End of training

README.md

@@ -0,0 +1,117 @@
+---
+library_name: transformers
+license: apache-2.0
+base_model: microsoft/swinv2-tiny-patch4-window8-256
+tags:
+- generated_from_trainer
+datasets:
+- imagefolder
+metrics:
+- accuracy
+model-index:
+- name: swinv2-tiny-patch4-window8-256-Ocular-Toxoplasmosis-DA
+ results:
+ - task:
+ name: Image Classification
+ type: image-classification
+ dataset:
+ name: imagefolder
+ type: imagefolder
+ config: default
+ split: validation
+ args: default
+ metrics:
+ - name: Accuracy
+ type: accuracy
+ value: 0.8548387096774194
+---
+
+<!-- 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. -->
+
+# swinv2-tiny-patch4-window8-256-Ocular-Toxoplasmosis-DA
+
+This model is a fine-tuned version of [microsoft/swinv2-tiny-patch4-window8-256](https://huggingface.co/microsoft/swinv2-tiny-patch4-window8-256) on the imagefolder dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.5075
+- Accuracy: 0.8548
+
+## 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: 5e-05
+- train_batch_size: 32
+- eval_batch_size: 32
+- seed: 42
+- gradient_accumulation_steps: 4
+- total_train_batch_size: 128
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 40
+
+### Training results
+
+| Training Loss | Epoch | Step | Validation Loss | Accuracy |
+|:-------------:|:-------:|:----:|:---------------:|:--------:|
+| 1.3402 | 0.9630 | 13 | 1.1682 | 0.5484 |
+| 1.1725 | 2.0 | 27 | 1.0025 | 0.6290 |
+| 0.8824 | 2.9630 | 40 | 0.7644 | 0.6613 |
+| 0.7342 | 4.0 | 54 | 0.5840 | 0.7258 |
+| 0.6734 | 4.9630 | 67 | 0.6754 | 0.6452 |
+| 0.5167 | 6.0 | 81 | 0.5904 | 0.6935 |
+| 0.5009 | 6.9630 | 94 | 0.5549 | 0.6935 |
+| 0.4988 | 8.0 | 108 | 0.6204 | 0.6774 |
+| 0.3856 | 8.9630 | 121 | 0.4463 | 0.8226 |
+| 0.4057 | 10.0 | 135 | 0.5232 | 0.7903 |
+| 0.3929 | 10.9630 | 148 | 0.4580 | 0.8387 |
+| 0.3638 | 12.0 | 162 | 0.5115 | 0.7742 |
+| 0.3248 | 12.9630 | 175 | 0.5313 | 0.7742 |
+| 0.2673 | 14.0 | 189 | 0.5203 | 0.7903 |
+| 0.2922 | 14.9630 | 202 | 0.4315 | 0.8387 |
+| 0.2803 | 16.0 | 216 | 0.4577 | 0.8387 |
+| 0.2735 | 16.9630 | 229 | 0.5467 | 0.8065 |
+| 0.2586 | 18.0 | 243 | 0.5236 | 0.8387 |
+| 0.2366 | 18.9630 | 256 | 0.5075 | 0.8548 |
+| 0.2347 | 20.0 | 270 | 0.5179 | 0.8387 |
+| 0.2046 | 20.9630 | 283 | 0.5428 | 0.8387 |
+| 0.2289 | 22.0 | 297 | 0.5748 | 0.8387 |
+| 0.2195 | 22.9630 | 310 | 0.5969 | 0.8226 |
+| 0.2224 | 24.0 | 324 | 0.6092 | 0.8226 |
+| 0.2167 | 24.9630 | 337 | 0.6333 | 0.8226 |
+| 0.1956 | 26.0 | 351 | 0.5993 | 0.8226 |
+| 0.2174 | 26.9630 | 364 | 0.6063 | 0.8548 |
+| 0.1999 | 28.0 | 378 | 0.6414 | 0.8387 |
+| 0.1667 | 28.9630 | 391 | 0.6297 | 0.8387 |
+| 0.1835 | 30.0 | 405 | 0.6149 | 0.8226 |
+| 0.186 | 30.9630 | 418 | 0.6430 | 0.8387 |
+| 0.1749 | 32.0 | 432 | 0.6678 | 0.8387 |
+| 0.1663 | 32.9630 | 445 | 0.6829 | 0.8387 |
+| 0.1557 | 34.0 | 459 | 0.6557 | 0.8387 |
+| 0.1913 | 34.9630 | 472 | 0.6275 | 0.8387 |
