Upload 8 files
Browse files- added_tokens.json +269 -0
- config.json +130 -0
- generation_config.json +79 -0
- pytorch_model.bin.index.json +0 -0
- special_tokens_map.json +24 -0
- tokenizer.model +3 -0
- tokenizer_config.json +35 -0
added_tokens.json
ADDED
|
@@ -0,0 +1,269 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"</mask>": 32008,
|
| 3 |
+
"</ref>": 32002,
|
| 4 |
+
"</region>": 32004,
|
| 5 |
+
"</sr>": 32006,
|
| 6 |
+
"<SEG>": 32000,
|
| 7 |
+
"<gen_100>": 32108,
|
| 8 |
+
"<gen_101>": 32109,
|
| 9 |
+
"<gen_102>": 32110,
|
| 10 |
+
"<gen_103>": 32111,
|
| 11 |
+
"<gen_104>": 32112,
|
| 12 |
+
"<gen_105>": 32113,
|
| 13 |
+
"<gen_106>": 32114,
|
| 14 |
+
"<gen_107>": 32115,
|
| 15 |
+
"<gen_108>": 32116,
|
| 16 |
+
"<gen_109>": 32117,
|
| 17 |
+
"<gen_10>": 32018,
|
| 18 |
+
"<gen_110>": 32118,
|
| 19 |
+
"<gen_111>": 32119,
|
| 20 |
+
"<gen_112>": 32120,
|
| 21 |
+
"<gen_113>": 32121,
|
| 22 |
+
"<gen_114>": 32122,
|
| 23 |
+
"<gen_115>": 32123,
|
| 24 |
+
"<gen_116>": 32124,
|
| 25 |
+
"<gen_117>": 32125,
|
| 26 |
+
"<gen_118>": 32126,
|
| 27 |
+
"<gen_119>": 32127,
|
| 28 |
+
"<gen_11>": 32019,
|
| 29 |
+
"<gen_120>": 32128,
|
| 30 |
+
"<gen_121>": 32129,
|
| 31 |
+
"<gen_122>": 32130,
|
| 32 |
+
"<gen_123>": 32131,
|
| 33 |
+
"<gen_124>": 32132,
|
| 34 |
+
"<gen_125>": 32133,
|
| 35 |
+
"<gen_126>": 32134,
|
| 36 |
+
"<gen_127>": 32135,
|
| 37 |
+
"<gen_128>": 32136,
|
| 38 |
+
"<gen_129>": 32137,
|
| 39 |
+
"<gen_12>": 32020,
|
| 40 |
+
"<gen_130>": 32138,
|
| 41 |
+
"<gen_131>": 32139,
|
| 42 |
+
"<gen_132>": 32140,
|
| 43 |
+
"<gen_133>": 32141,
|
| 44 |
+
"<gen_134>": 32142,
|
| 45 |
+
"<gen_135>": 32143,
|
| 46 |
+
"<gen_136>": 32144,
|
| 47 |
+
"<gen_137>": 32145,
|
| 48 |
+
"<gen_138>": 32146,
|
| 49 |
+
"<gen_139>": 32147,
|
| 50 |
+
"<gen_13>": 32021,
|
| 51 |
+
"<gen_140>": 32148,
|
| 52 |
+
"<gen_141>": 32149,
|
| 53 |
+
"<gen_142>": 32150,
|
| 54 |
+
"<gen_143>": 32151,
|
| 55 |
+
"<gen_144>": 32152,
|
| 56 |
+
"<gen_145>": 32153,
|
| 57 |
+
"<gen_146>": 32154,
|
| 58 |
+
"<gen_147>": 32155,
|
| 59 |
+
"<gen_148>": 32156,
|
| 60 |
+
"<gen_149>": 32157,
|
| 61 |
+
"<gen_14>": 32022,
|
| 62 |
+
"<gen_150>": 32158,
|
| 63 |
+
"<gen_151>": 32159,
|
| 64 |
+
"<gen_152>": 32160,
|
| 65 |
+
"<gen_153>": 32161,
|
| 66 |
+
"<gen_154>": 32162,
|
| 67 |
+
"<gen_155>": 32163,
|
| 68 |
+
"<gen_156>": 32164,
|
| 69 |
+
"<gen_157>": 32165,
|
| 70 |
