mjphayes/vit-elpv-augmented

README.md

@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.7721
-- Accuracy: 0.6929
+- Loss: 0.7687
+- Accuracy: 0.7259
 
 ## Model description
 
@@ -46,27 +46,31 @@ The following hyperparameters were used during training:
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_ratio: 0.1
-- num_epochs: 10
+- num_epochs: 14
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Accuracy |
 |:-------------:|:-----:|:----:|:---------------:|:--------:|
-| No log        | 1.0   | 69   | 1.0376          | 0.5939   |
-| No log        | 2.0   | 138  | 0.9578          | 0.5914   |
-| No log        | 3.0   | 207  | 0.9083          | 0.6447   |
-| No log        | 4.0   | 276  | 0.8320          | 0.6751   |
-| No log        | 5.0   | 345  | 0.8832          | 0.6371   |
-| No log        | 6.0   | 414  | 0.8498          | 0.6574   |
-| No log        | 7.0   | 483  | 0.8124          | 0.6701   |
-| 0.7809        | 8.0   | 552  | 0.8627          | 0.6701   |
-| 0.7809        | 9.0   | 621  | 0.8003          | 0.6701   |
-| 0.7809        | 10.0  | 690  | 0.7721          | 0.6929   |
+| No log        | 1.0   | 69   | 1.1146          | 0.5787   |
+| No log        | 2.0   | 138  | 0.9812          | 0.5787   |
+| No log        | 3.0   | 207  | 0.8885          | 0.6472   |
+| No log        | 4.0   | 276  | 0.7930          | 0.7081   |
+| No log        | 5.0   | 345  | 0.8019          | 0.6929   |
+| No log        | 6.0   | 414  | 0.8009          | 0.6878   |
+| No log        | 7.0   | 483  | 0.7984          | 0.6853   |
+| 0.8194        | 8.0   | 552  | 0.7714          | 0.7107   |
+| 0.8194        | 9.0   | 621  | 0.7667          | 0.7081   |
+| 0.8194        | 10.0  | 690  | 0.7303          | 0.7234   |
+| 0.8194        | 11.0  | 759  | 0.7321          | 0.7284   |
+| 0.8194        | 12.0  | 828  | 0.7373          | 0.7335   |
+| 0.8194        | 13.0  | 897  | 0.8051          | 0.6904   |
+| 0.8194        | 14.0  | 966  | 0.7687          | 0.7259   |
 
 
 ### Framework versions
 
-- Transformers 4.35.0
-- Pytorch 2.1.0
-- Datasets 2.14.6
-- Tokenizers 0.14.1
+- Transformers 4.35.2
+- Pytorch 2.1.1+cu121
+- Datasets 2.15.0
+- Tokenizers 0.15.0

all_results.json

@@ -1,7 +1,8 @@
 {
-    "epoch": 10.0,
-    "train_loss": 0.7178522997889025,
-    "train_runtime": 2833.88,
-    "train_samples_per_second": 6.514,
-    "train_steps_per_second": 0.102
+    "epoch": 14.0,
+    "total_flos": 4.790950643943604e+18,
+    "train_loss": 0.63200661163646,
+    "train_runtime": 1457.5958,
+    "train_samples_per_second": 42.415,
+    "train_steps_per_second": 0.663
 }

config.json

@@ -32,5 +32,5 @@
   "problem_type": "single_label_classification",
   "qkv_bias": true,
   "torch_dtype": "float32",
-  "transformers_version": "4.35.0"
+  "transformers_version": "4.35.2"
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b82d916b8466c58a9e22666cb75066ac16bfe1547114c247120c5dcf6d4a7f83
+oid sha256:d29546920076aeb62358d5f3b4a15298c89fc5581b645c9cb4015ab1832aeedd
 size 343230128

runs/Nov16_09-36-14_nk7ic4m731/events.out.tfevents.1700127380.nk7ic4m731.228.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:328f29abe32d555dc51c43e7580c4157a946ef8373f06e5e3f38e381640c3278
+size 9262

