Commit
•
a153bdf
1
Parent(s):
7172d85
🍻 cheers
Browse files- README.md +4 -2
- all_results.json +13 -0
- runs/Mar18_12-43-54_user-desktop/events.out.tfevents.1710780657.user-desktop.44603.1 +3 -0
- test_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +487 -0
README.md
CHANGED
|
@@ -2,6 +2,8 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: google/vit-base-patch16-224-in21k
|
| 4 |
tags:
|
|
|
|
|
|
|
| 5 |
- generated_from_trainer
|
| 6 |
datasets:
|
| 7 |
- imagefolder
|
|
@@ -14,7 +16,7 @@ model-index:
|
|
| 14 |
name: Image Classification
|
| 15 |
type: image-classification
|
| 16 |
dataset:
|
| 17 |
-
name:
|
| 18 |
type: imagefolder
|
| 19 |
config: default
|
| 20 |
split: train
|
|
@@ -30,7 +32,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 30 |
|
| 31 |
# image_classification_obipix_birdID
|
| 32 |
|
| 33 |
-
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the
|
| 34 |
It achieves the following results on the evaluation set:
|
| 35 |
- Loss: 0.1150
|
| 36 |
- Accuracy: 0.9720
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: google/vit-base-patch16-224-in21k
|
| 4 |
tags:
|
| 5 |
+
- image-classification
|
| 6 |
+
- bird species identification
|
| 7 |
- generated_from_trainer
|
| 8 |
datasets:
|
| 9 |
- imagefolder
|
|
|
|
| 16 |
name: Image Classification
|
| 17 |
type: image-classification
|
| 18 |
dataset:
|
| 19 |
+
name: private crawled images
|
| 20 |
type: imagefolder
|
| 21 |
config: default
|
| 22 |
split: train
|
|
|
|
| 32 |
|
| 33 |
# image_classification_obipix_birdID
|
| 34 |
|
| 35 |
+
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the private crawled images dataset.
|
| 36 |
It achieves the following results on the evaluation set:
|
| 37 |
- Loss: 0.1150
|
| 38 |
- Accuracy: 0.9720
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.9719696025912545,
|
| 4 |
+
"eval_loss": 0.11504850536584854,
|
| 5 |
+
"eval_runtime": 489.2058,
|
| 6 |
+
"eval_samples_per_second": 131.266,
|
| 7 |
+
"eval_steps_per_second": 8.205,
|
| 8 |
+
"total_flos": 1.4469915597827883e+20,
|
| 9 |
+
"train_loss": 0.618133510268158,
|
| 10 |
+
"train_runtime": 34089.5693,
|
| 11 |
+
"train_samples_per_second": 53.373,
|
| 12 |
+
"train_steps_per_second": 0.834
|
| 13 |
+
}
|
runs/Mar18_12-43-54_user-desktop/events.out.tfevents.1710780657.user-desktop.44603.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9ee4b441cbf620eebbe40f818f1f3851cf37b89ebab4ded0ebc108202ead8277
|
| 3 |
+
size 417
|
test_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.9719696025912545,
|
| 4 |
+
"eval_loss": 0.11504850536584854,
|
| 5 |
+
"eval_runtime": 489.2058,
|
| 6 |
+
"eval_samples_per_second": 131.266,
|
| 7 |
+
"eval_steps_per_second": 8.205
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 1.4469915597827883e+20,
|
| 4 |
+
"train_loss": 0.618133510268158,
|
| 5 |
+
"train_runtime": 34089.5693,
|
| 6 |
+
"train_samples_per_second": 53.373,
|
| 7 |
+
"train_steps_per_second": 0.834
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,487 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.11504850536584854,
