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: 0.5703
+- Loss: 0.5776
 - Accuracy: 0.9816
 - F1: 0.0
 - Precision: 0.0
@@ -60,15 +60,15 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss | Accuracy | F1     | Precision | Recall |
 |:-------------:|:------:|:----:|:---------------:|:--------:|:------:|:---------:|:------:|
-| 0.703         | 0.2857 | 1    | 0.7166          | 0.2310   | 0.0456 | 0.0233    | 1.0    |
-| 0.703         | 0.5714 | 2    | 0.7117          | 0.2677   | 0.0412 | 0.0211    | 0.8571 |
-| 0.703         | 0.8571 | 3    | 0.7020          | 0.3701   | 0.04   | 0.0206    | 0.7143 |
-| 0.703         | 1.1429 | 4    | 0.6875          | 0.5774   | 0.0359 | 0.0187    | 0.4286 |
-| 0.703         | 1.4286 | 5    | 0.6686          | 0.7874   | 0.0    | 0.0       | 0.0    |
-| 0.703         | 1.7143 | 6    | 0.6458          | 0.9423   | 0.0    | 0.0       | 0.0    |
-| 0.703         | 2.0    | 7    | 0.6198          | 0.9816   | 0.0    | 0.0       | 0.0    |
-| 0.703         | 2.2857 | 8    | 0.5946          | 0.9816   | 0.0    | 0.0       | 0.0    |
-| 0.703         | 2.5714 | 9    | 0.5703          | 0.9816   | 0.0    | 0.0       | 0.0    |
+| 0.7155        | 0.2857 | 1    | 0.7256          | 0.1969   | 0.0437 | 0.0224    | 1.0    |
+| 0.7155        | 0.5714 | 2    | 0.7204          | 0.2283   | 0.0455 | 0.0233    | 1.0    |
+| 0.7155        | 0.8571 | 3    | 0.7102          | 0.3333   | 0.0451 | 0.0232    | 0.8571 |
+| 0.7155        | 1.1429 | 4    | 0.6954          | 0.4803   | 0.0388 | 0.0201    | 0.5714 |
+| 0.7155        | 1.4286 | 5    | 0.6763          | 0.7008   | 0.0500 | 0.0265    | 0.4286 |
+| 0.7155        | 1.7143 | 6    | 0.6533          | 0.8530   | 0.0345 | 0.0196    | 0.1429 |
+| 0.7155        | 2.0    | 7    | 0.6268          | 0.9580   | 0.0    | 0.0       | 0.0    |
+| 0.7155        | 2.2857 | 8    | 0.6016          | 0.9816   | 0.0    | 0.0       | 0.0    |
+| 0.7155        | 2.5714 | 9    | 0.5776          | 0.9816   | 0.0    | 0.0       | 0.0    |
 
 
 ### Framework versions

logs/events.out.tfevents.1759931840.ip-172-31-12-22.95749.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:844f0c7b61f3f9d66c329092db238d9f75b13a488fd448688ce4957589d57b67
-size 9102
+oid sha256:7b2daa8b29668f6084d35a0ec29df6b4c5c5644a20d5c2c0e6423a6dcfe1bd5f
+size 9913

