d7259bd53252d5fb55a7bcbbdf41379ce877169d3237e7a3408820a220f6d8dc
Browse files- config.json +37 -37
- model.safetensors +1 -1
- smash_config.json +1 -1
config.json
CHANGED
@@ -1,12 +1,12 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "/covalent/.cache/models/
|
3 |
"activation_function": "gelu_new",
|
4 |
"all_reduce_scores": {
|
5 |
"0": "NON_PARTICIPATING",
|
6 |
"1": "NON_PARTICIPATING",
|
7 |
-
"10": "
|
8 |
"100": "NON_PARTICIPATING",
|
9 |
-
"101": "
|
10 |
"102": "NON_PARTICIPATING",
|
11 |
"103": "NON_PARTICIPATING",
|
12 |
"104": "NON_PARTICIPATING",
|
@@ -20,7 +20,7 @@
|
|
20 |
"111": "NON_PARTICIPATING",
|
21 |
"112": "NON_PARTICIPATING",
|
22 |
"113": "NON_PARTICIPATING",
|
23 |
-
"114": "
|
24 |
"115": "NON_PARTICIPATING",
|
25 |
"116": "NON_PARTICIPATING",
|
26 |
"117": "SUCCESS",
|
@@ -31,7 +31,7 @@
|
|
31 |
"121": "NON_PARTICIPATING",
|
32 |
"122": "NON_PARTICIPATING",
|
33 |
"123": "NON_PARTICIPATING",
|
34 |
-
"124": "
|
35 |
"125": "NON_PARTICIPATING",
|
36 |
"126": "NON_PARTICIPATING",
|
37 |
"127": "NON_PARTICIPATING",
|
@@ -51,13 +51,13 @@
|
|
51 |
"14": "NON_PARTICIPATING",
|
52 |
"140": "NON_PARTICIPATING",
|
53 |
"141": "SUCCESS",
|
54 |
-
"142": "
|
55 |
"143": "NON_PARTICIPATING",
|
56 |
"144": "NON_PARTICIPATING",
|
57 |
"145": "NON_PARTICIPATING",
|
58 |
"146": "NON_PARTICIPATING",
|
59 |
"147": "NON_PARTICIPATING",
|
60 |
-
"148": "
|
61 |
"149": "NON_PARTICIPATING",
|
62 |
"15": "NON_PARTICIPATING",
|
63 |
"150": "NON_PARTICIPATING",
|
@@ -68,23 +68,23 @@
|
|
68 |
"155": "SUCCESS",
|
69 |
"156": "SUCCESS",
|
70 |
"157": "NON_PARTICIPATING",
|
71 |
-
"158": "
|
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": "
|
79 |
"165": "NON_PARTICIPATING",
|
80 |
"166": "NON_PARTICIPATING",
|
81 |
-
"167": "
|
82 |
-
"168": "
|
83 |
"169": "SUCCESS",
|
84 |
"17": "NON_PARTICIPATING",
|
85 |
"170": "NON_PARTICIPATING",
|
86 |
"171": "NON_PARTICIPATING",
|
87 |
-
"172": "
|
88 |
"173": "NON_PARTICIPATING",
|
89 |
"174": "NON_PARTICIPATING",
|
90 |
"175": "NON_PARTICIPATING",
|
@@ -101,7 +101,7 @@
|
|
101 |
"185": "NON_PARTICIPATING",
|
102 |
"186": "NON_PARTICIPATING",
|
103 |
"187": "NON_PARTICIPATING",
|
104 |
-
"188": "
|
105 |
"189": "NON_PARTICIPATING",
|
106 |
"19": "NON_PARTICIPATING",
|
107 |
"190": "NON_PARTICIPATING",
|
@@ -133,8 +133,8 @@
|
|
133 |
"213": "NON_PARTICIPATING",
|
134 |
"214": "NON_PARTICIPATING",
|
135 |
"215": "NON_PARTICIPATING",
|
136 |
-
"216": "
|
137 |
-
"217": "
|
138 |
"218": "NON_PARTICIPATING",
|
139 |
"219": "NON_PARTICIPATING",
|
140 |
"22": "NON_PARTICIPATING",
|
@@ -143,24 +143,24 @@
|
|
143 |