|
| 3 |
+
"best_model_checkpoint": "./vit-base-obipix_birdID-v1/checkpoint-28000",
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 28430,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.18,
|
| 13 |
+
"grad_norm": 109984.9296875,
|
| 14 |
+
"learning_rate": 0.00019296517762926486,
|
| 15 |
+
"loss": 6.9257,
|
| 16 |
+
"step": 1000
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.18,
|
| 20 |
+
"eval_accuracy": 0.16375981063909306,
|
| 21 |
+
"eval_loss": 5.382956027984619,
|
| 22 |
+
"eval_runtime": 503.1163,
|
| 23 |
+
"eval_samples_per_second": 127.636,
|
| 24 |
+
"eval_steps_per_second": 7.978,
|
| 25 |
+
"step": 1000
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.35,
|
| 29 |
+
"grad_norm": 176902.25,
|
| 30 |
+
"learning_rate": 0.00018593035525852974,
|
| 31 |
+
"loss": 3.9727,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.35,
|
| 36 |
+
"eval_accuracy": 0.4797246792076741,
|
| 37 |
+
"eval_loss": 2.769526243209839,
|
| 38 |
+
"eval_runtime": 503.5958,
|
| 39 |
+
"eval_samples_per_second": 127.515,
|
| 40 |
+
"eval_steps_per_second": 7.971,
|
| 41 |
+
"step": 2000
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.53,
|
| 45 |
+
"grad_norm": 220305.0625,
|
| 46 |
+
"learning_rate": 0.0001788955328877946,
|
| 47 |
+
"loss": 2.057,
|
| 48 |
+
"step": 3000
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.53,
|
| 52 |
+
"eval_accuracy": 0.693612183879407,
|
| 53 |
+
"eval_loss": 1.5069622993469238,
|
| 54 |
+
"eval_runtime": 494.5536,
|
| 55 |
+
"eval_samples_per_second": 129.846,
|
| 56 |
+
"eval_steps_per_second": 8.116,
|
| 57 |
+
"step": 3000
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.7,
|
| 61 |
+
"grad_norm": 227490.640625,
|
| 62 |
+
"learning_rate": 0.00017186071051705947,
|
| 63 |
+
"loss": 1.2103,
|
| 64 |
+
"step": 4000
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.7,
|
| 68 |
+
"eval_accuracy": 0.7841815123956646,
|
| 69 |
+
"eval_loss": 0.9726743102073669,
|
| 70 |
+
"eval_runtime": 494.0168,
|
| 71 |
+
"eval_samples_per_second": 129.987,
|
| 72 |
+
"eval_steps_per_second": 8.125,
|
| 73 |
+
"step": 4000
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.88,
|
| 77 |
+
"grad_norm": 246928.109375,
|
| 78 |
+
"learning_rate": 0.00016482588814632432,
|
| 79 |
+
"loss": 0.8513,
|
| 80 |
+
"step": 5000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.88,
|
| 84 |
+
"eval_accuracy": 0.8318487604335368,
|
| 85 |
+
"eval_loss": 0.710105836391449,
|
| 86 |
+
"eval_runtime": 492.9112,
|
| 87 |
+
"eval_samples_per_second": 130.279,
|
| 88 |
+
"eval_steps_per_second": 8.143,
|
| 89 |
+
"step": 5000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 1.06,
|
| 93 |
+
"grad_norm": 182599.234375,
|
| 94 |
+
"learning_rate": 0.00015779106577558917,
|
| 95 |
+
"loss": 0.5836,
|
| 96 |
+
"step": 6000
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 1.06,
|
| 100 |
+
"eval_accuracy": 0.8561106266351065,
|
| 101 |
+
"eval_loss": 0.57969069480896,
|
| 102 |
+
"eval_runtime": 497.4445,
|
| 103 |
+
"eval_samples_per_second": 129.092,
|
| 104 |
+
"eval_steps_per_second": 8.069,
