Outer Step 1. Inner Step 0. Batch Size 0
Browse files- config.json +182 -188
- model.safetensors +1 -1
config.json
CHANGED
@@ -2,244 +2,244 @@
|
|
2 |
"_name_or_path": "distributed/optimized-gpt2-1b",
|
3 |
"activation_function": "gelu_new",
|
4 |
"all_reduce_scores": {
|
5 |
-
"0": "
|
6 |
"1": "SUCCESS",
|
7 |
"10": "SUCCESS",
|
8 |
-
"100": "
|
9 |
-
"101": "
|
10 |
"102": "SUCCESS",
|
11 |
-
"103": "
|
12 |
-
"104": "
|
13 |
"105": "SUCCESS",
|
14 |
"106": "SUCCESS",
|
15 |
-
"107": "
|
16 |
-
"108": "
|
17 |
"109": "SUCCESS",
|
18 |
-
"11": "
|
19 |
"110": "SUCCESS",
|
20 |
-
"111": "
|
21 |
-
"112": "
|
22 |
-
"113": "
|
23 |
-
"114": "
|
24 |
"115": "SUCCESS",
|
25 |
-
"116": "
|
26 |
-
"117": "
|
27 |
"118": "SUCCESS",
|
28 |
-
"119": "
|
29 |
-
"12": "
|
30 |
-
"120": "
|
31 |
"121": "SUCCESS",
|
32 |
"122": "SUCCESS",
|
33 |
-
"123": "
|
34 |
-
"124": "
|
35 |
-
"125": "
|
36 |
-
"126": "
|
37 |
-
"127": "
|
38 |
-
"128": "
|
39 |
-
"129": "
|
40 |
-
"13": "
|
41 |
-
"130": "
|
42 |
-
"131": "
|
43 |
-
"132": "
|
44 |
-
"133": "
|
45 |
"134": "SUCCESS",
|
46 |
-
"135": "
|
47 |
-
"136": "
|
48 |
-
"137": "
|
49 |
-
"138": "
|
50 |
-
"139": "
|
51 |
-
"14": "
|
52 |
-
"140": "
|
53 |
-
"141": "
|
54 |
-
"142": "
|
55 |
-
"143": "
|
56 |
"144": "SUCCESS",
|
57 |
"145": "SUCCESS",
|
58 |
"146": "SUCCESS",
|
59 |
-
"147": "
|
60 |
-
"148": "
|
61 |
-
"149": "
|
62 |
-
"15": "
|
63 |
-
"150": "
|
64 |
-
"151": "
|
65 |
-
"152": "
|
66 |
-
"153": "
|
67 |
-
"154": "
|
68 |
-
"155": "
|
69 |
"156": "SUCCESS",
|
70 |
"157": "SUCCESS",
|
71 |
-
"158": "
|
72 |
"159": "SUCCESS",
|
73 |
-
"16": "
|
74 |
-
"160": "
|
75 |
-
"161": "
|
76 |
-
"162": "
|
77 |
-
"163": "
|
78 |
-
"164": "
|
79 |
-
"165": "
|
80 |
-
"166": "
|
81 |
-
"167": "
|
82 |
-
"168": "
|
83 |
-
"169": "
|
84 |
-
"17": "
|
85 |
-
"170": "
|
86 |
-
"171": "
|
87 |
-
"172": "
|
88 |
"173": "SUCCESS",
|
89 |
-
"174": "
|
90 |
-
"175": "
|
91 |
-
"176": "
|
92 |
-
"177": "
|
93 |
-
"178": "
|
94 |
-
"179": "
|
95 |
-
"18": "
|
96 |
"180": "SUCCESS",
|
97 |
"181": "SUCCESS",
|
98 |
-
"182": "
|
99 |
-
"183": "
|
100 |
-
"184": "
|
101 |
-
"185": "
|
102 |
"186": "SUCCESS",
|
103 |
-
"187": "
|
104 |
"188": "SUCCESS",
|
105 |
-
"189": "
|
106 |
-
"19": "
|
107 |
-
"190": "
|
108 |
-
"191": "
|
109 |
-
"192": "
|
110 |
"193": "SUCCESS",
|
111 |
-
"194": "
|
112 |
"195": "SUCCESS",
|
113 |
-
"196": "
|
114 |
-
"197": "
|
115 |
"198": "SUCCESS",
|
116 |
-
"199": "
|
117 |
-
"2": "
|
118 |
-
"20": "
|
119 |
-
"200": "
|
