Joeran Bosma
commited on
Commit
•
efe9092
1
Parent(s):
8a647a1
Initial release
Browse files- .gitattributes +1 -0
- README.md +104 -3
- all_results.json +15 -0
- config.json +28 -0
- eval_results.json +10 -0
- pytorch_model.bin +3 -0
- sentencepiece.bpe.model +3 -0
- special_tokens_map.json +15 -0
- tokenizer.json +3 -0
- tokenizer_config.json +19 -0
- train_results.json +8 -0
- trainer_state.json +2935 -0
- training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
README.md
CHANGED
|
@@ -1,3 +1,104 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: cc-by-nc-sa-4.0
|
| 3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-nc-sa-4.0
|
| 3 |
+
---
|
| 4 |
+
|
| 5 |
+
# DRAGON RoBERTa base mixed-domain
|
| 6 |
+
|
| 7 |
+
Pretrained model on Dutch clinical reports using a masked language modeling (MLM) objective. It was introduced in [this](#pending) paper. The model was first pretrained using general domain data, as specified [here](https://huggingface.co/xlm-roberta-base). The pretrained model was taken from HuggingFace: [`xlm-roberta-base`](https://huggingface.co/xlm-roberta-base). Subsequently, the model was pretrained using domain-specific data (i.e., clinical reports). The tokenizer of [`xlm-roberta-base`](https://huggingface.co/xlm-roberta-base) was used.
|
| 8 |
+
|
| 9 |
+
## Model description
|
| 10 |
+
RoBERTa is a transformers model that was pretrained on a large corpus of Dutch clinical reports in a self-supervised fashion. This means it was pretrained on the raw texts only, with no humans labeling them in any way with an automatic process to generate inputs and labels from those texts.
|
| 11 |
+
|
| 12 |
+
This way, the model learns an inner representation of the Dutch medical language that can then be used to extract features useful for downstream tasks: if you have a dataset of labeled reports, for instance, you can train a standard classifier using the features produced by the BERT model as inputs.
|
| 13 |
+
|
| 14 |
+
## Model variations
|
| 15 |
+
Multiple architectures were pretrained for the DRAGON challenge.
|
| 16 |
+
|
| 17 |
+
| Model | #params | Language |
|
| 18 |
+
|------------------------|--------------------------------|-------|
|
| 19 |
+
| [`joeranbosma/dragon-bert-base-mixed-domain`](https://huggingface.co/joeranbosma/dragon-bert-base-mixed-domain) | 109M | Dutch → Dutch |
|
| 20 |
+
| [`joeranbosma/dragon-roberta-base-mixed-domain`](https://huggingface.co/joeranbosma/dragon-roberta-base-mixed-domain) | 278M | Multiple → Dutch |
|
| 21 |
+
| [`joeranbosma/dragon-roberta-large-mixed-domain`](https://huggingface.co/joeranbosma/dragon-roberta-large-mixed-domain) | 560M | Multiple → Dutch |
|
| 22 |
+
| [`joeranbosma/dragon-longformer-base-mixed-domain`](https://huggingface.co/joeranbosma/dragon-longformer-base-mixed-domain) | 149M | English → Dutch |
|
| 23 |
+
| [`joeranbosma/dragon-longformer-large-mixed-domain`](https://huggingface.co/joeranbosma/dragon-longformer-large-mixed-domain) | 435M | English → Dutch |
|
| 24 |
+
| [`joeranbosma/dragon-bert-base-domain-specific`](https://huggingface.co/joeranbosma/dragon-bert-base-domain-specific) | 109M | Dutch |
|
| 25 |
+
| [`joeranbosma/dragon-roberta-base-domain-specific`](https://huggingface.co/joeranbosma/dragon-roberta-base-domain-specific) | 278M | Dutch |
|
| 26 |
+
| [`joeranbosma/dragon-roberta-large-domain-specific`](https://huggingface.co/joeranbosma/dragon-roberta-large-domain-specific) | 560M | Dutch |
|
| 27 |
+
| [`joeranbosma/dragon-longformer-base-domain-specific`](https://huggingface.co/joeranbosma/dragon-longformer-base-domain-specific) | 149M | Dutch |
|
| 28 |
+
| [`joeranbosma/dragon-longformer-large-domain-specific`](https://huggingface.co/joeranbosma/dragon-longformer-large-domain-specific) | 435M | Dutch |
|
| 29 |
+
|
| 30 |
+
|
| 31 |
+
## Intended uses & limitations
|
| 32 |
+
You can use the raw model for masked language modeling, but it's mostly intended to be fine-tuned on a downstream task.
|
| 33 |
+
|
| 34 |
+
Note that this model is primarily aimed at being fine-tuned on tasks that use the whole text (e.g., a clinical report) to make decisions, such as sequence classification, token classification or question answering. For tasks such as text generation you should look at model like GPT2.
|
| 35 |
+
|
| 36 |
+
## How to use
|
| 37 |
+
You can use this model directly with a pipeline for masked language modeling:
|
| 38 |
+
|
| 39 |
+
```python
|
| 40 |
+
from transformers import pipeline
|
| 41 |
+
unmasker = pipeline("fill-mask", model="joeranbosma/dragon-roberta-base-mixed-domain")
|
| 42 |
+
unmasker("Dit onderzoek geen aanwijzingen voor significant carcinoom. PIRADS <mask>.")
|
| 43 |
+
```
|
| 44 |
+
|
| 45 |
+
Here is how to use this model to get the features of a given text in PyTorch:
|
| 46 |
+
|
| 47 |
+
```python
|
| 48 |
+
from transformers import AutoTokenizer, AutoModel
|
| 49 |
+
tokenizer = AutoTokenizer.from_pretrained("joeranbosma/dragon-roberta-base-mixed-domain")
|
| 50 |
+
model = AutoModel.from_pretrained("joeranbosma/dragon-roberta-base-mixed-domain")
|
| 51 |
+
text = "Replace me by any text you'd like."
|
| 52 |
+
encoded_input = tokenizer(text, return_tensors="pt")
|
| 53 |
+
output = model(**encoded_input)
|
| 54 |
+
```
|
| 55 |
+
|
| 56 |
+
## Limitations and bias
|
| 57 |
+
Even if the training data used for this model could be characterized as fairly neutral, this model can have biased predictions. This bias will also affect all fine-tuned versions of this model.
|
| 58 |
+
|
| 59 |
+
## Training data
|
| 60 |
+
For pretraining, 4,333,201 clinical reports (466,351 consecutive patients) were selected from Ziekenhuisgroep Twente from patients with a diagnostic or interventional visit between 13 July 2000 and 25 April 2023. 180,439 duplicate clinical reports (179,808 patients) were excluded, resulting in 4,152,762 included reports (463,692 patients). These reports were split into training (80%, 3,322,209 reports), validation (10%, 415,276 reports), and testing (10%, 415,277 reports). The testing reports were set aside for future analysis and are not used for pretraining.
|
| 61 |
+
|
| 62 |
+
## Training procedure
|
| 63 |
+
|
| 64 |
+
### Pretraining
|
| 65 |
+
The model was pretrained using masked language modeling (MLM): taking a sentence, the model randomly masks 15% of the words in the input then runs the entire masked sentence through the model and has to predict the masked words. This is different from traditional recurrent neural networks (RNNs) that usually see the words one after the other, or from autoregressive models like GPT which internally masks the future tokens. It allows the model to learn a bidirectional representation of the sentence.
|
| 66 |
+
|
| 67 |
+
The details of the masking procedure for each sentence are the following:
|
| 68 |
+
- 15% of the tokens are masked.
|
| 69 |
+
- In 80% of the cases, the masked tokens are replaced by `[MASK]`.
|
| 70 |
+
- In 10% of the cases, the masked tokens are replaced by a random token (different) from the one they replace.
|
| 71 |
+
- In the 10% remaining cases, the masked tokens are left as is.
|
| 72 |
+
|
| 73 |
+
The HuggingFace implementation was used for pretraining: [`run_mlm.py`](https://github.com/huggingface/transformers/blob/7c6ec195adbfcd22cb6baeee64dd3c24a4b80c74/examples/pytorch/language-modeling/run_mlm.py).
|
| 74 |
+
|
| 75 |
+
### Pretraining hyperparameters
|
| 76 |
+
|
| 77 |
+
The following hyperparameters were used during pretraining:
|
| 78 |
+
- `learning_rate`: 5e-05
|
| 79 |
+
- `train_batch_size`: 4
|
| 80 |
+
- `eval_batch_size`: 4
|
| 81 |
+
- `seed`: 42
|
| 82 |
+
- `gradient_accumulation_steps`: 4
|
| 83 |
+
- `total_train_batch_size`: 16
|
| 84 |
+
- `optimizer`: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 85 |
+
- `lr_scheduler_type`: linear
|
| 86 |
+
- `num_epochs`: 3.0
|
| 87 |
+
- `max_seq_length`: 512
|
| 88 |
+
|
| 89 |
+
### Framework versions
|
| 90 |
+
|
| 91 |
+
- Transformers 4.29.0.dev0
|
| 92 |
+
- Pytorch 2.0.0+cu117
|
| 93 |
+
- Datasets 2.11.0
|
| 94 |
+
- Tokenizers 0.13.3
|
| 95 |
+
|
| 96 |
+
## Evaluation results
|
| 97 |
+
|
| 98 |
+
Pending evaluation on the DRAGON benchmark.
|
| 99 |
+
|
| 100 |
+
### BibTeX entry and citation info
|
| 101 |
+
|
| 102 |
+
```bibtex
|
| 103 |
+
@article{PENDING}
|
| 104 |
+
```
|
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.9162802645968451,
|
| 4 |
+
"eval_loss": 0.33752548694610596,
|
| 5 |
+
"eval_runtime": 4485.3837,
|
| 6 |
+
"eval_samples": 140986,
|
| 7 |
+
"eval_samples_per_second": 31.432,
|
| 8 |
+
"eval_steps_per_second": 7.858,
|
| 9 |
+
"perplexity": 1.4014753272629237,
|
| 10 |
+
"train_loss": 0.4472081179925129,
|
| 11 |
+
"train_runtime": 493604.9481,
|
| 12 |
+
"train_samples": 1126219,
|
| 13 |
+
"train_samples_per_second": 6.845,
|
| 14 |
+
"train_steps_per_second": 0.428
|
| 15 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "xlm-roberta-base",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"XLMRobertaForMaskedLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": 0,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"hidden_act": "gelu",
|
| 11 |
+
"hidden_dropout_prob": 0.1,
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 3072,
|
| 15 |
+
"layer_norm_eps": 1e-05,
|
| 16 |
+
"max_position_embeddings": 514,
|
| 17 |
+
"model_type": "xlm-roberta",
|
| 18 |
+
"num_attention_heads": 12,
|
| 19 |
+
"num_hidden_layers": 12,
|
| 20 |
+
"output_past": true,
|
| 21 |
+
"pad_token_id": 1,
|
| 22 |
+
"position_embedding_type": "absolute",
|
| 23 |
+
"torch_dtype": "float32",
|
| 24 |
+
"transformers_version": "4.29.0.dev0",
|
| 25 |
+
"type_vocab_size": 1,
|
| 26 |
+
"use_cache": true,
|
| 27 |
+
"vocab_size": 250002
|
| 28 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.9162802645968451,
|
| 4 |
+
"eval_loss": 0.33752548694610596,
|
| 5 |
+
"eval_runtime": 4485.3837,
|
| 6 |
+
"eval_samples": 140986,
|
| 7 |
+
"eval_samples_per_second": 31.432,
|
| 8 |
+
"eval_steps_per_second": 7.858,
|
| 9 |
+
"perplexity": 1.4014753272629237
|
| 10 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:720933b287b58d466766bb378088b7ff49b363987e560983bda6c3f803d00a2b
|
| 3 |
+
size 1113254457
|
sentencepiece.bpe.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cfc8146abe2a0488e9e2a0c56de7952f7c11ab059eca145a0a727afce0db2865
|
| 3 |
+
size 5069051
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": {
|
| 6 |
+
"content": "<mask>",
|
| 7 |
+
"lstrip": true,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:62c24cdc13d4c9952d63718d6c9fa4c287974249e16b7ade6d5a85e7bbb75626
|
| 3 |
+
size 17082660
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"clean_up_tokenization_spaces": true,
|
| 4 |
+
"cls_token": "<s>",
|
| 5 |
+
"eos_token": "</s>",
|
| 6 |
+
"mask_token": {
|
| 7 |
+
"__type": "AddedToken",
|
| 8 |
+
"content": "<mask>",
|
| 9 |
+
"lstrip": true,
|
| 10 |
+
"normalized": true,
|
| 11 |
+
"rstrip": false,
|
| 12 |
+
"single_word": false
|
| 13 |
+
},
|
| 14 |
+
"model_max_length": 512,
|
| 15 |
+
"pad_token": "<pad>",
|
| 16 |
+
"sep_token": "</s>",
|
| 17 |
+
"tokenizer_class": "XLMRobertaTokenizer",
|
| 18 |
+
"unk_token": "<unk>"
|
| 19 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"train_loss": 0.4472081179925129,
|
| 4 |
+
"train_runtime": 493604.9481,
|
| 5 |
+
"train_samples": 1126219,
|
| 6 |
+
"train_samples_per_second": 6.845,
|
| 7 |
+
"train_steps_per_second": 0.428
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2935 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.3365662097930908,
|
| 3 |
+
