Upload folder using huggingface_hub

checkpoint-15740/config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "google/vit-base-patch16-224-in21k",
+  "_name_or_path": "dima806/facial_emotions_image_detection",
   "architectures": [
     "ViTForImageClassification"
   ],
@@ -38,5 +38,5 @@
   "problem_type": "single_label_classification",
   "qkv_bias": true,
   "torch_dtype": "float32",
-  "transformers_version": "4.33.3"
+  "transformers_version": "4.38.1"
 }

checkpoint-15740/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13e55d8b7149fc6fab265cf018036ec58a4c63c32b4bfc25f5f48b6425bd5c6c
+size 343239356

checkpoint-15740/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ac054760eb7be1735767627048f8575bd48082e39a0b4d2c28b4cda232cf91db
+oid sha256:5ed812d05d91fefab828d7a87a29455271710955ba68022c7b1c9148a838a1f9
 size 686599173

checkpoint-15740/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6da11932a00ba3c22d65cca3ce0eed71ee7fbb6229986476e7b37e57e16eb416
+oid sha256:274e4012e8c090cea5b5dd5fc3caed1814ac9189d6b2fb69a1a8db44f9ac3827
 size 14575

checkpoint-15740/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5652bb62c654fcdb2e0edd17c5199c98b9cb10ced1a2e8ec21174e5de298f889
+oid sha256:d5e9f4dfba22da0c3088c8ec57bcf6afb6af38ac229cb650d1ecc4b65c60bb31
 size 627