120 |
-
"201": "
|
121 |
-
"202": "
|
122 |
"203": "SUCCESS",
|
123 |
-
"204": "
|
124 |
-
"205": "
|
125 |
-
"206": "
|
126 |
-
"207": "
|
127 |
-
"208": "
|
128 |
-
"209": "
|
129 |
"21": "SUCCESS",
|
130 |
-
"210": "
|
131 |
-
"211": "
|
132 |
-
"212": "
|
133 |
-
"213": "
|
134 |
-
"214": "
|
135 |
-
"215": "
|
136 |
-
"216": "
|
137 |
-
"217": "
|
138 |
-
"218": "
|
139 |
"219": "SUCCESS",
|
140 |
-
"22": "
|
141 |
-
"220": "
|
142 |
"221": "SUCCESS",
|
143 |
-
"222": "
|
144 |
-
"223": "
|
145 |
-
"224": "
|
146 |
-
"225": "
|
147 |
-
"226": "
|
148 |
-
"227": "
|
149 |
-
"228": "
|
150 |
-
"229": "
|
151 |
-
"23": "
|
152 |
"230": "SUCCESS",
|
153 |
-
"231": "
|
154 |
-
"232": "
|
155 |
-
"233": "
|
156 |
"234": "SUCCESS",
|
157 |
"235": "SUCCESS",
|
158 |
"236": "SUCCESS",
|
159 |
"237": "SUCCESS",
|
160 |
-
"238": "
|
161 |
-
"239": "
|
162 |
-
"24": "
|
163 |
-
"240": "
|
164 |
-
"241": "
|
165 |
"242": "SUCCESS",
|
166 |
-
"243": "
|
167 |
-
"244": "
|
168 |
-
"245": "
|
169 |
"246": "SUCCESS",
|
170 |
-
"247": "
|
171 |
-
"248": "
|
172 |
-
"249": "
|
173 |
-
"25": "
|
174 |
-
"250": "
|
175 |
"251": "SUCCESS",
|
176 |
"252": "SUCCESS",
|
177 |
-
"253": "
|
178 |
-
"254": "
|
179 |
-
"255": "
|
180 |
-
"26": "
|
181 |
-
"27": "
|
182 |
-
"28": "
|
183 |
-
"29": "
|
184 |
-
"3": "
|
185 |
"30": "SUCCESS",
|
186 |
"31": "SUCCESS",
|
187 |
-
"32": "
|
188 |
-
"33": "
|
189 |
-
"34": "
|
190 |
-
"35": "
|
191 |
-
"36": "
|
192 |
-
"37": "
|
193 |
-
"38": "
|
194 |
-
"39": "
|
195 |
"4": "SUCCESS",
|
196 |
"40": "SUCCESS",
|
197 |
-
"41": "
|
198 |
-
"42": "
|
199 |
-
"43": "
|
200 |
"44": "SUCCESS",
|
201 |
-
"45": "
|
202 |
-
"46": "
|
203 |
"47": "SUCCESS",
|
204 |
-
"48": "
|
205 |
-
"49": "
|
206 |
-
"5": "
|
207 |
"50": "SUCCESS",
|
208 |
"51": "SUCCESS",
|
209 |
-
"52": "
|
210 |
-
"53": "
|
211 |
-
"54": "
|
212 |
-
"55": "
|
213 |
"56": "SUCCESS",
|
214 |
-
"57": "
|
215 |
"58": "SUCCESS",
|
216 |
-
"59": "
|
217 |
-
"6": "
|
218 |
-
"60": "
|
219 |
-
"61": "
|
220 |
"62": "SUCCESS",
|
221 |
"63": "SUCCESS",
|
222 |
"64": "SUCCESS",
|
223 |
"65": "SUCCESS",
|
224 |
-
"66": "
|
225 |
"67": "SUCCESS",
|
226 |
"68": "SUCCESS",
|
227 |
"69": "SUCCESS",
|
228 |
-
"7": "
|
229 |
-
"70": "
|
230 |
-
"71": "
|
231 |
"72": "SUCCESS",
|
232 |
-
"73": "
|
233 |
"74": "SUCCESS",
|
234 |
-
"75": "
|
235 |
"76": "SUCCESS",
|
236 |
-
"77": "
|
237 |
"78": "SUCCESS",
|
238 |
"79": "SUCCESS",
|
239 |
-
"8": "
|
240 |
"80": "SUCCESS",
|
241 |
"81": "SUCCESS",
|
242 |
-
"82": "
|
