Upload folder using huggingface_hub
Browse files- checkpoint-24210/config.json +34 -0
- checkpoint-24210/model.safetensors +3 -0
- checkpoint-24210/optimizer.pt +3 -0
- checkpoint-24210/preprocessor_config.json +22 -0
- checkpoint-24210/rng_state.pth +3 -0
- checkpoint-24210/scheduler.pt +3 -0
- checkpoint-24210/trainer_state.json +389 -0
- checkpoint-24210/training_args.bin +3 -0
- model.safetensors +1 -1
- training_args.bin +1 -1
checkpoint-24210/config.json
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": "E",
|
| 13 |
+
"1": "S",
|
| 14 |
+
"2": "SB"
|
| 15 |
+
},
|
| 16 |
+
"image_size": 224,
|
| 17 |
+
"initializer_range": 0.02,
|
| 18 |
+
"intermediate_size": 3072,
|
| 19 |
+
"label2id": {
|
| 20 |
+
"E": 0,
|
| 21 |
+
"S": 1,
|
| 22 |
+
"SB": 2
|
| 23 |
+
},
|
| 24 |
+
"layer_norm_eps": 1e-12,
|
| 25 |
+
"model_type": "vit",
|
| 26 |
+
"num_attention_heads": 12,
|
| 27 |
+
"num_channels": 3,
|
| 28 |
+
"num_hidden_layers": 12,
|
| 29 |
+
"patch_size": 16,
|
| 30 |
+
"problem_type": "single_label_classification",
|
| 31 |
+
"qkv_bias": true,
|
| 32 |
+
"torch_dtype": "float32",
|
| 33 |
+
"transformers_version": "4.44.2"
|
| 34 |
+
}
|
checkpoint-24210/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:91001fd450cf249ae5c849d6bc453dc30f7e594fa3c979545b1a5b8f115a0ed1
|
| 3 |
+
size 343227052
|
checkpoint-24210/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f1ed6281857a911e48ae5de7855cc5131736648638d7ea9c6aee660ac97069b6
|
| 3 |
+
size 686574597
|
checkpoint-24210/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 |
+
}
|
checkpoint-24210/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3af11ab9a843e81e9d44e6005748a0cf3748150753f1cbd67e098fd2b0e3a04d
|
| 3 |
+
size 14575
|
checkpoint-24210/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ca28fab56fe5c8ab4d1509bb1513b3a4ca2f62bef704f72a600e08d39e183f04
|
| 3 |
+
size 627
|
checkpoint-24210/trainer_state.json
ADDED
|
@@ -0,0 +1,389 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.5134674906730652,
|
| 3 |
+
"best_model_checkpoint": "galaxy_type_image_detection/checkpoint-24210",
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 24210,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.041305245766212306,
|
| 13 |
+
"grad_norm": 2.052537679672241,
|
| 14 |
+
"learning_rate": 9.813741721854305e-07,
|
| 15 |
+
"loss": 0.9706,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.08261049153242461,
|
| 20 |
+
"grad_norm": 1.7964216470718384,
|
| 21 |
+
"learning_rate": 9.606788079470197e-07,
|
| 22 |
+
"loss": 0.805,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.12391573729863693,
|
| 27 |
+
"grad_norm": 2.35636043548584,
|
| 28 |
+
"learning_rate": 9.399834437086093e-07,
|
| 29 |
+
"loss": 0.7313,
|
| 30 |
+
"step": 1500
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.16522098306484923,
|
| 34 |
+
"grad_norm": 4.531991958618164,
|
| 35 |
+
"learning_rate": 9.192880794701986e-07,
|
| 36 |
+
"loss": 0.6929,
|
| 37 |
+
"step": 2000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.20652622883106153,
|
| 41 |
+
"grad_norm": 2.8686130046844482,
|
| 42 |
+
"learning_rate": 8.98592715231788e-07,
|
| 43 |
+
"loss": 0.6575,
|
| 44 |
+
"step": 2500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.24783147459727387,
|
| 48 |
+
"grad_norm": 3.747077703475952,
|
| 49 |
+
"learning_rate": 8.778973509933775e-07,
|
| 50 |
+
"loss": 0.6312,
|
| 51 |
+
"step": 3000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.28913672036348614,
|
| 55 |
+
"grad_norm": 3.5423054695129395,
|
| 56 |
+
"learning_rate": 8.572019867549668e-07,
|
| 57 |
+
"loss": 0.6236,
|
| 58 |
+
"step": 3500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.33044196612969845,
