Run 4. Outer Step 50. Inner Step 71.
Browse files- config.json +55 -58
- inner_optimizer.pt +2 -2
- model.safetensors +1 -1
config.json
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "
|
3 |
"activation_function": "gelu_new",
|
4 |
"all_reduce_scores": {
|
5 |
"0": "NON_PARTICIPATING",
|
@@ -11,15 +11,15 @@
|
|
11 |
"103": "NON_PARTICIPATING",
|
12 |
"104": "NON_PARTICIPATING",
|
13 |
"105": "NON_PARTICIPATING",
|
14 |
-
"106": "
|
15 |
-
"107": "
|
16 |
"108": "NON_PARTICIPATING",
|
17 |
"109": "NON_PARTICIPATING",
|
18 |
"11": "SUCCESS",
|
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",
|
@@ -30,15 +30,15 @@
|
|
30 |
"120": "NON_PARTICIPATING",
|
31 |
"121": "NON_PARTICIPATING",
|
32 |
"122": "NON_PARTICIPATING",
|
33 |
-
"123": "
|
34 |
"124": "NON_PARTICIPATING",
|
35 |
"125": "NON_PARTICIPATING",
|
36 |
-
"126": "
|
37 |
"127": "NON_PARTICIPATING",
|
38 |
"128": "NON_PARTICIPATING",
|
39 |
-
"129": "
|
40 |
-
"13": "
|
41 |
-
"130": "
|
42 |
"131": "NON_PARTICIPATING",
|
43 |
"132": "NON_PARTICIPATING",
|
44 |
"133": "NON_PARTICIPATING",
|
@@ -48,26 +48,26 @@
|
|
48 |
"137": "NON_PARTICIPATING",
|
49 |
"138": "NON_PARTICIPATING",
|
50 |
"139": "NON_PARTICIPATING",
|
51 |
-
"14": "
|
52 |
"140": "NON_PARTICIPATING",
|
53 |
"141": "NON_PARTICIPATING",
|
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": "
|
62 |
"15": "SUCCESS",
|
63 |
-
"150": "
|
64 |
"151": "NON_PARTICIPATING",
|
65 |
"152": "NON_PARTICIPATING",
|
66 |
"153": "NON_PARTICIPATING",
|
67 |
-
"154": "
|
68 |
"155": "NON_PARTICIPATING",
|
69 |
-
"156": "
|
70 |
-
"157": "
|
71 |
"158": "NON_PARTICIPATING",
|
72 |
"159": "NON_PARTICIPATING",
|
73 |
"16": "SUCCESS",
|
@@ -76,37 +76,37 @@
|
|
76 |
"162": "NON_PARTICIPATING",
|
77 |
"163": "NON_PARTICIPATING",
|
78 |
"164": "NON_PARTICIPATING",
|
79 |
-
"165": "
|
80 |
"166": "NON_PARTICIPATING",
|
81 |
-
"167": "
|
82 |
"168": "NON_PARTICIPATING",
|
83 |
"169": "NON_PARTICIPATING",
|
84 |
"17": "SUCCESS",
|
85 |
-
"170": "
|
86 |
-
"171": "
|
87 |
"172": "NON_PARTICIPATING",
|
88 |
"173": "NON_PARTICIPATING",
|
89 |
"174": "NON_PARTICIPATING",
|
90 |
-
"175": "
|
91 |
"176": "NON_PARTICIPATING",
|
92 |
"177": "NON_PARTICIPATING",
|
93 |
"178": "NON_PARTICIPATING",
|
94 |
"179": "NON_PARTICIPATING",
|
95 |
"18": "SUCCESS",
|
96 |
-
"180": "
|
97 |
-
"181": "
|
98 |
"182": "NON_PARTICIPATING",
|
99 |
"183": "NON_PARTICIPATING",
|
100 |
-
"184": "
|
101 |
-
"185": "
|
102 |
-
"186": "
|
103 |
"187": "NON_PARTICIPATING",
|
104 |
"188": "NON_PARTICIPATING",
|
105 |
"189": "NON_PARTICIPATING",
|
106 |
"19": "SUCCESS",
|
107 |
"190": "NON_PARTICIPATING",
|
108 |
-
"191": "
|
109 |
-
"192": "
|
110 |
"193": "NON_PARTICIPATING",
|
111 |
"194": "NON_PARTICIPATING",
|
112 |
"195": "NON_PARTICIPATING",
|
@@ -119,7 +119,7 @@
|
|
119 |
"200": "NON_PARTICIPATING",
|
120 |
"201": "NON_PARTICIPATING",
|
121 |
"202": "NON_PARTICIPATING",
|
122 |
-
"203": "
|
123 |
"204": "NON_PARTICIPATING",
|
124 |
