🍻 cheers

README.md

@@ -2,6 +2,7 @@
 license: apache-2.0
 base_model: google/vit-base-patch16-224
 tags:
+- image-classification
 - generated_from_trainer
 metrics:
 - accuracy
@@ -18,13 +19,13 @@ should probably proofread and complete it, then remove this comment. -->
 
 # vit-lr-0.0001
 
-This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on an unknown dataset.
+This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on the skin-cancer dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6800
-- Accuracy: 0.8322
-- Precision: 0.8330
-- Recall: 0.8322
-- F1: 0.8229
+- Loss: 0.4886
+- Accuracy: 0.8273
+- Precision: 0.8287
+- Recall: 0.8273
+- F1: 0.8227
 
 ## Model description
 

all_results.json

@@ -1,13 +1,16 @@
 {
-    "epoch": 2.18,
-    "eval_accuracy": 0.8169209431345353,
-    "eval_loss": 0.49744343757629395,
-    "eval_runtime": 40.9787,
-    "eval_samples_per_second": 70.378,
-    "eval_steps_per_second": 8.809,
-    "total_flos": 8.667092503134536e+17,
-    "train_loss": 0.4869789886474609,
-    "train_runtime": 547.2513,
-    "train_samples_per_second": 937.047,
-    "train_steps_per_second": 58.657
+    "epoch": 3.43,
+    "eval_accuracy": 0.8273231622746186,
+    "eval_f1": 0.822737257017553,
+    "eval_loss": 0.4886268079280853,
+    "eval_precision": 0.8287258592840144,
+    "eval_recall": 0.8273231622746186,
+    "eval_runtime": 40.0459,
+    "eval_samples_per_second": 72.017,
+    "eval_steps_per_second": 9.015,
+    "total_flos": 1.362060245306622e+18,
+    "train_loss": 0.40586378964510833,
+    "train_runtime": 828.4507,
+    "train_samples_per_second": 618.987,
+    "train_steps_per_second": 38.747
 }

eval_results.json

@@ -1,8 +1,11 @@
 {
-    "epoch": 2.18,
-    "eval_accuracy": 0.8169209431345353,
-    "eval_loss": 0.49744343757629395,
-    "eval_runtime": 40.9787,
-    "eval_samples_per_second": 70.378,
-    "eval_steps_per_second": 8.809
+    "epoch": 3.43,
+    "eval_accuracy": 0.8273231622746186,
+    "eval_f1": 0.822737257017553,
+    "eval_loss": 0.4886268079280853,
+    "eval_precision": 0.8287258592840144,
+    "eval_recall": 0.8273231622746186,
+    "eval_runtime": 40.0459,
+    "eval_samples_per_second": 72.017,
+    "eval_steps_per_second": 9.015
 }

runs/Mar17_21-09-49_add18dc15495/events.out.tfevents.1710710837.add18dc15495.3839.6

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23901c00884c0ff9b4378a9daf6a29df8980be154cb078878ffe9f6c35e7cc21
+size 560

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.18,
-    "total_flos": 8.667092503134536e+17,
-    "train_loss": 0.4869789886474609,
-    "train_runtime": 547.2513,
-    "train_samples_per_second": 937.047,
-    "train_steps_per_second": 58.657
+    "epoch": 3.43,
+    "total_flos": 1.362060245306622e+18,
+    "train_loss": 0.40586378964510833,
+    "train_runtime": 828.4507,
+    "train_samples_per_second": 618.987,
+    "train_steps_per_second": 38.747
 }

