Upload folder using huggingface_hub
Browse files- added_tokens.json +1048 -0
- config.json +112 -0
- configuration_aimv2.py +62 -0
- generation_config.json +7 -0
- merges.txt +0 -0
- model-00001-of-00002.safetensors +3 -0
- model-00002-of-00002.safetensors +3 -0
- model.safetensors.index.json +524 -0
- modeling_aimv2.py +287 -0
- modeling_mono.py +370 -0
- preprocessor_config.json +32 -0
- processing_mono.py +259 -0
- processor_config.json +6 -0
- special_tokens_map.json +0 -0
- tokenizer.json +0 -0
- tokenizer_config.json +207 -0
- vocab.json +0 -0
added_tokens.json
ADDED
|
@@ -0,0 +1,1048 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"</cap>": 152670,
|
| 3 |
+
"</dcap>": 152674,
|
| 4 |
+
"</grounding>": 152676,
|
| 5 |
+
"</ncap>": 152672,
|
| 6 |
+
"</ocr>": 151668,
|
| 7 |
+
"</od>": 151666,
|
| 8 |
+
"</poly>": 152687,
|
| 9 |
+
"</proposal>": 152685,
|
| 10 |
+
"</region_cap>": 152681,
|
| 11 |
+
"</region_to_desciption>": 152683,
|
| 12 |
+
"</seg>": 152678,
|
| 13 |
+
"</tool_call>": 151658,
|
| 14 |
+
"<and>": 152688,
|
| 15 |
+
"<cap>": 152669,
|
| 16 |
+
"<dcap>": 152673,
|
| 17 |
+
"<grounding>": 152675,
|
| 18 |
+
"<loc_0>": 151669,
|
| 19 |
+
"<loc_100>": 151769,
|
| 20 |
+
"<loc_101>": 151770,
|
| 21 |
+
"<loc_102>": 151771,
|
| 22 |
+
"<loc_103>": 151772,
|
| 23 |
+
"<loc_104>": 151773,
|
| 24 |
+
"<loc_105>": 151774,
|
| 25 |
+
"<loc_106>": 151775,
|
| 26 |
+
"<loc_107>": 151776,
|
| 27 |
+
"<loc_108>": 151777,
|
| 28 |
+
"<loc_109>": 151778,
|
| 29 |
+
"<loc_10>": 151679,
|
| 30 |
+
"<loc_110>": 151779,
|
| 31 |
+
"<loc_111>": 151780,
|
| 32 |
+
"<loc_112>": 151781,
|
| 33 |
+
"<loc_113>": 151782,
|
| 34 |
+
"<loc_114>": 151783,
|
| 35 |
+
"<loc_115>": 151784,
|
| 36 |
+
"<loc_116>": 151785,
|
| 37 |
+
"<loc_117>": 151786,
|
| 38 |
+
"<loc_118>": 151787,
|
| 39 |
+
"<loc_119>": 151788,
|
| 40 |
+
"<loc_11>": 151680,
|
| 41 |
+
"<loc_120>": 151789,
|
| 42 |
+
"<loc_121>": 151790,
|
| 43 |
+
"<loc_122>": 151791,
|
| 44 |
+
"<loc_123>": 151792,
|
| 45 |
+
"<loc_124>": 151793,
|
| 46 |
+
"<loc_125>": 151794,
|
| 47 |
+
"<loc_126>": 151795,
|
| 48 |
+
"<loc_127>": 151796,
|
| 49 |
+
"<loc_128>": 151797,
|
| 50 |
+
"<loc_129>": 151798,
|
| 51 |
+
"<loc_12>": 151681,
|
| 52 |
+
"<loc_130>": 151799,
|
| 53 |
+
"<loc_131>": 151800,
|
| 54 |
+
"<loc_132>": 151801,
|
| 55 |
+
"<loc_133>": 151802,
|
| 56 |
+
"<loc_134>": 151803,
|
| 57 |
+
"<loc_135>": 151804,
|
| 58 |
+
"<loc_136>": 151805,
|
| 59 |
+
"<loc_137>": 151806,
|
| 60 |
+
"<loc_138>": 151807,
|
| 61 |
+
"<loc_139>": 151808,
|
| 62 |
+
"<loc_13>": 151682,
|
| 63 |
+
"<loc_140>": 151809,
|
| 64 |
+
"<loc_141>": 151810,
|
| 65 |
+
"<loc_142>": 151811,
|
| 66 |
+
"<loc_143>": 151812,
|
| 67 |
+
"<loc_144>": 151813,
|
| 68 |
+
"<loc_145>": 151814,
|
| 69 |
+
"<loc_146>": 151815,
|
| 70 |
+
"<loc_147>": 151816,
|
| 71 |
+
"<loc_148>": 151817,
|
| 72 |
+
"<loc_149>": 151818,
|
| 73 |
+
"<loc_14>": 151683,
|
| 74 |
+
"<loc_150>": 151819,
|
| 75 |
+
"<loc_151>": 151820,
|
| 76 |
+
"<loc_152>": 151821,
|
| 77 |
+
"<loc_153>": 151822,
|
| 78 |
+
"<loc_154>": 151823,
|
| 79 |
+
"<loc_155>": 151824,
|
| 80 |
+
"<loc_156>": 151825,
|
| 81 |
+
"<loc_157>": 151826,
|
| 82 |
+
"<loc_158>": 151827,
|
| 83 |
+
"<loc_159>": 151828,
|
| 84 |
+
"<loc_15>": 151684,
|
| 85 |
+
"<loc_160>": 151829,
|
| 86 |
+
"<loc_161>": 151830,
|
| 87 |
+
"<loc_162>": 151831,
|
| 88 |
+
"<loc_163>": 151832,
|
| 89 |
+
"<loc_164>": 151833,
|
| 90 |
+
"<loc_165>": 151834,
|
| 91 |
+
"<loc_166>": 151835,
|
| 92 |
+
"<loc_167>": 151836,
|
| 93 |
+
"<loc_168>": 151837,
|
| 94 |
+
"<loc_169>": 151838,
|
| 95 |
+
"<loc_16>": 151685,
|
| 96 |
+
"<loc_170>": 151839,
|
| 97 |
+
"<loc_171>": 151840,
|
| 98 |
+
"<loc_172>": 151841,
|
| 99 |
+
"<loc_173>": 151842,
|
| 100 |
+
"<loc_174>": 151843,
|
| 101 |
+
"<loc_175>": 151844,
|
| 102 |
+
"<loc_176>": 151845,
|
| 103 |
+
"<loc_177>": 151846,
|
| 104 |
+
"<loc_178>": 151847,
|
| 105 |
+
"<loc_179>": 151848,
|
| 106 |
+
"<loc_17>": 151686,
|
| 107 |
+
"<loc_180>": 151849,
|
| 108 |
+
"<loc_181>": 151850,
|
| 109 |
+
"<loc_182>": 151851,
|
| 110 |
+
"<loc_183>": 151852,
|
| 111 |
+
"<loc_184>": 151853,
|
| 112 |
+
"<loc_185>": 151854,
|
| 113 |
+
"<loc_186>": 151855,
|
| 114 |
+
"<loc_187>": 151856,
|
| 115 |
+
"<loc_188>": 151857,
|
| 116 |
+
"<loc_189>": 151858,
|
| 117 |
+
"<loc_18>": 151687,
|
| 118 |
+
"<loc_190>": 151859,
|
| 119 |
+
"<loc_191>": 151860,
|
| 120 |
+
"<loc_192>": 151861,
|
| 121 |
+
"<loc_193>": 151862,
|
| 122 |
+
"<loc_194>": 151863,
|
| 123 |
+
"<loc_195>": 151864,
|
| 124 |
+
"<loc_196>": 151865,
|
| 125 |
+
"<loc_197>": 151866,
|
| 126 |
+
"<loc_198>": 151867,
|
| 127 |
+
"<loc_199>": 151868,
|
| 128 |
+
"<loc_19>": 151688,
|
| 129 |
+
"<loc_1>": 151670,
|
| 130 |
+
"<loc_200>": 151869,
|
| 131 |
+
"<loc_201>": 151870,
|
| 132 |
+
"<loc_202>": 151871,
|
| 133 |
+
"<loc_203>": 151872,
|
| 134 |
+
"<loc_204>": 151873,
|
| 135 |
+
"<loc_205>": 151874,
|
| 136 |
+
"<loc_206>": 151875,
|
| 137 |
+
"<loc_207>": 151876,
|
| 138 |
+
"<loc_208>": 151877,
|
| 139 |
+
"<loc_209>": 151878,
|
| 140 |
+
"<loc_20>": 151689,
|
| 141 |
+
"<loc_210>": 151879,
|
| 142 |
+
"<loc_211>": 151880,
|
| 143 |
+
"<loc_212>": 151881,
|
| 144 |
+
"<loc_213>": 151882,
|
| 145 |
+
"<loc_214>": 151883,
|
| 146 |
+
"<loc_215>": 151884,
|
| 147 |
+
"<loc_216>": 151885,
|
| 148 |
+
"<loc_217>": 151886,
|
| 149 |
+
"<loc_218>": 151887,
|
| 150 |
+
"<loc_219>": 151888,
|
| 151 |
+
"<loc_21>": 151690,
|
| 152 |
+
"<loc_220>": 151889,
|
| 153 |
+
"<loc_221>": 151890,
|
| 154 |
+
"<loc_222>": 151891,
|
| 155 |
+
"<loc_223>": 151892,
|
| 156 |
+
"<loc_224>": 151893,
|
| 157 |
+
"<loc_225>": 151894,
|
| 158 |
+
"<loc_226>": 151895,
|
| 159 |
+
"<loc_227>": 151896,
|
| 160 |
+
"<loc_228>": 151897,
|
| 161 |
+
"<loc_229>": 151898,
|
| 162 |
+
"<loc_22>": 151691,
|
| 163 |
+
"<loc_230>": 151899,
|
| 164 |
+
"<loc_231>": 151900,
|
| 165 |
+
"<loc_232>": 151901,
|
| 166 |
+
"<loc_233>": 151902,
|
| 167 |
+
"<loc_234>": 151903,
|
| 168 |
+
"<loc_235>": 151904,
|
| 169 |
+
"<loc_236>": 151905,
|
| 170 |
+
"<loc_237>": 151906,
|
| 171 |
+
"<loc_238>": 151907,
|
| 172 |
+
"<loc_239>": 151908,
|
| 173 |
+
"<loc_23>": 151692,
|
| 174 |
+
"<loc_240>": 151909,
|
| 175 |
+
"<loc_241>": 151910,
|
| 176 |
+
"<loc_242>": 151911,
|
| 177 |
+
"<loc_243>": 151912,
|
| 178 |
+
"<loc_244>": 151913,
|
| 179 |
+
"<loc_245>": 151914,
|
| 180 |
+
"<loc_246>": 151915,
|
| 181 |
+
"<loc_247>": 151916,
|
| 182 |
+
"<loc_248>": 151917,
|
| 183 |
+
"<loc_249>": 151918,
|
| 184 |
+
"<loc_24>": 151693,
|
| 185 |
+
"<loc_250>": 151919,
|
| 186 |
+
"<loc_251>": 151920,
|
| 187 |
+
"<loc_252>": 151921,
|
| 188 |
+
"<loc_253>": 151922,
|
| 189 |
+
"<loc_254>": 151923,
|
| 190 |
+
"<loc_255>": 151924,
|
| 191 |
+
"<loc_256>": 151925,
|
| 192 |
+
"<loc_257>": 151926,
|
| 193 |
+
"<loc_258>": 151927,
|
| 194 |
+
"<loc_259>": 151928,
|
| 195 |
+
"<loc_25>": 151694,
|
| 196 |
+
"<loc_260>": 151929,
|
| 197 |
+
"<loc_261>": 151930,
|
| 198 |
+
"<loc_262>": 151931,
|
| 199 |
+
"<loc_263>": 151932,
|
| 200 |
+
"<loc_264>": 151933,
|
| 201 |
+
"<loc_265>": 151934,
|
| 202 |
+
"<loc_266>": 151935,
|
| 203 |
+
"<loc_267>": 151936,
|
| 204 |
+
"<loc_268>": 151937,
|
| 205 |
+
"<loc_269>": 151938,
|
| 206 |
+
"<loc_26>": 151695,
|
| 207 |
+
"<loc_270>": 151939,
|
| 208 |
+
"<loc_271>": 151940,
|
| 209 |
+
"<loc_272>": 151941,
|
| 210 |
+
"<loc_273>": 151942,
|
| 211 |
+
"<loc_274>": 151943,
|
| 212 |
+
"<loc_275>": 151944,
|
| 213 |
+
"<loc_276>": 151945,
|
| 214 |
+
"<loc_277>": 151946,
|
| 215 |
+
"<loc_278>": 151947,
|
| 216 |
+
"<loc_279>": 151948,
|
| 217 |
+
"<loc_27>": 151696,
|
| 218 |
+
"<loc_280>": 151949,
|
| 219 |
+
"<loc_281>": 151950,
|
| 220 |
+
"<loc_282>": 151951,
|
| 221 |
+
"<loc_283>": 151952,
|
| 222 |
+
"<loc_284>": 151953,
|
| 223 |
+
"<loc_285>": 151954,
|
| 224 |
+
"<loc_286>": 151955,
|
| 225 |
+
"<loc_287>": 151956,
|
| 226 |
+
"<loc_288>": 151957,
|
| 227 |
+
"<loc_289>": 151958,
|
| 228 |
+
"<loc_28>": 151697,
|
| 229 |
+
"<loc_290>": 151959,
|
| 230 |
+
"<loc_291>": 151960,
|
| 231 |
+
"<loc_292>": 151961,
|
| 232 |
+
"<loc_293>": 151962,
|
| 233 |
+
"<loc_294>": 151963,
|
| 234 |
+
"<loc_295>": 151964,
|
| 235 |
+
"<loc_296>": 151965,
|
| 236 |
+
"<loc_297>": 151966,
|
| 237 |
+
"<loc_298>": 151967,
|
| 238 |
+
"<loc_299>": 151968,
|
| 239 |
+
"<loc_29>": 151698,
|
| 240 |
+
"<loc_2>": 151671,
|
| 241 |
+
"<loc_300>": 151969,
|
| 242 |
+
"<loc_301>": 151970,
|
| 243 |
+
"<loc_302>": 151971,
|
| 244 |
+
"<loc_303>": 151972,
|
| 245 |
+
"<loc_304>": 151973,
|
| 246 |
+
"<loc_305>": 151974,
|
| 247 |
+
"<loc_306>": 151975,
|
| 248 |
+
"<loc_307>": 151976,
|
| 249 |
+
"<loc_308>": 151977,
|
| 250 |
+
"<loc_309>": 151978,
|
| 251 |
+
"<loc_30>": 151699,
|
| 252 |
+
"<loc_310>": 151979,
|
| 253 |
+
"<loc_311>": 151980,
|
| 254 |
+
"<loc_312>": 151981,
|
| 255 |
+
"<loc_313>": 151982,
|
| 256 |
+
"<loc_314>": 151983,
|
| 257 |
+
"<loc_315>": 151984,
|
| 258 |
+
"<loc_316>": 151985,
|
| 259 |
+
"<loc_317>": 151986,
|
| 260 |
+
"<loc_318>": 151987,
|
| 261 |
+
"<loc_319>": 151988,
|
| 262 |
+
"<loc_31>": 151700,
|
| 263 |
+
"<loc_320>": 151989,
|
| 264 |
+
"<loc_321>": 151990,
|
| 265 |
+
"<loc_322>": 151991,
|
| 266 |
+
"<loc_323>": 151992,
|
| 267 |
+
"<loc_324>": 151993,
|
| 268 |
+
"<loc_325>": 151994,
|
| 269 |
+
"<loc_326>": 151995,
|
| 270 |
+
"<loc_327>": 151996,
|
| 271 |
+
"<loc_328>": 151997,
|
| 272 |
+
"<loc_329>": 151998,
|
| 273 |
+
"<loc_32>": 151701,
|
| 274 |
+
"<loc_330>": 151999,
|
| 275 |
+
"<loc_331>": 152000,
|
| 276 |
+
"<loc_332>": 152001,
|
| 277 |
+
"<loc_333>": 152002,
|
| 278 |
+
"<loc_334>": 152003,
|
| 279 |
+
"<loc_335>": 152004,
|
| 280 |
+
"<loc_336>": 152005,
|
| 281 |
+
"<loc_337>": 152006,
|
| 282 |
+
"<loc_338>": 152007,
|
| 283 |
+
"<loc_339>": 152008,
|
| 284 |
+
"<loc_33>": 151702,
|
| 285 |
+
"<loc_340>": 152009,
|
| 286 |
+
"<loc_341>": 152010,
|
| 287 |
+
"<loc_342>": 152011,
|
| 288 |
+
"<loc_343>": 152012,
|
| 289 |
+
"<loc_344>": 152013,
|
| 290 |
+
"<loc_345>": 152014,
|
| 291 |
+
"<loc_346>": 152015,
|
| 292 |
+
"<loc_347>": 152016,
|
| 293 |
+
"<loc_348>": 152017,
|
| 294 |
+
"<loc_349>": 152018,
|
| 295 |
+
"<loc_34>": 151703,
|
| 296 |
+
"<loc_350>": 152019,
|
| 297 |
+
"<loc_351>": 152020,
|
| 298 |
+
"<loc_352>": 152021,
|
| 299 |
+
"<loc_353>": 152022,
|
| 300 |
+
"<loc_354>": 152023,
|
| 301 |
+
"<loc_355>": 152024,
|
| 302 |
+
"<loc_356>": 152025,
|
| 303 |
+
"<loc_357>": 152026,
|
| 304 |
+
"<loc_358>": 152027,
|
| 305 |
+
"<loc_359>": 152028,
|
| 306 |
+
"<loc_35>": 151704,
|
| 307 |
+
"<loc_360>": 152029,
|
| 308 |
+
"<loc_361>": 152030,
|
| 309 |
+
"<loc_362>": 152031,
|
| 310 |
+
"<loc_363>": 152032,
|
| 311 |
+
"<loc_364>": 152033,
|
| 312 |
+
"<loc_365>": 152034,
|
| 313 |
+
"<loc_366>": 152035,
|
| 314 |
+
"<loc_367>": 152036,
|
| 315 |
+
"<loc_368>": 152037,
|
| 316 |
+
"<loc_369>": 152038,
|
| 317 |
+
"<loc_36>": 151705,
|
| 318 |
+
"<loc_370>": 152039,
|
| 319 |
+
"<loc_371>": 152040,
|
| 320 |
+
"<loc_372>": 152041,
|
| 321 |
+
"<loc_373>": 152042,
|
| 322 |
+
"<loc_374>": 152043,
|
| 323 |
+
"<loc_375>": 152044,
|
| 324 |
+
"<loc_376>": 152045,
|
| 325 |
+
"<loc_377>": 152046,
|
| 326 |
+
"<loc_378>": 152047,
|
| 327 |
+
"<loc_379>": 152048,
|
| 328 |
+
"<loc_37>": 151706,
|
| 329 |
+
"<loc_380>": 152049,
|
| 330 |
+
"<loc_381>": 152050,
|
| 331 |
+
"<loc_382>": 152051,
|
| 332 |
+
"<loc_383>": 152052,
|
| 333 |
+
"<loc_384>": 152053,
|
| 334 |
+
"<loc_385>": 152054,
|
| 335 |
+
"<loc_386>": 152055,
|
| 336 |
+
"<loc_387>": 152056,
|
| 337 |
+
"<loc_388>": 152057,
|
| 338 |
+
"<loc_389>": 152058,
|
| 339 |
+
"<loc_38>": 151707,
|
| 340 |
+
"<loc_390>": 152059,
|
| 341 |
+
"<loc_391>": 152060,
|
| 342 |
+
"<loc_392>": 152061,
|
| 343 |
+
"<loc_393>": 152062,
|
| 344 |
+
"<loc_394>": 152063,
|
| 345 |
+
"<loc_395>": 152064,
|
| 346 |
+
"<loc_396>": 152065,
|
| 347 |
+
"<loc_397>": 152066,
|
| 348 |
+
"<loc_398>": 152067,
|
| 349 |
+
"<loc_399>": 152068,
|
| 350 |
+
"<loc_39>": 151708,
|
| 351 |
+
"<loc_3>": 151672,
|
| 352 |
+
"<loc_400>": 152069,
|
| 353 |
+
"<loc_401>": 152070,
|
| 354 |
+
"<loc_402>": 152071,
|
| 355 |
+
"<loc_403>": 152072,
|
| 356 |
+
"<loc_404>": 152073,
|
| 357 |
+
"<loc_405>": 152074,
|
| 358 |
+
"<loc_406>": 152075,
|
| 359 |
+
"<loc_407>": 152076,
|
| 360 |
+
"<loc_408>": 152077,
|
| 361 |
+
"<loc_409>": 152078,
|
| 362 |
+
"<loc_40>": 151709,
|
| 363 |
+
"<loc_410>": 152079,
|
| 364 |
+
"<loc_411>": 152080,
|
| 365 |
+
"<loc_412>": 152081,
|
| 366 |
+
"<loc_413>": 152082,
|
| 367 |
+
"<loc_414>": 152083,
|
| 368 |
+
"<loc_415>": 152084,
|
| 369 |
+
"<loc_416>": 152085,
|
| 370 |
+
"<loc_417>": 152086,
|
| 371 |
+
"<loc_418>": 152087,
|
| 372 |
+
"<loc_419>": 152088,
|
| 373 |
+
"<loc_41>": 151710,
|
| 374 |
+
"<loc_420>": 152089,
|
| 375 |
+
"<loc_421>": 152090,
|
| 376 |
+
"<loc_422>": 152091,
|
| 377 |
+
"<loc_423>": 152092,
|
| 378 |
+
"<loc_424>": 152093,
|
| 379 |
+
"<loc_425>": 152094,
|
| 380 |
+
"<loc_426>": 152095,
|
| 381 |
+
"<loc_427>": 152096,
|
| 382 |
+
"<loc_428>": 152097,
|
| 383 |
+
"<loc_429>": 152098,
|
| 384 |
+
"<loc_42>": 151711,
|
| 385 |
+
"<loc_430>": 152099,
|
| 386 |
+
"<loc_431>": 152100,
|
| 387 |
+
"<loc_432>": 152101,
|
| 388 |
+
"<loc_433>": 152102,
|
| 389 |
+
"<loc_434>": 152103,
|
| 390 |
+
"<loc_435>": 152104,
|
| 391 |
+
"<loc_436>": 152105,
|
| 392 |
+
"<loc_437>": 152106,
|
| 393 |
+
"<loc_438>": 152107,
|
| 394 |
+
"<loc_439>": 152108,
|
| 395 |
+
"<loc_43>": 151712,
|
| 396 |
+
"<loc_440>": 152109,
|
| 397 |
+
"<loc_441>": 152110,
|
| 398 |
+
"<loc_442>": 152111,
|
| 399 |
+
"<loc_443>": 152112,
|
| 400 |
+
"<loc_444>": 152113,
|
| 401 |
+
"<loc_445>": 152114,
|
| 402 |
+
"<loc_446>": 152115,
|
| 403 |
+
"<loc_447>": 152116,
|
| 404 |
+
"<loc_448>": 152117,
|
| 405 |
+
"<loc_449>": 152118,
|
| 406 |
+
"<loc_44>": 151713,
|
| 407 |
+
"<loc_450>": 152119,
|
| 408 |
+
"<loc_451>": 152120,
|
| 409 |
+
"<loc_452>": 152121,
|
| 410 |
+
"<loc_453>": 152122,
|
| 411 |
+
"<loc_454>": 152123,
|
| 412 |
+
"<loc_455>": 152124,
|
| 413 |
+
"<loc_456>": 152125,
|
| 414 |
+
"<loc_457>": 152126,
|
| 415 |
+
"<loc_458>": 152127,
|
| 416 |
+
"<loc_459>": 152128,
|
| 417 |
+
"<loc_45>": 151714,
|
| 418 |
+
"<loc_460>": 152129,
|
| 419 |
+
"<loc_461>": 152130,
|
| 420 |
+
"<loc_462>": 152131,
|
| 421 |
+
"<loc_463>": 152132,
|
| 422 |
+
"<loc_464>": 152133,
|
| 423 |
+
"<loc_465>": 152134,
|
| 424 |
+
"<loc_466>": 152135,
|
| 425 |
+
"<loc_467>": 152136,
|
| 426 |
+
"<loc_468>": 152137,
|
| 427 |
+
"<loc_469>": 152138,
|
| 428 |
+
"<loc_46>": 151715,
|
| 429 |
+
"<loc_470>": 152139,
|
| 430 |
+
"<loc_471>": 152140,
|
| 431 |
+
"<loc_472>": 152141,
|
| 432 |
+
"<loc_473>": 152142,
|
| 433 |
+
"<loc_474>": 152143,
|
| 434 |
+
"<loc_475>": 152144,
|
| 435 |
+
"<loc_476>": 152145,
|
| 436 |
+
"<loc_477>": 152146,
|
| 437 |
+
"<loc_478>": 152147,
|
| 438 |
+
"<loc_479>": 152148,
|
| 439 |
+
"<loc_47>": 151716,
|
| 440 |
+
"<loc_480>": 152149,
|
| 441 |
+
"<loc_481>": 152150,
|
| 442 |
+
"<loc_482>": 152151,
|
| 443 |
+
"<loc_483>": 152152,
|
| 444 |
+
"<loc_484>": 152153,
|
| 445 |
+
"<loc_485>": 152154,
|
| 446 |
+
"<loc_486>": 152155,
|
| 447 |
+
"<loc_487>": 152156,
|
| 448 |
+
"<loc_488>": 152157,
|
| 449 |
+
"<loc_489>": 152158,
|
| 450 |
+
"<loc_48>": 151717,
|
| 451 |
+
"<loc_490>": 152159,
|
| 452 |
+
"<loc_491>": 152160,
|
| 453 |
+
"<loc_492>": 152161,
|
| 454 |
+
"<loc_493>": 152162,
|
| 455 |
+
"<loc_494>": 152163,
|
| 456 |
+
"<loc_495>": 152164,
|
| 457 |
+
"<loc_496>": 152165,
|
| 458 |
+
"<loc_497>": 152166,
|
| 459 |
+
"<loc_498>": 152167,
|
| 460 |
+
"<loc_499>": 152168,
|
| 461 |
+
"<loc_49>": 151718,
|
| 462 |
+
"<loc_4>": 151673,
|
| 463 |
+
"<loc_500>": 152169,
|
| 464 |
+
"<loc_501>": 152170,
|
| 465 |
+
"<loc_502>": 152171,
|
| 466 |
+
"<loc_503>": 152172,
|
| 467 |
+
"<loc_504>": 152173,
|
| 468 |
+
"<loc_505>": 152174,
|
| 469 |
+
"<loc_506>": 152175,
|
| 470 |
+
"<loc_507>": 152176,
|
| 471 |
+
"<loc_508>": 152177,
|
| 472 |
+
"<loc_509>": 152178,
|
| 473 |
+
"<loc_50>": 151719,
|
| 474 |
+
"<loc_510>": 152179,
|
| 475 |
+
"<loc_511>": 152180,
|
| 476 |
+
"<loc_512>": 152181,
|
| 477 |
+
"<loc_513>": 152182,
|
| 478 |
+
"<loc_514>": 152183,
|
| 479 |
+
"<loc_515>": 152184,
|
| 480 |
+
"<loc_516>": 152185,
|
| 481 |
+
"<loc_517>": 152186,
|
| 482 |
+
"<loc_518>": 152187,
|
| 483 |
+
"<loc_519>": 152188,
|
| 484 |
+
"<loc_51>": 151720,
|
| 485 |
+
"<loc_520>": 152189,
|
| 486 |
+
"<loc_521>": 152190,
|
| 487 |
+
"<loc_522>": 152191,
|
| 488 |
+
"<loc_523>": 152192,
|
| 489 |
+
"<loc_524>": 152193,
|
| 490 |
+
"<loc_525>": 152194,
|
| 491 |
+
"<loc_526>": 152195,
|
| 492 |
+
"<loc_527>": 152196,
|
| 493 |
+
"<loc_528>": 152197,
|
| 494 |
+
"<loc_529>": 152198,
|
| 495 |
+
"<loc_52>": 151721,
|
| 496 |
+
"<loc_530>": 152199,
|
| 497 |
+
"<loc_531>": 152200,
|
| 498 |
+
"<loc_532>": 152201,
|
| 499 |
+
"<loc_533>": 152202,
|
| 500 |
+
"<loc_534>": 152203,
|
| 501 |
+
"<loc_535>": 152204,
|
| 502 |
+
"<loc_536>": 152205,
|
| 503 |
+
"<loc_537>": 152206,
|
| 504 |
+
"<loc_538>": 152207,
|
| 505 |
+
"<loc_539>": 152208,
|
| 506 |
+
"<loc_53>": 151722,
|
| 507 |
+
"<loc_540>": 152209,
|
| 508 |
+
"<loc_541>": 152210,
|
| 509 |
+
"<loc_542>": 152211,
|
| 510 |
+
"<loc_543>": 152212,
|
| 511 |
+
"<loc_544>": 152213,
|
| 512 |
+
"<loc_545>": 152214,
|
