Hugging Face's logo

Datasets:
Ian-L
/
syntheseus-runtime

Tags:
syntheseus
reaction-prediction
retrosynthesis
forward-reaction-prediction
License:
Dataset card Files
xet
 Community
Ian-L commited on
Commit
73fd193
·
verified ·
1 Parent(s): d15b015

Add files using upload-large-folder tool

Browse files
Files changed (50) hide show
  1. checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/SHA256SUMS +12 -0
  2. checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/aux/bart_vocab.json +539 -0
  3. checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/aux/bart_vocab_downstream.json +539 -0
  4. checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/aux/ds_config.json +18 -0
  5. checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/aux/hparams.yaml +25 -0
  6. checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/aux/hyper_parameters.json +29 -0
  7. checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/benchmarks/default_test.jsonl +0 -0
  8. checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/benchmarks/metadata.json +35 -0
  9. checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/manifest.json +154 -0
  10. checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/provenance.json +59 -0
  11. checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/wrapper_config.json +31 -0
  12. checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/SHA256SUMS +12 -0
  13. checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/aux/bart_vocab.json +539 -0
  14. checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/aux/bart_vocab_downstream.json +539 -0
  15. checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/aux/ds_config.json +18 -0
  16. checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/aux/hparams.yaml +25 -0
  17. checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/aux/hyper_parameters.json +29 -0
  18. checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/benchmarks/default_test.jsonl +0 -0
  19. checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/benchmarks/metadata.json +36 -0
  20. checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/manifest.json +155 -0
  21. checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/provenance.json +59 -0
  22. checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/wrapper_config.json +31 -0
  23. checkpoints/forward_normalize_report.json +69 -0
  24. checkpoints/forward_normalized_registry.json +440 -0
  25. checkpoints/forward_runtime_registry.json +519 -0
  26. checkpoints/normalize_report.json +269 -0
  27. checkpoints/normalized_registry.json +1385 -0
  28. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward/SHA256SUMS +26 -0
  29. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward/manifest.json +137 -0
  30. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward/provenance.json +495 -0
  31. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward/wrapper_config.json +40 -0
  32. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/SHA256SUMS +26 -0
  33. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/.gitattributes +35 -0
  34. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/README.md +108 -0
  35. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/UPSTREAM_URL.txt +2 -0
  36. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/config.json +34 -0
  37. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/generation_config.json +7 -0
  38. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/hf_model_info.json +97 -0
  39. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/provenance/README.txt +5 -0
  40. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/provenance/ReactionT5v2_source_README.md +237 -0
  41. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/provenance/reactiont5v2_forward_split_provenance.json +62 -0
  42. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/special_tokens_map.json +127 -0
  43. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/tokenizer.json +1947 -0
  44. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/tokenizer_config.json +1328 -0
  45. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/trainer_state.json +149 -0
  46. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/benchmarks/metadata.json +39 -0
  47. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/manifest.json +137 -0
  48. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/provenance.json +495 -0
  49. checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/wrapper_config.json +40 -0
  50. checkpoints/runtime_registry.json +1734 -0
checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/SHA256SUMS ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 8b17254dfde43cbac758e697d3939c0100ba2d5da341031e92fd6b8de6e87cfb aux/bart_vocab.json
2
+ 8c15ea5b60855344ddb24586b14c33811e89c93adf4189865045fcf3ad89459c aux/bart_vocab_downstream.json
3
+ ed5d259989f3d3e3d423ecc3c1f7f56dda31ea1debed0a29977be2270fc1b588 aux/ds_config.json
4
+ 8cdffda14b6bf90b5a93a090a73c6d34e8709739abad86fd0b74b48d2c0f249b aux/hparams.yaml
5
+ 0c7930b00473a1ef66b9d5062fa0bd394c1feec7ed7eb61ad8351abdd6644d61 aux/hyper_parameters.json
6
+ 23f8da8997717c3fce5bffece27832a3b8c91e177a249233c2935a8500a45215 benchmarks/default_test.jsonl
7
+ 045e5974fc693a241b89d4e8cefad2783be12b6b504409852e982a3b57c26855 benchmarks/metadata.json
8
+ 5ab57870f03c61697d463ed2aa509385f08d0295c7008e111271d928c6110c32 benchmarks/native/uspto_mixed.pickle
9
+ 7b9a63ccb905d1387f4b7d26b3fe9f529f50655bdea00cc034971d0455b18268 manifest.json
10
+ 79e867a06c063a3d09f009edd0c8f0dd1eb266034ae75864f175da3ee5ecbeb2 model.safetensors
11
+ a20cc9f310a9bcc6d2993e828f8b47201ef6c04871e9ee564f0f993406e82e2b provenance.json
12
+ 714da361b86c29cedef958d3cdb95c7d986b62a337d5649aab5a383fab3f5fab wrapper_config.json
checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/aux/bart_vocab.json ADDED
@@ -0,0 +1,539 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "properties": {
3
+ "regex": "(LogD_change_\\(-0.1, 0.1\\]|LogD_change_\\(0.1, 0.3\\]|LogD_change_\\(0.3, 0.5\\]|LogD_change_\\(0.5, 0.7\\]|LogD_change_\\(0.7, 0.9\\]|LogD_change_\\(0.9, 1.1\\]|LogD_change_\\(1.1, 1.3\\]|LogD_change_\\(1.3, 1.5\\]|LogD_change_\\(1.5, 1.7\\]|LogD_change_\\(1.7, 1.9\\]|LogD_change_\\(1.9, 2.1\\]|LogD_change_\\(2.1, 2.3\\]|LogD_change_\\(2.3, 2.5\\]|LogD_change_\\(2.5, 2.7\\]|LogD_change_\\(2.7, 2.9\\]|LogD_change_\\(2.9, 3.1\\]|LogD_change_\\(3.1, 3.3\\]|LogD_change_\\(3.3, 3.5\\]|LogD_change_\\(3.5, 3.7\\]|LogD_change_\\(3.7, 3.9\\]|LogD_change_\\(3.9, 4.1\\]|LogD_change_\\(4.1, 4.3\\]|LogD_change_\\(4.3, 4.5\\]|LogD_change_\\(4.5, 4.7\\]|LogD_change_\\(4.7, 4.9\\]|LogD_change_\\(4.9, 5.1\\]|LogD_change_\\(5.1, 5.3\\]|LogD_change_\\(5.3, 5.5\\]|LogD_change_\\(5.5, 5.7\\]|LogD_change_\\(5.7, 5.9\\]|LogD_change_\\(5.9, inf\\]|LogD_change_\\(-0.3, -0.1\\]|LogD_change_\\(-0.5, -0.3\\]|LogD_change_\\(-0.7, -0.5\\]|LogD_change_\\(-0.9, -0.7\\]|LogD_change_\\(-1.1, -0.9\\]|LogD_change_\\(-1.3, -1.1\\]|LogD_change_\\(-1.5, -1.3\\]|LogD_change_\\(-1.7, -1.5\\]|LogD_change_\\(-1.9, -1.7\\]|LogD_change_\\(-2.1, -1.9\\]|LogD_change_\\(-2.3, -2.1\\]|LogD_change_\\(-2.5, -2.3\\]|LogD_change_\\(-2.7, -2.5\\]|LogD_change_\\(-2.9, -2.7\\]|LogD_change_\\(-3.1, -2.9\\]|LogD_change_\\(-3.3, -3.1\\]|LogD_change_\\(-3.5, -3.3\\]|LogD_change_\\(-3.7, -3.5\\]|LogD_change_\\(-3.9, -3.7\\]|LogD_change_\\(-4.1, -3.9\\]|LogD_change_\\(-4.3, -4.1\\]|LogD_change_\\(-4.5, -4.3\\]|LogD_change_\\(-4.7, -4.5\\]|LogD_change_\\(-4.9, -4.7\\]|LogD_change_\\(-5.1, -4.9\\]|LogD_change_\\(-5.3, -5.1\\]|LogD_change_\\(-5.5, -5.3\\]|LogD_change_\\(-5.7, -5.5\\]|LogD_change_\\(-inf, -5.7\\]|Solubility_low->high|Solubility_high->low|Solubility_no_change|Clint_low->high|Clint_high->low|Clint_no_change|<UNUSED_0>|<UNUSED_1>|<UNUSED_2>|<UNUSED_3>|<UNUSED_4>|<UNUSED_5>|<UNUSED_6>|<UNUSED_7>|<UNUSED_8>|<UNUSED_9>|<UNUSED_10>|<UNUSED_11>|<UNUSED_12>|<UNUSED_13>|<UNUSED_14>|<UNUSED_15>|<UNUSED_16>|<UNUSED_17>|<UNUSED_18>|<UNUSED_19>|<UNUSED_20>|<UNUSED_21>|<UNUSED_22>|<UNUSED_23>|<UNUSED_24>|<UNUSED_25>|<UNUSED_26>|<UNUSED_27>|<UNUSED_28>|<UNUSED_29>|<UNUSED_30>|<UNUSED_31>|<UNUSED_32>|<UNUSED_33>|<UNUSED_34>|<UNUSED_35>|<UNUSED_36>|<UNUSED_37>|<UNUSED_38>|<UNUSED_39>|<UNUSED_40>|<UNUSED_41>|<UNUSED_42>|<UNUSED_43>|<UNUSED_44>|<UNUSED_45>|<UNUSED_46>|<UNUSED_47>|<UNUSED_48>|<UNUSED_49>|<UNUSED_50>|<UNUSED_51>|<UNUSED_52>|<UNUSED_53>|<UNUSED_54>|<UNUSED_55>|<UNUSED_56>|<UNUSED_57>|<UNUSED_58>|<UNUSED_59>|<UNUSED_60>|<UNUSED_61>|<UNUSED_62>|<UNUSED_63>|<UNUSED_64>|<UNUSED_65>|<UNUSED_66>|<UNUSED_67>|<UNUSED_68>|<UNUSED_69>|<UNUSED_70>|<UNUSED_71>|<UNUSED_72>|<UNUSED_73>|<UNUSED_74>|<UNUSED_75>|<UNUSED_76>|<UNUSED_77>|<UNUSED_78>|<UNUSED_79>|<UNUSED_80>|<UNUSED_81>|<UNUSED_82>|<UNUSED_83>|<UNUSED_84>|<UNUSED_85>|<UNUSED_86>|<UNUSED_87>|<UNUSED_88>|<UNUSED_89>|<UNUSED_90>|<UNUSED_91>|<UNUSED_92>|<UNUSED_93>|<UNUSED_94>|<UNUSED_95>|<UNUSED_96>|<UNUSED_97>|<UNUSED_98>|<UNUSED_99>|<UNUSED_100>|<UNUSED_101>|<UNUSED_102>|<UNUSED_103>|<UNUSED_104>|<UNUSED_105>|<UNUSED_106>|<UNUSED_107>|<UNUSED_108>|<UNUSED_109>|<UNUSED_110>|<UNUSED_111>|<UNUSED_112>|<UNUSED_113>|<UNUSED_114>|<UNUSED_115>|<UNUSED_116>|<UNUSED_117>|<UNUSED_118>|<UNUSED_119>|<UNUSED_120>|<UNUSED_121>|<UNUSED_122>|<UNUSED_123>|<UNUSED_124>|<UNUSED_125>|<UNUSED_126>|<UNUSED_127>|<UNUSED_128>|<UNUSED_129>|<UNUSED_130>|<UNUSED_131>|<UNUSED_132>|<UNUSED_133>|<UNUSED_134>|<UNUSED_135>|<UNUSED_136>|<UNUSED_137>|<UNUSED_138>|<UNUSED_139>|<UNUSED_140>|<UNUSED_141>|<UNUSED_142>|<UNUSED_143>|<UNUSED_144>|<UNUSED_145>|<UNUSED_146>|<UNUSED_147>|<UNUSED_148>|<UNUSED_149>|<UNUSED_150>|<UNUSED_151>|<UNUSED_152>|<UNUSED_153>|<UNUSED_154>|<UNUSED_155>|<UNUSED_156>|<UNUSED_157>|<UNUSED_158>|<UNUSED_159>|<UNUSED_160>|<UNUSED_161>|<UNUSED_162>|<UNUSED_163>|<UNUSED_164>|<UNUSED_165>|<UNUSED_166>|<UNUSED_167>|<UNUSED_168>|<UNUSED_169>|<UNUSED_170>|<UNUSED_171>|<UNUSED_172>|<UNUSED_173>|<UNUSED_174>|<UNUSED_175>|<UNUSED_176>|<UNUSED_177>|<UNUSED_178>|<UNUSED_179>|<UNUSED_180>|<UNUSED_181>|<UNUSED_182>|<UNUSED_183>|<UNUSED_184>|<UNUSED_185>|<UNUSED_186>|<UNUSED_187>|<UNUSED_188>|<UNUSED_189>|<UNUSED_190>|<UNUSED_191>|<UNUSED_192>|<UNUSED_193>|<UNUSED_194>|<UNUSED_195>|<UNUSED_196>|<UNUSED_197>|<UNUSED_198>|<UNUSED_199>|\\[[^\\]]+]|Br?|Cl?|N|O|S|P|F|I|b|c|n|o|s|p|\\(|\\)|\\.|=|#|-|\\+|\\\\|\\/|:|~|@|\\?|>|\\*|\\$|\\%[0-9]{2}|[0-9]|.)",
4
+ "special_tokens": {
5
+ "start": "^",
6
+ "end": "&",
7
+ "pad": "<PAD>",
8
+ "unknown": "?",
9
+ "mask": "<MASK>",
10
+ "sep": "<SEP>"
11
+ },
12
+ "chem_start_idx": 272
13
+ },
14
+ "vocabulary": [
15
+ "<PAD>",
16
+ "?",
17
+ "^",
18
+ "&",
19
+ "<MASK>",
20
+ "<SEP>",
21
+ "LogD_change_(-0.1, 0.1]",
22
+ "LogD_change_(0.1, 0.3]",
23
+ "LogD_change_(0.3, 0.5]",
24
+ "LogD_change_(0.5, 0.7]",
25
+ "LogD_change_(0.7, 0.9]",
26
+ "LogD_change_(0.9, 1.1]",
27
+ "LogD_change_(1.1, 1.3]",
28
+ "LogD_change_(1.3, 1.5]",
29
+ "LogD_change_(1.5, 1.7]",
30
+ "LogD_change_(1.7, 1.9]",
31
+ "LogD_change_(1.9, 2.1]",
32
+ "LogD_change_(2.1, 2.3]",
33
+ "LogD_change_(2.3, 2.5]",
34
+ "LogD_change_(2.5, 2.7]",
35
+ "LogD_change_(2.7, 2.9]",
36
+ "LogD_change_(2.9, 3.1]",
37
+ "LogD_change_(3.1, 3.3]",
38
+ "LogD_change_(3.3, 3.5]",
39
+ "LogD_change_(3.5, 3.7]",
40
+ "LogD_change_(3.7, 3.9]",
41
+ "LogD_change_(3.9, 4.1]",
42
+ "LogD_change_(4.1, 4.3]",
43
+ "LogD_change_(4.3, 4.5]",
44
+ "LogD_change_(4.5, 4.7]",
45
+ "LogD_change_(4.7, 4.9]",
46
+ "LogD_change_(4.9, 5.1]",
47
+ "LogD_change_(5.1, 5.3]",
48
+ "LogD_change_(5.3, 5.5]",
49
+ "LogD_change_(5.5, 5.7]",
50
+ "LogD_change_(5.7, 5.9]",
51
+ "LogD_change_(5.9, inf]",
52
+ "LogD_change_(-0.3, -0.1]",
53
+ "LogD_change_(-0.5, -0.3]",
54
+ "LogD_change_(-0.7, -0.5]",
55
+ "LogD_change_(-0.9, -0.7]",
56
+ "LogD_change_(-1.1, -0.9]",
57
+ "LogD_change_(-1.3, -1.1]",
58
+ "LogD_change_(-1.5, -1.3]",
59
+ "LogD_change_(-1.7, -1.5]",
60
+ "LogD_change_(-1.9, -1.7]",
61
+ "LogD_change_(-2.1, -1.9]",
62
+ "LogD_change_(-2.3, -2.1]",
63
+ "LogD_change_(-2.5, -2.3]",
64
+ "LogD_change_(-2.7, -2.5]",
65
+ "LogD_change_(-2.9, -2.7]",
66
+ "LogD_change_(-3.1, -2.9]",
67
+ "LogD_change_(-3.3, -3.1]",
68
+ "LogD_change_(-3.5, -3.3]",
69
+ "LogD_change_(-3.7, -3.5]",
70
+ "LogD_change_(-3.9, -3.7]",
71
+ "LogD_change_(-4.1, -3.9]",
72
+ "LogD_change_(-4.3, -4.1]",
73
+ "LogD_change_(-4.5, -4.3]",
74
+ "LogD_change_(-4.7, -4.5]",
75
+ "LogD_change_(-4.9, -4.7]",
76
+ "LogD_change_(-5.1, -4.9]",
77
+ "LogD_change_(-5.3, -5.1]",
78
+ "LogD_change_(-5.5, -5.3]",
79
+ "LogD_change_(-5.7, -5.5]",
80
+ "LogD_change_(-inf, -5.7]",
81
+ "Solubility_low->high",
82
+ "Solubility_high->low",
83
+ "Solubility_no_change",
84
+ "Clint_low->high",
85
+ "Clint_high->low",
86
+ "Clint_no_change",
87
+ "<UNUSED_0>",
88
+ "<UNUSED_1>",
89
+ "<UNUSED_2>",
90
+ "<UNUSED_3>",
91
+ "<UNUSED_4>",
92
+ "<UNUSED_5>",
93
+ "<UNUSED_6>",
94
+ "<UNUSED_7>",
95
+ "<UNUSED_8>",
96
+ "<UNUSED_9>",
97
+ "<UNUSED_10>",
98
+ "<UNUSED_11>",
99
+ "<UNUSED_12>",
100
+ "<UNUSED_13>",
101
+ "<UNUSED_14>",
102
+ "<UNUSED_15>",
103
+ "<UNUSED_16>",
104
+ "<UNUSED_17>",
105
+ "<UNUSED_18>",
106
+ "<UNUSED_19>",
107
+ "<UNUSED_20>",
108
+ "<UNUSED_21>",
109
+ "<UNUSED_22>",
110
+ "<UNUSED_23>",
111
+ "<UNUSED_24>",
112
+ "<UNUSED_25>",
113
+ "<UNUSED_26>",
114
+ "<UNUSED_27>",
115
+ "<UNUSED_28>",
116
+ "<UNUSED_29>",
117
+ "<UNUSED_30>",
118
+ "<UNUSED_31>",
119
+ "<UNUSED_32>",
120
+ "<UNUSED_33>",
121
+ "<UNUSED_34>",
122
+ "<UNUSED_35>",
123
+ "<UNUSED_36>",
124
+ "<UNUSED_37>",
125
+ "<UNUSED_38>",
126
+ "<UNUSED_39>",
127
+ "<UNUSED_40>",
128
+ "<UNUSED_41>",
129
+ "<UNUSED_42>",
130
+ "<UNUSED_43>",
131
+ "<UNUSED_44>",
132
+ "<UNUSED_45>",
133
+ "<UNUSED_46>",
134
+ "<UNUSED_47>",
135
+ "<UNUSED_48>",
136
+ "<UNUSED_49>",
137
+ "<UNUSED_50>",
138
+ "<UNUSED_51>",
139
+ "<UNUSED_52>",
140
+ "<UNUSED_53>",
141
+ "<UNUSED_54>",
142
+ "<UNUSED_55>",
143
+ "<UNUSED_56>",
144
+ "<UNUSED_57>",
145
+ "<UNUSED_58>",
146
+ "<UNUSED_59>",
147
+ "<UNUSED_60>",
148
+ "<UNUSED_61>",
149
+ "<UNUSED_62>",
150
+ "<UNUSED_63>",
151
+ "<UNUSED_64>",
152
+ "<UNUSED_65>",
153
+ "<UNUSED_66>",
154
+ "<UNUSED_67>",
155
+ "<UNUSED_68>",
156
+ "<UNUSED_69>",
157
+ "<UNUSED_70>",
158
+ "<UNUSED_71>",
159
+ "<UNUSED_72>",
160
+ "<UNUSED_73>",
161
+ "<UNUSED_74>",
162
+ "<UNUSED_75>",
163
+ "<UNUSED_76>",
164
+ "<UNUSED_77>",
165
+ "<UNUSED_78>",
166
+ "<UNUSED_79>",
167
+ "<UNUSED_80>",
168
+ "<UNUSED_81>",
169
+ "<UNUSED_82>",
170
+ "<UNUSED_83>",
171
+ "<UNUSED_84>",
172
+ "<UNUSED_85>",
173
+ "<UNUSED_86>",
174
+ "<UNUSED_87>",
175
+ "<UNUSED_88>",
176
+ "<UNUSED_89>",
177
+ "<UNUSED_90>",
178
+ "<UNUSED_91>",
179
+ "<UNUSED_92>",
180
+ "<UNUSED_93>",
181
+ "<UNUSED_94>",
182
+ "<UNUSED_95>",
183
+ "<UNUSED_96>",
184
+ "<UNUSED_97>",
185
+ "<UNUSED_98>",
186
+ "<UNUSED_99>",
187
+ "<UNUSED_100>",
188
+ "<UNUSED_101>",
189
+ "<UNUSED_102>",
190
+ "<UNUSED_103>",
191
+ "<UNUSED_104>",
192
+ "<UNUSED_105>",
193
+ "<UNUSED_106>",
194
+ "<UNUSED_107>",
195
+ "<UNUSED_108>",
196
+ "<UNUSED_109>",
197
+ "<UNUSED_110>",
198
+ "<UNUSED_111>",
199
+ "<UNUSED_112>",
200
+ "<UNUSED_113>",
201
+ "<UNUSED_114>",
202
+ "<UNUSED_115>",
203
+ "<UNUSED_116>",
204
+ "<UNUSED_117>",
205
+ "<UNUSED_118>",
206
+ "<UNUSED_119>",
207
+ "<UNUSED_120>",
208
+ "<UNUSED_121>",
209
+ "<UNUSED_122>",
210
+ "<UNUSED_123>",
211
+ "<UNUSED_124>",
212
+ "<UNUSED_125>",
213
+ "<UNUSED_126>",
214
+ "<UNUSED_127>",
215
+ "<UNUSED_128>",
216
+ "<UNUSED_129>",
217
+ "<UNUSED_130>",
218
+ "<UNUSED_131>",
219
+ "<UNUSED_132>",
220
+ "<UNUSED_133>",
221
+ "<UNUSED_134>",
222
+ "<UNUSED_135>",
223
+ "<UNUSED_136>",
224
+ "<UNUSED_137>",
225
+ "<UNUSED_138>",
226
+ "<UNUSED_139>",
227
+ "<UNUSED_140>",
228
+ "<UNUSED_141>",
229
+ "<UNUSED_142>",
230
+ "<UNUSED_143>",
231
+ "<UNUSED_144>",
232
+ "<UNUSED_145>",
233
+ "<UNUSED_146>",
234
+ "<UNUSED_147>",
235
+ "<UNUSED_148>",
236
+ "<UNUSED_149>",
237
+ "<UNUSED_150>",
238
+ "<UNUSED_151>",
239
+ "<UNUSED_152>",
240
+ "<UNUSED_153>",
241
+ "<UNUSED_154>",
242
+ "<UNUSED_155>",
243
+ "<UNUSED_156>",
244
+ "<UNUSED_157>",
245
+ "<UNUSED_158>",
246
+ "<UNUSED_159>",
247
+ "<UNUSED_160>",
248
+ "<UNUSED_161>",
249
+ "<UNUSED_162>",
250
+ "<UNUSED_163>",
251
+ "<UNUSED_164>",
252
+ "<UNUSED_165>",
253
+ "<UNUSED_166>",
254
+ "<UNUSED_167>",
255
+ "<UNUSED_168>",
256
+ "<UNUSED_169>",
257
+ "<UNUSED_170>",
258
+ "<UNUSED_171>",
259
+ "<UNUSED_172>",
260
+ "<UNUSED_173>",
261
+ "<UNUSED_174>",
262
+ "<UNUSED_175>",
263
+ "<UNUSED_176>",
264
+ "<UNUSED_177>",
265
+ "<UNUSED_178>",
266
+ "<UNUSED_179>",
267
+ "<UNUSED_180>",
268
+ "<UNUSED_181>",
269
+ "<UNUSED_182>",
270
+ "<UNUSED_183>",
271
+ "<UNUSED_184>",
272
+ "<UNUSED_185>",
273
+ "<UNUSED_186>",
274
+ "<UNUSED_187>",
275
+ "<UNUSED_188>",
276
+ "<UNUSED_189>",
277
+ "<UNUSED_190>",
278
+ "<UNUSED_191>",
279
+ "<UNUSED_192>",
280
+ "<UNUSED_193>",
281
+ "<UNUSED_194>",
282
+ "<UNUSED_195>",
283
+ "<UNUSED_196>",
284
+ "<UNUSED_197>",
285
+ "<UNUSED_198>",
286
+ "<UNUSED_199>",
287
+ "C",
288
+ "c",
289
+ "1",
290
+ "(",
291
+ "-",
292
+ "2",
293
+ "s",
294
+ "N",
295
+ "=",
296
+ ")",
297
+ "n",
298
+ "Br",
299
+ "3",
300
+ "O",
301
+ "[nH]",
302
+ "[C@H]",
303
+ "/",
304
+ "\\",
305
+ "[C@@H]",
306
+ "[C@]",
307
+ "o",
308
+ "4",
309
+ "[C@@]",
310
+ "5",
311
+ "6",
312
+ "7",
313
+ "[n+]",
314
+ ".",
315
+ "[Br-]",
316
+ "S",
317
+ "I",
318
+ "F",
319
+ "Cl",
320
+ "[O-]",
321
+ "#",
322
+ "P",
323
+ "[Na+]",
324
+ "[N+]",
325
+ "[Cl-]",
326
+ "8",
327
+ "[Si]",
328
+ "[N-]",
329
+ "[18F]",
330
+ "9",
331
+ "[I-]",
332
+ "B",
333
+ "[S+]",
334
+ "[2H]",
335
+ "[P+]",
336
+ "[125I]",
337
+ "%10",
338
+ "%11",
339
+ "[n-]",
340
+ "[B-]",
341
+ "[O]",
342
+ "[o+]",
343
+ "[N@@+]",
344
+ "[N@+]",
345
+ "[PH]",
346
+ "[se]",
347
+ "[Se]",
348
+ "[s+]",
349
+ "[Li+]",
350
+ "[P@]",
351
+ "[P@@]",
352
+ "[3H]",
353
+ "[K+]",
354
+ "[OH-]",
355
+ "[te+]",
356
+ "[se+]",
357
+ "[Te]",
358
+ "[S@+]",
359
+ "[BH2-]",
360
+ "[11CH3]",
361
+ "[11C]",
362
+ "[Cl+3]",
363
+ "[N@@]",
364
+ "[S-]",
365
+ "[C+]",
366
+ "[P@+]",
367
+ "[C-]",
368
+ "[Zn+2]",
369
+ "[Ca+2]",
370
+ "[Mg+2]",
371
+ "[SeH]",
372
+ "[BH-]",
373
+ "[SH2]",
374
+ "[TeH2]",
375
+ "[SiH4]",
376
+ "[N@]",
377
+ "[14CH2]",
378
+ "[Ag]",
379
+ "[S@@+]",
380
+ "[I+]",
381
+ "[MgH2]",
382
+ "[125IH]",
383
+ "[Se+]",
384
+ "[As]",
385
+ "[SiH2]",
386
+ "[Ra]",
387
+ "[IH2]",
388
+ "[P-]",
389
+ "[Na]",
390
+ "[NH-]",
391
+ "[Cs]",
392
+ "[Zn]",
393
+ "[Li]",
394
+ "[As+]",
395
+ "[te]",
396
+ "[131I]",
397
+ "[Ag+]",
398
+ "p",
399
+ "[Al-3]",
400
+ "[Rb+]",
401
+ "[13CH]",
402
+ "[11CH]",
403
+ "[11CH2]",
404
+ "[14c]",
405
+ "[14C@@]",
406
+ "[127I]",
407
+ "[Mg]",
408
+ "[14C]",
409
+ "[123I]",
410
+ "[124I]",
411
+ "[F-]",
412
+ "[14CH3]",
413
+ "[135I]",
414
+ "%12",
415
+ "[NH+]",
416
+ "[76Br]",
417
+ "[32PH]",
418
+ "[35S]",
419
+ "b",
420
+ "[73Se]",
421
+ "[11C@@H]",
422
+ "[Se-]",
423
+ "[c+]",
424
+ "[14C@H]",
425
+ "[18OH]",
426
+ "[SH]",
427
+ "[S@@]",
428
+ "[Cs+]",
429
+ "[He]",
430
+ "[O+]",
431
+ "[nH+]",
432
+ "[NH2+]",
433
+ "[32P]",
434
+ "[Zn+]",
435
+ "[BH3-]",
436
+ "[I+3]",
437
+ "[Si-]",
438
+ "[SH+]",
439
+ "[19F]",
440
+ "[Br+2]",
441
+ "[I+2]",
442
+ "[Al+3]",
443
+ "[123I-]",
444
+ "[131I-]",
445
+ "[127Xe]",
446
+ "[133Xe]",
447
+ "[89Sr+2]",
448
+ "[N]",
449
+ "[82Rb+]",
450
+ "[75Se]",
451
+ "[Rb]",
452
+ "[81Kr]",
453
+ "[18F-]",
454
+ "[13NH3]",
455
+ "[K]",
456
+ "[Cl+2]",
457
+ "[Zn-2]",
458
+ "[SeH2]",
459
+ "[AsH3]",
460
+ "[Kr]",
461
+ "[Xe]",
462
+ "[S@]",
463
+ "[cH-]",
464
+ "[NH4+]",
465
+ "[Al]",
466
+ "[Si@]",
467
+ "[15n]",
468
+ "[SH-]",
469
+ "[SiH]",
470
+ "[11c]",
471
+ "[OH]",
472
+ "[c-]",
473
+ "[18FH]",
474
+ "[123IH]",
475
+ "[13c]",
476
+ "[13cH]",
477
+ "[14cH]",
478
+ "[Cl+]",
479
+ "[Sr+2]",
480
+ "[CH-]",
481
+ "[Bi]",
482
+ "[B]",
483
+ "[Ba+2]",
484
+ "[Bi+3]",
485
+ "[SiH-]",
486
+ "[b-]",
487
+ "[H+]",
488
+ "[13C]",
489
+ "[OH+]",
490
+ "[14CH]",
491
+ "[15nH]",
492
+ "[CaH2]",
493
+ "[LiH]",
494
+ "[C]",
495
+ "[Ca]",
496
+ "[42K+]",
497
+ "[123Te]",
498
+ "[S-2]",
499
+ "[223Ra+2]",
500
+ "[S]",
501
+ "[Ra+2]",
502
+ "[22Na+]",
503
+ "[125I-]",
504
+ "[85Sr+2]",
505
+ "[PH2]",
506
+ "[SrH2]",
507
+ "[15OH2]",
508
+ "[47Ca+2]",
509
+ "[85SrH2]",
510
+ "[45Ca+2]",
511
+ "[B@@-]",
512
+ "[B@-]",
513
+ "[17F]",
514
+ "[PH2+]",
515
+ "[11C-]",
516
+ "[Mg+]",
517
+ "[NaH]",
518
+ "[P@@+]",
519
+ "[SiH3-]",
520
+ "[P-3]",
521
+ "[KH]",
522
+ "[Be+2]",
523
+ "[NH3+]",
524
+ "[Ag-4]",
525
+ "[18O]",
526
+ "[14C@@H]",
527
+ "[CH]",
528
+ "[CH2]",
529
+ "[O-2]",
530
+ "[124I-]",
531
+ "[As-]",
532
+ "[Ba]",
533
+ "[223Ra]",
534
+ "[82Rb]",
535
+ "[76BrH]",
536
+ "[AsH]",
537
+ "[131Cs]"
538
+ ]
539
+ }
checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/aux/bart_vocab_downstream.json ADDED
@@ -0,0 +1,539 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "properties": {
3
+ "regex": "(LogD_change_\\(-0\\.1, 0\\.1\\]|LogD_change_\\(0\\.1, 0\\.3\\]|LogD_change_\\(0\\.3, 0\\.5\\]|LogD_change_\\(0\\.5, 0\\.7\\]|LogD_change_\\(0\\.7, 0\\.9\\]|LogD_change_\\(0\\.9, 1\\.1\\]|LogD_change_\\(1\\.1, 1\\.3\\]|LogD_change_\\(1\\.3, 1\\.5\\]|LogD_change_\\(1\\.5, 1\\.7\\]|LogD_change_\\(1\\.7, 1\\.9\\]|LogD_change_\\(1\\.9, 2\\.1\\]|LogD_change_\\(2\\.1, 2\\.3\\]|LogD_change_\\(2\\.3, 2\\.5\\]|LogD_change_\\(2\\.5, 2\\.7\\]|LogD_change_\\(2\\.7, 2\\.9\\]|LogD_change_\\(2\\.9, 3\\.1\\]|LogD_change_\\(3\\.1, 3\\.3\\]|LogD_change_\\(3\\.3, 3\\.5\\]|LogD_change_\\(3\\.5, 3\\.7\\]|LogD_change_\\(3\\.7, 3\\.9\\]|LogD_change_\\(3\\.9, 4\\.1\\]|LogD_change_\\(4\\.1, 4\\.3\\]|LogD_change_\\(4\\.3, 4\\.5\\]|LogD_change_\\(4\\.5, 4\\.7\\]|LogD_change_\\(4\\.7, 4\\.9\\]|LogD_change_\\(4\\.9, 5\\.1\\]|LogD_change_\\(5\\.1, 5\\.3\\]|LogD_change_\\(5\\.3, 5\\.5\\]|LogD_change_\\(5\\.5, 5\\.7\\]|LogD_change_\\(5\\.7, 5\\.9\\]|LogD_change_\\(5\\.9, inf\\]|LogD_change_\\(-0\\.3, -0\\.1\\]|LogD_change_\\(-0\\.5, -0\\.3\\]|LogD_change_\\(-0\\.7, -0\\.5\\]|LogD_change_\\(-0\\.9, -0\\.7\\]|LogD_change_\\(-1\\.1, -0\\.9\\]|LogD_change_\\(-1\\.3, -1\\.1\\]|LogD_change_\\(-1\\.5, -1\\.3\\]|LogD_change_\\(-1\\.7, -1\\.5\\]|LogD_change_\\(-1\\.9, -1\\.7\\]|LogD_change_\\(-2\\.1, -1\\.9\\]|LogD_change_\\(-2\\.3, -2\\.1\\]|LogD_change_\\(-2\\.5, -2\\.3\\]|LogD_change_\\(-2\\.7, -2\\.5\\]|LogD_change_\\(-2\\.9, -2\\.7\\]|LogD_change_\\(-3\\.1, -2\\.9\\]|LogD_change_\\(-3\\.3, -3\\.1\\]|LogD_change_\\(-3\\.5, -3\\.3\\]|LogD_change_\\(-3\\.7, -3\\.5\\]|LogD_change_\\(-3\\.9, -3\\.7\\]|LogD_change_\\(-4\\.1, -3\\.9\\]|LogD_change_\\(-4\\.3, -4\\.1\\]|LogD_change_\\(-4\\.5, -4\\.3\\]|LogD_change_\\(-4\\.7, -4\\.5\\]|LogD_change_\\(-4\\.9, -4\\.7\\]|LogD_change_\\(-5\\.1, -4\\.9\\]|LogD_change_\\(-5\\.3, -5\\.1\\]|LogD_change_\\(-5\\.5, -5\\.3\\]|LogD_change_\\(-5\\.7, -5\\.5\\]|LogD_change_\\(-inf, -5\\.7\\]|Solubility_low->high|Solubility_high->low|Solubility_no_change|Clint_low->high|Clint_high->low|Clint_no_change|<UNUSED_0>|<UNUSED_1>|<UNUSED_2>|<UNUSED_3>|<UNUSED_4>|<UNUSED_5>|<UNUSED_6>|<UNUSED_7>|<UNUSED_8>|<UNUSED_9>|<UNUSED_10>|<UNUSED_11>|<UNUSED_12>|<UNUSED_13>|<UNUSED_14>|<UNUSED_15>|<UNUSED_16>|<UNUSED_17>|<UNUSED_18>|<UNUSED_19>|<UNUSED_20>|<UNUSED_21>|<UNUSED_22>|<UNUSED_23>|<UNUSED_24>|<UNUSED_25>|<UNUSED_26>|<UNUSED_27>|<UNUSED_28>|<UNUSED_29>|<UNUSED_30>|<UNUSED_31>|<UNUSED_32>|<UNUSED_33>|<UNUSED_34>|<UNUSED_35>|<UNUSED_36>|<UNUSED_37>|<UNUSED_38>|<UNUSED_39>|<UNUSED_40>|<UNUSED_41>|<UNUSED_42>|<UNUSED_43>|<UNUSED_44>|<UNUSED_45>|<UNUSED_46>|<UNUSED_47>|<UNUSED_48>|<UNUSED_49>|<UNUSED_50>|<UNUSED_51>|<UNUSED_52>|<UNUSED_53>|<UNUSED_54>|<UNUSED_55>|<UNUSED_56>|<UNUSED_57>|<UNUSED_58>|<UNUSED_59>|<UNUSED_60>|<UNUSED_61>|<UNUSED_62>|<UNUSED_63>|<UNUSED_64>|<UNUSED_65>|<UNUSED_66>|<UNUSED_67>|<UNUSED_68>|<UNUSED_69>|<UNUSED_70>|<UNUSED_71>|<UNUSED_72>|<UNUSED_73>|<UNUSED_74>|<UNUSED_75>|<UNUSED_76>|<UNUSED_77>|<UNUSED_78>|<UNUSED_79>|<UNUSED_80>|<UNUSED_81>|<UNUSED_82>|<UNUSED_83>|<UNUSED_84>|<UNUSED_85>|<UNUSED_86>|<UNUSED_87>|<UNUSED_88>|<UNUSED_89>|<UNUSED_90>|<UNUSED_91>|<UNUSED_92>|<UNUSED_93>|<UNUSED_94>|<UNUSED_95>|<UNUSED_96>|<UNUSED_97>|<UNUSED_98>|<UNUSED_99>|<UNUSED_100>|<UNUSED_101>|<UNUSED_102>|<UNUSED_103>|<UNUSED_104>|<UNUSED_105>|<UNUSED_106>|<UNUSED_107>|<UNUSED_108>|<UNUSED_109>|<UNUSED_110>|<UNUSED_111>|<UNUSED_112>|<UNUSED_113>|<UNUSED_114>|<UNUSED_115>|<UNUSED_116>|<UNUSED_117>|<UNUSED_118>|<UNUSED_119>|<UNUSED_120>|<UNUSED_121>|<UNUSED_122>|<UNUSED_123>|<UNUSED_124>|<UNUSED_125>|<UNUSED_126>|<UNUSED_127>|<UNUSED_128>|<UNUSED_129>|<UNUSED_130>|<UNUSED_131>|<UNUSED_132>|<UNUSED_133>|<UNUSED_134>|<UNUSED_135>|<UNUSED_136>|<UNUSED_137>|<UNUSED_138>|<UNUSED_139>|<UNUSED_140>|<UNUSED_141>|<UNUSED_142>|<UNUSED_143>|<UNUSED_144>|<UNUSED_145>|<UNUSED_146>|<UNUSED_147>|<UNUSED_148>|<UNUSED_149>|<UNUSED_150>|<UNUSED_151>|<UNUSED_152>|<UNUSED_153>|<UNUSED_154>|<UNUSED_155>|<UNUSED_156>|<UNUSED_157>|<UNUSED_158>|<UNUSED_159>|<UNUSED_160>|<UNUSED_161>|<UNUSED_162>|<UNUSED_163>|<UNUSED_164>|<UNUSED_165>|<UNUSED_166>|<UNUSED_167>|<UNUSED_168>|<UNUSED_169>|<UNUSED_170>|<UNUSED_171>|<UNUSED_172>|<UNUSED_173>|<UNUSED_174>|<UNUSED_175>|<UNUSED_176>|<UNUSED_177>|<UNUSED_178>|<UNUSED_179>|<UNUSED_180>|<UNUSED_181>|<UNUSED_182>|<UNUSED_183>|<UNUSED_184>|<UNUSED_185>|<UNUSED_186>|<UNUSED_187>|<UNUSED_188>|>|<RX_6>|<RX_2>|<RX_1>|<RX_3>|<RX_7>|<RX_9>|<RX_5>|<RX_10>|<RX_4>|<RX_8>|\\[[^\\]]+]|Br?|Cl?|N|O|S|P|F|I|b|c|n|o|s|p|\\(|\\)|\\.|=|#|-|\\+|\\\\|\\/|:|~|@|\\?|>|\\*|\\$|\\%[0-9]{2}|[0-9]|.)",
4
+ "special_tokens": {
5
+ "start": "^",
6
+ "end": "&",
7
+ "pad": "<PAD>",
8
+ "unknown": "?",
9
+ "mask": "<MASK>",
10
+ "sep": "<SEP>"
11
+ },
12
+ "chem_start_idx": 272
13
+ },
14
+ "vocabulary": [
15
+ "<PAD>",
16
+ "?",
17
+ "^",
18
+ "&",
19
+ "<MASK>",
20
+ "<SEP>",
21
+ "LogD_change_(-0.1, 0.1]",
22
+ "LogD_change_(0.1, 0.3]",
23
+ "LogD_change_(0.3, 0.5]",
24
+ "LogD_change_(0.5, 0.7]",
25
+ "LogD_change_(0.7, 0.9]",
26
+ "LogD_change_(0.9, 1.1]",
27
+ "LogD_change_(1.1, 1.3]",
28
+ "LogD_change_(1.3, 1.5]",
29
+ "LogD_change_(1.5, 1.7]",
30
+ "LogD_change_(1.7, 1.9]",
31
+ "LogD_change_(1.9, 2.1]",
32
+ "LogD_change_(2.1, 2.3]",
33
+ "LogD_change_(2.3, 2.5]",
34
+ "LogD_change_(2.5, 2.7]",
35
+ "LogD_change_(2.7, 2.9]",
36
+ "LogD_change_(2.9, 3.1]",
37
+ "LogD_change_(3.1, 3.3]",
38
+ "LogD_change_(3.3, 3.5]",
39
+ "LogD_change_(3.5, 3.7]",
40
+ "LogD_change_(3.7, 3.9]",
41
+ "LogD_change_(3.9, 4.1]",
42
+ "LogD_change_(4.1, 4.3]",
43
+ "LogD_change_(4.3, 4.5]",
44
+ "LogD_change_(4.5, 4.7]",
45
+ "LogD_change_(4.7, 4.9]",
46
+ "LogD_change_(4.9, 5.1]",
47
+ "LogD_change_(5.1, 5.3]",
48
+ "LogD_change_(5.3, 5.5]",
49
+ "LogD_change_(5.5, 5.7]",
50
+ "LogD_change_(5.7, 5.9]",
51
+ "LogD_change_(5.9, inf]",
52
+ "LogD_change_(-0.3, -0.1]",
53
+ "LogD_change_(-0.5, -0.3]",
54
+ "LogD_change_(-0.7, -0.5]",
55
+ "LogD_change_(-0.9, -0.7]",
56
+ "LogD_change_(-1.1, -0.9]",
57
+ "LogD_change_(-1.3, -1.1]",
58
+ "LogD_change_(-1.5, -1.3]",
59
+ "LogD_change_(-1.7, -1.5]",
60
+ "LogD_change_(-1.9, -1.7]",
61
+ "LogD_change_(-2.1, -1.9]",
62
+ "LogD_change_(-2.3, -2.1]",
63
+ "LogD_change_(-2.5, -2.3]",
64
+ "LogD_change_(-2.7, -2.5]",
65
+ "LogD_change_(-2.9, -2.7]",
66
+ "LogD_change_(-3.1, -2.9]",
67
+ "LogD_change_(-3.3, -3.1]",
68
+ "LogD_change_(-3.5, -3.3]",
69
+ "LogD_change_(-3.7, -3.5]",
70
+ "LogD_change_(-3.9, -3.7]",
71
+ "LogD_change_(-4.1, -3.9]",
72
+ "LogD_change_(-4.3, -4.1]",
73
+ "LogD_change_(-4.5, -4.3]",
74
+ "LogD_change_(-4.7, -4.5]",
75
+ "LogD_change_(-4.9, -4.7]",
76
+ "LogD_change_(-5.1, -4.9]",
77
+ "LogD_change_(-5.3, -5.1]",
78
+ "LogD_change_(-5.5, -5.3]",
79
+ "LogD_change_(-5.7, -5.5]",
80
+ "LogD_change_(-inf, -5.7]",
81
+ "Solubility_low->high",
82
+ "Solubility_high->low",
83
+ "Solubility_no_change",
84
+ "Clint_low->high",
85
+ "Clint_high->low",
86
+ "Clint_no_change",
87
+ "<UNUSED_0>",
88
+ "<UNUSED_1>",
89
+ "<UNUSED_2>",
90
+ "<UNUSED_3>",
91
+ "<UNUSED_4>",
92
+ "<UNUSED_5>",
93
+ "<UNUSED_6>",
94
+ "<UNUSED_7>",
95
+ "<UNUSED_8>",
96
+ "<UNUSED_9>",
97
+ "<UNUSED_10>",
98
+ "<UNUSED_11>",
99
+ "<UNUSED_12>",
100
+ "<UNUSED_13>",
101
+ "<UNUSED_14>",
102
+ "<UNUSED_15>",
103
+ "<UNUSED_16>",
104
+ "<UNUSED_17>",
105
+ "<UNUSED_18>",
106
+ "<UNUSED_19>",
107
+ "<UNUSED_20>",
108
+ "<UNUSED_21>",
109
+ "<UNUSED_22>",
110
+ "<UNUSED_23>",
111
+ "<UNUSED_24>",
112
+ "<UNUSED_25>",
113
+ "<UNUSED_26>",
114
+ "<UNUSED_27>",
115
+ "<UNUSED_28>",
116
+ "<UNUSED_29>",
117
+ "<UNUSED_30>",
118
+ "<UNUSED_31>",
119
+ "<UNUSED_32>",
120
+ "<UNUSED_33>",
121
+ "<UNUSED_34>",
122
+ "<UNUSED_35>",
123
+ "<UNUSED_36>",
124
+ "<UNUSED_37>",
125
+ "<UNUSED_38>",
126
+ "<UNUSED_39>",
127
+ "<UNUSED_40>",
128
+ "<UNUSED_41>",
129
+ "<UNUSED_42>",
130
+ "<UNUSED_43>",
131
+ "<UNUSED_44>",
132
+ "<UNUSED_45>",
133
+ "<UNUSED_46>",
134
+ "<UNUSED_47>",
135
+ "<UNUSED_48>",
136
+ "<UNUSED_49>",
137
+ "<UNUSED_50>",
138
+ "<UNUSED_51>",
139
+ "<UNUSED_52>",
140
+ "<UNUSED_53>",
141
+ "<UNUSED_54>",
142
+ "<UNUSED_55>",
143
+ "<UNUSED_56>",
144
+ "<UNUSED_57>",
145
+ "<UNUSED_58>",
146
+ "<UNUSED_59>",
147
+ "<UNUSED_60>",
148
+ "<UNUSED_61>",
149
+ "<UNUSED_62>",
150
+ "<UNUSED_63>",
151
+ "<UNUSED_64>",
152
+ "<UNUSED_65>",
153
+ "<UNUSED_66>",
154
+ "<UNUSED_67>",
155
+ "<UNUSED_68>",
156
+ "<UNUSED_69>",
157
+ "<UNUSED_70>",
158
+ "<UNUSED_71>",
159
+ "<UNUSED_72>",
160
+ "<UNUSED_73>",
161
+ "<UNUSED_74>",
162
+ "<UNUSED_75>",
163
+ "<UNUSED_76>",
164
+ "<UNUSED_77>",
165
+ "<UNUSED_78>",
166
+ "<UNUSED_79>",
167
+ "<UNUSED_80>",
168
+ "<UNUSED_81>",
169
+ "<UNUSED_82>",
170
+ "<UNUSED_83>",
171
+ "<UNUSED_84>",
172
+ "<UNUSED_85>",
173
+ "<UNUSED_86>",
174
+ "<UNUSED_87>",
175
+ "<UNUSED_88>",
176
+ "<UNUSED_89>",
177
+ "<UNUSED_90>",
178
+ "<UNUSED_91>",
179
+ "<UNUSED_92>",
180
+ "<UNUSED_93>",
181
+ "<UNUSED_94>",
182
+ "<UNUSED_95>",
183
+ "<UNUSED_96>",
184
+ "<UNUSED_97>",
185
+ "<UNUSED_98>",
186
+ "<UNUSED_99>",
187
+ "<UNUSED_100>",
188
+ "<UNUSED_101>",
189
+ "<UNUSED_102>",
190
+ "<UNUSED_103>",
191
+ "<UNUSED_104>",
192
+ "<UNUSED_105>",
193
+ "<UNUSED_106>",
194
+ "<UNUSED_107>",
195
+ "<UNUSED_108>",
196
+ "<UNUSED_109>",
197
+ "<UNUSED_110>",
198
+ "<UNUSED_111>",
199
+ "<UNUSED_112>",
200
+ "<UNUSED_113>",
201
+ "<UNUSED_114>",
202
+ "<UNUSED_115>",
203
+ "<UNUSED_116>",
204
+ "<UNUSED_117>",
205
+ "<UNUSED_118>",
206
+ "<UNUSED_119>",
207
+ "<UNUSED_120>",
208
+ "<UNUSED_121>",
209
+ "<UNUSED_122>",
210
+ "<UNUSED_123>",
211
+ "<UNUSED_124>",
212
+ "<UNUSED_125>",
213
+ "<UNUSED_126>",
214
+ "<UNUSED_127>",
215
+ "<UNUSED_128>",
216
+ "<UNUSED_129>",
217
+ "<UNUSED_130>",
218
+ "<UNUSED_131>",
219
+ "<UNUSED_132>",
220
+ "<UNUSED_133>",
221
+ "<UNUSED_134>",
222
+ "<UNUSED_135>",
223
+ "<UNUSED_136>",
224
+ "<UNUSED_137>",
225
+ "<UNUSED_138>",
226
+ "<UNUSED_139>",
227
+ "<UNUSED_140>",
228
+ "<UNUSED_141>",
229
+ "<UNUSED_142>",
230
+ "<UNUSED_143>",
231
+ "<UNUSED_144>",
232
+ "<UNUSED_145>",
233
+ "<UNUSED_146>",
234
+ "<UNUSED_147>",
235
+ "<UNUSED_148>",
236
+ "<UNUSED_149>",
237
+ "<UNUSED_150>",
238
+ "<UNUSED_151>",
239
+ "<UNUSED_152>",
240
+ "<UNUSED_153>",
241
+ "<UNUSED_154>",
242
+ "<UNUSED_155>",
243
+ "<UNUSED_156>",
244
+ "<UNUSED_157>",
245
+ "<UNUSED_158>",
246
+ "<UNUSED_159>",
247
+ "<UNUSED_160>",
248
+ "<UNUSED_161>",
249
+ "<UNUSED_162>",
250
+ "<UNUSED_163>",
251
+ "<UNUSED_164>",
252
+ "<UNUSED_165>",
253
+ "<UNUSED_166>",
254
+ "<UNUSED_167>",
255
+ "<UNUSED_168>",
256
+ "<UNUSED_169>",
257
+ "<UNUSED_170>",
258
+ "<UNUSED_171>",
259
+ "<UNUSED_172>",
260
+ "<UNUSED_173>",
261
+ "<UNUSED_174>",
262
+ "<UNUSED_175>",
263
+ "<UNUSED_176>",
264
+ "<UNUSED_177>",
265
+ "<UNUSED_178>",
266
+ "<UNUSED_179>",
267
+ "<UNUSED_180>",
268
+ "<UNUSED_181>",
269
+ "<UNUSED_182>",
270
+ "<UNUSED_183>",
271
+ "<UNUSED_184>",
272
+ "<UNUSED_185>",
273
+ "<UNUSED_186>",
274
+ "<UNUSED_187>",
275
+ "<UNUSED_188>",
276
+ ">",
277
+ "<RX_6>",
278
+ "<RX_2>",
279
+ "<RX_1>",
280
+ "<RX_3>",
281
+ "<RX_7>",
282
+ "<RX_9>",
283
+ "<RX_5>",
284
+ "<RX_10>",
285
+ "<RX_4>",
286
+ "<RX_8>",
287
+ "C",
288
+ "c",
289
+ "1",
290
+ "(",
291
+ "-",
292
+ "2",
293
+ "s",
294
+ "N",
295
+ "=",
296
+ ")",
297
+ "n",
298
+ "Br",
299
+ "3",
300
+ "O",
301
+ "[nH]",
302
+ "[C@H]",
303
+ "/",
304
+ "\\",
305
+ "[C@@H]",
306
+ "[C@]",
307
+ "o",
308
+ "4",
309
+ "[C@@]",
310
+ "5",
311
+ "6",
312
+ "7",
313
+ "[n+]",
314
+ ".",
315
+ "[Br-]",
316
+ "S",
317
+ "I",
318
+ "F",
319
+ "Cl",
320
+ "[O-]",
321
+ "#",
322
+ "P",
323
+ "[Na+]",
324
+ "[N+]",
325
+ "[Cl-]",
326
+ "8",
327
+ "[Si]",
328
+ "[N-]",
329
+ "[18F]",
330
+ "9",
331
+ "[I-]",
332
+ "B",
333
+ "[S+]",
334
+ "[2H]",
335
+ "[P+]",
336
+ "[125I]",
337
+ "%10",
338
+ "%11",
339
+ "[n-]",
340
+ "[B-]",
341
+ "[O]",
342
+ "[o+]",
343
+ "[N@@+]",
344
+ "[N@+]",
345
+ "[PH]",
346
+ "[se]",
347
+ "[Se]",
348
+ "[s+]",
349
+ "[Li+]",
350
+ "[P@]",
351
+ "[P@@]",
352
+ "[3H]",
353
+ "[K+]",
354
+ "[OH-]",
355
+ "[te+]",
356
+ "[se+]",
357
+ "[Te]",
358
+ "[S@+]",
359
+ "[BH2-]",
360
+ "[11CH3]",
361
+ "[11C]",
362
+ "[Cl+3]",
363
+ "[N@@]",
364
+ "[S-]",
365
+ "[C+]",
366
+ "[P@+]",
367
+ "[C-]",
368
+ "[Zn+2]",
369
+ "[Ca+2]",
370
+ "[Mg+2]",
371
+ "[SeH]",
372
+ "[BH-]",
373
+ "[SH2]",
374
+ "[TeH2]",
375
+ "[SiH4]",
376
+ "[N@]",
377
+ "[14CH2]",
378
+ "[Ag]",
379
+ "[S@@+]",
380
+ "[I+]",
381
+ "[MgH2]",
382
+ "[125IH]",
383
+ "[Se+]",
384
+ "[As]",
385
+ "[SiH2]",
386
+ "[Ra]",
387
+ "[IH2]",
388
+ "[P-]",
389
+ "[Na]",
390
+ "[NH-]",
391
+ "[Cs]",
392
+ "[Zn]",
393
+ "[Li]",
394
+ "[As+]",
395
+ "[te]",
396
+ "[131I]",
397
+ "[Ag+]",
398
+ "p",
399
+ "[Al-3]",
400
+ "[Rb+]",
401
+ "[13CH]",
402
+ "[11CH]",
403
+ "[11CH2]",
404
+ "[14c]",
405
+ "[14C@@]",
406
+ "[127I]",
407
+ "[Mg]",
408
+ "[14C]",
409
+ "[123I]",
410
+ "[124I]",
411
+ "[F-]",
412
+ "[14CH3]",
413
+ "[135I]",
414
+ "%12",
415
+ "[NH+]",
416
+ "[76Br]",
417
+ "[32PH]",
418
+ "[35S]",
419
+ "b",
420
+ "[73Se]",
421
+ "[11C@@H]",
422
+ "[Se-]",
423
+ "[c+]",
424
+ "[14C@H]",
425
+ "[18OH]",
426
+ "[SH]",
427
+ "[S@@]",
428
+ "[Cs+]",
429
+ "[He]",
430
+ "[O+]",
431
+ "[nH+]",
432
+ "[NH2+]",
433
+ "[32P]",
434
+ "[Zn+]",
435
+ "[BH3-]",
436
+ "[I+3]",
437
+ "[Si-]",
438
+ "[SH+]",
439
+ "[19F]",
440
+ "[Br+2]",
441
+ "[I+2]",
442
+ "[Al+3]",
443
+ "[123I-]",
444
+ "[131I-]",
445
+ "[127Xe]",
446
+ "[133Xe]",
447
+ "[89Sr+2]",
448
+ "[N]",
449
+ "[82Rb+]",
450
+ "[75Se]",
451
+ "[Rb]",
452
+ "[81Kr]",
453
+ "[18F-]",
454
+ "[13NH3]",
455
+ "[K]",
456
+ "[Cl+2]",
457
+ "[Zn-2]",
458
+ "[SeH2]",
459
+ "[AsH3]",
460
+ "[Kr]",
461
+ "[Xe]",
462
+ "[S@]",
463
+ "[cH-]",
464
+ "[NH4+]",
465
+ "[Al]",
466
+ "[Si@]",
467
+ "[15n]",
468
+ "[SH-]",
469
+ "[SiH]",
470
+ "[11c]",
471
+ "[OH]",
472
+ "[c-]",
473
+ "[18FH]",
474
+ "[123IH]",
475
+ "[13c]",
476
+ "[13cH]",
477
+ "[14cH]",
478
+ "[Cl+]",
479
+ "[Sr+2]",
480
+ "[CH-]",
481
+ "[Bi]",
482
+ "[B]",
483
+ "[Ba+2]",
484
+ "[Bi+3]",
485
+ "[SiH-]",
486
+ "[b-]",
487
+ "[H+]",
488
+ "[13C]",
489
+ "[OH+]",
490
+ "[14CH]",
491
+ "[15nH]",
492
+ "[CaH2]",
493
+ "[LiH]",
494
+ "[C]",
495
+ "[Ca]",
496
+ "[42K+]",
497
+ "[123Te]",
498
+ "[S-2]",
499
+ "[223Ra+2]",
500
+ "[S]",
501
+ "[Ra+2]",
502
+ "[22Na+]",
503
+ "[125I-]",
504
+ "[85Sr+2]",
505
+ "[PH2]",
506
+ "[SrH2]",
507
+ "[15OH2]",
508
+ "[47Ca+2]",
509
+ "[85SrH2]",
510
+ "[45Ca+2]",
511
+ "[B@@-]",
512
+ "[B@-]",
513
+ "[17F]",
514
+ "[PH2+]",
515
+ "[11C-]",
516
+ "[Mg+]",
517
+ "[NaH]",
518
+ "[P@@+]",
519
+ "[SiH3-]",
520
+ "[P-3]",
521
+ "[KH]",
522
+ "[Be+2]",
523
+ "[NH3+]",
524
+ "[Ag-4]",
525
+ "[18O]",
526
+ "[14C@@H]",
527
+ "[CH]",
528
+ "[CH2]",
529
+ "[O-2]",
530
+ "[124I-]",
531
+ "[As-]",
532
+ "[Ba]",
533
+ "[223Ra]",
534
+ "[82Rb]",
535
+ "[76BrH]",
536
+ "[AsH]",
537
+ "[131Cs]"
538
+ ]
539
+ }
checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/aux/ds_config.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "zero_optimization": {
3
+ "stage": 2,
4
+ "cpu_offload": false,
5
+ "contiguous_gradients": true,
6
+ "overlap_comm": true,
7
+ "allgather_bucket_size": 5e8,
8
+ "reduce_bucket_size": 5e8
9
+ },
10
+ "fp16": {
11
+ "enabled": true,
12
+ "loss_scale": 0.0,
13
+ "initial_scale_power": 32,
14
+ "loss_scale_window": 1000,
15
+ "hysteresis": 2,
16
+ "min_loss_scale": 1000
17
+ }
18
+ }
checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/aux/hparams.yaml ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ acc_batches: 8
2
+ activation: gelu
3
+ aug_prob: 0.5
4
+ augment: all
5
+ batch_size: 64
6
+ clip_grad: 1.0
7
+ d_feedforward: 2048
8
+ d_model: 512
9
+ dropout: 0.1
10
+ epochs: 150
11
+ limit_val_batches: 1.0
12
+ lr: 0.001
13
+ mask_prob: 0.1
14
+ max_seq_len: 512
15
+ num_buckets: 24
16
+ num_heads: 8
17
+ num_layers: 6
18
+ num_steps: 119851
19
+ pad_token_idx: 0
20
+ schedule: cycle
21
+ task: mask_aug
22
+ train_tokens: None
23
+ vocab_size: 523
24
+ warm_up_steps: 8000
25
+ weight_decay: 0.0
checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/aux/hyper_parameters.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "hyper_parameters": {
3
+ "acc_batches": 8,
4
+ "activation": "gelu",
5
+ "aug_prob": 0.5,
6
+ "augment": "all",
7
+ "batch_size": 64,
8
+ "clip_grad": 1.0,
9
+ "d_feedforward": 2048,
10
+ "d_model": 512,
11
+ "dropout": 0.1,
12
+ "epochs": 150,
13
+ "limit_val_batches": 1.0,
14
+ "lr": 0.001,
15
+ "mask_prob": 0.1,
16
+ "max_seq_len": 512,
17
+ "num_buckets": 24,
18
+ "num_heads": 8,
19
+ "num_layers": 6,
20
+ "num_steps": 119851,
21
+ "pad_token_idx": 0,
22
+ "schedule": "cycle",
23
+ "task": "mask_aug",
24
+ "train_tokens": "None",
25
+ "vocab_size": 523,
26
+ "warm_up_steps": 8000,
27
+ "weight_decay": 0.0
28
+ }
29
+ }
checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/benchmarks/default_test.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/benchmarks/metadata.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "benchmark_semantics": "forward_product_prediction_with_mixed_inputs",
3
+ "benchmark_variant": "forward_product_prediction",
4
+ "built_at": "1970-01-01T00:00:00+00:00",
5
+ "checkpoint_id": "uspto_mixed_forward_checkpoint",
6
+ "dataset_id": "uspto_mixed_forward_checkpoint_test",
7
+ "dataset_ref": "uspto_mixed_chemformer_project_release",
8
+ "native_input_format": "pandas_pickled_dataframe",
9
+ "native_sources": [
10
+ {
11
+ "notes": [
12
+ "Zip member: data/seq-to-seq_datasets/uspto_mixed.pickle",
13
+ "Archive: datasets/chemformer/Chemformer-selected.zip"
14
+ ],
15
+ "origin": "extracted_from_local_manual_download_zip",
16
+ "relative_path": "benchmarks/native/uspto_mixed.pickle",
17
+ "source_url": null
18
+ }
19
+ ],
20
+ "notes": [
21
+ "Native Chemformer forward benchmark source extracted from the public Box release uspto_mixed pickle.",
22
+ "benchmarks/default_test.jsonl preserves reactant-side input SMILES plus expected product SMILES from the test partition."
23
+ ],
24
+ "num_examples": 40000,
25
+ "record_schema": {
26
+ "inputs": [
27
+ "reactants"
28
+ ],
29
+ "targets": [
30
+ "expected_products"
31
+ ]
32
+ },
33
+ "split": "test",
34
+ "task_direction": "forward"
35
+ }
checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/manifest.json ADDED
@@ -0,0 +1,154 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auxiliary_files": [
3
+ "aux/bart_vocab.json",
4
+ "aux/bart_vocab_downstream.json",
5
+ "aux/ds_config.json",
6
+ "aux/hparams.yaml",
7
+ "aux/hyper_parameters.json",
8
+ "benchmarks/default_test.jsonl",
9
+ "benchmarks/metadata.json",
10
+ "benchmarks/native/uspto_mixed.pickle"
11
+ ],
12
+ "benchmark_datasets": [
13
+ {
14
+ "benchmark_semantics": "forward_product_prediction_with_mixed_inputs",
15
+ "benchmark_variant": "forward_product_prediction",
16
+ "dataset_id": "uspto_mixed_forward_checkpoint_test",
17
+ "dataset_ref": "uspto_mixed_chemformer_project_release",
18
+ "default": true,
19
+ "format": "jsonl",
20
+ "metadata_path": "benchmarks/metadata.json",
21
+ "native_input_format": "pandas_pickled_dataframe",
22
+ "native_source_paths": [
23
+ "benchmarks/native/uspto_mixed.pickle"
24
+ ],
25
+ "notes": [
26
+ "Native Chemformer forward benchmark source extracted from the public Box release uspto_mixed pickle.",
27
+ "benchmarks/default_test.jsonl preserves reactant-side input SMILES plus expected product SMILES from the test partition."
28
+ ],
29
+ "num_examples": 40000,
30
+ "path": "benchmarks/default_test.jsonl",
31
+ "record_schema": {
32
+ "inputs": [
33
+ "reactants"
34
+ ],
35
+ "targets": [
36
+ "expected_products"
37
+ ]
38
+ },
39
+ "split": "test",
40
+ "task_direction": "forward"
41
+ }
42
+ ],
43
+ "checkpoint_id": "uspto_mixed_forward_checkpoint",
44
+ "contained_artifact_bindings": [
45
+ {
46
+ "path": "hparams.yaml",
47
+ "role": "checkpoint_hyperparameters"
48
+ },
49
+ {
50
+ "path": "last.ckpt",
51
+ "role": "model_checkpoint"
52
+ }
53
+ ],
54
+ "conversion": {
55
+ "converted_at": "1970-01-01T00:00:00+00:00",
56
+ "source_format": "pytorch_lightning_ckpt",
57
+ "status": "converted",
58
+ "tool_version": "initial"
59
+ },
60
+ "display_name": "uspto_mixed forward checkpoint",
61
+ "embedding_supported": false,
62
+ "fine_tuning": {
63
+ "adapter_overlay_dir": "adapters",
64
+ "base_weights_path": "model.safetensors",
65
+ "notes": [
66
+ "This package preserves immutable base weights in model.safetensors and reserves adapters/ for future fine-tuning or LoRA overlays."
67
+ ]
68
+ },
69
+ "format_version": 1,
70
+ "generated_at": "1970-01-01T00:00:00+00:00",
71
+ "input_schema": {
72
+ "reactants_field": "reactants_mol",
73
+ "reagents_field": null,
74
+ "serialization": "dot_joined_smiles",
75
+ "source_dataset_member": "data/seq-to-seq_datasets/uspto_mixed.pickle",
76
+ "target_field": "products_mol"
77
+ },
78
+ "model_id": "chemformer_forward",
79
+ "normalized_checkpoint_id": "uspto_mixed_forward_checkpoint",
80
+ "package_layout": {
81
+ "adapters_dir": "adapters",
82
+ "aux_dir": "aux",
83
+ "benchmarks_dir": "benchmarks",
84
+ "weights_path": "model.safetensors"
85
+ },
86
+ "source_registry_refs": [
87
+ {
88
+ "checkpoint_id": "uspto_mixed_forward_checkpoint",
89
+ "model_id": "chemformer_forward"
90
+ }
91
+ ],
92
+ "split_provenance": [
93
+ {
94
+ "canonical_split_family": "chemformer_zinc15_100m_train_valid_test",
95
+ "dataset_ref": "zinc15_100m_chemformer_pretraining",
96
+ "split_names": [
97
+ "train",
98
+ "valid",
99
+ "test"
100
+ ],
101
+ "stage": "pretraining_initialization"
102
+ },
103
+ {
104
+ "canonical_split_family": "chemformer_project_usptomixed_pickle_train_val_test",
105
+ "dataset_ref": "uspto_mixed_chemformer_project_release",
106
+ "split_names": [
107
+ "train",
108
+ "val",
109
+ "test"
110
+ ],
111
+ "stage": "forward_fine_tuning_and_evaluation"
112
+ },
113
+ {
114
+ "path": "hparams.yaml"
115
+ },
116
+ {
117
+ "path": "last.ckpt"
118
+ }
119
+ ],
120
+ "task_direction": "forward",
121
+ "training_dataset_refs": [
122
+ "uspto_mixed_chemformer_project_release"
123
+ ],
124
+ "training_stage_bindings": [
125
+ {
126
+ "canonical_split_family": "chemformer_zinc15_100m_train_valid_test",
127
+ "dataset_ref": "zinc15_100m_chemformer_pretraining",
128
+ "split_names": [
129
+ "train",
130
+ "valid",
131
+ "test"
132
+ ],
133
+ "stage": "pretraining_initialization"
134
+ },
135
+ {
136
+ "canonical_split_family": "chemformer_project_usptomixed_pickle_train_val_test",
137
+ "dataset_ref": "uspto_mixed_chemformer_project_release",
138
+ "split_names": [
139
+ "train",
140
+ "val",
141
+ "test"
142
+ ],
143
+ "stage": "forward_fine_tuning_and_evaluation"
144
+ }
145
+ ],
146
+ "weights": {
147
+ "dtype_summary": "torch_saved_tensors",
148
+ "framework": "safetensors",
149
+ "path": "model.safetensors",
150
+ "tensor_namespace": "lightning_state_dict"
151
+ },
152
+ "wrapper_config": "wrapper_config.json",
153
+ "wrapper_family": "chemformer_bart"
154
+ }
checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/provenance.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "conversion_notes": [
3
+ "Extracted obj['state_dict'] from the Lightning checkpoint and wrote it to safetensors."
