🍻 cheers

README.md

@@ -2,6 +2,7 @@
 license: apache-2.0
 base_model: google/vit-base-patch16-224-in21k
 tags:
+- image-classification
 - generated_from_trainer
 model-index:
 - name: ryan_model314_3
@@ -13,11 +14,11 @@ should probably proofread and complete it, then remove this comment. -->
 
 # ryan_model314_3
 
-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.
+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 beans dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.2783
-- Na Accuracy: 0.9389
-- Ordinal Mae: 0.8154
+- Loss: 0.2652
+- Na Accuracy: 0.9372
+- Ordinal Mae: 1.0212
 
 ## Model description
 

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 2.4,
-    "eval_loss": 0.2546972334384918,
-    "eval_na_accuracy": 0.95,
-    "eval_ordinal_mae": 1.2089859222915764,
-    "eval_runtime": 8.0073,
-    "eval_samples_per_second": 24.977,
-    "eval_steps_per_second": 3.122,
-    "train_loss": 0.25249010701974234,
-    "train_runtime": 688.8214,
-    "train_samples_per_second": 11.614,
-    "train_steps_per_second": 0.726
+    "epoch": 0.51,
+    "eval_loss": 0.2651739716529846,
+    "eval_na_accuracy": 0.9371701432520734,
+    "eval_ordinal_mae": 1.021151511080242,
+    "eval_runtime": 162.3376,
+    "eval_samples_per_second": 24.511,
+    "eval_steps_per_second": 3.068,
+    "train_loss": 0.29898834055120294,
+    "train_runtime": 4048.1416,
+    "train_samples_per_second": 34.064,
+    "train_steps_per_second": 2.129
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 2.4,
-    "eval_loss": 0.2546972334384918,
-    "eval_na_accuracy": 0.95,
-    "eval_ordinal_mae": 1.2089859222915764,
-    "eval_runtime": 8.0073,
-    "eval_samples_per_second": 24.977,
-    "eval_steps_per_second": 3.122
+    "epoch": 0.51,
+    "eval_loss": 0.2651739716529846,
+    "eval_na_accuracy": 0.9371701432520734,
+    "eval_ordinal_mae": 1.021151511080242,
+    "eval_runtime": 162.3376,
+    "eval_samples_per_second": 24.511,
+    "eval_steps_per_second": 3.068
 }

runs/Mar26_23-38-05_ryanserver/events.out.tfevents.1711514950.ryanserver.18298.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6fdf06391994cdf0c1812f9860c0e3be96047ea32feaa533d718d042da3d933e
+size 469

train_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 2.4,
-    "train_loss": 0.25249010701974234,
-    "train_runtime": 688.8214,
-    "train_samples_per_second": 11.614,
-    "train_steps_per_second": 0.726
+    "epoch": 0.51,
+    "train_loss": 0.29898834055120294,
+    "train_runtime": 4048.1416,
+    "train_samples_per_second": 34.064,
+    "train_steps_per_second": 2.129
 }