train_results.json

@@ -1,7 +1,8 @@
 {
-    "epoch": 10.0,
-    "train_loss": 0.7178522997889025,
-    "train_runtime": 2833.88,
-    "train_samples_per_second": 6.514,
-    "train_steps_per_second": 0.102
+    "epoch": 14.0,
+    "total_flos": 4.790950643943604e+18,
+    "train_loss": 0.63200661163646,
+    "train_runtime": 1457.5958,
+    "train_samples_per_second": 42.415,
+    "train_steps_per_second": 0.663
 }

trainer_state.json

@@ -1,292 +1,160 @@
 {
-  "best_metric": 0.7518796992481203,
-  "best_model_checkpoint": "../results/elpv-vit/checkpoint-232",
-  "epoch": 10.0,
+  "best_metric": 0.733502538071066,
+  "best_model_checkpoint": "../results/elpv-vit/checkpoint-828",
+  "epoch": 14.0,
   "eval_steps": 500,
-  "global_step": 290,
+  "global_step": 966,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
-    {
-      "epoch": 0.34,
-      "learning_rate": 1.7241379310344828e-05,
-      "loss": 1.3357,
-      "step": 10
-    },
-    {
-      "epoch": 0.69,
-      "learning_rate": 3.4482758620689657e-05,
-      "loss": 1.1598,
-      "step": 20
-    },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.6954887218045113,
-      "eval_loss": 0.9091870188713074,
-      "eval_runtime": 10.3022,
-      "eval_samples_per_second": 25.82,
-      "eval_steps_per_second": 1.65,
-      "step": 29
-    },
-    {
-      "epoch": 1.03,
-      "learning_rate": 4.980842911877395e-05,
-      "loss": 0.9746,
-      "step": 30
-    },
-    {
-      "epoch": 1.38,
-      "learning_rate": 4.789272030651341e-05,
-      "loss": 0.8597,
-      "step": 40
-    },
-    {
-      "epoch": 1.72,
-      "learning_rate": 4.597701149425287e-05,
-      "loss": 0.8226,
-      "step": 50
+      "eval_accuracy": 0.5786802030456852,
+      "eval_loss": 1.1146228313446045,
+      "eval_runtime": 15.1595,
+      "eval_samples_per_second": 25.99,
+      "eval_steps_per_second": 1.649,
+      "step": 69
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.7368421052631579,
-      "eval_loss": 0.7852613925933838,
-      "eval_runtime": 10.1106,
-      "eval_samples_per_second": 26.309,
-      "eval_steps_per_second": 1.681,
-      "step": 58
-    },
-    {
-      "epoch": 2.07,
-      "learning_rate": 4.406130268199234e-05,
-      "loss": 0.7585,
-      "step": 60
-    },
-    {
-      "epoch": 2.41,
-      "learning_rate": 4.21455938697318e-05,
-      "loss": 0.7611,
-      "step": 70
-    },
-    {
-      "epoch": 2.76,
-      "learning_rate": 4.0229885057471265e-05,
-      "loss": 0.6841,
-      "step": 80
+      "eval_accuracy": 0.5786802030456852,
+      "eval_loss": 0.9812496304512024,
+      "eval_runtime": 4.4653,
+      "eval_samples_per_second": 88.237,
+      "eval_steps_per_second": 5.599,
+      "step": 138
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.7180451127819549,
-      "eval_loss": 0.811350405216217,
-      "eval_runtime": 13.3447,
-      "eval_samples_per_second": 19.933,
-      "eval_steps_per_second": 1.274,
-      "step": 87
-    },
-    {
-      "epoch": 3.1,
-      "learning_rate": 3.831417624521073e-05,
-      "loss": 0.7207,
-      "step": 90
-    },
-    {
-      "epoch": 3.45,
-      "learning_rate": 3.6398467432950195e-05,
-      "loss": 0.7422,
-      "step": 100
-    },
-    {
-      "epoch": 3.79,
-      "learning_rate": 3.4482758620689657e-05,
-      "loss": 0.73,
-      "step": 110
+      "eval_accuracy": 0.6472081218274112,
+      "eval_loss": 0.8884502053260803,
+      "eval_runtime": 2.1388,
+      "eval_samples_per_second": 184.219,
+      "eval_steps_per_second": 11.689,
+      "step": 207
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.7293233082706767,
-      "eval_loss": 0.7547905445098877,
-      "eval_runtime": 11.7612,
