c2c5a18196d024e5e56c85bd30143ebaa9643db4c8f467ad56ad2aab5cf22081
Browse files- config.json +35 -35
- 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": "NON_PARTICIPATING",
|
8 |
"100": "NON_PARTICIPATING",
|
9 |
-
"101": "
|
10 |
"102": "NON_PARTICIPATING",
|
11 |
"103": "NON_PARTICIPATING",
|
12 |
"104": "SUCCESS",
|
@@ -23,7 +23,7 @@
|
|
23 |
"114": "NON_PARTICIPATING",
|
24 |
"115": "NON_PARTICIPATING",
|
25 |
"116": "NON_PARTICIPATING",
|
26 |
-
"117": "
|
27 |
"118": "NON_PARTICIPATING",
|
28 |
"119": "NON_PARTICIPATING",
|
29 |
"12": "NON_PARTICIPATING",
|
@@ -49,38 +49,38 @@
|
|
49 |
"138": "NON_PARTICIPATING",
|
50 |
"139": "NON_PARTICIPATING",
|
51 |
"14": "NON_PARTICIPATING",
|
52 |
-
"140": "
|
53 |
-
"141": "
|
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": "
|
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": "
|
68 |
-
"155": "
|
69 |
"156": "NON_PARTICIPATING",
|
70 |
"157": "NON_PARTICIPATING",
|
71 |
"158": "NON_PARTICIPATING",
|
72 |
-
"159": "
|
73 |
"16": "SUCCESS",
|
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": "NON_PARTICIPATING",
|
83 |
-
"169": "
|
84 |
"17": "NON_PARTICIPATING",
|
85 |
"170": "NON_PARTICIPATING",
|
86 |
"171": "NON_PARTICIPATING",
|
@@ -90,7 +90,7 @@
|
|
90 |
"175": "NON_PARTICIPATING",
|
91 |
"176": "NON_PARTICIPATING",
|
92 |
"177": "NON_PARTICIPATING",
|
93 |
-
"178": "
|
94 |
"179": "NON_PARTICIPATING",
|
95 |
"18": "NON_PARTICIPATING",
|
96 |
"180": "NON_PARTICIPATING",
|
@@ -114,18 +114,18 @@
|
|
114 |
"197": "NON_PARTICIPATING",
|
115 |
"198": "NON_PARTICIPATING",
|
116 |
"199": "NON_PARTICIPATING",
|
117 |
-
"2": "
|
118 |
"20": "NON_PARTICIPATING",
|
119 |
"200": "NON_PARTICIPATING",
|
120 |
"201": "NON_PARTICIPATING",
|
121 |
"202": "NON_PARTICIPATING",
|
122 |
"203": "NON_PARTICIPATING",
|
123 |
-
"204": "
|
124 |
"205": "NON_PARTICIPATING",
|
125 |
"206": "NON_PARTICIPATING",
|
126 |
"207": "NON_PARTICIPATING",
|
127 |
-
"208": "
|
128 |
-
"209": "
|
129 |
"21": "NON_PARTICIPATING",
|
130 |
"210": "NON_PARTICIPATING",
|
131 |
"211": "NON_PARTICIPATING",
|
@@ -134,10 +134,10 @@
|
|
134 |
"214": "NON_PARTICIPATING",
|
135 |
"215": "NON_PARTICIPATING",
|
136 |
"216": "NON_PARTICIPATING",
|
137 |
-
"217": "
|
138 |
"218": "NON_PARTICIPATING",
|
139 |
"219": "NON_PARTICIPATING",
|
140 |
-
"22": "
|
141 |
"220": "NON_PARTICIPATING",
|
142 |
"221": "NON_PARTICIPATING",
|
143 |
"222": "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",
|
@@ -198,7 +198,7 @@
|
|
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",
|
@@ -208,53 +208,53 @@
|
|
208 |
"51": "NON_PARTICIPATING",
|
209 |
"52": "NON_PARTICIPATING",
|
210 |
"53": "NON_PARTICIPATING",
|
211 |
-
"54": "
|
212 |
"55": "NON_PARTICIPATING",
|
213 |
"56": "NON_PARTICIPATING",
|
214 |
"57": "NON_PARTICIPATING",
|
215 |
-
"58": "
|
216 |
"59": "NON_PARTICIPATING",
|
217 |
-
"6": "
|
218 |
"60": "NON_PARTICIPATING",
|
219 |
"61": "NON_PARTICIPATING",
|
220 |
-
"62": "
|
221 |
"63": "NON_PARTICIPATING",
|
222 |
"64": "NON_PARTICIPATING",
|
223 |
-
"65": "
|
224 |
"66": "NON_PARTICIPATING",
|
225 |
-
"67": "
|
226 |
"68": "NON_PARTICIPATING",
|
227 |
"69": "NON_PARTICIPATING",
|
228 |
"7": "NON_PARTICIPATING",
|
229 |
"70": "NON_PARTICIPATING",
|
230 |
"71": "NON_PARTICIPATING",
|
231 |
"72": "NON_PARTICIPATING",
|
232 |
-
"73": "
|
233 |
"74": "NON_PARTICIPATING",
|
234 |
"75": "SUCCESS",
|
235 |
"76": "NON_PARTICIPATING",
|
236 |
-
"77": "
|
237 |
"78": "NON_PARTICIPATING",
|
238 |
"79": "NON_PARTICIPATING",
|
239 |
"8": "NON_PARTICIPATING",
|
240 |
-
"80": "
|
241 |
"81": "NON_PARTICIPATING",
|
242 |
"82": "NON_PARTICIPATING",
|
243 |
"83": "NON_PARTICIPATING",
|
244 |
"84": "NON_PARTICIPATING",
|
245 |
-
"85": "
|
246 |
-
"86": "
|
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": "