+
"<gen_158>": 32166,
|
| 71 |
+
"<gen_159>": 32167,
|
| 72 |
+
"<gen_15>": 32023,
|
| 73 |
+
"<gen_160>": 32168,
|
| 74 |
+
"<gen_161>": 32169,
|
| 75 |
+
"<gen_162>": 32170,
|
| 76 |
+
"<gen_163>": 32171,
|
| 77 |
+
"<gen_164>": 32172,
|
| 78 |
+
"<gen_165>": 32173,
|
| 79 |
+
"<gen_166>": 32174,
|
| 80 |
+
"<gen_167>": 32175,
|
| 81 |
+
"<gen_168>": 32176,
|
| 82 |
+
"<gen_169>": 32177,
|
| 83 |
+
"<gen_16>": 32024,
|
| 84 |
+
"<gen_170>": 32178,
|
| 85 |
+
"<gen_171>": 32179,
|
| 86 |
+
"<gen_172>": 32180,
|
| 87 |
+
"<gen_173>": 32181,
|
| 88 |
+
"<gen_174>": 32182,
|
| 89 |
+
"<gen_175>": 32183,
|
| 90 |
+
"<gen_176>": 32184,
|
| 91 |
+
"<gen_177>": 32185,
|
| 92 |
+
"<gen_178>": 32186,
|
| 93 |
+
"<gen_179>": 32187,
|
| 94 |
+
"<gen_17>": 32025,
|
| 95 |
+
"<gen_180>": 32188,
|
| 96 |
+
"<gen_181>": 32189,
|
| 97 |
+
"<gen_182>": 32190,
|
| 98 |
+
"<gen_183>": 32191,
|
| 99 |
+
"<gen_184>": 32192,
|
| 100 |
+
"<gen_185>": 32193,
|
| 101 |
+
"<gen_186>": 32194,
|
| 102 |
+
"<gen_187>": 32195,
|
| 103 |
+
"<gen_188>": 32196,
|
| 104 |
+
"<gen_189>": 32197,
|
| 105 |
+
"<gen_18>": 32026,
|
| 106 |
+
"<gen_190>": 32198,
|
| 107 |
+
"<gen_191>": 32199,
|
| 108 |
+
"<gen_192>": 32200,
|
| 109 |
+
"<gen_193>": 32201,
|
| 110 |
+
"<gen_194>": 32202,
|
| 111 |
+
"<gen_195>": 32203,
|
| 112 |
+
"<gen_196>": 32204,
|
| 113 |
+
"<gen_197>": 32205,
|
| 114 |
+
"<gen_198>": 32206,
|
| 115 |
+
"<gen_199>": 32207,
|
| 116 |
+
"<gen_19>": 32027,
|
| 117 |
+
"<gen_1>": 32009,
|
| 118 |
+
"<gen_200>": 32208,
|
| 119 |
+
"<gen_201>": 32209,
|
| 120 |
+
"<gen_202>": 32210,
|
| 121 |
+
"<gen_203>": 32211,
|
| 122 |
+
"<gen_204>": 32212,
|
| 123 |
+
"<gen_205>": 32213,
|
| 124 |
+
"<gen_206>": 32214,
|
| 125 |
+
"<gen_207>": 32215,
|
| 126 |
+
"<gen_208>": 32216,
|
| 127 |
+
"<gen_209>": 32217,
|
| 128 |
+
"<gen_20>": 32028,
|
| 129 |
+
"<gen_210>": 32218,
|
| 130 |
+
"<gen_211>": 32219,
|
| 131 |
+
"<gen_212>": 32220,
|
| 132 |
+
"<gen_213>": 32221,
|
| 133 |
+
"<gen_214>": 32222,
|
| 134 |
+
"<gen_215>": 32223,
|
| 135 |
+
"<gen_216>": 32224,
|
| 136 |
+
"<gen_217>": 32225,
|
| 137 |
+
"<gen_218>": 32226,
|
| 138 |
+
"<gen_219>": 32227,
|
| 139 |
+
"<gen_21>": 32029,
|
| 140 |
+
"<gen_220>": 32228,
|
| 141 |
+
"<gen_221>": 32229,
|
| 142 |
+
"<gen_222>": 32230,
|
| 143 |
+
"<gen_223>": 32231,
|
| 144 |
+
"<gen_224>": 32232,
|
| 145 |
+
"<gen_225>": 32233,
|
| 146 |
+
"<gen_226>": 32234,