checkpoint-15740/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.5859972834587097,
+  "best_metric": 0.3781418800354004,
   "best_model_checkpoint": "facial_emotions_image_detection/checkpoint-15740",
   "epoch": 10.0,
   "eval_steps": 500,
@@ -10,286 +10,319 @@
   "log_history": [
     {
       "epoch": 0.32,
-      "learning_rate": 9.7131931166348e-06,
-      "loss": 1.5691,
+      "grad_norm": 11.892455101013184,
+      "learning_rate": 3.885277246653919e-06,
+      "loss": 0.5534,
       "step": 500
     },
     {
       "epoch": 0.64,
-      "learning_rate": 9.394518801784576e-06,
-      "loss": 1.1566,
+      "grad_norm": 12.262578964233398,
+      "learning_rate": 3.7578075207138302e-06,
+      "loss": 0.5424,
       "step": 1000
     },
     {
       "epoch": 0.95,
-      "learning_rate": 9.075844486934355e-06,
-      "loss": 1.005,
+      "grad_norm": 11.160516738891602,
+      "learning_rate": 3.630337794773741e-06,
+      "loss": 0.5402,
       "step": 1500
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.6603098927294399,
-      "eval_loss": 1.0014679431915283,
-      "eval_runtime": 113.4351,
-      "eval_samples_per_second": 110.944,
-      "eval_steps_per_second": 13.876,
+      "eval_accuracy": 0.8678585617798967,
+      "eval_loss": 0.4141866862773895,
+      "eval_runtime": 113.7657,
+      "eval_samples_per_second": 110.622,
+      "eval_steps_per_second": 13.835,
       "step": 1574
     },
     {
       "epoch": 1.27,
-      "learning_rate": 8.757170172084131e-06,
-      "loss": 0.9148,
+      "grad_norm": 12.081930160522461,
+      "learning_rate": 3.5028680688336517e-06,
+      "loss": 0.509,
       "step": 2000
     },
     {
       "epoch": 1.59,
-      "learning_rate": 8.438495857233908e-06,
-      "loss": 0.8522,
+      "grad_norm": 8.023213386535645,
+      "learning_rate": 3.375398342893563e-06,
+      "loss": 0.4984,
       "step": 2500
     },
     {
       "epoch": 1.91,
-      "learning_rate": 8.119821542383684e-06,
-      "loss": 0.8063,
+      "grad_norm": 11.671684265136719,
+      "learning_rate": 3.2479286169534735e-06,
+      "loss": 0.4817,
       "step": 3000
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.7133889551052841,
-      "eval_loss": 0.8330217003822327,
-      "eval_runtime": 108.2894,
-      "eval_samples_per_second": 116.216,
-      "eval_steps_per_second": 14.535,
+      "eval_accuracy": 0.8661899086213747,
+      "eval_loss": 0.41461509466171265,
+      "eval_runtime": 111.4496,
+      "eval_samples_per_second": 112.921,
+      "eval_steps_per_second": 14.123,
       "step": 3148
     },
     {
       "epoch": 2.22,
-      "learning_rate": 7.80114722753346e-06,
-      "loss": 0.7566,
+      "grad_norm": 11.465852737426758,
+      "learning_rate": 3.120458891013384e-06,
+      "loss": 0.4686,
       "step": 3500
     },
     {
       "epoch": 2.54,
-      "learning_rate": 7.482472912683239e-06,
-      "loss": 0.7097,
+      "grad_norm": 13.02270221710205,
+      "learning_rate": 2.992989165073295e-06,
+      "loss": 0.451,
       "step": 4000
     },
     {
       "epoch": 2.86,
-      "learning_rate": 7.163798597833015e-06,
-      "loss": 0.7016,
+      "grad_norm": 7.672973155975342,
+      "learning_rate": 2.8655194391332055e-06,
+      "loss": 0.4459,
       "step": 4500
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.7423917361938817,
-      "eval_loss": 0.7428602576255798,
-      "eval_runtime": 107.858,
-      "eval_samples_per_second": 116.681,
-      "eval_steps_per_second": 14.593,
+      "eval_accuracy": 0.8698450536352801,
+      "eval_loss": 0.40479913353919983,
+      "eval_runtime": 111.5158,
+      "eval_samples_per_second": 112.854,
+      "eval_steps_per_second": 14.115,
       "step": 4722
     },
     {
       "epoch": 3.18,
-      "learning_rate": 6.8451242829827925e-06,
-      "loss": 0.6608,
+      "grad_norm": 12.763694763183594,
+      "learning_rate": 2.7380497131931164e-06,
+      "loss": 0.4335,
       "step": 5000
     },
     {
       "epoch": 3.49,
-      "learning_rate": 6.526449968132569e-06,
-      "loss": 0.6235,
+      "grad_norm": 10.964895248413086,
+      "learning_rate": 2.6105799872530273e-06,
+      "loss": 0.4325,
       "step": 5500
     },
     {
       "epoch": 3.81,
-      "learning_rate": 6.207775653282346e-06,
-      "loss": 0.6138,
