Model save

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: 0.6019
+- Loss: 0.5007
 - 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 |
-|:-------------:|:------:|:----:|:---------------:|:--------:|:------:|:---------:|:------:|
-| 0.7327        | 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    |
+| Training Loss | Epoch  | Step | Validation Loss | Accuracy | F1  | Precision | Recall |
+|:-------------:|:------:|:----:|:---------------:|:--------:|:---:|:---------:|:------:|
+| 0.6651        | 0.2857 | 1    | 0.6412          | 0.9003   | 0.0 | 0.0       | 0.0    |
+| 0.6651        | 0.5714 | 2    | 0.6363          | 0.9239   | 0.0 | 0.0       | 0.0    |
+| 0.6651        | 0.8571 | 3    | 0.6264          | 0.9475   | 0.0 | 0.0       | 0.0    |
+| 0.6651        | 1.1429 | 4    | 0.6125          | 0.9606   | 0.0 | 0.0       | 0.0    |
+| 0.6651        | 1.4286 | 5    | 0.5944          | 0.9685   | 0.0 | 0.0       | 0.0    |
+| 0.6651        | 1.7143 | 6    | 0.5722          | 0.9738   | 0.0 | 0.0       | 0.0    |
+| 0.6651        | 2.0    | 7    | 0.5467          | 0.9790   | 0.0 | 0.0       | 0.0    |
+| 0.6651        | 2.2857 | 8    | 0.5230          | 0.9816   | 0.0 | 0.0       | 0.0    |
+| 0.6651        | 2.5714 | 9    | 0.5007          | 0.9816   | 0.0 | 0.0       | 0.0    |
 
 
 ### Framework versions

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 sha256:6725d7513e0d6bed9233b2693f6fdf0f3eb35ffea312bd982e7d86e7b9efe1a8
-size 9102
+oid sha256:a11e83f00225c05834b318ee327c15a0a7afbf22b06b17388b80551fff119d6c
+size 9913

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9973c72a478da0220790b93125c8a455209f350fceb5c666924c1784696ae544
+oid sha256:c270f4d5425845329b57b5a00429e669c0149bcf21a9e0c8303b71b8aa504852
 size 46634832

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
-0.7327;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
+0.6651;2.0492517948150635;3.3333333333333333e-06;0.2857142857142857;1;;;;;;;;;;;;;
+;;;0.2857142857142857;1;0.641178548336029;0.9002624671916011;0.0;0.0;0.0;5.428;70.191;4.422;;;;;
+;;;0.5714285714285714;2;0.6362672448158264;0.9238845144356955;0.0;0.0;0.0;5.345;71.282;4.49;;;;;
+;;;0.8571428571428571;3;0.6264415383338928;0.94750656167979;0.0;0.0;0.0;5.2138;73.076;4.603;;;;;
+;;;1.1428571428571428;4;0.6125248074531555;0.9606299212598425;0.0;0.0;0.0;5.2287;72.867;4.59;;;;;
+;;;1.4285714285714286;5;0.5944115519523621;0.968503937007874;0.0;0.0;0.0;5.3131;71.709;4.517;;;;;
+;;;1.7142857142857144;6;0.5722081661224365;0.973753280839895;0.0;0.0;0.0;5.3197;71.62;4.512;;;;;
+;;;2.0;7;0.5467000603675842;0.979002624671916;0.0;0.0;0.0;5.2338;72.796;4.586;;;;;
+;;;2.2857142857142856;8;0.5230435729026794;0.9816272965879265;0.0;0.0;0.0;5.3484;71.237;4.487;;;;;
+;;;2.571428571428571;9;0.5007050633430481;0.9816272965879265;0.0;0.0;0.0;5.4954;69.331;4.367;;;;;
+;;;2.571428571428571;9;;;;;;;;;92.7268;45.726;0.647;2140139840256.0;0.6428146097395155