| 513 |
+
"<loc_546>": 152215,
|
| 514 |
+
"<loc_547>": 152216,
|
| 515 |
+
"<loc_548>": 152217,
|
| 516 |
+
"<loc_549>": 152218,
|
| 517 |
+
"<loc_54>": 151723,
|
| 518 |
+
"<loc_550>": 152219,
|
| 519 |
+
"<loc_551>": 152220,
|
| 520 |
+
"<loc_552>": 152221,
|
| 521 |
+
"<loc_553>": 152222,
|
| 522 |
+
"<loc_554>": 152223,
|
| 523 |
+
"<loc_555>": 152224,
|
| 524 |
+
"<loc_556>": 152225,
|
| 525 |
+
"<loc_557>": 152226,
|
| 526 |
+
"<loc_558>": 152227,
|
| 527 |
+
"<loc_559>": 152228,
|
| 528 |
+
"<loc_55>": 151724,
|
| 529 |
+
"<loc_560>": 152229,
|
| 530 |
+
"<loc_561>": 152230,
|
| 531 |
+
"<loc_562>": 152231,
|
| 532 |
+
"<loc_563>": 152232,
|
| 533 |
+
"<loc_564>": 152233,
|
| 534 |
+
"<loc_565>": 152234,
|
| 535 |
+
"<loc_566>": 152235,
|
| 536 |
+
"<loc_567>": 152236,
|
| 537 |
+
"<loc_568>": 152237,
|
| 538 |
+
"<loc_569>": 152238,
|
| 539 |
+
"<loc_56>": 151725,
|
| 540 |
+
"<loc_570>": 152239,
|
| 541 |
+
"<loc_571>": 152240,
|
| 542 |
+
"<loc_572>": 152241,
|
| 543 |
+
"<loc_573>": 152242,
|
| 544 |
+
"<loc_574>": 152243,
|
| 545 |
+
"<loc_575>": 152244,
|
| 546 |
+
"<loc_576>": 152245,
|
| 547 |
+
"<loc_577>": 152246,
|
| 548 |
+
"<loc_578>": 152247,
|
| 549 |
+
"<loc_579>": 152248,
|
| 550 |
+
"<loc_57>": 151726,
|
| 551 |
+
"<loc_580>": 152249,
|
| 552 |
+
"<loc_581>": 152250,
|
| 553 |
+
"<loc_582>": 152251,
|
| 554 |
+
"<loc_583>": 152252,
|
| 555 |
+
"<loc_584>": 152253,
|
| 556 |
+
"<loc_585>": 152254,
|
| 557 |
+
"<loc_586>": 152255,
|
| 558 |
+
"<loc_587>": 152256,
|
| 559 |
+
"<loc_588>": 152257,
|
| 560 |
+
"<loc_589>": 152258,
|
| 561 |
+
"<loc_58>": 151727,
|
| 562 |
+
"<loc_590>": 152259,
|
| 563 |
+
"<loc_591>": 152260,
|
| 564 |
+
"<loc_592>": 152261,
|
| 565 |
+
"<loc_593>": 152262,
|
| 566 |
+
"<loc_594>": 152263,
|
| 567 |
+
"<loc_595>": 152264,
|
| 568 |
+
"<loc_596>": 152265,
|
| 569 |
+
"<loc_597>": 152266,
|
| 570 |
+
"<loc_598>": 152267,
|
| 571 |
+
"<loc_599>": 152268,
|
| 572 |
+
"<loc_59>": 151728,
|
| 573 |
+
"<loc_5>": 151674,
|
| 574 |
+
"<loc_600>": 152269,
|
| 575 |
+
"<loc_601>": 152270,
|
| 576 |
+
"<loc_602>": 152271,
|
| 577 |
+
"<loc_603>": 152272,
|
| 578 |
+
"<loc_604>": 152273,
|
| 579 |
+
"<loc_605>": 152274,
|
| 580 |
+
"<loc_606>": 152275,
|
| 581 |
+
"<loc_607>": 152276,
|
| 582 |
+
"<loc_608>": 152277,
|
| 583 |
+
"<loc_609>": 152278,
|
| 584 |
+
"<loc_60>": 151729,
|
| 585 |
+
"<loc_610>": 152279,
|
| 586 |
+
"<loc_611>": 152280,
|
| 587 |
+
"<loc_612>": 152281,
|
| 588 |
+
"<loc_613>": 152282,
|
| 589 |
+
"<loc_614>": 152283,
|
| 590 |
+
"<loc_615>": 152284,
|
| 591 |
+
"<loc_616>": 152285,
|
| 592 |
+
"<loc_617>": 152286,
|
| 593 |
+
"<loc_618>": 152287,
|
| 594 |
+
"<loc_619>": 152288,
|
| 595 |
+
"<loc_61>": 151730,
|
| 596 |
+
"<loc_620>": 152289,
|
| 597 |
+
"<loc_621>": 152290,
|
| 598 |
+
"<loc_622>": 152291,
|
| 599 |
+
"<loc_623>": 152292,
|
| 600 |
+
"<loc_624>": 152293,
|
| 601 |
+
"<loc_625>": 152294,
|
| 602 |
+
"<loc_626>": 152295,
|
| 603 |
+
"<loc_627>": 152296,
|
| 604 |
+
"<loc_628>": 152297,
|
| 605 |
+
"<loc_629>": 152298,
|
| 606 |
+
"<loc_62>": 151731,
|
| 607 |
+
"<loc_630>": 152299,
|
| 608 |
+
"<loc_631>": 152300,
|
| 609 |
+
"<loc_632>": 152301,
|
| 610 |
+
"<loc_633>": 152302,
|
| 611 |
+
"<loc_634>": 152303,
|
| 612 |
+
"<loc_635>": 152304,
|
| 613 |
+
"<loc_636>": 152305,
|
| 614 |
+
"<loc_637>": 152306,
|
| 615 |
+
"<loc_638>": 152307,
|
| 616 |
+
"<loc_639>": 152308,
|
| 617 |
+
"<loc_63>": 151732,
|
| 618 |
+
"<loc_640>": 152309,
|
| 619 |
+
"<loc_641>": 152310,
|
| 620 |
+
"<loc_642>": 152311,
|
| 621 |
+
"<loc_643>": 152312,
|
| 622 |
+
"<loc_644>": 152313,
|
| 623 |
+
"<loc_645>": 152314,
|
| 624 |
+
"<loc_646>": 152315,
|
| 625 |
+
"<loc_647>": 152316,
|
| 626 |
+
"<loc_648>": 152317,
|
| 627 |
+
"<loc_649>": 152318,
|
| 628 |
+
"<loc_64>": 151733,
|
| 629 |
+
"<loc_650>": 152319,
|
| 630 |
+
"<loc_651>": 152320,
|
| 631 |
+
"<loc_652>": 152321,
|
| 632 |
+
"<loc_653>": 152322,
|
| 633 |
+
"<loc_654>": 152323,
|
| 634 |
+
"<loc_655>": 152324,
|
| 635 |
+
"<loc_656>": 152325,
|
| 636 |
+
"<loc_657>": 152326,
|
| 637 |
+
"<loc_658>": 152327,
|
| 638 |
+
"<loc_659>": 152328,
|
| 639 |
+
"<loc_65>": 151734,
|
| 640 |
+
"<loc_660>": 152329,
|
| 641 |
+
"<loc_661>": 152330,
|
| 642 |
+
"<loc_662>": 152331,
|
| 643 |
+
"<loc_663>": 152332,
|
| 644 |
+
"<loc_664>": 152333,
|
| 645 |
+
"<loc_665>": 152334,
|
| 646 |
+
"<loc_666>": 152335,
|
| 647 |
+
"<loc_667>": 152336,
|
| 648 |
+
"<loc_668>": 152337,
|
| 649 |
+
"<loc_669>": 152338,
|
| 650 |
+
"<loc_66>": 151735,
|
| 651 |
+
"<loc_670>": 152339,
|
| 652 |
+
"<loc_671>": 152340,
|
| 653 |
+
"<loc_672>": 152341,
|
| 654 |
+
"<loc_673>": 152342,
|
| 655 |
+
"<loc_674>": 152343,
|
| 656 |
+
"<loc_675>": 152344,
|
| 657 |
+
"<loc_676>": 152345,
|
| 658 |
+
"<loc_677>": 152346,
|
| 659 |
+
"<loc_678>": 152347,
|
| 660 |
+
"<loc_679>": 152348,
|
| 661 |
+
"<loc_67>": 151736,
|
| 662 |
+
"<loc_680>": 152349,
|
| 663 |
+
"<loc_681>": 152350,
|
| 664 |
+
"<loc_682>": 152351,
|
| 665 |
+
"<loc_683>": 152352,
|
| 666 |
+
"<loc_684>": 152353,
|
| 667 |
+
"<loc_685>": 152354,
|
| 668 |
+
"<loc_686>": 152355,
|
| 669 |
+
"<loc_687>": 152356,
|
| 670 |
+
"<loc_688>": 152357,
|
| 671 |
+
"<loc_689>": 152358,
|
| 672 |
+
"<loc_68>": 151737,
|
| 673 |
+
"<loc_690>": 152359,
|
| 674 |
+
"<loc_691>": 152360,
|
| 675 |
+
"<loc_692>": 152361,
|
| 676 |
+
"<loc_693>": 152362,
|
| 677 |
+
"<loc_694>": 152363,
|
| 678 |
+
"<loc_695>": 152364,
|
| 679 |
+
"<loc_696>": 152365,
|
| 680 |
+
"<loc_697>": 152366,
|
| 681 |
+
"<loc_698>": 152367,
|
| 682 |
+
"<loc_699>": 152368,
|
| 683 |
+
"<loc_69>": 151738,
|
| 684 |
+
"<loc_6>": 151675,
|
| 685 |
+
"<loc_700>": 152369,
|
| 686 |
+
"<loc_701>": 152370,
|
| 687 |
+
"<loc_702>": 152371,
|
| 688 |
+
"<loc_703>": 152372,
|
| 689 |
+
"<loc_704>": 152373,
|
| 690 |
+
"<loc_705>": 152374,
|
| 691 |
+
"<loc_706>": 152375,
|
| 692 |
+
"<loc_707>": 152376,
|
| 693 |
+
"<loc_708>": 152377,
|
| 694 |
+
"<loc_709>": 152378,
|
| 695 |
+
"<loc_70>": 151739,
|
| 696 |
+
"<loc_710>": 152379,
|
| 697 |
+
"<loc_711>": 152380,
|
| 698 |
+
"<loc_712>": 152381,
|
| 699 |
+
"<loc_713>": 152382,
|
| 700 |
+
"<loc_714>": 152383,
|
| 701 |
+
"<loc_715>": 152384,
|
| 702 |
+
"<loc_716>": 152385,
|
| 703 |
+
"<loc_717>": 152386,
|
| 704 |
+
"<loc_718>": 152387,
|
| 705 |
+
"<loc_719>": 152388,
|
| 706 |
+
"<loc_71>": 151740,
|
| 707 |
+
"<loc_720>": 152389,
|
| 708 |
+
"<loc_721>": 152390,
|
| 709 |
+
"<loc_722>": 152391,
|
| 710 |
+
"<loc_723>": 152392,
|
| 711 |
+
"<loc_724>": 152393,
|
| 712 |
+
"<loc_725>": 152394,
|
| 713 |
+
"<loc_726>": 152395,
|
| 714 |
+
"<loc_727>": 152396,
|
| 715 |
+
"<loc_728>": 152397,
|
| 716 |
+
"<loc_729>": 152398,
|
| 717 |
+
"<loc_72>": 151741,
|
| 718 |
+
"<loc_730>": 152399,
|
| 719 |
+
"<loc_731>": 152400,
|
| 720 |
+
"<loc_732>": 152401,
|
| 721 |
+
"<loc_733>": 152402,
|
| 722 |
+
"<loc_734>": 152403,
|
| 723 |
+
"<loc_735>": 152404,
|
| 724 |
+
"<loc_736>": 152405,
|
| 725 |
+
"<loc_737>": 152406,
|
| 726 |
+
"<loc_738>": 152407,
|
| 727 |
+
"<loc_739>": 152408,
|
| 728 |
+
"<loc_73>": 151742,
|
| 729 |
+
"<loc_740>": 152409,
|
| 730 |
+
"<loc_741>": 152410,
|
| 731 |
+
"<loc_742>": 152411,
|
| 732 |
+
"<loc_743>": 152412,
|
| 733 |
+
"<loc_744>": 152413,
|
| 734 |
+
"<loc_745>": 152414,
|
| 735 |
+
"<loc_746>": 152415,
|
| 736 |
+
"<loc_747>": 152416,
|
| 737 |
+
"<loc_748>": 152417,
|
| 738 |
+
"<loc_749>": 152418,
|
| 739 |
+
"<loc_74>": 151743,
|
| 740 |
+
"<loc_750>": 152419,
|
| 741 |
+
"<loc_751>": 152420,
|
| 742 |
+
"<loc_752>": 152421,
|
| 743 |
+
"<loc_753>": 152422,
|
| 744 |
+
"<loc_754>": 152423,
|
| 745 |
+
"<loc_755>": 152424,
|
| 746 |
+
"<loc_756>": 152425,
|
| 747 |
+
"<loc_757>": 152426,
|
| 748 |
+
"<loc_758>": 152427,
|
| 749 |
+
"<loc_759>": 152428,
|
| 750 |
+
"<loc_75>": 151744,
|
| 751 |
+
"<loc_760>": 152429,
|
| 752 |
+
"<loc_761>": 152430,
|
| 753 |
+
"<loc_762>": 152431,
|
| 754 |
+
"<loc_763>": 152432,
|
| 755 |
+
"<loc_764>": 152433,
|
| 756 |
+
"<loc_765>": 152434,
|
| 757 |
+
"<loc_766>": 152435,
|
| 758 |
+
"<loc_767>": 152436,
|
| 759 |
+
"<loc_768>": 152437,
|
| 760 |
+
"<loc_769>": 152438,
|
| 761 |
+
"<loc_76>": 151745,
|
| 762 |
+
"<loc_770>": 152439,
|
| 763 |
+
"<loc_771>": 152440,
|
| 764 |
+
"<loc_772>": 152441,
|
| 765 |
+
"<loc_773>": 152442,
|
| 766 |
+
"<loc_774>": 152443,
|
| 767 |
+
"<loc_775>": 152444,
|
| 768 |
+
"<loc_776>": 152445,
|
| 769 |
+
"<loc_777>": 152446,
|
| 770 |
+
"<loc_778>": 152447,
|
| 771 |
+
"<loc_779>": 152448,
|
| 772 |
+
"<loc_77>": 151746,
|
| 773 |
+
"<loc_780>": 152449,
|
| 774 |
+
"<loc_781>": 152450,
|
| 775 |
+
"<loc_782>": 152451,
|
| 776 |
+
"<loc_783>": 152452,
|
| 777 |
+
"<loc_784>": 152453,
|
| 778 |
+
"<loc_785>": 152454,
|
| 779 |
+
"<loc_786>": 152455,
|
| 780 |
+
"<loc_787>": 152456,
|
| 781 |
+
"<loc_788>": 152457,
|
| 782 |
+
"<loc_789>": 152458,
|
| 783 |
+
"<loc_78>": 151747,
|
| 784 |
+
"<loc_790>": 152459,
|
| 785 |
+
"<loc_791>": 152460,
|
| 786 |
+
"<loc_792>": 152461,
|
| 787 |
+
"<loc_793>": 152462,
|
| 788 |
+
"<loc_794>": 152463,
|
| 789 |
+
"<loc_795>": 152464,
|
| 790 |
+
"<loc_796>": 152465,
|
| 791 |
+
"<loc_797>": 152466,
|
| 792 |
+
"<loc_798>": 152467,
|
| 793 |
+
"<loc_799>": 152468,
|
| 794 |
+
"<loc_79>": 151748,
|
| 795 |
+
"<loc_7>": 151676,
|
| 796 |
+
"<loc_800>": 152469,
|
| 797 |
+
"<loc_801>": 152470,
|
| 798 |
+
"<loc_802>": 152471,
|
| 799 |
+
"<loc_803>": 152472,
|
| 800 |
+
"<loc_804>": 152473,
|
| 801 |
+
"<loc_805>": 152474,
|
| 802 |
+
"<loc_806>": 152475,
|
| 803 |
+
"<loc_807>": 152476,
|
| 804 |
+
"<loc_808>": 152477,
|
| 805 |
+
"<loc_809>": 152478,
|
| 806 |
+
"<loc_80>": 151749,
|
| 807 |
+
"<loc_810>": 152479,
|
| 808 |
+
"<loc_811>": 152480,
|
| 809 |
+
"<loc_812>": 152481,
|
| 810 |
+
"<loc_813>": 152482,
|
| 811 |
+
"<loc_814>": 152483,
|
| 812 |
+
"<loc_815>": 152484,
|
| 813 |
+
"<loc_816>": 152485,
|
| 814 |
+
"<loc_817>": 152486,
|
| 815 |
+
"<loc_818>": 152487,
|
| 816 |
+
"<loc_819>": 152488,
|
| 817 |
+
"<loc_81>": 151750,
|
| 818 |
+
"<loc_820>": 152489,
|
| 819 |
+
"<loc_821>": 152490,
|
| 820 |
+
"<loc_822>": 152491,
|
| 821 |
+
"<loc_823>": 152492,
|
| 822 |
+
"<loc_824>": 152493,
|
| 823 |
+
"<loc_825>": 152494,
|
| 824 |
+
"<loc_826>": 152495,
|
| 825 |
+
"<loc_827>": 152496,
|
| 826 |
+
"<loc_828>": 152497,
|
| 827 |
+
"<loc_829>": 152498,
|
| 828 |
+
"<loc_82>": 151751,
|
| 829 |
+
"<loc_830>": 152499,
|
| 830 |
+
"<loc_831>": 152500,
|
| 831 |
+
"<loc_832>": 152501,
|
| 832 |
+
"<loc_833>": 152502,
|
| 833 |
+
"<loc_834>": 152503,
|
| 834 |
+
"<loc_835>": 152504,
|
| 835 |
+
"<loc_836>": 152505,
|
| 836 |
+
"<loc_837>": 152506,
|
| 837 |
+
"<loc_838>": 152507,
|
| 838 |
+
"<loc_839>": 152508,
|
| 839 |
+
"<loc_83>": 151752,
|
| 840 |
+
"<loc_840>": 152509,
|
| 841 |
+
"<loc_841>": 152510,
|
| 842 |
+
"<loc_842>": 152511,
|
| 843 |
+
"<loc_843>": 152512,
|
| 844 |
+
"<loc_844>": 152513,
|
| 845 |
+
"<loc_845>": 152514,
|
| 846 |
+
"<loc_846>": 152515,
|
| 847 |
+
"<loc_847>": 152516,
|
| 848 |
+
"<loc_848>": 152517,
|
| 849 |
+
"<loc_849>": 152518,
|
| 850 |
+
"<loc_84>": 151753,
|
| 851 |
+
"<loc_850>": 152519,
|
| 852 |
+
"<loc_851>": 152520,
|
| 853 |
+
"<loc_852>": 152521,
|
| 854 |
+
"<loc_853>": 152522,
|
| 855 |
+
"<loc_854>": 152523,
|
| 856 |
+
"<loc_855>": 152524,
|
| 857 |
+
"<loc_856>": 152525,
|
| 858 |
+
"<loc_857>": 152526,
|
| 859 |
+
"<loc_858>": 152527,
|
| 860 |
+
"<loc_859>": 152528,
|
| 861 |
+
"<loc_85>": 151754,
|
| 862 |
+
"<loc_860>": 152529,
|
| 863 |
+
"<loc_861>": 152530,
|
| 864 |
+
"<loc_862>": 152531,
|
| 865 |
+
"<loc_863>": 152532,
|
| 866 |
+
"<loc_864>": 152533,
|
| 867 |
+
"<loc_865>": 152534,
|
| 868 |
+
"<loc_866>": 152535,
|
| 869 |
+
"<loc_867>": 152536,
|
| 870 |
+
"<loc_868>": 152537,
|
| 871 |
+
"<loc_869>": 152538,
|
| 872 |
+
"<loc_86>": 151755,
|
| 873 |
+
"<loc_870>": 152539,
|
| 874 |
+
"<loc_871>": 152540,
|
| 875 |
+
"<loc_872>": 152541,
|
| 876 |
+
"<loc_873>": 152542,
|
| 877 |
+
"<loc_874>": 152543,
|
| 878 |
+
"<loc_875>": 152544,
|
| 879 |
+
"<loc_876>": 152545,
|
| 880 |
+
"<loc_877>": 152546,
|
| 881 |
+
"<loc_878>": 152547,
|
| 882 |
+
"<loc_879>": 152548,
|
| 883 |
+
"<loc_87>": 151756,
|
| 884 |
+
"<loc_880>": 152549,
|
| 885 |
+
"<loc_881>": 152550,
|
| 886 |
+
"<loc_882>": 152551,
|
| 887 |
+
"<loc_883>": 152552,
|
| 888 |
+
"<loc_884>": 152553,
|
| 889 |
+
"<loc_885>": 152554,
|
| 890 |
+
"<loc_886>": 152555,
|
| 891 |
+
"<loc_887>": 152556,
|
| 892 |
+
"<loc_888>": 152557,
|
| 893 |
+
"<loc_889>": 152558,
|
| 894 |
+
"<loc_88>": 151757,
|
| 895 |
+
"<loc_890>": 152559,
|
| 896 |
+
"<loc_891>": 152560,
|
| 897 |
+
"<loc_892>": 152561,
|
| 898 |
+
"<loc_893>": 152562,
|
| 899 |
+
"<loc_894>": 152563,
|
| 900 |
+
"<loc_895>": 152564,
|
| 901 |
+
"<loc_896>": 152565,
|
| 902 |
+
"<loc_897>": 152566,
|
| 903 |
+
"<loc_898>": 152567,
|
| 904 |
+
"<loc_899>": 152568,
|
| 905 |
+
"<loc_89>": 151758,
|
| 906 |
+
"<loc_8>": 151677,
|
| 907 |
+
"<loc_900>": 152569,
|
| 908 |
+
"<loc_901>": 152570,
|
| 909 |
+
"<loc_902>": 152571,
|
| 910 |
+
"<loc_903>": 152572,
|
| 911 |
+
"<loc_904>": 152573,
|
| 912 |
+
"<loc_905>": 152574,
|
| 913 |
+
"<loc_906>": 152575,
|
| 914 |
+
"<loc_907>": 152576,
|
| 915 |
+
"<loc_908>": 152577,
|
| 916 |
+
"<loc_909>": 152578,
|
| 917 |
+
"<loc_90>": 151759,
|
| 918 |
+
"<loc_910>": 152579,
|
| 919 |
+
"<loc_911>": 152580,
|
| 920 |
+
"<loc_912>": 152581,
|
| 921 |
+
"<loc_913>": 152582,
|
| 922 |
+
"<loc_914>": 152583,
|
| 923 |
+
"<loc_915>": 152584,
|
| 924 |
+
"<loc_916>": 152585,
|
| 925 |
+
"<loc_917>": 152586,
|
| 926 |
+
"<loc_918>": 152587,
|
| 927 |
+
"<loc_919>": 152588,
|
| 928 |
+
"<loc_91>": 151760,
|
| 929 |
+
"<loc_920>": 152589,
|
| 930 |
+
"<loc_921>": 152590,
|
| 931 |
+
"<loc_922>": 152591,
|
| 932 |
+
"<loc_923>": 152592,
|
| 933 |
+
"<loc_924>": 152593,
|
| 934 |
+
"<loc_925>": 152594,
|
| 935 |
+
"<loc_926>": 152595,
|
| 936 |
+
"<loc_927>": 152596,
|
| 937 |
+
"<loc_928>": 152597,
|
| 938 |
+
"<loc_929>": 152598,
|
| 939 |
+
"<loc_92>": 151761,
|
| 940 |
+
"<loc_930>": 152599,
|
| 941 |
+
"<loc_931>": 152600,
|
| 942 |
+
"<loc_932>": 152601,
|
| 943 |
+
"<loc_933>": 152602,
|
| 944 |
+
"<loc_934>": 152603,
|
| 945 |
+
"<loc_935>": 152604,
|
| 946 |
+
"<loc_936>": 152605,
|
| 947 |
+
"<loc_937>": 152606,
|
| 948 |
+
"<loc_938>": 152607,
|
| 949 |
+
"<loc_939>": 152608,
|
| 950 |
+
"<loc_93>": 151762,
|
| 951 |
+
"<loc_940>": 152609,
|
| 952 |
+
"<loc_941>": 152610,
|
| 953 |
+
"<loc_942>": 152611,
|
| 954 |
+
"<loc_943>": 152612,
|
| 955 |
+
"<loc_944>": 152613,
|
| 956 |
+
"<loc_945>": 152614,
|
| 957 |
+
"<loc_946>": 152615,
|
| 958 |
+
"<loc_947>": 152616,
|
| 959 |
+
"<loc_948>": 152617,
|
| 960 |
+
"<loc_949>": 152618,
|
| 961 |
+
"<loc_94>": 151763,
|
| 962 |
+
"<loc_950>": 152619,
|
| 963 |
+
"<loc_951>": 152620,
|
| 964 |
+
"<loc_952>": 152621,
|
| 965 |
+
"<loc_953>": 152622,
|
| 966 |
+
"<loc_954>": 152623,
|
| 967 |
+
"<loc_955>": 152624,
|
| 968 |
+
"<loc_956>": 152625,
|
| 969 |
+
"<loc_957>": 152626,
|
| 970 |
+
"<loc_958>": 152627,
|
| 971 |
+
"<loc_959>": 152628,
|
| 972 |
+
"<loc_95>": 151764,
|
| 973 |
+
"<loc_960>": 152629,
|
| 974 |
+
"<loc_961>": 152630,
|
| 975 |
+
"<loc_962>": 152631,
|
| 976 |
+
"<loc_963>": 152632,
|
| 977 |
+
"<loc_964>": 152633,
|
| 978 |
+
"<loc_965>": 152634,
|
| 979 |
+
"<loc_966>": 152635,
|
| 980 |
+
"<loc_967>": 152636,
|
| 981 |
+
"<loc_968>": 152637,
|
| 982 |
+
"<loc_969>": 152638,
|
| 983 |
+
"<loc_96>": 151765,
|
| 984 |
+
"<loc_970>": 152639,
|
| 985 |
+
"<loc_971>": 152640,
|
| 986 |
+
"<loc_972>": 152641,
|
| 987 |
+
"<loc_973>": 152642,
|
| 988 |
+
"<loc_974>": 152643,
|
| 989 |
+
"<loc_975>": 152644,
|
| 990 |
+
"<loc_976>": 152645,
|
| 991 |
+
"<loc_977>": 152646,
|
| 992 |
+
"<loc_978>": 152647,
|
| 993 |
+
"<loc_979>": 152648,
|
| 994 |
+
"<loc_97>": 151766,
|
| 995 |
+
"<loc_980>": 152649,
|
| 996 |
+
"<loc_981>": 152650,
|
| 997 |
+
"<loc_982>": 152651,
|
| 998 |
+
"<loc_983>": 152652,
|
| 999 |
+
"<loc_984>": 152653,
|
| 1000 |
+
"<loc_985>": 152654,
|
| 1001 |
+
"<loc_986>": 152655,
|
| 1002 |
+
"<loc_987>": 152656,
|
| 1003 |
+
"<loc_988>": 152657,
|
| 1004 |
+
"<loc_989>": 152658,
|
| 1005 |
+
"<loc_98>": 151767,
|
| 1006 |
+
"<loc_990>": 152659,
|
| 1007 |
+
"<loc_991>": 152660,
|
| 1008 |
+
"<loc_992>": 152661,
|
| 1009 |
+
"<loc_993>": 152662,
|
| 1010 |
+
"<loc_994>": 152663,
|
| 1011 |
+
"<loc_995>": 152664,
|
| 1012 |
+
"<loc_996>": 152665,
|
| 1013 |
+
"<loc_997>": 152666,
|
| 1014 |
+
"<loc_998>": 152667,
|
| 1015 |
+
"<loc_999>": 152668,
|
| 1016 |
+
"<loc_99>": 151768,
|
| 1017 |
+
"<loc_9>": 151678,
|
| 1018 |
+
"<ncap>": 152671,
|
| 1019 |
+
"<ocr>": 151667,
|
| 1020 |
+
"<od>": 151665,
|
| 1021 |
+
"<poly>": 152686,
|
| 1022 |
+
"<proposal>": 152684,
|
| 1023 |
+
"<region_cap>": 152680,
|
| 1024 |
+
"<region_to_desciption>": 152682,
|
| 1025 |
+
"<seg>": 152677,
|
| 1026 |
+
"<sep>": 152679,
|
| 1027 |
+
"<tool_call>": 151657,
|
| 1028 |
+
"<|box_end|>": 151649,
|
| 1029 |
+
"<|box_start|>": 151648,
|
| 1030 |
+
"<|endoftext|>": 151643,
|
| 1031 |
+
"<|file_sep|>": 151664,
|
| 1032 |
+
"<|fim_middle|>": 151660,
|
| 1033 |
+
"<|fim_pad|>": 151662,
|
| 1034 |
+
"<|fim_prefix|>": 151659,
|
| 1035 |
+
"<|fim_suffix|>": 151661,
|
| 1036 |
+
"<|im_end|>": 151645,
|
| 1037 |
+
"<|im_start|>": 151644,
|
| 1038 |