"best_model_checkpoint": "/output/xlm-roberta-base-finetuned-full/checkpoint-205000",
|
| 4 |
+
"epoch": 2.9999680346646302,
|
| 5 |
+
"global_step": 211164,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 4.9881608607527804e-05,
|
| 13 |
+
"loss": 1.5253,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 4.97632172150556e-05,
|
| 19 |
+
"loss": 1.1464,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.02,
|
| 24 |
+
"learning_rate": 4.964482582258339e-05,
|
| 25 |
+
"loss": 1.0217,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.03,
|
| 30 |
+
"learning_rate": 4.9526434430111195e-05,
|
| 31 |
+
"loss": 0.9542,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.04,
|
| 36 |
+
"learning_rate": 4.9408043037638996e-05,
|
| 37 |
+
"loss": 0.9065,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.04,
|
| 42 |
+
"learning_rate": 4.928965164516679e-05,
|
| 43 |
+
"loss": 0.8708,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.05,
|
| 48 |
+
"learning_rate": 4.917126025269459e-05,
|
| 49 |
+
"loss": 0.8342,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.06,
|
| 54 |
+
"learning_rate": 4.905286886022239e-05,
|
| 55 |
+
"loss": 0.806,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.06,
|
| 60 |
+
"learning_rate": 4.893447746775019e-05,
|
| 61 |
+
"loss": 0.79,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.07,
|
| 66 |
+
"learning_rate": 4.881608607527798e-05,
|
| 67 |
+
"loss": 0.7697,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.07,
|
| 72 |
+
"eval_accuracy": 0.8496130075487093,
|
| 73 |
+
"eval_loss": 0.6976155042648315,
|
| 74 |
+
"eval_runtime": 4472.5845,
|
| 75 |
+
"eval_samples_per_second": 31.522,
|
| 76 |
+
"eval_steps_per_second": 7.881,
|
| 77 |
+
"step": 5000
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.08,
|
| 81 |
+
"learning_rate": 4.8697694682805785e-05,
|
| 82 |
+
"loss": 0.7583,
|
| 83 |
+
"step": 5500
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.09,
|
| 87 |
+
"learning_rate": 4.8579303290333586e-05,
|
| 88 |
+
"loss": 0.7434,
|
| 89 |
+
"step": 6000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.09,
|
| 93 |
+
"learning_rate": 4.846091189786138e-05,
|
| 94 |
+
"loss": 0.7366,
|
| 95 |
+
"step": 6500
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 0.1,
|
| 99 |
+
"learning_rate": 4.8342520505389175e-05,
|
| 100 |
+
"loss": 0.7189,
|
| 101 |
+
"step": 7000
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.11,
|
| 105 |
+
"learning_rate": 4.822412911291698e-05,
|
| 106 |
+
"loss": 0.7056,
|
| 107 |
+
"step": 7500
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.11,
|
| 111 |
+
"learning_rate": 4.810573772044478e-05,
|
| 112 |
+
"loss": 0.6993,
|
| 113 |
+
"step": 8000
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.12,
|
| 117 |
+
"learning_rate": 4.798734632797257e-05,
|
| 118 |
+
"loss": 0.682,
|
| 119 |
+
"step": 8500
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.13,
|
| 123 |
+
"learning_rate": 4.786895493550037e-05,
|
| 124 |
+
"loss": 0.6792,
|
| 125 |
+
"step": 9000
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.13,
|
| 129 |
+
"learning_rate": 4.775056354302817e-05,
|
| 130 |
+
"loss": 0.6652,
|
| 131 |
+
"step": 9500
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.14,
|
| 135 |
+
"learning_rate": 4.763217215055597e-05,
|
| 136 |
+
"loss": 0.6638,
|
| 137 |
+
"step": 10000
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.14,
|
| 141 |
+
"eval_accuracy": 0.866376552224583,
|
| 142 |
+
"eval_loss": 0.6067455410957336,
|
| 143 |
+
"eval_runtime": 4473.7903,
|
| 144 |
+
"eval_samples_per_second": 31.514,
|
| 145 |
+
"eval_steps_per_second": 7.879,
|
| 146 |
+
"step": 10000
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.15,
|
| 150 |
+
"learning_rate": 4.7513780758083765e-05,
|
| 151 |
+
"loss": 0.6513,
|
| 152 |
+
"step": 10500
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.16,
|
| 156 |
+
"learning_rate": 4.739538936561157e-05,
|
| 157 |
+
"loss": 0.6492,
|
| 158 |
+
"step": 11000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.16,
|
| 162 |
+
"learning_rate": 4.727699797313936e-05,
|
| 163 |
+
"loss": 0.6443,
|
| 164 |
+
"step": 11500
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.17,
|
| 168 |
+
"learning_rate": 4.715860658066716e-05,
|
| 169 |
+
"loss": 0.6369,
|
| 170 |
+
"step": 12000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.18,
|
| 174 |
+
"learning_rate": 4.704021518819496e-05,
|
| 175 |
+
"loss": 0.6319,
|
| 176 |
+
"step": 12500
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.18,
|
| 180 |
+
"learning_rate": 4.692182379572276e-05,
|
| 181 |
+
"loss": 0.6247,
|
| 182 |
+
"step": 13000
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.19,
|
| 186 |
+
"learning_rate": 4.680343240325056e-05,
|
| 187 |
+
"loss": 0.6208,
|
| 188 |
+
"step": 13500
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.2,
|
| 192 |
+
"learning_rate": 4.6685041010778355e-05,
|
| 193 |
+
"loss": 0.619,
|
| 194 |
+
"step": 14000
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.21,
|
| 198 |
+
"learning_rate": 4.656664961830615e-05,
|
| 199 |
+
"loss": 0.6193,
|
| 200 |
+
"step": 14500
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.21,
|
| 204 |
+
"learning_rate": 4.644825822583395e-05,
|
| 205 |
+
"loss": 0.6101,
|
| 206 |
+
"step": 15000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.21,
|
| 210 |
+
"eval_accuracy": 0.8746108749481479,
|
| 211 |
+
"eval_loss": 0.5597154498100281,
|
| 212 |
+
"eval_runtime": 4477.6442,
|
| 213 |
+
"eval_samples_per_second": 31.487,
|
| 214 |
+
"eval_steps_per_second": 7.872,
|
| 215 |
+
"step": 15000
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.22,
|
| 219 |
+
"learning_rate": 4.632986683336175e-05,
|
| 220 |
+
"loss": 0.6035,
|
| 221 |
+
"step": 15500
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.23,
|
| 225 |
+
"learning_rate": 4.621147544088955e-05,
|
| 226 |
+
"loss": 0.6028,
|
| 227 |
+
"step": 16000
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.23,
|
| 231 |
+
"learning_rate": 4.609308404841734e-05,
|
| 232 |
+
"loss": 0.5945,
|
| 233 |
+
"step": 16500
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.24,
|
| 237 |
+
"learning_rate": 4.5974692655945144e-05,
|
| 238 |
+
"loss": 0.5905,
|
| 239 |
+
"step": 17000
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.25,
|
| 243 |
+
"learning_rate": 4.5856301263472945e-05,
|
| 244 |
+
"loss": 0.5869,
|
| 245 |
+
"step": 17500
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.26,
|
| 249 |
+
"learning_rate": 4.573790987100074e-05,
|
| 250 |
+
"loss": 0.5849,
|
| 251 |
+
"step": 18000
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.26,
|
| 255 |
+
"learning_rate": 4.561951847852854e-05,
|
| 256 |
+
"loss": 0.579,
|
| 257 |
+
"step": 18500
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.27,
|
| 261 |
+
"learning_rate": 4.5501127086056336e-05,
|
| 262 |
+
"loss": 0.5755,
|
| 263 |
+
"step": 19000
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.28,
|
| 267 |
+
"learning_rate": 4.538273569358414e-05,
|
| 268 |
+
"loss": 0.5809,
|
| 269 |
+
"step": 19500
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.28,
|
| 273 |
+
"learning_rate": 4.526434430111193e-05,
|
| 274 |
+
"loss": 0.5702,
|
| 275 |
+
"step": 20000
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.28,
|
| 279 |
+
"eval_accuracy": 0.8805759285133897,
|
| 280 |
+
"eval_loss": 0.5277929306030273,
|
| 281 |
+
"eval_runtime": 4476.5281,
|
| 282 |
+
"eval_samples_per_second": 31.494,
|
| 283 |
+
"eval_steps_per_second": 7.874,
|
| 284 |
+
"step": 20000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.29,
|
| 288 |
+
"learning_rate": 4.5145952908639734e-05,
|
| 289 |
+
"loss": 0.5763,
|
| 290 |
+
"step": 20500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.3,
|
| 294 |
+
"learning_rate": 4.5027561516167535e-05,
|
| 295 |
+
"loss": 0.564,
|
| 296 |
+
"step": 21000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.31,
|
| 300 |
+
"learning_rate": 4.490917012369533e-05,
|
| 301 |
+
"loss": 0.566,
|
| 302 |
+
"step": 21500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.31,
|
| 306 |
+
"learning_rate": 4.4790778731223124e-05,
|
| 307 |
+
"loss": 0.5614,
|
| 308 |
+
"step": 22000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.32,
|
| 312 |
+
"learning_rate": 4.4672387338750926e-05,
|
| 313 |
+
"loss": 0.5602,
|
| 314 |
+
"step": 22500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.33,
|
| 318 |
+
"learning_rate": 4.455399594627873e-05,
|
| 319 |
+
"loss": 0.5566,
|
| 320 |
+
"step": 23000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.33,
|
| 324 |
+
"learning_rate": 4.443560455380652e-05,
|
| 325 |
+
"loss": 0.5521,
|
| 326 |
+
"step": 23500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.34,
|
| 330 |
+
"learning_rate": 4.431721316133432e-05,
|
| 331 |
+
"loss": 0.5505,
|
| 332 |
+
"step": 24000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.35,
|
| 336 |
+
"learning_rate": 4.419882176886212e-05,
|
| 337 |
+
"loss": 0.5505,
|
| 338 |
+
"step": 24500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.36,
|
| 342 |
+
"learning_rate": 4.408043037638992e-05,
|
| 343 |
+
"loss": 0.5433,
|
| 344 |
+
"step": 25000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.36,
|
| 348 |
+
"eval_accuracy": 0.8846786107773023,
|
| 349 |
+
"eval_loss": 0.5012275576591492,
|
| 350 |
+
"eval_runtime": 4482.4883,
|
| 351 |
+
"eval_samples_per_second": 31.453,
|
| 352 |
+
"eval_steps_per_second": 7.863,
|
| 353 |
+
"step": 25000
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.36,
|
| 357 |
+
"learning_rate": 4.3962038983917714e-05,
|
| 358 |
+
"loss": 0.5462,
|
| 359 |
+
"step": 25500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.37,
|
| 363 |
+
"learning_rate": 4.3843647591445516e-05,
|
| 364 |
+
"loss": 0.5402,
|
| 365 |
+
"step": 26000
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 0.38,
|
| 369 |
+
"learning_rate": 4.372525619897331e-05,
|
| 370 |
+
"loss": 0.543,
|
| 371 |
+
"step": 26500
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 0.38,
|
| 375 |
+
"learning_rate": 4.3606864806501105e-05,
|
| 376 |
+
"loss": 0.5411,
|
| 377 |
+
"step": 27000
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.39,
|
| 381 |
+
"learning_rate": 4.348847341402891e-05,
|
| 382 |
+
"loss": 0.5359,
|
| 383 |
+
"step": 27500
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 0.4,
|
| 387 |
+
"learning_rate": 4.337008202155671e-05,
|
| 388 |
+
"loss": 0.5375,
|
| 389 |
+
"step": 28000
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.4,
|
| 393 |
+
"learning_rate": 4.325169062908451e-05,
|
| 394 |
+
"loss": 0.5293,
|
| 395 |
+
"step": 28500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.41,
|
| 399 |
+
"learning_rate": 4.3133299236612304e-05,
|
| 400 |
+
"loss": 0.5255,
|
| 401 |
+
"step": 29000
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.42,
|
| 405 |
+
"learning_rate": 4.30149078441401e-05,
|
| 406 |
+
"loss": 0.5281,
|
| 407 |
+
"step": 29500
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.43,
|
| 411 |
+
"learning_rate": 4.28965164516679e-05,
|
| 412 |
+
"loss": 0.5308,
|
| 413 |
+
"step": 30000
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.43,
|
| 417 |
+
"eval_accuracy": 0.8875591084578017,
|
| 418 |
+
"eval_loss": 0.48577040433883667,
|
| 419 |
+
"eval_runtime": 4481.7631,
|
| 420 |
+
"eval_samples_per_second": 31.458,
|
| 421 |
+
"eval_steps_per_second": 7.865,
|
| 422 |
+
"step": 30000
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.43,
|
| 426 |
+
"learning_rate": 4.27781250591957e-05,
|
| 427 |
+
"loss": 0.5304,
|
| 428 |
+
"step": 30500