"222": "NON_PARTICIPATING",
|
144 |
"223": "NON_PARTICIPATING",
|
145 |
"224": "NON_PARTICIPATING",
|
146 |
-
"225": "
|
147 |
"226": "NON_PARTICIPATING",
|
148 |
"227": "SUCCESS",
|
149 |
-
"228": "
|
150 |
"229": "NON_PARTICIPATING",
|
151 |
"23": "NON_PARTICIPATING",
|
152 |
"230": "SUCCESS",
|
153 |
-
"231": "
|
154 |
"232": "NON_PARTICIPATING",
|
155 |
"233": "NON_PARTICIPATING",
|
156 |
"234": "NON_PARTICIPATING",
|
157 |
"235": "NON_PARTICIPATING",
|
158 |
-
"236": "
|
159 |
"237": "SUCCESS",
|
160 |
"238": "SUCCESS",
|
161 |
-
"239": "
|
162 |
"24": "NON_PARTICIPATING",
|
163 |
-
"240": "
|
164 |
"241": "NON_PARTICIPATING",
|
165 |
"242": "NON_PARTICIPATING",
|
166 |
"243": "NON_PARTICIPATING",
|
@@ -173,13 +173,13 @@
|
|
173 |
"25": "SUCCESS",
|
174 |
"250": "NON_PARTICIPATING",
|
175 |
"251": "NON_PARTICIPATING",
|
176 |
-
"252": "
|
177 |
"253": "NON_PARTICIPATING",
|
178 |
"254": "NON_PARTICIPATING",
|
179 |
"255": "NON_PARTICIPATING",
|
180 |
"26": "NON_PARTICIPATING",
|
181 |
"27": "SUCCESS",
|
182 |
-
"28": "
|
183 |
"29": "NON_PARTICIPATING",
|
184 |
"3": "NON_PARTICIPATING",
|
185 |
"30": "NON_PARTICIPATING",
|
@@ -191,14 +191,14 @@
|
|
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": "NON_PARTICIPATING",
|
199 |
"43": "NON_PARTICIPATING",
|
200 |
"44": "NON_PARTICIPATING",
|
201 |
-
"45": "
|
202 |
"46": "NON_PARTICIPATING",
|
203 |
"47": "NON_PARTICIPATING",
|
204 |
"48": "SUCCESS",
|
@@ -208,44 +208,44 @@
|
|
208 |
"51": "NON_PARTICIPATING",
|
209 |
"52": "NON_PARTICIPATING",
|
210 |
"53": "NON_PARTICIPATING",
|
211 |
-
"54": "
|
212 |
"55": "NON_PARTICIPATING",
|
213 |
-
"56": "
|
214 |
"57": "NON_PARTICIPATING",
|
215 |
"58": "NON_PARTICIPATING",
|
216 |
"59": "NON_PARTICIPATING",
|
217 |
-
"6": "
|
218 |
"60": "NON_PARTICIPATING",
|
219 |
"61": "NON_PARTICIPATING",
|
220 |
"62": "SUCCESS",
|
221 |
"63": "NON_PARTICIPATING",
|
222 |
"64": "NON_PARTICIPATING",
|
223 |
-
"65": "
|
224 |
"66": "SUCCESS",
|
225 |
"67": "SUCCESS",
|
226 |
"68": "NON_PARTICIPATING",
|
227 |
"69": "NON_PARTICIPATING",
|
228 |
-
"7": "
|
229 |
"70": "NON_PARTICIPATING",
|
230 |
-
"71": "
|
231 |
"72": "NON_PARTICIPATING",
|
232 |
"73": "SUCCESS",
|
233 |
"74": "NON_PARTICIPATING",
|
234 |
"75": "NON_PARTICIPATING",
|
235 |
"76": "NON_PARTICIPATING",
|
236 |
-
"77": "
|
237 |
"78": "SUCCESS",
|
238 |
"79": "NON_PARTICIPATING",
|
239 |
"8": "NON_PARTICIPATING",
|
240 |
"80": "NON_PARTICIPATING",
|
241 |
"81": "NON_PARTICIPATING",
|
242 |
-
"82": "
|
243 |
"83": "NON_PARTICIPATING",
|
244 |
"84": "SUCCESS",
|
245 |
"85": "NON_PARTICIPATING",
|
246 |
"86": "NON_PARTICIPATING",
