Upload folder using huggingface_hub
Browse files- all_results.json +26 -0
- celeba_results.json +10 -0
- model.safetensors +3 -0
- per_sample.csv +0 -0
- pred_celeba.npz.npy +3 -0
- preprocessor_config.json +36 -0
- rep_celeba.pt +3 -0
- test_results.json +14 -0
- train_results.json +7 -0
- trainer_state.json +315 -0
- training_args.bin +3 -0
all_results.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"ave/id": 0.934262847938793,
|
| 3 |
+
"ave/ood": 0.5164835164835165,
|
| 4 |
+
"celeba": 0.934262847938793,
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_accuracy": 0.934262847938793,
|
| 7 |
+
"eval_combined_score": 0.7253731822111548,
|
| 8 |
+
"eval_loss": 0.2687215209007263,
|
| 9 |
+
"eval_ood": 0.5164835164835165,
|
| 10 |
+
"eval_runtime": 54.9748,
|
| 11 |
+
"eval_samples_per_second": 361.384,
|
| 12 |
+
"eval_steps_per_second": 1.419,
|
| 13 |
+
"gap/g1": 0.011005671606045486,
|
| 14 |
+
"gap/g2": -0.044833333791511554,
|
| 15 |
+
"gap/g3": 0.06996222163399135,
|
| 16 |
+
"gap/g4": 0.41777933145527646,
|
| 17 |
+
"gap/ood": 0.41777933145527646,
|
| 18 |
+
"ood/g1": 0.9232571763327475,
|
| 19 |
+
"ood/g2": 0.9790961817303045,
|
| 20 |
+
"ood/g3": 0.8643006263048016,
|
| 21 |
+
"ood/g4": 0.5164835164835165,
|
| 22 |
+
"train_loss": 0.09458405806583428,
|
| 23 |
+
"train_runtime": 12910.7556,
|
| 24 |
+
"train_samples_per_second": 126.073,
|
| 25 |
+
"train_steps_per_second": 0.985
|
| 26 |
+
}
|
celeba_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_accuracy": 0.934262847938793,
|
| 4 |
+
"eval_combined_score": 0.7253731822111548,
|
| 5 |
+
"eval_loss": 0.2687215209007263,
|
| 6 |
+
"eval_ood": 0.5164835164835165,
|
| 7 |
+
"eval_runtime": 54.9748,
|
| 8 |
+
"eval_samples_per_second": 361.384,
|
| 9 |
+
"eval_steps_per_second": 1.419
|
| 10 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99c9db3d9624014354d6f66e96a1170473a6119c8fd3a66c054d5e93423696db
|
| 3 |
+
size 345586928
|
per_sample.csv
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
pred_celeba.npz.npy
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f2c5d8730dd556caa2c7cfc636f0341792a15fbbdb192aae524f352d98e006bd
|
| 3 |
+
size 159064
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_valid_processor_keys": [
|
| 3 |
+
"images",
|
| 4 |
+
"do_resize",
|
| 5 |
+
"size",
|
| 6 |
+
"resample",
|
| 7 |
+
"do_rescale",
|
| 8 |
+
"rescale_factor",
|
| 9 |
+
"do_normalize",
|
| 10 |
+
"image_mean",
|
| 11 |
+
"image_std",
|
| 12 |
+
"return_tensors",
|
| 13 |
+
"data_format",
|
| 14 |
+
"input_data_format"
|
| 15 |
+
],
|
| 16 |
+
"do_normalize": true,
|
| 17 |
+
"do_rescale": true,
|
| 18 |
+
"do_resize": true,
|
| 19 |
+
"image_mean": [
|
| 20 |
+
0.5,
|
| 21 |
+
0.5,
|
| 22 |
+
0.5
|
| 23 |
+
],
|
| 24 |
+
"image_processor_type": "ViTImageProcessor",
|
| 25 |
+
"image_std": [
|
| 26 |
+
0.5,
|
| 27 |
+
0.5,
|
| 28 |
+
0.5
|
| 29 |
+
],
|
| 30 |
+
"resample": 2,
|
| 31 |
+
"rescale_factor": 0.00392156862745098,
|
| 32 |
+
"size": {
|
| 33 |
+
"height": 224,
|
| 34 |
+
"width": 224
|
| 35 |
+
}
|
| 36 |
+
}
|
rep_celeba.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:221ae1ab8890fc9d10d6a67fcc7c22c90363714ff02af0d3ee370c734d1816a2
|
| 3 |
+
size 732635110
|
test_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"ave/id": 0.934262847938793,
|
| 3 |
+
"ave/ood": 0.5164835164835165,
|
| 4 |
+
"celeba": 0.934262847938793,
|
| 5 |
+
"gap/g1": 0.011005671606045486,
|
| 6 |
+
"gap/g2": -0.044833333791511554,
|
| 7 |
+
"gap/g3": 0.06996222163399135,
|
| 8 |
+
"gap/g4": 0.41777933145527646,
|