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.44,
|
| 432 |
+
"learning_rate": 4.2659733666723497e-05,
|
| 433 |
+
"loss": 0.5272,
|
| 434 |
+
"step": 31000
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.45,
|
| 438 |
+
"learning_rate": 4.254134227425129e-05,
|
| 439 |
+
"loss": 0.523,
|
| 440 |
+
"step": 31500
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.45,
|
| 444 |
+
"learning_rate": 4.242295088177909e-05,
|
| 445 |
+
"loss": 0.5197,
|
| 446 |
+
"step": 32000
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.46,
|
| 450 |
+
"learning_rate": 4.2304559489306894e-05,
|
| 451 |
+
"loss": 0.5212,
|
| 452 |
+
"step": 32500
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.47,
|
| 456 |
+
"learning_rate": 4.218616809683469e-05,
|
| 457 |
+
"loss": 0.5203,
|
| 458 |
+
"step": 33000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.48,
|
| 462 |
+
"learning_rate": 4.206777670436249e-05,
|
| 463 |
+
"loss": 0.5132,
|
| 464 |
+
"step": 33500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.48,
|
| 468 |
+
"learning_rate": 4.1949385311890285e-05,
|
| 469 |
+
"loss": 0.5154,
|
| 470 |
+
"step": 34000
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.49,
|
| 474 |
+
"learning_rate": 4.183099391941808e-05,
|
| 475 |
+
"loss": 0.5193,
|
| 476 |
+
"step": 34500
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.5,
|
| 480 |
+
"learning_rate": 4.171260252694588e-05,
|
| 481 |
+
"loss": 0.5118,
|
| 482 |
+
"step": 35000
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.5,
|
| 486 |
+
"eval_accuracy": 0.890265619775884,
|
| 487 |
+
"eval_loss": 0.4702432155609131,
|
| 488 |
+
"eval_runtime": 4481.2911,
|
| 489 |
+
"eval_samples_per_second": 31.461,
|
| 490 |
+
"eval_steps_per_second": 7.865,
|
| 491 |
+
"step": 35000
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.5,
|
| 495 |
+
"learning_rate": 4.159421113447368e-05,
|
| 496 |
+
"loss": 0.5079,
|
| 497 |
+
"step": 35500
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.51,
|
| 501 |
+
"learning_rate": 4.1475819742001484e-05,
|
| 502 |
+
"loss": 0.5104,
|
| 503 |
+
"step": 36000
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 0.52,
|
| 507 |
+
"learning_rate": 4.135742834952928e-05,
|
| 508 |
+
"loss": 0.5047,
|
| 509 |
+
"step": 36500
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.53,
|
| 513 |
+
"learning_rate": 4.1239036957057073e-05,
|
| 514 |
+
"loss": 0.5102,
|
| 515 |
+
"step": 37000
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.53,
|
| 519 |
+
"learning_rate": 4.1120645564584875e-05,
|
| 520 |
+
"loss": 0.5054,
|
| 521 |
+
"step": 37500
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.54,
|
| 525 |
+
"learning_rate": 4.1002254172112676e-05,
|
| 526 |
+
"loss": 0.5052,
|
| 527 |
+
"step": 38000
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.55,
|
| 531 |
+
"learning_rate": 4.088386277964047e-05,
|
| 532 |
+
"loss": 0.502,
|
| 533 |
+
"step": 38500
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.55,
|
| 537 |
+
"learning_rate": 4.076547138716827e-05,
|
| 538 |
+
"loss": 0.5037,
|
| 539 |
+
"step": 39000
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 0.56,
|
| 543 |
+
"learning_rate": 4.064707999469607e-05,
|
| 544 |
+
"loss": 0.4998,
|
| 545 |
+
"step": 39500
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 0.57,
|
| 549 |
+
"learning_rate": 4.052868860222386e-05,
|
| 550 |
+
"loss": 0.4971,
|
| 551 |
+
"step": 40000
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 0.57,
|
| 555 |
+
"eval_accuracy": 0.8922914063763604,
|
| 556 |
+
"eval_loss": 0.4587230086326599,
|
| 557 |
+
"eval_runtime": 4481.3983,
|
| 558 |
+
"eval_samples_per_second": 31.46,
|
| 559 |
+
"eval_steps_per_second": 7.865,
|
| 560 |
+
"step": 40000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.58,
|
| 564 |
+
"learning_rate": 4.041029720975166e-05,
|
| 565 |
+
"loss": 0.5026,
|
| 566 |
+
"step": 40500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.58,
|
| 570 |
+
"learning_rate": 4.0291905817279465e-05,
|
| 571 |
+
"loss": 0.496,
|
| 572 |
+
"step": 41000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.59,
|
| 576 |
+
"learning_rate": 4.017351442480726e-05,
|
| 577 |
+
"loss": 0.4965,
|
| 578 |
+
"step": 41500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.6,
|
| 582 |
+
"learning_rate": 4.0055123032335054e-05,
|
| 583 |
+
"loss": 0.496,
|
| 584 |
+
"step": 42000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.6,
|
| 588 |
+
"learning_rate": 3.9936731639862856e-05,
|
| 589 |
+
"loss": 0.4873,
|
| 590 |
+
"step": 42500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.61,
|
| 594 |
+
"learning_rate": 3.981834024739066e-05,
|
| 595 |
+
"loss": 0.4977,
|
| 596 |
+
"step": 43000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.62,
|
| 600 |
+
"learning_rate": 3.969994885491846e-05,
|
| 601 |
+
"loss": 0.4896,
|
| 602 |
+
"step": 43500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.63,
|
| 606 |
+
"learning_rate": 3.958155746244625e-05,
|
| 607 |
+
"loss": 0.4858,
|
| 608 |
+
"step": 44000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.63,
|
| 612 |
+
"learning_rate": 3.946316606997405e-05,
|
| 613 |
+
"loss": 0.4877,
|
| 614 |
+
"step": 44500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.64,
|
| 618 |
+
"learning_rate": 3.934477467750185e-05,
|
| 619 |
+
"loss": 0.4857,
|
| 620 |
+
"step": 45000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.64,
|
| 624 |
+
"eval_accuracy": 0.894234712380511,
|
| 625 |
+
"eval_loss": 0.4484991431236267,
|
| 626 |
+
"eval_runtime": 4480.0256,
|
| 627 |
+
"eval_samples_per_second": 31.47,
|
| 628 |
+
"eval_steps_per_second": 7.868,
|
| 629 |
+
"step": 45000
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 0.65,
|
| 633 |
+
"learning_rate": 3.922638328502965e-05,
|
| 634 |
+
"loss": 0.4888,
|
| 635 |
+
"step": 45500
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 0.65,
|
| 639 |
+
"learning_rate": 3.9107991892557446e-05,
|
| 640 |
+
"loss": 0.4827,
|
| 641 |
+
"step": 46000
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 0.66,
|
| 645 |
+
"learning_rate": 3.898960050008525e-05,
|
| 646 |
+
"loss": 0.4799,
|
| 647 |
+
"step": 46500
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.67,
|
| 651 |
+
"learning_rate": 3.887120910761304e-05,
|
| 652 |
+
"loss": 0.4815,
|
| 653 |
+
"step": 47000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.67,
|
| 657 |
+
"learning_rate": 3.8752817715140836e-05,
|
| 658 |
+
"loss": 0.4847,
|
| 659 |
+
"step": 47500
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 0.68,
|
| 663 |
+
"learning_rate": 3.863442632266864e-05,
|
| 664 |
+
"loss": 0.4845,
|
| 665 |
+
"step": 48000
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 0.69,
|
| 669 |
+
"learning_rate": 3.851603493019644e-05,
|
| 670 |
+
"loss": 0.4832,
|
| 671 |
+
"step": 48500
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 0.7,
|
| 675 |
+
"learning_rate": 3.8397643537724234e-05,
|
| 676 |
+
"loss": 0.4783,
|
| 677 |
+
"step": 49000
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 0.7,
|
| 681 |
+
"learning_rate": 3.827925214525203e-05,
|
| 682 |
+
"loss": 0.4785,
|
| 683 |
+
"step": 49500
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.71,
|
| 687 |
+
"learning_rate": 3.816086075277983e-05,
|
| 688 |
+
"loss": 0.4812,
|
| 689 |
+
"step": 50000
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.71,
|
| 693 |
+
"eval_accuracy": 0.8959396424676428,
|
| 694 |
+
"eval_loss": 0.43902069330215454,
|
| 695 |
+
"eval_runtime": 4484.5246,
|
| 696 |
+
"eval_samples_per_second": 31.438,
|
| 697 |
+
"eval_steps_per_second": 7.86,
|
| 698 |
+
"step": 50000
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.72,
|
| 702 |
+
"learning_rate": 3.804246936030763e-05,
|
| 703 |
+
"loss": 0.4753,
|
| 704 |
+
"step": 50500
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.72,
|
| 708 |
+
"learning_rate": 3.792407796783543e-05,
|
| 709 |
+
"loss": 0.48,
|
| 710 |
+
"step": 51000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.73,
|
| 714 |
+
"learning_rate": 3.780568657536323e-05,
|
| 715 |
+
"loss": 0.4777,
|
| 716 |
+
"step": 51500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.74,
|
| 720 |
+
"learning_rate": 3.768729518289102e-05,
|
| 721 |
+
"loss": 0.4748,
|
| 722 |
+
"step": 52000
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.75,
|
| 726 |
+
"learning_rate": 3.7568903790418824e-05,
|
| 727 |
+
"loss": 0.4728,
|
| 728 |
+
"step": 52500
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 0.75,
|
| 732 |
+
"learning_rate": 3.745051239794662e-05,
|
| 733 |
+
"loss": 0.4677,
|
| 734 |
+
"step": 53000
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.76,
|
| 738 |
+
"learning_rate": 3.733212100547442e-05,
|
| 739 |
+
"loss": 0.4731,
|
| 740 |
+
"step": 53500
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.77,
|
| 744 |
+
"learning_rate": 3.721372961300222e-05,
|
| 745 |
+
"loss": 0.464,
|
| 746 |
+
"step": 54000
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.77,
|
| 750 |
+
"learning_rate": 3.7095338220530016e-05,
|
| 751 |
+
"loss": 0.4708,
|
| 752 |
+
"step": 54500
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.78,
|
| 756 |
+
"learning_rate": 3.697694682805781e-05,
|
| 757 |
+
"loss": 0.4716,
|
| 758 |
+
"step": 55000
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.78,
|
| 762 |
+
"eval_accuracy": 0.8970586319380052,
|
| 763 |
+
"eval_loss": 0.43407291173934937,
|
| 764 |
+
"eval_runtime": 4489.7395,
|
| 765 |
+
"eval_samples_per_second": 31.402,
|
| 766 |
+
"eval_steps_per_second": 7.851,
|
| 767 |
+
"step": 55000
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.79,
|
| 771 |
+
"learning_rate": 3.685855543558561e-05,
|
| 772 |
+
"loss": 0.469,
|
| 773 |
+
"step": 55500
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.8,
|
| 777 |
+
"learning_rate": 3.6740164043113414e-05,
|
| 778 |
+
"loss": 0.4684,
|
| 779 |
+
"step": 56000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.8,
|
| 783 |
+
"learning_rate": 3.662177265064121e-05,
|
| 784 |
+
"loss": 0.4682,
|
| 785 |
+
"step": 56500
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 0.81,
|
| 789 |
+
"learning_rate": 3.6503381258169e-05,
|
| 790 |
+
"loss": 0.466,
|
| 791 |
+
"step": 57000
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 0.82,
|
| 795 |
+
"learning_rate": 3.6384989865696805e-05,
|
| 796 |
+
"loss": 0.4629,
|
| 797 |
+
"step": 57500
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 0.82,
|
| 801 |
+
"learning_rate": 3.6266598473224606e-05,
|
| 802 |
+
"loss": 0.4641,
|
| 803 |
+
"step": 58000
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 0.83,
|
| 807 |
+
"learning_rate": 3.614820708075241e-05,
|
| 808 |
+
"loss": 0.4602,
|
| 809 |
+
"step": 58500
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.84,
|
| 813 |
+
"learning_rate": 3.60298156882802e-05,
|
| 814 |
+
"loss": 0.463,
|
| 815 |
+
"step": 59000
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.85,
|
| 819 |
+
"learning_rate": 3.5911424295808e-05,
|
| 820 |
+
"loss": 0.4586,
|
| 821 |
+
"step": 59500
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.85,
|
| 825 |
+
"learning_rate": 3.57930329033358e-05,
|
| 826 |
+
"loss": 0.4601,
|
| 827 |
+
"step": 60000
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 0.85,
|
| 831 |
+
"eval_accuracy": 0.8986926934131002,
|
| 832 |
+
"eval_loss": 0.42562565207481384,
|
| 833 |
+
"eval_runtime": 4489.1177,
|
| 834 |
+
"eval_samples_per_second": 31.406,
|
| 835 |
+
"eval_steps_per_second": 7.852,
|
| 836 |
+
"step": 60000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.86,
|
| 840 |
+
"learning_rate": 3.567464151086359e-05,
|
| 841 |