|
247 |
"87": "NON_PARTICIPATING",
|
248 |
-
"88": "
|
249 |
"89": "NON_PARTICIPATING",
|
250 |
"9": "SUCCESS",
|
251 |
"90": "NON_PARTICIPATING",
|
@@ -253,10 +253,10 @@
|
|
253 |
"92": "NON_PARTICIPATING",
|
254 |
"93": "SUCCESS",
|
255 |
"94": "NON_PARTICIPATING",
|
256 |
-
"95": "
|
257 |
-
"96": "
|
258 |
"97": "NON_PARTICIPATING",
|
259 |
-
"98": "
|
260 |
"99": "NON_PARTICIPATING"
|
261 |
},
|
262 |
"architectures": [
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "/covalent/.cache/models/tmpu0ztj0gtljdmn_rg",
|
3 |
"activation_function": "gelu_new",
|
4 |
"all_reduce_scores": {
|
5 |
"0": "NON_PARTICIPATING",
|
6 |
"1": "NON_PARTICIPATING",
|
7 |
+
"10": "NON_PARTICIPATING",
|
8 |
"100": "NON_PARTICIPATING",
|
9 |
+
"101": "NON_PARTICIPATING",
|
10 |
"102": "NON_PARTICIPATING",
|
11 |
"103": "NON_PARTICIPATING",
|
12 |
"104": "NON_PARTICIPATING",
|
|
|
20 |
"111": "NON_PARTICIPATING",
|
21 |
"112": "NON_PARTICIPATING",
|
22 |
"113": "NON_PARTICIPATING",
|
23 |
+
"114": "SUCCESS",
|
24 |
"115": "NON_PARTICIPATING",
|
25 |
"116": "NON_PARTICIPATING",
|
26 |
"117": "SUCCESS",
|
|
|
31 |
"121": "NON_PARTICIPATING",
|
32 |
"122": "NON_PARTICIPATING",
|
33 |
"123": "NON_PARTICIPATING",
|
34 |
+
"124": "SUCCESS",
|
35 |
"125": "NON_PARTICIPATING",
|
36 |
"126": "NON_PARTICIPATING",
|
37 |
"127": "NON_PARTICIPATING",
|
|
|
51 |
"14": "NON_PARTICIPATING",
|
52 |
"140": "NON_PARTICIPATING",
|
53 |
"141": "SUCCESS",
|
54 |
+
"142": "SUCCESS",
|
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": "NON_PARTICIPATING",
|
62 |
"15": "NON_PARTICIPATING",
|
63 |
"150": "NON_PARTICIPATING",
|
|
|
68 |
"155": "SUCCESS",
|
69 |
"156": "SUCCESS",
|
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": "SUCCESS",
|
79 |
"165": "NON_PARTICIPATING",
|
80 |
"166": "NON_PARTICIPATING",
|
81 |
+
"167": "NON_PARTICIPATING",
|
82 |
+
"168": "NON_PARTICIPATING",
|
83 |
"169": "SUCCESS",
|
84 |
"17": "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",
|
|
|
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",
|
|
|
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": "NON_PARTICIPATING",
|
140 |
"22": "NON_PARTICIPATING",
|
|
|
143 |
"222": "NON_PARTICIPATING",
|
144 |
"223": "NON_PARTICIPATING",
|
145 |
"224": "NON_PARTICIPATING",
|
146 |
+
"225": "NON_PARTICIPATING",
|
147 |
"226": "NON_PARTICIPATING",
|
148 |
"227": "SUCCESS",
|
149 |
+
"228": "NON_PARTICIPATING",
|
150 |
"229": "NON_PARTICIPATING",
|
151 |
"23": "NON_PARTICIPATING",
|
152 |
"230": "SUCCESS",
|
153 |
+
"231": "SUCCESS",
|
154 |
"232": "NON_PARTICIPATING",
|
155 |
"233": "NON_PARTICIPATING",
|
156 |
