upload model
Browse files- .gitignore +2 -0
- README.md +134 -0
- all_results.json +12 -0
- config.json +231 -0
- eval_results.json +8 -0
- preprocessor_config.json +22 -0
- pytorch_model.bin +3 -0
- train_results.json +7 -0
- trainer_state.json +1114 -0
- training_args.bin +3 -0
.gitignore
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
task_description.json
|
| 2 |
+
*.log
|
README.md
ADDED
|
@@ -0,0 +1,134 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
tags:
|
| 3 |
+
- image-classification
|
| 4 |
+
- vision
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
datasets:
|
| 7 |
+
- food101
|
| 8 |
+
metrics:
|
| 9 |
+
- accuracy
|
| 10 |
+
model-index:
|
| 11 |
+
- name: lr6e-05
|
| 12 |
+
results:
|
| 13 |
+
- task:
|
| 14 |
+
name: Image Classification
|
| 15 |
+
type: image-classification
|
| 16 |
+
dataset:
|
| 17 |
+
name: food101
|
| 18 |
+
type: food101
|
| 19 |
+
config: default
|
| 20 |
+
split: validation
|
| 21 |
+
args: default
|
| 22 |
+
metrics:
|
| 23 |
+
- name: Accuracy
|
| 24 |
+
type: accuracy
|
| 25 |
+
value: 0.8971089108910891
|
| 26 |
+
---
|
| 27 |
+
|
| 28 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 29 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 30 |
+
|
| 31 |
+
# vit-base-patch16-224-food101
|
| 32 |
+
|
| 33 |
+
This model is a fine-tuned version of [eslamxm/vit-base-food101](https://huggingface.co/eslamxm/vit-base-food101) on the food101 dataset.
|
| 34 |
+
It achieves the following results on the evaluation set:
|
| 35 |
+
- Loss: 0.3856
|
| 36 |
+
- Accuracy: 0.8971
|
| 37 |
+
|
| 38 |
+
## Model description
|
| 39 |
+
|
| 40 |
+
More information needed
|
| 41 |
+
|
| 42 |
+
## Intended uses & limitations
|
| 43 |
+
|
| 44 |
+
More information needed
|
| 45 |
+
|
| 46 |
+
## Training and evaluation data
|
| 47 |
+
|
| 48 |
+
More information needed
|
| 49 |
+
|
| 50 |
+
## Script
|
| 51 |
+
```python
|
| 52 |
+
"cmd_list": [
|
| 53 |
+
"python",
|
| 54 |
+
"run_image_classification.py",
|
| 55 |
+
"--model_name_or_path",
|
| 56 |
+
"eslamxm/vit-base-food101",
|
| 57 |
+
"--dataset_name",
|
| 58 |
+
"food101",
|
| 59 |
+
"--output_dir",
|
| 60 |
+
"<output_dir>",
|
| 61 |
+
"--overwrite_output_dir",
|
| 62 |
+
"--remove_unused_columns",
|
| 63 |
+
"False",
|
| 64 |
+
"--do_train",
|
| 65 |
+
"--do_eval",
|
| 66 |
+
"--optim",
|
| 67 |
+
"adamw_torch",
|
| 68 |
+
"--learning_rate",
|
| 69 |
+
"6e-05",
|
| 70 |
+
"--num_train_epochs",
|
| 71 |
+
"3",
|
| 72 |
+
"--dataloader_num_workers",
|
| 73 |
+
"10",
|
| 74 |
+
"--per_device_train_batch_size",
|
| 75 |
+
"64",
|
| 76 |
+
"--gradient_accumulation_steps",
|
| 77 |
+
"2",
|
| 78 |
+
"--per_device_eval_batch_size",
|
| 79 |
+
"128",
|
| 80 |
+
"--logging_strategy",
|
| 81 |
+
"steps",
|
| 82 |
+
"--logging_steps",
|
| 83 |
+
"10",
|
| 84 |
+
"--evaluation_strategy",
|
| 85 |
+
"steps",
|
| 86 |
+
"--eval_steps",
|
| 87 |
+
"500",
|
| 88 |
+
"--save_steps",
|
| 89 |
+
"500",
|
| 90 |
+
"--evaluation_strategy",
|
| 91 |
+
"epoch",
|
| 92 |
+
"--save_strategy",
|
| 93 |
+
"epoch",
|
| 94 |
+
"--load_best_model_at_end",
|
| 95 |
+
"False",
|
| 96 |
+
"--save_total_limit",
|
| 97 |
+
"1",
|
| 98 |
+
"--seed",
|
| 99 |
+
"42",
|
| 100 |
+
"--fp16"
|
| 101 |
+
]
|
| 102 |
+
```
|
| 103 |
+
|
| 104 |
+
## Training procedure
|
| 105 |
+
|
| 106 |
+
### Training hyperparameters
|
| 107 |
+
|
| 108 |
+
The following hyperparameters were used during training:
|
| 109 |
+
- learning_rate: 6e-05
|
| 110 |
+
- train_batch_size: 64
|
| 111 |
+
- eval_batch_size: 128
|
| 112 |
+
- seed: 42
|
| 113 |
+
- gradient_accumulation_steps: 2
|
| 114 |
+
- total_train_batch_size: 128
|
| 115 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 116 |
+
- lr_scheduler_type: linear
|
| 117 |
+
- num_epochs: 3.0
|
| 118 |
+
- mixed_precision_training: Native AMP
|
| 119 |
+
|
| 120 |
+
### Training results
|
| 121 |
+
|
| 122 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 123 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
| 124 |
+
| 0.3687 | 1.0 | 592 | 0.4044 | 0.8889 |
|
| 125 |
+
| 0.3422 | 2.0 | 1184 | 0.3911 | 0.8953 |
|
| 126 |
+
| 0.3808 | 3.0 | 1776 | 0.3856 | 0.8971 |
|
| 127 |
+
|
| 128 |
+
|
| 129 |
+
### Framework versions
|
| 130 |
+
|
| 131 |
+
- Transformers 4.27.4
|
| 132 |
+
- Pytorch 1.13.1
|
| 133 |
+
- Datasets 2.11.0
|
| 134 |
+
- Tokenizers 0.13.3
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.8971089108910891,
|
| 4 |
+
"eval_loss": 0.38559553027153015,
|
| 5 |
+
"eval_runtime": 33.4525,
|
| 6 |
+
"eval_samples_per_second": 754.801,
|
| 7 |
+
"eval_steps_per_second": 5.919,
|
| 8 |
+
"train_loss": 0.3829323179549999,
|
| 9 |
+
"train_runtime": 916.6333,
|
| 10 |
+
"train_samples_per_second": 247.918,
|
| 11 |
+
"train_steps_per_second": 1.938
|
| 12 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,231 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "/mnt/sh_flex_storage/home/yujiepan/vit-base-food101",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ViTForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.0,
|
| 7 |
+
"encoder_stride": 16,
|
| 8 |
+
"finetuning_task": "image-classification",
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.0,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "apple_pie",
|
| 14 |
+
"1": "baby_back_ribs",
|
| 15 |
+
"10": "bruschetta",
|
| 16 |
+
"100": "waffles",
|
| 17 |
+
"11": "caesar_salad",
|