trainer_state.json

@@ -1,133 +1,230 @@
 {
-  "best_metric": 0.49744343757629395,
-  "best_model_checkpoint": "./vit-lr-0.0001/checkpoint-400",
-  "epoch": 2.1806853582554515,
+  "best_metric": 0.4886268079280853,
+  "best_model_checkpoint": "./vit-lr-0.0001/checkpoint-800",
+  "epoch": 3.426791277258567,
   "eval_steps": 100,
-  "global_step": 700,
+  "global_step": 1100,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.31,
-      "grad_norm": 4.504175186157227,
-      "learning_rate": 9.999760543724762e-05,
-      "loss": 0.7774,
+      "grad_norm": 4.336536407470703,
+      "learning_rate": 9.999765308867361e-05,
+      "loss": 0.8247,
       "step": 100
     },
     {
       "epoch": 0.31,
-      "eval_accuracy": 0.7846740638002774,
-      "eval_loss": 0.5988100171089172,
-      "eval_runtime": 39.924,
-      "eval_samples_per_second": 72.237,
-      "eval_steps_per_second": 9.042,
+      "eval_accuracy": 0.7808599167822469,
+      "eval_f1": 0.7543146399483164,
+      "eval_loss": 0.6282410025596619,
+      "eval_precision": 0.761734030016078,
+      "eval_recall": 0.7808599167822469,
+      "eval_runtime": 37.9446,
+      "eval_samples_per_second": 76.006,
+      "eval_steps_per_second": 9.514,
       "step": 100
     },
     {
       "epoch": 0.62,
-      "grad_norm": 6.5739970207214355,
-      "learning_rate": 9.999042197834769e-05,
-      "loss": 0.59,
+      "grad_norm": 6.528007507324219,
+      "learning_rate": 9.999051751609367e-05,
+      "loss": 0.6235,
       "step": 200
     },
     {
       "epoch": 0.62,
-      "eval_accuracy": 0.7732316227461858,
-      "eval_loss": 0.6567931175231934,
-      "eval_runtime": 39.9411,
-      "eval_samples_per_second": 72.206,
-      "eval_steps_per_second": 9.038,
+      "eval_accuracy": 0.7510402219140083,
+      "eval_f1": 0.6980793391075255,
+      "eval_loss": 0.7094237208366394,
+      "eval_precision": 0.7832859172651316,
+      "eval_recall": 0.7510402219140083,
+      "eval_runtime": 38.1666,
+      "eval_samples_per_second": 75.564,
+      "eval_steps_per_second": 9.459,
       "step": 200
     },
     {
       "epoch": 0.93,
-      "grad_norm": 6.0171613693237305,
+      "grad_norm": 7.167332649230957,
       "learning_rate": 9.997859372626506e-05,
-      "loss": 0.584,
+      "loss": 0.5888,
       "step": 300
     },
     {
       "epoch": 0.93,
-      "eval_accuracy": 0.7562413314840499,
-      "eval_loss": 0.6686193346977234,
-      "eval_runtime": 40.0793,
-      "eval_samples_per_second": 71.957,
-      "eval_steps_per_second": 9.007,
+      "eval_accuracy": 0.7402912621359223,
+      "eval_f1": 0.7507721767549878,
+      "eval_loss": 0.687461793422699,
+      "eval_precision": 0.7995575456891281,
+      "eval_recall": 0.7402912621359223,
+      "eval_runtime": 38.3771,
+      "eval_samples_per_second": 75.149,
+      "eval_steps_per_second": 9.407,
       "step": 300
     },
     {
       "epoch": 1.25,
-      "grad_norm": 3.5295469760894775,
+      "grad_norm": 2.4604434967041016,
       "learning_rate": 9.996188286127832e-05,
-      "loss": 0.4275,
+      "loss": 0.454,
       "step": 400
     },
     {
       "epoch": 1.25,
-      "eval_accuracy": 0.8169209431345353,
-      "eval_loss": 0.49744343757629395,
-      "eval_runtime": 39.8506,
-      "eval_samples_per_second": 72.37,
-      "eval_steps_per_second": 9.059,
+      "eval_accuracy": 0.7739251040221914,
+      "eval_f1": 0.7841730444076129,
+      "eval_loss": 0.6142727136611938,
+      "eval_precision": 0.8039138713400071,
+      "eval_recall": 0.7739251040221914,
+      "eval_runtime": 38.6976,
+      "eval_samples_per_second": 74.527,
+      "eval_steps_per_second": 9.329,
       "step": 400
     },
     {
       "epoch": 1.56,
-      "grad_norm": 2.440728187561035,
+      "grad_norm": 2.9993300437927246,
       "learning_rate": 9.994038652174203e-05,
-      "loss": 0.3963,
+      "loss": 0.4104,
       "step": 500
     },
     {
       "epoch": 1.56,
-      "eval_accuracy": 0.8287101248266296,
-      "eval_loss": 0.507750391960144,
-      "eval_runtime": 39.6231,
-      "eval_samples_per_second": 72.786,
-      "eval_steps_per_second": 9.111,
+      "eval_accuracy": 0.808252427184466,
+      "eval_f1": 0.7911353922443626,
+      "eval_loss": 0.5673110485076904,
+      "eval_precision": 0.8033716514082728,
+      "eval_recall": 0.808252427184466,
+      "eval_runtime": 38.598,
+      "eval_samples_per_second": 74.719,
+      "eval_steps_per_second": 9.353,
       "step": 500
     },
     {