|
| 105 |
+
"step": 6000
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.23,
|
| 109 |
+
"grad_norm": 174586.96875,
|
| 110 |
+
"learning_rate": 0.00015075624340485404,
|
| 111 |
+
"loss": 0.3545,
|
| 112 |
+
"step": 7000
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.23,
|
| 116 |
+
"eval_accuracy": 0.8729755824093683,
|
| 117 |
+
"eval_loss": 0.5065863132476807,
|
| 118 |
+
"eval_runtime": 490.3364,
|
| 119 |
+
"eval_samples_per_second": 130.963,
|
| 120 |
+
"eval_steps_per_second": 8.186,
|
| 121 |
+
"step": 7000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 1.41,
|
| 125 |
+
"grad_norm": 164026.03125,
|
| 126 |
+
"learning_rate": 0.0001437214210341189,
|
| 127 |
+
"loss": 0.314,
|
| 128 |
+
"step": 8000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.41,
|
| 132 |
+
"eval_accuracy": 0.8818363024791329,
|
| 133 |
+
"eval_loss": 0.4521096646785736,
|
| 134 |
+
"eval_runtime": 487.972,
|
| 135 |
+
"eval_samples_per_second": 131.598,
|
| 136 |
+
"eval_steps_per_second": 8.226,
|
| 137 |
+
"step": 8000
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.58,
|
| 141 |
+
"grad_norm": 164116.1875,
|
| 142 |
+
"learning_rate": 0.00013668659866338377,
|
| 143 |
+
"loss": 0.2858,
|
| 144 |
+
"step": 9000
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.58,
|
| 148 |
+
"eval_accuracy": 0.8959760807275445,
|
| 149 |
+
"eval_loss": 0.39154955744743347,
|
| 150 |
+
"eval_runtime": 498.2235,
|
| 151 |
+
"eval_samples_per_second": 128.89,
|
| 152 |
+
"eval_steps_per_second": 8.057,
|
| 153 |
+
"step": 9000
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 1.76,
|
| 157 |
+
"grad_norm": 142166.859375,
|
| 158 |
+
"learning_rate": 0.00012965177629264862,
|
| 159 |
+
"loss": 0.2482,
|
| 160 |
+
"step": 10000
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 1.76,
|
| 164 |
+
"eval_accuracy": 0.9055687056185374,
|
| 165 |
+
"eval_loss": 0.3563849925994873,
|
| 166 |
+
"eval_runtime": 491.8161,
|
| 167 |
+
"eval_samples_per_second": 130.569,
|
| 168 |
+
"eval_steps_per_second": 8.162,
|
| 169 |
+
"step": 10000
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 1.93,
|
| 173 |
+
"grad_norm": 160608.578125,
|
| 174 |
+
"learning_rate": 0.00012261695392191347,
|
| 175 |
+
"loss": 0.2192,
|
| 176 |
+
"step": 11000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.93,
|
| 180 |
+
"eval_accuracy": 0.9148343092064283,
|
| 181 |
+
"eval_loss": 0.3131472170352936,
|
| 182 |
+
"eval_runtime": 494.6571,
|
| 183 |
+
"eval_samples_per_second": 129.819,
|
| 184 |
+
"eval_steps_per_second": 8.115,
|
| 185 |
+
"step": 11000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 2.11,
|
| 189 |
+
"grad_norm": 72832.3359375,
|
| 190 |
+
"learning_rate": 0.00011558213155117834,
|
| 191 |
+
"loss": 0.1271,
|
| 192 |
+
"step": 12000
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 2.11,
|
| 196 |
+
"eval_accuracy": 0.9206739753332502,
|
| 197 |
+
"eval_loss": 0.29156896471977234,
|
| 198 |
+
"eval_runtime": 490.4403,
|
| 199 |
+
"eval_samples_per_second": 130.935,
|
| 200 |
+
"eval_steps_per_second": 8.184,
|
| 201 |
