Upload folder using huggingface_hub
#32
by
sharpenb
- opened
- config.json +59 -59
- model.safetensors +1 -1
- smash_config.json +1 -1
config.json
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "/covalent/.cache/models/
|
3 |
"activation_function": "gelu_new",
|
4 |
"all_reduce_scores": {
|
5 |
"0": "NON_PARTICIPATING",
|
@@ -9,23 +9,23 @@
|
|
9 |
"101": "NON_PARTICIPATING",
|
10 |
"102": "NON_PARTICIPATING",
|
11 |
"103": "NON_PARTICIPATING",
|
12 |
-
"104": "
|
13 |
"105": "NON_PARTICIPATING",
|
14 |
"106": "NON_PARTICIPATING",
|
15 |
-
"107": "
|
16 |
"108": "NON_PARTICIPATING",
|
17 |
-
"109": "
|
18 |
"11": "NON_PARTICIPATING",
|
19 |
"110": "NON_PARTICIPATING",
|
20 |
-
"111": "
|
21 |
"112": "NON_PARTICIPATING",
|
22 |
"113": "NON_PARTICIPATING",
|
23 |
"114": "NON_PARTICIPATING",
|
24 |
"115": "NON_PARTICIPATING",
|
25 |
"116": "NON_PARTICIPATING",
|
26 |
"117": "NON_PARTICIPATING",
|
27 |
-
"118": "
|
28 |
-
"119": "
|
29 |
"12": "NON_PARTICIPATING",
|
30 |
"120": "NON_PARTICIPATING",
|
31 |
"121": "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",
|
@@ -52,7 +52,7 @@
|
|
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",
|
@@ -60,21 +60,21 @@
|
|
60 |
"148": "NON_PARTICIPATING",
|
61 |
"149": "NON_PARTICIPATING",
|
62 |
"15": "NON_PARTICIPATING",
|
63 |
-
"150": "
|
64 |
"151": "SUCCESS",
|
65 |
"152": "NON_PARTICIPATING",
|
66 |
"153": "NON_PARTICIPATING",
|
67 |
"154": "NON_PARTICIPATING",
|
68 |
"155": "NON_PARTICIPATING",
|
69 |
"156": "NON_PARTICIPATING",
|
70 |
-
"157": "
|
71 |
-
"158": "
|
72 |
"159": "NON_PARTICIPATING",
|
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",
|
@@ -85,32 +85,32 @@
|
|
85 |
"170": "NON_PARTICIPATING",
|
86 |
"171": "NON_PARTICIPATING",
|
87 |
"172": "NON_PARTICIPATING",
|
88 |
-
"173": "
|
89 |
-
"174": "
|
90 |
"175": "NON_PARTICIPATING",
|
91 |
"176": "NON_PARTICIPATING",
|
92 |
"177": "NON_PARTICIPATING",
|
93 |
"178": "SUCCESS",
|
94 |
-
"179": "
|
95 |
"18": "NON_PARTICIPATING",
|
96 |
"180": "NON_PARTICIPATING",
|
97 |
"181": "NON_PARTICIPATING",
|
98 |
-
"182": "
|
99 |
"183": "NON_PARTICIPATING",
|
100 |
-
"184": "
|
101 |
"185": "NON_PARTICIPATING",
|
102 |
-
"186": "
|
103 |
"187": "NON_PARTICIPATING",
|
104 |
-
"188": "
|
105 |
-
"189": "
|
106 |
"19": "NON_PARTICIPATING",
|
107 |
-
"190": "
|
108 |
-
"191": "
|
109 |
"192": "NON_PARTICIPATING",
|
110 |
"193": "NON_PARTICIPATING",
|
111 |
-
"194": "
|
112 |
"195": "NON_PARTICIPATING",
|
113 |
