Run 4. Outer Step 78. Inner Step 0. Peers 35.
Browse files- config.json +74 -74
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
- outer_optimizer.pt +1 -1
config.json
CHANGED
@@ -3,11 +3,11 @@
|
|
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": "NON_PARTICIPATING",
|
13 |
"105": "NON_PARTICIPATING",
|
@@ -15,7 +15,7 @@
|
|
15 |
"107": "NON_PARTICIPATING",
|
16 |
"108": "NON_PARTICIPATING",
|
17 |
"109": "NON_PARTICIPATING",
|
18 |
-
"11": "
|
19 |
"110": "NON_PARTICIPATING",
|
20 |
"111": "NON_PARTICIPATING",
|
21 |
"112": "NON_PARTICIPATING",
|
@@ -28,20 +28,20 @@
|
|
28 |
"119": "NON_PARTICIPATING",
|
29 |
"12": "SUCCESS",
|
30 |
"120": "NON_PARTICIPATING",
|
31 |
-
"121": "
|
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": "
|
39 |
"129": "NON_PARTICIPATING",
|
40 |
"13": "SUCCESS",
|
41 |
"130": "NON_PARTICIPATING",
|
42 |
"131": "NON_PARTICIPATING",
|
43 |
"132": "NON_PARTICIPATING",
|
44 |
-
"133": "
|
45 |
"134": "NON_PARTICIPATING",
|
46 |
"135": "NON_PARTICIPATING",
|
47 |
"136": "NON_PARTICIPATING",
|
@@ -54,7 +54,7 @@
|
|
54 |
"142": "NON_PARTICIPATING",
|
55 |
"143": "NON_PARTICIPATING",
|
56 |
"144": "NON_PARTICIPATING",
|
57 |
-
"145": "
|
58 |
"146": "NON_PARTICIPATING",
|
59 |
"147": "NON_PARTICIPATING",
|
60 |
"148": "NON_PARTICIPATING",
|
@@ -62,7 +62,7 @@
|
|
62 |
"15": "SUCCESS",
|
63 |
"150": "NON_PARTICIPATING",
|
64 |
"151": "NON_PARTICIPATING",
|
65 |
-
"152": "
|
66 |
"153": "NON_PARTICIPATING",
|
67 |
"154": "NON_PARTICIPATING",
|
68 |
"155": "NON_PARTICIPATING",
|
@@ -81,9 +81,9 @@
|
|
81 |
"167": "NON_PARTICIPATING",
|
82 |
"168": "NON_PARTICIPATING",
|
83 |
"169": "NON_PARTICIPATING",
|
84 |
-
"17": "
|
85 |
"170": "NON_PARTICIPATING",
|
86 |
-
"171": "
|
87 |
"172": "NON_PARTICIPATING",
|
88 |
"173": "NON_PARTICIPATING",
|
89 |
"174": "NON_PARTICIPATING",
|
@@ -92,7 +92,7 @@
|
|
92 |
"177": "NON_PARTICIPATING",
|
93 |
"178": "NON_PARTICIPATING",
|
94 |
"179": "NON_PARTICIPATING",
|
95 |
-
"18": "
|
96 |
"180": "NON_PARTICIPATING",
|
97 |
"181": "NON_PARTICIPATING",
|
98 |
"182": "NON_PARTICIPATING",
|
@@ -103,7 +103,7 @@
|
|
103 |
"187": "NON_PARTICIPATING",
|
104 |
"188": "NON_PARTICIPATING",
|
105 |
"189": "NON_PARTICIPATING",
|
106 |
-
"19": "
|
107 |
"190": "NON_PARTICIPATING",
|
108 |
"191": "NON_PARTICIPATING",
|
109 |
"192": "NON_PARTICIPATING",
|
@@ -115,8 +115,8 @@
|
|
115 |
"198": "NON_PARTICIPATING",
|
116 |
"199": "NON_PARTICIPATING",
|
117 |
"2": "SUCCESS",
|
118 |
-