training_artifacts/training_history.json

@@ -1,125 +1,125 @@
 [
   {
-    "loss": 0.7327,
-    "grad_norm": 2.7726080417633057,
+    "loss": 0.6651,
+    "grad_norm": 2.0492517948150635,
     "learning_rate": 3.3333333333333333e-06,
     "epoch": 0.2857142857142857,
     "step": 1
   },
   {
-    "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,
+    "eval_loss": 0.641178548336029,
+    "eval_accuracy": 0.9002624671916011,
+    "eval_f1": 0.0,
+    "eval_precision": 0.0,
+    "eval_recall": 0.0,
+    "eval_runtime": 5.428,
+    "eval_samples_per_second": 70.191,
+    "eval_steps_per_second": 4.422,
     "epoch": 0.2857142857142857,
     "step": 1
   },
   {
-    "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,
+    "eval_loss": 0.6362672448158264,
+    "eval_accuracy": 0.9238845144356955,
+    "eval_f1": 0.0,
+    "eval_precision": 0.0,
+    "eval_recall": 0.0,
+    "eval_runtime": 5.345,
+    "eval_samples_per_second": 71.282,
+    "eval_steps_per_second": 4.49,
     "epoch": 0.5714285714285714,
     "step": 2
   },
   {
-    "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,
+    "eval_loss": 0.6264415383338928,
+    "eval_accuracy": 0.94750656167979,
+    "eval_f1": 0.0,
+    "eval_precision": 0.0,
+    "eval_recall": 0.0,
+    "eval_runtime": 5.2138,
+    "eval_samples_per_second": 73.076,
+    "eval_steps_per_second": 4.603,
     "epoch": 0.8571428571428571,
     "step": 3
   },
   {
-    "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,
+    "eval_loss": 0.6125248074531555,
+    "eval_accuracy": 0.9606299212598425,
+    "eval_f1": 0.0,
+    "eval_precision": 0.0,
+    "eval_recall": 0.0,
+    "eval_runtime": 5.2287,
+    "eval_samples_per_second": 72.867,
+    "eval_steps_per_second": 4.59,
     "epoch": 1.1428571428571428,
     "step": 4
   },
   {
-    "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,
+    "eval_loss": 0.5944115519523621,
+    "eval_accuracy": 0.968503937007874,
+    "eval_f1": 0.0,
+    "eval_precision": 0.0,
+    "eval_recall": 0.0,
+    "eval_runtime": 5.3131,
+    "eval_samples_per_second": 71.709,
+    "eval_steps_per_second": 4.517,
     "epoch": 1.4285714285714286,
     "step": 5
   },
   {
-    "eval_loss": 0.6806334853172302,
-    "eval_accuracy": 0.6325459317585301,
+    "eval_loss": 0.5722081661224365,
+    "eval_accuracy": 0.973753280839895,
     "eval_f1": 0.0,
     "eval_precision": 0.0,
     "eval_recall": 0.0,
-    "eval_runtime": 5.773,
-    "eval_samples_per_second": 65.997,
-    "eval_steps_per_second": 4.157,
+    "eval_runtime": 5.3197,
+    "eval_samples_per_second": 71.62,
+    "eval_steps_per_second": 4.512,
     "epoch": 1.7142857142857144,
     "step": 6
   },
   {
-    "eval_loss": 0.6538448929786682,
-    "eval_accuracy": 0.868766404199475,
+    "eval_loss": 0.5467000603675842,
+    "eval_accuracy": 0.979002624671916,
     "eval_f1": 0.0,
     "eval_precision": 0.0,
     "eval_recall": 0.0,
-    "eval_runtime": 5.6817,
-    "eval_samples_per_second": 67.057,
-    "eval_steps_per_second": 4.224,
+    "eval_runtime": 5.2338,
+    "eval_samples_per_second": 72.796,
+    "eval_steps_per_second": 4.586,
     "epoch": 2.0,
     "step": 7
   },
   {
-    "eval_loss": 0.6274169683456421,
-    "eval_accuracy": 0.973753280839895,
+    "eval_loss": 0.5230435729026794,
+    "eval_accuracy": 0.9816272965879265,
     "eval_f1": 0.0,
     "eval_precision": 0.0,
     "eval_recall": 0.0,
-    "eval_runtime": 5.8624,
-    "eval_samples_per_second": 64.991,
-    "eval_steps_per_second": 4.094,
+    "eval_runtime": 5.3484,
+    "eval_samples_per_second": 71.237,
+    "eval_steps_per_second": 4.487,
     "epoch": 2.2857142857142856,
     "step": 8
   },
   {
-    "eval_loss": 0.6018883585929871,
+    "eval_loss": 0.5007050633430481,
     "eval_accuracy": 0.9816272965879265,
     "eval_f1": 0.0,
     "eval_precision": 0.0,
     "eval_recall": 0.0,
-    "eval_runtime": 5.644,
-    "eval_samples_per_second": 67.505,
-    "eval_steps_per_second": 4.252,
+    "eval_runtime": 5.4954,
+    "eval_samples_per_second": 69.331,
+    "eval_steps_per_second": 4.367,
     "epoch": 2.571428571428571,
     "step": 9
   },
   {
-    "train_runtime": 101.6832,
-    "train_samples_per_second": 41.698,
-    "train_steps_per_second": 0.59,
-    "total_flos": 2156353020864.0,
-    "train_loss": 0.6984957986407809,
+    "train_runtime": 92.7268,
+    "train_samples_per_second": 45.726,
+    "train_steps_per_second": 0.647,
+    "total_flos": 2140139840256.0,
+    "train_loss": 0.6428146097395155,
     "epoch": 2.571428571428571,
     "step": 9
   }

training_artifacts/training_summary.json

@@ -1,7 +1,7 @@
 {
   "total_epochs": 2.571428571428571,
   "total_steps": "9",
-  "final_train_loss": 0.7327,
-  "final_eval_loss": 0.6018883585929871,
-  "best_eval_loss": 0.6018883585929871
+  "final_train_loss": 0.6651,
+  "final_eval_loss": 0.5007050633430481,
+  "best_eval_loss": 0.5007050633430481
 }