c85dc975372c9aa706bfbb24499a2ac57e5b9c11564938822b7a8866a6afdcd6
Browse files- config.json +52 -52
- 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": "NON_PARTICIPATING",
|
| 8 |
"100": "NON_PARTICIPATING",
|
| 9 |
-
"101": "
|
| 10 |
"102": "NON_PARTICIPATING",
|
| 11 |
"103": "NON_PARTICIPATING",
|
| 12 |
-
"104": "
|
| 13 |
"105": "NON_PARTICIPATING",
|
| 14 |
"106": "NON_PARTICIPATING",
|
| 15 |
"107": "SUCCESS",
|
|
@@ -22,11 +22,11 @@
|
|
| 22 |
"113": "NON_PARTICIPATING",
|
| 23 |
"114": "NON_PARTICIPATING",
|
| 24 |
"115": "NON_PARTICIPATING",
|
| 25 |
-
"116": "
|
| 26 |
"117": "NON_PARTICIPATING",
|
| 27 |
"118": "NON_PARTICIPATING",
|
| 28 |
"119": "NON_PARTICIPATING",
|
| 29 |
-
"12": "
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
"122": "NON_PARTICIPATING",
|
|
@@ -36,15 +36,15 @@
|
|
| 36 |
"126": "NON_PARTICIPATING",
|
| 37 |
"127": "NON_PARTICIPATING",
|
| 38 |
"128": "NON_PARTICIPATING",
|
| 39 |
-
"129": "
|
| 40 |
"13": "NON_PARTICIPATING",
|
| 41 |
"130": "NON_PARTICIPATING",
|
| 42 |
"131": "NON_PARTICIPATING",
|
| 43 |
"132": "NON_PARTICIPATING",
|
| 44 |
"133": "NON_PARTICIPATING",
|
| 45 |
-
"134": "
|
| 46 |
"135": "NON_PARTICIPATING",
|
| 47 |
-
"136": "
|
| 48 |
"137": "NON_PARTICIPATING",
|
| 49 |
"138": "NON_PARTICIPATING",
|
| 50 |
"139": "NON_PARTICIPATING",
|
|
@@ -61,12 +61,12 @@
|
|
| 61 |
"149": "NON_PARTICIPATING",
|
| 62 |
"15": "NON_PARTICIPATING",
|
| 63 |
"150": "SUCCESS",
|
| 64 |
-
"151": "
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
"153": "NON_PARTICIPATING",
|
| 67 |
"154": "NON_PARTICIPATING",
|
| 68 |
-
"155": "
|
| 69 |
-
"156": "
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
"159": "NON_PARTICIPATING",
|
|
@@ -80,34 +80,34 @@
|
|
| 80 |
"166": "NON_PARTICIPATING",
|
| 81 |
"167": "NON_PARTICIPATING",
|
| 82 |
"168": "NON_PARTICIPATING",
|
| 83 |
-
"169": "
|
| 84 |
"17": "NON_PARTICIPATING",
|
| 85 |
"170": "NON_PARTICIPATING",
|
| 86 |
"171": "NON_PARTICIPATING",
|
| 87 |
"172": "NON_PARTICIPATING",
|
| 88 |
"173": "SUCCESS",
|
| 89 |
"174": "SUCCESS",
|
| 90 |
-
"175": "
|
| 91 |
-
"176": "
|
| 92 |
"177": "NON_PARTICIPATING",
|
| 93 |
-
"178": "
|
| 94 |
"179": "NON_PARTICIPATING",
|
| 95 |
"18": "NON_PARTICIPATING",
|
| 96 |
"180": "NON_PARTICIPATING",
|
| 97 |
"181": "NON_PARTICIPATING",
|
| 98 |
-
"182": "
|
| 99 |
-
"183": "
|
| 100 |
-
"184": "
|
| 101 |
"185": "NON_PARTICIPATING",
|
| 102 |
"186": "SUCCESS",
|
| 103 |
"187": "NON_PARTICIPATING",
|
| 104 |
-
"188": "
|
| 105 |
-
"189": "
|
| 106 |
"19": "NON_PARTICIPATING",
|
| 107 |
"190": "SUCCESS",
|
| 108 |
"191": "NON_PARTICIPATING",
|
| 109 |
"192": "NON_PARTICIPATING",
|
| 110 |
-
"193": "
|
| 111 |
"194": "NON_PARTICIPATING",
|
| 112 |
"195": "NON_PARTICIPATING",
|
| 113 |
"196": "NON_PARTICIPATING",
|
|
@@ -118,14 +118,14 @@
|
|
| 118 |
"20": "NON_PARTICIPATING",
|
| 119 |
"200": "NON_PARTICIPATING",
|
| 120 |
"201": "NON_PARTICIPATING",
|
| 121 |
-
"202": "
|
| 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",
|
|
@@ -134,7 +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",
|
|
@@ -144,8 +144,8 @@
|
|
| 144 |
"223": "NON_PARTICIPATING",
|
| 145 |
"224": "NON_PARTICIPATING",
|
| 146 |
"225": "NON_PARTICIPATING",
|
| 147 |
-
"226": "
|
| 148 |
-
"227": "
|
| 149 |
"228": "NON_PARTICIPATING",
|
