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: 0.1243
-- Accuracy: 0.9691
+- Loss: 0.1673
+- Accuracy: 0.9417
 
 ## Model description
 

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 9.1,
+    "eval_accuracy": 0.941717791411043,
+    "eval_loss": 0.16725970804691315,
+    "eval_runtime": 750.4557,
+    "eval_samples_per_second": 0.869,
+    "eval_steps_per_second": 0.434
+}

test_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 9.1,
+    "eval_accuracy": 0.941717791411043,
+    "eval_loss": 0.16725970804691315,
+    "eval_runtime": 750.4557,
+    "eval_samples_per_second": 0.869,
+    "eval_steps_per_second": 0.434
+}

trainer_state.json

@@ -0,0 +1,337 @@
+{
+  "best_metric": 0.972972972972973,
+  "best_model_checkpoint": "/media/cse/HDD/Shawon/shawon/10 fold timesformer/VideoMAE_default_fold__0__10_epoch_Aug_batch_2_4_LSA64/checkpoint-2592",
+  "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": 42.11737823486328,
+      "learning_rate": 1.736111111111111e-05,
+      "loss": 16.6298,
+      "step": 100
+    },
+    {
+      "epoch": 0.06944444444444445,
+      "grad_norm": 33.30021667480469,
+      "learning_rate": 3.472222222222222e-05,
+      "loss": 15.5442,
+      "step": 200
+    },
+    {
+      "epoch": 0.1,
+      "eval_accuracy": 0.08108108108108109,
+      "eval_loss": 3.505643129348755,
+      "eval_runtime": 283.7815,
+      "eval_samples_per_second": 0.913,
+      "eval_steps_per_second": 0.458,
+      "step": 288
+    },
+    {
+      "epoch": 1.0041666666666667,
+      "grad_norm": 178.25469970703125,
+      "learning_rate": 4.976851851851852e-05,
+      "loss": 14.6286,
+      "step": 300
+    },
+    {
+      "epoch": 1.038888888888889,
+      "grad_norm": 30.423072814941406,
+      "learning_rate": 4.783950617283951e-05,
+      "loss": 12.0331,
+      "step": 400
+    },
+    {
+      "epoch": 1.073611111111111,
+      "grad_norm": 32.779422760009766,
+      "learning_rate": 4.591049382716049e-05,
+      "loss": 7.6719,
+      "step": 500
+    },
+    {
+      "epoch": 1.1,
+      "eval_accuracy": 0.803088803088803,
+      "eval_loss": 1.0166326761245728,
+      "eval_runtime": 300.1952,
+      "eval_samples_per_second": 0.863,
+      "eval_steps_per_second": 0.433,
+      "step": 576
+    },
+    {
+      "epoch": 2.0083333333333333,
+      "grad_norm": 39.084625244140625,
+      "learning_rate": 4.3981481481481486e-05,
+      "loss": 4.3167,
+      "step": 600
+    },
+    {
+      "epoch": 2.0430555555555556,
+      "grad_norm": 24.13694190979004,
+      "learning_rate": 4.205246913580247e-05,
+      "loss": 2.4789,
+      "step": 700
+    },
+    {
+      "epoch": 2.077777777777778,
+      "grad_norm": 39.12590026855469,
+      "learning_rate": 4.012345679012346e-05,
+      "loss": 1.737,
+      "step": 800
+    },
+    {
+      "epoch": 2.1,
+      "eval_accuracy": 0.8803088803088803,
+      "eval_loss": 0.4851164221763611,
+      "eval_runtime": 299.8996,
+      "eval_samples_per_second": 0.864,
+      "eval_steps_per_second": 0.433,
+      "step": 864
+    },
+    {
+      "epoch": 3.0125,
+      "grad_norm": 13.427978515625,
+      "learning_rate": 3.8194444444444444e-05,
+      "loss": 1.2593,
+      "step": 900
+    },
+    {
+      "epoch": 3.047222222222222,
+      "grad_norm": 41.39619827270508,
+      "learning_rate": 3.626543209876543e-05,
+      "loss": 0.9449,
+      "step": 1000
+    },
+    {
+      "epoch": 3.0819444444444444,
+      "grad_norm": 2.619022846221924,
+      "learning_rate": 3.4336419753086427e-05,
+      "loss": 0.6836,
+      "step": 1100
+    },
+    {
+      "epoch": 3.1,
+      "eval_accuracy": 0.9305019305019305,
+      "eval_loss": 0.28447768092155457,
+      "eval_runtime": 298.4676,
+      "eval_samples_per_second": 0.868,
+      "eval_steps_per_second": 0.436,
+      "step": 1152
+    },
+    {
+      "epoch": 4.016666666666667,
+      "grad_norm": 6.768246173858643,
+      "learning_rate": 3.240740740740741e-05,
+      "loss": 0.728,
+      "step": 1200
+    },
+    {
+      "epoch": 4.051388888888889,
+      "grad_norm": 2.8624227046966553,
+      "learning_rate": 3.04783950617284e-05,
+      "loss": 0.3583,
+      "step": 1300
+    },
+    {
+      "epoch": 4.086111111111111,
+      "grad_norm": 34.53135299682617,
+      "learning_rate": 2.8549382716049384e-05,
+      "loss": 0.4155,
+      "step": 1400
+    },
+    {
+      "epoch": 4.1,
+      "eval_accuracy": 0.9498069498069498,
