0e91cd0cc260f9a317ba60aa3c52c7ad96c69e49bee5cc2245b6410888087dba
Browse files- config.json +28 -28
- model.safetensors +1 -1
- smash_config.json +1 -1
config.json
CHANGED
@@ -1,15 +1,15 @@
|
|
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": "SUCCESS",
|
10 |
"102": "NON_PARTICIPATING",
|
11 |
"103": "NON_PARTICIPATING",
|
12 |
-
"104": "
|
13 |
"105": "NON_PARTICIPATING",
|
14 |
"106": "NON_PARTICIPATING",
|
15 |
"107": "NON_PARTICIPATING",
|
@@ -20,12 +20,12 @@
|
|
20 |
"111": "NON_PARTICIPATING",
|
21 |
"112": "NON_PARTICIPATING",
|
22 |
"113": "NON_PARTICIPATING",
|
23 |
-
"114": "
|
24 |
"115": "NON_PARTICIPATING",
|
25 |
"116": "NON_PARTICIPATING",
|
26 |
"117": "NON_PARTICIPATING",
|
27 |
"118": "NON_PARTICIPATING",
|
28 |
-
"119": "
|
29 |
"12": "NON_PARTICIPATING",
|
30 |
"120": "NON_PARTICIPATING",
|
31 |
"121": "NON_PARTICIPATING",
|
@@ -36,8 +36,8 @@
|
|
36 |
"126": "NON_PARTICIPATING",
|
37 |
"127": "NON_PARTICIPATING",
|
38 |
"128": "NON_PARTICIPATING",
|
39 |
-
"129": "
|
40 |
-
"13": "
|
41 |
"130": "NON_PARTICIPATING",
|
42 |
"131": "NON_PARTICIPATING",
|
43 |
"132": "NON_PARTICIPATING",
|
@@ -48,7 +48,7 @@
|
|
48 |
"137": "NON_PARTICIPATING",
|
49 |
"138": "NON_PARTICIPATING",
|
50 |
"139": "NON_PARTICIPATING",
|
51 |
-
"14": "
|
52 |
"140": "SUCCESS",
|
53 |
"141": "NON_PARTICIPATING",
|
54 |
"142": "NON_PARTICIPATING",
|
@@ -64,13 +64,13 @@
|
|
64 |
"151": "NON_PARTICIPATING",
|
65 |
"152": "NON_PARTICIPATING",
|
66 |
"153": "NON_PARTICIPATING",
|
67 |
-
"154": "
|
68 |
"155": "NON_PARTICIPATING",
|
69 |
-
"156": "
|
70 |
"157": "NON_PARTICIPATING",
|
71 |
-
"158": "
|
72 |
-
"159": "
|
73 |
-
"16": "
|
74 |
"160": "NON_PARTICIPATING",
|
75 |
"161": "NON_PARTICIPATING",
|
76 |
"162": "NON_PARTICIPATING",
|
@@ -120,7 +120,7 @@
|
|
120 |
"201": "SUCCESS",
|
121 |
"202": "NON_PARTICIPATING",
|
122 |
"203": "NON_PARTICIPATING",
|
123 |
-
"204": "
|
124 |
"205": "NON_PARTICIPATING",
|
125 |
"206": "NON_PARTICIPATING",
|
126 |
"207": "NON_PARTICIPATING",
|
@@ -134,7 +134,7 @@
|
|
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",
|
@@ -178,7 +178,7 @@
|
|
178 |
"254": "NON_PARTICIPATING",
|
179 |
"255": "NON_PARTICIPATING",
|
180 |
"26": "NON_PARTICIPATING",
|
181 |
-
"27": "
|
182 |
"28": "NON_PARTICIPATING",
|
183 |
"29": "NON_PARTICIPATING",
|
184 |
"3": "NON_PARTICIPATING",
|
@@ -198,10 +198,10 @@
|
|
198 |
"42": "NON_PARTICIPATING",
|
199 |
"43": "NON_PARTICIPATING",
|
200 |
"44": "NON_PARTICIPATING",
|
201 |
-
"45": "
|
202 |
"46": "NON_PARTICIPATING",
|
203 |
"47": "NON_PARTICIPATING",
|
204 |
-
"48": "
|
205 |
"49": "NON_PARTICIPATING",
|
206 |
"5": "NON_PARTICIPATING",
|
207 |
"50": "NON_PARTICIPATING",
|
@@ -209,8 +209,8 @@
|
|
209 |
"52": "NON_PARTICIPATING",
|
210 |
"53": "NON_PARTICIPATING",
|
211 |
"54": "NON_PARTICIPATING",
|
212 |
-
"55": "
|
213 |
-
"56": "
|
214 |
"57": "NON_PARTICIPATING",
|
215 |
"58": "SUCCESS",
|
216 |
"59": "NON_PARTICIPATING",
|
@@ -229,34 +229,34 @@
|
|
229 |
"70": "NON_PARTICIPATING",
|
230 |
"71": "NON_PARTICIPATING",
|
231 |
"72": "NON_PARTICIPATING",
|
232 |
-
"73": "
|
233 |
-
"74": "
|
234 |
"75": "NON_PARTICIPATING",
|
235 |
"76": "NON_PARTICIPATING",
|
236 |
"77": "NON_PARTICIPATING",
|
237 |
-
"78": "
|
238 |
"79": "NON_PARTICIPATING",
|
239 |
"8": "NON_PARTICIPATING",
|
240 |
"80": "SUCCESS",
|
241 |
"81": "NON_PARTICIPATING",
|
242 |
-
"82": "
|
243 |
"83": "NON_PARTICIPATING",
|
244 |
"84": "NON_PARTICIPATING",
|
245 |
-