"205": "NON_PARTICIPATING",
|
125 |
"206": "NON_PARTICIPATING",
|
@@ -135,11 +135,11 @@
|
|
135 |
"215": "NON_PARTICIPATING",
|
136 |
"216": "NON_PARTICIPATING",
|
137 |
"217": "NON_PARTICIPATING",
|
138 |
-
"218": "
|
139 |
-
"219": "
|
140 |
"22": "SUCCESS",
|
141 |
-
"220": "
|
142 |
-
"221": "
|
143 |
"222": "NON_PARTICIPATING",
|
144 |
"223": "NON_PARTICIPATING",
|
145 |
"224": "NON_PARTICIPATING",
|
@@ -151,39 +151,39 @@
|
|
151 |
"23": "SUCCESS",
|
152 |
"230": "NON_PARTICIPATING",
|
153 |
"231": "NON_PARTICIPATING",
|
154 |
-
"232": "
|
155 |
"233": "NON_PARTICIPATING",
|
156 |
"234": "NON_PARTICIPATING",
|
157 |
"235": "NON_PARTICIPATING",
|
158 |
"236": "SUCCESS",
|
159 |
-
"237": "
|
160 |
"238": "NON_PARTICIPATING",
|
161 |
"239": "NON_PARTICIPATING",
|
162 |
"24": "SUCCESS",
|
163 |
"240": "NON_PARTICIPATING",
|
164 |
-
"241": "
|
165 |
"242": "NON_PARTICIPATING",
|
166 |
"243": "SUCCESS",
|
167 |
"244": "NON_PARTICIPATING",
|
168 |
"245": "NON_PARTICIPATING",
|
169 |
-
"246": "
|
170 |
-
"247": "
|
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": "
|
180 |
"26": "SUCCESS",
|
181 |
"27": "SUCCESS",
|
182 |
"28": "SUCCESS",
|
183 |
"29": "SUCCESS",
|
184 |
-
"3": "
|
185 |
"30": "NON_PARTICIPATING",
|
186 |
-
"31": "
|
187 |
"32": "SUCCESS",
|
188 |
"33": "NON_PARTICIPATING",
|
189 |
"34": "SUCCESS",
|
@@ -195,18 +195,18 @@
|
|
195 |
"4": "SUCCESS",
|
196 |
"40": "SUCCESS",
|
197 |
"41": "SUCCESS",
|
198 |
-
"42": "
|
199 |
"43": "SUCCESS",
|
200 |
"44": "SUCCESS",
|
201 |
"45": "SUCCESS",
|
202 |
"46": "NON_PARTICIPATING",
|
203 |
-
"47": "
|
204 |
"48": "NON_PARTICIPATING",
|
205 |
"49": "SUCCESS",
|
206 |
-
"5": "
|
207 |
"50": "SUCCESS",
|
208 |
"51": "SUCCESS",
|
209 |
-
"52": "
|
210 |
"53": "SUCCESS",
|
211 |
"54": "SUCCESS",
|
212 |
"55": "SUCCESS",
|
@@ -219,7 +219,7 @@
|
|
219 |
"61": "NON_PARTICIPATING",
|
220 |
"62": "SUCCESS",
|
221 |
"63": "SUCCESS",
|
222 |
-
"64": "
|
223 |
"65": "SUCCESS",
|
224 |
"66": "SUCCESS",
|
225 |
"67": "SUCCESS",
|
@@ -232,8 +232,8 @@
|
|
232 |
"73": "NON_PARTICIPATING",
|
233 |
"74": "SUCCESS",
|
234 |
"75": "SUCCESS",
|
235 |
-
"76": "
|
236 |
-
"77": "
|
237 |
"78": "SUCCESS",
|
238 |
"79": "NON_PARTICIPATING",
|
239 |
"8": "SUCCESS",
|
@@ -253,8 +253,8 @@
|
|
253 |
"92": "SUCCESS",
|
254 |
"93": "NON_PARTICIPATING",
|
255 |
"94": "SUCCESS",
|
256 |
-
"95": "
|
257 |
-
"96": "
|
258 |
"97": "SUCCESS",
|
259 |
"98": "SUCCESS",
|
260 |
"99": "SUCCESS"
|
@@ -268,20 +268,17 @@
|
|
268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
269 |
},
|
270 |
"block_list": [
|
271 |
-
|
272 |
-
|
273 |
-
5622049,
|
274 |
-
5622053,
|
275 |
-
5622057
|
276 |
],
|
277 |
"block_size": 1024,
|
278 |
"bos_token_id": 50256,
|
279 |
"embd_pdrop": 0.1,
|
280 |
"eos_token_id": 50256,
|
281 |
"initializer_range": 0.02,
|
282 |
-
"inner_step":
|
283 |
"inner_steps": 0,
|
284 |
-
"last_allreduce_block":
|
285 |
"layer_norm_epsilon": 1e-05,
|
286 |
"model_type": "gpt_optimized",
