End of training
Browse files- README.md +7 -45
- all_results.json +6 -7
- config.json +6 -13
- model.safetensors +2 -2
- optimizer.pt +2 -2
- rng_state.pth +2 -2
- scheduler.pt +2 -2
- test_results.json +6 -6
- trainer_state.json +305 -643
- training_args.bin +2 -2
README.md
CHANGED
|
@@ -1,11 +1,7 @@
|
|
| 1 |
---
|
| 2 |
library_name: transformers
|
| 3 |
-
license: cc-by-nc-4.0
|
| 4 |
-
base_model: MCG-NJU/videomae-base
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
| 7 |
-
metrics:
|
| 8 |
-
- accuracy
|
| 9 |
model-index:
|
| 10 |
- name: videomae-base-finetuned-dogBehavior
|
| 11 |
results: []
|
|
@@ -16,10 +12,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
# videomae-base-finetuned-dogBehavior
|
| 18 |
|
| 19 |
-
This model
|
| 20 |
-
It achieves the following results on the evaluation set:
|
| 21 |
-
- Loss: 0.7947
|
| 22 |
-
- Accuracy: 0.8387
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
@@ -42,45 +35,14 @@ The following hyperparameters were used during training:
|
|
| 42 |
- train_batch_size: 8
|
| 43 |
- eval_batch_size: 8
|
| 44 |
- seed: 42
|
| 45 |
-
- optimizer:
|
| 46 |
- lr_scheduler_type: linear
|
| 47 |
- lr_scheduler_warmup_ratio: 0.1
|
| 48 |
-
- training_steps:
|
| 49 |
-
|
| 50 |
-
### Training results
|
| 51 |
-
|
| 52 |
-
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 53 |
-
|:-------------:|:-------:|:----:|:---------------:|:--------:|
|
| 54 |
-
| 1.5395 | 0.0413 | 31 | 1.5124 | 0.3226 |
|
| 55 |
-
| 1.4347 | 1.0413 | 62 | 1.4659 | 0.3226 |
|
| 56 |
-
| 1.3176 | 2.0413 | 93 | 1.4264 | 0.4839 |
|
| 57 |
-
| 1.2765 | 3.0413 | 124 | 0.9989 | 0.7097 |
|
| 58 |
-
| 1.1538 | 4.0413 | 155 | 1.0743 | 0.6129 |
|
| 59 |
-
| 1.0894 | 5.0413 | 186 | 1.0309 | 0.6452 |
|
| 60 |
-
| 0.705 | 6.0413 | 217 | 0.8872 | 0.7097 |
|
| 61 |
-
| 0.8064 | 7.0413 | 248 | 1.0081 | 0.7097 |
|
| 62 |
-
| 0.6057 | 8.0413 | 279 | 0.8231 | 0.7419 |
|
| 63 |
-
| 0.493 | 9.0413 | 310 | 0.6791 | 0.7742 |
|
| 64 |
-
| 0.6431 | 10.0413 | 341 | 0.7946 | 0.6129 |
|
| 65 |
-
| 0.397 | 11.0413 | 372 | 0.8140 | 0.7742 |
|
| 66 |
-
| 0.405 | 12.0413 | 403 | 0.8367 | 0.7742 |
|
| 67 |
-
| 0.5463 | 13.0413 | 434 | 0.6394 | 0.7419 |
|
| 68 |
-
| 0.3068 | 14.0413 | 465 | 0.8780 | 0.7419 |
|
| 69 |
-
| 0.1043 | 15.0413 | 496 | 0.7384 | 0.7419 |
|
| 70 |
-
| 0.3512 | 16.0413 | 527 | 0.9134 | 0.7419 |
|
| 71 |
-
| 0.1996 | 17.0413 | 558 | 0.9459 | 0.7742 |
|
| 72 |
-
| 0.2886 | 18.0413 | 589 | 0.7947 | 0.8387 |
|
| 73 |
-
| 0.257 | 19.0413 | 620 | 0.9584 | 0.8065 |
|
| 74 |
-
| 0.3948 | 20.0413 | 651 | 0.9726 | 0.7742 |
|
| 75 |
-
| 0.1969 | 21.0413 | 682 | 0.8744 | 0.8065 |
|
| 76 |
-
| 0.1238 | 22.0413 | 713 | 0.8290 | 0.8065 |
|
| 77 |
-
| 0.1539 | 23.0413 | 744 | 0.8723 | 0.8065 |
|
| 78 |
-
| 0.2076 | 24.008 | 750 | 0.8719 | 0.8065 |
|
| 79 |
-
|
| 80 |
|
| 81 |
### Framework versions
|
| 82 |
|
| 83 |
-
- Transformers 4.
|
| 84 |
-
- Pytorch 2.
|
| 85 |
-
- Datasets
|
| 86 |
-
- Tokenizers 0.
|
|
|
|
| 1 |
---
|
| 2 |
library_name: transformers
|
|
|
|
|
|
|
| 3 |
tags:
|
| 4 |
- generated_from_trainer
|
|
|
|
|
|
|
| 5 |
model-index:
|
| 6 |
- name: videomae-base-finetuned-dogBehavior
|
| 7 |
results: []
|
|
|
|
| 12 |
|
| 13 |
# videomae-base-finetuned-dogBehavior
|
| 14 |
|
| 15 |
+
This model was trained from scratch on an unknown dataset.
|
|
|
|
|
|
|
|
|
|
| 16 |
|
| 17 |
## Model description
|
| 18 |
|
|
|
|
| 35 |
- train_batch_size: 8
|
| 36 |
- eval_batch_size: 8
|
| 37 |
- seed: 42
|
| 38 |
+
- optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 39 |
- lr_scheduler_type: linear
|
| 40 |
- lr_scheduler_warmup_ratio: 0.1
|
| 41 |
+
- training_steps: 125
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 42 |
|
| 43 |
### Framework versions
|
| 44 |
|
| 45 |
+
- Transformers 4.57.3
|
| 46 |
+
- Pytorch 2.9.1+cu130
|
| 47 |
+
- Datasets 4.4.2
|
| 48 |
+
- Tokenizers 0.22.2
|
all_results.json
CHANGED
|
@@ -1,9 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
| 3 |
-
"
|
| 4 |
-
"
|
| 5 |
-
"
|
| 6 |
-
"
|
| 7 |
-
"
|
| 8 |
-
"eval_steps_per_second": 0.075
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"eval_accuracy": 0.9393939393939394,
|
| 3 |
+
"eval_loss": 0.1900058090686798,
|
| 4 |
+
"eval_model_preparation_time": 0.0031,
|
| 5 |
+
"eval_runtime": 58.6079,
|
| 6 |
+
"eval_samples_per_second": 0.563,
|
| 7 |
+
"eval_steps_per_second": 0.085
|
|
|
|
| 8 |
}
|
config.json
CHANGED
|
@@ -1,5 +1,4 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "MCG-NJU/videomae-base",
|
| 3 |
"architectures": [
|
| 4 |
"VideoMAEForVideoClassification"
|
| 5 |
],
|
|
@@ -8,25 +7,20 @@
|
|
| 8 |
"decoder_intermediate_size": 1536,
|
| 9 |
"decoder_num_attention_heads": 6,
|
| 10 |
"decoder_num_hidden_layers": 4,
|
|
|
|
| 11 |
"hidden_act": "gelu",
|
| 12 |
"hidden_dropout_prob": 0.0,
|
| 13 |
"hidden_size": 768,
|
| 14 |
"id2label": {
|
| 15 |
-
"0": "
|
| 16 |
-
"1": "
|
| 17 |
-
"2": "scratch04_10",
|
| 18 |
-
"3": "specialsniff13_10",
|
| 19 |
-
"4": "turn04_10"
|
| 20 |
},
|
| 21 |
"image_size": 224,
|
| 22 |
"initializer_range": 0.02,
|
| 23 |
"intermediate_size": 3072,
|
| 24 |
"label2id": {
|
| 25 |
-
"
|
| 26 |
-
"
|
| 27 |
-
"scratch04_10": 2,
|
| 28 |
-
"specialsniff13_10": 3,
|
| 29 |
-
"turn04_10": 4
|
| 30 |
},
|
| 31 |
"layer_norm_eps": 1e-12,
|
| 32 |
"model_type": "videomae",
|
|
@@ -38,8 +32,7 @@
|
|
| 38 |
"patch_size": 16,
|
| 39 |
"problem_type": "single_label_classification",
|
| 40 |
"qkv_bias": true,
|
| 41 |
-
"
|
| 42 |
-
"transformers_version": "4.44.1",
|
| 43 |
"tubelet_size": 2,
|
| 44 |
"use_mean_pooling": false
|
| 45 |
}
|
|
|
|
| 1 |
{
|
|
|
|
| 2 |
"architectures": [
|
| 3 |
"VideoMAEForVideoClassification"
|
| 4 |
],
|
|
|
|
| 7 |
"decoder_intermediate_size": 1536,
|
| 8 |
"decoder_num_attention_heads": 6,
|
| 9 |
"decoder_num_hidden_layers": 4,
|
| 10 |
+
"dtype": "float32",
|
| 11 |
"hidden_act": "gelu",
|
| 12 |
"hidden_dropout_prob": 0.0,
|
| 13 |
"hidden_size": 768,
|
| 14 |
"id2label": {
|
| 15 |
+
"0": "justsniff",
|
| 16 |
+
"1": "lying"
|
|
|
|
|
|
|
|
|
|
| 17 |
},
|
| 18 |
"image_size": 224,
|
| 19 |
"initializer_range": 0.02,
|
| 20 |
"intermediate_size": 3072,
|
| 21 |
"label2id": {
|
| 22 |
+
"justsniff": 0,
|
| 23 |
+
"lying": 1
|
|
|
|
|
|
|
|
|
|
| 24 |
},
|
| 25 |
"layer_norm_eps": 1e-12,
|
| 26 |
"model_type": "videomae",
|
|
|
|
| 32 |
"patch_size": 16,
|
| 33 |
"problem_type": "single_label_classification",
|
| 34 |
"qkv_bias": true,
|
| 35 |
+
"transformers_version": "4.57.3",
|
|
|
|
| 36 |
"tubelet_size": 2,
|
| 37 |
"use_mean_pooling": false
|
| 38 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:94704761688e0370a2ad99340d995c6285595680d147cfce675b800a1f40cb95
|
| 3 |
+
size 344937368
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1b6c4bf485686dc2460f20adf662bf6ef21a6f5f482367135c6e12dae3122b67
|
| 3 |
+
size 689988619
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a33122a368dd627dd121c5da60672d47297ea002d4ed2ce3715b215c4f2e6ba4
|
| 3 |
+
size 14645
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:215f49c99657537ce1b0b1d277562ab616c1c5f0831e119eb6f1022948171c72
|
| 3 |
+
size 1465
|
test_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
| 3 |
-
"
|
| 4 |
-
"
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second": 0.
