Run 4. Outer Step 88. Inner Step 0. Peers 83.
Browse files- config.json +23 -23
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
- outer_optimizer.pt +1 -1
config.json
CHANGED
@@ -7,11 +7,11 @@
|
|
7 |
"10": "NON_PARTICIPATING",
|
8 |
"100": "NON_PARTICIPATING",
|
9 |
"101": "NON_PARTICIPATING",
|
10 |
-
"102": "
|
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",
|
@@ -28,15 +28,15 @@
|
|
28 |
"119": "NON_PARTICIPATING",
|
29 |
"12": "SUCCESS",
|
30 |
"120": "SUCCESS",
|
31 |
-
"121": "
|
32 |
"122": "NON_PARTICIPATING",
|
33 |
"123": "NON_PARTICIPATING",
|
34 |
"124": "NON_PARTICIPATING",
|
35 |
"125": "NON_PARTICIPATING",
|
36 |
"126": "NON_PARTICIPATING",
|
37 |
-
"127": "
|
38 |
"128": "NON_PARTICIPATING",
|
39 |
-
"129": "
|
40 |
"13": "SUCCESS",
|
41 |
"130": "NON_PARTICIPATING",
|
42 |
"131": "NON_PARTICIPATING",
|
@@ -54,10 +54,10 @@
|
|
54 |
"142": "NON_PARTICIPATING",
|
55 |
"143": "NON_PARTICIPATING",
|
56 |
"144": "NON_PARTICIPATING",
|
57 |
-
"145": "
|
58 |
"146": "NON_PARTICIPATING",
|
59 |
"147": "NON_PARTICIPATING",
|
60 |
-
"148": "
|
61 |
"149": "NON_PARTICIPATING",
|
62 |
"15": "SUCCESS",
|
63 |
"150": "NON_PARTICIPATING",
|
@@ -73,9 +73,9 @@
|
|
73 |
"16": "SUCCESS",
|
74 |
"160": "NON_PARTICIPATING",
|
75 |
"161": "NON_PARTICIPATING",
|
76 |
-
"162": "
|
77 |
"163": "NON_PARTICIPATING",
|
78 |
-
"164": "
|
79 |
"165": "NON_PARTICIPATING",
|
80 |
"166": "NON_PARTICIPATING",
|
81 |
"167": "NON_PARTICIPATING",
|
@@ -96,12 +96,12 @@
|
|
96 |
"180": "NON_PARTICIPATING",
|
97 |
"181": "NON_PARTICIPATING",
|
98 |
"182": "NON_PARTICIPATING",
|
99 |
-
"183": "
|
100 |
"184": "NON_PARTICIPATING",
|
101 |
"185": "NON_PARTICIPATING",
|
102 |
-
"186": "
|
103 |
"187": "NON_PARTICIPATING",
|
104 |
-
"188": "
|
105 |
"189": "NON_PARTICIPATING",
|
106 |
"19": "SUCCESS",
|
107 |
"190": "NON_PARTICIPATING",
|
@@ -110,7 +110,7 @@
|
|
110 |
"193": "NON_PARTICIPATING",
|
111 |
"194": "NON_PARTICIPATING",
|
112 |
"195": "NON_PARTICIPATING",
|
113 |
-
"196": "
|
114 |
"197": "SUCCESS",
|
115 |
"198": "NON_PARTICIPATING",
|
116 |
"199": "NON_PARTICIPATING",
|
@@ -124,7 +124,7 @@
|
|
124 |
"205": "NON_PARTICIPATING",
|
125 |
"206": "NON_PARTICIPATING",
|
126 |
"207": "NON_PARTICIPATING",
|
127 |
-
"208": "
|
128 |
"209": "NON_PARTICIPATING",
|
129 |
"21": "SUCCESS",
|
130 |
"210": "NON_PARTICIPATING",
|
@@ -133,9 +133,9 @@
|
|
133 |
"213": "NON_PARTICIPATING",
|
134 |
"214": "NON_PARTICIPATING",
|
135 |
"215": "NON_PARTICIPATING",
|
136 |
-
"216": "
|
137 |
"217": "NON_PARTICIPATING",
|
138 |
-
"218": "
|
139 |
"219": "NON_PARTICIPATING",
|
140 |
"22": "SUCCESS",
|
141 |
"220": "NON_PARTICIPATING",
|
@@ -146,13 +146,13 @@
|
|
146 |
"225": "NON_PARTICIPATING",
|
147 |
"226": "NON_PARTICIPATING",
|
148 |
"227": "NON_PARTICIPATING",
|
149 |
-
"228": "
|
150 |
-
"229": "
|
151 |
"23": "NON_PARTICIPATING",
|
152 |
"230": "NON_PARTICIPATING",
|
153 |
"231": "NON_PARTICIPATING",
|
154 |
"232": "NON_PARTICIPATING",
|
155 |
-
"233": "
|
156 |
"234": "NON_PARTICIPATING",
|
157 |
"235": "NON_PARTICIPATING",
|
158 |
"236": "NON_PARTICIPATING",
|
@@ -168,7 +168,7 @@
|
|
168 |
"245": "NON_PARTICIPATING",
|
169 |
"246": "NON_PARTICIPATING",
|
170 |
"247": "NON_PARTICIPATING",
|
171 |
-
"248": "
|
172 |
"249": "NON_PARTICIPATING",
|
173 |
"25": "SUCCESS",
|
174 |
"250": "NON_PARTICIPATING",
|
@@ -177,7 +177,7 @@
|
|
177 |
"253": "NON_PARTICIPATING",
|
178 |
"254": "NON_PARTICIPATING",
|
179 |
"255": "NON_PARTICIPATING",
|
180 |
-
"26": "
|
181 |
"27": "SUCCESS",
|
182 |
"28": "SUCCESS",
|
183 |
"29": "SUCCESS",
|
@@ -247,7 +247,7 @@
|
|
247 |
"87": "SUCCESS",
|
248 |
"88": "SUCCESS",
|
249 |
"89": "SUCCESS",
|