trainer_state.json

@@ -1,359 +1,216 @@
 {
-  "best_metric": 0.2546972334384918,
-  "best_model_checkpoint": "./ryan_model314_3/checkpoint-250",
-  "epoch": 2.4,
-  "eval_steps": 25,
-  "global_step": 300,
+  "best_metric": 0.2651739716529846,
+  "best_model_checkpoint": "./ryan_model314_3/checkpoint-1000",
+  "epoch": 0.5104408352668214,
+  "eval_steps": 100,
+  "global_step": 1100,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.08,
-      "grad_norm": 0.8899219036102295,
-      "learning_rate": 9.8e-05,
-      "loss": 0.6027,
-      "step": 10
-    },
-    {
-      "epoch": 0.16,
-      "grad_norm": 0.9669970273971558,
-      "learning_rate": 9.6e-05,
-      "loss": 0.4505,
-      "step": 20
-    },
-    {
-      "epoch": 0.2,
-      "eval_loss": 0.4262112081050873,
-      "eval_na_accuracy": 0.9,
-      "eval_ordinal_mae": 1.0091899644308249,
-      "eval_runtime": 24.8283,
-      "eval_samples_per_second": 8.055,
-      "eval_steps_per_second": 1.007,
-      "step": 25
-    },
-    {
-      "epoch": 0.24,
-      "grad_norm": 1.0023255348205566,
-      "learning_rate": 9.4e-05,
-      "loss": 0.4166,
-      "step": 30
-    },
-    {
-      "epoch": 0.32,
-      "grad_norm": 0.7017713189125061,
-      "learning_rate": 9.200000000000001e-05,
-      "loss": 0.3632,
-      "step": 40
-    },
-    {
-      "epoch": 0.4,
-      "grad_norm": 1.028002142906189,
-      "learning_rate": 9e-05,
-      "loss": 0.3847,
-      "step": 50
-    },
-    {
-      "epoch": 0.4,
-      "eval_loss": 0.367563396692276,
-      "eval_na_accuracy": 0.935,
-      "eval_ordinal_mae": 1.3719091470156184,
-      "eval_runtime": 7.6715,
-      "eval_samples_per_second": 26.071,
-      "eval_steps_per_second": 3.259,
-      "step": 50
-    },
-    {
-      "epoch": 0.48,
-      "grad_norm": 0.9666626453399658,
-      "learning_rate": 8.800000000000001e-05,
-      "loss": 0.3503,
-      "step": 60
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 1.6747545003890991,
-      "learning_rate": 8.6e-05,
-      "loss": 0.3061,
-      "step": 70
-    },
-    {
-      "epoch": 0.6,
-      "eval_loss": 0.32615897059440613,
-      "eval_na_accuracy": 0.945,
-      "eval_ordinal_mae": 0.7485670199170652,
-      "eval_runtime": 7.5813,
-      "eval_samples_per_second": 26.381,
-      "eval_steps_per_second": 3.298,
-      "step": 75
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 0.9667518138885498,
-      "learning_rate": 8.4e-05,
-      "loss": 0.3304,
-      "step": 80
-    },
-    {
-      "epoch": 0.72,
-      "grad_norm": 1.2316919565200806,
-      "learning_rate": 8.2e-05,
-      "loss": 0.3149,
-      "step": 90
-    },
-    {
-      "epoch": 0.8,
-      "grad_norm": 0.8591766953468323,
-      "learning_rate": 8e-05,
-      "loss": 0.2744,
+      "epoch": 0.05,
+      "grad_norm": 1.358393669128418,
+      "learning_rate": 0.00019767981438515082,
+      "loss": 0.3676,
       "step": 100
     },
     {
-      "epoch": 0.8,
-      "eval_loss": 0.3524325489997864,
-      "eval_na_accuracy": 0.905,
-      "eval_ordinal_mae": 1.1407896330054461,
-      "eval_runtime": 7.8965,
-      "eval_samples_per_second": 25.328,
-      "eval_steps_per_second": 3.166,
+      "epoch": 0.05,
+      "eval_loss": 0.34232085943222046,
+      "eval_na_accuracy": 0.9273686855993968,
+      "eval_ordinal_mae": 1.1293019706063574,
+      "eval_runtime": 381.3744,
+      "eval_samples_per_second": 10.433,
+      "eval_steps_per_second": 1.306,
       "step": 100
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 0.9472360014915466,