|
| 7 |
-
"eval_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"eval_accuracy": 0.9393939393939394,
|
| 3 |
+
"eval_loss": 0.1900058090686798,
|
| 4 |
+
"eval_model_preparation_time": 0.0031,
|
| 5 |
+
"eval_runtime": 58.6079,
|
| 6 |
+
"eval_samples_per_second": 0.563,
|
| 7 |
+
"eval_steps_per_second": 0.085
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,802 +1,464 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
| 3 |
-
"
|
| 4 |
-
"
|
|
|
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 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.
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss":
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate": 1.
|
| 22 |
-
"loss":
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.04,
|
| 27 |
-
"
|
| 28 |
-
"
|
| 29 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch":
|
| 34 |
-
"
|
| 35 |
-
"
|
| 36 |
-
"
|
| 37 |
-
"
|
| 38 |
-
"eval_steps_per_second": 0.082,
|
| 39 |
-
"step": 31
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 1.
|
| 43 |
-
"grad_norm":
|
| 44 |
-
"learning_rate":
|
| 45 |
-
"loss":
|
| 46 |
-
"step":
|
| 47 |
},
|
| 48 |
{
|
| 49 |
-
"epoch": 1.
|
| 50 |
-
"
|
| 51 |
-
"
|
| 52 |
-
"
|
|
|
|
|
|
|
| 53 |
"step": 50
|
| 54 |
},
|
| 55 |
{
|
| 56 |
-
"epoch":
|
| 57 |
-
"grad_norm":
|
| 58 |
-
"learning_rate":
|
| 59 |
-
"loss":
|
| 60 |
"step": 60
|
| 61 |
},
|
| 62 |
{
|
| 63 |
-
"epoch":
|
| 64 |
-
"
|
| 65 |
-
"
|
| 66 |
-
"
|
| 67 |
-
"
|
| 68 |
-
"eval_steps_per_second": 0.082,
|
| 69 |
-
"step": 62
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch": 2.
|
| 73 |
-
"
|
| 74 |
-
"
|
| 75 |
-
"
|
| 76 |
-
"
|
|
|
|
|
|
|
| 77 |
},
|
| 78 |
{
|
| 79 |
-
"epoch":
|
| 80 |
-
"grad_norm":
|
| 81 |
-
"learning_rate": 4.
|
| 82 |
-
"loss":
|
| 83 |
"step": 80
|
| 84 |
},
|
| 85 |
{
|
| 86 |
-
"epoch":
|
| 87 |
-
"grad_norm":
|
| 88 |
-
"learning_rate": 4.
|
| 89 |
-
"loss":
|
| 90 |
"step": 90
|
| 91 |
},
|
| 92 |
{
|
| 93 |
-
"epoch":
|
| 94 |
-
"
|
| 95 |
-
"
|
| 96 |
-
"
|
| 97 |
-
"
|
| 98 |
-
"eval_steps_per_second": 0.082,
|
| 99 |
-
"step": 93
|
| 100 |
},
|
| 101 |
{
|
| 102 |
-
"epoch": 3.
|
| 103 |
-
"
|
| 104 |
-
"
|
| 105 |
-
"
|
|
|
|
|
|
|
| 106 |
"step": 100
|
| 107 |
},
|
| 108 |
{
|
| 109 |
-
"epoch":
|
| 110 |
-
"grad_norm":
|
| 111 |
-
"learning_rate": 4.
|
| 112 |
-
"loss":
|
| 113 |
"step": 110
|
| 114 |
},
|
| 115 |
{
|
| 116 |
-
"epoch":
|
| 117 |
-
"grad_norm":
|
| 118 |
-
"learning_rate": 4.
|
| 119 |
-
"loss":
|
| 120 |
"step": 120
|
| 121 |
},
|
| 122 |
{
|
| 123 |
-
"epoch":
|
| 124 |
-
"eval_accuracy":
|
| 125 |
-
"eval_loss": 0.
|
| 126 |
-
"eval_runtime":
|
| 127 |
-
"eval_samples_per_second": 0.
|
| 128 |
-
"eval_steps_per_second": 0.
|
| 129 |
-
"step":
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch":
|
| 133 |
-
"grad_norm":
|
| 134 |
-
"learning_rate": 4.
|
| 135 |
-
"loss": 0.
|
| 136 |
"step": 130
|
| 137 |
},
|
| 138 |
{
|
| 139 |
-
"epoch":
|
| 140 |
-
"grad_norm":
|
| 141 |
-
"learning_rate": 4.
|
| 142 |
-
"loss": 0.
|
| 143 |
"step": 140
|
| 144 |
},
|
| 145 |
{
|
| 146 |
-
"epoch":
|
| 147 |
-
"grad_norm":
|
| 148 |
-
"learning_rate": 4.
|
| 149 |
-
"loss":
|
| 150 |
"step": 150
|
| 151 |
},
|
| 152 |
{
|
| 153 |
-
"epoch":
|
| 154 |
-
"eval_accuracy": 0.
|
| 155 |
-
"eval_loss":
|
| 156 |
-
"eval_runtime":
|
| 157 |
-
"eval_samples_per_second": 0.
|
| 158 |
-
"eval_steps_per_second": 0.
|
| 159 |
-
"step":
|
| 160 |
},
|
| 161 |
{
|
| 162 |
-
"epoch":
|
| 163 |
-
"grad_norm":
|
| 164 |
-
"learning_rate": 4.
|
| 165 |
-
"loss": 0.
|
| 166 |
"step": 160
|
| 167 |
},
|
| 168 |
{
|
| 169 |
-
"epoch":
|
| 170 |
-
"grad_norm":
|
| 171 |
-
"learning_rate": 4.
