Commit
·
be4f252
1
Parent(s):
ef58386
Training in progress, epoch 1
Browse files- all_results.json +8 -0
- config.json +40 -0
- model.safetensors +3 -0
- preprocessor_config.json +22 -0
- runs/Dec01_15-07-49_c33b0eaaad7d/events.out.tfevents.1701443278.c33b0eaaad7d.1499.0 +3 -0
- runs/Dec01_15-14-59_c33b0eaaad7d/events.out.tfevents.1701443706.c33b0eaaad7d.1499.1 +3 -0
- runs/Dec01_15-22-53_c33b0eaaad7d/events.out.tfevents.1701444177.c33b0eaaad7d.1499.2 +3 -0
- runs/Dec01_15-33-09_c33b0eaaad7d/events.out.tfevents.1701444805.c33b0eaaad7d.1499.3 +3 -0
- train_results.json +8 -0
- trainer_state.json +415 -0
- training_args.bin +3 -0
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 1.5064982852419584e+18,
|
| 4 |
+
"train_loss": 0.23008910012362627,
|
| 5 |
+
"train_runtime": 477.4194,
|
| 6 |
+
"train_samples_per_second": 40.719,
|
| 7 |
+
"train_steps_per_second": 1.276
|
| 8 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,40 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "google/vit-base-patch16-224-in21k",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ViTForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.0,
|
| 7 |
+
"encoder_stride": 16,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.0,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "crack",
|
| 13 |
+
"1": "scratch",
|
| 14 |
+
"2": "tire flat",
|
| 15 |
+
"3": "dent",
|
| 16 |
+
"4": "glass shatter",
|
| 17 |
+
"5": "lamp broken"
|
| 18 |
+
},
|
| 19 |
+
"image_size": 224,
|
| 20 |
+
"initializer_range": 0.02,
|
| 21 |
+
"intermediate_size": 3072,
|
| 22 |
+
"label2id": {
|
| 23 |
+
"crack": 0,
|
| 24 |
+
"dent": 3,
|
| 25 |
+
"glass shatter": 4,
|
| 26 |
+
"lamp broken": 5,
|
| 27 |
+
"scratch": 1,
|
| 28 |
+
"tire flat": 2
|
| 29 |
+
},
|
| 30 |
+
"layer_norm_eps": 1e-12,
|
| 31 |
+
"model_type": "vit",
|
| 32 |
+
"num_attention_heads": 12,
|
| 33 |
+
"num_channels": 3,
|
| 34 |
+
"num_hidden_layers": 12,
|
| 35 |
+
"patch_size": 16,
|
| 36 |
+
"problem_type": "single_label_classification",
|
| 37 |
+
"qkv_bias": true,
|
| 38 |
+
"torch_dtype": "float32",
|
| 39 |
+
"transformers_version": "4.35.2"
|
| 40 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c052c446e9fed118f4f7dc668ba6f41df6c765b7a76d4bd2fa9950ab3efa8d0b
|
| 3 |
+
size 343236280
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_normalize": true,
|
| 3 |
+
"do_rescale": true,
|
| 4 |
+
"do_resize": true,
|
| 5 |
+
"image_mean": [
|
| 6 |
+
0.5,
|
| 7 |
+
0.5,
|
| 8 |
+
0.5
|
| 9 |
+
],
|
| 10 |
+
"image_processor_type": "ViTImageProcessor",
|
| 11 |
+
"image_std": [
|
| 12 |
+
0.5,
|
| 13 |
+
0.5,
|
| 14 |
+
0.5
|
| 15 |
+
],
|
| 16 |
+
"resample": 2,
|
| 17 |
+
"rescale_factor": 0.00392156862745098,
|
| 18 |
+
"size": {
|
| 19 |
+
"height": 224,
|
| 20 |
+
"width": 224
|
| 21 |
+
}
|
| 22 |
+
}
|
runs/Dec01_15-07-49_c33b0eaaad7d/events.out.tfevents.1701443278.c33b0eaaad7d.1499.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f153a3fcebe98d7b47fe91ab8daee5a82ea78bdcf66cfa4764509a0bf202e95b
|
| 3 |
+
size 11458
|
runs/Dec01_15-14-59_c33b0eaaad7d/events.out.tfevents.1701443706.c33b0eaaad7d.1499.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1030200f468fda6127617e3d2bd273d8b8029b49cb44d2a1fa692995a802e4c8
|
| 3 |
+
size 11553
|
runs/Dec01_15-22-53_c33b0eaaad7d/events.out.tfevents.1701444177.c33b0eaaad7d.1499.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1820e01e6f6f93f0e229ff94e01dd89eff8bbfb7fd94e35f42b1a1829a59e2b4
|
| 3 |
+
size 15119
|
runs/Dec01_15-33-09_c33b0eaaad7d/events.out.tfevents.1701444805.c33b0eaaad7d.1499.3
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b1edde635b4a3a6d4f61206a7e531670eaf0105865832981527291774e17cc86
|
| 3 |
+
size 7931
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 1.5064982852419584e+18,
|
| 4 |
+
"train_loss": 0.23008910012362627,
|
| 5 |
+
"train_runtime": 477.4194,
|
| 6 |
+
"train_samples_per_second": 40.719,
