Epoch 1. Batch Size 41676. Peers 62.
Browse files- config.json +63 -63
- model-00001-of-00002.safetensors +1 -1
- model-00002-of-00002.safetensors +1 -1
- optimizer.pt +2 -2
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",
|
@@ -7,27 +7,27 @@
|
|
7 |
"10": "NON_PARTICIPATING",
|
8 |
"100": "NON_PARTICIPATING",
|
9 |
"101": "NON_PARTICIPATING",
|
10 |
-
"102": "
|
11 |
-
"103": "
|
12 |
"104": "NON_PARTICIPATING",
|
13 |
-
"105": "
|
14 |
"106": "NON_PARTICIPATING",
|
15 |
-
"107": "
|
16 |
"108": "NON_PARTICIPATING",
|
17 |
"109": "NON_PARTICIPATING",
|
18 |
-
"11": "
|
19 |
"110": "NON_PARTICIPATING",
|
20 |
"111": "NON_PARTICIPATING",
|
21 |
-
"112": "
|
22 |
-
"113": "
|
23 |
-
"114": "
|
24 |
"115": "NON_PARTICIPATING",
|
25 |
-
"116": "
|
26 |
"117": "NON_PARTICIPATING",
|
27 |
"118": "NON_PARTICIPATING",
|
28 |
"119": "NON_PARTICIPATING",
|
29 |
"12": "NON_PARTICIPATING",
|
30 |
-
"120": "
|
31 |
"121": "NON_PARTICIPATING",
|
32 |
"122": "NON_PARTICIPATING",
|
33 |
"123": "NON_PARTICIPATING",
|
@@ -44,7 +44,7 @@
|
|
44 |
"133": "NON_PARTICIPATING",
|
45 |
"134": "NON_PARTICIPATING",
|
46 |
"135": "NON_PARTICIPATING",
|
47 |
-
"136": "
|
48 |
"137": "NON_PARTICIPATING",
|
49 |
"138": "NON_PARTICIPATING",
|
50 |
"139": "NON_PARTICIPATING",
|
@@ -52,71 +52,71 @@
|
|
52 |
"140": "NON_PARTICIPATING",
|
53 |
"141": "NON_PARTICIPATING",
|
54 |
"142": "NON_PARTICIPATING",
|
55 |
-
"143": "
|
56 |
"144": "NON_PARTICIPATING",
|
57 |
"145": "NON_PARTICIPATING",
|
58 |
-
"146": "
|
59 |
-
"147": "
|
60 |
"148": "NON_PARTICIPATING",
|
61 |
"149": "NON_PARTICIPATING",
|
62 |
"15": "NON_PARTICIPATING",
|
63 |
"150": "NON_PARTICIPATING",
|
64 |
"151": "NON_PARTICIPATING",
|
65 |
-
"152": "
|
66 |
"153": "NON_PARTICIPATING",
|
67 |
"154": "NON_PARTICIPATING",
|
68 |
"155": "NON_PARTICIPATING",
|
69 |
"156": "NON_PARTICIPATING",
|
70 |
-
"157": "
|
71 |
"158": "NON_PARTICIPATING",
|
72 |
-
"159": "
|
73 |
"16": "NON_PARTICIPATING",
|
74 |
"160": "NON_PARTICIPATING",
|
75 |
"161": "NON_PARTICIPATING",
|
76 |
"162": "NON_PARTICIPATING",
|
77 |
-
"163": "
|
78 |
"164": "NON_PARTICIPATING",
|
79 |
"165": "NON_PARTICIPATING",
|
80 |
"166": "NON_PARTICIPATING",
|
81 |
"167": "NON_PARTICIPATING",
|
82 |
-
"168": "
|
83 |
"169": "NON_PARTICIPATING",
|
84 |
"17": "NON_PARTICIPATING",
|
85 |
-
"170": "
|
86 |
"171": "NON_PARTICIPATING",
|
87 |
"172": "NON_PARTICIPATING",
|
88 |
"173": "NON_PARTICIPATING",
|
89 |
-
"174": "
|
90 |
"175": "NON_PARTICIPATING",
|
91 |
"176": "NON_PARTICIPATING",
|
92 |
"177": "NON_PARTICIPATING",
|
93 |
"178": "NON_PARTICIPATING",
|
94 |
"179": "NON_PARTICIPATING",
