Commit
•
38513ec
1
Parent(s):
e956955
Subir modelo definitivo
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- README.md +72 -3
- checkpoint-100/config.json +85 -0
- checkpoint-100/model.safetensors +3 -0
- checkpoint-100/optimizer.pt +3 -0
- checkpoint-100/rng_state.pth +3 -0
- checkpoint-100/scheduler.pt +3 -0
- checkpoint-100/trainer_state.json +132 -0
- checkpoint-100/training_args.bin +3 -0
- checkpoint-103/config.json +85 -0
- checkpoint-103/model.safetensors +3 -0
- checkpoint-103/optimizer.pt +3 -0
- checkpoint-103/rng_state.pth +3 -0
- checkpoint-103/scheduler.pt +3 -0
- checkpoint-103/trainer_state.json +907 -0
- checkpoint-103/training_args.bin +3 -0
- checkpoint-108/config.json +85 -0
- checkpoint-108/model.safetensors +3 -0
- checkpoint-108/optimizer.pt +3 -0
- checkpoint-108/rng_state.pth +3 -0
- checkpoint-108/scheduler.pt +3 -0
- checkpoint-108/trainer_state.json +945 -0
- checkpoint-108/training_args.bin +3 -0
- checkpoint-112/config.json +85 -0
- checkpoint-112/model.safetensors +3 -0
- checkpoint-112/optimizer.pt +3 -0
- checkpoint-112/rng_state.pth +3 -0
- checkpoint-112/scheduler.pt +3 -0
- checkpoint-112/trainer_state.json +983 -0
- checkpoint-112/training_args.bin +3 -0
- checkpoint-117/config.json +85 -0
- checkpoint-117/model.safetensors +3 -0
- checkpoint-117/optimizer.pt +3 -0
- checkpoint-117/rng_state.pth +3 -0
- checkpoint-117/scheduler.pt +3 -0
- checkpoint-117/trainer_state.json +1021 -0
- checkpoint-117/training_args.bin +3 -0
- checkpoint-121/config.json +85 -0
- checkpoint-121/model.safetensors +3 -0
- checkpoint-121/optimizer.pt +3 -0
- checkpoint-121/rng_state.pth +3 -0
- checkpoint-121/scheduler.pt +3 -0
- checkpoint-121/trainer_state.json +1059 -0
- checkpoint-121/training_args.bin +3 -0
- checkpoint-126/config.json +85 -0
- checkpoint-126/model.safetensors +3 -0
- checkpoint-126/optimizer.pt +3 -0
- checkpoint-126/rng_state.pth +3 -0
- checkpoint-126/scheduler.pt +3 -0
- checkpoint-126/trainer_state.json +1097 -0
- checkpoint-126/training_args.bin +3 -0
README.md
CHANGED
|
@@ -1,3 +1,72 @@
|
|
| 1 |
-
---
|
| 2 |
-
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: transformers
|
| 3 |
+
license: apache-2.0
|
| 4 |
+
base_model: ntu-spml/distilhubert
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
metrics:
|
| 8 |
+
- accuracy
|
| 9 |
+
- f1
|
| 10 |
+
- precision
|
| 11 |
+
- recall
|
| 12 |
+
model-index:
|
| 13 |
+
- name: distilhubert-finetuned-mixed-data
|
| 14 |
+
results: []
|
| 15 |
+
---
|
| 16 |
+
|
| 17 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 18 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 19 |
+
|
| 20 |
+
# distilhubert-finetuned-mixed-data
|
| 21 |
+
|
| 22 |
+
This model is a fine-tuned version of [ntu-spml/distilhubert](https://huggingface.co/ntu-spml/distilhubert) on an unknown dataset.
|
| 23 |
+
It achieves the following results on the evaluation set:
|
| 24 |
+
- Loss: 0.8806
|
| 25 |
+
- Accuracy: 0.7912
|
| 26 |
+
- F1: 0.7772
|
| 27 |
+
- Precision: 0.8022
|
| 28 |
+
- Recall: 0.7912
|
| 29 |
+
- Confusion Matrix: [[59, 1, 1, 2], [20, 35, 22, 0], [2, 7, 68, 0], [2, 0, 0, 54]]
|
| 30 |
+
|
| 31 |
+
## Model description
|
| 32 |
+
|
| 33 |
+
More information needed
|
| 34 |
+
|
| 35 |
+
## Intended uses & limitations
|
| 36 |
+
|
| 37 |
+
More information needed
|
| 38 |
+
|
| 39 |
+
## Training and evaluation data
|
| 40 |
+
|
| 41 |
+
More information needed
|
| 42 |
+
|
| 43 |
+
## Training procedure
|
| 44 |
+
|
| 45 |
+
### Training hyperparameters
|
| 46 |
+
|
| 47 |
+
The following hyperparameters were used during training:
|
| 48 |
+
- learning_rate: 0.0005
|
| 49 |
+
- train_batch_size: 128
|
| 50 |
+
- eval_batch_size: 128
|
| 51 |
+
- seed: 123
|
| 52 |
+
- gradient_accumulation_steps: 2
|
| 53 |
+
- total_train_batch_size: 256
|
| 54 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 55 |
+
- lr_scheduler_type: cosine_with_restarts
|
| 56 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 57 |
+
- num_epochs: 40
|
| 58 |
+
- mixed_precision_training: Native AMP
|
| 59 |
+
- label_smoothing_factor: 0.1
|
| 60 |
+
|
| 61 |
+
### Training results
|
| 62 |
+
|
| 63 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall | Confusion Matrix |
|
| 64 |
+
|:-------------:|:-------:|:----:|:---------------:|:--------:|:------:|:---------:|:------:|:--------------------------------------------------------------:|
|
| 65 |
+
| 0.4221 | 22.2222 | 100 | 0.8806 | 0.7912 | 0.7772 | 0.8022 | 0.7912 | [[59, 1, 1, 2], [20, 35, 22, 0], [2, 7, 68, 0], [2, 0, 0, 54]] |
|
| 66 |
+
|
| 67 |
+
|
| 68 |
+
### Framework versions
|
| 69 |
+
|
| 70 |
+
- Transformers 4.44.2
|
| 71 |
+
- Pytorch 2.4.1+cu121
|
| 72 |
+
- Tokenizers 0.19.1
|
checkpoint-100/config.json
ADDED
|
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ntu-spml/distilhubert",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"HubertForSequenceClassification"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 1,
|
| 10 |
+
"classifier_proj_size": 256,
|
| 11 |
+
"conv_bias": false,
|
| 12 |
+
"conv_dim": [
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512
|
| 20 |
+
],
|
| 21 |
+
"conv_kernel": [
|
| 22 |
+
10,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
3,
|
| 27 |
+
2,
|
| 28 |
+
2
|
| 29 |
+
],
|
| 30 |
+
"conv_stride": [
|
| 31 |
+
5,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2,
|
| 37 |
+
2
|
| 38 |
+
],
|
| 39 |
+
"ctc_loss_reduction": "sum",
|
| 40 |
+
"ctc_zero_infinity": false,
|
| 41 |
+
"do_stable_layer_norm": false,
|
| 42 |
+
"eos_token_id": 2,
|
| 43 |
+
"feat_extract_activation": "gelu",
|
| 44 |
+
"feat_extract_norm": "group",
|
| 45 |
+
"feat_proj_dropout": 0.0,
|
| 46 |
+
"feat_proj_layer_norm": false,
|
| 47 |
+
"final_dropout": 0.0,
|
| 48 |
+
"finetuning_task": "audio-classification",
|
| 49 |
+
"hidden_act": "gelu",
|
| 50 |
+
"hidden_dropout": 0.1,
|
| 51 |
+
"hidden_size": 768,
|
| 52 |
+
"id2label": {
|
| 53 |
+
"0": "1s_normal",
|
| 54 |
+
"1": "1s_pain",
|
| 55 |
+
"2": "1s_hunger",
|
| 56 |
+
"3": "1s_asphyxia"
|
| 57 |
+
},
|
| 58 |
+
"initializer_range": 0.02,
|
| 59 |
+
"intermediate_size": 3072,
|
| 60 |
+
"label2id": {
|
| 61 |
+
"LABEL_0": 0,
|
| 62 |
+
"LABEL_1": 1,
|
| 63 |
+
"LABEL_2": 2,
|
| 64 |
+
"LABEL_3": 3
|
| 65 |
+
},
|
| 66 |
+
"layer_norm_eps": 1e-05,
|
| 67 |
+
"layerdrop": 0.0,
|
| 68 |
+
"mask_feature_length": 10,
|
| 69 |
+
"mask_feature_min_masks": 0,
|
| 70 |
+
"mask_feature_prob": 0.0,
|
| 71 |
+
"mask_time_length": 10,
|
| 72 |
+
"mask_time_min_masks": 2,
|
| 73 |
+
"mask_time_prob": 0.05,
|
| 74 |
+
"model_type": "hubert",
|
| 75 |
+
"num_attention_heads": 12,
|
| 76 |
+
"num_conv_pos_embedding_groups": 16,
|
| 77 |
+
"num_conv_pos_embeddings": 128,
|
| 78 |
+
"num_feat_extract_layers": 7,
|
| 79 |
+
"num_hidden_layers": 2,
|
| 80 |
+
"pad_token_id": 0,
|
| 81 |
+
"torch_dtype": "float32",
|
| 82 |
+
"transformers_version": "4.44.2",
|
| 83 |
+
"use_weighted_layer_sum": false,
|
| 84 |
+
"vocab_size": 32
|
| 85 |
+
}
|
checkpoint-100/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c2fd7ed3d85f5c5508fdba113cb218b717922e4e2b452800b8b2f7e86ce40400
|
| 3 |
+
size 94765560
|
checkpoint-100/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2d52014472918afa1cd037f4d14642147157140d55641d5cd90d54a9b6fd5639
|
| 3 |
+
size 189556666
|
checkpoint-100/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1571753b8b87733ddd5f0d1f9dee2fdc0442f683f5ae2636dd01c78bb407f35a
|
| 3 |
+
size 14308
|
checkpoint-100/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d2860514b4d60fa49e8171cee670c64927b3bd5b60ab32b25ae055e67c087f8
|
| 3 |
+
size 1064
|
checkpoint-100/trainer_state.json
ADDED
|
@@ -0,0 +1,132 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8240470006184293,
|
| 3 |
+
"best_model_checkpoint": "distilhubert-finetuned-mixed-data/checkpoint-100",
|
| 4 |
+
"epoch": 22.22222222222222,
|
| 5 |
+
"eval_steps": 50,
|
| 6 |
+
"global_step": 100,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 11.11111111111111,
|
| 13 |
+
"grad_norm": 3.250781297683716,
|
| 14 |
+
"learning_rate": 0.00027532317171194046,
|
| 15 |
+
"loss": 0.8288,
|
| 16 |
+
"step": 50
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 11.11111111111111,
|
| 20 |
+
"eval_accuracy": 0.8131868131868132,
|
| 21 |
+
"eval_confusion_matrix": [
|
| 22 |
+
[
|
| 23 |
+
60,
|
| 24 |
+
11,
|
| 25 |
+
1,
|
| 26 |
+
7
|
| 27 |
+
],
|
| 28 |
+
[
|
| 29 |
+
9,
|
| 30 |
+
40,
|
| 31 |
+
17,
|
| 32 |
+
0
|
| 33 |
+
],
|
| 34 |
+
[
|
| 35 |
+
0,
|
| 36 |
+
6,
|
| 37 |
+
63,
|
| 38 |
+
0
|
| 39 |
+
],
|
| 40 |
+
[
|
| 41 |
+
0,
|
| 42 |
+
0,
|
| 43 |
+
0,
|
| 44 |
+
59
|
| 45 |
+
]
|
| 46 |
+
],
|
| 47 |
+
"eval_f1": 0.8081937118034679,
|
| 48 |
+
"eval_loss": 0.7091822624206543,
|
| 49 |
+
"eval_precision": 0.8110642218193705,
|
| 50 |
+
"eval_recall": 0.8131868131868132,
|
| 51 |
+
"eval_runtime": 3.0071,
|
| 52 |
+
"eval_samples_per_second": 90.786,
|
| 53 |
+
"eval_steps_per_second": 0.998,
|
| 54 |
+
"step": 50
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 22.22222222222222,
|
| 58 |
+
"grad_norm": 2.9194233417510986,
|
| 59 |
+
"learning_rate": 0.00015290869976577364,
|
| 60 |
+
"loss": 0.4125,
|
| 61 |
+
"step": 100
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 22.22222222222222,
|
| 65 |
+
"eval_accuracy": 0.8278388278388278,
|
| 66 |
+
"eval_confusion_matrix": [
|
| 67 |
+
[
|
| 68 |
+
63,
|
| 69 |
+
7,
|
| 70 |
+
2,
|
| 71 |
+
7
|
| 72 |
+
],
|
| 73 |
+
[
|
| 74 |
+
5,
|
| 75 |
+
41,
|
| 76 |
+
20,
|
| 77 |
+
0
|
| 78 |
+
],
|
| 79 |
+
[
|
| 80 |
+
0,
|
| 81 |
+
6,
|
| 82 |
+
63,
|
| 83 |
+
0
|
| 84 |
+
],
|
| 85 |
+
[
|
| 86 |
+
0,
|
| 87 |
+
0,
|
| 88 |
+
0,
|
| 89 |
+
59
|
| 90 |
+
]
|
| 91 |
+
],
|
| 92 |
+
"eval_f1": 0.8240470006184293,
|
| 93 |
+
"eval_loss": 0.8097973465919495,
|
| 94 |
+
"eval_precision": 0.8321827410062705,
|
| 95 |
+
"eval_recall": 0.8278388278388278,
|
| 96 |
+
"eval_runtime": 3.0011,
|
| 97 |
+
"eval_samples_per_second": 90.968,
|
| 98 |
+
"eval_steps_per_second": 1.0,
|
| 99 |
+
"step": 100
|
| 100 |
+
}
|
| 101 |
+
],
|
| 102 |
+
"logging_steps": 50,
|
| 103 |
+
"max_steps": 180,
|
| 104 |
+
"num_input_tokens_seen": 0,
|
| 105 |
+
"num_train_epochs": 45,
|
| 106 |
+
"save_steps": 50,
|
| 107 |
+
"stateful_callbacks": {
|
| 108 |
+
"EarlyStoppingCallback": {
|
| 109 |
+
"args": {
|
| 110 |
+
"early_stopping_patience": 5,
|
| 111 |
+
"early_stopping_threshold": 0.001
|
| 112 |
+
},
|
| 113 |
+
"attributes": {
|
| 114 |
+
"early_stopping_patience_counter": 0
|
| 115 |
+
}
|
| 116 |
+
},
|
| 117 |
+
"TrainerControl": {
|
| 118 |
+
"args": {
|
| 119 |
+
"should_epoch_stop": false,
|
| 120 |
+
"should_evaluate": false,
|
| 121 |
+
"should_log": false,
|
| 122 |
+
"should_save": true,
|
| 123 |
+
"should_training_stop": false
|
| 124 |
+
},
|
| 125 |
+
"attributes": {}
|
| 126 |
+
}
|
| 127 |
+
},
|
| 128 |
+
"total_flos": 5.511815490816e+16,
|
| 129 |
+
"train_batch_size": 128,
|
| 130 |
+
"trial_name": null,
|
| 131 |
+
"trial_params": null
|
| 132 |
+
}
|
checkpoint-100/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b1a597595ff76a9a22ed372d7da43b8f98eac449e4a9a4f3f55925bca85cb26c
|
| 3 |
+
size 5240
|
checkpoint-103/config.json
ADDED
|
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ntu-spml/distilhubert",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"HubertForSequenceClassification"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 1,
|
| 10 |
+
"classifier_proj_size": 256,
|
| 11 |
+
"conv_bias": false,
|
| 12 |
+
"conv_dim": [
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512
|
| 20 |
+
],
|
| 21 |
+
"conv_kernel": [
|
| 22 |
+
10,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
3,
|
| 27 |
+
2,
|
| 28 |
+
2
|
| 29 |
+
],
|
| 30 |
+
"conv_stride": [
|
| 31 |
+
5,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2,
|
| 37 |
+
2
|
| 38 |
+
],
|
| 39 |
+
"ctc_loss_reduction": "sum",
|
| 40 |
+
"ctc_zero_infinity": false,
|
| 41 |
+
"do_stable_layer_norm": false,
|
| 42 |
+
"eos_token_id": 2,
|
| 43 |
+
"feat_extract_activation": "gelu",
|
| 44 |
+
"feat_extract_norm": "group",
|
| 45 |
+
"feat_proj_dropout": 0.0,
|
| 46 |
+
"feat_proj_layer_norm": false,
|
| 47 |
+
"final_dropout": 0.0,
|
| 48 |
+
"finetuning_task": "audio-classification",
|
| 49 |
+
"hidden_act": "gelu",
|
| 50 |
+
"hidden_dropout": 0.1,
|
| 51 |
+
"hidden_size": 768,
|
| 52 |
+
"id2label": {
|
| 53 |
+
"0": "1s_normal",
|
| 54 |
+
"1": "1s_pain",
|
| 55 |
+
"2": "1s_hunger",
|
| 56 |
+
"3": "1s_asphyxia"
|
| 57 |
+
},
|
| 58 |
+
"initializer_range": 0.02,
|
| 59 |
+
"intermediate_size": 3072,
|
| 60 |
+
"label2id": {
|
| 61 |
+
"LABEL_0": 0,
|
| 62 |
+
"LABEL_1": 1,
|
| 63 |
+
"LABEL_2": 2,
|
| 64 |
+
"LABEL_3": 3
|
| 65 |
+
},
|
| 66 |
+
"layer_norm_eps": 1e-05,
|
| 67 |
+
"layerdrop": 0.0,
|
| 68 |
+
"mask_feature_length": 10,
|
| 69 |
+
"mask_feature_min_masks": 0,
|
| 70 |
+
"mask_feature_prob": 0.0,
|
| 71 |
+
"mask_time_length": 10,
|
| 72 |
+
"mask_time_min_masks": 2,
|
| 73 |
+
"mask_time_prob": 0.05,
|
| 74 |
+
"model_type": "hubert",
|
| 75 |
+
"num_attention_heads": 12,
|
| 76 |
+
"num_conv_pos_embedding_groups": 16,
|
| 77 |
+
"num_conv_pos_embeddings": 128,
|
| 78 |
+
"num_feat_extract_layers": 7,
|
| 79 |
+
"num_hidden_layers": 2,
|
| 80 |
+
"pad_token_id": 0,
|
| 81 |
+
"torch_dtype": "float32",
|
| 82 |
+
"transformers_version": "4.44.2",
|
| 83 |
+
"use_weighted_layer_sum": false,
|
| 84 |
+
"vocab_size": 32
|
| 85 |
+
}
|
checkpoint-103/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4f7d6ad497eba26b3750aa84d3f2f51a6c75d441de28d6e5dffc2ba791ef85cb
|
| 3 |
+
size 94765560
|
checkpoint-103/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2e42ad9a6b8f81c14e7bf08d7fcd25f21c4525b5b243c2ac5dd8007a8ef28d51
|
| 3 |
+
size 189556666
|
checkpoint-103/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:758935c8432355b6f68b49d4629c80a8b5bf68f490598468627160be3b03f9d5
|
| 3 |
+
size 14308
|
checkpoint-103/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:76735059535a9df7daab319ea81a27335a904e3b4dea492d74215e6e4524b5f3
|
| 3 |
+
size 1064
|
checkpoint-103/trainer_state.json
ADDED
|
@@ -0,0 +1,907 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8627769756077204,
|
| 3 |
+
"best_model_checkpoint": "distilhubert-finetuned-mixed-data/checkpoint-81",
|
| 4 |
+
"epoch": 22.88888888888889,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 103,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.8888888888888888,
|
| 13 |
+
"eval_accuracy": 0.63003663003663,
|
| 14 |
+
"eval_confusion_matrix": [
|
| 15 |
+
[
|
| 16 |
+
53,
|
| 17 |
+
0,
|
| 18 |
+
16,
|
| 19 |
+
3
|
| 20 |
+
],
|
| 21 |
+
[
|
| 22 |
+
28,
|
| 23 |
+
1,
|
| 24 |
+
31,
|
| 25 |
+
0
|
| 26 |
+
],
|
| 27 |
+
[
|
| 28 |
+
13,
|
| 29 |
+
0,
|
| 30 |
+
62,
|
| 31 |
+
1
|
| 32 |
+
],
|
| 33 |
+
[
|
| 34 |
+
3,
|
| 35 |
+
0,
|
| 36 |
+
6,
|
| 37 |
+
56
|
| 38 |
+
]
|
| 39 |
+
],
|
| 40 |
+
"eval_f1": 0.566693372031096,
|
| 41 |
+
"eval_loss": 1.31327486038208,
|
| 42 |
+
"eval_precision": 0.7361933549293478,
|
| 43 |
+
"eval_recall": 0.63003663003663,
|
| 44 |
+
"eval_runtime": 2.8436,
|
| 45 |
+
"eval_samples_per_second": 96.005,
|
| 46 |
+
"eval_steps_per_second": 1.055,
|
| 47 |
+
"step": 4
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 2.0,
|
| 51 |
+
"eval_accuracy": 0.6593406593406593,
|
| 52 |
+
"eval_confusion_matrix": [
|
| 53 |
+
[
|
| 54 |
+
41,
|
| 55 |
+
10,
|
| 56 |
+
16,
|
| 57 |
+
5
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
21,
|
| 61 |
+
1,
|
| 62 |
+
38,
|
| 63 |
+
0
|
| 64 |
+
],
|
| 65 |
+
[
|
| 66 |
+
1,
|
| 67 |
+
2,
|
| 68 |
+
73,
|
| 69 |
+
0
|
| 70 |
+
],
|
| 71 |
+
[
|
| 72 |
+
0,
|
| 73 |
+
0,
|
| 74 |
+
0,
|
| 75 |
+
65
|
| 76 |
+
]
|
| 77 |
+
],
|
| 78 |
+
"eval_f1": 0.595713773130153,
|
| 79 |
+
"eval_loss": 0.9242589473724365,
|
| 80 |
+
"eval_precision": 0.5696509512811321,
|
| 81 |
+
"eval_recall": 0.6593406593406593,
|
| 82 |
+
"eval_runtime": 2.8668,
|
| 83 |
+
"eval_samples_per_second": 95.227,
|
| 84 |
+
"eval_steps_per_second": 1.046,
|
| 85 |
+
"step": 9
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 2.888888888888889,
|
| 89 |
+
"eval_accuracy": 0.717948717948718,
|
| 90 |
+
"eval_confusion_matrix": [
|
| 91 |
+
[
|
| 92 |
+
52,
|
| 93 |
+
8,
|
| 94 |
+
7,
|
| 95 |
+
5
|
| 96 |
+
],
|
| 97 |
+
[
|
| 98 |
+
18,
|
| 99 |
+
5,
|
| 100 |
+
37,
|
| 101 |
+
0
|
| 102 |
+
],
|
| 103 |
+
[
|
| 104 |
+
1,
|
| 105 |
+
0,
|
| 106 |
+
75,
|
| 107 |
+
0
|
| 108 |
+
],
|
| 109 |
+
[
|
| 110 |
+
1,
|
| 111 |
+
0,
|
| 112 |
+
0,
|
| 113 |
+
64
|
| 114 |
+
]
|
| 115 |
+
],
|
| 116 |
+
"eval_f1": 0.6621621567125869,
|
| 117 |
+
"eval_loss": 0.8232662081718445,
|
| 118 |
+
"eval_precision": 0.671303801513745,
|
| 119 |
+
"eval_recall": 0.717948717948718,
|
| 120 |
+
"eval_runtime": 2.8711,
|
| 121 |
+
"eval_samples_per_second": 95.084,
|
| 122 |
+
"eval_steps_per_second": 1.045,
|
| 123 |
+
"step": 13
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 4.0,
|
| 127 |
+
"eval_accuracy": 0.6959706959706959,
|
| 128 |
+
"eval_confusion_matrix": [
|
| 129 |
+
[
|
| 130 |
+
45,
|
| 131 |
+
9,
|
| 132 |
+
13,
|
| 133 |
+
5
|
| 134 |
+
],
|
| 135 |
+
[
|
| 136 |
+
13,
|
| 137 |
+
5,
|
| 138 |
+
42,
|
| 139 |
+
0
|
| 140 |
+
],
|
| 141 |
+
[
|
| 142 |
+
1,
|
| 143 |
+
0,
|
| 144 |
+
75,
|
| 145 |
+
0
|
| 146 |
+
],
|
| 147 |
+
[
|
| 148 |
+
0,
|
| 149 |
+
0,
|
| 150 |
+
0,
|
| 151 |
+
65
|
| 152 |
+
]
|
| 153 |
+
],
|
| 154 |
+
"eval_f1": 0.6428796900399459,
|
| 155 |
+
"eval_loss": 0.951453447341919,
|
| 156 |
+
"eval_precision": 0.6613447705829899,
|
| 157 |
+
"eval_recall": 0.6959706959706959,
|
| 158 |
+
"eval_runtime": 2.8813,
|
| 159 |
+
"eval_samples_per_second": 94.75,
|
| 160 |
+
"eval_steps_per_second": 1.041,
|
| 161 |
+
"step": 18
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 4.888888888888889,
|
| 165 |
+
"eval_accuracy": 0.6666666666666666,
|
| 166 |
+
"eval_confusion_matrix": [
|
| 167 |
+
[
|
| 168 |
+
16,
|
| 169 |
+
49,
|
| 170 |
+
3,
|
| 171 |
+
4
|
| 172 |
+
],
|
| 173 |
+
[
|
| 174 |
+
1,
|
| 175 |
+
26,
|
| 176 |
+
33,
|
| 177 |
+
0
|
| 178 |
+
],
|
| 179 |
+
[
|
| 180 |
+
0,
|
| 181 |
+
1,
|
| 182 |
+
75,
|
| 183 |
+
0
|
| 184 |
+
],
|
| 185 |
+
[
|
| 186 |
+
0,
|
| 187 |
+
0,
|
| 188 |
+
0,
|
| 189 |
+
65
|
| 190 |
+
]
|
| 191 |
+
],
|
| 192 |
+
"eval_f1": 0.6331541584523206,
|
| 193 |
+
"eval_loss": 1.0080645084381104,
|
| 194 |
+
"eval_precision": 0.7358031394913882,
|
| 195 |
+
"eval_recall": 0.6666666666666666,
|
| 196 |
+
"eval_runtime": 2.9108,
|
| 197 |
+
"eval_samples_per_second": 93.789,
|
| 198 |
+
"eval_steps_per_second": 1.031,
|
| 199 |
+
"step": 22
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 6.0,
|
| 203 |
+
"eval_accuracy": 0.652014652014652,
|
| 204 |
+
"eval_confusion_matrix": [
|
| 205 |
+
[
|
| 206 |
+
66,
|
| 207 |
+
0,
|
| 208 |
+
0,
|
| 209 |
+
6
|
| 210 |
+
],
|
| 211 |
+
[
|
| 212 |
+
36,
|
| 213 |
+
16,
|
| 214 |
+
8,
|
| 215 |
+
0
|
| 216 |
+
],
|
| 217 |
+
[
|
| 218 |
+
10,
|
| 219 |
+
35,
|
| 220 |
+
31,
|
| 221 |
+
0
|
| 222 |
+
],
|
| 223 |
+
[
|
| 224 |
+
0,
|
| 225 |
+
0,
|
| 226 |
+
0,
|
| 227 |
+
65
|
| 228 |
+
]
|
| 229 |
+
],
|
| 230 |
+
"eval_f1": 0.6302407955860642,
|
| 231 |
+
"eval_loss": 0.9598046541213989,
|
| 232 |
+
"eval_precision": 0.6636241740077424,
|
| 233 |
+
"eval_recall": 0.652014652014652,
|
| 234 |
+
"eval_runtime": 2.9773,
|
| 235 |
+
"eval_samples_per_second": 91.693,
|
| 236 |
+
"eval_steps_per_second": 1.008,
|
| 237 |
+
"step": 27
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 6.888888888888889,
|
| 241 |
+
"eval_accuracy": 0.7692307692307693,
|
| 242 |
+
"eval_confusion_matrix": [
|
| 243 |
+
[
|
| 244 |
+
66,
|
| 245 |
+
0,
|
| 246 |
+
0,
|
| 247 |
+
6
|
| 248 |
+
],
|
| 249 |
+
[
|
| 250 |
+
29,
|
| 251 |
+
21,
|
| 252 |
+
10,
|
| 253 |
+
0
|
| 254 |
+
],
|
| 255 |
+
[
|
| 256 |
+
3,
|
| 257 |
+
15,
|
| 258 |
+
58,
|
| 259 |
+
0
|
| 260 |
+
],
|
| 261 |
+
[
|
| 262 |
+
0,
|
| 263 |
+
0,
|
| 264 |
+
0,
|
| 265 |
+
65
|
| 266 |
+
]
|
| 267 |
+
],
|
| 268 |
+
"eval_f1": 0.7527855586679116,
|
| 269 |
+
"eval_loss": 0.7898163795471191,
|
| 270 |
+
"eval_precision": 0.7612467677056892,
|
| 271 |
+
"eval_recall": 0.7692307692307693,
|
| 272 |
+
"eval_runtime": 2.8918,
|
| 273 |
+
"eval_samples_per_second": 94.404,
|
| 274 |
+
"eval_steps_per_second": 1.037,
|
| 275 |
+
"step": 31
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 8.0,
|
| 279 |
+
"eval_accuracy": 0.7802197802197802,
|
| 280 |
+
"eval_confusion_matrix": [
|
| 281 |
+
[
|
| 282 |
+
66,
|
| 283 |
+
0,
|
| 284 |
+
0,
|
| 285 |
+
6
|
| 286 |
+
],
|
| 287 |
+
[
|
| 288 |
+
25,
|
| 289 |
+
15,
|
| 290 |
+
19,
|
| 291 |
+
1
|
| 292 |
+
],
|
| 293 |
+
[
|
| 294 |
+
1,
|
| 295 |
+
6,
|
| 296 |
+
68,
|
| 297 |
+
1
|
| 298 |
+
],
|
| 299 |
+
[
|
| 300 |
+
1,
|
| 301 |
+
0,
|
| 302 |
+
0,
|
| 303 |
+
64
|
| 304 |
+
]
|
| 305 |
+
],
|
| 306 |
+