4
+ ],
5
+ "copied_sidecars": [
6
+ "aux/bart_vocab.json",
7
+ "aux/bart_vocab_downstream.json",
8
+ "aux/ds_config.json",
9
+ "aux/hparams.yaml",
10
+ "aux/hyper_parameters.json"
11
+ ],
12
+ "inner_members": [
13
+ {
14
+ "archive_path": "datasets/chemformer/Chemformer-selected.zip",
15
+ "member": "models/fined-tuned/uspto_mixed/last.ckpt",
16
+ "role": "weights_source"
17
+ },
18
+ {
19
+ "archive_path": "datasets/chemformer/Chemformer-selected.zip",
20
+ "member": "models/fined-tuned/uspto_mixed/hparams.yaml",
21
+ "role": "hparams_source"
22
+ }
23
+ ],
24
+ "normalized_checkpoint_id": "uspto_mixed_forward_checkpoint",
25
+ "source_artifacts": [
26
+ {
27
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/chemformer/Chemformer-selected.zip",
28
+ "kind": "local_repo_file",
29
+ "path": "datasets/chemformer/Chemformer-selected.zip",
30
+ "role": "source_artifact",
31
+ "sha256": "84bce87299a05784b9141d7979f9f3a1b8762ede0f602db76e85e8975c1b464e",
32
+ "size": 14770784818
33
+ },
34
+ {
35
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/source_snapshots/chemformer/repo/bart_vocab.json",
36
+ "kind": "local_repo_file",
37
+ "path": "datasets/source_snapshots/chemformer/repo/bart_vocab.json",
38
+ "role": "source_artifact",
39
+ "sha256": "8b17254dfde43cbac758e697d3939c0100ba2d5da341031e92fd6b8de6e87cfb",
40
+ "size": 15954
41
+ },
42
+ {
43
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/source_snapshots/chemformer/repo/bart_vocab_downstream.json",
44
+ "kind": "local_repo_file",
45
+ "path": "datasets/source_snapshots/chemformer/repo/bart_vocab_downstream.json",
46
+ "role": "source_artifact",
47
+ "sha256": "8c15ea5b60855344ddb24586b14c33811e89c93adf4189865045fcf3ad89459c",
48
+ "size": 16050
49
+ },
50
+ {
51
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/source_snapshots/chemformer/repo/ds_config.json",
52
+ "kind": "local_repo_file",
53
+ "path": "datasets/source_snapshots/chemformer/repo/ds_config.json",
54
+ "role": "source_artifact",
55
+ "sha256": "ed5d259989f3d3e3d423ecc3c1f7f56dda31ea1debed0a29977be2270fc1b588",
56
+ "size": 426
57
+ }
58
+ ]
59
+ }
checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint/wrapper_config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architecture_class": "bart",
3
+ "device_supported": [
4
+ "cpu",
5
+ "cuda"
6
+ ],
7
+ "fine_tuning": {
8
+ "adapter_overlay_dir": "adapters",
9
+ "base_weights_path": "model.safetensors",
10
+ "notes": [
11
+ "Base weights stay immutable in model.safetensors; future fine-tuning or LoRA adapters should be added as overlays under adapters/."
12
+ ]
13
+ },
14
+ "hparams_file": "aux/hparams.yaml",
15
+ "hyper_parameters": "aux/hyper_parameters.json",
16
+ "loader": {
17
+ "state_dict_format": "safetensors",
18
+ "strict": true
19
+ },
20
+ "package_layout": {
21
+ "adapters_dir": "adapters",
22
+ "aux_dir": "aux",
23
+ "weights_path": "model.safetensors"
24
+ },
25
+ "task_direction": "forward",
26
+ "vocab_files": [
27
+ "aux/bart_vocab.json",
28
+ "aux/bart_vocab_downstream.json"
29
+ ],
30
+ "wrapper_family": "chemformer_bart"
31
+ }
checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/SHA256SUMS ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 8b17254dfde43cbac758e697d3939c0100ba2d5da341031e92fd6b8de6e87cfb aux/bart_vocab.json
2
+ 8c15ea5b60855344ddb24586b14c33811e89c93adf4189865045fcf3ad89459c aux/bart_vocab_downstream.json
3
+ ed5d259989f3d3e3d423ecc3c1f7f56dda31ea1debed0a29977be2270fc1b588 aux/ds_config.json
4
+ 8cdffda14b6bf90b5a93a090a73c6d34e8709739abad86fd0b74b48d2c0f249b aux/hparams.yaml
5
+ 0c7930b00473a1ef66b9d5062fa0bd394c1feec7ed7eb61ad8351abdd6644d61 aux/hyper_parameters.json
6
+ b43c8226e0105e825d75591cccfbc256703ffa285ef0fdb00b30635c07cf042f benchmarks/default_test.jsonl
7
+ 8f1a09618a6258f10ec07c5062b1c333c7f7f43e801a260e61d2e0828e51ea74 benchmarks/metadata.json
8
+ c676f7f247adcc1c70a6b3be66e2054d0dc827357f330291e3024f7b29ab61b1 benchmarks/native/uspto_sep.pickle
9
+ 3df37630a26ada94352cb2a85aed5eb4e626e20b2b34771c857479cd8d9af0cb manifest.json
10
+ c24ff3dbbec7acf4b21a4daadd705c97a975808442cd05cc96b8cdcc98c02bae model.safetensors
11
+ b83c80df068a7a5e94c7e9c4a93c8f041cb010c77f7f3383b9167a35f1212899 provenance.json
12
+ 714da361b86c29cedef958d3cdb95c7d986b62a337d5649aab5a383fab3f5fab wrapper_config.json
checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/aux/bart_vocab.json ADDED
@@ -0,0 +1,539 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "properties": {
3
+ "regex": "(LogD_change_\\(-0.1, 0.1\\]|LogD_change_\\(0.1, 0.3\\]|LogD_change_\\(0.3, 0.5\\]|LogD_change_\\(0.5, 0.7\\]|LogD_change_\\(0.7, 0.9\\]|LogD_change_\\(0.9, 1.1\\]|LogD_change_\\(1.1, 1.3\\]|LogD_change_\\(1.3, 1.5\\]|LogD_change_\\(1.5, 1.7\\]|LogD_change_\\(1.7, 1.9\\]|LogD_change_\\(1.9, 2.1\\]|LogD_change_\\(2.1, 2.3\\]|LogD_change_\\(2.3, 2.5\\]|LogD_change_\\(2.5, 2.7\\]|LogD_change_\\(2.7, 2.9\\]|LogD_change_\\(2.9, 3.1\\]|LogD_change_\\(3.1, 3.3\\]|LogD_change_\\(3.3, 3.5\\]|LogD_change_\\(3.5, 3.7\\]|LogD_change_\\(3.7, 3.9\\]|LogD_change_\\(3.9, 4.1\\]|LogD_change_\\(4.1, 4.3\\]|LogD_change_\\(4.3, 4.5\\]|LogD_change_\\(4.5, 4.7\\]|LogD_change_\\(4.7, 4.9\\]|LogD_change_\\(4.9, 5.1\\]|LogD_change_\\(5.1, 5.3\\]|LogD_change_\\(5.3, 5.5\\]|LogD_change_\\(5.5, 5.7\\]|LogD_change_\\(5.7, 5.9\\]|LogD_change_\\(5.9, inf\\]|LogD_change_\\(-0.3, -0.1\\]|LogD_change_\\(-0.5, -0.3\\]|LogD_change_\\(-0.7, -0.5\\]|LogD_change_\\(-0.9, -0.7\\]|LogD_change_\\(-1.1, -0.9\\]|LogD_change_\\(-1.3, -1.1\\]|LogD_change_\\(-1.5, -1.3\\]|LogD_change_\\(-1.7, -1.5\\]|LogD_change_\\(-1.9, -1.7\\]|LogD_change_\\(-2.1, -1.9\\]|LogD_change_\\(-2.3, -2.1\\]|LogD_change_\\(-2.5, -2.3\\]|LogD_change_\\(-2.7, -2.5\\]|LogD_change_\\(-2.9, -2.7\\]|LogD_change_\\(-3.1, -2.9\\]|LogD_change_\\(-3.3, -3.1\\]|LogD_change_\\(-3.5, -3.3\\]|LogD_change_\\(-3.7, -3.5\\]|LogD_change_\\(-3.9, -3.7\\]|LogD_change_\\(-4.1, -3.9\\]|LogD_change_\\(-4.3, -4.1\\]|LogD_change_\\(-4.5, -4.3\\]|LogD_change_\\(-4.7, -4.5\\]|LogD_change_\\(-4.9, -4.7\\]|LogD_change_\\(-5.1, -4.9\\]|LogD_change_\\(-5.3, -5.1\\]|LogD_change_\\(-5.5, -5.3\\]|LogD_change_\\(-5.7, -5.5\\]|LogD_change_\\(-inf, -5.7\\]|Solubility_low->high|Solubility_high->low|Solubility_no_change|Clint_low->high|Clint_high->low|Clint_no_change|<UNUSED_0>|<UNUSED_1>|<UNUSED_2>|<UNUSED_3>|<UNUSED_4>|<UNUSED_5>|<UNUSED_6>|<UNUSED_7>|<UNUSED_8>|<UNUSED_9>|<UNUSED_10>|<UNUSED_11>|<UNUSED_12>|<UNUSED_13>|<UNUSED_14>|<UNUSED_15>|<UNUSED_16>|<UNUSED_17>|<UNUSED_18>|<UNUSED_19>|<UNUSED_20>|<UNUSED_21>|<UNUSED_22>|<UNUSED_23>|<UNUSED_24>|<UNUSED_25>|<UNUSED_26>|<UNUSED_27>|<UNUSED_28>|<UNUSED_29>|<UNUSED_30>|<UNUSED_31>|<UNUSED_32>|<UNUSED_33>|<UNUSED_34>|<UNUSED_35>|<UNUSED_36>|<UNUSED_37>|<UNUSED_38>|<UNUSED_39>|<UNUSED_40>|<UNUSED_41>|<UNUSED_42>|<UNUSED_43>|<UNUSED_44>|<UNUSED_45>|<UNUSED_46>|<UNUSED_47>|<UNUSED_48>|<UNUSED_49>|<UNUSED_50>|<UNUSED_51>|<UNUSED_52>|<UNUSED_53>|<UNUSED_54>|<UNUSED_55>|<UNUSED_56>|<UNUSED_57>|<UNUSED_58>|<UNUSED_59>|<UNUSED_60>|<UNUSED_61>|<UNUSED_62>|<UNUSED_63>|<UNUSED_64>|<UNUSED_65>|<UNUSED_66>|<UNUSED_67>|<UNUSED_68>|<UNUSED_69>|<UNUSED_70>|<UNUSED_71>|<UNUSED_72>|<UNUSED_73>|<UNUSED_74>|<UNUSED_75>|<UNUSED_76>|<UNUSED_77>|<UNUSED_78>|<UNUSED_79>|<UNUSED_80>|<UNUSED_81>|<UNUSED_82>|<UNUSED_83>|<UNUSED_84>|<UNUSED_85>|<UNUSED_86>|<UNUSED_87>|<UNUSED_88>|<UNUSED_89>|<UNUSED_90>|<UNUSED_91>|<UNUSED_92>|<UNUSED_93>|<UNUSED_94>|<UNUSED_95>|<UNUSED_96>|<UNUSED_97>|<UNUSED_98>|<UNUSED_99>|<UNUSED_100>|<UNUSED_101>|<UNUSED_102>|<UNUSED_103>|<UNUSED_104>|<UNUSED_105>|<UNUSED_106>|<UNUSED_107>|<UNUSED_108>|<UNUSED_109>|<UNUSED_110>|<UNUSED_111>|<UNUSED_112>|<UNUSED_113>|<UNUSED_114>|<UNUSED_115>|<UNUSED_116>|<UNUSED_117>|<UNUSED_118>|<UNUSED_119>|<UNUSED_120>|<UNUSED_121>|<UNUSED_122>|<UNUSED_123>|<UNUSED_124>|<UNUSED_125>|<UNUSED_126>|<UNUSED_127>|<UNUSED_128>|<UNUSED_129>|<UNUSED_130>|<UNUSED_131>|<UNUSED_132>|<UNUSED_133>|<UNUSED_134>|<UNUSED_135>|<UNUSED_136>|<UNUSED_137>|<UNUSED_138>|<UNUSED_139>|<UNUSED_140>|<UNUSED_141>|<UNUSED_142>|<UNUSED_143>|<UNUSED_144>|<UNUSED_145>|<UNUSED_146>|<UNUSED_147>|<UNUSED_148>|<UNUSED_149>|<UNUSED_150>|<UNUSED_151>|<UNUSED_152>|<UNUSED_153>|<UNUSED_154>|<UNUSED_155>|<UNUSED_156>|<UNUSED_157>|<UNUSED_158>|<UNUSED_159>|<UNUSED_160>|<UNUSED_161>|<UNUSED_162>|<UNUSED_163>|<UNUSED_164>|<UNUSED_165>|<UNUSED_166>|<UNUSED_167>|<UNUSED_168>|<UNUSED_169>|<UNUSED_170>|<UNUSED_171>|<UNUSED_172>|<UNUSED_173>|<UNUSED_174>|<UNUSED_175>|<UNUSED_176>|<UNUSED_177>|<UNUSED_178>|<UNUSED_179>|<UNUSED_180>|<UNUSED_181>|<UNUSED_182>|<UNUSED_183>|<UNUSED_184>|<UNUSED_185>|<UNUSED_186>|<UNUSED_187>|<UNUSED_188>|<UNUSED_189>|<UNUSED_190>|<UNUSED_191>|<UNUSED_192>|<UNUSED_193>|<UNUSED_194>|<UNUSED_195>|<UNUSED_196>|<UNUSED_197>|<UNUSED_198>|<UNUSED_199>|\\[[^\\]]+]|Br?|Cl?|N|O|S|P|F|I|b|c|n|o|s|p|\\(|\\)|\\.|=|#|-|\\+|\\\\|\\/|:|~|@|\\?|>|\\*|\\$|\\%[0-9]{2}|[0-9]|.)",
4
+ "special_tokens": {
5
+ "start": "^",
6
+ "end": "&",
7
+ "pad": "<PAD>",
8
+ "unknown": "?",
9
+ "mask": "<MASK>",
10
+ "sep": "<SEP>"
11
+ },
12
+ "chem_start_idx": 272
13
+ },
14
+ "vocabulary": [
15
+ "<PAD>",
16
+ "?",
17
+ "^",
18
+ "&",
19
+ "<MASK>",
20
+ "<SEP>",
21
+ "LogD_change_(-0.1, 0.1]",
22
+ "LogD_change_(0.1, 0.3]",
23
+ "LogD_change_(0.3, 0.5]",
24
+ "LogD_change_(0.5, 0.7]",
25
+ "LogD_change_(0.7, 0.9]",
26
+ "LogD_change_(0.9, 1.1]",
27
+ "LogD_change_(1.1, 1.3]",
28
+ "LogD_change_(1.3, 1.5]",
29
+ "LogD_change_(1.5, 1.7]",
30
+ "LogD_change_(1.7, 1.9]",
31
+ "LogD_change_(1.9, 2.1]",
32
+ "LogD_change_(2.1, 2.3]",
33
+ "LogD_change_(2.3, 2.5]",
34
+ "LogD_change_(2.5, 2.7]",
35
+ "LogD_change_(2.7, 2.9]",
36
+ "LogD_change_(2.9, 3.1]",
37
+ "LogD_change_(3.1, 3.3]",
38
+ "LogD_change_(3.3, 3.5]",
39
+ "LogD_change_(3.5, 3.7]",
40
+ "LogD_change_(3.7, 3.9]",
41
+ "LogD_change_(3.9, 4.1]",
42
+ "LogD_change_(4.1, 4.3]",
43
+ "LogD_change_(4.3, 4.5]",
44
+ "LogD_change_(4.5, 4.7]",
45
+ "LogD_change_(4.7, 4.9]",
46
+ "LogD_change_(4.9, 5.1]",
47
+ "LogD_change_(5.1, 5.3]",
48
+ "LogD_change_(5.3, 5.5]",
49
+ "LogD_change_(5.5, 5.7]",
50
+ "LogD_change_(5.7, 5.9]",
51
+ "LogD_change_(5.9, inf]",
52
+ "LogD_change_(-0.3, -0.1]",
53
+ "LogD_change_(-0.5, -0.3]",
54
+ "LogD_change_(-0.7, -0.5]",
55
+ "LogD_change_(-0.9, -0.7]",
56
+ "LogD_change_(-1.1, -0.9]",
57
+ "LogD_change_(-1.3, -1.1]",
58
+ "LogD_change_(-1.5, -1.3]",
59
+ "LogD_change_(-1.7, -1.5]",
60
+ "LogD_change_(-1.9, -1.7]",
61
+ "LogD_change_(-2.1, -1.9]",
62
+ "LogD_change_(-2.3, -2.1]",
63
+ "LogD_change_(-2.5, -2.3]",
64
+ "LogD_change_(-2.7, -2.5]",
65
+ "LogD_change_(-2.9, -2.7]",
66
+ "LogD_change_(-3.1, -2.9]",
67
+ "LogD_change_(-3.3, -3.1]",
68
+ "LogD_change_(-3.5, -3.3]",
69
+ "LogD_change_(-3.7, -3.5]",
70
+ "LogD_change_(-3.9, -3.7]",
71
+ "LogD_change_(-4.1, -3.9]",
72
+ "LogD_change_(-4.3, -4.1]",
73
+ "LogD_change_(-4.5, -4.3]",
74
+ "LogD_change_(-4.7, -4.5]",
75
+ "LogD_change_(-4.9, -4.7]",
76
+ "LogD_change_(-5.1, -4.9]",
77
+ "LogD_change_(-5.3, -5.1]",
78
+ "LogD_change_(-5.5, -5.3]",
79
+ "LogD_change_(-5.7, -5.5]",
80
+ "LogD_change_(-inf, -5.7]",
81
+ "Solubility_low->high",
82
+ "Solubility_high->low",
83
+ "Solubility_no_change",
84
+ "Clint_low->high",
85
+ "Clint_high->low",
86
+ "Clint_no_change",
87
+ "<UNUSED_0>",
88
+ "<UNUSED_1>",
89
+ "<UNUSED_2>",
90
+ "<UNUSED_3>",
91
+ "<UNUSED_4>",
92
+ "<UNUSED_5>",
93
+ "<UNUSED_6>",
94
+ "<UNUSED_7>",
95
+ "<UNUSED_8>",
96
+ "<UNUSED_9>",
97
+ "<UNUSED_10>",
98
+ "<UNUSED_11>",
99
+ "<UNUSED_12>",
100
+ "<UNUSED_13>",
101
+ "<UNUSED_14>",
102
+ "<UNUSED_15>",
103
+ "<UNUSED_16>",
104
+ "<UNUSED_17>",
105
+ "<UNUSED_18>",
106
+ "<UNUSED_19>",
107
+ "<UNUSED_20>",
108
+ "<UNUSED_21>",
109
+ "<UNUSED_22>",
110
+ "<UNUSED_23>",
111
+ "<UNUSED_24>",
112
+ "<UNUSED_25>",
113
+ "<UNUSED_26>",
114
+ "<UNUSED_27>",
115
+ "<UNUSED_28>",
116
+ "<UNUSED_29>",
117
+ "<UNUSED_30>",
118
+ "<UNUSED_31>",
119
+ "<UNUSED_32>",
120
+ "<UNUSED_33>",
121
+ "<UNUSED_34>",
122
+ "<UNUSED_35>",
123
+ "<UNUSED_36>",
124
+ "<UNUSED_37>",
125
+ "<UNUSED_38>",
126
+ "<UNUSED_39>",
127
+ "<UNUSED_40>",
128
+ "<UNUSED_41>",
129
+ "<UNUSED_42>",
130
+ "<UNUSED_43>",
131
+ "<UNUSED_44>",
132
+ "<UNUSED_45>",
133
+ "<UNUSED_46>",
134
+ "<UNUSED_47>",
135
+ "<UNUSED_48>",
136
+ "<UNUSED_49>",
137
+ "<UNUSED_50>",
138
+ "<UNUSED_51>",
139
+ "<UNUSED_52>",
140
+ "<UNUSED_53>",
141
+ "<UNUSED_54>",
142
+ "<UNUSED_55>",
143
+ "<UNUSED_56>",
144
+ "<UNUSED_57>",
145
+ "<UNUSED_58>",
146
+ "<UNUSED_59>",
147
+ "<UNUSED_60>",
148
+ "<UNUSED_61>",
149
+ "<UNUSED_62>",
150
+ "<UNUSED_63>",
151
+ "<UNUSED_64>",
152
+ "<UNUSED_65>",
153
+ "<UNUSED_66>",
154
+ "<UNUSED_67>",
155
+ "<UNUSED_68>",
156
+ "<UNUSED_69>",
157
+ "<UNUSED_70>",
158
+ "<UNUSED_71>",
159
+ "<UNUSED_72>",
160
+ "<UNUSED_73>",
161
+ "<UNUSED_74>",
162
+ "<UNUSED_75>",
163
+ "<UNUSED_76>",
164
+ "<UNUSED_77>",
165
+ "<UNUSED_78>",
166
+ "<UNUSED_79>",
167
+ "<UNUSED_80>",
168
+ "<UNUSED_81>",
169
+ "<UNUSED_82>",
170
+ "<UNUSED_83>",
171
+ "<UNUSED_84>",
172
+ "<UNUSED_85>",
173
+ "<UNUSED_86>",
174
+ "<UNUSED_87>",
175
+ "<UNUSED_88>",
176
+ "<UNUSED_89>",
177
+ "<UNUSED_90>",
178
+ "<UNUSED_91>",
179
+ "<UNUSED_92>",
180
+ "<UNUSED_93>",
181
+ "<UNUSED_94>",
182
+ "<UNUSED_95>",
183
+ "<UNUSED_96>",
184
+ "<UNUSED_97>",
185
+ "<UNUSED_98>",
186
+ "<UNUSED_99>",
187
+ "<UNUSED_100>",
188
+ "<UNUSED_101>",
189
+ "<UNUSED_102>",
190
+ "<UNUSED_103>",
191
+ "<UNUSED_104>",
192
+ "<UNUSED_105>",
193
+ "<UNUSED_106>",
194
+ "<UNUSED_107>",
195
+ "<UNUSED_108>",
196
+ "<UNUSED_109>",
197
+ "<UNUSED_110>",
198
+ "<UNUSED_111>",
199
+ "<UNUSED_112>",
200
+ "<UNUSED_113>",
201
+ "<UNUSED_114>",
202
+ "<UNUSED_115>",
203
+ "<UNUSED_116>",
204
+ "<UNUSED_117>",
205
+ "<UNUSED_118>",
206
+ "<UNUSED_119>",
207
+ "<UNUSED_120>",
208
+ "<UNUSED_121>",
209
+ "<UNUSED_122>",
210
+ "<UNUSED_123>",
211
+ "<UNUSED_124>",
212
+ "<UNUSED_125>",
213
+ "<UNUSED_126>",
214
+ "<UNUSED_127>",
215
+ "<UNUSED_128>",
216
+ "<UNUSED_129>",
217
+ "<UNUSED_130>",
218
+ "<UNUSED_131>",
219
+ "<UNUSED_132>",
220
+ "<UNUSED_133>",
221
+ "<UNUSED_134>",
222
+ "<UNUSED_135>",
223
+ "<UNUSED_136>",
224
+ "<UNUSED_137>",
225
+ "<UNUSED_138>",
226
+ "<UNUSED_139>",
227
+ "<UNUSED_140>",
228
+ "<UNUSED_141>",
229
+ "<UNUSED_142>",
230
+ "<UNUSED_143>",
231
+ "<UNUSED_144>",
232
+ "<UNUSED_145>",
233
+ "<UNUSED_146>",
234
+ "<UNUSED_147>",
235
+ "<UNUSED_148>",
236
+ "<UNUSED_149>",
237
+ "<UNUSED_150>",
238
+ "<UNUSED_151>",
239
+ "<UNUSED_152>",
240
+ "<UNUSED_153>",
241
+ "<UNUSED_154>",
242
+ "<UNUSED_155>",
243
+ "<UNUSED_156>",
244
+ "<UNUSED_157>",
245
+ "<UNUSED_158>",
246
+ "<UNUSED_159>",
247
+ "<UNUSED_160>",
248
+ "<UNUSED_161>",
249
+ "<UNUSED_162>",
250
+ "<UNUSED_163>",
251
+ "<UNUSED_164>",
252
+ "<UNUSED_165>",
253
+ "<UNUSED_166>",
254
+ "<UNUSED_167>",
255
+ "<UNUSED_168>",
256
+ "<UNUSED_169>",
257
+ "<UNUSED_170>",
258
+ "<UNUSED_171>",
259
+ "<UNUSED_172>",
260
+ "<UNUSED_173>",
261
+ "<UNUSED_174>",
262
+ "<UNUSED_175>",
263
+ "<UNUSED_176>",
264
+ "<UNUSED_177>",
265
+ "<UNUSED_178>",
266
+ "<UNUSED_179>",
267
+ "<UNUSED_180>",
268
+ "<UNUSED_181>",
269
+ "<UNUSED_182>",
270
+ "<UNUSED_183>",
271
+ "<UNUSED_184>",
272
+ "<UNUSED_185>",
273
+ "<UNUSED_186>",
274
+ "<UNUSED_187>",
275
+ "<UNUSED_188>",
276
+ "<UNUSED_189>",
277
+ "<UNUSED_190>",
278
+ "<UNUSED_191>",
279
+ "<UNUSED_192>",
280
+ "<UNUSED_193>",
281
+ "<UNUSED_194>",
282
+ "<UNUSED_195>",
283
+ "<UNUSED_196>",
284
+ "<UNUSED_197>",
285
+ "<UNUSED_198>",
286
+ "<UNUSED_199>",
287
+ "C",
288
+ "c",
289
+ "1",
290
+ "(",
291
+ "-",
292
+ "2",
293
+ "s",
294
+ "N",
295
+ "=",
296
+ ")",
297
+ "n",
298
+ "Br",
299
+ "3",
300
+ "O",
301
+ "[nH]",
302
+ "[C@H]",
303
+ "/",
304
+ "\\",
305
+ "[C@@H]",
306
+ "[C@]",
307
+ "o",
308
+ "4",
309
+ "[C@@]",
310
+ "5",
311
+ "6",
312
+ "7",
313
+ "[n+]",
314
+ ".",
315
+ "[Br-]",
316
+ "S",
317
+ "I",
318
+ "F",
319
+ "Cl",
320
+ "[O-]",
321
+ "#",
322
+ "P",
323
+ "[Na+]",
324
+ "[N+]",
325
+ "[Cl-]",
326
+ "8",
327
+ "[Si]",
328
+ "[N-]",
329
+ "[18F]",
330
+ "9",
331
+ "[I-]",
332
+ "B",
333
+ "[S+]",
334
+ "[2H]",
335
+ "[P+]",
336
+ "[125I]",
337
+ "%10",
338
+ "%11",
339
+ "[n-]",
340
+ "[B-]",
341
+ "[O]",
342
+ "[o+]",
343
+ "[N@@+]",
344
+ "[N@+]",
345
+ "[PH]",
346
+ "[se]",
347
+ "[Se]",
348
+ "[s+]",
349
+ "[Li+]",
350
+ "[P@]",
351
+ "[P@@]",
352
+ "[3H]",
353
+ "[K+]",
354
+ "[OH-]",
355
+ "[te+]",
356
+ "[se+]",
357
+ "[Te]",
358
+ "[S@+]",
359
+ "[BH2-]",
360
+ "[11CH3]",
361
+ "[11C]",
362
+ "[Cl+3]",
363
+ "[N@@]",
364
+ "[S-]",
365
+ "[C+]",
366
+ "[P@+]",
367
+ "[C-]",
368
+ "[Zn+2]",
369
+ "[Ca+2]",
370
+ "[Mg+2]",
371
+ "[SeH]",
372
+ "[BH-]",
373
+ "[SH2]",
374
+ "[TeH2]",
375
+ "[SiH4]",
376
+ "[N@]",
377
+ "[14CH2]",
378
+ "[Ag]",
379
+ "[S@@+]",
380
+ "[I+]",
381
+ "[MgH2]",
382
+ "[125IH]",
383
+ "[Se+]",
384
+ "[As]",
385
+ "[SiH2]",
386
+ "[Ra]",
387
+ "[IH2]",
388
+ "[P-]",
389
+ "[Na]",
390
+ "[NH-]",
391
+ "[Cs]",
392
+ "[Zn]",
393
+ "[Li]",
394
+ "[As+]",
395
+ "[te]",
396
+ "[131I]",
397
+ "[Ag+]",
398
+ "p",
399
+ "[Al-3]",
400
+ "[Rb+]",
401
+ "[13CH]",
402
+ "[11CH]",
403
+ "[11CH2]",
404
+ "[14c]",
405
+ "[14C@@]",
406
+ "[127I]",
407
+ "[Mg]",
408
+ "[14C]",
409
+ "[123I]",
410
+ "[124I]",
411
+ "[F-]",
412
+ "[14CH3]",
413
+ "[135I]",
414
+ "%12",
415
+ "[NH+]",
416
+ "[76Br]",
417
+ "[32PH]",
418
+ "[35S]",
419
+ "b",
420
+ "[73Se]",
421
+ "[11C@@H]",
422
+ "[Se-]",
423
+ "[c+]",
424
+ "[14C@H]",
425
+ "[18OH]",
426
+ "[SH]",
427
+ "[S@@]",
428
+ "[Cs+]",
429
+ "[He]",
430
+ "[O+]",
431
+ "[nH+]",
432
+ "[NH2+]",
433
+ "[32P]",
434
+ "[Zn+]",
435
+ "[BH3-]",
436
+ "[I+3]",
437
+ "[Si-]",
438
+ "[SH+]",
439
+ "[19F]",
440
+ "[Br+2]",
441
+ "[I+2]",
442
+ "[Al+3]",
443
+ "[123I-]",
444
+ "[131I-]",
445
+ "[127Xe]",
446
+ "[133Xe]",
447
+ "[89Sr+2]",
448
+ "[N]",
449
+ "[82Rb+]",
450
+ "[75Se]",
451
+ "[Rb]",
452
+ "[81Kr]",
453
+ "[18F-]",
454
+ "[13NH3]",
455
+ "[K]",
456
+ "[Cl+2]",
457
+ "[Zn-2]",
458
+ "[SeH2]",
459
+ "[AsH3]",
460
+ "[Kr]",
461
+ "[Xe]",
462
+ "[S@]",
463
+ "[cH-]",
464
+ "[NH4+]",
465
+ "[Al]",
466
+ "[Si@]",
467
+ "[15n]",
468
+ "[SH-]",
469
+ "[SiH]",
470
+ "[11c]",
471
+ "[OH]",
472
+ "[c-]",
473
+ "[18FH]",
474
+ "[123IH]",
475
+ "[13c]",
476
+ "[13cH]",
477
+ "[14cH]",
478
+ "[Cl+]",
479
+ "[Sr+2]",
480
+ "[CH-]",
481
+ "[Bi]",
482
+ "[B]",
483
+ "[Ba+2]",
484
+ "[Bi+3]",
485
+ "[SiH-]",
486
+ "[b-]",
487
+ "[H+]",
488
+ "[13C]",
489
+ "[OH+]",
490
+ "[14CH]",
491
+ "[15nH]",
492
+ "[CaH2]",
493
+ "[LiH]",
494
+ "[C]",
495
+ "[Ca]",
496
+ "[42K+]",
497
+ "[123Te]",
498
+ "[S-2]",
499
+ "[223Ra+2]",
500
+ "[S]",
501
+ "[Ra+2]",
502
+ "[22Na+]",
503
+ "[125I-]",
504
+ "[85Sr+2]",
505
+ "[PH2]",
506
+ "[SrH2]",
507
+ "[15OH2]",
508
+ "[47Ca+2]",
509
+ "[85SrH2]",
510
+ "[45Ca+2]",
511
+ "[B@@-]",
512
+ "[B@-]",
513
+ "[17F]",
514
+ "[PH2+]",
515
+ "[11C-]",
516
+ "[Mg+]",
517
+ "[NaH]",
518
+ "[P@@+]",
519
+ "[SiH3-]",
520
+ "[P-3]",
521
+ "[KH]",
522
+ "[Be+2]",
523
+ "[NH3+]",
524
+ "[Ag-4]",
525
+ "[18O]",
526
+ "[14C@@H]",
527
+ "[CH]",
528
+ "[CH2]",
529
+ "[O-2]",
530
+ "[124I-]",
531
+ "[As-]",
532
+ "[Ba]",
533
+ "[223Ra]",
534
+ "[82Rb]",
535
+ "[76BrH]",
536
+ "[AsH]",
537
+ "[131Cs]"
538
+ ]
539
+ }
checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/aux/bart_vocab_downstream.json ADDED
@@ -0,0 +1,539 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "properties": {
3
+ "regex": "(LogD_change_\\(-0\\.1, 0\\.1\\]|LogD_change_\\(0\\.1, 0\\.3\\]|LogD_change_\\(0\\.3, 0\\.5\\]|LogD_change_\\(0\\.5, 0\\.7\\]|LogD_change_\\(0\\.7, 0\\.9\\]|LogD_change_\\(0\\.9, 1\\.1\\]|LogD_change_\\(1\\.1, 1\\.3\\]|LogD_change_\\(1\\.3, 1\\.5\\]|LogD_change_\\(1\\.5, 1\\.7\\]|LogD_change_\\(1\\.7, 1\\.9\\]|LogD_change_\\(1\\.9, 2\\.1\\]|LogD_change_\\(2\\.1, 2\\.3\\]|LogD_change_\\(2\\.3, 2\\.5\\]|LogD_change_\\(2\\.5, 2\\.7\\]|LogD_change_\\(2\\.7, 2\\.9\\]|LogD_change_\\(2\\.9, 3\\.1\\]|LogD_change_\\(3\\.1, 3\\.3\\]|LogD_change_\\(3\\.3, 3\\.5\\]|LogD_change_\\(3\\.5, 3\\.7\\]|LogD_change_\\(3\\.7, 3\\.9\\]|LogD_change_\\(3\\.9, 4\\.1\\]|LogD_change_\\(4\\.1, 4\\.3\\]|LogD_change_\\(4\\.3, 4\\.5\\]|LogD_change_\\(4\\.5, 4\\.7\\]|LogD_change_\\(4\\.7, 4\\.9\\]|LogD_change_\\(4\\.9, 5\\.1\\]|LogD_change_\\(5\\.1, 5\\.3\\]|LogD_change_\\(5\\.3, 5\\.5\\]|LogD_change_\\(5\\.5, 5\\.7\\]|LogD_change_\\(5\\.7, 5\\.9\\]|LogD_change_\\(5\\.9, inf\\]|LogD_change_\\(-0\\.3, -0\\.1\\]|LogD_change_\\(-0\\.5, -0\\.3\\]|LogD_change_\\(-0\\.7, -0\\.5\\]|LogD_change_\\(-0\\.9, -0\\.7\\]|LogD_change_\\(-1\\.1, -0\\.9\\]|LogD_change_\\(-1\\.3, -1\\.1\\]|LogD_change_\\(-1\\.5, -1\\.3\\]|LogD_change_\\(-1\\.7, -1\\.5\\]|LogD_change_\\(-1\\.9, -1\\.7\\]|LogD_change_\\(-2\\.1, -1\\.9\\]|LogD_change_\\(-2\\.3, -2\\.1\\]|LogD_change_\\(-2\\.5, -2\\.3\\]|LogD_change_\\(-2\\.7, -2\\.5\\]|LogD_change_\\(-2\\.9, -2\\.7\\]|LogD_change_\\(-3\\.1, -2\\.9\\]|LogD_change_\\(-3\\.3, -3\\.1\\]|LogD_change_\\(-3\\.5, -3\\.3\\]|LogD_change_\\(-3\\.7, -3\\.5\\]|LogD_change_\\(-3\\.9, -3\\.7\\]|LogD_change_\\(-4\\.1, -3\\.9\\]|LogD_change_\\(-4\\.3, -4\\.1\\]|LogD_change_\\(-4\\.5, -4\\.3\\]|LogD_change_\\(-4\\.7, -4\\.5\\]|LogD_change_\\(-4\\.9, -4\\.7\\]|LogD_change_\\(-5\\.1, -4\\.9\\]|LogD_change_\\(-5\\.3, -5\\.1\\]|LogD_change_\\(-5\\.5, -5\\.3\\]|LogD_change_\\(-5\\.7, -5\\.5\\]|LogD_change_\\(-inf, -5\\.7\\]|Solubility_low->high|Solubility_high->low|Solubility_no_change|Clint_low->high|Clint_high->low|Clint_no_change|<UNUSED_0>|<UNUSED_1>|<UNUSED_2>|<UNUSED_3>|<UNUSED_4>|<UNUSED_5>|<UNUSED_6>|<UNUSED_7>|<UNUSED_8>|<UNUSED_9>|<UNUSED_10>|<UNUSED_11>|<UNUSED_12>|<UNUSED_13>|<UNUSED_14>|<UNUSED_15>|<UNUSED_16>|<UNUSED_17>|<UNUSED_18>|<UNUSED_19>|<UNUSED_20>|<UNUSED_21>|<UNUSED_22>|<UNUSED_23>|<UNUSED_24>|<UNUSED_25>|<UNUSED_26>|<UNUSED_27>|<UNUSED_28>|<UNUSED_29>|<UNUSED_30>|<UNUSED_31>|<UNUSED_32>|<UNUSED_33>|<UNUSED_34>|<UNUSED_35>|<UNUSED_36>|<UNUSED_37>|<UNUSED_38>|<UNUSED_39>|<UNUSED_40>|<UNUSED_41>|<UNUSED_42>|<UNUSED_43>|<UNUSED_44>|<UNUSED_45>|<UNUSED_46>|<UNUSED_47>|<UNUSED_48>|<UNUSED_49>|<UNUSED_50>|<UNUSED_51>|<UNUSED_52>|<UNUSED_53>|<UNUSED_54>|<UNUSED_55>|<UNUSED_56>|<UNUSED_57>|<UNUSED_58>|<UNUSED_59>|<UNUSED_60>|<UNUSED_61>|<UNUSED_62>|<UNUSED_63>|<UNUSED_64>|<UNUSED_65>|<UNUSED_66>|<UNUSED_67>|<UNUSED_68>|<UNUSED_69>|<UNUSED_70>|<UNUSED_71>|<UNUSED_72>|<UNUSED_73>|<UNUSED_74>|<UNUSED_75>|<UNUSED_76>|<UNUSED_77>|<UNUSED_78>|<UNUSED_79>|<UNUSED_80>|<UNUSED_81>|<UNUSED_82>|<UNUSED_83>|<UNUSED_84>|<UNUSED_85>|<UNUSED_86>|<UNUSED_87>|<UNUSED_88>|<UNUSED_89>|<UNUSED_90>|<UNUSED_91>|<UNUSED_92>|<UNUSED_93>|<UNUSED_94>|<UNUSED_95>|<UNUSED_96>|<UNUSED_97>|<UNUSED_98>|<UNUSED_99>|<UNUSED_100>|<UNUSED_101>|<UNUSED_102>|<UNUSED_103>|<UNUSED_104>|<UNUSED_105>|<UNUSED_106>|<UNUSED_107>|<UNUSED_108>|<UNUSED_109>|<UNUSED_110>|<UNUSED_111>|<UNUSED_112>|<UNUSED_113>|<UNUSED_114>|<UNUSED_115>|<UNUSED_116>|<UNUSED_117>|<UNUSED_118>|<UNUSED_119>|<UNUSED_120>|<UNUSED_121>|<UNUSED_122>|<UNUSED_123>|<UNUSED_124>|<UNUSED_125>|<UNUSED_126>|<UNUSED_127>|<UNUSED_128>|<UNUSED_129>|<UNUSED_130>|<UNUSED_131>|<UNUSED_132>|<UNUSED_133>|<UNUSED_134>|<UNUSED_135>|<UNUSED_136>|<UNUSED_137>|<UNUSED_138>|<UNUSED_139>|<UNUSED_140>|<UNUSED_141>|<UNUSED_142>|<UNUSED_143>|<UNUSED_144>|<UNUSED_145>|<UNUSED_146>|<UNUSED_147>|<UNUSED_148>|<UNUSED_149>|<UNUSED_150>|<UNUSED_151>|<UNUSED_152>|<UNUSED_153>|<UNUSED_154>|<UNUSED_155>|<UNUSED_156>|<UNUSED_157>|<UNUSED_158>|<UNUSED_159>|<UNUSED_160>|<UNUSED_161>|<UNUSED_162>|<UNUSED_163>|<UNUSED_164>|<UNUSED_165>|<UNUSED_166>|<UNUSED_167>|<UNUSED_168>|<UNUSED_169>|<UNUSED_170>|<UNUSED_171>|<UNUSED_172>|<UNUSED_173>|<UNUSED_174>|<UNUSED_175>|<UNUSED_176>|<UNUSED_177>|<UNUSED_178>|<UNUSED_179>|<UNUSED_180>|<UNUSED_181>|<UNUSED_182>|<UNUSED_183>|<UNUSED_184>|<UNUSED_185>|<UNUSED_186>|<UNUSED_187>|<UNUSED_188>|>|<RX_6>|<RX_2>|<RX_1>|<RX_3>|<RX_7>|<RX_9>|<RX_5>|<RX_10>|<RX_4>|<RX_8>|\\[[^\\]]+]|Br?|Cl?|N|O|S|P|F|I|b|c|n|o|s|p|\\(|\\)|\\.|=|#|-|\\+|\\\\|\\/|:|~|@|\\?|>|\\*|\\$|\\%[0-9]{2}|[0-9]|.)",
4
+ "special_tokens": {
5
+ "start": "^",
6
+ "end": "&",
7
+ "pad": "<PAD>",
8
+ "unknown": "?",
9
+ "mask": "<MASK>",
10
+ "sep": "<SEP>"
11
+ },
12
+ "chem_start_idx": 272
13
+ },
14
+ "vocabulary": [
15
+ "<PAD>",
16
+ "?",
17
+ "^",
18
+ "&",
19
+ "<MASK>",
20
+ "<SEP>",
21
+ "LogD_change_(-0.1, 0.1]",
22
+ "LogD_change_(0.1, 0.3]",
23
+ "LogD_change_(0.3, 0.5]",
24
+ "LogD_change_(0.5, 0.7]",
25
+ "LogD_change_(0.7, 0.9]",
26
+ "LogD_change_(0.9, 1.1]",
27
+ "LogD_change_(1.1, 1.3]",
28
+ "LogD_change_(1.3, 1.5]",
29
+ "LogD_change_(1.5, 1.7]",
30
+ "LogD_change_(1.7, 1.9]",
31
+ "LogD_change_(1.9, 2.1]",
32
+ "LogD_change_(2.1, 2.3]",
33
+ "LogD_change_(2.3, 2.5]",
34
+ "LogD_change_(2.5, 2.7]",
35
+ "LogD_change_(2.7, 2.9]",
36
+ "LogD_change_(2.9, 3.1]",
37
+ "LogD_change_(3.1, 3.3]",
38
+ "LogD_change_(3.3, 3.5]",
39
+ "LogD_change_(3.5, 3.7]",
40
+ "LogD_change_(3.7, 3.9]",
41
+ "LogD_change_(3.9, 4.1]",
42
+ "LogD_change_(4.1, 4.3]",
43
+ "LogD_change_(4.3, 4.5]",
44
+ "LogD_change_(4.5, 4.7]",
45
+ "LogD_change_(4.7, 4.9]",
46
+ "LogD_change_(4.9, 5.1]",
47
+ "LogD_change_(5.1, 5.3]",
48
+ "LogD_change_(5.3, 5.5]",
49
+ "LogD_change_(5.5, 5.7]",
50
+ "LogD_change_(5.7, 5.9]",
51
+ "LogD_change_(5.9, inf]",
52
+ "LogD_change_(-0.3, -0.1]",
53
+ "LogD_change_(-0.5, -0.3]",
54
+ "LogD_change_(-0.7, -0.5]",
55
+ "LogD_change_(-0.9, -0.7]",
56
+ "LogD_change_(-1.1, -0.9]",
57
+ "LogD_change_(-1.3, -1.1]",
58
+ "LogD_change_(-1.5, -1.3]",
59
+ "LogD_change_(-1.7, -1.5]",
60
+ "LogD_change_(-1.9, -1.7]",
61
+ "LogD_change_(-2.1, -1.9]",
62
+ "LogD_change_(-2.3, -2.1]",
63
+ "LogD_change_(-2.5, -2.3]",
64
+ "LogD_change_(-2.7, -2.5]",
65
+ "LogD_change_(-2.9, -2.7]",
66
+ "LogD_change_(-3.1, -2.9]",
67
+ "LogD_change_(-3.3, -3.1]",
68
+ "LogD_change_(-3.5, -3.3]",
69
+ "LogD_change_(-3.7, -3.5]",
70
+ "LogD_change_(-3.9, -3.7]",
71
+ "LogD_change_(-4.1, -3.9]",
72
+ "LogD_change_(-4.3, -4.1]",
73
+ "LogD_change_(-4.5, -4.3]",
74
+ "LogD_change_(-4.7, -4.5]",
75
+ "LogD_change_(-4.9, -4.7]",
76
+ "LogD_change_(-5.1, -4.9]",
77
+ "LogD_change_(-5.3, -5.1]",
78
+ "LogD_change_(-5.5, -5.3]",
79
+ "LogD_change_(-5.7, -5.5]",
80
+ "LogD_change_(-inf, -5.7]",
81
+ "Solubility_low->high",
82
+ "Solubility_high->low",
83
+ "Solubility_no_change",
84
+ "Clint_low->high",
85
+ "Clint_high->low",
86
+ "Clint_no_change",
87
+ "<UNUSED_0>",
88
+ "<UNUSED_1>",
89
+ "<UNUSED_2>",
90
+ "<UNUSED_3>",
91
+ "<UNUSED_4>",
92
+ "<UNUSED_5>",
93
+ "<UNUSED_6>",
94
+ "<UNUSED_7>",
95
+ "<UNUSED_8>",
96
+ "<UNUSED_9>",
97
+ "<UNUSED_10>",
98
+ "<UNUSED_11>",
99
+ "<UNUSED_12>",
100
+ "<UNUSED_13>",
101
+ "<UNUSED_14>",
102
+ "<UNUSED_15>",
103
+ "<UNUSED_16>",
104
+ "<UNUSED_17>",
105
+ "<UNUSED_18>",
106
+ "<UNUSED_19>",
107
+ "<UNUSED_20>",
108
+ "<UNUSED_21>",
109
+ "<UNUSED_22>",
110
+ "<UNUSED_23>",
111
+ "<UNUSED_24>",
112
+ "<UNUSED_25>",
113
+ "<UNUSED_26>",
114
+ "<UNUSED_27>",
115
+ "<UNUSED_28>",
116
+ "<UNUSED_29>",
117
+ "<UNUSED_30>",
118
+ "<UNUSED_31>",
119
+ "<UNUSED_32>",
120
+ "<UNUSED_33>",
121
+ "<UNUSED_34>",
122
+ "<UNUSED_35>",
123
+ "<UNUSED_36>",
124
+ "<UNUSED_37>",
125
+ "<UNUSED_38>",
126
+ "<UNUSED_39>",
127
+ "<UNUSED_40>",
128
+ "<UNUSED_41>",
129
+ "<UNUSED_42>",
130
+ "<UNUSED_43>",
131
+ "<UNUSED_44>",
132
+ "<UNUSED_45>",
133
+ "<UNUSED_46>",
134
+ "<UNUSED_47>",
135
+ "<UNUSED_48>",
136
+ "<UNUSED_49>",
137
+ "<UNUSED_50>",
138
+ "<UNUSED_51>",
139
+ "<UNUSED_52>",
140
+ "<UNUSED_53>",
141
+ "<UNUSED_54>",
142
+ "<UNUSED_55>",
143
+ "<UNUSED_56>",
144
+ "<UNUSED_57>",
145
+ "<UNUSED_58>",
146
+ "<UNUSED_59>",
147
+ "<UNUSED_60>",
148
+ "<UNUSED_61>",
149
+ "<UNUSED_62>",
150
+ "<UNUSED_63>",
151
+ "<UNUSED_64>",
152
+ "<UNUSED_65>",
153
+ "<UNUSED_66>",
154
+ "<UNUSED_67>",
155
+ "<UNUSED_68>",
156
+ "<UNUSED_69>",
157
+ "<UNUSED_70>",
158
+ "<UNUSED_71>",
159
+ "<UNUSED_72>",
160
+ "<UNUSED_73>",
161
+ "<UNUSED_74>",
162
+ "<UNUSED_75>",
163
+ "<UNUSED_76>",
164
+ "<UNUSED_77>",
165
+ "<UNUSED_78>",
166
+ "<UNUSED_79>",
167
+ "<UNUSED_80>",
168
+ "<UNUSED_81>",
169
+ "<UNUSED_82>",
170
+ "<UNUSED_83>",
171
+ "<UNUSED_84>",
172
+ "<UNUSED_85>",
173
+ "<UNUSED_86>",
174
+ "<UNUSED_87>",
175
+ "<UNUSED_88>",
176
+ "<UNUSED_89>",
177
+ "<UNUSED_90>",
178
+ "<UNUSED_91>",
179
+ "<UNUSED_92>",
180
+ "<UNUSED_93>",
181
+ "<UNUSED_94>",
182
+ "<UNUSED_95>",
183
+ "<UNUSED_96>",
184
+ "<UNUSED_97>",
185
+ "<UNUSED_98>",
186
+ "<UNUSED_99>",
187
+ "<UNUSED_100>",
188
+ "<UNUSED_101>",
189
+ "<UNUSED_102>",
190
+ "<UNUSED_103>",
191
+ "<UNUSED_104>",
192
+ "<UNUSED_105>",
193
+ "<UNUSED_106>",
194
+ "<UNUSED_107>",
195
+ "<UNUSED_108>",
196
+ "<UNUSED_109>",
197
+ "<UNUSED_110>",
198
+ "<UNUSED_111>",
199
+ "<UNUSED_112>",
200
+ "<UNUSED_113>",
201
+ "<UNUSED_114>",
202
+ "<UNUSED_115>",
203
+ "<UNUSED_116>",
204
+ "<UNUSED_117>",
205
+ "<UNUSED_118>",
206
+ "<UNUSED_119>",
207
+ "<UNUSED_120>",
208
+ "<UNUSED_121>",
209
+ "<UNUSED_122>",
210
+ "<UNUSED_123>",
211
+ "<UNUSED_124>",
212
+ "<UNUSED_125>",
213
+ "<UNUSED_126>",
214
+ "<UNUSED_127>",
215
+ "<UNUSED_128>",
216
+ "<UNUSED_129>",
217
+ "<UNUSED_130>",
218
+ "<UNUSED_131>",
219
+ "<UNUSED_132>",
220
+ "<UNUSED_133>",
221
+ "<UNUSED_134>",
222
+ "<UNUSED_135>",
223
+ "<UNUSED_136>",
224
+ "<UNUSED_137>",
225
+ "<UNUSED_138>",
226
+ "<UNUSED_139>",
227
+ "<UNUSED_140>",
228
+ "<UNUSED_141>",
229
+ "<UNUSED_142>",
230
+ "<UNUSED_143>",
231
+ "<UNUSED_144>",
232
+ "<UNUSED_145>",
233
+ "<UNUSED_146>",
234
+ "<UNUSED_147>",
235
+ "<UNUSED_148>",
236
+ "<UNUSED_149>",
237
+ "<UNUSED_150>",
238
+ "<UNUSED_151>",
239
+ "<UNUSED_152>",
240
+ "<UNUSED_153>",
241
+ "<UNUSED_154>",
242
+ "<UNUSED_155>",
243
+ "<UNUSED_156>",
244
+ "<UNUSED_157>",
245
+ "<UNUSED_158>",
246
+ "<UNUSED_159>",
247
+ "<UNUSED_160>",
248
+ "<UNUSED_161>",
249
+ "<UNUSED_162>",
250
+ "<UNUSED_163>",
251
+ "<UNUSED_164>",
252
+ "<UNUSED_165>",
253
+ "<UNUSED_166>",
254
+ "<UNUSED_167>",
255
+ "<UNUSED_168>",
256
+ "<UNUSED_169>",
257
+ "<UNUSED_170>",
258
+ "<UNUSED_171>",
259
+ "<UNUSED_172>",
260
+ "<UNUSED_173>",
261
+ "<UNUSED_174>",
262
+ "<UNUSED_175>",
263
+ "<UNUSED_176>",
264
+ "<UNUSED_177>",
265
+ "<UNUSED_178>",
266
+ "<UNUSED_179>",
267
+ "<UNUSED_180>",
268
+ "<UNUSED_181>",
269
+ "<UNUSED_182>",
270
+ "<UNUSED_183>",
271
+ "<UNUSED_184>",
272
+ "<UNUSED_185>",
273
+ "<UNUSED_186>",
274
+ "<UNUSED_187>",
275
+ "<UNUSED_188>",
276
+ ">",
277
+ "<RX_6>",
278
+ "<RX_2>",
279
+ "<RX_1>",
280
+ "<RX_3>",
281
+ "<RX_7>",
282
+ "<RX_9>",
283
+ "<RX_5>",
284
+ "<RX_10>",
285
+ "<RX_4>",
286
+ "<RX_8>",
287
+ "C",
288
+ "c",
289
+ "1",
290
+ "(",
291
+ "-",
292
+ "2",
293
+ "s",
294
+ "N",
295
+ "=",
296
+ ")",
297
+ "n",
298
+ "Br",
299
+ "3",
300
+ "O",
301
+ "[nH]",
302
+ "[C@H]",
303
+ "/",
304
+ "\\",
305
+ "[C@@H]",
306
+ "[C@]",
307
+ "o",
308
+ "4",
309
+ "[C@@]",
310
+ "5",
311
+ "6",
312
+ "7",
313
+ "[n+]",
314
+ ".",
315
+ "[Br-]",
316
+ "S",
317
+ "I",
318
+ "F",
319
+ "Cl",
320
+ "[O-]",
321
+ "#",
322
+ "P",
323
+ "[Na+]",
324
+ "[N+]",
325
+ "[Cl-]",
326
+ "8",
327
+ "[Si]",
328
+ "[N-]",
329
+ "[18F]",
330
+ "9",
331
+ "[I-]",
332
+ "B",
333
+ "[S+]",
334
+ "[2H]",
335
+ "[P+]",
336
+ "[125I]",
337
+ "%10",
338
+ "%11",
339
+ "[n-]",
340
+ "[B-]",
341
+ "[O]",
342
+ "[o+]",
343
+ "[N@@+]",
344
+ "[N@+]",
345
+ "[PH]",
346
+ "[se]",
347
+ "[Se]",
348
+ "[s+]",
349
+ "[Li+]",
350
+ "[P@]",
351
+ "[P@@]",
352
+ "[3H]",
353
+ "[K+]",
354
+ "[OH-]",
355
+ "[te+]",
356
+ "[se+]",
357
+ "[Te]",
358
+ "[S@+]",
359
+ "[BH2-]",
360
+ "[11CH3]",
361
+ "[11C]",
362
+ "[Cl+3]",
363
+ "[N@@]",
364
+ "[S-]",
365
+ "[C+]",
366
+ "[P@+]",
367
+ "[C-]",
368
+ "[Zn+2]",
369
+ "[Ca+2]",
370
+ "[Mg+2]",
371
+ "[SeH]",
372
+ "[BH-]",
373
+ "[SH2]",
374
+ "[TeH2]",
375
+ "[SiH4]",
376
+ "[N@]",
377
+ "[14CH2]",
378
+ "[Ag]",
379
+ "[S@@+]",
380
+ "[I+]",
381
+ "[MgH2]",
382
+ "[125IH]",
383
+ "[Se+]",
384
+ "[As]",
385
+ "[SiH2]",
386
+ "[Ra]",
387
+ "[IH2]",
388
+ "[P-]",
389
+ "[Na]",
390
+ "[NH-]",
391
+ "[Cs]",
392
+ "[Zn]",
393
+ "[Li]",
394
+ "[As+]",
395
+ "[te]",
396
+ "[131I]",
397
+ "[Ag+]",
398
+ "p",
399
+ "[Al-3]",
400
+ "[Rb+]",
401
+ "[13CH]",
402
+ "[11CH]",
403
+ "[11CH2]",
404
+ "[14c]",
405
+ "[14C@@]",
406
+ "[127I]",
407
+ "[Mg]",
408
+ "[14C]",
409
+ "[123I]",
410
+ "[124I]",
411
+ "[F-]",
412
+ "[14CH3]",
413
+ "[135I]",
414
+ "%12",
415
+ "[NH+]",
416
+ "[76Br]",
417
+ "[32PH]",
418
+ "[35S]",
419
+ "b",
420
+ "[73Se]",
421
+ "[11C@@H]",
422
+ "[Se-]",
423
+ "[c+]",
424
+ "[14C@H]",
425
+ "[18OH]",
426
+ "[SH]",
427
+ "[S@@]",
428
+ "[Cs+]",
429
+ "[He]",
430
+ "[O+]",
431
+ "[nH+]",
432
+ "[NH2+]",
433
+ "[32P]",
434
+ "[Zn+]",
435
+ "[BH3-]",
436
+ "[I+3]",
437
+ "[Si-]",
438
+ "[SH+]",
439
+ "[19F]",
440
+ "[Br+2]",
441
+ "[I+2]",
442
+ "[Al+3]",
443
+ "[123I-]",
444
+ "[131I-]",
445
+ "[127Xe]",
446
+ "[133Xe]",
447
+ "[89Sr+2]",
448
+ "[N]",
449
+ "[82Rb+]",
450
+ "[75Se]",
451
+ "[Rb]",
452
+ "[81Kr]",
453
+ "[18F-]",
454
+ "[13NH3]",
455
+ "[K]",
456
+ "[Cl+2]",
457
+ "[Zn-2]",
458
+ "[SeH2]",
459
+ "[AsH3]",
460
+ "[Kr]",
461
+ "[Xe]",
462
+ "[S@]",
463
+ "[cH-]",
464
+ "[NH4+]",
465
+ "[Al]",
466
+ "[Si@]",
467
+ "[15n]",
468
+ "[SH-]",
469
+ "[SiH]",
470
+ "[11c]",
471
+ "[OH]",
472
+ "[c-]",
473
+ "[18FH]",
474
+ "[123IH]",
475
+ "[13c]",
476
+ "[13cH]",
477
+ "[14cH]",
478
+ "[Cl+]",
479
+ "[Sr+2]",
480
+ "[CH-]",
481
+ "[Bi]",
482
+ "[B]",
483
+ "[Ba+2]",
484
+ "[Bi+3]",
485
+ "[SiH-]",
486
+ "[b-]",
487
+ "[H+]",
488
+ "[13C]",
489
+ "[OH+]",
490
+ "[14CH]",
491
+ "[15nH]",
492
+ "[CaH2]",
493
+ "[LiH]",
494
+ "[C]",
495
+ "[Ca]",
496
+ "[42K+]",
497
+ "[123Te]",
498
+ "[S-2]",
499
+ "[223Ra+2]",
500
+ "[S]",
501
+ "[Ra+2]",
502
+ "[22Na+]",
503
+ "[125I-]",
504
+ "[85Sr+2]",
505
+ "[PH2]",
506
+ "[SrH2]",
507
+ "[15OH2]",
508
+ "[47Ca+2]",
509
+ "[85SrH2]",
510
+ "[45Ca+2]",
511
+ "[B@@-]",
512
+ "[B@-]",
513
+ "[17F]",
514
+ "[PH2+]",
515
+ "[11C-]",
516
+ "[Mg+]",
517
+ "[NaH]",
518
+ "[P@@+]",
519
+ "[SiH3-]",
520
+ "[P-3]",
521
+ "[KH]",
522
+ "[Be+2]",
523
+ "[NH3+]",
524
+ "[Ag-4]",
525
+ "[18O]",
526
+ "[14C@@H]",
527
+ "[CH]",
528
+ "[CH2]",
529
+ "[O-2]",
530
+ "[124I-]",
531
+ "[As-]",
532
+ "[Ba]",
533
+ "[223Ra]",
534
+ "[82Rb]",
535
+ "[76BrH]",
536
+ "[AsH]",
537
+ "[131Cs]"
538
+ ]
539
+ }
checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/aux/ds_config.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "zero_optimization": {
3
+ "stage": 2,
4
+ "cpu_offload": false,
5
+ "contiguous_gradients": true,
6
+ "overlap_comm": true,
7
+ "allgather_bucket_size": 5e8,
8
+ "reduce_bucket_size": 5e8
9
+ },
10
+ "fp16": {
11
+ "enabled": true,
12
+ "loss_scale": 0.0,
13
+ "initial_scale_power": 32,
14
+ "loss_scale_window": 1000,
15
+ "hysteresis": 2,
16
+ "min_loss_scale": 1000
17
+ }
18
+ }
checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/aux/hparams.yaml ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ acc_batches: 8
2
+ activation: gelu
3
+ aug_prob: 0.5
4
+ augment: all
5
+ batch_size: 64
6
+ clip_grad: 1.0
7
+ d_feedforward: 2048
8
+ d_model: 512
9
+ dropout: 0.1
10
+ epochs: 150
11
+ limit_val_batches: 1.0
12
+ lr: 0.001
13
+ mask_prob: 0.1
14
+ max_seq_len: 512
15
+ num_buckets: 24
16
+ num_heads: 8
17
+ num_layers: 6
18
+ num_steps: 119851
19
+ pad_token_idx: 0
20
+ schedule: cycle
21
+ task: mask_aug
22
+ train_tokens: None
23
+ vocab_size: 523
24
+ warm_up_steps: 8000
25
+ weight_decay: 0.0
checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/aux/hyper_parameters.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "hyper_parameters": {
3
+ "acc_batches": 8,
4
+ "activation": "gelu",
5
+ "aug_prob": 0.5,
6
+ "augment": "all",
7
+ "batch_size": 64,
8
+ "clip_grad": 1.0,
9
+ "d_feedforward": 2048,
10
+ "d_model": 512,
11
+ "dropout": 0.1,
12
+ "epochs": 150,
13
+ "limit_val_batches": 1.0,
14
+ "lr": 0.001,
15
+ "mask_prob": 0.1,
16
+ "max_seq_len": 512,
17
+ "num_buckets": 24,
18
+ "num_heads": 8,
19
+ "num_layers": 6,
20
+ "num_steps": 119851,
21
+ "pad_token_idx": 0,
22
+ "schedule": "cycle",
23
+ "task": "mask_aug",
24
+ "train_tokens": "None",
25
+ "vocab_size": 523,
26
+ "warm_up_steps": 8000,
27
+ "weight_decay": 0.0
28
+ }
29
+ }
checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/benchmarks/default_test.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/benchmarks/metadata.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "benchmark_semantics": "forward_product_prediction_with_separated_reagents",
3
+ "benchmark_variant": "forward_product_prediction",
4
+ "built_at": "1970-01-01T00:00:00+00:00",
5
+ "checkpoint_id": "uspto_sep_forward_checkpoint",
6
+ "dataset_id": "uspto_sep_forward_checkpoint_test",
7
+ "dataset_ref": "uspto_sep_chemformer_project_release",
8
+ "native_input_format": "pandas_pickled_dataframe",
9
+ "native_sources": [
10
+ {
11
+ "notes": [
12
+ "Zip member: data/seq-to-seq_datasets/uspto_sep.pickle",
13
+ "Archive: datasets/chemformer/Chemformer-selected.zip"
14
+ ],
15
+ "origin": "extracted_from_local_manual_download_zip",
16
+ "relative_path": "benchmarks/native/uspto_sep.pickle",
17
+ "source_url": null
18
+ }
19
+ ],
20
+ "notes": [
21
+ "Native Chemformer forward benchmark source extracted from the public Box release uspto_sep pickle.",
22
+ "benchmarks/default_test.jsonl preserves separated reactants/reagents fields plus expected product SMILES from the test partition."
23
+ ],
24
+ "num_examples": 40000,
25
+ "record_schema": {
26
+ "inputs": [
27
+ "reactants",
28
+ "reagents"
29
+ ],
30
+ "targets": [
31
+ "expected_products"
32
+ ]
33
+ },
34
+ "split": "test",
35
+ "task_direction": "forward"
36
+ }
checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/manifest.json ADDED
@@ -0,0 +1,155 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auxiliary_files": [
3
+ "aux/bart_vocab.json",
4
+ "aux/bart_vocab_downstream.json",
5
+ "aux/ds_config.json",
6
+ "aux/hparams.yaml",
7
+ "aux/hyper_parameters.json",
8
+ "benchmarks/default_test.jsonl",
9
+ "benchmarks/metadata.json",
10
+ "benchmarks/native/uspto_sep.pickle"
11
+ ],
12
+ "benchmark_datasets": [
13
+ {
14
+ "benchmark_semantics": "forward_product_prediction_with_separated_reagents",
15
+ "benchmark_variant": "forward_product_prediction",
16
+ "dataset_id": "uspto_sep_forward_checkpoint_test",
17
+ "dataset_ref": "uspto_sep_chemformer_project_release",
18
+ "default": true,
19
+ "format": "jsonl",
20
+ "metadata_path": "benchmarks/metadata.json",
21
+ "native_input_format": "pandas_pickled_dataframe",
22
+ "native_source_paths": [
23
+ "benchmarks/native/uspto_sep.pickle"
24
+ ],
25
+ "notes": [
26
+ "Native Chemformer forward benchmark source extracted from the public Box release uspto_sep pickle.",
27
+ "benchmarks/default_test.jsonl preserves separated reactants/reagents fields plus expected product SMILES from the test partition."
28
+ ],
29
+ "num_examples": 40000,
30
+ "path": "benchmarks/default_test.jsonl",
31
+ "record_schema": {
32
+ "inputs": [
33
+ "reactants",
34
+ "reagents"
35
+ ],
36
+ "targets": [
37
+ "expected_products"
38
+ ]
39
+ },
40
+ "split": "test",
41
+ "task_direction": "forward"
42
+ }
43
+ ],
44
+ "checkpoint_id": "uspto_sep_forward_checkpoint",
45
+ "contained_artifact_bindings": [
46
+ {
47
+ "path": "hparams.yaml",
48
+ "role": "checkpoint_hyperparameters"
49
+ },
50
+ {
51
+ "path": "last.ckpt",
52
+ "role": "model_checkpoint"
53
+ }
54
+ ],
55
+ "conversion": {
56
+ "converted_at": "1970-01-01T00:00:00+00:00",
57
+ "source_format": "pytorch_lightning_ckpt",
58
+ "status": "converted",
59
+ "tool_version": "initial"
60
+ },
61
+ "display_name": "uspto_sep forward checkpoint",
62
+ "embedding_supported": false,
63
+ "fine_tuning": {
64
+ "adapter_overlay_dir": "adapters",
65
+ "base_weights_path": "model.safetensors",
66
+ "notes": [
67
+ "This package preserves immutable base weights in model.safetensors and reserves adapters/ for future fine-tuning or LoRA overlays."