| 9 |
+
"gap/ood": 0.41777933145527646,
|
| 10 |
+
"ood/g1": 0.9232571763327475,
|
| 11 |
+
"ood/g2": 0.9790961817303045,
|
| 12 |
+
"ood/g3": 0.8643006263048016,
|
| 13 |
+
"ood/g4": 0.5164835164835165
|
| 14 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"train_loss": 0.09458405806583428,
|
| 4 |
+
"train_runtime": 12910.7556,
|
| 5 |
+
"train_samples_per_second": 126.073,
|
| 6 |
+
"train_steps_per_second": 0.985
|
| 7 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,315 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.7436654643823264,
|
| 3 |
+
"best_model_checkpoint": "outputs/vit-base/pgd/celeba_21/checkpoint-6360",
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 12720,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.39,
|
| 13 |
+
"grad_norm": 1.9278849363327026,
|
| 14 |
+
"learning_rate": 9.60691823899371e-05,
|
| 15 |
+
"loss": 0.4324,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.79,
|
| 20 |
+
"grad_norm": 1.8189257383346558,
|
| 21 |
+
"learning_rate": 9.213836477987423e-05,
|
| 22 |
+
"loss": 0.2609,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 1.0,
|
| 27 |
+
"eval_accuracy": 0.9457469191463781,
|
| 28 |
+
"eval_combined_score": 0.736762348462078,
|
| 29 |
+
"eval_loss": 0.17756877839565277,
|
| 30 |
+
"eval_ood": 0.5277777777777778,
|
| 31 |
+
"eval_runtime": 54.4552,
|
| 32 |
+
"eval_samples_per_second": 366.577,
|
| 33 |
+
"eval_steps_per_second": 1.432,
|
| 34 |
+
"step": 1272
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 1.18,
|
| 38 |
+
"grad_norm": 1.7144166231155396,
|
| 39 |
+
"learning_rate": 8.820754716981132e-05,
|
| 40 |
+
"loss": 0.1939,
|
| 41 |
+
"step": 1500
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 1.57,
|
| 45 |
+
"grad_norm": 1.43351411819458,
|
| 46 |
+
"learning_rate": 8.427672955974843e-05,
|
| 47 |
+
"loss": 0.1615,
|
| 48 |
+
"step": 2000
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 1.97,
|
| 52 |
+
"grad_norm": 1.4217851161956787,
|
| 53 |
+
"learning_rate": 8.034591194968554e-05,
|
| 54 |
+
"loss": 0.1397,
|
| 55 |
+
"step": 2500
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 2.0,
|
| 59 |
+
"eval_accuracy": 0.9457970143272217,
|
| 60 |
+
"eval_combined_score": 0.6978985071636109,
|
| 61 |
+
"eval_loss": 0.19592541456222534,
|
| 62 |
+
"eval_ood": 0.45,
|
| 63 |
+
"eval_runtime": 53.7806,
|
| 64 |
+
"eval_samples_per_second": 371.175,
|
| 65 |
+
"eval_steps_per_second": 1.45,
|
| 66 |
+
"step": 2544
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 2.36,
|
| 70 |
+
"grad_norm": 1.4929840564727783,
|
| 71 |
+
"learning_rate": 7.641509433962265e-05,
|
| 72 |
+
"loss": 0.1216,
|
| 73 |
+
"step": 3000
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 2.75,
|
| 77 |
+
"grad_norm": 1.1121305227279663,
|
| 78 |
+
"learning_rate": 7.248427672955975e-05,
|
| 79 |
+
"loss": 0.1121,
|
| 80 |
+
"step": 3500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 3.0,
|
| 84 |
+
"eval_accuracy": 0.9461476805931269,
|
| 85 |
+
"eval_combined_score": 0.6758516180743412,
|
| 86 |
+
"eval_loss": 0.2125873565673828,
|
| 87 |
+
"eval_ood": 0.40555555555555556,
|
| 88 |
+
"eval_runtime": 53.536,
|
| 89 |
+
"eval_samples_per_second": 372.871,
|
| 90 |
+
"eval_steps_per_second": 1.457,
|
| 91 |
+
"step": 3816
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 3.14,
|
| 95 |
+
"grad_norm": 1.3586345911026,
|
| 96 |
+
"learning_rate": 6.855345911949685e-05,
|
| 97 |
+
"loss": 0.0987,
|
| 98 |
+
"step": 4000
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 3.54,
|
| 102 |
+