+
"loss": 0.4612,
|
| 842 |
+
"step": 60500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.87,
|
| 846 |
+
"learning_rate": 3.5556250118391395e-05,
|
| 847 |
+
"loss": 0.4609,
|
| 848 |
+
"step": 61000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.87,
|
| 852 |
+
"learning_rate": 3.5437858725919196e-05,
|
| 853 |
+
"loss": 0.4552,
|
| 854 |
+
"step": 61500
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.88,
|
| 858 |
+
"learning_rate": 3.531946733344699e-05,
|
| 859 |
+
"loss": 0.4624,
|
| 860 |
+
"step": 62000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.89,
|
| 864 |
+
"learning_rate": 3.5201075940974785e-05,
|
| 865 |
+
"loss": 0.4572,
|
| 866 |
+
"step": 62500
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.9,
|
| 870 |
+
"learning_rate": 3.508268454850259e-05,
|
| 871 |
+
"loss": 0.4577,
|
| 872 |
+
"step": 63000
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.9,
|
| 876 |
+
"learning_rate": 3.496429315603039e-05,
|
| 877 |
+
"loss": 0.4527,
|
| 878 |
+
"step": 63500
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.91,
|
| 882 |
+
"learning_rate": 3.484590176355819e-05,
|
| 883 |
+
"loss": 0.4542,
|
| 884 |
+
"step": 64000
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.92,
|
| 888 |
+
"learning_rate": 3.472751037108598e-05,
|
| 889 |
+
"loss": 0.4549,
|
| 890 |
+
"step": 64500
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.92,
|
| 894 |
+
"learning_rate": 3.460911897861378e-05,
|
| 895 |
+
"loss": 0.4548,
|
| 896 |
+
"step": 65000
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.92,
|
| 900 |
+
"eval_accuracy": 0.9000358274024964,
|
| 901 |
+
"eval_loss": 0.418160617351532,
|
| 902 |
+
"eval_runtime": 4478.7532,
|
| 903 |
+
"eval_samples_per_second": 31.479,
|
| 904 |
+
"eval_steps_per_second": 7.87,
|
| 905 |
+
"step": 65000
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.93,
|
| 909 |
+
"learning_rate": 3.449072758614158e-05,
|
| 910 |
+
"loss": 0.4543,
|
| 911 |
+
"step": 65500
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.94,
|
| 915 |
+
"learning_rate": 3.4372336193669375e-05,
|
| 916 |
+
"loss": 0.4563,
|
| 917 |
+
"step": 66000
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 0.94,
|
| 921 |
+
"learning_rate": 3.425394480119718e-05,
|
| 922 |
+
"loss": 0.4534,
|
| 923 |
+
"step": 66500
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 0.95,
|
| 927 |
+
"learning_rate": 3.413555340872497e-05,
|
| 928 |
+
"loss": 0.4532,
|
| 929 |
+
"step": 67000
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.96,
|
| 933 |
+
"learning_rate": 3.401716201625277e-05,
|
| 934 |
+
"loss": 0.4512,
|
| 935 |
+
"step": 67500
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 0.97,
|
| 939 |
+
"learning_rate": 3.389877062378057e-05,
|
| 940 |
+
"loss": 0.4477,
|
| 941 |
+
"step": 68000
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.97,
|
| 945 |
+
"learning_rate": 3.378037923130837e-05,
|
| 946 |
+
"loss": 0.4487,
|
| 947 |
+
"step": 68500
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.98,
|
| 951 |
+
"learning_rate": 3.366198783883617e-05,
|
| 952 |
+
"loss": 0.4503,
|
| 953 |
+
"step": 69000
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 0.99,
|
| 957 |
+
"learning_rate": 3.3543596446363965e-05,
|
| 958 |
+
"loss": 0.445,
|
| 959 |
+
"step": 69500
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 0.99,
|
| 963 |
+
"learning_rate": 3.342520505389176e-05,
|
| 964 |
+
"loss": 0.4497,
|
| 965 |
+
"step": 70000
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 0.99,
|
| 969 |
+
"eval_accuracy": 0.9010131119019583,
|
| 970 |
+
"eval_loss": 0.4137929081916809,
|
| 971 |
+
"eval_runtime": 4478.1902,
|
| 972 |
+
"eval_samples_per_second": 31.483,
|
| 973 |
+
"eval_steps_per_second": 7.871,
|
| 974 |
+
"step": 70000
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 1.0,
|
| 978 |
+
"learning_rate": 3.330681366141956e-05,
|
| 979 |
+
"loss": 0.448,
|
| 980 |
+
"step": 70500
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 1.01,
|
| 984 |
+
"learning_rate": 3.318842226894736e-05,
|
| 985 |
+
"loss": 0.4412,
|
| 986 |
+
"step": 71000
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 1.02,
|
| 990 |
+
"learning_rate": 3.3070030876475164e-05,
|
| 991 |
+
"loss": 0.4441,
|
| 992 |
+
"step": 71500
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 1.02,
|
| 996 |
+
"learning_rate": 3.295163948400295e-05,
|
| 997 |
+
"loss": 0.4442,
|
| 998 |
+
"step": 72000
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.03,
|
| 1002 |
+
"learning_rate": 3.2833248091530754e-05,
|
| 1003 |
+
"loss": 0.4415,
|
| 1004 |
+
"step": 72500
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.04,
|
| 1008 |
+
"learning_rate": 3.2714856699058555e-05,
|
| 1009 |
+
"loss": 0.4457,
|
| 1010 |
+
"step": 73000
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.04,
|
| 1014 |
+
"learning_rate": 3.259646530658635e-05,
|
| 1015 |
+
"loss": 0.4438,
|
| 1016 |
+
"step": 73500
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 1.05,
|
| 1020 |
+
"learning_rate": 3.247807391411415e-05,
|
| 1021 |
+
"loss": 0.4453,
|
| 1022 |
+
"step": 74000
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 1.06,
|
| 1026 |
+
"learning_rate": 3.2359682521641946e-05,
|
| 1027 |
+
"loss": 0.4407,
|
| 1028 |
+
"step": 74500
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 1.07,
|
| 1032 |
+
"learning_rate": 3.224129112916975e-05,
|
| 1033 |
+
"loss": 0.4376,
|
| 1034 |
+
"step": 75000
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 1.07,
|
| 1038 |
+
"eval_accuracy": 0.9019829116607336,
|
| 1039 |
+
"eval_loss": 0.408815860748291,
|
| 1040 |
+
"eval_runtime": 4480.6277,
|
| 1041 |
+
"eval_samples_per_second": 31.466,
|
| 1042 |
+
"eval_steps_per_second": 7.867,
|
| 1043 |
+
"step": 75000
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 1.07,
|
| 1047 |
+
"learning_rate": 3.212289973669754e-05,
|
| 1048 |
+
"loss": 0.4371,
|
| 1049 |
+
"step": 75500
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 1.08,
|
| 1053 |
+
"learning_rate": 3.2004508344225344e-05,
|
| 1054 |
+
"loss": 0.4397,
|
| 1055 |
+
"step": 76000
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 1.09,
|
| 1059 |
+
"learning_rate": 3.1886116951753145e-05,
|
| 1060 |
+
"loss": 0.4312,
|
| 1061 |
+
"step": 76500
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.09,
|
| 1065 |
+
"learning_rate": 3.176772555928094e-05,
|
| 1066 |
+
"loss": 0.435,
|
| 1067 |
+
"step": 77000
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.1,
|
| 1071 |
+
"learning_rate": 3.1649334166808734e-05,
|
| 1072 |
+
"loss": 0.4398,
|
| 1073 |
+
"step": 77500
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.11,
|
| 1077 |
+
"learning_rate": 3.1530942774336536e-05,
|
| 1078 |
+
"loss": 0.4393,
|
| 1079 |
+
"step": 78000
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 1.12,
|
| 1083 |
+
"learning_rate": 3.141255138186434e-05,
|
| 1084 |
+
"loss": 0.4371,
|
| 1085 |
+
"step": 78500
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 1.12,
|
| 1089 |
+
"learning_rate": 3.129415998939213e-05,
|
| 1090 |
+
"loss": 0.4355,
|
| 1091 |
+
"step": 79000
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 1.13,
|
| 1095 |
+
"learning_rate": 3.117576859691993e-05,
|
| 1096 |
+
"loss": 0.4328,
|
| 1097 |
+
"step": 79500
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 1.14,
|
| 1101 |
+
"learning_rate": 3.105737720444773e-05,
|
| 1102 |
+
"loss": 0.4349,
|
| 1103 |
+
"step": 80000
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 1.14,
|
| 1107 |
+
"eval_accuracy": 0.9031310444806294,
|
| 1108 |
+
"eval_loss": 0.40232041478157043,
|
| 1109 |
+
"eval_runtime": 4482.6913,
|
| 1110 |
+
"eval_samples_per_second": 31.451,
|
| 1111 |
+
"eval_steps_per_second": 7.863,
|
| 1112 |
+
"step": 80000
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 1.14,
|
| 1116 |
+
"learning_rate": 3.093898581197553e-05,
|
| 1117 |
+
"loss": 0.4324,
|
| 1118 |
+
"step": 80500
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 1.15,
|
| 1122 |
+
"learning_rate": 3.0820594419503324e-05,
|
| 1123 |
+
"loss": 0.4326,
|
| 1124 |
+
"step": 81000
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 1.16,
|
| 1128 |
+
"learning_rate": 3.0702203027031126e-05,
|
| 1129 |
+
"loss": 0.4308,
|
| 1130 |
+
"step": 81500
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.16,
|
| 1134 |
+
"learning_rate": 3.058381163455892e-05,
|
| 1135 |
+
"loss": 0.4317,
|
| 1136 |
+
"step": 82000
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.17,
|
| 1140 |
+
"learning_rate": 3.0465420242086722e-05,
|
| 1141 |
+
"loss": 0.4322,
|
| 1142 |
+
"step": 82500
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 1.18,
|
| 1146 |
+
"learning_rate": 3.0347028849614517e-05,
|
| 1147 |
+
"loss": 0.43,
|
| 1148 |
+
"step": 83000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 1.19,
|
| 1152 |
+
"learning_rate": 3.0228637457142318e-05,
|
| 1153 |
+
"loss": 0.4321,
|
| 1154 |
+
"step": 83500
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 1.19,
|
| 1158 |
+
"learning_rate": 3.0110246064670116e-05,
|
| 1159 |
+
"loss": 0.4309,
|
| 1160 |
+
"step": 84000
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 1.2,
|
| 1164 |
+
"learning_rate": 2.9991854672197918e-05,
|
| 1165 |
+
"loss": 0.4274,
|
| 1166 |
+
"step": 84500
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 1.21,
|
| 1170 |
+
"learning_rate": 2.9873463279725712e-05,
|
| 1171 |
+
"loss": 0.4262,
|
| 1172 |
+
"step": 85000
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.21,
|
| 1176 |
+
"eval_accuracy": 0.9039536284020224,
|
| 1177 |
+
"eval_loss": 0.39793965220451355,
|
| 1178 |
+
"eval_runtime": 4483.3415,
|
| 1179 |
+
"eval_samples_per_second": 31.447,
|
| 1180 |
+
"eval_steps_per_second": 7.862,
|
| 1181 |
+
"step": 85000
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 1.21,
|
| 1185 |
+
"learning_rate": 2.975507188725351e-05,
|
| 1186 |
+
"loss": 0.4282,
|
| 1187 |
+
"step": 85500
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 1.22,
|
| 1191 |
+
"learning_rate": 2.9636680494781312e-05,
|
| 1192 |
+
"loss": 0.4283,
|
| 1193 |
+
"step": 86000
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.23,
|
| 1197 |
+
"learning_rate": 2.9518289102309103e-05,
|
| 1198 |
+
"loss": 0.4275,
|
| 1199 |
+
"step": 86500
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.24,
|
| 1203 |
+
"learning_rate": 2.9399897709836905e-05,
|
| 1204 |
+
"loss": 0.4251,
|
| 1205 |
+
"step": 87000
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"epoch": 1.24,
|
| 1209 |
+
"learning_rate": 2.9281506317364703e-05,
|
| 1210 |
+
"loss": 0.4272,
|
| 1211 |
+
"step": 87500
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 1.25,
|
| 1215 |
+
"learning_rate": 2.9163114924892504e-05,
|
| 1216 |
+
"loss": 0.428,
|
| 1217 |
+
"step": 88000
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 1.26,
|
| 1221 |
+
"learning_rate": 2.90447235324203e-05,
|
| 1222 |
+
"loss": 0.4281,
|
| 1223 |
+
"step": 88500
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 1.26,
|
| 1227 |
+
"learning_rate": 2.8926332139948097e-05,
|
| 1228 |
+
"loss": 0.4242,
|
| 1229 |
+
"step": 89000
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 1.27,
|
| 1233 |
+
"learning_rate": 2.8807940747475898e-05,
|
| 1234 |
+
"loss": 0.4233,
|
| 1235 |
+
"step": 89500
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.28,
|
| 1239 |
+
"learning_rate": 2.8689549355003696e-05,
|
| 1240 |
+
"loss": 0.4187,
|
| 1241 |
+
"step": 90000
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.28,
|
| 1245 |
+
"eval_accuracy": 0.905044954941931,
|
| 1246 |
+
"eval_loss": 0.3930475413799286,
