Evaluation on the test set completed on 2025_05_06.

README.md

@@ -1,4 +1,121 @@
 ---
+library_name: transformers
+license: apache-2.0
+base_model: facebook/dinov2-small
 tags:
-- hf-summary-writer
+- generated_from_trainer
+metrics:
+- accuracy
+model-index:
+- name: DinoAmoros-small-2025_05_06_36794-prova_bs16_freeze_monolabel
+  results: []
 ---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# DinoAmoros-small-2025_05_06_36794-prova_bs16_freeze_monolabel
+
+This model is a fine-tuned version of [facebook/dinov2-small](https://huggingface.co/facebook/dinov2-small) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 2.8192
+- F1 Micro: 0.4
+- F1 Macro: 0.2333
+- Accuracy: 0.4
+- Learning Rate: 1e-05
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 0.001
+- train_batch_size: 16
+- eval_batch_size: 16
+- seed: 42
+- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: linear
+- num_epochs: 150
+- mixed_precision_training: Native AMP
+
+### Training results
+
+| Training Loss | Epoch | Step | Validation Loss | F1 Micro | F1 Macro | Accuracy | Rate   |
+|:-------------:|:-----:|:----:|:---------------:|:--------:|:--------:|:--------:|:------:|
+| No log        | 1.0   | 1    | 3.1980          | 0.0      | 0.0      | 0.0      | 0.001  |
+| No log        | 2.0   | 2    | 3.1316          | 0.0      | 0.0      | 0.0      | 0.001  |
+| No log        | 3.0   | 3    | 3.125           | 0.0      | 0.0      | 0.0      | 0.001  |
+| No log        | 4.0   | 4    | 3.1275          | 0.0      | 0.0      | 0.0      | 0.001  |
+| No log        | 5.0   | 5    | 3.0676          | 0.1      | 0.0571   | 0.1      | 0.001  |
+| No log        | 6.0   | 6    | 3.0656          | 0.1      | 0.0571   | 0.1      | 0.001  |
+| No log        | 7.0   | 7    | 3.0043          | 0.1      | 0.05     | 0.1      | 0.001  |
+| No log        | 8.0   | 8    | 2.9486          | 0.3      | 0.1531   | 0.3      | 0.001  |
+| No log        | 9.0   | 9    | 2.8736          | 0.4      | 0.2816   | 0.4      | 0.001  |
+| No log        | 10.0  | 10   | 2.8121          | 0.4      | 0.2816   | 0.4      | 0.001  |
+| No log        | 11.0  | 11   | 2.7541          | 0.6      | 0.4028   | 0.6      | 0.001  |
+| No log        | 12.0  | 12   | 2.6967          | 0.6      | 0.4028   | 0.6      | 0.001  |
+| No log        | 13.0  | 13   | 2.6596          | 0.6      | 0.4841   | 0.6      | 0.001  |
+| No log        | 14.0  | 14   | 2.6483          | 0.5      | 0.3571   | 0.5      | 0.001  |
+| No log        | 15.0  | 15   | 2.6144          | 0.5      | 0.3571   | 0.5      | 0.001  |
+| No log        | 16.0  | 16   | 2.5909          | 0.5      | 0.3571   | 0.5      | 0.001  |
+| No log        | 17.0  | 17   | 2.5481          | 0.5      | 0.4286   | 0.5      | 0.001  |
+| No log        | 18.0  | 18   | 2.5126          | 0.5      | 0.3619   | 0.5      | 0.001  |
+| No log        | 19.0  | 19   | 2.4791          | 0.5      | 0.4286   | 0.5      | 0.001  |
+| No log        | 20.0  | 20   | 2.4738          | 0.5      | 0.4286   | 0.5      | 0.001  |
+| No log        | 21.0  | 21   | 2.4310          | 0.5      | 0.4286   | 0.5      | 0.001  |
+| No log        | 22.0  | 22   | 2.4030          | 0.5      | 0.4286   | 0.5      | 0.001  |
+| No log        | 23.0  | 23   | 2.4001          | 0.5      | 0.4286   | 0.5      | 0.001  |