"eval_f1": 0.7471179200524057,
|
| 307 |
+
"eval_loss": 0.7337484359741211,
|
| 308 |
+
"eval_precision": 0.7733845922309294,
|
| 309 |
+
"eval_recall": 0.7802197802197802,
|
| 310 |
+
"eval_runtime": 2.9006,
|
| 311 |
+
"eval_samples_per_second": 94.119,
|
| 312 |
+
"eval_steps_per_second": 1.034,
|
| 313 |
+
"step": 36
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 8.88888888888889,
|
| 317 |
+
"eval_accuracy": 0.7912087912087912,
|
| 318 |
+
"eval_confusion_matrix": [
|
| 319 |
+
[
|
| 320 |
+
57,
|
| 321 |
+
8,
|
| 322 |
+
3,
|
| 323 |
+
4
|
| 324 |
+
],
|
| 325 |
+
[
|
| 326 |
+
13,
|
| 327 |
+
24,
|
| 328 |
+
23,
|
| 329 |
+
0
|
| 330 |
+
],
|
| 331 |
+
[
|
| 332 |
+
0,
|
| 333 |
+
5,
|
| 334 |
+
71,
|
| 335 |
+
0
|
| 336 |
+
],
|
| 337 |
+
[
|
| 338 |
+
1,
|
| 339 |
+
0,
|
| 340 |
+
0,
|
| 341 |
+
64
|
| 342 |
+
]
|
| 343 |
+
],
|
| 344 |
+
"eval_f1": 0.7766564722186922,
|
| 345 |
+
"eval_loss": 0.7148727178573608,
|
| 346 |
+
"eval_precision": 0.7821504483074875,
|
| 347 |
+
"eval_recall": 0.7912087912087912,
|
| 348 |
+
"eval_runtime": 2.8866,
|
| 349 |
+
"eval_samples_per_second": 94.576,
|
| 350 |
+
"eval_steps_per_second": 1.039,
|
| 351 |
+
"step": 40
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 10.0,
|
| 355 |
+
"eval_accuracy": 0.8168498168498168,
|
| 356 |
+
"eval_confusion_matrix": [
|
| 357 |
+
[
|
| 358 |
+
51,
|
| 359 |
+
8,
|
| 360 |
+
1,
|
| 361 |
+
12
|
| 362 |
+
],
|
| 363 |
+
[
|
| 364 |
+
10,
|
| 365 |
+
36,
|
| 366 |
+
14,
|
| 367 |
+
0
|
| 368 |
+
],
|
| 369 |
+
[
|
| 370 |
+
0,
|
| 371 |
+
5,
|
| 372 |
+
71,
|
| 373 |
+
0
|
| 374 |
+
],
|
| 375 |
+
[
|
| 376 |
+
0,
|
| 377 |
+
0,
|
| 378 |
+
0,
|
| 379 |
+
65
|
| 380 |
+
]
|
| 381 |
+
],
|
| 382 |
+
"eval_f1": 0.8094336953840884,
|
| 383 |
+
"eval_loss": 0.7574812173843384,
|
| 384 |
+
"eval_precision": 0.8127936625684181,
|
| 385 |
+
"eval_recall": 0.8168498168498168,
|
| 386 |
+
"eval_runtime": 2.8967,
|
| 387 |
+
"eval_samples_per_second": 94.245,
|
| 388 |
+
"eval_steps_per_second": 1.036,
|
| 389 |
+
"step": 45
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 10.88888888888889,
|
| 393 |
+
"eval_accuracy": 0.8021978021978022,
|
| 394 |
+
"eval_confusion_matrix": [
|
| 395 |
+
[
|
| 396 |
+
48,
|
| 397 |
+
18,
|
| 398 |
+
2,
|
| 399 |
+
4
|
| 400 |
+
],
|
| 401 |
+
[
|
| 402 |
+
4,
|
| 403 |
+
41,
|
| 404 |
+
15,
|
| 405 |
+
0
|
| 406 |
+
],
|
| 407 |
+
[
|
| 408 |
+
0,
|
| 409 |
+
7,
|
| 410 |
+
69,
|
| 411 |
+
0
|
| 412 |
+
],
|
| 413 |
+
[
|
| 414 |
+
3,
|
| 415 |
+
0,
|
| 416 |
+
1,
|
| 417 |
+
61
|
| 418 |
+
]
|
| 419 |
+
],
|
| 420 |
+
"eval_f1": 0.801525180147331,
|
| 421 |
+
"eval_loss": 0.7140281200408936,
|
| 422 |
+
"eval_precision": 0.8109338936925145,
|
| 423 |
+
"eval_recall": 0.8021978021978022,
|
| 424 |
+
"eval_runtime": 2.917,
|
| 425 |
+
"eval_samples_per_second": 93.589,
|
| 426 |
+
"eval_steps_per_second": 1.028,
|
| 427 |
+
"step": 49
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 12.0,
|
| 431 |
+
"eval_accuracy": 0.8534798534798534,
|
| 432 |
+
"eval_confusion_matrix": [
|
| 433 |
+
[
|
| 434 |
+
56,
|
| 435 |
+
12,
|
| 436 |
+
0,
|
| 437 |
+
4
|
| 438 |
+
],
|
| 439 |
+
[
|
| 440 |
+
6,
|
| 441 |
+
43,
|
| 442 |
+
11,
|
| 443 |
+
0
|
| 444 |
+
],
|
| 445 |
+
[
|
| 446 |
+
0,
|
| 447 |
+
6,
|
| 448 |
+
70,
|
| 449 |
+
0
|
| 450 |
+
],
|
| 451 |
+
[
|
| 452 |
+
1,
|
| 453 |
+
0,
|
| 454 |
+
0,
|
| 455 |
+
64
|
| 456 |
+
]
|
| 457 |
+
],
|
| 458 |
+
"eval_f1": 0.8523994617102314,
|
| 459 |
+
"eval_loss": 0.6672152280807495,
|
| 460 |
+
"eval_precision": 0.8540313732642031,
|
| 461 |
+
"eval_recall": 0.8534798534798534,
|
| 462 |
+
"eval_runtime": 2.8969,
|
| 463 |
+
"eval_samples_per_second": 94.238,
|
| 464 |
+
"eval_steps_per_second": 1.036,
|
| 465 |
+
"step": 54
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 12.88888888888889,
|
| 469 |
+
"eval_accuracy": 0.8498168498168498,
|
| 470 |
+
"eval_confusion_matrix": [
|
| 471 |
+
[
|
| 472 |
+
60,
|
| 473 |
+
8,
|
| 474 |
+
0,
|
| 475 |
+
4
|
| 476 |
+
],
|
| 477 |
+
[
|
| 478 |
+
8,
|
| 479 |
+
38,
|
| 480 |
+
14,
|
| 481 |
+
0
|
| 482 |
+
],
|
| 483 |
+
[
|
| 484 |
+
0,
|
| 485 |
+
6,
|
| 486 |
+
70,
|
| 487 |
+
0
|
| 488 |
+
],
|
| 489 |
+
[
|
| 490 |
+
1,
|
| 491 |
+
0,
|
| 492 |
+
0,
|
| 493 |
+
64
|
| 494 |
+
]
|
| 495 |
+
],
|
| 496 |
+
"eval_f1": 0.8463270052615757,
|
| 497 |
+
"eval_loss": 0.6432910561561584,
|
| 498 |
+
"eval_precision": 0.8460243715014519,
|
| 499 |
+
"eval_recall": 0.8498168498168498,
|
| 500 |
+
"eval_runtime": 3.0281,
|
| 501 |
+
"eval_samples_per_second": 90.156,
|
| 502 |
+
"eval_steps_per_second": 0.991,
|
| 503 |
+
"step": 58
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 14.0,
|
| 507 |
+
"eval_accuracy": 0.8278388278388278,
|
| 508 |
+
"eval_confusion_matrix": [
|
| 509 |
+
[
|
| 510 |
+
54,
|
| 511 |
+
13,
|
| 512 |
+
1,
|
| 513 |
+
4
|
| 514 |
+
],
|
| 515 |
+
[
|
| 516 |
+
7,
|
| 517 |
+
44,
|
| 518 |
+
9,
|
| 519 |
+
0
|
| 520 |
+
],
|
| 521 |
+
[
|
| 522 |
+
0,
|
| 523 |
+
11,
|
| 524 |
+
65,
|
| 525 |
+
0
|
| 526 |
+
],
|
| 527 |
+
[
|
| 528 |
+
1,
|
| 529 |
+
1,
|
| 530 |
+
0,
|
| 531 |
+
63
|
| 532 |
+
]
|
| 533 |
+
],
|
| 534 |
+
"eval_f1": 0.82943590265942,
|
| 535 |
+
"eval_loss": 0.7395206093788147,
|
| 536 |
+
"eval_precision": 0.8350059217447382,
|
| 537 |
+
"eval_recall": 0.8278388278388278,
|
| 538 |
+
"eval_runtime": 2.9869,
|
| 539 |
+
"eval_samples_per_second": 91.399,
|
| 540 |
+
"eval_steps_per_second": 1.004,
|
| 541 |
+
"step": 63
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 14.88888888888889,
|
| 545 |
+
"eval_accuracy": 0.8315018315018315,
|
| 546 |
+
"eval_confusion_matrix": [
|
| 547 |
+
[
|
| 548 |
+
54,
|
| 549 |
+
13,
|
| 550 |
+
1,
|
| 551 |
+
4
|
| 552 |
+
],
|
| 553 |
+
[
|
| 554 |
+
10,
|
| 555 |
+
39,
|
| 556 |
+
11,
|
| 557 |
+
0
|
| 558 |
+
],
|
| 559 |
+
[
|
| 560 |
+
0,
|
| 561 |
+
7,
|
| 562 |
+
69,
|
| 563 |
+
0
|
| 564 |
+
],
|
| 565 |
+
[
|
| 566 |
+
0,
|
| 567 |
+
0,
|
| 568 |
+
0,
|
| 569 |
+
65
|
| 570 |
+
]
|
| 571 |
+
],
|
| 572 |
+
"eval_f1": 0.8291811389886823,
|
| 573 |
+
"eval_loss": 0.7115849852561951,
|
| 574 |
+
"eval_precision": 0.829244108966536,
|
| 575 |
+
"eval_recall": 0.8315018315018315,
|
| 576 |
+
"eval_runtime": 2.9616,
|
| 577 |
+
"eval_samples_per_second": 92.181,
|
| 578 |
+
"eval_steps_per_second": 1.013,
|
| 579 |
+
"step": 67
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 16.0,
|
| 583 |
+
"eval_accuracy": 0.8315018315018315,
|
| 584 |
+
"eval_confusion_matrix": [
|
| 585 |
+
[
|
| 586 |
+
60,
|
| 587 |
+
7,
|
| 588 |
+
1,
|
| 589 |
+
4
|
| 590 |
+
],
|
| 591 |
+
[
|
| 592 |
+
10,
|
| 593 |
+
39,
|
| 594 |
+
11,
|
| 595 |
+
0
|
| 596 |
+
],
|
| 597 |
+
[
|
| 598 |
+
0,
|
| 599 |
+
11,
|
| 600 |
+
65,
|
| 601 |
+
0
|
| 602 |
+
],
|
| 603 |
+
[
|
| 604 |
+
1,
|
| 605 |
+
1,
|
| 606 |
+
0,
|
| 607 |
+
63
|
| 608 |
+
]
|
| 609 |
+
],
|
| 610 |
+
"eval_f1": 0.8304073820984628,
|
| 611 |
+
"eval_loss": 0.7295921444892883,
|
| 612 |
+
"eval_precision": 0.8295426562258641,
|
| 613 |
+
"eval_recall": 0.8315018315018315,
|
| 614 |
+
"eval_runtime": 2.8758,
|
| 615 |
+
"eval_samples_per_second": 94.932,
|
| 616 |
+
"eval_steps_per_second": 1.043,
|
| 617 |
+
"step": 72
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 16.88888888888889,
|
| 621 |
+
"eval_accuracy": 0.8644688644688645,
|
| 622 |
+
"eval_confusion_matrix": [
|
| 623 |
+
[
|
| 624 |
+
62,
|
| 625 |
+
5,
|
| 626 |
+
1,
|
| 627 |
+
4
|
| 628 |
+
],
|
| 629 |
+
[
|
| 630 |
+
9,
|
| 631 |
+
37,
|
| 632 |
+
14,
|
| 633 |
+
0
|
| 634 |
+
],
|
| 635 |
+
[
|
| 636 |
+
0,
|
| 637 |
+
3,
|
| 638 |
+
73,
|
| 639 |
+
0
|
| 640 |
+
],
|
| 641 |
+
[
|
| 642 |
+
1,
|
| 643 |
+
0,
|
| 644 |
+
0,
|
| 645 |
+
64
|
| 646 |
+
]
|
| 647 |
+
],
|
| 648 |
+
"eval_f1": 0.8589767100678526,
|
| 649 |
+
"eval_loss": 0.7055637240409851,
|
| 650 |
+
"eval_precision": 0.8628397746044805,
|
| 651 |
+
"eval_recall": 0.8644688644688645,
|
| 652 |
+
"eval_runtime": 3.074,
|
| 653 |
+
"eval_samples_per_second": 88.809,
|
| 654 |
+
"eval_steps_per_second": 0.976,
|
| 655 |
+
"step": 76
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 18.0,
|
| 659 |
+
"eval_accuracy": 0.8644688644688645,
|
| 660 |
+
"eval_confusion_matrix": [
|
| 661 |
+
[
|
| 662 |
+
65,
|
| 663 |
+
2,
|
| 664 |
+
1,
|
| 665 |
+
4
|
| 666 |
+
],
|
| 667 |
+
[
|
| 668 |
+
13,
|
| 669 |
+
42,
|
| 670 |
+
5,
|
| 671 |
+
0
|
| 672 |
+
],
|
| 673 |
+
[
|
| 674 |
+
0,
|
| 675 |
+
12,
|
| 676 |
+
64,
|
| 677 |
+
0
|
| 678 |
+
],
|
| 679 |
+
[
|
| 680 |
+
0,
|
| 681 |
+
0,
|
| 682 |
+
0,
|
| 683 |
+
65
|
| 684 |
+
]
|
| 685 |
+
],
|
| 686 |
+
"eval_f1": 0.8627769756077204,
|
| 687 |
+
"eval_loss": 0.7563945651054382,
|
| 688 |
+
"eval_precision": 0.8634344261673453,
|
| 689 |
+
"eval_recall": 0.8644688644688645,
|
| 690 |
+
"eval_runtime": 3.0072,
|
| 691 |
+
"eval_samples_per_second": 90.783,
|
| 692 |
+
"eval_steps_per_second": 0.998,
|
| 693 |
+
"step": 81
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 18.88888888888889,
|
| 697 |
+
"eval_accuracy": 0.8424908424908425,
|
| 698 |
+
"eval_confusion_matrix": [
|
| 699 |
+
[
|
| 700 |
+
64,
|
| 701 |
+
4,
|
| 702 |
+
0,
|
| 703 |
+
4
|
| 704 |
+
],
|
| 705 |
+
[
|
| 706 |
+
11,
|
| 707 |
+
41,
|
| 708 |
+
8,
|
| 709 |
+
0
|
| 710 |
+
],
|
| 711 |
+
[
|
| 712 |
+
0,
|
| 713 |
+
10,
|
| 714 |
+
66,
|
| 715 |
+
0
|
| 716 |
+
],
|
| 717 |
+
[
|
| 718 |
+
6,
|
| 719 |
+
0,
|
| 720 |
+
0,
|
| 721 |
+
59
|
| 722 |
+
]
|
| 723 |
+
],
|
| 724 |
+
"eval_f1": 0.8418306879608031,
|
| 725 |
+
"eval_loss": 0.7825365662574768,
|
| 726 |
+
"eval_precision": 0.8434907006335578,
|
| 727 |
+
"eval_recall": 0.8424908424908425,
|
| 728 |
+
"eval_runtime": 3.0343,
|
| 729 |
+
"eval_samples_per_second": 89.972,
|
| 730 |
+
"eval_steps_per_second": 0.989,
|
| 731 |
+
"step": 85
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 20.0,
|
| 735 |
+
"eval_accuracy": 0.8058608058608059,
|
| 736 |
+
"eval_confusion_matrix": [
|
| 737 |
+
[
|
| 738 |
+
40,
|
| 739 |
+
28,
|
| 740 |
+
1,
|
| 741 |
+
3
|
| 742 |
+
],
|
| 743 |
+
[
|
| 744 |
+
2,
|
| 745 |
+
50,
|
| 746 |
+
8,
|
| 747 |
+
0
|
| 748 |
+
],
|
| 749 |
+
[
|
| 750 |
+
0,
|
| 751 |
+
11,
|
| 752 |
+
65,
|
| 753 |
+
0
|
| 754 |
+
],
|
| 755 |
+
[
|
| 756 |
+
0,
|
| 757 |
+
0,
|
| 758 |
+
0,
|
| 759 |
+
65
|
| 760 |
+
]
|
| 761 |
+
],
|
| 762 |
+
"eval_f1": 0.8065760931078588,
|
| 763 |
+
"eval_loss": 0.8426868915557861,
|
| 764 |
+
"eval_precision": 0.8467707085637404,
|
| 765 |
+
"eval_recall": 0.8058608058608059,
|
| 766 |
+
"eval_runtime": 3.0783,
|
| 767 |
+
"eval_samples_per_second": 88.686,
|
| 768 |
+
"eval_steps_per_second": 0.975,
|
| 769 |
+
"step": 90
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 20.88888888888889,
|
| 773 |
+
"eval_accuracy": 0.8498168498168498,
|
| 774 |
+
"eval_confusion_matrix": [
|
| 775 |
+
[
|
| 776 |
+
63,
|
| 777 |
+
4,
|
| 778 |
+
0,
|
| 779 |
+
5
|
| 780 |
+
],
|
| 781 |
+
[
|
| 782 |
+
13,
|
| 783 |
+
40,
|
| 784 |
+
7,
|
| 785 |
+
0
|
| 786 |
+
],
|
| 787 |
+
[
|
| 788 |
+
0,
|
| 789 |
+
12,
|
| 790 |
+
64,
|
| 791 |
+
0
|
| 792 |
+
],
|
| 793 |
+
[
|
| 794 |
+
0,
|
| 795 |
+
0,
|
| 796 |
+
0,
|
| 797 |
+
65
|
| 798 |
+
]
|
| 799 |
+
],
|
| 800 |
+
"eval_f1": 0.8477878057985963,
|
| 801 |
+
"eval_loss": 0.7440442442893982,
|
| 802 |
+
"eval_precision": 0.8476393351433065,
|
| 803 |
+
"eval_recall": 0.8498168498168498,
|
| 804 |
+
"eval_runtime": 2.9875,
|
| 805 |
+
"eval_samples_per_second": 91.38,
|
| 806 |
+
"eval_steps_per_second": 1.004,
|
| 807 |
+
"step": 94
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 22.0,
|
| 811 |
+
"eval_accuracy": 0.8608058608058609,
|
| 812 |
+
"eval_confusion_matrix": [
|
| 813 |
+
[
|
| 814 |
+
66,
|
| 815 |
+
1,
|
| 816 |
+
1,
|
| 817 |
+
4
|
| 818 |
+
],
|
| 819 |
+
[
|
| 820 |
+
13,
|
| 821 |
+
36,
|
| 822 |
+
11,
|
| 823 |
+
0
|
| 824 |
+
],
|
| 825 |
+
[
|
| 826 |
+
0,
|
| 827 |
+
5,
|
| 828 |
+
71,
|
| 829 |
+
0
|
| 830 |
+
],
|
| 831 |
+
[
|
| 832 |
+
3,
|
| 833 |
+
0,
|
| 834 |
+
0,
|
| 835 |
+
62
|
| 836 |
+
]
|
| 837 |
+
],
|
| 838 |
+
"eval_f1": 0.855194718990792,
|
| 839 |
+
"eval_loss": 0.7338178753852844,
|
| 840 |
+
"eval_precision": 0.8624631692093176,
|
| 841 |
+
"eval_recall": 0.8608058608058609,
|
| 842 |
+
"eval_runtime": 2.98,
|
| 843 |
+
"eval_samples_per_second": 91.611,
|
| 844 |
+
"eval_steps_per_second": 1.007,
|
| 845 |
+
"step": 99
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 22.88888888888889,
|
| 849 |
+
"eval_accuracy": 0.8498168498168498,
|
| 850 |
+
"eval_confusion_matrix": [
|
| 851 |
+
[
|
| 852 |
+
58,
|
| 853 |
+
10,
|
| 854 |
+
0,
|
| 855 |
+
4
|
| 856 |
+
],
|
| 857 |
+
[
|
| 858 |
+
10,
|
| 859 |
+
43,
|
| 860 |
+
7,
|
| 861 |
+
0
|
| 862 |
+
],
|
| 863 |
+
[
|
| 864 |
+
0,
|
| 865 |
+
7,
|
| 866 |
+
69,
|
| 867 |
+
0
|
| 868 |
+
],
|
| 869 |
+
[
|
| 870 |
+
2,
|
| 871 |
+
1,
|
| 872 |
+
0,
|
| 873 |
+
62
|
| 874 |
+
]
|
| 875 |
+
],
|
| 876 |
+
"eval_f1": 0.8497737987724402,
|
| 877 |
+
"eval_loss": 0.7231407761573792,
|
| 878 |
+
"eval_precision": 0.8498638493954653,
|
| 879 |
+
"eval_recall": 0.8498168498168498,
|
| 880 |
+
"eval_runtime": 2.9272,
|
| 881 |
+
"eval_samples_per_second": 93.262,
|
| 882 |
+
"eval_steps_per_second": 1.025,
|
| 883 |
+
"step": 103
|
| 884 |
+
}
|
| 885 |
+
],
|
| 886 |
+
"logging_steps": 500,
|
| 887 |
+
"max_steps": 180,
|
| 888 |
+
"num_input_tokens_seen": 0,
|
| 889 |
+
"num_train_epochs": 45,
|
| 890 |
+
"save_steps": 500,
|
| 891 |
+
"stateful_callbacks": {
|
| 892 |
+
"TrainerControl": {
|
| 893 |
+
"args": {
|
| 894 |
+
"should_epoch_stop": false,
|
| 895 |
+
"should_evaluate": false,
|
| 896 |
+
"should_log": false,
|
| 897 |
+
"should_save": true,
|
| 898 |
+
"should_training_stop": false
|
| 899 |
+
},
|
| 900 |
+
"attributes": {}
|
| 901 |
+
}
|
| 902 |
+
},
|
| 903 |
+
"total_flos": 5.70148598592e+16,
|
| 904 |
+
"train_batch_size": 128,
|
| 905 |
+
"trial_name": null,
|
| 906 |
+
"trial_params": null
|
| 907 |
+
}
|
checkpoint-103/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:09e217cad0464b642c54f8b7691100b62a93dceb32a1869f55e3f0eb3a54a79e
|
| 3 |
+
size 5240
|
checkpoint-108/config.json
ADDED
|
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ntu-spml/distilhubert",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"HubertForSequenceClassification"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 1,
|
| 10 |
+
"classifier_proj_size": 256,
|
| 11 |
+
"conv_bias": false,
|
| 12 |
+
"conv_dim": [
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512
|
| 20 |
+
],
|
| 21 |
+
"conv_kernel": [
|
| 22 |
+
10,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
3,
|
| 27 |
+
2,
|
| 28 |
+
2
|
| 29 |
+
],
|
| 30 |
+
"conv_stride": [
|
| 31 |
+
5,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2,
|
| 37 |
+
2
|
| 38 |
+
],
|
| 39 |
+
"ctc_loss_reduction": "sum",
|
| 40 |
+
"ctc_zero_infinity": false,
|
| 41 |
+
"do_stable_layer_norm": false,
|
| 42 |
+
"eos_token_id": 2,
|
| 43 |
+
"feat_extract_activation": "gelu",
|
| 44 |
+
"feat_extract_norm": "group",
|
| 45 |
+
"feat_proj_dropout": 0.0,
|
| 46 |
+
"feat_proj_layer_norm": false,
|
| 47 |
+
"final_dropout": 0.0,
|
| 48 |
+
"finetuning_task": "audio-classification",
|
| 49 |
+
"hidden_act": "gelu",
|
| 50 |
+
"hidden_dropout": 0.1,
|
| 51 |
+
"hidden_size": 768,
|
| 52 |
+
"id2label": {
|
| 53 |
+
"0": "1s_normal",
|
| 54 |
+
"1": "1s_pain",
|
| 55 |
+
"2": "1s_hunger",
|
| 56 |
+
"3": "1s_asphyxia"
|
| 57 |
+
},
|
| 58 |
+
"initializer_range": 0.02,
|
| 59 |
+
"intermediate_size": 3072,
|
| 60 |
+
"label2id": {
|
| 61 |
+
"LABEL_0": 0,
|
| 62 |
+
"LABEL_1": 1,
|
| 63 |
+
"LABEL_2": 2,
|
| 64 |
+
"LABEL_3": 3
|
| 65 |
+
},
|
| 66 |
+
"layer_norm_eps": 1e-05,
|
| 67 |
+
"layerdrop": 0.0,
|
| 68 |
+
"mask_feature_length": 10,
|
| 69 |
+
"mask_feature_min_masks": 0,
|
| 70 |
+
"mask_feature_prob": 0.0,
|
| 71 |
+
"mask_time_length": 10,
|
| 72 |
+
"mask_time_min_masks": 2,
|
| 73 |
+
"mask_time_prob": 0.05,
|
| 74 |
+
"model_type": "hubert",
|
| 75 |
+
"num_attention_heads": 12,
|
| 76 |
+
"num_conv_pos_embedding_groups": 16,
|
| 77 |
+
"num_conv_pos_embeddings": 128,
|
| 78 |
+
"num_feat_extract_layers": 7,
|
| 79 |
+
"num_hidden_layers": 2,
|
| 80 |
+
"pad_token_id": 0,
|
| 81 |
+
"torch_dtype": "float32",
|
| 82 |
+
"transformers_version": "4.44.2",
|
| 83 |
+
"use_weighted_layer_sum": false,
|
| 84 |
+
"vocab_size": 32
|
| 85 |
+
}
|
checkpoint-108/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5f193969e204945d23169464066234ad545dcdd00180959259c3b0d93230ee00
|
| 3 |
+
size 94765560
|
checkpoint-108/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7134ec55f60b7455d0f9bbdbb4efc5a3208c1fb45b53f2caa01c4a850e0af82c
|
| 3 |
+
size 189556666
|
checkpoint-108/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7910a20ae0ec73ddb102bf8728f5fbb2c63ac595644645b40b7095c74e739644
|
| 3 |
+
size 14308
|
checkpoint-108/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14a0f217e1cd7a6ee64cdcab8f02f2a72c157844aab523fbebe978c66b9af419
|
| 3 |
+
size 1064
|
checkpoint-108/trainer_state.json
ADDED
|
@@ -0,0 +1,945 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8627769756077204,
|
| 3 |
+
"best_model_checkpoint": "distilhubert-finetuned-mixed-data/checkpoint-81",
|
| 4 |
+
"epoch": 24.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 108,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.8888888888888888,
|
| 13 |
+
"eval_accuracy": 0.63003663003663,
|
| 14 |
+
"eval_confusion_matrix": [
|
| 15 |
+
[
|
| 16 |
+
53,
|
| 17 |
+
0,
|
| 18 |
+
16,
|
| 19 |
+
3
|
| 20 |
+
],
|
| 21 |
+
[
|
| 22 |
+
28,
|
| 23 |
+
1,
|
| 24 |
+
31,
|
| 25 |
+
0
|
| 26 |
+
],
|
| 27 |
+
[
|
| 28 |
+
13,
|
| 29 |
+
0,
|
| 30 |
+
62,
|
| 31 |
+
1
|
| 32 |
+
],
|
| 33 |
+
[
|
| 34 |
+
3,
|
| 35 |
+
0,
|
| 36 |
+
6,
|
| 37 |
+
56
|
| 38 |
+
]
|
| 39 |
+
],
|
| 40 |
+
"eval_f1": 0.566693372031096,
|
| 41 |
+
"eval_loss": 1.31327486038208,
|
| 42 |
+
"eval_precision": 0.7361933549293478,
|
| 43 |
+
"eval_recall": 0.63003663003663,
|
| 44 |
+
"eval_runtime": 2.8436,
|
| 45 |
+
"eval_samples_per_second": 96.005,
|
| 46 |
+
"eval_steps_per_second": 1.055,
|
| 47 |
+
"step": 4
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 2.0,
|
| 51 |
+
"eval_accuracy": 0.6593406593406593,
|
| 52 |
+
"eval_confusion_matrix": [
|
| 53 |
+
[
|
| 54 |
+
41,
|
| 55 |
+
10,
|
| 56 |
+
16,
|
| 57 |
+
5
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
21,
|
| 61 |
+
1,
|
| 62 |
+
38,
|
| 63 |
+
0
|
| 64 |
+
],
|
| 65 |
+
[
|
| 66 |
+
1,
|
| 67 |
+
2,
|
| 68 |
+
73,
|
| 69 |
+
0
|
| 70 |
+
],
|
| 71 |
+
[
|
| 72 |
+
0,
|
| 73 |
+
0,
|
| 74 |
+
0,
|
| 75 |
+
65
|
| 76 |
+
]
|
| 77 |
+
],
|
| 78 |
+
"eval_f1": 0.595713773130153,
|
| 79 |
+
"eval_loss": 0.9242589473724365,
|
| 80 |
+
"eval_precision": 0.5696509512811321,
|
| 81 |
+
"eval_recall": 0.6593406593406593,
|
| 82 |
+
"eval_runtime": 2.8668,
|
| 83 |
+
"eval_samples_per_second": 95.227,
|
| 84 |
+
"eval_steps_per_second": 1.046,
|
| 85 |
+
"step": 9
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 2.888888888888889,
|
| 89 |
+
"eval_accuracy": 0.717948717948718,
|
| 90 |
+
"eval_confusion_matrix": [
|
| 91 |
+
[
|
| 92 |
+
52,
|
| 93 |
+
8,
|
| 94 |
+
7,
|
| 95 |
+
5
|
| 96 |
+
],
|
| 97 |
+
[
|
| 98 |
+
18,
|
| 99 |
+
5,
|
| 100 |
+
37,
|
| 101 |
+
0
|
| 102 |
+
],
|
| 103 |
+
[
|
| 104 |
+
1,
|
| 105 |
+
0,
|
| 106 |
+
75,
|
| 107 |
+
0
|
| 108 |
+
],
|
| 109 |
+
[
|
| 110 |
+
1,
|
| 111 |
+
0,
|
| 112 |
+
0,
|
| 113 |
+
64
|
| 114 |
+
]
|
| 115 |
+
],
|
| 116 |
+
"eval_f1": 0.6621621567125869,
|
| 117 |
+
"eval_loss": 0.8232662081718445,
|
| 118 |
+
"eval_precision": 0.671303801513745,
|
| 119 |
+
"eval_recall": 0.717948717948718,
|
| 120 |
+
"eval_runtime": 2.8711,
|
| 121 |
+
"eval_samples_per_second": 95.084,
|
| 122 |
+
"eval_steps_per_second": 1.045,
|
| 123 |
+
"step": 13
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 4.0,
|
| 127 |
+
"eval_accuracy": 0.6959706959706959,
|
| 128 |
+
"eval_confusion_matrix": [
|
| 129 |
+
[
|
| 130 |
+
45,
|
| 131 |
+
9,
|
| 132 |
+
13,
|
| 133 |
+
5
|
| 134 |
+
],
|
| 135 |
+
[
|
| 136 |
+
13,
|
| 137 |
+
5,
|
| 138 |
+
42,
|
| 139 |
+
0
|
| 140 |
+
],
|
| 141 |
+
[
|
| 142 |
+
1,
|
| 143 |
+
0,
|
| 144 |
+
75,
|
| 145 |
+
0
|
| 146 |
+
],
|
| 147 |
+
[
|
| 148 |
+
0,
|
| 149 |
+
0,
|
| 150 |
+
0,
|
| 151 |
+
65
|
| 152 |
+
]
|
| 153 |
+
],
|
| 154 |
+
"eval_f1": 0.6428796900399459,
|
| 155 |
+
"eval_loss": 0.951453447341919,
|
| 156 |
+
"eval_precision": 0.6613447705829899,
|
| 157 |
+
"eval_recall": 0.6959706959706959,
|
| 158 |
+
"eval_runtime": 2.8813,
|
| 159 |
+
"eval_samples_per_second": 94.75,
|
| 160 |
+
"eval_steps_per_second": 1.041,
|
| 161 |
+
"step": 18
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 4.888888888888889,
|
| 165 |
+
"eval_accuracy": 0.6666666666666666,
|
| 166 |
+
"eval_confusion_matrix": [
|
| 167 |
+
[
|
| 168 |
+
16,
|
| 169 |
+
49,
|
| 170 |
