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: 0.5492
 - 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    |
+| 0.689         | 0.2857 | 1    | 0.6890          | 0.5564   | 0.0231 | 0.0120    | 0.2857 |
+| 0.689         | 0.5714 | 2    | 0.6843          | 0.6352   | 0.0    | 0.0       | 0.0    |
+| 0.689         | 0.8571 | 3    | 0.6749          | 0.7507   | 0.0    | 0.0       | 0.0    |
+| 0.689         | 1.1429 | 4    | 0.6614          | 0.8688   | 0.0    | 0.0       | 0.0    |
+| 0.689         | 1.4286 | 5    | 0.6435          | 0.9475   | 0.0    | 0.0       | 0.0    |
+| 0.689         | 1.7143 | 6    | 0.6220          | 0.9816   | 0.0    | 0.0       | 0.0    |
+| 0.689         | 2.0    | 7    | 0.5971          | 0.9816   | 0.0    | 0.0       | 0.0    |
+| 0.689         | 2.2857 | 8    | 0.5728          | 0.9816   | 0.0    | 0.0       | 0.0    |
+| 0.689         | 2.5714 | 9    | 0.5492          | 0.9816   | 0.0    | 0.0       | 0.0    |
 
 
 ### Framework versions

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f163bfefd249064226a7fa02fe22dde624b62e553bf4ba63375bfb00a734a92d
+oid sha256:140cdfee2958f1b237065f5364a532a0d8ad78e4847d35bdf445438cf1e09c69
 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
+0.689;2.7823822498321533;3.3333333333333333e-06;0.2857142857142857;1;;;;;;;;;;;;;
+;;;0.2857142857142857;1;0.6890391707420349;0.5564304461942258;0.023121387283236997;0.012048192771084338;0.2857142857142857;0.7451;511.33;32.21;;;;;
+;;;0.5714285714285714;2;0.6843268275260925;0.6351706036745407;0.0;0.0;0.0;0.7197;529.417;33.349;;;;;
+;;;0.8571428571428571;3;0.6749046444892883;0.7506561679790026;0.0;0.0;0.0;3.3581;113.455;7.147;;;;;
+;;;1.1428571428571428;4;0.6614006161689758;0.868766404199475;0.0;0.0;0.0;0.7108;535.986;33.763;;;;;
+;;;1.4285714285714286;5;0.6434675455093384;0.94750656167979;0.0;0.0;0.0;1.1845;321.664;20.262;;;;;
+;;;1.7142857142857144;6;0.6219972372055054;0.9816272965879265;0.0;0.0;0.0;0.7216;528.026;33.261;;;;;
+;;;2.0;7;0.5970718264579773;0.9816272965879265;0.0;0.0;0.0;0.7206;528.749;33.307;;;;;
+;;;2.2857142857142856;8;0.5727897882461548;0.9816272965879265;0.0;0.0;0.0;0.7052;540.284;34.034;;;;;
+;;;2.571428571428571;9;0.549211323261261;0.9816272965879265;0.0;0.0;0.0;0.6987;545.308;34.35;;;;;
+;;;2.571428571428571;9;;;;;;;;;18.2317;232.562;3.291;2148246430560.0;0.6723573472764757

training_artifacts/training_history.json

@@ -1,125 +1,125 @@
 [
   {
-    "loss": 0.703,
-    "grad_norm": 2.2138702869415283,
+    "loss": 0.689,
+    "grad_norm": 2.7823822498321533,
     "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_recall": 1.0,
-    "eval_runtime": 0.7454,
-    "eval_samples_per_second": 511.131,
-    "eval_steps_per_second": 32.197,
+    "eval_loss": 0.6890391707420349,
+    "eval_accuracy": 0.5564304461942258,
+    "eval_f1": 0.023121387283236997,
+    "eval_precision": 0.012048192771084338,
+    "eval_recall": 0.2857142857142857,
+    "eval_runtime": 0.7451,
+    "eval_samples_per_second": 511.33,
+    "eval_steps_per_second": 32.21,
     "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": 0.6843268275260925,
+    "eval_accuracy": 0.6351706036745407,
+    "eval_f1": 0.0,
+    "eval_precision": 0.0,
+    "eval_recall": 0.0,
+    "eval_runtime": 0.7197,
+    "eval_samples_per_second": 529.417,
+    "eval_steps_per_second": 33.349,
     "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": 0.6749046444892883,
+    "eval_accuracy": 0.7506561679790026,
+    "eval_f1": 0.0,
+    "eval_precision": 0.0,
+    "eval_recall": 0.0,
+    "eval_runtime": 3.3581,
+    "eval_samples_per_second": 113.455,
+    "eval_steps_per_second": 7.147,
     "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": 0.6614006161689758,
+    "eval_accuracy": 0.868766404199475,
+    "eval_f1": 0.0,
+    "eval_precision": 0.0,
+    "eval_recall": 0.0,
+    "eval_runtime": 0.7108,
+    "eval_samples_per_second": 535.986,
+    "eval_steps_per_second": 33.763,
     "epoch": 1.1428571428571428,
     "step": 4
   },
   {
-    "eval_loss": 0.6685800552368164,
-    "eval_accuracy": 0.7874015748031497,
+    "eval_loss": 0.6434675455093384,
+    "eval_accuracy": 0.94750656167979,
     "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_runtime": 1.1845,
+    "eval_samples_per_second": 321.664,
+    "eval_steps_per_second": 20.262,
     "epoch": 1.4285714285714286,
     "step": 5
   },
   {
-    "eval_loss": 0.6458192467689514,
-    "eval_accuracy": 0.9422572178477691,
+    "eval_loss": 0.6219972372055054,
+    "eval_accuracy": 0.9816272965879265,
     "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_runtime": 0.7216,
+    "eval_samples_per_second": 528.026,
+    "eval_steps_per_second": 33.261,
     "epoch": 1.7142857142857144,
     "step": 6
   },
   {
-    "eval_loss": 0.6197686195373535,
+    "eval_loss": 0.5970718264579773,
     "eval_accuracy": 0.9816272965879265,
     "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": 0.7206,
+    "eval_samples_per_second": 528.749,
+    "eval_steps_per_second": 33.307,
     "epoch": 2.0,
     "step": 7
   },
   {
-    "eval_loss": 0.5946394205093384,
+    "eval_loss": 0.5727897882461548,
     "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": 0.7052,
+    "eval_samples_per_second": 540.284,
+    "eval_steps_per_second": 34.034,
     "epoch": 2.2857142857142856,
     "step": 8
   },
   {
-    "eval_loss": 0.5703445672988892,
+    "eval_loss": 0.549211323261261,
     "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": 0.6987,
+    "eval_samples_per_second": 545.308,
+    "eval_steps_per_second": 34.35,
     "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": 18.2317,
+    "train_samples_per_second": 232.562,
+    "train_steps_per_second": 3.291,
+    "total_flos": 2148246430560.0,
+    "train_loss": 0.6723573472764757,
     "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": 0.689,
+  "final_eval_loss": 0.549211323261261,
+  "best_eval_loss": 0.549211323261261
 }