|
| 147 |
+
"<gen_227>": 32235,
|
| 148 |
+
"<gen_228>": 32236,
|
| 149 |
+
"<gen_229>": 32237,
|
| 150 |
+
"<gen_22>": 32030,
|
| 151 |
+
"<gen_230>": 32238,
|
| 152 |
+
"<gen_231>": 32239,
|
| 153 |
+
"<gen_232>": 32240,
|
| 154 |
+
"<gen_233>": 32241,
|
| 155 |
+
"<gen_234>": 32242,
|
| 156 |
+
"<gen_235>": 32243,
|
| 157 |
+
"<gen_236>": 32244,
|
| 158 |
+
"<gen_237>": 32245,
|
| 159 |
+
"<gen_238>": 32246,
|
| 160 |
+
"<gen_239>": 32247,
|
| 161 |
+
"<gen_23>": 32031,
|
| 162 |
+
"<gen_240>": 32248,
|
| 163 |
+
"<gen_241>": 32249,
|
| 164 |
+
"<gen_242>": 32250,
|
| 165 |
+
"<gen_243>": 32251,
|
| 166 |
+
"<gen_244>": 32252,
|
| 167 |
+
"<gen_245>": 32253,
|
| 168 |
+
"<gen_246>": 32254,
|
| 169 |
+
"<gen_247>": 32255,
|
| 170 |
+
"<gen_248>": 32256,
|
| 171 |
+
"<gen_249>": 32257,
|
| 172 |
+
"<gen_24>": 32032,
|
| 173 |
+
"<gen_250>": 32258,
|
| 174 |
+
"<gen_251>": 32259,
|
| 175 |
+
"<gen_252>": 32260,
|
| 176 |
+
"<gen_253>": 32261,
|
| 177 |
+
"<gen_254>": 32262,
|
| 178 |
+
"<gen_255>": 32263,
|
| 179 |
+
"<gen_256>": 32264,
|
| 180 |
+
"<gen_25>": 32033,
|
| 181 |
+
"<gen_26>": 32034,
|
| 182 |
+
"<gen_27>": 32035,
|
| 183 |
+
"<gen_28>": 32036,
|
| 184 |
+
"<gen_29>": 32037,
|
| 185 |
+
"<gen_2>": 32010,
|
| 186 |
+
"<gen_30>": 32038,
|
| 187 |
+
"<gen_31>": 32039,
|
| 188 |
+
"<gen_32>": 32040,
|
| 189 |
+
"<gen_33>": 32041,
|
| 190 |
+
"<gen_34>": 32042,
|
| 191 |
+
"<gen_35>": 32043,
|
| 192 |
+
"<gen_36>": 32044,
|
| 193 |
+
"<gen_37>": 32045,
|
| 194 |
+
"<gen_38>": 32046,
|
| 195 |
+
"<gen_39>": 32047,
|
| 196 |
+
"<gen_3>": 32011,
|
| 197 |
+
"<gen_40>": 32048,
|
| 198 |
+
"<gen_41>": 32049,
|
| 199 |
+
"<gen_42>": 32050,
|
| 200 |
+
"<gen_43>": 32051,
|
| 201 |
+
"<gen_44>": 32052,
|
| 202 |
+
"<gen_45>": 32053,
|
| 203 |
+
"<gen_46>": 32054,
|
| 204 |
+
"<gen_47>": 32055,
|
| 205 |
+
"<gen_48>": 32056,
|
| 206 |
+
"<gen_49>": 32057,
|
| 207 |
+
"<gen_4>": 32012,
|
| 208 |
+
"<gen_50>": 32058,
|
| 209 |
+
"<gen_51>": 32059,
|
| 210 |
+
"<gen_52>": 32060,
|
| 211 |
+
"<gen_53>": 32061,
|
| 212 |
+
"<gen_54>": 32062,
|
| 213 |
+
"<gen_55>": 32063,
|
| 214 |
+
"<gen_56>": 32064,
|
| 215 |
+
"<gen_57>": 32065,
|
| 216 |
+
"<gen_58>": 32066,
|
| 217 |
+
"<gen_59>": 32067,
|
| 218 |
+
"<gen_5>": 32013,
|
| 219 |
+
"<gen_60>": 32068,
|
| 220 |
+
"<gen_61>": 32069,
|
| 221 |
+
"<gen_62>": 32070,
|
| 222 |
+
"<gen_63>": 32071,
|
| 223 |
+