+| No log        | 24.0  | 24   | 2.3993          | 0.5      | 0.4286   | 0.5      | 0.001  |
+| No log        | 25.0  | 25   | 2.3928          | 0.5      | 0.4286   | 0.5      | 0.001  |
+| No log        | 26.0  | 26   | 2.3896          | 0.5      | 0.4286   | 0.5      | 0.001  |
+| No log        | 27.0  | 27   | 2.3909          | 0.5      | 0.4286   | 0.5      | 0.001  |
+| No log        | 28.0  | 28   | 2.3772          | 0.5      | 0.4286   | 0.5      | 0.001  |
+| No log        | 29.0  | 29   | 2.3432          | 0.5      | 0.4286   | 0.5      | 0.001  |
+| No log        | 30.0  | 30   | 2.3192          | 0.5      | 0.4286   | 0.5      | 0.001  |
+| No log        | 31.0  | 31   | 2.3088          | 0.6      | 0.5476   | 0.6      | 0.001  |
+| No log        | 32.0  | 32   | 2.3004          | 0.6      | 0.5476   | 0.6      | 0.001  |
+| No log        | 33.0  | 33   | 2.3044          | 0.6      | 0.5476   | 0.6      | 0.001  |
+| No log        | 34.0  | 34   | 2.2979          | 0.6      | 0.5476   | 0.6      | 0.001  |
+| No log        | 35.0  | 35   | 2.3048          | 0.6      | 0.5476   | 0.6      | 0.001  |
+| No log        | 36.0  | 36   | 2.2987          | 0.6      | 0.5476   | 0.6      | 0.001  |
+| No log        | 37.0  | 37   | 2.2997          | 0.6      | 0.5476   | 0.6      | 0.001  |
+| No log        | 38.0  | 38   | 2.3195          | 0.6      | 0.5476   | 0.6      | 0.001  |
+| No log        | 39.0  | 39   | 2.3158          | 0.5      | 0.4286   | 0.5      | 0.001  |
+| No log        | 40.0  | 40   | 2.3083          | 0.5      | 0.4286   | 0.5      | 0.001  |
+| No log        | 41.0  | 41   | 2.2830          | 0.5      | 0.4286   | 0.5      | 0.0001 |
+| No log        | 42.0  | 42   | 2.2719          | 0.5      | 0.4286   | 0.5      | 0.0001 |
+| No log        | 43.0  | 43   | 2.2404          | 0.5      | 0.4286   | 0.5      | 0.0001 |
+| No log        | 44.0  | 44   | 2.2439          | 0.5      | 0.4286   | 0.5      | 0.0001 |
+| No log        | 45.0  | 45   | 2.2249          | 0.5      | 0.4286   | 0.5      | 0.0001 |
+| No log        | 46.0  | 46   | 2.2116          | 0.5      | 0.4286   | 0.5      | 0.0001 |
+| No log        | 47.0  | 47   | 2.1979          | 0.5      | 0.4286   | 0.5      | 0.0001 |
+| No log        | 48.0  | 48   | 2.2088          | 0.5      | 0.4286   | 0.5      | 0.0001 |
+| No log        | 49.0  | 49   | 2.2075          | 0.5      | 0.4286   | 0.5      | 0.0001 |
+| No log        | 50.0  | 50   | 2.2067          | 0.5      | 0.4286   | 0.5      | 0.0001 |
+| No log        | 51.0  | 51   | 2.2182          | 0.5      | 0.4286   | 0.5      | 0.0001 |
+| No log        | 52.0  | 52   | 2.2243          | 0.5      | 0.4286   | 0.5      | 0.0001 |
+| No log        | 53.0  | 53   | 2.2344          | 0.5      | 0.4286   | 0.5      | 0.0001 |
+| No log        | 54.0  | 54   | 2.2222          | 0.5      | 0.4286   | 0.5      | 1e-05  |
+| No log        | 55.0  | 55   | 2.2211          | 0.5      | 0.4286   | 0.5      | 1e-05  |
+| No log        | 56.0  | 56   | 2.2072          | 0.5      | 0.4286   | 0.5      | 1e-05  |
+| No log        | 57.0  | 57   | 2.2094          | 0.5      | 0.4286   | 0.5      | 1e-05  |
+
+
+### Framework versions
+
+- Transformers 4.48.0
+- Pytorch 2.6.0+cu118
+- Datasets 3.0.2
+- Tokenizers 0.21.1

