teru00801 commited on
Commit
1929eef
·
verified ·
1 Parent(s): e071025

Upload folder using huggingface_hub

Browse files
20260228_014435/analysis_bundle/metrics/metrics.jsonl ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {"epoch": 1, "train_loss": 0.10506527318134698, "acc": 0.8101659751037344, "f1": 0.767175572519084, "precision": 0.6294363256784968, "recall": 0.9820846905537459, "selection_score": 0.7186459505770489, "score": 0.7186459505770489}
2
+ {"epoch": 2, "train_loss": 0.037222835986506865, "acc": 0.8739626556016598, "f1": 0.8311327310632384, "precision": 0.7248484848484849, "recall": 0.9739413680781759, "selection_score": 0.8076225068219838, "score": 0.8076225068219838}
3
+ {"epoch": 3, "train_loss": 0.022125127010995397, "acc": 0.9190871369294605, "f1": 0.8763866877971473, "precision": 0.8533950617283951, "recall": 0.9006514657980456, "selection_score": 0.9083997716936347, "score": 0.9083997716936347}
4
+ {"epoch": 4, "train_loss": 0.015069841979796758, "acc": 0.8895228215767634, "f1": 0.8486140724946695, "precision": 0.7528373266078184, "recall": 0.9723127035830619, "selection_score": 0.8329072028854942, "score": 0.8329072028854942}
5
+ {"epoch": 5, "train_loss": 0.009131610291980716, "acc": 0.8926348547717843, "f1": 0.8503253796095445, "precision": 0.764629388816645, "recall": 0.9576547231270358, "selection_score": 0.8413797102182641, "score": 0.8413797102182641}
20260228_014435/analysis_bundle/predictions/final_eval_predictions.csv ADDED
The diff for this file is too large to render. See raw diff
 
20260228_014435/analysis_bundle/predictions/final_eval_predictions_evalonly.csv ADDED
The diff for this file is too large to render. See raw diff
 
20260228_014435/analysis_bundle/summaries/final_eval_summary.csv ADDED
@@ -0,0 +1,2 @@
 
 
 
1
+ dataset,rows_evaluated,acc,f1,precision,recall,always_pick_acc,always_decline_acc
2
+ data_for_deberta/dedup_split/risk_val.csv,1928,0.9190871369294605,0.8763866877971473,0.8533950617283951,0.9006514657980456,0.3184647302904564,0.6815352697095436
20260228_014435/analysis_bundle/summaries/final_eval_summary_evalonly.csv ADDED
@@ -0,0 +1,2 @@
 
 
 
1
+ run_dir,model_path,dataset,rows_raw,rows_evaluated,acc,f1,precision,recall,always_pick_acc,always_decline_acc
2
+ outputs/task1_binary_runs/20260228_014435,outputs/task1_binary_runs/20260228_014435/best_model.pt,data_for_deberta/dedup_split/risk_test.csv,2893,2893,0.908745247148289,0.8674698795180723,0.818957345971564,0.9220917822838848,0.3238852402350501,0.6761147597649498
20260228_014435/analysis_bundle/summaries/summary.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_epoch": 3,
3
+ "best_score": 0.9083997716936347,
4
+ "best_selection_score": 0.9083997716936347,
5
+ "best_val_metrics_at_selection": {
6
+ "acc": 0.9190871369294605,
7
+ "f1": 0.8763866877971473,
8
+ "precision": 0.8533950617283951,
9
+ "recall": 0.9006514657980456
10
+ },
11
+ "model_selection": {
12
+ "metric": "recall-floor-constrained-fbeta",
13
+ "beta": 0.5,
14
+ "recall_floor": 0.9,
15
+ "recall_floor_penalty": 2.0,
16
+ "acc_weight": 0.05
17
+ },
18
+ "data_provenance": {
19
+ "train_csv": {
20
+ "path": "data_for_deberta/Risk-only-pick/task1_risk_training_20260225.csv",
21
+ "sha256_16": "0b22913a14eba047",
22
+ "size_bytes": 61730922,
23
+ "dataset_repo_id": "teru00801/New-Hawks-Data",
24
+ "dataset_file": "risk-only-pick/task1_risk_training_20260225.csv",
25
+ "hf_snapshot_revision": "4779fd5fec31f5f91f5a936be14e9c8f4ccd2030"
26
+ },
27
+ "val_csv_source": {
28
+ "path": "data_for_deberta/dedup_split/risk_val.csv",
29
+ "sha256_16": "75d6ac75345457d0",
30
+ "size_bytes": 6082107,
31
+ "dataset_repo_id": "teru00801/New-Hawks-Data",
32
+ "dataset_file": "dedup_split/risk_val.csv",
33
+ "hf_snapshot_revision": "4779fd5fec31f5f91f5a936be14e9c8f4ccd2030"
34
+ },
35
+ "final_eval_csv": {
36
+ "path": "data_for_deberta/dedup_split/risk_val.csv",
37
+ "sha256_16": "75d6ac75345457d0",
38
+ "size_bytes": 6082107,
39
+ "dataset_repo_id": "teru00801/New-Hawks-Data",
40
+ "dataset_file": "dedup_split/risk_val.csv",
41
+ "hf_snapshot_revision": "4779fd5fec31f5f91f5a936be14e9c8f4ccd2030"
42
+ }
43
+ },
44
+ "train_rows": 17394,
45
+ "val_rows": 1928,
46
+ "final_eval_rows": 1928,
47
+ "val_metrics": {
48
+ "acc": 0.9190871369294605,
49
+ "f1": 0.8763866877971473,
50
+ "precision": 0.8533950617283951,
51
+ "recall": 0.9006514657980456
52
+ },
53
+ "final_eval_metrics": {
54
+ "acc": 0.9190871369294605,
55
+ "f1": 0.8763866877971473,
56
+ "precision": 0.8533950617283951,
57
+ "recall": 0.9006514657980456
58
+ },
59
+ "final_eval_baseline": {
60
+ "always_pick_acc": 0.3184647302904564,
61
+ "always_decline_acc": 0.6815352697095436
62
+ }
63
+ }