|
| 1247 |
+
"eval_runtime": 4483.2316,
|
| 1248 |
+
"eval_samples_per_second": 31.447,
|
| 1249 |
+
"eval_steps_per_second": 7.862,
|
| 1250 |
+
"step": 90000
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 1.29,
|
| 1254 |
+
"learning_rate": 2.857115796253149e-05,
|
| 1255 |
+
"loss": 0.4248,
|
| 1256 |
+
"step": 90500
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.29,
|
| 1260 |
+
"learning_rate": 2.8452766570059293e-05,
|
| 1261 |
+
"loss": 0.4256,
|
| 1262 |
+
"step": 91000
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.3,
|
| 1266 |
+
"learning_rate": 2.833437517758709e-05,
|
| 1267 |
+
"loss": 0.4254,
|
| 1268 |
+
"step": 91500
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 1.31,
|
| 1272 |
+
"learning_rate": 2.8215983785114885e-05,
|
| 1273 |
+
"loss": 0.4213,
|
| 1274 |
+
"step": 92000
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 1.31,
|
| 1278 |
+
"learning_rate": 2.8097592392642687e-05,
|
| 1279 |
+
"loss": 0.4219,
|
| 1280 |
+
"step": 92500
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 1.32,
|
| 1284 |
+
"learning_rate": 2.7979201000170485e-05,
|
| 1285 |
+
"loss": 0.4231,
|
| 1286 |
+
"step": 93000
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 1.33,
|
| 1290 |
+
"learning_rate": 2.7860809607698286e-05,
|
| 1291 |
+
"loss": 0.4206,
|
| 1292 |
+
"step": 93500
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 1.34,
|
| 1296 |
+
"learning_rate": 2.7742418215226078e-05,
|
| 1297 |
+
"loss": 0.4203,
|
| 1298 |
+
"step": 94000
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.34,
|
| 1302 |
+
"learning_rate": 2.762402682275388e-05,
|
| 1303 |
+
"loss": 0.4174,
|
| 1304 |
+
"step": 94500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.35,
|
| 1308 |
+
"learning_rate": 2.7505635430281677e-05,
|
| 1309 |
+
"loss": 0.4183,
|
| 1310 |
+
"step": 95000
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 1.35,
|
| 1314 |
+
"eval_accuracy": 0.9058014412128497,
|
| 1315 |
+
"eval_loss": 0.3888256549835205,
|
| 1316 |
+
"eval_runtime": 4483.1222,
|
| 1317 |
+
"eval_samples_per_second": 31.448,
|
| 1318 |
+
"eval_steps_per_second": 7.862,
|
| 1319 |
+
"step": 95000
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.36,
|
| 1323 |
+
"learning_rate": 2.738724403780948e-05,
|
| 1324 |
+
"loss": 0.4226,
|
| 1325 |
+
"step": 95500
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.36,
|
| 1329 |
+
"learning_rate": 2.7268852645337273e-05,
|
| 1330 |
+
"loss": 0.4213,
|
| 1331 |
+
"step": 96000
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 1.37,
|
| 1335 |
+
"learning_rate": 2.715046125286507e-05,
|
| 1336 |
+
"loss": 0.4191,
|
| 1337 |
+
"step": 96500
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 1.38,
|
| 1341 |
+
"learning_rate": 2.7032069860392873e-05,
|
| 1342 |
+
"loss": 0.4156,
|
| 1343 |
+
"step": 97000
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 1.39,
|
| 1347 |
+
"learning_rate": 2.691367846792067e-05,
|
| 1348 |
+
"loss": 0.4179,
|
| 1349 |
+
"step": 97500
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 1.39,
|
| 1353 |
+
"learning_rate": 2.6795287075448466e-05,
|
| 1354 |
+
"loss": 0.4158,
|
| 1355 |
+
"step": 98000
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 1.4,
|
| 1359 |
+
"learning_rate": 2.6676895682976267e-05,
|
| 1360 |
+
"loss": 0.421,
|
| 1361 |
+
"step": 98500
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.41,
|
| 1365 |
+
"learning_rate": 2.6558504290504065e-05,
|
| 1366 |
+
"loss": 0.417,
|
| 1367 |
+
"step": 99000
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.41,
|
| 1371 |
+
"learning_rate": 2.644011289803186e-05,
|
| 1372 |
+
"loss": 0.415,
|
| 1373 |
+
"step": 99500
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 1.42,
|
| 1377 |
+
"learning_rate": 2.632172150555966e-05,
|
| 1378 |
+
"loss": 0.4126,
|
| 1379 |
+
"step": 100000
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 1.42,
|
| 1383 |
+
"eval_accuracy": 0.9066390531479637,
|
| 1384 |
+
"eval_loss": 0.3853781521320343,
|
| 1385 |
+
"eval_runtime": 4483.6399,
|
| 1386 |
+
"eval_samples_per_second": 31.445,
|
| 1387 |
+
"eval_steps_per_second": 7.861,
|
| 1388 |
+
"step": 100000
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.43,
|
| 1392 |
+
"learning_rate": 2.620333011308746e-05,
|
| 1393 |
+
"loss": 0.4131,
|
| 1394 |
+
"step": 100500
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 1.43,
|
| 1398 |
+
"learning_rate": 2.608493872061526e-05,
|
| 1399 |
+
"loss": 0.4168,
|
| 1400 |
+
"step": 101000
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 1.44,
|
| 1404 |
+
"learning_rate": 2.5966547328143055e-05,
|
| 1405 |
+
"loss": 0.4114,
|
| 1406 |
+
"step": 101500
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 1.45,
|
| 1410 |
+
"learning_rate": 2.5848155935670854e-05,
|
| 1411 |
+
"loss": 0.4164,
|
| 1412 |
+
"step": 102000
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 1.46,
|
| 1416 |
+
"learning_rate": 2.5729764543198655e-05,
|
| 1417 |
+
"loss": 0.4157,
|
| 1418 |
+
"step": 102500
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 1.46,
|
| 1422 |
+
"learning_rate": 2.5611373150726453e-05,
|
| 1423 |
+
"loss": 0.4155,
|
| 1424 |
+
"step": 103000
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.47,
|
| 1428 |
+
"learning_rate": 2.5492981758254248e-05,
|
| 1429 |
+
"loss": 0.4123,
|
| 1430 |
+
"step": 103500
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.48,
|
| 1434 |
+
"learning_rate": 2.5374590365782046e-05,
|
| 1435 |
+
"loss": 0.4124,
|
| 1436 |
+
"step": 104000
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 1.48,
|
| 1440 |
+
"learning_rate": 2.5256198973309847e-05,
|
| 1441 |
+
"loss": 0.4148,
|
| 1442 |
+
"step": 104500
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 1.49,
|
| 1446 |
+
"learning_rate": 2.5137807580837645e-05,
|
| 1447 |
+
"loss": 0.411,
|
| 1448 |
+
"step": 105000
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 1.49,
|
| 1452 |
+
"eval_accuracy": 0.907043654550547,
|
| 1453 |
+
"eval_loss": 0.38217270374298096,
|
| 1454 |
+
"eval_runtime": 4482.6373,
|
| 1455 |
+
"eval_samples_per_second": 31.452,
|
| 1456 |
+
"eval_steps_per_second": 7.863,
|
| 1457 |
+
"step": 105000
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 1.5,
|
| 1461 |
+
"learning_rate": 2.501941618836544e-05,
|
| 1462 |
+
"loss": 0.408,
|
| 1463 |
+
"step": 105500
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 1.51,
|
| 1467 |
+
"learning_rate": 2.490102479589324e-05,
|
| 1468 |
+
"loss": 0.4123,
|
| 1469 |
+
"step": 106000
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 1.51,
|
| 1473 |
+
"learning_rate": 2.4782633403421036e-05,
|
| 1474 |
+
"loss": 0.4125,
|
| 1475 |
+
"step": 106500
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 1.52,
|
| 1479 |
+
"learning_rate": 2.4664242010948838e-05,
|
| 1480 |
+
"loss": 0.4117,
|
| 1481 |
+
"step": 107000
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 1.53,
|
| 1485 |
+
"learning_rate": 2.4545850618476636e-05,
|
| 1486 |
+
"loss": 0.4083,
|
| 1487 |
+
"step": 107500
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.53,
|
| 1491 |
+
"learning_rate": 2.4427459226004434e-05,
|
| 1492 |
+
"loss": 0.4117,
|
| 1493 |
+
"step": 108000
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.54,
|
| 1497 |
+
"learning_rate": 2.4309067833532232e-05,
|
| 1498 |
+
"loss": 0.4079,
|
| 1499 |
+
"step": 108500
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 1.55,
|
| 1503 |
+
"learning_rate": 2.419067644106003e-05,
|
| 1504 |
+
"loss": 0.4092,
|
| 1505 |
+
"step": 109000
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 1.56,
|
| 1509 |
+
"learning_rate": 2.4072285048587828e-05,
|
| 1510 |
+
"loss": 0.4097,
|
| 1511 |
+
"step": 109500
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 1.56,
|
| 1515 |
+
"learning_rate": 2.395389365611563e-05,
|
| 1516 |
+
"loss": 0.409,
|
| 1517 |
+
"step": 110000
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 1.56,
|
| 1521 |
+
"eval_accuracy": 0.907909121265083,
|
| 1522 |
+
"eval_loss": 0.3783527910709381,
|
| 1523 |
+
"eval_runtime": 4481.8573,
|
| 1524 |
+
"eval_samples_per_second": 31.457,
|
| 1525 |
+
"eval_steps_per_second": 7.864,
|
| 1526 |
+
"step": 110000
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 1.57,
|
| 1530 |
+
"learning_rate": 2.3835502263643424e-05,
|
| 1531 |
+
"loss": 0.403,
|
| 1532 |
+
"step": 110500
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 1.58,
|
| 1536 |
+
"learning_rate": 2.3717110871171226e-05,
|
| 1537 |
+
"loss": 0.4109,
|
| 1538 |
+
"step": 111000
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 1.58,
|
| 1542 |
+
"learning_rate": 2.359871947869902e-05,
|
| 1543 |
+
"loss": 0.4072,
|
| 1544 |
+
"step": 111500
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 1.59,
|
| 1548 |
+
"learning_rate": 2.348032808622682e-05,
|
| 1549 |
+
"loss": 0.4063,
|
| 1550 |
+
"step": 112000
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.6,
|
| 1554 |
+
"learning_rate": 2.336193669375462e-05,
|
| 1555 |
+
"loss": 0.4087,
|
| 1556 |
+
"step": 112500
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.61,
|
| 1560 |
+
"learning_rate": 2.3243545301282415e-05,
|
| 1561 |
+
"loss": 0.4034,
|
| 1562 |
+
"step": 113000
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 1.61,
|
| 1566 |
+
"learning_rate": 2.3125153908810216e-05,
|
| 1567 |
+
"loss": 0.4041,
|
| 1568 |
+
"step": 113500
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 1.62,
|
| 1572 |
+
"learning_rate": 2.300676251633801e-05,
|
| 1573 |
+
"loss": 0.4044,
|
| 1574 |
+
"step": 114000
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 1.63,
|
| 1578 |
+
"learning_rate": 2.2888371123865812e-05,
|
| 1579 |
+
"loss": 0.4072,
|
| 1580 |
+
"step": 114500
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 1.63,
|
| 1584 |
+
"learning_rate": 2.276997973139361e-05,
|
| 1585 |
+
"loss": 0.4041,
|
| 1586 |
+
"step": 115000
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 1.63,
|
| 1590 |
+
"eval_accuracy": 0.9086945617090499,
|
| 1591 |
+
"eval_loss": 0.37441667914390564,
|
| 1592 |
+
"eval_runtime": 4484.0758,
|
| 1593 |
+
"eval_samples_per_second": 31.441,
|
| 1594 |
+
"eval_steps_per_second": 7.86,
|
| 1595 |
+
"step": 115000
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 1.64,
|
| 1599 |
+
"learning_rate": 2.265158833892141e-05,
|
| 1600 |
+
"loss": 0.4044,
|
| 1601 |
+
"step": 115500
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 1.65,
|
| 1605 |
+
"learning_rate": 2.2533196946449206e-05,
|
| 1606 |
+
"loss": 0.4038,
|
| 1607 |
+
"step": 116000
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 1.66,
|
| 1611 |
+
"learning_rate": 2.2414805553977004e-05,
|
| 1612 |
+
"loss": 0.4003,
|
| 1613 |
+
"step": 116500
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 1.66,
|
| 1617 |
+
"learning_rate": 2.2296414161504803e-05,
|
| 1618 |
+
"loss": 0.4024,
|
| 1619 |
+
"step": 117000
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 1.67,
|
| 1623 |
+
"learning_rate": 2.2178022769032604e-05,
|
| 1624 |
+
"loss": 0.4032,
|
| 1625 |
+
"step": 117500
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 1.68,
|
| 1629 |
+
"learning_rate": 2.20596313765604e-05,
|
| 1630 |
+
"loss": 0.3995,
|
| 1631 |
+
"step": 118000
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 1.68,
|
| 1635 |
+
"learning_rate": 2.1941239984088197e-05,
|
| 1636 |
+
"loss": 0.4052,
|
| 1637 |
+
"step": 118500
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 1.69,
|
| 1641 |
+
"learning_rate": 2.1822848591615995e-05,
|
| 1642 |
+
"loss": 0.4007,
|
| 1643 |
+
"step": 119000
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 1.7,