+
"<|image_pad|>": 151655,
|
| 1039 |
+
"<|object_ref_end|>": 151647,
|
| 1040 |
+
"<|object_ref_start|>": 151646,
|
| 1041 |
+
"<|quad_end|>": 151651,
|
| 1042 |
+
"<|quad_start|>": 151650,
|
| 1043 |
+
"<|repo_name|>": 151663,
|
| 1044 |
+
"<|video_pad|>": 151656,
|
| 1045 |
+
"<|vision_end|>": 151653,
|
| 1046 |
+
"<|vision_pad|>": 151654,
|
| 1047 |
+
"<|vision_start|>": 151652
|
| 1048 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,112 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "checkpoints/mono-ve-3b-v0/checkpoint-7400/",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"MonoForConditionalGeneration"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"auto_map": {
|
| 8 |
+
"AutoConfig": "configuration_aimv2.MonoConfig",
|
| 9 |
+
"AutoModel": "modeling_mono.MonoForConditionalGeneration",
|
| 10 |
+
"AutoModelForCausalLM": "modeling_mono.MonoForConditionalGeneration"
|
| 11 |
+
},
|
| 12 |
+
"bos_token_id": 151643,
|
| 13 |
+
"eos_token_id": 151645,
|
| 14 |
+
"hidden_act": "silu",
|
| 15 |
+
"hidden_size": 1536,
|
| 16 |
+
"ignore_index": -100,
|
| 17 |
+
"initializer_range": 0.02,
|
| 18 |
+
"intermediate_size": 8960,
|
| 19 |
+
"max_position_embeddings": 32768,
|
| 20 |
+
"max_window_layers": 21,
|
| 21 |
+
"model_type": "mono",
|
| 22 |
+
"num_attention_heads": 12,
|
| 23 |
+
"num_hidden_layers": 28,
|
| 24 |
+
"num_key_value_heads": 2,
|
| 25 |
+
"rms_norm_eps": 1e-06,
|
| 26 |
+
"rope_scaling": null,
|
| 27 |
+
"rope_theta": 1000000.0,
|
| 28 |
+
"sliding_window": null,
|
| 29 |
+
"tie_word_embeddings": true,
|
| 30 |
+
"torch_dtype": "bfloat16",
|
| 31 |
+
"transformers_version": "4.46.2",
|
| 32 |
+
"use_bias": false,
|
| 33 |
+
"use_cache": false,
|
| 34 |
+
"use_sliding_window": false,
|
| 35 |
+
"vision_config": {
|
| 36 |
+
"_attn_implementation_autoset": true,
|
| 37 |
+
"_name_or_path": "",
|
| 38 |
+
"add_cross_attention": false,
|
| 39 |
+
"architectures": null,
|
| 40 |
+
"attention_dropout": 0.0,
|
| 41 |
+
"bad_words_ids": null,
|
| 42 |
+
"begin_suppress_tokens": null,
|
| 43 |
+
"bos_token_id": null,
|
| 44 |
+
"chunk_size_feed_forward": 0,
|
| 45 |
+
"cross_attention_hidden_size": null,
|
| 46 |
+
"decoder_start_token_id": null,
|
| 47 |
+
"diversity_penalty": 0.0,
|
| 48 |
+
"do_sample": false,
|
| 49 |
+
"early_stopping": false,
|
| 50 |
+
"encoder_no_repeat_ngram_size": 0,
|
| 51 |
+
"eos_token_id": null,
|
| 52 |
+
"exponential_decay_length_penalty": null,
|
| 53 |
+
"finetuning_task": null,
|
| 54 |
+
"forced_bos_token_id": null,
|
| 55 |
+
"forced_eos_token_id": null,
|
| 56 |
+
"hidden_size": 3072,
|
| 57 |
+
"id2label": {
|
| 58 |
+
"0": "LABEL_0",
|
| 59 |
+
"1": "LABEL_1"
|
| 60 |
+
},
|
| 61 |
+
"image_size": 448,
|
| 62 |
+
"intermediate_size": 8192,
|
| 63 |
+
"is_decoder": false,
|
| 64 |
+
"is_encoder_decoder": false,
|
| 65 |
+
"label2id": {
|
| 66 |
+
"LABEL_0": 0,
|
| 67 |
+
"LABEL_1": 1
|
| 68 |
+
},
|
| 69 |
+
"length_penalty": 1.0,
|
| 70 |
+
"max_length": 20,
|
| 71 |
+
"min_length": 0,
|
| 72 |
+
"model_type": "aimv2",
|
| 73 |
+
"no_repeat_ngram_size": 0,
|
| 74 |
+
"num_attention_heads": 24,
|
| 75 |
+
"num_beam_groups": 1,
|
| 76 |
+
"num_beams": 1,
|
| 77 |
+
"num_channels": 3,
|
| 78 |
+
"num_hidden_layers": 24,
|
| 79 |
+
"num_return_sequences": 1,
|
| 80 |
+
"output_attentions": false,
|
| 81 |
+
"output_hidden_states": false,
|
| 82 |
+
"output_scores": false,
|
| 83 |
+
"pad_token_id": null,
|
| 84 |
+
"patch_size": 14,
|
| 85 |
+
"prefix": null,
|
| 86 |
+
"problem_type": null,
|
| 87 |
+
"projection_dropout": 0.0,
|
| 88 |
+
"pruned_heads": {},
|
| 89 |
+
"qkv_bias": false,
|
| 90 |
+
"remove_invalid_values": false,
|
| 91 |
+
"repetition_penalty": 1.0,
|
| 92 |
+
"return_dict": true,
|
| 93 |
+
"return_dict_in_generate": false,
|
| 94 |
+
"rms_norm_eps": 1e-05,
|
| 95 |
+
"sep_token_id": null,
|
| 96 |
+
"suppress_tokens": null,
|
| 97 |
+
"task_specific_params": null,
|
| 98 |
+
"temperature": 1.0,
|
| 99 |
+
"tf_legacy_loss": false,
|
| 100 |
+
"tie_encoder_decoder": false,
|
| 101 |
+
"tie_word_embeddings": true,
|
| 102 |
+
"tokenizer_class": null,
|
| 103 |
+
"top_k": 50,
|
| 104 |
+
"top_p": 1.0,
|
| 105 |
+
"torch_dtype": null,
|
| 106 |
+
"torchscript": false,
|
| 107 |
+
"typical_p": 1.0,
|
| 108 |
+
"use_bfloat16": false,
|
| 109 |
+
"use_bias": false
|
| 110 |
+
},
|
| 111 |
+
"vocab_size": 151936
|
| 112 |
+
}
|
configuration_aimv2.py
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
from typing import Any
|
| 2 |
+
|
| 3 |
+
from transformers.configuration_utils import PretrainedConfig
|
| 4 |
+
from transformers import Qwen2Config
|
| 5 |
+
|
| 6 |
+
__all__ = ["AIMv2Config", "MonoConfig"]
|
| 7 |
+
|
| 8 |
+
|
| 9 |
+
class AIMv2Config(PretrainedConfig):
|
| 10 |
+
|
| 11 |
+
model_type: str = "aimv2"
|
| 12 |
+
|
| 13 |
+
def __init__(
|
| 14 |
+
self,
|
| 15 |
+
hidden_size: int = 1024,
|
| 16 |
+
intermediate_size: int = 2816,
|
| 17 |
+
num_hidden_layers: int = 24,
|
| 18 |
+
num_attention_heads: int = 8,
|
| 19 |
+
num_channels: int = 3,
|
| 20 |
+
image_size: int = 224,
|
| 21 |
+
patch_size: int = 14,
|
| 22 |
+
rms_norm_eps: float = 1e-5,
|
| 23 |
+
attention_dropout: float = 0.0,
|
| 24 |
+
projection_dropout: float = 0.0,
|
| 25 |
+
qkv_bias: bool = False,
|
| 26 |
+
use_bias: bool = False,
|
| 27 |
+
text_config=None,
|
| 28 |
+
**kwargs: Any,
|
| 29 |
+
):
|
| 30 |
+
super().__init__(**kwargs)
|
| 31 |
+
self.hidden_size = hidden_size
|
| 32 |
+
self.intermediate_size = intermediate_size
|
| 33 |
+
self.num_hidden_layers = num_hidden_layers
|
| 34 |
+
self.num_attention_heads = num_attention_heads
|
| 35 |
+
self.num_channels = num_channels
|
| 36 |
+
self.patch_size = patch_size
|
| 37 |
+
self.image_size = image_size
|
| 38 |
+
self.attention_dropout = attention_dropout
|
| 39 |
+
self.rms_norm_eps = rms_norm_eps
|
| 40 |
+
|
| 41 |
+
self.projection_dropout = projection_dropout
|
| 42 |
+
self.qkv_bias = qkv_bias
|
| 43 |
+
self.use_bias = use_bias
|
| 44 |
+
|
| 45 |
+
|
| 46 |
+
class MonoConfig(Qwen2Config):
|
| 47 |
+
|
| 48 |
+
model_type = "mono"
|
| 49 |
+
is_composition = False
|
| 50 |
+
|
| 51 |
+
def __init__(
|
| 52 |
+
self,
|
| 53 |
+
vision_config=None,
|
| 54 |
+
ignore_index=-100,
|
| 55 |
+
**kwargs,
|
| 56 |
+
):
|
| 57 |
+
self.ignore_index = ignore_index
|
| 58 |
+
if vision_config is not None:
|
| 59 |
+
vision_config = AIMv2Config(**vision_config)
|
| 60 |
+
self.vision_config = vision_config
|
| 61 |
+
|
| 62 |
+
super().__init__(**kwargs)
|
generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 151643,
|
| 4 |
+
"eos_token_id": 151645,
|
| 5 |
+
"transformers_version": "4.46.2",
|
| 6 |
+
"use_cache": false
|
| 7 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
model-00001-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8a35a90d9e2f350fbed0c43707bbc51e1ca4b7605512f054b571bbb06bf1b45b
|
| 3 |
+
size 4993038848
|
model-00002-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bfee8731806f389a71319d02d93a67d6ab5520a27a440e3743117e8a55f14dc9
|
| 3 |
+
size 3634863200
|
model.safetensors.index.json
ADDED
|
@@ -0,0 +1,524 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"metadata": {
|
| 3 |
+
"total_size": 8627842048
|
| 4 |
+
},
|
| 5 |
+
"weight_map": {
|
| 6 |
+
"mm_projector.linear_proj.dense_4h_to_h.weight": "model-00002-of-00002.safetensors",
|
| 7 |
+
"mm_projector.linear_proj.dense_h_to_4h.weight": "model-00002-of-00002.safetensors",
|
| 8 |
+
"mm_projector.linear_proj.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 9 |
+
"mm_projector.linear_proj.linear_proj.weight": "model-00002-of-00002.safetensors",
|
| 10 |
+
"mm_projector.linear_proj.norm1.bias": "model-00002-of-00002.safetensors",
|
| 11 |
+
"mm_projector.linear_proj.norm1.weight": "model-00002-of-00002.safetensors",
|
| 12 |
+
"model.embed_tokens.weight": "model-00002-of-00002.safetensors",
|
| 13 |
+
"model.layers.0.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 14 |
+
"model.layers.0.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 15 |
+
"model.layers.0.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 16 |
+
"model.layers.0.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 17 |
+
"model.layers.0.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 18 |
+
"model.layers.0.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 19 |
+
"model.layers.0.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 20 |
+
"model.layers.0.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 21 |
+
"model.layers.0.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 22 |
+
"model.layers.0.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 23 |
+
"model.layers.0.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 24 |
+
"model.layers.0.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 25 |
+
"model.layers.1.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 26 |
+
"model.layers.1.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 27 |
+
"model.layers.1.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 28 |
+
"model.layers.1.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 29 |
+
"model.layers.1.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 30 |
+
"model.layers.1.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 31 |
+
"model.layers.1.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 32 |
+
"model.layers.1.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 33 |
+
"model.layers.1.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 34 |
+
"model.layers.1.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 35 |
+
"model.layers.1.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 36 |
+
"model.layers.1.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 37 |
+
"model.layers.10.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 38 |
+
"model.layers.10.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 39 |
+
"model.layers.10.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 40 |
+
"model.layers.10.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 41 |
+
"model.layers.10.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 42 |
+
"model.layers.10.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 43 |
+
"model.layers.10.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 44 |
+
"model.layers.10.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 45 |
+
"model.layers.10.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 46 |
+
"model.layers.10.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 47 |
+
"model.layers.10.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 48 |
+
"model.layers.10.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 49 |
+
"model.layers.11.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 50 |
+
"model.layers.11.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 51 |
+
"model.layers.11.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 52 |
+
"model.layers.11.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 53 |
+
"model.layers.11.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 54 |
+
"model.layers.11.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 55 |
+
"model.layers.11.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 56 |
+
"model.layers.11.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 57 |
+
"model.layers.11.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 58 |
+
"model.layers.11.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 59 |
+
"model.layers.11.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 60 |
+
"model.layers.11.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 61 |
+
"model.layers.12.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 62 |
+
"model.layers.12.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 63 |
+
"model.layers.12.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 64 |
+
"model.layers.12.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 65 |
+
"model.layers.12.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 66 |
+
"model.layers.12.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 67 |
+
"model.layers.12.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 68 |
+
"model.layers.12.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 69 |
+
"model.layers.12.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 70 |
+
"model.layers.12.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 71 |
+
"model.layers.12.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 72 |
+
"model.layers.12.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 73 |
+
"model.layers.13.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 74 |
+
"model.layers.13.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 75 |
+
"model.layers.13.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 76 |
+
"model.layers.13.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 77 |
+
"model.layers.13.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 78 |
+
"model.layers.13.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 79 |
+
"model.layers.13.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 80 |
+
"model.layers.13.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 81 |
+
"model.layers.13.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 82 |
+
"model.layers.13.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 83 |
+
"model.layers.13.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 84 |
+
"model.layers.13.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 85 |
+
"model.layers.14.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 86 |
+
"model.layers.14.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 87 |
+
"model.layers.14.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 88 |
+
"model.layers.14.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 89 |
+
"model.layers.14.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 90 |
+
"model.layers.14.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 91 |
+
"model.layers.14.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 92 |
+
"model.layers.14.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 93 |
+
"model.layers.14.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 94 |
+
"model.layers.14.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 95 |
+
"model.layers.14.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 96 |
+
"model.layers.14.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 97 |
+
"model.layers.15.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 98 |
+
"model.layers.15.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 99 |
+
"model.layers.15.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 100 |
+
"model.layers.15.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 101 |
+
"model.layers.15.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 102 |
+
"model.layers.15.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 103 |
+
"model.layers.15.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 104 |
+
"model.layers.15.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 105 |
+
"model.layers.15.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 106 |
+
"model.layers.15.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 107 |
+
"model.layers.15.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 108 |
+
"model.layers.15.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 109 |
+
"model.layers.16.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 110 |
+
"model.layers.16.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 111 |
+
"model.layers.16.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 112 |
+
"model.layers.16.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 113 |
+
"model.layers.16.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 114 |
+
"model.layers.16.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 115 |
+
"model.layers.16.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 116 |
+
"model.layers.16.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 117 |
+
"model.layers.16.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 118 |
+
"model.layers.16.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 119 |
+
"model.layers.16.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 120 |
+
"model.layers.16.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 121 |
+
"model.layers.17.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 122 |
+
"model.layers.17.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 123 |
+
"model.layers.17.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 124 |
+
"model.layers.17.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 125 |
+
"model.layers.17.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 126 |
+
"model.layers.17.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 127 |
+
"model.layers.17.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 128 |
+
"model.layers.17.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 129 |
+
"model.layers.17.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 130 |
+
"model.layers.17.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 131 |
+
"model.layers.17.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 132 |
+
"model.layers.17.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 133 |
+
"model.layers.18.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 134 |
+
