lgsilvaesilva commited on May 26

Commit

1efc71c

verified ·

1 Parent(s): f2e02af

Upload folder using huggingface_hub

Browse files

Files changed (27) hide show

README.md +31 -12
REPORT.md +28 -9
report.json +606 -54
transformer/checkpoint-1220/model.safetensors +1 -1
transformer/checkpoint-1220/optimizer.pt +1 -1
transformer/checkpoint-1220/scaler.pt +1 -1
transformer/checkpoint-1220/trainer_state.json +136 -136
transformer/checkpoint-1525/model.safetensors +1 -1
transformer/checkpoint-1525/optimizer.pt +1 -1
transformer/checkpoint-1525/scaler.pt +1 -1
transformer/checkpoint-1525/trainer_state.json +171 -171
transformer/checkpoint-305/model.safetensors +1 -1
transformer/checkpoint-305/optimizer.pt +1 -1
transformer/checkpoint-305/scaler.pt +1 -1
transformer/checkpoint-305/trainer_state.json +34 -34
transformer/checkpoint-610/model.safetensors +1 -1
transformer/checkpoint-610/optimizer.pt +1 -1
transformer/checkpoint-610/scaler.pt +1 -1
transformer/checkpoint-610/trainer_state.json +67 -67
transformer/checkpoint-915/model.safetensors +1 -1
transformer/checkpoint-915/optimizer.pt +1 -1
transformer/checkpoint-915/scaler.pt +1 -1
transformer/checkpoint-915/trainer_state.json +100 -100
transformer/config.json +5 -5
transformer/model.safetensors +1 -1
transformer/test_predictions.csv +0 -0
transformer/validation_predictions.csv +0 -0

README.md CHANGED Viewed

@@ -20,7 +20,7 @@ It includes the Transformer model, any configured TF-IDF or sentence-embedding b
 - Text column: `chunk_text`
 - Label column: `label`
 - Transformer: `FacebookAI/xlm-roberta-base`
-- Generated at: `2026-05-25T19:23:29.605062+00:00`
 ## Dataset Summary
@@ -30,6 +30,25 @@ It includes the Transformer model, any configured TF-IDF or sentence-embedding b
 | validation | 978 | 899 | 79 | 538 | 690.6 |
 | test | 1016 | 904 | 112 | 539 | 690.7 |
 ## Threshold Comparison on Test Split
 | Model | Threshold | Accuracy | Precision | Recall | F1 | ROC AUC | Average precision |
@@ -44,8 +63,8 @@ It includes the Transformer model, any configured TF-IDF or sentence-embedding b
 | embedding-svm_sentence_embeddings | 0.276 | 0.935 | 0.667 | 0.821 | 0.736 | 0.952 | 0.743 |
 | embedding-lightgbm_sentence_embeddings | 0.500 | 0.946 | 0.788 | 0.696 | 0.739 | 0.959 | 0.801 |
 | embedding-lightgbm_sentence_embeddings | 0.052 | 0.933 | 0.657 | 0.821 | 0.730 | 0.959 | 0.801 |
-| transformer | 0.500 | 0.950 | 0.748 | 0.821 | 0.783 | 0.951 | 0.785 |
-| transformer | 0.616 | 0.950 | 0.748 | 0.821 | 0.783 | 0.951 | 0.785 |
 ## Confusion Matrices on Test Split
@@ -125,15 +144,15 @@ Rows are true labels and columns are predicted labels.
 | True / Predicted | NOT_RELEVANT | RELEVANT |
 | --- | ---: | ---: |
-| NOT_RELEVANT | 873 | 31 |
-| RELEVANT | 20 | 92 |
-### transformer at threshold 0.616
 | True / Predicted | NOT_RELEVANT | RELEVANT |
 | --- | ---: | ---: |
-| NOT_RELEVANT | 873 | 31 |
-| RELEVANT | 20 | 92 |
 ## Validation-Tuned Thresholds
@@ -143,7 +162,7 @@ Rows are true labels and columns are predicted labels.
 - `embedding-logistic_sentence_embeddings`: threshold `0.616` (validation F1 `0.728`); test F1 change vs 0.5: `+0.074`.
 - `embedding-svm_sentence_embeddings`: threshold `0.276` (validation F1 `0.731`); test F1 change vs 0.5: `+0.024`.
 - `embedding-lightgbm_sentence_embeddings`: threshold `0.052` (validation F1 `0.739`); test F1 change vs 0.5: `-0.009`.
-- `transformer`: threshold `0.616` (validation F1 `0.807`); test F1 change vs 0.5: `+0.000`.
 ## Artifacts
@@ -168,7 +187,7 @@ pip install transformers torch huggingface_hub pandas joblib scikit-learn xgboos
 import torch
 from transformers import AutoModelForSequenceClassification, AutoTokenizer
-MODEL_ID = "faodl/agri-utilization-classifier"
 texts = [
     "Rice export prices increased after new procurement rules were announced.",
@@ -205,7 +224,7 @@ import json
 import joblib
 from huggingface_hub import hf_hub_download
-MODEL_ID = "faodl/agri-utilization-classifier"
 BASELINE = "logistic"
 texts = [
@@ -249,7 +268,7 @@ import joblib
 from huggingface_hub import hf_hub_download
 from sentence_transformers import SentenceTransformer
-MODEL_ID = "faodl/agri-utilization-classifier"
 BASELINE = "embedding-logistic"
 texts = [

 - Text column: `chunk_text`
 - Label column: `label`
 - Transformer: `FacebookAI/xlm-roberta-base`
+- Generated at: `2026-05-26T12:10:12.870005+00:00`
 ## Dataset Summary
 | validation | 978 | 899 | 79 | 538 | 690.6 |
 | test | 1016 | 904 | 112 | 539 | 690.7 |
+## Threshold Comparison on Validation Split
+Validation metrics document threshold selection and tuning behavior; test metrics remain the primary estimate of out-of-sample performance.
+| Model | Threshold | Accuracy | Precision | Recall | F1 | ROC AUC | Average precision |
+| --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
+| logistic_tfidf | 0.500 | 0.912 | 0.465 | 0.582 | 0.517 | 0.872 | 0.594 |
+| logistic_tfidf | 0.608 | 0.942 | 0.696 | 0.494 | 0.578 | 0.872 | 0.594 |
+| xgboost_tfidf | 0.500 | 0.945 | 0.931 | 0.342 | 0.500 | 0.823 | 0.588 |
+| xgboost_tfidf | 0.177 | 0.934 | 0.592 | 0.570 | 0.581 | 0.823 | 0.588 |
+| embedding-logistic_sentence_embeddings | 0.500 | 0.916 | 0.490 | 0.911 | 0.637 | 0.956 | 0.749 |
+| embedding-logistic_sentence_embeddings | 0.616 | 0.946 | 0.612 | 0.899 | 0.728 | 0.956 | 0.749 |
+| embedding-svm_sentence_embeddings | 0.500 | 0.957 | 0.803 | 0.620 | 0.700 | 0.958 | 0.743 |
+| embedding-svm_sentence_embeddings | 0.276 | 0.952 | 0.667 | 0.810 | 0.731 | 0.958 | 0.743 |
+| embedding-lightgbm_sentence_embeddings | 0.500 | 0.948 | 0.700 | 0.620 | 0.658 | 0.952 | 0.778 |
+| embedding-lightgbm_sentence_embeddings | 0.052 | 0.953 | 0.670 | 0.823 | 0.739 | 0.952 | 0.778 |
+| transformer | 0.500 | 0.967 | 0.776 | 0.835 | 0.805 | 0.974 | 0.874 |
+| transformer | 0.962 | 0.974 | 0.886 | 0.785 | 0.832 | 0.974 | 0.874 |
 ## Threshold Comparison on Test Split
 | Model | Threshold | Accuracy | Precision | Recall | F1 | ROC AUC | Average precision |
 | embedding-svm_sentence_embeddings | 0.276 | 0.935 | 0.667 | 0.821 | 0.736 | 0.952 | 0.743 |
 | embedding-lightgbm_sentence_embeddings | 0.500 | 0.946 | 0.788 | 0.696 | 0.739 | 0.959 | 0.801 |
 | embedding-lightgbm_sentence_embeddings | 0.052 | 0.933 | 0.657 | 0.821 | 0.730 | 0.959 | 0.801 |
+| transformer | 0.500 | 0.949 | 0.738 | 0.830 | 0.782 | 0.958 | 0.812 |
+| transformer | 0.962 | 0.948 | 0.798 | 0.705 | 0.749 | 0.958 | 0.812 |
 ## Confusion Matrices on Test Split
 | True / Predicted | NOT_RELEVANT | RELEVANT |
 | --- | ---: | ---: |
+| NOT_RELEVANT | 871 | 33 |
+| RELEVANT | 19 | 93 |
+### transformer at threshold 0.962
 | True / Predicted | NOT_RELEVANT | RELEVANT |
 | --- | ---: | ---: |
+| NOT_RELEVANT | 884 | 20 |
+| RELEVANT | 33 | 79 |
 ## Validation-Tuned Thresholds
 - `embedding-logistic_sentence_embeddings`: threshold `0.616` (validation F1 `0.728`); test F1 change vs 0.5: `+0.074`.
 - `embedding-svm_sentence_embeddings`: threshold `0.276` (validation F1 `0.731`); test F1 change vs 0.5: `+0.024`.
 - `embedding-lightgbm_sentence_embeddings`: threshold `0.052` (validation F1 `0.739`); test F1 change vs 0.5: `-0.009`.
+- `transformer`: threshold `0.962` (validation F1 `0.832`); test F1 change vs 0.5: `-0.033`.
 ## Artifacts
 import torch
 from transformers import AutoModelForSequenceClassification, AutoTokenizer
+MODEL_ID = "YOUR_USERNAME/YOUR_MODEL_REPO"
 texts = [
     "Rice export prices increased after new procurement rules were announced.",
 import joblib
 from huggingface_hub import hf_hub_download
+MODEL_ID = "YOUR_USERNAME/YOUR_MODEL_REPO"
 BASELINE = "logistic"
 texts = [
 from huggingface_hub import hf_hub_download
 from sentence_transformers import SentenceTransformer
+MODEL_ID = "YOUR_USERNAME/YOUR_MODEL_REPO"
 BASELINE = "embedding-logistic"
 texts = [

REPORT.md CHANGED Viewed

@@ -5,7 +5,7 @@
 - Text column: `chunk_text`
 - Label column: `label`
 - Transformer: `FacebookAI/xlm-roberta-base`
-- Generated at: `2026-05-25T19:23:29.605062+00:00`
 ## Dataset Summary
@@ -15,6 +15,25 @@
 | validation | 978 | 899 | 79 | 538 | 690.6 |
 | test | 1016 | 904 | 112 | 539 | 690.7 |
 ## Threshold Comparison on Test Split
 | Model | Threshold | Accuracy | Precision | Recall | F1 | ROC AUC | Average precision |
@@ -29,8 +48,8 @@
 | embedding-svm_sentence_embeddings | 0.276 | 0.935 | 0.667 | 0.821 | 0.736 | 0.952 | 0.743 |
 | embedding-lightgbm_sentence_embeddings | 0.500 | 0.946 | 0.788 | 0.696 | 0.739 | 0.959 | 0.801 |
 | embedding-lightgbm_sentence_embeddings | 0.052 | 0.933 | 0.657 | 0.821 | 0.730 | 0.959 | 0.801 |
-| transformer | 0.500 | 0.950 | 0.748 | 0.821 | 0.783 | 0.951 | 0.785 |
-| transformer | 0.616 | 0.950 | 0.748 | 0.821 | 0.783 | 0.951 | 0.785 |
 ## Confusion Matrices on Test Split
@@ -110,15 +129,15 @@ Rows are true labels and columns are predicted labels.
 | True / Predicted | NOT_RELEVANT | RELEVANT |
 | --- | ---: | ---: |
-| NOT_RELEVANT | 873 | 31 |
-| RELEVANT | 20 | 92 |
-### transformer at threshold 0.616
 | True / Predicted | NOT_RELEVANT | RELEVANT |
 | --- | ---: | ---: |
-| NOT_RELEVANT | 873 | 31 |
-| RELEVANT | 20 | 92 |
 ## Validation-Tuned Thresholds
@@ -128,7 +147,7 @@ Rows are true labels and columns are predicted labels.
 - `embedding-logistic_sentence_embeddings`: threshold `0.616` (validation F1 `0.728`); test F1 change vs 0.5: `+0.074`.
 - `embedding-svm_sentence_embeddings`: threshold `0.276` (validation F1 `0.731`); test F1 change vs 0.5: `+0.024`.
 - `embedding-lightgbm_sentence_embeddings`: threshold `0.052` (validation F1 `0.739`); test F1 change vs 0.5: `-0.009`.
-- `transformer`: threshold `0.616` (validation F1 `0.807`); test F1 change vs 0.5: `+0.000`.
 ## Artifacts

 - Text column: `chunk_text`
 - Label column: `label`
 - Transformer: `FacebookAI/xlm-roberta-base`
+- Generated at: `2026-05-26T12:10:12.870005+00:00`
 ## Dataset Summary
 | validation | 978 | 899 | 79 | 538 | 690.6 |
 | test | 1016 | 904 | 112 | 539 | 690.7 |
+## Threshold Comparison on Validation Split
+Validation metrics document threshold selection and tuning behavior; test metrics remain the primary estimate of out-of-sample performance.
+| Model | Threshold | Accuracy | Precision | Recall | F1 | ROC AUC | Average precision |
+| --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
+| logistic_tfidf | 0.500 | 0.912 | 0.465 | 0.582 | 0.517 | 0.872 | 0.594 |
+| logistic_tfidf | 0.608 | 0.942 | 0.696 | 0.494 | 0.578 | 0.872 | 0.594 |
+| xgboost_tfidf | 0.500 | 0.945 | 0.931 | 0.342 | 0.500 | 0.823 | 0.588 |
+| xgboost_tfidf | 0.177 | 0.934 | 0.592 | 0.570 | 0.581 | 0.823 | 0.588 |
+| embedding-logistic_sentence_embeddings | 0.500 | 0.916 | 0.490 | 0.911 | 0.637 | 0.956 | 0.749 |
+| embedding-logistic_sentence_embeddings | 0.616 | 0.946 | 0.612 | 0.899 | 0.728 | 0.956 | 0.749 |
+| embedding-svm_sentence_embeddings | 0.500 | 0.957 | 0.803 | 0.620 | 0.700 | 0.958 | 0.743 |
+| embedding-svm_sentence_embeddings | 0.276 | 0.952 | 0.667 | 0.810 | 0.731 | 0.958 | 0.743 |
+| embedding-lightgbm_sentence_embeddings | 0.500 | 0.948 | 0.700 | 0.620 | 0.658 | 0.952 | 0.778 |
+| embedding-lightgbm_sentence_embeddings | 0.052 | 0.953 | 0.670 | 0.823 | 0.739 | 0.952 | 0.778 |
+| transformer | 0.500 | 0.967 | 0.776 | 0.835 | 0.805 | 0.974 | 0.874 |
+| transformer | 0.962 | 0.974 | 0.886 | 0.785 | 0.832 | 0.974 | 0.874 |
 ## Threshold Comparison on Test Split
 | Model | Threshold | Accuracy | Precision | Recall | F1 | ROC AUC | Average precision |
 | embedding-svm_sentence_embeddings | 0.276 | 0.935 | 0.667 | 0.821 | 0.736 | 0.952 | 0.743 |
 | embedding-lightgbm_sentence_embeddings | 0.500 | 0.946 | 0.788 | 0.696 | 0.739 | 0.959 | 0.801 |
 | embedding-lightgbm_sentence_embeddings | 0.052 | 0.933 | 0.657 | 0.821 | 0.730 | 0.959 | 0.801 |
+| transformer | 0.500 | 0.949 | 0.738 | 0.830 | 0.782 | 0.958 | 0.812 |
+| transformer | 0.962 | 0.948 | 0.798 | 0.705 | 0.749 | 0.958 | 0.812 |
 ## Confusion Matrices on Test Split
 | True / Predicted | NOT_RELEVANT | RELEVANT |
 | --- | ---: | ---: |
+| NOT_RELEVANT | 871 | 33 |
+| RELEVANT | 19 | 93 |
+### transformer at threshold 0.962
 | True / Predicted | NOT_RELEVANT | RELEVANT |
 | --- | ---: | ---: |
+| NOT_RELEVANT | 884 | 20 |
+| RELEVANT | 33 | 79 |
 ## Validation-Tuned Thresholds
 - `embedding-logistic_sentence_embeddings`: threshold `0.616` (validation F1 `0.728`); test F1 change vs 0.5: `+0.074`.
 - `embedding-svm_sentence_embeddings`: threshold `0.276` (validation F1 `0.731`); test F1 change vs 0.5: `+0.024`.
 - `embedding-lightgbm_sentence_embeddings`: threshold `0.052` (validation F1 `0.739`); test F1 change vs 0.5: `-0.009`.
+- `transformer`: threshold `0.962` (validation F1 `0.832`); test F1 change vs 0.5: `-0.033`.
 ## Artifacts