+      "eval_loss": 0.1982465386390686,
+      "eval_runtime": 287.9216,
+      "eval_samples_per_second": 0.9,
+      "eval_steps_per_second": 0.452,
+      "step": 1440
+    },
+    {
+      "epoch": 5.020833333333333,
+      "grad_norm": 0.724148690700531,
+      "learning_rate": 2.6620370370370372e-05,
+      "loss": 0.2925,
+      "step": 1500
+    },
+    {
+      "epoch": 5.055555555555555,
+      "grad_norm": 12.181313514709473,
+      "learning_rate": 2.4691358024691357e-05,
+      "loss": 0.2779,
+      "step": 1600
+    },
+    {
+      "epoch": 5.090277777777778,
+      "grad_norm": 76.32938385009766,
+      "learning_rate": 2.2762345679012348e-05,
+      "loss": 0.3,
+      "step": 1700
+    },
+    {
+      "epoch": 5.1,
+      "eval_accuracy": 0.9613899613899614,
+      "eval_loss": 0.18946108222007751,
+      "eval_runtime": 289.0676,
+      "eval_samples_per_second": 0.896,
+      "eval_steps_per_second": 0.45,
+      "step": 1728
+    },
+    {
+      "epoch": 6.025,
+      "grad_norm": 2.48795485496521,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.174,
+      "step": 1800
+    },
+    {
+      "epoch": 6.059722222222222,
+      "grad_norm": 13.161165237426758,
+      "learning_rate": 1.8904320987654324e-05,
+      "loss": 0.2328,
+      "step": 1900
+    },
+    {
+      "epoch": 6.094444444444444,
+      "grad_norm": 0.14844463765621185,
+      "learning_rate": 1.697530864197531e-05,
+      "loss": 0.201,
+      "step": 2000
+    },
+    {
+      "epoch": 6.1,
+      "eval_accuracy": 0.9575289575289575,
+      "eval_loss": 0.13380590081214905,
+      "eval_runtime": 294.1938,
+      "eval_samples_per_second": 0.88,
+      "eval_steps_per_second": 0.442,
+      "step": 2016
+    },
+    {
+      "epoch": 7.029166666666667,
+      "grad_norm": 0.25279873609542847,
+      "learning_rate": 1.5046296296296297e-05,
+      "loss": 0.1906,
+      "step": 2100
+    },
+    {
+      "epoch": 7.063888888888889,
+      "grad_norm": 0.11046076565980911,
+      "learning_rate": 1.3117283950617285e-05,
+      "loss": 0.0744,
+      "step": 2200
+    },
+    {
+      "epoch": 7.098611111111111,
+      "grad_norm": 0.08352793753147125,
+      "learning_rate": 1.1188271604938271e-05,
+      "loss": 0.1054,
+      "step": 2300
+    },
+    {
+      "epoch": 7.1,
+      "eval_accuracy": 0.9613899613899614,
+      "eval_loss": 0.1466270387172699,
+      "eval_runtime": 299.2254,
+      "eval_samples_per_second": 0.866,
+      "eval_steps_per_second": 0.434,
+      "step": 2304
+    },
+    {
+      "epoch": 8.033333333333333,
+      "grad_norm": 0.0984601229429245,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 0.0637,
+      "step": 2400
+    },
+    {
+      "epoch": 8.068055555555556,
+      "grad_norm": 0.1953054666519165,
+      "learning_rate": 7.330246913580248e-06,
+      "loss": 0.0754,
+      "step": 2500
+    },
+    {
+      "epoch": 8.1,
+      "eval_accuracy": 0.972972972972973,
+      "eval_loss": 0.12713134288787842,
+      "eval_runtime": 289.1913,
+      "eval_samples_per_second": 0.896,
+      "eval_steps_per_second": 0.45,
+      "step": 2592
+    },
+    {
+      "epoch": 9.002777777777778,
+      "grad_norm": 0.14724653959274292,
+      "learning_rate": 5.401234567901234e-06,
+      "loss": 0.1152,
+      "step": 2600
+    },
+    {
+      "epoch": 9.0375,
+      "grad_norm": 1.0079474449157715,
+      "learning_rate": 3.4722222222222224e-06,
+      "loss": 0.077,
+      "step": 2700
+    },
+    {
+      "epoch": 9.072222222222223,
+      "grad_norm": 0.1408146470785141,
+      "learning_rate": 1.5432098765432098e-06,
+      "loss": 0.1212,
+      "step": 2800
+    },
+    {
+      "epoch": 9.1,
+      "eval_accuracy": 0.9691119691119691,
+      "eval_loss": 0.12428595870733261,
+      "eval_runtime": 291.2158,
+      "eval_samples_per_second": 0.889,
+      "eval_steps_per_second": 0.446,
+      "step": 2880
+    },
+    {
+      "epoch": 9.1,
+      "step": 2880,
+      "total_flos": 2.8725202238381752e+19,
+      "train_loss": 2.8408549391561086,
+      "train_runtime": 32027.1475,
+      "train_samples_per_second": 0.719,
+      "train_steps_per_second": 0.09
+    },
+    {
+      "epoch": 9.1,
+      "eval_accuracy": 0.941717791411043,
+      "eval_loss": 0.16725970804691315,
+      "eval_runtime": 750.4557,
+      "eval_samples_per_second": 0.869,
+      "eval_steps_per_second": 0.434,
+      "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
+}