|
95 |
-
"18": "
|
96 |
"180": "NON_PARTICIPATING",
|
97 |
-
"181": "
|
98 |
-
"182": "
|
99 |
"183": "NON_PARTICIPATING",
|
100 |
-
"184": "
|
101 |
"185": "NON_PARTICIPATING",
|
102 |
"186": "NON_PARTICIPATING",
|
103 |
-
"187": "
|
104 |
"188": "NON_PARTICIPATING",
|
105 |
-
"189": "
|
106 |
"19": "NON_PARTICIPATING",
|
107 |
"190": "NON_PARTICIPATING",
|
108 |
"191": "NON_PARTICIPATING",
|
109 |
"192": "NON_PARTICIPATING",
|
110 |
-
"193": "
|
111 |
"194": "NON_PARTICIPATING",
|
112 |
"195": "NON_PARTICIPATING",
|
113 |
"196": "NON_PARTICIPATING",
|
114 |
"197": "NON_PARTICIPATING",
|
115 |
-
"198": "
|
116 |
-
"199": "
|
117 |
"2": "NON_PARTICIPATING",
|
118 |
"20": "NON_PARTICIPATING",
|
119 |
-
"200": "
|
120 |
"201": "NON_PARTICIPATING",
|
121 |
"202": "NON_PARTICIPATING",
|
122 |
"203": "NON_PARTICIPATING",
|
@@ -126,24 +126,24 @@
|
|
126 |
"207": "NON_PARTICIPATING",
|
127 |
"208": "NON_PARTICIPATING",
|
128 |
"209": "NON_PARTICIPATING",
|
129 |
-
"21": "
|
130 |
"210": "NON_PARTICIPATING",
|
131 |
"211": "NON_PARTICIPATING",
|
132 |
"212": "NON_PARTICIPATING",
|
133 |
"213": "NON_PARTICIPATING",
|
134 |
-
"214": "
|
135 |
"215": "NON_PARTICIPATING",
|
136 |
"216": "NON_PARTICIPATING",
|
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": "NON_PARTICIPATING",
|
145 |
"224": "NON_PARTICIPATING",
|
146 |
-
"225": "
|
147 |
"226": "NON_PARTICIPATING",
|
148 |
"227": "NON_PARTICIPATING",
|
149 |
"228": "NON_PARTICIPATING",
|
@@ -151,26 +151,26 @@
|
|
151 |
"23": "NON_PARTICIPATING",
|
152 |
"230": "NON_PARTICIPATING",
|
153 |
"231": "NON_PARTICIPATING",
|
154 |
-
"232": "
|
155 |
"233": "NON_PARTICIPATING",
|
156 |
-
"234": "
|
157 |
"235": "NON_PARTICIPATING",
|
158 |
"236": "NON_PARTICIPATING",
|
159 |
"237": "NON_PARTICIPATING",
|
160 |
"238": "NON_PARTICIPATING",
|
161 |
"239": "NON_PARTICIPATING",
|
162 |
-
"24": "
|
163 |
"240": "NON_PARTICIPATING",
|
164 |
-
"241": "
|
165 |
-
"242": "
|
166 |
"243": "NON_PARTICIPATING",
|
167 |
"244": "NON_PARTICIPATING",
|
168 |
"245": "NON_PARTICIPATING",
|
169 |
-
"246": "
|
170 |
"247": "NON_PARTICIPATING",
|
171 |
"248": "NON_PARTICIPATING",
|
172 |
"249": "NON_PARTICIPATING",
|
173 |
-
"25": "
|
174 |
"250": "NON_PARTICIPATING",
|
175 |
"251": "NON_PARTICIPATING",
|
176 |
"252": "NON_PARTICIPATING",
|
@@ -184,78 +184,78 @@
|
|
184 |
"3": "NON_PARTICIPATING",
|
185 |
"30": "NON_PARTICIPATING",
|
186 |
"31": "NON_PARTICIPATING",
|
187 |
-
"32": "
|
188 |
"33": "NON_PARTICIPATING",
|
189 |
"34": "NON_PARTICIPATING",
|
190 |
"35": "NON_PARTICIPATING",
|
191 |
"36": "NON_PARTICIPATING",
|
192 |
"37": "NON_PARTICIPATING",