report.json CHANGED Viewed

@@ -1,5 +1,5 @@
 {
-  "created_at": "2026-05-25T19:23:29.605062+00:00",
   "config": {
     "hf_dataset": "faodl/amis-agri-utilization",
     "hf_subset": null,
@@ -37,8 +37,8 @@
     "embedding_batch_size": 64,
     "positive_label_name": "RELEVANT",
     "negative_label_name": "NOT_RELEVANT",
-    "push_to_hub": true,
-    "hub_model_id": "faodl/agri-utilization-classifier",
     "hub_private_repo": false
   },
   "dataset_summary": {
@@ -85,6 +85,98 @@
         "precision": 0.6964285714285714,
         "recall": 0.4936708860759494
       },
       "test_default_0_5": {
         "threshold": 0.5,
         "accuracy": 0.9261811023622047,
@@ -189,6 +281,98 @@
         "precision": 0.5921052631578947,
         "recall": 0.569620253164557
       },
       "test_default_0_5": {
         "threshold": 0.5,
         "accuracy": 0.9242125984251969,
@@ -294,6 +478,98 @@
         "precision": 0.6120689655172413,
         "recall": 0.8987341772151899
       },
       "test_default_0_5": {
         "threshold": 0.5,
         "accuracy": 0.8986220472440944,
@@ -399,6 +675,98 @@
         "precision": 0.6666666666666666,
         "recall": 0.810126582278481
       },
       "test_default_0_5": {
         "threshold": 0.5,
         "accuracy": 0.9409448818897638,
@@ -504,6 +872,98 @@
         "precision": 0.6701030927835051,
         "recall": 0.8227848101265823
       },
       "test_default_0_5": {
         "threshold": 0.5,
         "accuracy": 0.9458661417322834,
@@ -602,102 +1062,194 @@
       "model_name": "FacebookAI/xlm-roberta-base",
       "artifact_dir": "/content/agri-utilization-classifier/transformer",
       "validation_best_threshold": {
-        "threshold": 0.6156440377235413,
-        "f1": 0.8072289156626505,
-        "precision": 0.7701149425287356,
-        "recall": 0.8481012658227848
       },
       "test_default_0_5": {
         "threshold": 0.5,
-        "accuracy": 0.9498031496062992,
-        "precision": 0.7479674796747967,
-        "recall": 0.8214285714285714,
-        "f1": 0.7829787234042553,
         "confusion_matrix": [
           [
-            873,
-            31
           ],
           [
-            20,
-            92
           ]
         ],
         "classification_report": {
           "NOT_RELEVANT": {
-            "precision": 0.9776035834266518,
-            "recall": 0.9657079646017699,
-            "f1-score": 0.9716193656093489,
             "support": 904.0
           },
           "RELEVANT": {
-            "precision": 0.7479674796747967,
-            "recall": 0.8214285714285714,
-            "f1-score": 0.7829787234042553,
             "support": 112.0
           },
-          "accuracy": 0.9498031496062992,
           "macro avg": {
-            "precision": 0.8627855315507242,
-            "recall": 0.8935682680151706,
-            "f1-score": 0.8772990445068021,
             "support": 1016.0
           },
           "weighted avg": {
-            "precision": 0.9522893672650299,
-            "recall": 0.9498031496062992,
-            "f1-score": 0.9508243341851654,
             "support": 1016.0
           }
         },
-        "roc_auc": 0.9511694058154235,
-        "average_precision": 0.7846734208461954
       },
       "test_optimal_threshold": {
-        "threshold": 0.6156440377235413,
-        "accuracy": 0.9498031496062992,
-        "precision": 0.7479674796747967,
-        "recall": 0.8214285714285714,
-        "f1": 0.7829787234042553,
         "confusion_matrix": [
           [
-            873,
-            31
           ],
           [
-            20,
-            92
           ]
         ],
         "classification_report": {
           "NOT_RELEVANT": {
-            "precision": 0.9776035834266518,
-            "recall": 0.9657079646017699,
-            "f1-score": 0.9716193656093489,
             "support": 904.0
           },
           "RELEVANT": {
-            "precision": 0.7479674796747967,
-            "recall": 0.8214285714285714,
-            "f1-score": 0.7829787234042553,
             "support": 112.0
           },
-          "accuracy": 0.9498031496062992,
           "macro avg": {
-            "precision": 0.8627855315507242,
-            "recall": 0.8935682680151706,
-            "f1-score": 0.8772990445068021,
             "support": 1016.0
           },
           "weighted avg": {
-            "precision": 0.9522893672650299,
-            "recall": 0.9498031496062992,
-            "f1-score": 0.9508243341851654,
             "support": 1016.0
           }
         },
-        "roc_auc": 0.9511694058154235,
-        "average_precision": 0.7846734208461954
       }
     }
   ]

 {
+  "created_at": "2026-05-26T12:10:12.870005+00:00",
   "config": {
     "hf_dataset": "faodl/amis-agri-utilization",
     "hf_subset": null,
     "embedding_batch_size": 64,
     "positive_label_name": "RELEVANT",
     "negative_label_name": "NOT_RELEVANT",
+    "push_to_hub": false,
+    "hub_model_id": null,
     "hub_private_repo": false
   },
   "dataset_summary": {
         "precision": 0.6964285714285714,
         "recall": 0.4936708860759494
       },
+      "validation_default_0_5": {
+        "threshold": 0.5,
+        "accuracy": 0.9120654396728016,
+        "precision": 0.46464646464646464,
+        "recall": 0.5822784810126582,
+        "f1": 0.5168539325842697,
+        "confusion_matrix": [
+          [
+            846,
+            53
+          ],
+          [
+            33,
+            46
+          ]
+        ],
+        "classification_report": {
+          "NOT_RELEVANT": {
+            "precision": 0.962457337883959,
+            "recall": 0.9410456062291435,
+            "f1-score": 0.9516310461192351,
+            "support": 899.0
+          },
+          "RELEVANT": {
+            "precision": 0.46464646464646464,
+            "recall": 0.5822784810126582,
+            "f1-score": 0.5168539325842697,
+            "support": 79.0
+          },
+          "accuracy": 0.9120654396728016,
+          "macro avg": {
+            "precision": 0.7135519012652118,
+            "recall": 0.7616620436209008,
+            "f1-score": 0.7342424893517524,
+            "support": 978.0
+          },
+          "weighted avg": {
+            "precision": 0.9222456211296011,
+            "recall": 0.9120654396728016,
+            "f1-score": 0.9165110134308279,
+            "support": 978.0
+          }
+        },
+        "roc_auc": 0.871530955632841,
+        "average_precision": 0.5935308473185881
+      },
+      "validation_optimal_threshold": {
+        "threshold": 0.6076606929552563,
+        "accuracy": 0.941717791411043,
+        "precision": 0.6964285714285714,
+        "recall": 0.4936708860759494,
+        "f1": 0.5777777777777777,
+        "confusion_matrix": [
+          [
+            882,
+            17
+          ],
+          [
+            40,
+            39
+          ]
+        ],
+        "classification_report": {
+          "NOT_RELEVANT": {
+            "precision": 0.9566160520607375,
+            "recall": 0.9810901001112347,
+            "f1-score": 0.9686985172981878,
+            "support": 899.0
+          },
+          "RELEVANT": {
+            "precision": 0.6964285714285714,
+            "recall": 0.4936708860759494,
+            "f1-score": 0.5777777777777777,
+            "support": 79.0
+          },
+          "accuracy": 0.941717791411043,
+          "macro avg": {
+            "precision": 0.8265223117446545,
+            "recall": 0.737380493093592,
+            "f1-score": 0.7732381475379828,
+            "support": 978.0
+          },
+          "weighted avg": {
+            "precision": 0.9355988629299183,
+            "recall": 0.941717791411043,
+            "f1-score": 0.9371210751487886,
+            "support": 978.0
+          }
+        },
+        "roc_auc": 0.871530955632841,
+        "average_precision": 0.5935308473185881
+      },
       "test_default_0_5": {
         "threshold": 0.5,
         "accuracy": 0.9261811023622047,
         "precision": 0.5921052631578947,
         "recall": 0.569620253164557
       },
+      "validation_default_0_5": {
+        "threshold": 0.5,
+        "accuracy": 0.9447852760736196,
+        "precision": 0.9310344827586207,
+        "recall": 0.34177215189873417,
+        "f1": 0.5,
+        "confusion_matrix": [
+          [
+            897,
+            2
+          ],
+          [
+            52,
+            27
+          ]
+        ],
+        "classification_report": {
+          "NOT_RELEVANT": {
+            "precision": 0.9452054794520548,
+            "recall": 0.9977753058954394,
+            "f1-score": 0.9707792207792207,
+            "support": 899.0
+          },
+          "RELEVANT": {
+            "precision": 0.9310344827586207,
+            "recall": 0.34177215189873417,
+            "f1-score": 0.5,
+            "support": 79.0
+          },
+          "accuracy": 0.9447852760736196,
+          "macro avg": {
+            "precision": 0.9381199811053378,
+            "recall": 0.6697737288970868,
+            "f1-score": 0.7353896103896104,
+            "support": 978.0
+          },
+          "weighted avg": {
+            "precision": 0.9440607874901108,
+            "recall": 0.9447852760736196,
+            "f1-score": 0.9327510424136191,
+            "support": 978.0
+          }
+        },
+        "roc_auc": 0.822629926359809,
+        "average_precision": 0.5882293042162409
+      },
+      "validation_optimal_threshold": {
+        "threshold": 0.17728303372859955,
+        "accuracy": 0.9335378323108384,
+        "precision": 0.5921052631578947,
+        "recall": 0.569620253164557,
+        "f1": 0.5806451612903226,
+        "confusion_matrix": [
+          [
+            868,
+            31
+          ],
+          [
+            34,
+            45
+          ]
+        ],
+        "classification_report": {
+          "NOT_RELEVANT": {
+            "precision": 0.9623059866962306,
+            "recall": 0.9655172413793104,
+            "f1-score": 0.9639089394780678,
+            "support": 899.0
+          },
+          "RELEVANT": {
+            "precision": 0.5921052631578947,
+            "recall": 0.569620253164557,
+            "f1-score": 0.5806451612903226,
+            "support": 79.0
+          },
+          "accuracy": 0.9335378323108384,
+          "macro avg": {
+            "precision": 0.7772056249270627,
+            "recall": 0.7675687472719337,
+            "f1-score": 0.7722770503841951,
+            "support": 978.0
+          },
+          "weighted avg": {
+            "precision": 0.9324022472693098,
+            "recall": 0.9335378323108384,
+            "f1-score": 0.9329500044301825,
+            "support": 978.0
+          }
+        },
+        "roc_auc": 0.822629926359809,
+        "average_precision": 0.5882293042162409
+      },
       "test_default_0_5": {
         "threshold": 0.5,
         "accuracy": 0.9242125984251969,
         "precision": 0.6120689655172413,
         "recall": 0.8987341772151899
       },
+      "validation_default_0_5": {
+        "threshold": 0.5,
+        "accuracy": 0.9161554192229039,
+        "precision": 0.4897959183673469,
+        "recall": 0.9113924050632911,
+        "f1": 0.6371681415929203,
+        "confusion_matrix": [
+          [
+            824,
+            75
+          ],
+          [
+            7,
+            72
+          ]
+        ],
+        "classification_report": {
+          "NOT_RELEVANT": {
+            "precision": 0.9915764139590855,
+            "recall": 0.9165739710789766,
+            "f1-score": 0.9526011560693641,
+            "support": 899.0
+          },
+          "RELEVANT": {
+            "precision": 0.4897959183673469,
+            "recall": 0.9113924050632911,
+            "f1-score": 0.6371681415929203,
+            "support": 79.0
+          },
+          "accuracy": 0.9161554192229039,
+          "macro avg": {
+            "precision": 0.7406861661632163,
+            "recall": 0.9139831880711339,
+            "f1-score": 0.7948846488311423,
+            "support": 978.0
+          },
+          "weighted avg": {
+            "precision": 0.9510440426382804,
+            "recall": 0.9161554192229039,
+            "f1-score": 0.9271213931413078,
+            "support": 978.0
+          }
+        },
+        "roc_auc": 0.9563227777699554,
+        "average_precision": 0.7488532716951917
+      },
+      "validation_optimal_threshold": {
+        "threshold": 0.616087721531811,
+        "accuracy": 0.9458077709611452,
+        "precision": 0.6120689655172413,
+        "recall": 0.8987341772151899,
+        "f1": 0.7282051282051282,
+        "confusion_matrix": [
+          [
+            854,
+            45
+          ],
+          [
+            8,
+            71
+          ]
+        ],
+        "classification_report": {
+          "NOT_RELEVANT": {
+            "precision": 0.9907192575406032,
+            "recall": 0.949944382647386,
+            "f1-score": 0.9699034639409426,
+            "support": 899.0
+          },
+          "RELEVANT": {
+            "precision": 0.6120689655172413,
+            "recall": 0.8987341772151899,
+            "f1-score": 0.7282051282051282,
+            "support": 79.0
+          },
+          "accuracy": 0.9458077709611452,
+          "macro avg": {
+            "precision": 0.8013941115289223,
+            "recall": 0.924339279931288,
+            "f1-score": 0.8490542960730354,
+            "support": 978.0
+          },
+          "weighted avg": {
+            "precision": 0.9601329865080412,
+            "recall": 0.9458077709611452,
+            "f1-score": 0.9503797742444914,
+            "support": 978.0
+          }
+        },
+        "roc_auc": 0.9563227777699554,
+        "average_precision": 0.7488532716951917
+      },
       "test_default_0_5": {
         "threshold": 0.5,
         "accuracy": 0.8986220472440944,
         "precision": 0.6666666666666666,
         "recall": 0.810126582278481
       },
+      "validation_default_0_5": {
+        "threshold": 0.5,
+        "accuracy": 0.9570552147239264,
+        "precision": 0.8032786885245902,
+        "recall": 0.620253164556962,
+        "f1": 0.7,
+        "confusion_matrix": [
+          [
+            887,
+            12
+          ],
+          [
+            30,
+            49
+          ]
+        ],
+        "classification_report": {
+          "NOT_RELEVANT": {
+            "precision": 0.9672846237731734,
+            "recall": 0.9866518353726362,
+            "f1-score": 0.9768722466960352,
+            "support": 899.0
+          },
+          "RELEVANT": {
+            "precision": 0.8032786885245902,
+            "recall": 0.620253164556962,
+            "f1-score": 0.7,
+            "support": 79.0
+          },
+          "accuracy": 0.9570552147239264,
+          "macro avg": {
+            "precision": 0.8852816561488818,
+            "recall": 0.8034524999647992,
+            "f1-score": 0.8384361233480175,
+            "support": 978.0
+          },
+          "weighted avg": {
+            "precision": 0.9540367005782469,
+            "recall": 0.9570552147239264,
+            "f1-score": 0.9545073106132266,
+            "support": 978.0
+          }
+        },
+        "roc_auc": 0.9584911505047804,
+        "average_precision": 0.7425325495012566
+      },
+      "validation_optimal_threshold": {
+        "threshold": 0.27629376276966117,
+        "accuracy": 0.9519427402862985,
+        "precision": 0.6666666666666666,
+        "recall": 0.810126582278481,
+        "f1": 0.7314285714285714,
+        "confusion_matrix": [
+          [
+            867,
+            32
+          ],
+          [
+            15,
+            64
+          ]
+        ],
+        "classification_report": {
+          "NOT_RELEVANT": {
+            "precision": 0.9829931972789115,
+            "recall": 0.96440489432703,
+            "f1-score": 0.9736103312745649,
+            "support": 899.0
+          },
+          "RELEVANT": {
+            "precision": 0.6666666666666666,
+            "recall": 0.810126582278481,
+            "f1-score": 0.7314285714285714,
+            "support": 79.0
+          },
+          "accuracy": 0.9519427402862985,
+          "macro avg": {
+            "precision": 0.8248299319727891,
+            "recall": 0.8872657383027556,
+            "f1-score": 0.8525194513515681,
+            "support": 978.0
+          },
+          "weighted avg": {
+            "precision": 0.9574412587120738,
+            "recall": 0.9519427402862985,
+            "f1-score": 0.9540475919823016,
+            "support": 978.0
+          }
+        },
+        "roc_auc": 0.9584911505047804,
+        "average_precision": 0.7425325495012566
+      },
       "test_default_0_5": {
         "threshold": 0.5,
         "accuracy": 0.9409448818897638,
         "precision": 0.6701030927835051,
         "recall": 0.8227848101265823
       },
+      "validation_default_0_5": {
+        "threshold": 0.5,
+        "accuracy": 0.9478527607361963,
+        "precision": 0.7,
+        "recall": 0.620253164556962,
+        "f1": 0.6577181208053692,
+        "confusion_matrix": [
+          [
+            878,
+            21
+          ],
+          [
+            30,
+            49
+          ]
+        ],
+        "classification_report": {
+          "NOT_RELEVANT": {
+            "precision": 0.9669603524229075,
+            "recall": 0.9766407119021134,
+            "f1-score": 0.9717764250138351,
+            "support": 899.0
+          },
+          "RELEVANT": {
+            "precision": 0.7,
+            "recall": 0.620253164556962,
+            "f1-score": 0.6577181208053692,
+            "support": 79.0
+          },
+          "accuracy": 0.9478527607361963,
+          "macro avg": {
+            "precision": 0.8334801762114536,
+            "recall": 0.7984469382295377,
+            "f1-score": 0.8147472729096021,
+            "support": 978.0
+          },
+          "weighted avg": {
+            "precision": 0.9453960703764762,
+            "recall": 0.9478527607361963,
+            "f1-score": 0.9464077071892248,
+            "support": 978.0
+          }
+        },
+        "roc_auc": 0.952112755382211,
+        "average_precision": 0.777786126005225
+      },
+      "validation_optimal_threshold": {
+        "threshold": 0.05244099185733503,
+        "accuracy": 0.9529652351738241,
+        "precision": 0.6701030927835051,
+        "recall": 0.8227848101265823,
+        "f1": 0.7386363636363636,
+        "confusion_matrix": [
+          [
+            867,
+            32
+          ],
+          [
+            14,
+            65
+          ]
+        ],
+        "classification_report": {
+          "NOT_RELEVANT": {
+            "precision": 0.9841089670828603,
+            "recall": 0.96440489432703,
+            "f1-score": 0.9741573033707865,
+            "support": 899.0
+          },
+          "RELEVANT": {
+            "precision": 0.6701030927835051,
+            "recall": 0.8227848101265823,
+            "f1-score": 0.7386363636363636,
+            "support": 79.0
+          },
+          "accuracy": 0.9529652351738241,
+          "macro avg": {
+            "precision": 0.8271060299331827,
+            "recall": 0.8935948522268062,
+            "f1-score": 0.8563968335035751,
+            "support": 978.0
+          },
+          "weighted avg": {
+            "precision": 0.9587444843940576,
+            "recall": 0.9529652351738241,
+            "f1-score": 0.9551326057848771,
+            "support": 978.0
+          }
+        },
+        "roc_auc": 0.952112755382211,
+        "average_precision": 0.777786126005225
+      },
       "test_default_0_5": {
         "threshold": 0.5,
         "accuracy": 0.9458661417322834,
       "model_name": "FacebookAI/xlm-roberta-base",
       "artifact_dir": "/content/agri-utilization-classifier/transformer",
       "validation_best_threshold": {
+        "threshold": 0.961713969707489,
+        "f1": 0.8322147651006712,
+        "precision": 0.8857142857142857,
+        "recall": 0.7848101265822784
+      },
+      "validation_default_0_5": {
+        "threshold": 0.5,
+        "accuracy": 0.967280163599182,
+        "precision": 0.7764705882352941,
+        "recall": 0.8354430379746836,
+        "f1": 0.8048780487804879,
+        "confusion_matrix": [
+          [
+            880,
+            19
+          ],
+          [
+            13,
+            66
+          ]
+        ],
+        "classification_report": {
+          "NOT_RELEVANT": {
+            "precision": 0.9854423292273237,
+            "recall": 0.978865406006674,
+            "f1-score": 0.9821428571428571,
+            "support": 899.0
+          },
+          "RELEVANT": {
+            "precision": 0.7764705882352941,
+            "recall": 0.8354430379746836,
+            "f1-score": 0.8048780487804879,
+            "support": 79.0
+          },
+          "accuracy": 0.967280163599182,
+          "macro avg": {
+            "precision": 0.8809564587313089,
+            "recall": 0.9071542219906787,
+            "f1-score": 0.8935104529616724,
+            "support": 978.0
+          },
+          "weighted avg": {
+            "precision": 0.9685621988199921,
+            "recall": 0.967280163599182,
+            "f1-score": 0.9678239206800482,
+            "support": 978.0
+          }
+        },
+        "roc_auc": 0.9735078357105644,
+        "average_precision": 0.8740430936338875
+      },
+      "validation_optimal_threshold": {
+        "threshold": 0.961713969707489,
+        "accuracy": 0.9744376278118609,
+        "precision": 0.8857142857142857,
+        "recall": 0.7848101265822784,
+        "f1": 0.8322147651006712,
+        "confusion_matrix": [
+          [
+            891,
+            8
+          ],
+          [
+            17,
+            62
+          ]
+        ],
+        "classification_report": {
+          "NOT_RELEVANT": {
+            "precision": 0.9812775330396476,
+            "recall": 0.9911012235817576,
+            "f1-score": 0.9861649142224682,
+            "support": 899.0
+          },
+          "RELEVANT": {
+            "precision": 0.8857142857142857,
+            "recall": 0.7848101265822784,
+            "f1-score": 0.8322147651006712,
+            "support": 79.0
+          },
+          "accuracy": 0.9744376278118609,
+          "macro avg": {
+            "precision": 0.9334959093769666,
+            "recall": 0.887955675082018,
+            "f1-score": 0.9091898396615696,
+            "support": 978.0
+          },
+          "weighted avg": {
+            "precision": 0.9735582114254312,
+            "recall": 0.9744376278118609,
+            "f1-score": 0.9737292682300123,
+            "support": 978.0
+          }
+        },
+        "roc_auc": 0.9735078357105644,
+        "average_precision": 0.8740430936338875
       },
       "test_default_0_5": {
         "threshold": 0.5,
+        "accuracy": 0.9488188976377953,
+        "precision": 0.7380952380952381,
+        "recall": 0.8303571428571429,
+        "f1": 0.7815126050420168,
         "confusion_matrix": [
           [
+            871,
+            33
           ],
           [
+            19,
+            93
           ]
         ],
         "classification_report": {
           "NOT_RELEVANT": {
+            "precision": 0.9786516853932584,
+            "recall": 0.963495575221239,
+            "f1-score": 0.9710144927536232,
             "support": 904.0
           },
           "RELEVANT": {
+            "precision": 0.7380952380952381,
+            "recall": 0.8303571428571429,
+            "f1-score": 0.7815126050420168,
             "support": 112.0
           },
+          "accuracy": 0.9488188976377953,
           "macro avg": {
+            "precision": 0.8583734617442482,
+            "recall": 0.8969263590391909,
+            "f1-score": 0.87626354889782,
             "support": 1016.0
           },
           "weighted avg": {
+            "precision": 0.9521336518328467,
+            "recall": 0.9488188976377953,
+            "f1-score": 0.9501245208799027,
             "support": 1016.0
           }
         },
+        "roc_auc": 0.9575794089759799,
+        "average_precision": 0.8120836792897285
       },
       "test_optimal_threshold": {
+        "threshold": 0.961713969707489,
+        "accuracy": 0.9478346456692913,
+        "precision": 0.797979797979798,
+        "recall": 0.7053571428571429,
+        "f1": 0.7488151658767772,
         "confusion_matrix": [
           [
+            884,
+            20
           ],
           [
+            33,
+            79
           ]
         ],
         "classification_report": {
           "NOT_RELEVANT": {
+            "precision": 0.9640130861504908,
+            "recall": 0.9778761061946902,
+            "f1-score": 0.970895112575508,
             "support": 904.0
           },
           "RELEVANT": {
+            "precision": 0.797979797979798,
+            "recall": 0.7053571428571429,
+            "f1-score": 0.7488151658767772,
             "support": 112.0
           },
+          "accuracy": 0.9478346456692913,
           "macro avg": {
+            "precision": 0.8809964420651444,
+            "recall": 0.8416166245259166,
+            "f1-score": 0.8598551392261427,
             "support": 1016.0
           },
           "weighted avg": {
+            "precision": 0.945710203989942,
+            "recall": 0.9478346456692913,
+            "f1-score": 0.9464138586087188,
             "support": 1016.0
           }
         },
+        "roc_auc": 0.9575794089759799,
+        "average_precision": 0.8120836792897285
       }
     }
   ]

