Run 4. Outer Step 92. Inner Step 0. Peers 83.
Browse files- config.json +25 -25
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
- outer_optimizer.pt +1 -1
config.json
CHANGED
@@ -11,7 +11,7 @@
|
|
11 |
"103": "NON_PARTICIPATING",
|
12 |
"104": "NON_PARTICIPATING",
|
13 |
"105": "NON_PARTICIPATING",
|
14 |
-
"106": "
|
15 |
"107": "NON_PARTICIPATING",
|
16 |
"108": "NON_PARTICIPATING",
|
17 |
"109": "NON_PARTICIPATING",
|
@@ -19,7 +19,7 @@
|
|
19 |
"110": "NON_PARTICIPATING",
|
20 |
"111": "NON_PARTICIPATING",
|
21 |
"112": "NON_PARTICIPATING",
|
22 |
-
"113": "
|
23 |
"114": "NON_PARTICIPATING",
|
24 |
"115": "NON_PARTICIPATING",
|
25 |
"116": "NON_PARTICIPATING",
|
@@ -29,13 +29,13 @@
|
|
29 |
"12": "SUCCESS",
|
30 |
"120": "NON_PARTICIPATING",
|
31 |
"121": "NON_PARTICIPATING",
|
32 |
-
"122": "
|
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",
|
@@ -50,19 +50,19 @@
|
|
50 |
"139": "NON_PARTICIPATING",
|
51 |
"14": "SUCCESS",
|
52 |
"140": "NON_PARTICIPATING",
|
53 |
-
"141": "
|
54 |
"142": "NON_PARTICIPATING",
|
55 |
"143": "NON_PARTICIPATING",
|
56 |
"144": "NON_PARTICIPATING",
|
57 |
"145": "NON_PARTICIPATING",
|
58 |
"146": "NON_PARTICIPATING",
|
59 |
"147": "NON_PARTICIPATING",
|
60 |
-
"148": "
|
61 |
"149": "NON_PARTICIPATING",
|
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",
|
@@ -72,12 +72,12 @@
|
|
72 |
"159": "NON_PARTICIPATING",
|
73 |
"16": "SUCCESS",
|
74 |
"160": "NON_PARTICIPATING",
|
75 |
-
"161": "
|
76 |
"162": "SUCCESS",
|
77 |
"163": "NON_PARTICIPATING",
|
78 |
"164": "NON_PARTICIPATING",
|
79 |
"165": "NON_PARTICIPATING",
|
80 |
-
"166": "
|
81 |
"167": "NON_PARTICIPATING",
|
82 |
"168": "NON_PARTICIPATING",
|
83 |
"169": "NON_PARTICIPATING",
|
@@ -91,7 +91,7 @@
|
|
91 |
"176": "NON_PARTICIPATING",
|
92 |
"177": "SUCCESS",
|
93 |
"178": "NON_PARTICIPATING",
|
94 |
-
"179": "
|
95 |
"18": "SUCCESS",
|
96 |
"180": "NON_PARTICIPATING",
|
97 |
"181": "NON_PARTICIPATING",
|
@@ -120,14 +120,14 @@
|
|
120 |
"201": "NON_PARTICIPATING",
|
121 |
"202": "NON_PARTICIPATING",
|
122 |
"203": "NON_PARTICIPATING",
|
123 |
-
"204": "
|
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": "
|
131 |
"211": "NON_PARTICIPATING",
|
132 |
"212": "NON_PARTICIPATING",
|
133 |
"213": "NON_PARTICIPATING",
|
@@ -139,21 +139,21 @@
|
|
139 |
"219": "NON_PARTICIPATING",
|
140 |
"22": "SUCCESS",
|
141 |
"220": "NON_PARTICIPATING",
|
142 |
-
"221": "
|
143 |
"222": "NON_PARTICIPATING",
|
144 |
"223": "NON_PARTICIPATING",
|
145 |
"224": "NON_PARTICIPATING",
|
146 |
"225": "NON_PARTICIPATING",
|
147 |
-
"226": "
|
148 |
-
"227": "
|
149 |
"228": "NON_PARTICIPATING",
|
150 |
"229": "NON_PARTICIPATING",