| 18 |
+
"12": "cannoli",
|
| 19 |
+
"13": "caprese_salad",
|
| 20 |
+
"14": "carrot_cake",
|
| 21 |
+
"15": "ceviche",
|
| 22 |
+
"16": "cheesecake",
|
| 23 |
+
"17": "cheese_plate",
|
| 24 |
+
"18": "chicken_curry",
|
| 25 |
+
"19": "chicken_quesadilla",
|
| 26 |
+
"2": "baklava",
|
| 27 |
+
"20": "chicken_wings",
|
| 28 |
+
"21": "chocolate_cake",
|
| 29 |
+
"22": "chocolate_mousse",
|
| 30 |
+
"23": "churros",
|
| 31 |
+
"24": "clam_chowder",
|
| 32 |
+
"25": "club_sandwich",
|
| 33 |
+
"26": "crab_cakes",
|
| 34 |
+
"27": "creme_brulee",
|
| 35 |
+
"28": "croque_madame",
|
| 36 |
+
"29": "cup_cakes",
|
| 37 |
+
"3": "beef_carpaccio",
|
| 38 |
+
"30": "deviled_eggs",
|
| 39 |
+
"31": "donuts",
|
| 40 |
+
"32": "dumplings",
|
| 41 |
+
"33": "edamame",
|
| 42 |
+
"34": "eggs_benedict",
|
| 43 |
+
"35": "escargots",
|
| 44 |
+
"36": "falafel",
|
| 45 |
+
"37": "filet_mignon",
|
| 46 |
+
"38": "fish_and_chips",
|
| 47 |
+
"39": "foie_gras",
|
| 48 |
+
"4": "beef_tartare",
|
| 49 |
+
"40": "french_fries",
|
| 50 |
+
"41": "french_onion_soup",
|
| 51 |
+
"42": "french_toast",
|
| 52 |
+
"43": "fried_calamari",
|
| 53 |
+
"44": "fried_rice",
|
| 54 |
+
"45": "frozen_yogurt",
|
| 55 |
+
"46": "garlic_bread",
|
| 56 |
+
"47": "gnocchi",
|
| 57 |
+
"48": "greek_salad",
|
| 58 |
+
"49": "grilled_cheese_sandwich",
|
| 59 |
+
"5": "beet_salad",
|
| 60 |
+
"50": "grilled_salmon",
|
| 61 |
+
"51": "guacamole",
|
| 62 |
+
"52": "gyoza",
|
| 63 |
+
"53": "hamburger",
|
| 64 |
+
"54": "hot_and_sour_soup",
|
| 65 |
+
"55": "hot_dog",
|
| 66 |
+
"56": "huevos_rancheros",
|
| 67 |
+
"57": "hummus",
|
| 68 |
+
"58": "ice_cream",
|
| 69 |
+
"59": "lasagna",
|
| 70 |
+
"6": "beignets",
|
| 71 |
+
"60": "lobster_bisque",
|
| 72 |
+
"61": "lobster_roll_sandwich",
|
| 73 |
+
"62": "macaroni_and_cheese",
|
| 74 |
+
"63": "macarons",
|
| 75 |
+
"64": "miso_soup",
|
| 76 |
+
"65": "mussels",
|
| 77 |
+
"66": "nachos",
|
| 78 |
+
"67": "omelette",
|
| 79 |
+
"68": "onion_rings",
|
| 80 |
+
"69": "oysters",
|
| 81 |
+
"7": "bibimbap",
|
| 82 |
+
"70": "pad_thai",
|
| 83 |
+
"71": "paella",
|
| 84 |
+
"72": "pancakes",
|
| 85 |
+
"73": "panna_cotta",
|
| 86 |
+
"74": "peking_duck",
|
| 87 |
+
"75": "pho",
|
| 88 |
+
"76": "pizza",
|
| 89 |
+
"77": "pork_chop",
|
| 90 |
+
"78": "poutine",
|
| 91 |
+
"79": "prime_rib",
|
| 92 |
+
"8": "bread_pudding",
|
| 93 |
+
"80": "pulled_pork_sandwich",
|
| 94 |
+
"81": "ramen",
|
| 95 |
+
"82": "ravioli",
|
| 96 |
+
"83": "red_velvet_cake",
|
| 97 |
+
"84": "risotto",
|
| 98 |
+
"85": "samosa",
|
| 99 |
+
"86": "sashimi",
|
| 100 |
+
"87": "scallops",
|
| 101 |
+
"88": "seaweed_salad",
|
| 102 |
+
"89": "shrimp_and_grits",
|
| 103 |
+
"9": "breakfast_burrito",
|
| 104 |
+
"90": "spaghetti_bolognese",
|
| 105 |
+
"91": "spaghetti_carbonara",
|
| 106 |
+
"92": "spring_rolls",
|
| 107 |
+
"93": "steak",
|
| 108 |
+
"94": "strawberry_shortcake",
|
| 109 |
+
"95": "sushi",
|
| 110 |
+
"96": "tacos",
|
| 111 |
+
"97": "takoyaki",
|
| 112 |
+
"98": "tiramisu",
|
| 113 |
+
"99": "tuna_tartare"
|
| 114 |
+
},
|
| 115 |
+
"image_size": 224,
|
| 116 |
+
"initializer_range": 0.02,
|
| 117 |
+
"intermediate_size": 3072,
|
| 118 |
+
"label2id": {
|
| 119 |
+
"apple_pie": "0",
|
| 120 |
+
"baby_back_ribs": "1",
|
| 121 |
+
"baklava": "2",
|
| 122 |
+
"beef_carpaccio": "3",
|
| 123 |
+
"beef_tartare": "4",
|
| 124 |
+
"beet_salad": "5",
|
| 125 |
+
"beignets": "6",
|
| 126 |
+
"bibimbap": "7",
|
| 127 |
+
"bread_pudding": "8",
|
| 128 |
+
"breakfast_burrito": "9",
|
| 129 |
+
"bruschetta": "10",
|
| 130 |
+
"caesar_salad": "11",
|
| 131 |
+
"cannoli": "12",
|
| 132 |
+
"caprese_salad": "13",
|
| 133 |
+
"carrot_cake": "14",
|
| 134 |
+
"ceviche": "15",
|
| 135 |
+
"cheese_plate": "17",
|
| 136 |
+
"cheesecake": "16",
|
| 137 |
+
"chicken_curry": "18",
|
| 138 |
+
"chicken_quesadilla": "19",
|
| 139 |
+
"chicken_wings": "20",
|
| 140 |
+
"chocolate_cake": "21",
|
| 141 |
+
"chocolate_mousse": "22",
|
| 142 |
+
"churros": "23",
|
| 143 |
+
"clam_chowder": "24",
|
| 144 |
+
"club_sandwich": "25",
|
| 145 |
+
"crab_cakes": "26",
|
| 146 |
+
"creme_brulee": "27",
|
| 147 |
+
"croque_madame": "28",
|
| 148 |
+
"cup_cakes": "29",
|
| 149 |
+
"deviled_eggs": "30",
|
| 150 |
+
"donuts": "31",
|
| 151 |
+
"dumplings": "32",
|
| 152 |
+
"edamame": "33",
|
| 153 |
+
"eggs_benedict": "34",
|
| 154 |
+
"escargots": "35",
|
| 155 |
+
"falafel": "36",
|
| 156 |
+
"filet_mignon": "37",
|
| 157 |
+
"fish_and_chips": "38",
|
| 158 |
+
"foie_gras": "39",
|
| 159 |
+
"french_fries": "40",
|
| 160 |
+
"french_onion_soup": "41",
|
| 161 |
+
"french_toast": "42",
|
| 162 |
+
"fried_calamari": "43",
|
| 163 |
+
"fried_rice": "44",
|
| 164 |
+
"frozen_yogurt": "45",
|
| 165 |
+
"garlic_bread": "46",
|
| 166 |
+
"gnocchi": "47",
|
| 167 |
+
"greek_salad": "48",
|
| 168 |
+
"grilled_cheese_sandwich": "49",
|
| 169 |
+
"grilled_salmon": "50",
|
| 170 |
+
"guacamole": "51",
|
| 171 |
+
"gyoza": "52",
|
| 172 |
+
"hamburger": "53",
|
| 173 |
+
"hot_and_sour_soup": "54",
|
| 174 |
+
"hot_dog": "55",
|
| 175 |
+
"huevos_rancheros": "56",
|
| 176 |
+
"hummus": "57",
|
| 177 |
+
"ice_cream": "58",
|
| 178 |
+
"lasagna": "59",
|
| 179 |
+
"lobster_bisque": "60",
|
| 180 |
+
"lobster_roll_sandwich": "61",
|
| 181 |
+
"macaroni_and_cheese": "62",
|
| 182 |
+
"macarons": "63",
|
| 183 |
+
"miso_soup": "64",
|
| 184 |
+
"mussels": "65",