|
| 62 |
+
"grad_norm": 2.6661975383758545,
|
| 63 |
+
"learning_rate": 8.365066225165562e-07,
|
| 64 |
+
"loss": 0.6119,
|
| 65 |
+
"step": 4000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.37174721189591076,
|
| 69 |
+
"grad_norm": 2.6772751808166504,
|
| 70 |
+
"learning_rate": 8.158112582781457e-07,
|
| 71 |
+
"loss": 0.5979,
|
| 72 |
+
"step": 4500
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.41305245766212306,
|
| 76 |
+
"grad_norm": 4.741942405700684,
|
| 77 |
+
"learning_rate": 7.951158940397351e-07,
|
| 78 |
+
"loss": 0.5926,
|
| 79 |
+
"step": 5000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.4543577034283354,
|
| 83 |
+
"grad_norm": 4.627604007720947,
|
| 84 |
+
"learning_rate": 7.744205298013245e-07,
|
| 85 |
+
"loss": 0.5801,
|
| 86 |
+
"step": 5500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.49566294919454773,
|
| 90 |
+
"grad_norm": 3.0113418102264404,
|
| 91 |
+
"learning_rate": 7.537251655629138e-07,
|
| 92 |
+
"loss": 0.5736,
|
| 93 |
+
"step": 6000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.53696819496076,
|
| 97 |
+
"grad_norm": 4.815735340118408,
|
| 98 |
+
"learning_rate": 7.330298013245033e-07,
|
| 99 |
+
"loss": 0.5702,
|
| 100 |
+
"step": 6500
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.5782734407269723,
|
| 104 |
+
"grad_norm": 3.646869421005249,
|
| 105 |
+
"learning_rate": 7.123344370860927e-07,
|
| 106 |
+
"loss": 0.5651,
|
| 107 |
+
"step": 7000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.6195786864931846,
|
| 111 |
+
"grad_norm": 4.093480587005615,
|
| 112 |
+
"learning_rate": 6.916390728476821e-07,
|
| 113 |
+
"loss": 0.559,
|
| 114 |
+
"step": 7500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.6608839322593969,
|
| 118 |
+
"grad_norm": 3.088358163833618,
|
| 119 |
+
"learning_rate": 6.709437086092715e-07,
|
| 120 |
+
"loss": 0.561,
|
| 121 |
+
"step": 8000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.7021891780256092,
|
| 125 |
+
"grad_norm": 3.7480947971343994,
|
| 126 |
+
"learning_rate": 6.502483443708608e-07,
|
| 127 |
+
"loss": 0.5485,
|
| 128 |
+
"step": 8500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.7434944237918215,
|
| 132 |
+
"grad_norm": 3.2563865184783936,
|
| 133 |
+
"learning_rate": 6.295529801324503e-07,
|
| 134 |
+
"loss": 0.5479,
|
| 135 |
+
"step": 9000
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.7847996695580338,
|
| 139 |
+
"grad_norm": 3.8188157081604004,
|
| 140 |
+
"learning_rate": 6.088576158940398e-07,
|
| 141 |
+
"loss": 0.546,
|
| 142 |
+
"step": 9500
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.8261049153242461,
|
| 146 |
+
"grad_norm": 5.6608147621154785,
|
| 147 |
+
"learning_rate": 5.881622516556291e-07,
|
| 148 |
+
"loss": 0.5454,
|
| 149 |
+
"step": 10000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.8674101610904585,
|
| 153 |
+
"grad_norm": 3.919110059738159,
|
| 154 |
+
"learning_rate": 5.674668874172185e-07,
|
| 155 |
+
"loss": 0.5385,
|
| 156 |
+
"step": 10500
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.9087154068566708,
|
| 160 |
+
"grad_norm": 3.518577814102173,
|
| 161 |
+
"learning_rate": 5.46771523178808e-07,
|
| 162 |
+
"loss": 0.5407,
|
| 163 |
+
"step": 11000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.9500206526228832,
|
| 167 |
+
"grad_norm": 4.594565391540527,
|
| 168 |
+
"learning_rate": 5.260761589403974e-07,
|
| 169 |
+
"loss": 0.5378,
|
| 170 |
+
"step": 11500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.9913258983890955,
|
| 174 |
+
"grad_norm": 5.844511032104492,
|
| 175 |
+
"learning_rate": 5.053807947019867e-07,
|
| 176 |
+
"loss": 0.5347,
|
| 177 |
+
