
"230": "SUCCESS",
|
| 153 |
"231": "NON_PARTICIPATING",
|
| 154 |
+
"232": "SUCCESS",
|
| 155 |
"233": "SUCCESS",
|
| 156 |
"234": "NON_PARTICIPATING",
|
| 157 |
"235": "NON_PARTICIPATING",
|
|
|
|
| 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": "NON_PARTICIPATING",
|
| 171 |
"248": "NON_PARTICIPATING",
|
| 172 |
"249": "NON_PARTICIPATING",
|
|
|
|
| 175 |
"251": "NON_PARTICIPATING",
|
| 176 |
"252": "NON_PARTICIPATING",
|
| 177 |
"253": "NON_PARTICIPATING",
|
| 178 |
+
"254": "NON_PARTICIPATING",
|
| 179 |
"255": "NON_PARTICIPATING",
|
| 180 |
"26": "SUCCESS",
|
| 181 |
"27": "SUCCESS",
|
|
|
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "NON_PARTICIPATING",
|
| 190 |
"35": "SUCCESS",
|
| 191 |
+
"36": "SUCCESS",
|
| 192 |
"37": "SUCCESS",
|
| 193 |
"38": "SUCCESS",
|
| 194 |
"39": "SUCCESS",
|
| 195 |
+
"4": "SUCCESS",
|
| 196 |
"40": "SUCCESS",
|
| 197 |
"41": "SUCCESS",
|
| 198 |
"42": "SUCCESS",
|
|
|
|
| 220 |
"62": "SUCCESS",
|
| 221 |
"63": "SUCCESS",
|
| 222 |
"64": "SUCCESS",
|
| 223 |
+
"65": "SUCCESS",
|
| 224 |
"66": "NON_PARTICIPATING",
|
| 225 |
+
"67": "SUCCESS",
|
| 226 |
"68": "SUCCESS",
|
| 227 |
"69": "SUCCESS",
|
| 228 |
"7": "SUCCESS",
|
|
|
|
| 237 |
"78": "SUCCESS",
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "SUCCESS",
|
| 240 |
+
"80": "SUCCESS",
|
| 241 |
+
"81": "NON_PARTICIPATING",
|
| 242 |
"82": "SUCCESS",
|
| 243 |
"83": "SUCCESS",
|
| 244 |
"84": "SUCCESS",
|
|
|
|
| 249 |
"89": "SUCCESS",
|
| 250 |
"9": "SUCCESS",
|
| 251 |
"90": "SUCCESS",
|
| 252 |
+
"91": "SUCCESS",
|
| 253 |
"92": "SUCCESS",
|
| 254 |
"93": "NON_PARTICIPATING",
|
| 255 |
"94": "SUCCESS",
|
|
|
|
| 268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
| 269 |
},
|
| 270 |
"block_list": [
|
| 271 |
+
5643200,
|
| 272 |
+
5643251
|
| 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": 2,
|
| 280 |
"inner_steps": 0,
|
| 281 |
+
"last_allreduce_block": 5643114,
|
| 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 sha256:9d024e6a1d8d1025676a6399e1efb6d2684f1b2141c9632c76bd1a9e2a6bb56d
|
| 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:e6d8d41ef58bd7fdaac1c78c11805658106ccc659c53e6586222f300b1d46048
|
| 3 |
size 4040701744
|