68
+ ]
69
+ },
70
+ "format_version": 1,
71
+ "generated_at": "1970-01-01T00:00:00+00:00",
72
+ "input_schema": {
73
+ "reactants_field": "reactants_mol",
74
+ "reagents_field": "reagents_mol",
75
+ "serialization": "dot_joined_smiles",
76
+ "source_dataset_member": "data/seq-to-seq_datasets/uspto_sep.pickle",
77
+ "target_field": "products_mol"
78
+ },
79
+ "model_id": "chemformer_forward",
80
+ "normalized_checkpoint_id": "uspto_sep_forward_checkpoint",
81
+ "package_layout": {
82
+ "adapters_dir": "adapters",
83
+ "aux_dir": "aux",
84
+ "benchmarks_dir": "benchmarks",
85
+ "weights_path": "model.safetensors"
86
+ },
87
+ "source_registry_refs": [
88
+ {
89
+ "checkpoint_id": "uspto_sep_forward_checkpoint",
90
+ "model_id": "chemformer_forward"
91
+ }
92
+ ],
93
+ "split_provenance": [
94
+ {
95
+ "canonical_split_family": "chemformer_zinc15_100m_train_valid_test",
96
+ "dataset_ref": "zinc15_100m_chemformer_pretraining",
97
+ "split_names": [
98
+ "train",
99
+ "valid",
100
+ "test"
101
+ ],
102
+ "stage": "pretraining_initialization"
103
+ },
104
+ {
105
+ "canonical_split_family": "chemformer_project_usptosep_pickle_train_val_test",
106
+ "dataset_ref": "uspto_sep_chemformer_project_release",
107
+ "split_names": [
108
+ "train",
109
+ "val",
110
+ "test"
111
+ ],
112
+ "stage": "forward_fine_tuning_and_evaluation"
113
+ },
114
+ {
115
+ "path": "hparams.yaml"
116
+ },
117
+ {
118
+ "path": "last.ckpt"
119
+ }
120
+ ],
121
+ "task_direction": "forward",
122
+ "training_dataset_refs": [
123
+ "uspto_sep_chemformer_project_release"
124
+ ],
125
+ "training_stage_bindings": [
126
+ {
127
+ "canonical_split_family": "chemformer_zinc15_100m_train_valid_test",
128
+ "dataset_ref": "zinc15_100m_chemformer_pretraining",
129
+ "split_names": [
130
+ "train",
131
+ "valid",
132
+ "test"
133
+ ],
134
+ "stage": "pretraining_initialization"
135
+ },
136
+ {
137
+ "canonical_split_family": "chemformer_project_usptosep_pickle_train_val_test",
138
+ "dataset_ref": "uspto_sep_chemformer_project_release",
139
+ "split_names": [
140
+ "train",
141
+ "val",
142
+ "test"
143
+ ],
144
+ "stage": "forward_fine_tuning_and_evaluation"
145
+ }
146
+ ],
147
+ "weights": {
148
+ "dtype_summary": "torch_saved_tensors",
149
+ "framework": "safetensors",
150
+ "path": "model.safetensors",
151
+ "tensor_namespace": "lightning_state_dict"
152
+ },
153
+ "wrapper_config": "wrapper_config.json",
154
+ "wrapper_family": "chemformer_bart"
155
+ }
checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/provenance.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "conversion_notes": [
3
+ "Extracted obj['state_dict'] from the Lightning checkpoint and wrote it to safetensors."
4
+ ],
5
+ "copied_sidecars": [
6
+ "aux/bart_vocab.json",
7
+ "aux/bart_vocab_downstream.json",
8
+ "aux/ds_config.json",
9
+ "aux/hparams.yaml",
10
+ "aux/hyper_parameters.json"
11
+ ],
12
+ "inner_members": [
13
+ {
14
+ "archive_path": "datasets/chemformer/Chemformer-selected.zip",
15
+ "member": "models/fined-tuned/uspto_sep/last.ckpt",
16
+ "role": "weights_source"
17
+ },
18
+ {
19
+ "archive_path": "datasets/chemformer/Chemformer-selected.zip",
20
+ "member": "models/fined-tuned/uspto_sep/hparams.yaml",
21
+ "role": "hparams_source"
22
+ }
23
+ ],
24
+ "normalized_checkpoint_id": "uspto_sep_forward_checkpoint",
25
+ "source_artifacts": [
26
+ {
27
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/chemformer/Chemformer-selected.zip",
28
+ "kind": "local_repo_file",
29
+ "path": "datasets/chemformer/Chemformer-selected.zip",
30
+ "role": "source_artifact",
31
+ "sha256": "84bce87299a05784b9141d7979f9f3a1b8762ede0f602db76e85e8975c1b464e",
32
+ "size": 14770784818
33
+ },
34
+ {
35
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/source_snapshots/chemformer/repo/bart_vocab.json",
36
+ "kind": "local_repo_file",
37
+ "path": "datasets/source_snapshots/chemformer/repo/bart_vocab.json",
38
+ "role": "source_artifact",
39
+ "sha256": "8b17254dfde43cbac758e697d3939c0100ba2d5da341031e92fd6b8de6e87cfb",
40
+ "size": 15954
41
+ },
42
+ {
43
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/source_snapshots/chemformer/repo/bart_vocab_downstream.json",
44
+ "kind": "local_repo_file",
45
+ "path": "datasets/source_snapshots/chemformer/repo/bart_vocab_downstream.json",
46
+ "role": "source_artifact",
47
+ "sha256": "8c15ea5b60855344ddb24586b14c33811e89c93adf4189865045fcf3ad89459c",
48
+ "size": 16050
49
+ },
50
+ {
51
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/source_snapshots/chemformer/repo/ds_config.json",
52
+ "kind": "local_repo_file",
53
+ "path": "datasets/source_snapshots/chemformer/repo/ds_config.json",
54
+ "role": "source_artifact",
55
+ "sha256": "ed5d259989f3d3e3d423ecc3c1f7f56dda31ea1debed0a29977be2270fc1b588",
56
+ "size": 426
57
+ }
58
+ ]
59
+ }
checkpoints/chemformer_forward/uspto_sep_forward_checkpoint/wrapper_config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architecture_class": "bart",
3
+ "device_supported": [
4
+ "cpu",
5
+ "cuda"
6
+ ],
7
+ "fine_tuning": {
8
+ "adapter_overlay_dir": "adapters",
9
+ "base_weights_path": "model.safetensors",
10
+ "notes": [
11
+ "Base weights stay immutable in model.safetensors; future fine-tuning or LoRA adapters should be added as overlays under adapters/."
12
+ ]
13
+ },
14
+ "hparams_file": "aux/hparams.yaml",
15
+ "hyper_parameters": "aux/hyper_parameters.json",
16
+ "loader": {
17
+ "state_dict_format": "safetensors",
18
+ "strict": true
19
+ },
20
+ "package_layout": {
21
+ "adapters_dir": "adapters",
22
+ "aux_dir": "aux",
23
+ "weights_path": "model.safetensors"
24
+ },
25
+ "task_direction": "forward",
26
+ "vocab_files": [
27
+ "aux/bart_vocab.json",
28
+ "aux/bart_vocab_downstream.json"
29
+ ],
30
+ "wrapper_family": "chemformer_bart"
31
+ }
checkpoints/forward_normalize_report.json ADDED
@@ -0,0 +1,69 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "archive_root": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/DOES_NOT_EXIST",
3
+ "datasets_root": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets",
4
+ "generated_at": "1970-01-01T00:00:00+00:00",
5
+ "normalized_packages": {
6
+ "chemformer_forward/uspto_mixed_forward_checkpoint": {
7
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint",
8
+ "source_registry_refs": [
9
+ {
10
+ "checkpoint_id": "uspto_mixed_forward_checkpoint",
11
+ "model_id": "chemformer_forward"
12
+ }
13
+ ],
14
+ "status": "ok"
15
+ },
16
+ "chemformer_forward/uspto_sep_forward_checkpoint": {
17
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/chemformer_forward/uspto_sep_forward_checkpoint",
18
+ "source_registry_refs": [
19
+ {
20
+ "checkpoint_id": "uspto_sep_forward_checkpoint",
21
+ "model_id": "chemformer_forward"
22
+ }
23
+ ],
24
+ "status": "ok"
25
+ },
26
+ "molecular_transformer_forward/mit_mixed_augm_model_average_20_pt": {
27
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/molecular_transformer_forward/mit_mixed_augm_model_average_20_pt",
28
+ "source_registry_refs": [
29
+ {
30
+ "checkpoint_id": "mit_mixed_augm_model_average_20_pt",
31
+ "model_id": "molecular_transformer_forward"
32
+ }
33
+ ],
34
+ "status": "ok"
35
+ },
36
+ "molecular_transformer_forward/stereo_mixed_augm_model_average_20_pt": {
37
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/molecular_transformer_forward/stereo_mixed_augm_model_average_20_pt",
38
+ "source_registry_refs": [
39
+ {
40
+ "checkpoint_id": "stereo_mixed_augm_model_average_20_pt",
41
+ "model_id": "molecular_transformer_forward"
42
+ }
43
+ ],
44
+ "status": "ok"
45
+ },
46
+ "reactiont5v2_forward/sagawa_reactiont5v2_forward": {
47
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward",
48
+ "source_registry_refs": [
49
+ {
50
+ "checkpoint_id": "sagawa_reactiont5v2_forward",
51
+ "model_id": "reactiont5v2_forward"
52
+ }
53
+ ],
54
+ "status": "ok"
55
+ },
56
+ "reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit": {
57
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit",
58
+ "source_registry_refs": [
59
+ {
60
+ "checkpoint_id": "sagawa_reactiont5v2_forward_uspto_mit",
61
+ "model_id": "reactiont5v2_forward"
62
+ }
63
+ ],
64
+ "status": "ok"
65
+ }
66
+ },
67
+ "output_root": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints",
68
+ "source_root": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/source_snapshots"
69
+ }
checkpoints/forward_normalized_registry.json ADDED
@@ -0,0 +1,440 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "checkpoints": [
3
+ {
4
+ "auxiliary_files": [
5
+ "aux/bart_vocab.json",
6
+ "aux/bart_vocab_downstream.json",
7
+ "aux/ds_config.json",
8
+ "aux/hparams.yaml",
9
+ "aux/hyper_parameters.json",
10
+ "benchmarks/default_test.jsonl",
11
+ "benchmarks/metadata.json",
12
+ "benchmarks/native/uspto_mixed.pickle"
13
+ ],
14
+ "benchmark_datasets": [
15
+ {
16
+ "benchmark_semantics": "forward_product_prediction_with_mixed_inputs",
17
+ "benchmark_variant": "forward_product_prediction",
18
+ "dataset_id": "uspto_mixed_forward_checkpoint_test",
19
+ "dataset_ref": "uspto_mixed_chemformer_project_release",
20
+ "default": true,
21
+ "format": "jsonl",
22
+ "metadata_path": "benchmarks/metadata.json",
23
+ "native_input_format": "pandas_pickled_dataframe",
24
+ "native_source_paths": [
25
+ "benchmarks/native/uspto_mixed.pickle"
26
+ ],
27
+ "notes": [
28
+ "Native Chemformer forward benchmark source extracted from the public Box release uspto_mixed pickle.",
29
+ "benchmarks/default_test.jsonl preserves reactant-side input SMILES plus expected product SMILES from the test partition."
30
+ ],
31
+ "num_examples": 40000,
32
+ "path": "benchmarks/default_test.jsonl",
33
+ "record_schema": {
34
+ "inputs": [
35
+ "reactants"
36
+ ],
37
+ "targets": [
38
+ "expected_products"
39
+ ]
40
+ },
41
+ "split": "test",
42
+ "task_direction": "forward"
43
+ }
44
+ ],
45
+ "checkpoint_id": "uspto_mixed_forward_checkpoint",
46
+ "embedding_supported": false,
47
+ "input_schema": {
48
+ "reactants_field": "reactants_mol",
49
+ "reagents_field": null,
50
+ "serialization": "dot_joined_smiles",
51
+ "source_dataset_member": "data/seq-to-seq_datasets/uspto_mixed.pickle",
52
+ "target_field": "products_mol"
53
+ },
54
+ "model_id": "chemformer_forward",
55
+ "normalized_checkpoint_id": "uspto_mixed_forward_checkpoint",
56
+ "package_dir": "checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint",
57
+ "source_registry_refs": [
58
+ {
59
+ "checkpoint_id": "uspto_mixed_forward_checkpoint",
60
+ "model_id": "chemformer_forward"
61
+ }
62
+ ],
63
+ "task_direction": "forward",
64
+ "training_dataset_refs": [
65
+ "uspto_mixed_chemformer_project_release"
66
+ ],
67
+ "weights_path": "model.safetensors",
68
+ "wrapper_family": "chemformer_bart"
69
+ },
70
+ {
71
+ "auxiliary_files": [
72
+ "aux/bart_vocab.json",
73
+ "aux/bart_vocab_downstream.json",
74
+ "aux/ds_config.json",
75
+ "aux/hparams.yaml",
76
+ "aux/hyper_parameters.json",
77
+ "benchmarks/default_test.jsonl",
78
+ "benchmarks/metadata.json",
79
+ "benchmarks/native/uspto_sep.pickle"
80
+ ],
81
+ "benchmark_datasets": [
82
+ {
83
+ "benchmark_semantics": "forward_product_prediction_with_separated_reagents",
84
+ "benchmark_variant": "forward_product_prediction",
85
+ "dataset_id": "uspto_sep_forward_checkpoint_test",
86
+ "dataset_ref": "uspto_sep_chemformer_project_release",
87
+ "default": true,
88
+ "format": "jsonl",
89
+ "metadata_path": "benchmarks/metadata.json",
90
+ "native_input_format": "pandas_pickled_dataframe",
91
+ "native_source_paths": [
92
+ "benchmarks/native/uspto_sep.pickle"
93
+ ],
94
+ "notes": [
95
+ "Native Chemformer forward benchmark source extracted from the public Box release uspto_sep pickle.",
96
+ "benchmarks/default_test.jsonl preserves separated reactants/reagents fields plus expected product SMILES from the test partition."
97
+ ],
98
+ "num_examples": 40000,
99
+ "path": "benchmarks/default_test.jsonl",
100
+ "record_schema": {
101
+ "inputs": [
102
+ "reactants",
103
+ "reagents"
104
+ ],
105
+ "targets": [
106
+ "expected_products"
107
+ ]
108
+ },
109
+ "split": "test",
110
+ "task_direction": "forward"
111
+ }
112
+ ],
113
+ "checkpoint_id": "uspto_sep_forward_checkpoint",
114
+ "embedding_supported": false,
115
+ "input_schema": {
116
+ "reactants_field": "reactants_mol",
117
+ "reagents_field": "reagents_mol",
118
+ "serialization": "dot_joined_smiles",
119
+ "source_dataset_member": "data/seq-to-seq_datasets/uspto_sep.pickle",
120
+ "target_field": "products_mol"
121
+ },
122
+ "model_id": "chemformer_forward",
123
+ "normalized_checkpoint_id": "uspto_sep_forward_checkpoint",
124
+ "package_dir": "checkpoints/chemformer_forward/uspto_sep_forward_checkpoint",
125
+ "source_registry_refs": [
126
+ {
127
+ "checkpoint_id": "uspto_sep_forward_checkpoint",
128
+ "model_id": "chemformer_forward"
129
+ }
130
+ ],
131
+ "task_direction": "forward",
132
+ "training_dataset_refs": [
133
+ "uspto_sep_chemformer_project_release"
134
+ ],
135
+ "weights_path": "model.safetensors",
136
+ "wrapper_family": "chemformer_bart"
137
+ },
138
+ {
139
+ "auxiliary_files": [
140
+ "benchmarks/default_test.jsonl",
141
+ "benchmarks/metadata.json",
142
+ "benchmarks/native/src-test.txt",
143
+ "benchmarks/native/tgt-test.txt"
144
+ ],
145
+ "benchmark_datasets": [
146
+ {
147
+ "benchmark_semantics": "forward_product_prediction_with_mixed_inputs",
148
+ "benchmark_variant": "forward_product_prediction",
149
+ "dataset_id": "mit_mixed_augm_model_average_20_pt_test",
150
+ "dataset_ref": "uspto_mit_molecular_transformer_project_release",
151
+ "default": true,
152
+ "format": "jsonl",
153
+ "metadata_path": "benchmarks/metadata.json",
154
+ "native_input_format": "parallel_tokenized_smiles",
155
+ "native_source_paths": [
156
+ "benchmarks/native/src-test.txt",
157
+ "benchmarks/native/tgt-test.txt"
158
+ ],
159
+ "notes": [
160
+ "Native Molecular Transformer benchmark sources are the official tokenized test split files from the public Box release.",
161
+ "benchmarks/default_test.jsonl stores detokenized mixed-input reactant strings mapped to expected product SMILES for forward evaluation."
162
+ ],
163
+ "num_examples": 40000,
164
+ "path": "benchmarks/default_test.jsonl",
165
+ "record_schema": {
166
+ "inputs": [
167
+ "reactants"
168
+ ],
169
+ "targets": [
170
+ "expected_products"
171
+ ]
172
+ },
173
+ "serialized_input_format": "space_tokenized_smiles",
174
+ "split": "test",
175
+ "task_direction": "forward"
176
+ }
177
+ ],
178
+ "checkpoint_id": "mit_mixed_augm_model_average_20_pt",
179
+ "embedding_supported": false,
180
+ "input_schema": {
181
+ "benchmark_variant": "forward_product_prediction",
182
+ "reactants_field": "reactants_mixed",
183
+ "serialized_format": "space_tokenized_smiles",
184
+ "target_field": "PRODUCT",
185
+ "tokenizer": "smiles_regex_space_tokenizer"
186
+ },
187
+ "model_id": "molecular_transformer_forward",
188
+ "normalized_checkpoint_id": "mit_mixed_augm_model_average_20_pt",
189
+ "package_dir": "checkpoints/molecular_transformer_forward/mit_mixed_augm_model_average_20_pt",
190
+ "source_registry_refs": [
191
+ {
192
+ "checkpoint_id": "mit_mixed_augm_model_average_20_pt",
193
+ "model_id": "molecular_transformer_forward"
194
+ }
195
+ ],
196
+ "task_direction": "forward",
197
+ "training_dataset_refs": [
198
+ "uspto_mit_molecular_transformer_project_release"
199
+ ],
200
+ "weights_path": "model.pt",
201
+ "wrapper_family": "opennmt_transformer"
202
+ },
203
+ {
204
+ "auxiliary_files": [
205
+ "benchmarks/default_test.jsonl",
206
+ "benchmarks/metadata.json",
207
+ "benchmarks/native/src-test.txt",
208
+ "benchmarks/native/tgt-test.txt"
209
+ ],
210
+ "benchmark_datasets": [
211
+ {
212
+ "benchmark_semantics": "forward_product_prediction_with_mixed_inputs",
213
+ "benchmark_variant": "forward_product_prediction",
214
+ "dataset_id": "stereo_mixed_augm_model_average_20_pt_test",
215
+ "dataset_ref": "uspto_stereo_molecular_transformer_project_release",
216
+ "default": true,
217
+ "format": "jsonl",
218
+ "metadata_path": "benchmarks/metadata.json",
219
+ "native_input_format": "parallel_tokenized_smiles",
220
+ "native_source_paths": [
221
+ "benchmarks/native/src-test.txt",
222
+ "benchmarks/native/tgt-test.txt"
223
+ ],
224
+ "notes": [
225
+ "Native Molecular Transformer benchmark sources are the official tokenized test split files from the public Box release.",
226
+ "benchmarks/default_test.jsonl stores detokenized mixed-input reactant strings mapped to expected product SMILES for forward evaluation."
227
+ ],
228
+ "num_examples": 50258,
229
+ "path": "benchmarks/default_test.jsonl",
230
+ "record_schema": {
231
+ "inputs": [
232
+ "reactants"
233
+ ],
234
+ "targets": [
235
+ "expected_products"
236
+ ]
237
+ },
238
+ "serialized_input_format": "space_tokenized_smiles",
239
+ "split": "test",
240
+ "task_direction": "forward"
241
+ }
242
+ ],
243
+ "checkpoint_id": "stereo_mixed_augm_model_average_20_pt",
244
+ "embedding_supported": false,
245
+ "input_schema": {
246
+ "benchmark_variant": "forward_product_prediction",
247
+ "reactants_field": "reactants_mixed",
248
+ "serialized_format": "space_tokenized_smiles",
249
+ "target_field": "PRODUCT",
250
+ "tokenizer": "smiles_regex_space_tokenizer"
251
+ },
252
+ "model_id": "molecular_transformer_forward",
253
+ "normalized_checkpoint_id": "stereo_mixed_augm_model_average_20_pt",
254
+ "package_dir": "checkpoints/molecular_transformer_forward/stereo_mixed_augm_model_average_20_pt",
255
+ "source_registry_refs": [
256
+ {
257
+ "checkpoint_id": "stereo_mixed_augm_model_average_20_pt",
258
+ "model_id": "molecular_transformer_forward"
259
+ }
260
+ ],
261
+ "task_direction": "forward",
262
+ "training_dataset_refs": [
263
+ "uspto_stereo_molecular_transformer_project_release"
264
+ ],
265
+ "weights_path": "model.pt",
266
+ "wrapper_family": "opennmt_transformer"
267
+ },
268
+ {
269
+ "auxiliary_files": [
270
+ "aux/.gitattributes",
271
+ "aux/README.md",
272
+ "aux/UPSTREAM_URL.txt",
273
+ "aux/config.json",
274
+ "aux/generation_config.json",
275
+ "aux/hf_model_info.json",
276
+ "aux/provenance/README.txt",
277
+ "aux/provenance/ReactionT5v2_source_README.md",
278
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv",
279
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
280
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
281
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv",
282
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/train.csv",
283
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/val.csv",
284
+ "aux/provenance/reactiont5v2_forward_split_provenance.json",
285
+ "aux/special_tokens_map.json",
286
+ "aux/tokenizer.json",
287
+ "aux/tokenizer_config.json",
288
+ "aux/trainer_state.json",
289
+ "aux/training_args.bin",
290
+ "benchmarks/default_test.jsonl",
291
+ "benchmarks/metadata.json"
292
+ ],
293
+ "benchmark_datasets": [
294
+ {
295
+ "benchmark_semantics": "ord_trained_forward_checkpoint_evaluated_on_uspto_mit_holdout",
296
+ "benchmark_variant": "forward_product_prediction",
297
+ "blank_reagent_value": " ",
298
+ "dataset_id": "sagawa_reactiont5v2_forward_test",
299
+ "dataset_ref": "uspto_mit_forward_project_release",
300
+ "default": true,
301
+ "format": "jsonl",
302
+ "metadata_path": "benchmarks/metadata.json",
303
+ "native_input_format": "csv_forward_reaction_rows",
304
+ "native_source_paths": [
305
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv"
306
+ ],
307
+ "notes": [
308
+ "Native forward benchmark split for the public USPTO_MIT product-prediction task.",
309
+ "The normalized JSONL preserves reactants, reagents, and expected product(s) rather than the backward exact-match schema.",
310
+ "This ORD-trained checkpoint is evaluated on the USPTO_MIT holdout documented in the upstream README/model card."
311
+ ],
312
+ "num_examples": 40000,
313
+ "path": "benchmarks/default_test.jsonl",
314
+ "record_schema": {
315
+ "inputs": [
316
+ "reactants",
317
+ "reagents"
318
+ ],
319
+ "targets": [
320
+ "expected_products"
321
+ ]
322
+ },
323
+ "serialized_input_format": "REACTANT:<reactants>REAGENT:<reagents>",
324
+ "split": "test",
325
+ "task_direction": "forward"
326
+ }
327
+ ],
328
+ "checkpoint_id": "sagawa_reactiont5v2_forward",
329
+ "embedding_supported": false,
330
+ "input_schema": {
331
+ "blank_reagent_value": " ",
332
+ "reactants_field": "REACTANT",
333
+ "reagents_field": "REAGENT",
334
+ "serialized_format": "REACTANT:<reactants>REAGENT:<reagents>",
335
+ "target_field": "PRODUCT"
336
+ },
337
+ "model_id": "reactiont5v2_forward",
338
+ "normalized_checkpoint_id": "sagawa_reactiont5v2_forward",
339
+ "package_dir": "checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward",
340
+ "source_registry_refs": [
341
+ {
342
+ "checkpoint_id": "sagawa_reactiont5v2_forward",
343
+ "model_id": "reactiont5v2_forward"
344
+ }
345
+ ],
346
+ "task_direction": "forward",
347
+ "training_dataset_refs": [
348
+ "ord_forward_pretraining"
349
+ ],
350
+ "weights_path": "model.safetensors",
351
+ "wrapper_family": "seq2seq_t5"
352
+ },
353
+ {
354
+ "auxiliary_files": [
355
+ "aux/.gitattributes",
356
+ "aux/README.md",
357
+ "aux/UPSTREAM_URL.txt",
358
+ "aux/config.json",
359
+ "aux/generation_config.json",
360
+ "aux/hf_model_info.json",
361
+ "aux/provenance/README.txt",
362
+ "aux/provenance/ReactionT5v2_source_README.md",
363
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv",
364
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
365
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
366
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv",
367
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/train.csv",
368
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/val.csv",
369
+ "aux/provenance/reactiont5v2_forward_split_provenance.json",
370
+ "aux/special_tokens_map.json",
371
+ "aux/tokenizer.json",
372
+ "aux/tokenizer_config.json",
373
+ "aux/trainer_state.json",
374
+ "aux/training_args.bin",
375
+ "benchmarks/default_test.jsonl",
376
+ "benchmarks/metadata.json"
377
+ ],
378
+ "benchmark_datasets": [
379
+ {
380
+ "benchmark_semantics": "benchmark_specific_finetuned_forward_checkpoint",
381
+ "benchmark_variant": "forward_product_prediction",
382
+ "blank_reagent_value": " ",
383
+ "dataset_id": "sagawa_reactiont5v2_forward_uspto_mit_test",
384
+ "dataset_ref": "uspto_mit_forward_project_release",
385
+ "default": true,
386
+ "format": "jsonl",
387
+ "metadata_path": "benchmarks/metadata.json",
388
+ "native_input_format": "csv_forward_reaction_rows",
389
+ "native_source_paths": [
390
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv"
391
+ ],
392
+ "notes": [
393
+ "Native forward benchmark split for the public USPTO_MIT product-prediction task.",
394
+ "The normalized JSONL preserves reactants, reagents, and expected product(s) rather than the backward exact-match schema.",
395
+ "This is the native USPTO_MIT fine-tuned forward checkpoint benchmark split."
396
+ ],
397
+ "num_examples": 40000,
398
+ "path": "benchmarks/default_test.jsonl",
399
+ "record_schema": {
400
+ "inputs": [
401
+ "reactants",
402
+ "reagents"
403
+ ],
404
+ "targets": [
405
+ "expected_products"
406
+ ]
407
+ },
408
+ "serialized_input_format": "REACTANT:<reactants>REAGENT:<reagents>",
409
+ "split": "test",
410
+ "task_direction": "forward"
411
+ }
412
+ ],
413
+ "checkpoint_id": "sagawa_reactiont5v2_forward_uspto_mit",
414
+ "embedding_supported": false,
415
+ "input_schema": {
416
+ "blank_reagent_value": " ",
417
+ "reactants_field": "REACTANT",
418
+ "reagents_field": "REAGENT",
419
+ "serialized_format": "REACTANT:<reactants>REAGENT:<reagents>",
420
+ "target_field": "PRODUCT"
421
+ },
422
+ "model_id": "reactiont5v2_forward",
423
+ "normalized_checkpoint_id": "sagawa_reactiont5v2_forward_uspto_mit",
424
+ "package_dir": "checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit",
425
+ "source_registry_refs": [
426
+ {
427
+ "checkpoint_id": "sagawa_reactiont5v2_forward_uspto_mit",
428
+ "model_id": "reactiont5v2_forward"
429
+ }
430
+ ],
431
+ "task_direction": "forward",
432
+ "training_dataset_refs": [
433
+ "uspto_mit_forward_project_release"
434
+ ],
435
+ "weights_path": "model.safetensors",
436
+ "wrapper_family": "seq2seq_t5"
437
+ }
438
+ ],
439
+ "generated_at": "1970-01-01T00:00:00+00:00"
440
+ }
checkpoints/forward_runtime_registry.json ADDED
@@ -0,0 +1,519 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "checkpoints": [
3
+ {
4
+ "auxiliary_files": [
5
+ "aux/bart_vocab.json",
6
+ "aux/bart_vocab_downstream.json",
7
+ "aux/ds_config.json",
8
+ "aux/hparams.yaml",
9
+ "aux/hyper_parameters.json",
10
+ "benchmarks/default_test.jsonl",
11
+ "benchmarks/metadata.json",
12
+ "benchmarks/native/uspto_mixed.pickle"
13
+ ],
14
+ "benchmark_datasets": [
15
+ {
16
+ "benchmark_semantics": "forward_product_prediction_with_mixed_inputs",
17
+ "benchmark_variant": "forward_product_prediction",
18
+ "dataset_id": "uspto_mixed_forward_checkpoint_test",
19
+ "dataset_ref": "uspto_mixed_chemformer_project_release",
20
+ "default": true,
21
+ "format": "jsonl",
22
+ "metadata_path": "benchmarks/metadata.json",
23
+ "native_input_format": "pandas_pickled_dataframe",
24
+ "native_source_paths": [
25
+ "benchmarks/native/uspto_mixed.pickle"
26
+ ],
27
+ "notes": [
28
+ "Native Chemformer forward benchmark source extracted from the public Box release uspto_mixed pickle.",
29
+ "benchmarks/default_test.jsonl preserves reactant-side input SMILES plus expected product SMILES from the test partition."
30
+ ],
31
+ "num_examples": 40000,
32
+ "path": "benchmarks/default_test.jsonl",
33
+ "record_schema": {
34
+ "inputs": [
35
+ "reactants"
36
+ ],
37
+ "targets": [
38
+ "expected_products"
39
+ ]
40
+ },
41
+ "split": "test",
42
+ "task_direction": "forward"
43
+ }
44
+ ],
45
+ "checkpoint_id": "uspto_mixed_forward_checkpoint",
46
+ "display_name": "uspto_mixed forward checkpoint",
47
+ "embedding_modes": [],
48
+ "embedding_supported": false,
49
+ "fine_tuning_notes": "Forward Chemformer checkpoint fine-tuned on the public USPTO_mixed product-prediction split.",
50
+ "fine_tuning_role": "task_finetuned_seed",
51
+ "fine_tuning_supported": true,
52
+ "hf_model_name": null,
53
+ "host": "Box shared folder",
54
+ "inference_supported": true,
55
+ "input_schema": {
56
+ "reactants_field": "reactants_mol",
57
+ "reagents_field": null,
58
+ "serialization": "dot_joined_smiles",
59
+ "source_dataset_member": "data/seq-to-seq_datasets/uspto_mixed.pickle",
60
+ "target_field": "products_mol"
61
+ },
62
+ "larger_than_uspto50k": false,
63
+ "model_id": "chemformer_forward",
64
+ "model_name": "Chemformer Forward",
65
+ "normalized_checkpoint_id": "uspto_mixed_forward_checkpoint",
66
+ "package_dir": "checkpoints/chemformer_forward/uspto_mixed_forward_checkpoint",
67
+ "remote_url": "https://az.app.box.com/s/7eci3nd9vy0xplqniitpk02rbg9q2zcq/folder/145317555751",
68
+ "source_registry_refs": [
69
+ {
70
+ "checkpoint_id": "uspto_mixed_forward_checkpoint",
71
+ "model_id": "chemformer_forward"
72
+ }
73
+ ],
74
+ "task_direction": "forward",
75
+ "training_dataset_refs": [
76
+ "uspto_mixed_chemformer_project_release"
77
+ ],
78
+ "unsupported_reason": null,
79
+ "weights_path": "model.safetensors",
80
+ "wrapper_class": "syntheseus_wrappers.wrappers.chemformer_forward.ChemformerForwardModel",
81
+ "wrapper_family": "chemformer_bart"
82
+ },
83
+ {
84
+ "auxiliary_files": [
85
+ "aux/bart_vocab.json",
86
+ "aux/bart_vocab_downstream.json",
87
+ "aux/ds_config.json",
88
+ "aux/hparams.yaml",
89
+ "aux/hyper_parameters.json",
90
+ "benchmarks/default_test.jsonl",
91
+ "benchmarks/metadata.json",
92
+ "benchmarks/native/uspto_sep.pickle"
93
+ ],
94
+ "benchmark_datasets": [
95
+ {
96
+ "benchmark_semantics": "forward_product_prediction_with_separated_reagents",
97
+ "benchmark_variant": "forward_product_prediction",
98
+ "dataset_id": "uspto_sep_forward_checkpoint_test",
99
+ "dataset_ref": "uspto_sep_chemformer_project_release",
100
+ "default": true,
101
+ "format": "jsonl",
102
+ "metadata_path": "benchmarks/metadata.json",
103
+ "native_input_format": "pandas_pickled_dataframe",
104
+ "native_source_paths": [
105
+ "benchmarks/native/uspto_sep.pickle"
106
+ ],
107
+ "notes": [
108
+ "Native Chemformer forward benchmark source extracted from the public Box release uspto_sep pickle.",
109
+ "benchmarks/default_test.jsonl preserves separated reactants/reagents fields plus expected product SMILES from the test partition."
110
+ ],
111
+ "num_examples": 40000,
112
+ "path": "benchmarks/default_test.jsonl",
113
+ "record_schema": {
114
+ "inputs": [
115
+ "reactants",
116
+ "reagents"
117
+ ],
118
+ "targets": [
119
+ "expected_products"
120
+ ]
121
+ },
122
+ "split": "test",
123
+ "task_direction": "forward"
124
+ }
125
+ ],
126
+ "checkpoint_id": "uspto_sep_forward_checkpoint",
127
+ "display_name": "uspto_sep forward checkpoint",
128
+ "embedding_modes": [],
129
+ "embedding_supported": false,
130
+ "fine_tuning_notes": "Forward Chemformer checkpoint fine-tuned on the public USPTO_sep product-prediction split.",
131
+ "fine_tuning_role": "task_finetuned_seed",
132
+ "fine_tuning_supported": true,
133
+ "hf_model_name": null,
134
+ "host": "Box shared folder",
135
+ "inference_supported": true,
136
+ "input_schema": {
137
+ "reactants_field": "reactants_mol",
138
+ "reagents_field": "reagents_mol",
139
+ "serialization": "dot_joined_smiles",
140
+ "source_dataset_member": "data/seq-to-seq_datasets/uspto_sep.pickle",
141
+ "target_field": "products_mol"
142
+ },
143
+ "larger_than_uspto50k": false,
144
+ "model_id": "chemformer_forward",
145
+ "model_name": "Chemformer Forward",
146
+ "normalized_checkpoint_id": "uspto_sep_forward_checkpoint",
147
+ "package_dir": "checkpoints/chemformer_forward/uspto_sep_forward_checkpoint",
148
+ "remote_url": "https://az.app.box.com/s/7eci3nd9vy0xplqniitpk02rbg9q2zcq/folder/145317555751",
149
+ "source_registry_refs": [
150
+ {
151
+ "checkpoint_id": "uspto_sep_forward_checkpoint",
152
+ "model_id": "chemformer_forward"
153
+ }
154
+ ],
155
+ "task_direction": "forward",
156
+ "training_dataset_refs": [
157
+ "uspto_sep_chemformer_project_release"
158
+ ],
159
+ "unsupported_reason": null,
160
+ "weights_path": "model.safetensors",
161
+ "wrapper_class": "syntheseus_wrappers.wrappers.chemformer_forward.ChemformerForwardModel",
162
+ "wrapper_family": "chemformer_bart"
163
+ },
164
+ {
165
+ "auxiliary_files": [
166
+ "benchmarks/default_test.jsonl",
167
+ "benchmarks/metadata.json",
168
+ "benchmarks/native/src-test.txt",
169
+ "benchmarks/native/tgt-test.txt"
170
+ ],
171
+ "benchmark_datasets": [
172
+ {
173
+ "benchmark_semantics": "forward_product_prediction_with_mixed_inputs",
174
+ "benchmark_variant": "forward_product_prediction",
175
+ "dataset_id": "mit_mixed_augm_model_average_20_pt_test",
176
+ "dataset_ref": "uspto_mit_molecular_transformer_project_release",
177
+ "default": true,
178
+ "format": "jsonl",
179
+ "metadata_path": "benchmarks/metadata.json",
180
+ "native_input_format": "parallel_tokenized_smiles",
181
+ "native_source_paths": [
182
+ "benchmarks/native/src-test.txt",
183
+ "benchmarks/native/tgt-test.txt"
184
+ ],
185
+ "notes": [
186
+ "Native Molecular Transformer benchmark sources are the official tokenized test split files from the public Box release.",
187
+ "benchmarks/default_test.jsonl stores detokenized mixed-input reactant strings mapped to expected product SMILES for forward evaluation."
188
+ ],
189
+ "num_examples": 40000,
190
+ "path": "benchmarks/default_test.jsonl",
191
+ "record_schema": {
192
+ "inputs": [
193
+ "reactants"
194
+ ],
195
+ "targets": [
196
+ "expected_products"
197
+ ]
198
+ },
199
+ "serialized_input_format": "space_tokenized_smiles",
200
+ "split": "test",
201
+ "task_direction": "forward"
202
+ }
203
+ ],
204
+ "checkpoint_id": "mit_mixed_augm_model_average_20_pt",
205
+ "display_name": "MIT_mixed_augm_model_average_20.pt",
206
+ "embedding_modes": [],
207
+ "embedding_supported": false,
208
+ "fine_tuning_notes": null,
209
+ "fine_tuning_role": null,
210
+ "fine_tuning_supported": false,
211
+ "hf_model_name": null,
212
+ "host": "IBM Box shared folder",
213
+ "inference_supported": true,
214
+ "input_schema": {
215
+ "benchmark_variant": "forward_product_prediction",
216
+ "reactants_field": "reactants_mixed",
217
+ "serialized_format": "space_tokenized_smiles",
218
+ "target_field": "PRODUCT",
219
+ "tokenizer": "smiles_regex_space_tokenizer"
220
+ },
221
+ "larger_than_uspto50k": true,
222
+ "model_id": "molecular_transformer_forward",
223
+ "model_name": "Molecular Transformer Forward",
224
+ "normalized_checkpoint_id": "mit_mixed_augm_model_average_20_pt",
225
+ "package_dir": "checkpoints/molecular_transformer_forward/mit_mixed_augm_model_average_20_pt",
226
+ "remote_url": "https://ibm.box.com/v/MolecularTransformerModels",
227
+ "source_registry_refs": [
228
+ {
229
+ "checkpoint_id": "mit_mixed_augm_model_average_20_pt",
230
+ "model_id": "molecular_transformer_forward"
231
+ }
232
+ ],
233
+ "task_direction": "forward",
234
+ "training_dataset_refs": [
235
+ "uspto_mit_molecular_transformer_project_release"
236
+ ],
237
+ "unsupported_reason": null,
238
+ "weights_path": "model.pt",
239
+ "wrapper_class": "syntheseus_wrappers.wrappers.molecular_transformer_forward.MolecularTransformerForwardModel",
240
+ "wrapper_family": "opennmt_transformer"
241
+ },
242
+ {
243
+ "auxiliary_files": [
244
+ "benchmarks/default_test.jsonl",
245
+ "benchmarks/metadata.json",
246
+ "benchmarks/native/src-test.txt",
247
+ "benchmarks/native/tgt-test.txt"
248
+ ],
249
+ "benchmark_datasets": [
250
+ {
251
+ "benchmark_semantics": "forward_product_prediction_with_mixed_inputs",
252
+ "benchmark_variant": "forward_product_prediction",
253
+ "dataset_id": "stereo_mixed_augm_model_average_20_pt_test",
254
+ "dataset_ref": "uspto_stereo_molecular_transformer_project_release",
255
+ "default": true,
256
+ "format": "jsonl",
257
+ "metadata_path": "benchmarks/metadata.json",
258
+ "native_input_format": "parallel_tokenized_smiles",
259
+ "native_source_paths": [
260
+ "benchmarks/native/src-test.txt",
261
+ "benchmarks/native/tgt-test.txt"
262
+ ],
263
+ "notes": [
264
+ "Native Molecular Transformer benchmark sources are the official tokenized test split files from the public Box release.",
265
+ "benchmarks/default_test.jsonl stores detokenized mixed-input reactant strings mapped to expected product SMILES for forward evaluation."
266
+ ],
267
+ "num_examples": 50258,
268
+ "path": "benchmarks/default_test.jsonl",
269
+ "record_schema": {
270
+ "inputs": [
271
+ "reactants"
272
+ ],
273
+ "targets": [
274
+ "expected_products"
275
+ ]
276
+ },
277
+ "serialized_input_format": "space_tokenized_smiles",
278
+ "split": "test",
279
+ "task_direction": "forward"
280
+ }
281
+ ],
282
+ "checkpoint_id": "stereo_mixed_augm_model_average_20_pt",
283
+ "display_name": "STEREO_mixed_augm_model_average_20.pt",
284
+ "embedding_modes": [],
285
+ "embedding_supported": false,
286
+ "fine_tuning_notes": null,
287
+ "fine_tuning_role": null,
288
+ "fine_tuning_supported": false,
289
+ "hf_model_name": null,
290
+ "host": "IBM Box shared folder",
291
+ "inference_supported": true,
292
+ "input_schema": {
293
+ "benchmark_variant": "forward_product_prediction",
294
+ "reactants_field": "reactants_mixed",
295
+ "serialized_format": "space_tokenized_smiles",
296
+ "target_field": "PRODUCT",
297
+ "tokenizer": "smiles_regex_space_tokenizer"
298
+ },
299
+ "larger_than_uspto50k": true,
300
+ "model_id": "molecular_transformer_forward",
301
+ "model_name": "Molecular Transformer Forward",
302
+ "normalized_checkpoint_id": "stereo_mixed_augm_model_average_20_pt",
303
+ "package_dir": "checkpoints/molecular_transformer_forward/stereo_mixed_augm_model_average_20_pt",
304
+ "remote_url": "https://ibm.box.com/v/MolecularTransformerModels",
305
+ "source_registry_refs": [
306
+ {
307
+ "checkpoint_id": "stereo_mixed_augm_model_average_20_pt",
308
+ "model_id": "molecular_transformer_forward"
309
+ }
310
+ ],
311
+ "task_direction": "forward",
312
+ "training_dataset_refs": [
313
+ "uspto_stereo_molecular_transformer_project_release"
314
+ ],
315
+ "unsupported_reason": null,
316
+ "weights_path": "model.pt",
317
+ "wrapper_class": "syntheseus_wrappers.wrappers.molecular_transformer_forward.MolecularTransformerForwardModel",
318
+ "wrapper_family": "opennmt_transformer"
319
+ },
320
+ {
321
+ "auxiliary_files": [
322
+ "aux/.gitattributes",
323
+ "aux/README.md",
324
+ "aux/UPSTREAM_URL.txt",
325
+ "aux/config.json",
326
+ "aux/generation_config.json",
327
+ "aux/hf_model_info.json",
328
+ "aux/provenance/README.txt",
329
+ "aux/provenance/ReactionT5v2_source_README.md",
330
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv",
331
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
332
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
333
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv",
334
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/train.csv",
335
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/val.csv",
336
+ "aux/provenance/reactiont5v2_forward_split_provenance.json",
337
+ "aux/special_tokens_map.json",
338
+ "aux/tokenizer.json",
339
+ "aux/tokenizer_config.json",
340
+ "aux/trainer_state.json",
341
+ "aux/training_args.bin",
342
+ "benchmarks/default_test.jsonl",
343
+ "benchmarks/metadata.json"
344
+ ],
345
+ "benchmark_datasets": [
346
+ {
347
+ "benchmark_semantics": "ord_trained_forward_checkpoint_evaluated_on_uspto_mit_holdout",
348
+ "benchmark_variant": "forward_product_prediction",
349
+ "blank_reagent_value": " ",
350
+ "dataset_id": "sagawa_reactiont5v2_forward_test",
351
+ "dataset_ref": "uspto_mit_forward_project_release",
352
+ "default": true,
353
+ "format": "jsonl",
354
+ "metadata_path": "benchmarks/metadata.json",
355
+ "native_input_format": "csv_forward_reaction_rows",
356
+ "native_source_paths": [
357
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv"
358
+ ],
359
+ "notes": [
360
+ "Native forward benchmark split for the public USPTO_MIT product-prediction task.",
361
+ "The normalized JSONL preserves reactants, reagents, and expected product(s) rather than the backward exact-match schema.",
362
+ "This ORD-trained checkpoint is evaluated on the USPTO_MIT holdout documented in the upstream README/model card."
363
+ ],
364
+ "num_examples": 40000,
365
+ "path": "benchmarks/default_test.jsonl",
366
+ "record_schema": {
367
+ "inputs": [
368
+ "reactants",
369
+ "reagents"
370
+ ],
371
+ "targets": [
372
+ "expected_products"
373
+ ]
374
+ },
375
+ "serialized_input_format": "REACTANT:<reactants>REAGENT:<reagents>",
376
+ "split": "test",
377
+ "task_direction": "forward"
378
+ }
379
+ ],
380
+ "checkpoint_id": "sagawa_reactiont5v2_forward",
381
+ "display_name": "sagawa/ReactionT5v2-forward",
382
+ "embedding_modes": [],
383
+ "embedding_supported": false,
384
+ "fine_tuning_notes": "ORD-trained forward-synthesis checkpoint that can serve as a continued-training seed.",
385
+ "fine_tuning_role": "task_finetuned_seed",
386
+ "fine_tuning_supported": true,
387
+ "hf_model_name": "sagawa/ReactionT5v2-forward",
388
+ "host": "Hugging Face",
389
+ "inference_supported": true,
390
+ "input_schema": {
391
+ "blank_reagent_value": " ",
392
+ "reactants_field": "REACTANT",
393
+ "reagents_field": "REAGENT",
394
+ "serialized_format": "REACTANT:<reactants>REAGENT:<reagents>",
395
+ "target_field": "PRODUCT"
396
+ },
397
+ "larger_than_uspto50k": true,
398
+ "model_id": "reactiont5v2_forward",
399
+ "model_name": "ReactionT5v2 Forward",
400
+ "normalized_checkpoint_id": "sagawa_reactiont5v2_forward",
401
+ "package_dir": "checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward",
402
+ "remote_url": "https://huggingface.co/sagawa/ReactionT5v2-forward",
403
+ "source_registry_refs": [
404
+ {
405
+ "checkpoint_id": "sagawa_reactiont5v2_forward",
406
+ "model_id": "reactiont5v2_forward"
407
+ }
408
+ ],
409
+ "task_direction": "forward",
410
+ "training_dataset_refs": [
411
+ "ord_forward_pretraining"
412
+ ],
413
+ "unsupported_reason": null,
414
+ "weights_path": "model.safetensors",
415
+ "wrapper_class": "syntheseus_wrappers.wrappers.reactiont5v2_forward.ReactionT5v2ForwardModel",
416
+ "wrapper_family": "seq2seq_t5"
417
+ },
418
+ {
419
+ "auxiliary_files": [
420
+ "aux/.gitattributes",
421
+ "aux/README.md",
422
+ "aux/UPSTREAM_URL.txt",
423
+ "aux/config.json",
424
+ "aux/generation_config.json",
425
+ "aux/hf_model_info.json",
426
+ "aux/provenance/README.txt",
427
+ "aux/provenance/ReactionT5v2_source_README.md",
428
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv",
429
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
430
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
431
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv",
432
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/train.csv",
433
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/val.csv",
434
+ "aux/provenance/reactiont5v2_forward_split_provenance.json",
435
+ "aux/special_tokens_map.json",
436
+ "aux/tokenizer.json",
437
+ "aux/tokenizer_config.json",
438
+ "aux/trainer_state.json",
439
+ "aux/training_args.bin",
440
+ "benchmarks/default_test.jsonl",
441
+ "benchmarks/metadata.json"
442
+ ],
443
+ "benchmark_datasets": [
444
+ {
445
+ "benchmark_semantics": "benchmark_specific_finetuned_forward_checkpoint",
446
+ "benchmark_variant": "forward_product_prediction",
447
+ "blank_reagent_value": " ",
448
+ "dataset_id": "sagawa_reactiont5v2_forward_uspto_mit_test",
449
+ "dataset_ref": "uspto_mit_forward_project_release",
450
+ "default": true,
451
+ "format": "jsonl",
452
+ "metadata_path": "benchmarks/metadata.json",
453
+ "native_input_format": "csv_forward_reaction_rows",
454
+ "native_source_paths": [
455
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv"
456
+ ],
457
+ "notes": [
458
+ "Native forward benchmark split for the public USPTO_MIT product-prediction task.",
459
+ "The normalized JSONL preserves reactants, reagents, and expected product(s) rather than the backward exact-match schema.",
460
+ "This is the native USPTO_MIT fine-tuned forward checkpoint benchmark split."
461
+ ],
462
+ "num_examples": 40000,
463
+ "path": "benchmarks/default_test.jsonl",
464
+ "record_schema": {
465
+ "inputs": [
466
+ "reactants",
467
+ "reagents"
468
+ ],
469
+ "targets": [
470
+ "expected_products"
471
+ ]
472
+ },
473
+ "serialized_input_format": "REACTANT:<reactants>REAGENT:<reagents>",
474
+ "split": "test",
475
+ "task_direction": "forward"
476
+ }
477
+ ],
478
+ "checkpoint_id": "sagawa_reactiont5v2_forward_uspto_mit",
479
+ "display_name": "sagawa/ReactionT5v2-forward-USPTO_MIT",
480
+ "embedding_modes": [],
481
+ "embedding_supported": false,
482
+ "fine_tuning_notes": "USPTO_MIT fine-tuned forward-synthesis checkpoint that can be adapted further.",
483
+ "fine_tuning_role": "task_finetuned_seed",
484
+ "fine_tuning_supported": true,
485
+ "hf_model_name": "sagawa/ReactionT5v2-forward-USPTO_MIT",
486
+ "host": "Hugging Face",
487
+ "inference_supported": true,
488
+ "input_schema": {
489
+ "blank_reagent_value": " ",
490
+ "reactants_field": "REACTANT",
491
+ "reagents_field": "REAGENT",
492
+ "serialized_format": "REACTANT:<reactants>REAGENT:<reagents>",
493
+ "target_field": "PRODUCT"
494
+ },
495
+ "larger_than_uspto50k": true,
496
+ "model_id": "reactiont5v2_forward",
497
+ "model_name": "ReactionT5v2 Forward",
498
+ "normalized_checkpoint_id": "sagawa_reactiont5v2_forward_uspto_mit",
499
+ "package_dir": "checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit",
500
+ "remote_url": "https://huggingface.co/sagawa/ReactionT5v2-forward-USPTO_MIT",
501
+ "source_registry_refs": [
502
+ {
503
+ "checkpoint_id": "sagawa_reactiont5v2_forward_uspto_mit",
504
+ "model_id": "reactiont5v2_forward"
505
+ }
506
+ ],
507
+ "task_direction": "forward",
508
+ "training_dataset_refs": [
509
+ "uspto_mit_forward_project_release"
510
+ ],
511
+ "unsupported_reason": null,
512
+ "weights_path": "model.safetensors",
513
+ "wrapper_class": "syntheseus_wrappers.wrappers.reactiont5v2_forward.ReactionT5v2ForwardModel",
514
+ "wrapper_family": "seq2seq_t5"
515
+ }
516
+ ],
517
+ "registry_kind": "runtime",
518
+ "runtime_registry_version": 1
519
+ }
checkpoints/normalize_report.json ADDED
@@ -0,0 +1,269 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "archive_root": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/DOES_NOT_EXIST",
3
+ "datasets_root": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets",
4
+ "generated_at": "2026-06-28T00:57:27.214428+00:00",
5
+ "normalized_packages": {
6
+ "aizynthfinder/uspto_model_onnx": {
7
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/aizynthfinder/uspto_model_onnx",
8
+ "source_registry_refs": [
9
+ {
10
+ "checkpoint_id": "uspto_model_onnx",
11
+ "model_id": "aizynthfinder"
12
+ }
13
+ ],
14
+ "status": "ok"
15
+ },
16
+ "chemformer/combined_large_pre_trained_checkpoint": {
17
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/chemformer/combined_large_pre_trained_checkpoint",
18
+ "source_registry_refs": [
19
+ {
20
+ "checkpoint_id": "combined_large_pre_trained_checkpoint",
21
+ "model_id": "chemformer"
22
+ }
23
+ ],
24
+ "status": "ok"
25
+ },
26
+ "chemformer/combined_pre_trained_base_checkpoint": {
27
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/chemformer/combined_pre_trained_base_checkpoint",
28
+ "source_registry_refs": [
29
+ {
30
+ "checkpoint_id": "combined_pre_trained_base_checkpoint",
31
+ "model_id": "chemformer"
32
+ }
33
+ ],
34
+ "status": "ok"
35
+ },
36
+ "chemformer/uspto_50_fine_tuned_retrosynthesis_checkpoint": {
37
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/chemformer/uspto_50_fine_tuned_retrosynthesis_checkpoint",
38
+ "source_registry_refs": [
39
+ {
40
+ "checkpoint_id": "uspto_50_fine_tuned_retrosynthesis_checkpoint",
41
+ "model_id": "chemformer"
42
+ }
43
+ ],
44
+ "status": "ok"
45
+ },
46
+ "conretrobert/uspto_50k_ema_deep_epoch059_step0_val_top1_acc0_3521_ckpt": {
47
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/conretrobert/uspto_50k_ema_deep_epoch059_step0_val_top1_acc0_3521_ckpt",
48
+ "source_registry_refs": [
49
+ {
50
+ "checkpoint_id": "uspto_50k_ema_deep_epoch059_step0_val_top1_acc0_3521_ckpt",
51
+ "model_id": "conretrobert"
52
+ }
53
+ ],
54
+ "status": "ok"
55
+ },
56
+ "conretrobert/uspto_50k_ema_mid_epoch087_step0_val_top1_acc0_3547_val_loss_4_5051_ckpt": {
57
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/conretrobert/uspto_50k_ema_mid_epoch087_step0_val_top1_acc0_3547_val_loss_4_5051_ckpt",
58
+ "source_registry_refs": [
59
+ {
60
+ "checkpoint_id": "uspto_50k_ema_mid_epoch087_step0_val_top1_acc0_3547_val_loss_4_5051_ckpt",
61
+ "model_id": "conretrobert"
62
+ }
63
+ ],
64
+ "status": "ok"
65
+ },
66
+ "conretrobert/uspto_50k_frozen_te_epoch087_step54714_val_top1_acc0_3711_ckpt": {
67
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/conretrobert/uspto_50k_frozen_te_epoch087_step54714_val_top1_acc0_3711_ckpt",
68
+ "source_registry_refs": [
69
+ {
70
+ "checkpoint_id": "uspto_50k_frozen_te_epoch087_step54714_val_top1_acc0_3711_ckpt",
71
+ "model_id": "conretrobert"
72
+ }
73
+ ],
74
+ "status": "ok"
75
+ },
76
+ "conretrobert/uspto_full_2_uspto_50k_magic_epoch012_step0_weighted_val_accuracy0_8390_ckpt": {
77
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/conretrobert/uspto_full_2_uspto_50k_magic_epoch012_step0_weighted_val_accuracy0_8390_ckpt",
78
+ "source_registry_refs": [
79
+ {
80
+ "checkpoint_id": "uspto_full_2_uspto_50k_magic_epoch012_step0_weighted_val_accuracy0_8390_ckpt",
81
+ "model_id": "conretrobert"
82
+ }
83
+ ],
84
+ "status": "ok"
85
+ },
86
+ "conretrobert/uspto_full_frozen_te_epoch096_step129890_val_top1_acc0_6477_ckpt": {
87
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/conretrobert/uspto_full_frozen_te_epoch096_step129890_val_top1_acc0_6477_ckpt",
88
+ "source_registry_refs": [
89
+ {
90
+ "checkpoint_id": "uspto_full_frozen_te_epoch096_step129890_val_top1_acc0_6477_ckpt",
91
+ "model_id": "conretrobert"
92
+ }
93
+ ],
94
+ "status": "ok"
95
+ },
96
+ "conretrobert/uspto_full_transfer_uspto_50k_ft_epoch046_step29212_val_top1_acc0_7501_val_loss_1_3350_ckpt": {
97
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/conretrobert/uspto_full_transfer_uspto_50k_ft_epoch046_step29212_val_top1_acc0_7501_val_loss_1_3350_ckpt",
98
+ "source_registry_refs": [
99
+ {
100
+ "checkpoint_id": "uspto_full_transfer_uspto_50k_ft_epoch046_step29212_val_top1_acc0_7501_val_loss_1_3350_ckpt",
101
+ "model_id": "conretrobert"
102
+ }
103
+ ],
104
+ "status": "ok"
105
+ },
106
+ "reactiont5v2/sagawa_reactiont5v2_retrosynthesis": {
107
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/reactiont5v2/sagawa_reactiont5v2_retrosynthesis",
108
+ "source_registry_refs": [
109
+ {
110
+ "checkpoint_id": "sagawa_reactiont5v2_retrosynthesis",
111
+ "model_id": "reactiont5v2"
112
+ }
113
+ ],
114
+ "status": "ok"
115
+ },
116
+ "reactiont5v2/sagawa_reactiont5v2_retrosynthesis_uspto_50k": {
117
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/reactiont5v2/sagawa_reactiont5v2_retrosynthesis_uspto_50k",
118
+ "source_registry_refs": [
119
+ {
120
+ "checkpoint_id": "sagawa_reactiont5v2_retrosynthesis_uspto_50k",
121
+ "model_id": "reactiont5v2"
122
+ }
123
+ ],
124
+ "status": "ok"
125
+ },
126
+ "retrochimera_hybrid/pistachio": {
127
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/retrochimera_hybrid/pistachio",
128
+ "source_registry_refs": [
129
+ {
130
+ "checkpoint_id": "pistachio",
131
+ "model_id": "retrochimera_hybrid"
132
+ }
133
+ ],
134
+ "status": "ok"
135
+ },
136
+ "retrochimera_hybrid/uspto_50k": {
137
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/retrochimera_hybrid/uspto_50k",
138
+ "source_registry_refs": [
139
+ {
140
+ "checkpoint_id": "uspto_50k",
141
+ "model_id": "retrochimera_hybrid"
142
+ }
143
+ ],
144
+ "status": "ok"
145
+ },
146
+ "retrochimera_hybrid/uspto_full": {
147
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/retrochimera_hybrid/uspto_full",
148
+ "source_registry_refs": [
149
+ {
150
+ "checkpoint_id": "uspto_full",
151
+ "model_id": "retrochimera_hybrid"
152
+ }
153
+ ],
154
+ "status": "ok"
155
+ },
156
+ "retrochimera_template/pistachio": {
157
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/retrochimera_template/pistachio",
158
+ "source_registry_refs": [
159
+ {
160
+ "checkpoint_id": "pistachio",
161
+ "model_id": "retrochimera_template"
162
+ }
163
+ ],
164
+ "status": "ok"
165
+ },
166
+ "retrochimera_template/uspto_50k": {
167
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/retrochimera_template/uspto_50k",
168
+ "source_registry_refs": [
169
+ {
170
+ "checkpoint_id": "uspto_50k",
171
+ "model_id": "retrochimera_template"
172
+ }
173
+ ],
174
+ "status": "ok"
175
+ },
176
+ "retrochimera_template/uspto_full": {
177
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/retrochimera_template/uspto_full",
178
+ "source_registry_refs": [
179
+ {
180
+ "checkpoint_id": "uspto_full",
181
+ "model_id": "retrochimera_template"
182
+ }
183
+ ],
184
+ "status": "ok"
185
+ },
186
+ "retrochimera_transformer/pistachio": {
187
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/retrochimera_transformer/pistachio",
188
+ "source_registry_refs": [
189
+ {
190
+ "checkpoint_id": "pistachio",
191
+ "model_id": "retrochimera_transformer"
192
+ }
193
+ ],
194
+ "status": "ok"
195
+ },
196
+ "retrochimera_transformer/uspto_50k": {
197
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/retrochimera_transformer/uspto_50k",
198
+ "source_registry_refs": [
199
+ {
200
+ "checkpoint_id": "uspto_50k",
201
+ "model_id": "retrochimera_transformer"
202
+ }
203
+ ],
204
+ "status": "ok"
205
+ },
206
+ "retrochimera_transformer/uspto_full": {
207
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/retrochimera_transformer/uspto_full",
208
+ "source_registry_refs": [
209
+ {
210
+ "checkpoint_id": "uspto_full",
211
+ "model_id": "retrochimera_transformer"
212
+ }
213
+ ],
214
+ "status": "ok"
215
+ },
216
+ "rsgpt/finetune_50k_aug20_pth": {
217
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/rsgpt/finetune_50k_aug20_pth",
218
+ "source_registry_refs": [
219
+ {
220
+ "checkpoint_id": "finetune_50k_aug20_pth",
221
+ "model_id": "rsgpt"
222
+ }
223
+ ],
224
+ "status": "ok"
225
+ },
226
+ "rsgpt/finetune_50k_pth": {
227
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/rsgpt/finetune_50k_pth",
228
+ "source_registry_refs": [
229
+ {
230
+ "checkpoint_id": "finetune_50k_pth",
231
+ "model_id": "rsgpt"
232
+ }
233
+ ],
234
+ "status": "ok"
235
+ },
236
+ "rsgpt/finetune_full_pth": {
237
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/rsgpt/finetune_full_pth",
238
+ "source_registry_refs": [
239
+ {
240
+ "checkpoint_id": "finetune_full_pth",
241
+ "model_id": "rsgpt"
242
+ }
243
+ ],
244
+ "status": "ok"
245
+ },
246
+ "rsgpt/finetune_mit_pth": {
247
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/rsgpt/finetune_mit_pth",
248
+ "source_registry_refs": [
249
+ {
250
+ "checkpoint_id": "finetune_mit_pth",
251
+ "model_id": "rsgpt"
252
+ }
253
+ ],
254
+ "status": "ok"
255
+ },
256
+ "rsgpt/nopretrain_50k_pth": {
257
+ "package_dir": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints/rsgpt/nopretrain_50k_pth",
258
+ "source_registry_refs": [
259
+ {
260
+ "checkpoint_id": "nopretrain_50k_pth",
261
+ "model_id": "rsgpt"
262
+ }
263
+ ],
264
+ "status": "ok"
265
+ }
266
+ },
267
+ "output_root": "/home/ian-l/lucai/syntheseus-bundle-builder/generated/checkpoints",
268
+ "source_root": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/source_snapshots"
269
+ }
checkpoints/normalized_registry.json ADDED
@@ -0,0 +1,1385 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "checkpoints": [
3
+ {
4
+ "auxiliary_files": [
5
+ "aux/config.yml",
6
+ "aux/uspto_filter_model.onnx",
7
+ "aux/uspto_templates.csv.gz",
8
+ "benchmarks/default_test.jsonl",
9
+ "benchmarks/metadata.json",
10
+ "benchmarks/native/ref_routes_reaction_hashes.txt",
11
+ "benchmarks/native/selected_reactions.csv"
12
+ ],
13
+ "benchmark_datasets": [
14
+ {
15
+ "benchmark_variant": "paroutes_reference_route_holdout",
16
+ "dataset_id": "aizynthfinder_uspto_onnx_test",
17
+ "dataset_ref": "paroutes2_reference_route_reactions",
18
+ "default": true,
19
+ "format": "jsonl",
20
+ "metadata_path": "benchmarks/metadata.json",
21
+ "native_input_format": "tab_delimited_reactants_reagents_product_rows",
22
+ "native_source_paths": [
23
+ "benchmarks/native/selected_reactions.csv",
24
+ "benchmarks/native/ref_routes_reaction_hashes.txt"
25
+ ],
26
+ "notes": [
27
+ "Default bundled benchmark for the public AiZynthFinder USPTO ONNX policy.",
28
+ "The benchmark rows are reconstructed from PaRoutes selected_reactions.csv and ref_routes_reaction_hashes.txt so they match the held-out reference-route reactions used as the AiZynthTrain test set."
29
+ ],
30
+ "num_examples": 56657,
31
+ "path": "benchmarks/default_test.jsonl",
32
+ "split": "test"
33
+ }
34
+ ],
35
+ "checkpoint_id": "uspto_model_onnx",
36
+ "embedding_supported": false,
37
+ "input_schema": null,
38
+ "model_id": "aizynthfinder",
39
+ "normalized_checkpoint_id": "uspto_model_onnx",
40
+ "package_dir": "checkpoints/aizynthfinder/uspto_model_onnx",
41
+ "source_registry_refs": [
42
+ {
43
+ "checkpoint_id": "uspto_model_onnx",
44
+ "model_id": "aizynthfinder"
45
+ }
46
+ ],
47
+ "task_direction": null,
48
+ "training_dataset_refs": [
49
+ "aizynthtrain_uspto_selected_reactions_with_paroutes_holdout"
50
+ ],
51
+ "weights_path": "model.onnx",
52
+ "wrapper_family": "aizynthfinder_template_policy"
53
+ },
54
+ {
55
+ "auxiliary_files": [
56
+ "aux/bart_vocab.json",
57
+ "aux/bart_vocab_downstream.json",
58
+ "aux/ds_config.json",
59
+ "aux/hparams.yaml",
60
+ "aux/hyper_parameters.json"
61
+ ],
62
+ "checkpoint_id": "combined_large_pre_trained_checkpoint",
63
+ "embedding_supported": false,
64
+ "input_schema": null,
65
+ "model_id": "chemformer",
66
+ "normalized_checkpoint_id": "combined_large_pre_trained_checkpoint",
67
+ "package_dir": "checkpoints/chemformer/combined_large_pre_trained_checkpoint",
68
+ "source_registry_refs": [
69
+ {
70
+ "checkpoint_id": "combined_large_pre_trained_checkpoint",
71
+ "model_id": "chemformer"
72
+ }
73
+ ],
74
+ "task_direction": null,
75
+ "training_dataset_refs": [
76
+ "zinc15_100m_chemformer_pretraining"
77
+ ],
78
+ "weights_path": "model.safetensors",
79
+ "wrapper_family": "chemformer_bart"
80
+ },
81
+ {
82
+ "auxiliary_files": [
83
+ "aux/bart_vocab.json",
84
+ "aux/bart_vocab_downstream.json",
85
+ "aux/ds_config.json",
86
+ "aux/hparams.yaml",
87
+ "aux/hyper_parameters.json"
88
+ ],
89
+ "checkpoint_id": "combined_pre_trained_base_checkpoint",
90
+ "embedding_supported": false,
91
+ "input_schema": null,
92
+ "model_id": "chemformer",
93
+ "normalized_checkpoint_id": "combined_pre_trained_base_checkpoint",
94
+ "package_dir": "checkpoints/chemformer/combined_pre_trained_base_checkpoint",
95
+ "source_registry_refs": [
96
+ {
97
+ "checkpoint_id": "combined_pre_trained_base_checkpoint",
98
+ "model_id": "chemformer"
99
+ }
100
+ ],
101
+ "task_direction": null,
102
+ "training_dataset_refs": [
103
+ "zinc15_100m_chemformer_pretraining"
104
+ ],
105
+ "weights_path": "model.safetensors",
106
+ "wrapper_family": "chemformer_bart"
107
+ },
108
+ {
109
+ "auxiliary_files": [
110
+ "aux/bart_vocab.json",
111
+ "aux/bart_vocab_downstream.json",
112
+ "aux/ds_config.json",
113
+ "aux/hparams.yaml",
114
+ "aux/hyper_parameters.json",
115
+ "benchmarks/default_test.jsonl",
116
+ "benchmarks/metadata.json",
117
+ "benchmarks/native/uspto_50.pickle"
118
+ ],
119
+ "benchmark_datasets": [
120
+ {
121
+ "dataset_id": "uspto_50_fine_tuned_retrosynthesis_checkpoint_test",
122
+ "dataset_ref": "uspto_50k_chemformer_project_release",
123
+ "default": true,
124
+ "format": "jsonl",
125
+ "metadata_path": "benchmarks/metadata.json",
126
+ "native_input_format": "pandas_pickled_dataframe",
127
+ "native_source_paths": [
128
+ "benchmarks/native/uspto_50.pickle"
129
+ ],
130
+ "notes": [
131
+ "Native Chemformer benchmark source extracted from the public Box release pickle.",
132
+ "benchmarks/default_test.jsonl stores product/reactant SMILES converted from the pickled RDKit molecule objects in the test partition."