|
| 185 |
+
"nachos": "66",
|
| 186 |
+
"omelette": "67",
|
| 187 |
+
"onion_rings": "68",
|
| 188 |
+
"oysters": "69",
|
| 189 |
+
"pad_thai": "70",
|
| 190 |
+
"paella": "71",
|
| 191 |
+
"pancakes": "72",
|
| 192 |
+
"panna_cotta": "73",
|
| 193 |
+
"peking_duck": "74",
|
| 194 |
+
"pho": "75",
|
| 195 |
+
"pizza": "76",
|
| 196 |
+
"pork_chop": "77",
|
| 197 |
+
"poutine": "78",
|
| 198 |
+
"prime_rib": "79",
|
| 199 |
+
"pulled_pork_sandwich": "80",
|
| 200 |
+
"ramen": "81",
|
| 201 |
+
"ravioli": "82",
|
| 202 |
+
"red_velvet_cake": "83",
|
| 203 |
+
"risotto": "84",
|
| 204 |
+
"samosa": "85",
|
| 205 |
+
"sashimi": "86",
|
| 206 |
+
"scallops": "87",
|
| 207 |
+
"seaweed_salad": "88",
|
| 208 |
+
"shrimp_and_grits": "89",
|
| 209 |
+
"spaghetti_bolognese": "90",
|
| 210 |
+
"spaghetti_carbonara": "91",
|
| 211 |
+
"spring_rolls": "92",
|
| 212 |
+
"steak": "93",
|
| 213 |
+
"strawberry_shortcake": "94",
|
| 214 |
+
"sushi": "95",
|
| 215 |
+
"tacos": "96",
|
| 216 |
+
"takoyaki": "97",
|
| 217 |
+
"tiramisu": "98",
|
| 218 |
+
"tuna_tartare": "99",
|
| 219 |
+
"waffles": "100"
|
| 220 |
+
},
|
| 221 |
+
"layer_norm_eps": 1e-12,
|
| 222 |
+
"model_type": "vit",
|
| 223 |
+
"num_attention_heads": 12,
|
| 224 |
+
"num_channels": 3,
|
| 225 |
+
"num_hidden_layers": 12,
|
| 226 |
+
"patch_size": 16,
|
| 227 |
+
"problem_type": "single_label_classification",
|
| 228 |
+
"qkv_bias": true,
|
| 229 |
+
"torch_dtype": "float32",
|
| 230 |
+
"transformers_version": "4.27.4"
|
| 231 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.8971089108910891,
|
| 4 |
+
"eval_loss": 0.38559553027153015,
|
| 5 |
+
"eval_runtime": 33.4525,
|
| 6 |
+
"eval_samples_per_second": 754.801,
|
| 7 |
+
"eval_steps_per_second": 5.919
|
| 8 |
+
}
|
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 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:86a9f8e9e168929fa1bfb9ccaf6b0a3c543ccfc4345b28b7b895fb90b5e8c2a5
|
| 3 |
+
size 343573229
|
train_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"train_loss": 0.3829323179549999,
|
| 4 |
+
"train_runtime": 916.6333,
|
| 5 |
+
"train_samples_per_second": 247.918,
|
| 6 |
+
"train_steps_per_second": 1.938
|
| 7 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1114 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"global_step": 1776,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.02,
|
| 12 |
+
"learning_rate": 5.966216216216216e-05,
|
| 13 |
+
"loss": 0.426,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.03,
|
| 18 |
+
"learning_rate": 5.9324324324324324e-05,
|
| 19 |
+
"loss": 0.484,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.05,
|
| 24 |
+
"learning_rate": 5.8986486486486485e-05,
|
| 25 |
+
"loss": 0.4457,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.07,
|
| 30 |
+
"learning_rate": 5.8648648648648653e-05,
|
| 31 |
+
"loss": 0.4083,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.08,
|
| 36 |
+
"learning_rate": 5.83445945945946e-05,
|
| 37 |
+
"loss": 0.4695,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.1,
|
| 42 |
+
"learning_rate": 5.800675675675676e-05,
|
| 43 |
+
"loss": 0.5002,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.12,
|
| 48 |
+
"learning_rate": 5.766891891891892e-05,
|
| 49 |
+
"loss": 0.4436,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.14,
|
| 54 |
+
"learning_rate": 5.733108108108108e-05,
|
| 55 |
+
"loss": 0.3807,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.15,
|
| 60 |
+
"learning_rate": 5.699324324324325e-05,
|
| 61 |
+
"loss": 0.4198,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.17,
|
| 66 |
+
"learning_rate": 5.665540540540541e-05,
|
| 67 |
+
"loss": 0.4475,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.19,
|
| 72 |
+
"learning_rate": 5.631756756756757e-05,
|
| 73 |
+
"loss": 0.434,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.2,
|
| 78 |
+
"learning_rate": 5.5979729729729734e-05,
|
| 79 |
+
"loss": 0.4036,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.22,
|
| 84 |
+
"learning_rate": 5.5641891891891896e-05,
|
| 85 |
+
"loss": 0.4386,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.24,
|
| 90 |
+
"learning_rate": 5.530405405405406e-05,
|
| 91 |
+
"loss": 0.399,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.25,
|
| 96 |
+
"learning_rate": 5.496621621621622e-05,
|
| 97 |
+
"loss": 0.5032,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.27,
|
| 102 |
+
"learning_rate": 5.4628378378378386e-05,
|
| 103 |
+
"loss": 0.4435,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.29,
|
| 108 |
+
"learning_rate": 5.429054054054054e-05,
|
| 109 |
+
"loss": 0.4947,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.3,
|
| 114 |
+
"learning_rate": 5.39527027027027e-05,
|
| 115 |
+
"loss": 0.4708,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.32,
|
| 120 |
+
"learning_rate": 5.361486486486486e-05,
|
| 121 |
+
"loss": 0.4644,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.34,
|
| 126 |
+
"learning_rate": 5.3277027027027024e-05,
|
| 127 |
+
"loss": 0.4833,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.35,
|
| 132 |
+
"learning_rate": 5.2939189189189186e-05,
|
| 133 |
+
"loss": 0.4115,
|
| 134 |
+
"step": 210
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.37,
|
| 138 |
+
"learning_rate": 5.260135135135135e-05,
|
| 139 |
+
"loss": 0.5,
|
| 140 |
+
"step": 220
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.39,