-      "learning_rate": 7.800000000000001e-05,
-      "loss": 0.3389,
-      "step": 110
-    },
-    {
-      "epoch": 0.96,
-      "grad_norm": 0.5021968483924866,
-      "learning_rate": 7.6e-05,
-      "loss": 0.2384,
-      "step": 120
-    },
-    {
-      "epoch": 1.0,
-      "eval_loss": 0.36111611127853394,
-      "eval_na_accuracy": 0.93,
-      "eval_ordinal_mae": 0.6746639459123592,
-      "eval_runtime": 7.827,
-      "eval_samples_per_second": 25.553,
-      "eval_steps_per_second": 3.194,
-      "step": 125
-    },
-    {
-      "epoch": 1.04,
-      "grad_norm": 0.4840359687805176,
-      "learning_rate": 7.4e-05,
-      "loss": 0.1954,
-      "step": 130
-    },
-    {
-      "epoch": 1.12,
-      "grad_norm": 1.4421360492706299,
-      "learning_rate": 7.2e-05,
-      "loss": 0.223,
-      "step": 140
-    },
-    {
-      "epoch": 1.2,
-      "grad_norm": 0.36317479610443115,
-      "learning_rate": 7e-05,
-      "loss": 0.2021,
-      "step": 150
-    },
-    {
-      "epoch": 1.2,
-      "eval_loss": 0.31051769852638245,
-      "eval_na_accuracy": 0.95,
-      "eval_ordinal_mae": 1.0440793198453366,
-      "eval_runtime": 8.093,
-      "eval_samples_per_second": 24.713,
-      "eval_steps_per_second": 3.089,
-      "step": 150
-    },
-    {
-      "epoch": 1.28,
-      "grad_norm": 0.8046426773071289,
-      "learning_rate": 6.800000000000001e-05,
-      "loss": 0.21,
-      "step": 160
-    },
-    {
-      "epoch": 1.36,
-      "grad_norm": 0.8122718334197998,
-      "learning_rate": 6.6e-05,
-      "loss": 0.2234,
-      "step": 170
-    },
-    {
-      "epoch": 1.4,
-      "eval_loss": 0.27382710576057434,
-      "eval_na_accuracy": 0.955,
-      "eval_ordinal_mae": 1.4168444083902865,
-      "eval_runtime": 7.6337,
-      "eval_samples_per_second": 26.2,
-      "eval_steps_per_second": 3.275,
-      "step": 175
-    },
-    {
-      "epoch": 1.44,
-      "grad_norm": 1.5608426332473755,
-      "learning_rate": 6.400000000000001e-05,
-      "loss": 0.2306,
-      "step": 180
-    },
-    {
-      "epoch": 1.52,
-      "grad_norm": 0.9130323529243469,
-      "learning_rate": 6.2e-05,
-      "loss": 0.1506,
-      "step": 190
-    },
-    {
-      "epoch": 1.6,
-      "grad_norm": 0.6261627674102783,
-      "learning_rate": 6e-05,
-      "loss": 0.187,
+      "epoch": 0.09,
+      "grad_norm": 0.9884235262870789,
+      "learning_rate": 0.00019535962877030162,
+      "loss": 0.3329,
       "step": 200
     },
     {
-      "epoch": 1.6,
-      "eval_loss": 0.26879259943962097,
-      "eval_na_accuracy": 0.955,
-      "eval_ordinal_mae": 1.365302862794827,
-      "eval_runtime": 7.5957,
-      "eval_samples_per_second": 26.331,
-      "eval_steps_per_second": 3.291,
+      "epoch": 0.09,
+      "eval_loss": 0.31356191635131836,
+      "eval_na_accuracy": 0.9313897964312642,
+      "eval_ordinal_mae": 1.070580373245832,
+      "eval_runtime": 164.3074,
+      "eval_samples_per_second": 24.217,
+      "eval_steps_per_second": 3.031,
       "step": 200
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 0.7222486138343811,
-      "learning_rate": 5.8e-05,
-      "loss": 0.2111,
-      "step": 210
-    },
-    {
-      "epoch": 1.76,
-      "grad_norm": 0.7348190546035767,
-      "learning_rate": 5.6000000000000006e-05,
-      "loss": 0.2008,
-      "step": 220
-    },
-    {
-      "epoch": 1.8,
-      "eval_loss": 0.2668895423412323,
-      "eval_na_accuracy": 0.96,
-      "eval_ordinal_mae": 0.8935630971378721,
-      "eval_runtime": 7.6573,