|
| 1647 |
+
"learning_rate": 2.1704457199143793e-05,
|
| 1648 |
+
"loss": 0.4003,
|
| 1649 |
+
"step": 119500
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 1.7,
|
| 1653 |
+
"learning_rate": 2.1586065806671594e-05,
|
| 1654 |
+
"loss": 0.4025,
|
| 1655 |
+
"step": 120000
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.7,
|
| 1659 |
+
"eval_accuracy": 0.9094200337257503,
|
| 1660 |
+
"eval_loss": 0.37063175439834595,
|
| 1661 |
+
"eval_runtime": 4483.4685,
|
| 1662 |
+
"eval_samples_per_second": 31.446,
|
| 1663 |
+
"eval_steps_per_second": 7.862,
|
| 1664 |
+
"step": 120000
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 1.71,
|
| 1668 |
+
"learning_rate": 2.146767441419939e-05,
|
| 1669 |
+
"loss": 0.4016,
|
| 1670 |
+
"step": 120500
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 1.72,
|
| 1674 |
+
"learning_rate": 2.134928302172719e-05,
|
| 1675 |
+
"loss": 0.4001,
|
| 1676 |
+
"step": 121000
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 1.73,
|
| 1680 |
+
"learning_rate": 2.123089162925499e-05,
|
| 1681 |
+
"loss": 0.4006,
|
| 1682 |
+
"step": 121500
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 1.73,
|
| 1686 |
+
"learning_rate": 2.1112500236782787e-05,
|
| 1687 |
+
"loss": 0.3957,
|
| 1688 |
+
"step": 122000
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 1.74,
|
| 1692 |
+
"learning_rate": 2.0994108844310585e-05,
|
| 1693 |
+
"loss": 0.4004,
|
| 1694 |
+
"step": 122500
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 1.75,
|
| 1698 |
+
"learning_rate": 2.0875717451838383e-05,
|
| 1699 |
+
"loss": 0.3994,
|
| 1700 |
+
"step": 123000
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 1.75,
|
| 1704 |
+
"learning_rate": 2.075732605936618e-05,
|
| 1705 |
+
"loss": 0.3951,
|
| 1706 |
+
"step": 123500
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 1.76,
|
| 1710 |
+
"learning_rate": 2.063893466689398e-05,
|
| 1711 |
+
"loss": 0.3963,
|
| 1712 |
+
"step": 124000
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 1.77,
|
| 1716 |
+
"learning_rate": 2.0520543274421777e-05,
|
| 1717 |
+
"loss": 0.3955,
|
| 1718 |
+
"step": 124500
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 1.78,
|
| 1722 |
+
"learning_rate": 2.0402151881949575e-05,
|
| 1723 |
+
"loss": 0.398,
|
| 1724 |
+
"step": 125000
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 1.78,
|
| 1728 |
+
"eval_accuracy": 0.9097696447890194,
|
| 1729 |
+
"eval_loss": 0.3691096901893616,
|
| 1730 |
+
"eval_runtime": 4482.3141,
|
| 1731 |
+
"eval_samples_per_second": 31.454,
|
| 1732 |
+
"eval_steps_per_second": 7.864,
|
| 1733 |
+
"step": 125000
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 1.78,
|
| 1737 |
+
"learning_rate": 2.0283760489477373e-05,
|
| 1738 |
+
"loss": 0.402,
|
| 1739 |
+
"step": 125500
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 1.79,
|
| 1743 |
+
"learning_rate": 2.016536909700517e-05,
|
| 1744 |
+
"loss": 0.393,
|
| 1745 |
+
"step": 126000
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 1.8,
|
| 1749 |
+
"learning_rate": 2.004697770453297e-05,
|
| 1750 |
+
"loss": 0.3927,
|
| 1751 |
+
"step": 126500
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 1.8,
|
| 1755 |
+
"learning_rate": 1.9928586312060767e-05,
|
| 1756 |
+
"loss": 0.3931,
|
| 1757 |
+
"step": 127000
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 1.81,
|
| 1761 |
+
"learning_rate": 1.981019491958857e-05,
|
| 1762 |
+
"loss": 0.3922,
|
| 1763 |
+
"step": 127500
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 1.82,
|
| 1767 |
+
"learning_rate": 1.9691803527116364e-05,
|
| 1768 |
+
"loss": 0.3984,
|
| 1769 |
+
"step": 128000
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 1.83,
|
| 1773 |
+
"learning_rate": 1.9573412134644165e-05,
|
| 1774 |
+
"loss": 0.3958,
|
| 1775 |
+
"step": 128500
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 1.83,
|
| 1779 |
+
"learning_rate": 1.9455020742171963e-05,
|
| 1780 |
+
"loss": 0.3958,
|
| 1781 |
+
"step": 129000
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 1.84,
|
| 1785 |
+
"learning_rate": 1.933662934969976e-05,
|
| 1786 |
+
"loss": 0.3971,
|
| 1787 |
+
"step": 129500
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 1.85,
|
| 1791 |
+
"learning_rate": 1.921823795722756e-05,
|
| 1792 |
+
"loss": 0.3894,
|
| 1793 |
+
"step": 130000
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 1.85,
|
| 1797 |
+
"eval_accuracy": 0.9105847302510419,
|
| 1798 |
+
"eval_loss": 0.3647039532661438,
|
| 1799 |
+
"eval_runtime": 4483.1278,
|
| 1800 |
+
"eval_samples_per_second": 31.448,
|
| 1801 |
+
"eval_steps_per_second": 7.862,
|
| 1802 |
+
"step": 130000
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 1.85,
|
| 1806 |
+
"learning_rate": 1.9099846564755357e-05,
|
| 1807 |
+
"loss": 0.3912,
|
| 1808 |
+
"step": 130500
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 1.86,
|
| 1812 |
+
"learning_rate": 1.8981455172283155e-05,
|
| 1813 |
+
"loss": 0.3895,
|
| 1814 |
+
"step": 131000
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 1.87,
|
| 1818 |
+
"learning_rate": 1.8863063779810953e-05,
|
| 1819 |
+
"loss": 0.3906,
|
| 1820 |
+
"step": 131500
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 1.88,
|
| 1824 |
+
"learning_rate": 1.874467238733875e-05,
|
| 1825 |
+
"loss": 0.3893,
|
| 1826 |
+
"step": 132000
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 1.88,
|
| 1830 |
+
"learning_rate": 1.862628099486655e-05,
|
| 1831 |
+
"loss": 0.3941,
|
| 1832 |
+
"step": 132500
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 1.89,
|
| 1836 |
+
"learning_rate": 1.8507889602394348e-05,
|
| 1837 |
+
"loss": 0.394,
|
| 1838 |
+
"step": 133000
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 1.9,
|
| 1842 |
+
"learning_rate": 1.8389498209922146e-05,
|
| 1843 |
+
"loss": 0.3893,
|
| 1844 |
+
"step": 133500
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 1.9,
|
| 1848 |
+
"learning_rate": 1.8271106817449947e-05,
|
| 1849 |
+
"loss": 0.3911,
|
| 1850 |
+
"step": 134000
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 1.91,
|
| 1854 |
+
"learning_rate": 1.8152715424977742e-05,
|
| 1855 |
+
"loss": 0.3899,
|
| 1856 |
+
"step": 134500
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 1.92,
|
| 1860 |
+
"learning_rate": 1.8034324032505543e-05,
|
| 1861 |
+
"loss": 0.3898,
|
| 1862 |
+
"step": 135000
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 1.92,
|
| 1866 |
+
"eval_accuracy": 0.9108666750444402,
|
| 1867 |
+
"eval_loss": 0.36286085844039917,
|
| 1868 |
+
"eval_runtime": 4484.3794,
|
| 1869 |
+
"eval_samples_per_second": 31.439,
|
| 1870 |
+
"eval_steps_per_second": 7.86,
|
| 1871 |
+
"step": 135000
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 1.93,
|
| 1875 |
+
"learning_rate": 1.7915932640033338e-05,
|
| 1876 |
+
"loss": 0.389,
|
| 1877 |
+
"step": 135500
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 1.93,
|
| 1881 |
+
"learning_rate": 1.779754124756114e-05,
|
| 1882 |
+
"loss": 0.3878,
|
| 1883 |
+
"step": 136000
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 1.94,
|
| 1887 |
+
"learning_rate": 1.7679149855088938e-05,
|
| 1888 |
+
"loss": 0.3926,
|
| 1889 |
+
"step": 136500
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 1.95,
|
| 1893 |
+
"learning_rate": 1.7560758462616736e-05,
|
| 1894 |
+
"loss": 0.3916,
|
| 1895 |
+
"step": 137000
|
| 1896 |
+
},
|
| 1897 |
+
{
|
| 1898 |
+
"epoch": 1.95,
|
| 1899 |
+
"learning_rate": 1.7442367070144534e-05,
|
| 1900 |
+
"loss": 0.3903,
|
| 1901 |
+
"step": 137500
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 1.96,
|
| 1905 |
+
"learning_rate": 1.732397567767233e-05,
|
| 1906 |
+
"loss": 0.3932,
|
| 1907 |
+
"step": 138000
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 1.97,
|
| 1911 |
+
"learning_rate": 1.720558428520013e-05,
|
| 1912 |
+
"loss": 0.3894,
|
| 1913 |
+
"step": 138500
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 1.97,
|
| 1917 |
+
"learning_rate": 1.7087192892727928e-05,
|
| 1918 |
+
"loss": 0.3882,
|
| 1919 |
+
"step": 139000
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 1.98,
|
| 1923 |
+
"learning_rate": 1.6968801500255726e-05,
|
| 1924 |
+
"loss": 0.3874,
|
| 1925 |
+
"step": 139500
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 1.99,
|
| 1929 |
+
"learning_rate": 1.6850410107783524e-05,
|
| 1930 |
+
"loss": 0.3857,
|
| 1931 |
+
"step": 140000
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 1.99,
|
| 1935 |
+
"eval_accuracy": 0.9115911874403287,
|
| 1936 |
+
"eval_loss": 0.36069589853286743,
|
| 1937 |
+
"eval_runtime": 4496.781,
|
| 1938 |
+
"eval_samples_per_second": 31.353,
|
| 1939 |
+
"eval_steps_per_second": 7.838,
|
| 1940 |
+
"step": 140000
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 2.0,
|
| 1944 |
+
"learning_rate": 1.6732018715311322e-05,
|
| 1945 |
+
"loss": 0.3895,
|
| 1946 |
+
"step": 140500
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 2.0,
|
| 1950 |
+
"learning_rate": 1.661362732283912e-05,
|
| 1951 |
+
"loss": 0.3903,
|
| 1952 |
+
"step": 141000
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 2.01,
|
| 1956 |
+
"learning_rate": 1.6495235930366922e-05,
|
| 1957 |
+
"loss": 0.3858,
|
| 1958 |
+
"step": 141500
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 2.02,
|
| 1962 |
+
"learning_rate": 1.6376844537894716e-05,
|
| 1963 |
+
"loss": 0.3838,
|
| 1964 |
+
"step": 142000
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 2.02,
|
| 1968 |
+
"learning_rate": 1.6258453145422518e-05,
|
| 1969 |
+
"loss": 0.3811,
|
| 1970 |
+
"step": 142500
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 2.03,
|
| 1974 |
+
"learning_rate": 1.6140061752950313e-05,
|
| 1975 |
+
"loss": 0.3892,
|
| 1976 |
+
"step": 143000
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 2.04,
|
| 1980 |
+
"learning_rate": 1.6021670360478114e-05,
|
| 1981 |
+
"loss": 0.3857,
|
| 1982 |
+
"step": 143500
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 2.05,
|
| 1986 |
+
"learning_rate": 1.5903278968005912e-05,
|
| 1987 |
+
"loss": 0.3812,
|
| 1988 |
+
"step": 144000
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 2.05,
|
| 1992 |
+
"learning_rate": 1.5784887575533707e-05,
|
| 1993 |
+
"loss": 0.3862,
|
| 1994 |
+
"step": 144500
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 2.06,
|
| 1998 |
+
"learning_rate": 1.5666496183061508e-05,
|
| 1999 |
+
"loss": 0.3809,
|
| 2000 |
+
"step": 145000
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 2.06,
|
| 2004 |
+
"eval_accuracy": 0.9121499748273744,
|
| 2005 |
+
"eval_loss": 0.3579576909542084,
|
| 2006 |
+
"eval_runtime": 4496.0613,
|
| 2007 |
+
"eval_samples_per_second": 31.358,
|
| 2008 |
+
"eval_steps_per_second": 7.84,
|
| 2009 |
+
"step": 145000
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 2.07,
|
| 2013 |
+
"learning_rate": 1.5548104790589303e-05,
|
| 2014 |
+
"loss": 0.3831,
|
| 2015 |
+
"step": 145500
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 2.07,
|
| 2019 |
+
"learning_rate": 1.5429713398117104e-05,
|
| 2020 |
+
"loss": 0.3831,
|
| 2021 |
+
"step": 146000
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 2.08,
|
| 2025 |
+
"learning_rate": 1.5311322005644902e-05,
|
| 2026 |
+
"loss": 0.3822,
|
| 2027 |
+
"step": 146500
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 2.09,
|
| 2031 |
+
"learning_rate": 1.51929306131727e-05,
|
| 2032 |
+
"loss": 0.3817,
|
| 2033 |
+
"step": 147000
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 2.1,
|
| 2037 |
+
"learning_rate": 1.5074539220700499e-05,
|
| 2038 |
+
"loss": 0.3854,
|
| 2039 |
+
"step": 147500
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 2.1,
