Training in progress, step 2

README.md

@@ -21,7 +21,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [DedalusHealthCare/tinybert-mlm-en](https://huggingface.co/DedalusHealthCare/tinybert-mlm-en) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.5237
+- Loss: 0.6019
 - Accuracy: 0.9816
 - F1: 0.0
 - Precision: 0.0
@@ -58,17 +58,17 @@ The following hyperparameters were used during training:
 
 ### Training results
 
-| Training Loss | Epoch  | Step | Validation Loss | Accuracy | F1  | Precision | Recall |
-|:-------------:|:------:|:----:|:---------------:|:--------:|:---:|:---------:|:------:|
-| 0.6688        | 0.2857 | 1    | 0.6695          | 0.7690   | 0.0 | 0.0       | 0.0    |
-| 0.6688        | 0.5714 | 2    | 0.6647          | 0.8110   | 0.0 | 0.0       | 0.0    |
-| 0.6688        | 0.8571 | 3    | 0.6551          | 0.8635   | 0.0 | 0.0       | 0.0    |
-| 0.6688        | 1.1429 | 4    | 0.6407          | 0.9344   | 0.0 | 0.0       | 0.0    |
-| 0.6688        | 1.4286 | 5    | 0.6220          | 0.9659   | 0.0 | 0.0       | 0.0    |
-| 0.6688        | 1.7143 | 6    | 0.5993          | 0.9764   | 0.0 | 0.0       | 0.0    |
-| 0.6688        | 2.0    | 7    | 0.5731          | 0.9816   | 0.0 | 0.0       | 0.0    |
-| 0.6688        | 2.2857 | 8    | 0.5478          | 0.9816   | 0.0 | 0.0       | 0.0    |
-| 0.6688        | 2.5714 | 9    | 0.5237          | 0.9816   | 0.0 | 0.0       | 0.0    |
+| Training Loss | Epoch  | Step | Validation Loss | Accuracy | F1     | Precision | Recall |
+|:-------------:|:------:|:----:|:---------------:|:--------:|:------:|:---------:|:------:|
+| 0.7327        | 0.2857 | 1    | 0.7513          | 0.0341   | 0.0366 | 0.0187    | 1.0    |
+| 0.7327        | 0.5714 | 2    | 0.7464          | 0.0525   | 0.0373 | 0.0190    | 1.0    |
+| 0.7327        | 0.8571 | 3    | 0.7368          | 0.0892   | 0.0388 | 0.0198    | 1.0    |
+| 0.7327        | 1.1429 | 4    | 0.7225          | 0.1916   | 0.0375 | 0.0192    | 0.8571 |
+| 0.7327        | 1.4286 | 5    | 0.7037          | 0.3885   | 0.0251 | 0.0129    | 0.4286 |
+| 0.7327        | 1.7143 | 6    | 0.6806          | 0.6325   | 0.0    | 0.0       | 0.0    |
+| 0.7327        | 2.0    | 7    | 0.6538          | 0.8688   | 0.0    | 0.0       | 0.0    |
+| 0.7327        | 2.2857 | 8    | 0.6274          | 0.9738   | 0.0    | 0.0       | 0.0    |
+| 0.7327        | 2.5714 | 9    | 0.6019          | 0.9816   | 0.0    | 0.0       | 0.0    |
 
 
 ### Framework versions

evaluation_results_en/classification_report_goldset.json

@@ -1,27 +1,27 @@
 {
     "NOT_FAMILY": {
-        "precision": 0.9789029535864979,
-        "recall": 0.6010362694300518,
-        "f1-score": 0.7447833065810595,
+        "precision": 1.0,
+        "recall": 0.13212435233160622,
+        "f1-score": 0.2334096109839817,
         "support": 386.0
     },
     "FAMILY": {
-        "precision": 0.0,
-        "recall": 0.0,
-        "f1-score": 0.0,
+        "precision": 0.014705882352941176,
+        "recall": 1.0,
+        "f1-score": 0.028985507246376812,
         "support": 5.0
     },
-    "accuracy": 0.5933503836317136,
+    "accuracy": 0.1432225063938619,
     "macro avg": {
-        "precision": 0.48945147679324896,
-        "recall": 0.3005181347150259,
-        "f1-score": 0.37239165329052976,
+        "precision": 0.5073529411764706,
+        "recall": 0.5660621761658031,
+        "f1-score": 0.13119755911517925,
         "support": 391.0
     },
     "weighted avg": {
-        "precision": 0.9663850130035504,
-        "recall": 0.5933503836317136,
-        "f1-score": 0.7352592233766981,
+        "precision": 0.9874003309763802,
+        "recall": 0.1432225063938619,
+        "f1-score": 0.23079549201035504,
         "support": 391.0
     },
     "evaluation_metadata": {