| 150 |
"229": "NON_PARTICIPATING",
|
| 151 |
"23": "NON_PARTICIPATING",
|
|
@@ -157,14 +157,14 @@
|
|
| 157 |
"235": "NON_PARTICIPATING",
|
| 158 |
"236": "NON_PARTICIPATING",
|
| 159 |
"237": "NON_PARTICIPATING",
|
| 160 |
-
"238": "
|
| 161 |
-
"239": "
|
| 162 |
"24": "NON_PARTICIPATING",
|
| 163 |
"240": "NON_PARTICIPATING",
|
| 164 |
"241": "NON_PARTICIPATING",
|
| 165 |
"242": "SUCCESS",
|
| 166 |
-
"243": "
|
| 167 |
-
"244": "
|
| 168 |
"245": "NON_PARTICIPATING",
|
| 169 |
"246": "NON_PARTICIPATING",
|
| 170 |
"247": "NON_PARTICIPATING",
|
|
@@ -181,17 +181,17 @@
|
|
| 181 |
"27": "NON_PARTICIPATING",
|
| 182 |
"28": "NON_PARTICIPATING",
|
| 183 |
"29": "NON_PARTICIPATING",
|
| 184 |
-
"3": "
|
| 185 |
-
"30": "
|
| 186 |
-
"31": "
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "SUCCESS",
|
| 190 |
-
"35": "
|
| 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",
|
|
@@ -212,7 +212,7 @@
|
|
| 212 |
"55": "SUCCESS",
|
| 213 |
"56": "NON_PARTICIPATING",
|
| 214 |
"57": "NON_PARTICIPATING",
|
| 215 |
-
"58": "
|
| 216 |
"59": "NON_PARTICIPATING",
|
| 217 |
"6": "NON_PARTICIPATING",
|
| 218 |
"60": "NON_PARTICIPATING",
|
|
@@ -222,41 +222,41 @@
|
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
"65": "NON_PARTICIPATING",
|
| 224 |
"66": "NON_PARTICIPATING",
|
| 225 |
-
"67": "
|
| 226 |
"68": "NON_PARTICIPATING",
|
| 227 |
"69": "NON_PARTICIPATING",
|
| 228 |
-
"7": "
|
| 229 |
"70": "NON_PARTICIPATING",
|
| 230 |
-
"71": "
|
| 231 |
-
"72": "
|
| 232 |
-
"73": "
|
| 233 |
-
"74": "
|
| 234 |
"75": "NON_PARTICIPATING",
|
| 235 |
"76": "SUCCESS",
|
| 236 |
"77": "NON_PARTICIPATING",
|
| 237 |
-
"78": "
|
| 238 |
"79": "NON_PARTICIPATING",
|
| 239 |
"8": "NON_PARTICIPATING",
|
| 240 |
-
"80": "
|
| 241 |
"81": "NON_PARTICIPATING",
|
| 242 |
"82": "SUCCESS",
|
| 243 |
-
"83": "
|
| 244 |
"84": "NON_PARTICIPATING",
|
| 245 |
-
"85": "
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
"87": "NON_PARTICIPATING",
|
| 248 |
"88": "SUCCESS",
|
| 249 |
"89": "SUCCESS",
|
| 250 |
-
"9": "
|
| 251 |
-
"90": "
|
| 252 |
-
"91": "
|
| 253 |
"92": "NON_PARTICIPATING",
|
| 254 |
"93": "NON_PARTICIPATING",
|
| 255 |
"94": "NON_PARTICIPATING",
|
| 256 |
"95": "NON_PARTICIPATING",
|
| 257 |
"96": "NON_PARTICIPATING",
|
| 258 |
"97": "NON_PARTICIPATING",
|
| 259 |
-
"98": "
|
| 260 |
"99": "SUCCESS"
|
| 261 |
},
|
| 262 |
"architectures": [
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "/covalent/.cache/models/tmpzougtkly2l2m4z1p",
|
| 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": "NON_PARTICIPATING",
|
| 13 |
"105": "NON_PARTICIPATING",
|
| 14 |
"106": "NON_PARTICIPATING",
|
| 15 |
"107": "SUCCESS",
|
|
|
|
| 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": "SUCCESS",
|
| 30 |
"120": "NON_PARTICIPATING",
|
| 31 |
"121": "NON_PARTICIPATING",
|
| 32 |
"122": "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": "NON_PARTICIPATING",
|
| 46 |
"135": "NON_PARTICIPATING",
|
| 47 |
+
"136": "SUCCESS",
|
| 48 |
"137": "NON_PARTICIPATING",
|
| 49 |
"138": "NON_PARTICIPATING",
|
| 50 |
"139": "NON_PARTICIPATING",
|
|
|
|
| 61 |
"149": "NON_PARTICIPATING",
|
| 62 |
"15": "NON_PARTICIPATING",
|
| 63 |
"150": "SUCCESS",
|
| 64 |
+
"151": "SUCCESS",
|
| 65 |
"152": "NON_PARTICIPATING",
|
| 66 |
"153": "NON_PARTICIPATING",
|
| 67 |
