Training in progress, epoch 0

all_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 19.73,
-    "total_flos": 4.050283110270566e+16,
-    "train_loss": 0.24904835671186448,
-    "train_runtime": 1252.5827,
-    "train_samples_per_second": 10.921,
-    "train_steps_per_second": 0.032
+    "total_flos": 1.0463143975579976e+18,
+    "train_loss": 0.15636180818546563,
+    "train_runtime": 744.5374,
+    "train_samples_per_second": 18.374,
+    "train_steps_per_second": 0.054
 }

config.json

@@ -1,14 +1,10 @@
 {
-  "_name_or_path": "microsoft/beit-base-patch16-224",
+  "_name_or_path": "facebook/deit-base-patch16-224",
   "architectures": [
-    "BeitForImageClassification"
+    "ViTForImageClassification"
   ],
   "attention_probs_dropout_prob": 0.0,
-  "auxiliary_channels": 256,
-  "auxiliary_concat_input": false,
-  "auxiliary_loss_weight": 0.4,
-  "auxiliary_num_convs": 1,
-  "drop_path_rate": 0.1,
+  "encoder_stride": 16,
   "hidden_act": "gelu",
   "hidden_dropout_prob": 0.0,
   "hidden_size": 768,
@@ -24,33 +20,13 @@
     "1": "1"
   },
   "layer_norm_eps": 1e-12,
-  "layer_scale_init_value": 0.1,
-  "model_type": "beit",
+  "model_type": "vit",
   "num_attention_heads": 12,
   "num_channels": 3,
   "num_hidden_layers": 12,
-  "out_indices": [
-    3,
-    5,
-    7,
-    11
-  ],
   "patch_size": 16,
-  "pool_scales": [
-    1,
-    2,
-    3,
-    6
-  ],
   "problem_type": "single_label_classification",
-  "semantic_loss_ignore_index": 255,
+  "qkv_bias": true,
   "torch_dtype": "float32",
-  "transformers_version": "4.24.0",
-  "use_absolute_position_embeddings": false,
-  "use_auxiliary_head": true,
-  "use_mask_token": false,
-  "use_mean_pooling": true,
-  "use_relative_position_bias": true,
-  "use_shared_relative_position_bias": false,
-  "vocab_size": 8192
+  "transformers_version": "4.24.0"
 }

preprocessor_config.json

@@ -1,9 +1,7 @@
 {
-  "crop_size": 224,
-  "do_center_crop": false,
   "do_normalize": true,
   "do_resize": true,
-  "feature_extractor_type": "BeitFeatureExtractor",
+  "feature_extractor_type": "ViTFeatureExtractor",
   "image_mean": [
     0.5,
     0.5,
@@ -14,7 +12,6 @@
     0.5,
     0.5
   ],
-  "reduce_labels": false,
   "resample": 2,
   "size": 224
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6107934dd16374dafd647f11de87daab593abadcb701d311acbcd5aa56ea7fc1
-size 346860409
+oid sha256:f5a60e8db5ce0ed1a1f65e0017b7f4ebedb293cb52f8aefabaff42a39bfdbe89
+size 343268717

runs/Jul10_19-19-56_teesta/events.out.tfevents.1688997005.teesta.35180.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b3e03cba36a3085e592f1e63e5d312037da01fc88c5e3fa6fa3acf3e654ccb21
-size 12992
+oid sha256:73082cddfc5e135eb32f788629a190c4b7b0e87e63b46b22e5240097999944a7
+size 14420

runs/Jul10_19-45-51_teesta/1688998557.768573/events.out.tfevents.1688998557.teesta.35180.3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5d73d3467419920d314350c8538905f7814eaf2cec675d9418693e0d209313a
+size 5427

runs/Jul10_19-45-51_teesta/events.out.tfevents.1688998557.teesta.35180.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d0170a183c07052967337aa87d3bd905ea66097f17511a3f2cf247c755fa8a0
+size 4114