|
| 172 |
-
"loss":
|
| 173 |
"step": 170
|
| 174 |
},
|
| 175 |
{
|
| 176 |
-
"epoch":
|
| 177 |
-
"
|
| 178 |
-
"
|
| 179 |
-
"
|
| 180 |
-
"
|
|
|
|
|
|
|
| 181 |
},
|
| 182 |
{
|
| 183 |
-
"epoch":
|
| 184 |
-
"
|
| 185 |
-
"
|
| 186 |
-
"
|
| 187 |
-
"
|
| 188 |
-
"eval_steps_per_second": 0.081,
|
| 189 |
-
"step": 186
|
| 190 |
},
|
| 191 |
{
|
| 192 |
-
"epoch":
|
| 193 |
-
"grad_norm":
|
| 194 |
-
"learning_rate":
|
| 195 |
-
"loss":
|
| 196 |
"step": 190
|
| 197 |
},
|
| 198 |
{
|
| 199 |
-
"epoch":
|
| 200 |
-
"grad_norm":
|
| 201 |
-
"learning_rate":
|
| 202 |
-
"loss":
|
| 203 |
"step": 200
|
| 204 |
},
|
| 205 |
{
|
| 206 |
-
"epoch":
|
| 207 |
-
"
|
| 208 |
-
"
|
| 209 |
-
"
|
| 210 |
-
"
|
|
|
|
|
|
|
| 211 |
},
|
| 212 |
{
|
| 213 |
-
"epoch":
|
| 214 |
-
"
|
| 215 |
-
"
|
| 216 |
-
"
|
| 217 |
-
"
|
| 218 |
-
"eval_steps_per_second": 0.083,
|
| 219 |
-
"step": 217
|
| 220 |
},
|
| 221 |
{
|
| 222 |
-
"epoch":
|
| 223 |
-
"grad_norm":
|
| 224 |
-
"learning_rate": 3.
|
| 225 |
-
"loss": 0.
|
| 226 |
"step": 220
|
| 227 |
},
|
| 228 |
{
|
| 229 |
-
"epoch":
|
| 230 |
-
"
|
| 231 |
-
"
|
| 232 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 233 |
"step": 230
|
| 234 |
},
|
| 235 |
{
|
| 236 |
-
"epoch":
|
| 237 |
-
"grad_norm":
|
| 238 |
-
"learning_rate": 3.
|
| 239 |
-
"loss": 0.
|
| 240 |
"step": 240
|
| 241 |
},
|
| 242 |
{
|
| 243 |
-
"epoch":
|
| 244 |
-
"
|
| 245 |
-
"
|
| 246 |
-
"
|
| 247 |
-
"
|
| 248 |
-
"eval_steps_per_second": 0.083,
|
| 249 |
-
"step": 248
|
| 250 |
},
|
| 251 |
{
|
| 252 |
-
"epoch":
|
| 253 |
-
"
|
| 254 |
-
"
|
| 255 |
-
"
|
|
|
|
|
|
|
| 256 |
"step": 250
|
| 257 |
},
|
| 258 |
{
|
| 259 |
-
"epoch":
|
| 260 |
-
"grad_norm":
|
| 261 |
-
"learning_rate": 3.
|
| 262 |
-
"loss": 0.
|
| 263 |
"step": 260
|
| 264 |
},
|
| 265 |
{
|
| 266 |
-
"epoch":
|
| 267 |
-
"grad_norm":
|
| 268 |
-
"learning_rate": 3.
|
| 269 |
-
"loss": 0.
|
| 270 |
"step": 270
|
| 271 |
},
|
| 272 |
{
|
| 273 |
-
"epoch":
|
| 274 |
-
"eval_accuracy":
|
| 275 |
-
"eval_loss": 0.
|
| 276 |
-
"eval_runtime":
|
| 277 |
-
"eval_samples_per_second": 0.
|
| 278 |
-
"eval_steps_per_second": 0.
|
| 279 |
-
"step":
|
| 280 |
-
},
|
| 281 |
-
{
|
| 282 |
-
"epoch":
|
| 283 |
-
"grad_norm":
|
| 284 |
-
"learning_rate": 3.
|
| 285 |
-
"loss": 0.
|
| 286 |
"step": 280
|
| 287 |
},
|
| 288 |
{
|
| 289 |
-
"epoch":
|
| 290 |
-
"grad_norm":
|
| 291 |
-
"learning_rate":
|
| 292 |
-
"loss": 0.
|
| 293 |
"step": 290
|
| 294 |
},
|
| 295 |
{
|
| 296 |
-
"epoch":
|
| 297 |
-
"grad_norm":
|
| 298 |
-
"learning_rate":
|
| 299 |
-
"loss": 0.
|
| 300 |
"step": 300
|
| 301 |
},
|
| 302 |
{
|
| 303 |
-
"epoch":
|
| 304 |
-
"
|
| 305 |
-
"
|
| 306 |
-
"
|
| 307 |
-
"
|
|
|
|
|
|
|
| 308 |
},
|
| 309 |
{
|
| 310 |
-
"epoch":
|
| 311 |
-
"
|
| 312 |
-
"
|
| 313 |
-
"
|
| 314 |
-
"eval_samples_per_second": 0.644,
|
| 315 |
-
"eval_steps_per_second": 0.083,
|
| 316 |
"step": 310
|
| 317 |
},
|
| 318 |
{
|
| 319 |
-
"epoch":
|
| 320 |
-
"grad_norm":
|
| 321 |
-
"learning_rate":
|
| 322 |
-
"loss": 0.
|
| 323 |
"step": 320
|
| 324 |
},
|
| 325 |
{
|
| 326 |
-
"epoch":
|
| 327 |
-
"
|
| 328 |
-
"
|
| 329 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 330 |
"step": 330
|
| 331 |
},
|
| 332 |
{
|
| 333 |
-
"epoch":
|
| 334 |
-
"grad_norm":
|
| 335 |
-
"learning_rate":
|
| 336 |
-
"loss": 0.
|
| 337 |
"step": 340
|
| 338 |
},
|
| 339 |
{
|
| 340 |
-
"epoch":
|
| 341 |
-
"
|
| 342 |
-
"
|
| 343 |
-
"
|
| 344 |
-
"
|
| 345 |
-
"eval_steps_per_second": 0.042,
|
| 346 |
-
"step": 341
|
| 347 |
},
|
| 348 |
{
|
| 349 |
-
"epoch":
|
| 350 |
-
"
|
| 351 |
-
"
|
| 352 |
-
"
|
|
|
|
|
|
|
| 353 |
"step": 350
|
| 354 |
},
|
| 355 |
{
|
| 356 |
-
"epoch":
|
| 357 |
-
"grad_norm":
|
| 358 |
-
"learning_rate": 2.
|
| 359 |
-
"loss": 0.
|
| 360 |
"step": 360
|
| 361 |
},
|
| 362 |
{
|
| 363 |
-
"epoch":
|
| 364 |
-
"grad_norm":
|
| 365 |
-
"learning_rate": 2.
|
| 366 |
-
"loss": 0.
|
| 367 |
"step": 370
|
| 368 |
},
|
| 369 |
{
|
| 370 |
-
"epoch":
|
| 371 |
-
"eval_accuracy":
|
| 372 |
-
"eval_loss": 0.
|
| 373 |
-
"eval_runtime":
|
| 374 |
-
"eval_samples_per_second": 0.
|
| 375 |
-
"eval_steps_per_second": 0.
|
| 376 |
-
"step":
|
| 377 |
},
|
| 378 |
{
|
| 379 |
-
"epoch":
|
| 380 |
-
"grad_norm":
|
| 381 |
-
"learning_rate": 2.
|
| 382 |
-
"loss": 0.
|
| 383 |
"step": 380
|
| 384 |
},
|
| 385 |
{
|
| 386 |
-
"epoch":
|
| 387 |
-
"grad_norm":
|
| 388 |
-
"learning_rate": 2.
|
| 389 |
-
"loss": 0.
|
| 390 |
"step": 390
|
| 391 |
},
|
| 392 |
{
|
| 393 |
-
"epoch":
|
| 394 |
-
"grad_norm":
|
| 395 |
-
"learning_rate": 2.
|
| 396 |
-
"loss": 0.
|
| 397 |
"step": 400
|
| 398 |
},
|
| 399 |
{
|
| 400 |
-
"epoch":
|
| 401 |
-
"eval_accuracy":
|
| 402 |
-
"eval_loss": 0.
|
| 403 |
-
"eval_runtime":
|
| 404 |
-
"eval_samples_per_second": 0.
|
| 405 |
-
"eval_steps_per_second": 0.
|
| 406 |
-
"step":
|
| 407 |
},
|
| 408 |
{
|
| 409 |
-
"epoch":
|
| 410 |
-
"grad_norm":
|
| 411 |
-
"learning_rate":
|
| 412 |
-
"loss": 0.