133
+ ],
134
+ "num_examples": 5004,
135
+ "path": "benchmarks/default_test.jsonl",
136
+ "split": "test"
137
+ }
138
+ ],
139
+ "checkpoint_id": "uspto_50_fine_tuned_retrosynthesis_checkpoint",
140
+ "embedding_supported": false,
141
+ "input_schema": null,
142
+ "model_id": "chemformer",
143
+ "normalized_checkpoint_id": "uspto_50_fine_tuned_retrosynthesis_checkpoint",
144
+ "package_dir": "checkpoints/chemformer/uspto_50_fine_tuned_retrosynthesis_checkpoint",
145
+ "source_registry_refs": [
146
+ {
147
+ "checkpoint_id": "uspto_50_fine_tuned_retrosynthesis_checkpoint",
148
+ "model_id": "chemformer"
149
+ }
150
+ ],
151
+ "task_direction": null,
152
+ "training_dataset_refs": [
153
+ "uspto_50k_chemformer_project_release"
154
+ ],
155
+ "weights_path": "model.safetensors",
156
+ "wrapper_family": "chemformer_bart"
157
+ },
158
+ {
159
+ "auxiliary_files": [
160
+ "aux/README.md",
161
+ "aux/checkpoint_metadata.json",
162
+ "aux/configs/tokenizer.json",
163
+ "aux/configs/uspto_50k/base.yaml",
164
+ "aux/configs/uspto_50k/stage2_ema.yaml",
165
+ "aux/configs/uspto_50k/stage2_frozen.yaml",
166
+ "aux/provenance/uspto-50k/raw_test.jsonl",
167
+ "aux/retrieval.faiss",
168
+ "aux/template_cache.pt",
169
+ "aux/template_cache_metadata.json",
170
+ "aux/template_library.jsonl",
171
+ "benchmarks/default_test.jsonl",
172
+ "benchmarks/metadata.json",
173
+ "benchmarks/native/raw_test.jsonl"
174
+ ],
175
+ "benchmark_datasets": [
176
+ {
177
+ "dataset_id": "uspto_50k_ema_deep_epoch059_step0_val_top1_acc0_3521_ckpt_test",
178
+ "dataset_ref": "conretrobert_uspto_50k_packaged_release",
179
+ "default": true,
180
+ "format": "jsonl",
181
+ "metadata_path": "benchmarks/metadata.json",
182
+ "native_input_format": "jsonl_reaction_rows",
183
+ "native_source_paths": [
184
+ "benchmarks/native/raw_test.jsonl"
185
+ ],
186
+ "notes": [
187
+ "Native benchmark split for the packaged ConRetroBert USPTO-50K release.",
188
+ "benchmarks/default_test.jsonl preserves the native atom-mapped product strings for model inputs, while expected reactants are stored as atom-map-stripped canonical SMILES and the original mapped strings remain in per-record metadata."
189
+ ],
190
+ "num_examples": 5005,
191
+ "path": "benchmarks/default_test.jsonl",
192
+ "split": "test"
193
+ }
194
+ ],
195
+ "checkpoint_id": "uspto_50k_ema_deep_epoch059_step0_val_top1_acc0_3521_ckpt",
196
+ "embedding_supported": false,
197
+ "input_schema": null,
198
+ "model_id": "conretrobert",
199
+ "normalized_checkpoint_id": "uspto_50k_ema_deep_epoch059_step0_val_top1_acc0_3521_ckpt",
200
+ "package_dir": "checkpoints/conretrobert/uspto_50k_ema_deep_epoch059_step0_val_top1_acc0_3521_ckpt",
201
+ "source_registry_refs": [
202
+ {
203
+ "checkpoint_id": "uspto_50k_ema_deep_epoch059_step0_val_top1_acc0_3521_ckpt",
204
+ "model_id": "conretrobert"
205
+ }
206
+ ],
207
+ "task_direction": null,
208
+ "training_dataset_refs": [
209
+ "conretrobert_uspto_50k_packaged_release"
210
+ ],
211
+ "weights_path": "model.safetensors",
212
+ "wrapper_family": "conretrobert_template_ranker"
213
+ },
214
+ {
215
+ "auxiliary_files": [
216
+ "aux/README.md",
217
+ "aux/checkpoint_metadata.json",
218
+ "aux/configs/tokenizer.json",
219
+ "aux/configs/uspto_50k/base.yaml",
220
+ "aux/configs/uspto_50k/stage2_ema.yaml",
221
+ "aux/configs/uspto_50k/stage2_frozen.yaml",
222
+ "aux/provenance/uspto-50k/raw_test.jsonl",
223
+ "aux/retrieval.faiss",
224
+ "aux/template_cache.pt",
225
+ "aux/template_cache_metadata.json",
226
+ "aux/template_library.jsonl",
227
+ "benchmarks/default_test.jsonl",
228
+ "benchmarks/metadata.json",
229
+ "benchmarks/native/raw_test.jsonl"
230
+ ],
231
+ "benchmark_datasets": [
232
+ {
233
+ "dataset_id": "uspto_50k_ema_mid_epoch087_step0_val_top1_acc0_3547_val_loss_4_5051_ckpt_test",
234
+ "dataset_ref": "conretrobert_uspto_50k_packaged_release",
235
+ "default": true,
236
+ "format": "jsonl",
237
+ "metadata_path": "benchmarks/metadata.json",
238
+ "native_input_format": "jsonl_reaction_rows",
239
+ "native_source_paths": [
240
+ "benchmarks/native/raw_test.jsonl"
241
+ ],
242
+ "notes": [
243
+ "Native benchmark split for the packaged ConRetroBert USPTO-50K release.",
244
+ "benchmarks/default_test.jsonl preserves the native atom-mapped product strings for model inputs, while expected reactants are stored as atom-map-stripped canonical SMILES and the original mapped strings remain in per-record metadata."
245
+ ],
246
+ "num_examples": 5005,
247
+ "path": "benchmarks/default_test.jsonl",
248
+ "split": "test"
249
+ }
250
+ ],
251
+ "checkpoint_id": "uspto_50k_ema_mid_epoch087_step0_val_top1_acc0_3547_val_loss_4_5051_ckpt",
252
+ "embedding_supported": false,
253
+ "input_schema": null,
254
+ "model_id": "conretrobert",
255
+ "normalized_checkpoint_id": "uspto_50k_ema_mid_epoch087_step0_val_top1_acc0_3547_val_loss_4_5051_ckpt",
256
+ "package_dir": "checkpoints/conretrobert/uspto_50k_ema_mid_epoch087_step0_val_top1_acc0_3547_val_loss_4_5051_ckpt",
257
+ "source_registry_refs": [
258
+ {
259
+ "checkpoint_id": "uspto_50k_ema_mid_epoch087_step0_val_top1_acc0_3547_val_loss_4_5051_ckpt",
260
+ "model_id": "conretrobert"
261
+ }
262
+ ],
263
+ "task_direction": null,
264
+ "training_dataset_refs": [
265
+ "conretrobert_uspto_50k_packaged_release"
266
+ ],
267
+ "weights_path": "model.safetensors",
268
+ "wrapper_family": "conretrobert_template_ranker"
269
+ },
270
+ {
271
+ "auxiliary_files": [
272
+ "aux/README.md",
273
+ "aux/checkpoint_metadata.json",
274
+ "aux/configs/tokenizer.json",
275
+ "aux/configs/uspto_50k/base.yaml",
276
+ "aux/configs/uspto_50k/stage2_ema.yaml",
277
+ "aux/configs/uspto_50k/stage2_frozen.yaml",
278
+ "aux/provenance/uspto-50k/raw_test.jsonl",
279
+ "aux/retrieval.faiss",
280
+ "aux/template_cache.pt",
281
+ "aux/template_cache_metadata.json",
282
+ "aux/template_library.jsonl",
283
+ "benchmarks/default_test.jsonl",
284
+ "benchmarks/metadata.json",
285
+ "benchmarks/native/raw_test.jsonl"
286
+ ],
287
+ "benchmark_datasets": [
288
+ {
289
+ "dataset_id": "uspto_50k_frozen_te_epoch087_step54714_val_top1_acc0_3711_ckpt_test",
290
+ "dataset_ref": "conretrobert_uspto_50k_packaged_release",
291
+ "default": true,
292
+ "format": "jsonl",
293
+ "metadata_path": "benchmarks/metadata.json",
294
+ "native_input_format": "jsonl_reaction_rows",
295
+ "native_source_paths": [
296
+ "benchmarks/native/raw_test.jsonl"
297
+ ],
298
+ "notes": [
299
+ "Native benchmark split for the packaged ConRetroBert USPTO-50K release.",
300
+ "benchmarks/default_test.jsonl preserves the native atom-mapped product strings for model inputs, while expected reactants are stored as atom-map-stripped canonical SMILES and the original mapped strings remain in per-record metadata."
301
+ ],
302
+ "num_examples": 5005,
303
+ "path": "benchmarks/default_test.jsonl",
304
+ "split": "test"
305
+ }
306
+ ],
307
+ "checkpoint_id": "uspto_50k_frozen_te_epoch087_step54714_val_top1_acc0_3711_ckpt",
308
+ "embedding_supported": false,
309
+ "input_schema": null,
310
+ "model_id": "conretrobert",
311
+ "normalized_checkpoint_id": "uspto_50k_frozen_te_epoch087_step54714_val_top1_acc0_3711_ckpt",
312
+ "package_dir": "checkpoints/conretrobert/uspto_50k_frozen_te_epoch087_step54714_val_top1_acc0_3711_ckpt",
313
+ "source_registry_refs": [
314
+ {
315
+ "checkpoint_id": "uspto_50k_frozen_te_epoch087_step54714_val_top1_acc0_3711_ckpt",
316
+ "model_id": "conretrobert"
317
+ }
318
+ ],
319
+ "task_direction": null,
320
+ "training_dataset_refs": [
321
+ "conretrobert_uspto_50k_packaged_release"
322
+ ],
323
+ "weights_path": "model.safetensors",
324
+ "wrapper_family": "conretrobert_template_ranker"
325
+ },
326
+ {
327
+ "auxiliary_files": [
328
+ "aux/README.md",
329
+ "aux/checkpoint_metadata.json",
330
+ "aux/configs/tokenizer.json",
331
+ "aux/configs/uspto_50k/base.yaml",
332
+ "aux/configs/uspto_50k/stage2_ema.yaml",
333
+ "aux/configs/uspto_50k/stage2_frozen.yaml",
334
+ "aux/provenance/uspto-50k/raw_test.jsonl",
335
+ "aux/retrieval.faiss",
336
+ "aux/template_cache.pt",
337
+ "aux/template_cache_metadata.json",
338
+ "aux/template_library.jsonl",
339
+ "benchmarks/default_test.jsonl",
340
+ "benchmarks/metadata.json",
341
+ "benchmarks/native/raw_test.jsonl"
342
+ ],
343
+ "benchmark_datasets": [
344
+ {
345
+ "benchmark_semantics": "uspto_50k_transfer_compatibility_for_uspto_full_checkpoint",
346
+ "benchmark_variant": "transfer_compatibility",
347
+ "dataset_id": "uspto_full_2_uspto_50k_magic_epoch012_step0_weighted_val_accuracy0_8390_ckpt_uspto_50k_transfer_test",
348
+ "dataset_ref": "conretrobert_uspto_50k_packaged_release",
349
+ "default": true,
350
+ "format": "jsonl",
351
+ "metadata_path": "benchmarks/metadata.json",
352
+ "native_input_format": "jsonl_reaction_rows",
353
+ "native_source_paths": [
354
+ "benchmarks/native/raw_test.jsonl"
355
+ ],
356
+ "notes": [
357
+ "This checkpoint name indicates USPTO-Full training, but the public ConRetroBert release staged locally does not package a native USPTO-Full benchmark split or full template library. The default bundle benchmark therefore uses the packaged USPTO-50K raw_test.jsonl split as a transfer/compatibility evaluation only.",
358
+ "Native benchmark split for the packaged ConRetroBert USPTO-50K release.",
359
+ "benchmarks/default_test.jsonl preserves the native atom-mapped product strings for model inputs, while expected reactants are stored as atom-map-stripped canonical SMILES and the original mapped strings remain in per-record metadata."
360
+ ],
361
+ "num_examples": 5005,
362
+ "path": "benchmarks/default_test.jsonl",
363
+ "split": "test"
364
+ }
365
+ ],
366
+ "checkpoint_id": "uspto_full_2_uspto_50k_magic_epoch012_step0_weighted_val_accuracy0_8390_ckpt",
367
+ "embedding_supported": false,
368
+ "input_schema": null,
369
+ "model_id": "conretrobert",
370
+ "normalized_checkpoint_id": "uspto_full_2_uspto_50k_magic_epoch012_step0_weighted_val_accuracy0_8390_ckpt",
371
+ "package_dir": "checkpoints/conretrobert/uspto_full_2_uspto_50k_magic_epoch012_step0_weighted_val_accuracy0_8390_ckpt",
372
+ "source_registry_refs": [
373
+ {
374
+ "checkpoint_id": "uspto_full_2_uspto_50k_magic_epoch012_step0_weighted_val_accuracy0_8390_ckpt",
375
+ "model_id": "conretrobert"
376
+ }
377
+ ],
378
+ "task_direction": null,
379
+ "training_dataset_refs": [
380
+ "conretrobert_uspto_50k_packaged_release"
381
+ ],
382
+ "weights_path": "model.safetensors",
383
+ "wrapper_family": "conretrobert_template_ranker"
384
+ },
385
+ {
386
+ "auxiliary_files": [
387
+ "aux/README.md",
388
+ "aux/checkpoint_metadata.json",
389
+ "aux/configs/tokenizer.json",
390
+ "aux/configs/uspto_50k/base.yaml",
391
+ "aux/configs/uspto_50k/stage2_ema.yaml",
392
+ "aux/configs/uspto_50k/stage2_frozen.yaml",
393
+ "aux/provenance/uspto-50k/raw_test.jsonl",
394
+ "aux/retrieval.faiss",
395
+ "aux/template_cache.pt",
396
+ "aux/template_cache_metadata.json",
397
+ "aux/template_library.jsonl",
398
+ "benchmarks/default_test.jsonl",
399
+ "benchmarks/metadata.json",
400
+ "benchmarks/native/raw_test.jsonl"
401
+ ],
402
+ "benchmark_datasets": [
403
+ {
404
+ "benchmark_semantics": "uspto_50k_transfer_compatibility_for_uspto_full_checkpoint",
405
+ "benchmark_variant": "transfer_compatibility",
406
+ "dataset_id": "uspto_full_frozen_te_epoch096_step129890_val_top1_acc0_6477_ckpt_uspto_50k_transfer_test",
407
+ "dataset_ref": "conretrobert_uspto_50k_packaged_release",
408
+ "default": true,
409
+ "format": "jsonl",
410
+ "metadata_path": "benchmarks/metadata.json",
411
+ "native_input_format": "jsonl_reaction_rows",
412
+ "native_source_paths": [
413
+ "benchmarks/native/raw_test.jsonl"
414
+ ],
415
+ "notes": [
416
+ "This checkpoint name indicates USPTO-Full training, but the public ConRetroBert release staged locally does not package a native USPTO-Full benchmark split or full template library. The default bundle benchmark therefore uses the packaged USPTO-50K raw_test.jsonl split as a transfer/compatibility evaluation only.",
417
+ "Native benchmark split for the packaged ConRetroBert USPTO-50K release.",
418
+ "benchmarks/default_test.jsonl preserves the native atom-mapped product strings for model inputs, while expected reactants are stored as atom-map-stripped canonical SMILES and the original mapped strings remain in per-record metadata."
419
+ ],
420
+ "num_examples": 5005,
421
+ "path": "benchmarks/default_test.jsonl",
422
+ "split": "test"
423
+ }
424
+ ],
425
+ "checkpoint_id": "uspto_full_frozen_te_epoch096_step129890_val_top1_acc0_6477_ckpt",
426
+ "embedding_supported": false,
427
+ "input_schema": null,
428
+ "model_id": "conretrobert",
429
+ "normalized_checkpoint_id": "uspto_full_frozen_te_epoch096_step129890_val_top1_acc0_6477_ckpt",
430
+ "package_dir": "checkpoints/conretrobert/uspto_full_frozen_te_epoch096_step129890_val_top1_acc0_6477_ckpt",
431
+ "source_registry_refs": [
432
+ {
433
+ "checkpoint_id": "uspto_full_frozen_te_epoch096_step129890_val_top1_acc0_6477_ckpt",
434
+ "model_id": "conretrobert"
435
+ }
436
+ ],
437
+ "task_direction": null,
438
+ "training_dataset_refs": [],
439
+ "weights_path": "model.safetensors",
440
+ "wrapper_family": "conretrobert_template_ranker"
441
+ },
442
+ {
443
+ "auxiliary_files": [
444
+ "aux/README.md",
445
+ "aux/checkpoint_metadata.json",
446
+ "aux/configs/tokenizer.json",
447
+ "aux/configs/uspto_50k/base.yaml",
448
+ "aux/configs/uspto_50k/stage2_ema.yaml",
449
+ "aux/configs/uspto_50k/stage2_frozen.yaml",
450
+ "aux/provenance/uspto-50k/raw_test.jsonl",
451
+ "aux/retrieval.faiss",
452
+ "aux/template_cache.pt",
453
+ "aux/template_cache_metadata.json",
454
+ "aux/template_library.jsonl",
455
+ "benchmarks/default_test.jsonl",
456
+ "benchmarks/metadata.json",
457
+ "benchmarks/native/raw_test.jsonl"
458
+ ],
459
+ "benchmark_datasets": [
460
+ {
461
+ "benchmark_semantics": "uspto_50k_transfer_compatibility_for_uspto_full_checkpoint",
462
+ "benchmark_variant": "transfer_compatibility",
463
+ "dataset_id": "uspto_full_transfer_uspto_50k_ft_epoch046_step29212_val_top1_acc0_7501_val_loss_1_3350_ckpt_uspto_50k_transfer_test",
464
+ "dataset_ref": "conretrobert_uspto_50k_packaged_release",
465
+ "default": true,
466
+ "format": "jsonl",
467
+ "metadata_path": "benchmarks/metadata.json",
468
+ "native_input_format": "jsonl_reaction_rows",
469
+ "native_source_paths": [
470
+ "benchmarks/native/raw_test.jsonl"
471
+ ],
472
+ "notes": [
473
+ "This checkpoint name indicates USPTO-Full training, but the public ConRetroBert release staged locally does not package a native USPTO-Full benchmark split or full template library. The default bundle benchmark therefore uses the packaged USPTO-50K raw_test.jsonl split as a transfer/compatibility evaluation only.",
474
+ "Native benchmark split for the packaged ConRetroBert USPTO-50K release.",
475
+ "benchmarks/default_test.jsonl preserves the native atom-mapped product strings for model inputs, while expected reactants are stored as atom-map-stripped canonical SMILES and the original mapped strings remain in per-record metadata."
476
+ ],
477
+ "num_examples": 5005,
478
+ "path": "benchmarks/default_test.jsonl",
479
+ "split": "test"
480
+ }
481
+ ],
482
+ "checkpoint_id": "uspto_full_transfer_uspto_50k_ft_epoch046_step29212_val_top1_acc0_7501_val_loss_1_3350_ckpt",
483
+ "embedding_supported": false,
484
+ "input_schema": null,
485
+ "model_id": "conretrobert",
486
+ "normalized_checkpoint_id": "uspto_full_transfer_uspto_50k_ft_epoch046_step29212_val_top1_acc0_7501_val_loss_1_3350_ckpt",
487
+ "package_dir": "checkpoints/conretrobert/uspto_full_transfer_uspto_50k_ft_epoch046_step29212_val_top1_acc0_7501_val_loss_1_3350_ckpt",
488
+ "source_registry_refs": [
489
+ {
490
+ "checkpoint_id": "uspto_full_transfer_uspto_50k_ft_epoch046_step29212_val_top1_acc0_7501_val_loss_1_3350_ckpt",
491
+ "model_id": "conretrobert"
492
+ }
493
+ ],
494
+ "task_direction": null,
495
+ "training_dataset_refs": [
496
+ "conretrobert_uspto_50k_packaged_release"
497
+ ],
498
+ "weights_path": "model.safetensors",
499
+ "wrapper_family": "conretrobert_template_ranker"
500
+ },
501
+ {
502
+ "auxiliary_files": [
503
+ "aux/.gitattributes",
504
+ "aux/README.md",
505
+ "aux/UPSTREAM_URL.txt",
506
+ "aux/config.json",
507
+ "aux/generation_config.json",
508
+ "aux/hf_model_info.json",
509
+ "aux/provenance/README.txt",
510
+ "aux/provenance/ReactionT5v2_source_README.md",
511
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv",
512
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
513
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
514
+ "aux/provenance/datasets/USPTO_50k/test.csv",
515
+ "aux/provenance/datasets/USPTO_50k/train.csv",
516
+ "aux/provenance/datasets/USPTO_50k/val.csv",
517
+ "aux/provenance/reactiont5v2_split_provenance.json",
518
+ "aux/special_tokens_map.json",
519
+ "aux/tokenizer.json",
520
+ "aux/tokenizer_config.json",
521
+ "aux/trainer_state.json",
522
+ "aux/training_args.bin",
523
+ "benchmarks/default_test.jsonl",
524
+ "benchmarks/metadata.json"
525
+ ],
526
+ "benchmark_datasets": [
527
+ {
528
+ "dataset_id": "sagawa_reactiont5v2_retrosynthesis_test",
529
+ "dataset_ref": "ord_retrosynthesis_pretraining",
530
+ "default": true,
531
+ "format": "jsonl",
532
+ "metadata_path": "benchmarks/metadata.json",
533
+ "native_input_format": "csv_reaction_rows",
534
+ "native_source_paths": [
535
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv"
536
+ ],
537
+ "notes": [
538
+ "Native benchmark split for the public ORD-trained retrosynthesis checkpoint.",
539
+ "USPTO_50k/test.csv was used upstream for duplicate-removal / leakage control, not as the native training benchmark for this checkpoint."
540
+ ],
541
+ "num_examples": 225282,
542
+ "path": "benchmarks/default_test.jsonl",
543
+ "split": "test"
544
+ }
545
+ ],
546
+ "checkpoint_id": "sagawa_reactiont5v2_retrosynthesis",
547
+ "embedding_supported": true,
548
+ "input_schema": null,
549
+ "model_id": "reactiont5v2",
550
+ "normalized_checkpoint_id": "sagawa_reactiont5v2_retrosynthesis",
551
+ "package_dir": "checkpoints/reactiont5v2/sagawa_reactiont5v2_retrosynthesis",
552
+ "source_registry_refs": [
553
+ {
554
+ "checkpoint_id": "sagawa_reactiont5v2_retrosynthesis",
555
+ "model_id": "reactiont5v2"
556
+ }
557
+ ],
558
+ "task_direction": null,
559
+ "training_dataset_refs": [
560
+ "ord_retrosynthesis_pretraining"
561
+ ],
562
+ "weights_path": "model.safetensors",
563
+ "wrapper_family": "seq2seq_t5"
564
+ },
565
+ {
566
+ "auxiliary_files": [
567
+ "aux/.gitattributes",
568
+ "aux/README.md",
569
+ "aux/UPSTREAM_URL.txt",
570
+ "aux/config.json",
571
+ "aux/generation_config.json",
572
+ "aux/hf_model_info.json",
573
+ "aux/provenance/README.txt",
574
+ "aux/provenance/ReactionT5v2_source_README.md",
575
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv",
576
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
577
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
578
+ "aux/provenance/datasets/USPTO_50k/test.csv",
579
+ "aux/provenance/datasets/USPTO_50k/train.csv",
580
+ "aux/provenance/datasets/USPTO_50k/val.csv",
581
+ "aux/provenance/reactiont5v2_split_provenance.json",
582
+ "aux/special_tokens_map.json",
583
+ "aux/tokenizer.json",
584
+ "aux/tokenizer_config.json",
585
+ "aux/trainer_state.json",
586
+ "aux/training_args.bin",
587
+ "benchmarks/default_test.jsonl",
588
+ "benchmarks/metadata.json"
589
+ ],
590
+ "benchmark_datasets": [
591
+ {
592
+ "dataset_id": "sagawa_reactiont5v2_retrosynthesis_uspto_50k_test",
593
+ "dataset_ref": "uspto_50k_project_release",
594
+ "default": true,
595
+ "format": "jsonl",
596
+ "metadata_path": "benchmarks/metadata.json",
597
+ "native_input_format": "csv_reaction_rows",
598
+ "native_source_paths": [
599
+ "aux/provenance/datasets/USPTO_50k/test.csv"
600
+ ],
601
+ "notes": [
602
+ "Native benchmark split for the public USPTO-50K fine-tuned retrosynthesis checkpoint."
603
+ ],
604
+ "num_examples": 5004,
605
+ "path": "benchmarks/default_test.jsonl",
606
+ "split": "test"
607
+ }
608
+ ],
609
+ "checkpoint_id": "sagawa_reactiont5v2_retrosynthesis_uspto_50k",
610
+ "embedding_supported": true,
611
+ "input_schema": null,
612
+ "model_id": "reactiont5v2",
613
+ "normalized_checkpoint_id": "sagawa_reactiont5v2_retrosynthesis_uspto_50k",
614
+ "package_dir": "checkpoints/reactiont5v2/sagawa_reactiont5v2_retrosynthesis_uspto_50k",
615
+ "source_registry_refs": [
616
+ {
617
+ "checkpoint_id": "sagawa_reactiont5v2_retrosynthesis_uspto_50k",
618
+ "model_id": "reactiont5v2"
619
+ }
620
+ ],
621
+ "task_direction": null,
622
+ "training_dataset_refs": [
623
+ "uspto_50k_project_release"
624
+ ],
625
+ "weights_path": "model.safetensors",
626
+ "wrapper_family": "seq2seq_t5"
627
+ },
628
+ {
629
+ "auxiliary_files": [
630
+ "aux/models.json",
631
+ "benchmarks/default_test.jsonl",
632
+ "benchmarks/metadata.json",
633
+ "smiles_transformer/aux/checkpoint_metadata.json",
634
+ "smiles_transformer/aux/hyper_parameters.json",
635
+ "smiles_transformer/aux/vocab.txt",
636
+ "template_localization/aux/checkpoint_metadata.json",
637
+ "template_localization/aux/hyper_parameters.json",
638
+ "template_localization/aux/template_lib.json"
639
+ ],
640
+ "benchmark_datasets": [
641
+ {
642
+ "benchmark_variant": "benchmark_pending",
643
+ "dataset_id": "pistachio_test",
644
+ "dataset_ref": "pistachio_retrochimera_release",
645
+ "default": true,
646
+ "format": "jsonl",
647
+ "metadata_path": "benchmarks/metadata.json",
648
+ "native_input_format": "benchmark_not_available",
649
+ "native_source_paths": [],
650
+ "notes": [
651
+ "Benchmark companion for the end-to-end RetroChimera hybrid checkpoint on Pistachio.",
652
+ "These RetroChimera packages bundle both released ensemble components together and are intended for the upstream combined inference API.",
653
+ "No public benchmark split is bundled yet for this checkpoint; benchmarks/default_test.jsonl is intentionally empty for now.",
654
+ "The public RetroChimera checkpoint release documents Pistachio-trained checkpoints, but this repository does not yet bundle a public Pistachio evaluation split."
655
+ ],
656
+ "num_examples": 0,
657
+ "path": "benchmarks/default_test.jsonl",
658
+ "split": "test"
659
+ }
660
+ ],
661
+ "checkpoint_id": "pistachio",
662
+ "embedding_supported": false,
663
+ "input_schema": null,
664
+ "model_id": "retrochimera_hybrid",
665
+ "normalized_checkpoint_id": "pistachio",
666
+ "package_dir": "checkpoints/retrochimera_hybrid/pistachio",
667
+ "source_registry_refs": [
668
+ {
669
+ "checkpoint_id": "pistachio",
670
+ "model_id": "retrochimera_hybrid"
671
+ }
672
+ ],
673
+ "task_direction": null,
674
+ "training_dataset_refs": [
675
+ "pistachio_retrochimera_release"
676
+ ],
677
+ "weights_path": "template_localization/model.safetensors",
678
+ "wrapper_family": "retrochimera_hybrid"
679
+ },
680
+ {
681
+ "auxiliary_files": [
682
+ "aux/models.json",
683
+ "benchmarks/default_test.jsonl",
684
+ "benchmarks/metadata.json",
685
+ "benchmarks/native/src-test.txt",
686
+ "benchmarks/native/tgt-test.txt",
687
+ "smiles_transformer/aux/checkpoint_metadata.json",
688
+ "smiles_transformer/aux/hyper_parameters.json",
689
+ "smiles_transformer/aux/vocab.txt",
690
+ "template_localization/aux/checkpoint_metadata.json",
691
+ "template_localization/aux/hyper_parameters.json",
692
+ "template_localization/aux/template_lib.json"
693
+ ],
694
+ "benchmark_datasets": [
695
+ {
696
+ "benchmark_variant": "retrochimera_hybrid_gln_compatibility_view",
697
+ "dataset_id": "uspto_50k_test",
698
+ "dataset_ref": "uspto_50k_gln",
699
+ "default": true,
700
+ "format": "jsonl",
701
+ "metadata_path": "benchmarks/metadata.json",
702
+ "native_input_format": "tokenized_parallel_txt",
703
+ "native_source_paths": [
704
+ "benchmarks/native/src-test.txt",
705
+ "benchmarks/native/tgt-test.txt"
706
+ ],
707
+ "notes": [
708
+ "Benchmark companion for the end-to-end RetroChimera hybrid checkpoint on USPTO-50K via GLN/Coley split compatibility view.",
709
+ "These RetroChimera packages bundle both released ensemble components together and are intended for the upstream combined inference API.",
710
+ "The underlying test split is preserved from the local R-SMILES/GLN-compatible tokenized benchmark files already tracked by this repository.",
711
+ "The source files preserve rooted augmentation with factor 20; benchmarks/default_test.jsonl collapses each contiguous augmentation block back to one benchmark case per underlying reaction.",
712
+ "The native source files are tokenized parallel text; benchmarks/default_test.jsonl stores the corresponding de-tokenized product/reactant pairs.",
713
+ "RetroChimera cites Dai et al./GLN preprocessing for USPTO-50K; GLN in turn follows the Coley/RetroSim 80/10/10 split family."
714
+ ],
715
+ "num_examples": 5007,
716
+ "path": "benchmarks/default_test.jsonl",
717
+ "split": "test"
718
+ }
719
+ ],
720
+ "checkpoint_id": "uspto_50k",
721
+ "embedding_supported": false,
722
+ "input_schema": null,
723
+ "model_id": "retrochimera_hybrid",
724
+ "normalized_checkpoint_id": "uspto_50k",
725
+ "package_dir": "checkpoints/retrochimera_hybrid/uspto_50k",
726
+ "source_registry_refs": [
727
+ {
728
+ "checkpoint_id": "uspto_50k",
729
+ "model_id": "retrochimera_hybrid"
730
+ }
731
+ ],
732
+ "task_direction": null,
733
+ "training_dataset_refs": [
734
+ "uspto_50k_gln"
735
+ ],
736
+ "weights_path": "template_localization/model.safetensors",
737
+ "wrapper_family": "retrochimera_hybrid"
738
+ },
739
+ {
740
+ "auxiliary_files": [
741
+ "aux/models.json",
742
+ "benchmarks/default_test.jsonl",
743
+ "benchmarks/metadata.json",
744
+ "benchmarks/native/src-test.txt",
745
+ "benchmarks/native/tgt-test.txt",
746
+ "smiles_transformer/aux/checkpoint_metadata.json",
747
+ "smiles_transformer/aux/hyper_parameters.json",
748
+ "smiles_transformer/aux/vocab.txt",
749
+ "template_localization/aux/checkpoint_metadata.json",
750
+ "template_localization/aux/hyper_parameters.json",
751
+ "template_localization/aux/template_lib.json"
752
+ ],
753
+ "benchmark_datasets": [
754
+ {
755
+ "benchmark_variant": "retrochimera_hybrid_r_smiles_selected_full_view",
756
+ "dataset_id": "uspto_full_test",
757
+ "dataset_ref": "uspto_full_gln",
758
+ "default": true,
759
+ "format": "jsonl",
760
+ "metadata_path": "benchmarks/metadata.json",
761
+ "native_input_format": "tokenized_parallel_txt",
762
+ "native_source_paths": [
763
+ "benchmarks/native/src-test.txt",
764
+ "benchmarks/native/tgt-test.txt"
765
+ ],
766
+ "notes": [
767
+ "Benchmark companion for the end-to-end RetroChimera hybrid checkpoint on USPTO-FULL via RetroChimera\u2192R-SMILES\u2192GLN compatibility view.",
768
+ "These RetroChimera packages bundle both released ensemble components together and are intended for the upstream combined inference API.",
769
+ "The underlying test split is preserved from the local R-SMILES/GLN-compatible tokenized benchmark files already tracked by this repository.",
770
+ "The source files preserve rooted augmentation with factor 5; benchmarks/default_test.jsonl collapses each contiguous augmentation block back to one benchmark case per underlying reaction.",
771
+ "The native source files are tokenized parallel text; benchmarks/default_test.jsonl stores the corresponding de-tokenized product/reactant pairs.",
772
+ "RetroChimera notes that multiple USPTO-FULL variants exist and selects the R-SMILES/Zhong et al. version for fair comparison; the R-SMILES release points that data lineage back to GLN."
773
+ ],
774
+ "num_examples": 96023,
775
+ "path": "benchmarks/default_test.jsonl",
776
+ "split": "test"
777
+ }
778
+ ],
779
+ "checkpoint_id": "uspto_full",
780
+ "embedding_supported": false,
781
+ "input_schema": null,
782
+ "model_id": "retrochimera_hybrid",
783
+ "normalized_checkpoint_id": "uspto_full",
784
+ "package_dir": "checkpoints/retrochimera_hybrid/uspto_full",
785
+ "source_registry_refs": [
786
+ {
787
+ "checkpoint_id": "uspto_full",
788
+ "model_id": "retrochimera_hybrid"
789
+ }
790
+ ],
791
+ "task_direction": null,
792
+ "training_dataset_refs": [
793
+ "uspto_full_gln"
794
+ ],
795
+ "weights_path": "template_localization/model.safetensors",
796
+ "wrapper_family": "retrochimera_hybrid"
797
+ },
798
+ {
799
+ "auxiliary_files": [
800
+ "aux/checkpoint_metadata.json",
801
+ "aux/hyper_parameters.json",
802
+ "aux/models.json",
803
+ "aux/template_lib.json",
804
+ "benchmarks/default_test.jsonl",
805
+ "benchmarks/metadata.json"
806
+ ],
807
+ "benchmark_datasets": [
808
+ {
809
+ "benchmark_variant": "benchmark_pending",
810
+ "dataset_id": "pistachio_test",
811
+ "dataset_ref": "pistachio_retrochimera_release",
812
+ "default": true,
813
+ "format": "jsonl",
814
+ "metadata_path": "benchmarks/metadata.json",
815
+ "native_input_format": "benchmark_not_available",
816
+ "native_source_paths": [],
817
+ "notes": [
818
+ "Benchmark companion for RetroChimera component checkpoint on Pistachio.",
819
+ "These RetroChimera packages contain component checkpoints rather than the full end-to-end ensemble, so the bundled benchmark is best interpreted as a dataset compatibility view.",
820
+ "No public benchmark split is bundled yet for this checkpoint; benchmarks/default_test.jsonl is intentionally empty for now.",
821
+ "The public RetroChimera checkpoint release documents Pistachio-trained checkpoints, but this repository does not yet bundle a public Pistachio evaluation split."
822
+ ],
823
+ "num_examples": 0,
824
+ "path": "benchmarks/default_test.jsonl",
825
+ "split": "test"
826
+ }
827
+ ],
828
+ "checkpoint_id": "pistachio",
829
+ "embedding_supported": false,
830
+ "input_schema": null,
831
+ "model_id": "retrochimera_template",
832
+ "normalized_checkpoint_id": "pistachio",
833
+ "package_dir": "checkpoints/retrochimera_template/pistachio",
834
+ "source_registry_refs": [
835
+ {
836
+ "checkpoint_id": "pistachio",
837
+ "model_id": "retrochimera_template"
838
+ }
839
+ ],
840
+ "task_direction": null,
841
+ "training_dataset_refs": [
842
+ "pistachio_retrochimera_release"
843
+ ],
844
+ "weights_path": "model.safetensors",
845
+ "wrapper_family": "retrochimera_template_localization"
846
+ },
847
+ {
848
+ "auxiliary_files": [
849
+ "aux/checkpoint_metadata.json",
850
+ "aux/hyper_parameters.json",
851
+ "aux/models.json",
852
+ "aux/template_lib.json",
853
+ "benchmarks/default_test.jsonl",
854
+ "benchmarks/metadata.json",
855
+ "benchmarks/native/src-test.txt",
856
+ "benchmarks/native/tgt-test.txt"
857
+ ],
858
+ "benchmark_datasets": [
859
+ {
860
+ "benchmark_variant": "retrochimera_component_gln_compatibility_view",
861
+ "dataset_id": "uspto_50k_test",
862
+ "dataset_ref": "uspto_50k_gln",
863
+ "default": true,
864
+ "format": "jsonl",
865
+ "metadata_path": "benchmarks/metadata.json",
866
+ "native_input_format": "tokenized_parallel_txt",
867
+ "native_source_paths": [
868
+ "benchmarks/native/src-test.txt",
869
+ "benchmarks/native/tgt-test.txt"
870
+ ],
871
+ "notes": [
872
+ "Benchmark companion for RetroChimera component checkpoint on USPTO-50K via GLN/Coley split compatibility view.",
873
+ "These RetroChimera packages contain component checkpoints rather than the full end-to-end ensemble, so the bundled benchmark is best interpreted as a dataset compatibility view.",
874
+ "The underlying test split is preserved from the local R-SMILES/GLN-compatible tokenized benchmark files already tracked by this repository.",
875
+ "The source files preserve rooted augmentation with factor 20; benchmarks/default_test.jsonl collapses each contiguous augmentation block back to one benchmark case per underlying reaction.",
876
+ "The native source files are tokenized parallel text; benchmarks/default_test.jsonl stores the corresponding de-tokenized product/reactant pairs.",
877
+ "RetroChimera cites Dai et al./GLN preprocessing for USPTO-50K; GLN in turn follows the Coley/RetroSim 80/10/10 split family."
878
+ ],
879
+ "num_examples": 5007,
880
+ "path": "benchmarks/default_test.jsonl",
881
+ "split": "test"
882
+ }
883
+ ],
884
+ "checkpoint_id": "uspto_50k",
885
+ "embedding_supported": false,
886
+ "input_schema": null,
887
+ "model_id": "retrochimera_template",
888
+ "normalized_checkpoint_id": "uspto_50k",
889
+ "package_dir": "checkpoints/retrochimera_template/uspto_50k",
890
+ "source_registry_refs": [
891
+ {
892
+ "checkpoint_id": "uspto_50k",
893
+ "model_id": "retrochimera_template"
894
+ }
895
+ ],
896
+ "task_direction": null,
897
+ "training_dataset_refs": [
898
+ "uspto_50k_gln"
899
+ ],
900
+ "weights_path": "model.safetensors",
901
+ "wrapper_family": "retrochimera_template_localization"
902
+ },
903
+ {
904
+ "auxiliary_files": [
905
+ "aux/checkpoint_metadata.json",
906
+ "aux/hyper_parameters.json",
907
+ "aux/models.json",
908
+ "aux/template_lib.json",
909
+ "benchmarks/default_test.jsonl",
910
+ "benchmarks/metadata.json",
911
+ "benchmarks/native/src-test.txt",
912
+ "benchmarks/native/tgt-test.txt"
913
+ ],
914
+ "benchmark_datasets": [
915
+ {
916
+ "benchmark_variant": "retrochimera_component_r_smiles_selected_full_view",
917
+ "dataset_id": "uspto_full_test",
918
+ "dataset_ref": "uspto_full_gln",
919
+ "default": true,
920
+ "format": "jsonl",
921
+ "metadata_path": "benchmarks/metadata.json",
922
+ "native_input_format": "tokenized_parallel_txt",
923
+ "native_source_paths": [
924
+ "benchmarks/native/src-test.txt",
925
+ "benchmarks/native/tgt-test.txt"
926
+ ],
927
+ "notes": [
928
+ "Benchmark companion for RetroChimera component checkpoint on USPTO-FULL via RetroChimera\u2192R-SMILES\u2192GLN compatibility view.",
929
+ "These RetroChimera packages contain component checkpoints rather than the full end-to-end ensemble, so the bundled benchmark is best interpreted as a dataset compatibility view.",
930
+ "The underlying test split is preserved from the local R-SMILES/GLN-compatible tokenized benchmark files already tracked by this repository.",
931
+ "The source files preserve rooted augmentation with factor 5; benchmarks/default_test.jsonl collapses each contiguous augmentation block back to one benchmark case per underlying reaction.",
932
+ "The native source files are tokenized parallel text; benchmarks/default_test.jsonl stores the corresponding de-tokenized product/reactant pairs.",
933
+ "RetroChimera notes that multiple USPTO-FULL variants exist and selects the R-SMILES/Zhong et al. version for fair comparison; the R-SMILES release points that data lineage back to GLN."
934
+ ],
935
+ "num_examples": 96023,
936
+ "path": "benchmarks/default_test.jsonl",
937
+ "split": "test"
938
+ }
939
+ ],
940
+ "checkpoint_id": "uspto_full",
941
+ "embedding_supported": false,
942
+ "input_schema": null,
943
+ "model_id": "retrochimera_template",
944
+ "normalized_checkpoint_id": "uspto_full",
945
+ "package_dir": "checkpoints/retrochimera_template/uspto_full",
946
+ "source_registry_refs": [
947
+ {
948
+ "checkpoint_id": "uspto_full",
949
+ "model_id": "retrochimera_template"
950
+ }
951
+ ],
952
+ "task_direction": null,
953
+ "training_dataset_refs": [
954
+ "uspto_full_gln"
955
+ ],
956
+ "weights_path": "model.safetensors",
957
+ "wrapper_family": "retrochimera_template_localization"
958
+ },
959
+ {
960
+ "auxiliary_files": [
961
+ "aux/checkpoint_metadata.json",
962
+ "aux/hyper_parameters.json",
963
+ "aux/models.json",
964
+ "aux/vocab.txt",
965
+ "benchmarks/default_test.jsonl",
966
+ "benchmarks/metadata.json"
967
+ ],
968
+ "benchmark_datasets": [
969
+ {
970
+ "benchmark_variant": "benchmark_pending",
971
+ "dataset_id": "pistachio_test",
972
+ "dataset_ref": "pistachio_retrochimera_release",
973
+ "default": true,
974
+ "format": "jsonl",
975
+ "metadata_path": "benchmarks/metadata.json",
976
+ "native_input_format": "benchmark_not_available",
977
+ "native_source_paths": [],
978
+ "notes": [
979
+ "Benchmark companion for RetroChimera component checkpoint on Pistachio.",
980
+ "These RetroChimera packages contain component checkpoints rather than the full end-to-end ensemble, so the bundled benchmark is best interpreted as a dataset compatibility view.",
981
+ "No public benchmark split is bundled yet for this checkpoint; benchmarks/default_test.jsonl is intentionally empty for now.",
982
+ "The public RetroChimera checkpoint release documents Pistachio-trained checkpoints, but this repository does not yet bundle a public Pistachio evaluation split."
983
+ ],
984
+ "num_examples": 0,
985
+ "path": "benchmarks/default_test.jsonl",
986
+ "split": "test"
987
+ }
988
+ ],
989
+ "checkpoint_id": "pistachio",
990
+ "embedding_supported": false,
991
+ "input_schema": null,
992
+ "model_id": "retrochimera_transformer",
993
+ "normalized_checkpoint_id": "pistachio",
994
+ "package_dir": "checkpoints/retrochimera_transformer/pistachio",
995
+ "source_registry_refs": [
996
+ {
997
+ "checkpoint_id": "pistachio",
998
+ "model_id": "retrochimera_transformer"
999
+ }
1000
+ ],
1001
+ "task_direction": null,
1002
+ "training_dataset_refs": [
1003
+ "pistachio_retrochimera_release"
1004
+ ],
1005
+ "weights_path": "model.safetensors",
1006
+ "wrapper_family": "retrochimera_smiles_transformer"
1007
+ },
1008
+ {
1009
+ "auxiliary_files": [
1010
+ "aux/checkpoint_metadata.json",
1011
+ "aux/hyper_parameters.json",
1012
+ "aux/models.json",
1013
+ "aux/vocab.txt",
1014
+ "benchmarks/default_test.jsonl",
1015
+ "benchmarks/metadata.json",
1016
+ "benchmarks/native/src-test.txt",
1017
+ "benchmarks/native/tgt-test.txt"
1018
+ ],
1019
+ "benchmark_datasets": [
1020
+ {
1021
+ "benchmark_variant": "retrochimera_component_gln_compatibility_view",
1022
+ "dataset_id": "uspto_50k_test",
1023
+ "dataset_ref": "uspto_50k_gln",
1024
+ "default": true,
1025
+ "format": "jsonl",
1026
+ "metadata_path": "benchmarks/metadata.json",
1027
+ "native_input_format": "tokenized_parallel_txt",
1028
+ "native_source_paths": [
1029
+ "benchmarks/native/src-test.txt",
1030
+ "benchmarks/native/tgt-test.txt"
1031
+ ],
1032
+ "notes": [
1033
+ "Benchmark companion for RetroChimera component checkpoint on USPTO-50K via GLN/Coley split compatibility view.",
1034
+ "These RetroChimera packages contain component checkpoints rather than the full end-to-end ensemble, so the bundled benchmark is best interpreted as a dataset compatibility view.",
1035
+ "The underlying test split is preserved from the local R-SMILES/GLN-compatible tokenized benchmark files already tracked by this repository.",
1036
+ "The source files preserve rooted augmentation with factor 20; benchmarks/default_test.jsonl collapses each contiguous augmentation block back to one benchmark case per underlying reaction.",
1037
+ "The native source files are tokenized parallel text; benchmarks/default_test.jsonl stores the corresponding de-tokenized product/reactant pairs.",
1038
+ "RetroChimera cites Dai et al./GLN preprocessing for USPTO-50K; GLN in turn follows the Coley/RetroSim 80/10/10 split family."
1039
+ ],
1040
+ "num_examples": 5007,
1041
+ "path": "benchmarks/default_test.jsonl",
1042
+ "split": "test"
1043
+ }
1044
+ ],
1045
+ "checkpoint_id": "uspto_50k",
1046
+ "embedding_supported": false,
1047
+ "input_schema": null,
1048
+ "model_id": "retrochimera_transformer",
1049
+ "normalized_checkpoint_id": "uspto_50k",
1050
+ "package_dir": "checkpoints/retrochimera_transformer/uspto_50k",
1051
+ "source_registry_refs": [
1052
+ {
1053
+ "checkpoint_id": "uspto_50k",
1054
+ "model_id": "retrochimera_transformer"
1055
+ }
1056
+ ],
1057
+ "task_direction": null,
1058
+ "training_dataset_refs": [
1059
+ "uspto_50k_gln"
1060
+ ],
1061
+ "weights_path": "model.safetensors",
1062
+ "wrapper_family": "retrochimera_smiles_transformer"
1063
+ },
1064
+ {
1065
+ "auxiliary_files": [
1066
+ "aux/checkpoint_metadata.json",
1067
+ "aux/hyper_parameters.json",
1068
+ "aux/models.json",
1069
+ "aux/vocab.txt",
1070
+ "benchmarks/default_test.jsonl",
1071
+ "benchmarks/metadata.json",
1072
+ "benchmarks/native/src-test.txt",
1073
+ "benchmarks/native/tgt-test.txt"
1074
+ ],
1075
+ "benchmark_datasets": [
1076
+ {
1077
+ "benchmark_variant": "retrochimera_component_r_smiles_selected_full_view",
1078
+ "dataset_id": "uspto_full_test",
1079
+ "dataset_ref": "uspto_full_gln",
1080
+ "default": true,
1081
+ "format": "jsonl",
1082
+ "metadata_path": "benchmarks/metadata.json",
1083
+ "native_input_format": "tokenized_parallel_txt",
1084
+ "native_source_paths": [
1085
+ "benchmarks/native/src-test.txt",
1086
+ "benchmarks/native/tgt-test.txt"
1087
+ ],
1088
+ "notes": [
1089
+ "Benchmark companion for RetroChimera component checkpoint on USPTO-FULL via RetroChimera\u2192R-SMILES\u2192GLN compatibility view.",
1090
+ "These RetroChimera packages contain component checkpoints rather than the full end-to-end ensemble, so the bundled benchmark is best interpreted as a dataset compatibility view.",
1091
+ "The underlying test split is preserved from the local R-SMILES/GLN-compatible tokenized benchmark files already tracked by this repository.",
1092
+ "The source files preserve rooted augmentation with factor 5; benchmarks/default_test.jsonl collapses each contiguous augmentation block back to one benchmark case per underlying reaction.",
1093
+ "The native source files are tokenized parallel text; benchmarks/default_test.jsonl stores the corresponding de-tokenized product/reactant pairs.",
1094
+ "RetroChimera notes that multiple USPTO-FULL variants exist and selects the R-SMILES/Zhong et al. version for fair comparison; the R-SMILES release points that data lineage back to GLN."
1095
+ ],
1096
+ "num_examples": 96023,
1097
+ "path": "benchmarks/default_test.jsonl",
1098
+ "split": "test"
1099
+ }
1100
+ ],
1101
+ "checkpoint_id": "uspto_full",
1102
+ "embedding_supported": false,
1103
+ "input_schema": null,
1104
+ "model_id": "retrochimera_transformer",
1105
+ "normalized_checkpoint_id": "uspto_full",
1106
+ "package_dir": "checkpoints/retrochimera_transformer/uspto_full",
1107
+ "source_registry_refs": [
1108
+ {
1109
+ "checkpoint_id": "uspto_full",
1110
+ "model_id": "retrochimera_transformer"
1111
+ }
1112
+ ],
1113
+ "task_direction": null,
1114
+ "training_dataset_refs": [
1115
+ "uspto_full_gln"
1116
+ ],
1117
+ "weights_path": "model.safetensors",
1118
+ "wrapper_family": "retrochimera_smiles_transformer"
1119
+ },
1120
+ {
1121
+ "auxiliary_files": [
1122
+ "aux/configs/base.yml",
1123
+ "aux/configs/rxngpt_llama1B.json",
1124
+ "aux/deepspeed.yaml",
1125
+ "aux/tokenizer/special_tokens_map.json",
1126
+ "aux/tokenizer/tokenizer.json",
1127
+ "aux/tokenizer/tokenizer_config.json",
1128
+ "aux/vocab.json",
1129
+ "benchmarks/default_test.jsonl",
1130
+ "benchmarks/metadata.json",
1131
+ "benchmarks/native/USPTO-50k_test_tidy.txt"
1132
+ ],
1133
+ "benchmark_datasets": [
1134
+ {
1135
+ "benchmark_variant": "class_unknown_test_aug_best_effort_underlying_test_examples",
1136
+ "dataset_id": "finetune_50k_aug20_pth_test",
1137
+ "dataset_ref": "uspto_50k_rsgpt",
1138
+ "default": true,
1139
+ "format": "jsonl",
1140
+ "metadata_path": "benchmarks/metadata.json",
1141
+ "native_input_format": "reaction_smiles_txt",
1142
+ "native_source_paths": [
1143
+ "benchmarks/native/USPTO-50k_test_tidy.txt"
1144
+ ],
1145
+ "notes": [
1146
+ "The exact upstream test-time augmentation prefix file test_50k_prefixs20.txt is not present in the local manual artifacts or pinned source snapshot.",
1147
+ "This bundle therefore preserves the underlying standard USPTO-50K test reactions as the normalized benchmark truth set, while the checkpoint metadata still records the intended aug20 evaluation variant."
1148
+ ],
1149
+ "num_examples": 5007,
1150
+ "path": "benchmarks/default_test.jsonl",
1151
+ "split": "test"
1152
+ }
1153
+ ],
1154
+ "checkpoint_id": "finetune_50k_aug20_pth",
1155
+ "embedding_supported": false,
1156
+ "input_schema": null,
1157
+ "model_id": "rsgpt",
1158
+ "normalized_checkpoint_id": "finetune_50k_aug20_pth",
1159
+ "package_dir": "checkpoints/rsgpt/finetune_50k_aug20_pth",
1160
+ "source_registry_refs": [
1161
+ {
1162
+ "checkpoint_id": "finetune_50k_aug20_pth",
1163
+ "model_id": "rsgpt"
1164
+ }
1165
+ ],
1166
+ "task_direction": null,
1167
+ "training_dataset_refs": [
1168
+ "uspto_50k_rsgpt",
1169
+ "synthetic_10b_rsgpt"
1170
+ ],
1171
+ "weights_path": "model.safetensors",
1172
+ "wrapper_family": "rsgpt_llama_decoder"
1173
+ },
1174
+ {
1175
+ "auxiliary_files": [
1176
+ "aux/configs/base.yml",
1177
+ "aux/configs/rxngpt_llama1B.json",
1178
+ "aux/deepspeed.yaml",
1179
+ "aux/tokenizer/special_tokens_map.json",
1180
+ "aux/tokenizer/tokenizer.json",
1181
+ "aux/tokenizer/tokenizer_config.json",
1182
+ "aux/vocab.json",
1183
+ "benchmarks/default_test.jsonl",
1184
+ "benchmarks/metadata.json",
1185
+ "benchmarks/native/USPTO-50k_test_tidy_class1.csv"
1186
+ ],
1187
+ "benchmark_datasets": [
1188
+ {
1189
+ "benchmark_variant": "class_known",
1190
+ "dataset_id": "finetune_50k_pth_test",
1191
+ "dataset_ref": "uspto_50k_rsgpt",
1192
+ "default": true,
1193
+ "format": "jsonl",
1194
+ "metadata_path": "benchmarks/metadata.json",
1195
+ "native_input_format": "class_csv_reaction_smiles",
1196
+ "native_source_paths": [
1197
+ "benchmarks/native/USPTO-50k_test_tidy_class1.csv"
1198
+ ],
1199
+ "notes": [
1200
+ "This is the class-labeled USPTO-50K benchmark companion file corresponding to the best-effort class-known mapping for finetune_50k_pth."
1201
+ ],
1202
+ "num_examples": 5007,
1203
+ "path": "benchmarks/default_test.jsonl",
1204
+ "split": "test"
1205
+ }
1206
+ ],
1207
+ "checkpoint_id": "finetune_50k_pth",
1208
+ "embedding_supported": false,
1209
+ "input_schema": null,
1210
+ "model_id": "rsgpt",
1211
+ "normalized_checkpoint_id": "finetune_50k_pth",
1212
+ "package_dir": "checkpoints/rsgpt/finetune_50k_pth",
1213
+ "source_registry_refs": [
1214
+ {
1215
+ "checkpoint_id": "finetune_50k_pth",
1216
+ "model_id": "rsgpt"
1217
+ }
1218
+ ],
1219
+ "task_direction": null,
1220
+ "training_dataset_refs": [
1221
+ "uspto_50k_rsgpt",
1222
+ "synthetic_10b_rsgpt"
1223
+ ],
1224
+ "weights_path": "model.safetensors",
1225
+ "wrapper_family": "rsgpt_llama_decoder"
1226
+ },
1227
+ {
1228
+ "auxiliary_files": [
1229
+ "aux/configs/base.yml",
1230
+ "aux/configs/rxngpt_llama1B.json",
1231
+ "aux/deepspeed.yaml",
1232
+ "aux/tokenizer/special_tokens_map.json",
1233
+ "aux/tokenizer/tokenizer.json",
1234
+ "aux/tokenizer/tokenizer_config.json",
1235
+ "aux/vocab.json",
1236
+ "benchmarks/default_test.jsonl",
1237
+ "benchmarks/metadata.json",
1238
+ "benchmarks/native/USPTO-FULL_test_tidy.txt"
1239
+ ],
1240
+ "benchmark_datasets": [
1241
+ {
1242
+ "dataset_id": "finetune_full_pth_test",
1243
+ "dataset_ref": "uspto_full_rsgpt",
1244
+ "default": true,
1245
+ "format": "jsonl",
1246
+ "metadata_path": "benchmarks/metadata.json",
1247
+ "native_input_format": "reaction_smiles_txt",
1248
+ "native_source_paths": [
1249
+ "benchmarks/native/USPTO-FULL_test_tidy.txt"
1250
+ ],
1251
+ "notes": [
1252
+ "Native RSGPT benchmark source extracted from USPTO_data/USPTO-FULL_test_tidy.txt."
1253
+ ],
1254
+ "num_examples": 95389,
1255
+ "path": "benchmarks/default_test.jsonl",
1256
+ "split": "test"
1257
+ }
1258
+ ],
1259
+ "checkpoint_id": "finetune_full_pth",
1260
+ "embedding_supported": false,
1261
+ "input_schema": null,
1262
+ "model_id": "rsgpt",
1263
+ "normalized_checkpoint_id": "finetune_full_pth",
1264
+ "package_dir": "checkpoints/rsgpt/finetune_full_pth",
1265
+ "source_registry_refs": [
1266
+ {
1267
+ "checkpoint_id": "finetune_full_pth",
1268
+ "model_id": "rsgpt"
1269
+ }
1270
+ ],
1271
+ "task_direction": null,
1272
+ "training_dataset_refs": [
1273
+ "uspto_full_rsgpt",
1274
+ "synthetic_10b_rsgpt"
1275
+ ],
1276
+ "weights_path": "model.safetensors",
1277
+ "wrapper_family": "rsgpt_llama_decoder"
1278
+ },
1279
+ {
1280
+ "auxiliary_files": [
1281
+ "aux/configs/base.yml",
1282
+ "aux/configs/rxngpt_llama1B.json",
1283
+ "aux/deepspeed.yaml",
1284
+ "aux/tokenizer/special_tokens_map.json",
1285
+ "aux/tokenizer/tokenizer.json",
1286
+ "aux/tokenizer/tokenizer_config.json",
1287
+ "aux/vocab.json",
1288
+ "benchmarks/default_test.jsonl",
1289
+ "benchmarks/metadata.json",
1290
+ "benchmarks/native/USPTO-MIT_test_tidy.txt"
1291
+ ],
1292
+ "benchmark_datasets": [
1293
+ {
1294
+ "dataset_id": "finetune_mit_pth_test",
1295
+ "dataset_ref": "uspto_mit_rsgpt",
1296
+ "default": true,
1297
+ "format": "jsonl",
1298
+ "metadata_path": "benchmarks/metadata.json",
1299
+ "native_input_format": "reaction_smiles_txt",
1300
+ "native_source_paths": [
1301
+ "benchmarks/native/USPTO-MIT_test_tidy.txt"
1302
+ ],
1303
+ "notes": [
1304
+ "Native RSGPT benchmark source extracted from USPTO_data/USPTO-MIT_test_tidy.txt."
1305
+ ],
1306
+ "num_examples": 40000,
1307
+ "path": "benchmarks/default_test.jsonl",
1308
+ "split": "test"
1309
+ }
1310
+ ],
1311
+ "checkpoint_id": "finetune_mit_pth",
1312
+ "embedding_supported": false,
1313
+ "input_schema": null,
1314
+ "model_id": "rsgpt",
1315
+ "normalized_checkpoint_id": "finetune_mit_pth",
1316
+ "package_dir": "checkpoints/rsgpt/finetune_mit_pth",
1317
+ "source_registry_refs": [
1318
+ {
1319
+ "checkpoint_id": "finetune_mit_pth",
1320
+ "model_id": "rsgpt"
1321
+ }
1322
+ ],
1323
+ "task_direction": null,
1324
+ "training_dataset_refs": [
1325
+ "uspto_mit_rsgpt",
1326
+ "synthetic_10b_rsgpt"
1327
+ ],
1328
+ "weights_path": "model.safetensors",
1329
+ "wrapper_family": "rsgpt_llama_decoder"
1330
+ },
1331
+ {
1332
+ "auxiliary_files": [
1333
+ "aux/configs/base.yml",
1334
+ "aux/configs/rxngpt_llama1B.json",
1335
+ "aux/deepspeed.yaml",
1336
+ "aux/tokenizer/special_tokens_map.json",
1337
+ "aux/tokenizer/tokenizer.json",
1338
+ "aux/tokenizer/tokenizer_config.json",
1339
+ "aux/vocab.json",
1340
+ "benchmarks/default_test.jsonl",
1341
+ "benchmarks/metadata.json",
1342
+ "benchmarks/native/USPTO-50k_test_tidy.txt"
1343
+ ],
1344
+ "benchmark_datasets": [
1345
+ {
1346
+ "benchmark_variant": "class_unknown_no_pretraining_ablation",
1347
+ "dataset_id": "nopretrain_50k_pth_test",
1348
+ "dataset_ref": "uspto_50k_rsgpt",
1349
+ "default": true,
1350
+ "format": "jsonl",
1351
+ "metadata_path": "benchmarks/metadata.json",
1352
+ "native_input_format": "reaction_smiles_txt",
1353
+ "native_source_paths": [
1354
+ "benchmarks/native/USPTO-50k_test_tidy.txt"
1355
+ ],
1356
+ "notes": [
1357
+ "Native RSGPT benchmark source extracted from USPTO_data/USPTO-50k_test_tidy.txt."
1358
+ ],
1359
+ "num_examples": 5007,
1360
+ "path": "benchmarks/default_test.jsonl",
1361
+ "split": "test"
1362
+ }
1363
+ ],
1364
+ "checkpoint_id": "nopretrain_50k_pth",
1365
+ "embedding_supported": false,
1366
+ "input_schema": null,
1367
+ "model_id": "rsgpt",
1368
+ "normalized_checkpoint_id": "nopretrain_50k_pth",
1369
+ "package_dir": "checkpoints/rsgpt/nopretrain_50k_pth",
1370
+ "source_registry_refs": [
1371
+ {
1372
+ "checkpoint_id": "nopretrain_50k_pth",
1373
+ "model_id": "rsgpt"
1374
+ }
1375
+ ],
1376
+ "task_direction": null,
1377
+ "training_dataset_refs": [
1378
+ "uspto_50k_rsgpt"
1379
+ ],
1380
+ "weights_path": "model.safetensors",
1381
+ "wrapper_family": "rsgpt_llama_decoder"
1382
+ }
1383
+ ],
1384
+ "generated_at": "2026-06-28T00:57:27.202938+00:00"
1385
+ }
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward/SHA256SUMS ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 11ad7efa24975ee4b0c3c3a38ed18737f0658a5f75a0a96787b576a78a023361 aux/.gitattributes
2
+ 1a689168cf12a60fb49166cdc5245308f6ef18d28d8fec165295edf7404449c2 aux/README.md
3
+ 005ab587497141bf8a42fef5e820a3c2956893dbf444538722cbf010554f6a6a aux/UPSTREAM_URL.txt
4
+ a80eac6938e51ccabfc6a8b281497c6553626145e57b2268101f718fbaa034a2 aux/config.json
5
+ aed028489d41ae6a673920da9d103f2f58a650c3fbc50bd880b35e30a481dba6 aux/generation_config.json
6
+ 116da2894aad609845ba7977cbcdbbba5eddd02805fd2648419595464de98a91 aux/hf_model_info.json
7
+ d780354534a3a8d64e0cdb2ea1b69ca481ba3dcda6aa06c3f68ff46ae3b26b93 aux/provenance/README.txt
8
+ 7e03326292c50090f69e2ceefc5ab8ef78e07f13c6f76fa6a58975f11d4142e9 aux/provenance/ReactionT5v2_source_README.md
9
+ 84786cdb3d3abc34c98f56c76be1e72d28e9f78a4f463611367a42aaea9424aa aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv
10
+ 24e218c6b258f7624f1c9557678c29e7b6a18f2c9d725c2cc581d9cd57727e5b aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv
11
+ 6db5bdb900d801874a1728640ef8004e16381e71f7dad5700c7f1e4d8be573e9 aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv
12
+ b4d870d0550f6cf10a887dcee2bcfbaa46db14dbbdec734fa0f48b7ec661b5cb aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv
13
+ 4c906e960cd96a0f39e1790314db642742a0fe8152e4481cc7c81aac83f4ed14 aux/provenance/datasets/USPTO_MIT/MIT_separated/train.csv
14
+ 1712f93b057f783fb2fbb4f74a89ab9e302e19230d938c4f74c5ad959df8fc1d aux/provenance/datasets/USPTO_MIT/MIT_separated/val.csv
15
+ e4cf7b1b077d246605c1f2c7d4d2056d8d0d2dedbea0fc4e61748545a7a742cb aux/provenance/reactiont5v2_forward_split_provenance.json
16
+ 0f385cf62b07c1b63aaa84d4c86a313c751df7039d958a7d7bd7e723d933bcd9 aux/special_tokens_map.json
17
+ 511dfd8bb25c19cccae8d7d996c1ff0f1b5263207be4cdd275f23b75da8113b6 aux/tokenizer.json
18
+ ac38d16067df56f00dfa2c2e0d1546a6ff08aa37814662a6fc38dea66cd13d9d aux/tokenizer_config.json
19
+ f89bde13ce68202577b7eddf752376632a40ce296e5c005da51e37bba6d195ac aux/trainer_state.json
20
+ 0d40cc65896c8c30098177fdb6440c4900ea381e1fc01b3348193f689f02c9ba aux/training_args.bin
21
+ 162756dda7a3c5261a4905a1356d2e43362055a755a9e78d5909145d982a2a56 benchmarks/default_test.jsonl
22
+ f3f4c7133af1f5de7a2d8a5a0befaf66e746b6a1813cbe39089b3856829e425e benchmarks/metadata.json
23
+ bd330a7507780dc03bde5e95fb838b274c0bbb8b984913a41d1d7d00ac8201ac manifest.json
24
+ 69c23ca364e63e0e2be22d114652aad9f3df4e161d43936eee0dc3ce74a87fb7 model.safetensors
25
+ 93e76922d12ee2e5100b7a9c54279b80c01516633310e2979d19ede9d869ef34 provenance.json
26
+ 9c34f4d138ffe9d72b185b09856650809e2153522fbfeab8363db2017763ec6e wrapper_config.json
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward/manifest.json ADDED
@@ -0,0 +1,137 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auxiliary_files": [
3
+ "aux/.gitattributes",
4
+ "aux/README.md",
5
+ "aux/UPSTREAM_URL.txt",
6
+ "aux/config.json",
7
+ "aux/generation_config.json",
8
+ "aux/hf_model_info.json",
9
+ "aux/provenance/README.txt",
10
+ "aux/provenance/ReactionT5v2_source_README.md",
11
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv",
12
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
13
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
14
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv",
15
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/train.csv",
16
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/val.csv",
17
+ "aux/provenance/reactiont5v2_forward_split_provenance.json",
18
+ "aux/special_tokens_map.json",
19
+ "aux/tokenizer.json",
20
+ "aux/tokenizer_config.json",
21
+ "aux/trainer_state.json",
22
+ "aux/training_args.bin",
23
+ "benchmarks/default_test.jsonl",
24
+ "benchmarks/metadata.json"
25
+ ],
26
+ "benchmark_datasets": [
27
+ {
28
+ "benchmark_semantics": "ord_trained_forward_checkpoint_evaluated_on_uspto_mit_holdout",
29
+ "benchmark_variant": "forward_product_prediction",
30
+ "blank_reagent_value": " ",
31
+ "dataset_id": "sagawa_reactiont5v2_forward_test",
32
+ "dataset_ref": "uspto_mit_forward_project_release",
33
+ "default": true,
34
+ "format": "jsonl",
35
+ "metadata_path": "benchmarks/metadata.json",
36
+ "native_input_format": "csv_forward_reaction_rows",
37
+ "native_source_paths": [
38
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv"
39
+ ],
40
+ "notes": [
41
+ "Native forward benchmark split for the public USPTO_MIT product-prediction task.",
42
+ "The normalized JSONL preserves reactants, reagents, and expected product(s) rather than the backward exact-match schema.",
43
+ "This ORD-trained checkpoint is evaluated on the USPTO_MIT holdout documented in the upstream README/model card."
44
+ ],
45
+ "num_examples": 40000,
46
+ "path": "benchmarks/default_test.jsonl",
47
+ "record_schema": {
48
+ "inputs": [
49
+ "reactants",
50
+ "reagents"
51
+ ],
52
+ "targets": [
53
+ "expected_products"
54
+ ]
55
+ },
56
+ "serialized_input_format": "REACTANT:<reactants>REAGENT:<reagents>",
57
+ "split": "test",
58
+ "task_direction": "forward"
59
+ }
60
+ ],
61
+ "checkpoint_id": "sagawa_reactiont5v2_forward",
62
+ "contained_artifact_bindings": [],
63
+ "conversion": {
64
+ "converted_at": "1970-01-01T00:00:00+00:00",
65
+ "source_format": "huggingface_safetensors",
66
+ "status": "converted",
67
+ "tool_version": "initial"
68
+ },
69
+ "display_name": "sagawa/ReactionT5v2-forward",
70
+ "embedding_supported": false,
71
+ "fine_tuning": {
72
+ "adapter_overlay_dir": "adapters",
73
+ "base_weights_path": "model.safetensors",
74
+ "notes": [
75
+ "This package preserves immutable base weights in model.safetensors and reserves adapters/ for future fine-tuning or LoRA overlays."