|
151 |
"23": "NON_PARTICIPATING",
|
152 |
-
"230": "
|
153 |
"231": "NON_PARTICIPATING",
|
154 |
"232": "NON_PARTICIPATING",
|
155 |
"233": "NON_PARTICIPATING",
|
156 |
-
"234": "
|
157 |
"235": "NON_PARTICIPATING",
|
158 |
"236": "NON_PARTICIPATING",
|
159 |
"237": "NON_PARTICIPATING",
|
@@ -171,14 +171,14 @@
|
|
171 |
"248": "NON_PARTICIPATING",
|
172 |
"249": "NON_PARTICIPATING",
|
173 |
"25": "SUCCESS",
|
174 |
-
"250": "
|
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": "
|
182 |
"28": "SUCCESS",
|
183 |
"29": "SUCCESS",
|
184 |
"3": "NON_PARTICIPATING",
|
@@ -195,7 +195,7 @@
|
|
195 |
"4": "SUCCESS",
|
196 |
"40": "SUCCESS",
|
197 |
"41": "SUCCESS",
|
198 |
-
"42": "
|
199 |
"43": "NON_PARTICIPATING",
|
200 |
"44": "SUCCESS",
|
201 |
"45": "SUCCESS",
|
@@ -217,7 +217,7 @@
|
|
217 |
"6": "NON_PARTICIPATING",
|
218 |
"60": "SUCCESS",
|
219 |
"61": "SUCCESS",
|
220 |
-
"62": "
|
221 |
"63": "SUCCESS",
|
222 |
"64": "SUCCESS",
|
223 |
"65": "SUCCESS",
|
@@ -241,16 +241,16 @@
|
|
241 |
"81": "NON_PARTICIPATING",
|
242 |
"82": "SUCCESS",
|
243 |
"83": "SUCCESS",
|
244 |
-
"84": "
|
245 |
"85": "SUCCESS",
|
246 |
"86": "NON_PARTICIPATING",
|
247 |
"87": "SUCCESS",
|
248 |
"88": "SUCCESS",
|
249 |
-
"89": "
|
250 |
"9": "SUCCESS",
|
251 |
"90": "SUCCESS",
|
252 |
"91": "SUCCESS",
|
253 |
-
"92": "
|
254 |
"93": "NON_PARTICIPATING",
|
255 |
"94": "SUCCESS",
|
256 |
"95": "SUCCESS",
|
@@ -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,
|
|
|
11 |
"103": "NON_PARTICIPATING",
|
12 |
"104": "NON_PARTICIPATING",
|
13 |
"105": "NON_PARTICIPATING",
|
14 |
+
"106": "SUCCESS",
|
15 |
"107": "NON_PARTICIPATING",
|
16 |
"108": "NON_PARTICIPATING",
|
17 |
"109": "NON_PARTICIPATING",
|
|
|
19 |
"110": "NON_PARTICIPATING",
|
20 |
"111": "NON_PARTICIPATING",
|
21 |
"112": "NON_PARTICIPATING",
|
22 |
+
"113": "NON_PARTICIPATING",
|
23 |
"114": "NON_PARTICIPATING",
|
24 |
"115": "NON_PARTICIPATING",
|
25 |
"116": "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": "SUCCESS",
|
39 |
"129": "NON_PARTICIPATING",
|
40 |
"13": "SUCCESS",
|
41 |
"130": "NON_PARTICIPATING",
|
|
|
50 |
"139": "NON_PARTICIPATING",
|
51 |
"14": "SUCCESS",
|
52 |
"140": "NON_PARTICIPATING",
|
53 |
+
"141": "NON_PARTICIPATING",
|
54 |
"142": "NON_PARTICIPATING",
|
55 |
"143": "NON_PARTICIPATING",
|
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": "NON_PARTICIPATING",
|
64 |
"151": "NON_PARTICIPATING",
|
65 |
+
"152": "SUCCESS",
|
66 |
"153": "NON_PARTICIPATING",
|
67 |
"154": "NON_PARTICIPATING",
|
68 |
"155": "NON_PARTICIPATING",
|
|
|
72 |
"159": "NON_PARTICIPATING",
|
73 |
"16": "SUCCESS",
|
74 |
"160": "NON_PARTICIPATING",