+
3,
|
| 171 |
+
4
|
| 172 |
+
],
|
| 173 |
+
[
|
| 174 |
+
1,
|
| 175 |
+
26,
|
| 176 |
+
33,
|
| 177 |
+
0
|
| 178 |
+
],
|
| 179 |
+
[
|
| 180 |
+
0,
|
| 181 |
+
1,
|
| 182 |
+
75,
|
| 183 |
+
0
|
| 184 |
+
],
|
| 185 |
+
[
|
| 186 |
+
0,
|
| 187 |
+
0,
|
| 188 |
+
0,
|
| 189 |
+
65
|
| 190 |
+
]
|
| 191 |
+
],
|
| 192 |
+
"eval_f1": 0.6331541584523206,
|
| 193 |
+
"eval_loss": 1.0080645084381104,
|
| 194 |
+
"eval_precision": 0.7358031394913882,
|
| 195 |
+
"eval_recall": 0.6666666666666666,
|
| 196 |
+
"eval_runtime": 2.9108,
|
| 197 |
+
"eval_samples_per_second": 93.789,
|
| 198 |
+
"eval_steps_per_second": 1.031,
|
| 199 |
+
"step": 22
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 6.0,
|
| 203 |
+
"eval_accuracy": 0.652014652014652,
|
| 204 |
+
"eval_confusion_matrix": [
|
| 205 |
+
[
|
| 206 |
+
66,
|
| 207 |
+
0,
|
| 208 |
+
0,
|
| 209 |
+
6
|
| 210 |
+
],
|
| 211 |
+
[
|
| 212 |
+
36,
|
| 213 |
+
16,
|
| 214 |
+
8,
|
| 215 |
+
0
|
| 216 |
+
],
|
| 217 |
+
[
|
| 218 |
+
10,
|
| 219 |
+
35,
|
| 220 |
+
31,
|
| 221 |
+
0
|
| 222 |
+
],
|
| 223 |
+
[
|
| 224 |
+
0,
|
| 225 |
+
0,
|
| 226 |
+
0,
|
| 227 |
+
65
|
| 228 |
+
]
|
| 229 |
+
],
|
| 230 |
+
"eval_f1": 0.6302407955860642,
|
| 231 |
+
"eval_loss": 0.9598046541213989,
|
| 232 |
+
"eval_precision": 0.6636241740077424,
|
| 233 |
+
"eval_recall": 0.652014652014652,
|
| 234 |
+
"eval_runtime": 2.9773,
|
| 235 |
+
"eval_samples_per_second": 91.693,
|
| 236 |
+
"eval_steps_per_second": 1.008,
|
| 237 |
+
"step": 27
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 6.888888888888889,
|
| 241 |
+
"eval_accuracy": 0.7692307692307693,
|
| 242 |
+
"eval_confusion_matrix": [
|
| 243 |
+
[
|
| 244 |
+
66,
|
| 245 |
+
0,
|
| 246 |
+
0,
|
| 247 |
+
6
|
| 248 |
+
],
|
| 249 |
+
[
|
| 250 |
+
29,
|
| 251 |
+
21,
|
| 252 |
+
10,
|
| 253 |
+
0
|
| 254 |
+
],
|
| 255 |
+
[
|
| 256 |
+
3,
|
| 257 |
+
15,
|
| 258 |
+
58,
|
| 259 |
+
0
|
| 260 |
+
],
|
| 261 |
+
[
|
| 262 |
+
0,
|
| 263 |
+
0,
|
| 264 |
+
0,
|
| 265 |
+
65
|
| 266 |
+
]
|
| 267 |
+
],
|
| 268 |
+
"eval_f1": 0.7527855586679116,
|
| 269 |
+
"eval_loss": 0.7898163795471191,
|
| 270 |
+
"eval_precision": 0.7612467677056892,
|
| 271 |
+
"eval_recall": 0.7692307692307693,
|
| 272 |
+
"eval_runtime": 2.8918,
|
| 273 |
+
"eval_samples_per_second": 94.404,
|
| 274 |
+
"eval_steps_per_second": 1.037,
|
| 275 |
+
"step": 31
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 8.0,
|
| 279 |
+
"eval_accuracy": 0.7802197802197802,
|
| 280 |
+
"eval_confusion_matrix": [
|
| 281 |
+
[
|
| 282 |
+
66,
|
| 283 |
+
0,
|
| 284 |
+
0,
|
| 285 |
+
6
|
| 286 |
+
],
|
| 287 |
+
[
|
| 288 |
+
25,
|
| 289 |
+
15,
|
| 290 |
+
19,
|
| 291 |
+
1
|
| 292 |
+
],
|
| 293 |
+
[
|
| 294 |
+
1,
|
| 295 |
+
6,
|
| 296 |
+
68,
|
| 297 |
+
1
|
| 298 |
+
],
|
| 299 |
+
[
|
| 300 |
+
1,
|
| 301 |
+
0,
|
| 302 |
+
0,
|
| 303 |
+
64
|
| 304 |
+
]
|
| 305 |
+
],
|
| 306 |
+
"eval_f1": 0.7471179200524057,
|
| 307 |
+
"eval_loss": 0.7337484359741211,
|
| 308 |
+
"eval_precision": 0.7733845922309294,
|
| 309 |
+
"eval_recall": 0.7802197802197802,
|
| 310 |
+
"eval_runtime": 2.9006,
|
| 311 |
+
"eval_samples_per_second": 94.119,
|
| 312 |
+
"eval_steps_per_second": 1.034,
|
| 313 |
+
"step": 36
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 8.88888888888889,
|
| 317 |
+
"eval_accuracy": 0.7912087912087912,
|
| 318 |
+
"eval_confusion_matrix": [
|
| 319 |
+
[
|
| 320 |
+
57,
|
| 321 |
+
8,
|
| 322 |
+
3,
|
| 323 |
+
4
|
| 324 |
+
],
|
| 325 |
+
[
|
| 326 |
+
13,
|
| 327 |
+
24,
|
| 328 |
+
23,
|
| 329 |
+
0
|
| 330 |
+
],
|
| 331 |
+
[
|
| 332 |
+
0,
|
| 333 |
+
5,
|
| 334 |
+
71,
|
| 335 |
+
0
|
| 336 |
+
],
|
| 337 |
+
[
|
| 338 |
+
1,
|
| 339 |
+
0,
|
| 340 |
+
0,
|
| 341 |
+
64
|
| 342 |
+
]
|
| 343 |
+
],
|
| 344 |
+
"eval_f1": 0.7766564722186922,
|
| 345 |
+
"eval_loss": 0.7148727178573608,
|
| 346 |
+
"eval_precision": 0.7821504483074875,
|
| 347 |
+
"eval_recall": 0.7912087912087912,
|
| 348 |
+
"eval_runtime": 2.8866,
|
| 349 |
+
"eval_samples_per_second": 94.576,
|
| 350 |
+
"eval_steps_per_second": 1.039,
|
| 351 |
+
"step": 40
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 10.0,
|
| 355 |
+
"eval_accuracy": 0.8168498168498168,
|
| 356 |
+
"eval_confusion_matrix": [
|
| 357 |
+
[
|
| 358 |
+
51,
|
| 359 |
+
8,
|
| 360 |
+
1,
|
| 361 |
+
12
|
| 362 |
+
],
|
| 363 |
+
[
|
| 364 |
+
10,
|
| 365 |
+
36,
|
| 366 |
+
14,
|
| 367 |
+
0
|
| 368 |
+
],
|
| 369 |
+
[
|
| 370 |
+
0,
|
| 371 |
+
5,
|
| 372 |
+
71,
|
| 373 |
+
0
|
| 374 |
+
],
|
| 375 |
+
[
|
| 376 |
+
0,
|
| 377 |
+
0,
|
| 378 |
+
0,
|
| 379 |
+
65
|
| 380 |
+
]
|
| 381 |
+
],
|
| 382 |
+
"eval_f1": 0.8094336953840884,
|
| 383 |
+
"eval_loss": 0.7574812173843384,
|
| 384 |
+
"eval_precision": 0.8127936625684181,
|
| 385 |
+
"eval_recall": 0.8168498168498168,
|
| 386 |
+
"eval_runtime": 2.8967,
|
| 387 |
+
"eval_samples_per_second": 94.245,
|
| 388 |
+
"eval_steps_per_second": 1.036,
|
| 389 |
+
"step": 45
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 10.88888888888889,
|
| 393 |
+
"eval_accuracy": 0.8021978021978022,
|
| 394 |
+
"eval_confusion_matrix": [
|
| 395 |
+
[
|
| 396 |
+
48,
|
| 397 |
+
18,
|
| 398 |
+
2,
|
| 399 |
+
4
|
| 400 |
+
],
|
| 401 |
+
[
|
| 402 |
+
4,
|
| 403 |
+
41,
|
| 404 |
+
15,
|
| 405 |
+
0
|
| 406 |
+
],
|
| 407 |
+
[
|
| 408 |
+
0,
|
| 409 |
+
7,
|
| 410 |
+
69,
|
| 411 |
+
0
|
| 412 |
+
],
|
| 413 |
+
[
|
| 414 |
+
3,
|
| 415 |
+
0,
|
| 416 |
+
1,
|
| 417 |
+
61
|
| 418 |
+
]
|
| 419 |
+
],
|
| 420 |
+
"eval_f1": 0.801525180147331,
|
| 421 |
+
"eval_loss": 0.7140281200408936,
|
| 422 |
+
"eval_precision": 0.8109338936925145,
|
| 423 |
+
"eval_recall": 0.8021978021978022,
|
| 424 |
+
"eval_runtime": 2.917,
|
| 425 |
+
"eval_samples_per_second": 93.589,
|
| 426 |
+
"eval_steps_per_second": 1.028,
|
| 427 |
+
"step": 49
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 12.0,
|
| 431 |
+
"eval_accuracy": 0.8534798534798534,
|
| 432 |
+
"eval_confusion_matrix": [
|
| 433 |
+
[
|
| 434 |
+
56,
|
| 435 |
+
12,
|
| 436 |
+
0,
|
| 437 |
+
4
|
| 438 |
+
],
|
| 439 |
+
[
|
| 440 |
+
6,
|
| 441 |
+
43,
|
| 442 |
+
11,
|
| 443 |
+
0
|
| 444 |
+
],
|
| 445 |
+
[
|
| 446 |
+
0,
|
| 447 |
+
6,
|
| 448 |
+
70,
|
| 449 |
+
0
|
| 450 |
+
],
|
| 451 |
+
[
|
| 452 |
+
1,
|
| 453 |
+
0,
|
| 454 |
+
0,
|
| 455 |
+
64
|
| 456 |
+
]
|
| 457 |
+
],
|
| 458 |
+
"eval_f1": 0.8523994617102314,
|
| 459 |
+
"eval_loss": 0.6672152280807495,
|
| 460 |
+
"eval_precision": 0.8540313732642031,
|
| 461 |
+
"eval_recall": 0.8534798534798534,
|
| 462 |
+
"eval_runtime": 2.8969,
|
| 463 |
+
"eval_samples_per_second": 94.238,
|
| 464 |
+
"eval_steps_per_second": 1.036,
|
| 465 |
+
"step": 54
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 12.88888888888889,
|
| 469 |
+
"eval_accuracy": 0.8498168498168498,
|
| 470 |
+
"eval_confusion_matrix": [
|
| 471 |
+
[
|
| 472 |
+
60,
|
| 473 |
+
8,
|
| 474 |
+
0,
|
| 475 |
+
4
|
| 476 |
+
],
|
| 477 |
+
[
|
| 478 |
+
8,
|
| 479 |
+
38,
|
| 480 |
+
14,
|
| 481 |
+
0
|
| 482 |
+
],
|
| 483 |
+
[
|
| 484 |
+
0,
|
| 485 |
+
6,
|
| 486 |
+
70,
|
| 487 |
+
0
|
| 488 |
+
],
|
| 489 |
+
[
|
| 490 |
+
1,
|
| 491 |
+
0,
|
| 492 |
+
0,
|
| 493 |
+
64
|
| 494 |
+
]
|
| 495 |
+
],
|
| 496 |
+
"eval_f1": 0.8463270052615757,
|
| 497 |
+
"eval_loss": 0.6432910561561584,
|
| 498 |
+
"eval_precision": 0.8460243715014519,
|
| 499 |
+
"eval_recall": 0.8498168498168498,
|
| 500 |
+
"eval_runtime": 3.0281,
|
| 501 |
+
"eval_samples_per_second": 90.156,
|
| 502 |
+
"eval_steps_per_second": 0.991,
|
| 503 |
+
"step": 58
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 14.0,
|
| 507 |
+
"eval_accuracy": 0.8278388278388278,
|
| 508 |
+
"eval_confusion_matrix": [
|
| 509 |
+
[
|
| 510 |
+
54,
|
| 511 |
+
13,
|
| 512 |
+
1,
|
| 513 |
+
4
|
| 514 |
+
],
|
| 515 |
+
[
|
| 516 |
+
7,
|
| 517 |
+
44,
|
| 518 |
+
9,
|
| 519 |
+
0
|
| 520 |
+
],
|
| 521 |
+
[
|
| 522 |
+
0,
|
| 523 |
+
11,
|
| 524 |
+
65,
|
| 525 |
+
0
|
| 526 |
+
],
|
| 527 |
+
[
|
| 528 |
+
1,
|
| 529 |
+
1,
|
| 530 |
+
0,
|
| 531 |
+
63
|
| 532 |
+
]
|
| 533 |
+
],
|
| 534 |
+
"eval_f1": 0.82943590265942,
|
| 535 |
+
"eval_loss": 0.7395206093788147,
|
| 536 |
+
"eval_precision": 0.8350059217447382,
|
| 537 |
+
"eval_recall": 0.8278388278388278,
|
| 538 |
+
"eval_runtime": 2.9869,
|
| 539 |
+
"eval_samples_per_second": 91.399,
|
| 540 |
+
"eval_steps_per_second": 1.004,
|
| 541 |
+
"step": 63
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 14.88888888888889,
|
| 545 |
+
"eval_accuracy": 0.8315018315018315,
|
| 546 |
+
"eval_confusion_matrix": [
|
| 547 |
+
[
|
| 548 |
+
54,
|
| 549 |
+
13,
|
| 550 |
+
1,
|
| 551 |
+
4
|
| 552 |
+
],
|
| 553 |
+
[
|
| 554 |
+
10,
|
| 555 |
+
39,
|
| 556 |
+
11,
|
| 557 |
+
0
|
| 558 |
+
],
|
| 559 |
+
[
|
| 560 |
+
0,
|
| 561 |
+
7,
|
| 562 |
+
69,
|
| 563 |
+
0
|
| 564 |
+
],
|
| 565 |
+
[
|
| 566 |
+
0,
|
| 567 |
+
0,
|
| 568 |
+
0,
|
| 569 |
+
65
|
| 570 |
+
]
|
| 571 |
+
],
|
| 572 |
+
"eval_f1": 0.8291811389886823,
|
| 573 |
+
"eval_loss": 0.7115849852561951,
|
| 574 |
+
"eval_precision": 0.829244108966536,
|
| 575 |
+
"eval_recall": 0.8315018315018315,
|
| 576 |
+
"eval_runtime": 2.9616,
|
| 577 |
+
"eval_samples_per_second": 92.181,
|
| 578 |
+
"eval_steps_per_second": 1.013,
|
| 579 |
+
"step": 67
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 16.0,
|
| 583 |
+
"eval_accuracy": 0.8315018315018315,
|
| 584 |
+
"eval_confusion_matrix": [
|
| 585 |
+
[
|
| 586 |
+
60,
|
| 587 |
+
7,
|
| 588 |
+
1,
|
| 589 |
+
4
|
| 590 |
+
],
|
| 591 |
+
[
|
| 592 |
+
10,
|
| 593 |
+
39,
|
| 594 |
+
11,
|
| 595 |
+
0
|
| 596 |
+
],
|
| 597 |
+
[
|
| 598 |
+
0,
|
| 599 |
+
11,
|
| 600 |
+
65,
|
| 601 |
+
0
|
| 602 |
+
],
|
| 603 |
+
[
|
| 604 |
+
1,
|
| 605 |
+
1,
|
| 606 |
+
0,
|
| 607 |
+
63
|
| 608 |
+
]
|
| 609 |
+
],
|
| 610 |
+
"eval_f1": 0.8304073820984628,
|
| 611 |
+
"eval_loss": 0.7295921444892883,
|
| 612 |
+
"eval_precision": 0.8295426562258641,
|
| 613 |
+
"eval_recall": 0.8315018315018315,
|
| 614 |
+
"eval_runtime": 2.8758,
|
| 615 |
+
"eval_samples_per_second": 94.932,
|
| 616 |
+
"eval_steps_per_second": 1.043,
|
| 617 |
+
"step": 72
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 16.88888888888889,
|
| 621 |
+
"eval_accuracy": 0.8644688644688645,
|
| 622 |
+
"eval_confusion_matrix": [
|
| 623 |
+
[
|
| 624 |
+
62,
|
| 625 |
+
5,
|
| 626 |
+
1,
|
| 627 |
+
4
|
| 628 |
+
],
|
| 629 |
+
[
|
| 630 |
+
9,
|
| 631 |
+
37,
|
| 632 |
+
14,
|
| 633 |
+
0
|
| 634 |
+
],
|
| 635 |
+
[
|
| 636 |
+
0,
|
| 637 |
+
3,
|
| 638 |
+
73,
|
| 639 |
+
0
|
| 640 |
+
],
|
| 641 |
+
[
|
| 642 |
+
1,
|
| 643 |
+
0,
|
| 644 |
+
0,
|
| 645 |
+
64
|
| 646 |
+
]
|
| 647 |
+
],
|
| 648 |
+
"eval_f1": 0.8589767100678526,
|
| 649 |
+
"eval_loss": 0.7055637240409851,
|
| 650 |
+
"eval_precision": 0.8628397746044805,
|
| 651 |
+
"eval_recall": 0.8644688644688645,
|
| 652 |
+
"eval_runtime": 3.074,
|
| 653 |
+
"eval_samples_per_second": 88.809,
|
| 654 |
+
"eval_steps_per_second": 0.976,
|
| 655 |
+
"step": 76
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 18.0,
|
| 659 |
+
"eval_accuracy": 0.8644688644688645,
|
| 660 |
+
"eval_confusion_matrix": [
|
| 661 |
+
[
|
| 662 |
+
65,
|
| 663 |
+
2,
|
| 664 |
+
1,
|
| 665 |
+
4
|
| 666 |
+
],
|
| 667 |
+
[
|
| 668 |
+
13,
|
| 669 |
+
42,
|
| 670 |
+
5,
|
| 671 |
+
0
|
| 672 |
+
],
|
| 673 |
+
[
|
| 674 |
+
0,
|
| 675 |
+
12,
|
| 676 |
+
64,
|
| 677 |
+
0
|
| 678 |
+
],
|
| 679 |
+
[
|
| 680 |
+
0,
|
| 681 |
+
0,
|
| 682 |
+
0,
|
| 683 |
+
65
|
| 684 |
+
]
|
| 685 |
+
],
|
| 686 |
+
"eval_f1": 0.8627769756077204,
|
| 687 |
+
"eval_loss": 0.7563945651054382,
|
| 688 |
+
"eval_precision": 0.8634344261673453,
|
| 689 |
+
"eval_recall": 0.8644688644688645,
|
| 690 |
+
"eval_runtime": 3.0072,
|
| 691 |
+
"eval_samples_per_second": 90.783,
|
| 692 |
+
"eval_steps_per_second": 0.998,
|
| 693 |
+
"step": 81
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 18.88888888888889,
|
| 697 |
+
"eval_accuracy": 0.8424908424908425,
|
| 698 |
+
"eval_confusion_matrix": [
|
| 699 |
+
[
|
| 700 |
+
64,
|
| 701 |
+
4,
|
| 702 |
+
0,
|
| 703 |
+
4
|
| 704 |
+
],
|
| 705 |
+
[
|
| 706 |
+
11,
|
| 707 |
+
41,
|
| 708 |
+
8,
|
| 709 |
+
0
|
| 710 |
+
],
|
| 711 |
+
[
|
| 712 |
+
0,
|
| 713 |
+
10,
|
| 714 |
+
66,
|
| 715 |
+
0
|
| 716 |
+
],
|
| 717 |
+
[
|
| 718 |
+
6,
|
| 719 |
+
0,
|
| 720 |
+
0,
|
| 721 |
+
59
|
| 722 |
+
]
|
| 723 |
+
],
|
| 724 |
+
"eval_f1": 0.8418306879608031,
|
| 725 |
+
"eval_loss": 0.7825365662574768,
|
| 726 |
+
"eval_precision": 0.8434907006335578,
|
| 727 |
+
"eval_recall": 0.8424908424908425,
|
| 728 |
+
"eval_runtime": 3.0343,
|
| 729 |
+
"eval_samples_per_second": 89.972,
|
| 730 |
+
"eval_steps_per_second": 0.989,
|
| 731 |
+
"step": 85
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 20.0,
|
| 735 |
+
"eval_accuracy": 0.8058608058608059,
|
| 736 |
+
"eval_confusion_matrix": [
|
| 737 |
+
[
|
| 738 |
+
40,
|
| 739 |
+
28,
|
| 740 |
+
1,
|
| 741 |
+
3
|
| 742 |
+
],
|
| 743 |
+
[
|
| 744 |
+
2,
|
| 745 |
+
50,
|
| 746 |
+
8,
|
| 747 |
+
0
|
| 748 |
+
],
|
| 749 |
+
[
|
| 750 |
+
0,
|
| 751 |
+
11,
|
| 752 |
+
65,
|
| 753 |
+
0
|
| 754 |
+
],
|
| 755 |
+
[
|
| 756 |
+
0,
|
| 757 |
+
0,
|
| 758 |
+
0,
|
| 759 |
+
65
|
| 760 |
+
]
|
| 761 |
+
],
|
| 762 |
+
"eval_f1": 0.8065760931078588,
|
| 763 |
+
"eval_loss": 0.8426868915557861,
|
| 764 |
+
"eval_precision": 0.8467707085637404,
|
| 765 |
+
"eval_recall": 0.8058608058608059,
|
| 766 |
+
"eval_runtime": 3.0783,
|
| 767 |
+
"eval_samples_per_second": 88.686,
|
| 768 |
+
"eval_steps_per_second": 0.975,
|
| 769 |
+
"step": 90
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 20.88888888888889,
|
| 773 |
+
"eval_accuracy": 0.8498168498168498,
|
| 774 |
+
"eval_confusion_matrix": [
|
| 775 |
+
[
|
| 776 |
+
63,
|
| 777 |
+
4,
|
| 778 |
+
0,
|
| 779 |
+
5
|
| 780 |
+
],
|
| 781 |
+
[
|
| 782 |
+
13,
|
| 783 |
+
40,
|
| 784 |
+
7,
|
| 785 |
+
0
|
| 786 |
+
],
|
| 787 |
+
[
|
| 788 |
+
0,
|
| 789 |
+
12,
|
| 790 |
+
64,
|
| 791 |
+
0
|
| 792 |
+
],
|
| 793 |
+
[
|
| 794 |
+
0,
|
| 795 |
+
0,
|
| 796 |
+
0,
|
| 797 |
+
65
|
| 798 |
+
]
|
| 799 |
+
],
|
| 800 |
+
"eval_f1": 0.8477878057985963,
|
| 801 |
+
"eval_loss": 0.7440442442893982,
|
| 802 |
+
"eval_precision": 0.8476393351433065,
|
| 803 |
+
"eval_recall": 0.8498168498168498,
|
| 804 |
+
"eval_runtime": 2.9875,
|
| 805 |
+
"eval_samples_per_second": 91.38,
|
| 806 |
+
"eval_steps_per_second": 1.004,
|
| 807 |
+
"step": 94
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 22.0,
|
| 811 |
+
"eval_accuracy": 0.8608058608058609,
|
| 812 |
+
"eval_confusion_matrix": [
|
| 813 |
+
[
|
| 814 |
+
66,
|
| 815 |
+
1,
|
| 816 |
+
1,
|
| 817 |
+
4
|
| 818 |
+
],
|
| 819 |
+
[
|
| 820 |
+
13,
|
| 821 |
+
36,
|
| 822 |
+
11,
|
| 823 |
+
0
|
| 824 |
+
],
|
| 825 |
+
[
|
| 826 |
+
0,
|
| 827 |
+
5,
|
| 828 |
+
71,
|
| 829 |
+
0
|
| 830 |
+
],
|
| 831 |
+
[
|
| 832 |
+
3,
|
| 833 |
+
0,
|
| 834 |
+
0,
|
| 835 |
+
62
|
| 836 |
+
]
|
| 837 |
+
],
|
| 838 |
+
"eval_f1": 0.855194718990792,
|
| 839 |
+
"eval_loss": 0.7338178753852844,
|
| 840 |
+
"eval_precision": 0.8624631692093176,
|
| 841 |
+
"eval_recall": 0.8608058608058609,
|
| 842 |
+
"eval_runtime": 2.98,
|
| 843 |
+
"eval_samples_per_second": 91.611,
|
| 844 |
+
"eval_steps_per_second": 1.007,
|
| 845 |
+
"step": 99
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 22.88888888888889,
|
| 849 |
+
"eval_accuracy": 0.8498168498168498,
|
| 850 |
+
"eval_confusion_matrix": [
|
| 851 |
+
[
|
| 852 |
+
58,
|
| 853 |
+
10,
|
| 854 |
+
0,
|
| 855 |
+
4
|
| 856 |
+
],
|
| 857 |
+
[
|
| 858 |
+
10,
|
| 859 |
+
43,
|
| 860 |
+
7,
|
| 861 |
+
0
|
| 862 |
+
],
|
| 863 |
+
[
|
| 864 |
+
0,
|
| 865 |
+
7,
|
| 866 |
+
69,
|
| 867 |
+
0
|
| 868 |
+
],
|
| 869 |
+
[
|
| 870 |
+
2,
|
| 871 |
+
1,
|
| 872 |
+
0,
|
| 873 |
+
62
|
| 874 |
+
]
|
| 875 |
+
],
|
| 876 |
+
"eval_f1": 0.8497737987724402,
|
| 877 |
+
"eval_loss": 0.7231407761573792,
|
| 878 |
+
"eval_precision": 0.8498638493954653,
|
| 879 |
+
"eval_recall": 0.8498168498168498,
|
| 880 |
+
"eval_runtime": 2.9272,
|
| 881 |
+
"eval_samples_per_second": 93.262,
|
| 882 |
+
"eval_steps_per_second": 1.025,
|
| 883 |
+
"step": 103
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 24.0,
|
| 887 |
+
"eval_accuracy": 0.8424908424908425,
|
| 888 |
+
"eval_confusion_matrix": [
|
| 889 |
+
[
|
| 890 |
+
67,
|
| 891 |
+
1,
|
| 892 |
+
0,
|
| 893 |
+
4
|
| 894 |
+
],
|
| 895 |
+
[
|
| 896 |
+
12,
|
| 897 |
+
44,
|
| 898 |
+
4,
|
| 899 |
+
0
|
| 900 |
+
],
|
| 901 |
+
[
|
| 902 |
+
0,
|
| 903 |
+
18,
|
| 904 |
+
58,
|
| 905 |
+
0
|
| 906 |
+
],
|
| 907 |
+
[
|
| 908 |
+
2,
|
| 909 |
+
2,
|
| 910 |
+
0,
|
| 911 |
+
61
|
| 912 |
+
]
|
| 913 |
+
],
|
| 914 |
+
"eval_f1": 0.8431607380967995,
|
| 915 |
+
"eval_loss": 0.752363920211792,
|
| 916 |
+
"eval_precision": 0.8507974885146101,
|
| 917 |
+
"eval_recall": 0.8424908424908425,
|
| 918 |
+
"eval_runtime": 3.0257,
|
| 919 |
+
"eval_samples_per_second": 90.228,
|
| 920 |
+
"eval_steps_per_second": 0.992,
|
| 921 |
+
"step": 108
|
| 922 |
+
}
|
| 923 |
+
],
|
| 924 |
+
"logging_steps": 500,
|
| 925 |
+
"max_steps": 180,
|
| 926 |
+
"num_input_tokens_seen": 0,
|
| 927 |
+
"num_train_epochs": 45,
|
| 928 |
+
"save_steps": 500,
|
| 929 |
+
"stateful_callbacks": {
|
| 930 |
+
"TrainerControl": {
|
| 931 |
+
"args": {
|
| 932 |
+
"should_epoch_stop": false,
|
| 933 |
+
"should_evaluate": false,
|
| 934 |
+
"should_log": false,
|
| 935 |
+
"should_save": true,
|
| 936 |
+
"should_training_stop": false
|
| 937 |
+
},
|
| 938 |
+
"attributes": {}
|
| 939 |
+
}
|
| 940 |
+
},
|
| 941 |
+
"total_flos": 5.94937668096e+16,
|
| 942 |
+
"train_batch_size": 128,
|
| 943 |
+
"trial_name": null,
|
| 944 |
+
"trial_params": null
|
| 945 |
+
}
|
checkpoint-108/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:09e217cad0464b642c54f8b7691100b62a93dceb32a1869f55e3f0eb3a54a79e
|
| 3 |
+
size 5240
|
checkpoint-112/config.json
ADDED
|
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ntu-spml/distilhubert",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"HubertForSequenceClassification"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 1,
|
| 10 |
+
"classifier_proj_size": 256,
|
| 11 |
+
"conv_bias": false,
|
| 12 |
+
"conv_dim": [
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512
|
| 20 |
+
],
|
| 21 |
+
"conv_kernel": [
|
| 22 |
+
10,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
3,
|
| 27 |
+
2,
|
| 28 |
+
2
|
| 29 |
+
],
|
| 30 |
+
"conv_stride": [
|
| 31 |
+
5,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2,
|
| 37 |
+
2
|
| 38 |
+
],
|
| 39 |
+
"ctc_loss_reduction": "sum",
|
| 40 |
+
"ctc_zero_infinity": false,
|
| 41 |
+
"do_stable_layer_norm": false,
|
| 42 |
+
"eos_token_id": 2,
|
| 43 |
+
"feat_extract_activation": "gelu",
|
| 44 |
+
"feat_extract_norm": "group",
|
| 45 |
+
"feat_proj_dropout": 0.0,
|
| 46 |
+
"feat_proj_layer_norm": false,
|
| 47 |
+
"final_dropout": 0.0,
|
| 48 |
+
"finetuning_task": "audio-classification",
|
| 49 |
+
"hidden_act": "gelu",
|
| 50 |
+
"hidden_dropout": 0.1,
|
| 51 |
+
"hidden_size": 768,
|
| 52 |
+
"id2label": {
|
| 53 |
+
"0": "1s_normal",
|
| 54 |
+
"1": "1s_pain",
|
| 55 |
+
"2": "1s_hunger",
|
| 56 |
+
"3": "1s_asphyxia"
|
| 57 |
+
},
|
| 58 |
+
"initializer_range": 0.02,
|
| 59 |
+
"intermediate_size": 3072,
|
| 60 |
+
"label2id": {
|
| 61 |
+
"LABEL_0": 0,
|
| 62 |
+
"LABEL_1": 1,
|
| 63 |
+
"LABEL_2": 2,
|
| 64 |
+
"LABEL_3": 3
|
| 65 |
+
},
|
| 66 |
+
"layer_norm_eps": 1e-05,
|
| 67 |
+
"layerdrop": 0.0,
|
| 68 |
+
"mask_feature_length": 10,
|
| 69 |
+
"mask_feature_min_masks": 0,
|
| 70 |
+
"mask_feature_prob": 0.0,
|
| 71 |
+
"mask_time_length": 10,
|
| 72 |
+
"mask_time_min_masks": 2,
|
| 73 |
+
"mask_time_prob": 0.05,
|
| 74 |
+
"model_type": "hubert",
|
| 75 |
+
"num_attention_heads": 12,
|
| 76 |
+
"num_conv_pos_embedding_groups": 16,
|
| 77 |
+
"num_conv_pos_embeddings": 128,
|
| 78 |
+
"num_feat_extract_layers": 7,
|
| 79 |
+
"num_hidden_layers": 2,
|
| 80 |
+
"pad_token_id": 0,
|
| 81 |
+
"torch_dtype": "float32",
|
| 82 |
+
"transformers_version": "4.44.2",
|
| 83 |
+
"use_weighted_layer_sum": false,
|
| 84 |
+
"vocab_size": 32
|
| 85 |
+
}
|
checkpoint-112/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:55f1a1837a7c7f7a1f7cace5f5cae81f177ce53ed0076b7d4272fc34489ac246
|
| 3 |
+
size 94765560
|
checkpoint-112/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:52cbe97bfd5315f6a8903cf00166e6c8d6bf8f36344f973060f64a4bc98ceeaa
|
| 3 |
+
size 189556666
|
checkpoint-112/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3933a3c2e5fe0278cc211a89b2796b02c68fc53683a62ae85982e3d2c0703c13
|
| 3 |
+
size 14308
|
checkpoint-112/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2b6a6538f7897eef4a81ad512c8b5f7c4391b82687409a25f8f49aad850597e4
|
| 3 |
+
size 1064
|
checkpoint-112/trainer_state.json
ADDED
|
@@ -0,0 +1,983 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8627769756077204,