76
+ ]
77
+ },
78
+ "format_version": 1,
79
+ "generated_at": "1970-01-01T00:00:00+00:00",
80
+ "input_schema": {
81
+ "blank_reagent_value": " ",
82
+ "reactants_field": "REACTANT",
83
+ "reagents_field": "REAGENT",
84
+ "serialized_format": "REACTANT:<reactants>REAGENT:<reagents>",
85
+ "target_field": "PRODUCT"
86
+ },
87
+ "model_id": "reactiont5v2_forward",
88
+ "normalized_checkpoint_id": "sagawa_reactiont5v2_forward",
89
+ "package_layout": {
90
+ "adapters_dir": "adapters",
91
+ "aux_dir": "aux",
92
+ "benchmarks_dir": "benchmarks",
93
+ "weights_path": "model.safetensors"
94
+ },
95
+ "source_registry_refs": [
96
+ {
97
+ "checkpoint_id": "sagawa_reactiont5v2_forward",
98
+ "model_id": "reactiont5v2_forward"
99
+ }
100
+ ],
101
+ "split_provenance": [
102
+ {
103
+ "canonical_split_family": "reactiont5v2_project_ord_forward_train_valid_test",
104
+ "dataset_ref": "ord_forward_pretraining",
105
+ "split_names": [
106
+ "train",
107
+ "valid",
108
+ "test"
109
+ ],
110
+ "stage": "forward_training_or_pretraining"
111
+ }
112
+ ],
113
+ "task_direction": "forward",
114
+ "training_dataset_refs": [
115
+ "ord_forward_pretraining"
116
+ ],
117
+ "training_stage_bindings": [
118
+ {
119
+ "canonical_split_family": "reactiont5v2_project_ord_forward_train_valid_test",
120
+ "dataset_ref": "ord_forward_pretraining",
121
+ "split_names": [
122
+ "train",
123
+ "valid",
124
+ "test"
125
+ ],
126
+ "stage": "forward_training_or_pretraining"
127
+ }
128
+ ],
129
+ "weights": {
130
+ "dtype_summary": "upstream_preserved",
131
+ "framework": "safetensors",
132
+ "path": "model.safetensors",
133
+ "tensor_namespace": "native"
134
+ },
135
+ "wrapper_config": "wrapper_config.json",
136
+ "wrapper_family": "seq2seq_t5"
137
+ }
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward/provenance.json ADDED
@@ -0,0 +1,495 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "conversion_notes": [
3
+ "Upstream Hugging Face safetensors reused verbatim; no tensor conversion performed."
4
+ ],
5
+ "copied_sidecars": [
6
+ "aux/.gitattributes",
7
+ "aux/README.md",
8
+ "aux/UPSTREAM_URL.txt",
9
+ "aux/config.json",
10
+ "aux/generation_config.json",
11
+ "aux/hf_model_info.json",
12
+ "aux/provenance/README.txt",
13
+ "aux/provenance/ReactionT5v2_source_README.md",
14
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv",
15
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
16
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
17
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv",
18
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/train.csv",
19
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/val.csv",
20
+ "aux/provenance/reactiont5v2_forward_split_provenance.json",
21
+ "aux/special_tokens_map.json",
22
+ "aux/tokenizer.json",
23
+ "aux/tokenizer_config.json",
24
+ "aux/trainer_state.json",
25
+ "aux/training_args.bin"
26
+ ],
27
+ "inner_members": [],
28
+ "normalized_checkpoint_id": "sagawa_reactiont5v2_forward",
29
+ "source_artifacts": [
30
+ {
31
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/README.md",
32
+ "kind": "local_repo_file",
33
+ "path": "datasets/forward_models/reactionT5v2/README.md",
34
+ "role": "source_artifact",
35
+ "sha256": "8584b84f91f41356ee736fa7cfb797d99d82aeeade504ef0f4097388e3c9e890",
36
+ "size": 1414
37
+ },
38
+ {
39
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/.gitignore",
40
+ "kind": "local_repo_file",
41
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/.gitignore",
42
+ "role": "source_artifact",
43
+ "sha256": "684888c0ebb17f374298b65ee2807526c066094c701bcc7ebbe1c1095f494fc1",
44
+ "size": 1
45
+ },
46
+ {
47
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/.gitattributes.metadata",
48
+ "kind": "local_repo_file",
49
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/.gitattributes.metadata",
50
+ "role": "source_artifact",
51
+ "sha256": "8dbb80384bb198b63c09c92fa39320671fe97465152f8633c895ccdd95986236",
52
+ "size": 101
53
+ },
54
+ {
55
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/README.md.metadata",
56
+ "kind": "local_repo_file",
57
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/README.md.metadata",
58
+ "role": "source_artifact",
59
+ "sha256": "0495464d7459458caec3719798d56070718922d3bd6e1b4e6c00e76443d33d93",
60
+ "size": 101
61
+ },
62
+ {
63
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/config.json.metadata",
64
+ "kind": "local_repo_file",
65
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/config.json.metadata",
66
+ "role": "source_artifact",
67
+ "sha256": "34a1a79f11cdac899282589c1377c639fe1d923d36ab9987e23f7ba7c3f8a0e9",
68
+ "size": 101
69
+ },
70
+ {
71
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/generation_config.json.metadata",
72
+ "kind": "local_repo_file",
73
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/generation_config.json.metadata",
74
+ "role": "source_artifact",
75
+ "sha256": "2f59b6ddb10e05e11d0a053d261e3063e1207afe4a757153b15c8a1a4a5e07ce",
76
+ "size": 101
77
+ },
78
+ {
79
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/model.safetensors.metadata",
80
+ "kind": "local_repo_file",
81
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/model.safetensors.metadata",
82
+ "role": "source_artifact",
83
+ "sha256": "decd5fd8ab381657d595f428d8518fbf93a3d0fd1070079408ca588f0926a6a4",
84
+ "size": 125
85
+ },
86
+ {
87
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/special_tokens_map.json.metadata",
88
+ "kind": "local_repo_file",
89
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/special_tokens_map.json.metadata",
90
+ "role": "source_artifact",
91
+ "sha256": "703d4f5fe17f554f1bfcf4c9dacadb2670ec93660021c11c884c6e716013b539",
92
+ "size": 100
93
+ },
94
+ {
95
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/tokenizer.json.metadata",
96
+ "kind": "local_repo_file",
97
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/tokenizer.json.metadata",
98
+ "role": "source_artifact",
99
+ "sha256": "dcd06983dbea11ce92bdc72a7a109dd53936324020404f9c3a6e380892f82ed5",
100
+ "size": 101
101
+ },
102
+ {
103
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/tokenizer_config.json.metadata",
104
+ "kind": "local_repo_file",
105
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/tokenizer_config.json.metadata",
106
+ "role": "source_artifact",
107
+ "sha256": "e58e87ec0be6fdd7ab16a8eddf74b03a8e4b2cfe7c0ec17c1ce97a50303c908e",
108
+ "size": 101
109
+ },
110
+ {
111
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/trainer_state.json.metadata",
112
+ "kind": "local_repo_file",
113
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/trainer_state.json.metadata",
114
+ "role": "source_artifact",
115
+ "sha256": "23c346163e7f13692ee788f8576a84266aa436d70a5eae13367281375ab071b9",
116
+ "size": 101
117
+ },
118
+ {
119
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/training_args.bin.metadata",
120
+ "kind": "local_repo_file",
121
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/training_args.bin.metadata",
122
+ "role": "source_artifact",
123
+ "sha256": "e05c87b1eb8806e44385596c115f70806891932e31914a6040094fda76368c1d",
124
+ "size": 124
125
+ },
126
+ {
127
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.gitattributes",
128
+ "kind": "local_repo_file",
129
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.gitattributes",
130
+ "role": "source_artifact",
131
+ "sha256": "11ad7efa24975ee4b0c3c3a38ed18737f0658a5f75a0a96787b576a78a023361",
132
+ "size": 1519
133
+ },
134
+ {
135
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/README.md",
136
+ "kind": "local_repo_file",
137
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/README.md",
138
+ "role": "source_artifact",
139
+ "sha256": "1a689168cf12a60fb49166cdc5245308f6ef18d28d8fec165295edf7404449c2",
140
+ "size": 4822
141
+ },
142
+ {
143
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/UPSTREAM_URL.txt",
144
+ "kind": "local_repo_file",
145
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/UPSTREAM_URL.txt",
146
+ "role": "source_artifact",
147
+ "sha256": "005ab587497141bf8a42fef5e820a3c2956893dbf444538722cbf010554f6a6a",
148
+ "size": 101
149
+ },
150
+ {
151
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/config.json",
152
+ "kind": "local_repo_file",
153
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/config.json",
154
+ "role": "source_artifact",
155
+ "sha256": "a80eac6938e51ccabfc6a8b281497c6553626145e57b2268101f718fbaa034a2",
156
+ "size": 825
157
+ },
158
+ {
159
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/generation_config.json",
160
+ "kind": "local_repo_file",
161
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/generation_config.json",
162
+ "role": "source_artifact",
163
+ "sha256": "aed028489d41ae6a673920da9d103f2f58a650c3fbc50bd880b35e30a481dba6",
164
+ "size": 133
165
+ },
166
+ {
167
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/hf_model_info.json",
168
+ "kind": "local_repo_file",
169
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/hf_model_info.json",
170
+ "role": "source_artifact",
171
+ "sha256": "116da2894aad609845ba7977cbcdbbba5eddd02805fd2648419595464de98a91",
172
+ "size": 1967
173
+ },
174
+ {
175
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/model.safetensors",
176
+ "kind": "local_repo_file",
177
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/model.safetensors",
178
+ "role": "source_artifact",
179
+ "sha256": "69c23ca364e63e0e2be22d114652aad9f3df4e161d43936eee0dc3ce74a87fb7",
180
+ "size": 794597224
181
+ },
182
+ {
183
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/special_tokens_map.json",
184
+ "kind": "local_repo_file",
185
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/special_tokens_map.json",
186
+ "role": "source_artifact",
187
+ "sha256": "0f385cf62b07c1b63aaa84d4c86a313c751df7039d958a7d7bd7e723d933bcd9",
188
+ "size": 15428
189
+ },
190
+ {
191
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/tokenizer.json",
192
+ "kind": "local_repo_file",
193
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/tokenizer.json",
194
+ "role": "source_artifact",
195
+ "sha256": "511dfd8bb25c19cccae8d7d996c1ff0f1b5263207be4cdd275f23b75da8113b6",
196
+ "size": 36514
197
+ },
198
+ {
199
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/tokenizer_config.json",
200
+ "kind": "local_repo_file",
201
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/tokenizer_config.json",
202
+ "role": "source_artifact",
203
+ "sha256": "ac38d16067df56f00dfa2c2e0d1546a6ff08aa37814662a6fc38dea66cd13d9d",
204
+ "size": 28602
205
+ },
206
+ {
207
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/trainer_state.json",
208
+ "kind": "local_repo_file",
209
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/trainer_state.json",
210
+ "role": "source_artifact",
211
+ "sha256": "f89bde13ce68202577b7eddf752376632a40ce296e5c005da51e37bba6d195ac",
212
+ "size": 40605
213
+ },
214
+ {
215
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/training_args.bin",
216
+ "kind": "local_repo_file",
217
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/training_args.bin",
218
+ "role": "source_artifact",
219
+ "sha256": "0d40cc65896c8c30098177fdb6440c4900ea381e1fc01b3348193f689f02c9ba",
220
+ "size": 5112
221
+ },
222
+ {
223
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/.gitignore",
224
+ "kind": "local_repo_file",
225
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/.gitignore",
226
+ "role": "source_artifact",
227
+ "sha256": "684888c0ebb17f374298b65ee2807526c066094c701bcc7ebbe1c1095f494fc1",
228
+ "size": 1
229
+ },
230
+ {
231
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/.gitattributes.metadata",
232
+ "kind": "local_repo_file",
233
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/.gitattributes.metadata",
234
+ "role": "source_artifact",
235
+ "sha256": "d810874057b9fbbab79e9bddd2c6f7d92dbdcf2c2f61bff1c7cede7995b9c31d",
236
+ "size": 101
237
+ },
238
+ {
239
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/README.md.metadata",
240
+ "kind": "local_repo_file",
241
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/README.md.metadata",
242
+ "role": "source_artifact",
243
+ "sha256": "d079b773989ec3afa54b35f0ba650987123cd7e3e5c505803dee5a09a286abdc",
244
+ "size": 101
245
+ },
246
+ {
247
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/config.json.metadata",
248
+ "kind": "local_repo_file",
249
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/config.json.metadata",
250
+ "role": "source_artifact",
251
+ "sha256": "e29e915afc6ce29582e4fbb666f8ad53949f5c6af3b748fc0cde078759170a82",
252
+ "size": 101
253
+ },
254
+ {
255
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/generation_config.json.metadata",
256
+ "kind": "local_repo_file",
257
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/generation_config.json.metadata",
258
+ "role": "source_artifact",
259
+ "sha256": "baac117440c18d95e8669a85e39beb377a298addabee514b74879d3c9a817476",
260
+ "size": 101
261
+ },
262
+ {
263
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/model.safetensors.metadata",
264
+ "kind": "local_repo_file",
265
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/model.safetensors.metadata",
266
+ "role": "source_artifact",
267
+ "sha256": "c21b4b9b37243be2e068eb0a3c1c19803b573f3ebfbf78779b0512f501d784b8",
268
+ "size": 125
269
+ },
270
+ {
271
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/special_tokens_map.json.metadata",
272
+ "kind": "local_repo_file",
273
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/special_tokens_map.json.metadata",
274
+ "role": "source_artifact",
275
+ "sha256": "32272b480cb55f0eadfd294c1276638055b51654f48fe26670dcc30bd8ff94ae",
276
+ "size": 100
277
+ },
278
+ {
279
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/tokenizer.json.metadata",
280
+ "kind": "local_repo_file",
281
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/tokenizer.json.metadata",
282
+ "role": "source_artifact",
283
+ "sha256": "36b542e97a9f4e4e051d17eabe6428742f01574fe552b71b36500af960d2ee01",
284
+ "size": 101
285
+ },
286
+ {
287
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/tokenizer_config.json.metadata",
288
+ "kind": "local_repo_file",
289
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/tokenizer_config.json.metadata",
290
+ "role": "source_artifact",
291
+ "sha256": "bf546f252ad7756f049c9edc1d114ebd769d6c8f9a6f82c8ab059ef63d75a5af",
292
+ "size": 101
293
+ },
294
+ {
295
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/trainer_state.json.metadata",
296
+ "kind": "local_repo_file",
297
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/trainer_state.json.metadata",
298
+ "role": "source_artifact",
299
+ "sha256": "2b8ec757b5caa233136eb0cfefbb572d01f736cbefa7d47c78f6341deae6ef97",
300
+ "size": 100
301
+ },
302
+ {
303
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/training_args.bin.metadata",
304
+ "kind": "local_repo_file",
305
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/training_args.bin.metadata",
306
+ "role": "source_artifact",
307
+ "sha256": "799bfb2cf2602b85a922a7043a3787624d74ebd856299b20c8543536a2bb0f46",
308
+ "size": 125
309
+ },
310
+ {
311
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.gitattributes",
312
+ "kind": "local_repo_file",
313
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.gitattributes",
314
+ "role": "source_artifact",
315
+ "sha256": "11ad7efa24975ee4b0c3c3a38ed18737f0658a5f75a0a96787b576a78a023361",
316
+ "size": 1519
317
+ },
318
+ {
319
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/README.md",
320
+ "kind": "local_repo_file",
321
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/README.md",
322
+ "role": "source_artifact",
323
+ "sha256": "abc9ceaf22822bc2736909eac46f42f29a911c87bca6f29e1979ffe6e701904d",
324
+ "size": 4732
325
+ },
326
+ {
327
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/UPSTREAM_URL.txt",
328
+ "kind": "local_repo_file",
329
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/UPSTREAM_URL.txt",
330
+ "role": "source_artifact",
331
+ "sha256": "5a41716f340aff0d540fe7da15f12cd4b3818378a6922d8fbc242faf45c4998e",
332
+ "size": 111
333
+ },
334
+ {
335
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/config.json",
336
+ "kind": "local_repo_file",
337
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/config.json",
338
+ "role": "source_artifact",
339
+ "sha256": "c637cdad4f34d56a638ac6031e742a407b2ed6b9d2d674f1e8bc8cb02034ae90",
340
+ "size": 903
341
+ },
342
+ {
343
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/generation_config.json",
344
+ "kind": "local_repo_file",
345
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/generation_config.json",
346
+ "role": "source_artifact",
347
+ "sha256": "aed028489d41ae6a673920da9d103f2f58a650c3fbc50bd880b35e30a481dba6",
348
+ "size": 133
349
+ },
350
+ {
351
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/hf_model_info.json",
352
+ "kind": "local_repo_file",
353
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/hf_model_info.json",
354
+ "role": "source_artifact",
355
+ "sha256": "79debc90641b5144f54bd3e1c0199463b18c52a32408930964505883038e3f66",
356
+ "size": 1766
357
+ },
358
+ {
359
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/model.safetensors",
360
+ "kind": "local_repo_file",
361
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/model.safetensors",
362
+ "role": "source_artifact",
363
+ "sha256": "b5fdde934cb203742547527154c3af33560a143e606118dcea81076f0e4ab458",
364
+ "size": 794597224
365
+ },
366
+ {
367
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/special_tokens_map.json",
368
+ "kind": "local_repo_file",
369
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/special_tokens_map.json",
370
+ "role": "source_artifact",
371
+ "sha256": "a19f21a50f05b46cbbad085149d520ab1bc843a43976a08693df39f6ccc796d4",
372
+ "size": 2576
373
+ },
374
+ {
375
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/tokenizer.json",
376
+ "kind": "local_repo_file",
377
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/tokenizer.json",
378
+ "role": "source_artifact",
379
+ "sha256": "0ef5dc3e20dbfd2a3cafafbbbb61c3de27bc88372e0d96771d2b5fffe9cc0ed0",
380
+ "size": 36514
381
+ },
382
+ {
383
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/tokenizer_config.json",
384
+ "kind": "local_repo_file",
385
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/tokenizer_config.json",
386
+ "role": "source_artifact",
387
+ "sha256": "28ddd7e807d75d1b217963c1d15e8e5d57d3e5404999a25ef4e352de78daff5a",
388
+ "size": 28710
389
+ },
390
+ {
391
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/trainer_state.json",
392
+ "kind": "local_repo_file",
393
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/trainer_state.json",
394
+ "role": "source_artifact",
395
+ "sha256": "64a998131ae89d4405cef5d91614458885a49bbd6bc79733ae593cb8d10f5458",
396
+ "size": 3712
397
+ },
398
+ {
399
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/training_args.bin",
400
+ "kind": "local_repo_file",
401
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/training_args.bin",
402
+ "role": "source_artifact",
403
+ "sha256": "480806311c2fb18be472ad7ff49ac238235b6f54c2285fee10edef8206fd5bab",
404
+ "size": 5112
405
+ },
406
+ {
407
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/downloads/ORD.zip",
408
+ "kind": "local_repo_file",
409
+ "path": "datasets/forward_models/reactionT5v2/downloads/ORD.zip",
410
+ "role": "source_artifact",
411
+ "sha256": "05a999637c9e4a94673a6f8914e1b682cb7d90434fbc9678d296aa9285159c8b",
412
+ "size": 233136954
413
+ },
414
+ {
415
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/downloads/USPTO_MIT.zip",
416
+ "kind": "local_repo_file",
417
+ "path": "datasets/forward_models/reactionT5v2/downloads/USPTO_MIT.zip",
418
+ "role": "source_artifact",
419
+ "sha256": "3fb96bef12b0e87ef0f0ea243723a861b1fde161f6cd9dc45fe879eb98891c70",
420
+ "size": 14183273
421
+ },
422
+ {
423
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/README.txt",
424
+ "kind": "local_repo_file",
425
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/README.txt",
426
+ "role": "source_artifact",
427
+ "sha256": "d780354534a3a8d64e0cdb2ea1b69ca481ba3dcda6aa06c3f68ff46ae3b26b93",
428
+ "size": 307
429
+ },
430
+ {
431
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/ReactionT5v2_source_README.md",
432
+ "kind": "local_repo_file",
433
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/ReactionT5v2_source_README.md",
434
+ "role": "source_artifact",
435
+ "sha256": "7e03326292c50090f69e2ceefc5ab8ef78e07f13c6f76fa6a58975f11d4142e9",
436
+ "size": 12188
437
+ },
438
+ {
439
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv",
440
+ "kind": "local_repo_file",
441
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv",
442
+ "role": "source_artifact",
443
+ "sha256": "84786cdb3d3abc34c98f56c76be1e72d28e9f78a4f463611367a42aaea9424aa",
444
+ "size": 39188316
445
+ },
446
+ {
447
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
448
+ "kind": "local_repo_file",
449
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
450
+ "role": "source_artifact",
451
+ "sha256": "24e218c6b258f7624f1c9557678c29e7b6a18f2c9d725c2cc581d9cd57727e5b",
452
+ "size": 313922807
453
+ },
454
+ {
455
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
456
+ "kind": "local_repo_file",
457
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
458
+ "role": "source_artifact",
459
+ "sha256": "6db5bdb900d801874a1728640ef8004e16381e71f7dad5700c7f1e4d8be573e9",
460
+ "size": 39251481
461
+ },
462
+ {
463
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/datasets/USPTO_MIT/MIT_separated/test.csv",
464
+ "kind": "local_repo_file",
465
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/datasets/USPTO_MIT/MIT_separated/test.csv",
466
+ "role": "source_artifact",
467
+ "sha256": "b4d870d0550f6cf10a887dcee2bcfbaa46db14dbbdec734fa0f48b7ec661b5cb",
468
+ "size": 8612805
469
+ },
470
+ {
471
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/datasets/USPTO_MIT/MIT_separated/train.csv",
472
+ "kind": "local_repo_file",
473
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/datasets/USPTO_MIT/MIT_separated/train.csv",
474
+ "role": "source_artifact",
475
+ "sha256": "4c906e960cd96a0f39e1790314db642742a0fe8152e4481cc7c81aac83f4ed14",
476
+ "size": 48191398
477
+ },
478
+ {
479
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/datasets/USPTO_MIT/MIT_separated/val.csv",
480
+ "kind": "local_repo_file",
481
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/datasets/USPTO_MIT/MIT_separated/val.csv",
482
+ "role": "source_artifact",
483
+ "sha256": "1712f93b057f783fb2fbb4f74a89ab9e302e19230d938c4f74c5ad959df8fc1d",
484
+ "size": 3526860
485
+ },
486
+ {
487
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/reactiont5v2_forward_split_provenance.json",
488
+ "kind": "local_repo_file",
489
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/reactiont5v2_forward_split_provenance.json",
490
+ "role": "source_artifact",
491
+ "sha256": "e4cf7b1b077d246605c1f2c7d4d2056d8d0d2dedbea0fc4e61748545a7a742cb",
492
+ "size": 2252
493
+ }
494
+ ]
495
+ }
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward/wrapper_config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architecture_class": "T5ForConditionalGeneration",
3
+ "config_file": "aux/config.json",
4
+ "device_supported": [
5
+ "cpu",
6
+ "cuda"
7
+ ],
8
+ "fine_tuning": {
9
+ "adapter_overlay_dir": "adapters",
10
+ "base_weights_path": "model.safetensors",
11
+ "notes": [
12
+ "Base weights stay immutable in model.safetensors; future fine-tuning or LoRA adapters should be added as overlays under adapters/."
13
+ ]
14
+ },
15
+ "generation_config": "aux/generation_config.json",
16
+ "loader": {
17
+ "state_dict_format": "safetensors",
18
+ "strict": true
19
+ },
20
+ "package_layout": {
21
+ "adapters_dir": "adapters",
22
+ "aux_dir": "aux",
23
+ "weights_path": "model.safetensors"
24
+ },
25
+ "runtime_file_map": {
26
+ "aux/config.json": "config.json",
27
+ "aux/generation_config.json": "generation_config.json",
28
+ "aux/special_tokens_map.json": "special_tokens_map.json",
29
+ "aux/tokenizer.json": "tokenizer.json",
30
+ "aux/tokenizer_config.json": "tokenizer_config.json",
31
+ "model.safetensors": "model.safetensors"
32
+ },
33
+ "task_direction": "forward",
34
+ "tokenizer_files": [
35
+ "aux/tokenizer.json",
36
+ "aux/tokenizer_config.json",
37
+ "aux/special_tokens_map.json"
38
+ ],
39
+ "wrapper_family": "seq2seq_t5"
40
+ }
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/SHA256SUMS ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 11ad7efa24975ee4b0c3c3a38ed18737f0658a5f75a0a96787b576a78a023361 aux/.gitattributes
2
+ abc9ceaf22822bc2736909eac46f42f29a911c87bca6f29e1979ffe6e701904d aux/README.md
3
+ 5a41716f340aff0d540fe7da15f12cd4b3818378a6922d8fbc242faf45c4998e aux/UPSTREAM_URL.txt
4
+ c637cdad4f34d56a638ac6031e742a407b2ed6b9d2d674f1e8bc8cb02034ae90 aux/config.json
5
+ aed028489d41ae6a673920da9d103f2f58a650c3fbc50bd880b35e30a481dba6 aux/generation_config.json
6
+ 79debc90641b5144f54bd3e1c0199463b18c52a32408930964505883038e3f66 aux/hf_model_info.json
7
+ d780354534a3a8d64e0cdb2ea1b69ca481ba3dcda6aa06c3f68ff46ae3b26b93 aux/provenance/README.txt
8
+ 7e03326292c50090f69e2ceefc5ab8ef78e07f13c6f76fa6a58975f11d4142e9 aux/provenance/ReactionT5v2_source_README.md
9
+ 84786cdb3d3abc34c98f56c76be1e72d28e9f78a4f463611367a42aaea9424aa aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv
10
+ 24e218c6b258f7624f1c9557678c29e7b6a18f2c9d725c2cc581d9cd57727e5b aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv
11
+ 6db5bdb900d801874a1728640ef8004e16381e71f7dad5700c7f1e4d8be573e9 aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv
12
+ b4d870d0550f6cf10a887dcee2bcfbaa46db14dbbdec734fa0f48b7ec661b5cb aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv
13
+ 4c906e960cd96a0f39e1790314db642742a0fe8152e4481cc7c81aac83f4ed14 aux/provenance/datasets/USPTO_MIT/MIT_separated/train.csv
14
+ 1712f93b057f783fb2fbb4f74a89ab9e302e19230d938c4f74c5ad959df8fc1d aux/provenance/datasets/USPTO_MIT/MIT_separated/val.csv
15
+ e4cf7b1b077d246605c1f2c7d4d2056d8d0d2dedbea0fc4e61748545a7a742cb aux/provenance/reactiont5v2_forward_split_provenance.json
16
+ a19f21a50f05b46cbbad085149d520ab1bc843a43976a08693df39f6ccc796d4 aux/special_tokens_map.json
17
+ 0ef5dc3e20dbfd2a3cafafbbbb61c3de27bc88372e0d96771d2b5fffe9cc0ed0 aux/tokenizer.json
18
+ 28ddd7e807d75d1b217963c1d15e8e5d57d3e5404999a25ef4e352de78daff5a aux/tokenizer_config.json
19
+ 64a998131ae89d4405cef5d91614458885a49bbd6bc79733ae593cb8d10f5458 aux/trainer_state.json
20
+ 480806311c2fb18be472ad7ff49ac238235b6f54c2285fee10edef8206fd5bab aux/training_args.bin
21
+ 162756dda7a3c5261a4905a1356d2e43362055a755a9e78d5909145d982a2a56 benchmarks/default_test.jsonl
22
+ 984d2a1c06593068d98ec04b29fc3f7f074f6cc3a5faaf85de6b4c9e51bea15b benchmarks/metadata.json
23
+ 8b593c28203cf38f846685a958f2dddf9cf7a3d3b7205e557c54ce8b7e17f45b manifest.json
24
+ b5fdde934cb203742547527154c3af33560a143e606118dcea81076f0e4ab458 model.safetensors
25
+ 3ad245da00d952ee8ad24ab41f086c6b4db307c6ce5e5959b1e6a28a36112ba1 provenance.json
26
+ 9c34f4d138ffe9d72b185b09856650809e2153522fbfeab8363db2017763ec6e wrapper_config.json
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/.gitattributes ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ *.7z filter=lfs diff=lfs merge=lfs -text
2
+ *.arrow filter=lfs diff=lfs merge=lfs -text
3
+ *.bin filter=lfs diff=lfs merge=lfs -text
4
+ *.bz2 filter=lfs diff=lfs merge=lfs -text
5
+ *.ckpt filter=lfs diff=lfs merge=lfs -text
6
+ *.ftz filter=lfs diff=lfs merge=lfs -text
7
+ *.gz filter=lfs diff=lfs merge=lfs -text
8
+ *.h5 filter=lfs diff=lfs merge=lfs -text
9
+ *.joblib filter=lfs diff=lfs merge=lfs -text
10
+ *.lfs.* filter=lfs diff=lfs merge=lfs -text
11
+ *.mlmodel filter=lfs diff=lfs merge=lfs -text
12
+ *.model filter=lfs diff=lfs merge=lfs -text
13
+ *.msgpack filter=lfs diff=lfs merge=lfs -text
14
+ *.npy filter=lfs diff=lfs merge=lfs -text
15
+ *.npz filter=lfs diff=lfs merge=lfs -text
16
+ *.onnx filter=lfs diff=lfs merge=lfs -text
17
+ *.ot filter=lfs diff=lfs merge=lfs -text
18
+ *.parquet filter=lfs diff=lfs merge=lfs -text
19
+ *.pb filter=lfs diff=lfs merge=lfs -text
20
+ *.pickle filter=lfs diff=lfs merge=lfs -text
21
+ *.pkl filter=lfs diff=lfs merge=lfs -text
22
+ *.pt filter=lfs diff=lfs merge=lfs -text
23
+ *.pth filter=lfs diff=lfs merge=lfs -text
24
+ *.rar filter=lfs diff=lfs merge=lfs -text
25
+ *.safetensors filter=lfs diff=lfs merge=lfs -text
26
+ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
27
+ *.tar.* filter=lfs diff=lfs merge=lfs -text
28
+ *.tar filter=lfs diff=lfs merge=lfs -text
29
+ *.tflite filter=lfs diff=lfs merge=lfs -text
30
+ *.tgz filter=lfs diff=lfs merge=lfs -text
31
+ *.wasm filter=lfs diff=lfs merge=lfs -text
32
+ *.xz filter=lfs diff=lfs merge=lfs -text
33
+ *.zip filter=lfs diff=lfs merge=lfs -text
34
+ *.zst filter=lfs diff=lfs merge=lfs -text
35
+ *tfevents* filter=lfs diff=lfs merge=lfs -text
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/README.md ADDED
@@ -0,0 +1,108 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ license: mit
5
+ tags:
6
+ - chemistry
7
+ - SMILES
8
+ - product
9
+ datasets:
10
+ - ORD
11
+ metrics:
12
+ - accuracy
13
+ ---
14
+
15
+ # Model Card for ReactionT5v2-forward
16
+
17
+ This is a ReactionT5 pre-trained to predict the products of reactions. You can use the demo [here](https://huggingface.co/spaces/sagawa/ReactionT5_task_forward).
18
+ This is a ReactionT5 pre-trained to predict the products of reactions and fine-tuned on USPOT_50k's train split. Base model before fine-tuning is [here](https://huggingface.co/sagawa/ReactionT5v2-forward).
19
+
20
+
21
+ ### Model Sources
22
+
23
+ <!-- Provide the basic links for the model. -->
24
+
25
+ - **Repository:** https://github.com/sagawatatsuya/ReactionT5v2
26
+ - **Paper:** https://jcheminf.biomedcentral.com/articles/10.1186/s13321-025-01075-4
27
+ - **Demo:** https://huggingface.co/spaces/sagawa/ReactionT5
28
+
29
+ ## Uses
30
+
31
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
32
+ You can use this model for forward reaction prediction or fine-tune this model with your dataset.
33
+
34
+
35
+ ## How to Get Started with the Model
36
+
37
+ Use the code below to get started with the model.
38
+
39
+ ```python
40
+ from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
41
+
42
+ tokenizer = AutoTokenizer.from_pretrained("sagawa/ReactionT5v2-forward", return_tensors="pt")
43
+ model = AutoModelForSeq2SeqLM.from_pretrained("sagawa/ReactionT5v2-forward")
44
+
45
+ inp = tokenizer('REACTANT:COC(=O)C1=CCCN(C)C1.O.[Al+3].[H-].[Li+].[Na+].[OH-]REAGENT:C1CCOC1', return_tensors='pt')
46
+ output = model.generate(**inp, num_beams=1, num_return_sequences=1, return_dict_in_generate=True, output_scores=True)
47
+ output = tokenizer.decode(output['sequences'][0], skip_special_tokens=True).replace(' ', '').rstrip('.')
48
+ output # 'CN1CCC=C(CO)C1'
49
+ ```
50
+
51
+ ## Training Details
52
+
53
+ ### Training Procedure
54
+
55
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
56
+ We used the [USPTO_MIT dataset](https://drive.google.com/file/d/1NEb5DSwqFfg4gm0P2lPb19cqPt96-OSC/view?usp=drive_link) for model finetuning.
57
+ The command used for training is the following. For more information, please refer to the paper and GitHub repository.
58
+
59
+ ```python
60
+ cd task_forward
61
+ python finetune.py \
62
+ --output_dir='t5' \
63
+ --epochs=50 \
64
+ --lr=2e-5 \
65
+ --batch_size=32 \
66
+ --input_max_len=200 \
67
+ --target_max_len=150 \
68
+ --evaluation_strategy='epoch' \
69
+ --save_strategy='epoch' \
70
+ --logging_strategy='epoch' \
71
+ --save_total_limit=10 \
72
+ --train_data_path='../data/USPTO_MIT/MIT_separated/train.csv' \
73
+ --valid_data_path='../data/USPTO_MIT/MIT_separated/val.csv' \
74
+ --disable_tqdm \
75
+ --model_name_or_path='sagawa/ReactionT5v2-forward'
76
+ ```
77
+
78
+ ### Results
79
+
80
+
81
+ | Model | Training set | Test set | Top-1 [% acc.] | Top-2 [% acc.] | Top-3 [% acc.] | Top-5 [% acc.] |
82
+ |----------------------|---------------------------|----------|----------------|----------------|----------------|----------------|
83
+ | Sequence-to-sequence | USPTO_MIT | USPTO_MIT | 80.3 | 84.7 | 86.2 | 87.5 |
84
+ | WLDN | USPTO_MIT | USPTO_MIT | 80.6 (85.6) | 90.5 | 92.8 | 93.4 |
85
+ | Molecular Transformer| USPTO_MIT | USPTO_MIT | 88.8 | 92.6 | – | 94.4 |
86
+ | T5Chem | USPTO_MIT | USPTO_MIT | 90.4 | 94.2 | – | 96.4 |
87
+ | CompoundT5 | USPTO_MIT | USPTO_MIT | 86.6 | 89.5 | 90.4 | 91.2 |
88
+ | [ReactionT5](https://huggingface.co/sagawa/ReactionT5v2-forward) | - | USPTO_MIT | 92.8 | 95.6 | 96.4 | 97.1 |
89
+ | [ReactionT5 (This model)](https://huggingface.co/sagawa/ReactionT5v2-forward-USPTO_MIT) | USPTO_MIT | USPTO_MIT | 97.5 | 98.6 | 98.8 | 99.0 |
90
+
91
+ Performance comparison of Compound T5, ReactionT5, and other models in product prediction.
92
+
93
+ ## Citation
94
+
95
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
96
+ ```
97
+ @article{Sagawa2025,
98
+ title = {ReactionT5: a pre-trained transformer model for accurate chemical reaction prediction with limited data},
99
+ author = {Sagawa, Tatsuya and Kojima, Ryosuke},
100
+ journal = {Journal of Cheminformatics},
101
+ year = {2025},
102
+ volume = {17},
103
+ number = {1},
104
+ pages = {126},
105
+ doi = {10.1186/s13321-025-01075-4},
106
+ url = {https://doi.org/10.1186/s13321-025-01075-4}
107
+ }
108
+ ```
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/UPSTREAM_URL.txt ADDED
@@ -0,0 +1,2 @@
 
 
 
1
+ https://huggingface.co/sagawa/ReactionT5v2-forward-USPTO_MIT
2
+ revision=260cf4aa7e6f55b1a9db1e328f48701cd3fd7cfc
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/config.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/home/acf15718oa/ReactionT5_neword/forward_reaction_prediction_drop_dup/t5/checkpoint-486000",
3
+ "architectures": [
4
+ "T5ForConditionalGeneration"
5
+ ],
6
+ "classifier_dropout": 0.0,
7
+ "d_ff": 2048,
8
+ "d_kv": 64,
9
+ "d_model": 768,
10
+ "decoder_start_token_id": 0,
11
+ "dense_act_fn": "gelu_new",
12
+ "dropout_rate": 0.1,
13
+ "eos_token_id": 1,
14
+ "eval_beams": 5,
15
+ "feed_forward_proj": "gated-gelu",
16
+ "initializer_factor": 1.0,
17
+ "is_encoder_decoder": true,
18
+ "is_gated_act": true,
19
+ "layer_norm_epsilon": 1e-06,
20
+ "max_length": 150,
21
+ "model_type": "t5",
22
+ "num_decoder_layers": 12,
23
+ "num_heads": 12,
24
+ "num_layers": 12,
25
+ "output_past": true,
26
+ "pad_token_id": 0,
27
+ "relative_attention_max_distance": 128,
28
+ "relative_attention_num_buckets": 32,
29
+ "tie_word_embeddings": false,
30
+ "torch_dtype": "float32",
31
+ "transformers_version": "4.40.2",
32
+ "use_cache": true,
33
+ "vocab_size": 268
34
+ }
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "decoder_start_token_id": 0,
3
+ "eos_token_id": 1,
4
+ "max_length": 150,
5
+ "pad_token_id": 0,
6
+ "transformers_version": "4.40.2"
7
+ }
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/hf_model_info.json ADDED
@@ -0,0 +1,97 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_id": "66be0ec8399b14b08f121766",
3
+ "author": "sagawa",
4
+ "cardData": {
5
+ "datasets": [
6
+ "ORD"
7
+ ],
8
+ "language": [
9
+ "en"
10
+ ],
11
+ "license": "mit",
12
+ "metrics": [
13
+ "accuracy"
14
+ ],
15
+ "tags": [
16
+ "chemistry",
17
+ "SMILES",
18
+ "product"
19
+ ]
20
+ },
21
+ "config": {
22
+ "architectures": [
23
+ "T5ForConditionalGeneration"
24
+ ],
25
+ "model_type": "t5",
26
+ "tokenizer_config": {
27
+ "eos_token": "</s>",
28
+ "pad_token": "<pad>",
29
+ "unk_token": "<unk>"
30
+ }
31
+ },
32
+ "createdAt": "2024-08-15T14:20:56.000Z",
33
+ "disabled": false,
34
+ "downloads": 482,
35
+ "gated": false,
36
+ "id": "sagawa/ReactionT5v2-forward-USPTO_MIT",
37
+ "lastModified": "2025-08-22T10:47:29.000Z",
38
+ "likes": 1,
39
+ "model-index": null,
40
+ "modelId": "sagawa/ReactionT5v2-forward-USPTO_MIT",
41
+ "private": false,
42
+ "safetensors": {
43
+ "parameters": {
44
+ "F32": 198640896
45
+ },
46
+ "total": 198640896
47
+ },
48
+ "sha": "260cf4aa7e6f55b1a9db1e328f48701cd3fd7cfc",
49
+ "siblings": [
50
+ {
51
+ "rfilename": ".gitattributes"
52
+ },
53
+ {
54
+ "rfilename": "README.md"
55
+ },
56
+ {
57
+ "rfilename": "config.json"
58
+ },
59
+ {
60
+ "rfilename": "generation_config.json"
61
+ },
62
+ {
63
+ "rfilename": "model.safetensors"
64
+ },
65
+ {
66
+ "rfilename": "special_tokens_map.json"
67
+ },
68
+ {
69
+ "rfilename": "tokenizer.json"
70
+ },
71
+ {
72
+ "rfilename": "tokenizer_config.json"
73
+ },
74
+ {
75
+ "rfilename": "trainer_state.json"
76
+ },
77
+ {
78
+ "rfilename": "training_args.bin"
79
+ }
80
+ ],
81
+ "spaces": [
82
+ "sagawa/ReactionT5",
83
+ "alpata/ReactionPrediction"
84
+ ],
85
+ "tags": [
86
+ "safetensors",
87
+ "t5",
88
+ "chemistry",
89
+ "SMILES",
90
+ "product",
91
+ "en",
92
+ "dataset:ORD",
93
+ "license:mit",
94
+ "region:us"
95
+ ],
96
+ "usedStorage": 794602336
97
+ }
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/provenance/README.txt ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ ReactionT5v2 forward local inputs
2
+ - checkpoints/: pinned Hugging Face forward model snapshots
3
+ - downloads/: downloaded Google Drive dataset bundles
4
+ - shared_provenance/datasets/: extracted forward dataset bundles
5
+ - shared_provenance/reactiont5v2_forward_split_provenance.json: discovered split-file mapping
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/provenance/ReactionT5v2_source_README.md ADDED
@@ -0,0 +1,237 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # ReactionT5v2
2
+ [![Python](https://img.shields.io/badge/Python-3.11%2B-3776AB?logo=python&logoColor=white&style=flat)](https://www.python.org/downloads/)
3
+ [![Ruff](https://img.shields.io/endpoint?url=https://raw.githubusercontent.com/astral-sh/ruff/main/assets/badge/v2.json)](https://github.com/astral-sh/ruff)
4
+ [![Transformers](https://img.shields.io/badge/transformers-4.40.2-FFD21E?logo=huggingface&logoColor=white&style=flat)](https://huggingface.co/docs/transformers/en/installation)
5
+ [![License: MIT](https://img.shields.io/badge/License-MIT-green.svg)](https://github.com/sagawatatsuya/ReactionT5v2/blob/main/LICENSE.txt)
6
+ [![DOI](https://zenodo.org/badge/DOI/10.1186/s13321-025-01075-4.svg)](https://doi.org/10.1186/s13321-025-01075-4)
7
+ [![GoogleColab](https://img.shields.io/badge/Google%20Colab-Demo%20Notebook-F9AB00?logo=googlecolab&logoColor=F9AB00&style=flat)](https://colab.research.google.com/drive/10Hkx8YJTG5JGXj73XfnEYxZ8fDnsJRcs?usp=sharing)
8
+
9
+ ReactionT5 is a T5 model pretrained on a vast array of chemical reactions from the [Open Reaction Database (ORD)](https://github.com/open-reaction-database/ord-data). Unlike other models that are trained on smaller, potentially biased datasets (e.g. patent datasets or high-throughput reaction datasets created by a single reaction), ReactionT5 leverages the extyensive and diverse dataset provided by ORD. This ensures greater generalizability and performance, enabling ReactionT5 to condact product, retrosynthesis, and yield prediction of unseen chemical reactions with high accuracy. This makes it highly practical for real-world applications.
10
+
11
+ ![model image](https://github.com/sagawatatsuya/ReactionT5v2/blob/main/model-image.png)
12
+
13
+
14
+ In this repository, we will demonstrate how to use ReactionT5 for product prediction, retrosynthesis prediction, and yield prediction on your own datasets. The pretrained models and demo is available at [Hugging Face Hub](https://huggingface.co/collections/sagawa/reactiont5-67dbe0550cbb6886a85e348b) or [Google Colab](https://colab.research.google.com/drive/10Hkx8YJTG5JGXj73XfnEYxZ8fDnsJRcs?usp=sharing)
15
+
16
+ # Table of Contents
17
+ - [ReactionT5v2](#reactiont5v2)
18
+ - [Setup](#setup)
19
+ - [Dataset](#dataset)
20
+ - [Usage](#usage)
21
+ - [Fine-tuning](#fine-tuning)
22
+ - [Compound Pre-training](#compound-pre-training)
23
+ - [Reaction Pre-training](#reaction-pre-training)
24
+ - [Structure](#structure)
25
+ - [Authors](#authors)
26
+ - [Citation](#citation)
27
+
28
+
29
+ # Setup
30
+ ReactionT5 is based on the transformers library. Additionally, RDKit is used for validity check of predicted compounds. To install these and other necessary libraries, use the following commands:
31
+ ```
32
+ pip install rdkit
33
+ pip install torch
34
+ pip install tokenizers==0.19.1
35
+ pip install transformers==4.40.2
36
+ pip install datasets
37
+ pip install accelerate -U
38
+ pip install sentencepiece
39
+ ```
40
+
41
+ # Dataset
42
+ For model training and finetuning, we used the ORD dataset, USPTO_MIT dataset, USPTO_50k dataset, and C-N cross-coupling reactions dataset. Each dataset can be downloaded from the following links:
43
+ - [ORD](https://drive.google.com/file/d/1JozA2OlByfZ-ILt5H5YrTjLJvSvD8xdL/view?usp=drive_link)
44
+ - [USPTO_MIT](https://drive.google.com/file/d/1NEb5DSwqFfg4gm0P2lPb19cqPt96-OSC/view?usp=drive_link)
45
+ - [USPTO_50k](https://drive.google.com/file/d/15-E4eaxsUJ_aKxX0mnOrvYCTWKpqrLvf/view?usp=drive_link)
46
+ - [Buchwald-Hartwig C-N cross-coupling](https://drive.google.com/file/d/1qvsJk1_q1yfJvBYtEMGOcCuk2pRvsWvy/view?usp=drive_link)
47
+
48
+
49
+ # Usage
50
+ You can use ReactionT5 for product prediction, retrosynthesis prediction, and yield prediction.
51
+
52
+ ### Task: Forward
53
+ To predict the products of reactions from their inputs, use the following command. The code expects 'input_data' as a CSV file that contains columns named 'REACTANT', 'REAGENT', and 'PRODUCT'; each has SMILES information. If there is no reagent information, fill in the blank with ' '. For multiple compounds, concatenate them with ".".
54
+ ```
55
+ cd task_forward
56
+ python prediction.py \
57
+ --input_data="../data/demo_reaction_data.csv" \
58
+ --model_name_or_path="sagawa/ReactionT5v2-forward" \
59
+ --input_max_length=150 \
60
+ --num_beams=5 \
61
+ --num_return_sequences=5 \
62
+ --batch_size=16 \
63
+ --output_dir="output"
64
+ ```
65
+
66
+ ### Task: Retrosynthesis
67
+ To predict the reactants of reactions from their products, use the following command. The code expects 'input_data' as a CSV file that contains a "PRODUCT" column. The format of the contents of the column should be SMILES of each compound. For multiple compounds, concatenate them with ".".
68
+ ```
69
+ cd task_retrosynthesis
70
+ python prediction.py \
71
+ --input_data="../data/demo_reaction_data.csv" \
72
+ --model_name_or_path="sagawa/ReactionT5v2-retrosynthesis" \
73
+ --input_max_length=100 \
74
+ --num_beams=5 \
75
+ --num_return_sequences=5 \
76
+ --batch_size=16 \
77
+ --output_dir="output"
78
+ ```
79
+
80
+ ### Task: Yield
81
+ To predict the yields of reactions from their inputs, use the following command. The code expects 'input_data' as a CSV file that contains columns named 'REACTANT', 'REAGENT', 'PRODUCT', and 'YIELD'; except 'YIELD' have SMILES information, and 'YIELD' has numerical information. If there is no reagent information, fill in the blank with ' '. For multiple compounds, concatenate them with ".".
82
+ ```
83
+ cd task_yield
84
+ python prediction_with_PreTrainedModel.py \
85
+ --input_data="../data/demo_reaction_data.csv" \
86
+ --model_name_or_path="sagawa/ReactionT5v2-yield" \
87
+ --batch_size=16 \
88
+ --output_dir="output"
89
+ ```
90
+
91
+
92
+ # Fine-tuning
93
+ If your dataset is very specific and different from ORD's data, ReactionT5 may not predict well. In that case, you can fine-tune ReactionT5 on your dataset. From our study, ReactionT5's performance drastically improved its performance by fine-tuning using relatively small data (100 reactions).
94
+
95
+ ### Task: Forward
96
+ Specify your training and validation data used for fine-tuning and run the following command. The code expects train and valid data that contain columns named 'REACTANT', 'REAGENT', and 'PRODUCT'; each has SMILES information. If there is no reagent information, fill in the blank with ' '. For multiple compounds, concatenate them with ".".
97
+ ```
98
+ cd task_forward
99
+ python finetune.py \
100
+ --model_name_or_path='sagawa/ReactionT5v2-forward' \
101
+ --epochs=50 \
102
+ --batch_size=32 \
103
+ --train_data_path='../data/demo_reaction_data.csv' \
104
+ --valid_data_path='../data/demo_reaction_data.csv' \
105
+ --output_dir="output"
106
+ ```
107
+
108
+ ### Task: Retrosynthesis
109
+ Specify your training and validation data used for fine-tuning and run the following command. The code expects train and valid data that contain columns named 'REACTANT' and 'PRODUCT'; each has SMILES information. For multiple compounds, concatenate them with ".".
110
+ ```
111
+ cd task_retrosynthesis
112
+ python finetune.py \
113
+ --model_name_or_path='sagawa/ReactionT5v2-retrosynthesis' \
114
+ --epochs=20 \
115
+ --batch_size=32 \
116
+ --train_data_path='../data/demo_reaction_data.csv' \
117
+ --valid_data_path='../data/demo_reaction_data.csv' \
118
+ --output_dir='output'
119
+ ```
120
+
121
+ ### Task: Yield
122
+ Specify your training and validation data used for fine-tuning and run the following command. The code expects train and valid data that contain columns named 'REACTANT', 'REAGENT', 'PRODUCT', and 'YIELD'; except 'YIELD ' have SMILES information, and 'YIELD' has numerical information. If there is no reagent information, fill in the blank with ' '. For multiple compounds, concatenate them with ".".
123
+ ```
124
+ cd task_yield
125
+ python finetune.py \
126
+ --epochs=200 \
127
+ --batch_size=32 \
128
+ --train_data_path='../data/demo_reaction_data.csv' \
129
+ --valid_data_path='../data/demo_reaction_data.csv' \
130
+ --download_pretrained_model \
131
+ --output_dir='output'
132
+ ```
133
+
134
+ # Compound Pre-training
135
+ If you want to retrain CompoundT5 (compound pre-training), please refer to the [README.md](./CompoundT5/README.md) file located inside the CompoundT5 directory.
136
+
137
+
138
+ # Reaction Pre-training
139
+ If you want to retrain ReactionT5 from CompoundT5 (reaction pre-training), you can do so by running the following commands. These will train ReactionT5 on the Open Reaction Database (ORD) dataset.
140
+
141
+ ### Dataset
142
+ Download the preprocessed ORD dataset from [here](https://drive.google.com/file/d/1JozA2OlByfZ-ILt5H5YrTjLJvSvD8xdL/view?usp=drive_link) and place it in the data/ORD directory.
143
+ To prevent data leakage in downstream benchmarks, we specify either 'USPTO_test_data_path' or 'CN_test_data_path' in the training commands below and remove any duplicated reactions from the training data. You can also train ReactionT5 on full ORD data without specifying these parameters.
144
+ The corresponding test datasets can be downloaded from the following links:
145
+ - [USPTO_MIT](https://drive.google.com/file/d/1NEb5DSwqFfg4gm0P2lPb19cqPt96-OSC/view?usp=drive_link)
146
+ - [USPTO_50k](https://drive.google.com/file/d/15-E4eaxsUJ_aKxX0mnOrvYCTWKpqrLvf/view?usp=drive_link)
147
+ - [Buchwald-Hartwig C-N cross-coupling](https://drive.google.com/file/d/1qvsJk1_q1yfJvBYtEMGOcCuk2pRvsWvy/view?usp=drive_link)
148
+
149
+ ### Task: Forward
150
+ ```
151
+ cd task_forward
152
+ python train.py \
153
+ --output_dir='ReactionT5_forward' \
154
+ --epochs=100 \
155
+ --lr=1e-3 \
156
+ --batch_size=32 \
157
+ --input_max_len=150 \
158
+ --target_max_len=100 \
159
+ --weight_decay=0.01 \
160
+ --evaluation_strategy='epoch' \
161
+ --save_strategy='epoch' \
162
+ --logging_strategy='epoch' \
163
+ --save_total_limit=100 \
164
+ --train_data_path='../data/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv' \
165
+ --valid_data_path='../data/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv' \
166
+ --test_data_path='../data/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv' \
167
+ --USPTO_test_data_path='../data/USPTO_MIT/MIT_separated/test.csv' \
168
+ --disable_tqdm \
169
+ --pretrained_model_name_or_path='sagawa/CompoundT5'
170
+ ```
171
+ ### Task: Retrosynthesis
172
+ ```
173
+ cd task_retrosynthesis
174
+ python train.py \
175
+ --output_dir='ReactionT5_retrosynthesis' \
176
+ --epochs=100 \
177
+ --lr=2e-4 \
178
+ --batch_size=32 \
179
+ --input_max_len=100 \
180
+ --target_max_len=150 \
181
+ --weight_decay=0.01 \
182
+ --evaluation_strategy='epoch' \
183
+ --save_strategy='epoch' \
184
+ --logging_strategy='epoch' \
185
+ --save_total_limit=100 \
186
+ --train_data_path='../data/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv' \
187
+ --valid_data_path='../data/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv' \
188
+ --test_data_path='../data/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv' \
189
+ --USPTO_test_data_path='../data/USPTO_50k/test.csv' \
190
+ --disable_tqdm \
191
+ --pretrained_model_name_or_path='sagawa/CompoundT5'
192
+ ```
193
+ ### Task: Yield
194
+ ```
195
+ cd task_yield
196
+ python train.py \
197
+ --output_dir='ReactionT5_yield_CN_test1' \
198
+ --train_data_path='../data/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv' \
199
+ --valid_data_path='../data/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv' \
200
+ --test_data_path='../data/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv' \
201
+ --CN_test_data_path='../data/C_N_yield/MFF_Test1/test.csv' \
202
+ --epochs=100 \
203
+ --input_max_length=300 \
204
+ --pretrained_model_name_or_path='sagawa/CompoundT5' \
205
+ --batch_size=32
206
+ ```
207
+
208
+
209
+ # Structure
210
+ ```
211
+ ReactionT5v2/
212
+ ├── CompoundT5/ # Codes used for data processing and compound pretraining
213
+ │ └── README.md # More detailed README file for CompoundT5 creation
214
+ ├── data/ # Datasets
215
+ ├── task_forward/ # Forward prediction and finetuning
216
+ ├── task_retrosynthesis/ # Retrosynthesis prediction and finetuning
217
+ ├── task_yield/ # Yield prediction and finetuning
218
+ └── README.md # This README file
219
+ ```
220
+
221
+
222
+ # Authors
223
+ Tatsuya Sagawa, Ryosuke Kojima
224
+
225
+ # Citation
226
+ ```bibtex
227
+ @article{Sagawa2025,
228
+ title = {ReactionT5: a pre-trained transformer model for accurate chemical reaction prediction with limited data},
229
+ author = {Sagawa, Tatsuya and Kojima, Ryosuke},
230
+ journal = {Journal of Cheminformatics},
231
+ year = {2025},
232
+ volume = {17},
233
+ number = {1},
234
+ pages = {126},
235
+ doi = {10.1186/s13321-025-01075-4},
236
+ url = {https://doi.org/10.1186/s13321-025-01075-4}
237
+ }
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/provenance/reactiont5v2_forward_split_provenance.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "checkpoints": [
3
+ {
4
+ "local_dir": "checkpoints/sagawa_reactiont5v2-forward",
5
+ "repo_id": "sagawa/ReactionT5v2-forward",
6
+ "revision": "933114058cb2604dc1bf536dbebdfcefbe83d4fc",
7
+ "training_dataset_ref": "ord_forward_pretraining"
8
+ },
9
+ {
10
+ "local_dir": "checkpoints/sagawa_reactiont5v2-forward-uspto_mit",
11
+ "repo_id": "sagawa/ReactionT5v2-forward-USPTO_MIT",
12
+ "revision": "260cf4aa7e6f55b1a9db1e328f48701cd3fd7cfc",
13
+ "training_dataset_ref": "uspto_mit_forward_project_release"
14
+ }
15
+ ],
16
+ "datasets": {
17
+ "ord": {
18
+ "bundle": "downloads/ORD.zip",
19
+ "expected_split_filenames": [
20
+ "all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
21
+ "all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
22
+ "all_ord_reaction_uniq_with_attr20240506_v3_test.csv"
23
+ ],
24
+ "extract_dir": "shared_provenance/datasets/ORD",
25
+ "matched_files": {
26
+ "all_ord_reaction_uniq_with_attr20240506_v3_test.csv": [
27
+ "all_ord_reaction_uniq_with_attr20240506_v3_test.csv"
28
+ ],
29
+ "all_ord_reaction_uniq_with_attr20240506_v3_train.csv": [
30
+ "all_ord_reaction_uniq_with_attr20240506_v3_train.csv"
31
+ ],
32
+ "all_ord_reaction_uniq_with_attr20240506_v3_valid.csv": [
33
+ "all_ord_reaction_uniq_with_attr20240506_v3_valid.csv"
34
+ ]
35
+ },
36
+ "notes": "Forward-project ORD train/valid/test CSVs extracted by suffix match from the upstream ORD Google Drive bundle."
37
+ },
38
+ "uspto_mit": {
39
+ "bundle": "downloads/USPTO_MIT.zip",
40
+ "expected_split_filenames": [
41
+ "train.csv",
42
+ "val.csv",
43
+ "test.csv"
44
+ ],
45
+ "extract_dir": "shared_provenance/datasets/USPTO_MIT/MIT_separated",
46
+ "matched_files": {
47
+ "test.csv": [
48
+ "MIT_separated/test.csv"
49
+ ],
50
+ "train.csv": [
51
+ "MIT_separated/train.csv"
52
+ ],
53
+ "val.csv": [
54
+ "MIT_separated/val.csv"
55
+ ]
56
+ },
57
+ "notes": "Upstream USPTO_MIT bundle uses val.csv (not valid.csv); files were extracted by suffix match from the project Google Drive bundle."