+
"step": 12000
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 2.29,
|
| 205 |
+
"grad_norm": 63710.625,
|
| 206 |
+
"learning_rate": 0.0001085473091804432,
|
| 207 |
+
"loss": 0.0779,
|
| 208 |
+
"step": 13000
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 2.29,
|
| 212 |
+
"eval_accuracy": 0.9259997508409119,
|
| 213 |
+
"eval_loss": 0.272681325674057,
|
| 214 |
+
"eval_runtime": 493.6688,
|
| 215 |
+
"eval_samples_per_second": 130.079,
|
| 216 |
+
"eval_steps_per_second": 8.131,
|
| 217 |
+
"step": 13000
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 2.46,
|
| 221 |
+
"grad_norm": 108940.078125,
|
| 222 |
+
"learning_rate": 0.00010151248680970806,
|
| 223 |
+
"loss": 0.0749,
|
| 224 |
+
"step": 14000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 2.46,
|
| 228 |
+
"eval_accuracy": 0.9308894979444375,
|
| 229 |
+
"eval_loss": 0.259682297706604,
|
| 230 |
+
"eval_runtime": 489.3469,
|
| 231 |
+
"eval_samples_per_second": 131.228,
|
| 232 |
+
"eval_steps_per_second": 8.203,
|
| 233 |
+
"step": 14000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 2.64,
|
| 237 |
+
"grad_norm": 118623.28125,
|
| 238 |
+
"learning_rate": 9.447766443897293e-05,
|
| 239 |
+
"loss": 0.0682,
|
| 240 |
+
"step": 15000
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 2.64,
|
| 244 |
+
"eval_accuracy": 0.9354677961878659,
|
| 245 |
+
"eval_loss": 0.2414509356021881,
|
| 246 |
+
"eval_runtime": 498.6017,
|
| 247 |
+
"eval_samples_per_second": 128.792,
|
| 248 |
+
"eval_steps_per_second": 8.051,
|
| 249 |
+
"step": 15000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.81,
|
| 253 |
+
"grad_norm": 59107.12890625,
|
| 254 |
+
"learning_rate": 8.744284206823779e-05,
|
| 255 |
+
"loss": 0.0615,
|
| 256 |
+
"step": 16000
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.81,
|
| 260 |
+
"eval_accuracy": 0.9384577052447988,
|
| 261 |
+
"eval_loss": 0.226842001080513,
|
| 262 |
+
"eval_runtime": 489.9279,
|
| 263 |
+
"eval_samples_per_second": 131.072,
|
| 264 |
+
"eval_steps_per_second": 8.193,
|
| 265 |
+
"step": 16000
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 2.99,
|
| 269 |
+
"grad_norm": 104112.5390625,
|
| 270 |
+
"learning_rate": 8.040801969750264e-05,
|
| 271 |
+
"loss": 0.0566,
|
| 272 |
+
"step": 17000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 2.99,
|
| 276 |
+
"eval_accuracy": 0.9439703500685187,
|
| 277 |
+
"eval_loss": 0.2083783596754074,
|
| 278 |
+
"eval_runtime": 498.3818,
|
| 279 |
+
"eval_samples_per_second": 128.849,
|
| 280 |
+
"eval_steps_per_second": 8.054,
|
| 281 |
+
"step": 17000
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 3.17,
|
| 285 |
+
"grad_norm": 65992.59375,
|
| 286 |
+
"learning_rate": 7.33731973267675e-05,
|
| 287 |
+
"loss": 0.0197,
|
| 288 |
+
"step": 18000
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 3.17,
|
| 292 |
+
"eval_accuracy": 0.9474897221876168,
|
| 293 |
+
"eval_loss": 0.19510340690612793,
|
| 294 |
+
"eval_runtime": 491.467,
|
| 295 |
+
"eval_samples_per_second": 130.662,
|
| 296 |
+
"eval_steps_per_second": 8.167,
|
| 297 |
+
