Training in progress, epoch 1

all_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 9.23076923076923,
-    "eval_accuracy": 0.6086956521739131,
-    "eval_loss": 1.754765510559082,
-    "eval_runtime": 29.8242,
-    "eval_samples_per_second": 0.771,
-    "eval_steps_per_second": 0.101
+    "epoch": 7.384615384615385,
+    "eval_accuracy": 0.6956521739130435,
+    "eval_loss": 1.0403441190719604,
+    "eval_runtime": 25.3548,
+    "eval_samples_per_second": 0.907,
+    "eval_steps_per_second": 0.118
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9e2c17c444e9504dc5b7c2ff332d98e5deff4364522ebccdad09fbb99201cc32
+oid sha256:4ca87bf45a96137740aec1eafeab8faee9102ce90215e675f09f94174732f537
 size 344940444

test_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 9.23076923076923,
-    "eval_accuracy": 0.6086956521739131,
-    "eval_loss": 1.754765510559082,
-    "eval_runtime": 29.8242,
-    "eval_samples_per_second": 0.771,
-    "eval_steps_per_second": 0.101
+    "epoch": 7.384615384615385,
+    "eval_accuracy": 0.6956521739130435,
+    "eval_loss": 1.0403441190719604,
+    "eval_runtime": 25.3548,
+    "eval_samples_per_second": 0.907,
+    "eval_steps_per_second": 0.118
 }

trainer_state.json

@@ -1,210 +1,171 @@
 {
   "best_metric": 0.9166666666666666,
-  "best_model_checkpoint": "videomae-base-finetuned-soccer\\checkpoint-39",
-  "epoch": 9.23076923076923,
+  "best_model_checkpoint": "videomae-base-finetuned-soccer\\checkpoint-65",
+  "epoch": 7.384615384615385,
   "eval_steps": 500,
-  "global_step": 120,
+  "global_step": 96,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.7692307692307693,
-      "grad_norm": 1.7203185558319092,
-      "learning_rate": 4.166666666666667e-05,
-      "loss": 0.1996,
+      "grad_norm": 5.654183864593506,
+      "learning_rate": 5e-05,
+      "loss": 1.0527,
       "step": 10
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.8333333333333334,
-      "eval_loss": 0.4402236044406891,
-      "eval_runtime": 14.1724,
-      "eval_samples_per_second": 0.847,
-      "eval_steps_per_second": 0.141,
+      "eval_accuracy": 0.5833333333333334,
+      "eval_loss": 1.025175929069519,
+      "eval_runtime": 14.1316,
+      "eval_samples_per_second": 0.849,
+      "eval_steps_per_second": 0.142,
       "step": 13
     },
     {
       "epoch": 1.5384615384615383,
-      "grad_norm": 30.03859519958496,
-      "learning_rate": 4.62962962962963e-05,
-      "loss": 0.2638,
+      "grad_norm": 9.78184986114502,
+      "learning_rate": 4.418604651162791e-05,
+      "loss": 0.6937,
       "step": 20
     },
     {
       "epoch": 2.0,
       "eval_accuracy": 0.8333333333333334,
-      "eval_loss": 0.5667647123336792,
-      "eval_runtime": 14.0809,
-      "eval_samples_per_second": 0.852,
-      "eval_steps_per_second": 0.142,
+      "eval_loss": 0.5850326418876648,
+      "eval_runtime": 13.5201,
+      "eval_samples_per_second": 0.888,
+      "eval_steps_per_second": 0.148,
       "step": 26
     },
     {
       "epoch": 2.3076923076923075,
-      "grad_norm": 0.46670857071876526,
-      "learning_rate": 4.166666666666667e-05,
-      "loss": 0.0971,
+      "grad_norm": 2.8956298828125,
+      "learning_rate": 3.837209302325582e-05,
+      "loss": 0.4002,
       "step": 30
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.9166666666666666,
-      "eval_loss": 0.24797028303146362,