+| 0.1775 | 36.0 | 486 | 0.6555 | 0.8548 |
+| 0.152 | 36.9630 | 499 | 0.6653 | 0.8548 |
+| 0.1897 | 38.0 | 513 | 0.6682 | 0.8548 |
+| 0.1589 | 38.5185 | 520 | 0.6679 | 0.8548 |
+
+
+### Framework versions
+
+- Transformers 4.44.2
+- Pytorch 2.4.1+cu121
+- Datasets 3.0.1
+- Tokenizers 0.19.1

all_results.json

@@ -0,0 +1,13 @@
+{
+ "epoch": 38.51851851851852,
+ "eval_accuracy": 0.8548387096774194,
+ "eval_loss": 0.5075119137763977,
+ "eval_runtime": 2.3581,
+ "eval_samples_per_second": 26.293,
+ "eval_steps_per_second": 0.848,
+ "total_flos": 2.140878196703232e+18,
+ "train_loss": 0.35049390150950505,
+ "train_runtime": 3356.7171,
+ "train_samples_per_second": 20.353,
+ "train_steps_per_second": 0.155
+}

config.json

@@ -0,0 +1,71 @@
+{
+ "_name_or_path": "microsoft/swinv2-tiny-patch4-window8-256",
+ "architectures": [
+ "Swinv2ForImageClassification"
+ ],
+ "attention_probs_dropout_prob": 0.0,
+ "depths": [
+ 2,
+ 2,
+ 6,
+ 2
+ ],
+ "drop_path_rate": 0.1,
+ "embed_dim": 96,
+ "encoder_stride": 32,
+ "hidden_act": "gelu",
+ "hidden_dropout_prob": 0.0,
+ "hidden_size": 768,
+ "id2label": {
+ "0": "active",
+ "1": "active-inactive",
+ "2": "healthy",
+ "3": "inactive"
+ },
+ "image_size": 256,
+ "initializer_range": 0.02,
+ "label2id": {
+ "active": 0,
+ "active-inactive": 1,
+ "healthy": 2,
+ "inactive": 3
+ },
+ "layer_norm_eps": 1e-05,
+ "mlp_ratio": 4.0,
+ "model_type": "swinv2",
+ "num_channels": 3,
+ "num_heads": [
+ 3,
+ 6,
+ 12,
+ 24
+ ],
+ "num_layers": 4,
+ "out_features": [
+ "stage4"
+ ],
+ "out_indices": [
+ 4
+ ],
+ "patch_size": 4,
+ "path_norm": true,
+ "pretrained_window_sizes": [
+ 0,
+ 0,
+ 0,
+ 0
+ ],
+ "problem_type": "single_label_classification",
+ "qkv_bias": true,
+ "stage_names": [
+ "stem",
+ "stage1",
+ "stage2",
+ "stage3",
+ "stage4"
+ ],
+ "torch_dtype": "float32",
+ "transformers_version": "4.44.2",
+ "use_absolute_embeddings": false,
+ "window_size": 8
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+ "epoch": 38.51851851851852,
+ "eval_accuracy": 0.8548387096774194,
+ "eval_loss": 0.5075119137763977,
+ "eval_runtime": 2.3581,
+ "eval_samples_per_second": 26.293,
+ "eval_steps_per_second": 0.848
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:00b20367a51932debc487a0f982258dfd88b15d58596b4f5de73fdd897b2ee8e
+size 110356296

preprocessor_config.json

@@ -0,0 +1,22 @@
+{
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "image_mean": [
+ 0.485,
+ 0.456,
+ 0.406
+ ],
+ "image_processor_type": "ViTImageProcessor",
+ "image_std": [
+ 0.229,
+ 0.224,
+ 0.225
+ ],
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "height": 256,
+ "width": 256
+ }
+}

runs/Oct13_15-37-14_496ec48e7779/events.out.tfevents.1728833842.496ec48e7779.2126.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:930cd8a74a0a067ede15c21a4388bd1488e36af3369c4d196c3901505231464c
+size 29395