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 19.73,
-    "total_flos": 4.050283110270566e+16,
-    "train_loss": 0.24904835671186448,
-    "train_runtime": 1252.5827,
-    "train_samples_per_second": 10.921,
-    "train_steps_per_second": 0.032
+    "total_flos": 1.0463143975579976e+18,
+    "train_loss": 0.15636180818546563,
+    "train_runtime": 744.5374,
+    "train_samples_per_second": 18.374,
+    "train_steps_per_second": 0.054
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.9444444444444444,
-  "best_model_checkpoint": "test/checkpoint-14",
+  "best_metric": 0.9166666666666666,
+  "best_model_checkpoint": "test/checkpoint-20",
   "epoch": 19.727272727272727,
   "global_step": 40,
   "is_hyper_param_search": false,
@@ -9,281 +9,281 @@
   "log_history": [
     {
       "epoch": 0.73,
-      "eval_accuracy": 0.6111111111111112,
-      "eval_f1": 0.5416666666666666,
-      "eval_loss": 0.7271409630775452,
-      "eval_precision": 0.5208333333333334,
-      "eval_recall": 0.6111111111111112,
-      "eval_runtime": 6.3413,
-      "eval_samples_per_second": 5.677,
-      "eval_steps_per_second": 0.158,
+      "eval_accuracy": 0.7222222222222222,
+      "eval_f1": 0.6707818930041152,
+      "eval_loss": 0.561225950717926,
+      "eval_precision": 0.8090277777777778,
+      "eval_recall": 0.7222222222222222,
+      "eval_runtime": 5.7451,
+      "eval_samples_per_second": 6.266,
+      "eval_steps_per_second": 0.174,
       "step": 2
     },
     {
       "epoch": 1.73,
-      "eval_accuracy": 0.8333333333333334,
-      "eval_f1": 0.8148148148148149,
-      "eval_loss": 0.46258121728897095,
-      "eval_precision": 0.8666666666666667,
-      "eval_recall": 0.8333333333333334,
-      "eval_runtime": 5.8395,
-      "eval_samples_per_second": 6.165,
-      "eval_steps_per_second": 0.171,
+      "eval_accuracy": 0.7777777777777778,
+      "eval_f1": 0.7608080808080808,
+      "eval_loss": 0.3748733699321747,
+      "eval_precision": 0.7986111111111112,
+      "eval_recall": 0.7777777777777778,
+      "eval_runtime": 5.1904,
+      "eval_samples_per_second": 6.936,
+      "eval_steps_per_second": 0.193,
       "step": 4
     },
     {
       "epoch": 2.73,
-      "eval_accuracy": 0.9166666666666666,
-      "eval_f1": 0.9157054125998224,
-      "eval_loss": 0.3036719858646393,
-      "eval_precision": 0.9163636363636365,
-      "eval_recall": 0.9166666666666666,
-      "eval_runtime": 5.8411,
-      "eval_samples_per_second": 6.163,
-      "eval_steps_per_second": 0.171,
+      "eval_accuracy": 0.8055555555555556,
+      "eval_f1": 0.8066066066066067,
+      "eval_loss": 0.522664487361908,
+      "eval_precision": 0.8703703703703703,
+      "eval_recall": 0.8055555555555556,
+      "eval_runtime": 5.3959,
+      "eval_samples_per_second": 6.672,
+      "eval_steps_per_second": 0.185,
       "step": 6
     },
     {
       "epoch": 3.73,
-      "eval_accuracy": 0.9166666666666666,
-      "eval_f1": 0.9131652661064424,
-      "eval_loss": 0.24792712926864624,
-      "eval_precision": 0.9259259259259258,
-      "eval_recall": 0.9166666666666666,
-      "eval_runtime": 5.7041,
-      "eval_samples_per_second": 6.311,
-      "eval_steps_per_second": 0.175,
+      "eval_accuracy": 0.8055555555555556,
+      "eval_f1": 0.7864923747276689,
+      "eval_loss": 0.5397895574569702,
+      "eval_precision": 0.8524904214559387,
+      "eval_recall": 0.8055555555555556,
+      "eval_runtime": 5.524,
+      "eval_samples_per_second": 6.517,
+      "eval_steps_per_second": 0.181,
       "step": 8
     },
     {
       "epoch": 4.73,
       "learning_rate": 0.0001,
-      "loss": 0.6338,
+      "loss": 0.5735,
       "step": 10
     },
     {
       "epoch": 4.73,
-      "eval_accuracy": 0.9166666666666666,
-      "eval_f1": 0.9174468085106382,
-      "eval_loss": 0.23573820292949677,
-      "eval_precision": 0.919732441471572,
-      "eval_recall": 0.9166666666666666,