"step": 12000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.0,
|
| 181 |
+
"eval_accuracy": 0.774586746456075,
|
| 182 |
+
"eval_loss": 0.5334413647651672,
|
| 183 |
+
"eval_model_preparation_time": 0.0052,
|
| 184 |
+
"eval_runtime": 398.4848,
|
| 185 |
+
"eval_samples_per_second": 102.323,
|
| 186 |
+
"eval_steps_per_second": 3.2,
|
| 187 |
+
"step": 12105
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 1.0326311441553078,
|
| 191 |
+
"grad_norm": 4.9089179039001465,
|
| 192 |
+
"learning_rate": 4.846854304635761e-07,
|
| 193 |
+
"loss": 0.5274,
|
| 194 |
+
"step": 12500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 1.07393638992152,
|
| 198 |
+
"grad_norm": 5.5489325523376465,
|
| 199 |
+
"learning_rate": 4.6399006622516554e-07,
|
| 200 |
+
"loss": 0.5252,
|
| 201 |
+
"step": 13000
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 1.1152416356877324,
|
| 205 |
+
"grad_norm": 3.628629207611084,
|
| 206 |
+
"learning_rate": 4.4329470198675495e-07,
|
| 207 |
+
"loss": 0.5269,
|
| 208 |
+
"step": 13500
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 1.1565468814539446,
|
| 212 |
+
"grad_norm": 3.0863559246063232,
|
| 213 |
+
"learning_rate": 4.225993377483443e-07,
|
| 214 |
+
"loss": 0.5294,
|
| 215 |
+
"step": 14000
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 1.197852127220157,
|
| 219 |
+
"grad_norm": 3.804537773132324,
|
| 220 |
+
"learning_rate": 4.0190397350993377e-07,
|
| 221 |
+
"loss": 0.527,
|
| 222 |
+
"step": 14500
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 1.2391573729863692,
|
| 226 |
+
"grad_norm": 3.9841179847717285,
|
| 227 |
+
"learning_rate": 3.8120860927152313e-07,
|
| 228 |
+
"loss": 0.5199,
|
| 229 |
+
"step": 15000
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 1.2804626187525816,
|
| 233 |
+
"grad_norm": 6.212009906768799,
|
| 234 |
+
"learning_rate": 3.605132450331126e-07,
|
| 235 |
+
"loss": 0.5262,
|
| 236 |
+
"step": 15500
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.3217678645187938,
|
| 240 |
+
"grad_norm": 2.9575085639953613,
|
| 241 |
+
"learning_rate": 3.3981788079470195e-07,
|
| 242 |
+
"loss": 0.5271,
|
| 243 |
+
"step": 16000
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 1.3630731102850062,
|
| 247 |
+
"grad_norm": 6.535609722137451,
|
| 248 |
+
"learning_rate": 3.1912251655629136e-07,
|
| 249 |
+
"loss": 0.5128,
|
| 250 |
+
"step": 16500
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 1.4043783560512186,
|
| 254 |
+
"grad_norm": 3.2392661571502686,
|
| 255 |
+
"learning_rate": 2.9842715231788077e-07,
|
| 256 |
+
"loss": 0.5211,
|
| 257 |
+
"step": 17000
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 1.4456836018174308,
|
| 261 |
+
"grad_norm": 3.6159918308258057,
|
| 262 |
+
"learning_rate": 2.777317880794702e-07,
|
| 263 |
+
"loss": 0.5198,
|
| 264 |
+
"step": 17500
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 1.486988847583643,
|
| 268 |
+
"grad_norm": 4.245410919189453,
|
| 269 |
+
"learning_rate": 2.570364238410596e-07,
|
| 270 |
+
"loss": 0.5179,
|
| 271 |
+
"step": 18000
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 1.5282940933498554,
|
| 275 |
+
"grad_norm": 4.863912105560303,
|
| 276 |
+
"learning_rate": 2.36341059602649e-07,
|
| 277 |
+
"loss": 0.5189,
|
| 278 |
+
"step": 18500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 1.5695993391160679,
|
| 282 |
+
"grad_norm": 4.745334625244141,
|
| 283 |
+
"learning_rate": 2.156456953642384e-07,
|
| 284 |
+
"loss": 0.5141,
|
| 285 |
+
"step": 19000
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 1.61090458488228,
|
| 289 |
+
"grad_norm": 5.639151096343994,
|
| 290 |
+
"learning_rate": 1.9495033112582782e-07,
|
| 291 |
+