|
193 |
"38": "NON_PARTICIPATING",
|
194 |
-
"39": "
|
195 |
"4": "NON_PARTICIPATING",
|
196 |
"40": "NON_PARTICIPATING",
|
197 |
"41": "NON_PARTICIPATING",
|
198 |
-
"42": "
|
199 |
"43": "NON_PARTICIPATING",
|
200 |
"44": "NON_PARTICIPATING",
|
201 |
"45": "NON_PARTICIPATING",
|
202 |
"46": "NON_PARTICIPATING",
|
203 |
"47": "NON_PARTICIPATING",
|
204 |
-
"48": "
|
205 |
"49": "NON_PARTICIPATING",
|
206 |
-
"5": "
|
207 |
"50": "NON_PARTICIPATING",
|
208 |
"51": "NON_PARTICIPATING",
|
209 |
"52": "NON_PARTICIPATING",
|
210 |
"53": "NON_PARTICIPATING",
|
211 |
"54": "NON_PARTICIPATING",
|
212 |
"55": "NON_PARTICIPATING",
|
213 |
-
"56": "
|
214 |
"57": "NON_PARTICIPATING",
|
215 |
"58": "NON_PARTICIPATING",
|
216 |
"59": "NON_PARTICIPATING",
|
217 |
"6": "NON_PARTICIPATING",
|
218 |
"60": "NON_PARTICIPATING",
|
219 |
-
"61": "
|
220 |
"62": "NON_PARTICIPATING",
|
221 |
"63": "NON_PARTICIPATING",
|
222 |
"64": "NON_PARTICIPATING",
|
223 |
"65": "NON_PARTICIPATING",
|
224 |
-
"66": "
|
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": "
|
232 |
"73": "NON_PARTICIPATING",
|
233 |
"74": "NON_PARTICIPATING",
|
234 |
"75": "NON_PARTICIPATING",
|
235 |
"76": "NON_PARTICIPATING",
|
236 |
"77": "NON_PARTICIPATING",
|
237 |
"78": "NON_PARTICIPATING",
|
238 |
-
"79": "
|
239 |
"8": "NON_PARTICIPATING",
|
240 |
"80": "NON_PARTICIPATING",
|
241 |
-
"81": "
|
242 |
-
"82": "
|
243 |
-
"83": "
|
244 |
"84": "NON_PARTICIPATING",
|
245 |
"85": "NON_PARTICIPATING",
|
246 |
"86": "NON_PARTICIPATING",
|
247 |
"87": "NON_PARTICIPATING",
|
248 |
-
"88": "
|
249 |
"89": "NON_PARTICIPATING",
|
250 |
"9": "NON_PARTICIPATING",
|
251 |
"90": "NON_PARTICIPATING",
|
252 |
-
"91": "
|
253 |
"92": "NON_PARTICIPATING",
|
254 |
"93": "NON_PARTICIPATING",
|
255 |
-
"94": "
|
256 |
"95": "NON_PARTICIPATING",
|
257 |
"96": "NON_PARTICIPATING",
|
258 |
-
"97": "
|
259 |
"98": "NON_PARTICIPATING",
|
260 |
"99": "NON_PARTICIPATING"
|
261 |
},
|
@@ -264,8 +264,8 @@
|
|
264 |
],
|
265 |
"attn_pdrop": 0.1,
|
266 |
"auto_map": {
|
267 |
-
"AutoConfig": "configuration_gpt_optimized.GPTOptimConfig",
|
268 |
-
"AutoModelForCausalLM": "modeling_gpt_optimized.GPTOptim"
|
269 |
},
|
270 |
"block_size": 1024,
|
271 |
"bos_token_id": 50256,
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "distributed/optimized-gpt2-2b",
|
3 |
"activation_function": "gelu_new",
|
4 |
"all_reduce_scores": {
|
5 |
"0": "NON_PARTICIPATING",
|
|
|
7 |
"10": "NON_PARTICIPATING",
|
8 |
"100": "NON_PARTICIPATING",
|
9 |
"101": "NON_PARTICIPATING",
|
10 |
+
"102": "SUCCESS",
|
11 |
+
"103": "SUCCESS",
|
12 |
"104": "NON_PARTICIPATING",