-      "eval_runtime": 5.6473,
-      "eval_samples_per_second": 6.375,
-      "eval_steps_per_second": 0.177,
+      "eval_accuracy": 0.8611111111111112,
+      "eval_f1": 0.8600823045267489,
+      "eval_loss": 0.24970261752605438,
+      "eval_precision": 0.8604607952434039,
+      "eval_recall": 0.8611111111111112,
+      "eval_runtime": 5.1179,
+      "eval_samples_per_second": 7.034,
+      "eval_steps_per_second": 0.195,
       "step": 10
     },
     {
       "epoch": 5.73,
       "eval_accuracy": 0.8888888888888888,
-      "eval_f1": 0.890746934225195,
-      "eval_loss": 0.21319927275180817,
-      "eval_precision": 0.8982683982683982,
+      "eval_f1": 0.8899470899470899,
+      "eval_loss": 0.26077044010162354,
+      "eval_precision": 0.8965277777777777,
       "eval_recall": 0.8888888888888888,
-      "eval_runtime": 6.2525,
-      "eval_samples_per_second": 5.758,
-      "eval_steps_per_second": 0.16,
+      "eval_runtime": 4.7095,
+      "eval_samples_per_second": 7.644,
+      "eval_steps_per_second": 0.212,
       "step": 12
     },
     {
       "epoch": 6.73,
-      "eval_accuracy": 0.9444444444444444,
-      "eval_f1": 0.9430303030303031,
-      "eval_loss": 0.17125670611858368,
-      "eval_precision": 0.9487179487179487,
-      "eval_recall": 0.9444444444444444,
-      "eval_runtime": 5.7878,
-      "eval_samples_per_second": 6.22,
-      "eval_steps_per_second": 0.173,
+      "eval_accuracy": 0.8611111111111112,
+      "eval_f1": 0.8618907600463335,
+      "eval_loss": 0.2541900873184204,
+      "eval_precision": 0.8640211640211641,
+      "eval_recall": 0.8611111111111112,
+      "eval_runtime": 5.4129,
+      "eval_samples_per_second": 6.651,
+      "eval_steps_per_second": 0.185,
       "step": 14
     },
     {
       "epoch": 7.73,
-      "eval_accuracy": 0.9444444444444444,
-      "eval_f1": 0.9444444444444444,
-      "eval_loss": 0.15498512983322144,
-      "eval_precision": 0.9444444444444444,
-      "eval_recall": 0.9444444444444444,
-      "eval_runtime": 6.2641,
-      "eval_samples_per_second": 5.747,
-      "eval_steps_per_second": 0.16,
+      "eval_accuracy": 0.8611111111111112,
+      "eval_f1": 0.857210401891253,
+      "eval_loss": 0.2999325394630432,
+      "eval_precision": 0.8668686868686869,
+      "eval_recall": 0.8611111111111112,
+      "eval_runtime": 5.0159,
+      "eval_samples_per_second": 7.177,
+      "eval_steps_per_second": 0.199,
       "step": 16
     },
     {
       "epoch": 8.73,
-      "eval_accuracy": 0.9444444444444444,
-      "eval_f1": 0.9444444444444444,
-      "eval_loss": 0.14064311981201172,
-      "eval_precision": 0.9444444444444444,
-      "eval_recall": 0.9444444444444444,
-      "eval_runtime": 5.34,
-      "eval_samples_per_second": 6.742,
-      "eval_steps_per_second": 0.187,
+      "eval_accuracy": 0.8888888888888888,
+      "eval_f1": 0.8870308435525827,
+      "eval_loss": 0.2993186116218567,
+      "eval_precision": 0.8912037037037036,
+      "eval_recall": 0.8888888888888888,
+      "eval_runtime": 4.9756,
+      "eval_samples_per_second": 7.235,
+      "eval_steps_per_second": 0.201,
       "step": 18
     },
     {
       "epoch": 9.73,
       "learning_rate": 6.666666666666667e-05,
-      "loss": 0.217,
+      "loss": 0.0476,
       "step": 20
     },
     {
       "epoch": 9.73,
-      "eval_accuracy": 0.9444444444444444,
-      "eval_f1": 0.9444444444444444,
-      "eval_loss": 0.16372796893119812,
-      "eval_precision": 0.9444444444444444,
-      "eval_recall": 0.9444444444444444,
-      "eval_runtime": 5.2833,
-      "eval_samples_per_second": 6.814,
-      "eval_steps_per_second": 0.189,
+      "eval_accuracy": 0.9166666666666666,
+      "eval_f1": 0.9160493827160493,
+      "eval_loss": 0.2704654335975647,
+      "eval_precision": 0.9169453734671126,
+      "eval_recall": 0.9166666666666666,
+      "eval_runtime": 6.2342,
+      "eval_samples_per_second": 5.775,
+      "eval_steps_per_second": 0.16,
