End of training

README.md

@@ -18,8 +18,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [MCG-NJU/videomae-base](https://huggingface.co/MCG-NJU/videomae-base) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.1592
-- Accuracy: 0.9614
+- Loss: 0.2194
+- Accuracy: 0.9417
 
 ## Model description
 

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 9.1,
+    "eval_accuracy": 0.941717791411043,
+    "eval_loss": 0.2193548083305359,
+    "eval_runtime": 759.0426,
+    "eval_samples_per_second": 0.859,
+    "eval_steps_per_second": 0.429
+}

logs/events.out.tfevents.1734212381.DeepLearning2.1501842.7

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:786578fa79d3faa8400779ef1d3bef3a6d00c174e315c6e49756f5fbc6856c8e
+size 411

test_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 9.1,
+    "eval_accuracy": 0.941717791411043,
+    "eval_loss": 0.2193548083305359,
+    "eval_runtime": 759.0426,
+    "eval_samples_per_second": 0.859,
+    "eval_steps_per_second": 0.429
+}

trainer_state.json

@@ -0,0 +1,337 @@
+{
+  "best_metric": 0.9652509652509652,
+  "best_model_checkpoint": "/media/cse/HDD/Shawon/shawon/10 fold timesformer/VideoMAE_default_fold__3__10_epoch_Aug_batch_1_4_LSA64/checkpoint-2304",
+  "epoch": 9.1,
+  "eval_steps": 500,
+  "global_step": 2880,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.034722222222222224,
+      "grad_norm": 53.980350494384766,
+      "learning_rate": 1.736111111111111e-05,
+      "loss": 16.6245,
+      "step": 100
+    },
+    {
+      "epoch": 0.06944444444444445,
+      "grad_norm": 47.83296203613281,
+      "learning_rate": 3.472222222222222e-05,
+      "loss": 15.6268,
+      "step": 200
+    },
+    {
+      "epoch": 0.1,
+      "eval_accuracy": 0.03088803088803089,
+      "eval_loss": 3.6311283111572266,
+      "eval_runtime": 284.5829,
+      "eval_samples_per_second": 0.91,
+      "eval_steps_per_second": 0.457,
+      "step": 288
+    },
+    {
+      "epoch": 1.0041666666666667,
+      "grad_norm": 24.499309539794922,
+      "learning_rate": 4.976851851851852e-05,
+      "loss": 14.8414,
+      "step": 300
+    },
+    {
+      "epoch": 1.038888888888889,
+      "grad_norm": 30.084009170532227,
+      "learning_rate": 4.783950617283951e-05,
+      "loss": 12.8447,
+      "step": 400
+    },
+    {
+      "epoch": 1.073611111111111,
+      "grad_norm": 42.996299743652344,
+      "learning_rate": 4.591049382716049e-05,
+      "loss": 8.557,
+      "step": 500
+    },
+    {
+      "epoch": 1.1,
+      "eval_accuracy": 0.7104247104247104,
+      "eval_loss": 1.2254153490066528,
+      "eval_runtime": 288.3753,
+      "eval_samples_per_second": 0.898,
+      "eval_steps_per_second": 0.451,
+      "step": 576
+    },
+    {
+      "epoch": 2.0083333333333333,
+      "grad_norm": 35.8374137878418,
+      "learning_rate": 4.3981481481481486e-05,
+      "loss": 5.2507,
+      "step": 600
+    },
+    {
+      "epoch": 2.0430555555555556,
+      "grad_norm": 46.812416076660156,
+      "learning_rate": 4.205246913580247e-05,
+      "loss": 3.1729,
+      "step": 700
+    },
+    {
+      "epoch": 2.077777777777778,
+      "grad_norm": 26.94959831237793,
+      "learning_rate": 4.012345679012346e-05,
+      "loss": 2.4113,
+      "step": 800
+    },
+    {
+      "epoch": 2.1,
+      "eval_accuracy": 0.8494208494208494,
+      "eval_loss": 0.6185011863708496,
+      "eval_runtime": 289.9317,
+      "eval_samples_per_second": 0.893,
+      "eval_steps_per_second": 0.448,
+      "step": 864
+    },
+    {
+      "epoch": 3.0125,
+      "grad_norm": 10.739028930664062,
+      "learning_rate": 3.8194444444444444e-05,
+      "loss": 1.8377,
+      "step": 900
+    },
+    {
+      "epoch": 3.047222222222222,
+      "grad_norm": 11.71504020690918,
+      "learning_rate": 3.626543209876543e-05,
+      "loss": 1.4297,
+      "step": 1000
+    },
+    {
+      "epoch": 3.0819444444444444,
+      "grad_norm": 16.224809646606445,
+      "learning_rate": 3.4336419753086427e-05,
+      "loss": 0.9955,
+      "step": 1100
+    },
+    {
+      "epoch": 3.1,
+      "eval_accuracy": 0.8957528957528957,
+      "eval_loss": 0.3698173761367798,
+      "eval_runtime": 289.1485,
+      "eval_samples_per_second": 0.896,
+      "eval_steps_per_second": 0.45,
+      "step": 1152
+    },
+    {
+      "epoch": 4.016666666666667,
+      "grad_norm": 7.019224643707275,
+      "learning_rate": 3.240740740740741e-05,
+      "loss": 0.9355,
+      "step": 1200
+    },
+    {
+      "epoch": 4.051388888888889,
+      "grad_norm": 6.03433895111084,
+      "learning_rate": 3.04783950617284e-05,
+      "loss": 0.5441,
+      "step": 1300
+    },
+    {
+      "epoch": 4.086111111111111,
+      "grad_norm": 14.718276977539062,
+      "learning_rate": 2.8549382716049384e-05,