243 |
"83": "SUCCESS",
|
244 |
"84": "SUCCESS",
|
245 |
"85": "SUCCESS",
|
@@ -247,16 +247,16 @@
|
|
247 |
"87": "SUCCESS",
|
248 |
"88": "SUCCESS",
|
249 |
"89": "SUCCESS",
|
250 |
-
"9": "
|
251 |
"90": "SUCCESS",
|
252 |
-
"91": "
|
253 |
-
"92": "
|
254 |
"93": "SUCCESS",
|
255 |
"94": "SUCCESS",
|
256 |
"95": "SUCCESS",
|
257 |
"96": "SUCCESS",
|
258 |
"97": "SUCCESS",
|
259 |
-
"98": "
|
260 |
"99": "SUCCESS"
|
261 |
},
|
262 |
"architectures": [
|
@@ -267,21 +267,15 @@
|
|
267 |
"AutoConfig": "distributed/optimized-gpt2-500m--configuration_gpt_optimized.GPTOptimConfig",
|
268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
269 |
},
|
270 |
-
"block_list": [
|
271 |
-
5063044,
|
272 |
-
5063048,
|
273 |
-
5063050,
|
274 |
-
5063055,
|
275 |
-
5063059
|
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,
|
|
|
2 |
"_name_or_path": "distributed/optimized-gpt2-1b",
|
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": "NON_PARTICIPATING",
|
10 |
"102": "SUCCESS",
|
11 |
+
"103": "NON_PARTICIPATING",
|
12 |
+
"104": "NON_PARTICIPATING",
|
13 |
"105": "SUCCESS",
|
14 |
"106": "SUCCESS",
|
15 |
+
"107": "NON_PARTICIPATING",
|
16 |
+
"108": "NON_PARTICIPATING",
|
17 |
"109": "SUCCESS",
|
18 |
+
"11": "NON_PARTICIPATING",
|
19 |
"110": "SUCCESS",
|
20 |
+
"111": "NON_PARTICIPATING",
|
21 |
+
"112": "NON_PARTICIPATING",
|
22 |
+
"113": "NON_PARTICIPATING",
|
23 |
+
"114": "NON_PARTICIPATING",
|
24 |
"115": "SUCCESS",
|
25 |
+
"116": "NON_PARTICIPATING",
|
26 |
+
"117": "NON_PARTICIPATING",
|
27 |
"118": "SUCCESS",
|
28 |
+
"119": "NON_PARTICIPATING",
|
29 |
+
"12": "NON_PARTICIPATING",
|
30 |
+
"120": "NON_PARTICIPATING",
|
31 |
"121": "SUCCESS",
|
32 |
"122": "SUCCESS",
|
33 |
+
"123": "NON_PARTICIPATING",
|
34 |
+
"124": "NON_PARTICIPATING",
|
35 |
+
"125": "NON_PARTICIPATING",
|
36 |
+
"126": "NON_PARTICIPATING",
|
37 |
+
"127": "NON_PARTICIPATING",
|
38 |
+
"128": "NON_PARTICIPATING",
|
39 |
+
"129": "NON_PARTICIPATING",
|
40 |
+
"13": "NON_PARTICIPATING",
|
41 |
+
"130": "NON_PARTICIPATING",
|
42 |
+
"131": "NON_PARTICIPATING",
|
43 |
+
"132": "NON_PARTICIPATING",
|
44 |
+
"133": "NON_PARTICIPATING",
|
45 |
"134": "SUCCESS",
|
46 |
+
"135": "NON_PARTICIPATING",
|
47 |
+
"136": "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": "SUCCESS",
|
57 |
"145": "SUCCESS",
|
58 |
"146": "SUCCESS",
|
59 |
+
"147": "NON_PARTICIPATING",
|
60 |
+
"148": "NON_PARTICIPATING",
|
61 |
+
"149": "NON_PARTICIPATING",
|
62 |
+
"15": "NON_PARTICIPATING",
|