|
| 2043 |
+
"learning_rate": 1.4956147828228298e-05,
|
| 2044 |
+
"loss": 0.3851,
|
| 2045 |
+
"step": 148000
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 2.11,
|
| 2049 |
+
"learning_rate": 1.4837756435756095e-05,
|
| 2050 |
+
"loss": 0.3838,
|
| 2051 |
+
"step": 148500
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 2.12,
|
| 2055 |
+
"learning_rate": 1.4719365043283895e-05,
|
| 2056 |
+
"loss": 0.3812,
|
| 2057 |
+
"step": 149000
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 2.12,
|
| 2061 |
+
"learning_rate": 1.4600973650811693e-05,
|
| 2062 |
+
"loss": 0.3825,
|
| 2063 |
+
"step": 149500
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 2.13,
|
| 2067 |
+
"learning_rate": 1.4482582258339492e-05,
|
| 2068 |
+
"loss": 0.3819,
|
| 2069 |
+
"step": 150000
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 2.13,
|
| 2073 |
+
"eval_accuracy": 0.91259126809248,
|
| 2074 |
+
"eval_loss": 0.3549927771091461,
|
| 2075 |
+
"eval_runtime": 4493.1359,
|
| 2076 |
+
"eval_samples_per_second": 31.378,
|
| 2077 |
+
"eval_steps_per_second": 7.845,
|
| 2078 |
+
"step": 150000
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 2.14,
|
| 2082 |
+
"learning_rate": 1.4364190865867289e-05,
|
| 2083 |
+
"loss": 0.3793,
|
| 2084 |
+
"step": 150500
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 2.15,
|
| 2088 |
+
"learning_rate": 1.4245799473395085e-05,
|
| 2089 |
+
"loss": 0.3832,
|
| 2090 |
+
"step": 151000
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 2.15,
|
| 2094 |
+
"learning_rate": 1.4127408080922885e-05,
|
| 2095 |
+
"loss": 0.3806,
|
| 2096 |
+
"step": 151500
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 2.16,
|
| 2100 |
+
"learning_rate": 1.4009016688450683e-05,
|
| 2101 |
+
"loss": 0.3784,
|
| 2102 |
+
"step": 152000
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 2.17,
|
| 2106 |
+
"learning_rate": 1.3890625295978483e-05,
|
| 2107 |
+
"loss": 0.382,
|
| 2108 |
+
"step": 152500
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 2.17,
|
| 2112 |
+
"learning_rate": 1.3772233903506279e-05,
|
| 2113 |
+
"loss": 0.3788,
|
| 2114 |
+
"step": 153000
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 2.18,
|
| 2118 |
+
"learning_rate": 1.3653842511034079e-05,
|
| 2119 |
+
"loss": 0.3837,
|
| 2120 |
+
"step": 153500
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 2.19,
|
| 2124 |
+
"learning_rate": 1.3535451118561875e-05,
|
| 2125 |
+
"loss": 0.3843,
|
| 2126 |
+
"step": 154000
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 2.19,
|
| 2130 |
+
"learning_rate": 1.3417059726089675e-05,
|
| 2131 |
+
"loss": 0.3813,
|
| 2132 |
+
"step": 154500
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 2.2,
|
| 2136 |
+
"learning_rate": 1.3298668333617473e-05,
|
| 2137 |
+
"loss": 0.3795,
|
| 2138 |
+
"step": 155000
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 2.2,
|
| 2142 |
+
"eval_accuracy": 0.913045165928687,
|
| 2143 |
+
"eval_loss": 0.3533931076526642,
|
| 2144 |
+
"eval_runtime": 4479.8171,
|
| 2145 |
+
"eval_samples_per_second": 31.471,
|
| 2146 |
+
"eval_steps_per_second": 7.868,
|
| 2147 |
+
"step": 155000
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 2.21,
|
| 2151 |
+
"learning_rate": 1.3180276941145273e-05,
|
| 2152 |
+
"loss": 0.3806,
|
| 2153 |
+
"step": 155500
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 2.22,
|
| 2157 |
+
"learning_rate": 1.306188554867307e-05,
|
| 2158 |
+
"loss": 0.3763,
|
| 2159 |
+
"step": 156000
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 2.22,
|
| 2163 |
+
"learning_rate": 1.2943494156200869e-05,
|
| 2164 |
+
"loss": 0.371,
|
| 2165 |
+
"step": 156500
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 2.23,
|
| 2169 |
+
"learning_rate": 1.2825102763728667e-05,
|
| 2170 |
+
"loss": 0.3809,
|
| 2171 |
+
"step": 157000
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 2.24,
|
| 2175 |
+
"learning_rate": 1.2706711371256467e-05,
|
| 2176 |
+
"loss": 0.3775,
|
| 2177 |
+
"step": 157500
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 2.24,
|
| 2181 |
+
"learning_rate": 1.2588319978784263e-05,
|
| 2182 |
+
"loss": 0.3813,
|
| 2183 |
+
"step": 158000
|
| 2184 |
+
},
|
| 2185 |
+
{
|
| 2186 |
+
"epoch": 2.25,
|
| 2187 |
+
"learning_rate": 1.2469928586312061e-05,
|
| 2188 |
+
"loss": 0.3808,
|
| 2189 |
+
"step": 158500
|
| 2190 |
+
},
|
| 2191 |
+
{
|
| 2192 |
+
"epoch": 2.26,
|
| 2193 |
+
"learning_rate": 1.235153719383986e-05,
|
| 2194 |
+
"loss": 0.3771,
|
| 2195 |
+
"step": 159000
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 2.27,
|
| 2199 |
+
"learning_rate": 1.2233145801367659e-05,
|
| 2200 |
+
"loss": 0.3734,
|
| 2201 |
+
"step": 159500
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 2.27,
|
| 2205 |
+
"learning_rate": 1.2114754408895457e-05,
|
| 2206 |
+
"loss": 0.3772,
|
| 2207 |
+
"step": 160000
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 2.27,
|
| 2211 |
+
"eval_accuracy": 0.9134737242143287,
|
| 2212 |
+
"eval_loss": 0.3512367606163025,
|
| 2213 |
+
"eval_runtime": 4479.3521,
|
| 2214 |
+
"eval_samples_per_second": 31.475,
|
| 2215 |
+
"eval_steps_per_second": 7.869,
|
| 2216 |
+
"step": 160000
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 2.28,
|
| 2220 |
+
"learning_rate": 1.1996363016423255e-05,
|
| 2221 |
+
"loss": 0.3765,
|
| 2222 |
+
"step": 160500
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 2.29,
|
| 2226 |
+
"learning_rate": 1.1877971623951052e-05,
|
| 2227 |
+
"loss": 0.3771,
|
| 2228 |
+
"step": 161000
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 2.29,
|
| 2232 |
+
"learning_rate": 1.175958023147885e-05,
|
| 2233 |
+
"loss": 0.3757,
|
| 2234 |
+
"step": 161500
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 2.3,
|
| 2238 |
+
"learning_rate": 1.164118883900665e-05,
|
| 2239 |
+
"loss": 0.3742,
|
| 2240 |
+
"step": 162000
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 2.31,
|
| 2244 |
+
"learning_rate": 1.1522797446534448e-05,
|
| 2245 |
+
"loss": 0.3738,
|
| 2246 |
+
"step": 162500
|
| 2247 |
+
},
|
| 2248 |
+
{
|
| 2249 |
+
"epoch": 2.32,
|
| 2250 |
+
"learning_rate": 1.1404406054062246e-05,
|
| 2251 |
+
"loss": 0.3743,
|
| 2252 |
+
"step": 163000
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 2.32,
|
| 2256 |
+
"learning_rate": 1.1286014661590044e-05,
|
| 2257 |
+
"loss": 0.3738,
|
| 2258 |
+
"step": 163500
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 2.33,
|
| 2262 |
+
"learning_rate": 1.1167623269117842e-05,
|
| 2263 |
+
"loss": 0.3712,
|
| 2264 |
+
"step": 164000
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 2.34,
|
| 2268 |
+
"learning_rate": 1.1049231876645642e-05,
|
| 2269 |
+
"loss": 0.373,
|
| 2270 |
+
"step": 164500
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 2.34,
|
| 2274 |
+
"learning_rate": 1.093084048417344e-05,
|
| 2275 |
+
"loss": 0.3723,
|
| 2276 |
+
"step": 165000
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 2.34,
|
| 2280 |
+
"eval_accuracy": 0.9138787356297591,
|
| 2281 |
+
"eval_loss": 0.3488907516002655,
|
| 2282 |
+
"eval_runtime": 4481.5458,
|
| 2283 |
+
"eval_samples_per_second": 31.459,
|
| 2284 |
+
"eval_steps_per_second": 7.865,
|
| 2285 |
+
"step": 165000
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 2.35,
|
| 2289 |
+
"learning_rate": 1.0812449091701238e-05,
|
| 2290 |
+
"loss": 0.373,
|
| 2291 |
+
"step": 165500
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 2.36,
|
| 2295 |
+
"learning_rate": 1.0694057699229036e-05,
|
| 2296 |
+
"loss": 0.3715,
|
| 2297 |
+
"step": 166000
|
| 2298 |
+
},
|
| 2299 |
+
{
|
| 2300 |
+
"epoch": 2.37,
|
| 2301 |
+
"learning_rate": 1.0575666306756834e-05,
|
| 2302 |
+
"loss": 0.3753,
|
| 2303 |
+
"step": 166500
|
| 2304 |
+
},
|
| 2305 |
+
{
|
| 2306 |
+
"epoch": 2.37,
|
| 2307 |
+
"learning_rate": 1.0457274914284634e-05,
|
| 2308 |
+
"loss": 0.3755,
|
| 2309 |
+
"step": 167000
|
| 2310 |
+
},
|
| 2311 |
+
{
|
| 2312 |
+
"epoch": 2.38,
|
| 2313 |
+
"learning_rate": 1.033888352181243e-05,
|
| 2314 |
+
"loss": 0.3736,
|
| 2315 |
+
"step": 167500
|
| 2316 |
+
},
|
| 2317 |
+
{
|
| 2318 |
+
"epoch": 2.39,
|
| 2319 |
+
"learning_rate": 1.0220492129340228e-05,
|
| 2320 |
+
"loss": 0.3738,
|
| 2321 |
+
"step": 168000
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 2.39,
|
| 2325 |
+
"learning_rate": 1.0102100736868026e-05,
|
| 2326 |
+
"loss": 0.3718,
|
| 2327 |
+
"step": 168500
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 2.4,
|
| 2331 |
+
"learning_rate": 9.983709344395826e-06,
|
| 2332 |
+
"loss": 0.3701,
|
| 2333 |
+
"step": 169000
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 2.41,
|
| 2337 |
+
"learning_rate": 9.865317951923624e-06,
|
| 2338 |
+
"loss": 0.3747,
|
| 2339 |
+
"step": 169500
|
| 2340 |
+
},
|
| 2341 |
+
{
|
| 2342 |
+
"epoch": 2.42,
|
| 2343 |
+
"learning_rate": 9.746926559451422e-06,
|
| 2344 |
+
"loss": 0.3741,
|
| 2345 |
+
"step": 170000
|
| 2346 |
+
},
|
| 2347 |
+
{
|
| 2348 |
+
"epoch": 2.42,
|
| 2349 |
+
"eval_accuracy": 0.9141958812185558,
|
| 2350 |
+
"eval_loss": 0.34678730368614197,
|
| 2351 |
+
"eval_runtime": 4483.3675,
|
| 2352 |
+
"eval_samples_per_second": 31.446,
|
| 2353 |
+
"eval_steps_per_second": 7.862,
|
| 2354 |
+
"step": 170000
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 2.42,
|
| 2358 |
+
"learning_rate": 9.62853516697922e-06,
|
| 2359 |
+
"loss": 0.3731,
|
| 2360 |
+
"step": 170500
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 2.43,
|
| 2364 |
+
"learning_rate": 9.510143774507018e-06,
|
| 2365 |
+
"loss": 0.3709,
|
| 2366 |
+
"step": 171000
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 2.44,
|
| 2370 |
+
"learning_rate": 9.391752382034818e-06,
|
| 2371 |
+
"loss": 0.3674,
|
| 2372 |
+
"step": 171500
|
| 2373 |
+
},
|
| 2374 |
+
{
|
| 2375 |
+
"epoch": 2.44,
|
| 2376 |
+
"learning_rate": 9.273360989562616e-06,
|
| 2377 |
+
"loss": 0.3688,
|
| 2378 |
+
"step": 172000
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 2.45,
|
| 2382 |
+
"learning_rate": 9.154969597090414e-06,
|
| 2383 |
+
"loss": 0.3708,
|
| 2384 |
+
"step": 172500
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 2.46,
|
| 2388 |
+
"learning_rate": 9.036578204618212e-06,
|
| 2389 |
+
"loss": 0.3705,
|
| 2390 |
+
"step": 173000
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 2.46,
|
| 2394 |
+
"learning_rate": 8.91818681214601e-06,
|
| 2395 |
+
"loss": 0.3688,
|
| 2396 |
+
"step": 173500
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 2.47,
|
| 2400 |
+
"learning_rate": 8.799795419673808e-06,
|
| 2401 |
+
"loss": 0.3688,
|
| 2402 |
+
"step": 174000
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 2.48,
|
| 2406 |
+
"learning_rate": 8.681404027201606e-06,
|
| 2407 |
+
"loss": 0.3731,
|
| 2408 |
+
"step": 174500
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"epoch": 2.49,
|
| 2412 |
+
"learning_rate": 8.563012634729405e-06,
|
| 2413 |
+
"loss": 0.3678,
|
| 2414 |
+
"step": 175000
|
| 2415 |
+
},
|
| 2416 |
+
{
|
| 2417 |
+
"epoch": 2.49,
|
| 2418 |
+
"eval_accuracy": 0.9147443492541378,
|
| 2419 |
+
"eval_loss": 0.3443525731563568,
|
| 2420 |
+
"eval_runtime": 4487.0096,
|
| 2421 |
+
"eval_samples_per_second": 31.421,
|
| 2422 |
+
"eval_steps_per_second": 7.855,
|
| 2423 |
+
"step": 175000
|
| 2424 |
+
},
|
| 2425 |
+
{
|
| 2426 |
+
"epoch": 2.49,
|
| 2427 |
+
"learning_rate": 8.444621242257203e-06,
|
| 2428 |
+
"loss": 0.3712,
|
| 2429 |
+
"step": 175500
|
| 2430 |
+
},
|
| 2431 |
+
{
|
| 2432 |
+
"epoch": 2.5,
|
| 2433 |
+
"learning_rate": 8.326229849785e-06,
|
| 2434 |
+
"loss": 0.3707,
|
| 2435 |
+
"step": 176000
|
| 2436 |
+
},
|
| 2437 |
+
{
|
| 2438 |
+
"epoch": 2.51,