|
| 3 |
+
"best_model_checkpoint": "distilhubert-finetuned-mixed-data/checkpoint-81",
|
| 4 |
+
"epoch": 24.88888888888889,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 112,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.8888888888888888,
|
| 13 |
+
"eval_accuracy": 0.63003663003663,
|
| 14 |
+
"eval_confusion_matrix": [
|
| 15 |
+
[
|
| 16 |
+
53,
|
| 17 |
+
0,
|
| 18 |
+
16,
|
| 19 |
+
3
|
| 20 |
+
],
|
| 21 |
+
[
|
| 22 |
+
28,
|
| 23 |
+
1,
|
| 24 |
+
31,
|
| 25 |
+
0
|
| 26 |
+
],
|
| 27 |
+
[
|
| 28 |
+
13,
|
| 29 |
+
0,
|
| 30 |
+
62,
|
| 31 |
+
1
|
| 32 |
+
],
|
| 33 |
+
[
|
| 34 |
+
3,
|
| 35 |
+
0,
|
| 36 |
+
6,
|
| 37 |
+
56
|
| 38 |
+
]
|
| 39 |
+
],
|
| 40 |
+
"eval_f1": 0.566693372031096,
|
| 41 |
+
"eval_loss": 1.31327486038208,
|
| 42 |
+
"eval_precision": 0.7361933549293478,
|
| 43 |
+
"eval_recall": 0.63003663003663,
|
| 44 |
+
"eval_runtime": 2.8436,
|
| 45 |
+
"eval_samples_per_second": 96.005,
|
| 46 |
+
"eval_steps_per_second": 1.055,
|
| 47 |
+
"step": 4
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 2.0,
|
| 51 |
+
"eval_accuracy": 0.6593406593406593,
|
| 52 |
+
"eval_confusion_matrix": [
|
| 53 |
+
[
|
| 54 |
+
41,
|
| 55 |
+
10,
|
| 56 |
+
16,
|
| 57 |
+
5
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
21,
|
| 61 |
+
1,
|
| 62 |
+
38,
|
| 63 |
+
0
|
| 64 |
+
],
|
| 65 |
+
[
|
| 66 |
+
1,
|
| 67 |
+
2,
|
| 68 |
+
73,
|
| 69 |
+
0
|
| 70 |
+
],
|
| 71 |
+
[
|
| 72 |
+
0,
|
| 73 |
+
0,
|
| 74 |
+
0,
|
| 75 |
+
65
|
| 76 |
+
]
|
| 77 |
+
],
|
| 78 |
+
"eval_f1": 0.595713773130153,
|
| 79 |
+
"eval_loss": 0.9242589473724365,
|
| 80 |
+
"eval_precision": 0.5696509512811321,
|
| 81 |
+
"eval_recall": 0.6593406593406593,
|
| 82 |
+
"eval_runtime": 2.8668,
|
| 83 |
+
"eval_samples_per_second": 95.227,
|
| 84 |
+
"eval_steps_per_second": 1.046,
|
| 85 |
+
"step": 9
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 2.888888888888889,
|
| 89 |
+
"eval_accuracy": 0.717948717948718,
|
| 90 |
+
"eval_confusion_matrix": [
|
| 91 |
+
[
|
| 92 |
+
52,
|
| 93 |
+
8,
|
| 94 |
+
7,
|
| 95 |
+
5
|
| 96 |
+
],
|
| 97 |
+
[
|
| 98 |
+
18,
|
| 99 |
+
5,
|
| 100 |
+
37,
|
| 101 |
+
0
|
| 102 |
+
],
|
| 103 |
+
[
|
| 104 |
+
1,
|
| 105 |
+
0,
|
| 106 |
+
75,
|
| 107 |
+
0
|
| 108 |
+
],
|
| 109 |
+
[
|
| 110 |
+
1,
|
| 111 |
+
0,
|
| 112 |
+
0,
|
| 113 |
+
64
|
| 114 |
+
]
|
| 115 |
+
],
|
| 116 |
+
"eval_f1": 0.6621621567125869,
|
| 117 |
+
"eval_loss": 0.8232662081718445,
|
| 118 |
+
"eval_precision": 0.671303801513745,
|
| 119 |
+
"eval_recall": 0.717948717948718,
|
| 120 |
+
"eval_runtime": 2.8711,
|
| 121 |
+
"eval_samples_per_second": 95.084,
|
| 122 |
+
"eval_steps_per_second": 1.045,
|
| 123 |
+
"step": 13
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 4.0,
|
| 127 |
+
"eval_accuracy": 0.6959706959706959,
|
| 128 |
+
"eval_confusion_matrix": [
|
| 129 |
+
[
|
| 130 |
+
45,
|
| 131 |
+
9,
|
| 132 |
+
13,
|
| 133 |
+
5
|
| 134 |
+
],
|
| 135 |
+
[
|
| 136 |
+
13,
|
| 137 |
+
5,
|
| 138 |
+
42,
|
| 139 |
+
0
|
| 140 |
+
],
|
| 141 |
+
[
|
| 142 |
+
1,
|
| 143 |
+
0,
|
| 144 |
+
75,
|
| 145 |
+
0
|
| 146 |
+
],
|
| 147 |
+
[
|
| 148 |
+
0,
|
| 149 |
+
0,
|
| 150 |
+
0,
|
| 151 |
+
65
|
| 152 |
+
]
|
| 153 |
+
],
|
| 154 |
+
"eval_f1": 0.6428796900399459,
|
| 155 |
+
"eval_loss": 0.951453447341919,
|
| 156 |
+
"eval_precision": 0.6613447705829899,
|
| 157 |
+
"eval_recall": 0.6959706959706959,
|
| 158 |
+
"eval_runtime": 2.8813,
|
| 159 |
+
"eval_samples_per_second": 94.75,
|
| 160 |
+
"eval_steps_per_second": 1.041,
|
| 161 |
+
"step": 18
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 4.888888888888889,
|
| 165 |
+
"eval_accuracy": 0.6666666666666666,
|
| 166 |
+
"eval_confusion_matrix": [
|
| 167 |
+
[
|
| 168 |
+
16,
|
| 169 |
+
49,
|
| 170 |
+
3,
|
| 171 |
+
4
|
| 172 |
+
],
|
| 173 |
+
[
|
| 174 |
+
1,
|
| 175 |
+
26,
|
| 176 |
+
33,
|
| 177 |
+
0
|
| 178 |
+
],
|
| 179 |
+
[
|
| 180 |
+
0,
|
| 181 |
+
1,
|
| 182 |
+
75,
|
| 183 |
+
0
|
| 184 |
+
],
|
| 185 |
+
[
|
| 186 |
+
0,
|
| 187 |
+
0,
|
| 188 |
+
0,
|
| 189 |
+
65
|
| 190 |
+
]
|
| 191 |
+
],
|
| 192 |
+
"eval_f1": 0.6331541584523206,
|
| 193 |
+
"eval_loss": 1.0080645084381104,
|
| 194 |
+
"eval_precision": 0.7358031394913882,
|
| 195 |
+
"eval_recall": 0.6666666666666666,
|
| 196 |
+
"eval_runtime": 2.9108,
|
| 197 |
+
"eval_samples_per_second": 93.789,
|
| 198 |
+
"eval_steps_per_second": 1.031,
|
| 199 |
+
"step": 22
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 6.0,
|
| 203 |
+
"eval_accuracy": 0.652014652014652,
|
| 204 |
+
"eval_confusion_matrix": [
|
| 205 |
+
[
|
| 206 |
+
66,
|
| 207 |
+
0,
|
| 208 |
+
0,
|
| 209 |
+
6
|
| 210 |
+
],
|
| 211 |
+
[
|
| 212 |
+
36,
|
| 213 |
+
16,
|
| 214 |
+
8,
|
| 215 |
+
0
|
| 216 |
+
],
|
| 217 |
+
[
|
| 218 |
+
10,
|
| 219 |
+
35,
|
| 220 |
+
31,
|
| 221 |
+
0
|
| 222 |
+
],
|
| 223 |
+
[
|
| 224 |
+
0,
|
| 225 |
+
0,
|
| 226 |
+
0,
|
| 227 |
+
65
|
| 228 |
+
]
|
| 229 |
+
],
|
| 230 |
+
"eval_f1": 0.6302407955860642,
|
| 231 |
+
"eval_loss": 0.9598046541213989,
|
| 232 |
+
"eval_precision": 0.6636241740077424,
|
| 233 |
+
"eval_recall": 0.652014652014652,
|
| 234 |
+
"eval_runtime": 2.9773,
|
| 235 |
+
"eval_samples_per_second": 91.693,
|
| 236 |
+
"eval_steps_per_second": 1.008,
|
| 237 |
+
"step": 27
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 6.888888888888889,
|
| 241 |
+
"eval_accuracy": 0.7692307692307693,
|
| 242 |
+
"eval_confusion_matrix": [
|
| 243 |
+
[
|
| 244 |
+
66,
|
| 245 |
+
0,
|
| 246 |
+
0,
|
| 247 |
+
6
|
| 248 |
+
],
|
| 249 |
+
[
|
| 250 |
+
29,
|
| 251 |
+
21,
|
| 252 |
+
10,
|
| 253 |
+
0
|
| 254 |
+
],
|
| 255 |
+
[
|
| 256 |
+
3,
|
| 257 |
+
15,
|
| 258 |
+
58,
|
| 259 |
+
0
|
| 260 |
+
],
|
| 261 |
+
[
|
| 262 |
+
0,
|
| 263 |
+
0,
|
| 264 |
+
0,
|
| 265 |
+
65
|
| 266 |
+
]
|
| 267 |
+
],
|
| 268 |
+
"eval_f1": 0.7527855586679116,
|
| 269 |
+
"eval_loss": 0.7898163795471191,
|
| 270 |
+
"eval_precision": 0.7612467677056892,
|
| 271 |
+
"eval_recall": 0.7692307692307693,
|
| 272 |
+
"eval_runtime": 2.8918,
|
| 273 |
+
"eval_samples_per_second": 94.404,
|
| 274 |
+
"eval_steps_per_second": 1.037,
|
| 275 |
+
"step": 31
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 8.0,
|
| 279 |
+
"eval_accuracy": 0.7802197802197802,
|
| 280 |
+
"eval_confusion_matrix": [
|
| 281 |
+
[
|
| 282 |
+
66,
|
| 283 |
+
0,
|
| 284 |
+
0,
|
| 285 |
+
6
|
| 286 |
+
],
|
| 287 |
+
[
|
| 288 |
+
25,
|
| 289 |
+
15,
|
| 290 |
+
19,
|
| 291 |
+
1
|
| 292 |
+
],
|
| 293 |
+
[
|
| 294 |
+
1,
|
| 295 |
+
6,
|
| 296 |
+
68,
|
| 297 |
+
1
|
| 298 |
+
],
|
| 299 |
+
[
|
| 300 |
+
1,
|
| 301 |
+
0,
|
| 302 |
+
0,
|
| 303 |
+
64
|
| 304 |
+
]
|
| 305 |
+
],
|
| 306 |
+
"eval_f1": 0.7471179200524057,
|
| 307 |
+
"eval_loss": 0.7337484359741211,
|
| 308 |
+
"eval_precision": 0.7733845922309294,
|
| 309 |
+
"eval_recall": 0.7802197802197802,
|
| 310 |
+
"eval_runtime": 2.9006,
|
| 311 |
+
"eval_samples_per_second": 94.119,
|
| 312 |
+
"eval_steps_per_second": 1.034,
|
| 313 |
+
"step": 36
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 8.88888888888889,
|
| 317 |
+
"eval_accuracy": 0.7912087912087912,
|
| 318 |
+
"eval_confusion_matrix": [
|
| 319 |
+
[
|
| 320 |
+
57,
|
| 321 |
+
8,
|
| 322 |
+
3,
|
| 323 |
+
4
|
| 324 |
+
],
|
| 325 |
+
[
|
| 326 |
+
13,
|
| 327 |
+
24,
|
| 328 |
+
23,
|
| 329 |
+
0
|
| 330 |
+
],
|
| 331 |
+
[
|
| 332 |
+
0,
|
| 333 |
+
5,
|
| 334 |
+
71,
|
| 335 |
+
0
|
| 336 |
+
],
|
| 337 |
+
[
|
| 338 |
+
1,
|
| 339 |
+
0,
|
| 340 |
+
0,
|
| 341 |
+
64
|
| 342 |
+
]
|
| 343 |
+
],
|
| 344 |
+
"eval_f1": 0.7766564722186922,
|
| 345 |
+
"eval_loss": 0.7148727178573608,
|
| 346 |
+
"eval_precision": 0.7821504483074875,
|
| 347 |
+
"eval_recall": 0.7912087912087912,
|
| 348 |
+
"eval_runtime": 2.8866,
|
| 349 |
+
"eval_samples_per_second": 94.576,
|
| 350 |
+
"eval_steps_per_second": 1.039,
|
| 351 |
+
"step": 40
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 10.0,
|
| 355 |
+
"eval_accuracy": 0.8168498168498168,
|
| 356 |
+
"eval_confusion_matrix": [
|
| 357 |
+
[
|
| 358 |
+
51,
|
| 359 |
+
8,
|
| 360 |
+
1,
|
| 361 |
+
12
|
| 362 |
+
],
|
| 363 |
+
[
|
| 364 |
+
10,
|
| 365 |
+
36,
|
| 366 |
+
14,
|
| 367 |
+
0
|
| 368 |
+
],
|
| 369 |
+
[
|
| 370 |
+
0,
|
| 371 |
+
5,
|
| 372 |
+
71,
|
| 373 |
+
0
|
| 374 |
+
],
|
| 375 |
+
[
|
| 376 |
+
0,
|
| 377 |
+
0,
|
| 378 |
+
0,
|
| 379 |
+
65
|
| 380 |
+
]
|
| 381 |
+
],
|
| 382 |
+
"eval_f1": 0.8094336953840884,
|
| 383 |
+
"eval_loss": 0.7574812173843384,
|
| 384 |
+
"eval_precision": 0.8127936625684181,
|
| 385 |
+
"eval_recall": 0.8168498168498168,
|
| 386 |
+
"eval_runtime": 2.8967,
|
| 387 |
+
"eval_samples_per_second": 94.245,
|
| 388 |
+
"eval_steps_per_second": 1.036,
|
| 389 |
+
"step": 45
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 10.88888888888889,
|
| 393 |
+
"eval_accuracy": 0.8021978021978022,
|
| 394 |
+
"eval_confusion_matrix": [
|
| 395 |
+
[
|
| 396 |
+
48,
|
| 397 |
+
18,
|
| 398 |
+
2,
|
| 399 |
+
4
|
| 400 |
+
],
|
| 401 |
+
[
|
| 402 |
+
4,
|
| 403 |
+
41,
|
| 404 |
+
15,
|
| 405 |
+
0
|
| 406 |
+
],
|
| 407 |
+
[
|
| 408 |
+
0,
|
| 409 |
+
7,
|
| 410 |
+
69,
|
| 411 |
+
0
|
| 412 |
+
],
|
| 413 |
+
[
|
| 414 |
+
3,
|
| 415 |
+
0,
|
| 416 |
+
1,
|
| 417 |
+
61
|
| 418 |
+
]
|
| 419 |
+
],
|
| 420 |
+
"eval_f1": 0.801525180147331,
|
| 421 |
+
"eval_loss": 0.7140281200408936,
|
| 422 |
+
"eval_precision": 0.8109338936925145,
|
| 423 |
+
"eval_recall": 0.8021978021978022,
|
| 424 |
+
"eval_runtime": 2.917,
|
| 425 |
+
"eval_samples_per_second": 93.589,
|
| 426 |
+
"eval_steps_per_second": 1.028,
|
| 427 |
+
"step": 49
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 12.0,
|
| 431 |
+
"eval_accuracy": 0.8534798534798534,
|
| 432 |
+
"eval_confusion_matrix": [
|
| 433 |
+
[
|
| 434 |
+
56,
|
| 435 |
+
12,
|
| 436 |
+
0,
|
| 437 |
+
4
|
| 438 |
+
],
|
| 439 |
+
[
|
| 440 |
+
6,
|
| 441 |
+
43,
|
| 442 |
+
11,
|
| 443 |
+
0
|
| 444 |
+
],
|
| 445 |
+
[
|
| 446 |
+
0,
|
| 447 |
+
6,
|
| 448 |
+
70,
|
| 449 |
+
0
|
| 450 |
+
],
|
| 451 |
+
[
|
| 452 |
+
1,
|
| 453 |
+
0,
|
| 454 |
+
0,
|
| 455 |
+
64
|
| 456 |
+
]
|
| 457 |
+
],
|
| 458 |
+
"eval_f1": 0.8523994617102314,
|
| 459 |
+
"eval_loss": 0.6672152280807495,
|
| 460 |
+
"eval_precision": 0.8540313732642031,
|
| 461 |
+
"eval_recall": 0.8534798534798534,
|
| 462 |
+
"eval_runtime": 2.8969,
|
| 463 |
+
"eval_samples_per_second": 94.238,
|
| 464 |
+
"eval_steps_per_second": 1.036,
|
| 465 |
+
"step": 54
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 12.88888888888889,
|
| 469 |
+
"eval_accuracy": 0.8498168498168498,
|
| 470 |
+
"eval_confusion_matrix": [
|
| 471 |
+
[
|
| 472 |
+
60,
|
| 473 |
+
8,
|
| 474 |
+
0,
|
| 475 |
+
4
|
| 476 |
+
],
|
| 477 |
+
[
|
| 478 |
+
8,
|
| 479 |
+
38,
|
| 480 |
+
14,
|
| 481 |
+
0
|
| 482 |
+
],
|
| 483 |
+
[
|
| 484 |
+
0,
|
| 485 |
+
6,
|
| 486 |
+
70,
|
| 487 |
+
0
|
| 488 |
+
],
|
| 489 |
+
[
|
| 490 |
+
1,
|
| 491 |
+
0,
|
| 492 |
+
0,
|
| 493 |
+
64
|
| 494 |
+
]
|
| 495 |
+
],
|
| 496 |
+
"eval_f1": 0.8463270052615757,
|
| 497 |
+
"eval_loss": 0.6432910561561584,
|
| 498 |
+
"eval_precision": 0.8460243715014519,
|
| 499 |
+
"eval_recall": 0.8498168498168498,
|
| 500 |
+
"eval_runtime": 3.0281,
|
| 501 |
+
"eval_samples_per_second": 90.156,
|
| 502 |
+
"eval_steps_per_second": 0.991,
|
| 503 |
+
"step": 58
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 14.0,
|
| 507 |
+
"eval_accuracy": 0.8278388278388278,
|
| 508 |
+
"eval_confusion_matrix": [
|
| 509 |
+
[
|
| 510 |
+
54,
|
| 511 |
+
13,
|
| 512 |
+
1,
|
| 513 |
+
4
|
| 514 |
+
],
|
| 515 |
+
[
|
| 516 |
+
7,
|
| 517 |
+
44,
|
| 518 |
+
9,
|
| 519 |
+
0
|
| 520 |
+
],
|
| 521 |
+
[
|
| 522 |
+
0,
|
| 523 |
+
11,
|
| 524 |
+
65,
|
| 525 |
+
0
|
| 526 |
+
],
|
| 527 |
+
[
|
| 528 |
+
1,
|
| 529 |
+
1,
|
| 530 |
+
0,
|
| 531 |
+
63
|
| 532 |
+
]
|
| 533 |
+
],
|
| 534 |
+
"eval_f1": 0.82943590265942,
|
| 535 |
+
"eval_loss": 0.7395206093788147,
|
| 536 |
+
"eval_precision": 0.8350059217447382,
|
| 537 |
+
"eval_recall": 0.8278388278388278,
|
| 538 |
+
"eval_runtime": 2.9869,
|
| 539 |
+
"eval_samples_per_second": 91.399,
|
| 540 |
+
"eval_steps_per_second": 1.004,
|
| 541 |
+
"step": 63
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 14.88888888888889,
|
| 545 |
+
"eval_accuracy": 0.8315018315018315,
|
| 546 |
+
"eval_confusion_matrix": [
|
| 547 |
+
[
|
| 548 |
+
54,
|
| 549 |
+
13,
|
| 550 |
+
1,
|
| 551 |
+
4
|
| 552 |
+
],
|
| 553 |
+
[
|
| 554 |
+
10,
|
| 555 |
+
39,
|
| 556 |
+
11,
|
| 557 |
+
0
|
| 558 |
+
],
|
| 559 |
+
[
|
| 560 |
+
0,
|
| 561 |
+
7,
|
| 562 |
+
69,
|
| 563 |
+
0
|
| 564 |
+
],
|
| 565 |
+
[
|
| 566 |
+
0,
|
| 567 |
+
0,
|
| 568 |
+
0,
|
| 569 |
+
65
|
| 570 |
+
]
|
| 571 |
+
],
|
| 572 |
+
"eval_f1": 0.8291811389886823,
|
| 573 |
+
"eval_loss": 0.7115849852561951,
|
| 574 |
+
"eval_precision": 0.829244108966536,
|
| 575 |
+
"eval_recall": 0.8315018315018315,
|
| 576 |
+
"eval_runtime": 2.9616,
|
| 577 |
+
"eval_samples_per_second": 92.181,
|
| 578 |
+
"eval_steps_per_second": 1.013,
|
| 579 |
+
"step": 67
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 16.0,
|
| 583 |
+
"eval_accuracy": 0.8315018315018315,
|
| 584 |
+
"eval_confusion_matrix": [
|
| 585 |
+
[
|
| 586 |
+
60,
|
| 587 |
+
7,
|
| 588 |
+
1,
|
| 589 |
+
4
|
| 590 |
+
],
|
| 591 |
+
[
|
| 592 |
+
10,
|
| 593 |
+
39,
|
| 594 |
+
11,
|
| 595 |
+
0
|
| 596 |
+
],
|
| 597 |
+
[
|
| 598 |
+
0,
|
| 599 |
+
11,
|
| 600 |
+
65,
|
| 601 |
+
0
|
| 602 |
+
],
|
| 603 |
+
[
|
| 604 |
+
1,
|
| 605 |
+
1,
|
| 606 |
+
0,
|
| 607 |
+
63
|
| 608 |
+
]
|
| 609 |
+
],
|
| 610 |
+
"eval_f1": 0.8304073820984628,
|
| 611 |
+
"eval_loss": 0.7295921444892883,
|
| 612 |
+
"eval_precision": 0.8295426562258641,
|
| 613 |
+
"eval_recall": 0.8315018315018315,
|
| 614 |
+
"eval_runtime": 2.8758,
|
| 615 |
+
"eval_samples_per_second": 94.932,
|
| 616 |
+
"eval_steps_per_second": 1.043,
|
| 617 |
+
"step": 72
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 16.88888888888889,
|
| 621 |
+
"eval_accuracy": 0.8644688644688645,
|
| 622 |
+
"eval_confusion_matrix": [
|
| 623 |
+
[
|
| 624 |
+
62,
|
| 625 |
+
5,
|
| 626 |
+
1,
|
| 627 |
+
4
|
| 628 |
+
],
|
| 629 |
+
[
|
| 630 |
+
9,
|
| 631 |
+
37,
|
| 632 |
+
14,
|
| 633 |
+
0
|
| 634 |
+
],
|
| 635 |
+
[
|
| 636 |
+
0,
|
| 637 |
+
3,
|
| 638 |
+
73,
|
| 639 |
+
0
|
| 640 |
+
],
|
| 641 |
+
[
|
| 642 |
+
1,
|
| 643 |
+
0,
|
| 644 |
+
0,
|
| 645 |
+
64
|
| 646 |
+
]
|
| 647 |
+
],
|
| 648 |
+
"eval_f1": 0.8589767100678526,
|
| 649 |
+
"eval_loss": 0.7055637240409851,
|
| 650 |
+
"eval_precision": 0.8628397746044805,
|
| 651 |
+
"eval_recall": 0.8644688644688645,
|
| 652 |
+
"eval_runtime": 3.074,
|
| 653 |
+
"eval_samples_per_second": 88.809,
|
| 654 |
+
"eval_steps_per_second": 0.976,
|
| 655 |
+
"step": 76
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 18.0,
|
| 659 |
+
"eval_accuracy": 0.8644688644688645,
|
| 660 |
+
"eval_confusion_matrix": [
|
| 661 |
+
[
|
| 662 |
+
65,
|
| 663 |
+
2,
|
| 664 |
+
1,
|
| 665 |
+
4
|
| 666 |
+
],
|
| 667 |
+
[
|
| 668 |
+
13,
|
| 669 |
+
42,
|
| 670 |
+
5,
|
| 671 |
+
0
|
| 672 |
+
],
|
| 673 |
+
[
|
| 674 |
+
0,
|
| 675 |
+
12,
|
| 676 |
+
64,
|
| 677 |
+
0
|
| 678 |
+
],
|
| 679 |
+
[
|
| 680 |
+
0,
|
| 681 |
+
0,
|
| 682 |
+
0,
|
| 683 |
+
65
|
| 684 |
+
]
|
| 685 |
+
],
|
| 686 |
+
"eval_f1": 0.8627769756077204,
|
| 687 |
+
"eval_loss": 0.7563945651054382,
|
| 688 |
+
"eval_precision": 0.8634344261673453,
|
| 689 |
+
"eval_recall": 0.8644688644688645,
|
| 690 |
+
"eval_runtime": 3.0072,
|
| 691 |
+
"eval_samples_per_second": 90.783,
|
| 692 |
+
"eval_steps_per_second": 0.998,
|
| 693 |
+
"step": 81
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 18.88888888888889,
|
| 697 |
+
"eval_accuracy": 0.8424908424908425,
|
| 698 |
+
"eval_confusion_matrix": [
|
| 699 |
+
[
|
| 700 |
+
64,
|
| 701 |
+
4,
|
| 702 |
+
0,
|
| 703 |
+
4
|
| 704 |
+
],
|
| 705 |
+
[
|
| 706 |
+
11,
|
| 707 |
+
41,
|
| 708 |
+
8,
|
| 709 |
+
0
|
| 710 |
+
],
|
| 711 |
+
[
|
| 712 |
+
0,
|
| 713 |
+
10,
|
| 714 |
+
66,
|
| 715 |
+
0
|
| 716 |
+
],
|
| 717 |
+
[
|
| 718 |
+
6,
|
| 719 |
+
0,
|
| 720 |
+
0,
|
| 721 |
+
59
|
| 722 |
+
]
|
| 723 |
+
],
|
| 724 |
+
"eval_f1": 0.8418306879608031,
|
| 725 |
+
"eval_loss": 0.7825365662574768,
|
| 726 |
+
"eval_precision": 0.8434907006335578,
|
| 727 |
+
"eval_recall": 0.8424908424908425,
|
| 728 |
+
"eval_runtime": 3.0343,
|
| 729 |
+
"eval_samples_per_second": 89.972,
|
| 730 |
+
"eval_steps_per_second": 0.989,
|
| 731 |
+
"step": 85
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 20.0,
|
| 735 |
+
"eval_accuracy": 0.8058608058608059,
|
| 736 |
+
"eval_confusion_matrix": [
|
| 737 |
+
[
|
| 738 |
+
40,
|
| 739 |
+
28,
|
| 740 |
+
1,
|
| 741 |
+
3
|
| 742 |
+
],
|
| 743 |
+
[
|
| 744 |
+
2,
|
| 745 |
+
50,
|
| 746 |
+
8,
|
| 747 |
+
0
|
| 748 |
+
],
|
| 749 |
+
[
|
| 750 |
+
0,
|
| 751 |
+
11,
|
| 752 |
+
65,
|
| 753 |
+
0
|
| 754 |
+
],
|
| 755 |
+
[
|
| 756 |
+
0,
|
| 757 |
+
0,
|
| 758 |
+
0,
|
| 759 |
+
65
|
| 760 |
+
]
|
| 761 |
+
],
|
| 762 |
+
"eval_f1": 0.8065760931078588,
|
| 763 |
+
"eval_loss": 0.8426868915557861,
|
| 764 |
+
"eval_precision": 0.8467707085637404,
|
| 765 |
+
"eval_recall": 0.8058608058608059,
|
| 766 |
+
"eval_runtime": 3.0783,
|
| 767 |
+
"eval_samples_per_second": 88.686,
|
| 768 |
+
"eval_steps_per_second": 0.975,
|
| 769 |
+
"step": 90
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 20.88888888888889,
|
| 773 |
+
"eval_accuracy": 0.8498168498168498,
|
| 774 |
+
"eval_confusion_matrix": [
|
| 775 |
+
[
|
| 776 |
+
63,
|
| 777 |
+
4,
|
| 778 |
+
0,
|
| 779 |
+
5
|
| 780 |
+
],
|
| 781 |
+
[
|
| 782 |
+
13,
|
| 783 |
+
40,
|
| 784 |
+
7,
|
| 785 |
+
0
|
| 786 |
+
],
|
| 787 |
+
[
|
| 788 |
+
0,
|
| 789 |
+
12,
|
| 790 |
+
64,
|
| 791 |
+
0
|
| 792 |
+
],
|
| 793 |
+
[
|
| 794 |
+
0,
|
| 795 |
+
0,
|
| 796 |
+
0,
|
| 797 |
+
65
|
| 798 |
+
]
|
| 799 |
+
],
|
| 800 |
+
"eval_f1": 0.8477878057985963,
|
| 801 |
+
"eval_loss": 0.7440442442893982,
|
| 802 |
+
"eval_precision": 0.8476393351433065,
|
| 803 |
+
"eval_recall": 0.8498168498168498,
|
| 804 |
+
"eval_runtime": 2.9875,
|
| 805 |
+
"eval_samples_per_second": 91.38,
|
| 806 |
+
"eval_steps_per_second": 1.004,
|
| 807 |
+
"step": 94
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 22.0,
|
| 811 |
+
"eval_accuracy": 0.8608058608058609,
|
| 812 |
+
"eval_confusion_matrix": [
|
| 813 |
+
[
|
| 814 |
+
66,
|
| 815 |
+
1,
|
| 816 |
+
1,
|
| 817 |
+
4
|
| 818 |
+
],
|
| 819 |
+
[
|
| 820 |
+
13,
|
| 821 |
+
36,
|
| 822 |
+
11,
|
| 823 |
+
0
|
| 824 |
+
],
|
| 825 |
+
[
|
| 826 |
+
0,
|
| 827 |
+
5,
|
| 828 |
+
71,
|
| 829 |
+
0
|
| 830 |
+
],
|
| 831 |
+
[
|
| 832 |
+
3,
|
| 833 |
+
0,
|
| 834 |
+
0,
|
| 835 |
+
62
|
| 836 |
+
]
|
| 837 |
+
],
|
| 838 |
+
"eval_f1": 0.855194718990792,
|
| 839 |
+
"eval_loss": 0.7338178753852844,
|
| 840 |
+
"eval_precision": 0.8624631692093176,
|
| 841 |
+
"eval_recall": 0.8608058608058609,
|
| 842 |
+
"eval_runtime": 2.98,
|
| 843 |
+
"eval_samples_per_second": 91.611,
|
| 844 |
+
"eval_steps_per_second": 1.007,
|
| 845 |
+
"step": 99
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 22.88888888888889,
|
| 849 |
+
"eval_accuracy": 0.8498168498168498,
|
| 850 |
+
"eval_confusion_matrix": [
|
| 851 |
+
[
|
| 852 |
+
58,
|
| 853 |
+
10,
|
| 854 |
+
0,
|
| 855 |
+
4
|
| 856 |
+
],
|
| 857 |
+
[
|
| 858 |
+
10,
|
| 859 |
+
43,
|
| 860 |
+
7,
|
| 861 |
+
0
|
| 862 |
+
],
|
| 863 |
+
[
|
| 864 |
+
0,
|
| 865 |
+
7,
|
| 866 |
+
69,
|
| 867 |
+
0
|
| 868 |
+
],
|
| 869 |
+
[
|
| 870 |
+
2,
|
| 871 |
+
1,
|
| 872 |
+
0,
|
| 873 |
+
62
|
| 874 |
+
]
|
| 875 |
+
],
|
| 876 |
+
"eval_f1": 0.8497737987724402,
|
| 877 |
+
"eval_loss": 0.7231407761573792,
|
| 878 |
+
"eval_precision": 0.8498638493954653,
|
| 879 |
+
"eval_recall": 0.8498168498168498,
|
| 880 |
+
"eval_runtime": 2.9272,
|
| 881 |
+
"eval_samples_per_second": 93.262,
|
| 882 |
+
"eval_steps_per_second": 1.025,
|
| 883 |
+
"step": 103
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 24.0,
|
| 887 |
+
"eval_accuracy": 0.8424908424908425,
|
| 888 |
+
"eval_confusion_matrix": [
|
| 889 |
+
[
|
| 890 |
+
67,
|
| 891 |
+
1,
|
| 892 |
+
0,
|
| 893 |
+
4
|
| 894 |
+
],
|
| 895 |
+
[
|
| 896 |
+
12,
|
| 897 |
+
44,
|
| 898 |
+
4,
|
| 899 |
+
0
|
| 900 |
+
],
|
| 901 |
+
[
|
| 902 |
+
0,
|
| 903 |
+
18,
|
| 904 |
+
58,
|
| 905 |
+
0
|
| 906 |
+
],
|
| 907 |
+
[
|
| 908 |
+
2,
|
| 909 |
+
2,
|
| 910 |
+
0,
|
| 911 |
+
61
|
| 912 |
+
]
|
| 913 |
+
],
|
| 914 |
+
"eval_f1": 0.8431607380967995,
|
| 915 |
+
"eval_loss": 0.752363920211792,
|
| 916 |
+
"eval_precision": 0.8507974885146101,
|
| 917 |
+
"eval_recall": 0.8424908424908425,
|
| 918 |
+
"eval_runtime": 3.0257,
|
| 919 |
+
"eval_samples_per_second": 90.228,
|
| 920 |
+
"eval_steps_per_second": 0.992,
|
| 921 |
+
"step": 108
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 24.88888888888889,
|
| 925 |
+
