Run 4. Outer Step 14. Inner Step 0. Peers 87.
Browse files- config.json +45 -45
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
- outer_optimizer.pt +1 -1
config.json
CHANGED
@@ -1,17 +1,17 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "
|
3 |
"activation_function": "gelu_new",
|
4 |
"all_reduce_scores": {
|
5 |
"0": "NON_PARTICIPATING",
|
6 |
"1": "SUCCESS",
|
7 |
"10": "SUCCESS",
|
8 |
"100": "NON_PARTICIPATING",
|
9 |
-
"101": "
|
10 |
"102": "NON_PARTICIPATING",
|
11 |
"103": "NON_PARTICIPATING",
|
12 |
"104": "NON_PARTICIPATING",
|
13 |
"105": "NON_PARTICIPATING",
|
14 |
-
"106": "
|
15 |
"107": "SUCCESS",
|
16 |
"108": "NON_PARTICIPATING",
|
17 |
"109": "SUCCESS",
|
@@ -20,17 +20,17 @@
|
|
20 |
"111": "NON_PARTICIPATING",
|
21 |
"112": "NON_PARTICIPATING",
|
22 |
"113": "NON_PARTICIPATING",
|
23 |
-
"114": "
|
24 |
"115": "NON_PARTICIPATING",
|
25 |
"116": "NON_PARTICIPATING",
|
26 |
-
"117": "
|
27 |
"118": "NON_PARTICIPATING",
|
28 |
"119": "NON_PARTICIPATING",
|
29 |
"12": "SUCCESS",
|
30 |
"120": "SUCCESS",
|
31 |
"121": "NON_PARTICIPATING",
|
32 |
"122": "NON_PARTICIPATING",
|
33 |
-
"123": "
|
34 |
"124": "NON_PARTICIPATING",
|
35 |
"125": "NON_PARTICIPATING",
|
36 |
"126": "NON_PARTICIPATING",
|
@@ -51,16 +51,16 @@
|
|
51 |
"14": "NON_PARTICIPATING",
|
52 |
"140": "NON_PARTICIPATING",
|
53 |
"141": "NON_PARTICIPATING",
|
54 |
-
"142": "
|
55 |
"143": "NON_PARTICIPATING",
|
56 |
"144": "NON_PARTICIPATING",
|
57 |
"145": "NON_PARTICIPATING",
|
58 |
"146": "NON_PARTICIPATING",
|
59 |
"147": "SUCCESS",
|
60 |
-
"148": "
|
61 |
"149": "SUCCESS",
|
62 |
"15": "SUCCESS",
|
63 |
-
"150": "
|
64 |
"151": "NON_PARTICIPATING",
|
65 |
"152": "NON_PARTICIPATING",
|
66 |
"153": "NON_PARTICIPATING",
|
@@ -70,7 +70,7 @@
|
|
70 |
"157": "NON_PARTICIPATING",
|
71 |
"158": "NON_PARTICIPATING",
|
72 |
"159": "NON_PARTICIPATING",
|
73 |
-
"16": "
|
74 |
"160": "SUCCESS",
|
75 |
"161": "NON_PARTICIPATING",
|
76 |
"162": "NON_PARTICIPATING",
|
@@ -78,8 +78,8 @@
|
|
78 |
"164": "NON_PARTICIPATING",
|
79 |
"165": "NON_PARTICIPATING",
|
80 |
"166": "NON_PARTICIPATING",
|
81 |
-
"167": "
|
82 |
-
"168": "
|
83 |
"169": "NON_PARTICIPATING",
|
84 |
"17": "SUCCESS",
|
85 |
"170": "SUCCESS",
|
@@ -99,14 +99,14 @@
|
|
99 |
"183": "NON_PARTICIPATING",
|
100 |
"184": "NON_PARTICIPATING",
|
101 |
"185": "NON_PARTICIPATING",
|
102 |
-
"186": "
|
103 |
"187": "SUCCESS",
|
104 |
"188": "NON_PARTICIPATING",
|
105 |
"189": "NON_PARTICIPATING",
|
106 |
"19": "NON_PARTICIPATING",
|
107 |
"190": "NON_PARTICIPATING",
|
108 |
-
"191": "
|
109 |
-
"192": "
|
110 |
"193": "NON_PARTICIPATING",
|
111 |
"194": "NON_PARTICIPATING",
|
112 |
"195": "SUCCESS",
|
@@ -132,25 +132,25 @@
|
|
132 |
"212": "NON_PARTICIPATING",
|
133 |
"213": "NON_PARTICIPATING",
|
134 |
"214": "SUCCESS",
|
135 |
-
"215": "
|
136 |
"216": "NON_PARTICIPATING",
|
137 |
"217": "NON_PARTICIPATING",
|
138 |