       "step": 20
     },
     {
       "epoch": 10.73,
-      "eval_accuracy": 0.9166666666666666,
-      "eval_f1": 0.9157054125998224,
-      "eval_loss": 0.16719864308834076,
-      "eval_precision": 0.9163636363636365,
-      "eval_recall": 0.9166666666666666,
-      "eval_runtime": 6.4055,
-      "eval_samples_per_second": 5.62,
-      "eval_steps_per_second": 0.156,
+      "eval_accuracy": 0.8888888888888888,
+      "eval_f1": 0.8888888888888888,
+      "eval_loss": 0.28197962045669556,
+      "eval_precision": 0.8888888888888888,
+      "eval_recall": 0.8888888888888888,
+      "eval_runtime": 5.1606,
+      "eval_samples_per_second": 6.976,
+      "eval_steps_per_second": 0.194,
       "step": 22
     },
     {
       "epoch": 11.73,
-      "eval_accuracy": 0.9444444444444444,
-      "eval_f1": 0.9444444444444444,
-      "eval_loss": 0.13150224089622498,
-      "eval_precision": 0.9444444444444444,
-      "eval_recall": 0.9444444444444444,
-      "eval_runtime": 5.1218,
-      "eval_samples_per_second": 7.029,
-      "eval_steps_per_second": 0.195,
+      "eval_accuracy": 0.8611111111111112,
+      "eval_f1": 0.8618907600463335,
+      "eval_loss": 0.275021493434906,
+      "eval_precision": 0.8640211640211641,
+      "eval_recall": 0.8611111111111112,
+      "eval_runtime": 4.8026,
+      "eval_samples_per_second": 7.496,
+      "eval_steps_per_second": 0.208,
       "step": 24
     },
     {
       "epoch": 12.73,
-      "eval_accuracy": 0.9444444444444444,
-      "eval_f1": 0.9444444444444444,
-      "eval_loss": 0.1597292125225067,
-      "eval_precision": 0.9444444444444444,
-      "eval_recall": 0.9444444444444444,
-      "eval_runtime": 5.1268,
-      "eval_samples_per_second": 7.022,
-      "eval_steps_per_second": 0.195,
+      "eval_accuracy": 0.8888888888888888,
+      "eval_f1": 0.8888888888888888,
+      "eval_loss": 0.2932831346988678,
+      "eval_precision": 0.8888888888888888,
+      "eval_recall": 0.8888888888888888,
+      "eval_runtime": 4.6825,
+      "eval_samples_per_second": 7.688,
+      "eval_steps_per_second": 0.214,
       "step": 26
     },
     {
       "epoch": 13.73,
-      "eval_accuracy": 0.9444444444444444,
-      "eval_f1": 0.9444444444444444,
-      "eval_loss": 0.1412082463502884,
-      "eval_precision": 0.9444444444444444,
-      "eval_recall": 0.9444444444444444,
-      "eval_runtime": 5.6714,
-      "eval_samples_per_second": 6.348,
-      "eval_steps_per_second": 0.176,
+      "eval_accuracy": 0.8611111111111112,
+      "eval_f1": 0.8618907600463335,
+      "eval_loss": 0.2971099317073822,
+      "eval_precision": 0.8640211640211641,
+      "eval_recall": 0.8611111111111112,
+      "eval_runtime": 5.7546,
+      "eval_samples_per_second": 6.256,
+      "eval_steps_per_second": 0.174,
       "step": 28
     },
     {
       "epoch": 14.73,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 0.0906,
+      "loss": 0.003,
       "step": 30
     },
     {
       "epoch": 14.73,
-      "eval_accuracy": 0.9444444444444444,
-      "eval_f1": 0.9444444444444444,
-      "eval_loss": 0.1776580661535263,
-      "eval_precision": 0.9444444444444444,
-      "eval_recall": 0.9444444444444444,
-      "eval_runtime": 5.8702,
-      "eval_samples_per_second": 6.133,
-      "eval_steps_per_second": 0.17,
+      "eval_accuracy": 0.8888888888888888,
+      "eval_f1": 0.8888888888888888,
+      "eval_loss": 0.29043230414390564,
+      "eval_precision": 0.8888888888888888,
+      "eval_recall": 0.8888888888888888,
+      "eval_runtime": 9.7959,
+      "eval_samples_per_second": 3.675,
+      "eval_steps_per_second": 0.102,
       "step": 30
     },
     {
       "epoch": 15.73,
-      "eval_accuracy": 0.9444444444444444,
-      "eval_f1": 0.9444444444444444,
-      "eval_loss": 0.14607585966587067,
-      "eval_precision": 0.9444444444444444,
-      "eval_recall": 0.9444444444444444,
-      "eval_runtime": 7.113,
-      "eval_samples_per_second": 5.061,