all_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 57.0,
+    "eval_accuracy": 0.4,
+    "eval_f1_macro": 0.23333333333333334,
+    "eval_f1_micro": 0.4,
+    "eval_loss": 2.8191895484924316,
+    "eval_runtime": 0.3171,
+    "eval_samples_per_second": 31.535,
+    "eval_steps_per_second": 3.153,
+    "learning_rate": 1e-05,
+    "total_flos": 1.132788066729984e+16,
+    "train_loss": 1.0105553342584979,
+    "train_runtime": 99.3744,
+    "train_samples_per_second": 15.094,
+    "train_steps_per_second": 1.509
+}

config.json

@@ -0,0 +1,101 @@
+{
+  "_name_or_path": "facebook/dinov2-small",
+  "apply_layernorm": true,
+  "architectures": [
+    "Dinov2ForImageClassification"
+  ],
+  "attention_probs_dropout_prob": 0.0,
+  "drop_path_rate": 0.0,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.0,
+  "hidden_size": 384,
+  "id2label": {
+    "0": "ALGAE",
+    "1": "Acr",
+    "2": "Acr_Br",
+    "3": "Anem",
+    "4": "CCA",
+    "5": "Ech",
+    "6": "Fts",
+    "7": "Gal",
+    "8": "Gon",
+    "9": "H_oval",
+    "10": "H_uni",
+    "11": "Mtp",
+    "12": "P",
+    "13": "Poc",
+    "14": "Por",
+    "15": "R",
+    "16": "RDC",
+    "17": "S",
+    "18": "SG",
+    "19": "Sarg",
+    "20": "Ser",
+    "21": "Slt",
+    "22": "Sp",
+    "23": "Turf"
+  },
+  "image_size": 224,
+  "initializer_range": 0.02,
+  "label2id": {
+    "ALGAE": 0,
+    "Acr": 1,
+    "Acr_Br": 2,
+    "Anem": 3,
+    "CCA": 4,
+    "Ech": 5,
+    "Fts": 6,
+    "Gal": 7,
+    "Gon": 8,
+    "H_oval": 9,
+    "H_uni": 10,
+    "Mtp": 11,
+    "P": 12,
+    "Poc": 13,
+    "Por": 14,
+    "R": 15,
+    "RDC": 16,
+    "S": 17,
+    "SG": 18,
+    "Sarg": 19,
+    "Ser": 20,
+    "Slt": 21,
+    "Sp": 22,
+    "Turf": 23
+  },
+  "layer_norm_eps": 1e-06,
+  "layerscale_value": 1.0,
+  "mlp_ratio": 4,
+  "model_type": "dinov2",
+  "num_attention_heads": 6,
+  "num_channels": 3,
+  "num_hidden_layers": 12,
+  "out_features": [
+    "stage12"
+  ],
+  "out_indices": [
+    12
+  ],
+  "patch_size": 14,
+  "problem_type": "single_label_classification",
+  "qkv_bias": true,
+  "reshape_hidden_states": true,
+  "stage_names": [
+    "stem",
+    "stage1",
+    "stage2",
+    "stage3",
+    "stage4",
+    "stage5",
+    "stage6",
+    "stage7",
+    "stage8",
+    "stage9",
+    "stage10",
+    "stage11",
+    "stage12"
+  ],
+  "torch_dtype": "float32",
+  "transformers_version": "4.48.0",
+  "use_swiglu_ffn": false
+}

logs/events.out.tfevents.1746512000.Amoros

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e8b507054c6579dbccc00af0add7bdee2f8754f063e228549c3e7faae9df8049
+size 33795

logs/events.out.tfevents.1746512148.Amoros

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65eb708f7bc5997bb66803860785fef8b9619b6afb9c17f4f23a72d1f8d222b9
+size 40