"step": 18000
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 3.34,
|
| 301 |
+
"grad_norm": 27417.0,
|
| 302 |
+
"learning_rate": 6.633837495603237e-05,
|
| 303 |
+
"loss": 0.0158,
|
| 304 |
+
"step": 19000
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 3.34,
|
| 308 |
+
"eval_accuracy": 0.9513049707238072,
|
| 309 |
+
"eval_loss": 0.1843044012784958,
|
| 310 |
+
"eval_runtime": 491.4448,
|
| 311 |
+
"eval_samples_per_second": 130.668,
|
| 312 |
+
"eval_steps_per_second": 8.168,
|
| 313 |
+
"step": 19000
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 3.52,
|
| 317 |
+
"grad_norm": 9745.1337890625,
|
| 318 |
+
"learning_rate": 5.930355258529723e-05,
|
| 319 |
+
"loss": 0.0145,
|
| 320 |
+
"step": 20000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 3.52,
|
| 324 |
+
"eval_accuracy": 0.9541391553506914,
|
| 325 |
+
"eval_loss": 0.17461912333965302,
|
| 326 |
+
"eval_runtime": 497.3213,
|
| 327 |
+
"eval_samples_per_second": 129.124,
|
| 328 |
+
"eval_steps_per_second": 8.071,
|
| 329 |
+
"step": 20000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 3.69,
|
| 333 |
+
"grad_norm": 12192.4462890625,
|
| 334 |
+
"learning_rate": 5.226873021456209e-05,
|
| 335 |
+
"loss": 0.0118,
|
| 336 |
+
"step": 21000
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 3.69,
|
| 340 |
+
"eval_accuracy": 0.957269216394668,
|
| 341 |
+
"eval_loss": 0.1649465709924698,
|
| 342 |
+
"eval_runtime": 491.5529,
|
| 343 |
+
"eval_samples_per_second": 130.639,
|
| 344 |
+
"eval_steps_per_second": 8.166,
|
| 345 |
+
"step": 21000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 3.87,
|
| 349 |
+
"grad_norm": 18333.3828125,
|
| 350 |
+
"learning_rate": 4.523390784382695e-05,
|
| 351 |
+
"loss": 0.0103,
|
| 352 |
+
"step": 22000
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 3.87,
|
| 356 |
+
"eval_accuracy": 0.9599009592624891,
|
| 357 |
+
"eval_loss": 0.1530667096376419,
|
| 358 |
+
"eval_runtime": 499.5499,
|
| 359 |
+
"eval_samples_per_second": 128.548,
|
| 360 |
+
"eval_steps_per_second": 8.035,
|
| 361 |
+
"step": 22000
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 4.05,
|
| 365 |
+
"grad_norm": 1477.312744140625,
|
| 366 |
+
"learning_rate": 3.8199085473091805e-05,
|
| 367 |
+
"loss": 0.006,
|
| 368 |
+
"step": 23000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 4.05,
|
| 372 |
+
"eval_accuracy": 0.9643546779618787,
|
| 373 |
+
"eval_loss": 0.13792124390602112,
|
| 374 |
+
"eval_runtime": 499.2627,
|
| 375 |
+
"eval_samples_per_second": 128.622,
|
| 376 |
+
"eval_steps_per_second": 8.04,
|
| 377 |
+
"step": 23000
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 4.22,
|
| 381 |
+
"grad_norm": 903.4931030273438,
|
| 382 |
+
"learning_rate": 3.116426310235667e-05,
|
| 383 |
+
"loss": 0.0016,
|
| 384 |
+
"step": 24000
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 4.22,
|
| 388 |
+
"eval_accuracy": 0.9667528341846269,
|
| 389 |
+
"eval_loss": 0.13162368535995483,
|
| 390 |
+
"eval_runtime": 486.4405,
|
| 391 |
+
"eval_samples_per_second": 132.012,
|
| 392 |
+
"eval_steps_per_second": 8.252,
|