|
| 7 |
+
"train_steps_per_second": 1.276
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,415 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.030118444934487343,
|
| 3 |
+
"best_model_checkpoint": "./vit-base-crack-classification/checkpoint-609",
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 609,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.05,
|
| 13 |
+
"learning_rate": 0.00019671592775041052,
|
| 14 |
+
"loss": 1.4989,
|
| 15 |
+
"step": 10
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.1,
|
| 19 |
+
"learning_rate": 0.00019343185550082105,
|
| 20 |
+
"loss": 1.059,
|
| 21 |
+
"step": 20
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.15,
|
| 25 |
+
"learning_rate": 0.00019014778325123153,
|
| 26 |
+
"loss": 0.8111,
|
| 27 |
+
"step": 30
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.2,
|
| 31 |
+
"learning_rate": 0.00018686371100164203,
|
| 32 |
+
"loss": 0.6727,
|
| 33 |
+
"step": 40
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.25,
|
| 37 |
+
"learning_rate": 0.00018357963875205257,
|
| 38 |
+
"loss": 0.7606,
|
| 39 |
+
"step": 50
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.3,
|
| 43 |
+
"learning_rate": 0.00018029556650246307,
|
| 44 |
+
"loss": 0.688,
|
| 45 |
+
"step": 60
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.34,
|
| 49 |
+
"learning_rate": 0.00017701149425287358,
|
| 50 |
+
"loss": 0.6076,
|
| 51 |
+
"step": 70
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.39,
|
| 55 |
+
"learning_rate": 0.00017372742200328408,
|
| 56 |
+
"loss": 0.5333,
|
| 57 |
+
"step": 80
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.44,
|
| 61 |
+
"learning_rate": 0.0001704433497536946,
|
| 62 |
+
"loss": 0.4982,
|
| 63 |
+
"step": 90
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.49,
|
| 67 |
+
"learning_rate": 0.0001671592775041051,
|
| 68 |
+
"loss": 0.4319,
|
| 69 |
+
"step": 100
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.54,
|
| 73 |
+
"learning_rate": 0.0001638752052545156,
|
| 74 |
+
"loss": 0.5499,
|
| 75 |
+
"step": 110
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.59,
|
| 79 |
+
"learning_rate": 0.00016059113300492613,
|
| 80 |
+
"loss": 0.3999,
|
| 81 |
+
"step": 120
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.64,
|
| 85 |
+
"learning_rate": 0.00015730706075533664,
|
| 86 |
+
"loss": 0.443,
|
| 87 |
+
"step": 130
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.69,
|
| 91 |
+
"learning_rate": 0.00015402298850574712,
|
| 92 |
+
"loss": 0.417,
|
| 93 |
+
"step": 140
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.74,
|
| 97 |
+
"learning_rate": 0.00015073891625615765,
|
| 98 |
+
"loss": 0.3383,
|
| 99 |
+
"step": 150
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.79,
|
| 103 |
+
"learning_rate": 0.00014745484400656815,
|
| 104 |
+
"loss": 0.4005,
|
| 105 |
+
"step": 160
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.84,
|
| 109 |
+
"learning_rate": 0.00014417077175697866,
|
| 110 |
+
"loss": 0.3608,
|
| 111 |
+
"step": 170
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.89,
|
| 115 |
+
"learning_rate": 0.00014088669950738917,
|
| 116 |
+
"loss": 0.3782,
|
| 117 |
+
"step": 180
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.94,
|
| 121 |
+
"learning_rate": 0.0001376026272577997,
|
| 122 |
+
"loss": 0.3411,
|
| 123 |
+
"step": 190
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.99,
|
| 127 |
+
"learning_rate": 0.00013431855500821018,
|
| 128 |
+
"loss": 0.2138,
|
| 129 |
+
"step": 200
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.0,
|
| 133 |
+
"eval_accuracy": 0.9027777777777778,
|
| 134 |
+
"eval_loss": 0.2958821952342987,
|
| 135 |
+
"eval_runtime": 15.6222,
|
| 136 |
+
"eval_samples_per_second": 46.088,
|
| 137 |
+
"eval_steps_per_second": 5.761,
|
| 138 |
+
"step": 203
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 1.03,
|
| 142 |
+