-      "eval_samples_per_second": 22.617,
-      "eval_steps_per_second": 1.445,
-      "step": 116
-    },
-    {
-      "epoch": 4.14,
-      "learning_rate": 3.256704980842912e-05,
-      "loss": 0.6787,
-      "step": 120
-    },
-    {
-      "epoch": 4.48,
-      "learning_rate": 3.065134099616858e-05,
-      "loss": 0.6738,
-      "step": 130
-    },
-    {
-      "epoch": 4.83,
-      "learning_rate": 2.8735632183908045e-05,
-      "loss": 0.6725,
-      "step": 140
+      "eval_accuracy": 0.7081218274111675,
+      "eval_loss": 0.7930060625076294,
+      "eval_runtime": 2.1153,
+      "eval_samples_per_second": 186.265,
+      "eval_steps_per_second": 11.819,
+      "step": 276
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.7406015037593985,
-      "eval_loss": 0.75775146484375,
-      "eval_runtime": 12.9654,
-      "eval_samples_per_second": 20.516,
-      "eval_steps_per_second": 1.311,
-      "step": 145
-    },
-    {
-      "epoch": 5.17,
-      "learning_rate": 2.681992337164751e-05,
-      "loss": 0.6785,
-      "step": 150
-    },
-    {
-      "epoch": 5.52,
-      "learning_rate": 2.4904214559386975e-05,
-      "loss": 0.6259,
-      "step": 160
-    },
-    {
-      "epoch": 5.86,
-      "learning_rate": 2.2988505747126437e-05,
-      "loss": 0.6625,
-      "step": 170
+      "eval_accuracy": 0.6928934010152284,
+      "eval_loss": 0.8019062280654907,
+      "eval_runtime": 7.7818,
+      "eval_samples_per_second": 50.631,
+      "eval_steps_per_second": 3.213,
+      "step": 345
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.7368421052631579,
-      "eval_loss": 0.732495129108429,
-      "eval_runtime": 14.6428,
-      "eval_samples_per_second": 18.166,
-      "eval_steps_per_second": 1.161,
-      "step": 174
-    },
-    {
-      "epoch": 6.21,
-      "learning_rate": 2.10727969348659e-05,
-      "loss": 0.6335,
-      "step": 180
-    },
-    {
-      "epoch": 6.55,
-      "learning_rate": 1.9157088122605367e-05,
-      "loss": 0.6216,
-      "step": 190
-    },
-    {
-      "epoch": 6.9,
-      "learning_rate": 1.7241379310344828e-05,
-      "loss": 0.6214,
-      "step": 200
+      "eval_accuracy": 0.6878172588832487,
+      "eval_loss": 0.8009192943572998,
+      "eval_runtime": 2.0611,
+      "eval_samples_per_second": 191.162,
+      "eval_steps_per_second": 12.13,
+      "step": 414
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.7481203007518797,
-      "eval_loss": 0.6711514592170715,
-      "eval_runtime": 13.3792,
-      "eval_samples_per_second": 19.882,
-      "eval_steps_per_second": 1.271,
-      "step": 203
-    },
-    {
-      "epoch": 7.24,
-      "learning_rate": 1.532567049808429e-05,
-      "loss": 0.6385,
-      "step": 210
+      "eval_accuracy": 0.6852791878172588,
+      "eval_loss": 0.7984281182289124,
+      "eval_runtime": 12.8592,
+      "eval_samples_per_second": 30.639,
+      "eval_steps_per_second": 1.944,
+      "step": 483
     },
     {
-      "epoch": 7.59,
-      "learning_rate": 1.3409961685823755e-05,
-      "loss": 0.6202,
-      "step": 220
-    },
-    {
-      "epoch": 7.93,
-      "learning_rate": 1.1494252873563218e-05,
-      "loss": 0.5911,
-      "step": 230
+      "epoch": 7.25,
+      "learning_rate": 2.6812428078250863e-05,
+      "loss": 0.8194,
+      "step": 500
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.7518796992481203,
-      "eval_loss": 0.6812042593955994,
-      "eval_runtime": 14.3277,
-      "eval_samples_per_second": 18.565,
-      "eval_steps_per_second": 1.187,
-      "step": 232
-    },
-    {
-      "epoch": 8.28,
-      "learning_rate": 9.578544061302683e-06,
-      "loss": 0.5925,
-      "step": 240
-    },