"20": "
|
119 |
-
"200": "
|
120 |
"201": "NON_PARTICIPATING",
|
121 |
"202": "NON_PARTICIPATING",
|
122 |
"203": "NON_PARTICIPATING",
|
@@ -124,28 +124,28 @@
|
|
124 |
"205": "NON_PARTICIPATING",
|
125 |
"206": "NON_PARTICIPATING",
|
126 |
"207": "NON_PARTICIPATING",
|
127 |
-
"208": "
|
128 |
-
"209": "
|
129 |
"21": "SUCCESS",
|
130 |
"210": "NON_PARTICIPATING",
|
131 |
"211": "NON_PARTICIPATING",
|
132 |
-
"212": "
|
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",
|
140 |
-
"22": "
|
141 |
"220": "NON_PARTICIPATING",
|
142 |
"221": "NON_PARTICIPATING",
|
143 |
-
"222": "
|
144 |
-
"223": "
|
145 |
"224": "NON_PARTICIPATING",
|
146 |
"225": "NON_PARTICIPATING",
|
147 |
"226": "NON_PARTICIPATING",
|
148 |
-
"227": "
|
149 |
"228": "NON_PARTICIPATING",
|
150 |
"229": "NON_PARTICIPATING",
|
151 |
"23": "SUCCESS",
|
@@ -159,11 +159,11 @@
|
|
159 |
"237": "NON_PARTICIPATING",
|
160 |
"238": "NON_PARTICIPATING",
|
161 |
"239": "NON_PARTICIPATING",
|
162 |
-
"24": "
|
163 |
"240": "NON_PARTICIPATING",
|
164 |
"241": "NON_PARTICIPATING",
|
165 |
"242": "NON_PARTICIPATING",
|
166 |
-
"243": "
|
167 |
"244": "NON_PARTICIPATING",
|
168 |
"245": "NON_PARTICIPATING",
|
169 |
"246": "NON_PARTICIPATING",
|
@@ -176,88 +176,88 @@
|
|
176 |
"252": "NON_PARTICIPATING",
|
177 |
"253": "NON_PARTICIPATING",
|
178 |
"254": "NON_PARTICIPATING",
|
179 |
-
"255": "
|
180 |
"26": "NON_PARTICIPATING",
|
181 |
"27": "SUCCESS",
|
182 |
-
"28": "
|
183 |
-
"29": "
|
184 |
-
"3": "
|
185 |
"30": "NON_PARTICIPATING",
|
186 |
-
"31": "
|
187 |
-
"32": "
|
188 |
"33": "NON_PARTICIPATING",
|
189 |
-
"34": "
|
190 |
-
"35": "
|
191 |
-
"36": "
|
192 |
"37": "SUCCESS",
|
193 |
-
"38": "
|
194 |
"39": "SUCCESS",
|
195 |
-
"4": "
|
196 |
-
"40": "
|
197 |
"41": "SUCCESS",
|
198 |
-
"42": "
|
199 |
-
"43": "
|
200 |
"44": "SUCCESS",
|
201 |
-
"45": "
|
202 |
"46": "NON_PARTICIPATING",
|
203 |
-
"47": "
|
204 |
"48": "NON_PARTICIPATING",
|
205 |
"49": "SUCCESS",
|
206 |
-
"5": "
|
207 |
"50": "SUCCESS",
|
208 |
-
"51": "
|
209 |
-
"52": "
|
210 |
"53": "SUCCESS",
|
211 |
"54": "SUCCESS",
|
212 |
-
"55": "
|
213 |
-
"56": "
|
214 |
"57": "NON_PARTICIPATING",
|
215 |
"58": "NON_PARTICIPATING",
|
216 |
"59": "NON_PARTICIPATING",
|
217 |
"6": "SUCCESS",
|
218 |
"60": "NON_PARTICIPATING",
|
219 |
-
"61": "
|
220 |
-
"62": "
|
221 |
-
"63": "
|
222 |
-
"64": "
|
223 |
"65": "SUCCESS",
|
224 |
-
"66": "
|
225 |
-
"67": "
|
226 |
-
"68": "
|
227 |
-
"69": "
|
228 |
-
"7": "