"<gen_64>": 32072,
|
| 224 |
+
"<gen_65>": 32073,
|
| 225 |
+
"<gen_66>": 32074,
|
| 226 |
+
"<gen_67>": 32075,
|
| 227 |
+
"<gen_68>": 32076,
|
| 228 |
+
"<gen_69>": 32077,
|
| 229 |
+
"<gen_6>": 32014,
|
| 230 |
+
"<gen_70>": 32078,
|
| 231 |
+
"<gen_71>": 32079,
|
| 232 |
+
"<gen_72>": 32080,
|
| 233 |
+
"<gen_73>": 32081,
|
| 234 |
+
"<gen_74>": 32082,
|
| 235 |
+
"<gen_75>": 32083,
|
| 236 |
+
"<gen_76>": 32084,
|
| 237 |
+
"<gen_77>": 32085,
|
| 238 |
+
"<gen_78>": 32086,
|
| 239 |
+
"<gen_79>": 32087,
|
| 240 |
+
"<gen_7>": 32015,
|
| 241 |
+
"<gen_80>": 32088,
|
| 242 |
+
"<gen_81>": 32089,
|
| 243 |
+
"<gen_82>": 32090,
|
| 244 |
+
"<gen_83>": 32091,
|
| 245 |
+
"<gen_84>": 32092,
|
| 246 |
+
"<gen_85>": 32093,
|
| 247 |
+
"<gen_86>": 32094,
|
| 248 |
+
"<gen_87>": 32095,
|
| 249 |
+
"<gen_88>": 32096,
|
| 250 |
+
"<gen_89>": 32097,
|
| 251 |
+
"<gen_8>": 32016,
|
| 252 |
+
"<gen_90>": 32098,
|
| 253 |
+
"<gen_91>": 32099,
|
| 254 |
+
"<gen_92>": 32100,
|
| 255 |
+
"<gen_93>": 32101,
|
| 256 |
+
"<gen_94>": 32102,
|
| 257 |
+
"<gen_95>": 32103,
|
| 258 |
+
"<gen_96>": 32104,
|
| 259 |
+
"<gen_97>": 32105,
|
| 260 |
+
"<gen_98>": 32106,
|
| 261 |
+
"<gen_99>": 32107,
|
| 262 |
+
"<gen_9>": 32017,
|
| 263 |
+
"<im_end>": 32266,
|
| 264 |
+
"<im_start>": 32265,
|
| 265 |
+
"<mask>": 32007,
|
| 266 |
+
"<ref>": 32001,
|
| 267 |
+
"<region>": 32003,
|
| 268 |
+
"<sr>": 32005
|
| 269 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,130 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "/root/paddlejob/workspace/env_run/output/huangxiaoshuang/MedPLIB/runs/MedPLIB-7b-2e/hf",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"MedPLIBForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"bos_token_id": 1,
|
| 7 |
+
"eos_token_id": 2,
|
| 8 |
+
"freeze_mm_mlp_adapter": true,
|
| 9 |
+
"freeze_mm_vision_resampler": false,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 4096,
|
| 12 |
+
"image_aspect_ratio": "square",
|
| 13 |
+
"image_grid_pinpoints": null,
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 11008,
|
| 16 |
+
"max_length": 4096,
|
| 17 |
+
"max_position_embeddings": 4096,
|
| 18 |
+
"max_sample_point": 512,
|
| 19 |
+
"mm_hidden_size": 1024,
|
| 20 |
+
"mm_projector_type": "mlp2x_gelu",
|
| 21 |
+
"mm_resampler_type": null,
|
| 22 |
+
"mm_use_im_patch_token": false,
|
| 23 |
+
"mm_use_im_start_end": true,
|
| 24 |
+
"mm_vision_select_feature": "patch",