+      "grad_norm": 12.650814056396484,
+      "learning_rate": 2.4831102613129383e-06,
+      "loss": 0.4031,
       "step": 6000
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.7507350019864919,
-      "eval_loss": 0.7060629725456238,
-      "eval_runtime": 109.3785,
-      "eval_samples_per_second": 115.059,
-      "eval_steps_per_second": 14.39,
+      "eval_accuracy": 0.8707985697258641,
+      "eval_loss": 0.3987765610218048,
+      "eval_runtime": 110.9162,
+      "eval_samples_per_second": 113.464,
+      "eval_steps_per_second": 14.191,
       "step": 6296
     },
     {
       "epoch": 4.13,
-      "learning_rate": 5.889101338432123e-06,
-      "loss": 0.5723,
+      "grad_norm": 7.15053653717041,
+      "learning_rate": 2.3556405353728488e-06,
+      "loss": 0.4112,
       "step": 6500
     },
     {
       "epoch": 4.45,
-      "learning_rate": 5.5704270235819e-06,
-      "loss": 0.5429,
+      "grad_norm": 17.31258773803711,
+      "learning_rate": 2.2281708094327597e-06,
+      "loss": 0.3886,
       "step": 7000
     },
     {
       "epoch": 4.76,
-      "learning_rate": 5.251752708731677e-06,
-      "loss": 0.5403,
+      "grad_norm": 7.938920497894287,
+      "learning_rate": 2.1007010834926702e-06,
+      "loss": 0.3804,
       "step": 7500
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.7726658720699245,
-      "eval_loss": 0.6644242405891418,
-      "eval_runtime": 108.6283,
-      "eval_samples_per_second": 115.854,
-      "eval_steps_per_second": 14.49,
+      "eval_accuracy": 0.8742947953913389,
+      "eval_loss": 0.39545899629592896,
+      "eval_runtime": 110.9484,
+      "eval_samples_per_second": 113.431,
+      "eval_steps_per_second": 14.187,
       "step": 7870
     },
     {
       "epoch": 5.08,
-      "learning_rate": 4.933078393881454e-06,
-      "loss": 0.5112,
+      "grad_norm": 11.187897682189941,
+      "learning_rate": 1.973231357552581e-06,
+      "loss": 0.3856,
       "step": 8000
     },
     {
       "epoch": 5.4,
-      "learning_rate": 4.61440407903123e-06,
-      "loss": 0.4647,
+      "grad_norm": 12.73786735534668,
+      "learning_rate": 1.845761631612492e-06,
+      "loss": 0.3732,
       "step": 8500
     },
     {
       "epoch": 5.72,
-      "learning_rate": 4.295729764181008e-06,
-      "loss": 0.4584,
+      "grad_norm": 6.2277374267578125,
+      "learning_rate": 1.7182919056724028e-06,
+      "loss": 0.3733,
       "step": 9000
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.7871275327771157,
-      "eval_loss": 0.6358364820480347,
-      "eval_runtime": 108.2696,
-      "eval_samples_per_second": 116.238,
-      "eval_steps_per_second": 14.538,
+      "eval_accuracy": 0.8767580452920143,
+      "eval_loss": 0.3847721815109253,
+      "eval_runtime": 113.9748,
+      "eval_samples_per_second": 110.419,
+      "eval_steps_per_second": 13.81,
       "step": 9444
     },
     {
       "epoch": 6.04,
-      "learning_rate": 3.977055449330784e-06,
-      "loss": 0.4567,
+      "grad_norm": 9.49077320098877,
+      "learning_rate": 1.5908221797323135e-06,
+      "loss": 0.3552,
       "step": 9500
     },
     {
       "epoch": 6.35,
-      "learning_rate": 3.6583811344805616e-06,
-      "loss": 0.4134,
+      "grad_norm": 11.691384315490723,
+      "learning_rate": 1.4633524537922245e-06,
+      "loss": 0.355,
       "step": 10000
     },
     {
       "epoch": 6.67,
-      "learning_rate": 3.339706819630338e-06,
-      "loss": 0.3957,
+      "grad_norm": 7.24980354309082,
+      "learning_rate": 1.335882727852135e-06,
+      "loss": 0.347,
       "step": 10500
     },
     {
       "epoch": 6.99,
-      "learning_rate": 3.021032504780115e-06,
-      "loss": 0.3991,
+      "grad_norm": 15.983756065368652,
+      "learning_rate": 1.2084130019120457e-06,
+      "loss": 0.3462,
       "step": 11000
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.800238379022646,
-      "eval_loss": 0.6062497496604919,
-      "eval_runtime": 108.4345,
-      "eval_samples_per_second": 116.061,
-      "eval_steps_per_second": 14.516,
+      "eval_accuracy": 0.877870480731029,
+      "eval_loss": 0.38435637950897217,
+      "eval_runtime": 114.6747,
+      "eval_samples_per_second": 109.745,
+      "eval_steps_per_second": 13.726,
       "step": 11018
     },
     {
       "epoch": 7.31,