58
+ }
59
+ },
60
+ "generated_at": "2026-06-27T14:17:00.291340+00:00",
61
+ "source_repo_readme": "shared_provenance/ReactionT5v2_source_README.md"
62
+ }
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/special_tokens_map.json ADDED
@@ -0,0 +1,127 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<extra_id_0>",
4
+ "<extra_id_1>",
5
+ "<extra_id_2>",
6
+ "<extra_id_3>",
7
+ "<extra_id_4>",
8
+ "<extra_id_5>",
9
+ "<extra_id_6>",
10
+ "<extra_id_7>",
11
+ "<extra_id_8>",
12
+ "<extra_id_9>",
13
+ "<extra_id_10>",
14
+ "<extra_id_11>",
15
+ "<extra_id_12>",
16
+ "<extra_id_13>",
17
+ "<extra_id_14>",
18
+ "<extra_id_15>",
19
+ "<extra_id_16>",
20
+ "<extra_id_17>",
21
+ "<extra_id_18>",
22
+ "<extra_id_19>",
23
+ "<extra_id_20>",
24
+ "<extra_id_21>",
25
+ "<extra_id_22>",
26
+ "<extra_id_23>",
27
+ "<extra_id_24>",
28
+ "<extra_id_25>",
29
+ "<extra_id_26>",
30
+ "<extra_id_27>",
31
+ "<extra_id_28>",
32
+ "<extra_id_29>",
33
+ "<extra_id_30>",
34
+ "<extra_id_31>",
35
+ "<extra_id_32>",
36
+ "<extra_id_33>",
37
+ "<extra_id_34>",
38
+ "<extra_id_35>",
39
+ "<extra_id_36>",
40
+ "<extra_id_37>",
41
+ "<extra_id_38>",
42
+ "<extra_id_39>",
43
+ "<extra_id_40>",
44
+ "<extra_id_41>",
45
+ "<extra_id_42>",
46
+ "<extra_id_43>",
47
+ "<extra_id_44>",
48
+ "<extra_id_45>",
49
+ "<extra_id_46>",
50
+ "<extra_id_47>",
51
+ "<extra_id_48>",
52
+ "<extra_id_49>",
53
+ "<extra_id_50>",
54
+ "<extra_id_51>",
55
+ "<extra_id_52>",
56
+ "<extra_id_53>",
57
+ "<extra_id_54>",
58
+ "<extra_id_55>",
59
+ "<extra_id_56>",
60
+ "<extra_id_57>",
61
+ "<extra_id_58>",
62
+ "<extra_id_59>",
63
+ "<extra_id_60>",
64
+ "<extra_id_61>",
65
+ "<extra_id_62>",
66
+ "<extra_id_63>",
67
+ "<extra_id_64>",
68
+ "<extra_id_65>",
69
+ "<extra_id_66>",
70
+ "<extra_id_67>",
71
+ "<extra_id_68>",
72
+ "<extra_id_69>",
73
+ "<extra_id_70>",
74
+ "<extra_id_71>",
75
+ "<extra_id_72>",
76
+ "<extra_id_73>",
77
+ "<extra_id_74>",
78
+ "<extra_id_75>",
79
+ "<extra_id_76>",
80
+ "<extra_id_77>",
81
+ "<extra_id_78>",
82
+ "<extra_id_79>",
83
+ "<extra_id_80>",
84
+ "<extra_id_81>",
85
+ "<extra_id_82>",
86
+ "<extra_id_83>",
87
+ "<extra_id_84>",
88
+ "<extra_id_85>",
89
+ "<extra_id_86>",
90
+ "<extra_id_87>",
91
+ "<extra_id_88>",
92
+ "<extra_id_89>",
93
+ "<extra_id_90>",
94
+ "<extra_id_91>",
95
+ "<extra_id_92>",
96
+ "<extra_id_93>",
97
+ "<extra_id_94>",
98
+ "<extra_id_95>",
99
+ "<extra_id_96>",
100
+ "<extra_id_97>",
101
+ "<extra_id_98>",
102
+ "<extra_id_99>",
103
+ "REACTANT:",
104
+ "REAGENT:"
105
+ ],
106
+ "eos_token": {
107
+ "content": "</s>",
108
+ "lstrip": false,
109
+ "normalized": false,
110
+ "rstrip": false,
111
+ "single_word": false
112
+ },
113
+ "pad_token": {
114
+ "content": "<pad>",
115
+ "lstrip": false,
116
+ "normalized": false,
117
+ "rstrip": false,
118
+ "single_word": false
119
+ },
120
+ "unk_token": {
121
+ "content": "<unk>",
122
+ "lstrip": false,
123
+ "normalized": false,
124
+ "rstrip": false,
125
+ "single_word": false
126
+ }
127
+ }
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/tokenizer.json ADDED
@@ -0,0 +1,1947 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "version": "1.0",
3
+ "truncation": {
4
+ "direction": "Right",
5
+ "max_length": 150,
6
+ "strategy": "LongestFirst",
7
+ "stride": 0
8
+ },
9
+ "padding": null,
10
+ "added_tokens": [
11
+ {
12
+ "id": 0,
13
+ "content": "<pad>",
14
+ "single_word": false,
15
+ "lstrip": false,
16
+ "rstrip": false,
17
+ "normalized": false,
18
+ "special": true
19
+ },
20
+ {
21
+ "id": 1,
22
+ "content": "</s>",
23
+ "single_word": false,
24
+ "lstrip": false,
25
+ "rstrip": false,
26
+ "normalized": false,
27
+ "special": true
28
+ },
29
+ {
30
+ "id": 2,
31
+ "content": "<unk>",
32
+ "single_word": false,
33
+ "lstrip": false,
34
+ "rstrip": false,
35
+ "normalized": false,
36
+ "special": true
37
+ },
38
+ {
39
+ "id": 36,
40
+ "content": "Cl",
41
+ "single_word": false,
42
+ "lstrip": false,
43
+ "rstrip": false,
44
+ "normalized": true,
45
+ "special": false
46
+ },
47
+ {
48
+ "id": 121,
49
+ "content": "<extra_id_0>",
50
+ "single_word": false,
51
+ "lstrip": false,
52
+ "rstrip": false,
53
+ "normalized": false,
54
+ "special": true
55
+ },
56
+ {
57
+ "id": 122,
58
+ "content": "<extra_id_1>",
59
+ "single_word": false,
60
+ "lstrip": false,
61
+ "rstrip": false,
62
+ "normalized": false,
63
+ "special": true
64
+ },
65
+ {
66
+ "id": 123,
67
+ "content": "<extra_id_2>",
68
+ "single_word": false,
69
+ "lstrip": false,
70
+ "rstrip": false,
71
+ "normalized": false,
72
+ "special": true
73
+ },
74
+ {
75
+ "id": 124,
76
+ "content": "<extra_id_3>",
77
+ "single_word": false,
78
+ "lstrip": false,
79
+ "rstrip": false,
80
+ "normalized": false,
81
+ "special": true
82
+ },
83
+ {
84
+ "id": 125,
85
+ "content": "<extra_id_4>",
86
+ "single_word": false,
87
+ "lstrip": false,
88
+ "rstrip": false,
89
+ "normalized": false,
90
+ "special": true
91
+ },
92
+ {
93
+ "id": 126,
94
+ "content": "<extra_id_5>",
95
+ "single_word": false,
96
+ "lstrip": false,
97
+ "rstrip": false,
98
+ "normalized": false,
99
+ "special": true
100
+ },
101
+ {
102
+ "id": 127,
103
+ "content": "<extra_id_6>",
104
+ "single_word": false,
105
+ "lstrip": false,
106
+ "rstrip": false,
107
+ "normalized": false,
108
+ "special": true
109
+ },
110
+ {
111
+ "id": 128,
112
+ "content": "<extra_id_7>",
113
+ "single_word": false,
114
+ "lstrip": false,
115
+ "rstrip": false,
116
+ "normalized": false,
117
+ "special": true
118
+ },
119
+ {
120
+ "id": 129,
121
+ "content": "<extra_id_8>",
122
+ "single_word": false,
123
+ "lstrip": false,
124
+ "rstrip": false,
125
+ "normalized": false,
126
+ "special": true
127
+ },
128
+ {
129
+ "id": 130,
130
+ "content": "<extra_id_9>",
131
+ "single_word": false,
132
+ "lstrip": false,
133
+ "rstrip": false,
134
+ "normalized": false,
135
+ "special": true
136
+ },
137
+ {
138
+ "id": 131,
139
+ "content": "<extra_id_10>",
140
+ "single_word": false,
141
+ "lstrip": false,
142
+ "rstrip": false,
143
+ "normalized": false,
144
+ "special": true
145
+ },
146
+ {
147
+ "id": 132,
148
+ "content": "<extra_id_11>",
149
+ "single_word": false,
150
+ "lstrip": false,
151
+ "rstrip": false,
152
+ "normalized": false,
153
+ "special": true
154
+ },
155
+ {
156
+ "id": 133,
157
+ "content": "<extra_id_12>",
158
+ "single_word": false,
159
+ "lstrip": false,
160
+ "rstrip": false,
161
+ "normalized": false,
162
+ "special": true
163
+ },
164
+ {
165
+ "id": 134,
166
+ "content": "<extra_id_13>",
167
+ "single_word": false,
168
+ "lstrip": false,
169
+ "rstrip": false,
170
+ "normalized": false,
171
+ "special": true
172
+ },
173
+ {
174
+ "id": 135,
175
+ "content": "<extra_id_14>",
176
+ "single_word": false,
177
+ "lstrip": false,
178
+ "rstrip": false,
179
+ "normalized": false,
180
+ "special": true
181
+ },
182
+ {
183
+ "id": 136,
184
+ "content": "<extra_id_15>",
185
+ "single_word": false,
186
+ "lstrip": false,
187
+ "rstrip": false,
188
+ "normalized": false,
189
+ "special": true
190
+ },
191
+ {
192
+ "id": 137,
193
+ "content": "<extra_id_16>",
194
+ "single_word": false,
195
+ "lstrip": false,
196
+ "rstrip": false,
197
+ "normalized": false,
198
+ "special": true
199
+ },
200
+ {
201
+ "id": 138,
202
+ "content": "<extra_id_17>",
203
+ "single_word": false,
204
+ "lstrip": false,
205
+ "rstrip": false,
206
+ "normalized": false,
207
+ "special": true
208
+ },
209
+ {
210
+ "id": 139,
211
+ "content": "<extra_id_18>",
212
+ "single_word": false,
213
+ "lstrip": false,
214
+ "rstrip": false,
215
+ "normalized": false,
216
+ "special": true
217
+ },
218
+ {
219
+ "id": 140,
220
+ "content": "<extra_id_19>",
221
+ "single_word": false,
222
+ "lstrip": false,
223
+ "rstrip": false,
224
+ "normalized": false,
225
+ "special": true
226
+ },
227
+ {
228
+ "id": 141,
229
+ "content": "<extra_id_20>",
230
+ "single_word": false,
231
+ "lstrip": false,
232
+ "rstrip": false,
233
+ "normalized": false,
234
+ "special": true
235
+ },
236
+ {
237
+ "id": 142,
238
+ "content": "<extra_id_21>",
239
+ "single_word": false,
240
+ "lstrip": false,
241
+ "rstrip": false,
242
+ "normalized": false,
243
+ "special": true
244
+ },
245
+ {
246
+ "id": 143,
247
+ "content": "<extra_id_22>",
248
+ "single_word": false,
249
+ "lstrip": false,
250
+ "rstrip": false,
251
+ "normalized": false,
252
+ "special": true
253
+ },
254
+ {
255
+ "id": 144,
256
+ "content": "<extra_id_23>",
257
+ "single_word": false,
258
+ "lstrip": false,
259
+ "rstrip": false,
260
+ "normalized": false,
261
+ "special": true
262
+ },
263
+ {
264
+ "id": 145,
265
+ "content": "<extra_id_24>",
266
+ "single_word": false,
267
+ "lstrip": false,
268
+ "rstrip": false,
269
+ "normalized": false,
270
+ "special": true
271
+ },
272
+ {
273
+ "id": 146,
274
+ "content": "<extra_id_25>",
275
+ "single_word": false,
276
+ "lstrip": false,
277
+ "rstrip": false,
278
+ "normalized": false,
279
+ "special": true
280
+ },
281
+ {
282
+ "id": 147,
283
+ "content": "<extra_id_26>",
284
+ "single_word": false,
285
+ "lstrip": false,
286
+ "rstrip": false,
287
+ "normalized": false,
288
+ "special": true
289
+ },
290
+ {
291
+ "id": 148,
292
+ "content": "<extra_id_27>",
293
+ "single_word": false,
294
+ "lstrip": false,
295
+ "rstrip": false,
296
+ "normalized": false,
297
+ "special": true
298
+ },
299
+ {
300
+ "id": 149,
301
+ "content": "<extra_id_28>",
302
+ "single_word": false,
303
+ "lstrip": false,
304
+ "rstrip": false,
305
+ "normalized": false,
306
+ "special": true
307
+ },
308
+ {
309
+ "id": 150,
310
+ "content": "<extra_id_29>",
311
+ "single_word": false,
312
+ "lstrip": false,
313
+ "rstrip": false,
314
+ "normalized": false,
315
+ "special": true
316
+ },
317
+ {
318
+ "id": 151,
319
+ "content": "<extra_id_30>",
320
+ "single_word": false,
321
+ "lstrip": false,
322
+ "rstrip": false,
323
+ "normalized": false,
324
+ "special": true
325
+ },
326
+ {
327
+ "id": 152,
328
+ "content": "<extra_id_31>",
329
+ "single_word": false,
330
+ "lstrip": false,
331
+ "rstrip": false,
332
+ "normalized": false,
333
+ "special": true
334
+ },
335
+ {
336
+ "id": 153,
337
+ "content": "<extra_id_32>",
338
+ "single_word": false,
339
+ "lstrip": false,
340
+ "rstrip": false,
341
+ "normalized": false,
342
+ "special": true
343
+ },
344
+ {
345
+ "id": 154,
346
+ "content": "<extra_id_33>",
347
+ "single_word": false,
348
+ "lstrip": false,
349
+ "rstrip": false,
350
+ "normalized": false,
351
+ "special": true
352
+ },
353
+ {
354
+ "id": 155,
355
+ "content": "<extra_id_34>",
356
+ "single_word": false,
357
+ "lstrip": false,
358
+ "rstrip": false,
359
+ "normalized": false,
360
+ "special": true
361
+ },
362
+ {
363
+ "id": 156,
364
+ "content": "<extra_id_35>",
365
+ "single_word": false,
366
+ "lstrip": false,
367
+ "rstrip": false,
368
+ "normalized": false,
369
+ "special": true
370
+ },
371
+ {
372
+ "id": 157,
373
+ "content": "<extra_id_36>",
374
+ "single_word": false,
375
+ "lstrip": false,
376
+ "rstrip": false,
377
+ "normalized": false,
378
+ "special": true
379
+ },
380
+ {
381
+ "id": 158,
382
+ "content": "<extra_id_37>",
383
+ "single_word": false,
384
+ "lstrip": false,
385
+ "rstrip": false,
386
+ "normalized": false,
387
+ "special": true
388
+ },
389
+ {
390
+ "id": 159,
391
+ "content": "<extra_id_38>",
392
+ "single_word": false,
393
+ "lstrip": false,
394
+ "rstrip": false,
395
+ "normalized": false,
396
+ "special": true
397
+ },
398
+ {
399
+ "id": 160,
400
+ "content": "<extra_id_39>",
401
+ "single_word": false,
402
+ "lstrip": false,
403
+ "rstrip": false,
404
+ "normalized": false,
405
+ "special": true
406
+ },
407
+ {
408
+ "id": 161,
409
+ "content": "<extra_id_40>",
410
+ "single_word": false,
411
+ "lstrip": false,
412
+ "rstrip": false,
413
+ "normalized": false,
414
+ "special": true
415
+ },
416
+ {
417
+ "id": 162,
418
+ "content": "<extra_id_41>",
419
+ "single_word": false,
420
+ "lstrip": false,
421
+ "rstrip": false,
422
+ "normalized": false,
423
+ "special": true
424
+ },
425
+ {
426
+ "id": 163,
427
+ "content": "<extra_id_42>",
428
+ "single_word": false,
429
+ "lstrip": false,
430
+ "rstrip": false,
431
+ "normalized": false,
432
+ "special": true
433
+ },
434
+ {
435
+ "id": 164,
436
+ "content": "<extra_id_43>",
437
+ "single_word": false,
438
+ "lstrip": false,
439
+ "rstrip": false,
440
+ "normalized": false,
441
+ "special": true
442
+ },
443
+ {
444
+ "id": 165,
445
+ "content": "<extra_id_44>",
446
+ "single_word": false,
447
+ "lstrip": false,
448
+ "rstrip": false,
449
+ "normalized": false,
450
+ "special": true
451
+ },
452
+ {
453
+ "id": 166,
454
+ "content": "<extra_id_45>",
455
+ "single_word": false,
456
+ "lstrip": false,
457
+ "rstrip": false,
458
+ "normalized": false,
459
+ "special": true
460
+ },
461
+ {
462
+ "id": 167,
463
+ "content": "<extra_id_46>",
464
+ "single_word": false,
465
+ "lstrip": false,
466
+ "rstrip": false,
467
+ "normalized": false,
468
+ "special": true
469
+ },
470
+ {
471
+ "id": 168,
472
+ "content": "<extra_id_47>",
473
+ "single_word": false,
474
+ "lstrip": false,
475
+ "rstrip": false,
476
+ "normalized": false,
477
+ "special": true
478
+ },
479
+ {
480
+ "id": 169,
481
+ "content": "<extra_id_48>",
482
+ "single_word": false,
483
+ "lstrip": false,
484
+ "rstrip": false,
485
+ "normalized": false,
486
+ "special": true
487
+ },
488
+ {
489
+ "id": 170,
490
+ "content": "<extra_id_49>",
491
+ "single_word": false,
492
+ "lstrip": false,
493
+ "rstrip": false,
494
+ "normalized": false,
495
+ "special": true
496
+ },
497
+ {
498
+ "id": 171,
499
+ "content": "<extra_id_50>",
500
+ "single_word": false,
501
+ "lstrip": false,
502
+ "rstrip": false,
503
+ "normalized": false,
504
+ "special": true
505
+ },
506
+ {
507
+ "id": 172,
508
+ "content": "<extra_id_51>",
509
+ "single_word": false,
510
+ "lstrip": false,
511
+ "rstrip": false,
512
+ "normalized": false,
513
+ "special": true
514
+ },
515
+ {
516
+ "id": 173,
517
+ "content": "<extra_id_52>",
518
+ "single_word": false,
519
+ "lstrip": false,
520
+ "rstrip": false,
521
+ "normalized": false,
522
+ "special": true
523
+ },
524
+ {
525
+ "id": 174,
526
+ "content": "<extra_id_53>",
527
+ "single_word": false,
528
+ "lstrip": false,
529
+ "rstrip": false,
530
+ "normalized": false,
531
+ "special": true
532
+ },
533
+ {
534
+ "id": 175,
535
+ "content": "<extra_id_54>",
536
+ "single_word": false,
537
+ "lstrip": false,
538
+ "rstrip": false,
539
+ "normalized": false,
540
+ "special": true
541
+ },
542
+ {
543
+ "id": 176,
544
+ "content": "<extra_id_55>",
545
+ "single_word": false,
546
+ "lstrip": false,
547
+ "rstrip": false,
548
+ "normalized": false,
549
+ "special": true
550
+ },
551
+ {
552
+ "id": 177,
553
+ "content": "<extra_id_56>",
554
+ "single_word": false,
555
+ "lstrip": false,
556
+ "rstrip": false,
557
+ "normalized": false,
558
+ "special": true
559
+ },
560
+ {
561
+ "id": 178,
562
+ "content": "<extra_id_57>",
563
+ "single_word": false,
564
+ "lstrip": false,
565
+ "rstrip": false,
566
+ "normalized": false,
567
+ "special": true
568
+ },
569
+ {
570
+ "id": 179,
571
+ "content": "<extra_id_58>",
572
+ "single_word": false,
573
+ "lstrip": false,
574
+ "rstrip": false,
575
+ "normalized": false,
576
+ "special": true
577
+ },
578
+ {
579
+ "id": 180,
580
+ "content": "<extra_id_59>",
581
+ "single_word": false,
582
+ "lstrip": false,
583
+ "rstrip": false,
584
+ "normalized": false,
585
+ "special": true
586
+ },
587
+ {
588
+ "id": 181,
589
+ "content": "<extra_id_60>",
590
+ "single_word": false,
591
+ "lstrip": false,
592
+ "rstrip": false,
593
+ "normalized": false,
594
+ "special": true
595
+ },
596
+ {
597
+ "id": 182,
598
+ "content": "<extra_id_61>",
599
+ "single_word": false,
600
+ "lstrip": false,
601
+ "rstrip": false,
602
+ "normalized": false,
603
+ "special": true
604
+ },
605
+ {
606
+ "id": 183,
607
+ "content": "<extra_id_62>",
608
+ "single_word": false,
609
+ "lstrip": false,
610
+ "rstrip": false,
611
+ "normalized": false,
612
+ "special": true
613
+ },
614
+ {
615
+ "id": 184,
616
+ "content": "<extra_id_63>",
617
+ "single_word": false,
618
+ "lstrip": false,
619
+ "rstrip": false,
620
+ "normalized": false,
621
+ "special": true
622
+ },
623
+ {
624
+ "id": 185,
625
+ "content": "<extra_id_64>",
626
+ "single_word": false,
627
+ "lstrip": false,
628
+ "rstrip": false,
629
+ "normalized": false,
630
+ "special": true
631
+ },
632
+ {
633
+ "id": 186,
634
+ "content": "<extra_id_65>",
635
+ "single_word": false,
636
+ "lstrip": false,
637
+ "rstrip": false,
638
+ "normalized": false,
639
+ "special": true
640
+ },
641
+ {
642
+ "id": 187,
643
+ "content": "<extra_id_66>",
644
+ "single_word": false,
645
+ "lstrip": false,
646
+ "rstrip": false,
647
+ "normalized": false,
648
+ "special": true
649
+ },
650
+ {
651
+ "id": 188,
652
+ "content": "<extra_id_67>",
653
+ "single_word": false,
654
+ "lstrip": false,
655
+ "rstrip": false,
656
+ "normalized": false,
657
+ "special": true
658
+ },
659
+ {
660
+ "id": 189,
661
+ "content": "<extra_id_68>",
662
+ "single_word": false,
663
+ "lstrip": false,
664
+ "rstrip": false,
665
+ "normalized": false,
666
+ "special": true
667
+ },
668
+ {
669
+ "id": 190,
670
+ "content": "<extra_id_69>",
671
+ "single_word": false,
672
+ "lstrip": false,
673
+ "rstrip": false,
674
+ "normalized": false,
675
+ "special": true
676
+ },
677
+ {
678
+ "id": 191,
679
+ "content": "<extra_id_70>",
680
+ "single_word": false,
681
+ "lstrip": false,
682
+ "rstrip": false,
683
+ "normalized": false,
684
+ "special": true
685
+ },
686
+ {
687
+ "id": 192,
688
+ "content": "<extra_id_71>",
689
+ "single_word": false,
690
+ "lstrip": false,
691
+ "rstrip": false,
692
+ "normalized": false,
693
+ "special": true
694
+ },
695
+ {
696
+ "id": 193,
697
+ "content": "<extra_id_72>",
698
+ "single_word": false,
699
+ "lstrip": false,
700
+ "rstrip": false,
701
+ "normalized": false,
702
+ "special": true
703
+ },
704
+ {
705
+ "id": 194,
706
+ "content": "<extra_id_73>",
707
+ "single_word": false,
708
+ "lstrip": false,
709
+ "rstrip": false,
710
+ "normalized": false,
711
+ "special": true
712
+ },
713
+ {
714
+ "id": 195,
715
+ "content": "<extra_id_74>",
716
+ "single_word": false,
717
+ "lstrip": false,
718
+ "rstrip": false,
719
+ "normalized": false,
720
+ "special": true
721
+ },
722
+ {
723
+ "id": 196,
724
+ "content": "<extra_id_75>",
725
+ "single_word": false,
726
+ "lstrip": false,
727
+ "rstrip": false,
728
+ "normalized": false,
729
+ "special": true
730
+ },
731
+ {
732
+ "id": 197,
733
+ "content": "<extra_id_76>",
734
+ "single_word": false,
735
+ "lstrip": false,
736
+ "rstrip": false,
737
+ "normalized": false,
738
+ "special": true
739
+ },
740
+ {
741
+ "id": 198,
742
+ "content": "<extra_id_77>",
743
+ "single_word": false,
744
+ "lstrip": false,
745
+ "rstrip": false,
746
+ "normalized": false,
747
+ "special": true
748
+ },
749
+ {
750
+ "id": 199,
751
+ "content": "<extra_id_78>",
752
+ "single_word": false,
753
+ "lstrip": false,
754
+ "rstrip": false,
755
+ "normalized": false,
756
+ "special": true
757
+ },
758
+ {
759
+ "id": 200,
760
+ "content": "<extra_id_79>",
761
+ "single_word": false,
762
+ "lstrip": false,
763
+ "rstrip": false,
764
+ "normalized": false,
765
+ "special": true
766
+ },
767
+ {
768
+ "id": 201,
769
+ "content": "<extra_id_80>",
770
+ "single_word": false,
771
+ "lstrip": false,
772
+ "rstrip": false,
773
+ "normalized": false,
774
+ "special": true
775
+ },
776
+ {
777
+ "id": 202,
778
+ "content": "<extra_id_81>",
779
+ "single_word": false,
780
+ "lstrip": false,
781
+ "rstrip": false,
782
+ "normalized": false,
783
+ "special": true
784
+ },
785
+ {
786
+ "id": 203,
787
+ "content": "<extra_id_82>",
788
+ "single_word": false,
789
+ "lstrip": false,
790
+ "rstrip": false,
791
+ "normalized": false,
792
+ "special": true
793
+ },
794
+ {
795
+ "id": 204,
796
+ "content": "<extra_id_83>",
797
+ "single_word": false,
798
+ "lstrip": false,
799
+ "rstrip": false,
800
+ "normalized": false,
801
+ "special": true
802
+ },
803
+ {
804
+ "id": 205,
805
+ "content": "<extra_id_84>",
806
+ "single_word": false,
807
+ "lstrip": false,
808
+ "rstrip": false,
809
+ "normalized": false,
810
+ "special": true
811
+ },
812
+ {
813
+ "id": 206,
814
+ "content": "<extra_id_85>",
815
+ "single_word": false,
816
+ "lstrip": false,
817
+ "rstrip": false,
818
+ "normalized": false,
819
+ "special": true
820
+ },
821
+ {
822
+ "id": 207,
823
+ "content": "<extra_id_86>",
824
+ "single_word": false,
825
+ "lstrip": false,
826
+ "rstrip": false,
827
+ "normalized": false,
828
+ "special": true
829
+ },
830
+ {
831
+ "id": 208,
832
+ "content": "<extra_id_87>",
833
+ "single_word": false,
834
+ "lstrip": false,
835
+ "rstrip": false,
836
+ "normalized": false,
837
+ "special": true
838
+ },
839
+ {
840
+ "id": 209,
841
+ "content": "<extra_id_88>",
842
+ "single_word": false,
843
+ "lstrip": false,
844
+ "rstrip": false,
845
+ "normalized": false,
846
+ "special": true
847
+ },
848
+ {
849
+ "id": 210,
850
+ "content": "<extra_id_89>",
851
+ "single_word": false,
852
+ "lstrip": false,
853
+ "rstrip": false,
854
+ "normalized": false,
855
+ "special": true
856
+ },
857
+ {
858
+ "id": 211,
859
+ "content": "<extra_id_90>",
860
+ "single_word": false,
861
+ "lstrip": false,
862
+ "rstrip": false,
863
+ "normalized": false,
864
+ "special": true
865
+ },
866
+ {
867
+ "id": 212,
868
+ "content": "<extra_id_91>",
869
+ "single_word": false,
870
+ "lstrip": false,
871
+ "rstrip": false,
872
+ "normalized": false,
873
+ "special": true
874
+ },
875
+ {
876
+ "id": 213,
877
+ "content": "<extra_id_92>",
878
+ "single_word": false,
879
+ "lstrip": false,
880
+ "rstrip": false,
881
+ "normalized": false,
882
+ "special": true
883
+ },
884
+ {
885
+ "id": 214,
886
+ "content": "<extra_id_93>",
887
+ "single_word": false,
888
+ "lstrip": false,
889
+ "rstrip": false,
890
+ "normalized": false,
891
+ "special": true
892
+ },
893
+ {
894
+ "id": 215,
895
+ "content": "<extra_id_94>",
896
+ "single_word": false,
897
+ "lstrip": false,
898
+ "rstrip": false,
899
+ "normalized": false,
900
+ "special": true
901
+ },
902
+ {
903
+ "id": 216,
904
+ "content": "<extra_id_95>",
905
+ "single_word": false,
906
+ "lstrip": false,
907
+ "rstrip": false,
908
+ "normalized": false,
909
+ "special": true
910
+ },
911
+ {
912
+ "id": 217,
913
+ "content": "<extra_id_96>",
914
+ "single_word": false,
915
+ "lstrip": false,
916
+ "rstrip": false,
917
+ "normalized": false,
918
+ "special": true
919
+ },
920
+ {
921
+ "id": 218,
922
+ "content": "<extra_id_97>",
923
+ "single_word": false,
924
+ "lstrip": false,
925
+ "rstrip": false,
926
+ "normalized": false,
927
+ "special": true
928
+ },
929
+ {
930
+ "id": 219,
931
+ "content": "<extra_id_98>",
932
+ "single_word": false,
933
+ "lstrip": false,
934
+ "rstrip": false,
935
+ "normalized": false,
936
+ "special": true
937
+ },
938
+ {
939
+ "id": 220,
940
+ "content": "<extra_id_99>",
941
+ "single_word": false,
942
+ "lstrip": false,
943
+ "rstrip": false,
944
+ "normalized": false,
945
+ "special": true
946
+ },
947
+ {
948
+ "id": 221,
949
+ "content": ".",
950
+ "single_word": false,
951
+ "lstrip": false,
952
+ "rstrip": false,
953
+ "normalized": true,
954
+ "special": false
955
+ },
956
+ {
957
+ "id": 222,
958
+ "content": "6",
959
+ "single_word": false,
960
+ "lstrip": false,
961
+ "rstrip": false,
962
+ "normalized": true,
963
+ "special": false
964
+ },
965
+ {
966
+ "id": 223,
967
+ "content": "7",
968
+ "single_word": false,
969
+ "lstrip": false,
970
+ "rstrip": false,
971
+ "normalized": true,
972
+ "special": false
973
+ },
974
+ {
975
+ "id": 224,
976
+ "content": "8",
977
+ "single_word": false,
978
+ "lstrip": false,
979
+ "rstrip": false,
980
+ "normalized": true,
981
+ "special": false
982
+ },
983
+ {
984
+ "id": 225,
985
+ "content": "<",
986
+ "single_word": false,
987
+ "lstrip": false,
988
+ "rstrip": false,
989
+ "normalized": true,
990
+ "special": false
991
+ },
992
+ {
993
+ "id": 226,
994
+ "content": ">",
995
+ "single_word": false,
996
+ "lstrip": false,
997
+ "rstrip": false,
998
+ "normalized": true,
999
+ "special": false
1000
+ },
1001
+ {
1002
+ "id": 227,
1003
+ "content": "Ag",
1004
+ "single_word": false,
1005
+ "lstrip": false,
1006
+ "rstrip": false,
1007
+ "normalized": true,
1008
+ "special": false
1009
+ },
1010
+ {
1011
+ "id": 228,
1012
+ "content": "Al",
1013
+ "single_word": false,
1014
+ "lstrip": false,
1015
+ "rstrip": false,
1016
+ "normalized": true,
1017
+ "special": false
1018
+ },
1019
+ {
1020
+ "id": 229,
1021
+ "content": "Ar",
1022
+ "single_word": false,
1023
+ "lstrip": false,
1024
+ "rstrip": false,
1025
+ "normalized": true,
1026
+ "special": false
1027
+ },
1028
+ {
1029
+ "id": 230,
1030
+ "content": "As",
1031
+ "single_word": false,
1032
+ "lstrip": false,
1033
+ "rstrip": false,
1034
+ "normalized": true,
1035
+ "special": false
1036
+ },
1037
+ {
1038
+ "id": 231,
1039
+ "content": "Au",
1040
+ "single_word": false,
1041
+ "lstrip": false,
1042
+ "rstrip": false,
1043
+ "normalized": true,
1044
+ "special": false
1045
+ },
1046
+ {
1047
+ "id": 232,
1048
+ "content": "Ba",
1049
+ "single_word": false,
1050
+ "lstrip": false,
1051
+ "rstrip": false,
1052
+ "normalized": true,
1053
+ "special": false
1054
+ },
1055
+ {
1056
+ "id": 233,
1057
+ "content": "Bi",
1058
+ "single_word": false,
1059
+ "lstrip": false,
1060
+ "rstrip": false,
1061
+ "normalized": true,
1062
+ "special": false
1063
+ },
1064
+ {
1065
+ "id": 234,
1066
+ "content": "Ca",
1067
+ "single_word": false,
1068
+ "lstrip": false,
1069
+ "rstrip": false,
1070
+ "normalized": true,
1071
+ "special": false
1072
+ },
1073
+ {
1074
+ "id": 235,
1075
+ "content": "Cu",
1076
+ "single_word": false,
1077
+ "lstrip": false,
1078
+ "rstrip": false,
1079
+ "normalized": true,
1080
+ "special": false
1081
+ },
1082
+ {
1083
+ "id": 236,
1084
+ "content": "Fe",
1085
+ "single_word": false,
1086
+ "lstrip": false,
1087
+ "rstrip": false,
1088
+ "normalized": true,
1089
+ "special": false
1090
+ },
1091
+ {
1092
+ "id": 237,
1093
+ "content": "Ge",
1094
+ "single_word": false,
1095
+ "lstrip": false,
1096
+ "rstrip": false,
1097
+ "normalized": true,
1098
+ "special": false
1099
+ },
1100
+ {
1101
+ "id": 238,
1102
+ "content": "Hg",
1103
+ "single_word": false,
1104
+ "lstrip": false,
1105
+ "rstrip": false,
1106
+ "normalized": true,
1107
+ "special": false
1108
+ },
1109
+ {
1110
+ "id": 239,
1111
+ "content": "K",
1112
+ "single_word": false,
1113
+ "lstrip": false,
1114
+ "rstrip": false,
1115
+ "normalized": true,
1116
+ "special": false
1117
+ },
1118
+ {
1119
+ "id": 240,
1120
+ "content": "Li",
1121
+ "single_word": false,
1122
+ "lstrip": false,
1123
+ "rstrip": false,
1124
+ "normalized": true,
1125
+ "special": false
1126
+ },
1127
+ {
1128
+ "id": 241,
1129
+ "content": "Mg",
1130
+ "single_word": false,
1131
+ "lstrip": false,
1132
+ "rstrip": false,
1133
+ "normalized": true,
1134
+ "special": false
1135
+ },
1136
+ {
1137
+ "id": 242,
1138
+ "content": "Mn",
1139
+ "single_word": false,
1140
+ "lstrip": false,
1141
+ "rstrip": false,
1142
+ "normalized": true,
1143
+ "special": false
1144
+ },
1145
+ {
1146
+ "id": 243,
1147
+ "content": "Mo",
1148
+ "single_word": false,
1149
+ "lstrip": false,
1150
+ "rstrip": false,
1151
+ "normalized": true,
1152
+ "special": false
1153
+ },
1154
+ {
1155
+ "id": 244,
1156
+ "content": "Na",
1157
+ "single_word": false,
1158
+ "lstrip": false,
1159
+ "rstrip": false,
1160
+ "normalized": true,
1161
+ "special": false
1162
+ },
1163
+ {
1164
+ "id": 245,
1165
+ "content": "Nd",
1166
+ "single_word": false,
1167
+ "lstrip": false,
1168
+ "rstrip": false,
1169
+ "normalized": true,
1170
+ "special": false
1171
+ },
1172
+ {
1173
+ "id": 246,
1174
+ "content": "Ni",
1175
+ "single_word": false,
1176
+ "lstrip": false,
1177
+ "rstrip": false,
1178
+ "normalized": true,
1179
+ "special": false
1180
+ },
1181
+ {
1182
+ "id": 247,
1183
+ "content": "P",
1184
+ "single_word": false,
1185
+ "lstrip": false,
1186
+ "rstrip": false,
1187
+ "normalized": true,
1188
+ "special": false
1189
+ },
1190
+ {
1191
+ "id": 248,
1192
+ "content": "Pb",
1193
+ "single_word": false,
1194
+ "lstrip": false,
1195
+ "rstrip": false,
1196
+ "normalized": true,
1197
+ "special": false
1198
+ },
1199
+ {
1200
+ "id": 249,
1201
+ "content": "Pd",
1202
+ "single_word": false,
1203
+ "lstrip": false,
1204
+ "rstrip": false,
1205
+ "normalized": true,
1206
+ "special": false
1207
+ },
1208
+ {
1209
+ "id": 250,
1210
+ "content": "Pt",
1211
+ "single_word": false,
1212
+ "lstrip": false,
1213
+ "rstrip": false,
1214
+ "normalized": true,
1215
+ "special": false
1216
+ },
1217
+ {
1218
+ "id": 251,
1219
+ "content": "Re",
1220
+ "single_word": false,
1221
+ "lstrip": false,
1222
+ "rstrip": false,
1223
+ "normalized": true,
1224
+ "special": false
1225
+ },
1226
+ {
1227
+ "id": 252,
1228
+ "content": "Rh",
1229
+ "single_word": false,
1230
+ "lstrip": false,
1231
+ "rstrip": false,
1232
+ "normalized": true,
1233
+ "special": false
1234
+ },
1235
+ {
1236
+ "id": 253,
1237
+ "content": "Ru",
1238
+ "single_word": false,
1239
+ "lstrip": false,
1240
+ "rstrip": false,
1241
+ "normalized": true,
1242
+ "special": false
1243
+ },
1244
+ {
1245
+ "id": 254,
1246
+ "content": "Sb",
1247
+ "single_word": false,
1248
+ "lstrip": false,
1249
+ "rstrip": false,
1250
+ "normalized": true,
1251
+ "special": false
1252
+ },
1253
+ {
1254
+ "id": 255,
1255
+ "content": "Si",
1256
+ "single_word": false,
1257
+ "lstrip": false,
1258
+ "rstrip": false,
1259
+ "normalized": true,
1260
+ "special": false
1261
+ },
1262
+ {
1263
+ "id": 256,
1264
+ "content": "Sm",
1265
+ "single_word": false,
1266
+ "lstrip": false,
1267
+ "rstrip": false,
1268
+ "normalized": true,
1269
+ "special": false
1270
+ },
1271
+ {
1272
+ "id": 257,
1273
+ "content": "Ta",
1274
+ "single_word": false,
1275
+ "lstrip": false,
1276
+ "rstrip": false,
1277
+ "normalized": true,
1278
+ "special": false
1279
+ },
1280
+ {
1281
+ "id": 258,
1282
+ "content": "Ti",
1283
+ "single_word": false,
1284
+ "lstrip": false,
1285
+ "rstrip": false,
1286
+ "normalized": true,
1287
+ "special": false
1288
+ },
1289
+ {
1290
+ "id": 259,
1291
+ "content": "Tl",
1292
+ "single_word": false,
1293
+ "lstrip": false,
1294
+ "rstrip": false,
1295
+ "normalized": true,
1296
+ "special": false
1297
+ },
1298
+ {
1299
+ "id": 260,
1300
+ "content": "W",
1301
+ "single_word": false,
1302
+ "lstrip": false,
1303
+ "rstrip": false,
1304
+ "normalized": true,
1305
+ "special": false
1306
+ },
1307
+ {
1308
+ "id": 261,
1309
+ "content": "Yb",
1310
+ "single_word": false,
1311
+ "lstrip": false,
1312
+ "rstrip": false,
1313
+ "normalized": true,
1314
+ "special": false
1315
+ },
1316
+ {
1317
+ "id": 262,
1318
+ "content": "Zn",
1319
+ "single_word": false,
1320
+ "lstrip": false,
1321
+ "rstrip": false,
1322
+ "normalized": true,
1323
+ "special": false
1324
+ },
1325
+ {
1326
+ "id": 263,
1327
+ "content": "Zr",
1328
+ "single_word": false,
1329
+ "lstrip": false,
1330
+ "rstrip": false,
1331
+ "normalized": true,
1332
+ "special": false
1333
+ },
1334
+ {
1335
+ "id": 264,
1336
+ "content": "e",
1337
+ "single_word": false,
1338
+ "lstrip": false,
1339
+ "rstrip": false,
1340
+ "normalized": true,
1341
+ "special": false
1342
+ },
1343
+ {
1344
+ "id": 265,
1345
+ "content": "p",
1346
+ "single_word": false,
1347
+ "lstrip": false,
1348
+ "rstrip": false,
1349
+ "normalized": true,
1350
+ "special": false
1351
+ },
1352
+ {
1353
+ "id": 266,
1354
+ "content": "REACTANT:",
1355
+ "single_word": false,
1356
+ "lstrip": false,
1357
+ "rstrip": false,
1358
+ "normalized": false,
1359
+ "special": true
1360
+ },
1361
+ {
1362
+ "id": 267,
1363
+ "content": "REAGENT:",
1364
+ "single_word": false,
1365
+ "lstrip": false,
1366
+ "rstrip": false,
1367
+ "normalized": false,
1368
+ "special": true
1369
+ }
1370
+ ],
1371
+ "normalizer": {
1372
+ "type": "Sequence",
1373
+ "normalizers": [
1374
+ {
1375
+ "type": "Nmt"
1376
+ },
1377
+ {
1378
+ "type": "NFKC"
1379
+ },
1380
+ {
1381
+ "type": "Replace",
1382
+ "pattern": {
1383
+ "Regex": " {2,}"
1384
+ },
1385
+ "content": " "
1386
+ }
1387
+ ]
1388
+ },
1389
+ "pre_tokenizer": {
1390
+ "type": "Sequence",
1391
+ "pretokenizers": [
1392
+ {
1393
+ "type": "Metaspace",
1394
+ "replacement": "▁",
1395
+ "prepend_scheme": "always",
1396
+ "split": true
1397
+ },
1398
+ {
1399
+ "type": "Digits",
1400
+ "individual_digits": true
1401
+ },
1402
+ {
1403
+ "type": "Punctuation",
1404
+ "behavior": "Isolated"
1405
+ }
1406
+ ]
1407
+ },
1408
+ "post_processor": {
1409
+ "type": "TemplateProcessing",
1410
+ "single": [
1411
+ {
1412
+ "Sequence": {
1413
+ "id": "A",
1414
+ "type_id": 0
1415
+ }
1416
+ },
1417
+ {
1418
+ "SpecialToken": {
1419
+ "id": "</s>",
1420
+ "type_id": 0
1421
+ }
1422
+ }
1423
+ ],
1424
+ "pair": [
1425
+ {
1426
+ "Sequence": {
1427
+ "id": "A",
1428
+ "type_id": 0
1429
+ }
1430
+ },
1431
+ {
1432
+ "Sequence": {
1433
+ "id": "B",
1434
+ "type_id": 1
1435
+ }
1436
+ }
1437
+ ],
1438
+ "special_tokens": {
1439
+ "</s>": {
1440
+ "id": "</s>",
1441
+ "ids": [
1442
+ 1
1443
+ ],
1444
+ "tokens": [
1445
+ "</s>"
1446
+ ]
1447
+ }
1448
+ }
1449
+ },
1450
+ "decoder": {
1451
+ "type": "Metaspace",
1452
+ "replacement": "▁",
1453
+ "prepend_scheme": "always",
1454
+ "split": true
1455
+ },
1456
+ "model": {
1457
+ "type": "Unigram",
1458
+ "unk_id": 2,
1459
+ "vocab": [
1460
+ [
1461
+ "<pad>",
1462
+ 0.0
1463
+ ],
1464
+ [
1465
+ "</s>",
1466
+ 0.0
1467
+ ],
1468
+ [
1469
+ "<unk>",
1470
+ 0.0
1471
+ ],
1472
+ [
1473
+ "C",
1474
+ -1.934062026530869
1475
+ ],
1476
+ [
1477
+ "c",
1478
+ -2.005135123981896
1479
+ ],
1480
+ [
1481
+ ")",
1482
+ -2.3305416341016603
1483
+ ],
1484
+ [
1485
+ "(",
1486
+ -2.3305416341016603
1487
+ ],
1488
+ [
1489
+ "1",
1490
+ -2.659489340254079
1491
+ ],
1492
+ [
1493
+ "2",
1494
+ -3.0147314517826524
1495
+ ],
1496
+ [
1497
+ "O",
1498
+ -3.1705660462226426
1499
+ ],
1500
+ [
1501
+ "]",
1502
+ -3.306692943444802
1503
+ ],
1504
+ [
1505
+ "[",
1506
+ -3.306692943444802
1507
+ ],
1508
+ [
1509
+ "N",
1510
+ -3.3379525475363145
1511
+ ],
1512
+ [
1513
+ "@",
1514
+ -3.375613737475505
1515
+ ],
1516
+ [
1517
+ "H",
1518
+ -3.3960003181228675
1519
+ ],
1520
+ [
1521
+ "=",
1522
+ -3.4342358584336026
1523
+ ],
1524
+ [
1525
+ "n",
1526
+ -3.9996436484078535
1527
+ ],
1528
+ [
1529
+ "3",
1530
+ -4.06624150723795
1531
+ ],
1532
+ [
1533
+ "+",
1534
+ -4.576587821405814
1535
+ ],
1536
+ [
1537
+ "ccc",
1538
+ -4.581329068949029
1539
+ ],
1540
+ [
1541
+ "cccc",
1542
+ -4.794168477161054
1543
+ ],
1544
+ [
1545
+ "F",
1546
+ -4.844444469349959
1547
+ ],
1548
+ [
1549
+ "▁C",
1550
+ -4.930731418706828
1551
+ ],
1552
+ [
1553
+ "-",
1554
+ -5.128151794820781
1555
+ ],
1556
+ [
1557
+ "CC",
1558
+ -5.188054847421637
1559
+ ],
1560
+ [
1561
+ "▁",
1562
+ -5.243375183264384
1563
+ ],
1564
+ [
1565
+ "▁CC",
1566
+ -5.43665410318811
1567
+ ],
1568
+ [
1569
+ "cc",
1570
+ -5.5072446210190025
1571
+ ],
1572
+ [
1573
+ "S",
1574
+ -5.607990514129369
1575
+ ],
1576
+ [
1577
+ "CCC",
1578
+ -5.610845771534942
1579
+ ],
1580
+ [
1581
+ "4",
1582
+ -5.6728058623804705
1583
+ ],
1584
+ [
1585
+ "CCN",
1586
+ -5.984229431727735
1587
+ ],
1588
+ [
1589
+ "▁CO",
1590
+ -6.136725098692515
1591
+ ],
1592
+ [
1593
+ "s",
1594
+ -6.163972117410355
1595
+ ],
1596
+ [
1597
+ "l",
1598
+ -6.244242813422005
1599
+ ],
1600
+ [
1601
+ "nc",
1602
+ -6.260369976115332
1603
+ ],
1604
+ [
1605
+ "Cl",
1606
+ -6.29041030199948
1607
+ ],
1608
+ [
1609
+ "NC",
1610
+ -6.301741325285667
1611
+ ],
1612
+ [
1613
+ "#",
1614
+ -6.475695381808137
1615
+ ],
1616
+ [
1617
+ "CCCC",
1618
+ -6.630889749196982
1619
+ ],
1620
+ [
1621
+ "Br",
1622
+ -6.708882955827546
1623
+ ],
1624
+ [
1625
+ "o",
1626
+ -6.8821690990922
1627
+ ],
1628
+ [
1629
+ "no",
1630
+ -7.085301372799254
1631
+ ],
1632
+ [
1633
+ "nn",
1634
+ -7.118314757276033
1635
+ ],
1636
+ [
1637
+ "CN",
1638
+ -7.148023077707199
1639
+ ],
1640
+ [
1641
+ "▁O",
1642
+ -7.184905330255651
1643
+ ],
1644
+ [
1645
+ "CCO",
1646
+ -7.223262672570243
1647
+ ],
1648
+ [
1649
+ "▁CN",
1650
+ -7.340957119808042
1651
+ ],
1652
+ [
1653
+ "ccs",
1654
+ -7.358439269448935
1655
+ ],
1656
+ [
1657
+ "▁CCC",
1658
+ -7.384181818093239
1659
+ ],
1660
+ [
1661
+ "OC",
1662
+ -7.401588967543117
1663
+ ],
1664
+ [
1665
+ "/",
1666
+ -7.412130091887384
1667
+ ],
1668
+ [
1669
+ "▁COC",
1670
+ -7.453342081307827
1671
+ ],
1672
+ [
1673
+ "nnc",
1674
+ -7.47808245293119
1675
+ ],
1676
+ [
1677
+ "cccn",
1678
+ -7.526117466256242
1679
+ ],
1680
+ [
1681
+ "▁CCOC",
1682
+ -7.581460400684973
1683
+ ],
1684
+ [
1685
+ "sc",
1686
+ -7.587476370908391
1687
+ ],
1688
+ [
1689
+ "cco",
1690
+ -7.601812309424911
1691
+ ],
1692
+ [
1693
+ "CCOCC",
1694
+ -7.70962259415791
1695
+ ],
1696
+ [
1697
+ "CCCN",
1698
+ -7.716050546596486
1699
+ ],
1700
+ [
1701
+ "oc",
1702
+ -7.76950331715161
1703
+ ],
1704
+ [
1705
+ "ccnc",
1706
+ -7.885377075776418
1707
+ ],
1708
+ [
1709
+ "▁CCO",
1710
+ -7.958354042052764
1711
+ ],
1712
+ [
1713
+ "cs",
1714
+ -7.963618836898841
1715
+ ],
1716
+ [
1717
+ "▁CCN",
1718
+ -7.969745690077621
1719
+ ],
1720
+ [
1721
+ "NCC",
1722
+ -7.982831223650715
1723
+ ],
1724
+ [
1725
+ "CNC",
1726
+ -8.015812540478317
1727
+ ],
1728
+ [
1729
+ "CCCO",
1730
+ -8.02283744281797
1731
+ ],
1732
+ [
1733
+ "nccn",
1734
+ -8.023781204384559
1735
+ ],
1736
+ [
1737
+ "CCCCC",
1738
+ -8.113410349713947
1739
+ ],
1740
+ [
1741
+ "csc",
1742
+ -8.123340128859317
1743
+ ],
1744
+ [
1745
+ "nnn",
1746
+ -8.16923209155111
1747
+ ],
1748
+ [
1749
+ "CO",
1750
+ -8.197076591417558
1751
+ ],
1752
+ [
1753
+ "cnc",
1754
+ -8.24948897316689
1755
+ ],
1756
+ [
1757
+ "OCC",
1758
+ -8.337244843539168
1759
+ ],
1760
+ [
1761
+ "ncc",
1762
+ -8.35350988501787
1763
+ ],
1764
+ [
1765
+ "CCNC",
1766
+ -8.374584543880331
1767
+ ],
1768
+ [
1769
+ "cnn",
1770
+ -8.379764985534148
1771
+ ],
1772
+ [
1773
+ "▁CCCC",
1774
+ -8.408946579060395
1775
+ ],
1776
+ [
1777
+ "ccn",
1778
+ -8.441316367356151
1779
+ ],
1780
+ [
1781
+ "▁CCCN",
1782
+ -8.455940453267049
1783
+ ],
1784
+ [
1785
+ "ncn",
1786
+ -8.509281787409483
1787
+ ],
1788
+ [
1789
+ "CCOC",
1790
+ -8.545791064619086
1791
+ ],
1792
+ [
1793
+ "5",
1794
+ -8.595231126796552
1795
+ ],
1796
+ [
1797
+ "OCCO",
1798
+ -8.620568633537287
1799
+ ],
1800
+ [
1801
+ "SCC",
1802
+ -8.675945846432537
1803
+ ],
1804
+ [
1805
+ "\\",
1806
+ -8.738088269693655
1807
+ ],
1808
+ [
1809
+ "▁CS",
1810
+ -8.745760398405787
1811
+ ],
1812
+ [
1813
+ "cn",
1814
+ -8.81287462848524
1815
+ ],
1816
+ [
1817
+ "CS",
1818
+ -8.852680543942768
1819
+ ],
1820
+ [
1821
+ "COC",
1822
+ -8.901532120821837
1823
+ ],
1824
+ [
1825
+ "▁CCCO",
1826
+ -8.971729419021525
1827
+ ],
1828
+ [
1829
+ "▁COCC",
1830
+ -8.986602855501445
1831
+ ],
1832
+ [
1833
+ "▁F",
1834
+ -9.08172188898926
1835
+ ],
1836
+ [
1837
+ "▁CCNC",
1838
+ -9.213239929783953
1839
+ ],
1840
+ [
1841
+ "nccc",
1842
+ -9.267617565078302
1843
+ ],
1844
+ [
1845
+ "▁CCS",
1846
+ -9.284123840101111
1847
+ ],
1848
+ [
1849
+ "▁N",
1850
+ -9.46465900903722
1851
+ ],
1852
+ [
1853
+ "▁CCCCO",
1854
+ -9.549659454320029
1855
+ ],
1856
+ [
1857
+ "NCCC",
1858
+ -9.592808545839064
1859
+ ],
1860
+ [
1861
+ "NCCN",
1862
+ -9.616093114468232
1863
+ ],
1864
+ [
1865
+ "CCS",
1866
+ -9.68568843530331
1867
+ ],
1868
+ [
1869
+ "cncc",
1870
+ -9.884757958758378
1871
+ ],
1872
+ [
1873
+ "CCCS",
1874
+ -10.07490132384844
1875
+ ],
1876
+ [
1877
+ "▁NC",
1878
+ -10.174180411834037
1879
+ ],
1880
+ [
1881
+ "▁COCCN",
1882
+ -10.182867019690027
1883
+ ],
1884
+ [
1885
+ "▁COCCO",
1886
+ -10.210382702506402
1887
+ ],
1888
+ [
1889
+ "▁COCCC",
1890
+ -10.390960567106395
1891
+ ],
1892
+ [
1893
+ "OCO",
1894
+ -10.445584736697652
1895
+ ],
1896
+ [
1897
+ "COCC",
1898
+ -10.469925833149428
1899
+ ],
1900
+ [
1901
+ "▁CCOCCN",
1902
+ -10.487938027145828
1903
+ ],
1904
+ [
1905
+ "SC",
1906
+ -10.868655865451124
1907
+ ],
1908
+ [
1909
+ "co",
1910
+ -11.175046369147724
1911
+ ],
1912
+ [
1913
+ "I",
1914
+ -11.188088269693656
1915
+ ],
1916
+ [
1917
+ "on",
1918
+ -11.199838029113598
1919
+ ],
1920
+ [
1921
+ "CCCCN",
1922
+ -11.273835881601958
1923
+ ],
1924
+ [
1925
+ "▁CCCCN",
1926
+ -11.694782342260607
1927
+ ],
1928
+ [
1929
+ "CCOCCN",
1930
+ -11.98760417613926
1931
+ ],
1932
+ [
1933
+ "r",
1934
+ -12.313357692926305
1935
+ ],
1936
+ [
1937
+ "B",
1938
+ -12.313457692926304
1939
+ ],
1940
+ [
1941
+ "CCCCO",
1942
+ -12.313457692926304
1943
+ ]
1944
+ ],
1945
+ "byte_fallback": false
1946
+ }
1947
+ }
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/tokenizer_config.json ADDED
@@ -0,0 +1,1328 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<pad>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "</s>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "<unk>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "36": {
28
+ "content": "Cl",
29
+ "lstrip": false,
30
+ "normalized": true,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": false
34
+ },
35
+ "121": {
36
+ "content": "<extra_id_0>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ },
43
+ "122": {
44
+ "content": "<extra_id_1>",
45
+ "lstrip": false,
46
+ "normalized": false,
47
+ "rstrip": false,
48
+ "single_word": false,
49
+ "special": true
50
+ },
51
+ "123": {
52
+ "content": "<extra_id_2>",
53
+ "lstrip": false,
54
+ "normalized": false,
55
+ "rstrip": false,
56
+ "single_word": false,
57
+ "special": true
58
+ },
59
+ "124": {
60
+ "content": "<extra_id_3>",
61
+ "lstrip": false,
62
+ "normalized": false,
63
+ "rstrip": false,
64
+ "single_word": false,
65
+ "special": true
66
+ },
67
+ "125": {
68
+ "content": "<extra_id_4>",
69
+ "lstrip": false,
70
+ "normalized": false,
71
+ "rstrip": false,
72
+ "single_word": false,
73
+ "special": true
74
+ },
75
+ "126": {
76
+ "content": "<extra_id_5>",
77
+ "lstrip": false,
78
+ "normalized": false,
79
+ "rstrip": false,
80
+ "single_word": false,
81
+ "special": true
82
+ },
83
+ "127": {
84
+ "content": "<extra_id_6>",
85
+ "lstrip": false,
86
+ "normalized": false,
87
+ "rstrip": false,
88
+ "single_word": false,
89
+ "special": true
90
+ },
91
+ "128": {
92
+ "content": "<extra_id_7>",
93
+ "lstrip": false,
94
+ "normalized": false,
95
+ "rstrip": false,
96
+ "single_word": false,
97
+ "special": true
98
+ },
99
+ "129": {
100
+ "content": "<extra_id_8>",
101
+ "lstrip": false,
102
+ "normalized": false,
103
+ "rstrip": false,
104
+ "single_word": false,
105
+ "special": true
106
+ },
107
+ "130": {
108
+ "content": "<extra_id_9>",
109
+ "lstrip": false,
110
+ "normalized": false,
111
+ "rstrip": false,
112
+ "single_word": false,
113
+ "special": true
114
+ },
115
+ "131": {
116
+ "content": "<extra_id_10>",
117
+ "lstrip": false,
118
+ "normalized": false,
119
+ "rstrip": false,
120
+ "single_word": false,
121
+ "special": true
122
+ },
123
+ "132": {
124
+ "content": "<extra_id_11>",
125
+ "lstrip": false,
126
+ "normalized": false,
127
+ "rstrip": false,
128
+ "single_word": false,
129
+ "special": true
130
+ },
131
+ "133": {
132
+ "content": "<extra_id_12>",
133
+ "lstrip": false,
134
+ "normalized": false,
135
+ "rstrip": false,
136
+ "single_word": false,
137
+ "special": true
138
+ },
139
+ "134": {
140
+ "content": "<extra_id_13>",
141
+ "lstrip": false,
142
+ "normalized": false,
143
+ "rstrip": false,
144
+ "single_word": false,
145
+ "special": true
146
+ },
147
+ "135": {
148
+ "content": "<extra_id_14>",
149
+ "lstrip": false,
150
+ "normalized": false,
151
+ "rstrip": false,
152
+ "single_word": false,
153
+ "special": true
154
+ },
155
+ "136": {
156
+ "content": "<extra_id_15>",
157
+ "lstrip": false,
158
+ "normalized": false,
159
+ "rstrip": false,
160
+ "single_word": false,
161
+ "special": true
162
+ },
163
+ "137": {
164
+ "content": "<extra_id_16>",
165
+ "lstrip": false,
166
+ "normalized": false,
167
+ "rstrip": false,
168
+ "single_word": false,
169
+ "special": true
170
+ },
171
+ "138": {
172
+ "content": "<extra_id_17>",
173
+ "lstrip": false,
174
+ "normalized": false,
175
+ "rstrip": false,
176
+ "single_word": false,
177
+ "special": true
178
+ },
179
+ "139": {
180
+ "content": "<extra_id_18>",
181
+ "lstrip": false,
182
+ "normalized": false,
183
+ "rstrip": false,
184
+ "single_word": false,
185
+ "special": true
186
+ },
187
+ "140": {
188
+ "content": "<extra_id_19>",
189
+ "lstrip": false,
190
+ "normalized": false,
191
+ "rstrip": false,
192
+ "single_word": false,
193
+ "special": true
194
+ },
195
+ "141": {
196
+ "content": "<extra_id_20>",
197
+ "lstrip": false,
198
+ "normalized": false,
199
+ "rstrip": false,
200
+ "single_word": false,
201
+ "special": true
202
+ },
203
+ "142": {
204
+ "content": "<extra_id_21>",
205
+ "lstrip": false,
206
+ "normalized": false,
207
+ "rstrip": false,
208
+ "single_word": false,
209
+ "special": true
210
+ },
211
+ "143": {
212
+ "content": "<extra_id_22>",
213
+ "lstrip": false,
214
+ "normalized": false,
215
+ "rstrip": false,
216
+ "single_word": false,
217
+ "special": true
218
+ },
219
+ "144": {
220
+ "content": "<extra_id_23>",
221
+ "lstrip": false,
222
+ "normalized": false,
223
+ "rstrip": false,
224
+ "single_word": false,
225
+ "special": true
226
+ },
227
+ "145": {
228
+ "content": "<extra_id_24>",
229
+ "lstrip": false,
230
+ "normalized": false,
231
+ "rstrip": false,
232
+ "single_word": false,
233
+ "special": true
234
+ },
235
+ "146": {
236
+ "content": "<extra_id_25>",
237
+ "lstrip": false,
238
+ "normalized": false,
239
+ "rstrip": false,
240
+ "single_word": false,
241
+ "special": true
242
+ },
243
+ "147": {
244
+ "content": "<extra_id_26>",
245
+ "lstrip": false,
246
+ "normalized": false,
247
+ "rstrip": false,
248
+ "single_word": false,
249
+ "special": true
250
+ },
251
+ "148": {
252
+ "content": "<extra_id_27>",
253
+ "lstrip": false,
254
+ "normalized": false,
255
+ "rstrip": false,
256
+ "single_word": false,
257
+ "special": true
258
+ },
259
+ "149": {
260
+ "content": "<extra_id_28>",
261
+ "lstrip": false,
262
+ "normalized": false,
263
+ "rstrip": false,
264
+ "single_word": false,
265
+ "special": true
266
+ },
267
+ "150": {
268
+ "content": "<extra_id_29>",
269
+ "lstrip": false,
270
+ "normalized": false,
271
+ "rstrip": false,
272
+ "single_word": false,
273
+ "special": true
274
+ },
275
+ "151": {
276
+ "content": "<extra_id_30>",
277
+ "lstrip": false,
278
+ "normalized": false,
279
+ "rstrip": false,
280
+ "single_word": false,
281
+ "special": true
282
+ },
283
+ "152": {
284
+ "content": "<extra_id_31>",
285
+ "lstrip": false,
286
+ "normalized": false,
287
+ "rstrip": false,
288
+ "single_word": false,
289
+ "special": true
290
+ },
291
+ "153": {
292
+ "content": "<extra_id_32>",
293
+ "lstrip": false,
294
+ "normalized": false,
295
+ "rstrip": false,
296
+ "single_word": false,
297
+ "special": true
298
+ },
299
+ "154": {
300
+ "content": "<extra_id_33>",
301
+ "lstrip": false,
302
+ "normalized": false,
303
+ "rstrip": false,
304
+ "single_word": false,
305
+ "special": true
306
+ },
307
+ "155": {
308
+ "content": "<extra_id_34>",
309
+ "lstrip": false,
310
+ "normalized": false,
311
+ "rstrip": false,
312
+ "single_word": false,
313
+ "special": true
314
+ },
315
+ "156": {
316
+ "content": "<extra_id_35>",
317
+ "lstrip": false,
318
+ "normalized": false,
319
+ "rstrip": false,
320
+ "single_word": false,
321
+ "special": true
322
+ },
323
+ "157": {
324
+ "content": "<extra_id_36>",
325
+ "lstrip": false,
326
+ "normalized": false,
327
+ "rstrip": false,
328
+ "single_word": false,
329
+ "special": true
330
+ },
331
+ "158": {
332
+ "content": "<extra_id_37>",
333
+ "lstrip": false,
334
+ "normalized": false,
335
+ "rstrip": false,
336
+ "single_word": false,
337
+ "special": true
338
+ },
339
+ "159": {
340
+ "content": "<extra_id_38>",
341
+ "lstrip": false,
342
+ "normalized": false,
343
+ "rstrip": false,
344
+ "single_word": false,
345
+ "special": true
346
+ },
347
+ "160": {
348
+ "content": "<extra_id_39>",
349
+ "lstrip": false,
350
+ "normalized": false,
351
+ "rstrip": false,
352
+ "single_word": false,
353
+ "special": true
354
+ },
355
+ "161": {
356
+ "content": "<extra_id_40>",
357
+ "lstrip": false,
358
+ "normalized": false,
359
+ "rstrip": false,
360
+ "single_word": false,
361
+ "special": true
362
+ },
363
+ "162": {
364
+ "content": "<extra_id_41>",
365
+ "lstrip": false,
366
+ "normalized": false,
367
+ "rstrip": false,
368
+ "single_word": false,
369
+ "special": true
370
+ },
371
+ "163": {
372
+ "content": "<extra_id_42>",
373
+ "lstrip": false,
374
+ "normalized": false,
375
+ "rstrip": false,
376
+ "single_word": false,
377
+ "special": true
378
+ },
379
+ "164": {
380
+ "content": "<extra_id_43>",
381
+ "lstrip": false,
382
+ "normalized": false,
383
+ "rstrip": false,
384
+ "single_word": false,
385
+ "special": true
386
+ },
387
+ "165": {
388
+ "content": "<extra_id_44>",
389
+ "lstrip": false,
390
+ "normalized": false,
391
+ "rstrip": false,
392
+ "single_word": false,
393
+ "special": true
394
+ },
395
+ "166": {
396
+ "content": "<extra_id_45>",
397
+ "lstrip": false,
398
+ "normalized": false,
399
+ "rstrip": false,
400
+ "single_word": false,
401
+ "special": true
402
+ },
403
+ "167": {
404
+ "content": "<extra_id_46>",
405
+ "lstrip": false,
406
+ "normalized": false,
407
+ "rstrip": false,
408
+ "single_word": false,
409
+ "special": true
410
+ },
411
+ "168": {
412
+ "content": "<extra_id_47>",
413
+ "lstrip": false,
414
+ "normalized": false,
415
+ "rstrip": false,
416
+ "single_word": false,
417
+ "special": true
418
+ },
419
+ "169": {
420
+ "content": "<extra_id_48>",
421
+ "lstrip": false,
422
+ "normalized": false,
423
+ "rstrip": false,
424
+ "single_word": false,
425
+ "special": true
426
+ },
427
+ "170": {
428
+ "content": "<extra_id_49>",
429
+ "lstrip": false,
430
+ "normalized": false,
431
+ "rstrip": false,
432
+ "single_word": false,
433
+ "special": true
434
+ },
435
+ "171": {
436
+ "content": "<extra_id_50>",
437
+ "lstrip": false,
438
+ "normalized": false,
439
+ "rstrip": false,
440
+ "single_word": false,
441
+ "special": true
442
+ },
443
+ "172": {
444
+ "content": "<extra_id_51>",
445
+ "lstrip": false,
446
+ "normalized": false,
447
+ "rstrip": false,
448
+ "single_word": false,
449
+ "special": true
450
+ },
451
+ "173": {
452
+ "content": "<extra_id_52>",
453
+ "lstrip": false,
454
+ "normalized": false,
455
+ "rstrip": false,
456
+ "single_word": false,
457
+ "special": true
458
+ },
459
+ "174": {
460
+ "content": "<extra_id_53>",
461
+ "lstrip": false,
462
+ "normalized": false,
463
+ "rstrip": false,
464
+ "single_word": false,
465
+ "special": true
466
+ },
467
+ "175": {
468
+ "content": "<extra_id_54>",
469
+ "lstrip": false,
470
+ "normalized": false,
471
+ "rstrip": false,
472
+ "single_word": false,
473
+ "special": true
474
+ },
475
+ "176": {
476
+ "content": "<extra_id_55>",
477
+ "lstrip": false,
478
+ "normalized": false,
479
+ "rstrip": false,
480
+ "single_word": false,
481
+ "special": true
482
+ },
483
+ "177": {
484
+ "content": "<extra_id_56>",
485
+ "lstrip": false,
486
+ "normalized": false,
487
+ "rstrip": false,
488
+ "single_word": false,
489
+ "special": true
490
+ },
491
+ "178": {
492
+ "content": "<extra_id_57>",
493
+ "lstrip": false,
494
+ "normalized": false,
495
+ "rstrip": false,
496
+ "single_word": false,
497
+ "special": true
498
+ },
499
+ "179": {
500
+ "content": "<extra_id_58>",
501
+ "lstrip": false,
502
+ "normalized": false,
503
+ "rstrip": false,
504
+ "single_word": false,
505
+ "special": true
506
+ },
507
+ "180": {
508
+ "content": "<extra_id_59>",
509
+ "lstrip": false,
510
+ "normalized": false,
511
+ "rstrip": false,
512
+ "single_word": false,
513
+ "special": true
514
+ },
515
+ "181": {
516
+ "content": "<extra_id_60>",
517
+ "lstrip": false,
518
+ "normalized": false,
519
+ "rstrip": false,
520
+ "single_word": false,
521
+ "special": true
522
+ },
523
+ "182": {
524
+ "content": "<extra_id_61>",
525
+ "lstrip": false,
526
+ "normalized": false,
527
+ "rstrip": false,
528
+ "single_word": false,
529
+ "special": true
530
+ },
531
+ "183": {
532
+ "content": "<extra_id_62>",
533
+ "lstrip": false,
534
+ "normalized": false,
535
+ "rstrip": false,
536
+ "single_word": false,
537
+ "special": true
538
+ },
539
+ "184": {
540
+ "content": "<extra_id_63>",
541
+ "lstrip": false,
542
+ "normalized": false,
543
+ "rstrip": false,
544
+ "single_word": false,
545
+ "special": true
546
+ },
547
+ "185": {
548
+ "content": "<extra_id_64>",
549
+ "lstrip": false,
550
+ "normalized": false,
551
+ "rstrip": false,
552
+ "single_word": false,
553
+ "special": true
554
+ },
555
+ "186": {
556
+ "content": "<extra_id_65>",
557
+ "lstrip": false,
558
+ "normalized": false,
559
+ "rstrip": false,
560
+ "single_word": false,
561
+ "special": true
562
+ },
563
+ "187": {
564
+ "content": "<extra_id_66>",
565
+ "lstrip": false,
566
+ "normalized": false,
567
+ "rstrip": false,
568
+ "single_word": false,
569
+ "special": true
570
+ },
571
+ "188": {
572
+ "content": "<extra_id_67>",
573
+ "lstrip": false,
574
+ "normalized": false,
575
+ "rstrip": false,
576
+ "single_word": false,
577
+ "special": true
578
+ },
579
+ "189": {
580
+ "content": "<extra_id_68>",
581
+ "lstrip": false,
582
+ "normalized": false,
583
+ "rstrip": false,
584
+ "single_word": false,
585
+ "special": true
586
+ },
587
+ "190": {
588
+ "content": "<extra_id_69>",
589
+ "lstrip": false,
590
+ "normalized": false,
591
+ "rstrip": false,
592
+ "single_word": false,
593
+ "special": true
594
+ },
595
+ "191": {
596
+ "content": "<extra_id_70>",
597
+ "lstrip": false,
598
+ "normalized": false,
599
+ "rstrip": false,
600
+ "single_word": false,
601
+ "special": true
602
+ },
603
+ "192": {
604
+ "content": "<extra_id_71>",
605
+ "lstrip": false,
606
+ "normalized": false,
607
+ "rstrip": false,
608
+ "single_word": false,
609
+ "special": true
610
+ },
611
+ "193": {
612
+ "content": "<extra_id_72>",
613
+ "lstrip": false,
614
+ "normalized": false,
615
+ "rstrip": false,
616
+ "single_word": false,
617
+ "special": true
618
+ },
619
+ "194": {
620
+ "content": "<extra_id_73>",
621
+ "lstrip": false,
622
+ "normalized": false,
623
+ "rstrip": false,
624
+ "single_word": false,
625
+ "special": true
626
+ },
627
+ "195": {
628
+ "content": "<extra_id_74>",
629
+ "lstrip": false,
630
+ "normalized": false,
631
+ "rstrip": false,
632
+ "single_word": false,
633
+ "special": true
634
+ },
635
+ "196": {
636
+ "content": "<extra_id_75>",
637
+ "lstrip": false,
638
+ "normalized": false,
639
+ "rstrip": false,
640
+ "single_word": false,
641
+ "special": true
642
+ },
643
+ "197": {
644
+ "content": "<extra_id_76>",
645
+ "lstrip": false,
646
+ "normalized": false,
647
+ "rstrip": false,
648
+ "single_word": false,
649
+ "special": true
650
+ },
651
+ "198": {
652
+ "content": "<extra_id_77>",
653
+ "lstrip": false,
654
+ "normalized": false,
655
+ "rstrip": false,
656
+ "single_word": false,
657
+ "special": true
658
+ },
659
+ "199": {
660
+ "content": "<extra_id_78>",
661
+ "lstrip": false,
662
+ "normalized": false,
663
+ "rstrip": false,
664
+ "single_word": false,
665
+ "special": true
666
+ },
667
+ "200": {
668
+ "content": "<extra_id_79>",
669
+ "lstrip": false,
670
+ "normalized": false,
671
+ "rstrip": false,
672
+ "single_word": false,
673
+ "special": true
674
+ },
675
+ "201": {
676
+ "content": "<extra_id_80>",
677
+ "lstrip": false,
678
+ "normalized": false,
679
+ "rstrip": false,
680
+ "single_word": false,
681
+ "special": true
682
+ },
683
+ "202": {
684
+ "content": "<extra_id_81>",
685
+ "lstrip": false,
686
+ "normalized": false,
687
+ "rstrip": false,
688
+ "single_word": false,
689
+ "special": true
690
+ },
691
+ "203": {
692
+ "content": "<extra_id_82>",
693
+ "lstrip": false,
694
+ "normalized": false,
695
+ "rstrip": false,
696
+ "single_word": false,
697
+ "special": true
698
+ },
699
+ "204": {
700
+ "content": "<extra_id_83>",
701
+ "lstrip": false,
702
+ "normalized": false,
703
+ "rstrip": false,
704
+ "single_word": false,
705
+ "special": true
706
+ },
707
+ "205": {
708
+ "content": "<extra_id_84>",
709
+ "lstrip": false,
710
+ "normalized": false,
711
+ "rstrip": false,
712
+ "single_word": false,
713
+ "special": true
714
+ },
715
+ "206": {
716
+ "content": "<extra_id_85>",
717
+ "lstrip": false,
718
+ "normalized": false,
719
+ "rstrip": false,
720
+ "single_word": false,
721
+ "special": true
722
+ },
723
+ "207": {
724
+ "content": "<extra_id_86>",
725
+ "lstrip": false,
726
+ "normalized": false,
727
+ "rstrip": false,
728
+ "single_word": false,
729
+ "special": true
730
+ },
731
+ "208": {
732
+ "content": "<extra_id_87>",
733
+ "lstrip": false,
734
+ "normalized": false,
735
+ "rstrip": false,
736
+ "single_word": false,
737
+ "special": true
738
+ },
739
+ "209": {
740
+ "content": "<extra_id_88>",
741
+ "lstrip": false,
742
+ "normalized": false,
743
+ "rstrip": false,
744
+ "single_word": false,
745
+ "special": true
746
+ },
747
+ "210": {
748
+ "content": "<extra_id_89>",
749
+ "lstrip": false,
750
+ "normalized": false,
751
+ "rstrip": false,
752
+ "single_word": false,
753
+ "special": true
754
+ },
755
+ "211": {
756
+ "content": "<extra_id_90>",
757
+ "lstrip": false,
758
+ "normalized": false,
759
+ "rstrip": false,
760
+ "single_word": false,
761
+ "special": true
762
+ },
763
+ "212": {
764
+ "content": "<extra_id_91>",
765
+ "lstrip": false,
766
+ "normalized": false,
767
+ "rstrip": false,
768
+ "single_word": false,
769
+ "special": true
770
+ },
771
+ "213": {
772
+ "content": "<extra_id_92>",
773
+ "lstrip": false,
774
+ "normalized": false,
775
+ "rstrip": false,
776
+ "single_word": false,
777
+ "special": true
778
+ },
779
+ "214": {
780
+ "content": "<extra_id_93>",
781
+ "lstrip": false,
782
+ "normalized": false,
783
+ "rstrip": false,
784
+ "single_word": false,
785
+ "special": true
786
+ },
787
+ "215": {
788
+ "content": "<extra_id_94>",
789
+ "lstrip": false,
790
+ "normalized": false,
791
+ "rstrip": false,
792
+ "single_word": false,
793
+ "special": true
794
+ },
795
+ "216": {
796
+ "content": "<extra_id_95>",
797
+ "lstrip": false,
798
+ "normalized": false,
799
+ "rstrip": false,
800
+ "single_word": false,
801
+ "special": true
802
+ },
803
+ "217": {
804
+ "content": "<extra_id_96>",
805
+ "lstrip": false,
806
+ "normalized": false,
807
+ "rstrip": false,
808
+ "single_word": false,
809
+ "special": true
810
+ },
811
+ "218": {
812
+ "content": "<extra_id_97>",
813
+ "lstrip": false,
814
+ "normalized": false,
815
+ "rstrip": false,
816
+ "single_word": false,
817
+ "special": true
818
+ },
819
+ "219": {
820
+ "content": "<extra_id_98>",
821
+ "lstrip": false,
822
+ "normalized": false,
823
+ "rstrip": false,
824
+ "single_word": false,
825
+ "special": true
826
+ },
827
+ "220": {
828
+ "content": "<extra_id_99>",
829
+ "lstrip": false,
830
+ "normalized": false,
831
+ "rstrip": false,
832
+ "single_word": false,
833
+ "special": true
834
+ },
835
+ "221": {
836
+ "content": ".",
837
+ "lstrip": false,
838
+ "normalized": true,
839
+ "rstrip": false,
840
+ "single_word": false,
841
+ "special": false
842
+ },
843
+ "222": {
844
+ "content": "6",
845
+ "lstrip": false,
846
+ "normalized": true,
847
+ "rstrip": false,
848
+ "single_word": false,
849
+ "special": false
850
+ },
851
+ "223": {
852
+ "content": "7",
853
+ "lstrip": false,
854
+ "normalized": true,
855
+ "rstrip": false,
856
+ "single_word": false,
857
+ "special": false
858
+ },
859
+ "224": {
860
+ "content": "8",
861
+ "lstrip": false,
862
+ "normalized": true,
863
+ "rstrip": false,
864
+ "single_word": false,
865
+ "special": false
866
+ },
867
+ "225": {
868
+ "content": "<",
869
+ "lstrip": false,
870
+ "normalized": true,
871
+ "rstrip": false,
872
+ "single_word": false,
873
+ "special": false
874
+ },
875
+ "226": {
876
+ "content": ">",
877
+ "lstrip": false,
878
+ "normalized": true,
879
+ "rstrip": false,
880
+ "single_word": false,
881
+ "special": false
882
+ },
883
+ "227": {
884
+ "content": "Ag",
885
+ "lstrip": false,
886
+ "normalized": true,
887
+ "rstrip": false,
888
+ "single_word": false,
889
+ "special": false
890
+ },
891
+ "228": {
892
+ "content": "Al",
893
+ "lstrip": false,
894
+ "normalized": true,
895
+ "rstrip": false,
896
+ "single_word": false,
897
+ "special": false
898
+ },
899
+ "229": {
900
+ "content": "Ar",
901
+ "lstrip": false,
902
+ "normalized": true,
903
+ "rstrip": false,
904
+ "single_word": false,
905
+ "special": false
906
+ },
907
+ "230": {
908
+ "content": "As",
909
+ "lstrip": false,
910
+ "normalized": true,
911
+ "rstrip": false,
912
+ "single_word": false,
913
+ "special": false
914
+ },
915
+ "231": {
916
+ "content": "Au",
917
+ "lstrip": false,
918
+ "normalized": true,
919
+ "rstrip": false,
920
+ "single_word": false,
921
+ "special": false
922
+ },
923
+ "232": {
924
+ "content": "Ba",
925
+ "lstrip": false,
926
+ "normalized": true,
927
+ "rstrip": false,
928
+ "single_word": false,
929
+ "special": false
930
+ },
931
+ "233": {
932
+ "content": "Bi",
933
+ "lstrip": false,
934
+ "normalized": true,
935
+ "rstrip": false,
936
+ "single_word": false,
937
+ "special": false
938
+ },
939
+ "234": {
940
+ "content": "Ca",
941
+ "lstrip": false,
942
+ "normalized": true,
943
+ "rstrip": false,
944
+ "single_word": false,
945
+ "special": false
946
+ },
947
+ "235": {
948
+ "content": "Cu",
949
+ "lstrip": false,
950
+ "normalized": true,
951
+ "rstrip": false,
952
+ "single_word": false,
953
+ "special": false
954
+ },
955
+ "236": {
956
+ "content": "Fe",
957
+ "lstrip": false,
958
+ "normalized": true,
959
+ "rstrip": false,
960
+ "single_word": false,
961
+ "special": false
962
+ },
963
+ "237": {
964
+ "content": "Ge",
965
+ "lstrip": false,
966
+ "normalized": true,
967
+ "rstrip": false,
968
+ "single_word": false,
969
+ "special": false
970
+ },
971
+ "238": {
972
+ "content": "Hg",
973
+ "lstrip": false,
974
+ "normalized": true,
975
+ "rstrip": false,
976
+ "single_word": false,
977
+ "special": false
978
+ },
979
+ "239": {
980
+ "content": "K",
981
+ "lstrip": false,
982
+ "normalized": true,
983
+ "rstrip": false,
984
+ "single_word": false,
985
+ "special": false
986
+ },
987
+ "240": {
988
+ "content": "Li",
989
+ "lstrip": false,
990
+ "normalized": true,
991
+ "rstrip": false,
992
+ "single_word": false,
993
+ "special": false
994
+ },
995
+ "241": {
996
+ "content": "Mg",
997
+ "lstrip": false,
998
+ "normalized": true,
999
+ "rstrip": false,
1000
+ "single_word": false,
1001
+ "special": false
1002
+ },
1003
+ "242": {
1004
+ "content": "Mn",
1005
+ "lstrip": false,
1006
+ "normalized": true,
1007
+ "rstrip": false,
1008
+ "single_word": false,
1009
+ "special": false
1010
+ },
1011
+ "243": {
1012
+ "content": "Mo",
1013
+ "lstrip": false,
1014
+ "normalized": true,
1015
+ "rstrip": false,
1016
+ "single_word": false,
1017
+ "special": false
1018
+ },
1019
+ "244": {
1020
+ "content": "Na",
1021
+ "lstrip": false,
1022
+ "normalized": true,
1023
+ "rstrip": false,
1024
+ "single_word": false,
1025
+ "special": false
1026
+ },
1027
+ "245": {
1028
+ "content": "Nd",
1029
+ "lstrip": false,
1030
+ "normalized": true,
1031
+ "rstrip": false,
1032
+ "single_word": false,
1033
+ "special": false
1034
+ },
1035
+ "246": {
1036
+ "content": "Ni",
1037
+ "lstrip": false,
1038
+ "normalized": true,
1039
+ "rstrip": false,
1040
+ "single_word": false,
1041
+ "special": false
1042
+ },
1043
+ "247": {
1044
+ "content": "P",
1045
+ "lstrip": false,
1046
+ "normalized": true,
1047
+ "rstrip": false,
1048
+ "single_word": false,
1049
+ "special": false
1050
+ },
1051
+ "248": {
1052
+ "content": "Pb",
1053
+ "lstrip": false,
1054
+ "normalized": true,
1055
+ "rstrip": false,
1056
+ "single_word": false,
1057
+ "special": false
1058
+ },
1059
+ "249": {
1060
+ "content": "Pd",
1061
+ "lstrip": false,
1062
+ "normalized": true,
1063
+ "rstrip": false,
1064
+ "single_word": false,
1065
+ "special": false
1066
+ },
1067
+ "250": {
1068
+ "content": "Pt",
1069
+ "lstrip": false,
1070
+ "normalized": true,
1071
+ "rstrip": false,
1072
+ "single_word": false,
1073
+ "special": false
1074
+ },
1075
+ "251": {
1076
+ "content": "Re",
1077
+ "lstrip": false,
1078
+ "normalized": true,
1079
+ "rstrip": false,
1080
+ "single_word": false,
1081
+ "special": false
1082
+ },
1083
+ "252": {
1084
+ "content": "Rh",
1085
+ "lstrip": false,
1086
+ "normalized": true,
1087
+ "rstrip": false,
1088
+ "single_word": false,
1089
+ "special": false
1090
+ },
1091
+ "253": {
1092
+ "content": "Ru",
1093
+ "lstrip": false,
1094
+ "normalized": true,
1095
+ "rstrip": false,
1096
+ "single_word": false,
1097
+ "special": false
1098
+ },
1099
+ "254": {
1100
+ "content": "Sb",
1101
+ "lstrip": false,
1102
+ "normalized": true,
1103
+ "rstrip": false,
1104
+ "single_word": false,
1105
+ "special": false
1106
+ },
1107
+ "255": {
1108
+ "content": "Si",
1109
+ "lstrip": false,
1110
+ "normalized": true,
1111
+ "rstrip": false,
1112
+ "single_word": false,
1113
+ "special": false
1114
+ },
1115
+ "256": {
1116
+ "content": "Sm",
1117
+ "lstrip": false,
1118
+ "normalized": true,
1119
+ "rstrip": false,
1120
+ "single_word": false,
1121
+ "special": false
1122
+ },
1123
+ "257": {
1124
+ "content": "Ta",
1125
+ "lstrip": false,
1126
+ "normalized": true,
1127
+ "rstrip": false,
1128
+ "single_word": false,
1129
+ "special": false
1130
+ },
1131
+ "258": {
1132
+ "content": "Ti",
1133
+ "lstrip": false,
1134
+ "normalized": true,
1135
+ "rstrip": false,
1136
+ "single_word": false,
1137
+ "special": false
1138
+ },
1139
+ "259": {
1140
+ "content": "Tl",
1141
+ "lstrip": false,
1142
+ "normalized": true,
1143
+ "rstrip": false,
1144
+ "single_word": false,
1145
+ "special": false
1146
+ },
1147
+ "260": {
1148
+ "content": "W",
1149
+ "lstrip": false,
1150
+ "normalized": true,
1151
+ "rstrip": false,
1152
+ "single_word": false,
1153
+ "special": false
1154
+ },
1155
+ "261": {
1156
+ "content": "Yb",
1157
+ "lstrip": false,
1158
+ "normalized": true,
1159
+ "rstrip": false,
1160
+ "single_word": false,
1161
+ "special": false
1162
+ },
1163
+ "262": {
1164
+ "content": "Zn",
1165
+ "lstrip": false,
1166
+ "normalized": true,
1167
+ "rstrip": false,
1168
+ "single_word": false,
1169
+ "special": false
1170
+ },
1171
+ "263": {
1172
+ "content": "Zr",
1173
+ "lstrip": false,
1174
+ "normalized": true,
1175
+ "rstrip": false,
1176
+ "single_word": false,
1177
+ "special": false
1178
+ },
1179
+ "264": {
1180
+ "content": "e",
1181
+ "lstrip": false,
1182
+ "normalized": true,
1183
+ "rstrip": false,
1184
+ "single_word": false,
1185
+ "special": false
1186
+ },
1187
+ "265": {
1188
+ "content": "p",
1189
+ "lstrip": false,
1190
+ "normalized": true,
1191
+ "rstrip": false,
1192
+ "single_word": false,
1193
+ "special": false
1194
+ },
1195
+ "266": {
1196
+ "content": "REACTANT:",
1197
+ "lstrip": false,
1198
+ "normalized": false,
1199
+ "rstrip": false,
1200
+ "single_word": false,
1201
+ "special": true
1202
+ },
1203
+ "267": {
1204
+ "content": "REAGENT:",
1205
+ "lstrip": false,
1206
+ "normalized": false,
1207
+ "rstrip": false,
1208
+ "single_word": false,
1209
+ "special": true
1210
+ }
1211
+ },
1212
+ "additional_special_tokens": [
1213
+ "<extra_id_0>",
1214
+ "<extra_id_1>",
1215
+ "<extra_id_2>",
1216
+ "<extra_id_3>",
1217
+ "<extra_id_4>",
1218
+ "<extra_id_5>",
1219
+ "<extra_id_6>",
1220
+ "<extra_id_7>",
1221
+ "<extra_id_8>",
1222
+ "<extra_id_9>",
1223
+ "<extra_id_10>",
1224
+ "<extra_id_11>",
1225
+ "<extra_id_12>",
1226
+ "<extra_id_13>",
1227
+ "<extra_id_14>",
1228
+ "<extra_id_15>",
1229
+ "<extra_id_16>",
1230
+ "<extra_id_17>",
1231
+ "<extra_id_18>",
1232
+ "<extra_id_19>",
1233
+ "<extra_id_20>",
1234
+ "<extra_id_21>",
1235
+ "<extra_id_22>",
1236
+ "<extra_id_23>",
1237
+ "<extra_id_24>",
1238
+ "<extra_id_25>",
1239
+ "<extra_id_26>",
1240
+ "<extra_id_27>",
1241
+ "<extra_id_28>",
1242
+ "<extra_id_29>",
1243
+ "<extra_id_30>",
1244
+ "<extra_id_31>",
1245
+ "<extra_id_32>",
1246
+ "<extra_id_33>",
1247
+ "<extra_id_34>",
1248
+ "<extra_id_35>",
1249
+ "<extra_id_36>",
1250
+ "<extra_id_37>",
1251
+ "<extra_id_38>",
1252
+ "<extra_id_39>",
1253
+ "<extra_id_40>",
1254
+ "<extra_id_41>",
1255
+ "<extra_id_42>",
1256
+ "<extra_id_43>",
1257
+ "<extra_id_44>",
1258
+ "<extra_id_45>",
1259
+ "<extra_id_46>",
1260
+ "<extra_id_47>",
1261
+ "<extra_id_48>",
1262
+ "<extra_id_49>",
1263
+ "<extra_id_50>",
1264
+ "<extra_id_51>",
1265
+ "<extra_id_52>",
1266
+ "<extra_id_53>",
1267
+ "<extra_id_54>",
1268
+ "<extra_id_55>",
1269
+ "<extra_id_56>",
1270
+ "<extra_id_57>",
1271
+ "<extra_id_58>",
1272
+ "<extra_id_59>",
1273
+ "<extra_id_60>",
1274
+ "<extra_id_61>",
1275
+ "<extra_id_62>",
1276
+ "<extra_id_63>",
1277
+ "<extra_id_64>",
1278
+ "<extra_id_65>",
1279
+ "<extra_id_66>",
1280
+ "<extra_id_67>",
1281
+ "<extra_id_68>",
1282
+ "<extra_id_69>",
1283
+ "<extra_id_70>",
1284
+ "<extra_id_71>",
1285
+ "<extra_id_72>",
1286
+ "<extra_id_73>",
1287
+ "<extra_id_74>",
1288
+ "<extra_id_75>",
1289
+ "<extra_id_76>",
1290
+ "<extra_id_77>",
1291
+ "<extra_id_78>",
1292
+ "<extra_id_79>",
1293
+ "<extra_id_80>",
1294
+ "<extra_id_81>",
1295
+ "<extra_id_82>",
1296
+ "<extra_id_83>",
1297
+ "<extra_id_84>",
1298
+ "<extra_id_85>",
1299
+ "<extra_id_86>",
1300
+ "<extra_id_87>",
1301
+ "<extra_id_88>",
1302
+ "<extra_id_89>",
1303
+ "<extra_id_90>",
1304
+ "<extra_id_91>",
1305
+ "<extra_id_92>",
1306
+ "<extra_id_93>",
1307
+ "<extra_id_94>",
1308
+ "<extra_id_95>",
1309
+ "<extra_id_96>",
1310
+ "<extra_id_97>",
1311
+ "<extra_id_98>",
1312
+ "<extra_id_99>",
1313
+ "REACTANT:",
1314
+ "REAGENT:"
1315
+ ],
1316
+ "clean_up_tokenization_spaces": true,
1317
+ "eos_token": "</s>",
1318
+ "extra_ids": 100,
1319
+ "max_length": 100,
1320
+ "model_max_length": 1000000000000000019884624838656,
1321
+ "pad_token": "<pad>",
1322
+ "return_tensors": "pt",
1323
+ "stride": 0,
1324
+ "tokenizer_class": "T5Tokenizer",
1325
+ "truncation_side": "right",
1326
+ "truncation_strategy": "longest_first",
1327
+ "unk_token": "<unk>"
1328
+ }
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/aux/trainer_state.json ADDED
@@ -0,0 +1,149 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.002982336562126875,
3
+ "best_model_checkpoint": "t5/checkpoint-25566",
4
+ "epoch": 8.0,
5
+ "eval_steps": 500,
6
+ "global_step": 102264,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 1.0,
13
+ "grad_norm": 0.09756029397249222,
14
+ "learning_rate": 1.9600000000000002e-05,
15
+ "loss": 0.0059,
16
+ "step": 12783
17
+ },
18
+ {
19
+ "epoch": 1.0,
20
+ "eval_accuracy": 0.9804666666666667,
21
+ "eval_loss": 0.0031315775122493505,
22
+ "eval_runtime": 472.802,
23
+ "eval_samples_per_second": 63.452,
24
+ "eval_steps_per_second": 0.497,
25
+ "step": 12783
26
+ },
27
+ {
28
+ "epoch": 2.0,
29
+ "grad_norm": 0.0030228213872760534,
30
+ "learning_rate": 1.9200000000000003e-05,
31
+ "loss": 0.0022,
32
+ "step": 25566
33
+ },
34
+ {
35
+ "epoch": 2.0,
36
+ "eval_accuracy": 0.9818666666666667,
37
+ "eval_loss": 0.002982336562126875,
38
+ "eval_runtime": 474.0664,
39
+ "eval_samples_per_second": 63.282,
40
+ "eval_steps_per_second": 0.496,
41
+ "step": 25566
42
+ },
43
+ {
44
+ "epoch": 3.0,
45
+ "grad_norm": 0.00019585879635997117,
46
+ "learning_rate": 1.88e-05,
47
+ "loss": 0.0012,
48
+ "step": 38349
49
+ },
50
+ {
51
+ "epoch": 3.0,
52
+ "eval_accuracy": 0.9825,
53
+ "eval_loss": 0.0032824024092406034,
54
+ "eval_runtime": 478.2854,
55
+ "eval_samples_per_second": 62.724,
56
+ "eval_steps_per_second": 0.491,
57
+ "step": 38349
58
+ },
59
+ {
60
+ "epoch": 4.0,
61
+ "grad_norm": 0.004981159698218107,
62
+ "learning_rate": 1.8400000000000003e-05,
63
+ "loss": 0.0007,
64
+ "step": 51132
65
+ },
66
+ {
67
+ "epoch": 4.0,
68
+ "eval_accuracy": 0.9826,
69
+ "eval_loss": 0.003532745875418186,
70
+ "eval_runtime": 469.3109,
71
+ "eval_samples_per_second": 63.924,
72
+ "eval_steps_per_second": 0.501,
73
+ "step": 51132
74
+ },
75
+ {
76
+ "epoch": 5.0,
77
+ "grad_norm": 0.002326933667063713,
78
+ "learning_rate": 1.8e-05,
79
+ "loss": 0.0005,
80
+ "step": 63915
81
+ },
82
+ {
83
+ "epoch": 5.0,
84
+ "eval_accuracy": 0.9821666666666666,
85
+ "eval_loss": 0.003947438672184944,
86
+ "eval_runtime": 471.3239,
87
+ "eval_samples_per_second": 63.65,
88
+ "eval_steps_per_second": 0.499,
89
+ "step": 63915
90
+ },
91
+ {
92
+ "epoch": 6.0,
93
+ "grad_norm": 0.034791357815265656,
94
+ "learning_rate": 1.76e-05,
95
+ "loss": 0.0004,
96
+ "step": 76698
97
+ },
98
+ {
99
+ "epoch": 6.0,
100
+ "eval_accuracy": 0.9815666666666667,
101
+ "eval_loss": 0.0041623483411967754,
102
+ "eval_runtime": 470.7737,
103
+ "eval_samples_per_second": 63.725,
104
+ "eval_steps_per_second": 0.499,
105
+ "step": 76698
106
+ },
107
+ {
108
+ "epoch": 7.0,
109
+ "grad_norm": 9.868334018392488e-05,
110
+ "learning_rate": 1.72e-05,
111
+ "loss": 0.0003,
112
+ "step": 89481
113
+ },
114
+ {
115
+ "epoch": 7.0,
116
+ "eval_accuracy": 0.9822666666666666,
117
+ "eval_loss": 0.004203655291348696,
118
+ "eval_runtime": 471.0374,
119
+ "eval_samples_per_second": 63.689,
120
+ "eval_steps_per_second": 0.499,
121
+ "step": 89481
122
+ },
123
+ {
124
+ "epoch": 8.0,
125
+ "grad_norm": 0.002833570586517453,
126
+ "learning_rate": 1.6800000000000002e-05,
127
+ "loss": 0.0003,
128
+ "step": 102264
129
+ },
130
+ {
131
+ "epoch": 8.0,
132
+ "eval_accuracy": 0.9826,
133
+ "eval_loss": 0.0045397114008665085,
134
+ "eval_runtime": 469.6081,
135
+ "eval_samples_per_second": 63.883,
136
+ "eval_steps_per_second": 0.5,
137
+ "step": 102264
138
+ }
139
+ ],
140
+ "logging_steps": 500,
141
+ "max_steps": 639150,
142
+ "num_input_tokens_seen": 0,
143
+ "num_train_epochs": 50,
144
+ "save_steps": 500,
145
+ "total_flos": 5.937159320662856e+17,
146
+ "train_batch_size": 32,
147
+ "trial_name": null,
148
+ "trial_params": null
149
+ }
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/benchmarks/metadata.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "benchmark_semantics": "benchmark_specific_finetuned_forward_checkpoint",
3
+ "benchmark_variant": "forward_product_prediction",
4
+ "blank_reagent_value": " ",
5
+ "built_at": "1970-01-01T00:00:00+00:00",
6
+ "checkpoint_id": "sagawa_reactiont5v2_forward_uspto_mit",
7
+ "dataset_id": "sagawa_reactiont5v2_forward_uspto_mit_test",
8
+ "dataset_ref": "uspto_mit_forward_project_release",
9
+ "native_input_format": "csv_forward_reaction_rows",
10
+ "native_sources": [
11
+ {
12
+ "notes": [
13
+ "Exact project USPTO_MIT test CSV already preserved inside the normalized checkpoint bundle.",
14
+ "Forward CSV columns are REACTANT, REAGENT, and PRODUCT."