|
229 |
-
"70": "
|
230 |
-
"71": "
|
231 |
-
"72": "
|
232 |
"73": "SUCCESS",
|
233 |
-
"74": "
|
234 |
"75": "NON_PARTICIPATING",
|
235 |
-
"76": "
|
236 |
-
"77": "
|
237 |
-
"78": "
|
238 |
"79": "NON_PARTICIPATING",
|
239 |
-
"8": "
|
240 |
"80": "SUCCESS",
|
241 |
"81": "SUCCESS",
|
242 |
-
"82": "
|
243 |
-
"83": "
|
244 |
-
"84": "
|
245 |
"85": "NON_PARTICIPATING",
|
246 |
"86": "NON_PARTICIPATING",
|
247 |
"87": "SUCCESS",
|
248 |
-
"88": "
|
249 |
-
"89": "
|
250 |
"9": "SUCCESS",
|
251 |
"90": "SUCCESS",
|
252 |
-
"91": "
|
253 |
"92": "SUCCESS",
|
254 |
"93": "NON_PARTICIPATING",
|
255 |
-
"94": "
|
256 |
"95": "NON_PARTICIPATING",
|
257 |
-
"96": "
|
258 |
-
"97": "
|
259 |
"98": "SUCCESS",
|
260 |
-
"99": "
|
261 |
},
|
262 |
"architectures": [
|
263 |
"GPTOptim"
|
@@ -275,7 +275,7 @@
|
|
275 |
"initializer_range": 0.02,
|
276 |
"inner_step": 0,
|
277 |
"inner_steps": 0,
|
278 |
-
"last_allreduce_block":
|
279 |
"layer_norm_epsilon": 1e-05,
|
280 |
"model_type": "gpt_optimized",
|
281 |
"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": "SUCCESS",
|
11 |
"103": "NON_PARTICIPATING",
|
12 |
"104": "NON_PARTICIPATING",
|
13 |
"105": "NON_PARTICIPATING",
|
|
|
15 |
"107": "NON_PARTICIPATING",
|
16 |
"108": "NON_PARTICIPATING",
|
17 |
"109": "NON_PARTICIPATING",
|
18 |
+
"11": "NON_PARTICIPATING",
|
19 |
"110": "NON_PARTICIPATING",
|
20 |
"111": "NON_PARTICIPATING",
|
21 |
"112": "NON_PARTICIPATING",
|
|
|
28 |
"119": "NON_PARTICIPATING",
|
29 |
"12": "SUCCESS",
|
30 |
"120": "NON_PARTICIPATING",
|
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",
|
40 |
"13": "SUCCESS",
|
41 |
"130": "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",
|
|
|
54 |
"142": "NON_PARTICIPATING",
|
55 |
"143": "NON_PARTICIPATING",
|
56 |
"144": "NON_PARTICIPATING",
|
57 |
+
"145": "SUCCESS",
|
58 |
"146": "NON_PARTICIPATING",
|
59 |
"147": "NON_PARTICIPATING",
|
60 |
"148": "NON_PARTICIPATING",
|
|
|
62 |
"15": "SUCCESS",
|
63 |
"150": "NON_PARTICIPATING",
|
64 |
"151": "NON_PARTICIPATING",
|
65 |
+
"152": "NON_PARTICIPATING",
|
66 |
"153": "NON_PARTICIPATING",
|
67 |
"154": "NON_PARTICIPATING",
|
68 |
"155": "NON_PARTICIPATING",
|
|
|
81 |
"167": "NON_PARTICIPATING",
|
82 |
"168": "NON_PARTICIPATING",
|
83 |
"169": "NON_PARTICIPATING",
|
84 |
+
"17": "NON_PARTICIPATING",
|
85 |
"170": "NON_PARTICIPATING",
|
86 |
+
"171": "NON_PARTICIPATING",
|
87 |
"172": "NON_PARTICIPATING",
|
88 |
"173": "NON_PARTICIPATING",
|
89 |