"234": "NON_PARTICIPATING",
|
157 |
"235": "NON_PARTICIPATING",
|
158 |
+
"236": "NON_PARTICIPATING",
|
159 |
"237": "SUCCESS",
|
160 |
"238": "SUCCESS",
|
161 |
+
"239": "SUCCESS",
|
162 |
"24": "NON_PARTICIPATING",
|
163 |
+
"240": "NON_PARTICIPATING",
|
164 |
"241": "NON_PARTICIPATING",
|
165 |
"242": "NON_PARTICIPATING",
|
166 |
"243": "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": "SUCCESS",
|
183 |
"29": "NON_PARTICIPATING",
|
184 |
"3": "NON_PARTICIPATING",
|
185 |
"30": "NON_PARTICIPATING",
|
|
|
191 |
"36": "NON_PARTICIPATING",
|
192 |
"37": "NON_PARTICIPATING",
|
193 |
"38": "NON_PARTICIPATING",
|
194 |
+
"39": "NON_PARTICIPATING",
|
195 |
"4": "NON_PARTICIPATING",
|
196 |
"40": "NON_PARTICIPATING",
|
197 |
"41": "NON_PARTICIPATING",
|
198 |
"42": "NON_PARTICIPATING",
|
199 |
"43": "NON_PARTICIPATING",
|
200 |
"44": "NON_PARTICIPATING",
|
201 |
+
"45": "NON_PARTICIPATING",
|
202 |
"46": "NON_PARTICIPATING",
|
203 |
"47": "NON_PARTICIPATING",
|
204 |
"48": "SUCCESS",
|
|
|
208 |
"51": "NON_PARTICIPATING",
|
209 |
"52": "NON_PARTICIPATING",
|
210 |
"53": "NON_PARTICIPATING",
|
211 |
+
"54": "SUCCESS",
|
212 |
"55": "NON_PARTICIPATING",
|
213 |
+
"56": "SUCCESS",
|
214 |
"57": "NON_PARTICIPATING",
|
215 |
"58": "NON_PARTICIPATING",
|
216 |
"59": "NON_PARTICIPATING",
|
217 |
+
"6": "SUCCESS",
|
218 |
"60": "NON_PARTICIPATING",
|
219 |
"61": "NON_PARTICIPATING",
|
220 |
"62": "SUCCESS",
|
221 |
"63": "NON_PARTICIPATING",
|
222 |
"64": "NON_PARTICIPATING",
|
223 |
+
"65": "SUCCESS",
|
224 |
"66": "SUCCESS",
|
225 |
"67": "SUCCESS",
|
226 |
"68": "NON_PARTICIPATING",
|
227 |
"69": "NON_PARTICIPATING",
|
228 |
+
"7": "SUCCESS",
|
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": "SUCCESS",
|
237 |
"78": "SUCCESS",
|
238 |
"79": "NON_PARTICIPATING",
|
239 |
"8": "NON_PARTICIPATING",
|
240 |
"80": "NON_PARTICIPATING",
|
241 |
"81": "NON_PARTICIPATING",
|
242 |
+
"82": "SUCCESS",
|
243 |
"83": "NON_PARTICIPATING",
|
244 |
"84": "SUCCESS",
|
245 |
"85": "NON_PARTICIPATING",
|
246 |
"86": "NON_PARTICIPATING",
|
247 |
"87": "NON_PARTICIPATING",
|
248 |
+
"88": "SUCCESS",
|
249 |
"89": "NON_PARTICIPATING",
|
250 |
"9": "SUCCESS",
|
251 |
"90": "NON_PARTICIPATING",
|
|
|
253 |
"92": "NON_PARTICIPATING",
|
254 |
"93": "SUCCESS",
|
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": [
|
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:d260e04cf97694d965fc486a082b770635c2dc9e2edb272fbef94029d5d91e69
|
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/tmpu0ztj0gt",
|
32 |
"task": "",
|
33 |
"save_load_fn": "bitsandbytes",
|
34 |
"save_load_fn_args": {}
|