-    {
-      "epoch": 8.62,
-      "learning_rate": 7.662835249042145e-06,
-      "loss": 0.5711,
-      "step": 250
-    },
-    {
-      "epoch": 8.97,
-      "learning_rate": 5.747126436781609e-06,
-      "loss": 0.6169,
-      "step": 260
+      "eval_accuracy": 0.7106598984771574,
+      "eval_loss": 0.7714311480522156,
+      "eval_runtime": 2.123,
+      "eval_samples_per_second": 185.586,
+      "eval_steps_per_second": 11.776,
+      "step": 552
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.7481203007518797,
-      "eval_loss": 0.7047642469406128,
-      "eval_runtime": 17.1552,
-      "eval_samples_per_second": 15.505,
-      "eval_steps_per_second": 0.991,
-      "step": 261
-    },
-    {
-      "epoch": 9.31,
-      "learning_rate": 3.8314176245210725e-06,
-      "loss": 0.6163,
-      "step": 270
-    },
-    {
-      "epoch": 9.66,
-      "learning_rate": 1.9157088122605362e-06,
-      "loss": 0.5672,
-      "step": 280
-    },
-    {
-      "epoch": 10.0,
-      "learning_rate": 0.0,
-      "loss": 0.5863,
-      "step": 290
-    },
-    {
-      "epoch": 10.0,
-      "eval_accuracy": 0.7406015037593985,
-      "eval_loss": 0.7278200387954712,
-      "eval_runtime": 12.3926,
-      "eval_samples_per_second": 21.464,
-      "eval_steps_per_second": 1.372,
-      "step": 290
+      "eval_accuracy": 0.7081218274111675,
+      "eval_loss": 0.7666618824005127,
+      "eval_runtime": 2.021,
+      "eval_samples_per_second": 194.952,
+      "eval_steps_per_second": 12.37,
+      "step": 621
     },
     {
       "epoch": 10.0,
-      "step": 290,
-      "total_flos": 1.430527770561577e+18,
-      "train_loss": 0.7178522997889025,
-      "train_runtime": 2833.88,
-      "train_samples_per_second": 6.514,
-      "train_steps_per_second": 0.102
+      "eval_accuracy": 0.7233502538071066,
+      "eval_loss": 0.7302789688110352,
+      "eval_runtime": 14.4927,
+      "eval_samples_per_second": 27.186,
+      "eval_steps_per_second": 1.725,
+      "step": 690
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.7284263959390863,
+      "eval_loss": 0.7321494817733765,
+      "eval_runtime": 2.2455,
+      "eval_samples_per_second": 175.465,
+      "eval_steps_per_second": 11.134,
+      "step": 759
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.733502538071066,
+      "eval_loss": 0.7373437285423279,
+      "eval_runtime": 5.5421,
+      "eval_samples_per_second": 71.092,
+      "eval_steps_per_second": 4.511,
+      "step": 828
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.6903553299492385,
+      "eval_loss": 0.8051439523696899,
+      "eval_runtime": 12.7177,
+      "eval_samples_per_second": 30.98,
+      "eval_steps_per_second": 1.966,
+      "step": 897
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.7258883248730964,
+      "eval_loss": 0.7687421441078186,
+      "eval_runtime": 4.2192,
+      "eval_samples_per_second": 93.382,
+      "eval_steps_per_second": 5.925,
+      "step": 966
+    },
+    {
+      "epoch": 14.0,
+      "step": 966,
+      "total_flos": 4.790950643943604e+18,
+      "train_loss": 0.63200661163646,
+      "train_runtime": 1457.5958,
+      "train_samples_per_second": 42.415,
+      "train_steps_per_second": 0.663
     }
   ],
-  "logging_steps": 10,
-  "max_steps": 290,
-  "num_train_epochs": 10,
+  "logging_steps": 500,
+  "max_steps": 966,
+  "num_train_epochs": 14,
   "save_steps": 500,
-  "total_flos": 1.430527770561577e+18,
+  "total_flos": 4.790950643943604e+18,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:321bf62f0bb828ae9b9a88c70aa7a7f83bb62da583e16d3a70442dd7fa6e30ef
-size 4536
+oid sha256:ec40bcb32560657f17298036bfeeb617bf1339373552d659c9a5924eb30b7635
+size 4600