+      "loss": 0.5989,
+      "step": 1400
+    },
+    {
+      "epoch": 4.1,
+      "eval_accuracy": 0.9227799227799228,
+      "eval_loss": 0.2324027121067047,
+      "eval_runtime": 289.0049,
+      "eval_samples_per_second": 0.896,
+      "eval_steps_per_second": 0.45,
+      "step": 1440
+    },
+    {
+      "epoch": 5.020833333333333,
+      "grad_norm": 2.174286127090454,
+      "learning_rate": 2.6620370370370372e-05,
+      "loss": 0.3962,
+      "step": 1500
+    },
+    {
+      "epoch": 5.055555555555555,
+      "grad_norm": 72.37142181396484,
+      "learning_rate": 2.4691358024691357e-05,
+      "loss": 0.3747,
+      "step": 1600
+    },
+    {
+      "epoch": 5.090277777777778,
+      "grad_norm": 54.647560119628906,
+      "learning_rate": 2.2762345679012348e-05,
+      "loss": 0.3755,
+      "step": 1700
+    },
+    {
+      "epoch": 5.1,
+      "eval_accuracy": 0.9382239382239382,
+      "eval_loss": 0.2560836970806122,
+      "eval_runtime": 291.0772,
+      "eval_samples_per_second": 0.89,
+      "eval_steps_per_second": 0.447,
+      "step": 1728
+    },
+    {
+      "epoch": 6.025,
+      "grad_norm": 1.4880203008651733,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.3606,
+      "step": 1800
+    },
+    {
+      "epoch": 6.059722222222222,
+      "grad_norm": 21.25699234008789,
+      "learning_rate": 1.8904320987654324e-05,
+      "loss": 0.2581,
+      "step": 1900
+    },
+    {
+      "epoch": 6.094444444444444,
+      "grad_norm": 0.8457247018814087,
+      "learning_rate": 1.697530864197531e-05,
+      "loss": 0.2115,
+      "step": 2000
+    },
+    {
+      "epoch": 6.1,
+      "eval_accuracy": 0.9498069498069498,
+      "eval_loss": 0.1747695803642273,
+      "eval_runtime": 288.1554,
+      "eval_samples_per_second": 0.899,
+      "eval_steps_per_second": 0.451,
+      "step": 2016
+    },
+    {
+      "epoch": 7.029166666666667,
+      "grad_norm": 0.1845870167016983,
+      "learning_rate": 1.5046296296296297e-05,
+      "loss": 0.2672,
+      "step": 2100
+    },
+    {
+      "epoch": 7.063888888888889,
+      "grad_norm": 0.15726155042648315,
+      "learning_rate": 1.3117283950617285e-05,
+      "loss": 0.1068,
+      "step": 2200
+    },
+    {
+      "epoch": 7.098611111111111,
+      "grad_norm": 0.21128535270690918,
+      "learning_rate": 1.1188271604938271e-05,
+      "loss": 0.1498,
+      "step": 2300
+    },
+    {
+      "epoch": 7.1,
+      "eval_accuracy": 0.9652509652509652,
+      "eval_loss": 0.16693688929080963,
+      "eval_runtime": 291.4136,
+      "eval_samples_per_second": 0.889,
+      "eval_steps_per_second": 0.446,
+      "step": 2304
+    },
+    {
+      "epoch": 8.033333333333333,
+      "grad_norm": 0.13224439322948456,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 0.0899,
+      "step": 2400
+    },
+    {
+      "epoch": 8.068055555555556,
+      "grad_norm": 0.10548185557126999,
+      "learning_rate": 7.330246913580248e-06,
+      "loss": 0.1063,
+      "step": 2500
+    },
+    {
+      "epoch": 8.1,
+      "eval_accuracy": 0.9613899613899614,
+      "eval_loss": 0.15925812721252441,
+      "eval_runtime": 297.8407,
+      "eval_samples_per_second": 0.87,
+      "eval_steps_per_second": 0.436,
+      "step": 2592
+    },
+    {
+      "epoch": 9.002777777777778,
+      "grad_norm": 0.0979766771197319,
+      "learning_rate": 5.401234567901234e-06,
+      "loss": 0.1246,
+      "step": 2600
+    },
+    {
+      "epoch": 9.0375,
+      "grad_norm": 117.19750213623047,
+      "learning_rate": 3.4722222222222224e-06,
+      "loss": 0.1357,
+      "step": 2700
+    },
+    {
+      "epoch": 9.072222222222223,
+      "grad_norm": 0.16245272755622864,
+      "learning_rate": 1.5432098765432098e-06,
+      "loss": 0.0475,
+      "step": 2800
+    },
+    {
+      "epoch": 9.1,
+      "eval_accuracy": 0.9613899613899614,
+      "eval_loss": 0.15923245251178741,
+      "eval_runtime": 294.949,
+      "eval_samples_per_second": 0.878,
+      "eval_steps_per_second": 0.441,
+      "step": 2880
+    },
+    {
+      "epoch": 9.1,
+      "step": 2880,
+      "total_flos": 2.8725202238381752e+19,
+      "train_loss": 3.0831375959846707,
+      "train_runtime": 31748.8552,
+      "train_samples_per_second": 0.726,
+      "train_steps_per_second": 0.091
+    },
+    {
+      "epoch": 9.1,
+      "eval_accuracy": 0.941717791411043,
+      "eval_loss": 0.2193548083305359,
+      "eval_runtime": 759.0426,
+      "eval_samples_per_second": 0.859,
+      "eval_steps_per_second": 0.429,
+      "step": 2880
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 2880,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.8725202238381752e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}