|
287 |
"n_embd": 1280,
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "philip324mark/mark7",
|
3 |
"activation_function": "gelu_new",
|
4 |
"all_reduce_scores": {
|
5 |
"0": "NON_PARTICIPATING",
|
|
|
11 |
"103": "NON_PARTICIPATING",
|
12 |
"104": "NON_PARTICIPATING",
|
13 |
"105": "NON_PARTICIPATING",
|
14 |
+
"106": "NON_PARTICIPATING",
|
15 |
+
"107": "NON_PARTICIPATING",
|
16 |
"108": "NON_PARTICIPATING",
|
17 |
"109": "NON_PARTICIPATING",
|
18 |
"11": "SUCCESS",
|
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",
|
|
|
30 |
"120": "NON_PARTICIPATING",
|
31 |
"121": "NON_PARTICIPATING",
|
32 |
"122": "NON_PARTICIPATING",
|
33 |
+
"123": "SUCCESS",
|
34 |
"124": "NON_PARTICIPATING",
|
35 |
"125": "NON_PARTICIPATING",
|
36 |
+
"126": "NON_PARTICIPATING",
|
37 |
"127": "NON_PARTICIPATING",
|
38 |
"128": "NON_PARTICIPATING",
|
39 |
+
"129": "SUCCESS",
|
40 |
+
"13": "SUCCESS",
|
41 |
+
"130": "NON_PARTICIPATING",
|
42 |
"131": "NON_PARTICIPATING",
|
43 |
"132": "NON_PARTICIPATING",
|
44 |
"133": "NON_PARTICIPATING",
|
|
|
48 |
"137": "NON_PARTICIPATING",
|
49 |
"138": "NON_PARTICIPATING",
|
50 |
"139": "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": "NON_PARTICIPATING",
|
60 |
"148": "NON_PARTICIPATING",
|
61 |
+
"149": "SUCCESS",
|
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",
|
69 |
+
"156": "SUCCESS",
|
70 |
+
"157": "SUCCESS",
|
71 |
"158": "NON_PARTICIPATING",
|
72 |
"159": "NON_PARTICIPATING",
|
73 |
"16": "SUCCESS",
|
|
|
76 |
"162": "NON_PARTICIPATING",
|
77 |
"163": "NON_PARTICIPATING",
|
78 |
"164": "NON_PARTICIPATING",
|
79 |
+
"165": "NON_PARTICIPATING",
|
80 |
"166": "NON_PARTICIPATING",
|
81 |
+
"167": "SUCCESS",
|
82 |
"168": "NON_PARTICIPATING",
|
83 |
"169": "NON_PARTICIPATING",
|
84 |
"17": "SUCCESS",
|
85 |
+
"170": "SUCCESS",
|
86 |
+
"171": "SUCCESS",
|
87 |
"172": "NON_PARTICIPATING",
|
88 |
"173": "NON_PARTICIPATING",
|
89 |
"174": "NON_PARTICIPATING",
|
90 |
+
"175": "SUCCESS",
|
91 |
"176": "NON_PARTICIPATING",
|
92 |
"177": "NON_PARTICIPATING",
|
93 |
"178": "NON_PARTICIPATING",
|
94 |
"179": "NON_PARTICIPATING",
|
95 |
"18": "SUCCESS",
|
96 |
+
"180": "SUCCESS",
|
97 |
+
"181": "NON_PARTICIPATING",
|
98 |
"182": "NON_PARTICIPATING",
|
99 |
"183": "NON_PARTICIPATING",
|
100 |
+
"184": "SUCCESS",
|
101 |
+
"185": "SUCCESS",
|
102 |
+
"186": "NON_PARTICIPATING",
|
103 |
"187": "NON_PARTICIPATING",
|
104 |
"188": "NON_PARTICIPATING",
|
105 |
"189": "NON_PARTICIPATING",
|
106 |
"19": "SUCCESS",
|
107 |
"190": "NON_PARTICIPATING",
|
108 |
+
"191": "SUCCESS",
|
109 |
+
"192": "NON_PARTICIPATING",
|
110 |
"193": "NON_PARTICIPATING",
|
111 |
"194": "NON_PARTICIPATING",
|
112 |
"195": "NON_PARTICIPATING",
|
|
|
119 |
"200": "NON_PARTICIPATING",
|
120 |
"201": "NON_PARTICIPATING",
|
121 |
"202": "NON_PARTICIPATING",
|
122 |
+
"203": "SUCCESS",
|
123 |
"204": "NON_PARTICIPATING",
|
124 |
"205": "NON_PARTICIPATING",
|
125 |
"206": "NON_PARTICIPATING",
|
|
|
135 |