"218": "SUCCESS",
|
139 |
"219": "NON_PARTICIPATING",
|
140 |
-
"22": "
|
141 |
"220": "NON_PARTICIPATING",
|
142 |
-
"221": "
|
143 |
"222": "NON_PARTICIPATING",
|
144 |
-
"223": "
|
145 |
"224": "NON_PARTICIPATING",
|
146 |
-
"225": "
|
147 |
"226": "SUCCESS",
|
148 |
"227": "NON_PARTICIPATING",
|
149 |
"228": "NON_PARTICIPATING",
|
150 |
"229": "NON_PARTICIPATING",
|
151 |
"23": "SUCCESS",
|
152 |
"230": "NON_PARTICIPATING",
|
153 |
-
"231": "
|
154 |
"232": "NON_PARTICIPATING",
|
155 |
"233": "NON_PARTICIPATING",
|
156 |
"234": "NON_PARTICIPATING",
|
@@ -167,9 +167,9 @@
|
|
167 |
"244": "NON_PARTICIPATING",
|
168 |
"245": "NON_PARTICIPATING",
|
169 |
"246": "NON_PARTICIPATING",
|
170 |
-
"247": "
|
171 |
"248": "NON_PARTICIPATING",
|
172 |
-
"249": "
|
173 |
"25": "SUCCESS",
|
174 |
"250": "SUCCESS",
|
175 |
"251": "NON_PARTICIPATING",
|
@@ -180,9 +180,9 @@
|
|
180 |
"26": "SUCCESS",
|
181 |
"27": "SUCCESS",
|
182 |
"28": "NON_PARTICIPATING",
|
183 |
-
"29": "
|
184 |
"3": "NON_PARTICIPATING",
|
185 |
-
"30": "
|
186 |
"31": "NON_PARTICIPATING",
|
187 |
"32": "SUCCESS",
|
188 |
"33": "NON_PARTICIPATING",
|
@@ -193,12 +193,12 @@
|
|
193 |
"38": "NON_PARTICIPATING",
|
194 |
"39": "NON_PARTICIPATING",
|
195 |
"4": "NON_PARTICIPATING",
|
196 |
-
"40": "
|
197 |
-
"41": "
|
198 |
"42": "NON_PARTICIPATING",
|
199 |
-
"43": "
|
200 |
-
"44": "
|
201 |
-
"45": "
|
202 |
"46": "SUCCESS",
|
203 |
"47": "SUCCESS",
|
204 |
"48": "NON_PARTICIPATING",
|
@@ -210,19 +210,19 @@
|
|
210 |
"53": "NON_PARTICIPATING",
|
211 |
"54": "NON_PARTICIPATING",
|
212 |
"55": "SUCCESS",
|
213 |
-
"56": "
|
214 |
-
"57": "
|
215 |
-
"58": "
|
216 |
"59": "NON_PARTICIPATING",
|
217 |
"6": "SUCCESS",
|
218 |
"60": "NON_PARTICIPATING",
|
219 |
-
"61": "
|
220 |
-
"62": "
|
221 |
"63": "SUCCESS",
|
222 |
-
"64": "
|
223 |
"65": "NON_PARTICIPATING",
|
224 |
"66": "NON_PARTICIPATING",
|
225 |
-
"67": "
|
226 |
"68": "NON_PARTICIPATING",
|
227 |
"69": "NON_PARTICIPATING",
|
228 |
"7": "NON_PARTICIPATING",
|
@@ -236,27 +236,27 @@
|
|
236 |
"77": "NON_PARTICIPATING",
|
237 |
"78": "NON_PARTICIPATING",
|
238 |
"79": "NON_PARTICIPATING",
|
239 |
-
"8": "
|
240 |
"80": "SUCCESS",
|
241 |
"81": "NON_PARTICIPATING",
|
242 |
"82": "SUCCESS",
|
243 |
"83": "NON_PARTICIPATING",
|
244 |
-
"84": "
|
245 |
"85": "SUCCESS",
|
246 |
"86": "NON_PARTICIPATING",
|
247 |
"87": "SUCCESS",
|
248 |
-
"88": "
|
249 |
-
"89": "
|
250 |
"9": "NON_PARTICIPATING",
|
251 |
-
"90": "
|
252 |
"91": "SUCCESS",
|
253 |
"92": "NON_PARTICIPATING",
|
254 |
"93": "NON_PARTICIPATING",
|
255 |
"94": "NON_PARTICIPATING",
|
256 |
-
"95": "
|
257 |
"96": "SUCCESS",
|
258 |
"97": "NON_PARTICIPATING",
|
259 |
-
"98": "
|
260 |
"99": "SUCCESS"
|
261 |
},
|
262 |
"architectures": [
|
@@ -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,
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "wandb01/c66-h23",
|
3 |
"activation_function": "gelu_new",