|
| 144 |
+
"learning_rate": 5.2263513513513515e-05,
|
| 145 |
+
"loss": 0.4415,
|
| 146 |
+
"step": 230
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.41,
|
| 150 |
+
"learning_rate": 5.1925675675675676e-05,
|
| 151 |
+
"loss": 0.4143,
|
| 152 |
+
"step": 240
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.42,
|
| 156 |
+
"learning_rate": 5.158783783783784e-05,
|
| 157 |
+
"loss": 0.4295,
|
| 158 |
+
"step": 250
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.44,
|
| 162 |
+
"learning_rate": 5.125e-05,
|
| 163 |
+
"loss": 0.4441,
|
| 164 |
+
"step": 260
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.46,
|
| 168 |
+
"learning_rate": 5.091216216216216e-05,
|
| 169 |
+
"loss": 0.4273,
|
| 170 |
+
"step": 270
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.47,
|
| 174 |
+
"learning_rate": 5.057432432432432e-05,
|
| 175 |
+
"loss": 0.4175,
|
| 176 |
+
"step": 280
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.49,
|
| 180 |
+
"learning_rate": 5.023648648648648e-05,
|
| 181 |
+
"loss": 0.4219,
|
| 182 |
+
"step": 290
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.51,
|
| 186 |
+
"learning_rate": 4.989864864864865e-05,
|
| 187 |
+
"loss": 0.4526,
|
| 188 |
+
"step": 300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.52,
|
| 192 |
+
"learning_rate": 4.956081081081081e-05,
|
| 193 |
+
"loss": 0.4396,
|
| 194 |
+
"step": 310
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.54,
|
| 198 |
+
"learning_rate": 4.922297297297297e-05,
|
| 199 |
+
"loss": 0.444,
|
| 200 |
+
"step": 320
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.56,
|
| 204 |
+
"learning_rate": 4.8885135135135135e-05,
|
| 205 |
+
"loss": 0.4236,
|
| 206 |
+
"step": 330
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.57,
|
| 210 |
+
"learning_rate": 4.8547297297297296e-05,
|
| 211 |
+
"loss": 0.4228,
|
| 212 |
+
"step": 340
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.59,
|
| 216 |
+
"learning_rate": 4.820945945945946e-05,
|
| 217 |
+
"loss": 0.4055,
|
| 218 |
+
"step": 350
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.61,
|
| 222 |
+
"learning_rate": 4.787162162162162e-05,
|
| 223 |
+
"loss": 0.3954,
|
| 224 |
+
"step": 360
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.62,
|
| 228 |
+
"learning_rate": 4.7533783783783786e-05,
|
| 229 |
+
"loss": 0.3883,
|
| 230 |
+
"step": 370
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.64,
|
| 234 |
+
"learning_rate": 4.719594594594595e-05,
|
| 235 |
+
"loss": 0.4181,
|
| 236 |
+
"step": 380
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.66,
|
| 240 |
+
"learning_rate": 4.685810810810811e-05,
|
| 241 |
+
"loss": 0.4249,
|
| 242 |
+
"step": 390
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.68,
|
| 246 |
+
"learning_rate": 4.652027027027027e-05,
|
| 247 |
+
"loss": 0.4619,
|
| 248 |
+
"step": 400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.69,
|
| 252 |
+
"learning_rate": 4.618243243243243e-05,
|
| 253 |
+
"loss": 0.4298,
|
| 254 |
+
"step": 410
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.71,
|
| 258 |
+
"learning_rate": 4.584459459459459e-05,
|
| 259 |
+
"loss": 0.4184,
|
| 260 |
+
"step": 420
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.73,
|
| 264 |
+
"learning_rate": 4.550675675675676e-05,
|
| 265 |
+
"loss": 0.4062,
|
| 266 |
+
"step": 430
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.74,
|
| 270 |
+
"learning_rate": 4.516891891891892e-05,
|
| 271 |
+
"loss": 0.3768,
|
| 272 |
+
"step": 440
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.76,
|
| 276 |
+
"learning_rate": 4.4831081081081083e-05,
|
| 277 |
+
"loss": 0.4605,
|
| 278 |
+
"step": 450
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.78,
|
| 282 |
+
"learning_rate": 4.4493243243243245e-05,
|
| 283 |
+
"loss": 0.485,
|
| 284 |
+
"step": 460
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.79,
|
| 288 |
+
"learning_rate": 4.4155405405405406e-05,
|
| 289 |
+
"loss": 0.4169,
|
| 290 |
+
"step": 470
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.81,
|
| 294 |
+
"learning_rate": 4.381756756756757e-05,
|
| 295 |
+
"loss": 0.4136,
|
| 296 |
+
"step": 480
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.83,
|
| 300 |
+
"learning_rate": 4.347972972972973e-05,
|
| 301 |
+
"loss": 0.4337,
|
| 302 |
+
"step": 490
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.84,
|
| 306 |
+
"learning_rate": 4.31418918918919e-05,
|
| 307 |
+
"loss": 0.444,
|
| 308 |
+
"step": 500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.86,
|
| 312 |
+
"learning_rate": 4.280405405405406e-05,
|
| 313 |
+
"loss": 0.4044,
|
| 314 |
+
"step": 510
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.88,
|
| 318 |
+
"learning_rate": 4.246621621621622e-05,
|
| 319 |
+
"loss": 0.3761,
|
| 320 |
+
"step": 520
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.9,
|
| 324 |
+
"learning_rate": 4.212837837837838e-05,
|
| 325 |
+
"loss": 0.3932,
|
| 326 |
+
"step": 530
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.91,
|
| 330 |
+
"learning_rate": 4.179054054054054e-05,
|
| 331 |
+
"loss": 0.382,
|
| 332 |
+
"step": 540
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.93,
|
| 336 |
+
"learning_rate": 4.14527027027027e-05,
|
| 337 |
+
"loss": 0.3858,
|
| 338 |
+
"step": 550
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.95,
|
| 342 |
+
"learning_rate": 4.1114864864864864e-05,