       "epoch": 1.87,
-      "grad_norm": 4.625945568084717,
+      "grad_norm": 3.316398859024048,
       "learning_rate": 9.991410676662952e-05,
-      "loss": 0.3708,
+      "loss": 0.4039,
       "step": 600
     },
     {
       "epoch": 1.87,
-      "eval_accuracy": 0.794382801664355,
-      "eval_loss": 0.5709377527236938,
-      "eval_runtime": 39.6196,
-      "eval_samples_per_second": 72.792,
-      "eval_steps_per_second": 9.112,
+      "eval_accuracy": 0.8280166435506241,
+      "eval_f1": 0.8272377870020207,
+      "eval_loss": 0.5172362923622131,
+      "eval_precision": 0.838315056266821,
+      "eval_recall": 0.8280166435506241,
+      "eval_runtime": 38.241,
+      "eval_samples_per_second": 75.416,
+      "eval_steps_per_second": 9.44,
       "step": 600
     },
     {
       "epoch": 2.18,
-      "grad_norm": 2.9291675090789795,
+      "grad_norm": 3.1912124156951904,
       "learning_rate": 9.988304611308174e-05,
-      "loss": 0.2628,
+      "loss": 0.2785,
       "step": 700
     },
     {
       "epoch": 2.18,
-      "eval_accuracy": 0.812760055478502,
-      "eval_loss": 0.6022083163261414,
-      "eval_runtime": 39.3146,
-      "eval_samples_per_second": 73.357,
-      "eval_steps_per_second": 9.182,
+      "eval_accuracy": 0.826629680998613,
+      "eval_f1": 0.8306324023780052,
+      "eval_loss": 0.563904345035553,
+      "eval_precision": 0.8424957959799085,
+      "eval_recall": 0.826629680998613,
+      "eval_runtime": 38.3845,
+      "eval_samples_per_second": 75.134,
+      "eval_steps_per_second": 9.405,
       "step": 700
     },
     {
-      "epoch": 2.18,
-      "step": 700,
-      "total_flos": 8.667092503134536e+17,
-      "train_loss": 0.4869789886474609,
-      "train_runtime": 547.2513,
-      "train_samples_per_second": 937.047,
-      "train_steps_per_second": 58.657
+      "epoch": 2.49,
+      "grad_norm": 2.378309965133667,
+      "learning_rate": 9.984720753616604e-05,
+      "loss": 0.2744,
+      "step": 800
+    },
+    {
+      "epoch": 2.49,
+      "eval_accuracy": 0.8273231622746186,
+      "eval_f1": 0.822737257017553,
+      "eval_loss": 0.4886268079280853,
+      "eval_precision": 0.8287258592840144,
+      "eval_recall": 0.8273231622746186,
+      "eval_runtime": 38.1003,
+      "eval_samples_per_second": 75.695,
+      "eval_steps_per_second": 9.475,
+      "step": 800
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.579433441162109,
+      "learning_rate": 9.980659446859127e-05,
+      "loss": 0.2545,
+      "step": 900
+    },
+    {
+      "epoch": 2.8,
+      "eval_accuracy": 0.8495145631067961,
+      "eval_f1": 0.8472903587104604,
+      "eval_loss": 0.4898224472999573,
+      "eval_precision": 0.8510030222503284,
+      "eval_recall": 0.8495145631067961,
+      "eval_runtime": 38.8455,
+      "eval_samples_per_second": 74.243,
+      "eval_steps_per_second": 9.293,
+      "step": 900
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.6288774013519287,
+      "learning_rate": 9.976121080037899e-05,
+      "loss": 0.2197,
+      "step": 1000
+    },
+    {
+      "epoch": 3.12,
+      "eval_accuracy": 0.8574895977808599,
+      "eval_f1": 0.8550066256842739,
+      "eval_loss": 0.5342256426811218,
+      "eval_precision": 0.8561255195381418,
+      "eval_recall": 0.8574895977808599,
+      "eval_runtime": 39.0507,
+      "eval_samples_per_second": 73.853,
+      "eval_steps_per_second": 9.244,
+      "step": 1000
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 9.022770881652832,
+      "learning_rate": 9.971106087849084e-05,
+      "loss": 0.1321,
+      "step": 1100
+    },
+    {
+      "epoch": 3.43,
+      "eval_accuracy": 0.8321775312066574,
+      "eval_f1": 0.822903478849272,
+      "eval_loss": 0.6799635291099548,
+      "eval_precision": 0.8329985305708512,
+      "eval_recall": 0.8321775312066574,
+      "eval_runtime": 38.5736,
+      "eval_samples_per_second": 74.766,
+      "eval_steps_per_second": 9.359,
+      "step": 1100
+    },
+    {
+      "epoch": 3.43,
+      "step": 1100,
+      "total_flos": 1.362060245306622e+18,
+      "train_loss": 0.40586378964510833,
+      "train_runtime": 828.4507,
+      "train_samples_per_second": 618.987,
+      "train_steps_per_second": 38.747
     }
   ],
   "logging_steps": 100,
@@ -135,7 +232,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 100,
   "save_steps": 100,
-  "total_flos": 8.667092503134536e+17,
+  "total_flos": 1.362060245306622e+18,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null