"model.layers.18.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 135 |
+
"model.layers.18.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 136 |
+
"model.layers.18.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 137 |
+
"model.layers.18.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 138 |
+
"model.layers.18.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 139 |
+
"model.layers.18.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 140 |
+
"model.layers.18.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 141 |
+
"model.layers.18.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 142 |
+
"model.layers.18.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 143 |
+
"model.layers.18.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 144 |
+
"model.layers.18.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 145 |
+
"model.layers.19.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 146 |
+
"model.layers.19.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 147 |
+
"model.layers.19.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 148 |
+
"model.layers.19.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 149 |
+
"model.layers.19.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 150 |
+
"model.layers.19.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 151 |
+
"model.layers.19.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 152 |
+
"model.layers.19.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 153 |
+
"model.layers.19.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 154 |
+
"model.layers.19.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 155 |
+
"model.layers.19.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 156 |
+
"model.layers.19.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 157 |
+
"model.layers.2.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 158 |
+
"model.layers.2.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 159 |
+
"model.layers.2.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 160 |
+
"model.layers.2.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 161 |
+
"model.layers.2.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 162 |
+
"model.layers.2.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 163 |
+
"model.layers.2.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 164 |
+
"model.layers.2.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 165 |
+
"model.layers.2.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 166 |
+
"model.layers.2.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 167 |
+
"model.layers.2.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 168 |
+
"model.layers.2.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 169 |
+
"model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 170 |
+
"model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 171 |
+
"model.layers.20.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 172 |
+
"model.layers.20.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 173 |
+
"model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 174 |
+
"model.layers.20.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 175 |
+
"model.layers.20.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 176 |
+
"model.layers.20.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 177 |
+
"model.layers.20.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 178 |
+
"model.layers.20.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 179 |
+
"model.layers.20.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 180 |
+
"model.layers.20.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 181 |
+
"model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 182 |
+
"model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 183 |
+
"model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 184 |
+
"model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 185 |
+
"model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 186 |
+
"model.layers.21.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 187 |
+
"model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 188 |
+
"model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 189 |
+
"model.layers.21.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 190 |
+
"model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 191 |
+
"model.layers.21.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 192 |
+
"model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 193 |
+
"model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 194 |
+
"model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 195 |
+
"model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 196 |
+
"model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 197 |
+
"model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 198 |
+
"model.layers.22.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 199 |
+
"model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 200 |
+
"model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 201 |
+
"model.layers.22.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 202 |
+
"model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 203 |
+
"model.layers.22.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 204 |
+
"model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 205 |
+
"model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 206 |
+
"model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 207 |
+
"model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 208 |
+
"model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 209 |
+
"model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 210 |
+
"model.layers.23.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 211 |
+
"model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 212 |
+
"model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 213 |
+
"model.layers.23.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 214 |
+
"model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 215 |
+
"model.layers.23.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 216 |
+
"model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 217 |
+
"model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 218 |
+
"model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 219 |
+
"model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 220 |
+
"model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 221 |
+
"model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 222 |
+
"model.layers.24.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 223 |
+
"model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 224 |
+
"model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 225 |
+
"model.layers.24.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 226 |
+
"model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 227 |
+
"model.layers.24.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 228 |
+
"model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 229 |
+
"model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 230 |
+
"model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 231 |
+
"model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 232 |
+
"model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 233 |
+
"model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 234 |
+
"model.layers.25.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 235 |
+
"model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 236 |
+
"model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 237 |
+
"model.layers.25.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 238 |
+
"model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 239 |
+
"model.layers.25.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 240 |
+
"model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 241 |
+
"model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 242 |
+
"model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 243 |
+
"model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 244 |
+
"model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 245 |
+
"model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 246 |
+
"model.layers.26.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 247 |
+
"model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 248 |
+
"model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 249 |
+
"model.layers.26.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 250 |
+
"model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 251 |
+
"model.layers.26.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 252 |
+
"model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 253 |
+
"model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 254 |
+
"model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 255 |
+
"model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 256 |
+
"model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 257 |
+
"model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 258 |
+
"model.layers.27.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 259 |
+
"model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 260 |
+
"model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 261 |
+
"model.layers.27.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 262 |
+
"model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 263 |
+
"model.layers.27.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 264 |
+
"model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 265 |
+
"model.layers.3.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 266 |
+
"model.layers.3.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 267 |
+
"model.layers.3.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 268 |
+
"model.layers.3.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 269 |
+
"model.layers.3.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 270 |
+
"model.layers.3.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 271 |
+
"model.layers.3.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 272 |
+
"model.layers.3.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 273 |
+
"model.layers.3.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 274 |
+
"model.layers.3.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 275 |
+
"model.layers.3.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 276 |
+
"model.layers.3.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 277 |
+
"model.layers.4.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 278 |
+
"model.layers.4.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 279 |
+
"model.layers.4.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 280 |
+
"model.layers.4.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 281 |
+
"model.layers.4.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 282 |
+
"model.layers.4.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 283 |
+
"model.layers.4.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 284 |
+
"model.layers.4.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 285 |
+
"model.layers.4.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 286 |
+
"model.layers.4.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 287 |
+
"model.layers.4.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 288 |
+
"model.layers.4.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 289 |
+
"model.layers.5.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 290 |
+
"model.layers.5.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 291 |
+
"model.layers.5.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 292 |
+
"model.layers.5.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 293 |
+
"model.layers.5.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 294 |
+
"model.layers.5.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 295 |
+
"model.layers.5.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 296 |
+
"model.layers.5.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 297 |
+
"model.layers.5.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 298 |
+
"model.layers.5.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 299 |
+
"model.layers.5.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 300 |
+
"model.layers.5.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 301 |
+
"model.layers.6.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 302 |
+
"model.layers.6.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 303 |
+
"model.layers.6.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 304 |
+
"model.layers.6.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 305 |
+
"model.layers.6.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 306 |
+
"model.layers.6.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 307 |
+
"model.layers.6.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 308 |
+
"model.layers.6.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 309 |
+
"model.layers.6.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 310 |
+
"model.layers.6.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 311 |
+
"model.layers.6.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 312 |
+
"model.layers.6.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 313 |
+
"model.layers.7.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 314 |
+
"model.layers.7.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 315 |
+
"model.layers.7.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 316 |
+
"model.layers.7.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 317 |
+
"model.layers.7.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 318 |
+
"model.layers.7.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 319 |
+
"model.layers.7.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 320 |
+
"model.layers.7.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 321 |
+
"model.layers.7.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 322 |
+
"model.layers.7.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 323 |
+
"model.layers.7.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 324 |
+
"model.layers.7.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 325 |
+
"model.layers.8.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 326 |
+
"model.layers.8.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 327 |
+
"model.layers.8.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 328 |
+
"model.layers.8.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 329 |
+
"model.layers.8.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 330 |
+
"model.layers.8.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 331 |
+
"model.layers.8.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 332 |
+
"model.layers.8.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 333 |
+
"model.layers.8.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 334 |
+
"model.layers.8.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 335 |
+
"model.layers.8.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 336 |
+
"model.layers.8.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 337 |
+
"model.layers.9.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 338 |
+
"model.layers.9.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 339 |
+
"model.layers.9.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 340 |
+
"model.layers.9.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 341 |
+
"model.layers.9.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 342 |
+
"model.layers.9.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
|
| 343 |
+
"model.layers.9.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 344 |
+
"model.layers.9.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 345 |
+
"model.layers.9.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
|
| 346 |
+
"model.layers.9.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 347 |
+
"model.layers.9.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
|
| 348 |
+
"model.layers.9.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 349 |
+
"model.norm.weight": "model-00002-of-00002.safetensors",
|
| 350 |
+
"vision_tower.preprocessor.patchifier.norm.weight": "model-00001-of-00002.safetensors",
|
| 351 |
+
"vision_tower.preprocessor.patchifier.proj.bias": "model-00001-of-00002.safetensors",
|
| 352 |
+
"vision_tower.preprocessor.patchifier.proj.weight": "model-00001-of-00002.safetensors",
|
| 353 |
+
"vision_tower.preprocessor.pos_embed": "model-00001-of-00002.safetensors",
|
| 354 |
+
"vision_tower.trunk.blocks.0.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 355 |
+
"vision_tower.trunk.blocks.0.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 356 |
+
"vision_tower.trunk.blocks.0.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 357 |
+
"vision_tower.trunk.blocks.0.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 358 |
+
"vision_tower.trunk.blocks.0.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 359 |
+
"vision_tower.trunk.blocks.0.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 360 |
+
"vision_tower.trunk.blocks.0.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 361 |
+
"vision_tower.trunk.blocks.1.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 362 |
+