"grad_norm": 2.9185080528259277,
|
| 103 |
+
"learning_rate": 6.462264150943397e-05,
|
| 104 |
+
"loss": 0.0873,
|
| 105 |
+
"step": 4500
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 3.93,
|
| 109 |
+
"grad_norm": 1.238891363143921,
|
| 110 |
+
"learning_rate": 6.069182389937107e-05,
|
| 111 |
+
"loss": 0.0831,
|
| 112 |
+
"step": 5000
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 4.0,
|
| 116 |
+
"eval_accuracy": 0.9460474902314397,
|
| 117 |
+
"eval_combined_score": 0.6674681895601643,
|
| 118 |
+
"eval_loss": 0.24287694692611694,
|
| 119 |
+
"eval_ood": 0.3888888888888889,
|
| 120 |
+
"eval_runtime": 54.0501,
|
| 121 |
+
"eval_samples_per_second": 369.324,
|
| 122 |
+
"eval_steps_per_second": 1.443,
|
| 123 |
+
"step": 5088
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 4.32,
|
| 127 |
+
"grad_norm": 1.0512090921401978,
|
| 128 |
+
"learning_rate": 5.676100628930818e-05,
|
| 129 |
+
"loss": 0.0738,
|
| 130 |
+
"step": 5500
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 4.72,
|
| 134 |
+
"grad_norm": 1.0493574142456055,
|
| 135 |
+
"learning_rate": 5.283018867924528e-05,
|
| 136 |
+
"loss": 0.0735,
|
| 137 |
+
"step": 6000
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 5.0,
|
| 141 |
+
"eval_accuracy": 0.9373309287646528,
|
| 142 |
+
"eval_combined_score": 0.7436654643823264,
|
| 143 |
+
"eval_loss": 0.2648378312587738,
|
| 144 |
+
"eval_ood": 0.55,
|
| 145 |
+
"eval_runtime": 53.577,
|
| 146 |
+
"eval_samples_per_second": 372.586,
|
| 147 |
+
"eval_steps_per_second": 1.456,
|
| 148 |
+
"step": 6360
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 5.11,
|
| 152 |
+
"grad_norm": 1.6157053709030151,
|
| 153 |
+
"learning_rate": 4.8899371069182394e-05,
|
| 154 |
+
"loss": 0.0654,
|
| 155 |
+
"step": 6500
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 5.5,
|
| 159 |
+
"grad_norm": 1.053931713104248,
|
| 160 |
+
"learning_rate": 4.49685534591195e-05,
|
| 161 |
+
"loss": 0.0581,
|
| 162 |
+
"step": 7000
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 5.9,
|
| 166 |
+
"grad_norm": 1.1942349672317505,
|
| 167 |
+
"learning_rate": 4.103773584905661e-05,
|
| 168 |
+
"loss": 0.0559,
|
| 169 |
+
"step": 7500
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 6.0,
|
| 173 |
+
"eval_accuracy": 0.9455966336038473,
|
| 174 |
+
"eval_combined_score": 0.6839094279130348,
|
| 175 |
+
"eval_loss": 0.2476591318845749,
|
| 176 |
+
"eval_ood": 0.4222222222222222,
|
| 177 |
+
"eval_runtime": 53.9022,
|
| 178 |
+
"eval_samples_per_second": 370.337,
|
| 179 |
+
"eval_steps_per_second": 1.447,
|
| 180 |
+
"step": 7632
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 6.29,
|
| 184 |
+
"grad_norm": 1.2728811502456665,
|
| 185 |
+
"learning_rate": 3.710691823899371e-05,
|
| 186 |
+
"loss": 0.0498,
|
| 187 |
+
"step": 8000
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 6.68,
|
| 191 |
+
"grad_norm": 0.5797112584114075,
|
| 192 |
+
"learning_rate": 3.3176100628930816e-05,
|
| 193 |
+
"loss": 0.0469,
|
| 194 |
+
"step": 8500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 7.0,
|
| 198 |
+
"eval_accuracy": 0.9459973950505961,
|
| 199 |
+
"eval_combined_score": 0.7202209197475202,
|
| 200 |
+
"eval_loss": 0.27472054958343506,
|
| 201 |
+
"eval_ood": 0.49444444444444446,
|
| 202 |
+
"eval_runtime": 53.8167,
|
| 203 |
+
"eval_samples_per_second": 370.926,
|
| 204 |
+
"eval_steps_per_second": 1.449,
|
| 205 |
+
"step": 8904
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 7.08,
|
| 209 |
+
"grad_norm": 0.7845437526702881,
|
| 210 |
+