"eval_accuracy": 0.8498168498168498,
|
| 926 |
+
"eval_confusion_matrix": [
|
| 927 |
+
[
|
| 928 |
+
66,
|
| 929 |
+
0,
|
| 930 |
+
1,
|
| 931 |
+
5
|
| 932 |
+
],
|
| 933 |
+
[
|
| 934 |
+
15,
|
| 935 |
+
28,
|
| 936 |
+
17,
|
| 937 |
+
0
|
| 938 |
+
],
|
| 939 |
+
[
|
| 940 |
+
0,
|
| 941 |
+
2,
|
| 942 |
+
74,
|
| 943 |
+
0
|
| 944 |
+
],
|
| 945 |
+
[
|
| 946 |
+
1,
|
| 947 |
+
0,
|
| 948 |
+
0,
|
| 949 |
+
64
|
| 950 |
+
]
|
| 951 |
+
],
|
| 952 |
+
"eval_f1": 0.8354928653436116,
|
| 953 |
+
"eval_loss": 0.7849779725074768,
|
| 954 |
+
"eval_precision": 0.8621667009790022,
|
| 955 |
+
"eval_recall": 0.8498168498168498,
|
| 956 |
+
"eval_runtime": 2.8819,
|
| 957 |
+
"eval_samples_per_second": 94.73,
|
| 958 |
+
"eval_steps_per_second": 1.041,
|
| 959 |
+
"step": 112
|
| 960 |
+
}
|
| 961 |
+
],
|
| 962 |
+
"logging_steps": 500,
|
| 963 |
+
"max_steps": 180,
|
| 964 |
+
"num_input_tokens_seen": 0,
|
| 965 |
+
"num_train_epochs": 45,
|
| 966 |
+
"save_steps": 500,
|
| 967 |
+
"stateful_callbacks": {
|
| 968 |
+
"TrainerControl": {
|
| 969 |
+
"args": {
|
| 970 |
+
"should_epoch_stop": false,
|
| 971 |
+
"should_evaluate": false,
|
| 972 |
+
"should_log": false,
|
| 973 |
+
"should_save": true,
|
| 974 |
+
"should_training_stop": false
|
| 975 |
+
},
|
| 976 |
+
"attributes": {}
|
| 977 |
+
}
|
| 978 |
+
},
|
| 979 |
+
"total_flos": 6.197267376e+16,
|
| 980 |
+
"train_batch_size": 128,
|
| 981 |
+
"trial_name": null,
|
| 982 |
+
"trial_params": null
|
| 983 |
+
}
|
checkpoint-112/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:09e217cad0464b642c54f8b7691100b62a93dceb32a1869f55e3f0eb3a54a79e
|
| 3 |
+
size 5240
|
checkpoint-117/config.json
ADDED
|
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ntu-spml/distilhubert",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"HubertForSequenceClassification"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 1,
|
| 10 |
+
"classifier_proj_size": 256,
|
| 11 |
+
"conv_bias": false,
|
| 12 |
+
"conv_dim": [
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512
|
| 20 |
+
],
|
| 21 |
+
"conv_kernel": [
|
| 22 |
+
10,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
3,
|
| 27 |
+
2,
|
| 28 |
+
2
|
| 29 |
+
],
|
| 30 |
+
"conv_stride": [
|
| 31 |
+
5,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2,
|
| 37 |
+
2
|
| 38 |
+
],
|
| 39 |
+
"ctc_loss_reduction": "sum",
|
| 40 |
+
"ctc_zero_infinity": false,
|
| 41 |
+
"do_stable_layer_norm": false,
|
| 42 |
+
"eos_token_id": 2,
|
| 43 |
+
"feat_extract_activation": "gelu",
|
| 44 |
+
"feat_extract_norm": "group",
|
| 45 |
+
"feat_proj_dropout": 0.0,
|
| 46 |
+
"feat_proj_layer_norm": false,
|
| 47 |
+
"final_dropout": 0.0,
|
| 48 |
+
"finetuning_task": "audio-classification",
|
| 49 |
+
"hidden_act": "gelu",
|
| 50 |
+
"hidden_dropout": 0.1,
|
| 51 |
+
"hidden_size": 768,
|
| 52 |
+
"id2label": {
|
| 53 |
+
"0": "1s_normal",
|
| 54 |
+
"1": "1s_pain",
|
| 55 |
+
"2": "1s_hunger",
|
| 56 |
+
"3": "1s_asphyxia"
|
| 57 |
+
},
|
| 58 |
+
"initializer_range": 0.02,
|
| 59 |
+
"intermediate_size": 3072,
|
| 60 |
+
"label2id": {
|
| 61 |
+
"LABEL_0": 0,
|
| 62 |
+
"LABEL_1": 1,
|
| 63 |
+
"LABEL_2": 2,
|
| 64 |
+
"LABEL_3": 3
|
| 65 |
+
},
|
| 66 |
+
"layer_norm_eps": 1e-05,
|
| 67 |
+
"layerdrop": 0.0,
|
| 68 |
+
"mask_feature_length": 10,
|
| 69 |
+
"mask_feature_min_masks": 0,
|
| 70 |
+
"mask_feature_prob": 0.0,
|
| 71 |
+
"mask_time_length": 10,
|
| 72 |
+
"mask_time_min_masks": 2,
|
| 73 |
+
"mask_time_prob": 0.05,
|
| 74 |
+
"model_type": "hubert",
|
| 75 |
+
"num_attention_heads": 12,
|
| 76 |
+
"num_conv_pos_embedding_groups": 16,
|
| 77 |
+
"num_conv_pos_embeddings": 128,
|
| 78 |
+
"num_feat_extract_layers": 7,
|
| 79 |
+
"num_hidden_layers": 2,
|
| 80 |
+
"pad_token_id": 0,
|
| 81 |
+
"torch_dtype": "float32",
|
| 82 |
+
"transformers_version": "4.44.2",
|
| 83 |
+
"use_weighted_layer_sum": false,
|
| 84 |
+
"vocab_size": 32
|
| 85 |
+
}
|
checkpoint-117/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0050367da6fb440bd4b3a6a79626d4b674131390bf26c99429c28c09bcbd87aa
|
| 3 |
+
size 94765560
|
checkpoint-117/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1028db5ae3b0c3f80cdc19fbb2aacb98fac1fb05527149144068fa098cb632f2
|
| 3 |
+
size 189556666
|
checkpoint-117/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d855f52eaa4bfddfdd3009c65f5ca86f8146a821538e891832fe491d205d67d9
|
| 3 |
+
size 14308
|
checkpoint-117/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:82bcc86f54463eaec1c2a094b8d233f145e4d8b50ce541ffdcb92e8149a94060
|
| 3 |
+
size 1064
|
checkpoint-117/trainer_state.json
ADDED
|
@@ -0,0 +1,1021 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8627769756077204,
|
| 3 |
+
"best_model_checkpoint": "distilhubert-finetuned-mixed-data/checkpoint-81",
|
| 4 |
+
"epoch": 26.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 117,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.8888888888888888,
|
| 13 |
+
"eval_accuracy": 0.63003663003663,
|
| 14 |
+
"eval_confusion_matrix": [
|
| 15 |
+
[
|
| 16 |
+
53,
|
| 17 |
+
0,
|
| 18 |
+
16,
|
| 19 |
+
3
|
| 20 |
+
],
|
| 21 |
+
[
|
| 22 |
+
28,
|
| 23 |
+
1,
|
| 24 |
+
31,
|
| 25 |
+
0
|
| 26 |
+
],
|
| 27 |
+
[
|
| 28 |
+
13,
|
| 29 |
+
0,
|
| 30 |
+
62,
|
| 31 |
+
1
|
| 32 |
+
],
|
| 33 |
+
[
|
| 34 |
+
3,
|
| 35 |
+
0,
|
| 36 |
+
6,
|
| 37 |
+
56
|
| 38 |
+
]
|
| 39 |
+
],
|
| 40 |
+
"eval_f1": 0.566693372031096,
|
| 41 |
+
"eval_loss": 1.31327486038208,
|
| 42 |
+
"eval_precision": 0.7361933549293478,
|
| 43 |
+
"eval_recall": 0.63003663003663,
|
| 44 |
+
"eval_runtime": 2.8436,
|
| 45 |
+
"eval_samples_per_second": 96.005,
|
| 46 |
+
"eval_steps_per_second": 1.055,
|
| 47 |
+
"step": 4
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 2.0,
|
| 51 |
+
"eval_accuracy": 0.6593406593406593,
|
| 52 |
+
"eval_confusion_matrix": [
|
| 53 |
+
[
|
| 54 |
+
41,
|
| 55 |
+
10,
|
| 56 |
+
16,
|
| 57 |
+
5
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
21,
|
| 61 |
+
1,
|
| 62 |
+
38,
|
| 63 |
+
0
|
| 64 |
+
],
|
| 65 |
+
[
|
| 66 |
+
1,
|
| 67 |
+
2,
|
| 68 |
+
73,
|
| 69 |
+
0
|
| 70 |
+
],
|
| 71 |
+
[
|
| 72 |
+
0,
|
| 73 |
+
0,
|
| 74 |
+
0,
|
| 75 |
+
65
|
| 76 |
+
]
|
| 77 |
+
],
|
| 78 |
+
"eval_f1": 0.595713773130153,
|
| 79 |
+
"eval_loss": 0.9242589473724365,
|
| 80 |
+
"eval_precision": 0.5696509512811321,
|
| 81 |
+
"eval_recall": 0.6593406593406593,
|
| 82 |
+
"eval_runtime": 2.8668,
|
| 83 |
+
"eval_samples_per_second": 95.227,
|
| 84 |
+
"eval_steps_per_second": 1.046,
|
| 85 |
+
"step": 9
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 2.888888888888889,
|
| 89 |
+
"eval_accuracy": 0.717948717948718,
|
| 90 |
+
"eval_confusion_matrix": [
|
| 91 |
+
[
|
| 92 |
+
52,
|
| 93 |
+
8,
|
| 94 |
+
7,
|
| 95 |
+
5
|
| 96 |
+
],
|
| 97 |
+
[
|
| 98 |
+
18,
|
| 99 |
+
5,
|
| 100 |
+
37,
|
| 101 |
+
0
|
| 102 |
+
],
|
| 103 |
+
[
|
| 104 |
+
1,
|
| 105 |
+
0,
|
| 106 |
+
75,
|
| 107 |
+
0
|
| 108 |
+
],
|
| 109 |
+
[
|
| 110 |
+
1,
|
| 111 |
+
0,
|
| 112 |
+
0,
|
| 113 |
+
64
|
| 114 |
+
]
|
| 115 |
+
],
|
| 116 |
+
"eval_f1": 0.6621621567125869,
|
| 117 |
+
"eval_loss": 0.8232662081718445,
|
| 118 |
+
"eval_precision": 0.671303801513745,
|
| 119 |
+
"eval_recall": 0.717948717948718,
|
| 120 |
+
"eval_runtime": 2.8711,
|
| 121 |
+
"eval_samples_per_second": 95.084,
|
| 122 |
+
"eval_steps_per_second": 1.045,
|
| 123 |
+
"step": 13
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 4.0,
|
| 127 |
+
"eval_accuracy": 0.6959706959706959,
|
| 128 |
+
"eval_confusion_matrix": [
|
| 129 |
+
[
|
| 130 |
+
45,
|
| 131 |
+
9,
|
| 132 |
+
13,
|
| 133 |
+
5
|
| 134 |
+
],
|
| 135 |
+
[
|
| 136 |
+
13,
|
| 137 |
+
5,
|
| 138 |
+
42,
|
| 139 |
+
0
|
| 140 |
+
],
|
| 141 |
+
[
|
| 142 |
+
1,
|
| 143 |
+
0,
|
| 144 |
+
75,
|
| 145 |
+
0
|
| 146 |
+
],
|
| 147 |
+
[
|
| 148 |
+
0,
|
| 149 |
+
0,
|
| 150 |
+
0,
|
| 151 |
+
65
|
| 152 |
+
]
|
| 153 |
+
],
|
| 154 |
+
"eval_f1": 0.6428796900399459,
|
| 155 |
+
"eval_loss": 0.951453447341919,
|
| 156 |
+
"eval_precision": 0.6613447705829899,
|
| 157 |
+
"eval_recall": 0.6959706959706959,
|
| 158 |
+
"eval_runtime": 2.8813,
|
| 159 |
+
"eval_samples_per_second": 94.75,
|
| 160 |
+
"eval_steps_per_second": 1.041,
|
| 161 |
+
"step": 18
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 4.888888888888889,
|
| 165 |
+
"eval_accuracy": 0.6666666666666666,
|
| 166 |
+
"eval_confusion_matrix": [
|
| 167 |
+
[
|
| 168 |
+
16,
|
| 169 |
+
49,
|
| 170 |
+
3,
|
| 171 |
+
4
|
| 172 |
+
],
|
| 173 |
+
[
|
| 174 |
+
1,
|
| 175 |
+
26,
|
| 176 |
+
33,
|
| 177 |
+
0
|
| 178 |
+
],
|
| 179 |
+
[
|
| 180 |
+
0,
|
| 181 |
+
1,
|
| 182 |
+
75,
|
| 183 |
+
0
|
| 184 |
+
],
|
| 185 |
+
[
|
| 186 |
+
0,
|
| 187 |
+
0,
|
| 188 |
+
0,
|
| 189 |
+
65
|
| 190 |
+
]
|
| 191 |
+
],
|
| 192 |
+
"eval_f1": 0.6331541584523206,
|
| 193 |
+
"eval_loss": 1.0080645084381104,
|
| 194 |
+
"eval_precision": 0.7358031394913882,
|
| 195 |
+
"eval_recall": 0.6666666666666666,
|
| 196 |
+
"eval_runtime": 2.9108,
|
| 197 |
+
"eval_samples_per_second": 93.789,
|
| 198 |
+
"eval_steps_per_second": 1.031,
|
| 199 |
+
"step": 22
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 6.0,
|
| 203 |
+
"eval_accuracy": 0.652014652014652,
|
| 204 |
+
"eval_confusion_matrix": [
|
| 205 |
+
[
|
| 206 |
+
66,
|
| 207 |
+
0,
|
| 208 |
+
0,
|
| 209 |
+
6
|
| 210 |
+
],
|
| 211 |
+
[
|
| 212 |
+
36,
|
| 213 |
+
16,
|
| 214 |
+
8,
|
| 215 |
+
0
|
| 216 |
+
],
|
| 217 |
+
[
|
| 218 |
+
10,
|
| 219 |
+
35,
|
| 220 |
+
31,
|
| 221 |
+
0
|
| 222 |
+
],
|
| 223 |
+
[
|
| 224 |
+
0,
|
| 225 |
+
0,
|
| 226 |
+
0,
|
| 227 |
+
65
|
| 228 |
+
]
|
| 229 |
+
],
|
| 230 |
+
"eval_f1": 0.6302407955860642,
|
| 231 |
+
"eval_loss": 0.9598046541213989,
|
| 232 |
+
"eval_precision": 0.6636241740077424,
|
| 233 |
+
"eval_recall": 0.652014652014652,
|
| 234 |
+
"eval_runtime": 2.9773,
|
| 235 |
+
"eval_samples_per_second": 91.693,
|
| 236 |
+
"eval_steps_per_second": 1.008,
|
| 237 |
+
"step": 27
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 6.888888888888889,
|
| 241 |
+
"eval_accuracy": 0.7692307692307693,
|
| 242 |
+
"eval_confusion_matrix": [
|
| 243 |
+
[
|
| 244 |
+
66,
|
| 245 |
+
0,
|
| 246 |
+
0,
|
| 247 |
+
6
|
| 248 |
+
],
|
| 249 |
+
[
|
| 250 |
+
29,
|
| 251 |
+
21,
|
| 252 |
+
10,
|
| 253 |
+
0
|
| 254 |
+
],
|
| 255 |
+
[
|
| 256 |
+
3,
|
| 257 |
+
15,
|
| 258 |
+
58,
|
| 259 |
+
0
|
| 260 |
+
],
|
| 261 |
+
[
|
| 262 |
+
0,
|
| 263 |
+
0,
|
| 264 |
+
0,
|
| 265 |
+
65
|
| 266 |
+
]
|
| 267 |
+
],
|
| 268 |
+
"eval_f1": 0.7527855586679116,
|
| 269 |
+
"eval_loss": 0.7898163795471191,
|
| 270 |
+
"eval_precision": 0.7612467677056892,
|
| 271 |
+
"eval_recall": 0.7692307692307693,
|
| 272 |
+
"eval_runtime": 2.8918,
|
| 273 |
+
"eval_samples_per_second": 94.404,
|
| 274 |
+
"eval_steps_per_second": 1.037,
|
| 275 |
+
"step": 31
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 8.0,
|
| 279 |
+
"eval_accuracy": 0.7802197802197802,
|
| 280 |
+
"eval_confusion_matrix": [
|
| 281 |
+
[
|
| 282 |
+
66,
|
| 283 |
+
0,
|
| 284 |
+
0,
|
| 285 |
+
6
|
| 286 |
+
],
|
| 287 |
+
[
|
| 288 |
+
25,
|
| 289 |
+
15,
|
| 290 |
+
19,
|
| 291 |
+
1
|
| 292 |
+
],
|
| 293 |
+
[
|
| 294 |
+
1,
|
| 295 |
+
6,
|
| 296 |
+
68,
|
| 297 |
+
1
|
| 298 |
+
],
|
| 299 |
+
[
|
| 300 |
+
1,
|
| 301 |
+
0,
|
| 302 |
+
0,
|
| 303 |
+
64
|
| 304 |
+
]
|
| 305 |
+
],
|
| 306 |
+
"eval_f1": 0.7471179200524057,
|
| 307 |
+
"eval_loss": 0.7337484359741211,
|
| 308 |
+
"eval_precision": 0.7733845922309294,
|
| 309 |
+
"eval_recall": 0.7802197802197802,
|
| 310 |
+
"eval_runtime": 2.9006,
|
| 311 |
+
"eval_samples_per_second": 94.119,
|
| 312 |
+
"eval_steps_per_second": 1.034,
|
| 313 |
+
"step": 36
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 8.88888888888889,
|
| 317 |
+
"eval_accuracy": 0.7912087912087912,
|
| 318 |
+
"eval_confusion_matrix": [
|
| 319 |
+
[
|
| 320 |
+
57,
|
| 321 |
+
8,
|
| 322 |
+
3,
|
| 323 |
+
4
|
| 324 |
+
],
|
| 325 |
+
[
|
| 326 |
+
13,
|
| 327 |
+
24,
|
| 328 |
+
23,
|
| 329 |
+
0
|
| 330 |
+
],
|
| 331 |
+
[
|
| 332 |
+
0,
|
| 333 |
+
5,
|
| 334 |
+
71,
|
| 335 |
+
0
|
| 336 |
+
],
|
| 337 |
+
[
|
| 338 |
+
1,
|
| 339 |
+
0,
|
| 340 |
+
0,
|
| 341 |
+
64
|
| 342 |
+
]
|
| 343 |
+
],
|
| 344 |
+
"eval_f1": 0.7766564722186922,
|
| 345 |
+
"eval_loss": 0.7148727178573608,
|
| 346 |
+
"eval_precision": 0.7821504483074875,
|
| 347 |
+
"eval_recall": 0.7912087912087912,
|
| 348 |
+
"eval_runtime": 2.8866,
|
| 349 |
+
"eval_samples_per_second": 94.576,
|
| 350 |
+
"eval_steps_per_second": 1.039,
|
| 351 |
+
"step": 40
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 10.0,
|
| 355 |
+
"eval_accuracy": 0.8168498168498168,
|
| 356 |
+
"eval_confusion_matrix": [
|
| 357 |
+
[
|
| 358 |
+
51,
|
| 359 |
+
8,
|
| 360 |
+
1,
|
| 361 |
+
12
|
| 362 |
+
],
|
| 363 |
+
[
|
| 364 |
+
10,
|
| 365 |
+
36,
|
| 366 |
+
14,
|
| 367 |
+
0
|
| 368 |
+
],
|
| 369 |
+
[
|
| 370 |
+
0,
|
| 371 |
+
5,
|
| 372 |
+
71,
|
| 373 |
+
0
|
| 374 |
+
],
|
| 375 |
+
[
|
| 376 |
+
0,
|
| 377 |
+
0,
|
| 378 |
+
0,
|
| 379 |
+
65
|
| 380 |
+
]
|
| 381 |
+
],
|
| 382 |
+
"eval_f1": 0.8094336953840884,
|
| 383 |
+
"eval_loss": 0.7574812173843384,
|
| 384 |
+
"eval_precision": 0.8127936625684181,
|
| 385 |
+
"eval_recall": 0.8168498168498168,
|
| 386 |
+
"eval_runtime": 2.8967,
|
| 387 |
+
"eval_samples_per_second": 94.245,
|
| 388 |
+
"eval_steps_per_second": 1.036,
|
| 389 |
+
"step": 45
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 10.88888888888889,
|
| 393 |
+
"eval_accuracy": 0.8021978021978022,
|
| 394 |
+
"eval_confusion_matrix": [
|
| 395 |
+
[
|
| 396 |
+
48,
|
| 397 |
+
18,
|
| 398 |
+
2,
|
| 399 |
+
4
|
| 400 |
+
],
|
| 401 |
+
[
|
| 402 |
+
4,
|
| 403 |
+
41,
|
| 404 |
+
15,
|
| 405 |
+
0
|
| 406 |
+
],
|
| 407 |
+
[
|
| 408 |
+
0,
|
| 409 |
+
7,
|
| 410 |
+
69,
|
| 411 |
+
0
|
| 412 |
+
],
|
| 413 |
+
[
|
| 414 |
+
3,
|
| 415 |
+
0,
|
| 416 |
+
1,
|
| 417 |
+
61
|
| 418 |
+
]
|
| 419 |
+
],
|
| 420 |
+
"eval_f1": 0.801525180147331,
|
| 421 |
+
"eval_loss": 0.7140281200408936,
|
| 422 |
+
"eval_precision": 0.8109338936925145,
|
| 423 |
+
"eval_recall": 0.8021978021978022,
|
| 424 |
+
"eval_runtime": 2.917,
|
| 425 |
+
"eval_samples_per_second": 93.589,
|
| 426 |
+
"eval_steps_per_second": 1.028,
|
| 427 |
+
"step": 49
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 12.0,
|
| 431 |
+
"eval_accuracy": 0.8534798534798534,
|
| 432 |
+
"eval_confusion_matrix": [
|
| 433 |
+
[
|
| 434 |
+
56,
|
| 435 |
+
12,
|
| 436 |
+
0,
|
| 437 |
+
4
|
| 438 |
+
],
|
| 439 |
+
[
|
| 440 |
+
6,
|
| 441 |
+
43,
|
| 442 |
+
11,
|
| 443 |
+
0
|
| 444 |
+
],
|
| 445 |
+
[
|
| 446 |
+
0,
|
| 447 |
+
6,
|
| 448 |
+
70,
|
| 449 |
+
0
|
| 450 |
+
],
|
| 451 |
+
[
|
| 452 |
+
1,
|
| 453 |
+
0,
|
| 454 |
+
0,
|
| 455 |
+
64
|
| 456 |
+
]
|
| 457 |
+
],
|
| 458 |
+
"eval_f1": 0.8523994617102314,
|
| 459 |
+
"eval_loss": 0.6672152280807495,
|
| 460 |
+
"eval_precision": 0.8540313732642031,
|
| 461 |
+
"eval_recall": 0.8534798534798534,
|
| 462 |
+
"eval_runtime": 2.8969,
|
| 463 |
+
"eval_samples_per_second": 94.238,
|
| 464 |
+
"eval_steps_per_second": 1.036,
|
| 465 |
+
"step": 54
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 12.88888888888889,
|
| 469 |
+
"eval_accuracy": 0.8498168498168498,
|
| 470 |
+
"eval_confusion_matrix": [
|
| 471 |
+
[
|
| 472 |
+
60,
|
| 473 |
+
8,
|
| 474 |
+
0,
|
| 475 |
+
4
|
| 476 |
+
],
|
| 477 |
+
[
|
| 478 |
+
8,
|
| 479 |
+
38,
|
| 480 |
+
14,
|
| 481 |
+
0
|
| 482 |
+
],
|
| 483 |
+
[
|
| 484 |
+
0,
|
| 485 |
+
6,
|
| 486 |
+
70,
|
| 487 |
+
0
|
| 488 |
+
],
|
| 489 |
+
[
|
| 490 |
+
1,
|
| 491 |
+
0,
|
| 492 |
+
0,
|
| 493 |
+
64
|
| 494 |
+
]
|
| 495 |
+
],
|
| 496 |
+
"eval_f1": 0.8463270052615757,
|
| 497 |
+
"eval_loss": 0.6432910561561584,
|
| 498 |
+
"eval_precision": 0.8460243715014519,
|
| 499 |
+
"eval_recall": 0.8498168498168498,
|
| 500 |
+
"eval_runtime": 3.0281,
|
| 501 |
+
"eval_samples_per_second": 90.156,
|
| 502 |
+
"eval_steps_per_second": 0.991,
|
| 503 |
+
"step": 58
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 14.0,
|
| 507 |
+
"eval_accuracy": 0.8278388278388278,
|
| 508 |
+
"eval_confusion_matrix": [
|
| 509 |
+
[
|
| 510 |
+
54,
|
| 511 |
+
13,
|
| 512 |
+
1,
|
| 513 |
+
4
|
| 514 |
+
],
|
| 515 |
+
[
|
| 516 |
+
7,
|
| 517 |
+
44,
|
| 518 |
+
9,
|
| 519 |
+
0
|
| 520 |
+
],
|
| 521 |
+
[
|
| 522 |
+
0,
|
| 523 |
+
11,
|
| 524 |
+
65,
|
| 525 |
+
0
|
| 526 |
+
],
|
| 527 |
+
[
|
| 528 |
+
1,
|
| 529 |
+
1,
|
| 530 |
+
0,
|
| 531 |
+
63
|
| 532 |
+
]
|
| 533 |
+
],
|
| 534 |
+
"eval_f1": 0.82943590265942,
|
| 535 |
+
"eval_loss": 0.7395206093788147,
|
| 536 |
+
"eval_precision": 0.8350059217447382,
|
| 537 |
+
"eval_recall": 0.8278388278388278,
|
| 538 |
+
"eval_runtime": 2.9869,
|
| 539 |
+
"eval_samples_per_second": 91.399,
|
| 540 |
+
"eval_steps_per_second": 1.004,
|
| 541 |
+
"step": 63
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 14.88888888888889,
|
| 545 |
+
"eval_accuracy": 0.8315018315018315,
|
| 546 |
+
"eval_confusion_matrix": [
|
| 547 |
+
[
|
| 548 |
+
54,
|
| 549 |
+
13,
|
| 550 |
+
1,
|
| 551 |
+
4
|
| 552 |
+
],
|
| 553 |
+
[
|
| 554 |
+
10,
|
| 555 |
+
39,
|
| 556 |
+
11,
|
| 557 |
+
0
|
| 558 |
+
],
|
| 559 |
+
[
|
| 560 |
+
0,
|
| 561 |
+
7,
|
| 562 |
+
69,
|
| 563 |
+
0
|
| 564 |
+
],
|
| 565 |
+
[
|
| 566 |
+
0,
|
| 567 |
+
0,
|
| 568 |
+
0,
|
| 569 |
+
65
|
| 570 |
+
]
|
| 571 |
+
],
|
| 572 |
+
"eval_f1": 0.8291811389886823,
|
| 573 |
+
"eval_loss": 0.7115849852561951,
|
| 574 |
+
"eval_precision": 0.829244108966536,
|
| 575 |
+
"eval_recall": 0.8315018315018315,
|
| 576 |
+
"eval_runtime": 2.9616,
|
| 577 |
+
"eval_samples_per_second": 92.181,
|
| 578 |
+
"eval_steps_per_second": 1.013,
|
| 579 |
+
"step": 67
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 16.0,
|
| 583 |
+
"eval_accuracy": 0.8315018315018315,
|
| 584 |
+
"eval_confusion_matrix": [
|
| 585 |
+
[
|
| 586 |
+
60,
|
| 587 |
+
7,
|
| 588 |
+
1,
|
| 589 |
+
4
|
| 590 |
+
],
|
| 591 |
+
[
|
| 592 |
+
10,
|
| 593 |
+
39,
|
| 594 |
+
11,
|
| 595 |
+
0
|
| 596 |
+
],
|
| 597 |
+
[
|
| 598 |
+
0,
|
| 599 |
+
11,
|
| 600 |
+
65,
|
| 601 |
+
0
|
| 602 |
+
],
|
| 603 |
+
[
|
| 604 |
+
1,
|
| 605 |
+
1,
|
| 606 |
+
0,
|
| 607 |
+
63
|
| 608 |
+
]
|
| 609 |
+
],
|
| 610 |
+
"eval_f1": 0.8304073820984628,
|
| 611 |
+
"eval_loss": 0.7295921444892883,
|
| 612 |
+
"eval_precision": 0.8295426562258641,
|
| 613 |
+
"eval_recall": 0.8315018315018315,
|
| 614 |
+
"eval_runtime": 2.8758,
|
| 615 |
+
"eval_samples_per_second": 94.932,
|
| 616 |
+
"eval_steps_per_second": 1.043,
|
| 617 |
+
"step": 72
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 16.88888888888889,
|
| 621 |
+
"eval_accuracy": 0.8644688644688645,
|
| 622 |
+
"eval_confusion_matrix": [
|
| 623 |
+
[
|
| 624 |
+
62,
|
| 625 |
+
5,
|
| 626 |
+
1,
|
| 627 |
+
4
|
| 628 |
+
],
|
| 629 |
+
[
|
| 630 |
+
9,
|
| 631 |
+
37,
|
| 632 |
+
14,
|
| 633 |
+
0
|
| 634 |
+
],
|
| 635 |
+
[
|
| 636 |
+
0,
|
| 637 |
+
3,
|
| 638 |
+
73,
|
| 639 |
+
0
|
| 640 |
+
],
|
| 641 |
+
[
|
| 642 |
+
1,
|
| 643 |
+
0,
|
| 644 |
+
0,
|
| 645 |
+
64
|
| 646 |
+
]
|
| 647 |
+
],
|
| 648 |
+
"eval_f1": 0.8589767100678526,
|
| 649 |
+
"eval_loss": 0.7055637240409851,
|
| 650 |
+
"eval_precision": 0.8628397746044805,
|
| 651 |
+
"eval_recall": 0.8644688644688645,
|
| 652 |
+
"eval_runtime": 3.074,
|
| 653 |
+
"eval_samples_per_second": 88.809,
|
| 654 |
+
"eval_steps_per_second": 0.976,
|
| 655 |
+
"step": 76
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 18.0,
|
| 659 |
+
"eval_accuracy": 0.8644688644688645,
|
| 660 |
+
"eval_confusion_matrix": [
|
| 661 |
+
[
|
| 662 |
+
65,
|
| 663 |
+
2,
|
| 664 |
+
1,
|
| 665 |
+
4
|
| 666 |
+
],
|
| 667 |
+
[
|
| 668 |
+
13,
|
| 669 |
+
42,
|
| 670 |
+
5,
|
| 671 |
+
0
|
| 672 |
+
],
|
| 673 |
+
[
|
| 674 |
+
0,
|
| 675 |
+
12,
|
| 676 |
+
64,
|
| 677 |
+
0
|
| 678 |
+
],
|
| 679 |
+
[
|
| 680 |
+
0,
|
| 681 |
+
0,
|
| 682 |
+
0,
|
| 683 |
+
65
|
| 684 |
+
]
|
| 685 |
+
],
|
| 686 |
+
"eval_f1": 0.8627769756077204,
|
| 687 |
+
"eval_loss": 0.7563945651054382,
|
| 688 |
+
"eval_precision": 0.8634344261673453,
|
| 689 |
+
"eval_recall": 0.8644688644688645,
|
| 690 |
+
"eval_runtime": 3.0072,
|
| 691 |
+
"eval_samples_per_second": 90.783,
|
| 692 |
+
"eval_steps_per_second": 0.998,
|
| 693 |
+
"step": 81
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 18.88888888888889,
|
| 697 |
+
"eval_accuracy": 0.8424908424908425,
|
| 698 |
+
"eval_confusion_matrix": [
|
| 699 |
+
[
|
| 700 |
+
64,
|
| 701 |
+
4,
|
| 702 |
+
0,
|
| 703 |
+
4
|
| 704 |
+
],
|
| 705 |
+
[
|
| 706 |
+
11,
|
| 707 |
+
41,
|
| 708 |
+
8,
|
| 709 |
+
0
|
| 710 |
+
],
|
| 711 |
+
[
|
| 712 |
+
0,
|
| 713 |
+
10,
|
| 714 |
+
66,
|
| 715 |
+
0
|
| 716 |
+
],
|
| 717 |
+
[
|
| 718 |
+
6,
|
| 719 |
+
0,
|
| 720 |
+
0,
|
| 721 |
+
59
|
| 722 |
+
]
|
| 723 |
+
],
|
| 724 |
+
"eval_f1": 0.8418306879608031,
|
| 725 |
+
"eval_loss": 0.7825365662574768,
|
| 726 |