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": "SUCCESS",
|
73 |
+
"16": "NON_PARTICIPATING",
|
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",
|
82 |
+
"168": "NON_PARTICIPATING",
|
83 |
+
"169": "NON_PARTICIPATING",
|
84 |
+
"17": "NON_PARTICIPATING",
|
85 |
+
"170": "NON_PARTICIPATING",
|
86 |
+
"171": "NON_PARTICIPATING",
|
87 |
+
"172": "NON_PARTICIPATING",
|
88 |
"173": "SUCCESS",
|
89 |
+
"174": "FAIL",
|
90 |
+
"175": "NON_PARTICIPATING",
|
91 |
+
"176": "NON_PARTICIPATING",
|
92 |
+
"177": "NON_PARTICIPATING",
|
93 |
+
"178": "NON_PARTICIPATING",
|
94 |
+
"179": "NON_PARTICIPATING",
|
95 |
+
"18": "NON_PARTICIPATING",
|
96 |
"180": "SUCCESS",
|
97 |
"181": "SUCCESS",
|
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": "NON_PARTICIPATING",
|
107 |
+
"190": "NON_PARTICIPATING",
|
108 |
+
"191": "NON_PARTICIPATING",
|
109 |
+
"192": "NON_PARTICIPATING",
|
110 |
"193": "SUCCESS",
|
111 |
+
"194": "NON_PARTICIPATING",
|
112 |
"195": "SUCCESS",
|
113 |
+
"196": "NON_PARTICIPATING",
|
114 |
+
"197": "NON_PARTICIPATING",
|
115 |
"198": "SUCCESS",
|
116 |
+
"199": "NON_PARTICIPATING",
|
117 |
+
"2": "NON_PARTICIPATING",
|
118 |
+
"20": "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",
|
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": "NON_PARTICIPATING",
|
135 |
+
"215": "NON_PARTICIPATING",
|
136 |
+
"216": "NON_PARTICIPATING",
|
137 |
+
"217": "NON_PARTICIPATING",
|
138 |
+
"218": "NON_PARTICIPATING",
|
139 |
"219": "SUCCESS",
|
140 |
+
"22": "NON_PARTICIPATING",
|
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": "SUCCESS",
|
157 |
"235": "SUCCESS",
|
158 |
"236": "SUCCESS",
|
159 |
"237": "SUCCESS",
|
160 |
+
"238": "NON_PARTICIPATING",
|
161 |
+
"239": "NON_PARTICIPATING",
|
162 |
+
"24": "NON_PARTICIPATING",
|
163 |
+
"240": "NON_PARTICIPATING",
|
164 |
+
"241": "NON_PARTICIPATING",
|
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": "NON_PARTICIPATING",
|
174 |
+
"250": "NON_PARTICIPATING",
|
175 |
"251": "SUCCESS",
|
176 |
"252": "SUCCESS",
|
177 |
+
"253": "NON_PARTICIPATING",
|
178 |
+
"254": "NON_PARTICIPATING",
|
179 |
+
"255": "NON_PARTICIPATING",
|
180 |
+
"26": "NON_PARTICIPATING",
|
181 |
+
"27": "NON_PARTICIPATING",
|
182 |
+
"28": "NON_PARTICIPATING",
|
183 |
+
"29": "NON_PARTICIPATING",
|
184 |
+
"3": "NON_PARTICIPATING",
|
185 |
"30": "SUCCESS",
|
186 |
"31": "SUCCESS",
|
187 |
+
"32": "NON_PARTICIPATING",
|
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": "NON_PARTICIPATING",
|
195 |
"4": "SUCCESS",
|
196 |
"40": "SUCCESS",
|
197 |