|
4 |
"all_reduce_scores": {
|
5 |
"0": "NON_PARTICIPATING",
|
6 |
"1": "SUCCESS",
|
7 |
"10": "SUCCESS",
|
8 |
"100": "NON_PARTICIPATING",
|
9 |
+
"101": "SUCCESS",
|
10 |
"102": "NON_PARTICIPATING",
|
11 |
"103": "NON_PARTICIPATING",
|
12 |
"104": "NON_PARTICIPATING",
|
13 |
"105": "NON_PARTICIPATING",
|
14 |
+
"106": "NON_PARTICIPATING",
|
15 |
"107": "SUCCESS",
|
16 |
"108": "NON_PARTICIPATING",
|
17 |
"109": "SUCCESS",
|
|
|
20 |
"111": "NON_PARTICIPATING",
|
21 |
"112": "NON_PARTICIPATING",
|
22 |
"113": "NON_PARTICIPATING",
|
23 |
+
"114": "NON_PARTICIPATING",
|
24 |
"115": "NON_PARTICIPATING",
|
25 |
"116": "NON_PARTICIPATING",
|
26 |
+
"117": "SUCCESS",
|
27 |
"118": "NON_PARTICIPATING",
|
28 |
"119": "NON_PARTICIPATING",
|
29 |
"12": "SUCCESS",
|
30 |
"120": "SUCCESS",
|
31 |
"121": "NON_PARTICIPATING",
|
32 |
"122": "NON_PARTICIPATING",
|
33 |
+
"123": "SUCCESS",
|
34 |
"124": "NON_PARTICIPATING",
|
35 |
"125": "NON_PARTICIPATING",
|
36 |
"126": "NON_PARTICIPATING",
|
|
|
51 |
"14": "NON_PARTICIPATING",
|
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": "SUCCESS",
|
60 |
+
"148": "NON_PARTICIPATING",
|
61 |
"149": "SUCCESS",
|
62 |
"15": "SUCCESS",
|
63 |
+
"150": "SUCCESS",
|
64 |
"151": "NON_PARTICIPATING",
|
65 |
"152": "NON_PARTICIPATING",
|
66 |
"153": "NON_PARTICIPATING",
|
|
|
70 |
"157": "NON_PARTICIPATING",
|
71 |
"158": "NON_PARTICIPATING",
|
72 |
"159": "NON_PARTICIPATING",
|
73 |
+
"16": "SUCCESS",
|
74 |
"160": "SUCCESS",
|
75 |
"161": "NON_PARTICIPATING",
|
76 |
"162": "NON_PARTICIPATING",
|
|
|
78 |
"164": "NON_PARTICIPATING",
|
79 |
"165": "NON_PARTICIPATING",
|
80 |
"166": "NON_PARTICIPATING",
|
81 |
+
"167": "SUCCESS",
|
82 |
+
"168": "SUCCESS",
|
83 |
"169": "NON_PARTICIPATING",
|
84 |
"17": "SUCCESS",
|
85 |
"170": "SUCCESS",
|
|
|
99 |
"183": "NON_PARTICIPATING",
|
100 |
"184": "NON_PARTICIPATING",
|
101 |
"185": "NON_PARTICIPATING",
|
102 |
+
"186": "SUCCESS",
|
103 |
"187": "SUCCESS",
|
104 |
"188": "NON_PARTICIPATING",
|
105 |
"189": "NON_PARTICIPATING",
|
106 |
"19": "NON_PARTICIPATING",
|
107 |
"190": "NON_PARTICIPATING",
|
108 |
+
"191": "FAIL",
|
109 |
+
"192": "SUCCESS",
|
110 |
"193": "NON_PARTICIPATING",
|
111 |
"194": "NON_PARTICIPATING",
|
112 |
"195": "SUCCESS",
|
|
|
132 |
"212": "NON_PARTICIPATING",
|
133 |
"213": "NON_PARTICIPATING",
|
134 |
"214": "SUCCESS",
|
135 |
+
"215": "SUCCESS",
|
136 |
"216": "NON_PARTICIPATING",
|
137 |
"217": "NON_PARTICIPATING",
|
138 |
"218": "SUCCESS",
|
139 |
"219": "NON_PARTICIPATING",
|
140 |
+
"22": "SUCCESS",
|
141 |
"220": "NON_PARTICIPATING",
|
142 |
+
"221": "SUCCESS",
|
143 |
"222": "NON_PARTICIPATING",
|
144 |
+
"223": "SUCCESS",
|
145 |
"224": "NON_PARTICIPATING",
|
146 |
+
"225": "SUCCESS",
|
147 |
"226": "SUCCESS",
|
148 |
"227": "NON_PARTICIPATING",
|
149 |
"228": "NON_PARTICIPATING",
|
150 |
"229": "NON_PARTICIPATING",
|
151 |
"23": "SUCCESS",
|