-      "eval_runtime": 14.1385,
-      "eval_samples_per_second": 0.849,
-      "eval_steps_per_second": 0.141,
+      "eval_accuracy": 0.6666666666666666,
+      "eval_loss": 0.5491077303886414,
+      "eval_runtime": 13.511,
+      "eval_samples_per_second": 0.888,
+      "eval_steps_per_second": 0.148,
       "step": 39
     },
     {
       "epoch": 3.076923076923077,
-      "grad_norm": 0.8899642825126648,
-      "learning_rate": 3.7037037037037037e-05,
-      "loss": 0.0735,
+      "grad_norm": 5.322183609008789,
+      "learning_rate": 3.2558139534883724e-05,
+      "loss": 0.3079,
       "step": 40
     },
     {
       "epoch": 3.8461538461538463,
-      "grad_norm": 1.1780180931091309,
-      "learning_rate": 3.240740740740741e-05,
-      "loss": 0.0905,
+      "grad_norm": 8.3604097366333,
+      "learning_rate": 2.674418604651163e-05,
+      "loss": 0.216,
       "step": 50
     },
     {
       "epoch": 4.0,
       "eval_accuracy": 0.8333333333333334,
-      "eval_loss": 0.5412092804908752,
-      "eval_runtime": 14.4364,
-      "eval_samples_per_second": 0.831,
-      "eval_steps_per_second": 0.139,
+      "eval_loss": 0.4196215867996216,
+      "eval_runtime": 13.9122,
+      "eval_samples_per_second": 0.863,
+      "eval_steps_per_second": 0.144,
       "step": 52
     },
     {
       "epoch": 4.615384615384615,
-      "grad_norm": 13.846024513244629,
-      "learning_rate": 2.777777777777778e-05,
-      "loss": 0.0422,
+      "grad_norm": 9.210854530334473,
+      "learning_rate": 2.0930232558139536e-05,
+      "loss": 0.0978,
       "step": 60
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.8333333333333334,
-      "eval_loss": 0.49132850766181946,
-      "eval_runtime": 18.6583,
-      "eval_samples_per_second": 0.643,
-      "eval_steps_per_second": 0.107,
+      "eval_accuracy": 0.9166666666666666,
+      "eval_loss": 0.38431987166404724,
+      "eval_runtime": 13.7357,
+      "eval_samples_per_second": 0.874,
+      "eval_steps_per_second": 0.146,
       "step": 65
     },
     {
       "epoch": 5.384615384615385,
-      "grad_norm": 0.31654420495033264,
-      "learning_rate": 2.314814814814815e-05,
-      "loss": 0.0129,
+      "grad_norm": 2.0958056449890137,
+      "learning_rate": 1.5116279069767441e-05,
+      "loss": 0.0927,
       "step": 70
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.75,
-      "eval_loss": 0.46807315945625305,
-      "eval_runtime": 14.5391,
-      "eval_samples_per_second": 0.825,
-      "eval_steps_per_second": 0.138,
+      "eval_accuracy": 0.9166666666666666,
+      "eval_loss": 0.28933361172676086,
+      "eval_runtime": 13.6757,
+      "eval_samples_per_second": 0.877,
+      "eval_steps_per_second": 0.146,
       "step": 78
     },
     {
       "epoch": 6.153846153846154,
-      "grad_norm": 0.08448699116706848,
-      "learning_rate": 1.8518518518518518e-05,
-      "loss": 0.0072,
+      "grad_norm": 0.5293005108833313,
+      "learning_rate": 9.302325581395349e-06,
+      "loss": 0.0614,
       "step": 80
     },
     {
       "epoch": 6.923076923076923,
-      "grad_norm": 0.03759448230266571,
-      "learning_rate": 1.388888888888889e-05,
-      "loss": 0.0034,
+      "grad_norm": 0.14121398329734802,
+      "learning_rate": 3.488372093023256e-06,
+      "loss": 0.0588,
       "step": 90
     },
     {
       "epoch": 7.0,
       "eval_accuracy": 0.8333333333333334,