|
| 343 |
+
"loss": 0.4047,
|
| 344 |
+
"step": 560
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.96,
|
| 348 |
+
"learning_rate": 4.077702702702703e-05,
|
| 349 |
+
"loss": 0.4456,
|
| 350 |
+
"step": 570
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.98,
|
| 354 |
+
"learning_rate": 4.0439189189189194e-05,
|
| 355 |
+
"loss": 0.351,
|
| 356 |
+
"step": 580
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 1.0,
|
| 360 |
+
"learning_rate": 4.0101351351351355e-05,
|
| 361 |
+
"loss": 0.3687,
|
| 362 |
+
"step": 590
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.0,
|
| 366 |
+
"eval_accuracy": 0.8888712871287129,
|
| 367 |
+
"eval_loss": 0.4043530225753784,
|
| 368 |
+
"eval_runtime": 35.1515,
|
| 369 |
+
"eval_samples_per_second": 718.32,
|
| 370 |
+
"eval_steps_per_second": 5.633,
|
| 371 |
+
"step": 592
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.01,
|
| 375 |
+
"learning_rate": 3.9763513513513516e-05,
|
| 376 |
+
"loss": 0.3674,
|
| 377 |
+
"step": 600
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 1.03,
|
| 381 |
+
"learning_rate": 3.942567567567568e-05,
|
| 382 |
+
"loss": 0.3655,
|
| 383 |
+
"step": 610
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 1.05,
|
| 387 |
+
"learning_rate": 3.908783783783784e-05,
|
| 388 |
+
"loss": 0.4221,
|
| 389 |
+
"step": 620
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.06,
|
| 393 |
+
"learning_rate": 3.875e-05,
|
| 394 |
+
"loss": 0.3846,
|
| 395 |
+
"step": 630
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.08,
|
| 399 |
+
"learning_rate": 3.841216216216217e-05,
|
| 400 |
+
"loss": 0.4122,
|
| 401 |
+
"step": 640
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.1,
|
| 405 |
+
"learning_rate": 3.807432432432433e-05,
|
| 406 |
+
"loss": 0.4057,
|
| 407 |
+
"step": 650
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 1.11,
|
| 411 |
+
"learning_rate": 3.773648648648649e-05,
|
| 412 |
+
"loss": 0.4538,
|
| 413 |
+
"step": 660
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.13,
|
| 417 |
+
"learning_rate": 3.739864864864865e-05,
|
| 418 |
+
"loss": 0.3738,
|
| 419 |
+
"step": 670
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 1.15,
|
| 423 |
+
"learning_rate": 3.706081081081081e-05,
|
| 424 |
+
"loss": 0.4286,
|
| 425 |
+
"step": 680
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 1.17,
|
| 429 |
+
"learning_rate": 3.6722972972972974e-05,
|
| 430 |
+
"loss": 0.3568,
|
| 431 |
+
"step": 690
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.18,
|
| 435 |
+
"learning_rate": 3.638513513513514e-05,
|
| 436 |
+
"loss": 0.3664,
|
| 437 |
+
"step": 700
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.2,
|
| 441 |
+
"learning_rate": 3.6047297297297304e-05,
|
| 442 |
+
"loss": 0.3662,
|
| 443 |
+
"step": 710
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.22,
|
| 447 |
+
"learning_rate": 3.5709459459459465e-05,
|
| 448 |
+
"loss": 0.4049,
|
| 449 |
+
"step": 720
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 1.23,
|
| 453 |
+
"learning_rate": 3.537162162162162e-05,
|
| 454 |
+
"loss": 0.3396,
|
| 455 |
+
"step": 730
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 1.25,
|
| 459 |
+
"learning_rate": 3.503378378378378e-05,
|
| 460 |
+
"loss": 0.3986,
|
| 461 |
+
"step": 740
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 1.27,
|
| 465 |
+
"learning_rate": 3.469594594594594e-05,
|
| 466 |
+
"loss": 0.4362,
|
| 467 |
+
"step": 750
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 1.28,
|
| 471 |
+
"learning_rate": 3.43581081081081e-05,
|
| 472 |
+
"loss": 0.3804,
|
| 473 |
+
"step": 760
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.3,
|
| 477 |
+
"learning_rate": 3.402027027027027e-05,
|
| 478 |
+
"loss": 0.3586,
|
| 479 |
+
"step": 770
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.32,
|
| 483 |
+
"learning_rate": 3.368243243243243e-05,
|
| 484 |
+
"loss": 0.3433,
|
| 485 |
+
"step": 780
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.33,
|
| 489 |
+
"learning_rate": 3.3344594594594594e-05,
|
| 490 |
+
"loss": 0.4387,
|
| 491 |
+
"step": 790
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 1.35,
|
| 495 |
+
"learning_rate": 3.3006756756756755e-05,
|
| 496 |
+
"loss": 0.3773,
|
| 497 |
+
"step": 800
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 1.37,
|
| 501 |
+
"learning_rate": 3.2668918918918917e-05,
|
| 502 |
+
"loss": 0.3687,
|
| 503 |
+
"step": 810
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 1.39,
|
| 507 |
+
"learning_rate": 3.233108108108108e-05,
|
| 508 |
+
"loss": 0.3975,
|
| 509 |
+
"step": 820
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 1.4,
|
| 513 |
+
"learning_rate": 3.199324324324324e-05,
|
| 514 |
+
"loss": 0.3489,
|
| 515 |
+
"step": 830
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.42,
|
| 519 |
+
"learning_rate": 3.165540540540541e-05,
|
| 520 |
+
"loss": 0.3892,
|
| 521 |
+
"step": 840
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.44,
|
| 525 |
+
"learning_rate": 3.131756756756757e-05,
|
| 526 |
+
"loss": 0.3245,
|
| 527 |
+
"step": 850
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.45,
|
| 531 |
+
"learning_rate": 3.097972972972973e-05,
|
| 532 |
+
"loss": 0.3773,
|
| 533 |
+
"step": 860
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 1.47,
|
| 537 |
+
"learning_rate": 3.064189189189189e-05,
|
| 538 |
+
"loss": 0.377,
|