-      "eval_samples_per_second": 26.119,
-      "eval_steps_per_second": 3.265,
-      "step": 225
-    },
-    {
-      "epoch": 1.84,
-      "grad_norm": 1.3947832584381104,
-      "learning_rate": 5.4000000000000005e-05,
-      "loss": 0.2176,
-      "step": 230
-    },
-    {
-      "epoch": 1.92,
-      "grad_norm": 1.1390098333358765,
-      "learning_rate": 5.2000000000000004e-05,
-      "loss": 0.2037,
-      "step": 240
-    },
-    {
-      "epoch": 2.0,
-      "grad_norm": 0.7233979105949402,
-      "learning_rate": 5e-05,
-      "loss": 0.1541,
-      "step": 250
+      "epoch": 0.14,
+      "grad_norm": 1.2422517538070679,
+      "learning_rate": 0.00019303944315545243,
+      "loss": 0.3134,
+      "step": 300
     },
     {
-      "epoch": 2.0,
-      "eval_loss": 0.2546972334384918,
-      "eval_na_accuracy": 0.95,
-      "eval_ordinal_mae": 1.2089859222915764,
-      "eval_runtime": 7.685,
-      "eval_samples_per_second": 26.025,
-      "eval_steps_per_second": 3.253,
-      "step": 250
+      "epoch": 0.14,
+      "eval_loss": 0.3302006423473358,
+      "eval_na_accuracy": 0.9165619502387534,
+      "eval_ordinal_mae": 1.1219553900231423,
+      "eval_runtime": 164.3869,
+      "eval_samples_per_second": 24.205,
+      "eval_steps_per_second": 3.029,
+      "step": 300
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 0.5704214572906494,
-      "learning_rate": 4.8e-05,
-      "loss": 0.1104,
-      "step": 260
+      "epoch": 0.19,
+      "grad_norm": 1.1815475225448608,
+      "learning_rate": 0.00019071925754060324,
+      "loss": 0.314,
+      "step": 400
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 0.648725688457489,
-      "learning_rate": 4.600000000000001e-05,
-      "loss": 0.1201,
-      "step": 270
+      "epoch": 0.19,
+      "eval_loss": 0.2992381155490875,
+      "eval_na_accuracy": 0.9256094496104549,
+      "eval_ordinal_mae": 0.820222080006495,
+      "eval_runtime": 165.1522,
+      "eval_samples_per_second": 24.093,
+      "eval_steps_per_second": 3.015,
+      "step": 400
     },
     {
-      "epoch": 2.2,
-      "eval_loss": 0.2725123465061188,
-      "eval_na_accuracy": 0.95,
-      "eval_ordinal_mae": 0.7955228271403142,
-      "eval_runtime": 7.4815,
-      "eval_samples_per_second": 26.733,
-      "eval_steps_per_second": 3.342,
-      "step": 275
+      "epoch": 0.23,
+      "grad_norm": 0.47366032004356384,
+      "learning_rate": 0.00018839907192575407,
+      "loss": 0.2965,
+      "step": 500
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 0.5427641272544861,
-      "learning_rate": 4.4000000000000006e-05,
-      "loss": 0.096,
-      "step": 280
+      "epoch": 0.23,
+      "eval_loss": 0.319803923368454,
+      "eval_na_accuracy": 0.9248554913294798,
+      "eval_ordinal_mae": 1.2210362517446196,
+      "eval_runtime": 163.23,
+      "eval_samples_per_second": 24.377,
+      "eval_steps_per_second": 3.051,
+      "step": 500
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 1.687751293182373,
-      "learning_rate": 4.2e-05,
-      "loss": 0.155,
-      "step": 290
+      "epoch": 0.28,
+      "grad_norm": 1.8707951307296753,
+      "learning_rate": 0.00018607888631090488,
+      "loss": 0.3068,
+      "step": 600
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 0.6842709183692932,
-      "learning_rate": 4e-05,
-      "loss": 0.113,
-      "step": 300
+      "epoch": 0.28,
+      "eval_loss": 0.26730772852897644,
+      "eval_na_accuracy": 0.9371701432520734,
+      "eval_ordinal_mae": 1.103637127990878,