-
"196": "
|
114 |
"197": "NON_PARTICIPATING",
|
115 |
"198": "NON_PARTICIPATING",
|
116 |
"199": "NON_PARTICIPATING",
|
@@ -118,7 +118,7 @@
|
|
118 |
"20": "NON_PARTICIPATING",
|
119 |
"200": "NON_PARTICIPATING",
|
120 |
"201": "NON_PARTICIPATING",
|
121 |
-
"202": "
|
122 |
"203": "NON_PARTICIPATING",
|
123 |
"204": "NON_PARTICIPATING",
|
124 |
"205": "NON_PARTICIPATING",
|
@@ -134,18 +134,18 @@
|
|
134 |
"214": "NON_PARTICIPATING",
|
135 |
"215": "NON_PARTICIPATING",
|
136 |
"216": "NON_PARTICIPATING",
|
137 |
-
"217": "
|
138 |
"218": "NON_PARTICIPATING",
|
139 |
"219": "NON_PARTICIPATING",
|
140 |
"22": "NON_PARTICIPATING",
|
141 |
"220": "NON_PARTICIPATING",
|
142 |
-
"221": "
|
143 |
"222": "NON_PARTICIPATING",
|
144 |
"223": "NON_PARTICIPATING",
|
145 |
-
"224": "
|
146 |
"225": "NON_PARTICIPATING",
|
147 |
-
"226": "
|
148 |
-
"227": "
|
149 |
"228": "NON_PARTICIPATING",
|
150 |
"229": "NON_PARTICIPATING",
|
151 |
"23": "NON_PARTICIPATING",
|
@@ -154,17 +154,17 @@
|
|
154 |
"232": "NON_PARTICIPATING",
|
155 |
"233": "NON_PARTICIPATING",
|
156 |
"234": "NON_PARTICIPATING",
|
157 |
-
"235": "
|
158 |
"236": "NON_PARTICIPATING",
|
159 |
-
"237": "
|
160 |
"238": "NON_PARTICIPATING",
|
161 |
-
"239": "
|
162 |
"24": "NON_PARTICIPATING",
|
163 |
"240": "NON_PARTICIPATING",
|
164 |
"241": "NON_PARTICIPATING",
|
165 |
-
"242": "
|
166 |
-
"243": "
|
167 |
-
"244": "
|
168 |
"245": "NON_PARTICIPATING",
|
169 |
"246": "NON_PARTICIPATING",
|
170 |
"247": "NON_PARTICIPATING",
|
@@ -173,15 +173,15 @@
|
|
173 |
"25": "SUCCESS",
|
174 |
"250": "NON_PARTICIPATING",
|
175 |
"251": "NON_PARTICIPATING",
|
176 |
-
"252": "
|
177 |
"253": "NON_PARTICIPATING",
|
178 |
"254": "NON_PARTICIPATING",
|
179 |
-
"255": "
|
180 |
"26": "NON_PARTICIPATING",
|
181 |
-
"27": "
|
182 |
"28": "NON_PARTICIPATING",
|
183 |
"29": "NON_PARTICIPATING",
|
184 |
-
"3": "
|
185 |
"30": "NON_PARTICIPATING",
|
186 |
"31": "NON_PARTICIPATING",
|
187 |
"32": "NON_PARTICIPATING",
|
@@ -209,10 +209,10 @@
|
|
209 |
"52": "NON_PARTICIPATING",
|
210 |
"53": "NON_PARTICIPATING",
|
211 |
"54": "NON_PARTICIPATING",
|
212 |
-
"55": "
|
213 |
"56": "NON_PARTICIPATING",
|
214 |
"57": "NON_PARTICIPATING",
|
215 |
-
"58": "
|
216 |
"59": "NON_PARTICIPATING",
|
217 |
"6": "NON_PARTICIPATING",
|
218 |
"60": "NON_PARTICIPATING",
|
@@ -227,37 +227,37 @@
|
|
227 |
"69": "NON_PARTICIPATING",
|
228 |
"7": "NON_PARTICIPATING",
|
229 |
"70": "NON_PARTICIPATING",
|
230 |
-
"71": "