"vision_tower.trunk.blocks.1.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 363 |
+
"vision_tower.trunk.blocks.1.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 364 |
+
"vision_tower.trunk.blocks.1.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 365 |
+
"vision_tower.trunk.blocks.1.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 366 |
+
"vision_tower.trunk.blocks.1.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 367 |
+
"vision_tower.trunk.blocks.1.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 368 |
+
"vision_tower.trunk.blocks.10.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 369 |
+
"vision_tower.trunk.blocks.10.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 370 |
+
"vision_tower.trunk.blocks.10.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 371 |
+
"vision_tower.trunk.blocks.10.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 372 |
+
"vision_tower.trunk.blocks.10.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 373 |
+
"vision_tower.trunk.blocks.10.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 374 |
+
"vision_tower.trunk.blocks.10.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 375 |
+
"vision_tower.trunk.blocks.11.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 376 |
+
"vision_tower.trunk.blocks.11.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 377 |
+
"vision_tower.trunk.blocks.11.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 378 |
+
"vision_tower.trunk.blocks.11.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 379 |
+
"vision_tower.trunk.blocks.11.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 380 |
+
"vision_tower.trunk.blocks.11.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 381 |
+
"vision_tower.trunk.blocks.11.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 382 |
+
"vision_tower.trunk.blocks.12.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 383 |
+
"vision_tower.trunk.blocks.12.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 384 |
+
"vision_tower.trunk.blocks.12.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 385 |
+
"vision_tower.trunk.blocks.12.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 386 |
+
"vision_tower.trunk.blocks.12.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 387 |
+
"vision_tower.trunk.blocks.12.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 388 |
+
"vision_tower.trunk.blocks.12.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 389 |
+
"vision_tower.trunk.blocks.13.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 390 |
+
"vision_tower.trunk.blocks.13.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 391 |
+
"vision_tower.trunk.blocks.13.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 392 |
+
"vision_tower.trunk.blocks.13.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 393 |
+
"vision_tower.trunk.blocks.13.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 394 |
+
"vision_tower.trunk.blocks.13.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 395 |
+
"vision_tower.trunk.blocks.13.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 396 |
+
"vision_tower.trunk.blocks.14.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 397 |
+
"vision_tower.trunk.blocks.14.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 398 |
+
"vision_tower.trunk.blocks.14.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 399 |
+
"vision_tower.trunk.blocks.14.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 400 |
+
"vision_tower.trunk.blocks.14.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 401 |
+
"vision_tower.trunk.blocks.14.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 402 |
+
"vision_tower.trunk.blocks.14.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 403 |
+
"vision_tower.trunk.blocks.15.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 404 |
+
"vision_tower.trunk.blocks.15.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 405 |
+
"vision_tower.trunk.blocks.15.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 406 |
+
"vision_tower.trunk.blocks.15.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 407 |
+
"vision_tower.trunk.blocks.15.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 408 |
+
"vision_tower.trunk.blocks.15.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 409 |
+
"vision_tower.trunk.blocks.15.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 410 |
+
"vision_tower.trunk.blocks.16.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 411 |
+
"vision_tower.trunk.blocks.16.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 412 |
+
"vision_tower.trunk.blocks.16.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 413 |
+
"vision_tower.trunk.blocks.16.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 414 |
+
"vision_tower.trunk.blocks.16.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 415 |
+
"vision_tower.trunk.blocks.16.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 416 |
+
"vision_tower.trunk.blocks.16.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 417 |
+
"vision_tower.trunk.blocks.17.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 418 |
+
"vision_tower.trunk.blocks.17.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 419 |
+
"vision_tower.trunk.blocks.17.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 420 |
+
"vision_tower.trunk.blocks.17.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 421 |
+
"vision_tower.trunk.blocks.17.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 422 |
+
"vision_tower.trunk.blocks.17.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 423 |
+
"vision_tower.trunk.blocks.17.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 424 |
+
"vision_tower.trunk.blocks.18.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 425 |
+
"vision_tower.trunk.blocks.18.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 426 |
+
"vision_tower.trunk.blocks.18.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 427 |
+
"vision_tower.trunk.blocks.18.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 428 |
+
"vision_tower.trunk.blocks.18.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 429 |
+
"vision_tower.trunk.blocks.18.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 430 |
+
"vision_tower.trunk.blocks.18.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 431 |
+
"vision_tower.trunk.blocks.19.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 432 |
+
"vision_tower.trunk.blocks.19.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 433 |
+
"vision_tower.trunk.blocks.19.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 434 |
+
"vision_tower.trunk.blocks.19.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 435 |
+
"vision_tower.trunk.blocks.19.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 436 |
+
"vision_tower.trunk.blocks.19.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 437 |
+
"vision_tower.trunk.blocks.19.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 438 |
+
"vision_tower.trunk.blocks.2.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 439 |
+
"vision_tower.trunk.blocks.2.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 440 |
+
"vision_tower.trunk.blocks.2.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 441 |
+
"vision_tower.trunk.blocks.2.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 442 |
+
"vision_tower.trunk.blocks.2.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 443 |
+
"vision_tower.trunk.blocks.2.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 444 |
+
"vision_tower.trunk.blocks.2.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 445 |
+
"vision_tower.trunk.blocks.20.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 446 |
+
"vision_tower.trunk.blocks.20.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 447 |
+
"vision_tower.trunk.blocks.20.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 448 |
+
"vision_tower.trunk.blocks.20.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 449 |
+
"vision_tower.trunk.blocks.20.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 450 |
+
"vision_tower.trunk.blocks.20.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 451 |
+
"vision_tower.trunk.blocks.20.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 452 |
+
"vision_tower.trunk.blocks.21.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 453 |
+
"vision_tower.trunk.blocks.21.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 454 |
+
"vision_tower.trunk.blocks.21.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 455 |
+
"vision_tower.trunk.blocks.21.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 456 |
+
"vision_tower.trunk.blocks.21.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 457 |
+
"vision_tower.trunk.blocks.21.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 458 |
+
"vision_tower.trunk.blocks.21.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 459 |
+
"vision_tower.trunk.blocks.22.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 460 |
+
"vision_tower.trunk.blocks.22.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 461 |
+
"vision_tower.trunk.blocks.22.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 462 |
+
"vision_tower.trunk.blocks.22.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 463 |
+
"vision_tower.trunk.blocks.22.mlp.fc3.weight": "model-00002-of-00002.safetensors",
|
| 464 |
+
"vision_tower.trunk.blocks.22.norm_1.weight": "model-00002-of-00002.safetensors",
|
| 465 |
+
"vision_tower.trunk.blocks.22.norm_2.weight": "model-00002-of-00002.safetensors",
|
| 466 |
+
"vision_tower.trunk.blocks.23.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 467 |
+
"vision_tower.trunk.blocks.23.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 468 |
+
"vision_tower.trunk.blocks.23.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 469 |
+
"vision_tower.trunk.blocks.23.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 470 |
+
"vision_tower.trunk.blocks.23.mlp.fc3.weight": "model-00002-of-00002.safetensors",
|
| 471 |
+
"vision_tower.trunk.blocks.23.norm_1.weight": "model-00002-of-00002.safetensors",
|
| 472 |
+
"vision_tower.trunk.blocks.23.norm_2.weight": "model-00002-of-00002.safetensors",
|
| 473 |
+
"vision_tower.trunk.blocks.3.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 474 |
+
"vision_tower.trunk.blocks.3.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 475 |
+
"vision_tower.trunk.blocks.3.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 476 |
+
"vision_tower.trunk.blocks.3.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 477 |
+
"vision_tower.trunk.blocks.3.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 478 |
+
"vision_tower.trunk.blocks.3.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 479 |
+
"vision_tower.trunk.blocks.3.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 480 |
+
"vision_tower.trunk.blocks.4.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 481 |
+
"vision_tower.trunk.blocks.4.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 482 |
+
"vision_tower.trunk.blocks.4.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 483 |
+
"vision_tower.trunk.blocks.4.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 484 |
+
"vision_tower.trunk.blocks.4.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 485 |
+
"vision_tower.trunk.blocks.4.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 486 |
+
"vision_tower.trunk.blocks.4.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 487 |
+
"vision_tower.trunk.blocks.5.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 488 |
+
"vision_tower.trunk.blocks.5.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 489 |
+
"vision_tower.trunk.blocks.5.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 490 |
+
"vision_tower.trunk.blocks.5.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 491 |
+
"vision_tower.trunk.blocks.5.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 492 |
+
"vision_tower.trunk.blocks.5.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 493 |
+
"vision_tower.trunk.blocks.5.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 494 |
+
"vision_tower.trunk.blocks.6.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 495 |
+
"vision_tower.trunk.blocks.6.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 496 |
+
"vision_tower.trunk.blocks.6.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 497 |
+
"vision_tower.trunk.blocks.6.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 498 |
+
"vision_tower.trunk.blocks.6.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 499 |
+
"vision_tower.trunk.blocks.6.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 500 |
+
"vision_tower.trunk.blocks.6.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 501 |
+
"vision_tower.trunk.blocks.7.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 502 |
+
"vision_tower.trunk.blocks.7.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 503 |
+
"vision_tower.trunk.blocks.7.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 504 |
+
"vision_tower.trunk.blocks.7.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 505 |
+
"vision_tower.trunk.blocks.7.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 506 |
+
"vision_tower.trunk.blocks.7.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 507 |
+
"vision_tower.trunk.blocks.7.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 508 |
+
"vision_tower.trunk.blocks.8.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 509 |
+
"vision_tower.trunk.blocks.8.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 510 |
+
"vision_tower.trunk.blocks.8.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 511 |
+
"vision_tower.trunk.blocks.8.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 512 |
+
"vision_tower.trunk.blocks.8.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 513 |
+
"vision_tower.trunk.blocks.8.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 514 |
+
"vision_tower.trunk.blocks.8.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 515 |
+
"vision_tower.trunk.blocks.9.attn.proj.weight": "model-00001-of-00002.safetensors",
|
| 516 |
+
"vision_tower.trunk.blocks.9.attn.qkv.weight": "model-00001-of-00002.safetensors",
|
| 517 |
+
"vision_tower.trunk.blocks.9.mlp.fc1.weight": "model-00001-of-00002.safetensors",
|
| 518 |
+
"vision_tower.trunk.blocks.9.mlp.fc2.weight": "model-00001-of-00002.safetensors",
|
| 519 |
+
"vision_tower.trunk.blocks.9.mlp.fc3.weight": "model-00001-of-00002.safetensors",
|
| 520 |
+
"vision_tower.trunk.blocks.9.norm_1.weight": "model-00001-of-00002.safetensors",
|
| 521 |
+
"vision_tower.trunk.blocks.9.norm_2.weight": "model-00001-of-00002.safetensors",
|
| 522 |
+
"vision_tower.trunk.post_trunk_norm.weight": "model-00002-of-00002.safetensors"
|
| 523 |
+
}
|
| 524 |
+
}
|
modeling_aimv2.py
ADDED
|
@@ -0,0 +1,287 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
from typing import Optional, Tuple, Union
|
| 2 |
+
|
| 3 |
+
import torch
|
| 4 |
+
from .configuration_aimv2 import AIMv2Config
|
| 5 |
+
from torch import nn
|
| 6 |
+
from torch.nn import functional as F
|
| 7 |
+
from transformers.modeling_outputs import BaseModelOutputWithNoAttention
|
| 8 |
+
from transformers.modeling_utils import PreTrainedModel
|
| 9 |
+
|
| 10 |
+
__all__ = ["AIMv2Model"]
|
| 11 |
+
|
| 12 |
+
|
| 13 |
+
class RMSNorm(nn.Module):
|
| 14 |
+
def __init__(self, dim: int, eps: float = 1e-6):
|
| 15 |
+
super().__init__()
|
| 16 |
+
self.weight = nn.Parameter(torch.ones(dim))
|
| 17 |
+
self.eps = eps
|
| 18 |
+
|
| 19 |
+
def forward(self, x: torch.Tensor) -> torch.Tensor:
|
| 20 |
+
output = self._norm(x.float()).type_as(x)
|
| 21 |
+
return output * self.weight
|
| 22 |
+
|
| 23 |
+
def extra_repr(self) -> str:
|
| 24 |
+
return f"{tuple(self.weight.shape)}, eps={self.eps}"
|
| 25 |
+
|
| 26 |
+
def _norm(self, x: torch.Tensor) -> torch.Tensor:
|
| 27 |
+
return x * torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps)
|
| 28 |
+
|
| 29 |
+
|
| 30 |
+
class AIMv2SwiGLUFFN(nn.Module):
|
| 31 |
+
def __init__(self, config: AIMv2Config):
|
| 32 |
+
super().__init__()
|
| 33 |
+
hidden_features = config.intermediate_size
|
| 34 |
+
in_features = config.hidden_size
|
| 35 |
+
bias = config.use_bias
|
| 36 |
+
|
| 37 |
+
self.fc1 = nn.Linear(in_features, hidden_features, bias=bias)
|
| 38 |
+
self.fc2 = nn.Linear(hidden_features, in_features, bias=bias)
|
| 39 |
+
self.fc3 = nn.Linear(in_features, hidden_features, bias=bias)
|
| 40 |
+
|
| 41 |
+
def forward(self, x: torch.Tensor) -> torch.Tensor:
|
| 42 |
+
x = F.silu(self.fc1(x)) * self.fc3(x)
|
| 43 |
+
x = self.fc2(x)
|
| 44 |
+
return x
|
| 45 |
+
|
| 46 |
+
|
| 47 |
+
class AIMv2PatchEmbed(nn.Module):
|
| 48 |
+
def __init__(self, config: AIMv2Config):
|
| 49 |
+
super().__init__()
|
| 50 |
+
self.proj = nn.Conv2d(
|
| 51 |
+
config.num_channels,
|
| 52 |
+
config.hidden_size,
|
| 53 |
+
kernel_size=(config.patch_size, config.patch_size),
|
| 54 |
+
stride=(config.patch_size, config.patch_size),
|
| 55 |
+
)
|
| 56 |
+
self.norm = RMSNorm(config.hidden_size, eps=config.rms_norm_eps)
|
| 57 |
+
|
| 58 |
+
def forward(self, x: torch.Tensor) -> torch.Tensor:
|
| 59 |
+
x = self.proj(x).flatten(2).transpose(1, 2)
|
| 60 |
+
x = self.norm(x)
|
| 61 |
+
return x
|
| 62 |
+
|
| 63 |
+
|
| 64 |
+
class AIMv2ViTPreprocessor(nn.Module):