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a8b4876a6c262e5d4d0a68a9610f2f0e1c2e1c86f5e6a4f2831e4a46a054777c
+size 88047960

preprocessor_config.json

@@ -0,0 +1,28 @@
+{
+  "crop_size": {
+    "height": 224,
+    "width": 224
+  },
+  "do_center_crop": false,
+  "do_convert_rgb": true,
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.485,
+    0.456,
+    0.406
+  ],
+  "image_processor_type": "BitImageProcessor",
+  "image_std": [
+    0.229,
+    0.224,
+    0.225
+  ],
+  "resample": 3,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "height": 224,
+    "width": 224
+  }
+}

test_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 57.0,
+    "eval_accuracy": 0.4,
+    "eval_f1_macro": 0.23333333333333334,
+    "eval_f1_micro": 0.4,
+    "eval_loss": 2.8191895484924316,
+    "eval_runtime": 0.3171,
+    "eval_samples_per_second": 31.535,
+    "eval_steps_per_second": 3.153,
+    "learning_rate": 1e-05
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 57.0,
+    "learning_rate": 1e-05,
+    "total_flos": 1.132788066729984e+16,
+    "train_loss": 1.0105553342584979,
+    "train_runtime": 99.3744,
+    "train_samples_per_second": 15.094,
+    "train_steps_per_second": 1.509
+}