15
+ ],
16
+ "origin": "existing_bundle_file",
17
+ "relative_path": "aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv",
18
+ "source_url": null
19
+ }
20
+ ],
21
+ "notes": [
22
+ "Native forward benchmark split for the public USPTO_MIT product-prediction task.",
23
+ "The normalized JSONL preserves reactants, reagents, and expected product(s) rather than the backward exact-match schema.",
24
+ "This is the native USPTO_MIT fine-tuned forward checkpoint benchmark split."
25
+ ],
26
+ "num_examples": 40000,
27
+ "record_schema": {
28
+ "inputs": [
29
+ "reactants",
30
+ "reagents"
31
+ ],
32
+ "targets": [
33
+ "expected_products"
34
+ ]
35
+ },
36
+ "serialized_input_format": "REACTANT:<reactants>REAGENT:<reagents>",
37
+ "split": "test",
38
+ "task_direction": "forward"
39
+ }
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/manifest.json ADDED
@@ -0,0 +1,137 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auxiliary_files": [
3
+ "aux/.gitattributes",
4
+ "aux/README.md",
5
+ "aux/UPSTREAM_URL.txt",
6
+ "aux/config.json",
7
+ "aux/generation_config.json",
8
+ "aux/hf_model_info.json",
9
+ "aux/provenance/README.txt",
10
+ "aux/provenance/ReactionT5v2_source_README.md",
11
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv",
12
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
13
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
14
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv",
15
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/train.csv",
16
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/val.csv",
17
+ "aux/provenance/reactiont5v2_forward_split_provenance.json",
18
+ "aux/special_tokens_map.json",
19
+ "aux/tokenizer.json",
20
+ "aux/tokenizer_config.json",
21
+ "aux/trainer_state.json",
22
+ "aux/training_args.bin",
23
+ "benchmarks/default_test.jsonl",
24
+ "benchmarks/metadata.json"
25
+ ],
26
+ "benchmark_datasets": [
27
+ {
28
+ "benchmark_semantics": "benchmark_specific_finetuned_forward_checkpoint",
29
+ "benchmark_variant": "forward_product_prediction",
30
+ "blank_reagent_value": " ",
31
+ "dataset_id": "sagawa_reactiont5v2_forward_uspto_mit_test",
32
+ "dataset_ref": "uspto_mit_forward_project_release",
33
+ "default": true,
34
+ "format": "jsonl",
35
+ "metadata_path": "benchmarks/metadata.json",
36
+ "native_input_format": "csv_forward_reaction_rows",
37
+ "native_source_paths": [
38
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv"
39
+ ],
40
+ "notes": [
41
+ "Native forward benchmark split for the public USPTO_MIT product-prediction task.",
42
+ "The normalized JSONL preserves reactants, reagents, and expected product(s) rather than the backward exact-match schema.",
43
+ "This is the native USPTO_MIT fine-tuned forward checkpoint benchmark split."
44
+ ],
45
+ "num_examples": 40000,
46
+ "path": "benchmarks/default_test.jsonl",
47
+ "record_schema": {
48
+ "inputs": [
49
+ "reactants",
50
+ "reagents"
51
+ ],
52
+ "targets": [
53
+ "expected_products"
54
+ ]
55
+ },
56
+ "serialized_input_format": "REACTANT:<reactants>REAGENT:<reagents>",
57
+ "split": "test",
58
+ "task_direction": "forward"
59
+ }
60
+ ],
61
+ "checkpoint_id": "sagawa_reactiont5v2_forward_uspto_mit",
62
+ "contained_artifact_bindings": [],
63
+ "conversion": {
64
+ "converted_at": "1970-01-01T00:00:00+00:00",
65
+ "source_format": "huggingface_safetensors",
66
+ "status": "converted",
67
+ "tool_version": "initial"
68
+ },
69
+ "display_name": "sagawa/ReactionT5v2-forward-USPTO_MIT",
70
+ "embedding_supported": false,
71
+ "fine_tuning": {
72
+ "adapter_overlay_dir": "adapters",
73
+ "base_weights_path": "model.safetensors",
74
+ "notes": [
75
+ "This package preserves immutable base weights in model.safetensors and reserves adapters/ for future fine-tuning or LoRA overlays."
76
+ ]
77
+ },
78
+ "format_version": 1,
79
+ "generated_at": "1970-01-01T00:00:00+00:00",
80
+ "input_schema": {
81
+ "blank_reagent_value": " ",
82
+ "reactants_field": "REACTANT",
83
+ "reagents_field": "REAGENT",
84
+ "serialized_format": "REACTANT:<reactants>REAGENT:<reagents>",
85
+ "target_field": "PRODUCT"
86
+ },
87
+ "model_id": "reactiont5v2_forward",
88
+ "normalized_checkpoint_id": "sagawa_reactiont5v2_forward_uspto_mit",
89
+ "package_layout": {
90
+ "adapters_dir": "adapters",
91
+ "aux_dir": "aux",
92
+ "benchmarks_dir": "benchmarks",
93
+ "weights_path": "model.safetensors"
94
+ },
95
+ "source_registry_refs": [
96
+ {
97
+ "checkpoint_id": "sagawa_reactiont5v2_forward_uspto_mit",
98
+ "model_id": "reactiont5v2_forward"
99
+ }
100
+ ],
101
+ "split_provenance": [
102
+ {
103
+ "canonical_split_family": "reactiont5v2_project_uspto_mit_forward_train_valid_test",
104
+ "dataset_ref": "uspto_mit_forward_project_release",
105
+ "split_names": [
106
+ "train",
107
+ "valid",
108
+ "test"
109
+ ],
110
+ "stage": "fine_tuning_and_benchmark_evaluation"
111
+ }
112
+ ],
113
+ "task_direction": "forward",
114
+ "training_dataset_refs": [
115
+ "uspto_mit_forward_project_release"
116
+ ],
117
+ "training_stage_bindings": [
118
+ {
119
+ "canonical_split_family": "reactiont5v2_project_uspto_mit_forward_train_valid_test",
120
+ "dataset_ref": "uspto_mit_forward_project_release",
121
+ "split_names": [
122
+ "train",
123
+ "valid",
124
+ "test"
125
+ ],
126
+ "stage": "fine_tuning_and_benchmark_evaluation"
127
+ }
128
+ ],
129
+ "weights": {
130
+ "dtype_summary": "upstream_preserved",
131
+ "framework": "safetensors",
132
+ "path": "model.safetensors",
133
+ "tensor_namespace": "native"
134
+ },
135
+ "wrapper_config": "wrapper_config.json",
136
+ "wrapper_family": "seq2seq_t5"
137
+ }
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/provenance.json ADDED
@@ -0,0 +1,495 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "conversion_notes": [
3
+ "Upstream Hugging Face safetensors reused verbatim; no tensor conversion performed."
4
+ ],
5
+ "copied_sidecars": [
6
+ "aux/.gitattributes",
7
+ "aux/README.md",
8
+ "aux/UPSTREAM_URL.txt",
9
+ "aux/config.json",
10
+ "aux/generation_config.json",
11
+ "aux/hf_model_info.json",
12
+ "aux/provenance/README.txt",
13
+ "aux/provenance/ReactionT5v2_source_README.md",
14
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv",
15
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
16
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
17
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/test.csv",
18
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/train.csv",
19
+ "aux/provenance/datasets/USPTO_MIT/MIT_separated/val.csv",
20
+ "aux/provenance/reactiont5v2_forward_split_provenance.json",
21
+ "aux/special_tokens_map.json",
22
+ "aux/tokenizer.json",
23
+ "aux/tokenizer_config.json",
24
+ "aux/trainer_state.json",
25
+ "aux/training_args.bin"
26
+ ],
27
+ "inner_members": [],
28
+ "normalized_checkpoint_id": "sagawa_reactiont5v2_forward_uspto_mit",
29
+ "source_artifacts": [
30
+ {
31
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/README.md",
32
+ "kind": "local_repo_file",
33
+ "path": "datasets/forward_models/reactionT5v2/README.md",
34
+ "role": "source_artifact",
35
+ "sha256": "8584b84f91f41356ee736fa7cfb797d99d82aeeade504ef0f4097388e3c9e890",
36
+ "size": 1414
37
+ },
38
+ {
39
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/.gitignore",
40
+ "kind": "local_repo_file",
41
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/.gitignore",
42
+ "role": "source_artifact",
43
+ "sha256": "684888c0ebb17f374298b65ee2807526c066094c701bcc7ebbe1c1095f494fc1",
44
+ "size": 1
45
+ },
46
+ {
47
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/.gitattributes.metadata",
48
+ "kind": "local_repo_file",
49
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/.gitattributes.metadata",
50
+ "role": "source_artifact",
51
+ "sha256": "8dbb80384bb198b63c09c92fa39320671fe97465152f8633c895ccdd95986236",
52
+ "size": 101
53
+ },
54
+ {
55
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/README.md.metadata",
56
+ "kind": "local_repo_file",
57
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/README.md.metadata",
58
+ "role": "source_artifact",
59
+ "sha256": "0495464d7459458caec3719798d56070718922d3bd6e1b4e6c00e76443d33d93",
60
+ "size": 101
61
+ },
62
+ {
63
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/config.json.metadata",
64
+ "kind": "local_repo_file",
65
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/config.json.metadata",
66
+ "role": "source_artifact",
67
+ "sha256": "34a1a79f11cdac899282589c1377c639fe1d923d36ab9987e23f7ba7c3f8a0e9",
68
+ "size": 101
69
+ },
70
+ {
71
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/generation_config.json.metadata",
72
+ "kind": "local_repo_file",
73
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/generation_config.json.metadata",
74
+ "role": "source_artifact",
75
+ "sha256": "2f59b6ddb10e05e11d0a053d261e3063e1207afe4a757153b15c8a1a4a5e07ce",
76
+ "size": 101
77
+ },
78
+ {
79
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/model.safetensors.metadata",
80
+ "kind": "local_repo_file",
81
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/model.safetensors.metadata",
82
+ "role": "source_artifact",
83
+ "sha256": "decd5fd8ab381657d595f428d8518fbf93a3d0fd1070079408ca588f0926a6a4",
84
+ "size": 125
85
+ },
86
+ {
87
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/special_tokens_map.json.metadata",
88
+ "kind": "local_repo_file",
89
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/special_tokens_map.json.metadata",
90
+ "role": "source_artifact",
91
+ "sha256": "703d4f5fe17f554f1bfcf4c9dacadb2670ec93660021c11c884c6e716013b539",
92
+ "size": 100
93
+ },
94
+ {
95
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/tokenizer.json.metadata",
96
+ "kind": "local_repo_file",
97
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/tokenizer.json.metadata",
98
+ "role": "source_artifact",
99
+ "sha256": "dcd06983dbea11ce92bdc72a7a109dd53936324020404f9c3a6e380892f82ed5",
100
+ "size": 101
101
+ },
102
+ {
103
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/tokenizer_config.json.metadata",
104
+ "kind": "local_repo_file",
105
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/tokenizer_config.json.metadata",
106
+ "role": "source_artifact",
107
+ "sha256": "e58e87ec0be6fdd7ab16a8eddf74b03a8e4b2cfe7c0ec17c1ce97a50303c908e",
108
+ "size": 101
109
+ },
110
+ {
111
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/trainer_state.json.metadata",
112
+ "kind": "local_repo_file",
113
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/trainer_state.json.metadata",
114
+ "role": "source_artifact",
115
+ "sha256": "23c346163e7f13692ee788f8576a84266aa436d70a5eae13367281375ab071b9",
116
+ "size": 101
117
+ },
118
+ {
119
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/training_args.bin.metadata",
120
+ "kind": "local_repo_file",
121
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.cache/huggingface/download/training_args.bin.metadata",
122
+ "role": "source_artifact",
123
+ "sha256": "e05c87b1eb8806e44385596c115f70806891932e31914a6040094fda76368c1d",
124
+ "size": 124
125
+ },
126
+ {
127
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.gitattributes",
128
+ "kind": "local_repo_file",
129
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/.gitattributes",
130
+ "role": "source_artifact",
131
+ "sha256": "11ad7efa24975ee4b0c3c3a38ed18737f0658a5f75a0a96787b576a78a023361",
132
+ "size": 1519
133
+ },
134
+ {
135
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/README.md",
136
+ "kind": "local_repo_file",
137
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/README.md",
138
+ "role": "source_artifact",
139
+ "sha256": "1a689168cf12a60fb49166cdc5245308f6ef18d28d8fec165295edf7404449c2",
140
+ "size": 4822
141
+ },
142
+ {
143
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/UPSTREAM_URL.txt",
144
+ "kind": "local_repo_file",
145
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/UPSTREAM_URL.txt",
146
+ "role": "source_artifact",
147
+ "sha256": "005ab587497141bf8a42fef5e820a3c2956893dbf444538722cbf010554f6a6a",
148
+ "size": 101
149
+ },
150
+ {
151
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/config.json",
152
+ "kind": "local_repo_file",
153
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/config.json",
154
+ "role": "source_artifact",
155
+ "sha256": "a80eac6938e51ccabfc6a8b281497c6553626145e57b2268101f718fbaa034a2",
156
+ "size": 825
157
+ },
158
+ {
159
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/generation_config.json",
160
+ "kind": "local_repo_file",
161
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/generation_config.json",
162
+ "role": "source_artifact",
163
+ "sha256": "aed028489d41ae6a673920da9d103f2f58a650c3fbc50bd880b35e30a481dba6",
164
+ "size": 133
165
+ },
166
+ {
167
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/hf_model_info.json",
168
+ "kind": "local_repo_file",
169
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/hf_model_info.json",
170
+ "role": "source_artifact",
171
+ "sha256": "116da2894aad609845ba7977cbcdbbba5eddd02805fd2648419595464de98a91",
172
+ "size": 1967
173
+ },
174
+ {
175
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/model.safetensors",
176
+ "kind": "local_repo_file",
177
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/model.safetensors",
178
+ "role": "source_artifact",
179
+ "sha256": "69c23ca364e63e0e2be22d114652aad9f3df4e161d43936eee0dc3ce74a87fb7",
180
+ "size": 794597224
181
+ },
182
+ {
183
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/special_tokens_map.json",
184
+ "kind": "local_repo_file",
185
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/special_tokens_map.json",
186
+ "role": "source_artifact",
187
+ "sha256": "0f385cf62b07c1b63aaa84d4c86a313c751df7039d958a7d7bd7e723d933bcd9",
188
+ "size": 15428
189
+ },
190
+ {
191
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/tokenizer.json",
192
+ "kind": "local_repo_file",
193
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/tokenizer.json",
194
+ "role": "source_artifact",
195
+ "sha256": "511dfd8bb25c19cccae8d7d996c1ff0f1b5263207be4cdd275f23b75da8113b6",
196
+ "size": 36514
197
+ },
198
+ {
199
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/tokenizer_config.json",
200
+ "kind": "local_repo_file",
201
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/tokenizer_config.json",
202
+ "role": "source_artifact",
203
+ "sha256": "ac38d16067df56f00dfa2c2e0d1546a6ff08aa37814662a6fc38dea66cd13d9d",
204
+ "size": 28602
205
+ },
206
+ {
207
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/trainer_state.json",
208
+ "kind": "local_repo_file",
209
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/trainer_state.json",
210
+ "role": "source_artifact",
211
+ "sha256": "f89bde13ce68202577b7eddf752376632a40ce296e5c005da51e37bba6d195ac",
212
+ "size": 40605
213
+ },
214
+ {
215
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/training_args.bin",
216
+ "kind": "local_repo_file",
217
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward/training_args.bin",
218
+ "role": "source_artifact",
219
+ "sha256": "0d40cc65896c8c30098177fdb6440c4900ea381e1fc01b3348193f689f02c9ba",
220
+ "size": 5112
221
+ },
222
+ {
223
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/.gitignore",
224
+ "kind": "local_repo_file",
225
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/.gitignore",
226
+ "role": "source_artifact",
227
+ "sha256": "684888c0ebb17f374298b65ee2807526c066094c701bcc7ebbe1c1095f494fc1",
228
+ "size": 1
229
+ },
230
+ {
231
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/.gitattributes.metadata",
232
+ "kind": "local_repo_file",
233
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/.gitattributes.metadata",
234
+ "role": "source_artifact",
235
+ "sha256": "d810874057b9fbbab79e9bddd2c6f7d92dbdcf2c2f61bff1c7cede7995b9c31d",
236
+ "size": 101
237
+ },
238
+ {
239
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/README.md.metadata",
240
+ "kind": "local_repo_file",
241
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/README.md.metadata",
242
+ "role": "source_artifact",
243
+ "sha256": "d079b773989ec3afa54b35f0ba650987123cd7e3e5c505803dee5a09a286abdc",
244
+ "size": 101
245
+ },
246
+ {
247
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/config.json.metadata",
248
+ "kind": "local_repo_file",
249
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/config.json.metadata",
250
+ "role": "source_artifact",
251
+ "sha256": "e29e915afc6ce29582e4fbb666f8ad53949f5c6af3b748fc0cde078759170a82",
252
+ "size": 101
253
+ },
254
+ {
255
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/generation_config.json.metadata",
256
+ "kind": "local_repo_file",
257
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/generation_config.json.metadata",
258
+ "role": "source_artifact",
259
+ "sha256": "baac117440c18d95e8669a85e39beb377a298addabee514b74879d3c9a817476",
260
+ "size": 101
261
+ },
262
+ {
263
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/model.safetensors.metadata",
264
+ "kind": "local_repo_file",
265
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/model.safetensors.metadata",
266
+ "role": "source_artifact",
267
+ "sha256": "c21b4b9b37243be2e068eb0a3c1c19803b573f3ebfbf78779b0512f501d784b8",
268
+ "size": 125
269
+ },
270
+ {
271
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/special_tokens_map.json.metadata",
272
+ "kind": "local_repo_file",
273
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/special_tokens_map.json.metadata",
274
+ "role": "source_artifact",
275
+ "sha256": "32272b480cb55f0eadfd294c1276638055b51654f48fe26670dcc30bd8ff94ae",
276
+ "size": 100
277
+ },
278
+ {
279
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/tokenizer.json.metadata",
280
+ "kind": "local_repo_file",
281
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/tokenizer.json.metadata",
282
+ "role": "source_artifact",
283
+ "sha256": "36b542e97a9f4e4e051d17eabe6428742f01574fe552b71b36500af960d2ee01",
284
+ "size": 101
285
+ },
286
+ {
287
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/tokenizer_config.json.metadata",
288
+ "kind": "local_repo_file",
289
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/tokenizer_config.json.metadata",
290
+ "role": "source_artifact",
291
+ "sha256": "bf546f252ad7756f049c9edc1d114ebd769d6c8f9a6f82c8ab059ef63d75a5af",
292
+ "size": 101
293
+ },
294
+ {
295
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/trainer_state.json.metadata",
296
+ "kind": "local_repo_file",
297
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/trainer_state.json.metadata",
298
+ "role": "source_artifact",
299
+ "sha256": "2b8ec757b5caa233136eb0cfefbb572d01f736cbefa7d47c78f6341deae6ef97",
300
+ "size": 100
301
+ },
302
+ {
303
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/training_args.bin.metadata",
304
+ "kind": "local_repo_file",
305
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.cache/huggingface/download/training_args.bin.metadata",
306
+ "role": "source_artifact",
307
+ "sha256": "799bfb2cf2602b85a922a7043a3787624d74ebd856299b20c8543536a2bb0f46",
308
+ "size": 125
309
+ },
310
+ {
311
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.gitattributes",
312
+ "kind": "local_repo_file",
313
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/.gitattributes",
314
+ "role": "source_artifact",
315
+ "sha256": "11ad7efa24975ee4b0c3c3a38ed18737f0658a5f75a0a96787b576a78a023361",
316
+ "size": 1519
317
+ },
318
+ {
319
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/README.md",
320
+ "kind": "local_repo_file",
321
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/README.md",
322
+ "role": "source_artifact",
323
+ "sha256": "abc9ceaf22822bc2736909eac46f42f29a911c87bca6f29e1979ffe6e701904d",
324
+ "size": 4732
325
+ },
326
+ {
327
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/UPSTREAM_URL.txt",
328
+ "kind": "local_repo_file",
329
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/UPSTREAM_URL.txt",
330
+ "role": "source_artifact",
331
+ "sha256": "5a41716f340aff0d540fe7da15f12cd4b3818378a6922d8fbc242faf45c4998e",
332
+ "size": 111
333
+ },
334
+ {
335
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/config.json",
336
+ "kind": "local_repo_file",
337
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/config.json",
338
+ "role": "source_artifact",
339
+ "sha256": "c637cdad4f34d56a638ac6031e742a407b2ed6b9d2d674f1e8bc8cb02034ae90",
340
+ "size": 903
341
+ },
342
+ {
343
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/generation_config.json",
344
+ "kind": "local_repo_file",
345
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/generation_config.json",
346
+ "role": "source_artifact",
347
+ "sha256": "aed028489d41ae6a673920da9d103f2f58a650c3fbc50bd880b35e30a481dba6",
348
+ "size": 133
349
+ },
350
+ {
351
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/hf_model_info.json",
352
+ "kind": "local_repo_file",
353
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/hf_model_info.json",
354
+ "role": "source_artifact",
355
+ "sha256": "79debc90641b5144f54bd3e1c0199463b18c52a32408930964505883038e3f66",
356
+ "size": 1766
357
+ },
358
+ {
359
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/model.safetensors",
360
+ "kind": "local_repo_file",
361
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/model.safetensors",
362
+ "role": "source_artifact",
363
+ "sha256": "b5fdde934cb203742547527154c3af33560a143e606118dcea81076f0e4ab458",
364
+ "size": 794597224
365
+ },
366
+ {
367
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/special_tokens_map.json",
368
+ "kind": "local_repo_file",
369
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/special_tokens_map.json",
370
+ "role": "source_artifact",
371
+ "sha256": "a19f21a50f05b46cbbad085149d520ab1bc843a43976a08693df39f6ccc796d4",
372
+ "size": 2576
373
+ },
374
+ {
375
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/tokenizer.json",
376
+ "kind": "local_repo_file",
377
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/tokenizer.json",
378
+ "role": "source_artifact",
379
+ "sha256": "0ef5dc3e20dbfd2a3cafafbbbb61c3de27bc88372e0d96771d2b5fffe9cc0ed0",
380
+ "size": 36514
381
+ },
382
+ {
383
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/tokenizer_config.json",
384
+ "kind": "local_repo_file",
385
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/tokenizer_config.json",
386
+ "role": "source_artifact",
387
+ "sha256": "28ddd7e807d75d1b217963c1d15e8e5d57d3e5404999a25ef4e352de78daff5a",
388
+ "size": 28710
389
+ },
390
+ {
391
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/trainer_state.json",
392
+ "kind": "local_repo_file",
393
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/trainer_state.json",
394
+ "role": "source_artifact",
395
+ "sha256": "64a998131ae89d4405cef5d91614458885a49bbd6bc79733ae593cb8d10f5458",
396
+ "size": 3712
397
+ },
398
+ {
399
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/training_args.bin",
400
+ "kind": "local_repo_file",
401
+ "path": "datasets/forward_models/reactionT5v2/checkpoints/sagawa_reactiont5v2-forward-uspto_mit/training_args.bin",
402
+ "role": "source_artifact",
403
+ "sha256": "480806311c2fb18be472ad7ff49ac238235b6f54c2285fee10edef8206fd5bab",
404
+ "size": 5112
405
+ },
406
+ {
407
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/downloads/ORD.zip",
408
+ "kind": "local_repo_file",
409
+ "path": "datasets/forward_models/reactionT5v2/downloads/ORD.zip",
410
+ "role": "source_artifact",
411
+ "sha256": "05a999637c9e4a94673a6f8914e1b682cb7d90434fbc9678d296aa9285159c8b",
412
+ "size": 233136954
413
+ },
414
+ {
415
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/downloads/USPTO_MIT.zip",
416
+ "kind": "local_repo_file",
417
+ "path": "datasets/forward_models/reactionT5v2/downloads/USPTO_MIT.zip",
418
+ "role": "source_artifact",
419
+ "sha256": "3fb96bef12b0e87ef0f0ea243723a861b1fde161f6cd9dc45fe879eb98891c70",
420
+ "size": 14183273
421
+ },
422
+ {
423
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/README.txt",
424
+ "kind": "local_repo_file",
425
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/README.txt",
426
+ "role": "source_artifact",
427
+ "sha256": "d780354534a3a8d64e0cdb2ea1b69ca481ba3dcda6aa06c3f68ff46ae3b26b93",
428
+ "size": 307
429
+ },
430
+ {
431
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/ReactionT5v2_source_README.md",
432
+ "kind": "local_repo_file",
433
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/ReactionT5v2_source_README.md",
434
+ "role": "source_artifact",
435
+ "sha256": "7e03326292c50090f69e2ceefc5ab8ef78e07f13c6f76fa6a58975f11d4142e9",
436
+ "size": 12188
437
+ },
438
+ {
439
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv",
440
+ "kind": "local_repo_file",
441
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv",
442
+ "role": "source_artifact",
443
+ "sha256": "84786cdb3d3abc34c98f56c76be1e72d28e9f78a4f463611367a42aaea9424aa",
444
+ "size": 39188316
445
+ },
446
+ {
447
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
448
+ "kind": "local_repo_file",
449
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
450
+ "role": "source_artifact",
451
+ "sha256": "24e218c6b258f7624f1c9557678c29e7b6a18f2c9d725c2cc581d9cd57727e5b",
452
+ "size": 313922807
453
+ },
454
+ {
455
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
456
+ "kind": "local_repo_file",
457
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
458
+ "role": "source_artifact",
459
+ "sha256": "6db5bdb900d801874a1728640ef8004e16381e71f7dad5700c7f1e4d8be573e9",
460
+ "size": 39251481
461
+ },
462
+ {
463
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/datasets/USPTO_MIT/MIT_separated/test.csv",
464
+ "kind": "local_repo_file",
465
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/datasets/USPTO_MIT/MIT_separated/test.csv",
466
+ "role": "source_artifact",
467
+ "sha256": "b4d870d0550f6cf10a887dcee2bcfbaa46db14dbbdec734fa0f48b7ec661b5cb",
468
+ "size": 8612805
469
+ },
470
+ {
471
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/datasets/USPTO_MIT/MIT_separated/train.csv",
472
+ "kind": "local_repo_file",
473
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/datasets/USPTO_MIT/MIT_separated/train.csv",
474
+ "role": "source_artifact",
475
+ "sha256": "4c906e960cd96a0f39e1790314db642742a0fe8152e4481cc7c81aac83f4ed14",
476
+ "size": 48191398
477
+ },
478
+ {
479
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/datasets/USPTO_MIT/MIT_separated/val.csv",
480
+ "kind": "local_repo_file",
481
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/datasets/USPTO_MIT/MIT_separated/val.csv",
482
+ "role": "source_artifact",
483
+ "sha256": "1712f93b057f783fb2fbb4f74a89ab9e302e19230d938c4f74c5ad959df8fc1d",
484
+ "size": 3526860
485
+ },
486
+ {
487
+ "absolute_path": "/home/ian-l/lucai/syntheseus-bundle-builder/datasets/forward_models/reactionT5v2/shared_provenance/reactiont5v2_forward_split_provenance.json",
488
+ "kind": "local_repo_file",
489
+ "path": "datasets/forward_models/reactionT5v2/shared_provenance/reactiont5v2_forward_split_provenance.json",
490
+ "role": "source_artifact",
491
+ "sha256": "e4cf7b1b077d246605c1f2c7d4d2056d8d0d2dedbea0fc4e61748545a7a742cb",
492
+ "size": 2252
493
+ }
494
+ ]
495
+ }
checkpoints/reactiont5v2_forward/sagawa_reactiont5v2_forward_uspto_mit/wrapper_config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architecture_class": "T5ForConditionalGeneration",
3
+ "config_file": "aux/config.json",
4
+ "device_supported": [
5
+ "cpu",
6
+ "cuda"
7
+ ],
8
+ "fine_tuning": {
9
+ "adapter_overlay_dir": "adapters",
10
+ "base_weights_path": "model.safetensors",
11
+ "notes": [
12
+ "Base weights stay immutable in model.safetensors; future fine-tuning or LoRA adapters should be added as overlays under adapters/."
13
+ ]
14
+ },
15
+ "generation_config": "aux/generation_config.json",
16
+ "loader": {
17
+ "state_dict_format": "safetensors",
18
+ "strict": true
19
+ },
20
+ "package_layout": {
21
+ "adapters_dir": "adapters",
22
+ "aux_dir": "aux",
23
+ "weights_path": "model.safetensors"
24
+ },
25
+ "runtime_file_map": {
26
+ "aux/config.json": "config.json",
27
+ "aux/generation_config.json": "generation_config.json",
28
+ "aux/special_tokens_map.json": "special_tokens_map.json",
29
+ "aux/tokenizer.json": "tokenizer.json",
30
+ "aux/tokenizer_config.json": "tokenizer_config.json",
31
+ "model.safetensors": "model.safetensors"
32
+ },
33
+ "task_direction": "forward",
34
+ "tokenizer_files": [
35
+ "aux/tokenizer.json",
36
+ "aux/tokenizer_config.json",
37
+ "aux/special_tokens_map.json"
38
+ ],
39
+ "wrapper_family": "seq2seq_t5"
40
+ }
checkpoints/runtime_registry.json ADDED
@@ -0,0 +1,1734 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "checkpoints": [
3
+ {
4
+ "auxiliary_files": [
5
+ "aux/config.yml",
6
+ "aux/uspto_filter_model.onnx",
7
+ "aux/uspto_templates.csv.gz",
8
+ "benchmarks/default_test.jsonl",
9
+ "benchmarks/metadata.json",
10
+ "benchmarks/native/ref_routes_reaction_hashes.txt",
11
+ "benchmarks/native/selected_reactions.csv"
12
+ ],
13
+ "benchmark_datasets": [
14
+ {
15
+ "benchmark_variant": "paroutes_reference_route_holdout",
16
+ "dataset_id": "aizynthfinder_uspto_onnx_test",
17
+ "dataset_ref": "paroutes2_reference_route_reactions",
18
+ "default": true,
19
+ "format": "jsonl",
20
+ "metadata_path": "benchmarks/metadata.json",
21
+ "native_input_format": "tab_delimited_reactants_reagents_product_rows",
22
+ "native_source_paths": [
23
+ "benchmarks/native/selected_reactions.csv",
24
+ "benchmarks/native/ref_routes_reaction_hashes.txt"
25
+ ],
26
+ "notes": [
27
+ "Default bundled benchmark for the public AiZynthFinder USPTO ONNX policy.",
28
+ "The benchmark rows are reconstructed from PaRoutes selected_reactions.csv and ref_routes_reaction_hashes.txt so they match the held-out reference-route reactions used as the AiZynthTrain test set."
29
+ ],
30
+ "num_examples": 56657,
31
+ "path": "benchmarks/default_test.jsonl",
32
+ "split": "test"
33
+ }
34
+ ],
35
+ "checkpoint_id": "uspto_model_onnx",
36
+ "display_name": "uspto_model.onnx",
37
+ "embedding_modes": [],
38
+ "embedding_supported": false,
39
+ "fine_tuning_notes": null,
40
+ "fine_tuning_role": null,
41
+ "fine_tuning_supported": false,
42
+ "hf_model_name": null,
43
+ "host": "Zenodo",
44
+ "inference_supported": true,
45
+ "input_schema": null,
46
+ "larger_than_uspto50k": true,
47
+ "model_id": "aizynthfinder",
48
+ "model_name": "AiZynthFinder",
49
+ "normalized_checkpoint_id": "uspto_model_onnx",
50
+ "package_dir": "checkpoints/aizynthfinder/uspto_model_onnx",
51
+ "remote_url": "https://zenodo.org/records/11430881",
52
+ "source_registry_refs": [
53
+ {
54
+ "checkpoint_id": "uspto_model_onnx",
55
+ "model_id": "aizynthfinder"
56
+ }
57
+ ],
58
+ "task_direction": null,
59
+ "training_dataset_refs": [
60
+ "aizynthtrain_uspto_selected_reactions_with_paroutes_holdout"
61
+ ],
62
+ "unsupported_reason": null,
63
+ "weights_path": "model.onnx",
64
+ "wrapper_class": "syntheseus_wrappers.wrappers.aizynthfinder.AiZynthFinderModel",
65
+ "wrapper_family": "aizynthfinder_template_policy"
66
+ },
67
+ {
68
+ "auxiliary_files": [
69
+ "aux/bart_vocab.json",
70
+ "aux/bart_vocab_downstream.json",
71
+ "aux/ds_config.json",
72
+ "aux/hparams.yaml",
73
+ "aux/hyper_parameters.json"
74
+ ],
75
+ "checkpoint_id": "combined_large_pre_trained_checkpoint",
76
+ "display_name": "combined-large pre-trained checkpoint",
77
+ "embedding_modes": [],
78
+ "embedding_supported": false,
79
+ "fine_tuning_notes": null,
80
+ "fine_tuning_role": "base_pretraining",
81
+ "fine_tuning_supported": true,
82
+ "hf_model_name": null,
83
+ "host": "Box shared folder",
84
+ "inference_supported": false,
85
+ "input_schema": null,
86
+ "larger_than_uspto50k": true,
87
+ "model_id": "chemformer",
88
+ "model_name": "Chemformer",
89
+ "normalized_checkpoint_id": "combined_large_pre_trained_checkpoint",
90
+ "package_dir": "checkpoints/chemformer/combined_large_pre_trained_checkpoint",
91
+ "remote_url": "https://az.app.box.com/s/7eci3nd9vy0xplqniitpk02rbg9q2zcq/folder/144881806154",
92
+ "source_registry_refs": [
93
+ {
94
+ "checkpoint_id": "combined_large_pre_trained_checkpoint",
95
+ "model_id": "chemformer"
96
+ }
97
+ ],
98
+ "task_direction": null,
99
+ "training_dataset_refs": [
100
+ "zinc15_100m_chemformer_pretraining"
101
+ ],
102
+ "unsupported_reason": "pretraining-only Chemformer large bundle; not a retrosynthesis fine-tuned backward reaction checkpoint",
103
+ "weights_path": "model.safetensors",
104
+ "wrapper_class": null,
105
+ "wrapper_family": "chemformer_bart"
106
+ },
107
+ {
108
+ "auxiliary_files": [
109
+ "aux/bart_vocab.json",
110
+ "aux/bart_vocab_downstream.json",
111
+ "aux/ds_config.json",
112
+ "aux/hparams.yaml",
113
+ "aux/hyper_parameters.json"
114
+ ],
115
+ "checkpoint_id": "combined_pre_trained_base_checkpoint",
116
+ "display_name": "combined pre-trained base checkpoint",
117
+ "embedding_modes": [],
118
+ "embedding_supported": false,
119
+ "fine_tuning_notes": null,
120
+ "fine_tuning_role": "base_pretraining",
121
+ "fine_tuning_supported": true,
122
+ "hf_model_name": null,
123
+ "host": "Box shared folder",
124
+ "inference_supported": false,
125
+ "input_schema": null,
126
+ "larger_than_uspto50k": true,
127
+ "model_id": "chemformer",
128
+ "model_name": "Chemformer",
129
+ "normalized_checkpoint_id": "combined_pre_trained_base_checkpoint",
130
+ "package_dir": "checkpoints/chemformer/combined_pre_trained_base_checkpoint",
131
+ "remote_url": "https://az.app.box.com/s/7eci3nd9vy0xplqniitpk02rbg9q2zcq/folder/144881804954",
132
+ "source_registry_refs": [
133
+ {
134
+ "checkpoint_id": "combined_pre_trained_base_checkpoint",
135
+ "model_id": "chemformer"
136
+ }
137
+ ],
138
+ "task_direction": null,
139
+ "training_dataset_refs": [
140
+ "zinc15_100m_chemformer_pretraining"
141
+ ],
142
+ "unsupported_reason": "pretraining-only Chemformer base bundle; not a retrosynthesis fine-tuned backward reaction checkpoint",
143
+ "weights_path": "model.safetensors",
144
+ "wrapper_class": null,
145
+ "wrapper_family": "chemformer_bart"
146
+ },
147
+ {
148
+ "auxiliary_files": [
149
+ "aux/bart_vocab.json",
150
+ "aux/bart_vocab_downstream.json",
151
+ "aux/ds_config.json",
152
+ "aux/hparams.yaml",
153
+ "aux/hyper_parameters.json",
154
+ "benchmarks/default_test.jsonl",
155
+ "benchmarks/metadata.json",
156
+ "benchmarks/native/uspto_50.pickle"
157
+ ],
158
+ "benchmark_datasets": [
159
+ {
160
+ "dataset_id": "uspto_50_fine_tuned_retrosynthesis_checkpoint_test",
161
+ "dataset_ref": "uspto_50k_chemformer_project_release",
162
+ "default": true,
163
+ "format": "jsonl",
164
+ "metadata_path": "benchmarks/metadata.json",
165
+ "native_input_format": "pandas_pickled_dataframe",
166
+ "native_source_paths": [
167
+ "benchmarks/native/uspto_50.pickle"
168
+ ],
169
+ "notes": [
170
+ "Native Chemformer benchmark source extracted from the public Box release pickle.",
171
+ "benchmarks/default_test.jsonl stores product/reactant SMILES converted from the pickled RDKit molecule objects in the test partition."
172
+ ],
173
+ "num_examples": 5004,
174
+ "path": "benchmarks/default_test.jsonl",
175
+ "split": "test"
176
+ }
177
+ ],
178
+ "checkpoint_id": "uspto_50_fine_tuned_retrosynthesis_checkpoint",
179
+ "display_name": "uspto_50 fine-tuned retrosynthesis checkpoint",
180
+ "embedding_modes": [],
181
+ "embedding_supported": false,
182
+ "fine_tuning_notes": null,
183
+ "fine_tuning_role": "task_finetuned_seed",
184
+ "fine_tuning_supported": true,
185
+ "hf_model_name": null,
186
+ "host": "Box shared folder",
187
+ "inference_supported": true,
188
+ "input_schema": null,
189
+ "larger_than_uspto50k": false,
190
+ "model_id": "chemformer",
191
+ "model_name": "Chemformer",
192
+ "normalized_checkpoint_id": "uspto_50_fine_tuned_retrosynthesis_checkpoint",
193
+ "package_dir": "checkpoints/chemformer/uspto_50_fine_tuned_retrosynthesis_checkpoint",
194
+ "remote_url": "https://az.app.box.com/s/7eci3nd9vy0xplqniitpk02rbg9q2zcq/folder/145317555751",
195
+ "source_registry_refs": [
196
+ {
197
+ "checkpoint_id": "uspto_50_fine_tuned_retrosynthesis_checkpoint",
198
+ "model_id": "chemformer"
199
+ }
200
+ ],
201
+ "task_direction": null,
202
+ "training_dataset_refs": [
203
+ "uspto_50k_chemformer_project_release"
204
+ ],
205
+ "unsupported_reason": null,
206
+ "weights_path": "model.safetensors",
207
+ "wrapper_class": "syntheseus_wrappers.wrappers.chemformer.ChemformerModel",
208
+ "wrapper_family": "chemformer_bart"
209
+ },
210
+ {
211
+ "auxiliary_files": [
212
+ "aux/README.md",
213
+ "aux/checkpoint_metadata.json",
214
+ "aux/configs/tokenizer.json",
215
+ "aux/configs/uspto_50k/base.yaml",
216
+ "aux/configs/uspto_50k/stage2_ema.yaml",
217
+ "aux/configs/uspto_50k/stage2_frozen.yaml",
218
+ "aux/provenance/uspto-50k/raw_test.jsonl",
219
+ "aux/retrieval.faiss",
220
+ "aux/template_cache.pt",
221
+ "aux/template_cache_metadata.json",
222
+ "aux/template_library.jsonl",
223
+ "benchmarks/default_test.jsonl",
224
+ "benchmarks/metadata.json",
225
+ "benchmarks/native/raw_test.jsonl"
226
+ ],
227
+ "benchmark_datasets": [
228
+ {
229
+ "dataset_id": "uspto_50k_ema_deep_epoch059_step0_val_top1_acc0_3521_ckpt_test",
230
+ "dataset_ref": "conretrobert_uspto_50k_packaged_release",
231
+ "default": true,
232
+ "format": "jsonl",
233
+ "metadata_path": "benchmarks/metadata.json",
234
+ "native_input_format": "jsonl_reaction_rows",
235
+ "native_source_paths": [
236
+ "benchmarks/native/raw_test.jsonl"
237
+ ],
238
+ "notes": [
239
+ "Native benchmark split for the packaged ConRetroBert USPTO-50K release.",
240
+ "benchmarks/default_test.jsonl preserves the native atom-mapped product strings for model inputs, while expected reactants are stored as atom-map-stripped canonical SMILES and the original mapped strings remain in per-record metadata."
241
+ ],
242
+ "num_examples": 5005,
243
+ "path": "benchmarks/default_test.jsonl",
244
+ "split": "test"
245
+ }
246
+ ],
247
+ "checkpoint_id": "uspto_50k_ema_deep_epoch059_step0_val_top1_acc0_3521_ckpt",
248
+ "display_name": "USPTO_50K_EMA_DEEP/epoch059-step0-val_top1_acc0.3521.ckpt",
249
+ "embedding_modes": [],
250
+ "embedding_supported": false,
251
+ "fine_tuning_notes": null,
252
+ "fine_tuning_role": null,
253
+ "fine_tuning_supported": false,
254
+ "hf_model_name": "USPTO_50K_EMA_DEEP/epoch059-step0-val_top1_acc0.3521.ckpt",
255
+ "host": "Google Drive folder",
256
+ "inference_supported": true,
257
+ "input_schema": null,
258
+ "larger_than_uspto50k": false,
259
+ "model_id": "conretrobert",
260
+ "model_name": "ConRetroBert",
261
+ "normalized_checkpoint_id": "uspto_50k_ema_deep_epoch059_step0_val_top1_acc0_3521_ckpt",
262
+ "package_dir": "checkpoints/conretrobert/uspto_50k_ema_deep_epoch059_step0_val_top1_acc0_3521_ckpt",
263
+ "remote_url": "https://drive.google.com/drive/folders/1iSKIUiZZ8uA7KdYR62S6YjhV2eVBKKNj?usp=drive_link",
264
+ "source_registry_refs": [
265
+ {
266
+ "checkpoint_id": "uspto_50k_ema_deep_epoch059_step0_val_top1_acc0_3521_ckpt",
267
+ "model_id": "conretrobert"
268
+ }
269
+ ],
270
+ "task_direction": null,
271
+ "training_dataset_refs": [
272
+ "conretrobert_uspto_50k_packaged_release"
273
+ ],
274
+ "unsupported_reason": null,
275
+ "weights_path": "model.safetensors",
276
+ "wrapper_class": "syntheseus_wrappers.wrappers.conretrobert.ConRetroBertModel",
277
+ "wrapper_family": "conretrobert_template_ranker"
278
+ },
279
+ {
280
+ "auxiliary_files": [
281
+ "aux/README.md",
282
+ "aux/checkpoint_metadata.json",
283
+ "aux/configs/tokenizer.json",
284
+ "aux/configs/uspto_50k/base.yaml",
285
+ "aux/configs/uspto_50k/stage2_ema.yaml",
286
+ "aux/configs/uspto_50k/stage2_frozen.yaml",
287
+ "aux/provenance/uspto-50k/raw_test.jsonl",
288
+ "aux/retrieval.faiss",
289
+ "aux/template_cache.pt",
290
+ "aux/template_cache_metadata.json",
291
+ "aux/template_library.jsonl",
292
+ "benchmarks/default_test.jsonl",
293
+ "benchmarks/metadata.json",
294
+ "benchmarks/native/raw_test.jsonl"
295
+ ],
296
+ "benchmark_datasets": [
297
+ {
298
+ "dataset_id": "uspto_50k_ema_mid_epoch087_step0_val_top1_acc0_3547_val_loss_4_5051_ckpt_test",
299
+ "dataset_ref": "conretrobert_uspto_50k_packaged_release",
300
+ "default": true,
301
+ "format": "jsonl",
302
+ "metadata_path": "benchmarks/metadata.json",
303
+ "native_input_format": "jsonl_reaction_rows",
304
+ "native_source_paths": [
305
+ "benchmarks/native/raw_test.jsonl"
306
+ ],
307
+ "notes": [
308
+ "Native benchmark split for the packaged ConRetroBert USPTO-50K release.",
309
+ "benchmarks/default_test.jsonl preserves the native atom-mapped product strings for model inputs, while expected reactants are stored as atom-map-stripped canonical SMILES and the original mapped strings remain in per-record metadata."
310
+ ],
311
+ "num_examples": 5005,
312
+ "path": "benchmarks/default_test.jsonl",
313
+ "split": "test"
314
+ }
315
+ ],
316
+ "checkpoint_id": "uspto_50k_ema_mid_epoch087_step0_val_top1_acc0_3547_val_loss_4_5051_ckpt",
317
+ "display_name": "USPTO_50K_EMA_MID/epoch087-step0-val_top1_acc0.3547_val_loss_4.5051.ckpt",
318
+ "embedding_modes": [],
319
+ "embedding_supported": false,
320
+ "fine_tuning_notes": null,
321
+ "fine_tuning_role": null,
322
+ "fine_tuning_supported": false,
323
+ "hf_model_name": "USPTO_50K_EMA_MID/epoch087-step0-val_top1_acc0.3547_val_loss_4.5051.ckpt",
324
+ "host": "Google Drive folder",
325
+ "inference_supported": true,
326
+ "input_schema": null,
327
+ "larger_than_uspto50k": false,
328
+ "model_id": "conretrobert",
329
+ "model_name": "ConRetroBert",
330
+ "normalized_checkpoint_id": "uspto_50k_ema_mid_epoch087_step0_val_top1_acc0_3547_val_loss_4_5051_ckpt",
331
+ "package_dir": "checkpoints/conretrobert/uspto_50k_ema_mid_epoch087_step0_val_top1_acc0_3547_val_loss_4_5051_ckpt",
332
+ "remote_url": "https://drive.google.com/drive/folders/1iSKIUiZZ8uA7KdYR62S6YjhV2eVBKKNj?usp=drive_link",
333
+ "source_registry_refs": [
334
+ {
335
+ "checkpoint_id": "uspto_50k_ema_mid_epoch087_step0_val_top1_acc0_3547_val_loss_4_5051_ckpt",
336
+ "model_id": "conretrobert"
337
+ }
338
+ ],
339
+ "task_direction": null,
340
+ "training_dataset_refs": [
341
+ "conretrobert_uspto_50k_packaged_release"
342
+ ],
343
+ "unsupported_reason": null,
344
+ "weights_path": "model.safetensors",
345
+ "wrapper_class": "syntheseus_wrappers.wrappers.conretrobert.ConRetroBertModel",
346
+ "wrapper_family": "conretrobert_template_ranker"
347
+ },
348
+ {
349
+ "auxiliary_files": [
350
+ "aux/README.md",
351
+ "aux/checkpoint_metadata.json",
352
+ "aux/configs/tokenizer.json",
353
+ "aux/configs/uspto_50k/base.yaml",
354
+ "aux/configs/uspto_50k/stage2_ema.yaml",
355
+ "aux/configs/uspto_50k/stage2_frozen.yaml",
356
+ "aux/provenance/uspto-50k/raw_test.jsonl",
357
+ "aux/retrieval.faiss",
358
+ "aux/template_cache.pt",
359
+ "aux/template_cache_metadata.json",
360
+ "aux/template_library.jsonl",
361
+ "benchmarks/default_test.jsonl",
362
+ "benchmarks/metadata.json",
363
+ "benchmarks/native/raw_test.jsonl"
364
+ ],
365
+ "benchmark_datasets": [
366
+ {
367
+ "dataset_id": "uspto_50k_frozen_te_epoch087_step54714_val_top1_acc0_3711_ckpt_test",
368
+ "dataset_ref": "conretrobert_uspto_50k_packaged_release",
369
+ "default": true,
370
+ "format": "jsonl",
371
+ "metadata_path": "benchmarks/metadata.json",
372
+ "native_input_format": "jsonl_reaction_rows",
373
+ "native_source_paths": [
374
+ "benchmarks/native/raw_test.jsonl"
375
+ ],
376
+ "notes": [
377
+ "Native benchmark split for the packaged ConRetroBert USPTO-50K release.",
378
+ "benchmarks/default_test.jsonl preserves the native atom-mapped product strings for model inputs, while expected reactants are stored as atom-map-stripped canonical SMILES and the original mapped strings remain in per-record metadata."