runs/Oct13_15-37-14_496ec48e7779/events.out.tfevents.1728837202.496ec48e7779.2126.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b91f5baca1ecf4952c704c3b9e1b037a88008e9411beac13764f5df0ec37d18f
+size 411

train_results.json

@@ -0,0 +1,8 @@
+{
+ "epoch": 38.51851851851852,
+ "total_flos": 2.140878196703232e+18,
+ "train_loss": 0.35049390150950505,
+ "train_runtime": 3356.7171,
+ "train_samples_per_second": 20.353,
+ "train_steps_per_second": 0.155
+}

trainer_state.json

@@ -0,0 +1,757 @@
+{
+ "best_metric": 0.8548387096774194,
+ "best_model_checkpoint": "swinv2-tiny-patch4-window8-256-Ocular-Toxoplasmosis-DA/checkpoint-256",
+ "epoch": 38.51851851851852,
+ "eval_steps": 500,
+ "global_step": 520,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.7407407407407407,
+ "grad_norm": 4.370074272155762,
+ "learning_rate": 9.615384615384616e-06,
+ "loss": 1.3402,
+ "step": 10
+ },
+ {
+ "epoch": 0.9629629629629629,
+ "eval_accuracy": 0.5483870967741935,
+ "eval_loss": 1.1682088375091553,
+ "eval_runtime": 2.3908,
+ "eval_samples_per_second": 25.932,
+ "eval_steps_per_second": 0.837,
+ "step": 13
+ },
+ {
+ "epoch": 1.4814814814814814,
+ "grad_norm": 7.86944580078125,
+ "learning_rate": 1.923076923076923e-05,
+ "loss": 1.1725,
+ "step": 20
+ },
+ {
+ "epoch": 2.0,
+ "eval_accuracy": 0.6290322580645161,
+ "eval_loss": 1.0024793148040771,
+ "eval_runtime": 3.324,
+ "eval_samples_per_second": 18.652,
+ "eval_steps_per_second": 0.602,
+ "step": 27
+ },
+ {
+ "epoch": 2.2222222222222223,
+ "grad_norm": 4.94896936416626,
+ "learning_rate": 2.8846153846153845e-05,
+ "loss": 1.0671,
+ "step": 30
+ },
+ {
+ "epoch": 2.962962962962963,
+ "grad_norm": 4.811951160430908,
+ "learning_rate": 3.846153846153846e-05,
+ "loss": 0.8824,
+ "step": 40
+ },
+ {
+ "epoch": 2.962962962962963,
+ "eval_accuracy": 0.6612903225806451,
+ "eval_loss": 0.7644360065460205,
+ "eval_runtime": 3.388,
+ "eval_samples_per_second": 18.3,
+ "eval_steps_per_second": 0.59,
+ "step": 40
+ },
+ {
+ "epoch": 3.7037037037037037,
+ "grad_norm": 7.974093914031982,
+ "learning_rate": 4.8076923076923084e-05,
+ "loss": 0.7342,
+ "step": 50
+ },
+ {
+ "epoch": 4.0,
+ "eval_accuracy": 0.7258064516129032,
+ "eval_loss": 0.5839676260948181,
+ "eval_runtime": 3.0543,
+ "eval_samples_per_second": 20.299,
+ "eval_steps_per_second": 0.655,
+ "step": 54
+ },
+ {
+ "epoch": 4.444444444444445,
+ "grad_norm": 8.472794532775879,
+ "learning_rate": 4.9145299145299147e-05,
+ "loss": 0.6734,
+ "step": 60
+ },
+ {
+ "epoch": 4.962962962962963,
+ "eval_accuracy": 0.6451612903225806,
+ "eval_loss": 0.6753666400909424,
+ "eval_runtime": 2.3642,
+ "eval_samples_per_second": 26.225,
+ "eval_steps_per_second": 0.846,
+ "step": 67
+ },
+ {
+ "epoch": 5.185185185185185,
+ "grad_norm": 9.15774917602539,
+ "learning_rate": 4.8076923076923084e-05,
+ "loss": 0.6373,
+ "step": 70
+ },
+ {
+ "epoch": 5.925925925925926,
+ "grad_norm": 12.02450942993164,
+ "learning_rate": 4.700854700854701e-05,
