End of training

README.md

@@ -2,6 +2,8 @@
 license: apache-2.0
 base_model: google/vit-huge-patch14-224-in21k
 tags:
+- image-classification
+- vision
 - generated_from_trainer
 datasets:
 - imagefolder
@@ -14,7 +16,7 @@ model-index:
       name: Image Classification
       type: image-classification
     dataset:
-      name: imagefolder
+      name: touchtech/fashion-images-perspectives
       type: imagefolder
       config: default
       split: train
@@ -30,9 +32,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # fashion-images-perspectives-vit-huge-patch14-224-in21k
 
-This model is a fine-tuned version of [google/vit-huge-patch14-224-in21k](https://huggingface.co/google/vit-huge-patch14-224-in21k) on the imagefolder dataset.
+This model is a fine-tuned version of [google/vit-huge-patch14-224-in21k](https://huggingface.co/google/vit-huge-patch14-224-in21k) on the touchtech/fashion-images-perspectives dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.2710
+- Loss: 0.2604
 - Accuracy: 0.9150
 
 ## Model description

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 5.0,
+    "eval_accuracy": 0.9149976711690732,
+    "eval_loss": 0.26038262248039246,
+    "eval_runtime": 354.8337,
+    "eval_samples_per_second": 12.101,
+    "eval_steps_per_second": 1.513,
+    "train_loss": 0.4780347077960955,
+    "train_runtime": 18614.0668,
+    "train_samples_per_second": 6.535,
+    "train_steps_per_second": 0.817
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 5.0,
+    "eval_accuracy": 0.9149976711690732,
+    "eval_loss": 0.26038262248039246,
+    "eval_runtime": 354.8337,
+    "eval_samples_per_second": 12.101,
+    "eval_steps_per_second": 1.513
+}

train_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 5.0,
+    "train_loss": 0.4780347077960955,
+    "train_runtime": 18614.0668,
+    "train_samples_per_second": 6.535,
+    "train_steps_per_second": 0.817
+}

trainer_state.json

@@ -0,0 +1,253 @@
+{
+  "best_metric": 0.26038262248039246,
+  "best_model_checkpoint": "/workspace/training_output/perspectives-vit-huge-patch14-224-in21k/checkpoint-12168",
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 15210,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9342537804076265e-05,
+      "loss": 1.561,
+      "step": 500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.8685075608152533e-05,
+      "loss": 1.008,
+      "step": 1000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.80276134122288e-05,
+      "loss": 0.8167,
+      "step": 1500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7370151216305063e-05,
+      "loss": 0.717,
+      "step": 2000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.671268902038133e-05,
+      "loss": 0.6546,
+      "step": 2500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.6055226824457594e-05,
+      "loss": 0.5932,
+      "step": 3000
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.8889147647880764,
+      "eval_loss": 0.3862580955028534,
+      "eval_runtime": 426.1609,
+      "eval_samples_per_second": 10.076,
+      "eval_steps_per_second": 1.26,
+      "step": 3042
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.539776462853386e-05,
+      "loss": 0.5426,
+      "step": 3500
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.4740302432610125e-05,
+      "loss": 0.5235,
+      "step": 4000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.4082840236686392e-05,
+      "loss": 0.5008,
+      "step": 4500
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.3425378040762658e-05,
+      "loss": 0.4795,
+      "step": 5000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.2767915844838923e-05,
+      "loss": 0.4353,
+      "step": 5500
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.2110453648915189e-05,
+      "loss": 0.4286,
+      "step": 6000
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8642291569632045,
+      "eval_loss": 0.4201112985610962,
+      "eval_runtime": 405.9091,
+      "eval_samples_per_second": 10.579,
+      "eval_steps_per_second": 1.323,
+      "step": 6084
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.1452991452991454e-05,
+      "loss": 0.4019,
+      "step": 6500
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.0795529257067721e-05,
+      "loss": 0.4,
+      "step": 7000
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.0138067061143987e-05,
+      "loss": 0.3997,
+      "step": 7500
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 9.48060486522025e-06,
+      "loss": 0.3644,
+      "step": 8000
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 8.823142669296516e-06,
+      "loss": 0.3684,
+      "step": 8500
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 8.165680473372781e-06,
+      "loss": 0.3628,
+      "step": 9000
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9101071262226362,
+      "eval_loss": 0.28201931715011597,
+      "eval_runtime": 434.5616,
+      "eval_samples_per_second": 9.881,
+      "eval_steps_per_second": 1.236,
+      "step": 9126
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 7.508218277449047e-06,
+      "loss": 0.3414,
+      "step": 9500
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 6.850756081525313e-06,
+      "loss": 0.3458,
+      "step": 10000
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 6.193293885601579e-06,
+      "loss": 0.3484,
+      "step": 10500
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 5.535831689677844e-06,
+      "loss": 0.3679,
+      "step": 11000
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 4.878369493754109e-06,
+      "loss": 0.328,
+      "step": 11500
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 4.220907297830375e-06,
+      "loss": 0.3183,
+      "step": 12000
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.9149976711690732,
+      "eval_loss": 0.26038262248039246,
+      "eval_runtime": 392.7229,
+      "eval_samples_per_second": 10.934,
+      "eval_steps_per_second": 1.367,
+      "step": 12168
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 3.563445101906641e-06,
+      "loss": 0.3123,
+      "step": 12500
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 2.9059829059829063e-06,
+      "loss": 0.309,
+      "step": 13000
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 2.2485207100591717e-06,
+      "loss": 0.2992,
+      "step": 13500
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1.5910585141354374e-06,
+      "loss": 0.3119,
+      "step": 14000
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 9.335963182117029e-07,
+      "loss": 0.3192,
+      "step": 14500
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 2.7613412228796843e-07,
+      "loss": 0.2648,
+      "step": 15000
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.9149976711690732,
+      "eval_loss": 0.27095553278923035,
+      "eval_runtime": 390.7206,
+      "eval_samples_per_second": 10.99,
+      "eval_steps_per_second": 1.374,
+      "step": 15210
+    },
+    {
+      "epoch": 5.0,
+      "step": 15210,
+      "total_flos": 6.9304343223654605e+19,
+      "train_loss": 0.4780347077960955,
+      "train_runtime": 18614.0668,
+      "train_samples_per_second": 6.535,
+      "train_steps_per_second": 0.817
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 15210,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "total_flos": 6.9304343223654605e+19,
+  "trial_name": null,
+  "trial_params": null
+}