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b1b8179425a4646feb09a1f50f2c11f78d1fd85bb7835d740a1e8a4f87a2f071
+oid sha256:3ac7f9b65871eb9359db4ad027a35c1da1b260202ec0695cb721ba7e7d007e98
 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
-0.703;2.2138702869415283;3.3333333333333333e-06;0.2857142857142857;1;;;;;;;;;;;;;
-;;;0.2857142857142857;1;0.7165892720222473;0.23097112860892388;0.04560260586319218;0.023333333333333334;1.0;0.7454;511.131;32.197;;;;;
-;;;0.5714285714285714;2;0.7117474675178528;0.2677165354330709;0.041237113402061855;0.02112676056338028;0.8571428571428571;1.5578;244.576;15.406;;;;;
-;;;0.8571428571428571;3;0.7019844055175781;0.3700787401574803;0.04;0.0205761316872428;0.7142857142857143;2.8847;132.075;8.32;;;;;
-;;;1.1428571428571428;4;0.6875269412994385;0.5774278215223098;0.03592814371257485;0.01875;0.42857142857142855;2.1358;178.386;11.237;;;;;
-;;;1.4285714285714286;5;0.6685800552368164;0.7874015748031497;0.0;0.0;0.0;0.7158;532.274;33.529;;;;;
-;;;1.7142857142857144;6;0.6458192467689514;0.9422572178477691;0.0;0.0;0.0;1.2139;313.874;19.772;;;;;
-;;;2.0;7;0.6197686195373535;0.9816272965879265;0.0;0.0;0.0;0.7728;493.017;31.056;;;;;
-;;;2.2857142857142856;8;0.5946394205093384;0.9816272965879265;0.0;0.0;0.0;0.7291;522.579;32.918;;;;;
-;;;2.571428571428571;9;0.5703445672988892;0.9816272965879265;0.0;0.0;0.0;0.9339;407.972;25.699;;;;;
-;;;2.571428571428571;9;;;;;;;;;22.1708;191.242;2.706;2172566201472.0;0.6836379369099935
+0.7155;3.170656681060791;3.3333333333333333e-06;0.2857142857142857;1;;;;;;;;;;;;;
+;;;0.2857142857142857;1;0.7255510687828064;0.1968503937007874;0.04375;0.022364217252396165;1.0;7.3525;51.819;3.264;;;;;
+;;;0.5714285714285714;2;0.7203730940818787;0.2283464566929134;0.04545454545454545;0.023255813953488372;1.0;7.5704;50.328;3.17;;;;;
+;;;0.8571428571428571;3;0.7101789712905884;0.3333333333333333;0.04511278195488722;0.023166023166023165;0.8571428571428571;7.7608;49.093;3.092;;;;;
+;;;1.1428571428571428;4;0.6954057216644287;0.48031496062992124;0.038834951456310676;0.020100502512562814;0.5714285714285714;7.798;48.859;3.078;;;;;
+;;;1.4285714285714286;5;0.6762979626655579;0.7007874015748031;0.049999999999999996;0.02654867256637168;0.42857142857142855;8.2567;46.144;2.907;;;;;
+;;;1.7142857142857144;6;0.6532518267631531;0.8530183727034121;0.034482758620689655;0.0196078431372549;0.14285714285714285;7.127;53.459;3.367;;;;;
+;;;2.0;7;0.6268358826637268;0.958005249343832;0.0;0.0;0.0;7.3211;52.041;3.278;;;;;
+;;;2.2857142857142856;8;0.6016007661819458;0.9816272965879265;0.0;0.0;0.0;8.4314;45.188;2.847;;;;;
+;;;2.571428571428571;9;0.5776201486587524;0.9816272965879265;0.0;0.0;0.0;7.5633;50.375;3.173;;;;;
+;;;2.571428571428571;9;;;;;;;;;133.4632;31.769;0.45;2156353020864.0;0.68766188621521