+ "loss": 0.5167,
+ "step": 80
+ },
+ {
+ "epoch": 6.0,
+ "eval_accuracy": 0.6935483870967742,
+ "eval_loss": 0.5904402136802673,
+ "eval_runtime": 2.3866,
+ "eval_samples_per_second": 25.979,
+ "eval_steps_per_second": 0.838,
+ "step": 81
+ },
+ {
+ "epoch": 6.666666666666667,
+ "grad_norm": 7.348090648651123,
+ "learning_rate": 4.594017094017094e-05,
+ "loss": 0.5009,
+ "step": 90
+ },
+ {
+ "epoch": 6.962962962962963,
+ "eval_accuracy": 0.6935483870967742,
+ "eval_loss": 0.5549384355545044,
+ "eval_runtime": 2.9982,
+ "eval_samples_per_second": 20.679,
+ "eval_steps_per_second": 0.667,
+ "step": 94
+ },
+ {
+ "epoch": 7.407407407407407,
+ "grad_norm": 5.642479419708252,
+ "learning_rate": 4.4871794871794874e-05,
+ "loss": 0.4988,
+ "step": 100
+ },
+ {
+ "epoch": 8.0,
+ "eval_accuracy": 0.6774193548387096,
+ "eval_loss": 0.620449423789978,
+ "eval_runtime": 2.4283,
+ "eval_samples_per_second": 25.532,
+ "eval_steps_per_second": 0.824,
+ "step": 108
+ },
+ {
+ "epoch": 8.148148148148149,
+ "grad_norm": 6.128896713256836,
+ "learning_rate": 4.3803418803418805e-05,
+ "loss": 0.4619,
+ "step": 110
+ },
+ {
+ "epoch": 8.88888888888889,
+ "grad_norm": 7.555347919464111,
+ "learning_rate": 4.2735042735042735e-05,
+ "loss": 0.3856,
+ "step": 120
+ },
+ {
+ "epoch": 8.962962962962964,
+ "eval_accuracy": 0.8225806451612904,
+ "eval_loss": 0.44631102681159973,
+ "eval_runtime": 2.3506,
+ "eval_samples_per_second": 26.376,
+ "eval_steps_per_second": 0.851,
+ "step": 121
+ },
+ {
+ "epoch": 9.62962962962963,
+ "grad_norm": 9.627432823181152,
+ "learning_rate": 4.166666666666667e-05,
+ "loss": 0.4057,
+ "step": 130
+ },
+ {
+ "epoch": 10.0,
+ "eval_accuracy": 0.7903225806451613,
+ "eval_loss": 0.5231879353523254,
+ "eval_runtime": 3.1544,
+ "eval_samples_per_second": 19.655,
+ "eval_steps_per_second": 0.634,
+ "step": 135
+ },
+ {
+ "epoch": 10.37037037037037,
+ "grad_norm": 8.669109344482422,
+ "learning_rate": 4.05982905982906e-05,
+ "loss": 0.3929,
+ "step": 140
+ },
+ {
+ "epoch": 10.962962962962964,
+ "eval_accuracy": 0.8387096774193549,
+ "eval_loss": 0.45801177620887756,
+ "eval_runtime": 2.3878,
+ "eval_samples_per_second": 25.965,
+ "eval_steps_per_second": 0.838,
+ "step": 148
+ },
+ {
+ "epoch": 11.11111111111111,
+ "grad_norm": 6.289756774902344,
+ "learning_rate": 3.952991452991453e-05,
+ "loss": 0.3673,
+ "step": 150
+ },
+ {
+ "epoch": 11.851851851851851,
+ "grad_norm": 12.90579605102539,
+ "learning_rate": 3.846153846153846e-05,
+ "loss": 0.3638,
+ "step": 160
+ },
+ {
+ "epoch": 12.0,
+ "eval_accuracy": 0.7741935483870968,
+ "eval_loss": 0.5114619135856628,
+ "eval_runtime": 3.3569,
+ "eval_samples_per_second": 18.47,
+ "eval_steps_per_second": 0.596,
+ "step": 162
+ },
+ {
+ "epoch": 12.592592592592592,
+ "grad_norm": 10.698553085327148,
+ "learning_rate": 3.739316239316239e-05,
+ "loss": 0.3248,
+ "step": 170
+ },
+ {