|
| 65 |
+
def __init__(self, config: AIMv2Config):
|
| 66 |
+
super().__init__()
|
| 67 |
+
num_patches = (config.image_size // config.patch_size) ** 2
|
| 68 |
+
|
| 69 |
+
self.patchifier = AIMv2PatchEmbed(config)
|
| 70 |
+
self.pos_embed = nn.Parameter(torch.zeros((1, num_patches, config.hidden_size)))
|
| 71 |
+
|
| 72 |
+
def forward(self, x: torch.Tensor) -> torch.Tensor:
|
| 73 |
+
tokens = self.patchifier(x)
|
| 74 |
+
_, N, _ = tokens.shape
|
| 75 |
+
pos_embed = self.pos_embed.to(tokens.device)
|
| 76 |
+
tokens = tokens + pos_embed[:, :N]
|
| 77 |
+
return tokens
|
| 78 |
+
|
| 79 |
+
|
| 80 |
+
class AIMv2Attention(nn.Module):
|
| 81 |
+
def __init__(self, config: AIMv2Config):
|
| 82 |
+
super().__init__()
|
| 83 |
+
dim = config.hidden_size
|
| 84 |
+
|
| 85 |
+
self.num_heads = config.num_attention_heads
|
| 86 |
+
self.qkv = nn.Linear(dim, dim * 3, bias=config.qkv_bias)
|
| 87 |
+
self.attn_drop = nn.Dropout(config.attention_dropout)
|
| 88 |
+
self.proj = nn.Linear(dim, dim, bias=config.use_bias)
|
| 89 |
+
self.proj_drop = nn.Dropout(config.projection_dropout)
|
| 90 |
+
|
| 91 |
+
def forward(
|
| 92 |
+
self, x: torch.Tensor, mask: Optional[torch.Tensor] = None
|
| 93 |
+
) -> torch.Tensor:
|
| 94 |
+
B, N, C = x.shape
|
| 95 |
+
qkv = (
|
| 96 |
+
self.qkv(x)
|
| 97 |
+
.reshape(B, N, 3, self.num_heads, C // self.num_heads)
|
| 98 |
+
.permute(2, 0, 3, 1, 4)
|
| 99 |
+
)
|
| 100 |
+
q, k, v = qkv.unbind(0)
|
| 101 |
+
|
| 102 |
+
x = F.scaled_dot_product_attention(q, k, v, attn_mask=mask)
|
| 103 |
+
x = x.transpose(1, 2).contiguous().reshape(B, N, C)
|
| 104 |
+
x = self.proj(x)
|
| 105 |
+
x = self.proj_drop(x)
|
| 106 |
+
return x
|
| 107 |
+
|
| 108 |
+
|
| 109 |
+
class AIMv2Block(nn.Module):
|
| 110 |
+
def __init__(self, config: AIMv2Config):
|
| 111 |
+
super().__init__()
|
| 112 |
+
self.attn = AIMv2Attention(config)
|
| 113 |
+
self.norm_1 = RMSNorm(config.hidden_size, eps=config.rms_norm_eps)
|
| 114 |
+
self.mlp = AIMv2SwiGLUFFN(config)
|
| 115 |
+
self.norm_2 = RMSNorm(config.hidden_size, eps=config.rms_norm_eps)
|
| 116 |
+
|
| 117 |
+
def forward(
|
| 118 |
+
self, x: torch.Tensor, mask: Optional[torch.Tensor] = None
|
| 119 |
+
) -> torch.Tensor:
|
| 120 |
+
x = x + self.attn(self.norm_1(x), mask)
|
| 121 |
+
x = x + self.mlp(self.norm_2(x))
|
| 122 |
+
return x
|
| 123 |
+
|
| 124 |
+
|
| 125 |
+
class AIMv2Transformer(nn.Module):
|
| 126 |
+
def __init__(self, config: AIMv2Config):
|
| 127 |
+
super().__init__()
|
| 128 |
+
self.blocks = nn.ModuleList(
|
| 129 |
+
[AIMv2Block(config) for _ in range(config.num_hidden_layers)]
|
| 130 |
+
)
|
| 131 |
+
self.post_trunk_norm = RMSNorm(config.hidden_size, eps=config.rms_norm_eps)
|
| 132 |
+
|
| 133 |
+
def forward(
|
| 134 |
+
self,
|
| 135 |
+
tokens: torch.Tensor,
|
| 136 |
+
mask: Optional[torch.Tensor] = None,
|
| 137 |
+
output_hidden_states: bool = False,
|
| 138 |
+
) -> Tuple[torch.Tensor, Optional[Tuple[torch.Tensor, ...]]]:
|
| 139 |
+
hidden_states = () if output_hidden_states else None
|
| 140 |
+
for block in self.blocks:
|
| 141 |
+
tokens = block(tokens, mask)
|
| 142 |
+
if output_hidden_states:
|
| 143 |
+
hidden_states += (tokens,)
|
| 144 |
+
tokens = self.post_trunk_norm(tokens)
|
| 145 |
+
return tokens, hidden_states
|
| 146 |
+
|
| 147 |
+
|
| 148 |
+
class AIMv2PretrainedModel(PreTrainedModel):
|
| 149 |
+
config_class = AIMv2Config
|
| 150 |
+
base_model_prefix = "aimv2"
|
| 151 |
+
main_input_name = "pixel_values"
|
| 152 |
+
_supports_sdpa = True
|
| 153 |
+
|
| 154 |
+
|
| 155 |
+
class AIMv2Model(AIMv2PretrainedModel):
|
| 156 |
+
def __init__(self, config: AIMv2Config):
|
| 157 |
+
super().__init__(config)
|
| 158 |
+
self.preprocessor = AIMv2ViTPreprocessor(config)
|
| 159 |
+
self.trunk = AIMv2Transformer(config)
|
| 160 |
+
|
| 161 |
+
def forward(
|
| 162 |
+
self,
|
| 163 |
+
pixel_values: torch.Tensor,
|
| 164 |
+
mask: Optional[torch.Tensor] = None,
|
| 165 |
+
output_hidden_states: Optional[bool] = None,
|
| 166 |
+
return_dict: Optional[bool] = None,
|
| 167 |
+
) -> Union[
|
| 168 |
+
Tuple[torch.Tensor],
|
| 169 |
+
Tuple[torch.Tensor, Tuple[torch.Tensor, ...]],
|
| 170 |
+
BaseModelOutputWithNoAttention,
|
| 171 |
+
]:
|
| 172 |
+
if output_hidden_states is None:
|
| 173 |
+
output_hidden_states = self.config.output_hidden_states
|
| 174 |
+
if return_dict is None:
|
| 175 |
+
return_dict = self.config.use_return_dict
|
| 176 |
+
|
| 177 |
+
x = self.preprocessor(pixel_values)
|
| 178 |
+
x, hidden_states = self.trunk(
|
| 179 |
+
x, mask, output_hidden_states=output_hidden_states
|
| 180 |
+
)
|
| 181 |
+
|
| 182 |
+
if not return_dict:
|
| 183 |
+
res = (x,)
|
| 184 |
+
res += (hidden_states,) if output_hidden_states else ()
|
| 185 |
+
return res
|
| 186 |
+
|
| 187 |
+
return BaseModelOutputWithNoAttention(
|
| 188 |
+
last_hidden_state=x,
|
| 189 |
+
hidden_states=hidden_states,
|
| 190 |
+
)
|
| 191 |
+
|
| 192 |
+
|
| 193 |
+
|
| 194 |
+
from functools import partial
|
| 195 |
+
from torch import nn
|
| 196 |
+
import torch.nn.functional as F
|
| 197 |
+
from transformers.activations import ACT2FN
|
| 198 |
+
import math
|
| 199 |
+
import torch
|
| 200 |
+
|
| 201 |
+
|
| 202 |
+
class GLU(nn.Module):
|
| 203 |
+
def __init__(self, hidden_size, ffn_hidden_size, in_features):
|
| 204 |
+
super().__init__()
|
| 205 |
+
self.linear_proj = nn.Linear(in_features, hidden_size, bias=False)
|
| 206 |
+
self.norm1 = nn.LayerNorm(hidden_size)
|
| 207 |
+
self.act1 = nn.GELU()
|
| 208 |
+
self.act2 = nn.functional.silu
|
| 209 |
+
self.dense_h_to_4h = nn.Linear(hidden_size, ffn_hidden_size, bias=False)
|
| 210 |
+
self.gate_proj = nn.Linear(hidden_size, ffn_hidden_size, bias=False)
|
| 211 |
+
self.dense_4h_to_h = nn.Linear(ffn_hidden_size, hidden_size, bias=False)
|
| 212 |
+
|
| 213 |
+
def forward(self, x):
|
| 214 |
+
x = self.linear_proj(x)
|
| 215 |
+
x = self.act1(self.norm1(x))
|
| 216 |
+
x = self.act2(self.gate_proj(x)) * self.dense_h_to_4h(x)
|
| 217 |
+
x = self.dense_4h_to_h(x)
|
| 218 |
+
return x
|
| 219 |
+
|
| 220 |
+
|
| 221 |
+
class MlpGLU(nn.Module):
|
| 222 |
+
def __init__(self, in_hidden_size, out_hidden_size):
|
| 223 |
+
super(MlpGLU, self).__init__()
|
| 224 |
+
|
| 225 |
+
ffn_hidden_size = out_hidden_size * 4 # out_hidden_size * 4 3584 * 4 = 14336
|
| 226 |
+
self.linear_proj = GLU(
|
| 227 |
+
hidden_size=out_hidden_size,
|
| 228 |
+
ffn_hidden_size=ffn_hidden_size,
|
| 229 |
+
in_features=in_hidden_size,
|
| 230 |
+
)
|
| 231 |
+
|
| 232 |
+
def forward(self, x, attention_mask: torch.Tensor = None):
|
| 233 |
+
x = self.linear_proj(x)
|
| 234 |
+
return x
|
| 235 |
+
|
| 236 |
+
|
| 237 |
+
class PixelShuffleLayer(nn.Module):
|
| 238 |
+
|
| 239 |
+
def __init(self):
|
| 240 |
+
super(PixelShuffleLayer, self).__init__()
|
| 241 |
+
|
| 242 |
+
def forward(self, x, scale_factor=0.5):
|
| 243 |
+
# print(f'in pixelshuffle: {x.shape}')
|
| 244 |
+
n, w, h, c = x.size()
|
| 245 |
+
# N, W, H, C --> N, W, H * scale, C // scale
|
| 246 |
+
x = x.reshape(n, w, int(h * scale_factor), int(c / scale_factor))
|
| 247 |
+
# N, W, H * scale, C // scale --> N, H * scale, W, C // scale
|
| 248 |
+
x = x.permute(0, 2, 1, 3).contiguous()
|
| 249 |
+
# N, H * scale, W, C // scale --> N, H * scale, W * scale, C // (scale ** 2)
|
| 250 |
+
x = x.view(n, int(h * scale_factor), int(w * scale_factor),
|
| 251 |
+
int(c / (scale_factor * scale_factor)))
|
| 252 |
+
x = x.permute(0, 2, 1, 3).contiguous()
|
| 253 |
+
return x
|
| 254 |
+
|
| 255 |
+
|
| 256 |
+
|
| 257 |
+
class PixelShuffleConnector(nn.Module):
|
| 258 |
+
|
| 259 |
+
def __init__(self, in_hidden_size, out_hidden_size, down_rate=2):
|
| 260 |
+
super(PixelShuffleConnector, self).__init__()
|
| 261 |
+
# ffn_hidden_size = 13696
|
| 262 |
+
ffn_hidden_size = out_hidden_size * 4 # out_hidden_size * 4 3584 * 4 = 14336
|
| 263 |
+
self.linear_proj = GLU(
|
| 264 |
+
hidden_size=out_hidden_size,
|
| 265 |
+
ffn_hidden_size=ffn_hidden_size,
|
| 266 |
+
in_features=in_hidden_size * 4,
|
| 267 |
+
)
|
| 268 |
+
self.down_rate = down_rate
|
| 269 |
+
if self.down_rate == 2:
|
| 270 |
+
down = PixelShuffleLayer()
|
| 271 |
+
self.downsample = nn.Sequential(*[down])
|
| 272 |
+
else:
|
| 273 |
+
print(f"unsupported downsample rate for now!")
|
| 274 |
+
self.scaling_factor = 8
|
| 275 |
+
|
| 276 |
+
|
| 277 |
+
def forward(self, x, attention_mask: torch.Tensor = None):
|
| 278 |
+
# print(f'xin: {x.shape}')
|
| 279 |
+
b, s, h = x.shape
|
| 280 |
+
grid_size = int(s**0.5)
|
| 281 |
+
x = x.reshape(b, grid_size, grid_size, h)
|
| 282 |
+
x = self.downsample(x)
|
| 283 |
+
# print(f'x: {x.shape}')
|
| 284 |
+
# [11, 16, 16, 4608]
|
| 285 |
+
x = x.reshape(x.shape[0], -1, x.shape[-1])
|
| 286 |
+
x = self.linear_proj(x)
|
| 287 |
+
return x
|
modeling_mono.py
ADDED
|
@@ -0,0 +1,370 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
from dataclasses import dataclass
|
| 2 |
+
from typing import List, Optional, Tuple, Union
|
| 3 |
+
import torch
|
| 4 |
+
from torch.nn import CrossEntropyLoss
|
| 5 |
+
|
| 6 |
+
from transformers import Qwen2Config, Qwen2Model, Qwen2ForCausalLM, PreTrainedModel
|
| 7 |
+
from transformers.modeling_outputs import CausalLMOutputWithPast, ModelOutput
|
| 8 |
+
from torch import nn
|
| 9 |
+
import torch.nn.functional as F
|
| 10 |
+
from .configuration_aimv2 import MonoConfig
|
| 11 |
+
from .modeling_aimv2 import AIMv2Model, PixelShuffleConnector
|
| 12 |
+
from transformers.generation import GenerationMixin
|
| 13 |
+
|
| 14 |
+
"""
|
| 15 |
+
|
| 16 |
+
Simple arch of Mono, used for pretrain vision encoder.
|
| 17 |
+
|
| 18 |
+
"""
|
| 19 |
+
|
| 20 |
+
|
| 21 |
+
@dataclass
|
| 22 |
+
class MonoCausalLMOutputWithPast(ModelOutput):
|
| 23 |
+
|
| 24 |
+
loss: Optional[torch.FloatTensor] = None
|
| 25 |
+
logits: torch.FloatTensor = None
|
| 26 |
+
past_key_values: Optional[List[torch.FloatTensor]] = None
|
| 27 |
+
hidden_states: Optional[Tuple[torch.FloatTensor]] = None
|
| 28 |
+
attentions: Optional[Tuple[torch.FloatTensor]] = None
|
| 29 |
+
image_hidden_states: Optional[Tuple[torch.FloatTensor]] = None
|
| 30 |
+
|
| 31 |
+
|
| 32 |
+
class MonoPretrainedModel(PreTrainedModel):
|
| 33 |
+
config_class = MonoConfig
|
| 34 |
+
base_model_prefix = "mono"
|
| 35 |
+
# main_input_name = "pixel_values"
|
| 36 |
+
_supports_sdpa = True
|
| 37 |
+
_supports_flash_attn_2 = True
|
| 38 |
+
_supports_cache_class = True
|
| 39 |
+
supports_gradient_checkpointing = True
|
| 40 |
+
|
| 41 |
+
|
| 42 |
+
# class MonoForConditionalGeneration(MonoPretrainedModel, Qwen2ForCausalLM):
|
| 43 |
+
class MonoForConditionalGeneration(MonoPretrainedModel, GenerationMixin):
|
| 44 |
+
_tied_weights_keys = ["lm_head.weight"]
|
| 45 |
+
|
| 46 |
+
def __init__(self, config: MonoConfig):
|
| 47 |
+
# super().__init__(config)
|
| 48 |
+
MonoPretrainedModel.__init__(self, config)
|
| 49 |
+
# super(Qwen2ForCausalLM, self).__init__(config)
|
| 50 |
+
|
| 51 |
+
self.vision_tower = AIMv2Model(config=config.vision_config)
|
| 52 |
+
self._attn_implementation = config._attn_implementation
|
| 53 |
+
|
| 54 |
+
self._build_image_projection_layers(config)
|
| 55 |
+
|
| 56 |
+
self.model = Qwen2Model(config)
|
| 57 |
+
self.vocab_size = config.vocab_size
|
| 58 |
+
self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
|
| 59 |
+
|
| 60 |
+
self.pad_token_id = config.pad_token_id
|
| 61 |
+
print(f"==> pad_token_id: {self.pad_token_id}")
|
| 62 |
+
self.post_init()
|
| 63 |
+
|
| 64 |
+
def _build_image_projection_layers(self, config):
|
| 65 |
+
image_dim_out = config.vision_config.hidden_size
|
| 66 |
+
dim_projection = config.hidden_size
|
| 67 |
+
# self.mm_projector = nn.Linear(image_dim_out, dim_projection)
|
| 68 |
+
self.mm_projector = PixelShuffleConnector(image_dim_out, dim_projection)
|
| 69 |
+
print(f"==> build mm_projector: {image_dim_out} -> {dim_projection}")
|
| 70 |
+
|
| 71 |
+
def get_vision_tower(self):
|
| 72 |
+
return self.vision_tower
|
| 73 |
+
|
| 74 |
+
def get_input_embeddings(self):
|
| 75 |
+
return self.model.get_input_embeddings()
|
| 76 |
+
|
| 77 |
+
def resize_token_embeddings(
|
| 78 |
+
self, new_num_tokens: Optional[int] = None, pad_to_multiple_of=None
|
| 79 |
+
) -> nn.Embedding:
|
| 80 |
+
model_embeds = self.model.resize_token_embeddings(
|
| 81 |
+
new_num_tokens, pad_to_multiple_of
|
| 82 |
+
)
|
| 83 |
+
# update vocab size
|
| 84 |
+
self.config.text_config.vocab_size = model_embeds.num_embeddings
|
| 85 |
+
self.config.vocab_size = model_embeds.num_embeddings
|
| 86 |
+
self.vocab_size = model_embeds.num_embeddings
|
| 87 |
+
return model_embeds
|
| 88 |
+
|
| 89 |
+
def _encode_image(self, pixel_values):
|
| 90 |
+
# print(f"pixel_values: {pixel_values}")
|
| 91 |
+
batch_size, C, H, W = pixel_values.shape
|
| 92 |
+
x = self.vision_tower(pixel_values, output_hidden_states=True)
|
| 93 |
+
x = x[-2]
|
| 94 |
+
# print(x)
|
| 95 |
+
x = self.mm_projector(x)
|
| 96 |
+
# print(f"image features: {x}")
|
| 97 |
+
return x
|
| 98 |
+
|
| 99 |
+
def forward(
|
| 100 |
+
self,
|
| 101 |
+
input_ids: Optional[torch.LongTensor] = None,
|
| 102 |
+
pixel_values: torch.FloatTensor = None,
|
| 103 |
+
attention_mask: Optional[torch.FloatTensor] = None,
|
| 104 |
+
position_ids: Optional[torch.LongTensor] = None,
|
| 105 |
+
past_key_values: Optional[Tuple[Tuple[torch.Tensor]]] = None,
|
| 106 |
+
inputs_embeds: Optional[torch.FloatTensor] = None,
|
| 107 |
+
labels: Optional[torch.LongTensor] = None,
|
| 108 |
+
use_cache: Optional[bool] = None,
|
| 109 |
+
output_attentions: Optional[bool] = None,
|
| 110 |
+
output_hidden_states: Optional[bool] = None,
|
| 111 |
+
return_dict: Optional[bool] = None,
|
| 112 |
+
cache_position=None,
|
| 113 |
+
) -> Union[Tuple, CausalLMOutputWithPast]:
|
| 114 |
+
output_attentions = (
|
| 115 |
+
output_attentions
|
| 116 |
+
if output_attentions is not None
|
| 117 |
+
else self.config.output_attentions
|
| 118 |
+
)
|
| 119 |
+
output_hidden_states = (
|
| 120 |
+
output_hidden_states
|
| 121 |
+
if output_hidden_states is not None
|
| 122 |
+
else self.config.output_hidden_states
|
| 123 |
+
)
|
| 124 |
+
return_dict = (
|
| 125 |
+
return_dict if return_dict is not None else self.config.use_return_dict
|
| 126 |
+
)
|
| 127 |
+
|
| 128 |
+
image_features = None
|
| 129 |
+
if inputs_embeds is None:
|
| 130 |
+
if pixel_values is not None:
|
| 131 |
+
# (batch_size, num_image_tokens, hidden_size)
|
| 132 |
+
image_features = self._encode_image(pixel_values)
|
| 133 |
+
|
| 134 |
+
if input_ids is not None:
|
| 135 |
+
inputs_embeds, attention_mask, labels = (
|
| 136 |
+
self._get_input_embeds_with_image(input_ids, image_features, labels)
|
| 137 |
+
)
|
| 138 |
+
|
| 139 |
+
# print(f'before inputs_embeds: {inputs_embeds.shape}')
|
| 140 |
+
# print(f'before labels: {labels.shape}')
|
| 141 |
+
|
| 142 |
+
# padding all to normal sequence length only train
|
| 143 |
+
# if labels is not None:
|
| 144 |
+
# input_length = inputs_embeds.shape[1]
|
| 145 |
+
# label_length = labels.shape[1]
|
| 146 |
+
|
| 147 |
+
# if labels is not None:
|
| 148 |
+
# labels = F.pad(labels, (input_length, 0), value=-100)
|
| 149 |
+
|
| 150 |
+
# if inputs_embeds is not None:
|
| 151 |
+
# # append embeds and attn_mask to labels length
|
| 152 |
+
# padding = torch.zeros(
|
| 153 |
+
# inputs_embeds.shape[0],
|
| 154 |
+
# label_length,
|
| 155 |
+
# inputs_embeds.shape[2],
|
| 156 |
+
# dtype=inputs_embeds.dtype,
|
| 157 |
+
# device=inputs_embeds.device,
|
| 158 |
+
# )
|
| 159 |
+
# inputs_embeds = torch.cat([inputs_embeds, padding], dim=1)
|
| 160 |
+
# attention_mask = attention_mask.to(inputs_embeds.dtype)
|
| 161 |
+
# attention_mask = F.pad(attention_mask, (0, label_length), value=0)
|
| 162 |
+
|
| 163 |
+
# if position_ids is None:
|
| 164 |
+
# position_ids = torch.arange(
|
| 165 |
+
# input_length + label_length, device=inputs_embeds.device
|
| 166 |
+
# )
|
| 167 |
+
# position_ids = position_ids.unsqueeze(0).expand(
|
| 168 |
+
# inputs_embeds.shape[0], -1
|
| 169 |
+
# )
|
| 170 |
+
# position_ids[input_length:] = 0
|
| 171 |
+
|
| 172 |
+
# print(f"position_ids {position_ids}")
|
| 173 |
+
# print(f"labels {labels.shape}")
|
| 174 |
+
# print(f"labels {labels}")
|
| 175 |
+
# print(f"inputs_embeds {inputs_embeds.shape}")
|
| 176 |
+
# print(f"inputs_embeds {inputs_embeds}")
|
| 177 |
+
# print(f"attention_mask {attention_mask.shape}")
|
| 178 |
+
# print(f"attention_mask {attention_mask}")
|
| 179 |
+
|
| 180 |
+
outputs = self.model(
|
| 181 |
+
input_ids=None,
|
| 182 |
+
attention_mask=attention_mask,
|
| 183 |
+
position_ids=position_ids,
|
| 184 |
+
past_key_values=past_key_values,
|
| 185 |
+
inputs_embeds=inputs_embeds,
|
| 186 |
+
use_cache=use_cache,
|
| 187 |
+
output_attentions=output_attentions,
|
| 188 |
+
output_hidden_states=output_hidden_states,
|
| 189 |
+
return_dict=return_dict,
|
| 190 |
+
)
|
| 191 |
+
|
| 192 |
+
hidden_states = outputs[0]
|
| 193 |
+
logits = self.lm_head(hidden_states)
|
| 194 |
+
|
| 195 |
+
loss = None
|
| 196 |
+
if labels is not None:
|
| 197 |
+
# Upcast to float if we need to compute the loss to avoid potential precision issues
|
| 198 |
+
logits = logits.float()
|
| 199 |
+
labels = labels.to(logits.device)
|
| 200 |
+
# Shift so that tokens < n predict n
|
| 201 |
+
if attention_mask is not None:
