File size: 5,609 Bytes
f12d393
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
{
  "model_version": "0.3.0-nli-experimental",
  "model_revision": "sha256:058be8f2224e2f995de5fd3c7b4b14bc3df19abc7874be82793727af2c1b68bf",
  "nli": true,
  "training": {
    "config": {
      "epochs": 2,
      "batch_size": 4,
      "gradient_accumulation": 4,
      "learning_rate": 2.8431918937187533e-05,
      "freeze_embeddings": true,
      "warmup_ratio": 0.1,
      "weight_decay": 0.09721437278134223,
      "label_smoothing": 0.0,
      "patience": 2,
      "bf16": true,
      "layer_lr_decay": 0.95,
      "seed": 42,
      "rdrop": 0.1,
      "freeze_lower_layers": 6,
      "gradient_checkpointing": false
    },
    "history": [
      {
        "epoch": 1,
        "train_loss": 0.5870543534764819,
        "validation": {
          "support": 132,
          "positive_support": 57,
          "precision": 0.9473684210526315,
          "recall": 0.9473684210526315,
          "f1": 0.9473684210526315,
          "macro_f1": 0.9536842105263157,
          "average_precision": 0.9783504049341503,
          "confusion_matrix": [
            [
              72,
              3
            ],
            [
              3,
              54
            ]
          ],
          "coverage": 0.9924242424242424,
          "review_rate": 0.007575757575757576,
          "review_count": 1,
          "redundant_count": 61,
          "not_redundant_count": 70,
          "selective_precision": 0.9016393442622951,
          "negative_predictive_value": 0.9714285714285714
        }
      },
      {
        "epoch": 2,
        "train_loss": 0.22414214792911724,
        "validation": {
          "support": 132,
          "positive_support": 57,
          "precision": 0.9473684210526315,
          "recall": 0.9473684210526315,
          "f1": 0.9473684210526315,
          "macro_f1": 0.9536842105263157,
          "average_precision": 0.978941878575674,
          "confusion_matrix": [
            [
              72,
              3
            ],
            [
              3,
              54
            ]
          ],
          "coverage": 0.9924242424242424,
          "review_rate": 0.007575757575757576,
          "review_count": 1,
          "redundant_count": 61,
          "not_redundant_count": 70,
          "selective_precision": 0.9016393442622951,
          "negative_predictive_value": 0.9714285714285714
        }
      }
    ],
    "selected_epoch": 2,
    "selected_parameter_delta_l2": 0.11925771832466125,
    "optimizer_steps": 94,
    "pruned": false,
    "policy": {
      "binary_threshold": 0.4262998402118683,
      "thresholds": {
        "low": 0.04250901937484741,
        "high": 0.08026428520679474,
        "positive_enabled": true,
        "negative_enabled": true
      },
      "selection_split": "validation",
      "precision_target": 0.9,
      "negative_predictive_value_target": 0.9,
      "minimum_support": 5,
      "decision_policy_version": "0.1.0"
    },
    "training_pairs": 744,
    "validation_pairs": 132,
    "trainable_parameters": 43514883,
    "total_parameters": 278811651,
    "duration_seconds": 80.389151964,
    "environment": {
      "python": "3.11.14",
      "platform": "Linux-7.1.8+deb13-amd64-x86_64-with-glibc2.41",
      "code_commit": "b17b4735f0c4b4cc05fc63afd04ce0d61ca64580",
      "source_sha256": {
        "cli.py": "8462dd1c90e06165826a5053c740c2e4379bc4db1a3fafacb38eccc866f007ab",
        "decision.py": "ce69d18701fb687f0fd94217a121f7ebdd29e9397c0c7a18c899dcede96f5e2a",
        "calibration.py": "d999744c3ed24167540740e010d98deb96ab23e01570f2de24d2e12d8758a044",
        "pipeline.py": "780d3fb9e827ab65a9aa21ec09eaecf1f4289fb2492c23a006659f08f0384933",
        "encoder.py": "1536cb1330fe00f6c30e41ee0ec393f6fb964da71c7e70a0448e286d7493a6c1",
        "nli.py": "f74b06c83c66bb6d6f500b01ee9831f56422e8aad44e2355bff48ab1c1b27558",
        "research_scores.py": "4a55de8f34cf82185e66d5192d5db120f3b27893b314f2fe51d4ed6df99d77d1",
        "pseudo.py": "b974fd870e6bf614131d097fa7f49aba77696d3b29f164a3ea1c094d7383ef74",
        "data.py": "5b8143e617f9166f13159f96081b9048b24fe6d80289b181fc165f1c307b7dde",
        "experiment.py": "fd3468655604e01ea8867e0e5aec266281239f5835d290e929728c1af2922e68",
        "score_graph.py": "252d95a1ba60fe26cfab1e9e57eb47b51ddf04d7b706ea52ced1f5376bd0f9ed",
        "domain_adaptation.py": "cab4001a9da57af5aad758d43ffcb1cb76c5cc6c97eac2c7a6c92f45a076ed8f",
        "training.py": "2a375b311b3a9e3e5d773e0b0f974ff859dd4f232d639c8176302128142aa8c0",
        "__main__.py": "b466d3224c1cfa96b82f891d496e36e7bd4ad1eb70fcc530b45c29ce62d3f9c4",
        "cross_encoder.py": "2fd31598b2ae3e039ffcbab908c1c6b190aa6aba74d117f5658fc2fde89b5437",
        "schemas.py": "361128ca989504938a2bf8977cbf0b1be49bee9296c5da4903b05736e149535c",
        "evaluation.py": "4880d0ee9fcf6fc15163204b2edcf4856dc06dfe5aaa3930a9dca236b16d0fd0",
        "__init__.py": "8f3ef65a08808435e85992bcb80379e0a060dab973ffbe2b5e193bc924766681",
        "v03_data.py": "989e17936dcc772dc414d6ff918929b4dd961ee279e4f337bec82579844f8479",
        "dataset.py": "bf9681d9a0ca6a1e283b7423d4eb367c478cffd8c9adfab5979e64b825f7648a"
      },
      "deterministic_warn_only": true,
      "cublas_workspace_config": ":4096:8",
      "gpu": "NVIDIA RTX A1000 6GB Laptop GPU",
      "packages": {
        "torch": "2.6.0",
        "sentence-transformers": "5.1.2",
        "transformers": "4.57.1",
        "numpy": "2.3.4",
        "scikit-learn": "1.7.2"
      }
    },
    "objective": "multitask directional NLI CE and symmetric redundancy BCE",
    "score": "minimum bidirectional entailment softmax; uncalibrated"
  }
}