"154": "NON_PARTICIPATING",
|
| 68 |
+
"155": "SUCCESS",
|
| 69 |
+
"156": "SUCCESS",
|
| 70 |
"157": "NON_PARTICIPATING",
|
| 71 |
"158": "NON_PARTICIPATING",
|
| 72 |
"159": "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": "SUCCESS",
|
| 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": "NON_PARTICIPATING",
|
| 97 |
"181": "NON_PARTICIPATING",
|
| 98 |
+
"182": "NON_PARTICIPATING",
|
| 99 |
+
"183": "SUCCESS",
|
| 100 |
+
"184": "NON_PARTICIPATING",
|
| 101 |
"185": "NON_PARTICIPATING",
|
| 102 |
"186": "SUCCESS",
|
| 103 |
"187": "NON_PARTICIPATING",
|
| 104 |
+
"188": "NON_PARTICIPATING",
|
| 105 |
+
"189": "SUCCESS",
|
| 106 |
"19": "NON_PARTICIPATING",
|
| 107 |
"190": "SUCCESS",
|
| 108 |
"191": "NON_PARTICIPATING",
|
| 109 |
"192": "NON_PARTICIPATING",
|
| 110 |
+
"193": "SUCCESS",
|
| 111 |
"194": "NON_PARTICIPATING",
|
| 112 |
"195": "NON_PARTICIPATING",
|
| 113 |
"196": "NON_PARTICIPATING",
|
|
|
|
| 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": "NON_PARTICIPATING",
|
| 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": "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",
|
|
|
|
| 157 |
"235": "NON_PARTICIPATING",
|
| 158 |
"236": "NON_PARTICIPATING",
|
| 159 |
"237": "NON_PARTICIPATING",
|
| 160 |
+
"238": "SUCCESS",
|
| 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": "SUCCESS",
|
| 168 |
"245": "NON_PARTICIPATING",
|
| 169 |
"246": "NON_PARTICIPATING",
|
| 170 |
"247": "NON_PARTICIPATING",
|
|
|
|
| 181 |
"27": "NON_PARTICIPATING",
|
| 182 |
"28": "NON_PARTICIPATING",
|
| 183 |
"29": "NON_PARTICIPATING",
|
| 184 |
+
"3": "SUCCESS",
|
| 185 |
+
"30": "SUCCESS",
|
| 186 |
+
"31": "NON_PARTICIPATING",
|
| 187 |
"32": "SUCCESS",
|
| 188 |
"33": "NON_PARTICIPATING",
|
| 189 |
"34": "SUCCESS",
|
| 190 |
+
"35": "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",
|
|
|
|
| 212 |
"55": "SUCCESS",
|
| 213 |
"56": "NON_PARTICIPATING",
|
| 214 |
"57": "NON_PARTICIPATING",
|
| 215 |
+
"58": "NON_PARTICIPATING",
|
| 216 |
"59": "NON_PARTICIPATING",
|
| 217 |
"6": "NON_PARTICIPATING",
|
| 218 |
"60": "NON_PARTICIPATING",
|
|
|
|
| 222 |
"64": "NON_PARTICIPATING",
|
| 223 |
"65": "NON_PARTICIPATING",
|
| 224 |
"66": "NON_PARTICIPATING",
|
| 225 |
+
"67": "NON_PARTICIPATING",
|
| 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": "NON_PARTICIPATING",
|
| 235 |
"76": "SUCCESS",
|
| 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": "SUCCESS",
|
| 244 |
"84": "NON_PARTICIPATING",
|
| 245 |
+
"85": "NON_PARTICIPATING",
|
| 246 |
"86": "NON_PARTICIPATING",
|
| 247 |
"87": "NON_PARTICIPATING",
|
| 248 |
"88": "SUCCESS",
|
| 249 |
"89": "SUCCESS",
|
| 250 |
+
"9": "SUCCESS",
|
| 251 |
+
"90": "NON_PARTICIPATING",
|
| 252 |
+
"91": "SUCCESS",
|
| 253 |
"92": "NON_PARTICIPATING",
|
| 254 |
"93": "NON_PARTICIPATING",
|
| 255 |
"94": "NON_PARTICIPATING",
|
| 256 |
"95": "NON_PARTICIPATING",
|
| 257 |
"96": "NON_PARTICIPATING",
|
| 258 |
"97": "NON_PARTICIPATING",
|
| 259 |
+
"98": "NON_PARTICIPATING",
|
| 260 |
"99": "SUCCESS"
|
| 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 sha256:5e02b75babc21d54280ad083cf287c36bb9674c798aea3b5a91fa8aa911817aa
|
| 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/tmpzougtkly",
|
| 32 |
"task": "",
|
| 33 |
"save_load_fn": "bitsandbytes",
|
| 34 |
"save_load_fn_args": {}
|