trainer_state.json

@@ -0,0 +1,736 @@
+{
+  "best_metric": 2.19789457321167,
+  "best_model_checkpoint": "models\\DinoAmoros-small-2025_05_06_36794-prova_bs16_freeze_monolabel\\checkpoint-47",
+  "epoch": 57.0,
+  "eval_steps": 500,
+  "global_step": 57,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.0,
+      "eval_f1_macro": 0.0,
+      "eval_f1_micro": 0.0,
+      "eval_loss": 3.198047161102295,
+      "eval_runtime": 0.2491,
+      "eval_samples_per_second": 40.14,
+      "eval_steps_per_second": 4.014,
+      "learning_rate": 0.001,
+      "step": 1
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.0,
+      "eval_f1_macro": 0.0,
+      "eval_f1_micro": 0.0,
+      "eval_loss": 3.131640911102295,
+      "eval_runtime": 0.1122,
+      "eval_samples_per_second": 89.154,
+      "eval_steps_per_second": 8.915,
+      "learning_rate": 0.001,
+      "step": 2
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.0,
+      "eval_f1_macro": 0.0,
+      "eval_f1_micro": 0.0,
+      "eval_loss": 3.125,
+      "eval_runtime": 0.1717,
+      "eval_samples_per_second": 58.232,
+      "eval_steps_per_second": 5.823,
+      "learning_rate": 0.001,
+      "step": 3
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.0,
+      "eval_f1_macro": 0.0,
+      "eval_f1_micro": 0.0,
+      "eval_loss": 3.1275391578674316,
+      "eval_runtime": 0.1416,
+      "eval_samples_per_second": 70.64,
+      "eval_steps_per_second": 7.064,
+      "learning_rate": 0.001,
+      "step": 4
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.1,
+      "eval_f1_macro": 0.05714285714285715,
+      "eval_f1_micro": 0.1,
+      "eval_loss": 3.067577838897705,
+      "eval_runtime": 0.1405,
+      "eval_samples_per_second": 71.164,
+      "eval_steps_per_second": 7.116,
+      "learning_rate": 0.001,
+      "step": 5
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.1,
+      "eval_f1_macro": 0.05714285714285715,
+      "eval_f1_micro": 0.1,
+      "eval_loss": 3.065624713897705,
+      "eval_runtime": 0.173,
+      "eval_samples_per_second": 57.811,
+      "eval_steps_per_second": 5.781,
+      "learning_rate": 0.001,
+      "step": 6
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.1,
+      "eval_f1_macro": 0.05,
+      "eval_f1_micro": 0.1,
+      "eval_loss": 3.0042967796325684,
+      "eval_runtime": 0.1325,
+      "eval_samples_per_second": 75.457,
+      "eval_steps_per_second": 7.546,
+      "learning_rate": 0.001,
+      "step": 7
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.3,
+      "eval_f1_macro": 0.15306122448979592,
+      "eval_f1_micro": 0.3,
+      "eval_loss": 2.9486327171325684,
+      "eval_runtime": 0.1302,
+      "eval_samples_per_second": 76.779,
+      "eval_steps_per_second": 7.678,
+      "learning_rate": 0.001,
+      "step": 8
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.4,
+      "eval_f1_macro": 0.2816326530612245,
+      "eval_f1_micro": 0.4,
+      "eval_loss": 2.8736329078674316,
+      "eval_runtime": 0.1384,
+      "eval_samples_per_second": 72.254,
+      "eval_steps_per_second": 7.225,
+      "learning_rate": 0.001,
+      "step": 9
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.4,
+      "eval_f1_macro": 0.2816326530612245,
+      "eval_f1_micro": 0.4,
+      "eval_loss": 2.8121094703674316,
+      "eval_runtime": 0.1367,
+      "eval_samples_per_second": 73.148,
+      "eval_steps_per_second": 7.315,
+      "learning_rate": 0.001,
+      "step": 10
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.6,
+      "eval_f1_macro": 0.40277777777777773,
+      "eval_f1_micro": 0.6,
+      "eval_loss": 2.754101276397705,
+      "eval_runtime": 0.164,
+      "eval_samples_per_second": 60.964,
+      "eval_steps_per_second": 6.096,
+      "learning_rate": 0.001,
+      "step": 11
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.6,
+      "eval_f1_macro": 0.40277777777777773,
+      "eval_f1_micro": 0.6,
+      "eval_loss": 2.6966795921325684,
+      "eval_runtime": 0.1396,
+      "eval_samples_per_second": 71.647,
+      "eval_steps_per_second": 7.165,
+      "learning_rate": 0.001,
+      "step": 12
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.6,
+      "eval_f1_macro": 0.484126984126984,
+      "eval_f1_micro": 0.6,
+      "eval_loss": 2.6595702171325684,
+      "eval_runtime": 0.1263,
+      "eval_samples_per_second": 79.178,
+      "eval_steps_per_second": 7.918,
+      "learning_rate": 0.001,
+      "step": 13
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.35714285714285715,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.6483397483825684,
+      "eval_runtime": 0.1255,
+      "eval_samples_per_second": 79.679,