+ "epoch": 12.962962962962964,
+ "eval_accuracy": 0.7741935483870968,
+ "eval_loss": 0.5312773585319519,
+ "eval_runtime": 2.4335,
+ "eval_samples_per_second": 25.478,
+ "eval_steps_per_second": 0.822,
+ "step": 175
+ },
+ {
+ "epoch": 13.333333333333334,
+ "grad_norm": 6.529489994049072,
+ "learning_rate": 3.6324786324786323e-05,
+ "loss": 0.2673,
+ "step": 180
+ },
+ {
+ "epoch": 14.0,
+ "eval_accuracy": 0.7903225806451613,
+ "eval_loss": 0.5203306674957275,
+ "eval_runtime": 3.4828,
+ "eval_samples_per_second": 17.802,
+ "eval_steps_per_second": 0.574,
+ "step": 189
+ },
+ {
+ "epoch": 14.074074074074074,
+ "grad_norm": 6.994911193847656,
+ "learning_rate": 3.525641025641026e-05,
+ "loss": 0.3216,
+ "step": 190
+ },
+ {
+ "epoch": 14.814814814814815,
+ "grad_norm": 9.194233894348145,
+ "learning_rate": 3.418803418803419e-05,
+ "loss": 0.2922,
+ "step": 200
+ },
+ {
+ "epoch": 14.962962962962964,
+ "eval_accuracy": 0.8387096774193549,
+ "eval_loss": 0.4315454959869385,
+ "eval_runtime": 2.3822,
+ "eval_samples_per_second": 26.026,
+ "eval_steps_per_second": 0.84,
+ "step": 202
+ },
+ {
+ "epoch": 15.555555555555555,
+ "grad_norm": 6.076256275177002,
+ "learning_rate": 3.311965811965812e-05,
+ "loss": 0.2803,
+ "step": 210
+ },
+ {
+ "epoch": 16.0,
+ "eval_accuracy": 0.8387096774193549,
+ "eval_loss": 0.4577220380306244,
+ "eval_runtime": 2.9439,
+ "eval_samples_per_second": 21.06,
+ "eval_steps_per_second": 0.679,
+ "step": 216
+ },
+ {
+ "epoch": 16.296296296296298,
+ "grad_norm": 12.038761138916016,
+ "learning_rate": 3.205128205128206e-05,
+ "loss": 0.2735,
+ "step": 220
+ },
+ {
+ "epoch": 16.962962962962962,
+ "eval_accuracy": 0.8064516129032258,
+ "eval_loss": 0.5466907024383545,
+ "eval_runtime": 2.3229,
+ "eval_samples_per_second": 26.691,
+ "eval_steps_per_second": 0.861,
+ "step": 229
+ },
+ {
+ "epoch": 17.037037037037038,
+ "grad_norm": 8.897506713867188,
+ "learning_rate": 3.098290598290599e-05,
+ "loss": 0.2776,
+ "step": 230
+ },
+ {
+ "epoch": 17.77777777777778,
+ "grad_norm": 9.66178035736084,
+ "learning_rate": 2.9914529914529915e-05,
+ "loss": 0.2586,
+ "step": 240
+ },
+ {
+ "epoch": 18.0,
+ "eval_accuracy": 0.8387096774193549,
+ "eval_loss": 0.5236416459083557,
+ "eval_runtime": 3.4253,
+ "eval_samples_per_second": 18.101,
+ "eval_steps_per_second": 0.584,
+ "step": 243
+ },
+ {
+ "epoch": 18.51851851851852,
+ "grad_norm": 7.729655742645264,
+ "learning_rate": 2.8846153846153845e-05,
+ "loss": 0.2366,
+ "step": 250
+ },
+ {
+ "epoch": 18.962962962962962,
+ "eval_accuracy": 0.8548387096774194,
+ "eval_loss": 0.5075119137763977,
+ "eval_runtime": 2.3943,
+ "eval_samples_per_second": 25.895,
+ "eval_steps_per_second": 0.835,
+ "step": 256
+ },
+ {
+ "epoch": 19.25925925925926,
+ "grad_norm": 11.543585777282715,
+ "learning_rate": 2.777777777777778e-05,
+ "loss": 0.252,
+ "step": 260
+ },
+ {
+ "epoch": 20.0,
+ "grad_norm": 7.877120494842529,