|
231 |
"72": "NON_PARTICIPATING",
|
232 |
"73": "SUCCESS",
|
233 |
-
"74": "
|
234 |
"75": "NON_PARTICIPATING",
|
235 |
-
"76": "
|
236 |
"77": "NON_PARTICIPATING",
|
237 |
-
"78": "
|
238 |
"79": "NON_PARTICIPATING",
|
239 |
"8": "NON_PARTICIPATING",
|
240 |
-
"80": "
|
241 |
"81": "NON_PARTICIPATING",
|
242 |
-
"82": "
|
243 |
"83": "NON_PARTICIPATING",
|
244 |
-
"84": "
|
245 |
-
"85": "
|
246 |
-
"86": "
|
247 |
"87": "NON_PARTICIPATING",
|
248 |
-
"88": "
|
249 |
-
"89": "
|
250 |
"9": "NON_PARTICIPATING",
|
251 |
"90": "NON_PARTICIPATING",
|
252 |
-
"91": "
|
253 |
"92": "NON_PARTICIPATING",
|
254 |
-
"93": "
|
255 |
-
"94": "
|
256 |
"95": "NON_PARTICIPATING",
|
257 |
-
"96": "
|
258 |
"97": "NON_PARTICIPATING",
|
259 |
"98": "NON_PARTICIPATING",
|
260 |
-
"99": "
|
261 |
},
|
262 |
"architectures": [
|
263 |
"GPTOptim"
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "/covalent/.cache/models/tmpzhu59ngnwn1ymqki",
|
3 |
"activation_function": "gelu_new",
|
4 |
"all_reduce_scores": {
|
5 |
"0": "NON_PARTICIPATING",
|
|
|
9 |
"101": "NON_PARTICIPATING",
|
10 |
"102": "NON_PARTICIPATING",
|
11 |
"103": "NON_PARTICIPATING",
|
12 |
+
"104": "SUCCESS",
|
13 |
"105": "NON_PARTICIPATING",
|
14 |
"106": "NON_PARTICIPATING",
|
15 |
+
"107": "NON_PARTICIPATING",
|
16 |
"108": "NON_PARTICIPATING",
|
17 |
+
"109": "NON_PARTICIPATING",
|
18 |
"11": "NON_PARTICIPATING",
|
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",
|
26 |
"117": "NON_PARTICIPATING",
|
27 |
+
"118": "NON_PARTICIPATING",
|
28 |
+
"119": "SUCCESS",
|
29 |
"12": "NON_PARTICIPATING",
|
30 |
"120": "NON_PARTICIPATING",
|
31 |
"121": "NON_PARTICIPATING",
|
|
|
44 |
"133": "NON_PARTICIPATING",
|
45 |
"134": "NON_PARTICIPATING",
|
46 |
"135": "NON_PARTICIPATING",
|
47 |
+
"136": "NON_PARTICIPATING",
|
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": "NON_PARTICIPATING",
|
56 |
"144": "NON_PARTICIPATING",
|
57 |
"145": "NON_PARTICIPATING",
|
58 |
"146": "NON_PARTICIPATING",
|
|
|
60 |
"148": "NON_PARTICIPATING",
|
61 |
"149": "NON_PARTICIPATING",
|
62 |
"15": "NON_PARTICIPATING",
|
63 |
+
"150": "NON_PARTICIPATING",
|
64 |
"151": "SUCCESS",
|
65 |
"152": "NON_PARTICIPATING",
|
66 |
"153": "NON_PARTICIPATING",
|
67 |
"154": "NON_PARTICIPATING",
|
68 |
"155": "NON_PARTICIPATING",
|
69 |
"156": "NON_PARTICIPATING",
|
70 |
+
"157": "NON_PARTICIPATING",
|
71 |
+
"158": "NON_PARTICIPATING",
|
72 |
"159": "NON_PARTICIPATING",
|
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",