-      "eval_loss": 0.5456988215446472,
-      "eval_runtime": 14.3366,
-      "eval_samples_per_second": 0.837,
-      "eval_steps_per_second": 0.14,
-      "step": 91
-    },
-    {
-      "epoch": 7.6923076923076925,
-      "grad_norm": 0.026149122044444084,
-      "learning_rate": 9.259259259259259e-06,
-      "loss": 0.0025,
-      "step": 100
-    },
-    {
-      "epoch": 8.0,
-      "eval_accuracy": 0.8333333333333334,
-      "eval_loss": 0.5663144588470459,
-      "eval_runtime": 18.3063,
-      "eval_samples_per_second": 0.656,
-      "eval_steps_per_second": 0.109,
-      "step": 104
-    },
-    {
-      "epoch": 8.461538461538462,
-      "grad_norm": 0.036780279129743576,
-      "learning_rate": 4.6296296296296296e-06,
-      "loss": 0.0022,
-      "step": 110
-    },
-    {
-      "epoch": 9.0,
-      "eval_accuracy": 0.8333333333333334,
-      "eval_loss": 0.5768366456031799,
-      "eval_runtime": 13.6506,
-      "eval_samples_per_second": 0.879,
+      "eval_loss": 0.351638525724411,
+      "eval_runtime": 13.5976,
+      "eval_samples_per_second": 0.883,
       "eval_steps_per_second": 0.147,
-      "step": 117
-    },
-    {
-      "epoch": 9.23076923076923,
-      "grad_norm": 0.027885640040040016,
-      "learning_rate": 0.0,
-      "loss": 0.002,
-      "step": 120
+      "step": 91
     },
     {
-      "epoch": 9.23076923076923,
+      "epoch": 7.384615384615385,
       "eval_accuracy": 0.8333333333333334,
-      "eval_loss": 0.5772767066955566,
-      "eval_runtime": 15.1473,
-      "eval_samples_per_second": 0.792,
-      "eval_steps_per_second": 0.132,
-      "step": 120
-    },
-    {
-      "epoch": 9.23076923076923,
-      "step": 120,
-      "total_flos": 1.117729904739582e+18,
-      "train_loss": 0.0664138655954351,
-      "train_runtime": 1626.004,
-      "train_samples_per_second": 0.59,
-      "train_steps_per_second": 0.074
-    },
-    {
-      "epoch": 9.23076923076923,
-      "eval_accuracy": 0.6086956521739131,
-      "eval_loss": 1.754765510559082,
-      "eval_runtime": 29.8242,
-      "eval_samples_per_second": 0.771,
-      "eval_steps_per_second": 0.101,
-      "step": 120
+      "eval_loss": 0.35855498909950256,
+      "eval_runtime": 14.7901,
+      "eval_samples_per_second": 0.811,
+      "eval_steps_per_second": 0.135,
+      "step": 96
+    },
+    {
+      "epoch": 7.384615384615385,
+      "step": 96,
+      "total_flos": 8.959284297745367e+17,
+      "train_loss": 0.31444934258858365,
+      "train_runtime": 1126.2966,
+      "train_samples_per_second": 0.682,
+      "train_steps_per_second": 0.085
+    },
+    {
+      "epoch": 7.384615384615385,
+      "eval_accuracy": 0.6956521739130435,
+      "eval_loss": 1.0403441190719604,
+      "eval_runtime": 25.3548,
+      "eval_samples_per_second": 0.907,
+      "eval_steps_per_second": 0.118,
+      "step": 96
     }
   ],
   "logging_steps": 10,
-  "max_steps": 120,
+  "max_steps": 96,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 10,
+  "num_train_epochs": 8,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -218,7 +179,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.117729904739582e+18,
+  "total_flos": 8.959284297745367e+17,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6443fe9f698c13a093227a508cc3abfbd53a0433b89867195dca97ca58c600ed
+oid sha256:58fcc67e1cda0d5be48dc9fda44ccd7ec38782e45dcb68bebda93c16e9d0bd8e
 size 5368