"learning_rate": 2.9245283018867926e-05,
|
| 211 |
+
"loss": 0.0439,
|
| 212 |
+
"step": 9000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 7.47,
|
| 216 |
+
"grad_norm": 2.0326216220855713,
|
| 217 |
+
"learning_rate": 2.5314465408805033e-05,
|
| 218 |
+
"loss": 0.0402,
|
| 219 |
+
"step": 9500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 7.86,
|
| 223 |
+
"grad_norm": 1.2046332359313965,
|
| 224 |
+
"learning_rate": 2.138364779874214e-05,
|
| 225 |
+
"loss": 0.0372,
|
| 226 |
+
"step": 10000
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 8.0,
|
| 230 |
+
"eval_accuracy": 0.9455465384230037,
|
| 231 |
+
"eval_combined_score": 0.6783288247670574,
|
| 232 |
+
"eval_loss": 0.2704795002937317,
|
| 233 |
+
"eval_ood": 0.4111111111111111,
|
| 234 |
+
"eval_runtime": 53.9509,
|
| 235 |
+
"eval_samples_per_second": 370.003,
|
| 236 |
+
"eval_steps_per_second": 1.446,
|
| 237 |
+
"step": 10176
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 8.25,
|
| 241 |
+
"grad_norm": 2.038419246673584,
|
| 242 |
+
"learning_rate": 1.7452830188679244e-05,
|
| 243 |
+
"loss": 0.0352,
|
| 244 |
+
"step": 10500
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 8.65,
|
| 248 |
+
"grad_norm": 1.4056214094161987,
|
| 249 |
+
"learning_rate": 1.3522012578616353e-05,
|
| 250 |
+
"loss": 0.0336,
|
| 251 |
+
"step": 11000
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 9.0,
|
| 255 |
+
"eval_accuracy": 0.9451457769762549,
|
| 256 |
+
"eval_combined_score": 0.6920173329325718,
|
| 257 |
+
"eval_loss": 0.28445887565612793,
|
| 258 |
+
"eval_ood": 0.4388888888888889,
|
| 259 |
+
"eval_runtime": 53.8553,
|
| 260 |
+
"eval_samples_per_second": 370.66,
|
| 261 |
+
"eval_steps_per_second": 1.448,
|
| 262 |
+
"step": 11448
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 9.04,
|
| 266 |
+
"grad_norm": 0.5328162908554077,
|
| 267 |
+
"learning_rate": 9.59119496855346e-06,
|
| 268 |
+
"loss": 0.0309,
|
| 269 |
+
"step": 11500
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 9.43,
|
| 273 |
+
"grad_norm": 1.3629292249679565,
|
| 274 |
+
"learning_rate": 5.660377358490566e-06,
|
| 275 |
+
"loss": 0.0303,
|
| 276 |
+
"step": 12000
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 9.83,
|
| 280 |
+
"grad_norm": 0.6843909025192261,
|
| 281 |
+
"learning_rate": 1.729559748427673e-06,
|
| 282 |
+
"loss": 0.0287,
|
| 283 |
+
"step": 12500
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 10.0,
|
| 287 |
+
"eval_accuracy": 0.9453461576996293,
|
| 288 |
+
"eval_combined_score": 0.7004508566275924,
|
| 289 |
+
"eval_loss": 0.29068228602409363,
|
| 290 |
+
"eval_ood": 0.45555555555555555,
|
| 291 |
+
"eval_runtime": 53.9886,
|
| 292 |
+
"eval_samples_per_second": 369.745,
|
| 293 |
+
"eval_steps_per_second": 1.445,
|
| 294 |
+
"step": 12720
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 10.0,
|
| 298 |
+
"step": 12720,
|
| 299 |
+
"total_flos": 0.0,
|
| 300 |
+
"train_loss": 0.09458405806583428,
|
| 301 |
+
"train_runtime": 12910.7556,
|
| 302 |
+
"train_samples_per_second": 126.073,
|
| 303 |
+
"train_steps_per_second": 0.985
|
| 304 |
+
}
|
| 305 |
+
],
|
| 306 |
+
"logging_steps": 500,
|
| 307 |
+
"max_steps": 12720,
|
| 308 |
+
"num_input_tokens_seen": 0,
|
| 309 |
+
"num_train_epochs": 10,
|
| 310 |
+
"save_steps": 500,
|
| 311 |
+
"total_flos": 0.0,
|
| 312 |
+
"train_batch_size": 128,
|
| 313 |
+
"trial_name": null,
|
| 314 |
+
"trial_params": null
|
| 315 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72e2370287940ceeb07ed9c39cb4196002b9feb7ada42511cfb549f2ee462ae8
|
| 3 |
+
size 5816
|