|
| 413 |
"step": 410
|
| 414 |
},
|
| 415 |
-
{
|
| 416 |
-
"epoch": 13.022666666666666,
|
| 417 |
-
"grad_norm": 2.9040560722351074,
|
| 418 |
-
"learning_rate": 2.4444444444444445e-05,
|
| 419 |
-
"loss": 0.3197,
|
| 420 |
-
"step": 420
|
| 421 |
-
},
|
| 422 |
-
{
|
| 423 |
-
"epoch": 13.036,
|
| 424 |
-
"grad_norm": 6.378393650054932,
|
| 425 |
-
"learning_rate": 2.3703703703703707e-05,
|
| 426 |
-
"loss": 0.5463,
|
| 427 |
-
"step": 430
|
| 428 |
-
},
|
| 429 |
-
{
|
| 430 |
-
"epoch": 13.041333333333334,
|
| 431 |
-
"eval_accuracy": 0.7419354838709677,
|
| 432 |
-
"eval_loss": 0.6393604278564453,
|
| 433 |
-
"eval_runtime": 96.7879,
|
| 434 |
-
"eval_samples_per_second": 0.32,
|
| 435 |
-
"eval_steps_per_second": 0.041,
|
| 436 |
-
"step": 434
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 14.008,
|
| 440 |
-
"grad_norm": 6.623013496398926,
|
| 441 |
-
"learning_rate": 2.2962962962962965e-05,
|
| 442 |
-
"loss": 0.4882,
|
| 443 |
-
"step": 440
|
| 444 |
-
},
|
| 445 |
-
{
|
| 446 |
-
"epoch": 14.021333333333333,
|
| 447 |
-
"grad_norm": 10.516149520874023,
|
| 448 |
-
"learning_rate": 2.2222222222222223e-05,
|
| 449 |
-
"loss": 0.3528,
|
| 450 |
-
"step": 450
|
| 451 |
-
},
|
| 452 |
-
{
|
| 453 |
-
"epoch": 14.034666666666666,
|
| 454 |
-
"grad_norm": 0.7739436626434326,
|
| 455 |
-
"learning_rate": 2.148148148148148e-05,
|
| 456 |
-
"loss": 0.3068,
|
| 457 |
-
"step": 460
|
| 458 |
-
},
|
| 459 |
-
{
|
| 460 |
-
"epoch": 14.041333333333334,
|
| 461 |
-
"eval_accuracy": 0.7419354838709677,
|
| 462 |
-
"eval_loss": 0.8780257701873779,
|
| 463 |
-
"eval_runtime": 94.12,
|
| 464 |
-
"eval_samples_per_second": 0.329,
|
| 465 |
-
"eval_steps_per_second": 0.042,
|
| 466 |
-
"step": 465
|
| 467 |
-
},
|
| 468 |
-
{
|
| 469 |
-
"epoch": 15.006666666666666,
|
| 470 |
-
"grad_norm": 13.473437309265137,
|
| 471 |
-
"learning_rate": 2.074074074074074e-05,
|
| 472 |
-
"loss": 0.4486,
|
| 473 |
-
"step": 470
|
| 474 |
-
},
|
| 475 |
-
{
|
| 476 |
-
"epoch": 15.02,
|
| 477 |
-
"grad_norm": 11.189105987548828,
|
| 478 |
-
"learning_rate": 2e-05,
|
| 479 |
-
"loss": 0.2921,
|
| 480 |
-
"step": 480
|
| 481 |
-
},
|
| 482 |
-
{
|
| 483 |
-
"epoch": 15.033333333333333,
|
| 484 |
-
"grad_norm": 5.033076763153076,
|
| 485 |
-
"learning_rate": 1.925925925925926e-05,
|
| 486 |
-
"loss": 0.1043,
|
| 487 |
-
"step": 490
|
| 488 |
-
},
|
| 489 |
-
{
|
| 490 |
-
"epoch": 15.041333333333334,
|
| 491 |
-
"eval_accuracy": 0.7419354838709677,
|
| 492 |
-
"eval_loss": 0.7384097576141357,
|
| 493 |
-
"eval_runtime": 98.6686,
|
| 494 |
-
"eval_samples_per_second": 0.314,
|
| 495 |
-
"eval_steps_per_second": 0.041,
|
| 496 |
-
"step": 496
|
| 497 |
-
},
|
| 498 |
-
{
|
| 499 |
-
"epoch": 16.005333333333333,
|
| 500 |
-
"grad_norm": 5.1849589347839355,
|
| 501 |
-
"learning_rate": 1.8518518518518518e-05,
|
| 502 |
-
"loss": 0.1256,
|
| 503 |
-
"step": 500
|
| 504 |
-
},
|
| 505 |
-
{
|
| 506 |
-
"epoch": 16.018666666666668,
|
| 507 |
-
"grad_norm": 15.674521446228027,
|
| 508 |
-
"learning_rate": 1.777777777777778e-05,
|
| 509 |
-
"loss": 0.2178,
|
| 510 |
-
"step": 510
|
| 511 |
-
},
|
| 512 |
{
|
| 513 |
"epoch": 16.032,
|
| 514 |
-
"grad_norm":
|
| 515 |
-
"learning_rate": 1.
|
| 516 |
-
"loss": 0.
|
| 517 |
-
"step":
|
| 518 |
-
},
|
| 519 |
-
{
|
| 520 |
-
"epoch": 16.041333333333334,
|
| 521 |
-
"eval_accuracy": 0.7419354838709677,
|
| 522 |
-
"eval_loss": 0.9134318232536316,
|
| 523 |
-
"eval_runtime": 105.2236,
|
| 524 |
-
"eval_samples_per_second": 0.295,
|
| 525 |
-
"eval_steps_per_second": 0.038,
|
| 526 |
-
"step": 527
|
| 527 |
-
},
|
| 528 |
-
{
|
| 529 |
-
"epoch": 17.004,
|
| 530 |
-
"grad_norm": 1.709756851196289,
|
| 531 |
-
"learning_rate": 1.62962962962963e-05,
|
| 532 |
-
"loss": 0.1933,
|
| 533 |
-
"step": 530
|
| 534 |
-
},
|
| 535 |
-
{
|
| 536 |
-
"epoch": 17.017333333333333,
|
| 537 |
-
"grad_norm": 4.699275970458984,
|
| 538 |
-
"learning_rate": 1.5555555555555555e-05,
|
| 539 |
-
"loss": 0.2147,
|
| 540 |
-
"step": 540
|
| 541 |
-
},
|
| 542 |
-
{
|
| 543 |
-
"epoch": 17.030666666666665,
|
| 544 |
-
"grad_norm": 19.59837532043457,
|
| 545 |
-
"learning_rate": 1.4814814814814815e-05,
|
| 546 |
-
"loss": 0.1996,
|
| 547 |
-
"step": 550
|
| 548 |
-
},
|
| 549 |
-
{
|
| 550 |
-
"epoch": 17.041333333333334,
|
| 551 |
-
"eval_accuracy": 0.7741935483870968,
|
| 552 |
-
"eval_loss": 0.9459213018417358,
|
| 553 |
-
"eval_runtime": 104.9758,
|
| 554 |
-
"eval_samples_per_second": 0.295,
|
| 555 |
-
"eval_steps_per_second": 0.038,
|
| 556 |
-
"step": 558
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 18.002666666666666,
|
| 560 |
-
"grad_norm": 6.645630836486816,
|
| 561 |
-
"learning_rate": 1.4074074074074075e-05,
|
| 562 |
-
"loss": 0.5076,
|
| 563 |
-
"step": 560
|
| 564 |
-
},
|
| 565 |
-
{
|
| 566 |
-
"epoch": 18.016,
|
| 567 |
-
"grad_norm": 14.730420112609863,
|
| 568 |
-
"learning_rate": 1.3333333333333333e-05,
|
| 569 |
-
"loss": 0.2173,
|
| 570 |
-
"step": 570
|
| 571 |
-
},
|
| 572 |
-
{
|
| 573 |
-
"epoch": 18.029333333333334,
|
| 574 |
-
"grad_norm": 1.0937589406967163,
|
| 575 |
-
"learning_rate": 1.2592592592592592e-05,
|
| 576 |
-
"loss": 0.2886,
|
| 577 |
-
"step": 580
|
| 578 |
-
},
|
| 579 |
-
{
|
| 580 |
-
"epoch": 18.041333333333334,
|
| 581 |
-
"eval_accuracy": 0.8387096774193549,
|
| 582 |
-
"eval_loss": 0.7947275042533875,
|
| 583 |
-
"eval_runtime": 118.0527,
|
| 584 |
-
"eval_samples_per_second": 0.263,
|
| 585 |
-
"eval_steps_per_second": 0.034,
|
| 586 |
-
"step": 589
|
| 587 |
-
},
|
| 588 |
-
{
|
| 589 |
-
"epoch": 19.001333333333335,
|
| 590 |
-
"grad_norm": 0.08054836094379425,
|
| 591 |
-
"learning_rate": 1.1851851851851853e-05,
|
| 592 |
-
"loss": 0.1476,
|
| 593 |
-
"step": 590
|
| 594 |
-
},
|
| 595 |
-
{
|
| 596 |
-
"epoch": 19.014666666666667,
|
| 597 |
-
"grad_norm": 10.522760391235352,
|
| 598 |
-
"learning_rate": 1.1111111111111112e-05,
|
| 599 |
-
"loss": 0.1061,
|
| 600 |
-
"step": 600
|
| 601 |
-
},
|
| 602 |
-
{
|
| 603 |
-
"epoch": 19.028,
|
| 604 |
-