|
|
|
85 |
"170": "NON_PARTICIPATING",
|
86 |
"171": "NON_PARTICIPATING",
|
87 |
"172": "NON_PARTICIPATING",
|
88 |
+
"173": "NON_PARTICIPATING",
|
89 |
+
"174": "NON_PARTICIPATING",
|
90 |
"175": "NON_PARTICIPATING",
|
91 |
"176": "NON_PARTICIPATING",
|
92 |
"177": "NON_PARTICIPATING",
|
93 |
"178": "SUCCESS",
|
94 |
+
"179": "NON_PARTICIPATING",
|
95 |
"18": "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": "NON_PARTICIPATING",
|
103 |
"187": "NON_PARTICIPATING",
|
104 |
+
"188": "NON_PARTICIPATING",
|
105 |
+
"189": "NON_PARTICIPATING",
|
106 |
"19": "NON_PARTICIPATING",
|
107 |
+
"190": "NON_PARTICIPATING",
|
108 |
+
"191": "NON_PARTICIPATING",
|
109 |
"192": "NON_PARTICIPATING",
|
110 |
"193": "NON_PARTICIPATING",
|
111 |
+
"194": "NON_PARTICIPATING",
|
112 |
"195": "NON_PARTICIPATING",
|
113 |
+
"196": "NON_PARTICIPATING",
|
114 |
"197": "NON_PARTICIPATING",
|
115 |
"198": "NON_PARTICIPATING",
|
116 |
"199": "NON_PARTICIPATING",
|
|
|
118 |
"20": "NON_PARTICIPATING",
|
119 |
"200": "NON_PARTICIPATING",
|
120 |
"201": "NON_PARTICIPATING",
|
121 |
+
"202": "NON_PARTICIPATING",
|
122 |
"203": "NON_PARTICIPATING",
|
123 |
"204": "NON_PARTICIPATING",
|
124 |
"205": "NON_PARTICIPATING",
|
|
|
134 |
"214": "NON_PARTICIPATING",
|
135 |
"215": "NON_PARTICIPATING",
|
136 |
"216": "NON_PARTICIPATING",
|
137 |
+
"217": "NON_PARTICIPATING",
|
138 |
"218": "NON_PARTICIPATING",
|
139 |
"219": "NON_PARTICIPATING",
|
140 |
"22": "NON_PARTICIPATING",
|
141 |
"220": "NON_PARTICIPATING",
|
142 |
+
"221": "NON_PARTICIPATING",
|
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",
|
|
|
154 |
"232": "NON_PARTICIPATING",
|
155 |
"233": "NON_PARTICIPATING",
|
156 |
"234": "NON_PARTICIPATING",
|
157 |
+
"235": "NON_PARTICIPATING",
|
158 |
"236": "NON_PARTICIPATING",
|
159 |
+
"237": "NON_PARTICIPATING",
|
160 |
"238": "NON_PARTICIPATING",
|
161 |
+
"239": "SUCCESS",
|
162 |
"24": "NON_PARTICIPATING",
|
163 |
"240": "NON_PARTICIPATING",
|
164 |
"241": "NON_PARTICIPATING",
|
165 |
+
"242": "NON_PARTICIPATING",
|
166 |
+
"243": "NON_PARTICIPATING",
|
167 |
+
"244": "SUCCESS",
|
168 |
"245": "NON_PARTICIPATING",
|
169 |
"246": "NON_PARTICIPATING",
|
170 |
"247": "NON_PARTICIPATING",
|
|
|
173 |
"25": "SUCCESS",
|
174 |
"250": "NON_PARTICIPATING",
|
175 |
"251": "NON_PARTICIPATING",
|
176 |
+
"252": "NON_PARTICIPATING",
|
177 |
"253": "NON_PARTICIPATING",
|
178 |
"254": "NON_PARTICIPATING",
|
179 |
+