+
"eval_precision": 0.8434907006335578,
|
| 727 |
+
"eval_recall": 0.8424908424908425,
|
| 728 |
+
"eval_runtime": 3.0343,
|
| 729 |
+
"eval_samples_per_second": 89.972,
|
| 730 |
+
"eval_steps_per_second": 0.989,
|
| 731 |
+
"step": 85
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 20.0,
|
| 735 |
+
"eval_accuracy": 0.8058608058608059,
|
| 736 |
+
"eval_confusion_matrix": [
|
| 737 |
+
[
|
| 738 |
+
40,
|
| 739 |
+
28,
|
| 740 |
+
1,
|
| 741 |
+
3
|
| 742 |
+
],
|
| 743 |
+
[
|
| 744 |
+
2,
|
| 745 |
+
50,
|
| 746 |
+
8,
|
| 747 |
+
0
|
| 748 |
+
],
|
| 749 |
+
[
|
| 750 |
+
0,
|
| 751 |
+
11,
|
| 752 |
+
65,
|
| 753 |
+
0
|
| 754 |
+
],
|
| 755 |
+
[
|
| 756 |
+
0,
|
| 757 |
+
0,
|
| 758 |
+
0,
|
| 759 |
+
65
|
| 760 |
+
]
|
| 761 |
+
],
|
| 762 |
+
"eval_f1": 0.8065760931078588,
|
| 763 |
+
"eval_loss": 0.8426868915557861,
|
| 764 |
+
"eval_precision": 0.8467707085637404,
|
| 765 |
+
"eval_recall": 0.8058608058608059,
|
| 766 |
+
"eval_runtime": 3.0783,
|
| 767 |
+
"eval_samples_per_second": 88.686,
|
| 768 |
+
"eval_steps_per_second": 0.975,
|
| 769 |
+
"step": 90
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 20.88888888888889,
|
| 773 |
+
"eval_accuracy": 0.8498168498168498,
|
| 774 |
+
"eval_confusion_matrix": [
|
| 775 |
+
[
|
| 776 |
+
63,
|
| 777 |
+
4,
|
| 778 |
+
0,
|
| 779 |
+
5
|
| 780 |
+
],
|
| 781 |
+
[
|
| 782 |
+
13,
|
| 783 |
+
40,
|
| 784 |
+
7,
|
| 785 |
+
0
|
| 786 |
+
],
|
| 787 |
+
[
|
| 788 |
+
0,
|
| 789 |
+
12,
|
| 790 |
+
64,
|
| 791 |
+
0
|
| 792 |
+
],
|
| 793 |
+
[
|
| 794 |
+
0,
|
| 795 |
+
0,
|
| 796 |
+
0,
|
| 797 |
+
65
|
| 798 |
+
]
|
| 799 |
+
],
|
| 800 |
+
"eval_f1": 0.8477878057985963,
|
| 801 |
+
"eval_loss": 0.7440442442893982,
|
| 802 |
+
"eval_precision": 0.8476393351433065,
|
| 803 |
+
"eval_recall": 0.8498168498168498,
|
| 804 |
+
"eval_runtime": 2.9875,
|
| 805 |
+
"eval_samples_per_second": 91.38,
|
| 806 |
+
"eval_steps_per_second": 1.004,
|
| 807 |
+
"step": 94
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 22.0,
|
| 811 |
+
"eval_accuracy": 0.8608058608058609,
|
| 812 |
+
"eval_confusion_matrix": [
|
| 813 |
+
[
|
| 814 |
+
66,
|
| 815 |
+
1,
|
| 816 |
+
1,
|
| 817 |
+
4
|
| 818 |
+
],
|
| 819 |
+
[
|
| 820 |
+
13,
|
| 821 |
+
36,
|
| 822 |
+
11,
|
| 823 |
+
0
|
| 824 |
+
],
|
| 825 |
+
[
|
| 826 |
+
0,
|
| 827 |
+
5,
|
| 828 |
+
71,
|
| 829 |
+
0
|
| 830 |
+
],
|
| 831 |
+
[
|
| 832 |
+
3,
|
| 833 |
+
0,
|
| 834 |
+
0,
|
| 835 |
+
62
|
| 836 |
+
]
|
| 837 |
+
],
|
| 838 |
+
"eval_f1": 0.855194718990792,
|
| 839 |
+
"eval_loss": 0.7338178753852844,
|
| 840 |
+
"eval_precision": 0.8624631692093176,
|
| 841 |
+
"eval_recall": 0.8608058608058609,
|
| 842 |
+
"eval_runtime": 2.98,
|
| 843 |
+
"eval_samples_per_second": 91.611,
|
| 844 |
+
"eval_steps_per_second": 1.007,
|
| 845 |
+
"step": 99
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 22.88888888888889,
|
| 849 |
+
"eval_accuracy": 0.8498168498168498,
|
| 850 |
+
"eval_confusion_matrix": [
|
| 851 |
+
[
|
| 852 |
+
58,
|
| 853 |
+
10,
|
| 854 |
+
0,
|
| 855 |
+
4
|
| 856 |
+
],
|
| 857 |
+
[
|
| 858 |
+
10,
|
| 859 |
+
43,
|
| 860 |
+
7,
|
| 861 |
+
0
|
| 862 |
+
],
|
| 863 |
+
[
|
| 864 |
+
0,
|
| 865 |
+
7,
|
| 866 |
+
69,
|
| 867 |
+
0
|
| 868 |
+
],
|
| 869 |
+
[
|
| 870 |
+
2,
|
| 871 |
+
1,
|
| 872 |
+
0,
|
| 873 |
+
62
|
| 874 |
+
]
|
| 875 |
+
],
|
| 876 |
+
"eval_f1": 0.8497737987724402,
|
| 877 |
+
"eval_loss": 0.7231407761573792,
|
| 878 |
+
"eval_precision": 0.8498638493954653,
|
| 879 |
+
"eval_recall": 0.8498168498168498,
|
| 880 |
+
"eval_runtime": 2.9272,
|
| 881 |
+
"eval_samples_per_second": 93.262,
|
| 882 |
+
"eval_steps_per_second": 1.025,
|
| 883 |
+
"step": 103
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 24.0,
|
| 887 |
+
"eval_accuracy": 0.8424908424908425,
|
| 888 |
+
"eval_confusion_matrix": [
|
| 889 |
+
[
|
| 890 |
+
67,
|
| 891 |
+
1,
|
| 892 |
+
0,
|
| 893 |
+
4
|
| 894 |
+
],
|
| 895 |
+
[
|
| 896 |
+
12,
|
| 897 |
+
44,
|
| 898 |
+
4,
|
| 899 |
+
0
|
| 900 |
+
],
|
| 901 |
+
[
|
| 902 |
+
0,
|
| 903 |
+
18,
|
| 904 |
+
58,
|
| 905 |
+
0
|
| 906 |
+
],
|
| 907 |
+
[
|
| 908 |
+
2,
|
| 909 |
+
2,
|
| 910 |
+
0,
|
| 911 |
+
61
|
| 912 |
+
]
|
| 913 |
+
],
|
| 914 |
+
"eval_f1": 0.8431607380967995,
|
| 915 |
+
"eval_loss": 0.752363920211792,
|
| 916 |
+
"eval_precision": 0.8507974885146101,
|
| 917 |
+
"eval_recall": 0.8424908424908425,
|
| 918 |
+
"eval_runtime": 3.0257,
|
| 919 |
+
"eval_samples_per_second": 90.228,
|
| 920 |
+
"eval_steps_per_second": 0.992,
|
| 921 |
+
"step": 108
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 24.88888888888889,
|
| 925 |
+
"eval_accuracy": 0.8498168498168498,
|
| 926 |
+
"eval_confusion_matrix": [
|
| 927 |
+
[
|
| 928 |
+
66,
|
| 929 |
+
0,
|
| 930 |
+
1,
|
| 931 |
+
5
|
| 932 |
+
],
|
| 933 |
+
[
|
| 934 |
+
15,
|
| 935 |
+
28,
|
| 936 |
+
17,
|
| 937 |
+
0
|
| 938 |
+
],
|
| 939 |
+
[
|
| 940 |
+
0,
|
| 941 |
+
2,
|
| 942 |
+
74,
|
| 943 |
+
0
|
| 944 |
+
],
|
| 945 |
+
[
|
| 946 |
+
1,
|
| 947 |
+
0,
|
| 948 |
+
0,
|
| 949 |
+
64
|
| 950 |
+
]
|
| 951 |
+
],
|
| 952 |
+
"eval_f1": 0.8354928653436116,
|
| 953 |
+
"eval_loss": 0.7849779725074768,
|
| 954 |
+
"eval_precision": 0.8621667009790022,
|
| 955 |
+
"eval_recall": 0.8498168498168498,
|
| 956 |
+
"eval_runtime": 2.8819,
|
| 957 |
+
"eval_samples_per_second": 94.73,
|
| 958 |
+
"eval_steps_per_second": 1.041,
|
| 959 |
+
"step": 112
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 26.0,
|
| 963 |
+
"eval_accuracy": 0.8131868131868132,
|
| 964 |
+
"eval_confusion_matrix": [
|
| 965 |
+
[
|
| 966 |
+
62,
|
| 967 |
+
6,
|
| 968 |
+
0,
|
| 969 |
+
4
|
| 970 |
+
],
|
| 971 |
+
[
|
| 972 |
+
11,
|
| 973 |
+
44,
|
| 974 |
+
5,
|
| 975 |
+
0
|
| 976 |
+
],
|
| 977 |
+
[
|
| 978 |
+
0,
|
| 979 |
+
24,
|
| 980 |
+
52,
|
| 981 |
+
0
|
| 982 |
+
],
|
| 983 |
+
[
|
| 984 |
+
1,
|
| 985 |
+
0,
|
| 986 |
+
0,
|
| 987 |
+
64
|
| 988 |
+
]
|
| 989 |
+
],
|
| 990 |
+
"eval_f1": 0.8151598256855266,
|
| 991 |
+
"eval_loss": 0.7896661162376404,
|
| 992 |
+
"eval_precision": 0.8297062414709473,
|
| 993 |
+
"eval_recall": 0.8131868131868132,
|
| 994 |
+
"eval_runtime": 2.8953,
|
| 995 |
+
"eval_samples_per_second": 94.291,
|
| 996 |
+
"eval_steps_per_second": 1.036,
|
| 997 |
+
"step": 117
|
| 998 |
+
}
|
| 999 |
+
],
|
| 1000 |
+
"logging_steps": 500,
|
| 1001 |
+
"max_steps": 180,
|
| 1002 |
+
"num_input_tokens_seen": 0,
|
| 1003 |
+
"num_train_epochs": 45,
|
| 1004 |
+
"save_steps": 500,
|
| 1005 |
+
"stateful_callbacks": {
|
| 1006 |
+
"TrainerControl": {
|
| 1007 |
+
"args": {
|
| 1008 |
+
"should_epoch_stop": false,
|
| 1009 |
+
"should_evaluate": false,
|
| 1010 |
+
"should_log": false,
|
| 1011 |
+
"should_save": true,
|
| 1012 |
+
"should_training_stop": false
|
| 1013 |
+
},
|
| 1014 |
+
"attributes": {}
|
| 1015 |
+
}
|
| 1016 |
+
},
|
| 1017 |
+
"total_flos": 6.44515807104e+16,
|
| 1018 |
+
"train_batch_size": 128,
|
| 1019 |
+
"trial_name": null,
|
| 1020 |
+
"trial_params": null
|
| 1021 |
+
}
|
checkpoint-117/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:09e217cad0464b642c54f8b7691100b62a93dceb32a1869f55e3f0eb3a54a79e
|
| 3 |
+
size 5240
|
checkpoint-121/config.json
ADDED
|
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ntu-spml/distilhubert",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"HubertForSequenceClassification"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 1,
|
| 10 |
+
"classifier_proj_size": 256,
|
| 11 |
+
"conv_bias": false,
|
| 12 |
+
"conv_dim": [
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512
|
| 20 |
+
],
|
| 21 |
+
"conv_kernel": [
|
| 22 |
+
10,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
3,
|
| 27 |
+
2,
|
| 28 |
+
2
|
| 29 |
+
],
|
| 30 |
+
"conv_stride": [
|
| 31 |
+
5,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2,
|
| 37 |
+
2
|
| 38 |
+
],
|
| 39 |
+
"ctc_loss_reduction": "sum",
|
| 40 |
+
"ctc_zero_infinity": false,
|
| 41 |
+
"do_stable_layer_norm": false,
|
| 42 |
+
"eos_token_id": 2,
|
| 43 |
+
"feat_extract_activation": "gelu",
|
| 44 |
+
"feat_extract_norm": "group",
|
| 45 |
+
"feat_proj_dropout": 0.0,
|
| 46 |
+
"feat_proj_layer_norm": false,
|
| 47 |
+
"final_dropout": 0.0,
|
| 48 |
+
"finetuning_task": "audio-classification",
|
| 49 |
+
"hidden_act": "gelu",
|
| 50 |
+
"hidden_dropout": 0.1,
|
| 51 |
+
"hidden_size": 768,
|
| 52 |
+
"id2label": {
|
| 53 |
+
"0": "1s_normal",
|
| 54 |
+
"1": "1s_pain",
|
| 55 |
+
"2": "1s_hunger",
|
| 56 |
+
"3": "1s_asphyxia"
|
| 57 |
+
},
|
| 58 |
+
"initializer_range": 0.02,
|
| 59 |
+
"intermediate_size": 3072,
|
| 60 |
+
"label2id": {
|
| 61 |
+
"LABEL_0": 0,
|
| 62 |
+
"LABEL_1": 1,
|
| 63 |
+
"LABEL_2": 2,
|
| 64 |
+
"LABEL_3": 3
|
| 65 |
+
},
|
| 66 |
+
"layer_norm_eps": 1e-05,
|
| 67 |
+
"layerdrop": 0.0,
|
| 68 |
+
"mask_feature_length": 10,
|
| 69 |
+
"mask_feature_min_masks": 0,
|
| 70 |
+
"mask_feature_prob": 0.0,
|
| 71 |
+
"mask_time_length": 10,
|
| 72 |
+
"mask_time_min_masks": 2,
|
| 73 |
+
"mask_time_prob": 0.05,
|
| 74 |
+
"model_type": "hubert",
|
| 75 |
+
"num_attention_heads": 12,
|
| 76 |
+
"num_conv_pos_embedding_groups": 16,
|
| 77 |
+
"num_conv_pos_embeddings": 128,
|
| 78 |
+
"num_feat_extract_layers": 7,
|
| 79 |
+
"num_hidden_layers": 2,
|
| 80 |
+
"pad_token_id": 0,
|
| 81 |
+
"torch_dtype": "float32",
|
| 82 |
+
"transformers_version": "4.44.2",
|
| 83 |
+
"use_weighted_layer_sum": false,
|
| 84 |
+
"vocab_size": 32
|
| 85 |
+
}
|
checkpoint-121/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:00956f5c7a7b4b257696dd508120e7f8c293d6eae2868350246f64214485eec0
|
| 3 |
+
size 94765560
|
checkpoint-121/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3bf7362c0ebbcddc9516c4702d04e42283fe14ca75340a0e003469d351fa1c2b
|
| 3 |
+
size 189556666
|
checkpoint-121/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3d46bac788524b95c2eb6a503c884354aedaaf5845a246dad40a99742be55b9f
|
| 3 |
+
size 14308
|
checkpoint-121/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:baac56bdbea76b1ed3b409eadae05b655886a9f39f011f0f648a4eb3fa804071
|
| 3 |
+
size 1064
|
checkpoint-121/trainer_state.json
ADDED
|
@@ -0,0 +1,1059 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8754390108936493,
|
| 3 |
+
"best_model_checkpoint": "distilhubert-finetuned-mixed-data/checkpoint-121",
|
| 4 |
+
"epoch": 26.88888888888889,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 121,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.8888888888888888,
|
| 13 |
+
"eval_accuracy": 0.63003663003663,
|
| 14 |
+
"eval_confusion_matrix": [
|
| 15 |
+
[
|
| 16 |
+
53,
|
| 17 |
+
0,
|
| 18 |
+
16,
|
| 19 |
+
3
|
| 20 |
+
],
|
| 21 |
+
[
|
| 22 |
+
28,
|
| 23 |
+
1,
|
| 24 |
+
31,
|
| 25 |
+
0
|
| 26 |
+
],
|
| 27 |
+
[
|
| 28 |
+
13,
|
| 29 |
+
0,
|
| 30 |
+
62,
|
| 31 |
+
1
|
| 32 |
+
],
|
| 33 |
+
[
|
| 34 |
+
3,
|
| 35 |
+
0,
|
| 36 |
+
6,
|
| 37 |
+
56
|
| 38 |
+
]
|
| 39 |
+
],
|
| 40 |
+
"eval_f1": 0.566693372031096,
|
| 41 |
+
"eval_loss": 1.31327486038208,
|
| 42 |
+
"eval_precision": 0.7361933549293478,
|
| 43 |
+
"eval_recall": 0.63003663003663,
|
| 44 |
+
"eval_runtime": 2.8436,
|
| 45 |
+
"eval_samples_per_second": 96.005,
|
| 46 |
+
"eval_steps_per_second": 1.055,
|
| 47 |
+
"step": 4
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 2.0,
|
| 51 |
+
"eval_accuracy": 0.6593406593406593,
|
| 52 |
+
"eval_confusion_matrix": [
|
| 53 |
+
[
|
| 54 |
+
41,
|
| 55 |
+
10,
|
| 56 |
+
16,
|
| 57 |
+
5
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
21,
|
| 61 |
+
1,
|
| 62 |
+
38,
|
| 63 |
+
0
|
| 64 |
+
],
|
| 65 |
+
[
|
| 66 |
+
1,
|
| 67 |
+
2,
|
| 68 |
+
73,
|
| 69 |
+
0
|
| 70 |
+
],
|
| 71 |
+
[
|
| 72 |
+
0,
|
| 73 |
+
0,
|
| 74 |
+
0,
|
| 75 |
+
65
|
| 76 |
+
]
|
| 77 |
+
],
|
| 78 |
+
"eval_f1": 0.595713773130153,
|
| 79 |
+
"eval_loss": 0.9242589473724365,
|
| 80 |
+
"eval_precision": 0.5696509512811321,
|
| 81 |
+
"eval_recall": 0.6593406593406593,
|
| 82 |
+
"eval_runtime": 2.8668,
|
| 83 |
+
"eval_samples_per_second": 95.227,
|
| 84 |
+
"eval_steps_per_second": 1.046,
|
| 85 |
+
"step": 9
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 2.888888888888889,
|
| 89 |
+
"eval_accuracy": 0.717948717948718,
|
| 90 |
+
"eval_confusion_matrix": [
|
| 91 |
+
[
|
| 92 |
+
52,
|
| 93 |
+
8,
|
| 94 |
+
7,
|
| 95 |
+
5
|
| 96 |
+
],
|
| 97 |
+
[
|
| 98 |
+
18,
|
| 99 |
+
5,
|
| 100 |
+
37,
|
| 101 |
+
0
|
| 102 |
+
],
|
| 103 |
+
[
|
| 104 |
+
1,
|
| 105 |
+
0,
|
| 106 |
+
75,
|
| 107 |
+
0
|
| 108 |
+
],
|
| 109 |
+
[
|
| 110 |
+
1,
|
| 111 |
+
0,
|
| 112 |
+
0,
|
| 113 |
+
64
|
| 114 |
+
]
|
| 115 |
+
],
|
| 116 |
+
"eval_f1": 0.6621621567125869,
|
| 117 |
+
"eval_loss": 0.8232662081718445,
|
| 118 |
+
"eval_precision": 0.671303801513745,
|
| 119 |
+
"eval_recall": 0.717948717948718,
|
| 120 |
+
"eval_runtime": 2.8711,
|
| 121 |
+
"eval_samples_per_second": 95.084,
|
| 122 |
+
"eval_steps_per_second": 1.045,
|
| 123 |
+
"step": 13
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 4.0,
|
| 127 |
+
"eval_accuracy": 0.6959706959706959,
|
| 128 |
+
"eval_confusion_matrix": [
|
| 129 |
+
[
|
| 130 |
+
45,
|
| 131 |
+
9,
|
| 132 |
+
13,
|
| 133 |
+
5
|
| 134 |
+
],
|
| 135 |
+
[
|
| 136 |
+
13,
|
| 137 |
+
5,
|
| 138 |
+
42,
|
| 139 |
+
0
|
| 140 |
+
],
|
| 141 |
+
[
|
| 142 |
+
1,
|
| 143 |
+
0,
|
| 144 |
+
75,
|
| 145 |
+
0
|
| 146 |
+
],
|
| 147 |
+
[
|
| 148 |
+
0,
|
| 149 |
+
0,
|
| 150 |
+
0,
|
| 151 |
+
65
|
| 152 |
+
]
|
| 153 |
+
],
|
| 154 |
+
"eval_f1": 0.6428796900399459,
|
| 155 |
+
"eval_loss": 0.951453447341919,
|
| 156 |
+
"eval_precision": 0.6613447705829899,
|
| 157 |
+
"eval_recall": 0.6959706959706959,
|
| 158 |
+
"eval_runtime": 2.8813,
|
| 159 |
+
"eval_samples_per_second": 94.75,
|
| 160 |
+
"eval_steps_per_second": 1.041,
|
| 161 |
+
"step": 18
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 4.888888888888889,
|
| 165 |
+
"eval_accuracy": 0.6666666666666666,
|
| 166 |
+
"eval_confusion_matrix": [
|
| 167 |
+
[
|
| 168 |
+
16,
|
| 169 |
+
49,
|
| 170 |
+
3,
|
| 171 |
+
4
|
| 172 |
+
],
|
| 173 |
+
[
|
| 174 |
+
1,
|
| 175 |
+
26,
|
| 176 |
+
33,
|
| 177 |
+
0
|
| 178 |
+
],
|
| 179 |
+
[
|
| 180 |
+
0,
|
| 181 |
+
1,
|
| 182 |
+
75,
|
| 183 |
+
0
|
| 184 |
+
],
|
| 185 |
+
[
|
| 186 |
+
0,
|
| 187 |
+
0,
|
| 188 |
+
0,
|
| 189 |
+
65
|
| 190 |
+
]
|
| 191 |
+
],
|
| 192 |
+
"eval_f1": 0.6331541584523206,
|
| 193 |
+
"eval_loss": 1.0080645084381104,
|
| 194 |
+
"eval_precision": 0.7358031394913882,
|
| 195 |
+
"eval_recall": 0.6666666666666666,
|
| 196 |
+
"eval_runtime": 2.9108,
|
| 197 |
+
"eval_samples_per_second": 93.789,
|
| 198 |
+
"eval_steps_per_second": 1.031,
|
| 199 |
+
"step": 22
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 6.0,
|
| 203 |
+
"eval_accuracy": 0.652014652014652,
|
| 204 |
+
"eval_confusion_matrix": [
|
| 205 |
+
[
|
| 206 |
+
66,
|
| 207 |
+
0,
|
| 208 |
+
0,
|
| 209 |
+
6
|
| 210 |
+
],
|
| 211 |
+
[
|
| 212 |
+
36,
|
| 213 |
+
16,
|
| 214 |
+
8,
|
| 215 |
+
0
|
| 216 |
+
],
|
| 217 |
+
[
|
| 218 |
+
10,
|
| 219 |
+
35,
|
| 220 |
+
31,
|
| 221 |
+
0
|
| 222 |
+
],
|
| 223 |
+
[
|
| 224 |
+
0,
|
| 225 |
+
0,
|
| 226 |
+
0,
|
| 227 |
+
65
|
| 228 |
+
]
|
| 229 |
+
],
|
| 230 |
+
"eval_f1": 0.6302407955860642,
|
| 231 |
+
"eval_loss": 0.9598046541213989,
|
| 232 |
+
"eval_precision": 0.6636241740077424,
|
| 233 |
+
"eval_recall": 0.652014652014652,
|
| 234 |
+
"eval_runtime": 2.9773,
|
| 235 |
+
"eval_samples_per_second": 91.693,
|
| 236 |
+
"eval_steps_per_second": 1.008,
|
| 237 |
+
"step": 27
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 6.888888888888889,
|
| 241 |
+
"eval_accuracy": 0.7692307692307693,
|
| 242 |
+
"eval_confusion_matrix": [
|
| 243 |
+
[
|
| 244 |
+
66,
|
| 245 |
+
0,
|
| 246 |
+
0,
|
| 247 |
+
6
|
| 248 |
+
],
|
| 249 |
+
[
|
| 250 |
+
29,
|
| 251 |
+
21,
|
| 252 |
+
10,
|
| 253 |
+
0
|
| 254 |
+
],
|
| 255 |
+
[
|
| 256 |
+
3,
|
| 257 |
+
15,
|
| 258 |
+
58,
|
| 259 |
+
0
|
| 260 |
+
],
|
| 261 |
+
[
|
| 262 |
+
0,
|
| 263 |
+
0,
|
| 264 |
+
0,
|
| 265 |
+
65
|
| 266 |
+
]
|
| 267 |
+
],
|
| 268 |
+
"eval_f1": 0.7527855586679116,
|
| 269 |
+
"eval_loss": 0.7898163795471191,
|
| 270 |
+
"eval_precision": 0.7612467677056892,
|
| 271 |
+
"eval_recall": 0.7692307692307693,
|
| 272 |
+
"eval_runtime": 2.8918,
|
| 273 |
+
"eval_samples_per_second": 94.404,
|
| 274 |
+
"eval_steps_per_second": 1.037,
|
| 275 |
+
"step": 31
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 8.0,
|
| 279 |
+
"eval_accuracy": 0.7802197802197802,
|
| 280 |
+
"eval_confusion_matrix": [
|
| 281 |
+
[
|
| 282 |
+
66,
|
| 283 |
+
0,
|
| 284 |
+
0,
|
| 285 |
+
6
|
| 286 |
+
],
|
| 287 |
+
[
|
| 288 |
+
25,
|
| 289 |
+
15,
|
| 290 |
+
19,
|
| 291 |
+
1
|
| 292 |
+
],
|
| 293 |
+
[
|
| 294 |
+
1,
|
| 295 |
+
6,
|
| 296 |
+
68,
|
| 297 |
+
1
|
| 298 |
+
],
|
| 299 |
+
[
|
| 300 |
+
1,
|
| 301 |
+
0,
|
| 302 |
+
0,
|
| 303 |
+
64
|
| 304 |
+
]
|
| 305 |
+
],
|
| 306 |
+
"eval_f1": 0.7471179200524057,
|
| 307 |
+
"eval_loss": 0.7337484359741211,
|
| 308 |
+
"eval_precision": 0.7733845922309294,
|
| 309 |
+
"eval_recall": 0.7802197802197802,
|
| 310 |
+
"eval_runtime": 2.9006,
|
| 311 |
+
"eval_samples_per_second": 94.119,
|
| 312 |
+
"eval_steps_per_second": 1.034,
|
| 313 |
+
"step": 36
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 8.88888888888889,
|
| 317 |
+
"eval_accuracy": 0.7912087912087912,
|
| 318 |
+
"eval_confusion_matrix": [
|
| 319 |
+
[
|
| 320 |
+
57,
|
| 321 |
+
8,
|
| 322 |
+
3,
|
| 323 |
+
4
|
| 324 |
+
],
|
| 325 |
+
[
|
| 326 |
+
13,
|
| 327 |
+
24,
|
| 328 |
+
23,
|
| 329 |
+
0
|
| 330 |
+
],
|
| 331 |
+
[
|
| 332 |
+
0,
|
| 333 |
+
5,
|
| 334 |
+
71,
|
| 335 |
+
0
|
| 336 |
+
],
|
| 337 |
+
[
|
| 338 |
+
1,
|
| 339 |
+
0,
|
| 340 |
+
0,
|
| 341 |
+
64
|
| 342 |
+
]
|
| 343 |
+
],
|
| 344 |
+
"eval_f1": 0.7766564722186922,
|
| 345 |
+
"eval_loss": 0.7148727178573608,
|
| 346 |
+
"eval_precision": 0.7821504483074875,
|
| 347 |
+
"eval_recall": 0.7912087912087912,
|
| 348 |
+
"eval_runtime": 2.8866,
|
| 349 |
+
"eval_samples_per_second": 94.576,
|
| 350 |
+
"eval_steps_per_second": 1.039,
|
| 351 |
+
"step": 40
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 10.0,
|
| 355 |
+
"eval_accuracy": 0.8168498168498168,
|
| 356 |
+
"eval_confusion_matrix": [
|
| 357 |
+
[
|
| 358 |
+
51,
|
| 359 |
+
8,
|
| 360 |
+
1,
|
| 361 |
+
12
|
| 362 |
+
],
|
| 363 |
+
[
|
| 364 |
+
10,
|
| 365 |
+
36,
|
| 366 |
+
14,
|
| 367 |
+
0
|
| 368 |
+
],
|
| 369 |
+
[
|
| 370 |
+
0,
|
| 371 |
+
5,
|
| 372 |
+
71,
|
| 373 |
+
0
|
| 374 |
+
],
|
| 375 |
+
[
|
| 376 |
+
0,
|
| 377 |
+
0,
|
| 378 |
+
0,
|
| 379 |
+
65
|
| 380 |
+
]
|
| 381 |
+
],
|
| 382 |
+
"eval_f1": 0.8094336953840884,
|
| 383 |
+
"eval_loss": 0.7574812173843384,
|
| 384 |
+
"eval_precision": 0.8127936625684181,
|
| 385 |
+
"eval_recall": 0.8168498168498168,
|
| 386 |
+
"eval_runtime": 2.8967,
|
| 387 |
+
"eval_samples_per_second": 94.245,
|
| 388 |
+
"eval_steps_per_second": 1.036,
|
| 389 |
+
"step": 45
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 10.88888888888889,
|
| 393 |
+
"eval_accuracy": 0.8021978021978022,
|
| 394 |
+
"eval_confusion_matrix": [
|
| 395 |
+
[
|
| 396 |
+
48,
|
| 397 |
+
18,
|
| 398 |
+
2,
|
| 399 |
+
4
|
| 400 |
+
],
|
| 401 |
+
[
|
| 402 |
+
4,
|
| 403 |
+
41,
|
| 404 |
+
15,
|
| 405 |
+
0
|
| 406 |
+
],
|
| 407 |
+
[
|
| 408 |
+
0,
|
| 409 |
+
7,
|
| 410 |
+
69,
|
| 411 |
+
0
|
| 412 |
+
],
|
| 413 |
+
[
|
| 414 |
+
3,
|
| 415 |
+
0,
|
| 416 |
+
1,
|
| 417 |
+
61
|
| 418 |
+
]
|
| 419 |
+
],
|
| 420 |
+
"eval_f1": 0.801525180147331,
|
| 421 |
+
"eval_loss": 0.7140281200408936,
|
| 422 |
+
"eval_precision": 0.8109338936925145,
|
| 423 |
+
"eval_recall": 0.8021978021978022,
|
| 424 |
+
"eval_runtime": 2.917,
|
| 425 |
+
"eval_samples_per_second": 93.589,
|
| 426 |
+
"eval_steps_per_second": 1.028,
|
| 427 |
+
"step": 49
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 12.0,
|
| 431 |
+
"eval_accuracy": 0.8534798534798534,
|
| 432 |
+
"eval_confusion_matrix": [
|
| 433 |
+
[
|
| 434 |
+
56,
|
| 435 |
+
12,
|
| 436 |
+
0,
|
| 437 |
+
4
|
| 438 |
+
],
|
| 439 |
+
[
|
| 440 |
+
6,
|
| 441 |
+
43,
|
| 442 |
+
11,
|
| 443 |
+
0
|
| 444 |
+
],
|
| 445 |
+
[
|
| 446 |
+
0,
|
| 447 |
+
6,
|
| 448 |
+
70,
|
| 449 |
+
0
|
| 450 |
+
],
|
| 451 |
+
[
|
| 452 |
+
1,
|
| 453 |
+
0,
|
| 454 |
+
0,
|
| 455 |
+
64
|
| 456 |
+
]
|
| 457 |
+
],
|
| 458 |
+
"eval_f1": 0.8523994617102314,
|
| 459 |
+
"eval_loss": 0.6672152280807495,
|
| 460 |
+
"eval_precision": 0.8540313732642031,
|
| 461 |
+
"eval_recall": 0.8534798534798534,
|
| 462 |
+
"eval_runtime": 2.8969,
|
| 463 |
+
"eval_samples_per_second": 94.238,
|
| 464 |
+
"eval_steps_per_second": 1.036,
|
| 465 |
+
"step": 54
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 12.88888888888889,
|
| 469 |
+
"eval_accuracy": 0.8498168498168498,
|
| 470 |
+
"eval_confusion_matrix": [
|
| 471 |
+
[
|
| 472 |
+
60,
|
| 473 |
+
8,
|
| 474 |
+
0,
|
| 475 |
+
4
|
| 476 |
+
],
|
| 477 |