training_artifacts/training_history.json

@@ -1,125 +1,125 @@
 [
   {
-    "loss": 0.703,
-    "grad_norm": 2.2138702869415283,
+    "loss": 0.7155,
+    "grad_norm": 3.170656681060791,
     "learning_rate": 3.3333333333333333e-06,
     "epoch": 0.2857142857142857,
     "step": 1
   },
   {
-    "eval_loss": 0.7165892720222473,
-    "eval_accuracy": 0.23097112860892388,
-    "eval_f1": 0.04560260586319218,
-    "eval_precision": 0.023333333333333334,
+    "eval_loss": 0.7255510687828064,
+    "eval_accuracy": 0.1968503937007874,
+    "eval_f1": 0.04375,
+    "eval_precision": 0.022364217252396165,
     "eval_recall": 1.0,
-    "eval_runtime": 0.7454,
-    "eval_samples_per_second": 511.131,
-    "eval_steps_per_second": 32.197,
+    "eval_runtime": 7.3525,
+    "eval_samples_per_second": 51.819,
+    "eval_steps_per_second": 3.264,
     "epoch": 0.2857142857142857,
     "step": 1
   },
   {
-    "eval_loss": 0.7117474675178528,
-    "eval_accuracy": 0.2677165354330709,
-    "eval_f1": 0.041237113402061855,
-    "eval_precision": 0.02112676056338028,
-    "eval_recall": 0.8571428571428571,
-    "eval_runtime": 1.5578,
-    "eval_samples_per_second": 244.576,
-    "eval_steps_per_second": 15.406,
+    "eval_loss": 0.7203730940818787,
+    "eval_accuracy": 0.2283464566929134,
+    "eval_f1": 0.04545454545454545,
+    "eval_precision": 0.023255813953488372,
+    "eval_recall": 1.0,
+    "eval_runtime": 7.5704,
+    "eval_samples_per_second": 50.328,
+    "eval_steps_per_second": 3.17,
     "epoch": 0.5714285714285714,
     "step": 2
   },
   {
-    "eval_loss": 0.7019844055175781,
-    "eval_accuracy": 0.3700787401574803,
-    "eval_f1": 0.04,
-    "eval_precision": 0.0205761316872428,
-    "eval_recall": 0.7142857142857143,
-    "eval_runtime": 2.8847,
-    "eval_samples_per_second": 132.075,
-    "eval_steps_per_second": 8.32,
+    "eval_loss": 0.7101789712905884,
+    "eval_accuracy": 0.3333333333333333,
+    "eval_f1": 0.04511278195488722,
+    "eval_precision": 0.023166023166023165,
+    "eval_recall": 0.8571428571428571,
+    "eval_runtime": 7.7608,
+    "eval_samples_per_second": 49.093,
+    "eval_steps_per_second": 3.092,
     "epoch": 0.8571428571428571,
     "step": 3
   },
   {
-    "eval_loss": 0.6875269412994385,
-    "eval_accuracy": 0.5774278215223098,
-    "eval_f1": 0.03592814371257485,
-    "eval_precision": 0.01875,
-    "eval_recall": 0.42857142857142855,
-    "eval_runtime": 2.1358,
-    "eval_samples_per_second": 178.386,
-    "eval_steps_per_second": 11.237,
+    "eval_loss": 0.6954057216644287,
+    "eval_accuracy": 0.48031496062992124,
+    "eval_f1": 0.038834951456310676,
+    "eval_precision": 0.020100502512562814,
+    "eval_recall": 0.5714285714285714,
+    "eval_runtime": 7.798,
+    "eval_samples_per_second": 48.859,
+    "eval_steps_per_second": 3.078,
     "epoch": 1.1428571428571428,
     "step": 4
   },
   {
-    "eval_loss": 0.6685800552368164,
-    "eval_accuracy": 0.7874015748031497,
-    "eval_f1": 0.0,
-    "eval_precision": 0.0,
-    "eval_recall": 0.0,
-    "eval_runtime": 0.7158,
-    "eval_samples_per_second": 532.274,
-    "eval_steps_per_second": 33.529,
+    "eval_loss": 0.6762979626655579,
+    "eval_accuracy": 0.7007874015748031,
+    "eval_f1": 0.049999999999999996,
+    "eval_precision": 0.02654867256637168,
+    "eval_recall": 0.42857142857142855,
+    "eval_runtime": 8.2567,
+    "eval_samples_per_second": 46.144,
+    "eval_steps_per_second": 2.907,
     "epoch": 1.4285714285714286,
     "step": 5
   },
   {
-    "eval_loss": 0.6458192467689514,
-    "eval_accuracy": 0.9422572178477691,
-    "eval_f1": 0.0,
-    "eval_precision": 0.0,
-    "eval_recall": 0.0,
-    "eval_runtime": 1.2139,
-    "eval_samples_per_second": 313.874,
-    "eval_steps_per_second": 19.772,
+    "eval_loss": 0.6532518267631531,
+    "eval_accuracy": 0.8530183727034121,
+    "eval_f1": 0.034482758620689655,
+    "eval_precision": 0.0196078431372549,
+    "eval_recall": 0.14285714285714285,
+    "eval_runtime": 7.127,
+    "eval_samples_per_second": 53.459,
+    "eval_steps_per_second": 3.367,
     "epoch": 1.7142857142857144,
     "step": 6
   },
   {
-    "eval_loss": 0.6197686195373535,
-    "eval_accuracy": 0.9816272965879265,
+    "eval_loss": 0.6268358826637268,
+    "eval_accuracy": 0.958005249343832,
     "eval_f1": 0.0,
     "eval_precision": 0.0,
     "eval_recall": 0.0,
-    "eval_runtime": 0.7728,
-    "eval_samples_per_second": 493.017,
-    "eval_steps_per_second": 31.056,
+    "eval_runtime": 7.3211,
+    "eval_samples_per_second": 52.041,
+    "eval_steps_per_second": 3.278,
     "epoch": 2.0,
     "step": 7
   },
   {
-    "eval_loss": 0.5946394205093384,
+    "eval_loss": 0.6016007661819458,
     "eval_accuracy": 0.9816272965879265,
     "eval_f1": 0.0,
     "eval_precision": 0.0,
     "eval_recall": 0.0,
-    "eval_runtime": 0.7291,
-    "eval_samples_per_second": 522.579,
-    "eval_steps_per_second": 32.918,
+    "eval_runtime": 8.4314,
+    "eval_samples_per_second": 45.188,
+    "eval_steps_per_second": 2.847,
     "epoch": 2.2857142857142856,
     "step": 8
   },
   {
-    "eval_loss": 0.5703445672988892,
+    "eval_loss": 0.5776201486587524,
     "eval_accuracy": 0.9816272965879265,
     "eval_f1": 0.0,
     "eval_precision": 0.0,
     "eval_recall": 0.0,
-    "eval_runtime": 0.9339,
-    "eval_samples_per_second": 407.972,
-    "eval_steps_per_second": 25.699,
+    "eval_runtime": 7.5633,
+    "eval_samples_per_second": 50.375,
+    "eval_steps_per_second": 3.173,
     "epoch": 2.571428571428571,
     "step": 9
   },
   {
-    "train_runtime": 22.1708,
-    "train_samples_per_second": 191.242,
-    "train_steps_per_second": 2.706,
-    "total_flos": 2172566201472.0,
-    "train_loss": 0.6836379369099935,
+    "train_runtime": 133.4632,
+    "train_samples_per_second": 31.769,
+    "train_steps_per_second": 0.45,
+    "total_flos": 2156353020864.0,
+    "train_loss": 0.68766188621521,
     "epoch": 2.571428571428571,
     "step": 9
   }

training_artifacts/training_summary.json

@@ -1,7 +1,7 @@
 {
   "total_epochs": 2.571428571428571,
   "total_steps": "9",
-  "final_train_loss": 0.703,
-  "final_eval_loss": 0.5703445672988892,
-  "best_eval_loss": 0.5703445672988892
+  "final_train_loss": 0.7155,
+  "final_eval_loss": 0.5776201486587524,
+  "best_eval_loss": 0.5776201486587524
 }