|
| 202 |
+
# we use the input attention mask to shift the logits and labels, because it is 2D.
|
| 203 |
+
# we also crop attn mask in case it is longer, which happens in PrefixTuning with peft
|
| 204 |
+
shift_attention_mask = attention_mask[:, -(logits.shape[1] - 1) :].to(
|
| 205 |
+
logits.device
|
| 206 |
+
)
|
| 207 |
+
shift_logits = logits[..., :-1, :][
|
| 208 |
+
shift_attention_mask != 0
|
| 209 |
+
].contiguous()
|
| 210 |
+
# print(f"shift_logits: {shift_logits.shape}")
|
| 211 |
+
shift_labels = labels[..., 1:][shift_attention_mask != 0].contiguous()
|
| 212 |
+
# print(f"shift_labels: {shift_labels.shape}")
|
| 213 |
+
else:
|
| 214 |
+
shift_logits = logits[..., :-1, :].contiguous()
|
| 215 |
+
shift_labels = labels[..., 1:].contiguous()
|
| 216 |
+
# Flatten the tokens
|
| 217 |
+
loss_fct = CrossEntropyLoss()
|
| 218 |
+
loss = loss_fct(
|
| 219 |
+
shift_logits.view(-1, shift_logits.size(-1)), shift_labels.view(-1)
|
| 220 |
+
)
|
| 221 |
+
|
| 222 |
+
if not return_dict:
|
| 223 |
+
output = (logits,) + outputs[1:]
|
| 224 |
+
return (loss,) + output if loss is not None else output
|
| 225 |
+
|
| 226 |
+
return MonoCausalLMOutputWithPast(
|
| 227 |
+
loss=loss,
|
| 228 |
+
logits=logits,
|
| 229 |
+
past_key_values=outputs.past_key_values,
|
| 230 |
+
hidden_states=outputs.hidden_states,
|
| 231 |
+
attentions=outputs.attentions,
|
| 232 |
+
)
|
| 233 |
+
|
| 234 |
+
def _get_input_embeds_with_image(self, input_ids, image_features, labels=None):
|
| 235 |
+
# 1. replace image token with features; 2. replace -100 in input_ids into zeroes
|
| 236 |
+
# 3. handling right attention_mask
|
| 237 |
+
# not complicated, you can understand.
|
| 238 |
+
batch_size = input_ids.size(0)
|
| 239 |
+
processed_embeds = []
|
| 240 |
+
processed_masks = []
|
| 241 |
+
labels_ignored_im = []
|
| 242 |
+
|
| 243 |
+
max_seq_len = 0
|
| 244 |
+
for idx in range(batch_size):
|
| 245 |
+
seq = input_ids[idx]
|
| 246 |
+
im_pos = (seq == -200).nonzero(as_tuple=True)[0]
|
| 247 |
+
|
| 248 |
+
if im_pos.numel() > 0:
|
| 249 |
+
im_pos = im_pos.item()
|
| 250 |
+
before = seq[:im_pos]
|
| 251 |
+
after = seq[im_pos + 1 :]
|
| 252 |
+
# Exclude -100 tokens (maybe, input_ids padding with -100 intentionly)
|
| 253 |
+
before = before[before != -100]
|
| 254 |
+
after = after[after != -100]
|
| 255 |
+
# Get embeddings for before and after
|
| 256 |
+
before_embed = self.get_input_embeddings()(before)
|
| 257 |
+
after_embed = self.get_input_embeddings()(after)
|
| 258 |
+
# Concatenate before, image features, and after
|
| 259 |
+
seq_embed = torch.cat(
|
| 260 |
+
[before_embed, image_features[idx], after_embed], dim=0
|
| 261 |
+
)
|
| 262 |
+
new_seq_len = seq_embed.size(0)
|
| 263 |
+
|
| 264 |
+
# if labels not None, change image token into -100, keep image tokens length
|
| 265 |
+
if labels is not None:
|
| 266 |
+
image_token_ignore = torch.full(
|
| 267 |
+
(image_features[idx].shape[0],),
|
| 268 |
+
-100,
|
| 269 |
+
dtype=torch.long,
|
| 270 |
+
device=labels.device,
|
| 271 |
+
)
|
| 272 |
+
labels_ignored_im.append(
|
| 273 |
+
torch.cat(
|
| 274 |
+
(
|
| 275 |
+
labels[idx][:im_pos],
|
| 276 |
+
image_token_ignore,
|
| 277 |
+
labels[idx][im_pos + 1 :],
|
| 278 |
+
),
|
| 279 |
+
dim=0,
|
| 280 |
+
)
|
| 281 |
+
)
|
| 282 |
+
|
| 283 |
+
else:
|
| 284 |
+
# Exclude -100 tokens
|
| 285 |
+
valid_tokens = seq[seq != -100]
|
| 286 |
+
seq_embed = self.get_input_embeddings()(valid_tokens)
|
| 287 |
+
new_seq_len = seq_embed.size(0)
|
| 288 |
+
|
| 289 |
+
# Update the maximum sequence length
|
| 290 |
+
if new_seq_len > max_seq_len:
|
| 291 |
+
max_seq_len = new_seq_len
|
| 292 |
+
|
| 293 |
+
processed_embeds.append(seq_embed)
|
| 294 |
+
attn_mask = torch.ones(new_seq_len, dtype=torch.bool, device=seq.device)
|
| 295 |
+
processed_masks.append(attn_mask)
|
| 296 |
+
|
| 297 |
+
# rest embedding is 0, rest mask is False, just padding it
|
| 298 |
+
inputs_embeds = torch.nn.utils.rnn.pad_sequence(
|
| 299 |
+
processed_embeds, batch_first=True, padding_value=0.0
|
| 300 |
+
)
|
| 301 |
+
attn_masks = torch.nn.utils.rnn.pad_sequence(
|
| 302 |
+
processed_masks, batch_first=True, padding_value=0
|
| 303 |
+
)
|
| 304 |
+
if labels is not None:
|
| 305 |
+
labels_ignored_im = torch.stack(labels_ignored_im, dim=0)
|
| 306 |
+
return inputs_embeds, attn_masks, labels_ignored_im
|
| 307 |
+
return inputs_embeds, attn_masks, None
|
| 308 |
+
|
| 309 |
+
@torch.no_grad()
|
| 310 |
+
def generate(self, input_ids, pixel_values=None, **kwargs):
|
| 311 |
+
# print(input_ids)
|
| 312 |
+
# print(f"pixel_values {pixel_values}")
|
| 313 |
+
if pixel_values is not None:
|
| 314 |
+
image_features = self._encode_image(pixel_values)
|
| 315 |
+
# print(f"image_features {image_features}")
|
| 316 |
+
inputs_embeds, attention_mask, _ = self._get_input_embeds_with_image(
|
| 317 |
+
input_ids, image_features
|
| 318 |
+
)
|
| 319 |
+
else:
|
| 320 |
+
if input_ids is not None:
|
| 321 |
+
inputs_embeds = self.get_input_embeddings()(input_ids)
|
| 322 |
+
attention_mask = torch.ones(
|
| 323 |
+
inputs_embeds.size(0),
|
| 324 |
+
inputs_embeds.size(1),
|
| 325 |
+
dtype=torch.bool,
|
| 326 |
+
device=inputs_embeds.device,
|
| 327 |
+
)
|
| 328 |
+
|
| 329 |
+
# print(f"inputs_embeds: {inputs_embeds}")
|
| 330 |
+
return super().generate(
|
| 331 |
+
input_ids=None,
|
| 332 |
+
inputs_embeds=inputs_embeds,
|
| 333 |
+
attention_mask=attention_mask,
|
| 334 |
+
**kwargs,
|
| 335 |
+
)
|
| 336 |
+
|
| 337 |
+
def prepare_inputs_for_generation(
|
| 338 |
+
self,
|
| 339 |
+
input_ids,
|
| 340 |
+
past_key_values=None,
|
| 341 |
+
inputs_embeds=None,
|
| 342 |
+
attention_mask=None,
|
| 343 |
+
**kwargs,
|
| 344 |
+
):
|
| 345 |
+
# cut input_ids if past_key_values is used
|
| 346 |
+
# if past_key_values is not None:
|
| 347 |
+
# past_length = past_key_values[0][0].shape[2]
|
| 348 |
+
|
| 349 |
+
# # Some generation methods already pass only the last input ID
|
| 350 |
+
# if input_ids.shape[1] > past_length:
|
| 351 |
+
# input_ids = input_ids[:, -1:]
|
| 352 |
+
# elif input_ids.shape[1] == 1:
|
| 353 |
+
# pass
|
| 354 |
+
# else:
|
| 355 |
+
# # Default to old behavior: keep only final ID
|
| 356 |
+
# input_ids = input_ids[:, -1:]
|
| 357 |
+
|
| 358 |
+
model_inputs = super().prepare_inputs_for_generation(
|
| 359 |
+
input_ids,
|
| 360 |
+
past_key_values=past_key_values,
|
| 361 |
+
inputs_embeds=inputs_embeds,
|
| 362 |
+
**kwargs,
|
| 363 |
+
)
|
| 364 |
+
return model_inputs
|
| 365 |
+
|
| 366 |
+
def prepare_decoder_input_ids_from_labels(self, labels: torch.Tensor):
|
| 367 |
+
return self.model.shift_tokens_right(labels)
|
| 368 |
+
|
| 369 |
+
def _reorder_cache(self, *args, **kwargs):
|
| 370 |
+
return self.model._reorder_cache(*args, **kwargs)
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_map": {
|
| 3 |
+
"AutoProcessor": "processing_mono.MonoProcessor"
|
| 4 |
+
},
|
| 5 |
+
"crop_size": {
|
| 6 |
+
"height": 448,
|
| 7 |
+
"width": 448
|
| 8 |
+
},
|
| 9 |
+
"do_center_crop": true,
|
| 10 |
+
"do_convert_rgb": true,
|
| 11 |
+
"do_normalize": true,
|
| 12 |
+
"do_rescale": true,
|
| 13 |
+
"do_resize": true,
|
| 14 |
+
"image_mean": [
|
| 15 |
+
0.48145466,
|
| 16 |
+
0.4578275,
|
| 17 |
+
0.40821073
|
| 18 |
+
],
|
| 19 |
+
"image_processor_type": "CLIPImageProcessor",
|
| 20 |
+
"image_seq_length": 577,
|
| 21 |
+
"image_std": [
|
| 22 |
+
0.26862954,
|
| 23 |
+
0.26130258,
|
| 24 |
+
0.27577711
|
| 25 |
+
],
|
| 26 |
+
"processor_class": "MonoProcessor",
|
| 27 |
+
"resample": 3,
|
| 28 |
+
"rescale_factor": 0.00392156862745098,
|
| 29 |
+
"size": {
|
| 30 |
+
"shortest_edge": 448
|
| 31 |
+
}
|
| 32 |
+
}
|
processing_mono.py
ADDED
|
@@ -0,0 +1,259 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import re
|
| 2 |
+
import logging
|
| 3 |
+
from typing import List, Optional, Union
|
| 4 |
+
import numpy as np
|
| 5 |
+
|
| 6 |
+
import torch
|
| 7 |
+
|
| 8 |
+
from transformers.feature_extraction_utils import BatchFeature
|
| 9 |
+
from transformers.image_utils import ImageInput, is_valid_image
|
| 10 |
+
from transformers.processing_utils import ProcessorMixin
|
| 11 |
+
from transformers.tokenization_utils_base import (
|
| 12 |
+
PaddingStrategy,
|
| 13 |
+
PreTokenizedInput,
|
| 14 |
+
TextInput,
|
| 15 |
+
TruncationStrategy,
|
| 16 |
+
)
|
| 17 |
+
from transformers.utils import TensorType
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
logger = logging.getLogger(__name__)
|
| 21 |
+
|
| 22 |
+
|
| 23 |
+
# Copied from transformers.models.idefics2.processing_idefics2.is_url
|
| 24 |
+
def is_url(val) -> bool:
|
| 25 |
+
return isinstance(val, str) and val.startswith("http")
|
| 26 |
+
|
| 27 |
+
|
| 28 |
+
# Copied from transformers.models.idefics2.processing_idefics2.is_image_or_image_url
|
| 29 |
+
def is_image_or_image_url(elem):
|
| 30 |
+
return is_url(elem) or is_valid_image(elem)
|
| 31 |
+
|
| 32 |
+
|
| 33 |
+
def _is_str_or_image(elem):
|
| 34 |
+
return isinstance(elem, (str)) or is_image_or_image_url(elem)
|
| 35 |
+
|
| 36 |
+
|
| 37 |
+
class MonoProcessor(ProcessorMixin):
|
| 38 |
+
|
| 39 |
+
attributes = ["image_processor", "tokenizer"]
|
| 40 |
+
image_processor_class = "CLIPImageProcessor"
|
| 41 |
+
# tokenizer_class = ("BartTokenizer", "BartTokenizerFast")
|
| 42 |
+
tokenizer_class = ("Qwen2Tokenizer", "Qwen2TokenizerFast")
|
| 43 |
+
|
| 44 |
+
def __init__(
|
| 45 |
+
self,
|
| 46 |
+
image_processor=None,
|
| 47 |
+
tokenizer=None,
|
| 48 |
+
):
|
| 49 |
+
if image_processor is None:
|
| 50 |
+
raise ValueError("You need to specify an `image_processor`.")
|
| 51 |
+
if tokenizer is None:
|
| 52 |
+
raise ValueError("You need to specify a `tokenizer`.")