|
| 25 |
+
"mm_vision_select_layer": -2,
|
| 26 |
+
"mm_vision_tower": "/public/home/s20213081508/huangxiaoshuang/huggingface/clip-vit-large-patch14-336",
|
| 27 |
+
"model_type": "medplib_moe_llama",
|
| 28 |
+
"moe": {
|
| 29 |
+
"capacity_factor": 2.0,
|
| 30 |
+
"ep_size": 1,
|
| 31 |
+
"eval_capacity_factor": 2,
|
| 32 |
+
"min_capacity": 0,
|
| 33 |
+
"moe_enable": true,
|
| 34 |
+
"moe_layers_idx": [
|
| 35 |
+
0,
|
| 36 |
+
1,
|
| 37 |
+
2,
|
| 38 |
+
3,
|
| 39 |
+
4,
|
| 40 |
+
5,
|
| 41 |
+
6,
|
| 42 |
+
7,
|
| 43 |
+
8,
|
| 44 |
+
9,
|
| 45 |
+
10,
|
| 46 |
+
11,
|
| 47 |
+
12,
|
| 48 |
+
13,
|
| 49 |
+
14,
|
| 50 |
+
15,
|
| 51 |
+
16,
|
| 52 |
+
17,
|
| 53 |
+
18,
|
| 54 |
+
19,
|
| 55 |
+
20,
|
| 56 |
+
21,
|
| 57 |
+
22,
|
| 58 |
+
23,
|
| 59 |
+
24,
|
| 60 |
+
25,
|
| 61 |
+
26,
|
| 62 |
+
27,
|
| 63 |
+
28,
|
| 64 |
+
29,
|
| 65 |
+
30,
|
| 66 |
+
31
|
| 67 |
+
],
|
| 68 |
+
"moe_mode": "dense",
|
| 69 |
+
"num_experts": [
|
| 70 |
+
2,
|
| 71 |
+
2,
|
| 72 |
+
2,
|
| 73 |
+
2,
|
| 74 |
+
2,
|
| 75 |
+
2,
|
| 76 |
+
2,
|
| 77 |
+
2,
|
| 78 |
+
2,
|
| 79 |
+
2,
|
| 80 |
+
2,
|
| 81 |
+
2,
|
| 82 |
+
2,
|
| 83 |
+
2,
|
| 84 |
+
2,
|
| 85 |
+
2,
|
| 86 |
+
2,
|
| 87 |
+
2,
|
| 88 |
+
2,
|
| 89 |
+
2,
|
| 90 |
+
2,
|
| 91 |
+
2,
|
| 92 |
+
2,
|
| 93 |
+
2,
|
| 94 |
+
2,
|
| 95 |
+
2,
|
| 96 |
+
2,
|
| 97 |
+
2,
|
| 98 |
+
2,
|
| 99 |
+
2,
|
| 100 |
+
2,
|
| 101 |
+
2
|
| 102 |
+
],
|
| 103 |
+
"router_aux_loss_coef": 0.0,
|
| 104 |
+
"top_k_experts": 1,
|
| 105 |
+
"use_residual": false
|
| 106 |
+
},
|
| 107 |
+
"num_attention_heads": 32,
|
| 108 |
+
"num_hidden_layers": 32,
|
| 109 |
+
"num_key_value_heads": 32,
|
| 110 |
+
"out_dim": 256,
|
| 111 |
+
"pad_token_id": 0,
|
| 112 |
+
"pretrain_mm_mlp_adapter": null,
|
| 113 |
+
"pretraining_tp": 1,
|
| 114 |
+
"region_fea_adapter": true,
|
| 115 |
+
"region_geo_sampler": false,
|
| 116 |
+
"rms_norm_eps": 1e-05,
|
| 117 |
+
"rope_scaling": null,
|
| 118 |
+
"sampler_pooler_mode": null,
|
| 119 |
+
"tie_word_embeddings": false,
|
| 120 |
+
"torch_dtype": "bfloat16",
|
| 121 |
+
"train_mask_decoder": true,
|
| 122 |
+
"transformers_version": "4.31.0",
|
| 123 |
+
"tune_mm_mlp_adapter": false,
|
| 124 |
+
"tune_mm_vision_resampler": false,
|
| 125 |
+
"unfreeze_mm_vision_tower": false,
|
| 126 |
+
"use_cache": false,
|
| 127 |
+
"use_mm_proj": true,
|
| 128 |
+