+      "eval_steps_per_second": 7.968,
+      "learning_rate": 0.001,
+      "step": 14
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.35714285714285715,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.6143555641174316,
+      "eval_runtime": 0.1245,
+      "eval_samples_per_second": 80.34,
+      "eval_steps_per_second": 8.034,
+      "learning_rate": 0.001,
+      "step": 15
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.35714285714285715,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.5909180641174316,
+      "eval_runtime": 0.141,
+      "eval_samples_per_second": 70.936,
+      "eval_steps_per_second": 7.094,
+      "learning_rate": 0.001,
+      "step": 16
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.548144817352295,
+      "eval_runtime": 0.1409,
+      "eval_samples_per_second": 70.958,
+      "eval_steps_per_second": 7.096,
+      "learning_rate": 0.001,
+      "step": 17
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.3619047619047619,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.5125975608825684,
+      "eval_runtime": 0.1285,
+      "eval_samples_per_second": 77.801,
+      "eval_steps_per_second": 7.78,
+      "learning_rate": 0.001,
+      "step": 18
+    },
+    {
+      "epoch": 19.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.4791016578674316,
+      "eval_runtime": 0.1252,
+      "eval_samples_per_second": 79.893,
+      "eval_steps_per_second": 7.989,
+      "learning_rate": 0.001,
+      "step": 19
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.473827838897705,
+      "eval_runtime": 0.1649,
+      "eval_samples_per_second": 60.639,
+      "eval_steps_per_second": 6.064,
+      "learning_rate": 0.001,
+      "step": 20
+    },
+    {
+      "epoch": 21.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.4310059547424316,
+      "eval_runtime": 0.1318,
+      "eval_samples_per_second": 75.857,
+      "eval_steps_per_second": 7.586,
+      "learning_rate": 0.001,
+      "step": 21
+    },
+    {
+      "epoch": 22.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.403027057647705,
+      "eval_runtime": 0.1782,
+      "eval_samples_per_second": 56.119,
+      "eval_steps_per_second": 5.612,
+      "learning_rate": 0.001,
+      "step": 22
+    },
+    {
+      "epoch": 23.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.400097370147705,
+      "eval_runtime": 0.1295,
+      "eval_samples_per_second": 77.208,
+      "eval_steps_per_second": 7.721,
+      "learning_rate": 0.001,
+      "step": 23
+    },
+    {
+      "epoch": 24.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.3993163108825684,
+      "eval_runtime": 0.1217,
+      "eval_samples_per_second": 82.203,
+      "eval_steps_per_second": 8.22,
+      "learning_rate": 0.001,
+      "step": 24
+    },
+    {
+      "epoch": 25.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.392773151397705,
+      "eval_runtime": 0.1217,
+      "eval_samples_per_second": 82.192,
+      "eval_steps_per_second": 8.219,
+      "learning_rate": 0.001,
+      "step": 25
+    },
+    {
+      "epoch": 26.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.3896484375,
+      "eval_runtime": 0.1371,
+      "eval_samples_per_second": 72.942,
+      "eval_steps_per_second": 7.294,
+      "learning_rate": 0.001,
+      "step": 26
+    },
+    {
+      "epoch": 27.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.390869140625,
+      "eval_runtime": 0.1358,
+      "eval_samples_per_second": 73.642,
+      "eval_steps_per_second": 7.364,
+      "learning_rate": 0.001,
+      "step": 27
+    },
+    {
+      "epoch": 28.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.377246379852295,
+      "eval_runtime": 0.1254,
+      "eval_samples_per_second": 79.747,
+      "eval_steps_per_second": 7.975,
+      "learning_rate": 0.001,
+      "step": 28
+    },
+    {
+      "epoch": 29.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.3432374000549316,
+      "eval_runtime": 0.1266,
+      "eval_samples_per_second": 78.973,
+      "eval_steps_per_second": 7.897,
+      "learning_rate": 0.001,
+      "step": 29
+    },
+    {
+      "epoch": 30.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.3191895484924316,
+      "eval_runtime": 0.1195,
+      "eval_samples_per_second": 83.69,
+      "eval_steps_per_second": 8.369,
+      "learning_rate": 0.001,
+      "step": 30
+    },
+    {
+      "epoch": 31.0,
+      "eval_accuracy": 0.6,
+      "eval_f1_macro": 0.5476190476190477,
+      "eval_f1_micro": 0.6,
+      "eval_loss": 2.308764934539795,
+      "eval_runtime": 0.1395,
+      "eval_samples_per_second": 71.709,