|
13 |
+
"105": "SUCCESS",
|
14 |
"106": "NON_PARTICIPATING",
|
15 |
+
"107": "SUCCESS",
|
16 |
"108": "NON_PARTICIPATING",
|
17 |
"109": "NON_PARTICIPATING",
|
18 |
+
"11": "SUCCESS",
|
19 |
"110": "NON_PARTICIPATING",
|
20 |
"111": "NON_PARTICIPATING",
|
21 |
+
"112": "SUCCESS",
|
22 |
+
"113": "SUCCESS",
|
23 |
+
"114": "SUCCESS",
|
24 |
"115": "NON_PARTICIPATING",
|
25 |
+
"116": "SUCCESS",
|
26 |
"117": "NON_PARTICIPATING",
|
27 |
"118": "NON_PARTICIPATING",
|
28 |
"119": "NON_PARTICIPATING",
|
29 |
"12": "NON_PARTICIPATING",
|
30 |
+
"120": "SUCCESS",
|
31 |
"121": "NON_PARTICIPATING",
|
32 |
"122": "NON_PARTICIPATING",
|
33 |
"123": "NON_PARTICIPATING",
|
|
|
44 |
"133": "NON_PARTICIPATING",
|
45 |
"134": "NON_PARTICIPATING",
|
46 |
"135": "NON_PARTICIPATING",
|
47 |
+
"136": "SUCCESS",
|
48 |
"137": "NON_PARTICIPATING",
|
49 |
"138": "NON_PARTICIPATING",
|
50 |
"139": "NON_PARTICIPATING",
|
|
|
52 |
"140": "NON_PARTICIPATING",
|
53 |
"141": "NON_PARTICIPATING",
|
54 |
"142": "NON_PARTICIPATING",
|
55 |
+
"143": "SUCCESS",
|
56 |
"144": "NON_PARTICIPATING",
|
57 |
"145": "NON_PARTICIPATING",
|
58 |
+
"146": "SUCCESS",
|
59 |
+
"147": "SUCCESS",
|
60 |
"148": "NON_PARTICIPATING",
|
61 |
"149": "NON_PARTICIPATING",
|
62 |
"15": "NON_PARTICIPATING",
|
63 |
"150": "NON_PARTICIPATING",
|
64 |
"151": "NON_PARTICIPATING",
|
65 |
+
"152": "SUCCESS",
|
66 |
"153": "NON_PARTICIPATING",
|
67 |
"154": "NON_PARTICIPATING",
|
68 |
"155": "NON_PARTICIPATING",
|
69 |
"156": "NON_PARTICIPATING",
|
70 |
+
"157": "SUCCESS",
|
71 |
"158": "NON_PARTICIPATING",
|
72 |
+
"159": "SUCCESS",
|
73 |
"16": "NON_PARTICIPATING",
|
74 |
"160": "NON_PARTICIPATING",
|
75 |
"161": "NON_PARTICIPATING",
|
76 |
"162": "NON_PARTICIPATING",
|
77 |
+
"163": "SUCCESS",
|
78 |
"164": "NON_PARTICIPATING",
|
79 |
"165": "NON_PARTICIPATING",
|
80 |
"166": "NON_PARTICIPATING",
|
81 |
"167": "NON_PARTICIPATING",
|
82 |
+
"168": "SUCCESS",
|
83 |
"169": "NON_PARTICIPATING",
|
84 |
"17": "NON_PARTICIPATING",
|
85 |
+
"170": "SUCCESS",
|
86 |
"171": "NON_PARTICIPATING",
|
87 |
"172": "NON_PARTICIPATING",
|
88 |
"173": "NON_PARTICIPATING",
|
89 |
+
"174": "SUCCESS",
|
90 |
"175": "NON_PARTICIPATING",
|
91 |
"176": "NON_PARTICIPATING",
|
92 |
"177": "NON_PARTICIPATING",
|
93 |
"178": "NON_PARTICIPATING",
|
94 |
"179": "NON_PARTICIPATING",
|
95 |
+
"18": "SUCCESS",
|
96 |
"180": "NON_PARTICIPATING",
|
97 |
+
"181": "SUCCESS",
|
98 |
+
"182": "SUCCESS",
|
99 |
"183": "NON_PARTICIPATING",
|
100 |
+
"184": "SUCCESS",
|
101 |
"185": "NON_PARTICIPATING",
|
102 |
"186": "NON_PARTICIPATING",
|
103 |
+
"187": "SUCCESS",
|
104 |