|
255 |
"94": "NON_PARTICIPATING",
|
256 |
"95": "SUCCESS",
|
257 |
-
"96": "
|
258 |
"97": "NON_PARTICIPATING",
|
259 |
"98": "SUCCESS",
|
260 |
"99": "NON_PARTICIPATING"
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "/covalent/.cache/models/tmpgxt7xdyndnmadags",
|
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": "SUCCESS",
|
10 |
"102": "NON_PARTICIPATING",
|
11 |
"103": "NON_PARTICIPATING",
|
12 |
"104": "SUCCESS",
|
|
|
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",
|
|
|
49 |
"138": "NON_PARTICIPATING",
|
50 |
"139": "NON_PARTICIPATING",
|
51 |
"14": "NON_PARTICIPATING",
|
52 |
+
"140": "SUCCESS",
|
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",
|
64 |
"151": "NON_PARTICIPATING",
|
65 |
"152": "NON_PARTICIPATING",
|
66 |
"153": "NON_PARTICIPATING",
|
67 |
+
"154": "SUCCESS",
|
68 |
+
"155": "NON_PARTICIPATING",
|
69 |
"156": "NON_PARTICIPATING",
|
70 |
"157": "NON_PARTICIPATING",
|
71 |
"158": "NON_PARTICIPATING",
|
72 |
+
"159": "SUCCESS",
|
73 |
"16": "SUCCESS",
|
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": "SUCCESS",
|
82 |
"168": "NON_PARTICIPATING",
|
83 |
+
"169": "NON_PARTICIPATING",
|
84 |
"17": "NON_PARTICIPATING",
|
85 |
"170": "NON_PARTICIPATING",
|
86 |
"171": "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",
|
|
|
114 |
"197": "NON_PARTICIPATING",
|
115 |
"198": "NON_PARTICIPATING",
|
116 |
"199": "NON_PARTICIPATING",
|
117 |
+
"2": "SUCCESS",
|
118 |
"20": "NON_PARTICIPATING",
|
119 |
"200": "NON_PARTICIPATING",
|
120 |
"201": "NON_PARTICIPATING",
|
121 |
"202": "NON_PARTICIPATING",
|
122 |
"203": "NON_PARTICIPATING",
|
123 |
+
"204": "SUCCESS",
|
124 |
"205": "NON_PARTICIPATING",
|
125 |
"206": "NON_PARTICIPATING",
|
126 |
"207": "NON_PARTICIPATING",
|
127 |
+
"208": "NON_PARTICIPATING",
|
128 |
+
"209": "SUCCESS",
|
129 |
"21": "NON_PARTICIPATING",
|
130 |
"210": "NON_PARTICIPATING",
|
131 |
"211": "NON_PARTICIPATING",
|
|
|
134 |
"214": "NON_PARTICIPATING",
|
135 |
"215": "NON_PARTICIPATING",
|
136 |
"216": "NON_PARTICIPATING",
|
137 |
+
"217": "SUCCESS",
|
138 |
"218": "NON_PARTICIPATING",
|
139 |
"219": "NON_PARTICIPATING",
|
140 |
+
"22": "SUCCESS",
|
141 |
"220": "NON_PARTICIPATING",
|
142 |
"221": "NON_PARTICIPATING",
|
143 |
"222": "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",
|
|
|
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",
|
|
|
208 |
"51": "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",
|
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": "NON_PARTICIPATING",
|
225 |
+
"67": "SUCCESS",
|
226 |
"68": "NON_PARTICIPATING",
|
227 |
"69": "NON_PARTICIPATING",
|
228 |
"7": "NON_PARTICIPATING",
|
229 |
"70": "NON_PARTICIPATING",
|
230 |
"71": "NON_PARTICIPATING",
|
231 |
"72": "NON_PARTICIPATING",
|
232 |
+
"73": "NON_PARTICIPATING",
|
233 |
"74": "NON_PARTICIPATING",
|
234 |
"75": "SUCCESS",
|
235 |
"76": "NON_PARTICIPATING",
|
236 |
+
"77": "NON_PARTICIPATING",
|
237 |
"78": "NON_PARTICIPATING",
|
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": "SUCCESS",
|
246 |
+
"86": "SUCCESS",
|
247 |
"87": "NON_PARTICIPATING",
|
248 |
"88": "NON_PARTICIPATING",
|
249 |
"89": "NON_PARTICIPATING",
|
250 |
+
"9": "SUCCESS",
|
251 |
"90": "NON_PARTICIPATING",
|
252 |
"91": "NON_PARTICIPATING",
|
253 |
"92": "NON_PARTICIPATING",
|
254 |
+
"93": "SUCCESS",
|
255 |
"94": "NON_PARTICIPATING",
|
256 |
"95": "SUCCESS",
|
257 |
+
"96": "NON_PARTICIPATING",
|
258 |
"97": "NON_PARTICIPATING",
|
259 |
"98": "SUCCESS",
|
260 |
"99": "NON_PARTICIPATING"
|
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:b141718d95274c42f1cc2d7a1ab502efca3ffcd1c1d66519e42fc6ff02dc2b49
|
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/tmpgxt7xdyn",
|
32 |
"task": "",
|
33 |
"save_load_fn": "bitsandbytes",
|
34 |
"save_load_fn_args": {}
|