-      "learning_rate": 2.702358189929892e-06,
-      "loss": 0.3545,
+      "grad_norm": 6.091674327850342,
+      "learning_rate": 1.0809432759719566e-06,
+      "loss": 0.3481,
       "step": 11500
     },
     {
       "epoch": 7.62,
-      "learning_rate": 2.3836838750796687e-06,
-      "loss": 0.3538,
+      "grad_norm": 14.106987953186035,
+      "learning_rate": 9.534735500318673e-07,
+      "loss": 0.3373,
       "step": 12000
     },
     {
       "epoch": 7.94,
-      "learning_rate": 2.0650095602294456e-06,
-      "loss": 0.3544,
+      "grad_norm": 8.448904991149902,
+      "learning_rate": 8.260038240917782e-07,
+      "loss": 0.324,
       "step": 12500
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.8045292014302742,
-      "eval_loss": 0.6008493900299072,
-      "eval_runtime": 110.3834,
-      "eval_samples_per_second": 114.012,
-      "eval_steps_per_second": 14.259,
+      "eval_accuracy": 0.876201827572507,
+      "eval_loss": 0.3851085305213928,
+      "eval_runtime": 113.668,
+      "eval_samples_per_second": 110.717,
+      "eval_steps_per_second": 13.847,
       "step": 12592
     },
     {
       "epoch": 8.26,
-      "learning_rate": 1.7463352453792225e-06,
-      "loss": 0.3183,
+      "grad_norm": 6.959225177764893,
+      "learning_rate": 6.98534098151689e-07,
+      "loss": 0.3254,
       "step": 13000
     },
     {
       "epoch": 8.58,
-      "learning_rate": 1.4276609305289997e-06,
-      "loss": 0.3218,
+      "grad_norm": 10.736397743225098,
+      "learning_rate": 5.710643722115998e-07,
+      "loss": 0.3179,
       "step": 13500
     },
     {
       "epoch": 8.89,
-      "learning_rate": 1.1089866156787763e-06,
-      "loss": 0.3184,
+      "grad_norm": 10.716438293457031,
+      "learning_rate": 4.435946462715105e-07,
+      "loss": 0.3217,
       "step": 14000
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.8116805721096544,
-      "eval_loss": 0.590340256690979,
-      "eval_runtime": 108.9419,
-      "eval_samples_per_second": 115.52,
-      "eval_steps_per_second": 14.448,
+      "eval_accuracy": 0.8787445371473976,
+      "eval_loss": 0.3801954388618469,
+      "eval_runtime": 113.1501,
+      "eval_samples_per_second": 111.224,
+      "eval_steps_per_second": 13.911,
       "step": 14166
     },
     {
       "epoch": 9.21,
-      "learning_rate": 7.903123008285534e-07,
-      "loss": 0.3054,
+      "grad_norm": 12.26369571685791,
+      "learning_rate": 3.161249203314213e-07,
+      "loss": 0.309,
       "step": 14500
     },
     {
       "epoch": 9.53,
-      "learning_rate": 4.7163798597833015e-07,
-      "loss": 0.2813,
+      "grad_norm": 6.435930252075195,
+      "learning_rate": 1.8865519439133203e-07,
+      "loss": 0.3158,
       "step": 15000
     },
     {
       "epoch": 9.85,
-      "learning_rate": 1.5296367112810708e-07,
-      "loss": 0.2984,
+      "grad_norm": 15.131240844726562,
+      "learning_rate": 6.118546845124282e-08,
+      "loss": 0.3105,
       "step": 15500
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 0.8162097735399285,
-      "eval_loss": 0.5859972834587097,
-      "eval_runtime": 111.4622,
-      "eval_samples_per_second": 112.908,
-      "eval_steps_per_second": 14.121,
+      "eval_accuracy": 0.8789829161700436,
+      "eval_loss": 0.3781418800354004,
+      "eval_runtime": 112.3381,
+      "eval_samples_per_second": 112.028,
+      "eval_steps_per_second": 14.011,
       "step": 15740
     }
   ],
   "logging_steps": 500,
   "max_steps": 15740,
+  "num_input_tokens_seen": 0,
   "num_train_epochs": 10,
   "save_steps": 500,
   "total_flos": 3.900966581033497e+19,
+  "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null
 }

checkpoint-15740/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:88f8d355cab9ef739f5fc476aba7f6e893c6c8f92be3fff90b9d4091956a9dd9
-size 4027
+oid sha256:412104b61f174a6b32d7fb7f0dee0c8ea25fe50495d4c2050346fa634d05d144
+size 4411

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a6c60a65625a5858cb798dc3334d437083ee9b580c83d48a51f373240a7a1787
+oid sha256:13e55d8b7149fc6fab265cf018036ec58a4c63c32b4bfc25f5f48b6425bd5c6c
 size 343239356

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f4a27ac1cf21ac69961911d3f69cc32618d3c82fd492d9d41337dd8d2d59a7a0
+oid sha256:412104b61f174a6b32d7fb7f0dee0c8ea25fe50495d4c2050346fa634d05d144
 size 4411