| 539 |
+
"step": 870
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 1.49,
|
| 543 |
+
"learning_rate": 3.0304054054054052e-05,
|
| 544 |
+
"loss": 0.3985,
|
| 545 |
+
"step": 880
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 1.5,
|
| 549 |
+
"learning_rate": 2.9966216216216217e-05,
|
| 550 |
+
"loss": 0.3336,
|
| 551 |
+
"step": 890
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 1.52,
|
| 555 |
+
"learning_rate": 2.9628378378378378e-05,
|
| 556 |
+
"loss": 0.3807,
|
| 557 |
+
"step": 900
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.54,
|
| 561 |
+
"learning_rate": 2.929054054054054e-05,
|
| 562 |
+
"loss": 0.3827,
|
| 563 |
+
"step": 910
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.55,
|
| 567 |
+
"learning_rate": 2.8952702702702704e-05,
|
| 568 |
+
"loss": 0.3531,
|
| 569 |
+
"step": 920
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.57,
|
| 573 |
+
"learning_rate": 2.8614864864864865e-05,
|
| 574 |
+
"loss": 0.3814,
|
| 575 |
+
"step": 930
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 1.59,
|
| 579 |
+
"learning_rate": 2.8277027027027027e-05,
|
| 580 |
+
"loss": 0.3634,
|
| 581 |
+
"step": 940
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 1.6,
|
| 585 |
+
"learning_rate": 2.7939189189189188e-05,
|
| 586 |
+
"loss": 0.3309,
|
| 587 |
+
"step": 950
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 1.62,
|
| 591 |
+
"learning_rate": 2.7601351351351353e-05,
|
| 592 |
+
"loss": 0.3763,
|
| 593 |
+
"step": 960
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 1.64,
|
| 597 |
+
"learning_rate": 2.7263513513513514e-05,
|
| 598 |
+
"loss": 0.3521,
|
| 599 |
+
"step": 970
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.66,
|
| 603 |
+
"learning_rate": 2.6925675675675675e-05,
|
| 604 |
+
"loss": 0.364,
|
| 605 |
+
"step": 980
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.67,
|
| 609 |
+
"learning_rate": 2.658783783783784e-05,
|
| 610 |
+
"loss": 0.4004,
|
| 611 |
+
"step": 990
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.69,
|
| 615 |
+
"learning_rate": 2.625e-05,
|
| 616 |
+
"loss": 0.3656,
|
| 617 |
+
"step": 1000
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 1.71,
|
| 621 |
+
"learning_rate": 2.5912162162162162e-05,
|
| 622 |
+
"loss": 0.3789,
|
| 623 |
+
"step": 1010
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 1.72,
|
| 627 |
+
"learning_rate": 2.5574324324324324e-05,
|
| 628 |
+
"loss": 0.3441,
|
| 629 |
+
"step": 1020
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 1.74,
|
| 633 |
+
"learning_rate": 2.5236486486486488e-05,
|
| 634 |
+
"loss": 0.3743,
|
| 635 |
+
"step": 1030
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 1.76,
|
| 639 |
+
"learning_rate": 2.489864864864865e-05,
|
| 640 |
+
"loss": 0.3844,
|
| 641 |
+
"step": 1040
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.77,
|
| 645 |
+
"learning_rate": 2.456081081081081e-05,
|
| 646 |
+
"loss": 0.4033,
|
| 647 |
+
"step": 1050
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.79,
|
| 651 |
+
"learning_rate": 2.4222972972972976e-05,
|
| 652 |
+
"loss": 0.3749,
|
| 653 |
+
"step": 1060
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.81,
|
| 657 |
+
"learning_rate": 2.3885135135135137e-05,
|
| 658 |
+
"loss": 0.3956,
|
| 659 |
+
"step": 1070
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 1.82,
|
| 663 |
+
"learning_rate": 2.3547297297297298e-05,
|
| 664 |
+
"loss": 0.3968,
|
| 665 |
+
"step": 1080
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 1.84,
|
| 669 |
+
"learning_rate": 2.3209459459459463e-05,
|
| 670 |
+
"loss": 0.3801,
|
| 671 |
+
"step": 1090
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 1.86,
|
| 675 |
+
"learning_rate": 2.2871621621621624e-05,
|
| 676 |
+
"loss": 0.321,
|
| 677 |
+
"step": 1100
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 1.88,
|
| 681 |
+
"learning_rate": 2.2533783783783785e-05,
|
| 682 |
+
"loss": 0.3457,
|
| 683 |
+
"step": 1110
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.89,
|
| 687 |
+
"learning_rate": 2.2195945945945947e-05,
|
| 688 |
+
"loss": 0.3789,
|
| 689 |
+
"step": 1120
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.91,
|
| 693 |
+
"learning_rate": 2.185810810810811e-05,
|
| 694 |
+
"loss": 0.3738,
|
| 695 |
+
"step": 1130
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.93,
|
| 699 |
+
"learning_rate": 2.152027027027027e-05,
|
| 700 |
+
"loss": 0.3785,
|
| 701 |
+
"step": 1140
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 1.94,
|
| 705 |
+
"learning_rate": 2.118243243243243e-05,
|
| 706 |
+
"loss": 0.329,
|
| 707 |
+
"step": 1150
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 1.96,
|
| 711 |
+
"learning_rate": 2.0844594594594595e-05,
|
| 712 |
+
"loss": 0.3649,
|
| 713 |
+
"step": 1160
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 1.98,
|
| 717 |
+
"learning_rate": 2.0506756756756756e-05,
|
| 718 |
+
"loss": 0.4189,
|
| 719 |
+
"step": 1170
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 1.99,
|
| 723 |
+
"learning_rate": 2.0168918918918918e-05,
|
| 724 |
+
"loss": 0.3422,
|
| 725 |
+
"step": 1180
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 2.0,
|
| 729 |
+
"eval_accuracy": 0.8953267326732673,
|
| 730 |
+
"eval_loss": 0.3911304175853729,
|
| 731 |
+
"eval_runtime": 34.9199,
|
| 732 |
+
"eval_samples_per_second": 723.084,
|
| 733 |
+
"eval_steps_per_second": 5.67,