logs/events.out.tfevents.1760011961.ip-172-31-12-22.10247.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5cf63626a466d75dc5371a61b6bb529168ff6904ed43931995aea17dd46de900
-size 5861
+oid sha256:a13171870ca1b5dab2e9056526c53f21f5ea5f5ec978465df9a61df8400950d4
+size 6324

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dad234a2b33b93334794723105ac1864560d83f4af65d915679ffa594b1f7a4d
+oid sha256:c6ce88543f6e1070478bae37f47905a56b32a61575d8e04d9c5ac28e2a35c4f0
 size 46634832

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7a4cc4e9eff3450f9d5c31bd5bf3c299968002ed6b50bab6a7a9d4b76137fbc6
-size 5176
+oid sha256:dd6ab76cdb6290642263495ef9ccbac4bce2acbc40c68819c1dc3fde0b088978
+size 5240

training_artifacts/training_history.csv

@@ -1,12 +1,12 @@
 loss;grad_norm;learning_rate;epoch;step;eval_loss;eval_accuracy;eval_f1;eval_precision;eval_recall;eval_runtime;eval_samples_per_second;eval_steps_per_second;train_runtime;train_samples_per_second;train_steps_per_second;total_flos;train_loss
-0.6688;2.3227410316467285;3.3333333333333333e-06;0.2857142857142857;1;;;;;;;;;;;;;
-;;;0.2857142857142857;1;0.6694528460502625;0.7690288713910761;0.0;0.0;0.0;0.7051;540.382;34.04;;;;;
-;;;0.5714285714285714;2;0.6646609902381897;0.8110236220472441;0.0;0.0;0.0;0.7126;534.676;33.68;;;;;
-;;;0.8571428571428571;3;0.6550504565238953;0.863517060367454;0.0;0.0;0.0;2.6308;144.825;9.123;;;;;
-;;;1.1428571428571428;4;0.6407275199890137;0.9343832020997376;0.0;0.0;0.0;0.7121;535.065;33.705;;;;;
-;;;1.4285714285714286;5;0.6219562292098999;0.9658792650918635;0.0;0.0;0.0;0.7078;538.256;33.906;;;;;
-;;;1.7142857142857144;6;0.599317193031311;0.9763779527559056;0.0;0.0;0.0;1.1518;330.775;20.836;;;;;
-;;;2.0;7;0.5731121301651001;0.9816272965879265;0.0;0.0;0.0;0.7338;519.238;32.708;;;;;
-;;;2.2857142857142856;8;0.5478406548500061;0.9816272965879265;0.0;0.0;0.0;0.7067;539.157;33.963;;;;;
-;;;2.571428571428571;9;0.5236592292785645;0.9816272965879265;0.0;0.0;0.0;0.7083;537.937;33.886;;;;;
-;;;2.571428571428571;9;;;;;;;;;18.534;228.769;3.237;2148246430560.0;0.6554172303941515
+0.7327;2.7726080417633057;3.3333333333333333e-06;0.2857142857142857;1;;;;;;;;;;;;;
+;;;0.2857142857142857;1;0.7512667775154114;0.03412073490813648;0.03664921465968587;0.018666666666666668;1.0;5.3956;70.613;4.448;;;;;
+;;;0.5714285714285714;2;0.7464109659194946;0.05249343832020997;0.037333333333333336;0.019021739130434784;1.0;5.5933;68.117;4.291;;;;;
+;;;0.8571428571428571;3;0.7368069291114807;0.08923884514435695;0.038781163434903045;0.01977401129943503;1.0;5.5621;68.5;4.315;;;;;
+;;;1.1428571428571428;4;0.7224730849266052;0.19160104986876642;0.03749999999999999;0.019169329073482427;0.8571428571428571;5.6279;67.698;4.264;;;;;
+;;;1.4285714285714286;5;0.7036912441253662;0.3884514435695538;0.02510460251046025;0.01293103448275862;0.42857142857142855;5.703;66.807;4.208;;;;;
+;;;1.7142857142857144;6;0.6806334853172302;0.6325459317585301;0.0;0.0;0.0;5.773;65.997;4.157;;;;;
+;;;2.0;7;0.6538448929786682;0.868766404199475;0.0;0.0;0.0;5.6817;67.057;4.224;;;;;
+;;;2.2857142857142856;8;0.6274169683456421;0.973753280839895;0.0;0.0;0.0;5.8624;64.991;4.094;;;;;
+;;;2.571428571428571;9;0.6018883585929871;0.9816272965879265;0.0;0.0;0.0;5.644;67.505;4.252;;;;;
+;;;2.571428571428571;9;;;;;;;;;101.6832;41.698;0.59;2156353020864.0;0.6984957986407809