| 393 |
+
"step": 24000
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 4.4,
|
| 397 |
+
"grad_norm": 1672.1170654296875,
|
| 398 |
+
"learning_rate": 2.412944073162153e-05,
|
| 399 |
+
"loss": 0.0013,
|
| 400 |
+
"step": 25000
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 4.4,
|
| 404 |
+
"eval_accuracy": 0.9685903824592001,
|
| 405 |
+
"eval_loss": 0.12647932767868042,
|
| 406 |
+
"eval_runtime": 488.945,
|
| 407 |
+
"eval_samples_per_second": 131.336,
|
| 408 |
+
"eval_steps_per_second": 8.21,
|
| 409 |
+
"step": 25000
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 4.57,
|
| 413 |
+
"grad_norm": 179.9572296142578,
|
| 414 |
+
"learning_rate": 1.709461836088639e-05,
|
| 415 |
+
"loss": 0.0014,
|
| 416 |
+
"step": 26000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 4.57,
|
| 420 |
+
"eval_accuracy": 0.9697427432415597,
|
| 421 |
+
"eval_loss": 0.1232375055551529,
|
| 422 |
+
"eval_runtime": 488.3862,
|
| 423 |
+
"eval_samples_per_second": 131.486,
|
| 424 |
+
"eval_steps_per_second": 8.219,
|
| 425 |
+
"step": 26000
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 4.75,
|
| 429 |
+
"grad_norm": 538.2374877929688,
|
| 430 |
+
"learning_rate": 1.005979599015125e-05,
|
| 431 |
+
"loss": 0.0009,
|
| 432 |
+
"step": 27000
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 4.75,
|
| 436 |
+
"eval_accuracy": 0.9711598355550018,
|
| 437 |
+
"eval_loss": 0.11887777596712112,
|
| 438 |
+
"eval_runtime": 490.273,
|
| 439 |
+
"eval_samples_per_second": 130.98,
|
| 440 |
+
"eval_steps_per_second": 8.187,
|
| 441 |
+
"step": 27000
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 4.92,
|
| 445 |
+
"grad_norm": 845.1385498046875,
|
| 446 |
+
"learning_rate": 3.02497361941611e-06,
|
| 447 |
+
"loss": 0.001,
|
| 448 |
+
"step": 28000
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 4.92,
|
| 452 |
+
"eval_accuracy": 0.9719696025912545,
|
| 453 |
+
"eval_loss": 0.11504850536584854,
|
| 454 |
+
"eval_runtime": 495.362,
|
| 455 |
+
"eval_samples_per_second": 129.634,
|
| 456 |
+
"eval_steps_per_second": 8.103,
|
| 457 |
+
"step": 28000
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 5.0,
|
| 461 |
+
"step": 28430,
|
| 462 |
+
"total_flos": 1.4469915597827883e+20,
|
| 463 |
+
"train_loss": 0.618133510268158,
|
| 464 |
+
"train_runtime": 34089.5693,
|
| 465 |
+
"train_samples_per_second": 53.373,
|
| 466 |
+
"train_steps_per_second": 0.834
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 5.0,
|
| 470 |
+
"eval_accuracy": 0.9719696025912545,
|
| 471 |
+
"eval_loss": 0.11504850536584854,
|
| 472 |
+
"eval_runtime": 489.2058,
|
| 473 |
+
"eval_samples_per_second": 131.266,
|
| 474 |
+
"eval_steps_per_second": 8.205,
|
| 475 |
+
"step": 28430
|
| 476 |
+
}
|
| 477 |
+
],
|
| 478 |
+
"logging_steps": 1000,
|
| 479 |
+
"max_steps": 28430,
|
| 480 |
+
"num_input_tokens_seen": 0,
|
| 481 |
+
"num_train_epochs": 5,
|
| 482 |
+
"save_steps": 1000,
|
| 483 |
+
"total_flos": 1.4469915597827883e+20,
|
| 484 |
+
"train_batch_size": 64,
|
| 485 |
+
"trial_name": null,
|
| 486 |
+
"trial_params": null
|
| 487 |
+
}
|