+      "eval_steps_per_second": 7.171,
+      "learning_rate": 0.001,
+      "step": 31
+    },
+    {
+      "epoch": 32.0,
+      "eval_accuracy": 0.6,
+      "eval_f1_macro": 0.5476190476190477,
+      "eval_f1_micro": 0.6,
+      "eval_loss": 2.3003783226013184,
+      "eval_runtime": 0.1233,
+      "eval_samples_per_second": 81.112,
+      "eval_steps_per_second": 8.111,
+      "learning_rate": 0.001,
+      "step": 32
+    },
+    {
+      "epoch": 33.0,
+      "eval_accuracy": 0.6,
+      "eval_f1_macro": 0.5476190476190477,
+      "eval_f1_micro": 0.6,
+      "eval_loss": 2.304394245147705,
+      "eval_runtime": 0.1231,
+      "eval_samples_per_second": 81.231,
+      "eval_steps_per_second": 8.123,
+      "learning_rate": 0.001,
+      "step": 33
+    },
+    {
+      "epoch": 34.0,
+      "eval_accuracy": 0.6,
+      "eval_f1_macro": 0.5476190476190477,
+      "eval_f1_micro": 0.6,
+      "eval_loss": 2.297924518585205,
+      "eval_runtime": 0.136,
+      "eval_samples_per_second": 73.545,
+      "eval_steps_per_second": 7.355,
+      "learning_rate": 0.001,
+      "step": 34
+    },
+    {
+      "epoch": 35.0,
+      "eval_accuracy": 0.6,
+      "eval_f1_macro": 0.5476190476190477,
+      "eval_f1_micro": 0.6,
+      "eval_loss": 2.304821491241455,
+      "eval_runtime": 0.1332,
+      "eval_samples_per_second": 75.078,
+      "eval_steps_per_second": 7.508,
+      "learning_rate": 0.001,
+      "step": 35
+    },
+    {
+      "epoch": 36.0,
+      "eval_accuracy": 0.6,
+      "eval_f1_macro": 0.5476190476190477,
+      "eval_f1_micro": 0.6,
+      "eval_loss": 2.2987122535705566,
+      "eval_runtime": 0.1444,
+      "eval_samples_per_second": 69.246,
+      "eval_steps_per_second": 6.925,
+      "learning_rate": 0.001,
+      "step": 36
+    },
+    {
+      "epoch": 37.0,
+      "eval_accuracy": 0.6,
+      "eval_f1_macro": 0.5476190476190477,
+      "eval_f1_micro": 0.6,
+      "eval_loss": 2.299713134765625,
+      "eval_runtime": 0.175,
+      "eval_samples_per_second": 57.129,
+      "eval_steps_per_second": 5.713,
+      "learning_rate": 0.001,
+      "step": 37
+    },
+    {
+      "epoch": 38.0,
+      "eval_accuracy": 0.6,
+      "eval_f1_macro": 0.5476190476190477,
+      "eval_f1_micro": 0.6,
+      "eval_loss": 2.319528102874756,
+      "eval_runtime": 0.1336,
+      "eval_samples_per_second": 74.854,
+      "eval_steps_per_second": 7.485,
+      "learning_rate": 0.001,
+      "step": 38
+    },
+    {
+      "epoch": 39.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.3158326148986816,
+      "eval_runtime": 0.1144,
+      "eval_samples_per_second": 87.426,
+      "eval_steps_per_second": 8.743,
+      "learning_rate": 0.001,
+      "step": 39
+    },
+    {
+      "epoch": 40.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.3082613945007324,
+      "eval_runtime": 0.144,
+      "eval_samples_per_second": 69.427,
+      "eval_steps_per_second": 6.943,
+      "learning_rate": 0.001,
+      "step": 40
+    },
+    {
+      "epoch": 41.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.28303861618042,
+      "eval_runtime": 0.1397,
+      "eval_samples_per_second": 71.6,
+      "eval_steps_per_second": 7.16,
+      "learning_rate": 0.0001,
+      "step": 41
+    },
+    {
+      "epoch": 42.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.2719388008117676,
+      "eval_runtime": 0.1221,
+      "eval_samples_per_second": 81.871,
+      "eval_steps_per_second": 8.187,
+      "learning_rate": 0.0001,
+      "step": 42
+    },
+    {
+      "epoch": 43.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.240405559539795,
+      "eval_runtime": 0.1323,
+      "eval_samples_per_second": 75.566,
+      "eval_steps_per_second": 7.557,
+      "learning_rate": 0.0001,
+      "step": 43
+    },
+    {
+      "epoch": 44.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.2438597679138184,
+      "eval_runtime": 0.1759,
+      "eval_samples_per_second": 56.864,
+      "eval_steps_per_second": 5.686,
+      "learning_rate": 0.0001,
+      "step": 44
+    },
+    {
+      "epoch": 45.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.22489595413208,
+      "eval_runtime": 0.1278,
+      "eval_samples_per_second": 78.272,
+      "eval_steps_per_second": 7.827,
+      "learning_rate": 0.0001,
+      "step": 45
+    },
+    {
+      "epoch": 46.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.21156644821167,
+      "eval_runtime": 0.1529,
+      "eval_samples_per_second": 65.404,
+      "eval_steps_per_second": 6.54,
+      "learning_rate": 0.0001,
+      "step": 46
+    },
+    {
+      "epoch": 47.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.19789457321167,