"188": "NON_PARTICIPATING",
|
105 |
+
"189": "SUCCESS",
|
106 |
"19": "NON_PARTICIPATING",
|
107 |
"190": "NON_PARTICIPATING",
|
108 |
"191": "NON_PARTICIPATING",
|
109 |
"192": "NON_PARTICIPATING",
|
110 |
+
"193": "SUCCESS",
|
111 |
"194": "NON_PARTICIPATING",
|
112 |
"195": "NON_PARTICIPATING",
|
113 |
"196": "NON_PARTICIPATING",
|
114 |
"197": "NON_PARTICIPATING",
|
115 |
+
"198": "SUCCESS",
|
116 |
+
"199": "SUCCESS",
|
117 |
"2": "NON_PARTICIPATING",
|
118 |
"20": "NON_PARTICIPATING",
|
119 |
+
"200": "SUCCESS",
|
120 |
"201": "NON_PARTICIPATING",
|
121 |
"202": "NON_PARTICIPATING",
|
122 |
"203": "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": "NON_PARTICIPATING",
|
133 |
"213": "NON_PARTICIPATING",
|
134 |
+
"214": "SUCCESS",
|
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": "SUCCESS",
|
144 |
"223": "NON_PARTICIPATING",
|
145 |
"224": "NON_PARTICIPATING",
|
146 |
+
"225": "SUCCESS",
|
147 |
"226": "NON_PARTICIPATING",
|
148 |
"227": "NON_PARTICIPATING",
|
149 |
"228": "NON_PARTICIPATING",
|
|
|
151 |
"23": "NON_PARTICIPATING",
|
152 |
"230": "NON_PARTICIPATING",
|
153 |
"231": "NON_PARTICIPATING",
|
154 |
+
"232": "SUCCESS",
|
155 |
"233": "NON_PARTICIPATING",
|
156 |
+
"234": "SUCCESS",
|
157 |
"235": "NON_PARTICIPATING",
|
158 |
"236": "NON_PARTICIPATING",
|
159 |
"237": "NON_PARTICIPATING",
|
160 |
"238": "NON_PARTICIPATING",
|
161 |
"239": "NON_PARTICIPATING",
|
162 |
+
"24": "SUCCESS",
|
163 |
"240": "NON_PARTICIPATING",
|
164 |
+
"241": "SUCCESS",
|
165 |
+
"242": "SUCCESS",
|
166 |
"243": "NON_PARTICIPATING",
|
167 |
"244": "NON_PARTICIPATING",
|
168 |
"245": "NON_PARTICIPATING",
|
169 |
+
"246": "SUCCESS",
|
170 |
"247": "NON_PARTICIPATING",
|
171 |
"248": "NON_PARTICIPATING",
|
172 |
"249": "NON_PARTICIPATING",
|
173 |
+
"25": "SUCCESS",
|
174 |
"250": "NON_PARTICIPATING",
|
175 |
"251": "NON_PARTICIPATING",
|
176 |
"252": "NON_PARTICIPATING",
|
|
|
184 |
"3": "NON_PARTICIPATING",
|
185 |
"30": "NON_PARTICIPATING",
|
186 |
"31": "NON_PARTICIPATING",
|
187 |
+
"32": "SUCCESS",
|
188 |
"33": "NON_PARTICIPATING",
|
189 |
"34": "NON_PARTICIPATING",
|
190 |
"35": "NON_PARTICIPATING",
|
191 |
"36": "NON_PARTICIPATING",
|
192 |
"37": "NON_PARTICIPATING",
|
193 |
"38": "NON_PARTICIPATING",
|
194 |
+
"39": "SUCCESS",
|
195 |
"4": "NON_PARTICIPATING",
|
196 |
"40": "NON_PARTICIPATING",
|
197 |
"41": "NON_PARTICIPATING",
|
198 |
+
"42": "SUCCESS",
|
199 |
"43": "NON_PARTICIPATING",
|
200 |
"44": "NON_PARTICIPATING",
|
201 |
"45": "NON_PARTICIPATING",
|
202 |
"46": "NON_PARTICIPATING",
|
203 |
"47": "NON_PARTICIPATING",
|
204 |
+
"48": "SUCCESS",
|
205 |