"grad_norm": 28.20208168029785,
|
| 605 |
-
"learning_rate": 1.037037037037037e-05,
|
| 606 |
-
"loss": 0.2122,
|
| 607 |
-
"step": 610
|
| 608 |
-
},
|
| 609 |
-
{
|
| 610 |
-
"epoch": 19.041333333333334,
|
| 611 |
-
"grad_norm": 1.5529413223266602,
|
| 612 |
-
"learning_rate": 9.62962962962963e-06,
|
| 613 |
-
"loss": 0.257,
|
| 614 |
-
"step": 620
|
| 615 |
-
},
|
| 616 |
-
{
|
| 617 |
-
"epoch": 19.041333333333334,
|
| 618 |
-
"eval_accuracy": 0.8064516129032258,
|
| 619 |
-
"eval_loss": 0.9583818316459656,
|
| 620 |
-
"eval_runtime": 103.5542,
|
| 621 |
-
"eval_samples_per_second": 0.299,
|
| 622 |
-
"eval_steps_per_second": 0.039,
|
| 623 |
-
"step": 620
|
| 624 |
-
},
|
| 625 |
-
{
|
| 626 |
-
"epoch": 20.013333333333332,
|
| 627 |
-
"grad_norm": 0.6693050265312195,
|
| 628 |
-
"learning_rate": 8.88888888888889e-06,
|
| 629 |
-
"loss": 0.0574,
|
| 630 |
-
"step": 630
|
| 631 |
-
},
|
| 632 |
-
{
|
| 633 |
-
"epoch": 20.026666666666667,
|
| 634 |
-
"grad_norm": 0.20252086222171783,
|
| 635 |
-
"learning_rate": 8.14814814814815e-06,
|
| 636 |
-
"loss": 0.3624,
|
| 637 |
-
"step": 640
|
| 638 |
-
},
|
| 639 |
-
{
|
| 640 |
-
"epoch": 20.04,
|
| 641 |
-
"grad_norm": 46.063262939453125,
|
| 642 |
-
"learning_rate": 7.4074074074074075e-06,
|
| 643 |
-
"loss": 0.3948,
|
| 644 |
-
"step": 650
|
| 645 |
-
},
|
| 646 |
-
{
|
| 647 |
-
"epoch": 20.041333333333334,
|
| 648 |
-
"eval_accuracy": 0.7741935483870968,
|
| 649 |
-
"eval_loss": 0.9725757837295532,
|
| 650 |
-
"eval_runtime": 99.1534,
|
| 651 |
-
"eval_samples_per_second": 0.313,
|
| 652 |
-
"eval_steps_per_second": 0.04,
|
| 653 |
-
"step": 651
|
| 654 |
-
},
|
| 655 |
-
{
|
| 656 |
-
"epoch": 21.012,
|
| 657 |
-
"grad_norm": 16.540903091430664,
|
| 658 |
-
"learning_rate": 6.666666666666667e-06,
|
| 659 |
-
"loss": 0.14,
|
| 660 |
-
"step": 660
|
| 661 |
-
},
|
| 662 |
-
{
|
| 663 |
-
"epoch": 21.025333333333332,
|
| 664 |
-
"grad_norm": 0.8179630041122437,
|
| 665 |
-
"learning_rate": 5.925925925925927e-06,
|
| 666 |
-
"loss": 0.1677,
|
| 667 |
-
"step": 670
|
| 668 |
-
},
|
| 669 |
-
{
|
| 670 |
-
"epoch": 21.038666666666668,
|
| 671 |
-
"grad_norm": 6.616237163543701,
|
| 672 |
-
"learning_rate": 5.185185185185185e-06,
|
| 673 |
-
"loss": 0.1969,
|
| 674 |
-
"step": 680
|
| 675 |
-
},
|
| 676 |
-
{
|
| 677 |
-
"epoch": 21.041333333333334,
|
| 678 |
-
"eval_accuracy": 0.8064516129032258,
|
| 679 |
-
"eval_loss": 0.8744015097618103,
|
| 680 |
-
"eval_runtime": 100.2476,
|
| 681 |
-
"eval_samples_per_second": 0.309,
|
| 682 |
-
"eval_steps_per_second": 0.04,
|
| 683 |
-
"step": 682
|
| 684 |
-
},
|
| 685 |
-
{
|
| 686 |
-
"epoch": 22.010666666666665,
|
| 687 |
-
"grad_norm": 0.2741946578025818,
|
| 688 |
-
"learning_rate": 4.444444444444445e-06,
|
| 689 |
-
"loss": 0.1749,
|
| 690 |
-
"step": 690
|
| 691 |
-
},
|
| 692 |
-
{
|
| 693 |
-
"epoch": 22.024,
|
| 694 |
-
"grad_norm": 4.550361633300781,
|
| 695 |
-
"learning_rate": 3.7037037037037037e-06,
|
| 696 |
-
"loss": 0.1865,
|
| 697 |
-
"step": 700
|
| 698 |
-
},
|
| 699 |
-
{
|
| 700 |
-
"epoch": 22.037333333333333,
|
| 701 |
-
"grad_norm": 1.1673641204833984,
|
| 702 |
-
"learning_rate": 2.9629629629629633e-06,
|
| 703 |
-
"loss": 0.1238,
|
| 704 |
-
"step": 710
|
| 705 |
-
},
|
| 706 |
-
{
|
| 707 |
-
"epoch": 22.041333333333334,
|
| 708 |
-
"eval_accuracy": 0.8064516129032258,
|
| 709 |
-
"eval_loss": 0.8289940357208252,
|
| 710 |
-
"eval_runtime": 107.0685,
|
| 711 |
-
"eval_samples_per_second": 0.29,
|
| 712 |
-
"eval_steps_per_second": 0.037,
|
| 713 |
-
"step": 713
|
| 714 |
-
},
|
| 715 |
-
{
|
| 716 |
-
"epoch": 23.009333333333334,
|
| 717 |
-
"grad_norm": 0.2685479521751404,
|
| 718 |
-
"learning_rate": 2.2222222222222225e-06,
|
| 719 |
-
"loss": 0.1107,
|
| 720 |
-
"step": 720
|
| 721 |
-
},
|
| 722 |
-
{
|
| 723 |
-
"epoch": 23.022666666666666,
|
| 724 |
-
"grad_norm": 8.248332023620605,
|
| 725 |
-
"learning_rate": 1.4814814814814817e-06,
|
| 726 |
-
"loss": 0.1607,
|
| 727 |
-
"step": 730
|
| 728 |
-
},
|
| 729 |
-
{
|
| 730 |
-
"epoch": 23.036,
|
| 731 |
-
"grad_norm": 0.12218274176120758,
|
| 732 |
-
"learning_rate": 7.407407407407408e-07,
|
| 733 |
-
"loss": 0.1539,
|
| 734 |
-
"step": 740
|
| 735 |
-
},
|
| 736 |
-
{
|
| 737 |
-
"epoch": 23.041333333333334,
|
| 738 |
-
"eval_accuracy": 0.8064516129032258,
|
| 739 |
-
"eval_loss": 0.8722597360610962,
|
| 740 |
-
"eval_runtime": 101.2734,
|
| 741 |
-
"eval_samples_per_second": 0.306,
|
| 742 |
-
"eval_steps_per_second": 0.039,
|
| 743 |
-
"step": 744
|
| 744 |
-
},
|
| 745 |
-
{
|
| 746 |
-
"epoch": 24.008,
|
| 747 |
-
"grad_norm": 21.209386825561523,
|
| 748 |
-
"learning_rate": 0.0,
|
| 749 |
-
"loss": 0.2076,
|
| 750 |
-
"step": 750
|
| 751 |
-
},
|
| 752 |
-
{
|
| 753 |
-
"epoch": 24.008,
|
| 754 |
-
"eval_accuracy": 0.8064516129032258,
|
| 755 |
-
"eval_loss": 0.8719311356544495,
|
| 756 |
-
"eval_runtime": 97.5374,
|
| 757 |
-
"eval_samples_per_second": 0.318,
|
| 758 |
-
"eval_steps_per_second": 0.041,
|
| 759 |
-
"step": 750
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 24.008,
|
| 763 |
-
"step": 750,
|
| 764 |
-
"total_flos": 7.41677486274773e+18,
|
| 765 |
-
"train_loss": 0.6099979252020518,
|
| 766 |
-
"train_runtime": 27777.5595,
|
| 767 |
-
"train_samples_per_second": 0.216,
|
| 768 |
-
"train_steps_per_second": 0.027
|
| 769 |
-
},
|
| 770 |
-
{
|
| 771 |
-
"epoch": 24.008,
|
| 772 |
-
"eval_accuracy": 0.8387096774193549,
|
| 773 |
-
"eval_loss": 0.7947275042533875,
|
| 774 |
-
"eval_runtime": 43.5309,
|
| 775 |
-
"eval_samples_per_second": 0.712,
|
| 776 |
-
"eval_steps_per_second": 0.092,
|
| 777 |
-
"step": 750
|
| 778 |
-
},
|
| 779 |
-
{
|
| 780 |
-
"epoch": 24.008,
|
| 781 |
-
"eval_accuracy": 0.8387096774193549,
|
| 782 |
-
"eval_loss": 0.7947273850440979,
|
| 783 |
-
"eval_runtime": 43.7384,
|
| 784 |
-
"eval_samples_per_second": 0.709,
|
| 785 |
-
"eval_steps_per_second": 0.091,
|
| 786 |
-
"step": 750
|
| 787 |
},
|
| 788 |
{
|
| 789 |
-
"epoch":
|
| 790 |
-
"eval_accuracy":
|
| 791 |
-
"eval_loss": 0.