+ "learning_rate": 2.670940170940171e-05,
+ "loss": 0.2347,
+ "step": 270
+ },
+ {
+ "epoch": 20.0,
+ "eval_accuracy": 0.8387096774193549,
+ "eval_loss": 0.5178562998771667,
+ "eval_runtime": 2.4124,
+ "eval_samples_per_second": 25.701,
+ "eval_steps_per_second": 0.829,
+ "step": 270
+ },
+ {
+ "epoch": 20.74074074074074,
+ "grad_norm": 7.83768892288208,
+ "learning_rate": 2.564102564102564e-05,
+ "loss": 0.2046,
+ "step": 280
+ },
+ {
+ "epoch": 20.962962962962962,
+ "eval_accuracy": 0.8387096774193549,
+ "eval_loss": 0.5427502393722534,
+ "eval_runtime": 3.4728,
+ "eval_samples_per_second": 17.853,
+ "eval_steps_per_second": 0.576,
+ "step": 283
+ },
+ {
+ "epoch": 21.48148148148148,
+ "grad_norm": 7.919957637786865,
+ "learning_rate": 2.4572649572649573e-05,
+ "loss": 0.2289,
+ "step": 290
+ },
+ {
+ "epoch": 22.0,
+ "eval_accuracy": 0.8387096774193549,
+ "eval_loss": 0.57480788230896,
+ "eval_runtime": 2.4021,
+ "eval_samples_per_second": 25.811,
+ "eval_steps_per_second": 0.833,
+ "step": 297
+ },
+ {
+ "epoch": 22.22222222222222,
+ "grad_norm": 8.665252685546875,
+ "learning_rate": 2.3504273504273504e-05,
+ "loss": 0.2394,
+ "step": 300
+ },
+ {
+ "epoch": 22.962962962962962,
+ "grad_norm": 7.902819633483887,
+ "learning_rate": 2.2435897435897437e-05,
+ "loss": 0.2195,
+ "step": 310
+ },
+ {
+ "epoch": 22.962962962962962,
+ "eval_accuracy": 0.8225806451612904,
+ "eval_loss": 0.5968937277793884,
+ "eval_runtime": 3.4133,
+ "eval_samples_per_second": 18.164,
+ "eval_steps_per_second": 0.586,
+ "step": 310
+ },
+ {
+ "epoch": 23.703703703703702,
+ "grad_norm": 9.844597816467285,
+ "learning_rate": 2.1367521367521368e-05,
+ "loss": 0.2224,
+ "step": 320
+ },
+ {
+ "epoch": 24.0,
+ "eval_accuracy": 0.8225806451612904,
+ "eval_loss": 0.6092303991317749,
+ "eval_runtime": 2.3949,
+ "eval_samples_per_second": 25.888,
+ "eval_steps_per_second": 0.835,
+ "step": 324
+ },
+ {
+ "epoch": 24.444444444444443,
+ "grad_norm": 6.439063549041748,
+ "learning_rate": 2.02991452991453e-05,
+ "loss": 0.2167,
+ "step": 330
+ },
+ {
+ "epoch": 24.962962962962962,
+ "eval_accuracy": 0.8225806451612904,
+ "eval_loss": 0.6333113312721252,
+ "eval_runtime": 2.4482,
+ "eval_samples_per_second": 25.325,
+ "eval_steps_per_second": 0.817,
+ "step": 337
+ },
+ {
+ "epoch": 25.185185185185187,
+ "grad_norm": 8.865224838256836,
+ "learning_rate": 1.923076923076923e-05,
+ "loss": 0.2323,
+ "step": 340
+ },
+ {
+ "epoch": 25.925925925925927,
+ "grad_norm": 6.462991237640381,
+ "learning_rate": 1.8162393162393162e-05,
+ "loss": 0.1956,
+ "step": 350
+ },
+ {
+ "epoch": 26.0,
+ "eval_accuracy": 0.8225806451612904,
+ "eval_loss": 0.5993022322654724,
+ "eval_runtime": 2.3358,
+ "eval_samples_per_second": 26.543,
+ "eval_steps_per_second": 0.856,
+ "step": 351
+ },
+ {
+ "epoch": 26.666666666666668,
+ "grad_norm": 6.978143692016602,
+ "learning_rate": 1.7094017094017095e-05,