training_artifacts/training_history.json

@@ -1,125 +1,125 @@
 [
   {
-    "loss": 0.6688,
-    "grad_norm": 2.3227410316467285,
+    "loss": 0.7327,
+    "grad_norm": 2.7726080417633057,
     "learning_rate": 3.3333333333333333e-06,
     "epoch": 0.2857142857142857,
     "step": 1
   },
   {
-    "eval_loss": 0.6694528460502625,
-    "eval_accuracy": 0.7690288713910761,
-    "eval_f1": 0.0,
-    "eval_precision": 0.0,
-    "eval_recall": 0.0,
-    "eval_runtime": 0.7051,
-    "eval_samples_per_second": 540.382,
-    "eval_steps_per_second": 34.04,
+    "eval_loss": 0.7512667775154114,
+    "eval_accuracy": 0.03412073490813648,
+    "eval_f1": 0.03664921465968587,
+    "eval_precision": 0.018666666666666668,
+    "eval_recall": 1.0,
+    "eval_runtime": 5.3956,
+    "eval_samples_per_second": 70.613,
+    "eval_steps_per_second": 4.448,
     "epoch": 0.2857142857142857,
     "step": 1
   },
   {
-    "eval_loss": 0.6646609902381897,
-    "eval_accuracy": 0.8110236220472441,
-    "eval_f1": 0.0,
-    "eval_precision": 0.0,
-    "eval_recall": 0.0,
-    "eval_runtime": 0.7126,
-    "eval_samples_per_second": 534.676,
-    "eval_steps_per_second": 33.68,
+    "eval_loss": 0.7464109659194946,
+    "eval_accuracy": 0.05249343832020997,
+    "eval_f1": 0.037333333333333336,
+    "eval_precision": 0.019021739130434784,
+    "eval_recall": 1.0,
+    "eval_runtime": 5.5933,
+    "eval_samples_per_second": 68.117,
+    "eval_steps_per_second": 4.291,
     "epoch": 0.5714285714285714,
     "step": 2
   },
   {
-    "eval_loss": 0.6550504565238953,
-    "eval_accuracy": 0.863517060367454,
-    "eval_f1": 0.0,
-    "eval_precision": 0.0,
-    "eval_recall": 0.0,
-    "eval_runtime": 2.6308,
-    "eval_samples_per_second": 144.825,
-    "eval_steps_per_second": 9.123,
+    "eval_loss": 0.7368069291114807,
+    "eval_accuracy": 0.08923884514435695,
+    "eval_f1": 0.038781163434903045,
+    "eval_precision": 0.01977401129943503,
+    "eval_recall": 1.0,
+    "eval_runtime": 5.5621,
+    "eval_samples_per_second": 68.5,
+    "eval_steps_per_second": 4.315,
     "epoch": 0.8571428571428571,
     "step": 3
   },
   {
-    "eval_loss": 0.6407275199890137,
-    "eval_accuracy": 0.9343832020997376,
-    "eval_f1": 0.0,
-    "eval_precision": 0.0,
-    "eval_recall": 0.0,
-    "eval_runtime": 0.7121,
-    "eval_samples_per_second": 535.065,
-    "eval_steps_per_second": 33.705,
+    "eval_loss": 0.7224730849266052,
+    "eval_accuracy": 0.19160104986876642,
+    "eval_f1": 0.03749999999999999,
+    "eval_precision": 0.019169329073482427,
+    "eval_recall": 0.8571428571428571,
+    "eval_runtime": 5.6279,
+    "eval_samples_per_second": 67.698,
+    "eval_steps_per_second": 4.264,
     "epoch": 1.1428571428571428,
     "step": 4
   },
   {
-    "eval_loss": 0.6219562292098999,
-    "eval_accuracy": 0.9658792650918635,
-    "eval_f1": 0.0,
-    "eval_precision": 0.0,
-    "eval_recall": 0.0,
-    "eval_runtime": 0.7078,
-    "eval_samples_per_second": 538.256,
-    "eval_steps_per_second": 33.906,
+    "eval_loss": 0.7036912441253662,