|
| 734 |
+
"step": 1184
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 2.01,
|
| 738 |
+
"learning_rate": 1.983108108108108e-05,
|
| 739 |
+
"loss": 0.3691,
|
| 740 |
+
"step": 1190
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 2.03,
|
| 744 |
+
"learning_rate": 1.9493243243243244e-05,
|
| 745 |
+
"loss": 0.3039,
|
| 746 |
+
"step": 1200
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 2.04,
|
| 750 |
+
"learning_rate": 1.9155405405405405e-05,
|
| 751 |
+
"loss": 0.307,
|
| 752 |
+
"step": 1210
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 2.06,
|
| 756 |
+
"learning_rate": 1.8817567567567566e-05,
|
| 757 |
+
"loss": 0.3207,
|
| 758 |
+
"step": 1220
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.08,
|
| 762 |
+
"learning_rate": 1.847972972972973e-05,
|
| 763 |
+
"loss": 0.3473,
|
| 764 |
+
"step": 1230
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 2.09,
|
| 768 |
+
"learning_rate": 1.8141891891891892e-05,
|
| 769 |
+
"loss": 0.3155,
|
| 770 |
+
"step": 1240
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 2.11,
|
| 774 |
+
"learning_rate": 1.7804054054054053e-05,
|
| 775 |
+
"loss": 0.3217,
|
| 776 |
+
"step": 1250
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 2.13,
|
| 780 |
+
"learning_rate": 1.7466216216216218e-05,
|
| 781 |
+
"loss": 0.3473,
|
| 782 |
+
"step": 1260
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 2.15,
|
| 786 |
+
"learning_rate": 1.712837837837838e-05,
|
| 787 |
+
"loss": 0.2966,
|
| 788 |
+
"step": 1270
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 2.16,
|
| 792 |
+
"learning_rate": 1.679054054054054e-05,
|
| 793 |
+
"loss": 0.3411,
|
| 794 |
+
"step": 1280
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 2.18,
|
| 798 |
+
"learning_rate": 1.6452702702702702e-05,
|
| 799 |
+
"loss": 0.3876,
|
| 800 |
+
"step": 1290
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.2,
|
| 804 |
+
"learning_rate": 1.6114864864864866e-05,
|
| 805 |
+
"loss": 0.2818,
|
| 806 |
+
"step": 1300
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 2.21,
|
| 810 |
+
"learning_rate": 1.5777027027027028e-05,
|
| 811 |
+
"loss": 0.351,
|
| 812 |
+
"step": 1310
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 2.23,
|
| 816 |
+
"learning_rate": 1.543918918918919e-05,
|
| 817 |
+
"loss": 0.3947,
|
| 818 |
+
"step": 1320
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 2.25,
|
| 822 |
+
"learning_rate": 1.5101351351351352e-05,
|
| 823 |
+
"loss": 0.3157,
|
| 824 |
+
"step": 1330
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 2.26,
|
| 828 |
+
"learning_rate": 1.4763513513513515e-05,
|
| 829 |
+
"loss": 0.4128,
|
| 830 |
+
"step": 1340
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 2.28,
|
| 834 |
+
"learning_rate": 1.4425675675675675e-05,
|
| 835 |
+
"loss": 0.3189,
|
| 836 |
+
"step": 1350
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 2.3,
|
| 840 |
+
"learning_rate": 1.4087837837837838e-05,
|
| 841 |
+
"loss": 0.3791,
|
| 842 |
+
"step": 1360
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.31,
|
| 846 |
+
"learning_rate": 1.375e-05,
|
| 847 |
+
"loss": 0.3299,
|
| 848 |
+
"step": 1370
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 2.33,
|
| 852 |
+
"learning_rate": 1.3412162162162162e-05,
|
| 853 |
+
"loss": 0.3351,
|
| 854 |
+
"step": 1380
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 2.35,
|
| 858 |
+
"learning_rate": 1.3074324324324325e-05,
|
| 859 |
+
"loss": 0.3197,
|
| 860 |
+
"step": 1390
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 2.36,
|
| 864 |
+
"learning_rate": 1.2736486486486486e-05,
|
| 865 |
+
"loss": 0.3191,
|
| 866 |
+
"step": 1400
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 2.38,
|
| 870 |
+
"learning_rate": 1.2398648648648649e-05,
|
| 871 |
+
"loss": 0.3765,
|
| 872 |
+
"step": 1410
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 2.4,
|
| 876 |
+
"learning_rate": 1.2060810810810812e-05,
|
| 877 |
+
"loss": 0.3019,
|
| 878 |
+
"step": 1420
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 2.42,
|
| 882 |
+
"learning_rate": 1.1722972972972973e-05,
|
| 883 |
+
"loss": 0.3662,
|
| 884 |
+
"step": 1430
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.43,
|
| 888 |
+
"learning_rate": 1.1385135135135136e-05,
|
| 889 |
+
"loss": 0.3232,
|
| 890 |
+
"step": 1440
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 2.45,
|
| 894 |
+
"learning_rate": 1.1047297297297297e-05,
|
| 895 |
+
"loss": 0.3149,
|
| 896 |
+
"step": 1450
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 2.47,
|
| 900 |
+
"learning_rate": 1.070945945945946e-05,
|
| 901 |
+
"loss": 0.3516,
|
| 902 |
+
"step": 1460
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 2.48,
|
| 906 |
+
"learning_rate": 1.0371621621621622e-05,
|
| 907 |
+
"loss": 0.3339,
|
| 908 |
+
"step": 1470
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 2.5,
|
| 912 |
+
"learning_rate": 1.0033783783783785e-05,
|
| 913 |
+
"loss": 0.3046,
|
| 914 |
+
"step": 1480
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 2.52,
|
| 918 |
+
"learning_rate": 9.695945945945946e-06,
|
| 919 |
+
"loss": 0.3794,
|
| 920 |
+
"step": 1490
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.53,
|
| 924 |
+
"learning_rate": 9.358108108108107e-06,
|
| 925 |
+
"loss": 0.4017,
|
| 926 |
+
"step": 1500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.55,
|
| 930 |
+
"learning_rate": 9.02027027027027e-06,