"49": "NON_PARTICIPATING",
|
206 |
+
"5": "SUCCESS",
|
207 |
"50": "NON_PARTICIPATING",
|
208 |
"51": "NON_PARTICIPATING",
|
209 |
"52": "NON_PARTICIPATING",
|
210 |
"53": "NON_PARTICIPATING",
|
211 |
"54": "NON_PARTICIPATING",
|
212 |
"55": "NON_PARTICIPATING",
|
213 |
+
"56": "SUCCESS",
|
214 |
"57": "NON_PARTICIPATING",
|
215 |
"58": "NON_PARTICIPATING",
|
216 |
"59": "NON_PARTICIPATING",
|
217 |
"6": "NON_PARTICIPATING",
|
218 |
"60": "NON_PARTICIPATING",
|
219 |
+
"61": "SUCCESS",
|
220 |
"62": "NON_PARTICIPATING",
|
221 |
"63": "NON_PARTICIPATING",
|
222 |
"64": "NON_PARTICIPATING",
|
223 |
"65": "NON_PARTICIPATING",
|
224 |
+
"66": "SUCCESS",
|
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": "SUCCESS",
|
232 |
"73": "NON_PARTICIPATING",
|
233 |
"74": "NON_PARTICIPATING",
|
234 |
"75": "NON_PARTICIPATING",
|
235 |
"76": "NON_PARTICIPATING",
|
236 |
"77": "NON_PARTICIPATING",
|
237 |
"78": "NON_PARTICIPATING",
|
238 |
+
"79": "SUCCESS",
|
239 |
"8": "NON_PARTICIPATING",
|
240 |
"80": "NON_PARTICIPATING",
|
241 |
+
"81": "SUCCESS",
|
242 |
+
"82": "SUCCESS",
|
243 |
+
"83": "SUCCESS",
|
244 |
"84": "NON_PARTICIPATING",
|
245 |
"85": "NON_PARTICIPATING",
|
246 |
"86": "NON_PARTICIPATING",
|
247 |
"87": "NON_PARTICIPATING",
|
248 |
+
"88": "SUCCESS",
|
249 |
"89": "NON_PARTICIPATING",
|
250 |
"9": "NON_PARTICIPATING",
|
251 |
"90": "NON_PARTICIPATING",
|
252 |
+
"91": "SUCCESS",
|
253 |
"92": "NON_PARTICIPATING",
|
254 |
"93": "NON_PARTICIPATING",
|
255 |
+
"94": "SUCCESS",
|
256 |
"95": "NON_PARTICIPATING",
|
257 |
"96": "NON_PARTICIPATING",
|
258 |
+
"97": "SUCCESS",
|
259 |
"98": "NON_PARTICIPATING",
|
260 |
"99": "NON_PARTICIPATING"
|
261 |
},
|
|
|
264 |
],
|
265 |
"attn_pdrop": 0.1,
|
266 |
"auto_map": {
|
267 |
+
"AutoConfig": "distributed/optimized-gpt2-2b--configuration_gpt_optimized.GPTOptimConfig",
|
268 |
+
"AutoModelForCausalLM": "distributed/optimized-gpt2-2b--modeling_gpt_optimized.GPTOptim"
|
269 |
},
|
270 |
"block_size": 1024,
|
271 |
"bos_token_id": 50256,
|
model-00001-of-00002.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 4975530824
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3f251f64644ff47de7681be609433a70744a826cefec1c76dabf46bd6c0b432f
|
3 |
size 4975530824
|
model-00002-of-00002.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 3075942944
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2d12f4f51a917bc253a75c94912aab5e8e27c16acae2a4b6111f1f25ede58780
|
3 |
size 3075942944
|
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:e8310153bcb53c95027e5f6745f92ba6c6a0ac91a66783788363fa4ee79d597a
|
3 |
+
size 4675070692
|