"215": "NON_PARTICIPATING",
|
136 |
"216": "NON_PARTICIPATING",
|
137 |
"217": "NON_PARTICIPATING",
|
138 |
+
"218": "NON_PARTICIPATING",
|
139 |
+
"219": "NON_PARTICIPATING",
|
140 |
"22": "SUCCESS",
|
141 |
+
"220": "NON_PARTICIPATING",
|
142 |
+
"221": "NON_PARTICIPATING",
|
143 |
"222": "NON_PARTICIPATING",
|
144 |
"223": "NON_PARTICIPATING",
|
145 |
"224": "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",
|
157 |
"235": "NON_PARTICIPATING",
|
158 |
"236": "SUCCESS",
|
159 |
+
"237": "SUCCESS",
|
160 |
"238": "NON_PARTICIPATING",
|
161 |
"239": "NON_PARTICIPATING",
|
162 |
"24": "SUCCESS",
|
163 |
"240": "NON_PARTICIPATING",
|
164 |
+
"241": "NON_PARTICIPATING",
|
165 |
"242": "NON_PARTICIPATING",
|
166 |
"243": "SUCCESS",
|
167 |
"244": "NON_PARTICIPATING",
|
168 |
"245": "NON_PARTICIPATING",
|
169 |
+
"246": "NON_PARTICIPATING",
|
170 |
+
"247": "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": "SUCCESS",
|
180 |
"26": "SUCCESS",
|
181 |
"27": "SUCCESS",
|
182 |
"28": "SUCCESS",
|
183 |
"29": "SUCCESS",
|
184 |
+
"3": "SUCCESS",
|
185 |
"30": "NON_PARTICIPATING",
|
186 |
+
"31": "SUCCESS",
|
187 |
"32": "SUCCESS",
|
188 |
"33": "NON_PARTICIPATING",
|
189 |
"34": "SUCCESS",
|
|
|
195 |
"4": "SUCCESS",
|
196 |
"40": "SUCCESS",
|
197 |
"41": "SUCCESS",
|
198 |
+
"42": "NON_PARTICIPATING",
|
199 |
"43": "SUCCESS",
|
200 |
"44": "SUCCESS",
|
201 |
"45": "SUCCESS",
|
202 |
"46": "NON_PARTICIPATING",
|
203 |
+
"47": "SUCCESS",
|
204 |
"48": "NON_PARTICIPATING",
|
205 |
"49": "SUCCESS",
|
206 |
+
"5": "NON_PARTICIPATING",
|
207 |
"50": "SUCCESS",
|
208 |
"51": "SUCCESS",
|
209 |
+
"52": "SUCCESS",
|
210 |
"53": "SUCCESS",
|
211 |
"54": "SUCCESS",
|
212 |
"55": "SUCCESS",
|
|
|
219 |
"61": "NON_PARTICIPATING",
|
220 |
"62": "SUCCESS",
|
221 |
"63": "SUCCESS",
|
222 |
+
"64": "NON_PARTICIPATING",
|
223 |
"65": "SUCCESS",
|
224 |
"66": "SUCCESS",
|
225 |
"67": "SUCCESS",
|
|
|
232 |
"73": "NON_PARTICIPATING",
|
233 |
"74": "SUCCESS",
|
234 |
"75": "SUCCESS",
|
235 |
+
"76": "SUCCESS",
|
236 |
+
"77": "SUCCESS",
|
237 |
"78": "SUCCESS",
|
238 |
"79": "NON_PARTICIPATING",
|
239 |
"8": "SUCCESS",
|
|
|
253 |
"92": "SUCCESS",
|
254 |
"93": "NON_PARTICIPATING",
|
255 |
"94": "SUCCESS",
|
256 |
+
"95": "SUCCESS",
|
257 |
+
"96": "SUCCESS",
|
258 |
"97": "SUCCESS",
|
259 |
"98": "SUCCESS",
|
260 |
"99": "SUCCESS"
|
|
|
268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
269 |
},
|
270 |
"block_list": [
|
271 |
+
5622191,
|
272 |
+
5622220
|
|
|
|
|
|
|
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": 71,
|
280 |
"inner_steps": 0,
|
281 |
+
"last_allreduce_block": 5568918,
|
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
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:21ed344ae7ddcf3910c4074eff5c2f0524375ea578e852f02be67f74d54de181
|
3 |
+
size 8081782503
|
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 sha256:21d2b8e80f38746134120719cfdff47ddd24036b3b07e520bdf6d7c2be6e4065
|
3 |
size 4040701744
|