"vision_tower": "/public/home/s20213081508/huangxiaoshuang/huggingface/clip-vit-large-patch14-336",
|
| 129 |
+
"vocab_size": 32267
|
| 130 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,79 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_resume": true,
|
| 3 |
+
"batch_size": 4,
|
| 4 |
+
"bce_loss_weight": 1.0,
|
| 5 |
+
"beta1": 0.9,
|
| 6 |
+
"beta2": 0.95,
|
| 7 |
+
"bos_token_id": 1,
|
| 8 |
+
"capacity_factor": 2.0,
|
| 9 |
+
"ce_loss_weight": 2.0,
|
| 10 |
+
"conv_type": "llava_v1",
|
| 11 |
+
"data_path": "/root/paddlejob/workspace/env_run/data/huangxiaoshuang/jsons/seg/0301_CombinedAll_7.5w_train1_chatlm_img_coord2_rand5k_seg_cat_origin.json",
|
| 12 |
+
"dice_loss_weight": 1.0,
|
| 13 |
+
"eos_token_id": 2,
|
| 14 |
+
"ep_size": 1,
|
| 15 |
+
"epochs": 1,
|
| 16 |
+
"eval_capacity_factor": 2,
|
| 17 |
+
"eval_only": true,
|
| 18 |
+
"exclude_val": false,
|
| 19 |
+
"exp_name": "",
|
| 20 |
+
"expert_pretrained_path": "/root/paddlejob/workspace/env_run/output/huangxiaoshuang/BiRD2/runs/medplib-7b-stage3/hf,/root/paddlejob/workspace/env_run/output/huangxiaoshuang/BiRD2/runs/medplib-7b-stage2/hf",
|
| 21 |
+
"explanatory": 0.1,
|
| 22 |
+
"finetune_moe": false,
|
| 23 |
+
"focal_loss_weight": 1.0,
|
| 24 |
+
"grad_accumulation_steps": 8,
|
| 25 |
+
"gradient_checkpointing": true,
|
| 26 |
+
"image_aspect_ratio": "pad",
|
| 27 |
+
"image_folder": "/tmp/v2_mnt/HCG/huangxiaoshuang/SAMed2D_v1",
|
| 28 |
+
"iou_loss_weight": 2.0,
|
| 29 |
+
"is_multimodal": true,
|
| 30 |
+
"local_rank": 0,
|
| 31 |
+
"log_base_dir": "./runs",
|
| 32 |
+
"log_dir": "./runs/",
|
| 33 |
+
"lora_alpha": 16,
|
| 34 |
+
"lora_dropout": 0.05,
|
| 35 |
+
"lora_r": 8,
|
| 36 |
+
"lora_target_modules": "gate_proj,up_proj,down_proj,q_proj,v_proj",
|
| 37 |
+
"lr": 2e-05,
|
| 38 |
+
"max_length": 4096,
|
| 39 |
+
"max_sample_point": 512,
|
| 40 |
+
"min_capacity": 0,
|
| 41 |
+
"model_max_length": 1024,
|
| 42 |
+
"moe_enable": true,
|
| 43 |
+
"moe_layers_idx": null,
|
| 44 |
+
"moe_mode": "dense",
|
| 45 |
+
"no_eval": false,
|
| 46 |
+
"num_classes_per_sample": 3,
|
| 47 |
+
"num_experts": [
|
| 48 |
+
2
|
| 49 |
+
],
|
| 50 |
+
"out_dim": 256,
|
| 51 |
+
"pad_token_id": 0,
|
| 52 |
+
"precision": "bf16",
|
| 53 |
+
"pretrain_mm_mlp_adapter": null,
|
| 54 |
+
"print_freq": 1,
|
| 55 |
+
"region_fea_adapter": true,
|
| 56 |
+
"region_geo_sampler": false,
|
| 57 |
+
"resume": "",
|
| 58 |
+
"router_aux_loss_coef": 0.0,