379
+ ],
380
+ "num_examples": 5005,
381
+ "path": "benchmarks/default_test.jsonl",
382
+ "split": "test"
383
+ }
384
+ ],
385
+ "checkpoint_id": "uspto_50k_frozen_te_epoch087_step54714_val_top1_acc0_3711_ckpt",
386
+ "display_name": "USPTO_50K_FROZEN_TE/epoch087-step54714-val_top1_acc0.3711.ckpt",
387
+ "embedding_modes": [],
388
+ "embedding_supported": false,
389
+ "fine_tuning_notes": null,
390
+ "fine_tuning_role": null,
391
+ "fine_tuning_supported": false,
392
+ "hf_model_name": "USPTO_50K_FROZEN_TE/epoch087-step54714-val_top1_acc0.3711.ckpt",
393
+ "host": "Google Drive folder",
394
+ "inference_supported": true,
395
+ "input_schema": null,
396
+ "larger_than_uspto50k": false,
397
+ "model_id": "conretrobert",
398
+ "model_name": "ConRetroBert",
399
+ "normalized_checkpoint_id": "uspto_50k_frozen_te_epoch087_step54714_val_top1_acc0_3711_ckpt",
400
+ "package_dir": "checkpoints/conretrobert/uspto_50k_frozen_te_epoch087_step54714_val_top1_acc0_3711_ckpt",
401
+ "remote_url": "https://drive.google.com/drive/folders/1iSKIUiZZ8uA7KdYR62S6YjhV2eVBKKNj?usp=drive_link",
402
+ "source_registry_refs": [
403
+ {
404
+ "checkpoint_id": "uspto_50k_frozen_te_epoch087_step54714_val_top1_acc0_3711_ckpt",
405
+ "model_id": "conretrobert"
406
+ }
407
+ ],
408
+ "task_direction": null,
409
+ "training_dataset_refs": [
410
+ "conretrobert_uspto_50k_packaged_release"
411
+ ],
412
+ "unsupported_reason": null,
413
+ "weights_path": "model.safetensors",
414
+ "wrapper_class": "syntheseus_wrappers.wrappers.conretrobert.ConRetroBertModel",
415
+ "wrapper_family": "conretrobert_template_ranker"
416
+ },
417
+ {
418
+ "auxiliary_files": [
419
+ "aux/README.md",
420
+ "aux/checkpoint_metadata.json",
421
+ "aux/configs/tokenizer.json",
422
+ "aux/configs/uspto_50k/base.yaml",
423
+ "aux/configs/uspto_50k/stage2_ema.yaml",
424
+ "aux/configs/uspto_50k/stage2_frozen.yaml",
425
+ "aux/provenance/uspto-50k/raw_test.jsonl",
426
+ "aux/retrieval.faiss",
427
+ "aux/template_cache.pt",
428
+ "aux/template_cache_metadata.json",
429
+ "aux/template_library.jsonl",
430
+ "benchmarks/default_test.jsonl",
431
+ "benchmarks/metadata.json",
432
+ "benchmarks/native/raw_test.jsonl"
433
+ ],
434
+ "benchmark_datasets": [
435
+ {
436
+ "benchmark_semantics": "uspto_50k_transfer_compatibility_for_uspto_full_checkpoint",
437
+ "benchmark_variant": "transfer_compatibility",
438
+ "dataset_id": "uspto_full_2_uspto_50k_magic_epoch012_step0_weighted_val_accuracy0_8390_ckpt_uspto_50k_transfer_test",
439
+ "dataset_ref": "conretrobert_uspto_50k_packaged_release",
440
+ "default": true,
441
+ "format": "jsonl",
442
+ "metadata_path": "benchmarks/metadata.json",
443
+ "native_input_format": "jsonl_reaction_rows",
444
+ "native_source_paths": [
445
+ "benchmarks/native/raw_test.jsonl"
446
+ ],
447
+ "notes": [
448
+ "This checkpoint name indicates USPTO-Full training, but the public ConRetroBert release staged locally does not package a native USPTO-Full benchmark split or full template library. The default bundle benchmark therefore uses the packaged USPTO-50K raw_test.jsonl split as a transfer/compatibility evaluation only.",
449
+ "Native benchmark split for the packaged ConRetroBert USPTO-50K release.",
450
+ "benchmarks/default_test.jsonl preserves the native atom-mapped product strings for model inputs, while expected reactants are stored as atom-map-stripped canonical SMILES and the original mapped strings remain in per-record metadata."
451
+ ],
452
+ "num_examples": 5005,
453
+ "path": "benchmarks/default_test.jsonl",
454
+ "split": "test"
455
+ }
456
+ ],
457
+ "checkpoint_id": "uspto_full_2_uspto_50k_magic_epoch012_step0_weighted_val_accuracy0_8390_ckpt",
458
+ "display_name": "USPTO_FULL_2_USPTO_50K_MAGIC/epoch012-step0-weighted_val_accuracy0.8390.ckpt",
459
+ "embedding_modes": [],
460
+ "embedding_supported": false,
461
+ "fine_tuning_notes": null,
462
+ "fine_tuning_role": null,
463
+ "fine_tuning_supported": false,
464
+ "hf_model_name": "USPTO_FULL_2_USPTO_50K_MAGIC/epoch012-step0-weighted_val_accuracy0.8390.ckpt",
465
+ "host": "Google Drive folder",
466
+ "inference_supported": true,
467
+ "input_schema": null,
468
+ "larger_than_uspto50k": false,
469
+ "model_id": "conretrobert",
470
+ "model_name": "ConRetroBert",
471
+ "normalized_checkpoint_id": "uspto_full_2_uspto_50k_magic_epoch012_step0_weighted_val_accuracy0_8390_ckpt",
472
+ "package_dir": "checkpoints/conretrobert/uspto_full_2_uspto_50k_magic_epoch012_step0_weighted_val_accuracy0_8390_ckpt",
473
+ "remote_url": "https://drive.google.com/drive/folders/1iSKIUiZZ8uA7KdYR62S6YjhV2eVBKKNj?usp=drive_link",
474
+ "source_registry_refs": [
475
+ {
476
+ "checkpoint_id": "uspto_full_2_uspto_50k_magic_epoch012_step0_weighted_val_accuracy0_8390_ckpt",
477
+ "model_id": "conretrobert"
478
+ }
479
+ ],
480
+ "task_direction": null,
481
+ "training_dataset_refs": [
482
+ "conretrobert_uspto_50k_packaged_release"
483
+ ],
484
+ "unsupported_reason": null,
485
+ "weights_path": "model.safetensors",
486
+ "wrapper_class": "syntheseus_wrappers.wrappers.conretrobert.ConRetroBertModel",
487
+ "wrapper_family": "conretrobert_template_ranker"
488
+ },
489
+ {
490
+ "auxiliary_files": [
491
+ "aux/README.md",
492
+ "aux/checkpoint_metadata.json",
493
+ "aux/configs/tokenizer.json",
494
+ "aux/configs/uspto_50k/base.yaml",
495
+ "aux/configs/uspto_50k/stage2_ema.yaml",
496
+ "aux/configs/uspto_50k/stage2_frozen.yaml",
497
+ "aux/provenance/uspto-50k/raw_test.jsonl",
498
+ "aux/retrieval.faiss",
499
+ "aux/template_cache.pt",
500
+ "aux/template_cache_metadata.json",
501
+ "aux/template_library.jsonl",
502
+ "benchmarks/default_test.jsonl",
503
+ "benchmarks/metadata.json",
504
+ "benchmarks/native/raw_test.jsonl"
505
+ ],
506
+ "benchmark_datasets": [
507
+ {
508
+ "benchmark_semantics": "uspto_50k_transfer_compatibility_for_uspto_full_checkpoint",
509
+ "benchmark_variant": "transfer_compatibility",
510
+ "dataset_id": "uspto_full_frozen_te_epoch096_step129890_val_top1_acc0_6477_ckpt_uspto_50k_transfer_test",
511
+ "dataset_ref": "conretrobert_uspto_50k_packaged_release",
512
+ "default": true,
513
+ "format": "jsonl",
514
+ "metadata_path": "benchmarks/metadata.json",
515
+ "native_input_format": "jsonl_reaction_rows",
516
+ "native_source_paths": [
517
+ "benchmarks/native/raw_test.jsonl"
518
+ ],
519
+ "notes": [
520
+ "This checkpoint name indicates USPTO-Full training, but the public ConRetroBert release staged locally does not package a native USPTO-Full benchmark split or full template library. The default bundle benchmark therefore uses the packaged USPTO-50K raw_test.jsonl split as a transfer/compatibility evaluation only.",
521
+ "Native benchmark split for the packaged ConRetroBert USPTO-50K release.",
522
+ "benchmarks/default_test.jsonl preserves the native atom-mapped product strings for model inputs, while expected reactants are stored as atom-map-stripped canonical SMILES and the original mapped strings remain in per-record metadata."
523
+ ],
524
+ "num_examples": 5005,
525
+ "path": "benchmarks/default_test.jsonl",
526
+ "split": "test"
527
+ }
528
+ ],
529
+ "checkpoint_id": "uspto_full_frozen_te_epoch096_step129890_val_top1_acc0_6477_ckpt",
530
+ "display_name": "USPTO_FULL_FROZEN_TE/epoch096-step129890-val_top1_acc0.6477.ckpt",
531
+ "embedding_modes": [],
532
+ "embedding_supported": false,
533
+ "fine_tuning_notes": null,
534
+ "fine_tuning_role": null,
535
+ "fine_tuning_supported": false,
536
+ "hf_model_name": "USPTO_FULL_FROZEN_TE/epoch096-step129890-val_top1_acc0.6477.ckpt",
537
+ "host": "Google Drive folder",
538
+ "inference_supported": true,
539
+ "input_schema": null,
540
+ "larger_than_uspto50k": true,
541
+ "model_id": "conretrobert",
542
+ "model_name": "ConRetroBert",
543
+ "normalized_checkpoint_id": "uspto_full_frozen_te_epoch096_step129890_val_top1_acc0_6477_ckpt",
544
+ "package_dir": "checkpoints/conretrobert/uspto_full_frozen_te_epoch096_step129890_val_top1_acc0_6477_ckpt",
545
+ "remote_url": "https://drive.google.com/drive/folders/1iSKIUiZZ8uA7KdYR62S6YjhV2eVBKKNj?usp=drive_link",
546
+ "source_registry_refs": [
547
+ {
548
+ "checkpoint_id": "uspto_full_frozen_te_epoch096_step129890_val_top1_acc0_6477_ckpt",
549
+ "model_id": "conretrobert"
550
+ }
551
+ ],
552
+ "task_direction": null,
553
+ "training_dataset_refs": [],
554
+ "unsupported_reason": null,
555
+ "weights_path": "model.safetensors",
556
+ "wrapper_class": "syntheseus_wrappers.wrappers.conretrobert.ConRetroBertModel",
557
+ "wrapper_family": "conretrobert_template_ranker"
558
+ },
559
+ {
560
+ "auxiliary_files": [
561
+ "aux/README.md",
562
+ "aux/checkpoint_metadata.json",
563
+ "aux/configs/tokenizer.json",
564
+ "aux/configs/uspto_50k/base.yaml",
565
+ "aux/configs/uspto_50k/stage2_ema.yaml",
566
+ "aux/configs/uspto_50k/stage2_frozen.yaml",
567
+ "aux/provenance/uspto-50k/raw_test.jsonl",
568
+ "aux/retrieval.faiss",
569
+ "aux/template_cache.pt",
570
+ "aux/template_cache_metadata.json",
571
+ "aux/template_library.jsonl",
572
+ "benchmarks/default_test.jsonl",
573
+ "benchmarks/metadata.json",
574
+ "benchmarks/native/raw_test.jsonl"
575
+ ],
576
+ "benchmark_datasets": [
577
+ {
578
+ "benchmark_semantics": "uspto_50k_transfer_compatibility_for_uspto_full_checkpoint",
579
+ "benchmark_variant": "transfer_compatibility",
580
+ "dataset_id": "uspto_full_transfer_uspto_50k_ft_epoch046_step29212_val_top1_acc0_7501_val_loss_1_3350_ckpt_uspto_50k_transfer_test",
581
+ "dataset_ref": "conretrobert_uspto_50k_packaged_release",
582
+ "default": true,
583
+ "format": "jsonl",
584
+ "metadata_path": "benchmarks/metadata.json",
585
+ "native_input_format": "jsonl_reaction_rows",
586
+ "native_source_paths": [
587
+ "benchmarks/native/raw_test.jsonl"
588
+ ],
589
+ "notes": [
590
+ "This checkpoint name indicates USPTO-Full training, but the public ConRetroBert release staged locally does not package a native USPTO-Full benchmark split or full template library. The default bundle benchmark therefore uses the packaged USPTO-50K raw_test.jsonl split as a transfer/compatibility evaluation only.",
591
+ "Native benchmark split for the packaged ConRetroBert USPTO-50K release.",
592
+ "benchmarks/default_test.jsonl preserves the native atom-mapped product strings for model inputs, while expected reactants are stored as atom-map-stripped canonical SMILES and the original mapped strings remain in per-record metadata."
593
+ ],
594
+ "num_examples": 5005,
595
+ "path": "benchmarks/default_test.jsonl",
596
+ "split": "test"
597
+ }
598
+ ],
599
+ "checkpoint_id": "uspto_full_transfer_uspto_50k_ft_epoch046_step29212_val_top1_acc0_7501_val_loss_1_3350_ckpt",
600
+ "display_name": "USPTO_FULL_TRANSFER_USPTO_50K_FT/epoch046-step29212-val_top1_acc0.7501_val_loss_1.3350.ckpt",
601
+ "embedding_modes": [],
602
+ "embedding_supported": false,
603
+ "fine_tuning_notes": null,
604
+ "fine_tuning_role": null,
605
+ "fine_tuning_supported": false,
606
+ "hf_model_name": "USPTO_FULL_TRANSFER_USPTO_50K_FT/epoch046-step29212-val_top1_acc0.7501_val_loss_1.3350.ckpt",
607
+ "host": "Google Drive folder",
608
+ "inference_supported": true,
609
+ "input_schema": null,
610
+ "larger_than_uspto50k": false,
611
+ "model_id": "conretrobert",
612
+ "model_name": "ConRetroBert",
613
+ "normalized_checkpoint_id": "uspto_full_transfer_uspto_50k_ft_epoch046_step29212_val_top1_acc0_7501_val_loss_1_3350_ckpt",
614
+ "package_dir": "checkpoints/conretrobert/uspto_full_transfer_uspto_50k_ft_epoch046_step29212_val_top1_acc0_7501_val_loss_1_3350_ckpt",
615
+ "remote_url": "https://drive.google.com/drive/folders/1iSKIUiZZ8uA7KdYR62S6YjhV2eVBKKNj?usp=drive_link",
616
+ "source_registry_refs": [
617
+ {
618
+ "checkpoint_id": "uspto_full_transfer_uspto_50k_ft_epoch046_step29212_val_top1_acc0_7501_val_loss_1_3350_ckpt",
619
+ "model_id": "conretrobert"
620
+ }
621
+ ],
622
+ "task_direction": null,
623
+ "training_dataset_refs": [
624
+ "conretrobert_uspto_50k_packaged_release"
625
+ ],
626
+ "unsupported_reason": null,
627
+ "weights_path": "model.safetensors",
628
+ "wrapper_class": "syntheseus_wrappers.wrappers.conretrobert.ConRetroBertModel",
629
+ "wrapper_family": "conretrobert_template_ranker"
630
+ },
631
+ {
632
+ "auxiliary_files": [
633
+ "aux/.gitattributes",
634
+ "aux/README.md",
635
+ "aux/UPSTREAM_URL.txt",
636
+ "aux/config.json",
637
+ "aux/generation_config.json",
638
+ "aux/hf_model_info.json",
639
+ "aux/provenance/README.txt",
640
+ "aux/provenance/ReactionT5v2_source_README.md",
641
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv",
642
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
643
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
644
+ "aux/provenance/datasets/USPTO_50k/test.csv",
645
+ "aux/provenance/datasets/USPTO_50k/train.csv",
646
+ "aux/provenance/datasets/USPTO_50k/val.csv",
647
+ "aux/provenance/reactiont5v2_split_provenance.json",
648
+ "aux/special_tokens_map.json",
649
+ "aux/tokenizer.json",
650
+ "aux/tokenizer_config.json",
651
+ "aux/trainer_state.json",
652
+ "aux/training_args.bin",
653
+ "benchmarks/default_test.jsonl",
654
+ "benchmarks/metadata.json"
655
+ ],
656
+ "benchmark_datasets": [
657
+ {
658
+ "dataset_id": "sagawa_reactiont5v2_retrosynthesis_test",
659
+ "dataset_ref": "ord_retrosynthesis_pretraining",
660
+ "default": true,
661
+ "format": "jsonl",
662
+ "metadata_path": "benchmarks/metadata.json",
663
+ "native_input_format": "csv_reaction_rows",
664
+ "native_source_paths": [
665
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv"
666
+ ],
667
+ "notes": [
668
+ "Native benchmark split for the public ORD-trained retrosynthesis checkpoint.",
669
+ "USPTO_50k/test.csv was used upstream for duplicate-removal / leakage control, not as the native training benchmark for this checkpoint."
670
+ ],
671
+ "num_examples": 225282,
672
+ "path": "benchmarks/default_test.jsonl",
673
+ "split": "test"
674
+ }
675
+ ],
676
+ "checkpoint_id": "sagawa_reactiont5v2_retrosynthesis",
677
+ "display_name": "sagawa/ReactionT5v2-retrosynthesis",
678
+ "embedding_modes": [
679
+ "product_encoder",
680
+ "reactant_decoder",
681
+ "reaction_pair",
682
+ "token"
683
+ ],
684
+ "embedding_supported": true,
685
+ "fine_tuning_notes": "ORD-trained retrosynthesis checkpoint that can serve as a continued-training seed.",
686
+ "fine_tuning_role": "task_finetuned_seed",
687
+ "fine_tuning_supported": true,
688
+ "hf_model_name": "sagawa/ReactionT5v2-retrosynthesis",
689
+ "host": "Hugging Face",
690
+ "inference_supported": true,
691
+ "input_schema": null,
692
+ "larger_than_uspto50k": true,
693
+ "model_id": "reactiont5v2",
694
+ "model_name": "ReactionT5v2",
695
+ "normalized_checkpoint_id": "sagawa_reactiont5v2_retrosynthesis",
696
+ "package_dir": "checkpoints/reactiont5v2/sagawa_reactiont5v2_retrosynthesis",
697
+ "remote_url": "https://huggingface.co/sagawa/ReactionT5v2-retrosynthesis",
698
+ "source_registry_refs": [
699
+ {
700
+ "checkpoint_id": "sagawa_reactiont5v2_retrosynthesis",
701
+ "model_id": "reactiont5v2"
702
+ }
703
+ ],
704
+ "task_direction": null,
705
+ "training_dataset_refs": [
706
+ "ord_retrosynthesis_pretraining"
707
+ ],
708
+ "unsupported_reason": null,
709
+ "weights_path": "model.safetensors",
710
+ "wrapper_class": "syntheseus_wrappers.wrappers.reactiont5v2.ReactionT5v2Model",
711
+ "wrapper_family": "seq2seq_t5"
712
+ },
713
+ {
714
+ "auxiliary_files": [
715
+ "aux/.gitattributes",
716
+ "aux/README.md",
717
+ "aux/UPSTREAM_URL.txt",
718
+ "aux/config.json",
719
+ "aux/generation_config.json",
720
+ "aux/hf_model_info.json",
721
+ "aux/provenance/README.txt",
722
+ "aux/provenance/ReactionT5v2_source_README.md",
723
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_test.csv",
724
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_train.csv",
725
+ "aux/provenance/datasets/ORD/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv",
726
+ "aux/provenance/datasets/USPTO_50k/test.csv",
727
+ "aux/provenance/datasets/USPTO_50k/train.csv",
728
+ "aux/provenance/datasets/USPTO_50k/val.csv",
729
+ "aux/provenance/reactiont5v2_split_provenance.json",
730
+ "aux/special_tokens_map.json",
731
+ "aux/tokenizer.json",
732
+ "aux/tokenizer_config.json",
733
+ "aux/trainer_state.json",
734
+ "aux/training_args.bin",
735
+ "benchmarks/default_test.jsonl",
736
+ "benchmarks/metadata.json"
737
+ ],
738
+ "benchmark_datasets": [
739
+ {
740
+ "dataset_id": "sagawa_reactiont5v2_retrosynthesis_uspto_50k_test",
741
+ "dataset_ref": "uspto_50k_project_release",
742
+ "default": true,
743
+ "format": "jsonl",
744
+ "metadata_path": "benchmarks/metadata.json",
745
+ "native_input_format": "csv_reaction_rows",
746
+ "native_source_paths": [
747
+ "aux/provenance/datasets/USPTO_50k/test.csv"
748
+ ],
749
+ "notes": [
750
+ "Native benchmark split for the public USPTO-50K fine-tuned retrosynthesis checkpoint."
751
+ ],
752
+ "num_examples": 5004,
753
+ "path": "benchmarks/default_test.jsonl",
754
+ "split": "test"
755
+ }
756
+ ],
757
+ "checkpoint_id": "sagawa_reactiont5v2_retrosynthesis_uspto_50k",
758
+ "display_name": "sagawa/ReactionT5v2-retrosynthesis-USPTO_50k",
759
+ "embedding_modes": [
760
+ "product_encoder",
761
+ "reactant_decoder",
762
+ "reaction_pair",
763
+ "token"
764
+ ],
765
+ "embedding_supported": true,
766
+ "fine_tuning_notes": "USPTO-50K fine-tuned retrosynthesis checkpoint that can be adapted further.",
767
+ "fine_tuning_role": "task_finetuned_seed",
768
+ "fine_tuning_supported": true,
769
+ "hf_model_name": "sagawa/ReactionT5v2-retrosynthesis-USPTO_50k",
770
+ "host": "Hugging Face",
771
+ "inference_supported": true,
772
+ "input_schema": null,
773
+ "larger_than_uspto50k": false,
774
+ "model_id": "reactiont5v2",
775
+ "model_name": "ReactionT5v2",
776
+ "normalized_checkpoint_id": "sagawa_reactiont5v2_retrosynthesis_uspto_50k",
777
+ "package_dir": "checkpoints/reactiont5v2/sagawa_reactiont5v2_retrosynthesis_uspto_50k",
778
+ "remote_url": "https://huggingface.co/sagawa/ReactionT5v2-retrosynthesis-USPTO_50k",
779
+ "source_registry_refs": [
780
+ {
781
+ "checkpoint_id": "sagawa_reactiont5v2_retrosynthesis_uspto_50k",
782
+ "model_id": "reactiont5v2"
783
+ }
784
+ ],
785
+ "task_direction": null,
786
+ "training_dataset_refs": [
787
+ "uspto_50k_project_release"
788
+ ],
789
+ "unsupported_reason": null,
790
+ "weights_path": "model.safetensors",
791
+ "wrapper_class": "syntheseus_wrappers.wrappers.reactiont5v2.ReactionT5v2Model",
792
+ "wrapper_family": "seq2seq_t5"
793
+ },
794
+ {
795
+ "auxiliary_files": [
796
+ "aux/models.json",
797
+ "benchmarks/default_test.jsonl",
798
+ "benchmarks/metadata.json",
799
+ "smiles_transformer/aux/checkpoint_metadata.json",
800
+ "smiles_transformer/aux/hyper_parameters.json",
801
+ "smiles_transformer/aux/vocab.txt",
802
+ "template_localization/aux/checkpoint_metadata.json",
803
+ "template_localization/aux/hyper_parameters.json",
804
+ "template_localization/aux/template_lib.json"
805
+ ],
806
+ "benchmark_datasets": [
807
+ {
808
+ "benchmark_variant": "benchmark_pending",
809
+ "dataset_id": "pistachio_test",
810
+ "dataset_ref": "pistachio_retrochimera_release",
811
+ "default": true,
812
+ "format": "jsonl",
813
+ "metadata_path": "benchmarks/metadata.json",
814
+ "native_input_format": "benchmark_not_available",
815
+ "native_source_paths": [],
816
+ "notes": [
817
+ "Benchmark companion for the end-to-end RetroChimera hybrid checkpoint on Pistachio.",
818
+ "These RetroChimera packages bundle both released ensemble components together and are intended for the upstream combined inference API.",
819
+ "No public benchmark split is bundled yet for this checkpoint; benchmarks/default_test.jsonl is intentionally empty for now.",
820
+ "The public RetroChimera checkpoint release documents Pistachio-trained checkpoints, but this repository does not yet bundle a public Pistachio evaluation split."
821
+ ],
822
+ "num_examples": 0,
823
+ "path": "benchmarks/default_test.jsonl",
824
+ "split": "test"
825
+ }
826
+ ],
827
+ "checkpoint_id": "pistachio",
828
+ "display_name": "pistachio",
829
+ "embedding_modes": [],
830
+ "embedding_supported": false,
831
+ "fine_tuning_notes": null,
832
+ "fine_tuning_role": null,
833
+ "fine_tuning_supported": false,
834
+ "hf_model_name": null,
835
+ "host": "Figshare file",
836
+ "inference_supported": true,
837
+ "input_schema": null,
838
+ "larger_than_uspto50k": true,
839
+ "model_id": "retrochimera_hybrid",
840
+ "model_name": "RetroChimera Hybrid",
841
+ "normalized_checkpoint_id": "pistachio",
842
+ "package_dir": "checkpoints/retrochimera_hybrid/pistachio",
843
+ "remote_url": "https://figshare.com/ndownloader/files/59468882",
844
+ "source_registry_refs": [
845
+ {
846
+ "checkpoint_id": "pistachio",
847
+ "model_id": "retrochimera_hybrid"
848
+ }
849
+ ],
850
+ "task_direction": null,
851
+ "training_dataset_refs": [
852
+ "pistachio_retrochimera_release"
853
+ ],
854
+ "unsupported_reason": null,
855
+ "weights_path": "template_localization/model.safetensors",
856
+ "wrapper_class": "syntheseus_wrappers.wrappers.retrochimera_hybrid.RetroChimeraHybridModel",
857
+ "wrapper_family": "retrochimera_hybrid"
858
+ },
859
+ {
860
+ "auxiliary_files": [
861
+ "aux/models.json",
862
+ "benchmarks/default_test.jsonl",
863
+ "benchmarks/metadata.json",
864
+ "benchmarks/native/src-test.txt",
865
+ "benchmarks/native/tgt-test.txt",
866
+ "smiles_transformer/aux/checkpoint_metadata.json",
867
+ "smiles_transformer/aux/hyper_parameters.json",
868
+ "smiles_transformer/aux/vocab.txt",
869
+ "template_localization/aux/checkpoint_metadata.json",
870
+ "template_localization/aux/hyper_parameters.json",
871
+ "template_localization/aux/template_lib.json"
872
+ ],
873
+ "benchmark_datasets": [
874
+ {
875
+ "benchmark_variant": "retrochimera_hybrid_gln_compatibility_view",
876
+ "dataset_id": "uspto_50k_test",
877
+ "dataset_ref": "uspto_50k_gln",
878
+ "default": true,
879
+ "format": "jsonl",
880
+ "metadata_path": "benchmarks/metadata.json",
881
+ "native_input_format": "tokenized_parallel_txt",
882
+ "native_source_paths": [
883
+ "benchmarks/native/src-test.txt",
884
+ "benchmarks/native/tgt-test.txt"
885
+ ],
886
+ "notes": [
887
+ "Benchmark companion for the end-to-end RetroChimera hybrid checkpoint on USPTO-50K via GLN/Coley split compatibility view.",
888
+ "These RetroChimera packages bundle both released ensemble components together and are intended for the upstream combined inference API.",
889
+ "The underlying test split is preserved from the local R-SMILES/GLN-compatible tokenized benchmark files already tracked by this repository.",
890
+ "The source files preserve rooted augmentation with factor 20; benchmarks/default_test.jsonl collapses each contiguous augmentation block back to one benchmark case per underlying reaction.",
891
+ "The native source files are tokenized parallel text; benchmarks/default_test.jsonl stores the corresponding de-tokenized product/reactant pairs.",
892
+ "RetroChimera cites Dai et al./GLN preprocessing for USPTO-50K; GLN in turn follows the Coley/RetroSim 80/10/10 split family."
893
+ ],
894
+ "num_examples": 5007,
895
+ "path": "benchmarks/default_test.jsonl",
896
+ "split": "test"
897
+ }
898
+ ],
899
+ "checkpoint_id": "uspto_50k",
900
+ "display_name": "uspto_50k",
901
+ "embedding_modes": [],
902
+ "embedding_supported": false,
903
+ "fine_tuning_notes": null,
904
+ "fine_tuning_role": null,
905
+ "fine_tuning_supported": false,
906
+ "hf_model_name": null,
907
+ "host": "Figshare file",
908
+ "inference_supported": true,
909
+ "input_schema": null,
910
+ "larger_than_uspto50k": false,
911
+ "model_id": "retrochimera_hybrid",
912
+ "model_name": "RetroChimera Hybrid",
913
+ "normalized_checkpoint_id": "uspto_50k",
914
+ "package_dir": "checkpoints/retrochimera_hybrid/uspto_50k",
915
+ "remote_url": "https://figshare.com/ndownloader/files/59511926",
916
+ "source_registry_refs": [
917
+ {
918
+ "checkpoint_id": "uspto_50k",
919
+ "model_id": "retrochimera_hybrid"
920
+ }
921
+ ],
922
+ "task_direction": null,
923
+ "training_dataset_refs": [
924
+ "uspto_50k_gln"
925
+ ],
926
+ "unsupported_reason": null,
927
+ "weights_path": "template_localization/model.safetensors",
928
+ "wrapper_class": "syntheseus_wrappers.wrappers.retrochimera_hybrid.RetroChimeraHybridModel",
929
+ "wrapper_family": "retrochimera_hybrid"
930
+ },
931
+ {
932
+ "auxiliary_files": [
933
+ "aux/models.json",
934
+ "benchmarks/default_test.jsonl",
935
+ "benchmarks/metadata.json",
936
+ "benchmarks/native/src-test.txt",
937
+ "benchmarks/native/tgt-test.txt",
938
+ "smiles_transformer/aux/checkpoint_metadata.json",
939
+ "smiles_transformer/aux/hyper_parameters.json",
940
+ "smiles_transformer/aux/vocab.txt",
941
+ "template_localization/aux/checkpoint_metadata.json",
942
+ "template_localization/aux/hyper_parameters.json",
943
+ "template_localization/aux/template_lib.json"
944
+ ],
945
+ "benchmark_datasets": [
946
+ {
947
+ "benchmark_variant": "retrochimera_hybrid_r_smiles_selected_full_view",
948
+ "dataset_id": "uspto_full_test",
949
+ "dataset_ref": "uspto_full_gln",
950
+ "default": true,
951
+ "format": "jsonl",
952
+ "metadata_path": "benchmarks/metadata.json",
953
+ "native_input_format": "tokenized_parallel_txt",
954
+ "native_source_paths": [
955
+ "benchmarks/native/src-test.txt",
956
+ "benchmarks/native/tgt-test.txt"
957
+ ],
958
+ "notes": [
959
+ "Benchmark companion for the end-to-end RetroChimera hybrid checkpoint on USPTO-FULL via RetroChimera\u2192R-SMILES\u2192GLN compatibility view.",
960
+ "These RetroChimera packages bundle both released ensemble components together and are intended for the upstream combined inference API.",
961
+ "The underlying test split is preserved from the local R-SMILES/GLN-compatible tokenized benchmark files already tracked by this repository.",
962
+ "The source files preserve rooted augmentation with factor 5; benchmarks/default_test.jsonl collapses each contiguous augmentation block back to one benchmark case per underlying reaction.",
963
+ "The native source files are tokenized parallel text; benchmarks/default_test.jsonl stores the corresponding de-tokenized product/reactant pairs.",
964
+ "RetroChimera notes that multiple USPTO-FULL variants exist and selects the R-SMILES/Zhong et al. version for fair comparison; the R-SMILES release points that data lineage back to GLN."
965
+ ],
966
+ "num_examples": 96023,
967
+ "path": "benchmarks/default_test.jsonl",
968
+ "split": "test"
969
+ }
970
+ ],
971
+ "checkpoint_id": "uspto_full",
972
+ "display_name": "uspto_full",
973
+ "embedding_modes": [],
974
+ "embedding_supported": false,
975
+ "fine_tuning_notes": null,
976
+ "fine_tuning_role": null,
977
+ "fine_tuning_supported": false,
978
+ "hf_model_name": null,
979
+ "host": "Figshare file",
980
+ "inference_supported": true,
981
+ "input_schema": null,
982
+ "larger_than_uspto50k": true,
983
+ "model_id": "retrochimera_hybrid",
984
+ "model_name": "RetroChimera Hybrid",
985
+ "normalized_checkpoint_id": "uspto_full",
986
+ "package_dir": "checkpoints/retrochimera_hybrid/uspto_full",
987
+ "remote_url": "https://figshare.com/ndownloader/files/59494598",
988
+ "source_registry_refs": [
989
+ {
990
+ "checkpoint_id": "uspto_full",
991
+ "model_id": "retrochimera_hybrid"
992
+ }
993
+ ],
994
+ "task_direction": null,
995
+ "training_dataset_refs": [
996
+ "uspto_full_gln"
997
+ ],
998
+ "unsupported_reason": null,
999
+ "weights_path": "template_localization/model.safetensors",
1000
+ "wrapper_class": "syntheseus_wrappers.wrappers.retrochimera_hybrid.RetroChimeraHybridModel",
1001
+ "wrapper_family": "retrochimera_hybrid"
1002
+ },
1003
+ {
1004
+ "auxiliary_files": [
1005
+ "aux/checkpoint_metadata.json",
1006
+ "aux/hyper_parameters.json",
1007
+ "aux/models.json",
1008
+ "aux/template_lib.json",
1009
+ "benchmarks/default_test.jsonl",
1010
+ "benchmarks/metadata.json"
1011
+ ],
1012
+ "benchmark_datasets": [
1013
+ {
1014
+ "benchmark_variant": "benchmark_pending",
1015
+ "dataset_id": "pistachio_test",
1016
+ "dataset_ref": "pistachio_retrochimera_release",
1017
+ "default": true,
1018
+ "format": "jsonl",
1019
+ "metadata_path": "benchmarks/metadata.json",
1020
+ "native_input_format": "benchmark_not_available",
1021
+ "native_source_paths": [],
1022
+ "notes": [
1023
+ "Benchmark companion for RetroChimera component checkpoint on Pistachio.",
1024
+ "These RetroChimera packages contain component checkpoints rather than the full end-to-end ensemble, so the bundled benchmark is best interpreted as a dataset compatibility view.",
1025
+ "No public benchmark split is bundled yet for this checkpoint; benchmarks/default_test.jsonl is intentionally empty for now.",
1026
+ "The public RetroChimera checkpoint release documents Pistachio-trained checkpoints, but this repository does not yet bundle a public Pistachio evaluation split."
1027
+ ],
1028
+ "num_examples": 0,
1029
+ "path": "benchmarks/default_test.jsonl",
1030
+ "split": "test"
1031
+ }
1032
+ ],
1033
+ "checkpoint_id": "pistachio",
1034
+ "display_name": "pistachio",
1035
+ "embedding_modes": [],
1036
+ "embedding_supported": false,
1037
+ "fine_tuning_notes": null,
1038
+ "fine_tuning_role": "task_finetuned_seed",
1039
+ "fine_tuning_supported": true,
1040
+ "hf_model_name": null,
1041
+ "host": "Figshare file",
1042
+ "inference_supported": true,
1043
+ "input_schema": null,
1044
+ "larger_than_uspto50k": true,
1045
+ "model_id": "retrochimera_template",
1046
+ "model_name": "RetroChimera Template Localization",
1047
+ "normalized_checkpoint_id": "pistachio",
1048
+ "package_dir": "checkpoints/retrochimera_template/pistachio",
1049
+ "remote_url": "https://figshare.com/ndownloader/files/59468882",
1050
+ "source_registry_refs": [
1051
+ {
1052
+ "checkpoint_id": "pistachio",
1053
+ "model_id": "retrochimera_template"
1054
+ }
1055
+ ],
1056
+ "task_direction": null,
1057
+ "training_dataset_refs": [
1058
+ "pistachio_retrochimera_release"
1059
+ ],
1060
+ "unsupported_reason": null,
1061
+ "weights_path": "model.safetensors",
1062
+ "wrapper_class": "syntheseus_wrappers.wrappers.retrochimera_template.RetroChimeraTemplateModel",
1063
+ "wrapper_family": "retrochimera_template_localization"
1064
+ },
1065
+ {
1066
+ "auxiliary_files": [
1067
+ "aux/checkpoint_metadata.json",
1068
+ "aux/hyper_parameters.json",
1069
+ "aux/models.json",
1070
+ "aux/template_lib.json",
1071
+ "benchmarks/default_test.jsonl",
1072
+ "benchmarks/metadata.json",
1073
+ "benchmarks/native/src-test.txt",
1074
+ "benchmarks/native/tgt-test.txt"
1075
+ ],
1076
+ "benchmark_datasets": [
1077
+ {
1078
+ "benchmark_variant": "retrochimera_component_gln_compatibility_view",
1079
+ "dataset_id": "uspto_50k_test",
1080
+ "dataset_ref": "uspto_50k_gln",
1081
+ "default": true,
1082
+ "format": "jsonl",
1083
+ "metadata_path": "benchmarks/metadata.json",
1084
+ "native_input_format": "tokenized_parallel_txt",
1085
+ "native_source_paths": [
1086
+ "benchmarks/native/src-test.txt",
1087
+ "benchmarks/native/tgt-test.txt"
1088
+ ],
1089
+ "notes": [
1090
+ "Benchmark companion for RetroChimera component checkpoint on USPTO-50K via GLN/Coley split compatibility view.",
1091
+ "These RetroChimera packages contain component checkpoints rather than the full end-to-end ensemble, so the bundled benchmark is best interpreted as a dataset compatibility view.",
1092
+ "The underlying test split is preserved from the local R-SMILES/GLN-compatible tokenized benchmark files already tracked by this repository.",
1093
+ "The source files preserve rooted augmentation with factor 20; benchmarks/default_test.jsonl collapses each contiguous augmentation block back to one benchmark case per underlying reaction.",
1094
+ "The native source files are tokenized parallel text; benchmarks/default_test.jsonl stores the corresponding de-tokenized product/reactant pairs.",
1095
+ "RetroChimera cites Dai et al./GLN preprocessing for USPTO-50K; GLN in turn follows the Coley/RetroSim 80/10/10 split family."
1096
+ ],
1097
+ "num_examples": 5007,
1098
+ "path": "benchmarks/default_test.jsonl",
1099
+ "split": "test"
1100
+ }
1101
+ ],
1102
+ "checkpoint_id": "uspto_50k",
1103
+ "display_name": "uspto_50k",
1104
+ "embedding_modes": [],
1105
+ "embedding_supported": false,
1106
+ "fine_tuning_notes": null,
1107
+ "fine_tuning_role": "task_finetuned_seed",
1108
+ "fine_tuning_supported": true,
1109
+ "hf_model_name": null,
1110
+ "host": "Figshare file",
1111
+ "inference_supported": true,
1112
+ "input_schema": null,
1113
+ "larger_than_uspto50k": false,
1114
+ "model_id": "retrochimera_template",
1115
+ "model_name": "RetroChimera Template Localization",
1116
+ "normalized_checkpoint_id": "uspto_50k",
1117
+ "package_dir": "checkpoints/retrochimera_template/uspto_50k",
1118
+ "remote_url": "https://figshare.com/ndownloader/files/59511926",
1119
+ "source_registry_refs": [
1120
+ {
1121
+ "checkpoint_id": "uspto_50k",
1122
+ "model_id": "retrochimera_template"
1123
+ }
1124
+ ],
1125
+ "task_direction": null,
1126
+ "training_dataset_refs": [
1127
+ "uspto_50k_gln"
1128
+ ],
1129
+ "unsupported_reason": null,
1130
+ "weights_path": "model.safetensors",
1131
+ "wrapper_class": "syntheseus_wrappers.wrappers.retrochimera_template.RetroChimeraTemplateModel",
1132
+ "wrapper_family": "retrochimera_template_localization"
1133
+ },
1134
+ {
1135
+ "auxiliary_files": [
1136
+ "aux/checkpoint_metadata.json",
1137
+ "aux/hyper_parameters.json",
1138
+ "aux/models.json",
1139
+ "aux/template_lib.json",
1140
+ "benchmarks/default_test.jsonl",
1141
+ "benchmarks/metadata.json",
1142
+ "benchmarks/native/src-test.txt",
1143
+ "benchmarks/native/tgt-test.txt"
1144
+ ],
1145
+ "benchmark_datasets": [
1146
+ {
1147
+ "benchmark_variant": "retrochimera_component_r_smiles_selected_full_view",
1148
+ "dataset_id": "uspto_full_test",
1149
+ "dataset_ref": "uspto_full_gln",
1150
+ "default": true,
1151
+ "format": "jsonl",
1152
+ "metadata_path": "benchmarks/metadata.json",
1153
+ "native_input_format": "tokenized_parallel_txt",
1154
+ "native_source_paths": [
1155
+ "benchmarks/native/src-test.txt",
1156
+ "benchmarks/native/tgt-test.txt"
1157
+ ],
1158
+ "notes": [
1159
+ "Benchmark companion for RetroChimera component checkpoint on USPTO-FULL via RetroChimera\u2192R-SMILES\u2192GLN compatibility view.",
1160
+ "These RetroChimera packages contain component checkpoints rather than the full end-to-end ensemble, so the bundled benchmark is best interpreted as a dataset compatibility view.",
1161
+ "The underlying test split is preserved from the local R-SMILES/GLN-compatible tokenized benchmark files already tracked by this repository.",
1162
+ "The source files preserve rooted augmentation with factor 5; benchmarks/default_test.jsonl collapses each contiguous augmentation block back to one benchmark case per underlying reaction.",
1163
+ "The native source files are tokenized parallel text; benchmarks/default_test.jsonl stores the corresponding de-tokenized product/reactant pairs.",
1164
+ "RetroChimera notes that multiple USPTO-FULL variants exist and selects the R-SMILES/Zhong et al. version for fair comparison; the R-SMILES release points that data lineage back to GLN."
1165
+ ],
1166
+ "num_examples": 96023,
1167
+ "path": "benchmarks/default_test.jsonl",
1168
+ "split": "test"
1169
+ }
1170
+ ],
1171
+ "checkpoint_id": "uspto_full",
1172
+ "display_name": "uspto_full",
1173
+ "embedding_modes": [],
1174
+ "embedding_supported": false,
1175
+ "fine_tuning_notes": null,
1176
+ "fine_tuning_role": "task_finetuned_seed",
1177
+ "fine_tuning_supported": true,
1178
+ "hf_model_name": null,
1179
+ "host": "Figshare file",
1180
+ "inference_supported": true,
1181
+ "input_schema": null,
1182
+ "larger_than_uspto50k": true,
1183
+ "model_id": "retrochimera_template",
1184
+ "model_name": "RetroChimera Template Localization",
1185
+ "normalized_checkpoint_id": "uspto_full",
1186
+ "package_dir": "checkpoints/retrochimera_template/uspto_full",
1187
+ "remote_url": "https://figshare.com/ndownloader/files/59494598",
1188
+ "source_registry_refs": [
1189
+ {
1190
+ "checkpoint_id": "uspto_full",
1191
+ "model_id": "retrochimera_template"
1192
+ }
1193
+ ],
1194
+ "task_direction": null,
1195
+ "training_dataset_refs": [
1196
+ "uspto_full_gln"
1197
+ ],
1198
+ "unsupported_reason": null,
1199
+ "weights_path": "model.safetensors",
1200
+ "wrapper_class": "syntheseus_wrappers.wrappers.retrochimera_template.RetroChimeraTemplateModel",
1201
+ "wrapper_family": "retrochimera_template_localization"
1202
+ },
1203
+ {
1204
+ "auxiliary_files": [
1205
+ "aux/checkpoint_metadata.json",
1206
+ "aux/hyper_parameters.json",
1207
+ "aux/models.json",
1208
+ "aux/vocab.txt",
1209
+ "benchmarks/default_test.jsonl",
1210
+ "benchmarks/metadata.json"
1211
+ ],
1212
+ "benchmark_datasets": [
1213
+ {
1214
+ "benchmark_variant": "benchmark_pending",
1215
+ "dataset_id": "pistachio_test",
1216
+ "dataset_ref": "pistachio_retrochimera_release",
1217
+ "default": true,
1218
+ "format": "jsonl",
1219
+ "metadata_path": "benchmarks/metadata.json",
1220
+ "native_input_format": "benchmark_not_available",
1221
+ "native_source_paths": [],
1222
+ "notes": [
1223
+ "Benchmark companion for RetroChimera component checkpoint on Pistachio.",
1224
+ "These RetroChimera packages contain component checkpoints rather than the full end-to-end ensemble, so the bundled benchmark is best interpreted as a dataset compatibility view.",
1225
+ "No public benchmark split is bundled yet for this checkpoint; benchmarks/default_test.jsonl is intentionally empty for now.",
1226
+ "The public RetroChimera checkpoint release documents Pistachio-trained checkpoints, but this repository does not yet bundle a public Pistachio evaluation split."
1227
+ ],
1228
+ "num_examples": 0,
1229
+ "path": "benchmarks/default_test.jsonl",
1230
+ "split": "test"
1231
+ }
1232
+ ],
1233
+ "checkpoint_id": "pistachio",
1234
+ "display_name": "pistachio",
1235
+ "embedding_modes": [],
1236
+ "embedding_supported": false,
1237
+ "fine_tuning_notes": null,
1238
+ "fine_tuning_role": "task_finetuned_seed",
1239
+ "fine_tuning_supported": true,
1240
+ "hf_model_name": null,
1241
+ "host": "Figshare file",
1242
+ "inference_supported": true,
1243
+ "input_schema": null,
1244
+ "larger_than_uspto50k": true,
1245
+ "model_id": "retrochimera_transformer",
1246
+ "model_name": "RetroChimera Smiles Transformer",
1247
+ "normalized_checkpoint_id": "pistachio",
1248
+ "package_dir": "checkpoints/retrochimera_transformer/pistachio",
1249
+ "remote_url": "https://figshare.com/ndownloader/files/59468882",
1250
+ "source_registry_refs": [
1251
+ {
1252
+ "checkpoint_id": "pistachio",
1253
+ "model_id": "retrochimera_transformer"
1254
+ }
1255
+ ],
1256
+ "task_direction": null,
1257
+ "training_dataset_refs": [
1258
+ "pistachio_retrochimera_release"
1259
+ ],
1260
+ "unsupported_reason": null,
1261
+ "weights_path": "model.safetensors",
1262
+ "wrapper_class": "syntheseus_wrappers.wrappers.retrochimera_transformer.RetroChimeraTransformerModel",
1263
+ "wrapper_family": "retrochimera_smiles_transformer"
1264
+ },
1265
+ {
1266
+ "auxiliary_files": [
1267
+ "aux/checkpoint_metadata.json",
1268
+ "aux/hyper_parameters.json",
1269
+ "aux/models.json",
1270
+ "aux/vocab.txt",
1271
+ "benchmarks/default_test.jsonl",
1272
+ "benchmarks/metadata.json",
1273
+ "benchmarks/native/src-test.txt",
1274
+ "benchmarks/native/tgt-test.txt"
1275
+ ],
1276
+ "benchmark_datasets": [
1277
+ {
1278
+ "benchmark_variant": "retrochimera_component_gln_compatibility_view",
1279
+ "dataset_id": "uspto_50k_test",
1280
+ "dataset_ref": "uspto_50k_gln",
1281
+ "default": true,
1282
+ "format": "jsonl",
1283
+ "metadata_path": "benchmarks/metadata.json",
1284
+ "native_input_format": "tokenized_parallel_txt",
1285
+ "native_source_paths": [
1286
+ "benchmarks/native/src-test.txt",
1287
+ "benchmarks/native/tgt-test.txt"
1288
+ ],
1289
+ "notes": [
1290
+ "Benchmark companion for RetroChimera component checkpoint on USPTO-50K via GLN/Coley split compatibility view.",
1291
+ "These RetroChimera packages contain component checkpoints rather than the full end-to-end ensemble, so the bundled benchmark is best interpreted as a dataset compatibility view.",
1292
+ "The underlying test split is preserved from the local R-SMILES/GLN-compatible tokenized benchmark files already tracked by this repository.",
1293
+ "The source files preserve rooted augmentation with factor 20; benchmarks/default_test.jsonl collapses each contiguous augmentation block back to one benchmark case per underlying reaction.",
1294
+ "The native source files are tokenized parallel text; benchmarks/default_test.jsonl stores the corresponding de-tokenized product/reactant pairs.",
1295
+ "RetroChimera cites Dai et al./GLN preprocessing for USPTO-50K; GLN in turn follows the Coley/RetroSim 80/10/10 split family."
1296
+ ],
1297
+ "num_examples": 5007,
1298
+ "path": "benchmarks/default_test.jsonl",
1299
+ "split": "test"
1300
+ }
1301
+ ],
1302
+ "checkpoint_id": "uspto_50k",
1303
+ "display_name": "uspto_50k",
1304
+ "embedding_modes": [],
1305
+ "embedding_supported": false,
1306
+ "fine_tuning_notes": null,
1307
+ "fine_tuning_role": "task_finetuned_seed",
1308
+ "fine_tuning_supported": true,
1309
+ "hf_model_name": null,
1310
+ "host": "Figshare file",
1311
+ "inference_supported": true,
1312
+ "input_schema": null,
1313
+ "larger_than_uspto50k": false,
1314
+ "model_id": "retrochimera_transformer",
1315
+ "model_name": "RetroChimera Smiles Transformer",
1316
+ "normalized_checkpoint_id": "uspto_50k",
1317
+ "package_dir": "checkpoints/retrochimera_transformer/uspto_50k",
1318
+ "remote_url": "https://figshare.com/ndownloader/files/59511926",
1319
+ "source_registry_refs": [
1320
+ {
1321
+ "checkpoint_id": "uspto_50k",
1322
+ "model_id": "retrochimera_transformer"
1323
+ }
1324
+ ],
1325
+ "task_direction": null,
1326
+ "training_dataset_refs": [
1327
+ "uspto_50k_gln"
1328
+ ],
1329
+ "unsupported_reason": null,
1330
+ "weights_path": "model.safetensors",
1331
+ "wrapper_class": "syntheseus_wrappers.wrappers.retrochimera_transformer.RetroChimeraTransformerModel",
1332
+ "wrapper_family": "retrochimera_smiles_transformer"
1333
+ },
1334
+ {
1335
+ "auxiliary_files": [
1336
+ "aux/checkpoint_metadata.json",
1337
+ "aux/hyper_parameters.json",
1338
+ "aux/models.json",
1339
+ "aux/vocab.txt",
1340
+ "benchmarks/default_test.jsonl",
1341
+ "benchmarks/metadata.json",
1342
+ "benchmarks/native/src-test.txt",
1343
+ "benchmarks/native/tgt-test.txt"
1344
+ ],
1345
+ "benchmark_datasets": [
1346
+ {
1347
+ "benchmark_variant": "retrochimera_component_r_smiles_selected_full_view",
1348
+ "dataset_id": "uspto_full_test",
1349
+ "dataset_ref": "uspto_full_gln",
1350
+ "default": true,
1351
+ "format": "jsonl",
1352
+ "metadata_path": "benchmarks/metadata.json",
1353
+ "native_input_format": "tokenized_parallel_txt",
1354
+ "native_source_paths": [
1355
+ "benchmarks/native/src-test.txt",
1356
+ "benchmarks/native/tgt-test.txt"
1357
+ ],
1358
+ "notes": [
1359
+ "Benchmark companion for RetroChimera component checkpoint on USPTO-FULL via RetroChimera\u2192R-SMILES\u2192GLN compatibility view.",
1360
+ "These RetroChimera packages contain component checkpoints rather than the full end-to-end ensemble, so the bundled benchmark is best interpreted as a dataset compatibility view.",
1361
+ "The underlying test split is preserved from the local R-SMILES/GLN-compatible tokenized benchmark files already tracked by this repository.",
1362
+ "The source files preserve rooted augmentation with factor 5; benchmarks/default_test.jsonl collapses each contiguous augmentation block back to one benchmark case per underlying reaction.",
1363
+ "The native source files are tokenized parallel text; benchmarks/default_test.jsonl stores the corresponding de-tokenized product/reactant pairs.",
1364
+ "RetroChimera notes that multiple USPTO-FULL variants exist and selects the R-SMILES/Zhong et al. version for fair comparison; the R-SMILES release points that data lineage back to GLN."
1365
+ ],
1366
+ "num_examples": 96023,
1367
+ "path": "benchmarks/default_test.jsonl",
1368
+ "split": "test"
1369
+ }
1370
+ ],
1371
+ "checkpoint_id": "uspto_full",
1372
+ "display_name": "uspto_full",
1373
+ "embedding_modes": [],
1374
+ "embedding_supported": false,
1375
+ "fine_tuning_notes": null,
1376
+ "fine_tuning_role": "task_finetuned_seed",
1377
+ "fine_tuning_supported": true,
1378
+ "hf_model_name": null,
1379
+ "host": "Figshare file",
1380
+ "inference_supported": true,
1381
+ "input_schema": null,
1382
+ "larger_than_uspto50k": true,
1383
+ "model_id": "retrochimera_transformer",
1384
+ "model_name": "RetroChimera Smiles Transformer",
1385
+ "normalized_checkpoint_id": "uspto_full",
1386
+ "package_dir": "checkpoints/retrochimera_transformer/uspto_full",
1387
+ "remote_url": "https://figshare.com/ndownloader/files/59494598",
1388
+ "source_registry_refs": [
1389
+ {
1390
+ "checkpoint_id": "uspto_full",
1391
+ "model_id": "retrochimera_transformer"
1392
+ }
1393
+ ],
1394
+ "task_direction": null,
1395
+ "training_dataset_refs": [
1396
+ "uspto_full_gln"
1397
+ ],
1398
+ "unsupported_reason": null,
1399
+ "weights_path": "model.safetensors",
1400
+ "wrapper_class": "syntheseus_wrappers.wrappers.retrochimera_transformer.RetroChimeraTransformerModel",
1401
+ "wrapper_family": "retrochimera_smiles_transformer"
1402
+ },
1403
+ {
1404
+ "auxiliary_files": [
1405
+ "aux/configs/base.yml",
1406
+ "aux/configs/rxngpt_llama1B.json",
1407
+ "aux/deepspeed.yaml",
1408
+ "aux/tokenizer/special_tokens_map.json",
1409
+ "aux/tokenizer/tokenizer.json",
1410
+ "aux/tokenizer/tokenizer_config.json",
1411
+ "aux/vocab.json",
1412
+ "benchmarks/default_test.jsonl",
1413
+ "benchmarks/metadata.json",
1414
+ "benchmarks/native/USPTO-50k_test_tidy.txt"
1415
+ ],
1416
+ "benchmark_datasets": [
1417
+ {
1418
+ "benchmark_variant": "class_unknown_test_aug_best_effort_underlying_test_examples",
1419
+ "dataset_id": "finetune_50k_aug20_pth_test",
1420
+ "dataset_ref": "uspto_50k_rsgpt",
1421
+ "default": true,
1422
+ "format": "jsonl",
1423
+ "metadata_path": "benchmarks/metadata.json",
1424
+ "native_input_format": "reaction_smiles_txt",
1425
+ "native_source_paths": [
1426
+ "benchmarks/native/USPTO-50k_test_tidy.txt"
1427
+ ],
1428
+ "notes": [
1429
+ "The exact upstream test-time augmentation prefix file test_50k_prefixs20.txt is not present in the local manual artifacts or pinned source snapshot.",
1430
+ "This bundle therefore preserves the underlying standard USPTO-50K test reactions as the normalized benchmark truth set, while the checkpoint metadata still records the intended aug20 evaluation variant."
1431
+ ],
1432
+ "num_examples": 5007,
1433
+ "path": "benchmarks/default_test.jsonl",
1434
+ "split": "test"
1435
+ }
1436
+ ],
1437
+ "checkpoint_id": "finetune_50k_aug20_pth",
1438
+ "display_name": "finetune_50k_aug20.pth",
1439
+ "embedding_modes": [],
1440
+ "embedding_supported": false,
1441
+ "fine_tuning_notes": null,
1442
+ "fine_tuning_role": "task_finetuned_seed",
1443
+ "fine_tuning_supported": true,
1444
+ "hf_model_name": null,
1445
+ "host": "Zenodo Sandbox",
1446
+ "inference_supported": true,
1447
+ "input_schema": null,
1448
+ "larger_than_uspto50k": false,
1449
+ "model_id": "rsgpt",
1450
+ "model_name": "RSGPT",
1451
+ "normalized_checkpoint_id": "finetune_50k_aug20_pth",
1452
+ "package_dir": "checkpoints/rsgpt/finetune_50k_aug20_pth",
1453
+ "remote_url": "https://sandbox.zenodo.org/records/203391",
1454
+ "source_registry_refs": [
1455
+ {
1456
+ "checkpoint_id": "finetune_50k_aug20_pth",
1457
+ "model_id": "rsgpt"
1458
+ }
1459
+ ],
1460
+ "task_direction": null,
1461
+ "training_dataset_refs": [
1462
+ "uspto_50k_rsgpt",
1463
+ "synthetic_10b_rsgpt"
1464
+ ],
1465
+ "unsupported_reason": null,
1466
+ "weights_path": "model.safetensors",
1467
+ "wrapper_class": "syntheseus_wrappers.wrappers.rsgpt.RSGPTModel",
1468
+ "wrapper_family": "rsgpt_llama_decoder"
1469
+ },
1470
+ {
1471
+ "auxiliary_files": [
1472
+ "aux/configs/base.yml",
1473
+ "aux/configs/rxngpt_llama1B.json",
1474
+ "aux/deepspeed.yaml",
1475
+ "aux/tokenizer/special_tokens_map.json",
1476
+ "aux/tokenizer/tokenizer.json",
1477
+ "aux/tokenizer/tokenizer_config.json",
1478
+ "aux/vocab.json",
1479
+ "benchmarks/default_test.jsonl",
1480
+ "benchmarks/metadata.json",
1481
+ "benchmarks/native/USPTO-50k_test_tidy_class1.csv"
1482
+ ],
1483
+ "benchmark_datasets": [
1484
+ {
1485
+ "benchmark_variant": "class_known",
1486
+ "dataset_id": "finetune_50k_pth_test",
1487
+ "dataset_ref": "uspto_50k_rsgpt",
1488
+ "default": true,
1489
+ "format": "jsonl",
1490
+ "metadata_path": "benchmarks/metadata.json",
1491
+ "native_input_format": "class_csv_reaction_smiles",
1492
+ "native_source_paths": [
1493
+ "benchmarks/native/USPTO-50k_test_tidy_class1.csv"
1494
+ ],
1495
+ "notes": [
1496
+ "This is the class-labeled USPTO-50K benchmark companion file corresponding to the best-effort class-known mapping for finetune_50k_pth."
1497
+ ],
1498
+ "num_examples": 5007,
1499
+ "path": "benchmarks/default_test.jsonl",
1500
+ "split": "test"
1501
+ }
1502
+ ],
1503
+ "checkpoint_id": "finetune_50k_pth",
1504
+ "display_name": "finetune_50k.pth",
1505
+ "embedding_modes": [],
1506
+ "embedding_supported": false,
1507
+ "fine_tuning_notes": null,
1508
+ "fine_tuning_role": "task_finetuned_seed",
1509
+ "fine_tuning_supported": true,
1510
+ "hf_model_name": null,
1511
+ "host": "Zenodo Sandbox",
1512
+ "inference_supported": true,
1513
+ "input_schema": null,
1514
+ "larger_than_uspto50k": false,
1515
+ "model_id": "rsgpt",
1516
+ "model_name": "RSGPT",
1517
+ "normalized_checkpoint_id": "finetune_50k_pth",
1518
+ "package_dir": "checkpoints/rsgpt/finetune_50k_pth",
1519
+ "remote_url": "https://sandbox.zenodo.org/records/203391",
1520
+ "source_registry_refs": [
1521
+ {
1522
+ "checkpoint_id": "finetune_50k_pth",
1523
+ "model_id": "rsgpt"
1524
+ }
1525
+ ],
1526
+ "task_direction": null,
1527
+ "training_dataset_refs": [
1528
+ "uspto_50k_rsgpt",
1529
+ "synthetic_10b_rsgpt"
1530
+ ],
1531
+ "unsupported_reason": null,
1532
+ "weights_path": "model.safetensors",
1533
+ "wrapper_class": "syntheseus_wrappers.wrappers.rsgpt.RSGPTModel",
1534
+ "wrapper_family": "rsgpt_llama_decoder"
1535
+ },
1536
+ {
1537
+ "auxiliary_files": [
1538
+ "aux/configs/base.yml",
1539
+ "aux/configs/rxngpt_llama1B.json",
1540
+ "aux/deepspeed.yaml",
1541
+ "aux/tokenizer/special_tokens_map.json",
1542
+ "aux/tokenizer/tokenizer.json",
1543
+ "aux/tokenizer/tokenizer_config.json",
1544
+ "aux/vocab.json",
1545
+ "benchmarks/default_test.jsonl",
1546
+ "benchmarks/metadata.json",
1547
+ "benchmarks/native/USPTO-FULL_test_tidy.txt"
1548
+ ],
1549
+ "benchmark_datasets": [
1550
+ {
1551
+ "dataset_id": "finetune_full_pth_test",
1552
+ "dataset_ref": "uspto_full_rsgpt",
1553
+ "default": true,
1554
+ "format": "jsonl",
1555
+ "metadata_path": "benchmarks/metadata.json",
1556
+ "native_input_format": "reaction_smiles_txt",
1557
+ "native_source_paths": [
1558
+ "benchmarks/native/USPTO-FULL_test_tidy.txt"
1559
+ ],
1560
+ "notes": [
1561
+ "Native RSGPT benchmark source extracted from USPTO_data/USPTO-FULL_test_tidy.txt."
1562
+ ],
1563
+ "num_examples": 95389,
1564
+ "path": "benchmarks/default_test.jsonl",
1565
+ "split": "test"
1566
+ }
1567
+ ],
1568
+ "checkpoint_id": "finetune_full_pth",
1569
+ "display_name": "finetune_full.pth",
1570
+ "embedding_modes": [],
1571
+ "embedding_supported": false,
1572
+ "fine_tuning_notes": null,
1573
+ "fine_tuning_role": "task_finetuned_seed",
1574
+ "fine_tuning_supported": true,
1575
+ "hf_model_name": null,
1576
+ "host": "Zenodo Sandbox",
1577
+ "inference_supported": true,
1578
+ "input_schema": null,
1579
+ "larger_than_uspto50k": true,
1580
+ "model_id": "rsgpt",
1581
+ "model_name": "RSGPT",
1582
+ "normalized_checkpoint_id": "finetune_full_pth",
1583
+ "package_dir": "checkpoints/rsgpt/finetune_full_pth",
1584
+ "remote_url": "https://sandbox.zenodo.org/records/203391",
1585
+ "source_registry_refs": [
1586
+ {
1587
+ "checkpoint_id": "finetune_full_pth",
1588
+ "model_id": "rsgpt"
1589
+ }
1590
+ ],
1591
+ "task_direction": null,
1592
+ "training_dataset_refs": [
1593
+ "uspto_full_rsgpt",
1594
+ "synthetic_10b_rsgpt"
1595
+ ],
1596
+ "unsupported_reason": null,
1597
+ "weights_path": "model.safetensors",
1598
+ "wrapper_class": "syntheseus_wrappers.wrappers.rsgpt.RSGPTModel",
1599
+ "wrapper_family": "rsgpt_llama_decoder"
1600
+ },
1601
+ {
1602
+ "auxiliary_files": [
1603
+ "aux/configs/base.yml",
1604
+ "aux/configs/rxngpt_llama1B.json",
1605
+ "aux/deepspeed.yaml",
1606
+ "aux/tokenizer/special_tokens_map.json",
1607
+ "aux/tokenizer/tokenizer.json",
1608
+ "aux/tokenizer/tokenizer_config.json",
1609
+ "aux/vocab.json",
1610
+ "benchmarks/default_test.jsonl",
1611
+ "benchmarks/metadata.json",
1612
+ "benchmarks/native/USPTO-MIT_test_tidy.txt"
1613
+ ],
1614
+ "benchmark_datasets": [
1615
+ {
1616
+ "dataset_id": "finetune_mit_pth_test",
1617
+ "dataset_ref": "uspto_mit_rsgpt",
1618
+ "default": true,
1619
+ "format": "jsonl",
1620
+ "metadata_path": "benchmarks/metadata.json",
1621
+ "native_input_format": "reaction_smiles_txt",
1622
+ "native_source_paths": [
1623
+ "benchmarks/native/USPTO-MIT_test_tidy.txt"
1624
+ ],
1625
+ "notes": [
1626
+ "Native RSGPT benchmark source extracted from USPTO_data/USPTO-MIT_test_tidy.txt."
1627
+ ],
1628
+ "num_examples": 40000,
1629
+ "path": "benchmarks/default_test.jsonl",
1630
+ "split": "test"
1631
+ }
1632
+ ],
1633
+ "checkpoint_id": "finetune_mit_pth",
1634
+ "display_name": "finetune_mit.pth",
1635
+ "embedding_modes": [],
1636
+ "embedding_supported": false,
1637
+ "fine_tuning_notes": null,
1638
+ "fine_tuning_role": "task_finetuned_seed",
1639
+ "fine_tuning_supported": true,
1640
+ "hf_model_name": null,
1641
+ "host": "Zenodo Sandbox",
1642
+ "inference_supported": true,
1643
+ "input_schema": null,
1644
+ "larger_than_uspto50k": true,
1645
+ "model_id": "rsgpt",
1646
+ "model_name": "RSGPT",
1647
+ "normalized_checkpoint_id": "finetune_mit_pth",
1648
+ "package_dir": "checkpoints/rsgpt/finetune_mit_pth",
1649
+ "remote_url": "https://sandbox.zenodo.org/records/203391",
1650
+ "source_registry_refs": [
1651
+ {
1652
+ "checkpoint_id": "finetune_mit_pth",
1653
+ "model_id": "rsgpt"
1654
+ }
1655
+ ],
1656
+ "task_direction": null,
1657
+ "training_dataset_refs": [
1658
+ "uspto_mit_rsgpt",
1659
+ "synthetic_10b_rsgpt"
1660
+ ],
1661
+ "unsupported_reason": null,
1662
+ "weights_path": "model.safetensors",
1663
+ "wrapper_class": "syntheseus_wrappers.wrappers.rsgpt.RSGPTModel",
1664
+ "wrapper_family": "rsgpt_llama_decoder"
1665
+ },
1666
+ {
1667
+ "auxiliary_files": [
1668
+ "aux/configs/base.yml",
1669
+ "aux/configs/rxngpt_llama1B.json",
1670
+ "aux/deepspeed.yaml",
1671
+ "aux/tokenizer/special_tokens_map.json",
1672
+ "aux/tokenizer/tokenizer.json",
1673
+ "aux/tokenizer/tokenizer_config.json",
1674
+ "aux/vocab.json",
1675
+ "benchmarks/default_test.jsonl",
1676
+ "benchmarks/metadata.json",
1677
+ "benchmarks/native/USPTO-50k_test_tidy.txt"
1678
+ ],
1679
+ "benchmark_datasets": [
1680
+ {
1681
+ "benchmark_variant": "class_unknown_no_pretraining_ablation",
1682
+ "dataset_id": "nopretrain_50k_pth_test",
1683
+ "dataset_ref": "uspto_50k_rsgpt",
1684
+ "default": true,
1685
+ "format": "jsonl",
1686
+ "metadata_path": "benchmarks/metadata.json",
1687
+ "native_input_format": "reaction_smiles_txt",
1688
+ "native_source_paths": [
1689
+ "benchmarks/native/USPTO-50k_test_tidy.txt"
1690
+ ],
1691
+ "notes": [
1692
+ "Native RSGPT benchmark source extracted from USPTO_data/USPTO-50k_test_tidy.txt."
1693
+ ],
1694
+ "num_examples": 5007,
1695
+ "path": "benchmarks/default_test.jsonl",
1696
+ "split": "test"
1697
+ }
1698
+ ],
1699
+ "checkpoint_id": "nopretrain_50k_pth",
1700
+ "display_name": "nopretrain_50k.pth",
1701
+ "embedding_modes": [],
1702
+ "embedding_supported": false,
1703
+ "fine_tuning_notes": "Task-specific 50K ablation checkpoint without the synthetic pretraining stage.",
1704
+ "fine_tuning_role": "task_finetuned_seed",
1705
+ "fine_tuning_supported": true,
1706
+ "hf_model_name": null,
1707
+ "host": "Zenodo Sandbox",
1708
+ "inference_supported": true,
1709
+ "input_schema": null,
1710
+ "larger_than_uspto50k": false,
1711
+ "model_id": "rsgpt",
1712
+ "model_name": "RSGPT",
1713
+ "normalized_checkpoint_id": "nopretrain_50k_pth",
1714
+ "package_dir": "checkpoints/rsgpt/nopretrain_50k_pth",
1715
+ "remote_url": "https://sandbox.zenodo.org/records/203391",
1716
+ "source_registry_refs": [
1717
+ {
1718
+ "checkpoint_id": "nopretrain_50k_pth",
1719
+ "model_id": "rsgpt"
1720
+ }
1721
+ ],
1722
+ "task_direction": null,
1723
+ "training_dataset_refs": [
1724
+ "uspto_50k_rsgpt"
1725
+ ],
1726
+ "unsupported_reason": null,
1727
+ "weights_path": "model.safetensors",
1728
+ "wrapper_class": "syntheseus_wrappers.wrappers.rsgpt.RSGPTModel",
1729
+ "wrapper_family": "rsgpt_llama_decoder"
1730
+ }
1731
+ ],
1732
+ "registry_kind": "runtime",
1733
+ "runtime_registry_version": 1
1734
+ }