"255": "NON_PARTICIPATING",
|
180 |
"26": "NON_PARTICIPATING",
|
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": "NON_PARTICIPATING",
|
|
|
209 |
"52": "NON_PARTICIPATING",
|
210 |
"53": "NON_PARTICIPATING",
|
211 |
"54": "NON_PARTICIPATING",
|
212 |
+
"55": "NON_PARTICIPATING",
|
213 |
"56": "NON_PARTICIPATING",
|
214 |
"57": "NON_PARTICIPATING",
|
215 |
+
"58": "NON_PARTICIPATING",
|
216 |
"59": "NON_PARTICIPATING",
|
217 |
"6": "NON_PARTICIPATING",
|
218 |
"60": "NON_PARTICIPATING",
|
|
|
227 |
"69": "NON_PARTICIPATING",
|
228 |
"7": "NON_PARTICIPATING",
|
229 |
"70": "NON_PARTICIPATING",
|
230 |
+
"71": "NON_PARTICIPATING",
|
231 |
"72": "NON_PARTICIPATING",
|
232 |
"73": "SUCCESS",
|
233 |
+
"74": "NON_PARTICIPATING",
|
234 |
"75": "NON_PARTICIPATING",
|
235 |
+
"76": "NON_PARTICIPATING",
|
236 |
"77": "NON_PARTICIPATING",
|
237 |
+
"78": "SUCCESS",
|
238 |
"79": "NON_PARTICIPATING",
|
239 |
"8": "NON_PARTICIPATING",
|
240 |
+
"80": "NON_PARTICIPATING",
|
241 |
"81": "NON_PARTICIPATING",
|
242 |
+
"82": "NON_PARTICIPATING",
|
243 |
"83": "NON_PARTICIPATING",
|
244 |
+
"84": "NON_PARTICIPATING",
|
245 |
+
"85": "NON_PARTICIPATING",
|
246 |
+
"86": "NON_PARTICIPATING",
|
247 |
"87": "NON_PARTICIPATING",
|
248 |
+
"88": "NON_PARTICIPATING",
|
249 |
+
"89": "NON_PARTICIPATING",
|
250 |
"9": "NON_PARTICIPATING",
|
251 |
"90": "NON_PARTICIPATING",
|
252 |
+
"91": "NON_PARTICIPATING",
|
253 |
"92": "NON_PARTICIPATING",
|
254 |
+
"93": "NON_PARTICIPATING",
|
255 |
+
"94": "NON_PARTICIPATING",
|
256 |
"95": "NON_PARTICIPATING",
|
257 |
+
"96": "SUCCESS",
|
258 |
"97": "NON_PARTICIPATING",
|
259 |
"98": "NON_PARTICIPATING",
|
260 |
+
"99": "NON_PARTICIPATING"
|
261 |
},
|
262 |
"architectures": [
|
263 |
"GPTOptim"
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1207575528
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:bb6e39c04e12bba1f9b95c52f3dff7549b4d23a98636baa290c0549f3b1d0b6b
|
3 |
size 1207575528
|
smash_config.json
CHANGED
@@ -28,7 +28,7 @@
|
|
28 |
"quant_llm-int8_weight_bits": 8,
|
29 |
"max_batch_size": 1,
|
30 |
"device": "cuda",
|
31 |
-
"cache_dir": "/covalent/.cache/models/
|
32 |
"task": "",
|
33 |
"save_load_fn": "bitsandbytes",
|
34 |
"save_load_fn_args": {}
|
|
|
28 |
"quant_llm-int8_weight_bits": 8,
|
29 |
"max_batch_size": 1,
|
30 |
"device": "cuda",
|
31 |
+
"cache_dir": "/covalent/.cache/models/tmpzhu59ngn",
|
32 |
"task": "",
|
33 |
"save_load_fn": "bitsandbytes",
|
34 |
"save_load_fn_args": {}
|