+
[
|
| 478 |
+
8,
|
| 479 |
+
38,
|
| 480 |
+
14,
|
| 481 |
+
0
|
| 482 |
+
],
|
| 483 |
+
[
|
| 484 |
+
0,
|
| 485 |
+
6,
|
| 486 |
+
70,
|
| 487 |
+
0
|
| 488 |
+
],
|
| 489 |
+
[
|
| 490 |
+
1,
|
| 491 |
+
0,
|
| 492 |
+
0,
|
| 493 |
+
64
|
| 494 |
+
]
|
| 495 |
+
],
|
| 496 |
+
"eval_f1": 0.8463270052615757,
|
| 497 |
+
"eval_loss": 0.6432910561561584,
|
| 498 |
+
"eval_precision": 0.8460243715014519,
|
| 499 |
+
"eval_recall": 0.8498168498168498,
|
| 500 |
+
"eval_runtime": 3.0281,
|
| 501 |
+
"eval_samples_per_second": 90.156,
|
| 502 |
+
"eval_steps_per_second": 0.991,
|
| 503 |
+
"step": 58
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 14.0,
|
| 507 |
+
"eval_accuracy": 0.8278388278388278,
|
| 508 |
+
"eval_confusion_matrix": [
|
| 509 |
+
[
|
| 510 |
+
54,
|
| 511 |
+
13,
|
| 512 |
+
1,
|
| 513 |
+
4
|
| 514 |
+
],
|
| 515 |
+
[
|
| 516 |
+
7,
|
| 517 |
+
44,
|
| 518 |
+
9,
|
| 519 |
+
0
|
| 520 |
+
],
|
| 521 |
+
[
|
| 522 |
+
0,
|
| 523 |
+
11,
|
| 524 |
+
65,
|
| 525 |
+
0
|
| 526 |
+
],
|
| 527 |
+
[
|
| 528 |
+
1,
|
| 529 |
+
1,
|
| 530 |
+
0,
|
| 531 |
+
63
|
| 532 |
+
]
|
| 533 |
+
],
|
| 534 |
+
"eval_f1": 0.82943590265942,
|
| 535 |
+
"eval_loss": 0.7395206093788147,
|
| 536 |
+
"eval_precision": 0.8350059217447382,
|
| 537 |
+
"eval_recall": 0.8278388278388278,
|
| 538 |
+
"eval_runtime": 2.9869,
|
| 539 |
+
"eval_samples_per_second": 91.399,
|
| 540 |
+
"eval_steps_per_second": 1.004,
|
| 541 |
+
"step": 63
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 14.88888888888889,
|
| 545 |
+
"eval_accuracy": 0.8315018315018315,
|
| 546 |
+
"eval_confusion_matrix": [
|
| 547 |
+
[
|
| 548 |
+
54,
|
| 549 |
+
13,
|
| 550 |
+
1,
|
| 551 |
+
4
|
| 552 |
+
],
|
| 553 |
+
[
|
| 554 |
+
10,
|
| 555 |
+
39,
|
| 556 |
+
11,
|
| 557 |
+
0
|
| 558 |
+
],
|
| 559 |
+
[
|
| 560 |
+
0,
|
| 561 |
+
7,
|
| 562 |
+
69,
|
| 563 |
+
0
|
| 564 |
+
],
|
| 565 |
+
[
|
| 566 |
+
0,
|
| 567 |
+
0,
|
| 568 |
+
0,
|
| 569 |
+
65
|
| 570 |
+
]
|
| 571 |
+
],
|
| 572 |
+
"eval_f1": 0.8291811389886823,
|
| 573 |
+
"eval_loss": 0.7115849852561951,
|
| 574 |
+
"eval_precision": 0.829244108966536,
|
| 575 |
+
"eval_recall": 0.8315018315018315,
|
| 576 |
+
"eval_runtime": 2.9616,
|
| 577 |
+
"eval_samples_per_second": 92.181,
|
| 578 |
+
"eval_steps_per_second": 1.013,
|
| 579 |
+
"step": 67
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 16.0,
|
| 583 |
+
"eval_accuracy": 0.8315018315018315,
|
| 584 |
+
"eval_confusion_matrix": [
|
| 585 |
+
[
|
| 586 |
+
60,
|
| 587 |
+
7,
|
| 588 |
+
1,
|
| 589 |
+
4
|
| 590 |
+
],
|
| 591 |
+
[
|
| 592 |
+
10,
|
| 593 |
+
39,
|
| 594 |
+
11,
|
| 595 |
+
0
|
| 596 |
+
],
|
| 597 |
+
[
|
| 598 |
+
0,
|
| 599 |
+
11,
|
| 600 |
+
65,
|
| 601 |
+
0
|
| 602 |
+
],
|
| 603 |
+
[
|
| 604 |
+
1,
|
| 605 |
+
1,
|
| 606 |
+
0,
|
| 607 |
+
63
|
| 608 |
+
]
|
| 609 |
+
],
|
| 610 |
+
"eval_f1": 0.8304073820984628,
|
| 611 |
+
"eval_loss": 0.7295921444892883,
|
| 612 |
+
"eval_precision": 0.8295426562258641,
|
| 613 |
+
"eval_recall": 0.8315018315018315,
|
| 614 |
+
"eval_runtime": 2.8758,
|
| 615 |
+
"eval_samples_per_second": 94.932,
|
| 616 |
+
"eval_steps_per_second": 1.043,
|
| 617 |
+
"step": 72
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 16.88888888888889,
|
| 621 |
+
"eval_accuracy": 0.8644688644688645,
|
| 622 |
+
"eval_confusion_matrix": [
|
| 623 |
+
[
|
| 624 |
+
62,
|
| 625 |
+
5,
|
| 626 |
+
1,
|
| 627 |
+
4
|
| 628 |
+
],
|
| 629 |
+
[
|
| 630 |
+
9,
|
| 631 |
+
37,
|
| 632 |
+
14,
|
| 633 |
+
0
|
| 634 |
+
],
|
| 635 |
+
[
|
| 636 |
+
0,
|
| 637 |
+
3,
|
| 638 |
+
73,
|
| 639 |
+
0
|
| 640 |
+
],
|
| 641 |
+
[
|
| 642 |
+
1,
|
| 643 |
+
0,
|
| 644 |
+
0,
|
| 645 |
+
64
|
| 646 |
+
]
|
| 647 |
+
],
|
| 648 |
+
"eval_f1": 0.8589767100678526,
|
| 649 |
+
"eval_loss": 0.7055637240409851,
|
| 650 |
+
"eval_precision": 0.8628397746044805,
|
| 651 |
+
"eval_recall": 0.8644688644688645,
|
| 652 |
+
"eval_runtime": 3.074,
|
| 653 |
+
"eval_samples_per_second": 88.809,
|
| 654 |
+
"eval_steps_per_second": 0.976,
|
| 655 |
+
"step": 76
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 18.0,
|
| 659 |
+
"eval_accuracy": 0.8644688644688645,
|
| 660 |
+
"eval_confusion_matrix": [
|
| 661 |
+
[
|
| 662 |
+
65,
|
| 663 |
+
2,
|
| 664 |
+
1,
|
| 665 |
+
4
|
| 666 |
+
],
|
| 667 |
+
[
|
| 668 |
+
13,
|
| 669 |
+
42,
|
| 670 |
+
5,
|
| 671 |
+
0
|
| 672 |
+
],
|
| 673 |
+
[
|
| 674 |
+
0,
|
| 675 |
+
12,
|
| 676 |
+
64,
|
| 677 |
+
0
|
| 678 |
+
],
|
| 679 |
+
[
|
| 680 |
+
0,
|
| 681 |
+
0,
|
| 682 |
+
0,
|
| 683 |
+
65
|
| 684 |
+
]
|
| 685 |
+
],
|
| 686 |
+
"eval_f1": 0.8627769756077204,
|
| 687 |
+
"eval_loss": 0.7563945651054382,
|
| 688 |
+
"eval_precision": 0.8634344261673453,
|
| 689 |
+
"eval_recall": 0.8644688644688645,
|
| 690 |
+
"eval_runtime": 3.0072,
|
| 691 |
+
"eval_samples_per_second": 90.783,
|
| 692 |
+
"eval_steps_per_second": 0.998,
|
| 693 |
+
"step": 81
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 18.88888888888889,
|
| 697 |
+
"eval_accuracy": 0.8424908424908425,
|
| 698 |
+
"eval_confusion_matrix": [
|
| 699 |
+
[
|
| 700 |
+
64,
|
| 701 |
+
4,
|
| 702 |
+
0,
|
| 703 |
+
4
|
| 704 |
+
],
|
| 705 |
+
[
|
| 706 |
+
11,
|
| 707 |
+
41,
|
| 708 |
+
8,
|
| 709 |
+
0
|
| 710 |
+
],
|
| 711 |
+
[
|
| 712 |
+
0,
|
| 713 |
+
10,
|
| 714 |
+
66,
|
| 715 |
+
0
|
| 716 |
+
],
|
| 717 |
+
[
|
| 718 |
+
6,
|
| 719 |
+
0,
|
| 720 |
+
0,
|
| 721 |
+
59
|
| 722 |
+
]
|
| 723 |
+
],
|
| 724 |
+
"eval_f1": 0.8418306879608031,
|
| 725 |
+
"eval_loss": 0.7825365662574768,
|
| 726 |
+
"eval_precision": 0.8434907006335578,
|
| 727 |
+
"eval_recall": 0.8424908424908425,
|
| 728 |
+
"eval_runtime": 3.0343,
|
| 729 |
+
"eval_samples_per_second": 89.972,
|
| 730 |
+
"eval_steps_per_second": 0.989,
|
| 731 |
+
"step": 85
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 20.0,
|
| 735 |
+
"eval_accuracy": 0.8058608058608059,
|
| 736 |
+
"eval_confusion_matrix": [
|
| 737 |
+
[
|
| 738 |
+
40,
|
| 739 |
+
28,
|
| 740 |
+
1,
|
| 741 |
+
3
|
| 742 |
+
],
|
| 743 |
+
[
|
| 744 |
+
2,
|
| 745 |
+
50,
|
| 746 |
+
8,
|
| 747 |
+
0
|
| 748 |
+
],
|
| 749 |
+
[
|
| 750 |
+
0,
|
| 751 |
+
11,
|
| 752 |
+
65,
|
| 753 |
+
0
|
| 754 |
+
],
|
| 755 |
+
[
|
| 756 |
+
0,
|
| 757 |
+
0,
|
| 758 |
+
0,
|
| 759 |
+
65
|
| 760 |
+
]
|
| 761 |
+
],
|
| 762 |
+
"eval_f1": 0.8065760931078588,
|
| 763 |
+
"eval_loss": 0.8426868915557861,
|
| 764 |
+
"eval_precision": 0.8467707085637404,
|
| 765 |
+
"eval_recall": 0.8058608058608059,
|
| 766 |
+
"eval_runtime": 3.0783,
|
| 767 |
+
"eval_samples_per_second": 88.686,
|
| 768 |
+
"eval_steps_per_second": 0.975,
|
| 769 |
+
"step": 90
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 20.88888888888889,
|
| 773 |
+
"eval_accuracy": 0.8498168498168498,
|
| 774 |
+
"eval_confusion_matrix": [
|
| 775 |
+
[
|
| 776 |
+
63,
|
| 777 |
+
4,
|
| 778 |
+
0,
|
| 779 |
+
5
|
| 780 |
+
],
|
| 781 |
+
[
|
| 782 |
+
13,
|
| 783 |
+
40,
|
| 784 |
+
7,
|
| 785 |
+
0
|
| 786 |
+
],
|
| 787 |
+
[
|
| 788 |
+
0,
|
| 789 |
+
12,
|
| 790 |
+
64,
|
| 791 |
+
0
|
| 792 |
+
],
|
| 793 |
+
[
|
| 794 |
+
0,
|
| 795 |
+
0,
|
| 796 |
+
0,
|
| 797 |
+
65
|
| 798 |
+
]
|
| 799 |
+
],
|
| 800 |
+
"eval_f1": 0.8477878057985963,
|
| 801 |
+
"eval_loss": 0.7440442442893982,
|
| 802 |
+
"eval_precision": 0.8476393351433065,
|
| 803 |
+
"eval_recall": 0.8498168498168498,
|
| 804 |
+
"eval_runtime": 2.9875,
|
| 805 |
+
"eval_samples_per_second": 91.38,
|
| 806 |
+
"eval_steps_per_second": 1.004,
|
| 807 |
+
"step": 94
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 22.0,
|
| 811 |
+
"eval_accuracy": 0.8608058608058609,
|
| 812 |
+
"eval_confusion_matrix": [
|
| 813 |
+
[
|
| 814 |
+
66,
|
| 815 |
+
1,
|
| 816 |
+
1,
|
| 817 |
+
4
|
| 818 |
+
],
|
| 819 |
+
[
|
| 820 |
+
13,
|
| 821 |
+
36,
|
| 822 |
+
11,
|
| 823 |
+
0
|
| 824 |
+
],
|
| 825 |
+
[
|
| 826 |
+
0,
|
| 827 |
+
5,
|
| 828 |
+
71,
|
| 829 |
+
0
|
| 830 |
+
],
|
| 831 |
+
[
|
| 832 |
+
3,
|
| 833 |
+
0,
|
| 834 |
+
0,
|
| 835 |
+
62
|
| 836 |
+
]
|
| 837 |
+
],
|
| 838 |
+
"eval_f1": 0.855194718990792,
|
| 839 |
+
"eval_loss": 0.7338178753852844,
|
| 840 |
+
"eval_precision": 0.8624631692093176,
|
| 841 |
+
"eval_recall": 0.8608058608058609,
|
| 842 |
+
"eval_runtime": 2.98,
|
| 843 |
+
"eval_samples_per_second": 91.611,
|
| 844 |
+
"eval_steps_per_second": 1.007,
|
| 845 |
+
"step": 99
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 22.88888888888889,
|
| 849 |
+
"eval_accuracy": 0.8498168498168498,
|
| 850 |
+
"eval_confusion_matrix": [
|
| 851 |
+
[
|
| 852 |
+
58,
|
| 853 |
+
10,
|
| 854 |
+
0,
|
| 855 |
+
4
|
| 856 |
+
],
|
| 857 |
+
[
|
| 858 |
+
10,
|
| 859 |
+
43,
|
| 860 |
+
7,
|
| 861 |
+
0
|
| 862 |
+
],
|
| 863 |
+
[
|
| 864 |
+
0,
|
| 865 |
+
7,
|
| 866 |
+
69,
|
| 867 |
+
0
|
| 868 |
+
],
|
| 869 |
+
[
|
| 870 |
+
2,
|
| 871 |
+
1,
|
| 872 |
+
0,
|
| 873 |
+
62
|
| 874 |
+
]
|
| 875 |
+
],
|
| 876 |
+
"eval_f1": 0.8497737987724402,
|
| 877 |
+
"eval_loss": 0.7231407761573792,
|
| 878 |
+
"eval_precision": 0.8498638493954653,
|
| 879 |
+
"eval_recall": 0.8498168498168498,
|
| 880 |
+
"eval_runtime": 2.9272,
|
| 881 |
+
"eval_samples_per_second": 93.262,
|
| 882 |
+
"eval_steps_per_second": 1.025,
|
| 883 |
+
"step": 103
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 24.0,
|
| 887 |
+
"eval_accuracy": 0.8424908424908425,
|
| 888 |
+
"eval_confusion_matrix": [
|
| 889 |
+
[
|
| 890 |
+
67,
|
| 891 |
+
1,
|
| 892 |
+
0,
|
| 893 |
+
4
|
| 894 |
+
],
|
| 895 |
+
[
|
| 896 |
+
12,
|
| 897 |
+
44,
|
| 898 |
+
4,
|
| 899 |
+
0
|
| 900 |
+
],
|
| 901 |
+
[
|
| 902 |
+
0,
|
| 903 |
+
18,
|
| 904 |
+
58,
|
| 905 |
+
0
|
| 906 |
+
],
|
| 907 |
+
[
|
| 908 |
+
2,
|
| 909 |
+
2,
|
| 910 |
+
0,
|
| 911 |
+
61
|
| 912 |
+
]
|
| 913 |
+
],
|
| 914 |
+
"eval_f1": 0.8431607380967995,
|
| 915 |
+
"eval_loss": 0.752363920211792,
|
| 916 |
+
"eval_precision": 0.8507974885146101,
|
| 917 |
+
"eval_recall": 0.8424908424908425,
|
| 918 |
+
"eval_runtime": 3.0257,
|
| 919 |
+
"eval_samples_per_second": 90.228,
|
| 920 |
+
"eval_steps_per_second": 0.992,
|
| 921 |
+
"step": 108
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 24.88888888888889,
|
| 925 |
+
"eval_accuracy": 0.8498168498168498,
|
| 926 |
+
"eval_confusion_matrix": [
|
| 927 |
+
[
|
| 928 |
+
66,
|
| 929 |
+
0,
|
| 930 |
+
1,
|
| 931 |
+
5
|
| 932 |
+
],
|
| 933 |
+
[
|
| 934 |
+
15,
|
| 935 |
+
28,
|
| 936 |
+
17,
|
| 937 |
+
0
|
| 938 |
+
],
|
| 939 |
+
[
|
| 940 |
+
0,
|
| 941 |
+
2,
|
| 942 |
+
74,
|
| 943 |
+
0
|
| 944 |
+
],
|
| 945 |
+
[
|
| 946 |
+
1,
|
| 947 |
+
0,
|
| 948 |
+
0,
|
| 949 |
+
64
|
| 950 |
+
]
|
| 951 |
+
],
|
| 952 |
+
"eval_f1": 0.8354928653436116,
|
| 953 |
+
"eval_loss": 0.7849779725074768,
|
| 954 |
+
"eval_precision": 0.8621667009790022,
|
| 955 |
+
"eval_recall": 0.8498168498168498,
|
| 956 |
+
"eval_runtime": 2.8819,
|
| 957 |
+
"eval_samples_per_second": 94.73,
|
| 958 |
+
"eval_steps_per_second": 1.041,
|
| 959 |
+
"step": 112
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 26.0,
|
| 963 |
+
"eval_accuracy": 0.8131868131868132,
|
| 964 |
+
"eval_confusion_matrix": [
|
| 965 |
+
[
|
| 966 |
+
62,
|
| 967 |
+
6,
|
| 968 |
+
0,
|
| 969 |
+
4
|
| 970 |
+
],
|
| 971 |
+
[
|
| 972 |
+
11,
|
| 973 |
+
44,
|
| 974 |
+
5,
|
| 975 |
+
0
|
| 976 |
+
],
|
| 977 |
+
[
|
| 978 |
+
0,
|
| 979 |
+
24,
|
| 980 |
+
52,
|
| 981 |
+
0
|
| 982 |
+
],
|
| 983 |
+
[
|
| 984 |
+
1,
|
| 985 |
+
0,
|
| 986 |
+
0,
|
| 987 |
+
64
|
| 988 |
+
]
|
| 989 |
+
],
|
| 990 |
+
"eval_f1": 0.8151598256855266,
|
| 991 |
+
"eval_loss": 0.7896661162376404,
|
| 992 |
+
"eval_precision": 0.8297062414709473,
|
| 993 |
+
"eval_recall": 0.8131868131868132,
|
| 994 |
+
"eval_runtime": 2.8953,
|
| 995 |
+
"eval_samples_per_second": 94.291,
|
| 996 |
+
"eval_steps_per_second": 1.036,
|
| 997 |
+
"step": 117
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 26.88888888888889,
|
| 1001 |
+
"eval_accuracy": 0.8791208791208791,
|
| 1002 |
+
"eval_confusion_matrix": [
|
| 1003 |
+
[
|
| 1004 |
+
66,
|
| 1005 |
+
1,
|
| 1006 |
+
1,
|
| 1007 |
+
4
|
| 1008 |
+
],
|
| 1009 |
+
[
|
| 1010 |
+
13,
|
| 1011 |
+
40,
|
| 1012 |
+
7,
|
| 1013 |
+
0
|
| 1014 |
+
],
|
| 1015 |
+
[
|
| 1016 |
+
1,
|
| 1017 |
+
6,
|
| 1018 |
+
69,
|
| 1019 |
+
0
|
| 1020 |
+
],
|
| 1021 |
+
[
|
| 1022 |
+
0,
|
| 1023 |
+
0,
|
| 1024 |
+
0,
|
| 1025 |
+
65
|
| 1026 |
+
]
|
| 1027 |
+
],
|
| 1028 |
+
"eval_f1": 0.8754390108936493,
|
| 1029 |
+
"eval_loss": 0.7321063876152039,
|
| 1030 |
+
"eval_precision": 0.878386849630407,
|
| 1031 |
+
"eval_recall": 0.8791208791208791,
|
| 1032 |
+
"eval_runtime": 2.8941,
|
| 1033 |
+
"eval_samples_per_second": 94.329,
|
| 1034 |
+
"eval_steps_per_second": 1.037,
|
| 1035 |
+
"step": 121
|
| 1036 |
+
}
|
| 1037 |
+
],
|
| 1038 |
+
"logging_steps": 500,
|
| 1039 |
+
"max_steps": 180,
|
| 1040 |
+
"num_input_tokens_seen": 0,
|
| 1041 |
+
"num_train_epochs": 45,
|
| 1042 |
+
"save_steps": 500,
|
| 1043 |
+
"stateful_callbacks": {
|
| 1044 |
+
"TrainerControl": {
|
| 1045 |
+
"args": {
|
| 1046 |
+
"should_epoch_stop": false,
|
| 1047 |
+
"should_evaluate": false,
|
| 1048 |
+
"should_log": false,
|
| 1049 |
+
"should_save": true,
|
| 1050 |
+
"should_training_stop": false
|
| 1051 |
+
},
|
| 1052 |
+
"attributes": {}
|
| 1053 |
+
}
|
| 1054 |
+
},
|
| 1055 |
+
"total_flos": 6.69304876608e+16,
|
| 1056 |
+
"train_batch_size": 128,
|
| 1057 |
+
"trial_name": null,
|
| 1058 |
+
"trial_params": null
|
| 1059 |
+
}
|
checkpoint-121/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:09e217cad0464b642c54f8b7691100b62a93dceb32a1869f55e3f0eb3a54a79e
|
| 3 |
+
size 5240
|
checkpoint-126/config.json
ADDED
|
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ntu-spml/distilhubert",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"HubertForSequenceClassification"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 1,
|
| 10 |
+
"classifier_proj_size": 256,
|
| 11 |
+
"conv_bias": false,
|
| 12 |
+
"conv_dim": [
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512
|
| 20 |
+
],
|
| 21 |
+
"conv_kernel": [
|
| 22 |
+
10,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
3,
|
| 27 |
+
2,
|
| 28 |
+
2
|
| 29 |
+
],
|
| 30 |
+
"conv_stride": [
|
| 31 |
+
5,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2,
|
| 37 |
+
2
|
| 38 |
+
],
|
| 39 |
+
"ctc_loss_reduction": "sum",
|
| 40 |
+
"ctc_zero_infinity": false,
|
| 41 |
+
"do_stable_layer_norm": false,
|
| 42 |
+
"eos_token_id": 2,
|
| 43 |
+
"feat_extract_activation": "gelu",
|
| 44 |
+
"feat_extract_norm": "group",
|
| 45 |
+
"feat_proj_dropout": 0.0,
|
| 46 |
+
"feat_proj_layer_norm": false,
|
| 47 |
+
"final_dropout": 0.0,
|
| 48 |
+
"finetuning_task": "audio-classification",
|
| 49 |
+
"hidden_act": "gelu",
|
| 50 |
+
"hidden_dropout": 0.1,
|
| 51 |
+
"hidden_size": 768,
|
| 52 |
+
"id2label": {
|
| 53 |
+
"0": "1s_normal",
|
| 54 |
+
"1": "1s_pain",
|
| 55 |
+
"2": "1s_hunger",
|
| 56 |
+
"3": "1s_asphyxia"
|
| 57 |
+
},
|
| 58 |
+
"initializer_range": 0.02,
|
| 59 |
+
"intermediate_size": 3072,
|
| 60 |
+
"label2id": {
|
| 61 |
+
"LABEL_0": 0,
|
| 62 |
+
"LABEL_1": 1,
|
| 63 |
+
"LABEL_2": 2,
|
| 64 |
+
"LABEL_3": 3
|
| 65 |
+
},
|
| 66 |
+
"layer_norm_eps": 1e-05,
|
| 67 |
+
"layerdrop": 0.0,
|
| 68 |
+
"mask_feature_length": 10,
|
| 69 |
+
"mask_feature_min_masks": 0,
|
| 70 |
+
"mask_feature_prob": 0.0,
|
| 71 |
+
"mask_time_length": 10,
|
| 72 |
+
"mask_time_min_masks": 2,
|
| 73 |
+
"mask_time_prob": 0.05,
|
| 74 |
+
"model_type": "hubert",
|
| 75 |
+
"num_attention_heads": 12,
|
| 76 |
+
"num_conv_pos_embedding_groups": 16,
|
| 77 |
+
"num_conv_pos_embeddings": 128,
|
| 78 |
+
"num_feat_extract_layers": 7,
|
| 79 |
+
"num_hidden_layers": 2,
|
| 80 |
+
"pad_token_id": 0,
|
| 81 |
+
"torch_dtype": "float32",
|
| 82 |
+
"transformers_version": "4.44.2",
|
| 83 |
+
"use_weighted_layer_sum": false,
|
| 84 |
+
"vocab_size": 32
|
| 85 |
+
}
|
checkpoint-126/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:06f0d1b8ea136d0eeebfeb45abb3a0bd540dc213dbae5c03e5bf2ebfe926bc7b
|
| 3 |
+
size 94765560
|
checkpoint-126/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b3390789d74ae0285ec7c077478bca5275fd78b948bd51a8210c4e4547eba1ae
|
| 3 |
+
size 189556666
|
checkpoint-126/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:63288390848aaf8c678d1ad2c771d062d656c8caedcb8079034b6b7b2b3c0359
|
| 3 |
+
size 14308
|
checkpoint-126/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:00fa2c49ac4b95860fb0f8bfb07178e641fa9e4de37635dd229a5022d1dcfb6e
|
| 3 |
+
size 1064
|
checkpoint-126/trainer_state.json
ADDED
|
@@ -0,0 +1,1097 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8754390108936493,
|
| 3 |
+
"best_model_checkpoint": "distilhubert-finetuned-mixed-data/checkpoint-121",
|
| 4 |
+
"epoch": 28.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 126,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.8888888888888888,
|
| 13 |
+
"eval_accuracy": 0.63003663003663,
|
| 14 |
+
"eval_confusion_matrix": [
|
| 15 |
+
[
|
| 16 |
+
53,
|
| 17 |
+
0,
|
| 18 |
+
16,
|
| 19 |
+
3
|
| 20 |
+
],
|
| 21 |
+
[
|
| 22 |
+
28,
|
| 23 |
+
1,
|
| 24 |
+
31,
|
| 25 |
+
0
|
| 26 |
+
],
|
| 27 |
+
[
|
| 28 |
+
13,
|
| 29 |
+
0,
|
| 30 |
+
62,
|
| 31 |
+
1
|
| 32 |
+
],
|
| 33 |
+
[
|
| 34 |
+
3,
|
| 35 |
+
0,
|
| 36 |
+
6,
|
| 37 |
+
56
|
| 38 |
+
]
|
| 39 |
+
],
|
| 40 |
+
"eval_f1": 0.566693372031096,
|
| 41 |
+
"eval_loss": 1.31327486038208,
|
| 42 |
+
"eval_precision": 0.7361933549293478,
|
| 43 |
+
"eval_recall": 0.63003663003663,
|
| 44 |
+
"eval_runtime": 2.8436,
|
| 45 |
+
"eval_samples_per_second": 96.005,
|
| 46 |
+
"eval_steps_per_second": 1.055,
|
| 47 |
+
"step": 4
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 2.0,
|
| 51 |
+
"eval_accuracy": 0.6593406593406593,
|
| 52 |
+
"eval_confusion_matrix": [
|
| 53 |
+
[
|
| 54 |
+
41,
|
| 55 |
+
10,
|
| 56 |
+
16,
|
| 57 |
+
5
|
| 58 |
+
],
|
| 59 |
+
[
|
| 60 |
+
21,
|
| 61 |
+
1,
|
| 62 |
+
38,
|
| 63 |
+
0
|
| 64 |
+
],
|
| 65 |
+
[
|
| 66 |
+
1,
|
| 67 |
+
2,
|
| 68 |
+
73,
|
| 69 |
+
0
|
| 70 |
+
],
|
| 71 |
+
[
|
| 72 |
+
0,
|
| 73 |
+
0,
|
| 74 |
+
0,
|
| 75 |
+
65
|
| 76 |
+
]
|
| 77 |
+
],
|
| 78 |
+
"eval_f1": 0.595713773130153,
|
| 79 |
+
"eval_loss": 0.9242589473724365,
|
| 80 |
+
"eval_precision": 0.5696509512811321,
|
| 81 |
+
"eval_recall": 0.6593406593406593,
|
| 82 |
+
"eval_runtime": 2.8668,
|
| 83 |
+
"eval_samples_per_second": 95.227,
|
| 84 |
+
"eval_steps_per_second": 1.046,
|
| 85 |
+
"step": 9
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 2.888888888888889,
|
| 89 |
+
"eval_accuracy": 0.717948717948718,
|
| 90 |
+
"eval_confusion_matrix": [
|
| 91 |
+
[
|
| 92 |
+
52,
|
| 93 |
+
8,
|
| 94 |
+
7,
|
| 95 |
+
5
|
| 96 |
+
],
|
| 97 |
+
[
|
| 98 |
+
18,
|
| 99 |
+
5,
|
| 100 |
+
37,
|
| 101 |
+
0
|
| 102 |
+
],
|
| 103 |
+
[
|
| 104 |
+
1,
|
| 105 |
+
0,
|
| 106 |
+
75,
|
| 107 |
+
0
|
| 108 |
+
],
|
| 109 |
+
[
|
| 110 |
+
1,
|
| 111 |
+
0,
|
| 112 |
+
0,
|
| 113 |
+
64
|
| 114 |
+
]
|
| 115 |
+
],
|
| 116 |
+
"eval_f1": 0.6621621567125869,
|
| 117 |
+
"eval_loss": 0.8232662081718445,
|
| 118 |
+
"eval_precision": 0.671303801513745,
|
| 119 |
+
"eval_recall": 0.717948717948718,
|
| 120 |
+
"eval_runtime": 2.8711,
|
| 121 |
+
"eval_samples_per_second": 95.084,
|
| 122 |
+
"eval_steps_per_second": 1.045,
|
| 123 |
+
"step": 13
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 4.0,
|
| 127 |
+
"eval_accuracy": 0.6959706959706959,
|
| 128 |
+
"eval_confusion_matrix": [
|
| 129 |
+
[
|
| 130 |
+
45,
|
| 131 |
+
9,
|
| 132 |
+
13,
|
| 133 |
+
5
|
| 134 |
+
],
|
| 135 |
+
[
|
| 136 |
+
13,
|
| 137 |
+
5,
|
| 138 |
+
42,
|
| 139 |
+
0
|
| 140 |
+
],
|
| 141 |
+
[
|
| 142 |
+
1,
|
| 143 |
+
0,
|
| 144 |
+
75,
|
| 145 |
+
0
|
| 146 |
+
],
|
| 147 |
+
[
|
| 148 |
+
0,
|
| 149 |
+
0,
|
| 150 |
+
0,
|
| 151 |
+
65
|
| 152 |
+
]
|
| 153 |
+
],
|
| 154 |
+
"eval_f1": 0.6428796900399459,
|
| 155 |
+
"eval_loss": 0.951453447341919,
|
| 156 |
+
"eval_precision": 0.6613447705829899,
|
| 157 |
+
"eval_recall": 0.6959706959706959,
|
| 158 |
+
"eval_runtime": 2.8813,
|
| 159 |
+
"eval_samples_per_second": 94.75,
|
| 160 |
+
"eval_steps_per_second": 1.041,
|
| 161 |
+
"step": 18
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 4.888888888888889,
|
| 165 |
+
"eval_accuracy": 0.6666666666666666,
|
| 166 |
+
"eval_confusion_matrix": [
|
| 167 |
+
[
|
| 168 |
+
16,
|
| 169 |
+
49,
|
| 170 |
+
3,
|
| 171 |
+
4
|
| 172 |
+
],
|
| 173 |
+
[
|
| 174 |
+
1,
|
| 175 |
+
26,
|
| 176 |
+
33,
|
| 177 |
+
0
|
| 178 |
+
],
|
| 179 |
+
[
|
| 180 |
+
0,
|
| 181 |
+
1,
|
| 182 |
+
75,
|
| 183 |
+
0
|
| 184 |
+
],
|
| 185 |
+
[
|
| 186 |
+
0,
|
| 187 |
+
0,
|
| 188 |
+
0,
|
| 189 |
+
65
|
| 190 |
+
]
|
| 191 |
+
],
|
| 192 |
+
"eval_f1": 0.6331541584523206,
|
| 193 |
+
"eval_loss": 1.0080645084381104,
|
| 194 |
+
"eval_precision": 0.7358031394913882,
|
| 195 |
+
"eval_recall": 0.6666666666666666,
|
| 196 |
+