transformer/checkpoint-1220/model.safetensors CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:23260da79af693d20fe8414504a730a2e1fae0128c96e784d5a388b934f65ef8
 size 1112205008

 version https://git-lfs.github.com/spec/v1
+oid sha256:10fb0ab01e1be47d3c6e92b6cda92cf113dcdb8bb660ad8c297c6dffcdb53187
 size 1112205008

transformer/checkpoint-1220/optimizer.pt CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f5656e89640a42e9a162840684d7e143689ced913e28cbc8709b4e3fcc33ea94
 size 2224532875

 version https://git-lfs.github.com/spec/v1
+oid sha256:68014421c88c3dc4aff836ffde24d87d96bcdaf3fec73670a1a8aa36229ab905
 size 2224532875

transformer/checkpoint-1220/scaler.pt CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7ad54f7b6fbdb8393459d5a595aa5dfdf4cf4c483f044be07288464f573b4d8e
 size 1383

 version https://git-lfs.github.com/spec/v1
+oid sha256:bb2d7099cfd24be7d85605d300cde0de65ab0a9c5c1ef4722d0de46d4564d89a
 size 1383

transformer/checkpoint-1220/trainer_state.json CHANGED Viewed

@@ -1,7 +1,7 @@
 {
-  "best_global_step": 1220,
-  "best_metric": 0.8,
-  "best_model_checkpoint": "/content/agri-utilization-classifier/transformer/checkpoint-1220",
   "epoch": 4.0,
   "eval_steps": 500,
   "global_step": 1220,
@@ -11,390 +11,390 @@
   "log_history": [
     {
       "epoch": 0.08196721311475409,
-      "grad_norm": 6.984184741973877,
       "learning_rate": 3.157894736842105e-06,
-      "loss": 0.7167730712890625,
       "step": 25
     },
     {
       "epoch": 0.16393442622950818,
-      "grad_norm": 9.77598762512207,
       "learning_rate": 6.447368421052632e-06,
-      "loss": 0.5636273193359375,
       "step": 50
     },
     {
       "epoch": 0.2459016393442623,
-      "grad_norm": 8.441609382629395,
       "learning_rate": 9.736842105263159e-06,
-      "loss": 0.37406421661376954,
       "step": 75
     },
     {
       "epoch": 0.32786885245901637,
-      "grad_norm": 5.313694953918457,
       "learning_rate": 1.3026315789473684e-05,
-      "loss": 0.2965927886962891,
       "step": 100
     },
     {
       "epoch": 0.4098360655737705,
-      "grad_norm": 7.240467548370361,
       "learning_rate": 1.6315789473684213e-05,
-      "loss": 0.29742313385009767,
       "step": 125
     },
     {
       "epoch": 0.4918032786885246,
-      "grad_norm": 6.100603103637695,
       "learning_rate": 1.960526315789474e-05,
-      "loss": 0.2068590545654297,
       "step": 150
     },
     {
       "epoch": 0.5737704918032787,
-      "grad_norm": 4.169040679931641,
       "learning_rate": 1.9679533867443555e-05,
-      "loss": 0.21712726593017578,
       "step": 175
     },
     {
       "epoch": 0.6557377049180327,
-      "grad_norm": 4.665876865386963,
       "learning_rate": 1.9315367807720323e-05,
-      "loss": 0.2889243125915527,
       "step": 200
     },
     {
       "epoch": 0.7377049180327869,
-      "grad_norm": 1.094870924949646,
       "learning_rate": 1.8951201747997088e-05,
-      "loss": 0.2303921699523926,
       "step": 225
     },
     {
       "epoch": 0.819672131147541,
-      "grad_norm": 1.2164329290390015,
       "learning_rate": 1.8587035688273852e-05,
-      "loss": 0.16723501205444335,
       "step": 250
     },
     {
       "epoch": 0.9016393442622951,
-      "grad_norm": 2.6314468383789062,
       "learning_rate": 1.822286962855062e-05,
-      "loss": 0.15685997009277344,
       "step": 275
     },
     {
       "epoch": 0.9836065573770492,
-      "grad_norm": 14.927972793579102,
       "learning_rate": 1.7858703568827385e-05,
-      "loss": 0.1979808807373047,
       "step": 300
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.9642126789366053,
-      "eval_f1": 0.7852760736196319,
-      "eval_loss": 0.10679091513156891,
-      "eval_precision": 0.7619047619047619,
-      "eval_recall": 0.810126582278481,
-      "eval_roc_auc": 0.9606665634108222,
-      "eval_runtime": 3.9998,
-      "eval_samples_per_second": 244.514,
-      "eval_steps_per_second": 7.75,
       "step": 305
     },
     {
       "epoch": 1.0655737704918034,
-      "grad_norm": 22.521757125854492,
       "learning_rate": 1.7494537509104153e-05,
-      "loss": 0.1316550636291504,
       "step": 325
     },
     {
       "epoch": 1.1475409836065573,
-      "grad_norm": 0.22687062621116638,
       "learning_rate": 1.7130371449380918e-05,
-      "loss": 0.2059168815612793,
       "step": 350
     },
     {
       "epoch": 1.2295081967213115,
-      "grad_norm": 0.657261312007904,
       "learning_rate": 1.6766205389657686e-05,
-      "loss": 0.09930330276489258,
       "step": 375
     },
     {
       "epoch": 1.3114754098360657,
-      "grad_norm": 4.796896457672119,
       "learning_rate": 1.640203932993445e-05,
-      "loss": 0.13251757621765137,
       "step": 400
     },
     {
       "epoch": 1.3934426229508197,
-      "grad_norm": 0.5394258499145508,
       "learning_rate": 1.603787327021122e-05,
-      "loss": 0.17834033966064453,
       "step": 425
     },
     {
       "epoch": 1.4754098360655736,
-      "grad_norm": 0.4655781388282776,
       "learning_rate": 1.5673707210487983e-05,
-      "loss": 0.122637300491333,
       "step": 450
     },
     {
       "epoch": 1.5573770491803278,
-      "grad_norm": 0.42849695682525635,
       "learning_rate": 1.530954115076475e-05,
-      "loss": 0.15983641624450684,
       "step": 475
     },
     {
       "epoch": 1.639344262295082,
-      "grad_norm": 0.45386794209480286,
       "learning_rate": 1.4945375091041516e-05,
-      "loss": 0.14264726638793945,
       "step": 500
     },
     {
       "epoch": 1.721311475409836,
-      "grad_norm": 0.6170782446861267,
       "learning_rate": 1.4581209031318282e-05,
-      "loss": 0.18886091232299804,
       "step": 525
     },
     {
       "epoch": 1.8032786885245902,
-      "grad_norm": 5.4546122550964355,
       "learning_rate": 1.4217042971595047e-05,
-      "loss": 0.14393989562988282,
       "step": 550
     },
     {
       "epoch": 1.8852459016393444,
-      "grad_norm": 17.419189453125,
       "learning_rate": 1.3852876911871815e-05,
-      "loss": 0.10272212982177735,
       "step": 575
     },
     {
       "epoch": 1.9672131147540983,
-      "grad_norm": 1.6267497539520264,
       "learning_rate": 1.3488710852148582e-05,
-      "loss": 0.1810975456237793,
       "step": 600
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.9580777096114519,
-      "eval_f1": 0.7759562841530054,
-      "eval_loss": 0.17126062512397766,
-      "eval_precision": 0.6826923076923077,
-      "eval_recall": 0.8987341772151899,
-      "eval_roc_auc": 0.9631939848777121,
-      "eval_runtime": 3.84,
-      "eval_samples_per_second": 254.687,
-      "eval_steps_per_second": 8.073,
       "step": 610
     },
     {
       "epoch": 2.0491803278688523,
-      "grad_norm": 1.3180276155471802,
       "learning_rate": 1.3124544792425346e-05,
-      "loss": 0.05921304225921631,
       "step": 625
     },
     {
       "epoch": 2.1311475409836067,
-      "grad_norm": 5.675038814544678,
       "learning_rate": 1.2760378732702113e-05,
-      "loss": 0.16824769973754883,
       "step": 650
     },
     {
       "epoch": 2.2131147540983607,
-      "grad_norm": 0.16993092000484467,
       "learning_rate": 1.239621267297888e-05,
-      "loss": 0.12186273574829101,
       "step": 675
     },
     {
       "epoch": 2.2950819672131146,
-      "grad_norm": 1.1791695356369019,
       "learning_rate": 1.2032046613255645e-05,
-      "loss": 0.08795836448669433,
       "step": 700
     },
     {
       "epoch": 2.3770491803278686,
-      "grad_norm": 0.07541065663099289,
       "learning_rate": 1.1667880553532412e-05,
-      "loss": 0.10129087448120117,
       "step": 725
     },
     {
       "epoch": 2.459016393442623,
-      "grad_norm": 3.394912004470825,
       "learning_rate": 1.1303714493809176e-05,
-      "loss": 0.14056243896484374,
       "step": 750
     },
     {
       "epoch": 2.540983606557377,
-      "grad_norm": 8.074258804321289,
       "learning_rate": 1.0939548434085944e-05,
-      "loss": 0.06563093185424805,
       "step": 775
     },
     {
       "epoch": 2.6229508196721314,
-      "grad_norm": 12.472029685974121,
       "learning_rate": 1.057538237436271e-05,
-      "loss": 0.09851057052612305,
       "step": 800
     },
     {
       "epoch": 2.7049180327868854,
-      "grad_norm": 0.10368915647268295,
       "learning_rate": 1.0211216314639475e-05,
-      "loss": 0.11658324241638184,
       "step": 825
     },
     {
       "epoch": 2.7868852459016393,
-      "grad_norm": 44.263092041015625,
       "learning_rate": 9.847050254916243e-06,
-      "loss": 0.13634946823120117,
       "step": 850
     },
     {
       "epoch": 2.8688524590163933,
-      "grad_norm": 0.07709958404302597,
       "learning_rate": 9.482884195193008e-06,
-      "loss": 0.12144805908203125,
       "step": 875
     },
     {
       "epoch": 2.9508196721311473,
-      "grad_norm": 0.11255892366170883,
       "learning_rate": 9.118718135469774e-06,
-      "loss": 0.11815821647644043,
       "step": 900
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.9601226993865031,
-      "eval_f1": 0.7868852459016393,
-      "eval_loss": 0.14711864292621613,
-      "eval_precision": 0.6923076923076923,
-      "eval_recall": 0.9113924050632911,
-      "eval_roc_auc": 0.9719449177003984,
-      "eval_runtime": 3.9043,
-      "eval_samples_per_second": 250.495,
-      "eval_steps_per_second": 7.94,
       "step": 915
     },
     {
       "epoch": 3.0327868852459017,
-      "grad_norm": 0.2808685302734375,
       "learning_rate": 8.754552075746541e-06,
-      "loss": 0.1051255989074707,
       "step": 925
     },
     {
       "epoch": 3.1147540983606556,
-      "grad_norm": 0.07298991084098816,
       "learning_rate": 8.390386016023307e-06,
-      "loss": 0.08817357063293457,
       "step": 950
     },
     {
       "epoch": 3.19672131147541,
-      "grad_norm": 0.049921419471502304,
       "learning_rate": 8.026219956300074e-06,
-      "loss": 0.11110530853271484,
       "step": 975
     },
     {
       "epoch": 3.278688524590164,
-      "grad_norm": 1.874350905418396,
       "learning_rate": 7.66205389657684e-06,
-      "loss": 0.09003183364868164,
       "step": 1000
     },
     {
       "epoch": 3.360655737704918,
-      "grad_norm": 0.09576287865638733,
       "learning_rate": 7.2978878368536055e-06,
-      "loss": 0.05897871017456055,
       "step": 1025
     },
     {
       "epoch": 3.442622950819672,
-      "grad_norm": 20.84284019470215,
       "learning_rate": 6.933721777130372e-06,
-      "loss": 0.06021720886230469,
       "step": 1050
     },
     {
       "epoch": 3.5245901639344264,
-      "grad_norm": 0.06452233344316483,
       "learning_rate": 6.569555717407138e-06,
-      "loss": 0.06818977355957032,
       "step": 1075
     },
     {
       "epoch": 3.6065573770491803,
-      "grad_norm": 0.2655308246612549,
       "learning_rate": 6.2053896576839045e-06,
-      "loss": 0.07051475524902344,
       "step": 1100
     },
     {
       "epoch": 3.6885245901639343,
-      "grad_norm": 0.05852988734841347,
       "learning_rate": 5.84122359796067e-06,
-      "loss": 0.08089996337890625,
       "step": 1125
     },
     {
       "epoch": 3.7704918032786887,
-      "grad_norm": 13.798267364501953,
       "learning_rate": 5.477057538237437e-06,
-      "loss": 0.0827936840057373,
       "step": 1150
     },
     {
       "epoch": 3.8524590163934427,
-      "grad_norm": 6.363399982452393,
       "learning_rate": 5.112891478514203e-06,
-      "loss": 0.06787658214569092,
       "step": 1175
     },
     {
       "epoch": 3.9344262295081966,
-      "grad_norm": 0.059008605778217316,
       "learning_rate": 4.748725418790969e-06,
-      "loss": 0.05120136260986328,
       "step": 1200
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.9662576687116564,
-      "eval_f1": 0.8,
-      "eval_loss": 0.1406129151582718,
-      "eval_precision": 0.7674418604651163,
-      "eval_recall": 0.8354430379746836,
-      "eval_roc_auc": 0.9636797566916827,
-      "eval_runtime": 3.9045,
-      "eval_samples_per_second": 250.478,
-      "eval_steps_per_second": 7.939,
       "step": 1220
     }
   ],
@@ -410,7 +410,7 @@
         "early_stopping_threshold": 0.0
       },
       "attributes": {
-        "early_stopping_patience_counter": 0
       }
     },
     "TrainerControl": {