|
| 59 |
+
"sam_img_size": 256,
|
| 60 |
+
"sampler_pooler_mode": "max",
|
| 61 |
+
"save_path": "/root/paddlejob/workspace/env_run/output/huangxiaoshuang/BiRD2/runs/lisa-7b-bird-0514_1/hf",
|
| 62 |
+
"save_steps": 200,
|
| 63 |
+
"seg_token_idx": 32000,
|
| 64 |
+
"sft_modules": "wg,lm_head,embed_tokens,mask_decoder,text_hidden_fcs,region_fea_adapter",
|
| 65 |
+
"start_epoch": 0,
|
| 66 |
+
"top_k_experts": 1,
|
| 67 |
+
"train_mask_decoder": true,
|
| 68 |
+
"transformers_version": "4.31.0",
|
| 69 |
+
"use_mm_start_end": true,
|
| 70 |
+
"use_residual": false,
|
| 71 |
+
"val_batch_size": 1,
|
| 72 |
+
"val_data_path": "/root/paddlejob/workspace/env_run/data/huangxiaoshuang/jsons/seg/0301_CombinedAll_7.5w_train1_chatlm_img_coord2_rand5k_seg.json",
|
| 73 |
+
"version": "/root/paddlejob/workspace/env_run/output/huangxiaoshuang/BiRD2/runs/lisa-7b-bird-0504/hf",
|
| 74 |
+
"vis_save_path": "./vis_output",
|
| 75 |
+
"vision_pretrained": "/root/paddlejob/workspace/env_run/output/huangxiaoshuang/huggingface_models/sam-med2d_b.pth",
|
| 76 |
+
"vision_tower": "/root/paddlejob/workspace/env_run/output/huangxiaoshuang/huggingface_models/clip-vit-large-patch14-336",
|
| 77 |
+
"weight": "/root/paddlejob/workspace/env_run/output/huangxiaoshuang/BiRD2/runs/lisa-7b-bird-0514_1/pytorch_model.bin",
|
| 78 |
+
"workers": 8
|
| 79 |
+
}
|
pytorch_model.bin.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "<unk>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<unk>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"bos_token": {
|
| 5 |
+
"__type": "AddedToken",
|
| 6 |
+
"content": "<s>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"clean_up_tokenization_spaces": false,
|
| 13 |
+
"eos_token": {
|
| 14 |
+
"__type": "AddedToken",
|
| 15 |
+
"content": "</s>",
|
| 16 |
+
"lstrip": false,
|
| 17 |
+
"normalized": false,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"single_word": false
|
| 20 |
+
},
|
| 21 |
+
"legacy": true,
|
| 22 |
+
"model_max_length": 1024,
|
| 23 |
+
"pad_token": null,
|
| 24 |
+
"padding_side": "right",
|
| 25 |
+
"sp_model_kwargs": {},
|
| 26 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 27 |
+
"unk_token": {
|
| 28 |
+
"__type": "AddedToken",
|
| 29 |
+
"content": "<unk>",
|
| 30 |
+
"lstrip": false,
|
| 31 |
+
"normalized": false,
|
| 32 |
+
"rstrip": false,
|
| 33 |
+
"single_word": false
|
| 34 |
+
}
|
| 35 |
+
}
|