|
| 792 |
-
"eval_runtime":
|
| 793 |
-
"eval_samples_per_second": 0.
|
| 794 |
"eval_steps_per_second": 0.075,
|
| 795 |
-
"step":
|
| 796 |
}
|
| 797 |
],
|
| 798 |
"logging_steps": 10,
|
| 799 |
-
"max_steps":
|
| 800 |
"num_input_tokens_seen": 0,
|
| 801 |
"num_train_epochs": 9223372036854775807,
|
| 802 |
"save_steps": 500,
|
|
@@ -807,12 +469,12 @@
|
|
| 807 |
"should_evaluate": false,
|
| 808 |
"should_log": false,
|
| 809 |
"should_save": true,
|
| 810 |
-
"should_training_stop":
|
| 811 |
},
|
| 812 |
"attributes": {}
|
| 813 |
}
|
| 814 |
},
|
| 815 |
-
"total_flos":
|
| 816 |
"train_batch_size": 8,
|
| 817 |
"trial_name": null,
|
| 818 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_global_step": 25,
|
| 3 |
+
"best_metric": 1.0,
|
| 4 |
+
"best_model_checkpoint": "videomae-base-finetuned-dogBehavior/checkpoint-25",
|
| 5 |
+
"epoch": 16.04,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 425,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"epoch": 0.016,
|
| 14 |
+
"grad_norm": 0.08260558545589447,
|
| 15 |
+
"learning_rate": 7.142857142857143e-06,
|
| 16 |
+
"loss": 0.0785,
|
| 17 |
"step": 10
|
| 18 |
},
|
| 19 |
{
|
| 20 |
+
"epoch": 0.032,
|
| 21 |
+
"grad_norm": 0.0643758624792099,
|
| 22 |
+
"learning_rate": 1.5079365079365079e-05,
|
| 23 |
+
"loss": 0.1621,
|
| 24 |
"step": 20
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.04,
|
| 28 |
+
"eval_accuracy": 1.0,
|
| 29 |
+
"eval_loss": 0.003746605012565851,
|
| 30 |
+
"eval_runtime": 65.8623,
|
| 31 |
+
"eval_samples_per_second": 0.501,
|
| 32 |
+
"eval_steps_per_second": 0.076,
|
| 33 |
+
"step": 25
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 1.008,
|
| 37 |
+
"grad_norm": 5.545346736907959,
|
| 38 |
+
"learning_rate": 2.3015873015873015e-05,
|
| 39 |
+
"loss": 0.1879,
|
| 40 |
"step": 30
|
| 41 |
},
|
| 42 |
{
|
| 43 |
+
"epoch": 1.024,
|
| 44 |
+
"grad_norm": 0.2720630466938019,
|
| 45 |
+
"learning_rate": 3.095238095238095e-05,
|
| 46 |
+
"loss": 0.129,
|
| 47 |
+
"step": 40
|
|
|
|
|
|
|
| 48 |
},
|
| 49 |
{
|
| 50 |
+
"epoch": 1.04,
|
| 51 |
+
"grad_norm": 0.02173990197479725,
|
| 52 |
+
"learning_rate": 3.888888888888889e-05,
|
| 53 |
+
"loss": 0.0774,
|
| 54 |
+
"step": 50
|
| 55 |
},
|
| 56 |
{
|
| 57 |
+
"epoch": 1.04,
|
| 58 |
+
"eval_accuracy": 1.0,
|
| 59 |
+
"eval_loss": 0.0037209123838692904,
|
| 60 |
+
"eval_runtime": 68.3183,
|
| 61 |
+
"eval_samples_per_second": 0.483,
|
| 62 |
+
"eval_steps_per_second": 0.073,
|
| 63 |
"step": 50
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 2.016,
|
| 67 |
+
"grad_norm": 0.1353379189968109,
|
| 68 |
+
"learning_rate": 4.682539682539683e-05,
|
| 69 |
+
"loss": 0.2455,
|
| 70 |
"step": 60
|
| 71 |
},
|
| 72 |
{
|
| 73 |
+
"epoch": 2.032,
|
| 74 |
+
"grad_norm": 0.08677935600280762,
|
| 75 |
+
"learning_rate": 4.946619217081851e-05,
|
| 76 |
+
"loss": 0.2208,
|
| 77 |
+
"step": 70
|
|
|
|
|
|
|
| 78 |
},
|
| 79 |
{
|
| 80 |
+
"epoch": 2.04,
|
| 81 |
+
"eval_accuracy": 1.0,
|
| 82 |
+
"eval_loss": 0.0056559075601398945,
|
| 83 |
+
"eval_runtime": 58.9935,
|
| 84 |
+
"eval_samples_per_second": 0.559,
|
| 85 |
+
"eval_steps_per_second": 0.085,
|
| 86 |
+
"step": 75
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 3.008,
|
| 90 |
+
"grad_norm": 5.721180438995361,
|
| 91 |
+
"learning_rate": 4.8576512455516014e-05,
|
| 92 |
+
"loss": 0.1613,
|
| 93 |
"step": 80
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 3.024,
|
| 97 |
+
"grad_norm": 13.376359939575195,
|
| 98 |
+
"learning_rate": 4.768683274021353e-05,
|
| 99 |
+
"loss": 0.2327,
|
| 100 |
"step": 90
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"epoch": 3.04,
|
| 104 |
+
"grad_norm": 0.04108860716223717,
|
| 105 |
+
"learning_rate": 4.6797153024911034e-05,
|
| 106 |
+
"loss": 0.0413,
|
| 107 |
+
"step": 100
|
|
|
|
|
|
|
| 108 |
},
|
| 109 |
{
|
| 110 |
+
"epoch": 3.04,
|
| 111 |
+
"eval_accuracy": 0.9090909090909091,
|
| 112 |
+
"eval_loss": 0.3500887453556061,
|
| 113 |
+
"eval_runtime": 70.4068,
|
| 114 |
+
"eval_samples_per_second": 0.469,
|
| 115 |
+
"eval_steps_per_second": 0.071,
|
| 116 |
"step": 100
|
| 117 |
},
|
| 118 |
{
|
| 119 |
+
"epoch": 4.016,
|
| 120 |
+
"grad_norm": 0.014622594229876995,
|
| 121 |
+
"learning_rate": 4.590747330960855e-05,
|
| 122 |
+
"loss": 0.0045,
|
| 123 |
"step": 110
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 4.032,
|
| 127 |
+
"grad_norm": 0.011623167432844639,
|
| 128 |
+
"learning_rate": 4.501779359430605e-05,
|
| 129 |
+
"loss": 0.1179,
|
| 130 |
"step": 120
|
| 131 |
},
|
| 132 |
{
|
| 133 |
+
"epoch": 4.04,
|
| 134 |
+
"eval_accuracy": 1.0,
|
| 135 |
+
"eval_loss": 0.011436140164732933,
|
| 136 |
+
"eval_runtime": 70.1018,
|
| 137 |
+
"eval_samples_per_second": 0.471,
|
| 138 |
+
"eval_steps_per_second": 0.071,
|
| 139 |
+
"step": 125
|
| 140 |
},
|
| 141 |
{
|
| 142 |
+
"epoch": 5.008,
|
| 143 |
+
"grad_norm": 0.16675230860710144,
|
| 144 |
+
"learning_rate": 4.4128113879003566e-05,
|
| 145 |
+
"loss": 0.0699,
|
| 146 |
"step": 130
|
| 147 |
},
|
| 148 |
{
|
| 149 |
+
"epoch": 5.024,
|
| 150 |
+