"174": "NON_PARTICIPATING",
|
|
|
92 |
"177": "NON_PARTICIPATING",
|
93 |
"178": "NON_PARTICIPATING",
|
94 |
"179": "NON_PARTICIPATING",
|
95 |
+
"18": "NON_PARTICIPATING",
|
96 |
"180": "NON_PARTICIPATING",
|
97 |
"181": "NON_PARTICIPATING",
|
98 |
"182": "NON_PARTICIPATING",
|
|
|
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",
|
|
|
115 |
"198": "NON_PARTICIPATING",
|
116 |
"199": "NON_PARTICIPATING",
|
117 |
"2": "SUCCESS",
|
118 |
+
"20": "NON_PARTICIPATING",
|
119 |
+
"200": "NON_PARTICIPATING",
|
120 |
"201": "NON_PARTICIPATING",
|
121 |
"202": "NON_PARTICIPATING",
|
122 |
"203": "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": "SUCCESS",
|
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",
|
140 |
+
"22": "NON_PARTICIPATING",
|
141 |
"220": "NON_PARTICIPATING",
|
142 |
"221": "NON_PARTICIPATING",
|
143 |
+
"222": "NON_PARTICIPATING",
|
144 |
+
"223": "SUCCESS",
|
145 |
"224": "NON_PARTICIPATING",
|
146 |
"225": "NON_PARTICIPATING",
|
147 |
"226": "NON_PARTICIPATING",
|
148 |
+
"227": "NON_PARTICIPATING",
|
149 |
"228": "NON_PARTICIPATING",
|
150 |
"229": "NON_PARTICIPATING",
|
151 |
"23": "SUCCESS",
|
|
|
159 |
"237": "NON_PARTICIPATING",
|
160 |
"238": "NON_PARTICIPATING",
|
161 |
"239": "NON_PARTICIPATING",
|
162 |
+
"24": "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",
|
|
|
176 |
"252": "NON_PARTICIPATING",
|
177 |
"253": "NON_PARTICIPATING",
|
178 |
"254": "NON_PARTICIPATING",
|
179 |
+
"255": "NON_PARTICIPATING",
|
180 |
"26": "NON_PARTICIPATING",
|
181 |
"27": "SUCCESS",
|
182 |
+
"28": "NON_PARTICIPATING",
|
183 |
+
"29": "NON_PARTICIPATING",
|
184 |
+
"3": "NON_PARTICIPATING",
|
185 |
"30": "NON_PARTICIPATING",
|
186 |
+
"31": "NON_PARTICIPATING",
|
187 |
+
"32": "NON_PARTICIPATING",
|
188 |
"33": "NON_PARTICIPATING",
|
189 |
+
"34": "SUCCESS",
|
190 |
+
"35": "NON_PARTICIPATING",
|
191 |
+
"36": "NON_PARTICIPATING",
|
192 |
"37": "SUCCESS",
|
193 |
+
"38": "NON_PARTICIPATING",
|
194 |
"39": "SUCCESS",
|
195 |
+
"4": "NON_PARTICIPATING",
|
196 |
+
"40": "NON_PARTICIPATING",
|
197 |
"41": "SUCCESS",
|
198 |
+
"42": "NON_PARTICIPATING",
|
199 |
+
"43": "NON_PARTICIPATING",
|
200 |
"44": "SUCCESS",
|
201 |
+
"45": "NON_PARTICIPATING",
|
202 |
"46": "NON_PARTICIPATING",
|
203 |
+
"47": "NON_PARTICIPATING",
|
204 |
"48": "NON_PARTICIPATING",
|
205 |
"49": "SUCCESS",
|
206 |
+
"5": "NON_PARTICIPATING",
|
207 |
"50": "SUCCESS",
|
208 |
+
"51": "NON_PARTICIPATING",
|
209 |
+
"52": "NON_PARTICIPATING",