|
75 |
+
"161": "SUCCESS",
|
76 |
"162": "SUCCESS",
|
77 |
"163": "NON_PARTICIPATING",
|
78 |
"164": "NON_PARTICIPATING",
|
79 |
"165": "NON_PARTICIPATING",
|
80 |
+
"166": "SUCCESS",
|
81 |
"167": "NON_PARTICIPATING",
|
82 |
"168": "NON_PARTICIPATING",
|
83 |
"169": "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",
|
|
|
120 |
"201": "NON_PARTICIPATING",
|
121 |
"202": "NON_PARTICIPATING",
|
122 |
"203": "NON_PARTICIPATING",
|
123 |
+
"204": "SUCCESS",
|
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": "SUCCESS",
|
131 |
"211": "NON_PARTICIPATING",
|
132 |
"212": "NON_PARTICIPATING",
|
133 |
"213": "NON_PARTICIPATING",
|
|
|
139 |
"219": "NON_PARTICIPATING",
|
140 |
"22": "SUCCESS",
|
141 |
"220": "NON_PARTICIPATING",
|
142 |
+
"221": "SUCCESS",
|
143 |
"222": "NON_PARTICIPATING",
|
144 |
"223": "NON_PARTICIPATING",
|
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": "NON_PARTICIPATING",
|
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",
|
|
|
171 |
"248": "NON_PARTICIPATING",
|
172 |
"249": "NON_PARTICIPATING",
|
173 |
"25": "SUCCESS",
|
174 |
+
"250": "SUCCESS",
|
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": "NON_PARTICIPATING",
|
182 |
"28": "SUCCESS",
|
183 |
"29": "SUCCESS",
|
184 |
"3": "NON_PARTICIPATING",
|
|
|
195 |
"4": "SUCCESS",
|
196 |
"40": "SUCCESS",
|
197 |
"41": "SUCCESS",
|
198 |
+
"42": "NON_PARTICIPATING",
|
199 |
"43": "NON_PARTICIPATING",
|
200 |
"44": "SUCCESS",
|
201 |
"45": "SUCCESS",
|
|
|
217 |
"6": "NON_PARTICIPATING",
|
218 |
"60": "SUCCESS",
|
219 |
"61": "SUCCESS",
|
220 |
+
"62": "SUCCESS",
|
221 |
"63": "SUCCESS",
|
222 |
"64": "SUCCESS",
|
223 |
"65": "SUCCESS",
|
|
|
241 |
"81": "NON_PARTICIPATING",
|
242 |
"82": "SUCCESS",
|
243 |
"83": "SUCCESS",
|
244 |
+
"84": "SUCCESS",
|
245 |
"85": "SUCCESS",
|
246 |
"86": "NON_PARTICIPATING",
|
247 |
"87": "SUCCESS",
|
248 |
"88": "SUCCESS",
|
249 |
+
"89": "NON_PARTICIPATING",
|
250 |
"9": "SUCCESS",
|
251 |
"90": "SUCCESS",
|
252 |
"91": "SUCCESS",
|
253 |
+
"92": "SUCCESS",
|
254 |
"93": "NON_PARTICIPATING",
|
255 |
"94": "SUCCESS",
|
256 |
"95": "SUCCESS",
|
|
|
275 |
"initializer_range": 0.02,
|
276 |
"inner_step": 0,
|
277 |
"inner_steps": 0,
|
278 |
+
"last_allreduce_block": 5684850,
|
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 sha256:4e7aadb404966b2d3dc06f5278259b3c872af9b429f62d4764f02cc486354f40
|
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:a1af80aed97ca489fff929b95cdd4b7fedeefa6c429ebfcbc143b3e6f111bd69
|
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 sha256:804bb683b89f96c1ccc4f9687b6431679bb7e0b57a46255b392bd7610e0d0aab
|
3 |
size 4040805354
|