"eval_runtime": 2.9108,
|
| 197 |
+
"eval_samples_per_second": 93.789,
|
| 198 |
+
"eval_steps_per_second": 1.031,
|
| 199 |
+
"step": 22
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 6.0,
|
| 203 |
+
"eval_accuracy": 0.652014652014652,
|
| 204 |
+
"eval_confusion_matrix": [
|
| 205 |
+
[
|
| 206 |
+
66,
|
| 207 |
+
0,
|
| 208 |
+
0,
|
| 209 |
+
6
|
| 210 |
+
],
|
| 211 |
+
[
|
| 212 |
+
36,
|
| 213 |
+
16,
|
| 214 |
+
8,
|
| 215 |
+
0
|
| 216 |
+
],
|
| 217 |
+
[
|
| 218 |
+
10,
|
| 219 |
+
35,
|
| 220 |
+
31,
|
| 221 |
+
0
|
| 222 |
+
],
|
| 223 |
+
[
|
| 224 |
+
0,
|
| 225 |
+
0,
|
| 226 |
+
0,
|
| 227 |
+
65
|
| 228 |
+
]
|
| 229 |
+
],
|
| 230 |
+
"eval_f1": 0.6302407955860642,
|
| 231 |
+
"eval_loss": 0.9598046541213989,
|
| 232 |
+
"eval_precision": 0.6636241740077424,
|
| 233 |
+
"eval_recall": 0.652014652014652,
|
| 234 |
+
"eval_runtime": 2.9773,
|
| 235 |
+
"eval_samples_per_second": 91.693,
|
| 236 |
+
"eval_steps_per_second": 1.008,
|
| 237 |
+
"step": 27
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 6.888888888888889,
|
| 241 |
+
"eval_accuracy": 0.7692307692307693,
|
| 242 |
+
"eval_confusion_matrix": [
|
| 243 |
+
[
|
| 244 |
+
66,
|
| 245 |
+
0,
|
| 246 |
+
0,
|
| 247 |
+
6
|
| 248 |
+
],
|
| 249 |
+
[
|
| 250 |
+
29,
|
| 251 |
+
21,
|
| 252 |
+
10,
|
| 253 |
+
0
|
| 254 |
+
],
|
| 255 |
+
[
|
| 256 |
+
3,
|
| 257 |
+
15,
|
| 258 |
+
58,
|
| 259 |
+
0
|
| 260 |
+
],
|
| 261 |
+
[
|
| 262 |
+
0,
|
| 263 |
+
0,
|
| 264 |
+
0,
|
| 265 |
+
65
|
| 266 |
+
]
|
| 267 |
+
],
|
| 268 |
+
"eval_f1": 0.7527855586679116,
|
| 269 |
+
"eval_loss": 0.7898163795471191,
|
| 270 |
+
"eval_precision": 0.7612467677056892,
|
| 271 |
+
"eval_recall": 0.7692307692307693,
|
| 272 |
+
"eval_runtime": 2.8918,
|
| 273 |
+
"eval_samples_per_second": 94.404,
|
| 274 |
+
"eval_steps_per_second": 1.037,
|
| 275 |
+
"step": 31
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 8.0,
|
| 279 |
+
"eval_accuracy": 0.7802197802197802,
|
| 280 |
+
"eval_confusion_matrix": [
|
| 281 |
+
[
|
| 282 |
+
66,
|
| 283 |
+
0,
|
| 284 |
+
0,
|
| 285 |
+
6
|
| 286 |
+
],
|
| 287 |
+
[
|
| 288 |
+
25,
|
| 289 |
+
15,
|
| 290 |
+
19,
|
| 291 |
+
1
|
| 292 |
+
],
|
| 293 |
+
[
|
| 294 |
+
1,
|
| 295 |
+
6,
|
| 296 |
+
68,
|
| 297 |
+
1
|
| 298 |
+
],
|
| 299 |
+
[
|
| 300 |
+
1,
|
| 301 |
+
0,
|
| 302 |
+
0,
|
| 303 |
+
64
|
| 304 |
+
]
|
| 305 |
+
],
|
| 306 |
+
"eval_f1": 0.7471179200524057,
|
| 307 |
+
"eval_loss": 0.7337484359741211,
|
| 308 |
+
"eval_precision": 0.7733845922309294,
|
| 309 |
+
"eval_recall": 0.7802197802197802,
|
| 310 |
+
"eval_runtime": 2.9006,
|
| 311 |
+
"eval_samples_per_second": 94.119,
|
| 312 |
+
"eval_steps_per_second": 1.034,
|
| 313 |
+
"step": 36
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 8.88888888888889,
|
| 317 |
+
"eval_accuracy": 0.7912087912087912,
|
| 318 |
+
"eval_confusion_matrix": [
|
| 319 |
+
[
|
| 320 |
+
57,
|
| 321 |
+
8,
|
| 322 |
+
3,
|
| 323 |
+
4
|
| 324 |
+
],
|
| 325 |
+
[
|
| 326 |
+
13,
|
| 327 |
+
24,
|
| 328 |
+
23,
|
| 329 |
+
0
|
| 330 |
+
],
|
| 331 |
+
[
|
| 332 |
+
0,
|
| 333 |
+
5,
|
| 334 |
+
71,
|
| 335 |
+
0
|
| 336 |
+
],
|
| 337 |
+
[
|
| 338 |
+
1,
|
| 339 |
+
0,
|
| 340 |
+
0,
|
| 341 |
+
64
|
| 342 |
+
]
|
| 343 |
+
],
|
| 344 |
+
"eval_f1": 0.7766564722186922,
|
| 345 |
+
"eval_loss": 0.7148727178573608,
|
| 346 |
+
"eval_precision": 0.7821504483074875,
|
| 347 |
+
"eval_recall": 0.7912087912087912,
|
| 348 |
+
"eval_runtime": 2.8866,
|
| 349 |
+
"eval_samples_per_second": 94.576,
|
| 350 |
+
"eval_steps_per_second": 1.039,
|
| 351 |
+
"step": 40
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 10.0,
|
| 355 |
+
"eval_accuracy": 0.8168498168498168,
|
| 356 |
+
"eval_confusion_matrix": [
|
| 357 |
+
[
|
| 358 |
+
51,
|
| 359 |
+
8,
|
| 360 |
+
1,
|
| 361 |
+
12
|
| 362 |
+
],
|
| 363 |
+
[
|
| 364 |
+
10,
|
| 365 |
+
36,
|
| 366 |
+
14,
|
| 367 |
+
0
|
| 368 |
+
],
|
| 369 |
+
[
|
| 370 |
+
0,
|
| 371 |
+
5,
|
| 372 |
+
71,
|
| 373 |
+
0
|
| 374 |
+
],
|
| 375 |
+
[
|
| 376 |
+
0,
|
| 377 |
+
0,
|
| 378 |
+
0,
|
| 379 |
+
65
|
| 380 |
+
]
|
| 381 |
+
],
|
| 382 |
+
"eval_f1": 0.8094336953840884,
|
| 383 |
+
"eval_loss": 0.7574812173843384,
|
| 384 |
+
"eval_precision": 0.8127936625684181,
|
| 385 |
+
"eval_recall": 0.8168498168498168,
|
| 386 |
+
"eval_runtime": 2.8967,
|
| 387 |
+
"eval_samples_per_second": 94.245,
|
| 388 |
+
"eval_steps_per_second": 1.036,
|
| 389 |
+
"step": 45
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 10.88888888888889,
|
| 393 |
+
"eval_accuracy": 0.8021978021978022,
|
| 394 |
+
"eval_confusion_matrix": [
|
| 395 |
+
[
|
| 396 |
+
48,
|
| 397 |
+
18,
|
| 398 |
+
2,
|
| 399 |
+
4
|
| 400 |
+
],
|
| 401 |
+
[
|
| 402 |
+
4,
|
| 403 |
+
41,
|
| 404 |
+
15,
|
| 405 |
+
0
|
| 406 |
+
],
|
| 407 |
+
[
|
| 408 |
+
0,
|
| 409 |
+
7,
|
| 410 |
+
69,
|
| 411 |
+
0
|
| 412 |
+
],
|
| 413 |
+
[
|
| 414 |
+
3,
|
| 415 |
+
0,
|
| 416 |
+
1,
|
| 417 |
+
61
|
| 418 |
+
]
|
| 419 |
+
],
|
| 420 |
+
"eval_f1": 0.801525180147331,
|
| 421 |
+
"eval_loss": 0.7140281200408936,
|
| 422 |
+
"eval_precision": 0.8109338936925145,
|
| 423 |
+
"eval_recall": 0.8021978021978022,
|
| 424 |
+
"eval_runtime": 2.917,
|
| 425 |
+
"eval_samples_per_second": 93.589,
|
| 426 |
+
"eval_steps_per_second": 1.028,
|
| 427 |
+
"step": 49
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 12.0,
|
| 431 |
+
"eval_accuracy": 0.8534798534798534,
|
| 432 |
+
"eval_confusion_matrix": [
|
| 433 |
+
[
|
| 434 |
+
56,
|
| 435 |
+
12,
|
| 436 |
+
0,
|
| 437 |
+
4
|
| 438 |
+
],
|
| 439 |
+
[
|
| 440 |
+
6,
|
| 441 |
+
43,
|
| 442 |
+
11,
|
| 443 |
+
0
|
| 444 |
+
],
|
| 445 |
+
[
|
| 446 |
+
0,
|
| 447 |
+
6,
|
| 448 |
+
70,
|
| 449 |
+
0
|
| 450 |
+
],
|
| 451 |
+
[
|
| 452 |
+
1,
|
| 453 |
+
0,
|
| 454 |
+
0,
|
| 455 |
+
64
|
| 456 |
+
]
|
| 457 |
+
],
|
| 458 |
+
"eval_f1": 0.8523994617102314,
|
| 459 |
+
"eval_loss": 0.6672152280807495,
|
| 460 |
+
"eval_precision": 0.8540313732642031,
|
| 461 |
+
"eval_recall": 0.8534798534798534,
|
| 462 |
+
"eval_runtime": 2.8969,
|
| 463 |
+
"eval_samples_per_second": 94.238,
|
| 464 |
+
"eval_steps_per_second": 1.036,
|
| 465 |
+
"step": 54
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 12.88888888888889,
|
| 469 |
+
"eval_accuracy": 0.8498168498168498,
|
| 470 |
+
"eval_confusion_matrix": [
|
| 471 |
+
[
|
| 472 |
+
60,
|
| 473 |
+
8,
|
| 474 |
+
0,
|
| 475 |
+
4
|
| 476 |
+
],
|
| 477 |
+
[
|
| 478 |
+
8,
|
| 479 |
+
38,
|
| 480 |
+
14,
|
| 481 |
+
0
|
| 482 |
+
],
|
| 483 |
+
[
|
| 484 |
+
0,
|
| 485 |
+
6,
|
| 486 |
+
70,
|
| 487 |
+
0
|
| 488 |
+
],
|
| 489 |
+
[
|
| 490 |
+
1,
|
| 491 |
+
0,
|
| 492 |
+
0,
|
| 493 |
+
64
|
| 494 |
+
]
|
| 495 |
+
],
|
| 496 |
+
"eval_f1": 0.8463270052615757,
|
| 497 |
+
"eval_loss": 0.6432910561561584,
|
| 498 |
+
"eval_precision": 0.8460243715014519,
|
| 499 |
+
"eval_recall": 0.8498168498168498,
|
| 500 |
+
"eval_runtime": 3.0281,
|
| 501 |
+
"eval_samples_per_second": 90.156,
|
| 502 |
+
"eval_steps_per_second": 0.991,
|
| 503 |
+
"step": 58
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 14.0,
|
| 507 |
+
"eval_accuracy": 0.8278388278388278,
|
| 508 |
+
"eval_confusion_matrix": [
|
| 509 |
+
[
|
| 510 |
+
54,
|
| 511 |
+
13,
|
| 512 |
+
1,
|
| 513 |
+
4
|
| 514 |
+
],
|
| 515 |
+
[
|
| 516 |
+
7,
|
| 517 |
+
44,
|
| 518 |
+
9,
|
| 519 |
+
0
|
| 520 |
+
],
|
| 521 |
+
[
|
| 522 |
+
0,
|
| 523 |
+
11,
|
| 524 |
+
65,
|
| 525 |
+
0
|
| 526 |
+
],
|
| 527 |
+
[
|
| 528 |
+
1,
|
| 529 |
+
1,
|
| 530 |
+
0,
|
| 531 |
+
63
|
| 532 |
+
]
|
| 533 |
+
],
|
| 534 |
+
"eval_f1": 0.82943590265942,
|
| 535 |
+
"eval_loss": 0.7395206093788147,
|
| 536 |
+
"eval_precision": 0.8350059217447382,
|
| 537 |
+
"eval_recall": 0.8278388278388278,
|
| 538 |
+
"eval_runtime": 2.9869,
|
| 539 |
+
"eval_samples_per_second": 91.399,
|
| 540 |
+
"eval_steps_per_second": 1.004,
|
| 541 |
+
"step": 63
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 14.88888888888889,
|
| 545 |
+
"eval_accuracy": 0.8315018315018315,
|
| 546 |
+
"eval_confusion_matrix": [
|
| 547 |
+
[
|
| 548 |
+
54,
|
| 549 |
+
13,
|
| 550 |
+
1,
|
| 551 |
+
4
|
| 552 |
+
],
|
| 553 |
+
[
|
| 554 |
+
10,
|
| 555 |
+
39,
|
| 556 |
+
11,
|
| 557 |
+
0
|
| 558 |
+
],
|
| 559 |
+
[
|
| 560 |
+
0,
|
| 561 |
+
7,
|
| 562 |
+
69,
|
| 563 |
+
0
|
| 564 |
+
],
|
| 565 |
+
[
|
| 566 |
+
0,
|
| 567 |
+
0,
|
| 568 |
+
0,
|
| 569 |
+
65
|
| 570 |
+
]
|
| 571 |
+
],
|
| 572 |
+
"eval_f1": 0.8291811389886823,
|
| 573 |
+
"eval_loss": 0.7115849852561951,
|
| 574 |
+
"eval_precision": 0.829244108966536,
|
| 575 |
+
"eval_recall": 0.8315018315018315,
|
| 576 |
+
"eval_runtime": 2.9616,
|
| 577 |
+
"eval_samples_per_second": 92.181,
|
| 578 |
+
"eval_steps_per_second": 1.013,
|
| 579 |
+
"step": 67
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 16.0,
|
| 583 |
+
"eval_accuracy": 0.8315018315018315,
|
| 584 |
+
"eval_confusion_matrix": [
|
| 585 |
+
[
|
| 586 |
+
60,
|
| 587 |
+
7,
|
| 588 |
+
1,
|
| 589 |
+
4
|
| 590 |
+
],
|
| 591 |
+
[
|
| 592 |
+
10,
|
| 593 |
+
39,
|
| 594 |
+
11,
|
| 595 |
+
0
|
| 596 |
+
],
|
| 597 |
+
[
|
| 598 |
+
0,
|
| 599 |
+
11,
|
| 600 |
+
65,
|
| 601 |
+
0
|
| 602 |
+
],
|
| 603 |
+
[
|
| 604 |
+
1,
|
| 605 |
+
1,
|
| 606 |
+
0,
|
| 607 |
+
63
|
| 608 |
+
]
|
| 609 |
+
],
|
| 610 |
+
"eval_f1": 0.8304073820984628,
|
| 611 |
+
"eval_loss": 0.7295921444892883,
|
| 612 |
+
"eval_precision": 0.8295426562258641,
|
| 613 |
+
"eval_recall": 0.8315018315018315,
|
| 614 |
+
"eval_runtime": 2.8758,
|
| 615 |
+
"eval_samples_per_second": 94.932,
|
| 616 |
+
"eval_steps_per_second": 1.043,
|
| 617 |
+
"step": 72
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 16.88888888888889,
|
| 621 |
+
"eval_accuracy": 0.8644688644688645,
|
| 622 |
+
"eval_confusion_matrix": [
|
| 623 |
+
[
|
| 624 |
+
62,
|
| 625 |
+
5,
|
| 626 |
+
1,
|
| 627 |
+
4
|
| 628 |
+
],
|
| 629 |
+
[
|
| 630 |
+
9,
|
| 631 |
+
37,
|
| 632 |
+
14,
|
| 633 |
+
0
|
| 634 |
+
],
|
| 635 |
+
[
|
| 636 |
+
0,
|
| 637 |
+
3,
|
| 638 |
+
73,
|
| 639 |
+
0
|
| 640 |
+
],
|
| 641 |
+
[
|
| 642 |
+
1,
|
| 643 |
+
0,
|
| 644 |
+
0,
|
| 645 |
+
64
|
| 646 |
+
]
|
| 647 |
+
],
|
| 648 |
+
"eval_f1": 0.8589767100678526,
|
| 649 |
+
"eval_loss": 0.7055637240409851,
|
| 650 |
+
"eval_precision": 0.8628397746044805,
|
| 651 |
+
"eval_recall": 0.8644688644688645,
|
| 652 |
+
"eval_runtime": 3.074,
|
| 653 |
+
"eval_samples_per_second": 88.809,
|
| 654 |
+
"eval_steps_per_second": 0.976,
|
| 655 |
+
"step": 76
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 18.0,
|
| 659 |
+
"eval_accuracy": 0.8644688644688645,
|
| 660 |
+
"eval_confusion_matrix": [
|
| 661 |
+
[
|
| 662 |
+
65,
|
| 663 |
+
2,
|
| 664 |
+
1,
|
| 665 |
+
4
|
| 666 |
+
],
|
| 667 |
+
[
|
| 668 |
+
13,
|
| 669 |
+
42,
|
| 670 |
+
5,
|
| 671 |
+
0
|
| 672 |
+
],
|
| 673 |
+
[
|
| 674 |
+
0,
|
| 675 |
+
12,
|
| 676 |
+
64,
|
| 677 |
+
0
|
| 678 |
+
],
|
| 679 |
+
[
|
| 680 |
+
0,
|
| 681 |
+
0,
|
| 682 |
+
0,
|
| 683 |
+
65
|
| 684 |
+
]
|
| 685 |
+
],
|
| 686 |
+
"eval_f1": 0.8627769756077204,
|
| 687 |
+
"eval_loss": 0.7563945651054382,
|
| 688 |
+
"eval_precision": 0.8634344261673453,
|
| 689 |
+
"eval_recall": 0.8644688644688645,
|
| 690 |
+
"eval_runtime": 3.0072,
|
| 691 |
+
"eval_samples_per_second": 90.783,
|
| 692 |
+
"eval_steps_per_second": 0.998,
|
| 693 |
+
"step": 81
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 18.88888888888889,
|
| 697 |
+
"eval_accuracy": 0.8424908424908425,
|
| 698 |
+
"eval_confusion_matrix": [
|
| 699 |
+
[
|
| 700 |
+
64,
|
| 701 |
+
4,
|
| 702 |
+
0,
|
| 703 |
+
4
|
| 704 |
+
],
|
| 705 |
+
[
|
| 706 |
+
11,
|
| 707 |
+
41,
|
| 708 |
+
8,
|
| 709 |
+
0
|
| 710 |
+
],
|
| 711 |
+
[
|
| 712 |
+
0,
|
| 713 |
+
10,
|
| 714 |
+
66,
|
| 715 |
+
0
|
| 716 |
+
],
|
| 717 |
+
[
|
| 718 |
+
6,
|
| 719 |
+
0,
|
| 720 |
+
0,
|
| 721 |
+
59
|
| 722 |
+
]
|
| 723 |
+
],
|
| 724 |
+
"eval_f1": 0.8418306879608031,
|
| 725 |
+
"eval_loss": 0.7825365662574768,
|
| 726 |
+
"eval_precision": 0.8434907006335578,
|
| 727 |
+
"eval_recall": 0.8424908424908425,
|
| 728 |
+
"eval_runtime": 3.0343,
|
| 729 |
+
"eval_samples_per_second": 89.972,
|
| 730 |
+
"eval_steps_per_second": 0.989,
|
| 731 |
+
"step": 85
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 20.0,
|
| 735 |
+
"eval_accuracy": 0.8058608058608059,
|
| 736 |
+
"eval_confusion_matrix": [
|
| 737 |
+
[
|
| 738 |
+
40,
|
| 739 |
+
28,
|
| 740 |
+
1,
|
| 741 |
+
3
|
| 742 |
+
],
|
| 743 |
+
[
|
| 744 |
+
2,
|
| 745 |
+
50,
|
| 746 |
+
8,
|
| 747 |
+
0
|
| 748 |
+
],
|
| 749 |
+
[
|
| 750 |
+
0,
|
| 751 |
+
11,
|
| 752 |
+
65,
|
| 753 |
+
0
|
| 754 |
+
],
|
| 755 |
+
[
|
| 756 |
+
0,
|
| 757 |
+
0,
|
| 758 |
+
0,
|
| 759 |
+
65
|
| 760 |
+
]
|
| 761 |
+
],
|
| 762 |
+
"eval_f1": 0.8065760931078588,
|
| 763 |
+
"eval_loss": 0.8426868915557861,
|
| 764 |
+
"eval_precision": 0.8467707085637404,
|
| 765 |
+
"eval_recall": 0.8058608058608059,
|
| 766 |
+
"eval_runtime": 3.0783,
|
| 767 |
+
"eval_samples_per_second": 88.686,
|
| 768 |
+
"eval_steps_per_second": 0.975,
|
| 769 |
+
"step": 90
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 20.88888888888889,
|
| 773 |
+
"eval_accuracy": 0.8498168498168498,
|
| 774 |
+
"eval_confusion_matrix": [
|
| 775 |
+
[
|
| 776 |
+
63,
|
| 777 |
+
4,
|
| 778 |
+
0,
|
| 779 |
+
5
|
| 780 |
+
],
|
| 781 |
+
[
|
| 782 |
+
13,
|
| 783 |
+
40,
|
| 784 |
+
7,
|
| 785 |
+
0
|
| 786 |
+
],
|
| 787 |
+
[
|
| 788 |
+
0,
|
| 789 |
+
12,
|
| 790 |
+
64,
|
| 791 |
+
0
|
| 792 |
+
],
|
| 793 |
+
[
|
| 794 |
+
0,
|
| 795 |
+
0,
|
| 796 |
+
0,
|
| 797 |
+
65
|
| 798 |
+
]
|
| 799 |
+
],
|
| 800 |
+
"eval_f1": 0.8477878057985963,
|
| 801 |
+
"eval_loss": 0.7440442442893982,
|
| 802 |
+
"eval_precision": 0.8476393351433065,
|
| 803 |
+
"eval_recall": 0.8498168498168498,
|
| 804 |
+
"eval_runtime": 2.9875,
|
| 805 |
+
"eval_samples_per_second": 91.38,
|
| 806 |
+
"eval_steps_per_second": 1.004,
|
| 807 |
+
"step": 94
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 22.0,
|
| 811 |
+
"eval_accuracy": 0.8608058608058609,
|
| 812 |
+
"eval_confusion_matrix": [
|
| 813 |
+
[
|
| 814 |
+
66,
|
| 815 |
+
1,
|
| 816 |
+
1,
|
| 817 |
+
4
|
| 818 |
+
],
|
| 819 |
+
[
|
| 820 |
+
13,
|
| 821 |
+
36,
|
| 822 |
+
11,
|
| 823 |
+
0
|
| 824 |
+
],
|
| 825 |
+
[
|
| 826 |
+
0,
|
| 827 |
+
5,
|
| 828 |
+
71,
|
| 829 |
+
0
|
| 830 |
+
],
|
| 831 |
+
[
|
| 832 |
+
3,
|
| 833 |
+
0,
|
| 834 |
+
0,
|
| 835 |
+
62
|
| 836 |
+
]
|
| 837 |
+
],
|
| 838 |
+
"eval_f1": 0.855194718990792,
|
| 839 |
+
"eval_loss": 0.7338178753852844,
|
| 840 |
+
"eval_precision": 0.8624631692093176,
|
| 841 |
+
"eval_recall": 0.8608058608058609,
|
| 842 |
+
"eval_runtime": 2.98,
|
| 843 |
+
"eval_samples_per_second": 91.611,
|
| 844 |
+
"eval_steps_per_second": 1.007,
|
| 845 |
+
"step": 99
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 22.88888888888889,
|
| 849 |
+
"eval_accuracy": 0.8498168498168498,
|
| 850 |
+
"eval_confusion_matrix": [
|
| 851 |
+
[
|
| 852 |
+
58,
|
| 853 |
+
10,
|
| 854 |
+
0,
|
| 855 |
+
4
|
| 856 |
+
],
|
| 857 |
+
[
|
| 858 |
+
10,
|
| 859 |
+
43,
|
| 860 |
+
7,
|
| 861 |
+
0
|
| 862 |
+
],
|
| 863 |
+
[
|
| 864 |
+
0,
|
| 865 |
+
7,
|
| 866 |
+
69,
|
| 867 |
+
0
|
| 868 |
+
],
|
| 869 |
+
[
|
| 870 |
+
2,
|
| 871 |
+
1,
|
| 872 |
+
0,
|
| 873 |
+
62
|
| 874 |
+
]
|
| 875 |
+
],
|
| 876 |
+
"eval_f1": 0.8497737987724402,
|
| 877 |
+
"eval_loss": 0.7231407761573792,
|
| 878 |
+
"eval_precision": 0.8498638493954653,
|
| 879 |
+
"eval_recall": 0.8498168498168498,
|
| 880 |
+
"eval_runtime": 2.9272,
|
| 881 |
+
"eval_samples_per_second": 93.262,
|
| 882 |
+
"eval_steps_per_second": 1.025,
|
| 883 |
+
"step": 103
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 24.0,
|
| 887 |
+
"eval_accuracy": 0.8424908424908425,
|
| 888 |
+
"eval_confusion_matrix": [
|
| 889 |
+
[
|
| 890 |
+
67,
|
| 891 |
+
1,
|
| 892 |
+
0,
|
| 893 |
+
4
|
| 894 |
+
],
|
| 895 |
+
[
|
| 896 |
+
12,
|
| 897 |
+
44,
|
| 898 |
+
4,
|
| 899 |
+
0
|
| 900 |
+
],
|
| 901 |
+
[
|
| 902 |
+
0,
|
| 903 |
+
18,
|
| 904 |
+
58,
|
| 905 |
+
0
|
| 906 |
+
],
|
| 907 |
+
[
|
| 908 |
+
2,
|
| 909 |
+
2,
|
| 910 |
+
0,
|
| 911 |
+
61
|
| 912 |
+
]
|
| 913 |
+
],
|
| 914 |
+
"eval_f1": 0.8431607380967995,
|
| 915 |
+
"eval_loss": 0.752363920211792,
|
| 916 |
+
"eval_precision": 0.8507974885146101,
|
| 917 |
+
"eval_recall": 0.8424908424908425,
|
| 918 |
+
"eval_runtime": 3.0257,
|
| 919 |
+
"eval_samples_per_second": 90.228,
|
| 920 |
+
"eval_steps_per_second": 0.992,
|
| 921 |
+
"step": 108
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 24.88888888888889,
|
| 925 |
+
"eval_accuracy": 0.8498168498168498,
|
| 926 |
+
"eval_confusion_matrix": [
|
| 927 |
+
[
|
| 928 |
+
66,
|
| 929 |
+
0,
|
| 930 |
+
1,
|
| 931 |
+
5
|
| 932 |
+
],
|
| 933 |
+
[
|
| 934 |
+
15,
|
| 935 |
+
28,
|
| 936 |
+
17,
|
| 937 |
+
0
|
| 938 |
+
],
|
| 939 |
+
[
|
| 940 |
+
0,
|
| 941 |
+
2,
|
| 942 |
+
74,
|
| 943 |
+
0
|
| 944 |
+
],
|
| 945 |
+
[
|
| 946 |
+
1,
|
| 947 |
+
0,
|
| 948 |
+
0,
|
| 949 |
+
64
|
| 950 |
+
]
|
| 951 |
+
],
|
| 952 |
+
"eval_f1": 0.8354928653436116,
|
| 953 |
+
"eval_loss": 0.7849779725074768,
|
| 954 |
+
"eval_precision": 0.8621667009790022,
|
| 955 |
+
"eval_recall": 0.8498168498168498,
|
| 956 |
+
"eval_runtime": 2.8819,
|
| 957 |
+
"eval_samples_per_second": 94.73,
|
| 958 |
+
"eval_steps_per_second": 1.041,
|
| 959 |
+
"step": 112
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 26.0,
|
| 963 |
+
"eval_accuracy": 0.8131868131868132,
|
| 964 |
+
"eval_confusion_matrix": [
|
| 965 |
+
[
|
| 966 |
+
62,
|
| 967 |
+
6,
|
| 968 |
+
0,
|
| 969 |
+
4
|
| 970 |
+
],
|
| 971 |
+
[
|
| 972 |
+
11,
|
| 973 |
+
44,
|
| 974 |
+
5,
|
| 975 |
+
0
|
| 976 |
+
],
|
| 977 |
+
[
|
| 978 |
+
0,
|
| 979 |
+
24,
|
| 980 |
+
52,
|
| 981 |
+
0
|
| 982 |
+
],
|
| 983 |
+
[
|
| 984 |
+
1,
|
| 985 |
+
0,
|
| 986 |
+
0,
|
| 987 |
+
64
|
| 988 |
+
]
|
| 989 |
+
],
|
| 990 |
+
"eval_f1": 0.8151598256855266,
|
| 991 |
+
"eval_loss": 0.7896661162376404,
|
| 992 |
+
"eval_precision": 0.8297062414709473,
|
| 993 |
+
"eval_recall": 0.8131868131868132,
|
| 994 |
+
"eval_runtime": 2.8953,
|
| 995 |
+
"eval_samples_per_second": 94.291,
|
| 996 |
+
"eval_steps_per_second": 1.036,
|
| 997 |
+
"step": 117
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 26.88888888888889,
|
| 1001 |
+
"eval_accuracy": 0.8791208791208791,
|
| 1002 |
+
"eval_confusion_matrix": [
|
| 1003 |
+
[
|
| 1004 |
+
66,
|
| 1005 |
+
1,
|
| 1006 |
+
1,
|
| 1007 |
+
4
|
| 1008 |
+
],
|
| 1009 |
+
[
|
| 1010 |
+
13,
|
| 1011 |
+
40,
|
| 1012 |
+
7,
|
| 1013 |
+
0
|
| 1014 |
+
],
|
| 1015 |
+
[
|
| 1016 |
+
1,
|
| 1017 |
+
6,
|
| 1018 |
+
69,
|
| 1019 |
+
0
|
| 1020 |
+
],
|
| 1021 |
+
[
|
| 1022 |
+
0,
|
| 1023 |
+
0,
|
| 1024 |
+
0,
|
| 1025 |
+
65
|
| 1026 |
+
]
|
| 1027 |
+
],
|
| 1028 |
+
"eval_f1": 0.8754390108936493,
|
| 1029 |
+
"eval_loss": 0.7321063876152039,
|
| 1030 |
+
"eval_precision": 0.878386849630407,
|
| 1031 |
+
"eval_recall": 0.8791208791208791,
|
| 1032 |
+
"eval_runtime": 2.8941,
|
| 1033 |
+
"eval_samples_per_second": 94.329,
|
| 1034 |
+
"eval_steps_per_second": 1.037,
|
| 1035 |
+
"step": 121
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 28.0,
|
| 1039 |
+
"eval_accuracy": 0.8608058608058609,
|
| 1040 |
+
"eval_confusion_matrix": [
|
| 1041 |
+
[
|
| 1042 |
+
66,
|
| 1043 |
+
2,
|
| 1044 |
+
0,
|
| 1045 |
+
4
|
| 1046 |
+
],
|
| 1047 |
+
[
|
| 1048 |
+
15,
|
| 1049 |
+
40,
|
| 1050 |
+
5,
|
| 1051 |
+
0
|
| 1052 |
+
],
|
| 1053 |
+
[
|
| 1054 |
+
1,
|
| 1055 |
+
11,
|
| 1056 |
+
64,
|
| 1057 |
+
0
|
| 1058 |
+
],
|
| 1059 |
+
[
|
| 1060 |
+
0,
|
| 1061 |
+
0,
|
| 1062 |
+
0,
|
| 1063 |
+
65
|
| 1064 |
+
]
|
| 1065 |
+
],
|
| 1066 |
+
"eval_f1": 0.8583938406059761,
|
| 1067 |
+
"eval_loss": 0.7691925168037415,
|
| 1068 |
+
"eval_precision": 0.8606552236676889,
|
| 1069 |
+
"eval_recall": 0.8608058608058609,
|
| 1070 |
+
"eval_runtime": 2.9043,
|
| 1071 |
+
"eval_samples_per_second": 93.997,
|
| 1072 |
+
"eval_steps_per_second": 1.033,
|
| 1073 |
+
"step": 126
|
| 1074 |
+
}
|
| 1075 |
+
],
|
| 1076 |
+
"logging_steps": 500,
|
| 1077 |
+
"max_steps": 180,
|
| 1078 |
+
"num_input_tokens_seen": 0,
|
| 1079 |
+
"num_train_epochs": 45,
|
| 1080 |
+
"save_steps": 500,
|
| 1081 |
+
"stateful_callbacks": {
|
| 1082 |
+
"TrainerControl": {
|
| 1083 |
+
"args": {
|
| 1084 |
+
"should_epoch_stop": false,
|
| 1085 |
+
"should_evaluate": false,
|
| 1086 |
+
"should_log": false,
|
| 1087 |
+
"should_save": true,
|
| 1088 |
+
"should_training_stop": false
|
| 1089 |
+
},
|
| 1090 |
+
"attributes": {}
|
| 1091 |
+
}
|
| 1092 |
+
},
|
| 1093 |
+
"total_flos": 6.94093946112e+16,
|
| 1094 |
+
"train_batch_size": 128,
|
| 1095 |
+
"trial_name": null,
|
| 1096 |
+
"trial_params": null
|
| 1097 |
+
}
|
checkpoint-126/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:09e217cad0464b642c54f8b7691100b62a93dceb32a1869f55e3f0eb3a54a79e
|
| 3 |
+
size 5240
|