+      "eval_runtime": 0.1469,
+      "eval_samples_per_second": 68.081,
+      "eval_steps_per_second": 6.808,
+      "learning_rate": 0.0001,
+      "step": 47
+    },
+    {
+      "epoch": 48.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.208752155303955,
+      "eval_runtime": 0.1318,
+      "eval_samples_per_second": 75.849,
+      "eval_steps_per_second": 7.585,
+      "learning_rate": 0.0001,
+      "step": 48
+    },
+    {
+      "epoch": 49.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.2074952125549316,
+      "eval_runtime": 0.1432,
+      "eval_samples_per_second": 69.811,
+      "eval_steps_per_second": 6.981,
+      "learning_rate": 0.0001,
+      "step": 49
+    },
+    {
+      "epoch": 50.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.2066588401794434,
+      "eval_runtime": 0.1397,
+      "eval_samples_per_second": 71.588,
+      "eval_steps_per_second": 7.159,
+      "learning_rate": 0.0001,
+      "step": 50
+    },
+    {
+      "epoch": 51.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.218231201171875,
+      "eval_runtime": 0.1212,
+      "eval_samples_per_second": 82.531,
+      "eval_steps_per_second": 8.253,
+      "learning_rate": 0.0001,
+      "step": 51
+    },
+    {
+      "epoch": 52.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.2242674827575684,
+      "eval_runtime": 0.17,
+      "eval_samples_per_second": 58.815,
+      "eval_steps_per_second": 5.881,
+      "learning_rate": 0.0001,
+      "step": 52
+    },
+    {
+      "epoch": 53.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.23439359664917,
+      "eval_runtime": 0.1367,
+      "eval_samples_per_second": 73.138,
+      "eval_steps_per_second": 7.314,
+      "learning_rate": 0.0001,
+      "step": 53
+    },
+    {
+      "epoch": 54.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.2221617698669434,
+      "eval_runtime": 0.1651,
+      "eval_samples_per_second": 60.581,
+      "eval_steps_per_second": 6.058,
+      "learning_rate": 1e-05,
+      "step": 54
+    },
+    {
+      "epoch": 55.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.2211365699768066,
+      "eval_runtime": 0.1366,
+      "eval_samples_per_second": 73.229,
+      "eval_steps_per_second": 7.323,
+      "learning_rate": 1e-05,
+      "step": 55
+    },
+    {
+      "epoch": 56.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.207244873046875,
+      "eval_runtime": 0.1238,
+      "eval_samples_per_second": 80.78,
+      "eval_steps_per_second": 8.078,
+      "learning_rate": 1e-05,
+      "step": 56
+    },
+    {
+      "epoch": 57.0,
+      "eval_accuracy": 0.5,
+      "eval_f1_macro": 0.42857142857142855,
+      "eval_f1_micro": 0.5,
+      "eval_loss": 2.2094483375549316,
+      "eval_runtime": 0.1349,
+      "eval_samples_per_second": 74.14,
+      "eval_steps_per_second": 7.414,
+      "learning_rate": 1e-05,
+      "step": 57
+    },
+    {
+      "epoch": 57.0,
+      "learning_rate": 1e-05,
+      "step": 57,
+      "total_flos": 1.132788066729984e+16,
+      "train_loss": 1.0105553342584979,
+      "train_runtime": 99.3744,
+      "train_samples_per_second": 15.094,
+      "train_steps_per_second": 1.509
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 150,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 150,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 10,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 10
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.132788066729984e+16,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:93911835f2fcc78a8008b53309a0eaf76340cffa8f4a9dfdff7297059bad1270
+size 5432

transforms.json

@@ -0,0 +1,44 @@
+{
+    "train_transforms": [
+        {
+            "operation": "PreProcess"
+        },
+        {
+            "operation": "Resize",
+            "probability": 1.0
+        },
+        {
+            "operation": "RandomHorizontalFlip",
+            "probability": 0.25
+        },
+        {
+            "operation": "RandomVerticalFlip",
+            "probability": 0.25
+        },
+        {
+            "operation": "ColorJiggle",
+            "probability": 0.25
+        },
+        {
+            "operation": "RandomPerspective",
+            "probability": 0.25
+        },
+        {
+            "operation": "Normalize",
+            "probability": 1.0
+        }
+    ],
+    "val_transforms": [
+        {
+            "operation": "PreProcess"
+        },
+        {
+            "operation": "Resize",
+            "probability": 1.0
+        },
+        {
+            "operation": "Normalize",
+            "probability": 1.0
+        }
+    ]
+}