+ "loss": 0.2174,
+ "step": 360
+ },
+ {
+ "epoch": 26.962962962962962,
+ "eval_accuracy": 0.8548387096774194,
+ "eval_loss": 0.6063364744186401,
+ "eval_runtime": 2.3579,
+ "eval_samples_per_second": 26.295,
+ "eval_steps_per_second": 0.848,
+ "step": 364
+ },
+ {
+ "epoch": 27.40740740740741,
+ "grad_norm": 8.283989906311035,
+ "learning_rate": 1.602564102564103e-05,
+ "loss": 0.1999,
+ "step": 370
+ },
+ {
+ "epoch": 28.0,
+ "eval_accuracy": 0.8387096774193549,
+ "eval_loss": 0.6413679718971252,
+ "eval_runtime": 3.4435,
+ "eval_samples_per_second": 18.005,
+ "eval_steps_per_second": 0.581,
+ "step": 378
+ },
+ {
+ "epoch": 28.14814814814815,
+ "grad_norm": 5.77383279800415,
+ "learning_rate": 1.4957264957264958e-05,
+ "loss": 0.1783,
+ "step": 380
+ },
+ {
+ "epoch": 28.88888888888889,
+ "grad_norm": 7.4615654945373535,
+ "learning_rate": 1.388888888888889e-05,
+ "loss": 0.1667,
+ "step": 390
+ },
+ {
+ "epoch": 28.962962962962962,
+ "eval_accuracy": 0.8387096774193549,
+ "eval_loss": 0.6296666860580444,
+ "eval_runtime": 2.3485,
+ "eval_samples_per_second": 26.4,
+ "eval_steps_per_second": 0.852,
+ "step": 391
+ },
+ {
+ "epoch": 29.62962962962963,
+ "grad_norm": 9.373270034790039,
+ "learning_rate": 1.282051282051282e-05,
+ "loss": 0.1835,
+ "step": 400
+ },
+ {
+ "epoch": 30.0,
+ "eval_accuracy": 0.8225806451612904,
+ "eval_loss": 0.6148854494094849,
+ "eval_runtime": 3.1829,
+ "eval_samples_per_second": 19.479,
+ "eval_steps_per_second": 0.628,
+ "step": 405
+ },
+ {
+ "epoch": 30.37037037037037,
+ "grad_norm": 8.87562370300293,
+ "learning_rate": 1.1752136752136752e-05,
+ "loss": 0.186,
+ "step": 410
+ },
+ {
+ "epoch": 30.962962962962962,
+ "eval_accuracy": 0.8387096774193549,
+ "eval_loss": 0.6429581642150879,
+ "eval_runtime": 2.4503,
+ "eval_samples_per_second": 25.303,
+ "eval_steps_per_second": 0.816,
+ "step": 418
+ },
+ {
+ "epoch": 31.11111111111111,
+ "grad_norm": 5.281705856323242,
+ "learning_rate": 1.0683760683760684e-05,
+ "loss": 0.1706,
+ "step": 420
+ },
+ {
+ "epoch": 31.85185185185185,
+ "grad_norm": 4.753020286560059,
+ "learning_rate": 9.615384615384616e-06,
+ "loss": 0.1749,
+ "step": 430
+ },
+ {
+ "epoch": 32.0,
+ "eval_accuracy": 0.8387096774193549,
+ "eval_loss": 0.6677759885787964,
+ "eval_runtime": 2.3885,
+ "eval_samples_per_second": 25.957,
+ "eval_steps_per_second": 0.837,
+ "step": 432
+ },
+ {
+ "epoch": 32.592592592592595,
+ "grad_norm": 7.2512526512146,
+ "learning_rate": 8.547008547008548e-06,
+ "loss": 0.1663,
+ "step": 440
+ },
+ {
+ "epoch": 32.96296296296296,
+ "eval_accuracy": 0.8387096774193549,
+ "eval_loss": 0.6828835010528564,
+ "eval_runtime": 2.3483,
+ "eval_samples_per_second": 26.402,
+ "eval_steps_per_second": 0.852,
+ "step": 445
+ },
+ {
+ "epoch": 33.333333333333336,
+ "grad_norm": 9.678658485412598,
+ "learning_rate": 7.478632478632479e-06,
+ "loss": 0.1557,
+ "step": 450
+ },
+ {