"loss": 0.5157,
|
| 292 |
+
"step": 19500
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 1.6522098306484923,
|
| 296 |
+
"grad_norm": 3.5343689918518066,
|
| 297 |
+
"learning_rate": 1.742549668874172e-07,
|
| 298 |
+
"loss": 0.5172,
|
| 299 |
+
"step": 20000
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 1.6935150764147047,
|
| 303 |
+
"grad_norm": 3.778491735458374,
|
| 304 |
+
"learning_rate": 1.5355960264900662e-07,
|
| 305 |
+
"loss": 0.5173,
|
| 306 |
+
"step": 20500
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 1.734820322180917,
|
| 310 |
+
"grad_norm": 3.17097806930542,
|
| 311 |
+
"learning_rate": 1.3286423841059603e-07,
|
| 312 |
+
"loss": 0.5119,
|
| 313 |
+
"step": 21000
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 1.7761255679471293,
|
| 317 |
+
"grad_norm": 4.683322906494141,
|
| 318 |
+
"learning_rate": 1.1216887417218542e-07,
|
| 319 |
+
"loss": 0.5172,
|
| 320 |
+
"step": 21500
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.8174308137133415,
|
| 324 |
+
"grad_norm": 5.71618127822876,
|
| 325 |
+
"learning_rate": 9.147350993377482e-08,
|
| 326 |
+
"loss": 0.5105,
|
| 327 |
+
"step": 22000
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 1.858736059479554,
|
| 331 |
+
"grad_norm": 4.671562671661377,
|
| 332 |
+
"learning_rate": 7.077814569536423e-08,
|
| 333 |
+
"loss": 0.5154,
|
| 334 |
+
"step": 22500
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 1.9000413052457663,
|
| 338 |
+
"grad_norm": 4.385097503662109,
|
| 339 |
+
"learning_rate": 5.008278145695364e-08,
|
| 340 |
+
"loss": 0.5126,
|
| 341 |
+
"step": 23000
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 1.9413465510119785,
|
| 345 |
+
"grad_norm": 3.7954232692718506,
|
| 346 |
+
"learning_rate": 2.9387417218543044e-08,
|
| 347 |
+
"loss": 0.511,
|
| 348 |
+
"step": 23500
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 1.9826517967781907,
|
| 352 |
+
"grad_norm": 5.682091236114502,
|
| 353 |
+
"learning_rate": 8.69205298013245e-09,
|
| 354 |
+
"loss": 0.5168,
|
| 355 |
+
"step": 24000
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 2.0,
|
| 359 |
+
"eval_accuracy": 0.7810859861676559,
|
| 360 |
+
"eval_loss": 0.5134674906730652,
|
| 361 |
+
"eval_model_preparation_time": 0.0052,
|
| 362 |
+
"eval_runtime": 395.5533,
|
| 363 |
+
"eval_samples_per_second": 103.081,
|
| 364 |
+
"eval_steps_per_second": 3.223,
|
| 365 |
+
"step": 24210
|
| 366 |
+
}
|
| 367 |
+
],
|
| 368 |
+
"logging_steps": 500,
|
| 369 |
+
"max_steps": 24210,
|
| 370 |
+
"num_input_tokens_seen": 0,
|
| 371 |
+
"num_train_epochs": 2,
|
| 372 |
+
"save_steps": 500,
|
| 373 |
+
"stateful_callbacks": {
|
| 374 |
+
"TrainerControl": {
|
| 375 |
+
"args": {
|
| 376 |
+
"should_epoch_stop": false,
|
| 377 |
+
"should_evaluate": false,
|
| 378 |
+
"should_log": false,
|
| 379 |
+
"should_save": true,
|
| 380 |
+
"should_training_stop": true
|
| 381 |
+
},
|
| 382 |
+
"attributes": {}
|
| 383 |
+
}
|
| 384 |
+
},
|
| 385 |
+
"total_flos": 1.2006545998417342e+20,
|
| 386 |
+
"train_batch_size": 64,
|
| 387 |
+
"trial_name": null,
|
| 388 |
+
"trial_params": null
|
| 389 |
+
}
|
checkpoint-24210/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:200be95460fb1d796a36a6316c5c1e2982d1d3cc913cf467667e59fa3ddc9cbe
|
| 3 |
+
size 4667
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 343227052
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:91001fd450cf249ae5c849d6bc453dc30f7e594fa3c979545b1a5b8f115a0ed1
|
| 3 |
size 343227052
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4667
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:200be95460fb1d796a36a6316c5c1e2982d1d3cc913cf467667e59fa3ddc9cbe
|
| 3 |
size 4667
|