250 |
-
"9": "
|
251 |
"90": "SUCCESS",
|
252 |
"91": "SUCCESS",
|
253 |
"92": "NON_PARTICIPATING",
|
@@ -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,
|
|
|
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",
|
14 |
+
"106": "SUCCESS",
|
15 |
"107": "NON_PARTICIPATING",
|
16 |
"108": "NON_PARTICIPATING",
|
17 |
"109": "NON_PARTICIPATING",
|
|
|
28 |
"119": "NON_PARTICIPATING",
|
29 |
"12": "SUCCESS",
|
30 |
"120": "SUCCESS",
|
31 |
+
"121": "SUCCESS",
|
32 |
"122": "NON_PARTICIPATING",
|
33 |
"123": "NON_PARTICIPATING",
|
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": "SUCCESS",
|
41 |
"130": "NON_PARTICIPATING",
|
42 |
"131": "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",
|
|
|
73 |
"16": "SUCCESS",
|
74 |
"160": "NON_PARTICIPATING",
|
75 |
"161": "NON_PARTICIPATING",
|
76 |
+
"162": "NON_PARTICIPATING",
|
77 |
"163": "NON_PARTICIPATING",
|
78 |
+
"164": "NON_PARTICIPATING",
|
79 |
"165": "NON_PARTICIPATING",
|
80 |
"166": "NON_PARTICIPATING",
|
81 |
"167": "NON_PARTICIPATING",
|
|
|
96 |
"180": "NON_PARTICIPATING",
|
97 |
"181": "NON_PARTICIPATING",
|
98 |
"182": "NON_PARTICIPATING",
|
99 |
+
"183": "NON_PARTICIPATING",
|
100 |
"184": "NON_PARTICIPATING",
|
101 |
"185": "NON_PARTICIPATING",
|
102 |
+
"186": "SUCCESS",
|
103 |
"187": "NON_PARTICIPATING",
|
104 |
+
"188": "SUCCESS",
|
105 |
"189": "NON_PARTICIPATING",
|
106 |
"19": "SUCCESS",
|
107 |
"190": "NON_PARTICIPATING",
|
|
|
110 |
"193": "NON_PARTICIPATING",
|
111 |
"194": "NON_PARTICIPATING",
|
112 |
"195": "NON_PARTICIPATING",
|
113 |
+
"196": "NON_PARTICIPATING",
|
114 |
"197": "SUCCESS",
|
115 |
"198": "NON_PARTICIPATING",
|
116 |
"199": "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",
|
|
|
133 |
"213": "NON_PARTICIPATING",
|
134 |
"214": "NON_PARTICIPATING",
|
135 |
"215": "NON_PARTICIPATING",
|
136 |
+
"216": "SUCCESS",
|
137 |
"217": "NON_PARTICIPATING",
|
138 |
+
"218": "SUCCESS",
|
139 |
"219": "NON_PARTICIPATING",
|
140 |
"22": "SUCCESS",
|
141 |
"220": "NON_PARTICIPATING",
|
|
|
146 |
"225": "NON_PARTICIPATING",
|
147 |
"226": "NON_PARTICIPATING",
|
148 |
"227": "NON_PARTICIPATING",
|
149 |
+
"228": "NON_PARTICIPATING",
|
150 |
+
"229": "SUCCESS",
|
151 |
"23": "NON_PARTICIPATING",
|
152 |
"230": "NON_PARTICIPATING",
|
153 |
"231": "NON_PARTICIPATING",
|
154 |
"232": "NON_PARTICIPATING",
|
155 |
+
"233": "NON_PARTICIPATING",
|
156 |
"234": "NON_PARTICIPATING",
|
157 |
"235": "NON_PARTICIPATING",
|
158 |
"236": "NON_PARTICIPATING",
|
|
|
168 |
"245": "NON_PARTICIPATING",
|
169 |
"246": "NON_PARTICIPATING",
|
170 |
"247": "NON_PARTICIPATING",
|
171 |
+
"248": "SUCCESS",
|
172 |
"249": "NON_PARTICIPATING",
|
173 |
"25": "SUCCESS",
|
174 |
"250": "NON_PARTICIPATING",
|
|
|
177 |
"253": "NON_PARTICIPATING",
|
178 |
"254": "NON_PARTICIPATING",
|
179 |
"255": "NON_PARTICIPATING",
|
180 |
+
"26": "NON_PARTICIPATING",
|
181 |
"27": "SUCCESS",
|
182 |
"28": "SUCCESS",
|
183 |
"29": "SUCCESS",
|
|
|
247 |
"87": "SUCCESS",
|
248 |
"88": "SUCCESS",
|
249 |
"89": "SUCCESS",
|
250 |
+
"9": "SUCCESS",
|
251 |
"90": "SUCCESS",
|
252 |
"91": "SUCCESS",
|
253 |
"92": "NON_PARTICIPATING",
|
|
|
275 |
"initializer_range": 0.02,
|
276 |
"inner_step": 0,
|
277 |
"inner_steps": 0,
|
278 |
+
"last_allreduce_block": 5678972,
|
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:c3e72494a699d03393d570ae6979acc9bd9dad135ee39c241be748befb6871a1
|
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 sha256:442c1b443802216df8991501a3fa2f09453849fc69a4bb88a7216a182b149f97
|
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 sha256:7549f4015acbd63467a4ad746f2f03f64d6590e75cc20ae69bdc12f795d6b0fa
|
3 |
size 4040805354
|