|
| 53 |
+
|
| 54 |
+
tokens_to_add = {
|
| 55 |
+
"additional_special_tokens": tokenizer.additional_special_tokens
|
| 56 |
+
+ ["<od>", "</od>", "<ocr>", "</ocr>"]
|
| 57 |
+
+ [f"<loc_{x}>" for x in range(1000)]
|
| 58 |
+
+ [
|
| 59 |
+
"<cap>",
|
| 60 |
+
"</cap>",
|
| 61 |
+
"<ncap>",
|
| 62 |
+
"</ncap>",
|
| 63 |
+
"<dcap>",
|
| 64 |
+
"</dcap>",
|
| 65 |
+
"<grounding>",
|
| 66 |
+
"</grounding>",
|
| 67 |
+
"<seg>",
|
| 68 |
+
"</seg>",
|
| 69 |
+
"<sep>",
|
| 70 |
+
"<region_cap>",
|
| 71 |
+
"</region_cap>",
|
| 72 |
+
"<region_to_desciption>",
|
| 73 |
+
"</region_to_desciption>",
|
| 74 |
+
"<proposal>",
|
| 75 |
+
"</proposal>",
|
| 76 |
+
"<poly>",
|
| 77 |
+
"</poly>",
|
| 78 |
+
"<and>",
|
| 79 |
+
]
|
| 80 |
+
}
|
| 81 |
+
tokenizer.add_special_tokens(tokens_to_add)
|
| 82 |
+
|
| 83 |
+
self.tasks_answer_post_processing_type = {
|
| 84 |
+
"<OCR>": "pure_text",
|
| 85 |
+
"<OCR_WITH_REGION>": "ocr",
|
| 86 |
+
"<CAPTION>": "pure_text",
|
| 87 |
+
"<DETAILED_CAPTION>": "pure_text",
|
| 88 |
+
"<MORE_DETAILED_CAPTION>": "pure_text",
|
| 89 |
+
"<OD>": "description_with_bboxes",
|
| 90 |
+
"<DENSE_REGION_CAPTION>": "description_with_bboxes",
|
| 91 |
+
"<CAPTION_TO_PHRASE_GROUNDING>": "phrase_grounding",
|
| 92 |
+
"<REFERRING_EXPRESSION_SEGMENTATION>": "polygons",
|
| 93 |
+
"<REGION_TO_SEGMENTATION>": "polygons",
|
| 94 |
+
"<OPEN_VOCABULARY_DETECTION>": "description_with_bboxes_or_polygons",
|
| 95 |
+
"<REGION_TO_CATEGORY>": "pure_text",
|
| 96 |
+
"<REGION_TO_DESCRIPTION>": "pure_text",
|
| 97 |
+
"<REGION_TO_OCR>": "pure_text",
|
| 98 |
+
"<REGION_PROPOSAL>": "bboxes",
|
| 99 |
+
}
|
| 100 |
+
|
| 101 |
+
self.task_prompts_without_inputs = {
|
| 102 |
+
"<OCR>": "What is the text in the image?",
|
| 103 |
+
"<OCR_WITH_REGION>": "What is the text in the image, with regions?",
|
| 104 |
+
"<CAPTION>": "What does the image describe?",
|
| 105 |
+
"<DETAILED_CAPTION>": "Describe in detail what is shown in the image.",
|
| 106 |
+
"<MORE_DETAILED_CAPTION>": "Describe with a paragraph what is shown in the image.",
|
| 107 |
+
"<OD>": "Locate the objects with category name in the image.",
|
| 108 |
+
"<DENSE_REGION_CAPTION>": "Locate the objects in the image, with their descriptions.",
|
| 109 |
+
"<REGION_PROPOSAL>": "Locate the region proposals in the image.",
|
| 110 |
+
}
|
| 111 |
+
|
| 112 |
+
self.task_prompts_with_input = {
|
| 113 |
+
"<CAPTION_TO_PHRASE_GROUNDING>": "Locate the phrases in the caption: {input}",
|
| 114 |
+
"<REFERRING_EXPRESSION_SEGMENTATION>": "Locate {input} in the image with mask",
|
| 115 |
+
"<REGION_TO_SEGMENTATION>": "What is the polygon mask of region {input}",
|
| 116 |
+
"<OPEN_VOCABULARY_DETECTION>": "Locate {input} in the image.",
|
| 117 |
+
"<REGION_TO_CATEGORY>": "What is the region {input}?",
|
| 118 |
+
"<REGION_TO_DESCRIPTION>": "What does the region {input} describe?",
|
| 119 |
+
"<REGION_TO_OCR>": "What text is in the region {input}?",
|
| 120 |
+
}
|
| 121 |
+
|
| 122 |
+
super().__init__(image_processor, tokenizer)
|
| 123 |
+
|
| 124 |
+
def construct_prompts(self, text):
|
| 125 |
+
# replace the task tokens with the task prompts if task token is in the text
|
| 126 |
+
if isinstance(text, str):
|
| 127 |
+
for task_token, task_prompt in self.task_prompts_without_inputs.items():
|
| 128 |
+
if task_token in text:
|
| 129 |
+
_text = task_prompt
|
| 130 |
+
break
|
| 131 |
+
return _text
|
| 132 |
+
prompts = []
|
| 133 |
+
for _text in text:
|
| 134 |
+
# 1. fixed task prompts without additional inputs
|
| 135 |
+
for task_token, task_prompt in self.task_prompts_without_inputs.items():
|
| 136 |
+
if task_token in _text:
|
| 137 |
+
assert (
|
| 138 |
+
_text == task_token
|
| 139 |
+
), f"Task token {task_token} should be the only token in the text."
|
| 140 |
+
_text = task_prompt
|
| 141 |
+
break
|
| 142 |
+
# 2. task prompts with additional inputs
|
| 143 |
+
for task_token, task_prompt in self.task_prompts_with_input.items():
|
| 144 |
+
if task_token in _text:
|
| 145 |
+
_text = task_prompt.format(input=_text.replace(task_token, ""))
|
| 146 |
+
break
|
| 147 |
+
prompts.append(_text)
|
| 148 |
+
return prompts
|
| 149 |
+
|
| 150 |
+
def __call__(
|
| 151 |
+
self,
|
| 152 |
+
text: Union[
|
| 153 |
+
TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]
|
| 154 |
+
] = None,
|
| 155 |
+
images: ImageInput = None,
|
| 156 |
+
tokenize_newline_separately: bool = True,
|
| 157 |
+
padding: Union[bool, str, PaddingStrategy] = False,
|
| 158 |
+
truncation: Union[bool, str, TruncationStrategy] = None,
|
| 159 |
+
max_length=None,
|
| 160 |
+
return_tensors: Optional[Union[str, TensorType]] = TensorType.PYTORCH,
|
| 161 |
+
do_resize: bool = None,
|
| 162 |
+
size=None,
|
| 163 |
+
do_normalize: bool = None,
|
| 164 |
+
image_mean: Optional[Union[float, List[float]]] = None,
|
| 165 |
+
image_std: Optional[Union[float, List[float]]] = None,
|
| 166 |
+
data_format: Optional["ChannelDimension"] = "channels_first", # noqa: F821
|
| 167 |
+
input_data_format: Optional[
|
| 168 |
+
Union[str, "ChannelDimension"] # noqa: F821
|
| 169 |
+
] = None,
|
| 170 |
+
resample: "PILImageResampling" = None, # noqa: F821
|
| 171 |
+
do_convert_rgb: bool = None,
|
| 172 |
+
do_thumbnail: bool = None,
|
| 173 |
+
do_align_long_axis: bool = None,
|
| 174 |
+
do_rescale: bool = None,
|
| 175 |
+
) -> BatchFeature:
|
| 176 |
+
return_token_type_ids = False
|
| 177 |
+
|
| 178 |
+
if text is None:
|
| 179 |
+
logger.warning_once("You are using Florence-2 without a text prompt.")
|
| 180 |
+
text = ""
|
| 181 |
+
|
| 182 |
+
if isinstance(text, List) and isinstance(images, List):
|
| 183 |
+
if len(images) < len(text):
|
| 184 |
+
raise ValueError(
|
| 185 |
+
f"Received {len(images)} images for {len(text)} prompts. Each prompt should be associated with an image."
|
| 186 |
+
)
|
| 187 |
+
if _is_str_or_image(text):
|
| 188 |
+
text = [text]
|
| 189 |
+
elif isinstance(text, list) and _is_str_or_image(text[0]):
|
| 190 |
+
pass
|
| 191 |
+
|
| 192 |
+
if images is not None:
|
| 193 |
+
pixel_values = self.image_processor(
|
| 194 |
+
images,
|
| 195 |
+
size=size,
|
| 196 |
+
do_resize=do_resize,
|
| 197 |
+
do_normalize=do_normalize,
|
| 198 |
+
return_tensors=return_tensors,
|
| 199 |
+
image_mean=image_mean,
|
| 200 |
+
image_std=image_std,
|
| 201 |
+
input_data_format=input_data_format,
|
| 202 |
+
data_format=data_format,
|
| 203 |
+
resample=resample,
|
| 204 |
+
do_convert_rgb=do_convert_rgb,
|
| 205 |
+
)["pixel_values"]
|
| 206 |
+
|
| 207 |
+
# text = self.construct_prompts(text)
|
| 208 |
+
|
| 209 |
+
inputs = self.tokenizer(
|
| 210 |
+
text,
|
| 211 |
+
return_tensors=return_tensors,
|
| 212 |
+
padding=padding,
|
| 213 |
+
max_length=max_length,
|
| 214 |
+
truncation=truncation,
|
| 215 |
+
return_token_type_ids=return_token_type_ids,
|
| 216 |
+
)
|
| 217 |
+
|
| 218 |
+
if images is not None:
|
| 219 |
+
# print(inputs)
|
| 220 |
+
# add IMAGE_TOKEN
|
| 221 |
+
inputs_with_image = [
|
| 222 |
+
torch.cat((torch.tensor([-200]), b), dim=0) for b in inputs["input_ids"]
|
| 223 |
+
]
|
| 224 |
+
# inputs["input_ids"] = torch.stack(inputs_with_image)
|
| 225 |
+
inputs["input_ids"] = inputs_with_image
|
| 226 |
+
|
| 227 |
+
return_data = {**inputs, "pixel_values": pixel_values}
|
| 228 |
+
else:
|
| 229 |
+
return_data = {**inputs, "pixel_values": None}
|
| 230 |
+
|
| 231 |
+
if return_token_type_ids:
|
| 232 |
+
labels = inputs["input_ids"].masked_fill(
|
| 233 |
+
inputs["token_type_ids"] == 0, -100
|
| 234 |
+
)
|
| 235 |
+
return_data.update({"labels": labels})
|
| 236 |
+
return BatchFeature(data=return_data)
|
| 237 |
+
|
| 238 |
+
# Copied from transformers.models.clip.processing_clip.CLIPProcessor.batch_decode with CLIP->Florence2
|
| 239 |
+
def batch_decode(self, *args, **kwargs):
|
| 240 |
+
"""
|
| 241 |
+
This method forwards all its arguments to BartTokenizerFast's [`~PreTrainedTokenizer.batch_decode`]. Please
|
| 242 |
+
refer to the docstring of this method for more information.
|
| 243 |
+
"""
|
| 244 |
+
return self.tokenizer.batch_decode(*args, **kwargs)
|
| 245 |
+
|
| 246 |
+
# Copied from transformers.models.clip.processing_clip.CLIPProcessor.decode with CLIP->Florence2
|
| 247 |
+
def decode(self, *args, **kwargs):
|
| 248 |
+
"""
|
| 249 |
+
This method forwards all its arguments to BartTokenizerFast's [`~PreTrainedTokenizer.decode`]. Please refer to
|
| 250 |
+
the docstring of this method for more information.
|
| 251 |
+
"""
|
| 252 |
+
return self.tokenizer.decode(*args, **kwargs)
|
| 253 |
+
|
| 254 |
+
@property
|
| 255 |
+
# Copied from transformers.models.clip.processing_clip.CLIPProcessor.model_input_names with CLIP->Florence2
|
| 256 |
+
def model_input_names(self):
|
| 257 |
+
tokenizer_input_names = self.tokenizer.model_input_names
|
| 258 |
+
image_processor_input_names = self.image_processor.model_input_names
|
| 259 |
+
return list(dict.fromkeys(tokenizer_input_names + image_processor_input_names))
|
processor_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_map": {
|
| 3 |
+
"AutoProcessor": "processing_mono.MonoProcessor"
|
| 4 |
+
},
|
| 5 |
+
"processor_class": "MonoProcessor"
|
| 6 |
+
}
|
special_tokens_map.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,207 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_prefix_space": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"151643": {
|
| 6 |
+
"content": "<|endoftext|>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
},
|
| 13 |
+
"151644": {
|
| 14 |
+
"content": "<|im_start|>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false,
|
| 19 |
+
"special": true
|
| 20 |
+
},
|
| 21 |
+
"151645": {
|
| 22 |
+
"content": "<|im_end|>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": false,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false,
|
| 27 |
+
"special": true
|
| 28 |
+
},
|
| 29 |
+
"151646": {
|
| 30 |
+
"content": "<|object_ref_start|>",
|
| 31 |
+
"lstrip": false,
|
| 32 |
+
"normalized": false,
|
| 33 |
+
"rstrip": false,
|
| 34 |
+
"single_word": false,
|
| 35 |
+
"special": true
|
| 36 |
+
},
|
| 37 |
+
"151647": {
|
| 38 |
+
"content": "<|object_ref_end|>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false,
|
| 43 |
+
"special": true
|
| 44 |
+
},
|
| 45 |
+
"151648": {
|
| 46 |
+
"content": "<|box_start|>",
|
| 47 |
+
"lstrip": false,
|
| 48 |
+
"normalized": false,
|
| 49 |
+
"rstrip": false,
|
| 50 |
+
"single_word": false,
|
| 51 |
+
"special": true
|
| 52 |
+
},
|
| 53 |
+
"151649": {
|
| 54 |
+
"content": "<|box_end|>",
|
| 55 |
+
"lstrip": false,
|
| 56 |
+
"normalized": false,
|
| 57 |
+
"rstrip": false,
|
| 58 |
+
"single_word": false,
|
| 59 |
+
"special": true
|
| 60 |
+
},
|
| 61 |
+
"151650": {
|
| 62 |
+
"content": "<|quad_start|>",
|
| 63 |
+
"lstrip": false,
|
| 64 |
+
"normalized": false,
|
| 65 |
+
"rstrip": false,
|
| 66 |
+
"single_word": false,
|
| 67 |
+
"special": true
|
| 68 |
+
},
|
| 69 |
+
"151651": {
|
| 70 |
+
"content": "<|quad_end|>",
|
| 71 |
+
"lstrip": false,
|
| 72 |
+
"normalized": false,
|
| 73 |
+
"rstrip": false,
|
| 74 |
+
"single_word": false,
|
| 75 |
+
"special": true
|
| 76 |
+
},
|
| 77 |
+
"151652": {
|
| 78 |
+
"content": "<|vision_start|>",
|
| 79 |
+
"lstrip": false,
|
| 80 |
+
"normalized": false,
|
| 81 |
+
"rstrip": false,
|
| 82 |
+
"single_word": false,
|
| 83 |
+
"special": true
|
| 84 |
+
},
|
| 85 |
+
"151653": {
|
| 86 |
+
"content": "<|vision_end|>",
|
| 87 |
+
"lstrip": false,
|
| 88 |
+
"normalized": false,
|
| 89 |
+
"rstrip": false,
|
| 90 |
+
"single_word": false,
|
| 91 |
+
"special": true
|
| 92 |
+
},
|
| 93 |
+
"151654": {
|
| 94 |
+
"content": "<|vision_pad|>",
|
| 95 |
+
"lstrip": false,
|
| 96 |
+
"normalized": false,
|
| 97 |
+
"rstrip": false,
|
| 98 |
+
"single_word": false,
|
| 99 |
+
"special": true
|
| 100 |
+
},
|
| 101 |
+
"151655": {
|
| 102 |
+
"content": "<|image_pad|>",
|
| 103 |
+
"lstrip": false,
|
| 104 |
+
"normalized": false,
|
| 105 |
+
"rstrip": false,
|
| 106 |
+
"single_word": false,
|
| 107 |
+
"special": true
|
| 108 |
+
},
|
| 109 |
+
"151656": {
|
| 110 |
+
"content": "<|video_pad|>",
|
| 111 |
+
"lstrip": false,
|
| 112 |
+
"normalized": false,
|
| 113 |
+
"rstrip": false,
|
| 114 |
+
"single_word": false,
|
| 115 |
+
"special": true
|
| 116 |
+
},
|
| 117 |
+
"151657": {
|
| 118 |
+
"content": "<tool_call>",
|
| 119 |
+
"lstrip": false,
|
| 120 |
+
"normalized": false,
|
| 121 |
+
"rstrip": false,
|
| 122 |
+
"single_word": false,
|
| 123 |
+
"special": false
|
| 124 |
+
},
|
| 125 |
+
"151658": {
|
| 126 |
+
"content": "</tool_call>",
|
| 127 |
+
"lstrip": false,
|
| 128 |
+
"normalized": false,
|
| 129 |
+
"rstrip": false,
|
| 130 |
+
"single_word": false,
|
| 131 |
+
"special": false
|
| 132 |
+
},
|
| 133 |
+
"151659": {
|
| 134 |
+
"content": "<|fim_prefix|>",
|
| 135 |
+
"lstrip": false,
|
| 136 |
+
"normalized": false,
|
| 137 |
+
"rstrip": false,
|
| 138 |
+
"single_word": false,
|
| 139 |
+
"special": false
|
| 140 |
+
},
|
| 141 |
+
"151660": {
|
| 142 |
+
"content": "<|fim_middle|>",
|
| 143 |
+
"lstrip": false,
|
| 144 |
+
"normalized": false,
|
| 145 |
+
"rstrip": false,
|
| 146 |
+
"single_word": false,
|
| 147 |
+
"special": false
|
| 148 |
+
},
|
| 149 |
+
"151661": {
|
| 150 |
+
"content": "<|fim_suffix|>",
|
| 151 |
+
"lstrip": false,
|
| 152 |
+
"normalized": false,
|
| 153 |
+
"rstrip": false,
|
| 154 |
+
"single_word": false,
|
| 155 |
+
"special": false
|
| 156 |
+
},
|
| 157 |
+
"151662": {
|
| 158 |
+
"content": "<|fim_pad|>",
|
| 159 |
+
"lstrip": false,
|
| 160 |
+
"normalized": false,
|
| 161 |
+
"rstrip": false,
|
| 162 |
+
"single_word": false,
|
| 163 |
+
"special": false
|
| 164 |
+
},
|
| 165 |
+
"151663": {
|
| 166 |
+
"content": "<|repo_name|>",
|
| 167 |
+
"lstrip": false,
|
| 168 |
+
"normalized": false,
|
| 169 |
+
"rstrip": false,
|
| 170 |
+
"single_word": false,
|
| 171 |
+
"special": false
|
| 172 |
+
},
|
| 173 |
+
"151664": {
|
| 174 |
+
"content": "<|file_sep|>",
|
| 175 |
+
"lstrip": false,
|
| 176 |
+
"normalized": false,
|
| 177 |
+
"rstrip": false,
|
| 178 |
+
"single_word": false,
|
| 179 |
+
"special": false
|
| 180 |
+
}
|
| 181 |
+
},
|
| 182 |
+
"additional_special_tokens": [
|
| 183 |
+
"<|im_start|>",
|
| 184 |
+
"<|im_end|>",
|
| 185 |
+
"<|object_ref_start|>",
|
| 186 |
+
"<|object_ref_end|>",
|
| 187 |
+
"<|box_start|>",
|
| 188 |
+
"<|box_end|>",
|
| 189 |
+
"<|quad_start|>",
|
| 190 |
+
"<|quad_end|>",
|
| 191 |
+
"<|vision_start|>",
|
| 192 |
+
"<|vision_end|>",
|
| 193 |
+
"<|vision_pad|>",
|
| 194 |
+
"<|image_pad|>",
|
| 195 |
+
"<|video_pad|>"
|
| 196 |
+
],
|
| 197 |
+
"bos_token": null,
|
| 198 |
+
"chat_template": "{%- if tools %}\n {{- '<|im_start|>system\\n' }}\n {%- if messages[0]['role'] == 'system' %}\n {{- messages[0]['content'] }}\n {%- else %}\n {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n {%- endif %}\n {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n {%- for tool in tools %}\n {{- \"\\n\" }}\n {{- tool | tojson }}\n {%- endfor %}\n {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n {%- if messages[0]['role'] == 'system' %}\n {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n {%- else %}\n {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n {%- elif message.role == \"assistant\" %}\n {{- '<|im_start|>' + message.role }}\n {%- if message.content %}\n {{- '\\n' + message.content }}\n {%- endif %}\n {%- for tool_call in message.tool_calls %}\n {%- if tool_call.function is defined %}\n {%- set tool_call = tool_call.function %}\n {%- endif %}\n {{- '\\n<tool_call>\\n{\"name\": \"' }}\n {{- tool_call.name }}\n {{- '\", \"arguments\": ' }}\n {{- tool_call.arguments | tojson }}\n {{- '}\\n</tool_call>' }}\n {%- endfor %}\n {{- '<|im_end|>\\n' }}\n {%- elif message.role == \"tool\" %}\n {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n {{- '<|im_start|>user' }}\n {%- endif %}\n {{- '\\n<tool_response>\\n' }}\n {{- message.content }}\n {{- '\\n</tool_response>' }}\n {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n {{- '<|im_end|>\\n' }}\n {%- endif %}\n {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
|
| 199 |
+
"clean_up_tokenization_spaces": false,
|
| 200 |
+
"eos_token": "<|im_end|>",
|
| 201 |
+
"errors": "replace",
|
| 202 |
+
"model_max_length": 131072,
|
| 203 |
+
"pad_token": "<|endoftext|>",
|
| 204 |
+
"split_special_tokens": false,
|
| 205 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 206 |
+
"unk_token": null
|
| 207 |
+
}
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|