 {
+  "best_global_step": 915,
+  "best_metric": 0.8048780487804879,
+  "best_model_checkpoint": "/content/agri-utilization-classifier/transformer/checkpoint-915",
   "epoch": 4.0,
   "eval_steps": 500,
   "global_step": 1220,
   "log_history": [
     {
       "epoch": 0.08196721311475409,
+      "grad_norm": 4.966984272003174,
       "learning_rate": 3.157894736842105e-06,
+      "loss": 0.570748291015625,
       "step": 25
     },
     {
       "epoch": 0.16393442622950818,
+      "grad_norm": 14.035679817199707,
       "learning_rate": 6.447368421052632e-06,
+      "loss": 0.4261627197265625,
       "step": 50
     },
     {
       "epoch": 0.2459016393442623,
+      "grad_norm": 6.1731109619140625,
       "learning_rate": 9.736842105263159e-06,
+      "loss": 0.35031925201416014,
       "step": 75
     },
     {
       "epoch": 0.32786885245901637,
+      "grad_norm": 3.6704306602478027,
       "learning_rate": 1.3026315789473684e-05,
+      "loss": 0.32047801971435547,
       "step": 100
     },
     {
       "epoch": 0.4098360655737705,
+      "grad_norm": 20.360071182250977,
       "learning_rate": 1.6315789473684213e-05,
+      "loss": 0.21136175155639647,
       "step": 125
     },
     {
       "epoch": 0.4918032786885246,
+      "grad_norm": 2.045907497406006,
       "learning_rate": 1.960526315789474e-05,
+      "loss": 0.21009126663208008,
       "step": 150
     },
     {
       "epoch": 0.5737704918032787,
+      "grad_norm": 7.755379676818848,
       "learning_rate": 1.9679533867443555e-05,
+      "loss": 0.19892181396484376,
       "step": 175
     },
     {
       "epoch": 0.6557377049180327,
+      "grad_norm": 0.6007249355316162,
       "learning_rate": 1.9315367807720323e-05,
+      "loss": 0.19585962295532228,
       "step": 200
     },
     {
       "epoch": 0.7377049180327869,
+      "grad_norm": 0.7606620192527771,
       "learning_rate": 1.8951201747997088e-05,
+      "loss": 0.282861213684082,
       "step": 225
     },
     {
       "epoch": 0.819672131147541,
+      "grad_norm": 5.813822269439697,
       "learning_rate": 1.8587035688273852e-05,
+      "loss": 0.16557788848876953,
       "step": 250
     },
     {
       "epoch": 0.9016393442622951,
+      "grad_norm": 3.329322099685669,
       "learning_rate": 1.822286962855062e-05,
+      "loss": 0.1400923728942871,
       "step": 275
     },
     {
       "epoch": 0.9836065573770492,
+      "grad_norm": 10.887845993041992,
       "learning_rate": 1.7858703568827385e-05,
+      "loss": 0.1979439926147461,
       "step": 300
     },
     {
       "epoch": 1.0,
+      "eval_accuracy": 0.9693251533742331,
+      "eval_f1": 0.7916666666666666,
+      "eval_loss": 0.1259259283542633,
+      "eval_precision": 0.8769230769230769,
+      "eval_recall": 0.7215189873417721,
+      "eval_roc_auc": 0.9533659058588304,
+      "eval_runtime": 3.8981,
+      "eval_samples_per_second": 250.889,
+      "eval_steps_per_second": 7.953,
       "step": 305
     },
     {
       "epoch": 1.0655737704918034,
+      "grad_norm": 4.172677516937256,
       "learning_rate": 1.7494537509104153e-05,
+      "loss": 0.13631089210510253,
       "step": 325
     },
     {
       "epoch": 1.1475409836065573,
+      "grad_norm": 0.07315228134393692,
       "learning_rate": 1.7130371449380918e-05,
+      "loss": 0.15998102188110352,
       "step": 350
     },
     {
       "epoch": 1.2295081967213115,
+      "grad_norm": 0.4799639582633972,
       "learning_rate": 1.6766205389657686e-05,
+      "loss": 0.10631587982177734,
       "step": 375
     },
     {
       "epoch": 1.3114754098360657,
+      "grad_norm": 1.1866878271102905,
       "learning_rate": 1.640203932993445e-05,
+      "loss": 0.11795808792114258,
       "step": 400
     },
     {
       "epoch": 1.3934426229508197,
+      "grad_norm": 0.6632266044616699,
       "learning_rate": 1.603787327021122e-05,
+      "loss": 0.15535160064697265,
       "step": 425
     },
     {
       "epoch": 1.4754098360655736,
+      "grad_norm": 7.679214000701904,
       "learning_rate": 1.5673707210487983e-05,
+      "loss": 0.13739371299743652,
       "step": 450
     },
     {
       "epoch": 1.5573770491803278,
+      "grad_norm": 0.5809745192527771,
       "learning_rate": 1.530954115076475e-05,
+      "loss": 0.15218215942382812,
       "step": 475
     },
     {
       "epoch": 1.639344262295082,
+      "grad_norm": 0.14357534050941467,
       "learning_rate": 1.4945375091041516e-05,
+      "loss": 0.11159341812133788,
       "step": 500
     },
     {
       "epoch": 1.721311475409836,
+      "grad_norm": 0.15218983590602875,
       "learning_rate": 1.4581209031318282e-05,
+      "loss": 0.22156574249267577,
       "step": 525
     },
     {
       "epoch": 1.8032786885245902,
+      "grad_norm": 9.121078491210938,
       "learning_rate": 1.4217042971595047e-05,
+      "loss": 0.15829898834228515,
       "step": 550
     },
     {
       "epoch": 1.8852459016393444,
+      "grad_norm": 11.138046264648438,
       "learning_rate": 1.3852876911871815e-05,
+      "loss": 0.10829511642456055,
       "step": 575
     },
     {
       "epoch": 1.9672131147540983,
+      "grad_norm": 5.579981803894043,
       "learning_rate": 1.3488710852148582e-05,
+      "loss": 0.12963885307312012,
       "step": 600
     },
     {
       "epoch": 2.0,
+      "eval_accuracy": 0.9631901840490797,
+      "eval_f1": 0.7906976744186046,
+      "eval_loss": 0.12336025387048721,
+      "eval_precision": 0.7311827956989247,
+      "eval_recall": 0.8607594936708861,
+      "eval_roc_auc": 0.968502273975303,
+      "eval_runtime": 3.941,
+      "eval_samples_per_second": 248.162,
+      "eval_steps_per_second": 7.866,
       "step": 610
     },
     {
       "epoch": 2.0491803278688523,
+      "grad_norm": 23.79497718811035,
       "learning_rate": 1.3124544792425346e-05,
+      "loss": 0.07728540420532226,
       "step": 625
     },
     {
       "epoch": 2.1311475409836067,
+      "grad_norm": 0.3032953143119812,
       "learning_rate": 1.2760378732702113e-05,
+      "loss": 0.15112388610839844,
       "step": 650
     },
     {
       "epoch": 2.2131147540983607,
+      "grad_norm": 0.15015727281570435,
       "learning_rate": 1.239621267297888e-05,
+      "loss": 0.11781220436096192,
       "step": 675
     },
     {
       "epoch": 2.2950819672131146,
+      "grad_norm": 35.07762145996094,
       "learning_rate": 1.2032046613255645e-05,
+      "loss": 0.06247422218322754,
       "step": 700
     },
     {
       "epoch": 2.3770491803278686,
+      "grad_norm": 0.06300532817840576,
       "learning_rate": 1.1667880553532412e-05,
+      "loss": 0.07382740020751953,
       "step": 725
     },
     {
       "epoch": 2.459016393442623,
+      "grad_norm": 0.0386500246822834,
       "learning_rate": 1.1303714493809176e-05,
+      "loss": 0.10034572601318359,
       "step": 750
     },
     {
       "epoch": 2.540983606557377,
+      "grad_norm": 38.45136642456055,
       "learning_rate": 1.0939548434085944e-05,
+      "loss": 0.0807158374786377,
       "step": 775
     },
     {
       "epoch": 2.6229508196721314,
+      "grad_norm": 73.75131225585938,
       "learning_rate": 1.057538237436271e-05,
+      "loss": 0.11720505714416504,
       "step": 800
     },
     {
       "epoch": 2.7049180327868854,
+      "grad_norm": 3.530329704284668,
       "learning_rate": 1.0211216314639475e-05,
+      "loss": 0.07547494888305664,
       "step": 825
     },
     {
       "epoch": 2.7868852459016393,
+      "grad_norm": 2.0287606716156006,
       "learning_rate": 9.847050254916243e-06,
+      "loss": 0.07913532733917236,
       "step": 850
     },
     {
       "epoch": 2.8688524590163933,
+      "grad_norm": 0.07940487563610077,
       "learning_rate": 9.482884195193008e-06,
+      "loss": 0.0882901382446289,
       "step": 875
     },
     {
       "epoch": 2.9508196721311473,
+      "grad_norm": 0.03960009664297104,
       "learning_rate": 9.118718135469774e-06,
+      "loss": 0.10696781158447266,
       "step": 900
     },
     {
       "epoch": 3.0,
+      "eval_accuracy": 0.967280163599182,
+      "eval_f1": 0.8048780487804879,
+      "eval_loss": 0.13383829593658447,
+      "eval_precision": 0.7764705882352941,
+      "eval_recall": 0.8354430379746836,
+      "eval_roc_auc": 0.9735078357105644,
+      "eval_runtime": 4.0602,
+      "eval_samples_per_second": 240.875,
+      "eval_steps_per_second": 7.635,
       "step": 915
     },
     {
       "epoch": 3.0327868852459017,
+      "grad_norm": 0.08427837491035461,
       "learning_rate": 8.754552075746541e-06,
+      "loss": 0.07147216320037841,
       "step": 925
     },
     {
       "epoch": 3.1147540983606556,
+      "grad_norm": 0.027753490954637527,
       "learning_rate": 8.390386016023307e-06,
+      "loss": 0.060765981674194336,
       "step": 950
     },
     {
       "epoch": 3.19672131147541,
+      "grad_norm": 0.027190053835511208,
       "learning_rate": 8.026219956300074e-06,
+      "loss": 0.08444092750549316,
       "step": 975
     },
     {
       "epoch": 3.278688524590164,
+      "grad_norm": 6.747233867645264,
       "learning_rate": 7.66205389657684e-06,
+      "loss": 0.05561910152435303,
       "step": 1000
     },
     {
       "epoch": 3.360655737704918,
+      "grad_norm": 0.07523014396429062,
       "learning_rate": 7.2978878368536055e-06,
+      "loss": 0.05905795097351074,
       "step": 1025
     },
     {
       "epoch": 3.442622950819672,
+      "grad_norm": 8.827320098876953,
       "learning_rate": 6.933721777130372e-06,
+      "loss": 0.06066291809082031,
       "step": 1050
     },
     {
       "epoch": 3.5245901639344264,
+      "grad_norm": 0.2744128108024597,
       "learning_rate": 6.569555717407138e-06,
+      "loss": 0.07047759056091309,
       "step": 1075
     },
     {
       "epoch": 3.6065573770491803,
+      "grad_norm": 0.13911227881908417,
       "learning_rate": 6.2053896576839045e-06,
+      "loss": 0.049641780853271485,
       "step": 1100
     },
     {
       "epoch": 3.6885245901639343,
+      "grad_norm": 0.06898390501737595,
       "learning_rate": 5.84122359796067e-06,
+      "loss": 0.054807920455932614,
       "step": 1125
     },
     {
       "epoch": 3.7704918032786887,
+      "grad_norm": 37.68058395385742,
       "learning_rate": 5.477057538237437e-06,
+      "loss": 0.07445104598999024,
       "step": 1150
     },
     {
       "epoch": 3.8524590163934427,
+      "grad_norm": 0.046950697898864746,
       "learning_rate": 5.112891478514203e-06,
+      "loss": 0.03682142734527588,
       "step": 1175
     },
     {
       "epoch": 3.9344262295081966,
+      "grad_norm": 0.025340111926198006,
       "learning_rate": 4.748725418790969e-06,
+      "loss": 0.038784558773040774,
       "step": 1200
     },
     {
       "epoch": 4.0,
+      "eval_accuracy": 0.9642126789366053,
+      "eval_f1": 0.7770700636942676,
+      "eval_loss": 0.18827994167804718,
+      "eval_precision": 0.782051282051282,
+      "eval_recall": 0.7721518987341772,
+      "eval_roc_auc": 0.9623491643316765,
+      "eval_runtime": 4.0232,
+      "eval_samples_per_second": 243.092,
+      "eval_steps_per_second": 7.705,
       "step": 1220
     }
   ],
         "early_stopping_threshold": 0.0
       },
       "attributes": {
+        "early_stopping_patience_counter": 1
       }
     },
     "TrainerControl": {

transformer/checkpoint-1525/model.safetensors CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8f09e4daa3612b0a08afd0486aa5ce25a384c7e9c3abe05df4b6d1f68e5033b8
 size 1112205008

 version https://git-lfs.github.com/spec/v1
+oid sha256:c71d5f126ca24288aedac641eab5145ed59b0b4c91dceecb99619843d17d4eba
 size 1112205008

transformer/checkpoint-1525/optimizer.pt CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a0ce2fdbe5bad58a74bc9f05881c3817b46567f474f1c364e65e73b116534a13
 size 2224532875

 version https://git-lfs.github.com/spec/v1
+oid sha256:f235748cb6eb59173928bfbcaa332051d8a6fc9211b117c1725ef2e32de9a484
 size 2224532875

transformer/checkpoint-1525/scaler.pt CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cbcaa50b1076392191a1f8665b7b77d819d3cc73803b90b8509f590e4b16af02
 size 1383

 version https://git-lfs.github.com/spec/v1
+oid sha256:e30c7885b0ea805c04a948da2a40fb77ce17cb4449de4250c5d46514ee869ba0
 size 1383