"grad_norm": 0.1757970154285431,
|
| 151 |
+
"learning_rate": 4.3238434163701066e-05,
|
| 152 |
+
"loss": 0.069,
|
| 153 |
"step": 140
|
| 154 |
},
|
| 155 |
{
|
| 156 |
+
"epoch": 5.04,
|
| 157 |
+
"grad_norm": 0.02306324616074562,
|
| 158 |
+
"learning_rate": 4.234875444839858e-05,
|
| 159 |
+
"loss": 0.0921,
|
| 160 |
"step": 150
|
| 161 |
},
|
| 162 |
{
|
| 163 |
+
"epoch": 5.04,
|
| 164 |
+
"eval_accuracy": 0.9696969696969697,
|
| 165 |
+
"eval_loss": 0.08965646475553513,
|
| 166 |
+
"eval_runtime": 58.4708,
|
| 167 |
+
"eval_samples_per_second": 0.564,
|
| 168 |
+
"eval_steps_per_second": 0.086,
|
| 169 |
+
"step": 150
|
| 170 |
},
|
| 171 |
{
|
| 172 |
+
"epoch": 6.016,
|
| 173 |
+
"grad_norm": 0.05490000173449516,
|
| 174 |
+
"learning_rate": 4.1459074733096085e-05,
|
| 175 |
+
"loss": 0.2398,
|
| 176 |
"step": 160
|
| 177 |
},
|
| 178 |
{
|
| 179 |
+
"epoch": 6.032,
|
| 180 |
+
"grad_norm": 82.93873596191406,
|
| 181 |
+
"learning_rate": 4.05693950177936e-05,
|
| 182 |
+
"loss": 0.0557,
|
| 183 |
"step": 170
|
| 184 |
},
|
| 185 |
{
|
| 186 |
+
"epoch": 6.04,
|
| 187 |
+
"eval_accuracy": 1.0,
|
| 188 |
+
"eval_loss": 0.0020419044885784388,
|
| 189 |
+
"eval_runtime": 66.481,
|
| 190 |
+
"eval_samples_per_second": 0.496,
|
| 191 |
+
"eval_steps_per_second": 0.075,
|
| 192 |
+
"step": 175
|
| 193 |
},
|
| 194 |
{
|
| 195 |
+
"epoch": 7.008,
|
| 196 |
+
"grad_norm": 0.34412676095962524,
|
| 197 |
+
"learning_rate": 3.9679715302491105e-05,
|
| 198 |
+
"loss": 0.2047,
|
| 199 |
+
"step": 180
|
|
|
|
|
|
|
| 200 |
},
|
| 201 |
{
|
| 202 |
+
"epoch": 7.024,
|
| 203 |
+
"grad_norm": 2.353379249572754,
|
| 204 |
+
"learning_rate": 3.879003558718861e-05,
|
| 205 |
+
"loss": 0.038,
|
| 206 |
"step": 190
|
| 207 |
},
|
| 208 |
{
|
| 209 |
+
"epoch": 7.04,
|
| 210 |
+
"grad_norm": 0.10385355353355408,
|
| 211 |
+
"learning_rate": 3.7900355871886125e-05,
|
| 212 |
+
"loss": 0.0366,
|
| 213 |
"step": 200
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 7.04,
|
| 217 |
+
"eval_accuracy": 1.0,
|
| 218 |
+
"eval_loss": 0.0015560887986794114,
|
| 219 |
+
"eval_runtime": 65.9693,
|
| 220 |
+
"eval_samples_per_second": 0.5,
|
| 221 |
+
"eval_steps_per_second": 0.076,
|
| 222 |
+
"step": 200
|
| 223 |
},
|
| 224 |
{
|
| 225 |
+
"epoch": 8.016,
|
| 226 |
+
"grad_norm": 0.23118802905082703,
|
| 227 |
+
"learning_rate": 3.701067615658363e-05,
|
| 228 |
+
"loss": 0.2039,
|
| 229 |
+
"step": 210
|
|
|
|
|
|
|
| 230 |
},
|
| 231 |
{
|
| 232 |
+
"epoch": 8.032,
|
| 233 |
+
"grad_norm": 0.016625450924038887,
|
| 234 |
+
"learning_rate": 3.6120996441281144e-05,
|
| 235 |
+
"loss": 0.0667,
|
| 236 |
"step": 220
|
| 237 |
},
|
| 238 |
{
|
| 239 |
+
"epoch": 8.04,
|
| 240 |
+
"eval_accuracy": 0.9696969696969697,
|
| 241 |
+
"eval_loss": 0.06851887702941895,
|
| 242 |
+
"eval_runtime": 59.1794,
|
| 243 |
+
"eval_samples_per_second": 0.558,
|
| 244 |
+
"eval_steps_per_second": 0.084,
|
| 245 |
+
"step": 225
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 9.008,
|
| 249 |
+
"grad_norm": 0.011781550943851471,
|
| 250 |
+
"learning_rate": 3.5231316725978644e-05,
|
| 251 |
+
"loss": 0.0016,
|
| 252 |
"step": 230
|
| 253 |
},
|
| 254 |
{
|
| 255 |
+
"epoch": 9.024,
|
| 256 |
+
"grad_norm": 0.008836457505822182,
|
| 257 |
+
"learning_rate": 3.434163701067616e-05,
|
| 258 |
+
"loss": 0.0037,
|
| 259 |
"step": 240
|
| 260 |
},
|
| 261 |
{
|
| 262 |
+
"epoch": 9.04,
|
| 263 |
+
"grad_norm": 0.0074387528002262115,
|
| 264 |
+
"learning_rate": 3.345195729537366e-05,
|
| 265 |
+
"loss": 0.1505,
|
| 266 |
+
"step": 250
|
|
|
|
|
|
|
| 267 |
},
|
| 268 |
{
|
| 269 |
+
"epoch": 9.04,
|
| 270 |
+
"eval_accuracy": 0.9696969696969697,
|
| 271 |
+
"eval_loss": 0.12975242733955383,
|
| 272 |
+
"eval_runtime": 66.9783,
|
| 273 |
+
"eval_samples_per_second": 0.493,
|
| 274 |
+
"eval_steps_per_second": 0.075,
|
| 275 |
"step": 250
|
| 276 |
},
|
| 277 |
{
|
| 278 |
+
"epoch": 10.016,
|
| 279 |
+
"grad_norm": 0.030738357454538345,
|
| 280 |
+
"learning_rate": 3.2562277580071177e-05,
|
| 281 |
+
"loss": 0.081,
|
| 282 |
"step": 260
|
| 283 |
},
|
| 284 |
{
|
| 285 |
+
"epoch": 10.032,
|
| 286 |
+
"grad_norm": 0.011689607053995132,
|
| 287 |
+
"learning_rate": 3.167259786476868e-05,
|
| 288 |
+
"loss": 0.0589,
|
| 289 |
"step": 270
|
| 290 |
},
|
| 291 |
{
|
| 292 |
+
"epoch": 10.04,
|
| 293 |
+
"eval_accuracy": 1.0,
|
| 294 |
+
"eval_loss": 0.005353024695068598,
|
| 295 |
+
"eval_runtime": 64.1994,
|
| 296 |
+
"eval_samples_per_second": 0.514,
|
| 297 |
+
"eval_steps_per_second": 0.078,
|
| 298 |
+
"step": 275
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 11.008,
|
| 302 |
+
"grad_norm": 4.4128546714782715,
|
| 303 |
+
"learning_rate": 3.0782918149466196e-05,
|
| 304 |
+
"loss": 0.0977,
|
| 305 |
"step": 280
|
| 306 |
},
|
| 307 |
{
|
| 308 |
+
"epoch": 11.024,
|
| 309 |
+
"grad_norm": 0.013472197577357292,
|
| 310 |
+
"learning_rate": 2.98932384341637e-05,
|
| 311 |
+
"loss": 0.0509,
|
| 312 |
"step": 290
|
| 313 |
},
|
| 314 |
{
|
| 315 |
+
"epoch": 11.04,
|
| 316 |
+
"grad_norm": 0.01684253290295601,
|
| 317 |