|
| 931 |
+
"loss": 0.3851,
|
| 932 |
+
"step": 1510
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 2.57,
|
| 936 |
+
"learning_rate": 8.682432432432431e-06,
|
| 937 |
+
"loss": 0.3449,
|
| 938 |
+
"step": 1520
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 2.58,
|
| 942 |
+
"learning_rate": 8.344594594594594e-06,
|
| 943 |
+
"loss": 0.3558,
|
| 944 |
+
"step": 1530
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 2.6,
|
| 948 |
+
"learning_rate": 8.006756756756757e-06,
|
| 949 |
+
"loss": 0.3176,
|
| 950 |
+
"step": 1540
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 2.62,
|
| 954 |
+
"learning_rate": 7.668918918918919e-06,
|
| 955 |
+
"loss": 0.3305,
|
| 956 |
+
"step": 1550
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 2.64,
|
| 960 |
+
"learning_rate": 7.331081081081082e-06,
|
| 961 |
+
"loss": 0.3171,
|
| 962 |
+
"step": 1560
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.65,
|
| 966 |
+
"learning_rate": 6.993243243243244e-06,
|
| 967 |
+
"loss": 0.332,
|
| 968 |
+
"step": 1570
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.67,
|
| 972 |
+
"learning_rate": 6.655405405405406e-06,
|
| 973 |
+
"loss": 0.2867,
|
| 974 |
+
"step": 1580
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 2.69,
|
| 978 |
+
"learning_rate": 6.317567567567567e-06,
|
| 979 |
+
"loss": 0.3478,
|
| 980 |
+
"step": 1590
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 2.7,
|
| 984 |
+
"learning_rate": 5.979729729729729e-06,
|
| 985 |
+
"loss": 0.3609,
|
| 986 |
+
"step": 1600
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 2.72,
|
| 990 |
+
"learning_rate": 5.6418918918918914e-06,
|
| 991 |
+
"loss": 0.3591,
|
| 992 |
+
"step": 1610
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 2.74,
|
| 996 |
+
"learning_rate": 5.304054054054054e-06,
|
| 997 |
+
"loss": 0.3131,
|
| 998 |
+
"step": 1620
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 2.75,
|
| 1002 |
+
"learning_rate": 4.9662162162162165e-06,
|
| 1003 |
+
"loss": 0.3224,
|
| 1004 |
+
"step": 1630
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.77,
|
| 1008 |
+
"learning_rate": 4.628378378378379e-06,
|
| 1009 |
+
"loss": 0.3849,
|
| 1010 |
+
"step": 1640
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.79,
|
| 1014 |
+
"learning_rate": 4.290540540540541e-06,
|
| 1015 |
+
"loss": 0.4212,
|
| 1016 |
+
"step": 1650
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 2.8,
|
| 1020 |
+
"learning_rate": 3.952702702702702e-06,
|
| 1021 |
+
"loss": 0.3889,
|
| 1022 |
+
"step": 1660
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 2.82,
|
| 1026 |
+
"learning_rate": 3.614864864864865e-06,
|
| 1027 |
+
"loss": 0.3444,
|
| 1028 |
+
"step": 1670
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.84,
|
| 1032 |
+
"learning_rate": 3.277027027027027e-06,
|
| 1033 |
+
"loss": 0.3445,
|
| 1034 |
+
"step": 1680
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 2.85,
|
| 1038 |
+
"learning_rate": 2.9391891891891893e-06,
|
| 1039 |
+
"loss": 0.3322,
|
| 1040 |
+
"step": 1690
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 2.87,
|
| 1044 |
+
"learning_rate": 2.6013513513513514e-06,
|
| 1045 |
+
"loss": 0.3587,
|
| 1046 |
+
"step": 1700
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.89,
|
| 1050 |
+
"learning_rate": 2.263513513513514e-06,
|
| 1051 |
+
"loss": 0.3127,
|
| 1052 |
+
"step": 1710
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.91,
|
| 1056 |
+
"learning_rate": 1.9256756756756756e-06,
|
| 1057 |
+
"loss": 0.3552,
|
| 1058 |
+
"step": 1720
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 2.92,
|
| 1062 |
+
"learning_rate": 1.587837837837838e-06,
|
| 1063 |
+
"loss": 0.3377,
|
| 1064 |
+
"step": 1730
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 2.94,
|
| 1068 |
+
"learning_rate": 1.2499999999999999e-06,
|
| 1069 |
+
"loss": 0.3269,
|
| 1070 |
+
"step": 1740
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.96,
|
| 1074 |
+
"learning_rate": 9.121621621621622e-07,
|
| 1075 |
+
"loss": 0.3098,
|
| 1076 |
+
"step": 1750
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 2.97,
|
| 1080 |
+
"learning_rate": 5.743243243243243e-07,
|
| 1081 |
+
"loss": 0.3506,
|
| 1082 |
+
"step": 1760
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 2.99,
|
| 1086 |
+
"learning_rate": 2.3648648648648647e-07,
|
| 1087 |
+
"loss": 0.3808,
|
| 1088 |
+
"step": 1770
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 3.0,
|
| 1092 |
+
"eval_accuracy": 0.8971089108910891,
|
| 1093 |
+
"eval_loss": 0.38559553027153015,
|
| 1094 |
+
"eval_runtime": 33.8726,
|
| 1095 |
+
"eval_samples_per_second": 745.441,
|
| 1096 |
+
"eval_steps_per_second": 5.845,
|
| 1097 |
+
"step": 1776
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 3.0,
|
| 1101 |
+
"step": 1776,
|
| 1102 |
+
"total_flos": 1.76256801415296e+19,
|
| 1103 |
+
"train_loss": 0.3829323179549999,
|
| 1104 |
+
"train_runtime": 916.6333,
|
| 1105 |
+
"train_samples_per_second": 247.918,
|
| 1106 |
+
"train_steps_per_second": 1.938
|
| 1107 |
+
}
|
| 1108 |
+
],
|
| 1109 |
+
"max_steps": 1776,
|
| 1110 |
+
"num_train_epochs": 3,
|
| 1111 |
+
"total_flos": 1.76256801415296e+19,
|
| 1112 |
+
"trial_name": null,
|
| 1113 |
+
"trial_params": null
|
| 1114 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99375d127eeeb0528bc62baffa4c5e80febdd57a4666587ea2223f1d43d64cf0
|
| 3 |
+
size 3771
|