+ "epoch": 34.0,
+ "eval_accuracy": 0.8387096774193549,
+ "eval_loss": 0.655702531337738,
+ "eval_runtime": 3.1084,
+ "eval_samples_per_second": 19.946,
+ "eval_steps_per_second": 0.643,
+ "step": 459
+ },
+ {
+ "epoch": 34.074074074074076,
+ "grad_norm": 5.886323928833008,
+ "learning_rate": 6.41025641025641e-06,
+ "loss": 0.2095,
+ "step": 460
+ },
+ {
+ "epoch": 34.81481481481482,
+ "grad_norm": 5.312963485717773,
+ "learning_rate": 5.341880341880342e-06,
+ "loss": 0.1913,
+ "step": 470
+ },
+ {
+ "epoch": 34.96296296296296,
+ "eval_accuracy": 0.8387096774193549,
+ "eval_loss": 0.6274862885475159,
+ "eval_runtime": 3.2878,
+ "eval_samples_per_second": 18.858,
+ "eval_steps_per_second": 0.608,
+ "step": 472
+ },
+ {
+ "epoch": 35.55555555555556,
+ "grad_norm": 7.064798355102539,
+ "learning_rate": 4.273504273504274e-06,
+ "loss": 0.1775,
+ "step": 480
+ },
+ {
+ "epoch": 36.0,
+ "eval_accuracy": 0.8548387096774194,
+ "eval_loss": 0.6554756760597229,
+ "eval_runtime": 2.4759,
+ "eval_samples_per_second": 25.041,
+ "eval_steps_per_second": 0.808,
+ "step": 486
+ },
+ {
+ "epoch": 36.2962962962963,
+ "grad_norm": 5.463845729827881,
+ "learning_rate": 3.205128205128205e-06,
+ "loss": 0.152,
+ "step": 490
+ },
+ {
+ "epoch": 36.96296296296296,
+ "eval_accuracy": 0.8548387096774194,
+ "eval_loss": 0.6653042435646057,
+ "eval_runtime": 3.3751,
+ "eval_samples_per_second": 18.37,
+ "eval_steps_per_second": 0.593,
+ "step": 499
+ },
+ {
+ "epoch": 37.03703703703704,
+ "grad_norm": 5.512512683868408,
+ "learning_rate": 2.136752136752137e-06,
+ "loss": 0.1681,
+ "step": 500
+ },
+ {
+ "epoch": 37.77777777777778,
+ "grad_norm": 6.535687446594238,
+ "learning_rate": 1.0683760683760685e-06,
+ "loss": 0.1897,
+ "step": 510
+ },
+ {
+ "epoch": 38.0,
+ "eval_accuracy": 0.8548387096774194,
+ "eval_loss": 0.6681959629058838,
+ "eval_runtime": 2.3421,
+ "eval_samples_per_second": 26.472,
+ "eval_steps_per_second": 0.854,
+ "step": 513
+ },
+ {
+ "epoch": 38.51851851851852,
+ "grad_norm": 8.290581703186035,
+ "learning_rate": 0.0,
+ "loss": 0.1589,
+ "step": 520
+ },
+ {
+ "epoch": 38.51851851851852,
+ "eval_accuracy": 0.8548387096774194,
+ "eval_loss": 0.6678970456123352,
+ "eval_runtime": 2.3455,
+ "eval_samples_per_second": 26.434,
+ "eval_steps_per_second": 0.853,
+ "step": 520
+ },
+ {
+ "epoch": 38.51851851851852,
+ "step": 520,
+ "total_flos": 2.140878196703232e+18,
+ "train_loss": 0.35049390150950505,
+ "train_runtime": 3356.7171,
+ "train_samples_per_second": 20.353,
+ "train_steps_per_second": 0.155
+ }
+ ],
+ "logging_steps": 10,
+ "max_steps": 520,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 40,
+ "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": 2.140878196703232e+18,
+ "train_batch_size": 32,
+ "trial_name": null,
+ "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9ed5fae65789d986275c6f5b5304c2bbfd8e92ec07663c5d5e7b7917f8ab2ef
+size 5304