+      "eval_runtime": 160.1368,
+      "eval_samples_per_second": 24.848,
+      "eval_steps_per_second": 3.11,
+      "step": 600
     },
     {
-      "epoch": 2.4,
-      "eval_loss": 0.2817830443382263,
-      "eval_na_accuracy": 0.955,
-      "eval_ordinal_mae": 1.2378182741668735,
-      "eval_runtime": 8.0351,
-      "eval_samples_per_second": 24.891,
-      "eval_steps_per_second": 3.111,
-      "step": 300
+      "epoch": 0.32,
+      "grad_norm": 1.1020443439483643,
+      "learning_rate": 0.0001837587006960557,
+      "loss": 0.2824,
+      "step": 700
     },
     {
-      "epoch": 2.4,
-      "step": 300,
-      "total_flos": 3.71974885244928e+17,
-      "train_loss": 0.25249010701974234,
-      "train_runtime": 688.8214,
-      "train_samples_per_second": 11.614,
-      "train_steps_per_second": 0.726
+      "epoch": 0.32,
+      "eval_loss": 0.2921823561191559,
+      "eval_na_accuracy": 0.9371701432520734,
+      "eval_ordinal_mae": 1.4977284913306896,
+      "eval_runtime": 163.7628,
+      "eval_samples_per_second": 24.297,
+      "eval_steps_per_second": 3.041,
+      "step": 700
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2855117321014404,
+      "learning_rate": 0.0001814385150812065,
+      "loss": 0.2914,
+      "step": 800
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 0.2798122763633728,
+      "eval_na_accuracy": 0.9384267403870319,
+      "eval_ordinal_mae": 0.7788859930601368,
+      "eval_runtime": 163.524,
+      "eval_samples_per_second": 24.333,
+      "eval_steps_per_second": 3.045,
+      "step": 800
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4634666442871094,
+      "learning_rate": 0.00017911832946635733,
+      "loss": 0.2968,
+      "step": 900
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 0.27096793055534363,
+      "eval_na_accuracy": 0.9369188238250816,
+      "eval_ordinal_mae": 0.9694435305190251,
+      "eval_runtime": 163.2761,
+      "eval_samples_per_second": 24.37,
+      "eval_steps_per_second": 3.05,
+      "step": 900
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3547471761703491,
+      "learning_rate": 0.00017679814385150814,
+      "loss": 0.2433,
+      "step": 1000
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 0.2651739716529846,
+      "eval_na_accuracy": 0.9371701432520734,
+      "eval_ordinal_mae": 1.021151511080242,
+      "eval_runtime": 165.2163,
+      "eval_samples_per_second": 24.084,
+      "eval_steps_per_second": 3.014,
+      "step": 1000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6242479085922241,
+      "learning_rate": 0.00017447795823665894,
+      "loss": 0.2438,
+      "step": 1100
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 0.2782880365848541,
+      "eval_na_accuracy": 0.9389293792410154,
+      "eval_ordinal_mae": 0.8154305260353155,
+      "eval_runtime": 161.9425,
+      "eval_samples_per_second": 24.57,
+      "eval_steps_per_second": 3.075,
+      "step": 1100
+    },
+    {
+      "epoch": 0.51,
+      "step": 1100,
+      "total_flos": 1.363907912564736e+18,
+      "train_loss": 0.29898834055120294,
+      "train_runtime": 4048.1416,
+      "train_samples_per_second": 34.064,
+      "train_steps_per_second": 2.129
     }
   ],
-  "logging_steps": 10,
-  "max_steps": 500,
+  "logging_steps": 100,
+  "max_steps": 8620,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 4,
-  "save_steps": 25,
-  "total_flos": 3.71974885244928e+17,
+  "save_steps": 100,
+  "total_flos": 1.363907912564736e+18,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null