"85": "
|
246 |
"86": "SUCCESS",
|
247 |
"87": "NON_PARTICIPATING",
|
248 |
"88": "NON_PARTICIPATING",
|
249 |
"89": "NON_PARTICIPATING",
|
250 |
-
"9": "
|
251 |
"90": "NON_PARTICIPATING",
|
252 |
"91": "NON_PARTICIPATING",
|
253 |
"92": "NON_PARTICIPATING",
|
254 |
"93": "NON_PARTICIPATING",
|
255 |
"94": "SUCCESS",
|
256 |
"95": "NON_PARTICIPATING",
|
257 |
-
"96": "
|
258 |
"97": "NON_PARTICIPATING",
|
259 |
-
"98": "
|
260 |
"99": "NON_PARTICIPATING"
|
261 |
},
|
262 |
"architectures": [
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "/covalent/.cache/models/tmpgb9zdxe9xo01w5z2",
|
3 |
"activation_function": "gelu_new",
|
4 |
"all_reduce_scores": {
|
5 |
"0": "NON_PARTICIPATING",
|
6 |
"1": "NON_PARTICIPATING",
|
7 |
+
"10": "SUCCESS",
|
8 |
"100": "NON_PARTICIPATING",
|
9 |
"101": "SUCCESS",
|
10 |
"102": "NON_PARTICIPATING",
|
11 |
"103": "NON_PARTICIPATING",
|
12 |
+
"104": "NON_PARTICIPATING",
|
13 |
"105": "NON_PARTICIPATING",
|
14 |
"106": "NON_PARTICIPATING",
|
15 |
"107": "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": "NON_PARTICIPATING",
|
29 |
"12": "NON_PARTICIPATING",
|
30 |
"120": "NON_PARTICIPATING",
|
31 |
"121": "NON_PARTICIPATING",
|
|
|
36 |
"126": "NON_PARTICIPATING",
|
37 |
"127": "NON_PARTICIPATING",
|
38 |
"128": "NON_PARTICIPATING",
|
39 |
+
"129": "NON_PARTICIPATING",
|
40 |
+
"13": "SUCCESS",
|
41 |
"130": "NON_PARTICIPATING",
|
42 |
"131": "NON_PARTICIPATING",
|
43 |
"132": "NON_PARTICIPATING",
|
|
|
48 |
"137": "NON_PARTICIPATING",
|
49 |
"138": "NON_PARTICIPATING",
|
50 |
"139": "NON_PARTICIPATING",
|
51 |
+
"14": "SUCCESS",
|
52 |
"140": "SUCCESS",
|
53 |
"141": "NON_PARTICIPATING",
|
54 |
"142": "NON_PARTICIPATING",
|
|
|
64 |
"151": "NON_PARTICIPATING",
|
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": "SUCCESS",
|
74 |
"160": "NON_PARTICIPATING",
|
75 |
"161": "NON_PARTICIPATING",
|
76 |
"162": "NON_PARTICIPATING",
|
|
|
120 |
"201": "SUCCESS",
|
121 |
"202": "NON_PARTICIPATING",
|
122 |
"203": "NON_PARTICIPATING",
|
123 |
+
"204": "NON_PARTICIPATING",
|
124 |
"205": "NON_PARTICIPATING",
|
125 |
"206": "NON_PARTICIPATING",
|
126 |
"207": "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",
|
|
|
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",
|
|
|
198 |
"42": "NON_PARTICIPATING",
|
199 |
"43": "NON_PARTICIPATING",
|
200 |
"44": "NON_PARTICIPATING",
|
201 |
+
"45": "SUCCESS",
|
202 |
"46": "NON_PARTICIPATING",
|
203 |
"47": "NON_PARTICIPATING",
|
204 |
+
"48": "SUCCESS",
|
205 |
"49": "NON_PARTICIPATING",
|
206 |
"5": "NON_PARTICIPATING",
|
207 |
"50": "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": "SUCCESS",
|
216 |
"59": "NON_PARTICIPATING",
|
|
|
229 |
"70": "NON_PARTICIPATING",
|
230 |
"71": "NON_PARTICIPATING",
|
231 |
"72": "NON_PARTICIPATING",
|
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": "NON_PARTICIPATING",
|
239 |
"8": "NON_PARTICIPATING",
|
240 |
"80": "SUCCESS",
|
241 |
"81": "NON_PARTICIPATING",
|
242 |
+
"82": "SUCCESS",
|
243 |
"83": "NON_PARTICIPATING",
|
244 |
"84": "NON_PARTICIPATING",
|
245 |
+
"85": "NON_PARTICIPATING",
|
246 |
"86": "SUCCESS",
|
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": "SUCCESS",
|
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:be116a07ffa8ae2c80a7464f42f82fa1226f3f6743c09614d3977c0330c06f38
|
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/tmpgb9zdxe9",
|
32 |
"task": "",
|
33 |
"save_load_fn": "bitsandbytes",
|
34 |
"save_load_fn_args": {}
|