+    "eval_accuracy": 0.3884514435695538,
+    "eval_f1": 0.02510460251046025,
+    "eval_precision": 0.01293103448275862,
+    "eval_recall": 0.42857142857142855,
+    "eval_runtime": 5.703,
+    "eval_samples_per_second": 66.807,
+    "eval_steps_per_second": 4.208,
     "epoch": 1.4285714285714286,
     "step": 5
   },
   {
-    "eval_loss": 0.599317193031311,
-    "eval_accuracy": 0.9763779527559056,
+    "eval_loss": 0.6806334853172302,
+    "eval_accuracy": 0.6325459317585301,
     "eval_f1": 0.0,
     "eval_precision": 0.0,
     "eval_recall": 0.0,
-    "eval_runtime": 1.1518,
-    "eval_samples_per_second": 330.775,
-    "eval_steps_per_second": 20.836,
+    "eval_runtime": 5.773,
+    "eval_samples_per_second": 65.997,
+    "eval_steps_per_second": 4.157,
     "epoch": 1.7142857142857144,
     "step": 6
   },
   {
-    "eval_loss": 0.5731121301651001,
-    "eval_accuracy": 0.9816272965879265,
+    "eval_loss": 0.6538448929786682,
+    "eval_accuracy": 0.868766404199475,
     "eval_f1": 0.0,
     "eval_precision": 0.0,
     "eval_recall": 0.0,
-    "eval_runtime": 0.7338,
-    "eval_samples_per_second": 519.238,
-    "eval_steps_per_second": 32.708,
+    "eval_runtime": 5.6817,
+    "eval_samples_per_second": 67.057,
+    "eval_steps_per_second": 4.224,
     "epoch": 2.0,
     "step": 7
   },
   {
-    "eval_loss": 0.5478406548500061,
-    "eval_accuracy": 0.9816272965879265,
+    "eval_loss": 0.6274169683456421,
+    "eval_accuracy": 0.973753280839895,
     "eval_f1": 0.0,
     "eval_precision": 0.0,
     "eval_recall": 0.0,
-    "eval_runtime": 0.7067,
-    "eval_samples_per_second": 539.157,
-    "eval_steps_per_second": 33.963,
+    "eval_runtime": 5.8624,
+    "eval_samples_per_second": 64.991,
+    "eval_steps_per_second": 4.094,
     "epoch": 2.2857142857142856,
     "step": 8
   },
   {
-    "eval_loss": 0.5236592292785645,
+    "eval_loss": 0.6018883585929871,
     "eval_accuracy": 0.9816272965879265,
     "eval_f1": 0.0,
     "eval_precision": 0.0,
     "eval_recall": 0.0,
-    "eval_runtime": 0.7083,
-    "eval_samples_per_second": 537.937,
-    "eval_steps_per_second": 33.886,
+    "eval_runtime": 5.644,
+    "eval_samples_per_second": 67.505,
+    "eval_steps_per_second": 4.252,
     "epoch": 2.571428571428571,
     "step": 9
   },
   {
-    "train_runtime": 18.534,
-    "train_samples_per_second": 228.769,
-    "train_steps_per_second": 3.237,
-    "total_flos": 2148246430560.0,
-    "train_loss": 0.6554172303941515,
+    "train_runtime": 101.6832,
+    "train_samples_per_second": 41.698,
+    "train_steps_per_second": 0.59,
+    "total_flos": 2156353020864.0,
+    "train_loss": 0.6984957986407809,
     "epoch": 2.571428571428571,
     "step": 9
   }

training_artifacts/training_summary.json

@@ -1,7 +1,7 @@
 {
   "total_epochs": 2.571428571428571,
   "total_steps": "9",
-  "final_train_loss": 0.6688,
-  "final_eval_loss": 0.5236592292785645,
-  "best_eval_loss": 0.5236592292785645
+  "final_train_loss": 0.7327,
+  "final_eval_loss": 0.6018883585929871,
+  "best_eval_loss": 0.6018883585929871
 }