transformer/checkpoint-1525/trainer_state.json CHANGED Viewed

@@ -1,7 +1,7 @@
 {
-  "best_global_step": 1525,
-  "best_metric": 0.8072289156626506,
-  "best_model_checkpoint": "/content/agri-utilization-classifier/transformer/checkpoint-1525",
   "epoch": 5.0,
   "eval_steps": 500,
   "global_step": 1525,
@@ -11,494 +11,494 @@
   "log_history": [
     {
       "epoch": 0.08196721311475409,
-      "grad_norm": 6.984184741973877,
       "learning_rate": 3.157894736842105e-06,
-      "loss": 0.7167730712890625,
       "step": 25
     },
     {
       "epoch": 0.16393442622950818,
-      "grad_norm": 9.77598762512207,
       "learning_rate": 6.447368421052632e-06,
-      "loss": 0.5636273193359375,
       "step": 50
     },
     {
       "epoch": 0.2459016393442623,
-      "grad_norm": 8.441609382629395,
       "learning_rate": 9.736842105263159e-06,
-      "loss": 0.37406421661376954,
       "step": 75
     },
     {
       "epoch": 0.32786885245901637,
-      "grad_norm": 5.313694953918457,
       "learning_rate": 1.3026315789473684e-05,
-      "loss": 0.2965927886962891,
       "step": 100
     },
     {
       "epoch": 0.4098360655737705,
-      "grad_norm": 7.240467548370361,
       "learning_rate": 1.6315789473684213e-05,
-      "loss": 0.29742313385009767,
       "step": 125
     },
     {
       "epoch": 0.4918032786885246,
-      "grad_norm": 6.100603103637695,
       "learning_rate": 1.960526315789474e-05,
-      "loss": 0.2068590545654297,
       "step": 150
     },
     {
       "epoch": 0.5737704918032787,
-      "grad_norm": 4.169040679931641,
       "learning_rate": 1.9679533867443555e-05,
-      "loss": 0.21712726593017578,
       "step": 175
     },
     {
       "epoch": 0.6557377049180327,
-      "grad_norm": 4.665876865386963,
       "learning_rate": 1.9315367807720323e-05,
-      "loss": 0.2889243125915527,
       "step": 200
     },
     {
       "epoch": 0.7377049180327869,
-      "grad_norm": 1.094870924949646,
       "learning_rate": 1.8951201747997088e-05,
-      "loss": 0.2303921699523926,
       "step": 225
     },
     {
       "epoch": 0.819672131147541,
-      "grad_norm": 1.2164329290390015,
       "learning_rate": 1.8587035688273852e-05,
-      "loss": 0.16723501205444335,
       "step": 250
     },
     {
       "epoch": 0.9016393442622951,
-      "grad_norm": 2.6314468383789062,
       "learning_rate": 1.822286962855062e-05,
-      "loss": 0.15685997009277344,
       "step": 275
     },
     {
       "epoch": 0.9836065573770492,
-      "grad_norm": 14.927972793579102,
       "learning_rate": 1.7858703568827385e-05,
-      "loss": 0.1979808807373047,
       "step": 300
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.9642126789366053,
-      "eval_f1": 0.7852760736196319,
-      "eval_loss": 0.10679091513156891,
-      "eval_precision": 0.7619047619047619,
-      "eval_recall": 0.810126582278481,
-      "eval_roc_auc": 0.9606665634108222,
-      "eval_runtime": 3.9998,
-      "eval_samples_per_second": 244.514,
-      "eval_steps_per_second": 7.75,
       "step": 305
     },
     {
       "epoch": 1.0655737704918034,
-      "grad_norm": 22.521757125854492,
       "learning_rate": 1.7494537509104153e-05,
-      "loss": 0.1316550636291504,
       "step": 325
     },
     {
       "epoch": 1.1475409836065573,
-      "grad_norm": 0.22687062621116638,
       "learning_rate": 1.7130371449380918e-05,
-      "loss": 0.2059168815612793,
       "step": 350
     },
     {
       "epoch": 1.2295081967213115,
-      "grad_norm": 0.657261312007904,
       "learning_rate": 1.6766205389657686e-05,
-      "loss": 0.09930330276489258,
       "step": 375
     },
     {
       "epoch": 1.3114754098360657,
-      "grad_norm": 4.796896457672119,
       "learning_rate": 1.640203932993445e-05,
-      "loss": 0.13251757621765137,
       "step": 400
     },
     {
       "epoch": 1.3934426229508197,
-      "grad_norm": 0.5394258499145508,
       "learning_rate": 1.603787327021122e-05,
-      "loss": 0.17834033966064453,
       "step": 425
     },
     {
       "epoch": 1.4754098360655736,
-      "grad_norm": 0.4655781388282776,
       "learning_rate": 1.5673707210487983e-05,
-      "loss": 0.122637300491333,
       "step": 450
     },
     {
       "epoch": 1.5573770491803278,
-      "grad_norm": 0.42849695682525635,
       "learning_rate": 1.530954115076475e-05,
-      "loss": 0.15983641624450684,
       "step": 475
     },
     {
       "epoch": 1.639344262295082,
-      "grad_norm": 0.45386794209480286,
       "learning_rate": 1.4945375091041516e-05,
-      "loss": 0.14264726638793945,
       "step": 500
     },
     {
       "epoch": 1.721311475409836,
-      "grad_norm": 0.6170782446861267,
       "learning_rate": 1.4581209031318282e-05,
-      "loss": 0.18886091232299804,
       "step": 525
     },
     {
       "epoch": 1.8032786885245902,
-      "grad_norm": 5.4546122550964355,
       "learning_rate": 1.4217042971595047e-05,
-      "loss": 0.14393989562988282,
       "step": 550
     },
     {
       "epoch": 1.8852459016393444,
-      "grad_norm": 17.419189453125,
       "learning_rate": 1.3852876911871815e-05,
-      "loss": 0.10272212982177735,
       "step": 575
     },
     {
       "epoch": 1.9672131147540983,
-      "grad_norm": 1.6267497539520264,
       "learning_rate": 1.3488710852148582e-05,
-      "loss": 0.1810975456237793,
       "step": 600
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.9580777096114519,
-      "eval_f1": 0.7759562841530054,
-      "eval_loss": 0.17126062512397766,
-      "eval_precision": 0.6826923076923077,
-      "eval_recall": 0.8987341772151899,
-      "eval_roc_auc": 0.9631939848777121,
-      "eval_runtime": 3.84,
-      "eval_samples_per_second": 254.687,
-      "eval_steps_per_second": 8.073,
       "step": 610
     },
     {
       "epoch": 2.0491803278688523,
-      "grad_norm": 1.3180276155471802,
       "learning_rate": 1.3124544792425346e-05,
-      "loss": 0.05921304225921631,
       "step": 625
     },
     {
       "epoch": 2.1311475409836067,
-      "grad_norm": 5.675038814544678,
       "learning_rate": 1.2760378732702113e-05,
-      "loss": 0.16824769973754883,
       "step": 650
     },
     {
       "epoch": 2.2131147540983607,
-      "grad_norm": 0.16993092000484467,
       "learning_rate": 1.239621267297888e-05,
-      "loss": 0.12186273574829101,
       "step": 675
     },
     {
       "epoch": 2.2950819672131146,
-      "grad_norm": 1.1791695356369019,
       "learning_rate": 1.2032046613255645e-05,
-      "loss": 0.08795836448669433,
       "step": 700
     },
     {
       "epoch": 2.3770491803278686,
-      "grad_norm": 0.07541065663099289,
       "learning_rate": 1.1667880553532412e-05,
-      "loss": 0.10129087448120117,
       "step": 725
     },
     {
       "epoch": 2.459016393442623,
-      "grad_norm": 3.394912004470825,
       "learning_rate": 1.1303714493809176e-05,
-      "loss": 0.14056243896484374,
       "step": 750
     },
     {
       "epoch": 2.540983606557377,
-      "grad_norm": 8.074258804321289,
       "learning_rate": 1.0939548434085944e-05,
-      "loss": 0.06563093185424805,
       "step": 775
     },
     {
       "epoch": 2.6229508196721314,
-      "grad_norm": 12.472029685974121,
       "learning_rate": 1.057538237436271e-05,
-      "loss": 0.09851057052612305,
       "step": 800
     },
     {
       "epoch": 2.7049180327868854,
-      "grad_norm": 0.10368915647268295,
       "learning_rate": 1.0211216314639475e-05,
-      "loss": 0.11658324241638184,
       "step": 825
     },
     {
       "epoch": 2.7868852459016393,
-      "grad_norm": 44.263092041015625,
       "learning_rate": 9.847050254916243e-06,
-      "loss": 0.13634946823120117,
       "step": 850
     },
     {
       "epoch": 2.8688524590163933,
-      "grad_norm": 0.07709958404302597,
       "learning_rate": 9.482884195193008e-06,
-      "loss": 0.12144805908203125,
       "step": 875
     },
     {
       "epoch": 2.9508196721311473,
-      "grad_norm": 0.11255892366170883,
       "learning_rate": 9.118718135469774e-06,
-      "loss": 0.11815821647644043,
       "step": 900
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.9601226993865031,
-      "eval_f1": 0.7868852459016393,
-      "eval_loss": 0.14711864292621613,
-      "eval_precision": 0.6923076923076923,
-      "eval_recall": 0.9113924050632911,
-      "eval_roc_auc": 0.9719449177003984,
-      "eval_runtime": 3.9043,
-      "eval_samples_per_second": 250.495,
-      "eval_steps_per_second": 7.94,
       "step": 915
     },
     {
       "epoch": 3.0327868852459017,
-      "grad_norm": 0.2808685302734375,
       "learning_rate": 8.754552075746541e-06,
-      "loss": 0.1051255989074707,
       "step": 925
     },
     {
       "epoch": 3.1147540983606556,
-      "grad_norm": 0.07298991084098816,
       "learning_rate": 8.390386016023307e-06,
-      "loss": 0.08817357063293457,
       "step": 950
     },
     {
       "epoch": 3.19672131147541,
-      "grad_norm": 0.049921419471502304,
       "learning_rate": 8.026219956300074e-06,
-      "loss": 0.11110530853271484,
       "step": 975
     },
     {
       "epoch": 3.278688524590164,
-      "grad_norm": 1.874350905418396,
       "learning_rate": 7.66205389657684e-06,
-      "loss": 0.09003183364868164,
       "step": 1000
     },
     {
       "epoch": 3.360655737704918,
-      "grad_norm": 0.09576287865638733,
       "learning_rate": 7.2978878368536055e-06,
-      "loss": 0.05897871017456055,
       "step": 1025
     },
     {
       "epoch": 3.442622950819672,
-      "grad_norm": 20.84284019470215,
       "learning_rate": 6.933721777130372e-06,
-      "loss": 0.06021720886230469,
       "step": 1050
     },
     {
       "epoch": 3.5245901639344264,
-      "grad_norm": 0.06452233344316483,
       "learning_rate": 6.569555717407138e-06,
-      "loss": 0.06818977355957032,
       "step": 1075
     },
     {
       "epoch": 3.6065573770491803,
-      "grad_norm": 0.2655308246612549,
       "learning_rate": 6.2053896576839045e-06,
-      "loss": 0.07051475524902344,
       "step": 1100
     },
     {
       "epoch": 3.6885245901639343,
-      "grad_norm": 0.05852988734841347,
       "learning_rate": 5.84122359796067e-06,
-      "loss": 0.08089996337890625,
       "step": 1125
     },
     {
       "epoch": 3.7704918032786887,
-      "grad_norm": 13.798267364501953,
       "learning_rate": 5.477057538237437e-06,
-      "loss": 0.0827936840057373,
       "step": 1150
     },
     {
       "epoch": 3.8524590163934427,
-      "grad_norm": 6.363399982452393,
       "learning_rate": 5.112891478514203e-06,
-      "loss": 0.06787658214569092,
       "step": 1175
     },
     {
       "epoch": 3.9344262295081966,
-      "grad_norm": 0.059008605778217316,
       "learning_rate": 4.748725418790969e-06,
-      "loss": 0.05120136260986328,
       "step": 1200
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.9662576687116564,
-      "eval_f1": 0.8,
-      "eval_loss": 0.1406129151582718,
-      "eval_precision": 0.7674418604651163,
-      "eval_recall": 0.8354430379746836,
-      "eval_roc_auc": 0.9636797566916827,
-      "eval_runtime": 3.9045,
-      "eval_samples_per_second": 250.478,
-      "eval_steps_per_second": 7.939,
       "step": 1220
     },
     {
       "epoch": 4.016393442622951,
-      "grad_norm": 7.579391002655029,
       "learning_rate": 4.3845593590677355e-06,
-      "loss": 0.07703603267669677,
       "step": 1225
     },
     {
       "epoch": 4.098360655737705,
-      "grad_norm": 2.102841377258301,
       "learning_rate": 4.020393299344502e-06,
-      "loss": 0.032845423221588135,
       "step": 1250
     },
     {
       "epoch": 4.180327868852459,
-      "grad_norm": 0.07024268805980682,
       "learning_rate": 3.656227239621268e-06,
-      "loss": 0.03449820995330811,
       "step": 1275
     },
     {
       "epoch": 4.262295081967213,
-      "grad_norm": 0.02977728098630905,
       "learning_rate": 3.292061179898034e-06,
-      "loss": 0.06648642539978028,
       "step": 1300
     },
     {
       "epoch": 4.344262295081967,
-      "grad_norm": 0.02086547203361988,
       "learning_rate": 2.9278951201748e-06,
-      "loss": 0.06277695178985596,
       "step": 1325
     },
     {
       "epoch": 4.426229508196721,
-      "grad_norm": 0.15774419903755188,
       "learning_rate": 2.5637290604515665e-06,
-      "loss": 0.02380265951156616,
       "step": 1350
     },
     {
       "epoch": 4.508196721311475,
-      "grad_norm": 0.030775833874940872,
       "learning_rate": 2.1995630007283324e-06,
-      "loss": 0.06929955959320068,
       "step": 1375
     },
     {
       "epoch": 4.590163934426229,
-      "grad_norm": 0.07542883604764938,
       "learning_rate": 1.8353969410050983e-06,
-      "loss": 0.027865142822265626,
       "step": 1400
     },
     {
       "epoch": 4.672131147540983,
-      "grad_norm": 6.874780178070068,
       "learning_rate": 1.4712308812818645e-06,
-      "loss": 0.0548116397857666,
       "step": 1425
     },
     {
       "epoch": 4.754098360655737,
-      "grad_norm": 37.22102355957031,
       "learning_rate": 1.1070648215586309e-06,
-      "loss": 0.04357499122619629,
       "step": 1450
     },
     {
       "epoch": 4.836065573770492,
-      "grad_norm": 0.03843735158443451,
       "learning_rate": 7.428987618353969e-07,
-      "loss": 0.03130460500717163,
       "step": 1475
     },
     {
       "epoch": 4.918032786885246,
-      "grad_norm": 0.028637070208787918,
       "learning_rate": 3.787327021121632e-07,
-      "loss": 0.02157698631286621,
       "step": 1500
     },
     {
       "epoch": 5.0,
-      "grad_norm": 0.021514365449547768,
       "learning_rate": 1.4566642388929353e-08,
-      "loss": 0.04632264614105225,
       "step": 1525
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.967280163599182,
-      "eval_f1": 0.8072289156626506,
-      "eval_loss": 0.15767407417297363,
-      "eval_precision": 0.7701149425287356,
-      "eval_recall": 0.8481012658227848,
-      "eval_roc_auc": 0.9619549147435266,
-      "eval_runtime": 3.8427,
-      "eval_samples_per_second": 254.511,
-      "eval_steps_per_second": 8.067,
       "step": 1525
     }
   ],
@@ -514,7 +514,7 @@
         "early_stopping_threshold": 0.0
       },
       "attributes": {
-        "early_stopping_patience_counter": 0
       }
     },
     "TrainerControl": {