-      "eval_steps_per_second": 0.141,
+      "eval_accuracy": 0.9166666666666666,
+      "eval_f1": 0.9160493827160493,
+      "eval_loss": 0.305411696434021,
+      "eval_precision": 0.9169453734671126,
+      "eval_recall": 0.9166666666666666,
+      "eval_runtime": 5.2177,
+      "eval_samples_per_second": 6.9,
+      "eval_steps_per_second": 0.192,
       "step": 32
     },
     {
       "epoch": 16.73,
       "eval_accuracy": 0.9166666666666666,
-      "eval_f1": 0.9174468085106382,
-      "eval_loss": 0.19147923588752747,
-      "eval_precision": 0.919732441471572,
+      "eval_f1": 0.9160493827160493,
+      "eval_loss": 0.29818451404571533,
+      "eval_precision": 0.9169453734671126,
       "eval_recall": 0.9166666666666666,
-      "eval_runtime": 5.3662,
-      "eval_samples_per_second": 6.709,
-      "eval_steps_per_second": 0.186,
+      "eval_runtime": 4.9626,
+      "eval_samples_per_second": 7.254,
+      "eval_steps_per_second": 0.202,
       "step": 34
     },
     {
       "epoch": 17.73,
       "eval_accuracy": 0.9166666666666666,
-      "eval_f1": 0.9174468085106382,
-      "eval_loss": 0.13753658533096313,
-      "eval_precision": 0.919732441471572,
+      "eval_f1": 0.9160493827160493,
+      "eval_loss": 0.30663102865219116,
+      "eval_precision": 0.9169453734671126,
       "eval_recall": 0.9166666666666666,
-      "eval_runtime": 5.3487,
-      "eval_samples_per_second": 6.731,
-      "eval_steps_per_second": 0.187,
+      "eval_runtime": 4.7376,
+      "eval_samples_per_second": 7.599,
+      "eval_steps_per_second": 0.211,
       "step": 36
     },
     {
       "epoch": 18.73,
-      "eval_accuracy": 0.9166666666666666,
-      "eval_f1": 0.9174468085106382,
-      "eval_loss": 0.1448293775320053,
-      "eval_precision": 0.919732441471572,
-      "eval_recall": 0.9166666666666666,
-      "eval_runtime": 5.5619,
-      "eval_samples_per_second": 6.473,
-      "eval_steps_per_second": 0.18,
+      "eval_accuracy": 0.8888888888888888,
+      "eval_f1": 0.8870308435525827,
+      "eval_loss": 0.30919915437698364,
+      "eval_precision": 0.8912037037037036,
+      "eval_recall": 0.8888888888888888,
+      "eval_runtime": 6.3085,
+      "eval_samples_per_second": 5.707,
+      "eval_steps_per_second": 0.159,
       "step": 38
     },
     {
       "epoch": 19.73,
       "learning_rate": 0.0,
-      "loss": 0.0548,
+      "loss": 0.0013,
       "step": 40
     },
     {
       "epoch": 19.73,
-      "eval_accuracy": 0.9444444444444444,
-      "eval_f1": 0.9444444444444444,
-      "eval_loss": 0.12921589612960815,
-      "eval_precision": 0.9444444444444444,
-      "eval_recall": 0.9444444444444444,
-      "eval_runtime": 5.8187,
-      "eval_samples_per_second": 6.187,
-      "eval_steps_per_second": 0.172,
+      "eval_accuracy": 0.8888888888888888,
+      "eval_f1": 0.8870308435525827,
+      "eval_loss": 0.3030144274234772,
+      "eval_precision": 0.8912037037037036,
+      "eval_recall": 0.8888888888888888,
+      "eval_runtime": 4.7135,
+      "eval_samples_per_second": 7.638,
+      "eval_steps_per_second": 0.212,
       "step": 40
     },
     {
       "epoch": 19.73,
       "step": 40,
-      "total_flos": 4.050283110270566e+16,
-      "train_loss": 0.24904835671186448,
-      "train_runtime": 1252.5827,
-      "train_samples_per_second": 10.921,
-      "train_steps_per_second": 0.032
+      "total_flos": 1.0463143975579976e+18,
+      "train_loss": 0.15636180818546563,
+      "train_runtime": 744.5374,
+      "train_samples_per_second": 18.374,
+      "train_steps_per_second": 0.054
     }
   ],
   "max_steps": 40,
   "num_train_epochs": 20,
-  "total_flos": 4.050283110270566e+16,
+  "total_flos": 1.0463143975579976e+18,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6e79370bcf724d04d0deeebfd7cb730102e73af0e10cf634a2778707017455f8
+oid sha256:7461b0cb9b4b366e4551fef7c56538fb52d6a1008d314c875bdc23478b386ac2
 size 3387