"learning_rate": 0.00013103448275862068,
|
| 143 |
+
"loss": 0.1832,
|
| 144 |
+
"step": 210
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.08,
|
| 148 |
+
"learning_rate": 0.00012775041050903122,
|
| 149 |
+
"loss": 0.1517,
|
| 150 |
+
"step": 220
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.13,
|
| 154 |
+
"learning_rate": 0.00012446633825944172,
|
| 155 |
+
"loss": 0.1099,
|
| 156 |
+
"step": 230
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 1.18,
|
| 160 |
+
"learning_rate": 0.00012118226600985223,
|
| 161 |
+
"loss": 0.1762,
|
| 162 |
+
"step": 240
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 1.23,
|
| 166 |
+
"learning_rate": 0.00011789819376026272,
|
| 167 |
+
"loss": 0.1496,
|
| 168 |
+
"step": 250
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.28,
|
| 172 |
+
"learning_rate": 0.00011461412151067324,
|
| 173 |
+
"loss": 0.1287,
|
| 174 |
+
"step": 260
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 1.33,
|
| 178 |
+
"learning_rate": 0.00011133004926108374,
|
| 179 |
+
"loss": 0.1389,
|
| 180 |
+
"step": 270
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 1.38,
|
| 184 |
+
"learning_rate": 0.00010804597701149426,
|
| 185 |
+
"loss": 0.1583,
|
| 186 |
+
"step": 280
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 1.43,
|
| 190 |
+
"learning_rate": 0.00010476190476190477,
|
| 191 |
+
"loss": 0.1214,
|
| 192 |
+
"step": 290
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.48,
|
| 196 |
+
"learning_rate": 0.00010147783251231529,
|
| 197 |
+
"loss": 0.1566,
|
| 198 |
+
"step": 300
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.53,
|
| 202 |
+
"learning_rate": 9.819376026272578e-05,
|
| 203 |
+
"loss": 0.1256,
|
| 204 |
+
"step": 310
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 1.58,
|
| 208 |
+
"learning_rate": 9.490968801313629e-05,
|
| 209 |
+
"loss": 0.1123,
|
| 210 |
+
"step": 320
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.63,
|
| 214 |
+
"learning_rate": 9.16256157635468e-05,
|
| 215 |
+
"loss": 0.0853,
|
| 216 |
+
"step": 330
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 1.67,
|
| 220 |
+
"learning_rate": 8.834154351395731e-05,
|
| 221 |
+
"loss": 0.067,
|
| 222 |
+
"step": 340
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 1.72,
|
| 226 |
+
"learning_rate": 8.505747126436782e-05,
|
| 227 |
+
"loss": 0.1035,
|
| 228 |
+
"step": 350
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.77,
|
| 232 |
+
"learning_rate": 8.177339901477834e-05,
|
| 233 |
+
"loss": 0.0681,
|
| 234 |
+
"step": 360
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.82,
|
| 238 |
+
"learning_rate": 7.848932676518884e-05,
|
| 239 |
+
"loss": 0.0382,
|
| 240 |
+
"step": 370
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.87,
|
| 244 |
+
"learning_rate": 7.520525451559935e-05,
|
| 245 |
+
"loss": 0.0537,
|
| 246 |
+
"step": 380
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 1.92,
|
| 250 |
+
"learning_rate": 7.192118226600985e-05,
|
| 251 |
+
"loss": 0.0283,
|
| 252 |
+
"step": 390
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.97,
|
| 256 |
+
"learning_rate": 6.863711001642037e-05,
|
| 257 |
+
"loss": 0.0291,
|
| 258 |
+
"step": 400
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 2.0,
|
| 262 |
+
"eval_accuracy": 0.9833333333333333,
|
| 263 |
+
"eval_loss": 0.06349290162324905,
|
| 264 |
+
"eval_runtime": 15.4242,
|
| 265 |
+
"eval_samples_per_second": 46.68,
|
| 266 |
+
"eval_steps_per_second": 5.835,
|
| 267 |
+
"step": 406
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 2.02,
|
| 271 |
+
"learning_rate": 6.535303776683088e-05,
|
| 272 |
+
"loss": 0.0275,
|
| 273 |
+
"step": 410
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 2.07,
|
| 277 |
+
"learning_rate": 6.206896551724138e-05,
|
| 278 |
+
"loss": 0.0318,
|
| 279 |
+
"step": 420
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 2.12,
|
| 283 |
+