 {
+  "best_global_step": 915,
+  "best_metric": 0.8048780487804879,
+  "best_model_checkpoint": "/content/agri-utilization-classifier/transformer/checkpoint-915",
   "epoch": 5.0,
   "eval_steps": 500,
   "global_step": 1525,
   "log_history": [
     {
       "epoch": 0.08196721311475409,
+      "grad_norm": 4.966984272003174,
       "learning_rate": 3.157894736842105e-06,
+      "loss": 0.570748291015625,
       "step": 25
     },
     {
       "epoch": 0.16393442622950818,
+      "grad_norm": 14.035679817199707,
       "learning_rate": 6.447368421052632e-06,
+      "loss": 0.4261627197265625,
       "step": 50
     },
     {
       "epoch": 0.2459016393442623,
+      "grad_norm": 6.1731109619140625,
       "learning_rate": 9.736842105263159e-06,
+      "loss": 0.35031925201416014,
       "step": 75
     },
     {
       "epoch": 0.32786885245901637,
+      "grad_norm": 3.6704306602478027,
       "learning_rate": 1.3026315789473684e-05,
+      "loss": 0.32047801971435547,
       "step": 100
     },
     {
       "epoch": 0.4098360655737705,
+      "grad_norm": 20.360071182250977,
       "learning_rate": 1.6315789473684213e-05,
+      "loss": 0.21136175155639647,
       "step": 125
     },
     {
       "epoch": 0.4918032786885246,
+      "grad_norm": 2.045907497406006,
       "learning_rate": 1.960526315789474e-05,
+      "loss": 0.21009126663208008,
       "step": 150
     },
     {
       "epoch": 0.5737704918032787,
+      "grad_norm": 7.755379676818848,
       "learning_rate": 1.9679533867443555e-05,
+      "loss": 0.19892181396484376,
       "step": 175
     },
     {
       "epoch": 0.6557377049180327,
+      "grad_norm": 0.6007249355316162,
       "learning_rate": 1.9315367807720323e-05,
+      "loss": 0.19585962295532228,
       "step": 200
     },
     {
       "epoch": 0.7377049180327869,
+      "grad_norm": 0.7606620192527771,
       "learning_rate": 1.8951201747997088e-05,
+      "loss": 0.282861213684082,
       "step": 225
     },
     {
       "epoch": 0.819672131147541,
+      "grad_norm": 5.813822269439697,
       "learning_rate": 1.8587035688273852e-05,
+      "loss": 0.16557788848876953,
       "step": 250
     },
     {
       "epoch": 0.9016393442622951,
+      "grad_norm": 3.329322099685669,
       "learning_rate": 1.822286962855062e-05,
+      "loss": 0.1400923728942871,
       "step": 275
     },
     {
       "epoch": 0.9836065573770492,
+      "grad_norm": 10.887845993041992,
       "learning_rate": 1.7858703568827385e-05,
+      "loss": 0.1979439926147461,
       "step": 300
     },
     {
       "epoch": 1.0,
+      "eval_accuracy": 0.9693251533742331,
+      "eval_f1": 0.7916666666666666,
+      "eval_loss": 0.1259259283542633,
+      "eval_precision": 0.8769230769230769,
+      "eval_recall": 0.7215189873417721,
+      "eval_roc_auc": 0.9533659058588304,
+      "eval_runtime": 3.8981,
+      "eval_samples_per_second": 250.889,
+      "eval_steps_per_second": 7.953,
       "step": 305
     },
     {
       "epoch": 1.0655737704918034,
+      "grad_norm": 4.172677516937256,
       "learning_rate": 1.7494537509104153e-05,
+      "loss": 0.13631089210510253,
       "step": 325
     },
     {
       "epoch": 1.1475409836065573,
+      "grad_norm": 0.07315228134393692,
       "learning_rate": 1.7130371449380918e-05,
+      "loss": 0.15998102188110352,
       "step": 350
     },
     {
       "epoch": 1.2295081967213115,
+      "grad_norm": 0.4799639582633972,
       "learning_rate": 1.6766205389657686e-05,
+      "loss": 0.10631587982177734,
       "step": 375
     },
     {
       "epoch": 1.3114754098360657,
+      "grad_norm": 1.1866878271102905,
       "learning_rate": 1.640203932993445e-05,
+      "loss": 0.11795808792114258,
       "step": 400
     },
     {
       "epoch": 1.3934426229508197,
+      "grad_norm": 0.6632266044616699,
       "learning_rate": 1.603787327021122e-05,
+      "loss": 0.15535160064697265,
       "step": 425
     },
     {
       "epoch": 1.4754098360655736,
+      "grad_norm": 7.679214000701904,
       "learning_rate": 1.5673707210487983e-05,
+      "loss": 0.13739371299743652,
       "step": 450
     },
     {
       "epoch": 1.5573770491803278,
+      "grad_norm": 0.5809745192527771,
       "learning_rate": 1.530954115076475e-05,
+      "loss": 0.15218215942382812,
       "step": 475
     },
     {
       "epoch": 1.639344262295082,
+      "grad_norm": 0.14357534050941467,
       "learning_rate": 1.4945375091041516e-05,
+      "loss": 0.11159341812133788,
       "step": 500
     },
     {
       "epoch": 1.721311475409836,
+      "grad_norm": 0.15218983590602875,
       "learning_rate": 1.4581209031318282e-05,
+      "loss": 0.22156574249267577,
       "step": 525
     },
     {
       "epoch": 1.8032786885245902,
+      "grad_norm": 9.121078491210938,
       "learning_rate": 1.4217042971595047e-05,
+      "loss": 0.15829898834228515,
       "step": 550
     },
     {
       "epoch": 1.8852459016393444,
+      "grad_norm": 11.138046264648438,
       "learning_rate": 1.3852876911871815e-05,
+      "loss": 0.10829511642456055,
       "step": 575
     },
     {
       "epoch": 1.9672131147540983,
+      "grad_norm": 5.579981803894043,
       "learning_rate": 1.3488710852148582e-05,
+      "loss": 0.12963885307312012,
       "step": 600
     },
     {
       "epoch": 2.0,
+      "eval_accuracy": 0.9631901840490797,
+      "eval_f1": 0.7906976744186046,
+      "eval_loss": 0.12336025387048721,
+      "eval_precision": 0.7311827956989247,
+      "eval_recall": 0.8607594936708861,
+      "eval_roc_auc": 0.968502273975303,
+      "eval_runtime": 3.941,
+      "eval_samples_per_second": 248.162,
+      "eval_steps_per_second": 7.866,
       "step": 610
     },
     {
       "epoch": 2.0491803278688523,
+      "grad_norm": 23.79497718811035,
       "learning_rate": 1.3124544792425346e-05,
+      "loss": 0.07728540420532226,
       "step": 625
     },
     {
       "epoch": 2.1311475409836067,
+      "grad_norm": 0.3032953143119812,
       "learning_rate": 1.2760378732702113e-05,
+      "loss": 0.15112388610839844,
       "step": 650
     },
     {
       "epoch": 2.2131147540983607,
+      "grad_norm": 0.15015727281570435,
       "learning_rate": 1.239621267297888e-05,
+      "loss": 0.11781220436096192,
       "step": 675
     },
     {
       "epoch": 2.2950819672131146,
+      "grad_norm": 35.07762145996094,
       "learning_rate": 1.2032046613255645e-05,
+      "loss": 0.06247422218322754,
       "step": 700
     },
     {
       "epoch": 2.3770491803278686,
+      "grad_norm": 0.06300532817840576,
       "learning_rate": 1.1667880553532412e-05,
+      "loss": 0.07382740020751953,
       "step": 725
     },
     {
       "epoch": 2.459016393442623,
+      "grad_norm": 0.0386500246822834,
       "learning_rate": 1.1303714493809176e-05,
+      "loss": 0.10034572601318359,
       "step": 750
     },
     {
       "epoch": 2.540983606557377,
+      "grad_norm": 38.45136642456055,
       "learning_rate": 1.0939548434085944e-05,
+      "loss": 0.0807158374786377,
       "step": 775
     },
     {
       "epoch": 2.6229508196721314,
+      "grad_norm": 73.75131225585938,
       "learning_rate": 1.057538237436271e-05,
+      "loss": 0.11720505714416504,
       "step": 800
     },
     {
       "epoch": 2.7049180327868854,
+      "grad_norm": 3.530329704284668,
       "learning_rate": 1.0211216314639475e-05,
+      "loss": 0.07547494888305664,
       "step": 825
     },
     {
       "epoch": 2.7868852459016393,
+      "grad_norm": 2.0287606716156006,
       "learning_rate": 9.847050254916243e-06,
+      "loss": 0.07913532733917236,
       "step": 850
     },
     {
       "epoch": 2.8688524590163933,
+      "grad_norm": 0.07940487563610077,
       "learning_rate": 9.482884195193008e-06,
+      "loss": 0.0882901382446289,
       "step": 875
     },
     {
       "epoch": 2.9508196721311473,
+      "grad_norm": 0.03960009664297104,
       "learning_rate": 9.118718135469774e-06,
+      "loss": 0.10696781158447266,
       "step": 900
     },
     {
       "epoch": 3.0,
+      "eval_accuracy": 0.967280163599182,
+      "eval_f1": 0.8048780487804879,
+      "eval_loss": 0.13383829593658447,
+      "eval_precision": 0.7764705882352941,
+      "eval_recall": 0.8354430379746836,
+      "eval_roc_auc": 0.9735078357105644,
+      "eval_runtime": 4.0602,
+      "eval_samples_per_second": 240.875,
+      "eval_steps_per_second": 7.635,
       "step": 915
     },
     {
       "epoch": 3.0327868852459017,
+      "grad_norm": 0.08427837491035461,
       "learning_rate": 8.754552075746541e-06,
+      "loss": 0.07147216320037841,
       "step": 925
     },
     {
       "epoch": 3.1147540983606556,
+      "grad_norm": 0.027753490954637527,
       "learning_rate": 8.390386016023307e-06,
+      "loss": 0.060765981674194336,
       "step": 950
     },
     {
       "epoch": 3.19672131147541,
+      "grad_norm": 0.027190053835511208,
       "learning_rate": 8.026219956300074e-06,
+      "loss": 0.08444092750549316,
       "step": 975
     },
     {
       "epoch": 3.278688524590164,
+      "grad_norm": 6.747233867645264,
       "learning_rate": 7.66205389657684e-06,
+      "loss": 0.05561910152435303,
       "step": 1000
     },
     {
       "epoch": 3.360655737704918,
+      "grad_norm": 0.07523014396429062,
       "learning_rate": 7.2978878368536055e-06,
+      "loss": 0.05905795097351074,
       "step": 1025
     },
     {
       "epoch": 3.442622950819672,
+      "grad_norm": 8.827320098876953,
       "learning_rate": 6.933721777130372e-06,
+      "loss": 0.06066291809082031,
       "step": 1050
     },
     {
       "epoch": 3.5245901639344264,
+      "grad_norm": 0.2744128108024597,
       "learning_rate": 6.569555717407138e-06,
+      "loss": 0.07047759056091309,
       "step": 1075
     },
     {
       "epoch": 3.6065573770491803,
+      "grad_norm": 0.13911227881908417,
       "learning_rate": 6.2053896576839045e-06,
+      "loss": 0.049641780853271485,
       "step": 1100
     },
     {
       "epoch": 3.6885245901639343,
+      "grad_norm": 0.06898390501737595,
       "learning_rate": 5.84122359796067e-06,
+      "loss": 0.054807920455932614,
       "step": 1125
     },
     {
       "epoch": 3.7704918032786887,
+      "grad_norm": 37.68058395385742,
       "learning_rate": 5.477057538237437e-06,
+      "loss": 0.07445104598999024,
       "step": 1150
     },
     {
       "epoch": 3.8524590163934427,
+      "grad_norm": 0.046950697898864746,
       "learning_rate": 5.112891478514203e-06,
+      "loss": 0.03682142734527588,
       "step": 1175
     },
     {
       "epoch": 3.9344262295081966,
+      "grad_norm": 0.025340111926198006,
       "learning_rate": 4.748725418790969e-06,
+      "loss": 0.038784558773040774,
       "step": 1200
     },
     {
       "epoch": 4.0,
+      "eval_accuracy": 0.9642126789366053,
+      "eval_f1": 0.7770700636942676,
+      "eval_loss": 0.18827994167804718,
+      "eval_precision": 0.782051282051282,
+      "eval_recall": 0.7721518987341772,
+      "eval_roc_auc": 0.9623491643316765,
+      "eval_runtime": 4.0232,
+      "eval_samples_per_second": 243.092,
+      "eval_steps_per_second": 7.705,
       "step": 1220
     },
     {
       "epoch": 4.016393442622951,
+      "grad_norm": 0.04186977073550224,
       "learning_rate": 4.3845593590677355e-06,
+      "loss": 0.04547274112701416,
       "step": 1225
     },
     {
       "epoch": 4.098360655737705,
+      "grad_norm": 0.05808034911751747,
       "learning_rate": 4.020393299344502e-06,
+      "loss": 0.021595537662506104,
       "step": 1250
     },
     {
       "epoch": 4.180327868852459,
+      "grad_norm": 0.0242879930883646,
       "learning_rate": 3.656227239621268e-06,
+      "loss": 0.00493621289730072,
       "step": 1275
     },
     {
       "epoch": 4.262295081967213,
+      "grad_norm": 0.01932583376765251,
       "learning_rate": 3.292061179898034e-06,
+      "loss": 0.04504331111907959,
       "step": 1300
     },
     {
       "epoch": 4.344262295081967,
+      "grad_norm": 0.030308857560157776,
       "learning_rate": 2.9278951201748e-06,
+      "loss": 0.019644882678985596,
       "step": 1325
     },
     {
       "epoch": 4.426229508196721,
+      "grad_norm": 0.05657515302300453,
       "learning_rate": 2.5637290604515665e-06,
+      "loss": 0.024864859580993652,
       "step": 1350
     },
     {
       "epoch": 4.508196721311475,
+      "grad_norm": 0.025285128504037857,
       "learning_rate": 2.1995630007283324e-06,
+      "loss": 0.048575634956359866,
       "step": 1375
     },
     {
       "epoch": 4.590163934426229,
+      "grad_norm": 0.06963524967432022,
       "learning_rate": 1.8353969410050983e-06,
+      "loss": 0.04116048812866211,
       "step": 1400
     },
     {
       "epoch": 4.672131147540983,
+      "grad_norm": 69.19009399414062,
       "learning_rate": 1.4712308812818645e-06,
+      "loss": 0.047010822296142576,
       "step": 1425
     },
     {
       "epoch": 4.754098360655737,
+      "grad_norm": 0.08654582500457764,
       "learning_rate": 1.1070648215586309e-06,
+      "loss": 0.06569741249084472,
       "step": 1450
     },
     {
       "epoch": 4.836065573770492,
+      "grad_norm": 0.015758372843265533,
       "learning_rate": 7.428987618353969e-07,
+      "loss": 0.02632326602935791,
       "step": 1475
     },
     {
       "epoch": 4.918032786885246,
+      "grad_norm": 0.01882295496761799,
       "learning_rate": 3.787327021121632e-07,
+      "loss": 0.0074628710746765135,
       "step": 1500
     },
     {
       "epoch": 5.0,
+      "grad_norm": 0.04482486471533775,
       "learning_rate": 1.4566642388929353e-08,
+      "loss": 0.09066429138183593,
       "step": 1525
     },
     {
       "epoch": 5.0,
+      "eval_accuracy": 0.9611451942740287,
+      "eval_f1": 0.7625,
+      "eval_loss": 0.20975340902805328,
+      "eval_precision": 0.7530864197530864,
+      "eval_recall": 0.7721518987341772,
+      "eval_roc_auc": 0.9564776615367285,
+      "eval_runtime": 4.0531,
+      "eval_samples_per_second": 241.296,
+      "eval_steps_per_second": 7.648,
       "step": 1525
     }
   ],
         "early_stopping_threshold": 0.0
       },
       "attributes": {
+        "early_stopping_patience_counter": 2
       }
     },
     "TrainerControl": {

transformer/checkpoint-305/model.safetensors CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bb20e73a65d9daab2db9b645e0bd0608878ef7b24edb7ac88b5558863e0c5ecc
 size 1112205008

 version https://git-lfs.github.com/spec/v1
+oid sha256:200a48db20ed88a2035cb80d6b7d829284b98d0c8533b39ac5e28b87454f084f
 size 1112205008

transformer/checkpoint-305/optimizer.pt CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:48f419be6f85e4a9a74197d6249f1efe7127ac7668e97007ff458f4b5034cff4
 size 2224532875

 version https://git-lfs.github.com/spec/v1
+oid sha256:cc82673d0b9371c401f758333f509b134588b8d418658fae3ec8bc997290a087
 size 2224532875

transformer/checkpoint-305/scaler.pt CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9b479a091e6482cdb8085a5931fe45bae60997ab8939a20b371d61d6be6f7199
 size 1383

 version https://git-lfs.github.com/spec/v1
+oid sha256:58229eb0f47284e612f425b77489160034644a8dc962f6fd1132ab3848e5340b
 size 1383

transformer/checkpoint-305/trainer_state.json CHANGED Viewed

@@ -1,6 +1,6 @@
 {
   "best_global_step": 305,
-  "best_metric": 0.7852760736196319,
   "best_model_checkpoint": "/content/agri-utilization-classifier/transformer/checkpoint-305",
   "epoch": 1.0,
   "eval_steps": 500,
@@ -11,99 +11,99 @@
   "log_history": [
     {
       "epoch": 0.08196721311475409,
-      "grad_norm": 6.984184741973877,
       "learning_rate": 3.157894736842105e-06,
-      "loss": 0.7167730712890625,
       "step": 25
     },
     {
       "epoch": 0.16393442622950818,
-      "grad_norm": 9.77598762512207,
       "learning_rate": 6.447368421052632e-06,
-      "loss": 0.5636273193359375,
       "step": 50
     },
     {
       "epoch": 0.2459016393442623,
-      "grad_norm": 8.441609382629395,
       "learning_rate": 9.736842105263159e-06,
-      "loss": 0.37406421661376954,
       "step": 75
     },
     {
       "epoch": 0.32786885245901637,
-      "grad_norm": 5.313694953918457,
       "learning_rate": 1.3026315789473684e-05,
-      "loss": 0.2965927886962891,
       "step": 100
     },
     {
       "epoch": 0.4098360655737705,
-      "grad_norm": 7.240467548370361,
       "learning_rate": 1.6315789473684213e-05,
-      "loss": 0.29742313385009767,
       "step": 125
     },
     {
       "epoch": 0.4918032786885246,
-      "grad_norm": 6.100603103637695,
       "learning_rate": 1.960526315789474e-05,
-      "loss": 0.2068590545654297,
       "step": 150
     },
     {
       "epoch": 0.5737704918032787,
-      "grad_norm": 4.169040679931641,
       "learning_rate": 1.9679533867443555e-05,
-      "loss": 0.21712726593017578,
       "step": 175
     },
     {
       "epoch": 0.6557377049180327,
-      "grad_norm": 4.665876865386963,
       "learning_rate": 1.9315367807720323e-05,
-      "loss": 0.2889243125915527,
       "step": 200
     },
     {
       "epoch": 0.7377049180327869,
-      "grad_norm": 1.094870924949646,
       "learning_rate": 1.8951201747997088e-05,
-      "loss": 0.2303921699523926,
       "step": 225
     },
     {
       "epoch": 0.819672131147541,
-      "grad_norm": 1.2164329290390015,
       "learning_rate": 1.8587035688273852e-05,
-      "loss": 0.16723501205444335,
       "step": 250
     },
     {
       "epoch": 0.9016393442622951,
-      "grad_norm": 2.6314468383789062,
       "learning_rate": 1.822286962855062e-05,
-      "loss": 0.15685997009277344,
       "step": 275
     },
     {
       "epoch": 0.9836065573770492,
-      "grad_norm": 14.927972793579102,
       "learning_rate": 1.7858703568827385e-05,
-      "loss": 0.1979808807373047,
       "step": 300
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.9642126789366053,
-      "eval_f1": 0.7852760736196319,
-      "eval_loss": 0.10679091513156891,
-      "eval_precision": 0.7619047619047619,
-      "eval_recall": 0.810126582278481,
-      "eval_roc_auc": 0.9606665634108222,
-      "eval_runtime": 3.9998,
-      "eval_samples_per_second": 244.514,
-      "eval_steps_per_second": 7.75,
       "step": 305
     }
   ],

 {
   "best_global_step": 305,
+  "best_metric": 0.7916666666666666,
   "best_model_checkpoint": "/content/agri-utilization-classifier/transformer/checkpoint-305",
   "epoch": 1.0,
   "eval_steps": 500,
   "log_history": [
     {
       "epoch": 0.08196721311475409,
+      "grad_norm": 4.966984272003174,
       "learning_rate": 3.157894736842105e-06,
+      "loss": 0.570748291015625,
       "step": 25
     },
     {
       "epoch": 0.16393442622950818,
+      "grad_norm": 14.035679817199707,
       "learning_rate": 6.447368421052632e-06,
+      "loss": 0.4261627197265625,
       "step": 50
     },
     {
       "epoch": 0.2459016393442623,
+      "grad_norm": 6.1731109619140625,
       "learning_rate": 9.736842105263159e-06,
+      "loss": 0.35031925201416014,
       "step": 75
     },
     {
       "epoch": 0.32786885245901637,
+      "grad_norm": 3.6704306602478027,
       "learning_rate": 1.3026315789473684e-05,
+      "loss": 0.32047801971435547,
       "step": 100
     },
     {
       "epoch": 0.4098360655737705,
+      "grad_norm": 20.360071182250977,
       "learning_rate": 1.6315789473684213e-05,
+      "loss": 0.21136175155639647,
       "step": 125
     },
     {
       "epoch": 0.4918032786885246,
+      "grad_norm": 2.045907497406006,
       "learning_rate": 1.960526315789474e-05,
+      "loss": 0.21009126663208008,
       "step": 150
     },
     {
       "epoch": 0.5737704918032787,
+      "grad_norm": 7.755379676818848,
       "learning_rate": 1.9679533867443555e-05,
+      "loss": 0.19892181396484376,
       "step": 175
     },
     {
       "epoch": 0.6557377049180327,
+      "grad_norm": 0.6007249355316162,
       "learning_rate": 1.9315367807720323e-05,
+      "loss": 0.19585962295532228,
       "step": 200
     },
     {
       "epoch": 0.7377049180327869,
+      "grad_norm": 0.7606620192527771,
       "learning_rate": 1.8951201747997088e-05,
+      "loss": 0.282861213684082,
       "step": 225
     },
     {
       "epoch": 0.819672131147541,
+      "grad_norm": 5.813822269439697,
       "learning_rate": 1.8587035688273852e-05,
+      "loss": 0.16557788848876953,
       "step": 250
     },
     {
       "epoch": 0.9016393442622951,
+      "grad_norm": 3.329322099685669,
       "learning_rate": 1.822286962855062e-05,
+      "loss": 0.1400923728942871,
       "step": 275
     },
     {
       "epoch": 0.9836065573770492,
+      "grad_norm": 10.887845993041992,
       "learning_rate": 1.7858703568827385e-05,
+      "loss": 0.1979439926147461,
       "step": 300
     },
     {
       "epoch": 1.0,
+      "eval_accuracy": 0.9693251533742331,
+      "eval_f1": 0.7916666666666666,
+      "eval_loss": 0.1259259283542633,
+      "eval_precision": 0.8769230769230769,
+      "eval_recall": 0.7215189873417721,
+      "eval_roc_auc": 0.9533659058588304,
+      "eval_runtime": 3.8981,
+      "eval_samples_per_second": 250.889,
+      "eval_steps_per_second": 7.953,
       "step": 305
     }
   ],