152 |
"230": "NON_PARTICIPATING",
|
153 |
+
"231": "NON_PARTICIPATING",
|
154 |
"232": "NON_PARTICIPATING",
|
155 |
"233": "NON_PARTICIPATING",
|
156 |
"234": "NON_PARTICIPATING",
|
|
|
167 |
"244": "NON_PARTICIPATING",
|
168 |
"245": "NON_PARTICIPATING",
|
169 |
"246": "NON_PARTICIPATING",
|
170 |
+
"247": "SUCCESS",
|
171 |
"248": "NON_PARTICIPATING",
|
172 |
+
"249": "SUCCESS",
|
173 |
"25": "SUCCESS",
|
174 |
"250": "SUCCESS",
|
175 |
"251": "NON_PARTICIPATING",
|
|
|
180 |
"26": "SUCCESS",
|
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": "SUCCESS",
|
188 |
"33": "NON_PARTICIPATING",
|
|
|
193 |
"38": "NON_PARTICIPATING",
|
194 |
"39": "NON_PARTICIPATING",
|
195 |
"4": "NON_PARTICIPATING",
|
196 |
+
"40": "SUCCESS",
|
197 |
+
"41": "SUCCESS",
|
198 |
"42": "NON_PARTICIPATING",
|
199 |
+
"43": "SUCCESS",
|
200 |
+
"44": "SUCCESS",
|
201 |
+
"45": "SUCCESS",
|
202 |
"46": "SUCCESS",
|
203 |
"47": "SUCCESS",
|
204 |
"48": "NON_PARTICIPATING",
|
|
|
210 |
"53": "NON_PARTICIPATING",
|
211 |
"54": "NON_PARTICIPATING",
|
212 |
"55": "SUCCESS",
|
213 |
+
"56": "NON_PARTICIPATING",
|
214 |
+
"57": "NON_PARTICIPATING",
|
215 |
+
"58": "SUCCESS",
|
216 |
"59": "NON_PARTICIPATING",
|
217 |
"6": "SUCCESS",
|
218 |
"60": "NON_PARTICIPATING",
|
219 |
+
"61": "SUCCESS",
|
220 |
+
"62": "SUCCESS",
|
221 |
"63": "SUCCESS",
|
222 |
+
"64": "SUCCESS",
|
223 |
"65": "NON_PARTICIPATING",
|
224 |
"66": "NON_PARTICIPATING",
|
225 |
+
"67": "SUCCESS",
|
226 |
"68": "NON_PARTICIPATING",
|
227 |
"69": "NON_PARTICIPATING",
|
228 |
"7": "NON_PARTICIPATING",
|
|
|
236 |
"77": "NON_PARTICIPATING",
|
237 |
"78": "NON_PARTICIPATING",
|
238 |
"79": "NON_PARTICIPATING",
|
239 |
+
"8": "NON_PARTICIPATING",
|
240 |
"80": "SUCCESS",
|
241 |
"81": "NON_PARTICIPATING",
|
242 |
"82": "SUCCESS",
|
243 |
"83": "NON_PARTICIPATING",
|
244 |
+
"84": "SUCCESS",
|
245 |
"85": "SUCCESS",
|
246 |
"86": "NON_PARTICIPATING",
|
247 |
"87": "SUCCESS",
|
248 |
+
"88": "NON_PARTICIPATING",
|
249 |
+
"89": "SUCCESS",
|
250 |
"9": "NON_PARTICIPATING",
|
251 |
+
"90": "SUCCESS",
|
252 |
"91": "SUCCESS",
|
253 |
"92": "NON_PARTICIPATING",
|
254 |
"93": "NON_PARTICIPATING",
|
255 |
"94": "NON_PARTICIPATING",
|
256 |
+
"95": "SUCCESS",
|
257 |
"96": "SUCCESS",
|
258 |
"97": "NON_PARTICIPATING",
|
259 |
+
"98": "SUCCESS",
|
260 |
"99": "SUCCESS"
|
261 |
},
|
262 |
"architectures": [
|
|
|
275 |
"initializer_range": 0.02,
|
276 |
"inner_step": 0,
|
277 |
"inner_steps": 0,
|
278 |
+
"last_allreduce_block": 5511564,
|
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 sha256:4b275692b8227d489a4e64506a414b1382a2aca5ee7edf452d608399f69faf60
|
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:e7adcd34c6067d981582d836e0856f5c97d0bbc72aadbebeac5595209cabae33
|
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 sha256:60417667cb2008bd26037f55e2a74c817d02448de47c5d5af9cff2d19ec3cb6b
|
3 |
size 4040805354
|