"learning_rate": 5.878489326765189e-05,
|
| 284 |
+
"loss": 0.0312,
|
| 285 |
+
"step": 430
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 2.17,
|
| 289 |
+
"learning_rate": 5.55008210180624e-05,
|
| 290 |
+
"loss": 0.0284,
|
| 291 |
+
"step": 440
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.22,
|
| 295 |
+
"learning_rate": 5.2216748768472914e-05,
|
| 296 |
+
"loss": 0.0217,
|
| 297 |
+
"step": 450
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 2.27,
|
| 301 |
+
"learning_rate": 4.893267651888342e-05,
|
| 302 |
+
"loss": 0.014,
|
| 303 |
+
"step": 460
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 2.32,
|
| 307 |
+
"learning_rate": 4.5648604269293925e-05,
|
| 308 |
+
"loss": 0.02,
|
| 309 |
+
"step": 470
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 2.36,
|
| 313 |
+
"learning_rate": 4.236453201970443e-05,
|
| 314 |
+
"loss": 0.0148,
|
| 315 |
+
"step": 480
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 2.41,
|
| 319 |
+
"learning_rate": 3.908045977011495e-05,
|
| 320 |
+
"loss": 0.0136,
|
| 321 |
+
"step": 490
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 2.46,
|
| 325 |
+
"learning_rate": 3.5796387520525456e-05,
|
| 326 |
+
"loss": 0.0129,
|
| 327 |
+
"step": 500
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 2.51,
|
| 331 |
+
"learning_rate": 3.251231527093596e-05,
|
| 332 |
+
"loss": 0.0375,
|
| 333 |
+
"step": 510
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 2.56,
|
| 337 |
+
"learning_rate": 2.9228243021346467e-05,
|
| 338 |
+
"loss": 0.0124,
|
| 339 |
+
"step": 520
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.61,
|
| 343 |
+
"learning_rate": 2.5944170771756983e-05,
|
| 344 |
+
"loss": 0.0129,
|
| 345 |
+
"step": 530
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.66,
|
| 349 |
+
"learning_rate": 2.266009852216749e-05,
|
| 350 |
+
"loss": 0.0189,
|
| 351 |
+
"step": 540
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 2.71,
|
| 355 |
+
"learning_rate": 1.9376026272577998e-05,
|
| 356 |
+
"loss": 0.0142,
|
| 357 |
+
"step": 550
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 2.76,
|
| 361 |
+
"learning_rate": 1.6091954022988507e-05,
|
| 362 |
+
"loss": 0.0332,
|
| 363 |
+
"step": 560
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 2.81,
|
| 367 |
+
"learning_rate": 1.2807881773399016e-05,
|
| 368 |
+
"loss": 0.0116,
|
| 369 |
+
"step": 570
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 2.86,
|
| 373 |
+
"learning_rate": 9.523809523809523e-06,
|
| 374 |
+
"loss": 0.0311,
|
| 375 |
+
"step": 580
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 2.91,
|
| 379 |
+
"learning_rate": 6.239737274220033e-06,
|
| 380 |
+
"loss": 0.012,
|
| 381 |
+
"step": 590
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 2.96,
|
| 385 |
+
"learning_rate": 2.955665024630542e-06,
|
| 386 |
+
"loss": 0.0128,
|
| 387 |
+
"step": 600
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 3.0,
|
| 391 |
+
"eval_accuracy": 0.9944444444444445,
|
| 392 |
+
"eval_loss": 0.030118444934487343,
|
| 393 |
+
"eval_runtime": 15.4638,
|
| 394 |
+
"eval_samples_per_second": 46.56,
|
| 395 |
+
"eval_steps_per_second": 5.82,
|
| 396 |
+
"step": 609
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 3.0,
|
| 400 |
+
"step": 609,
|
| 401 |
+
"total_flos": 1.5064982852419584e+18,
|
| 402 |
+
"train_loss": 0.23008910012362627,
|
| 403 |
+
"train_runtime": 477.4194,
|
| 404 |
+
"train_samples_per_second": 40.719,
|
| 405 |
+
"train_steps_per_second": 1.276
|
| 406 |
+
}
|
| 407 |
+
],
|
| 408 |
+
"logging_steps": 10,
|
| 409 |
+
"max_steps": 609,
|
| 410 |
+
"num_train_epochs": 3,
|
| 411 |
+
"save_steps": 500,
|
| 412 |
+
"total_flos": 1.5064982852419584e+18,
|
| 413 |
+
"trial_name": null,
|
| 414 |
+
"trial_params": null
|
| 415 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cedec33f9dc5a2bd63a2ed4b6f28140757c38b6223ce3865056836798d9b5e88
|
| 3 |
+
size 4600
|