transformer/checkpoint-610/model.safetensors CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0987d912d862daec9556dfa11ee4741f1f682f9ab67dfb40e0fb50fd6e0dd487
 size 1112205008

 version https://git-lfs.github.com/spec/v1
+oid sha256:959d44f755b56ade2cade52e51b7bd7fefad198f3c2b73494c05738161ffe28b
 size 1112205008

transformer/checkpoint-610/optimizer.pt CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:86b08b50cc64bad13c80cad58510d1315d9e730f97c4bb407691e44e8de8a71c
 size 2224532875

 version https://git-lfs.github.com/spec/v1
+oid sha256:64454423648f5f1d8cbb7a5d4aa9b57cc031c285bd51408afb20ef9fca929c39
 size 2224532875

transformer/checkpoint-610/scaler.pt CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:69a4619e39f20a9c2cf678e758d60086fc21c975ac26125b10b8caf00ffbba70
 size 1383

 version https://git-lfs.github.com/spec/v1
+oid sha256:03779557801a26e18eddf49e1c693f0a8452f6f1087c4117b7b6a9ee5260d6a3
 size 1383

transformer/checkpoint-610/trainer_state.json CHANGED Viewed

@@ -1,6 +1,6 @@
 {
   "best_global_step": 305,
-  "best_metric": 0.7852760736196319,
   "best_model_checkpoint": "/content/agri-utilization-classifier/transformer/checkpoint-305",
   "epoch": 2.0,
   "eval_steps": 500,
@@ -11,196 +11,196 @@
   "log_history": [
     {
       "epoch": 0.08196721311475409,
-      "grad_norm": 6.984184741973877,
       "learning_rate": 3.157894736842105e-06,
-      "loss": 0.7167730712890625,
       "step": 25
     },
     {
       "epoch": 0.16393442622950818,
-      "grad_norm": 9.77598762512207,
       "learning_rate": 6.447368421052632e-06,
-      "loss": 0.5636273193359375,
       "step": 50
     },
     {
       "epoch": 0.2459016393442623,
-      "grad_norm": 8.441609382629395,
       "learning_rate": 9.736842105263159e-06,
-      "loss": 0.37406421661376954,
       "step": 75
     },
     {
       "epoch": 0.32786885245901637,
-      "grad_norm": 5.313694953918457,
       "learning_rate": 1.3026315789473684e-05,
-      "loss": 0.2965927886962891,
       "step": 100
     },
     {
       "epoch": 0.4098360655737705,
-      "grad_norm": 7.240467548370361,
       "learning_rate": 1.6315789473684213e-05,
-      "loss": 0.29742313385009767,
       "step": 125
     },
     {
       "epoch": 0.4918032786885246,
-      "grad_norm": 6.100603103637695,
       "learning_rate": 1.960526315789474e-05,
-      "loss": 0.2068590545654297,
       "step": 150
     },
     {
       "epoch": 0.5737704918032787,
-      "grad_norm": 4.169040679931641,
       "learning_rate": 1.9679533867443555e-05,
-      "loss": 0.21712726593017578,
       "step": 175
     },
     {
       "epoch": 0.6557377049180327,
-      "grad_norm": 4.665876865386963,
       "learning_rate": 1.9315367807720323e-05,
-      "loss": 0.2889243125915527,
       "step": 200
     },
     {
       "epoch": 0.7377049180327869,
-      "grad_norm": 1.094870924949646,
       "learning_rate": 1.8951201747997088e-05,
-      "loss": 0.2303921699523926,
       "step": 225
     },
     {
       "epoch": 0.819672131147541,
-      "grad_norm": 1.2164329290390015,
       "learning_rate": 1.8587035688273852e-05,
-      "loss": 0.16723501205444335,
       "step": 250
     },
     {
       "epoch": 0.9016393442622951,
-      "grad_norm": 2.6314468383789062,
       "learning_rate": 1.822286962855062e-05,
-      "loss": 0.15685997009277344,
       "step": 275
     },
     {
       "epoch": 0.9836065573770492,
-      "grad_norm": 14.927972793579102,
       "learning_rate": 1.7858703568827385e-05,
-      "loss": 0.1979808807373047,
       "step": 300
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.9642126789366053,
-      "eval_f1": 0.7852760736196319,
-      "eval_loss": 0.10679091513156891,
-      "eval_precision": 0.7619047619047619,
-      "eval_recall": 0.810126582278481,
-      "eval_roc_auc": 0.9606665634108222,
-      "eval_runtime": 3.9998,
-      "eval_samples_per_second": 244.514,
-      "eval_steps_per_second": 7.75,
       "step": 305
     },
     {
       "epoch": 1.0655737704918034,
-      "grad_norm": 22.521757125854492,
       "learning_rate": 1.7494537509104153e-05,
-      "loss": 0.1316550636291504,
       "step": 325
     },
     {
       "epoch": 1.1475409836065573,
-      "grad_norm": 0.22687062621116638,
       "learning_rate": 1.7130371449380918e-05,
-      "loss": 0.2059168815612793,
       "step": 350
     },
     {
       "epoch": 1.2295081967213115,
-      "grad_norm": 0.657261312007904,
       "learning_rate": 1.6766205389657686e-05,
-      "loss": 0.09930330276489258,
       "step": 375
     },
     {
       "epoch": 1.3114754098360657,
-      "grad_norm": 4.796896457672119,
       "learning_rate": 1.640203932993445e-05,
-      "loss": 0.13251757621765137,
       "step": 400
     },
     {
       "epoch": 1.3934426229508197,
-      "grad_norm": 0.5394258499145508,
       "learning_rate": 1.603787327021122e-05,
-      "loss": 0.17834033966064453,
       "step": 425
     },
     {
       "epoch": 1.4754098360655736,
-      "grad_norm": 0.4655781388282776,
       "learning_rate": 1.5673707210487983e-05,
-      "loss": 0.122637300491333,
       "step": 450
     },
     {
       "epoch": 1.5573770491803278,
-      "grad_norm": 0.42849695682525635,
       "learning_rate": 1.530954115076475e-05,
-      "loss": 0.15983641624450684,
       "step": 475
     },
     {
       "epoch": 1.639344262295082,
-      "grad_norm": 0.45386794209480286,
       "learning_rate": 1.4945375091041516e-05,
-      "loss": 0.14264726638793945,
       "step": 500
     },
     {
       "epoch": 1.721311475409836,
-      "grad_norm": 0.6170782446861267,
       "learning_rate": 1.4581209031318282e-05,
-      "loss": 0.18886091232299804,
       "step": 525
     },
     {
       "epoch": 1.8032786885245902,
-      "grad_norm": 5.4546122550964355,
       "learning_rate": 1.4217042971595047e-05,
-      "loss": 0.14393989562988282,
       "step": 550
     },
     {
       "epoch": 1.8852459016393444,
-      "grad_norm": 17.419189453125,
       "learning_rate": 1.3852876911871815e-05,
-      "loss": 0.10272212982177735,
       "step": 575
     },
     {
       "epoch": 1.9672131147540983,
-      "grad_norm": 1.6267497539520264,
       "learning_rate": 1.3488710852148582e-05,
-      "loss": 0.1810975456237793,
       "step": 600
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.9580777096114519,
-      "eval_f1": 0.7759562841530054,
-      "eval_loss": 0.17126062512397766,
-      "eval_precision": 0.6826923076923077,
-      "eval_recall": 0.8987341772151899,
-      "eval_roc_auc": 0.9631939848777121,
-      "eval_runtime": 3.84,
-      "eval_samples_per_second": 254.687,
-      "eval_steps_per_second": 8.073,
       "step": 610
     }
   ],

 {
   "best_global_step": 305,
+  "best_metric": 0.7916666666666666,
   "best_model_checkpoint": "/content/agri-utilization-classifier/transformer/checkpoint-305",
   "epoch": 2.0,
   "eval_steps": 500,
   "log_history": [
     {
       "epoch": 0.08196721311475409,
+      "grad_norm": 4.966984272003174,
       "learning_rate": 3.157894736842105e-06,
+      "loss": 0.570748291015625,
       "step": 25
     },
     {
       "epoch": 0.16393442622950818,
+      "grad_norm": 14.035679817199707,
       "learning_rate": 6.447368421052632e-06,
+      "loss": 0.4261627197265625,
       "step": 50
     },
     {
       "epoch": 0.2459016393442623,
+      "grad_norm": 6.1731109619140625,
       "learning_rate": 9.736842105263159e-06,
+      "loss": 0.35031925201416014,
       "step": 75
     },
     {
       "epoch": 0.32786885245901637,
+      "grad_norm": 3.6704306602478027,
       "learning_rate": 1.3026315789473684e-05,
+      "loss": 0.32047801971435547,
       "step": 100
     },
     {
       "epoch": 0.4098360655737705,
+      "grad_norm": 20.360071182250977,
       "learning_rate": 1.6315789473684213e-05,
+      "loss": 0.21136175155639647,
       "step": 125
     },
     {
       "epoch": 0.4918032786885246,
+      "grad_norm": 2.045907497406006,
       "learning_rate": 1.960526315789474e-05,
+      "loss": 0.21009126663208008,
       "step": 150
     },
     {
       "epoch": 0.5737704918032787,
+      "grad_norm": 7.755379676818848,
       "learning_rate": 1.9679533867443555e-05,
+      "loss": 0.19892181396484376,
       "step": 175
     },
     {
       "epoch": 0.6557377049180327,
+      "grad_norm": 0.6007249355316162,
       "learning_rate": 1.9315367807720323e-05,
+      "loss": 0.19585962295532228,
       "step": 200
     },
     {
       "epoch": 0.7377049180327869,
+      "grad_norm": 0.7606620192527771,
       "learning_rate": 1.8951201747997088e-05,
+      "loss": 0.282861213684082,
       "step": 225
     },
     {
       "epoch": 0.819672131147541,
+      "grad_norm": 5.813822269439697,
       "learning_rate": 1.8587035688273852e-05,
+      "loss": 0.16557788848876953,
       "step": 250
     },
     {
       "epoch": 0.9016393442622951,
+      "grad_norm": 3.329322099685669,
       "learning_rate": 1.822286962855062e-05,
+      "loss": 0.1400923728942871,
       "step": 275
     },
     {
       "epoch": 0.9836065573770492,
+      "grad_norm": 10.887845993041992,
       "learning_rate": 1.7858703568827385e-05,
+      "loss": 0.1979439926147461,
       "step": 300
     },
     {
       "epoch": 1.0,
+      "eval_accuracy": 0.9693251533742331,
+      "eval_f1": 0.7916666666666666,
+      "eval_loss": 0.1259259283542633,
+      "eval_precision": 0.8769230769230769,
+      "eval_recall": 0.7215189873417721,
+      "eval_roc_auc": 0.9533659058588304,
+      "eval_runtime": 3.8981,
+      "eval_samples_per_second": 250.889,
+      "eval_steps_per_second": 7.953,
       "step": 305
     },
     {
       "epoch": 1.0655737704918034,
+      "grad_norm": 4.172677516937256,
       "learning_rate": 1.7494537509104153e-05,
+      "loss": 0.13631089210510253,
       "step": 325
     },
     {
       "epoch": 1.1475409836065573,
+      "grad_norm": 0.07315228134393692,
       "learning_rate": 1.7130371449380918e-05,
+      "loss": 0.15998102188110352,
       "step": 350
     },
     {
       "epoch": 1.2295081967213115,
+      "grad_norm": 0.4799639582633972,
       "learning_rate": 1.6766205389657686e-05,
+      "loss": 0.10631587982177734,
       "step": 375
     },
     {
       "epoch": 1.3114754098360657,
+      "grad_norm": 1.1866878271102905,
       "learning_rate": 1.640203932993445e-05,
+      "loss": 0.11795808792114258,
       "step": 400
     },
     {
       "epoch": 1.3934426229508197,
+      "grad_norm": 0.6632266044616699,
       "learning_rate": 1.603787327021122e-05,
+      "loss": 0.15535160064697265,
       "step": 425
     },
     {
       "epoch": 1.4754098360655736,
+      "grad_norm": 7.679214000701904,
       "learning_rate": 1.5673707210487983e-05,
+      "loss": 0.13739371299743652,
       "step": 450
     },
     {
       "epoch": 1.5573770491803278,
+      "grad_norm": 0.5809745192527771,
       "learning_rate": 1.530954115076475e-05,
+      "loss": 0.15218215942382812,
       "step": 475
     },
     {
       "epoch": 1.639344262295082,
+      "grad_norm": 0.14357534050941467,
       "learning_rate": 1.4945375091041516e-05,
+      "loss": 0.11159341812133788,
       "step": 500
     },
     {
       "epoch": 1.721311475409836,
+      "grad_norm": 0.15218983590602875,
       "learning_rate": 1.4581209031318282e-05,
+      "loss": 0.22156574249267577,
       "step": 525
     },
     {
       "epoch": 1.8032786885245902,
+      "grad_norm": 9.121078491210938,
       "learning_rate": 1.4217042971595047e-05,
+      "loss": 0.15829898834228515,
       "step": 550
     },
     {
       "epoch": 1.8852459016393444,
+      "grad_norm": 11.138046264648438,
       "learning_rate": 1.3852876911871815e-05,
+      "loss": 0.10829511642456055,
       "step": 575
     },
     {
       "epoch": 1.9672131147540983,
+      "grad_norm": 5.579981803894043,
       "learning_rate": 1.3488710852148582e-05,
+      "loss": 0.12963885307312012,
       "step": 600
     },
     {
       "epoch": 2.0,
+      "eval_accuracy": 0.9631901840490797,
+      "eval_f1": 0.7906976744186046,
+      "eval_loss": 0.12336025387048721,
+      "eval_precision": 0.7311827956989247,
+      "eval_recall": 0.8607594936708861,
+      "eval_roc_auc": 0.968502273975303,
+      "eval_runtime": 3.941,
+      "eval_samples_per_second": 248.162,
+      "eval_steps_per_second": 7.866,
       "step": 610
     }
   ],

transformer/checkpoint-915/model.safetensors CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d3728e86863d86207f22f0d07fcdb0cdefb8ea0edac728a54f99e18b7d45449f
 size 1112205008

 version https://git-lfs.github.com/spec/v1
+oid sha256:18f129b4e4685fa8d1bb28430377232612ff45044f3801b6322b1ab5b76e8b6c
 size 1112205008

transformer/checkpoint-915/optimizer.pt CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9b61445e4f3f6c58912fc7900494d712e518b57d94365301594eab7debc2f82f
 size 2224532875

 version https://git-lfs.github.com/spec/v1
+oid sha256:e98e47dfa6571159478c43aacef8750b2b563202e0541be73a70743e64d9f090
 size 2224532875

transformer/checkpoint-915/scaler.pt CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d5de3f7c64db116b19b953c696752ac11438d1a3bafe7b6ff41955b33f8fba9c
 size 1383

 version https://git-lfs.github.com/spec/v1
+oid sha256:e4f3a4f326f16471e5d5c4e3138e7cc78982ebb595c20a5ac90094aba45c9fa8
 size 1383