|
210 |
"53": "SUCCESS",
|
211 |
"54": "SUCCESS",
|
212 |
+
"55": "NON_PARTICIPATING",
|
213 |
+
"56": "NON_PARTICIPATING",
|
214 |
"57": "NON_PARTICIPATING",
|
215 |
"58": "NON_PARTICIPATING",
|
216 |
"59": "NON_PARTICIPATING",
|
217 |
"6": "SUCCESS",
|
218 |
"60": "NON_PARTICIPATING",
|
219 |
+
"61": "NON_PARTICIPATING",
|
220 |
+
"62": "NON_PARTICIPATING",
|
221 |
+
"63": "NON_PARTICIPATING",
|
222 |
+
"64": "SUCCESS",
|
223 |
"65": "SUCCESS",
|
224 |
+
"66": "NON_PARTICIPATING",
|
225 |
+
"67": "NON_PARTICIPATING",
|
226 |
+
"68": "NON_PARTICIPATING",
|
227 |
+
"69": "NON_PARTICIPATING",
|
228 |
+
"7": "NON_PARTICIPATING",
|
229 |
+
"70": "NON_PARTICIPATING",
|
230 |
+
"71": "NON_PARTICIPATING",
|
231 |
+
"72": "NON_PARTICIPATING",
|
232 |
"73": "SUCCESS",
|
233 |
+
"74": "NON_PARTICIPATING",
|
234 |
"75": "NON_PARTICIPATING",
|
235 |
+
"76": "NON_PARTICIPATING",
|
236 |
+
"77": "NON_PARTICIPATING",
|
237 |
+
"78": "NON_PARTICIPATING",
|
238 |
"79": "NON_PARTICIPATING",
|
239 |
+
"8": "NON_PARTICIPATING",
|
240 |
"80": "SUCCESS",
|
241 |
"81": "SUCCESS",
|
242 |
+
"82": "NON_PARTICIPATING",
|
243 |
+
"83": "NON_PARTICIPATING",
|
244 |
+
"84": "NON_PARTICIPATING",
|
245 |
"85": "NON_PARTICIPATING",
|
246 |
"86": "NON_PARTICIPATING",
|
247 |
"87": "SUCCESS",
|
248 |
+
"88": "NON_PARTICIPATING",
|
249 |
+
"89": "NON_PARTICIPATING",
|
250 |
"9": "SUCCESS",
|
251 |
"90": "SUCCESS",
|
252 |
+
"91": "NON_PARTICIPATING",
|
253 |
"92": "SUCCESS",
|
254 |
"93": "NON_PARTICIPATING",
|
255 |
+
"94": "NON_PARTICIPATING",
|
256 |
"95": "NON_PARTICIPATING",
|
257 |
+
"96": "NON_PARTICIPATING",
|
258 |
+
"97": "NON_PARTICIPATING",
|
259 |
"98": "SUCCESS",
|
260 |
+
"99": "NON_PARTICIPATING"
|
261 |
},
|
262 |
"architectures": [
|
263 |
"GPTOptim"
|
|
|
275 |
"initializer_range": 0.02,
|
276 |
"inner_step": 0,
|
277 |
"inner_steps": 0,
|
278 |
+
"last_allreduce_block": 5665748,
|
279 |
"layer_norm_epsilon": 1e-05,
|
280 |
"model_type": "gpt_optimized",
|
281 |
"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 sha256:58d247a2ff38a1668c26ed4b2c5eee13664bcfaf79f27b9167d3ca85fc7cf999
|
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 sha256:011617872d27740638fa544a890b2f4ffb1ef4a2ba59167afbe52cfa4244321d
|
3 |
size 4040701744
|
outer_optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 4040805354
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f5fe374380cebeaef48d7a2fdc9b6414bc91a7cd2a5cf288a8c2acf097b6335b
|
3 |
size 4040805354
|