+
"learning_rate": 2.9003558718861212e-05,
|
| 318 |
+
"loss": 0.0669,
|
| 319 |
"step": 300
|
| 320 |
},
|
| 321 |
{
|
| 322 |
+
"epoch": 11.04,
|
| 323 |
+
"eval_accuracy": 0.9696969696969697,
|
| 324 |
+
"eval_loss": 0.026497945189476013,
|
| 325 |
+
"eval_runtime": 114.7091,
|
| 326 |
+
"eval_samples_per_second": 0.288,
|
| 327 |
+
"eval_steps_per_second": 0.044,
|
| 328 |
+
"step": 300
|
| 329 |
},
|
| 330 |
{
|
| 331 |
+
"epoch": 12.016,
|
| 332 |
+
"grad_norm": 42.653953552246094,
|
| 333 |
+
"learning_rate": 2.811387900355872e-05,
|
| 334 |
+
"loss": 0.0402,
|
|
|
|
|
|
|
| 335 |
"step": 310
|
| 336 |
},
|
| 337 |
{
|
| 338 |
+
"epoch": 12.032,
|
| 339 |
+
"grad_norm": 0.9748566150665283,
|
| 340 |
+
"learning_rate": 2.7224199288256232e-05,
|
| 341 |
+
"loss": 0.0094,
|
| 342 |
"step": 320
|
| 343 |
},
|
| 344 |
{
|
| 345 |
+
"epoch": 12.04,
|
| 346 |
+
"eval_accuracy": 0.9696969696969697,
|
| 347 |
+
"eval_loss": 0.12569771707057953,
|
| 348 |
+
"eval_runtime": 67.8328,
|
| 349 |
+
"eval_samples_per_second": 0.486,
|
| 350 |
+
"eval_steps_per_second": 0.074,
|
| 351 |
+
"step": 325
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 13.008,
|
| 355 |
+
"grad_norm": 0.06350671499967575,
|
| 356 |
+
"learning_rate": 2.6334519572953735e-05,
|
| 357 |
+
"loss": 0.0117,
|
| 358 |
"step": 330
|
| 359 |
},
|
| 360 |
{
|
| 361 |
+
"epoch": 13.024,
|
| 362 |
+
"grad_norm": 0.02096625417470932,
|
| 363 |
+
"learning_rate": 2.5444839857651248e-05,
|
| 364 |
+
"loss": 0.0003,
|
| 365 |
"step": 340
|
| 366 |
},
|
| 367 |
{
|
| 368 |
+
"epoch": 13.04,
|
| 369 |
+
"grad_norm": 13.891782760620117,
|
| 370 |
+
"learning_rate": 2.4555160142348754e-05,
|
| 371 |
+
"loss": 0.1737,
|
| 372 |
+
"step": 350
|
|
|
|
|
|
|
| 373 |
},
|
| 374 |
{
|
| 375 |
+
"epoch": 13.04,
|
| 376 |
+
"eval_accuracy": 1.0,
|
| 377 |
+
"eval_loss": 0.00027246223180554807,
|
| 378 |
+
"eval_runtime": 66.3634,
|
| 379 |
+
"eval_samples_per_second": 0.497,
|
| 380 |
+
"eval_steps_per_second": 0.075,
|
| 381 |
"step": 350
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 14.016,
|
| 385 |
+
"grad_norm": 0.00640275189653039,
|
| 386 |
+
"learning_rate": 2.3665480427046264e-05,
|
| 387 |
+
"loss": 0.0056,
|
| 388 |
"step": 360
|
| 389 |
},
|
| 390 |
{
|
| 391 |
+
"epoch": 14.032,
|
| 392 |
+
"grad_norm": 0.2554427683353424,
|
| 393 |
+
"learning_rate": 2.277580071174377e-05,
|
| 394 |
+
"loss": 0.0775,
|
| 395 |
"step": 370
|
| 396 |
},
|
| 397 |
{
|
| 398 |
+
"epoch": 14.04,
|
| 399 |
+
"eval_accuracy": 1.0,
|
| 400 |
+
"eval_loss": 0.0005332967266440392,
|
| 401 |
+
"eval_runtime": 57.622,
|
| 402 |
+
"eval_samples_per_second": 0.573,
|
| 403 |
+
"eval_steps_per_second": 0.087,
|
| 404 |
+
"step": 375
|
| 405 |
},
|
| 406 |
{
|
| 407 |
+
"epoch": 15.008,
|
| 408 |
+
"grad_norm": 0.010061250999569893,
|
| 409 |
+
"learning_rate": 2.188612099644128e-05,
|
| 410 |
+
"loss": 0.0076,
|
| 411 |
"step": 380
|
| 412 |
},
|
| 413 |
{
|
| 414 |
+
"epoch": 15.024,
|
| 415 |
+
"grad_norm": 0.009161919355392456,
|
| 416 |
+
"learning_rate": 2.099644128113879e-05,
|
| 417 |
+
"loss": 0.0775,
|
| 418 |
"step": 390
|
| 419 |
},
|
| 420 |
{
|
| 421 |
+
"epoch": 15.04,
|
| 422 |
+
"grad_norm": 0.030138863250613213,
|
| 423 |
+
"learning_rate": 2.01067615658363e-05,
|
| 424 |
+
"loss": 0.0805,
|
| 425 |
"step": 400
|
| 426 |
},
|
| 427 |
{
|
| 428 |
+
"epoch": 15.04,
|
| 429 |
+
"eval_accuracy": 1.0,
|
| 430 |
+
"eval_loss": 0.0005667555378749967,
|
| 431 |
+
"eval_runtime": 65.0511,
|
| 432 |
+
"eval_samples_per_second": 0.507,
|
| 433 |
+
"eval_steps_per_second": 0.077,
|
| 434 |
+
"step": 400
|
| 435 |
},
|
| 436 |
{
|
| 437 |
+
"epoch": 16.016,
|
| 438 |
+
"grad_norm": 0.02383618988096714,
|
| 439 |
+
"learning_rate": 1.9217081850533806e-05,
|
| 440 |
+
"loss": 0.0705,
|
| 441 |
"step": 410
|
| 442 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 443 |
{
|
| 444 |
"epoch": 16.032,
|
| 445 |
+
"grad_norm": 0.027891239151358604,
|
| 446 |
+
"learning_rate": 1.8327402135231316e-05,
|
| 447 |
+
"loss": 0.0035,
|
| 448 |
+
"step": 420
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 449 |
},
|
| 450 |
{
|
| 451 |
+
"epoch": 16.04,
|
| 452 |
+
"eval_accuracy": 1.0,
|
| 453 |
+
"eval_loss": 0.0008408486610278487,
|
| 454 |
+
"eval_runtime": 66.7117,
|
| 455 |
+
"eval_samples_per_second": 0.495,
|
| 456 |
"eval_steps_per_second": 0.075,
|
| 457 |
+
"step": 425
|
| 458 |
}
|
| 459 |
],
|
| 460 |
"logging_steps": 10,
|
| 461 |
+
"max_steps": 625,
|
| 462 |
"num_input_tokens_seen": 0,
|
| 463 |
"num_train_epochs": 9223372036854775807,
|
| 464 |
"save_steps": 500,
|
|
|
|
| 469 |
"should_evaluate": false,
|
| 470 |
"should_log": false,
|
| 471 |
"should_save": true,
|
| 472 |
+
"should_training_stop": false
|
| 473 |
},
|
| 474 |
"attributes": {}
|
| 475 |
}
|
| 476 |
},
|
| 477 |
+
"total_flos": 4.2366196042039296e+18,
|
| 478 |
"train_batch_size": 8,
|
| 479 |
"trial_name": null,
|
| 480 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aeeec3c2e5d7c9acdb45dfc40a15b0dfa96eadede48209631aad893ee4584cea
|
| 3 |
+
size 5841
|