+
"41": "NON_PARTICIPATING",
|
198 |
+
"42": "NON_PARTICIPATING",
|
199 |
+
"43": "NON_PARTICIPATING",
|
200 |
"44": "SUCCESS",
|
201 |
+
"45": "NON_PARTICIPATING",
|
202 |
+
"46": "NON_PARTICIPATING",
|
203 |
"47": "SUCCESS",
|
204 |
+
"48": "NON_PARTICIPATING",
|
205 |
+
"49": "NON_PARTICIPATING",
|
206 |
+
"5": "NON_PARTICIPATING",
|
207 |
"50": "SUCCESS",
|
208 |
"51": "SUCCESS",
|
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": "SUCCESS",
|
216 |
+
"59": "NON_PARTICIPATING",
|
217 |
+
"6": "NON_PARTICIPATING",
|
218 |
+
"60": "NON_PARTICIPATING",
|
219 |
+
"61": "NON_PARTICIPATING",
|
220 |
"62": "SUCCESS",
|
221 |
"63": "SUCCESS",
|
222 |
"64": "SUCCESS",
|
223 |
"65": "SUCCESS",
|
224 |
+
"66": "NON_PARTICIPATING",
|
225 |
"67": "SUCCESS",
|
226 |
"68": "SUCCESS",
|
227 |
"69": "SUCCESS",
|
228 |
+
"7": "NON_PARTICIPATING",
|
229 |
+
"70": "NON_PARTICIPATING",
|
230 |
+
"71": "NON_PARTICIPATING",
|
231 |
"72": "SUCCESS",
|
232 |
+
"73": "NON_PARTICIPATING",
|
233 |
"74": "SUCCESS",
|
234 |
+
"75": "NON_PARTICIPATING",
|
235 |
"76": "SUCCESS",
|
236 |
+
"77": "NON_PARTICIPATING",
|
237 |
"78": "SUCCESS",
|
238 |
"79": "SUCCESS",
|
239 |
+
"8": "NON_PARTICIPATING",
|
240 |
"80": "SUCCESS",
|
241 |
"81": "SUCCESS",
|
242 |
+
"82": "NON_PARTICIPATING",
|
243 |
"83": "SUCCESS",
|
244 |
"84": "SUCCESS",
|
245 |
"85": "SUCCESS",
|
|
|
247 |
"87": "SUCCESS",
|
248 |
"88": "SUCCESS",
|
249 |
"89": "SUCCESS",
|
250 |
+
"9": "NON_PARTICIPATING",
|
251 |
"90": "SUCCESS",
|
252 |
+
"91": "NON_PARTICIPATING",
|
253 |
+
"92": "NON_PARTICIPATING",
|
254 |
"93": "SUCCESS",
|
255 |
"94": "SUCCESS",
|
256 |
"95": "SUCCESS",
|
257 |
"96": "SUCCESS",
|
258 |
"97": "SUCCESS",
|
259 |
+
"98": "NON_PARTICIPATING",
|
260 |
"99": "SUCCESS"
|
261 |
},
|
262 |
"architectures": [
|
|
|
267 |
"AutoConfig": "distributed/optimized-gpt2-500m--configuration_gpt_optimized.GPTOptimConfig",
|
268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
269 |
},
|
270 |
+
"block_list": [],
|
|
|
|
|
|
|
|
|
|
|
|
|
271 |
"block_size": 1024,
|
272 |
"bos_token_id": 50256,
|
273 |
"embd_pdrop": 0.1,
|
274 |
"eos_token_id": 50256,
|
275 |
"initializer_range": 0.02,
|
276 |
+
"inner_step": 0,
|
277 |
"inner_steps": 0,
|
278 |
+
"last_allreduce_block": 5063053,
|
279 |
"layer_norm_epsilon": 1e-05,
|
280 |
"model_type": "gpt_optimized",
|
281 |
"n_embd": 1280,
|
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 sha256:5faee8098e9190b4acd3c97a1564e15de3e0cc1b07e9b0741b26cf92e98fcfbe
|
3 |
size 4040701744
|