transformer/checkpoint-915/trainer_state.json CHANGED Viewed

@@ -1,6 +1,6 @@
 {
   "best_global_step": 915,
-  "best_metric": 0.7868852459016393,
   "best_model_checkpoint": "/content/agri-utilization-classifier/transformer/checkpoint-915",
   "epoch": 3.0,
   "eval_steps": 500,
@@ -11,293 +11,293 @@
   "log_history": [
     {
       "epoch": 0.08196721311475409,
-      "grad_norm": 6.984184741973877,
       "learning_rate": 3.157894736842105e-06,
-      "loss": 0.7167730712890625,
       "step": 25
     },
     {
       "epoch": 0.16393442622950818,
-      "grad_norm": 9.77598762512207,
       "learning_rate": 6.447368421052632e-06,
-      "loss": 0.5636273193359375,
       "step": 50
     },
     {
       "epoch": 0.2459016393442623,
-      "grad_norm": 8.441609382629395,
       "learning_rate": 9.736842105263159e-06,
-      "loss": 0.37406421661376954,
       "step": 75
     },
     {
       "epoch": 0.32786885245901637,
-      "grad_norm": 5.313694953918457,
       "learning_rate": 1.3026315789473684e-05,
-      "loss": 0.2965927886962891,
       "step": 100
     },
     {
       "epoch": 0.4098360655737705,
-      "grad_norm": 7.240467548370361,
       "learning_rate": 1.6315789473684213e-05,
-      "loss": 0.29742313385009767,
       "step": 125
     },
     {
       "epoch": 0.4918032786885246,
-      "grad_norm": 6.100603103637695,
       "learning_rate": 1.960526315789474e-05,
-      "loss": 0.2068590545654297,
       "step": 150
     },
     {
       "epoch": 0.5737704918032787,
-      "grad_norm": 4.169040679931641,
       "learning_rate": 1.9679533867443555e-05,
-      "loss": 0.21712726593017578,
       "step": 175
     },
     {
       "epoch": 0.6557377049180327,
-      "grad_norm": 4.665876865386963,
       "learning_rate": 1.9315367807720323e-05,
-      "loss": 0.2889243125915527,
       "step": 200
     },
     {
       "epoch": 0.7377049180327869,
-      "grad_norm": 1.094870924949646,
       "learning_rate": 1.8951201747997088e-05,
-      "loss": 0.2303921699523926,
       "step": 225
     },
     {
       "epoch": 0.819672131147541,
-      "grad_norm": 1.2164329290390015,
       "learning_rate": 1.8587035688273852e-05,
-      "loss": 0.16723501205444335,
       "step": 250
     },
     {
       "epoch": 0.9016393442622951,
-      "grad_norm": 2.6314468383789062,
       "learning_rate": 1.822286962855062e-05,
-      "loss": 0.15685997009277344,
       "step": 275
     },
     {
       "epoch": 0.9836065573770492,
-      "grad_norm": 14.927972793579102,
       "learning_rate": 1.7858703568827385e-05,
-      "loss": 0.1979808807373047,
       "step": 300
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.9642126789366053,
-      "eval_f1": 0.7852760736196319,
-      "eval_loss": 0.10679091513156891,
-      "eval_precision": 0.7619047619047619,
-      "eval_recall": 0.810126582278481,
-      "eval_roc_auc": 0.9606665634108222,
-      "eval_runtime": 3.9998,
-      "eval_samples_per_second": 244.514,
-      "eval_steps_per_second": 7.75,
       "step": 305
     },
     {
       "epoch": 1.0655737704918034,
-      "grad_norm": 22.521757125854492,
       "learning_rate": 1.7494537509104153e-05,
-      "loss": 0.1316550636291504,
       "step": 325
     },
     {
       "epoch": 1.1475409836065573,
-      "grad_norm": 0.22687062621116638,
       "learning_rate": 1.7130371449380918e-05,
-      "loss": 0.2059168815612793,
       "step": 350
     },
     {
       "epoch": 1.2295081967213115,
-      "grad_norm": 0.657261312007904,
       "learning_rate": 1.6766205389657686e-05,
-      "loss": 0.09930330276489258,
       "step": 375
     },
     {
       "epoch": 1.3114754098360657,
-      "grad_norm": 4.796896457672119,
       "learning_rate": 1.640203932993445e-05,
-      "loss": 0.13251757621765137,
       "step": 400
     },
     {
       "epoch": 1.3934426229508197,
-      "grad_norm": 0.5394258499145508,
       "learning_rate": 1.603787327021122e-05,
-      "loss": 0.17834033966064453,
       "step": 425
     },
     {
       "epoch": 1.4754098360655736,
-      "grad_norm": 0.4655781388282776,
       "learning_rate": 1.5673707210487983e-05,
-      "loss": 0.122637300491333,
       "step": 450
     },
     {
       "epoch": 1.5573770491803278,
-      "grad_norm": 0.42849695682525635,
       "learning_rate": 1.530954115076475e-05,
-      "loss": 0.15983641624450684,
       "step": 475
     },
     {
       "epoch": 1.639344262295082,
-      "grad_norm": 0.45386794209480286,
       "learning_rate": 1.4945375091041516e-05,
-      "loss": 0.14264726638793945,
       "step": 500
     },
     {
       "epoch": 1.721311475409836,
-      "grad_norm": 0.6170782446861267,
       "learning_rate": 1.4581209031318282e-05,
-      "loss": 0.18886091232299804,
       "step": 525
     },
     {
       "epoch": 1.8032786885245902,
-      "grad_norm": 5.4546122550964355,
       "learning_rate": 1.4217042971595047e-05,
-      "loss": 0.14393989562988282,
       "step": 550
     },
     {
       "epoch": 1.8852459016393444,
-      "grad_norm": 17.419189453125,
       "learning_rate": 1.3852876911871815e-05,
-      "loss": 0.10272212982177735,
       "step": 575
     },
     {
       "epoch": 1.9672131147540983,
-      "grad_norm": 1.6267497539520264,
       "learning_rate": 1.3488710852148582e-05,
-      "loss": 0.1810975456237793,
       "step": 600
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.9580777096114519,
-      "eval_f1": 0.7759562841530054,
-      "eval_loss": 0.17126062512397766,
-      "eval_precision": 0.6826923076923077,
-      "eval_recall": 0.8987341772151899,
-      "eval_roc_auc": 0.9631939848777121,
-      "eval_runtime": 3.84,
-      "eval_samples_per_second": 254.687,
-      "eval_steps_per_second": 8.073,
       "step": 610
     },
     {
       "epoch": 2.0491803278688523,
-      "grad_norm": 1.3180276155471802,
       "learning_rate": 1.3124544792425346e-05,
-      "loss": 0.05921304225921631,
       "step": 625
     },
     {
       "epoch": 2.1311475409836067,
-      "grad_norm": 5.675038814544678,
       "learning_rate": 1.2760378732702113e-05,
-      "loss": 0.16824769973754883,
       "step": 650
     },
     {
       "epoch": 2.2131147540983607,
-      "grad_norm": 0.16993092000484467,
       "learning_rate": 1.239621267297888e-05,
-      "loss": 0.12186273574829101,
       "step": 675
     },
     {
       "epoch": 2.2950819672131146,
-      "grad_norm": 1.1791695356369019,
       "learning_rate": 1.2032046613255645e-05,
-      "loss": 0.08795836448669433,
       "step": 700
     },
     {
       "epoch": 2.3770491803278686,
-      "grad_norm": 0.07541065663099289,
       "learning_rate": 1.1667880553532412e-05,
-      "loss": 0.10129087448120117,
       "step": 725
     },
     {
       "epoch": 2.459016393442623,
-      "grad_norm": 3.394912004470825,
       "learning_rate": 1.1303714493809176e-05,
-      "loss": 0.14056243896484374,
       "step": 750
     },
     {
       "epoch": 2.540983606557377,
-      "grad_norm": 8.074258804321289,
       "learning_rate": 1.0939548434085944e-05,
-      "loss": 0.06563093185424805,
       "step": 775
     },
     {
       "epoch": 2.6229508196721314,
-      "grad_norm": 12.472029685974121,
       "learning_rate": 1.057538237436271e-05,
-      "loss": 0.09851057052612305,
       "step": 800
     },
     {
       "epoch": 2.7049180327868854,
-      "grad_norm": 0.10368915647268295,
       "learning_rate": 1.0211216314639475e-05,
-      "loss": 0.11658324241638184,
       "step": 825
     },
     {
       "epoch": 2.7868852459016393,
-      "grad_norm": 44.263092041015625,
       "learning_rate": 9.847050254916243e-06,
-      "loss": 0.13634946823120117,
       "step": 850
     },
     {
       "epoch": 2.8688524590163933,
-      "grad_norm": 0.07709958404302597,
       "learning_rate": 9.482884195193008e-06,
-      "loss": 0.12144805908203125,
       "step": 875
     },
     {
       "epoch": 2.9508196721311473,
-      "grad_norm": 0.11255892366170883,
       "learning_rate": 9.118718135469774e-06,
-      "loss": 0.11815821647644043,
       "step": 900
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.9601226993865031,
-      "eval_f1": 0.7868852459016393,
-      "eval_loss": 0.14711864292621613,
-      "eval_precision": 0.6923076923076923,
-      "eval_recall": 0.9113924050632911,
-      "eval_roc_auc": 0.9719449177003984,
-      "eval_runtime": 3.9043,
-      "eval_samples_per_second": 250.495,
-      "eval_steps_per_second": 7.94,
       "step": 915
     }
   ],

 {
   "best_global_step": 915,
+  "best_metric": 0.8048780487804879,
   "best_model_checkpoint": "/content/agri-utilization-classifier/transformer/checkpoint-915",
   "epoch": 3.0,
   "eval_steps": 500,
   "log_history": [
     {
       "epoch": 0.08196721311475409,
+      "grad_norm": 4.966984272003174,
       "learning_rate": 3.157894736842105e-06,
+      "loss": 0.570748291015625,
       "step": 25
     },
     {
       "epoch": 0.16393442622950818,
+      "grad_norm": 14.035679817199707,
       "learning_rate": 6.447368421052632e-06,
+      "loss": 0.4261627197265625,
       "step": 50
     },
     {
       "epoch": 0.2459016393442623,
+      "grad_norm": 6.1731109619140625,
       "learning_rate": 9.736842105263159e-06,
+      "loss": 0.35031925201416014,
       "step": 75
     },
     {
       "epoch": 0.32786885245901637,
+      "grad_norm": 3.6704306602478027,
       "learning_rate": 1.3026315789473684e-05,
+      "loss": 0.32047801971435547,
       "step": 100
     },
     {
       "epoch": 0.4098360655737705,
+      "grad_norm": 20.360071182250977,
       "learning_rate": 1.6315789473684213e-05,
+      "loss": 0.21136175155639647,
       "step": 125
     },
     {
       "epoch": 0.4918032786885246,
+      "grad_norm": 2.045907497406006,
       "learning_rate": 1.960526315789474e-05,
+      "loss": 0.21009126663208008,
       "step": 150
     },
     {
       "epoch": 0.5737704918032787,
+      "grad_norm": 7.755379676818848,
       "learning_rate": 1.9679533867443555e-05,
+      "loss": 0.19892181396484376,
       "step": 175
     },
     {
       "epoch": 0.6557377049180327,
+      "grad_norm": 0.6007249355316162,
       "learning_rate": 1.9315367807720323e-05,
+      "loss": 0.19585962295532228,
       "step": 200
     },
     {
       "epoch": 0.7377049180327869,
+      "grad_norm": 0.7606620192527771,
       "learning_rate": 1.8951201747997088e-05,
+      "loss": 0.282861213684082,
       "step": 225
     },
     {
       "epoch": 0.819672131147541,
+      "grad_norm": 5.813822269439697,
       "learning_rate": 1.8587035688273852e-05,
+      "loss": 0.16557788848876953,
       "step": 250
     },
     {
       "epoch": 0.9016393442622951,
+      "grad_norm": 3.329322099685669,
       "learning_rate": 1.822286962855062e-05,
+      "loss": 0.1400923728942871,
       "step": 275
     },
     {
       "epoch": 0.9836065573770492,
+      "grad_norm": 10.887845993041992,
       "learning_rate": 1.7858703568827385e-05,
+      "loss": 0.1979439926147461,
       "step": 300
     },
     {
       "epoch": 1.0,
+      "eval_accuracy": 0.9693251533742331,
+      "eval_f1": 0.7916666666666666,
+      "eval_loss": 0.1259259283542633,
+      "eval_precision": 0.8769230769230769,
+      "eval_recall": 0.7215189873417721,
+      "eval_roc_auc": 0.9533659058588304,
+      "eval_runtime": 3.8981,
+      "eval_samples_per_second": 250.889,
+      "eval_steps_per_second": 7.953,
       "step": 305
     },
     {
       "epoch": 1.0655737704918034,
+      "grad_norm": 4.172677516937256,
       "learning_rate": 1.7494537509104153e-05,
+      "loss": 0.13631089210510253,
       "step": 325
     },
     {
       "epoch": 1.1475409836065573,
+      "grad_norm": 0.07315228134393692,
       "learning_rate": 1.7130371449380918e-05,
+      "loss": 0.15998102188110352,
       "step": 350
     },
     {
       "epoch": 1.2295081967213115,
+      "grad_norm": 0.4799639582633972,
       "learning_rate": 1.6766205389657686e-05,
+      "loss": 0.10631587982177734,
       "step": 375
     },
     {
       "epoch": 1.3114754098360657,
+      "grad_norm": 1.1866878271102905,
       "learning_rate": 1.640203932993445e-05,
+      "loss": 0.11795808792114258,
       "step": 400
     },
     {
       "epoch": 1.3934426229508197,
+      "grad_norm": 0.6632266044616699,
       "learning_rate": 1.603787327021122e-05,
+      "loss": 0.15535160064697265,
       "step": 425
     },
     {
       "epoch": 1.4754098360655736,
+      "grad_norm": 7.679214000701904,
       "learning_rate": 1.5673707210487983e-05,
+      "loss": 0.13739371299743652,
       "step": 450
     },
     {
       "epoch": 1.5573770491803278,
+      "grad_norm": 0.5809745192527771,
       "learning_rate": 1.530954115076475e-05,
+      "loss": 0.15218215942382812,
       "step": 475
     },
     {
       "epoch": 1.639344262295082,
+      "grad_norm": 0.14357534050941467,
       "learning_rate": 1.4945375091041516e-05,
+      "loss": 0.11159341812133788,
       "step": 500
     },
     {
       "epoch": 1.721311475409836,
+      "grad_norm": 0.15218983590602875,
       "learning_rate": 1.4581209031318282e-05,
+      "loss": 0.22156574249267577,
       "step": 525
     },
     {
       "epoch": 1.8032786885245902,
+      "grad_norm": 9.121078491210938,
       "learning_rate": 1.4217042971595047e-05,
+      "loss": 0.15829898834228515,
       "step": 550
     },
     {
       "epoch": 1.8852459016393444,
+      "grad_norm": 11.138046264648438,
       "learning_rate": 1.3852876911871815e-05,
+      "loss": 0.10829511642456055,
       "step": 575
     },
     {
       "epoch": 1.9672131147540983,
+      "grad_norm": 5.579981803894043,
       "learning_rate": 1.3488710852148582e-05,
+      "loss": 0.12963885307312012,
       "step": 600
     },
     {
       "epoch": 2.0,
+      "eval_accuracy": 0.9631901840490797,
+      "eval_f1": 0.7906976744186046,
+      "eval_loss": 0.12336025387048721,
+      "eval_precision": 0.7311827956989247,
+      "eval_recall": 0.8607594936708861,
+      "eval_roc_auc": 0.968502273975303,
+      "eval_runtime": 3.941,
+      "eval_samples_per_second": 248.162,
+      "eval_steps_per_second": 7.866,
       "step": 610
     },
     {
       "epoch": 2.0491803278688523,
+      "grad_norm": 23.79497718811035,
       "learning_rate": 1.3124544792425346e-05,
+      "loss": 0.07728540420532226,
       "step": 625
     },
     {
       "epoch": 2.1311475409836067,
+      "grad_norm": 0.3032953143119812,
       "learning_rate": 1.2760378732702113e-05,
+      "loss": 0.15112388610839844,
       "step": 650
     },
     {
       "epoch": 2.2131147540983607,
+      "grad_norm": 0.15015727281570435,
       "learning_rate": 1.239621267297888e-05,
+      "loss": 0.11781220436096192,
       "step": 675
     },
     {
       "epoch": 2.2950819672131146,
+      "grad_norm": 35.07762145996094,
       "learning_rate": 1.2032046613255645e-05,
+      "loss": 0.06247422218322754,
       "step": 700
     },
     {
       "epoch": 2.3770491803278686,
+      "grad_norm": 0.06300532817840576,
       "learning_rate": 1.1667880553532412e-05,
+      "loss": 0.07382740020751953,
       "step": 725
     },
     {
       "epoch": 2.459016393442623,
+      "grad_norm": 0.0386500246822834,
       "learning_rate": 1.1303714493809176e-05,
+      "loss": 0.10034572601318359,
       "step": 750
     },
     {
       "epoch": 2.540983606557377,
+      "grad_norm": 38.45136642456055,
       "learning_rate": 1.0939548434085944e-05,
+      "loss": 0.0807158374786377,
       "step": 775
     },
     {
       "epoch": 2.6229508196721314,
+      "grad_norm": 73.75131225585938,
       "learning_rate": 1.057538237436271e-05,
+      "loss": 0.11720505714416504,
       "step": 800
     },
     {
       "epoch": 2.7049180327868854,
+      "grad_norm": 3.530329704284668,
       "learning_rate": 1.0211216314639475e-05,
+      "loss": 0.07547494888305664,
       "step": 825
     },
     {
       "epoch": 2.7868852459016393,
+      "grad_norm": 2.0287606716156006,
       "learning_rate": 9.847050254916243e-06,
+      "loss": 0.07913532733917236,
       "step": 850
     },
     {
       "epoch": 2.8688524590163933,
+      "grad_norm": 0.07940487563610077,
       "learning_rate": 9.482884195193008e-06,
+      "loss": 0.0882901382446289,
       "step": 875
     },
     {
       "epoch": 2.9508196721311473,
+      "grad_norm": 0.03960009664297104,
       "learning_rate": 9.118718135469774e-06,
+      "loss": 0.10696781158447266,
       "step": 900
     },
     {
       "epoch": 3.0,
+      "eval_accuracy": 0.967280163599182,
+      "eval_f1": 0.8048780487804879,
+      "eval_loss": 0.13383829593658447,
+      "eval_precision": 0.7764705882352941,
+      "eval_recall": 0.8354430379746836,
+      "eval_roc_auc": 0.9735078357105644,
+      "eval_runtime": 4.0602,
+      "eval_samples_per_second": 240.875,
+      "eval_steps_per_second": 7.635,
       "step": 915
     }
   ],

transformer/config.json CHANGED Viewed

@@ -31,16 +31,16 @@
   "pad_token_id": 1,
   "position_embedding_type": "absolute",
   "problem_type": "single_label_classification",
-  "threshold": 0.6156440377235413,
   "tie_word_embeddings": true,
   "transformers_version": "5.9.0",
   "type_vocab_size": 1,
   "use_cache": false,
   "validation_threshold_report": {
-    "f1": 0.8072289156626505,
-    "precision": 0.7701149425287356,
-    "recall": 0.8481012658227848,
-    "threshold": 0.6156440377235413
   },
   "vocab_size": 250002
 }

   "pad_token_id": 1,
   "position_embedding_type": "absolute",
   "problem_type": "single_label_classification",
+  "threshold": 0.961713969707489,
   "tie_word_embeddings": true,
   "transformers_version": "5.9.0",
   "type_vocab_size": 1,
   "use_cache": false,
   "validation_threshold_report": {
+    "f1": 0.8322147651006712,
+    "precision": 0.8857142857142857,
+    "recall": 0.7848101265822784,
+    "threshold": 0.961713969707489
   },
   "vocab_size": 250002
 }

transformer/model.safetensors CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8f09e4daa3612b0a08afd0486aa5ce25a384c7e9c3abe05df4b6d1f68e5033b8
 size 1112205008

 version https://git-lfs.github.com/spec/v1
+oid sha256:18f129b4e4685fa8d1bb28430377232612ff45044f3801b6322b1ab5b76e8b6c
 size 1112205008

transformer/test_predictions.csv CHANGED Viewed

The diff for this file is too large to render. See raw diff

transformer/validation_predictions.csv CHANGED Viewed

The diff for this file is too large to render. See raw diff