Add evaluation results and output files to model card

README.md

@@ -5,18 +5,18 @@ tags:
   - model
   - evaluation
 metrics:
-  - average_precision: 0.96
-  - roc_auc: 0.94
-  - best threshold according to F1: 0.30
+  - average_precision: 0.95
+  - roc_auc: 0.93
+  - best threshold according to F1: 0.44
 ---
 
 # Binary Classification Model
 
 ## Evaluation Results
 
-**Average Precision:** 0.96  
-**ROC AUC:** 0.94
-**best threshold according to F1: 0.30
+**Average Precision:** 0.95  
+**ROC AUC:** 0.93
+**best threshold according to F1: 0.44
 
 
 ## Visualizations

checkpoint-8/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:958f354c0a4a7b17ee2f1630478630a3a890eaf792fbaaa6127973903ec18b95
+oid sha256:fced4ab76217d28bcd7de9a83e6fc619c1576a41165fd3b91280418c68ced5d1
 size 8760128

checkpoint-8/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f2774fd9f37a8d9cff1d919819d8448613233296dff71df0ef1e7413b0c44f16
+oid sha256:6f581e008d5914466c5395d23b18e45299d3ac0fb3aeef93ba811aaab6c7114b
 size 17564858

checkpoint-8/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a0ed227609d668f32c29a88e864644bbc2d719856cd6e14fe5b57623aeddb19d
+oid sha256:698379ffc8ebaa82bdd977f5c57de7891b341867f48cc2cee1320c2e7fe9fa12
 size 1064

checkpoint-8/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.7894614129167639,
+  "best_metric": 0.8425618780301097,
   "best_model_checkpoint": "/tmp/logs/binary_classification_model_v3.1.5_Junction_batch2/checkpoint-8",
   "epoch": 1.0,
   "eval_steps": 500,
@@ -10,85 +10,85 @@
   "log_history": [
     {
       "epoch": 0.13333333333333333,
-      "grad_norm": 28787.228515625,
-      "learning_rate": 0.00035714285714285714,
-      "loss": 0.3856,
+      "grad_norm": 13324.8955078125,
+      "learning_rate": 0.00014285714285714287,
+      "loss": 0.3452,
       "step": 1
     },
     {
       "epoch": 0.26666666666666666,
-      "grad_norm": 27319.6640625,
-      "learning_rate": 0.0007142857142857143,
-      "loss": 0.3913,
+      "grad_norm": 13969.9384765625,
+      "learning_rate": 0.00028571428571428574,
+      "loss": 0.3389,
       "step": 2
     },
     {
       "epoch": 0.4,
-      "grad_norm": 764010.0625,
-      "learning_rate": 0.0010714285714285715,
-      "loss": 0.657,
+      "grad_norm": 25136.65234375,
+      "learning_rate": 0.0004285714285714286,
+      "loss": 0.3301,
       "step": 3
     },
     {
       "epoch": 0.5333333333333333,
-      "grad_norm": 460677.375,
-      "learning_rate": 0.0014285714285714286,
-      "loss": 0.6085,
+      "grad_norm": 21701.81640625,
+      "learning_rate": 0.0005714285714285715,
+      "loss": 0.3312,
       "step": 4
     },
     {
       "epoch": 0.6666666666666666,
-      "grad_norm": 349255.0625,
-      "learning_rate": 0.0017857142857142859,
-      "loss": 0.6177,
+      "grad_norm": 24685.49609375,
+      "learning_rate": 0.0007142857142857143,
+      "loss": 0.323,
       "step": 5
     },
     {
       "epoch": 0.8,
-      "grad_norm": 162546.3125,
-      "learning_rate": 0.002142857142857143,
-      "loss": 0.5485,
+      "grad_norm": 45404.9921875,
+      "learning_rate": 0.0008571428571428572,
+      "loss": 0.358,
       "step": 6
     },
     {
       "epoch": 0.9333333333333333,
-      "grad_norm": 441874.4375,
-      "learning_rate": 0.0025,
-      "loss": 0.7531,
+      "grad_norm": 20938.6328125,
+      "learning_rate": 0.001,
+      "loss": 0.3457,
       "step": 7
     },
     {
       "epoch": 1.0,
-      "grad_norm": 153196.75,
-      "learning_rate": 0.002857142857142857,
-      "loss": 0.3116,
+      "grad_norm": 30029.5546875,
+      "learning_rate": 0.001142857142857143,
+      "loss": 0.1635,
       "step": 8
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7538838920686836,
-      "eval_best_threshold": 0.5175232887268066,
-      "eval_f1": 0.7894614129167639,
-      "eval_loss": 0.5186505913734436,
-      "eval_pr_auc": 0.8433328368486523,
-      "eval_precision": 0.7086647132691503,
-      "eval_recall": 0.8910526315789473,
-      "eval_roc_auc": 0.8432376305376215,
-      "eval_runtime": 2.5983,
-      "eval_samples_per_second": 1412.087,
-      "eval_steps_per_second": 3.079,
+      "eval_accuracy": 0.8318342872717361,
+      "eval_best_threshold": 0.3127627372741699,
+      "eval_f1": 0.8425618780301097,
+      "eval_loss": 0.368960976600647,
+      "eval_pr_auc": 0.9260693646178895,
+      "eval_precision": 0.8177315502724121,
+      "eval_recall": 0.8689473684210526,
+      "eval_roc_auc": 0.9189140459968462,
+      "eval_runtime": 2.7482,
+      "eval_samples_per_second": 1335.042,
+      "eval_steps_per_second": 2.911,
       "step": 8
     }
   ],
   "logging_steps": 1,
-  "max_steps": 1400,
+  "max_steps": 3500,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 200,
+  "num_train_epochs": 500,
   "save_steps": 500,
   "stateful_callbacks": {
     "EarlyStoppingCallback": {
       "args": {
-        "early_stopping_patience": 20,
+        "early_stopping_patience": 50,
         "early_stopping_threshold": 0.0
       },
       "attributes": {

checkpoint-8/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:99c19006956dac46c4f9af44fd85d92055df85ce6d52f69565d2637981ffd87f
-size 5496
+oid sha256:4056f3d0beca21dab5fc55a93dc339d6de9a32168cdb0039f8dffef443e6b27f
+size 5432

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eb87fedae2e556c6ed432f3f6304b5fbcb6f36bfcf1a1b86ab7742ebf8c70899
+oid sha256:5c12e90cdacbeb9a95697e94c9c5cd32d029cf14517dd16c5b49332d0eac0c61
 size 8760128

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:533942e4c51378a0ec0b57c4b2c984691303160f66d5ab43e15a964dc09f965e
+oid sha256:4056f3d0beca21dab5fc55a93dc339d6de9a32168cdb0039f8dffef443e6b27f
 size 5432