|
| 2439 |
+
"learning_rate": 8.2078384573128e-06,
|
| 2440 |
+
"loss": 0.3693,
|
| 2441 |
+
"step": 176500
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 2.51,
|
| 2445 |
+
"learning_rate": 8.089447064840599e-06,
|
| 2446 |
+
"loss": 0.371,
|
| 2447 |
+
"step": 177000
|
| 2448 |
+
},
|
| 2449 |
+
{
|
| 2450 |
+
"epoch": 2.52,
|
| 2451 |
+
"learning_rate": 7.971055672368397e-06,
|
| 2452 |
+
"loss": 0.371,
|
| 2453 |
+
"step": 177500
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 2.53,
|
| 2457 |
+
"learning_rate": 7.852664279896195e-06,
|
| 2458 |
+
"loss": 0.3724,
|
| 2459 |
+
"step": 178000
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 2.54,
|
| 2463 |
+
"learning_rate": 7.734272887423993e-06,
|
| 2464 |
+
"loss": 0.3707,
|
| 2465 |
+
"step": 178500
|
| 2466 |
+
},
|
| 2467 |
+
{
|
| 2468 |
+
"epoch": 2.54,
|
| 2469 |
+
"learning_rate": 7.615881494951792e-06,
|
| 2470 |
+
"loss": 0.3673,
|
| 2471 |
+
"step": 179000
|
| 2472 |
+
},
|
| 2473 |
+
{
|
| 2474 |
+
"epoch": 2.55,
|
| 2475 |
+
"learning_rate": 7.4974901024795906e-06,
|
| 2476 |
+
"loss": 0.3684,
|
| 2477 |
+
"step": 179500
|
| 2478 |
+
},
|
| 2479 |
+
{
|
| 2480 |
+
"epoch": 2.56,
|
| 2481 |
+
"learning_rate": 7.379098710007389e-06,
|
| 2482 |
+
"loss": 0.3698,
|
| 2483 |
+
"step": 180000
|
| 2484 |
+
},
|
| 2485 |
+
{
|
| 2486 |
+
"epoch": 2.56,
|
| 2487 |
+
"eval_accuracy": 0.9149897136925411,
|
| 2488 |
+
"eval_loss": 0.3437068462371826,
|
| 2489 |
+
"eval_runtime": 4486.3551,
|
| 2490 |
+
"eval_samples_per_second": 31.426,
|
| 2491 |
+
"eval_steps_per_second": 7.856,
|
| 2492 |
+
"step": 180000
|
| 2493 |
+
},
|
| 2494 |
+
{
|
| 2495 |
+
"epoch": 2.56,
|
| 2496 |
+
"learning_rate": 7.260707317535187e-06,
|
| 2497 |
+
"loss": 0.3644,
|
| 2498 |
+
"step": 180500
|
| 2499 |
+
},
|
| 2500 |
+
{
|
| 2501 |
+
"epoch": 2.57,
|
| 2502 |
+
"learning_rate": 7.142315925062984e-06,
|
| 2503 |
+
"loss": 0.3696,
|
| 2504 |
+
"step": 181000
|
| 2505 |
+
},
|
| 2506 |
+
{
|
| 2507 |
+
"epoch": 2.58,
|
| 2508 |
+
"learning_rate": 7.023924532590782e-06,
|
| 2509 |
+
"loss": 0.3682,
|
| 2510 |
+
"step": 181500
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 2.59,
|
| 2514 |
+
"learning_rate": 6.905533140118581e-06,
|
| 2515 |
+
"loss": 0.3703,
|
| 2516 |
+
"step": 182000
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 2.59,
|
| 2520 |
+
"learning_rate": 6.787141747646379e-06,
|
| 2521 |
+
"loss": 0.3686,
|
| 2522 |
+
"step": 182500
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 2.6,
|
| 2526 |
+
"learning_rate": 6.668750355174178e-06,
|
| 2527 |
+
"loss": 0.3638,
|
| 2528 |
+
"step": 183000
|
| 2529 |
+
},
|
| 2530 |
+
{
|
| 2531 |
+
"epoch": 2.61,
|
| 2532 |
+
"learning_rate": 6.550358962701976e-06,
|
| 2533 |
+
"loss": 0.3652,
|
| 2534 |
+
"step": 183500
|
| 2535 |
+
},
|
| 2536 |
+
{
|
| 2537 |
+
"epoch": 2.61,
|
| 2538 |
+
"learning_rate": 6.431967570229774e-06,
|
| 2539 |
+
"loss": 0.367,
|
| 2540 |
+
"step": 184000
|
| 2541 |
+
},
|
| 2542 |
+
{
|
| 2543 |
+
"epoch": 2.62,
|
| 2544 |
+
"learning_rate": 6.313576177757573e-06,
|
| 2545 |
+
"loss": 0.3665,
|
| 2546 |
+
"step": 184500
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"epoch": 2.63,
|
| 2550 |
+
"learning_rate": 6.195184785285371e-06,
|
| 2551 |
+
"loss": 0.3685,
|
| 2552 |
+
"step": 185000
|
| 2553 |
+
},
|
| 2554 |
+
{
|
| 2555 |
+
"epoch": 2.63,
|
| 2556 |
+
"eval_accuracy": 0.9152310790161654,
|
| 2557 |
+
"eval_loss": 0.34259435534477234,
|
| 2558 |
+
"eval_runtime": 4485.6065,
|
| 2559 |
+
"eval_samples_per_second": 31.431,
|
| 2560 |
+
"eval_steps_per_second": 7.858,
|
| 2561 |
+
"step": 185000
|
| 2562 |
+
},
|
| 2563 |
+
{
|
| 2564 |
+
"epoch": 2.64,
|
| 2565 |
+
"learning_rate": 6.076793392813169e-06,
|
| 2566 |
+
"loss": 0.3635,
|
| 2567 |
+
"step": 185500
|
| 2568 |
+
},
|
| 2569 |
+
{
|
| 2570 |
+
"epoch": 2.64,
|
| 2571 |
+
"learning_rate": 5.958402000340967e-06,
|
| 2572 |
+
"loss": 0.3645,
|
| 2573 |
+
"step": 186000
|
| 2574 |
+
},
|
| 2575 |
+
{
|
| 2576 |
+
"epoch": 2.65,
|
| 2577 |
+
"learning_rate": 5.840010607868765e-06,
|
| 2578 |
+
"loss": 0.3697,
|
| 2579 |
+
"step": 186500
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 2.66,
|
| 2583 |
+
"learning_rate": 5.721619215396564e-06,
|
| 2584 |
+
"loss": 0.3693,
|
| 2585 |
+
"step": 187000
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 2.66,
|
| 2589 |
+
"learning_rate": 5.603227822924362e-06,
|
| 2590 |
+
"loss": 0.3638,
|
| 2591 |
+
"step": 187500
|
| 2592 |
+
},
|
| 2593 |
+
{
|
| 2594 |
+
"epoch": 2.67,
|
| 2595 |
+
"learning_rate": 5.484836430452161e-06,
|
| 2596 |
+
"loss": 0.3621,
|
| 2597 |
+
"step": 188000
|
| 2598 |
+
},
|
| 2599 |
+
{
|
| 2600 |
+
"epoch": 2.68,
|
| 2601 |
+
"learning_rate": 5.366445037979959e-06,
|
| 2602 |
+
"loss": 0.3602,
|
| 2603 |
+
"step": 188500
|
| 2604 |
+
},
|
| 2605 |
+
{
|
| 2606 |
+
"epoch": 2.69,
|
| 2607 |
+
"learning_rate": 5.248053645507757e-06,
|
| 2608 |
+
"loss": 0.369,
|
| 2609 |
+
"step": 189000
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"epoch": 2.69,
|
| 2613 |
+
"learning_rate": 5.1296622530355554e-06,
|
| 2614 |
+
"loss": 0.3678,
|
| 2615 |
+
"step": 189500
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 2.7,
|
| 2619 |
+
"learning_rate": 5.0112708605633535e-06,
|
| 2620 |
+
"loss": 0.3663,
|
| 2621 |
+
"step": 190000
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 2.7,
|
| 2625 |
+
"eval_accuracy": 0.9157103027094395,
|
| 2626 |
+
"eval_loss": 0.3402460515499115,
|
| 2627 |
+
"eval_runtime": 4483.4518,
|
| 2628 |
+
"eval_samples_per_second": 31.446,
|
| 2629 |
+
"eval_steps_per_second": 7.862,
|
| 2630 |
+
"step": 190000
|
| 2631 |
+
},
|
| 2632 |
+
{
|
| 2633 |
+
"epoch": 2.71,
|
| 2634 |
+
"learning_rate": 4.8928794680911524e-06,
|
| 2635 |
+
"loss": 0.366,
|
| 2636 |
+
"step": 190500
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 2.71,
|
| 2640 |
+
"learning_rate": 4.7744880756189505e-06,
|
| 2641 |
+
"loss": 0.3676,
|
| 2642 |
+
"step": 191000
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 2.72,
|
| 2646 |
+
"learning_rate": 4.656096683146749e-06,
|
| 2647 |
+
"loss": 0.3653,
|
| 2648 |
+
"step": 191500
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 2.73,
|
| 2652 |
+
"learning_rate": 4.5377052906745475e-06,
|
| 2653 |
+
"loss": 0.3649,
|
| 2654 |
+
"step": 192000
|
| 2655 |
+
},
|
| 2656 |
+
{
|
| 2657 |
+
"epoch": 2.73,
|
| 2658 |
+
"learning_rate": 4.419313898202345e-06,
|
| 2659 |
+
"loss": 0.3626,
|
| 2660 |
+
"step": 192500
|
| 2661 |
+
},
|
| 2662 |
+
{
|
| 2663 |
+
"epoch": 2.74,
|
| 2664 |
+
"learning_rate": 4.300922505730144e-06,
|
| 2665 |
+
"loss": 0.3603,
|
| 2666 |
+
"step": 193000
|
| 2667 |
+
},
|
| 2668 |
+
{
|
| 2669 |
+
"epoch": 2.75,
|
| 2670 |
+
"learning_rate": 4.182531113257942e-06,
|
| 2671 |
+
"loss": 0.3597,
|
| 2672 |
+
"step": 193500
|
| 2673 |
+
},
|
| 2674 |
+
{
|
| 2675 |
+
"epoch": 2.76,
|
| 2676 |
+
"learning_rate": 4.064139720785741e-06,
|
| 2677 |
+
"loss": 0.3666,
|
| 2678 |
+
"step": 194000
|
| 2679 |
+
},
|
| 2680 |
+
{
|
| 2681 |
+
"epoch": 2.76,
|
| 2682 |
+
"learning_rate": 3.945748328313539e-06,
|
| 2683 |
+
"loss": 0.3626,
|
| 2684 |
+
"step": 194500
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 2.77,
|
| 2688 |
+
"learning_rate": 3.827356935841337e-06,
|
| 2689 |
+
"loss": 0.3636,
|
| 2690 |
+
"step": 195000
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 2.77,
|
| 2694 |
+
"eval_accuracy": 0.9157984246825175,
|
| 2695 |
+
"eval_loss": 0.33979707956314087,
|
| 2696 |
+
"eval_runtime": 4482.6546,
|
| 2697 |
+
"eval_samples_per_second": 31.451,
|
| 2698 |
+
"eval_steps_per_second": 7.863,
|
| 2699 |
+
"step": 195000
|
| 2700 |
+
},
|
| 2701 |
+
{
|
| 2702 |
+
"epoch": 2.78,
|
| 2703 |
+
"learning_rate": 3.708965543369135e-06,
|
| 2704 |
+
"loss": 0.3641,
|
| 2705 |
+
"step": 195500
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 2.78,
|
| 2709 |
+
"learning_rate": 3.5905741508969334e-06,
|
| 2710 |
+
"loss": 0.3637,
|
| 2711 |
+
"step": 196000
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 2.79,
|
| 2715 |
+
"learning_rate": 3.4721827584247314e-06,
|
| 2716 |
+
"loss": 0.3646,
|
| 2717 |
+
"step": 196500
|
| 2718 |
+
},
|
| 2719 |
+
{
|
| 2720 |
+
"epoch": 2.8,
|
| 2721 |
+
"learning_rate": 3.35379136595253e-06,
|
| 2722 |
+
"loss": 0.3603,
|
| 2723 |
+
"step": 197000
|
| 2724 |
+
},
|
| 2725 |
+
{
|
| 2726 |
+
"epoch": 2.81,
|
| 2727 |
+
"learning_rate": 3.2353999734803284e-06,
|
| 2728 |
+
"loss": 0.3663,
|
| 2729 |
+
"step": 197500
|
| 2730 |
+
},
|
| 2731 |
+
{
|
| 2732 |
+
"epoch": 2.81,
|
| 2733 |
+
"learning_rate": 3.1170085810081265e-06,
|
| 2734 |
+
"loss": 0.3667,
|
| 2735 |
+
"step": 198000
|
| 2736 |
+
},
|
| 2737 |
+
{
|
| 2738 |
+
"epoch": 2.82,
|
| 2739 |
+
"learning_rate": 2.998617188535925e-06,
|
| 2740 |
+
"loss": 0.3638,
|
| 2741 |
+
"step": 198500
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 2.83,
|
| 2745 |
+
"learning_rate": 2.880225796063723e-06,
|
| 2746 |
+
"loss": 0.3597,
|
| 2747 |
+
"step": 199000
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 2.83,
|
| 2751 |
+
"learning_rate": 2.761834403591521e-06,
|
| 2752 |
+
"loss": 0.3639,
|
| 2753 |
+
"step": 199500
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 2.84,
|
| 2757 |
+
"learning_rate": 2.6434430111193197e-06,
|
| 2758 |
+
"loss": 0.3618,
|
| 2759 |
+
"step": 200000
|
| 2760 |
+
},
|
| 2761 |
+
{
|
| 2762 |
+
"epoch": 2.84,
|
| 2763 |
+
"eval_accuracy": 0.9161026342896044,
|
| 2764 |
+
"eval_loss": 0.33811721205711365,
|
| 2765 |
+
"eval_runtime": 4483.081,
|
| 2766 |
+
"eval_samples_per_second": 31.448,
|
| 2767 |
+
"eval_steps_per_second": 7.862,
|
| 2768 |
+
"step": 200000
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 2.85,
|
| 2772 |
+
"learning_rate": 2.525051618647118e-06,
|
| 2773 |
+
"loss": 0.3648,
|
| 2774 |
+
"step": 200500
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 2.86,
|
| 2778 |
+
"learning_rate": 2.4066602261749166e-06,
|
| 2779 |
+
"loss": 0.3614,
|
| 2780 |
+
"step": 201000
|
| 2781 |
+
},
|
| 2782 |
+
{
|
| 2783 |
+
"epoch": 2.86,
|
| 2784 |
+
"learning_rate": 2.2882688337027147e-06,
|
| 2785 |
+
"loss": 0.3607,
|
| 2786 |
+
"step": 201500
|
| 2787 |
+
},
|
| 2788 |
+
{
|
| 2789 |
+
"epoch": 2.87,
|
| 2790 |
+
"learning_rate": 2.169877441230513e-06,
|
| 2791 |
+
"loss": 0.3638,
|
| 2792 |
+
"step": 202000
|
| 2793 |
+
},
|
| 2794 |
+
{
|
| 2795 |
+
"epoch": 2.88,
|
| 2796 |
+
"learning_rate": 2.0514860487583113e-06,
|
| 2797 |
+
"loss": 0.3634,
|
| 2798 |
+
"step": 202500
|
| 2799 |
+
},
|
| 2800 |
+
{
|
| 2801 |
+
"epoch": 2.88,
|
| 2802 |
+
"learning_rate": 1.9330946562861094e-06,
|
| 2803 |
+
"loss": 0.3628,
|
| 2804 |
+
"step": 203000
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"epoch": 2.89,
|
| 2808 |
+
"learning_rate": 1.8147032638139079e-06,
|
| 2809 |
+
"loss": 0.358,
|
| 2810 |
+
"step": 203500
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 2.9,
|
| 2814 |
+
"learning_rate": 1.6963118713417062e-06,
|
| 2815 |
+
"loss": 0.3586,
|
| 2816 |
+
"step": 204000
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 2.91,
|
| 2820 |
+
"learning_rate": 1.5779204788695042e-06,
|
| 2821 |
+
"loss": 0.364,
|
| 2822 |
+
"step": 204500
|
| 2823 |
+
},
|
| 2824 |
+
{
|
| 2825 |
+
"epoch": 2.91,
|
| 2826 |
+
"learning_rate": 1.4595290863973025e-06,
|
| 2827 |
+
"loss": 0.3613,
|
| 2828 |
+
"step": 205000
|
| 2829 |
+
},
|
| 2830 |
+
{
|
| 2831 |
+
"epoch": 2.91,
|
| 2832 |
+
"eval_accuracy": 0.9164505308425883,
|
| 2833 |
+
"eval_loss": 0.3365662097930908,
|
| 2834 |
+
"eval_runtime": 4483.3011,
|
| 2835 |
+
"eval_samples_per_second": 31.447,
|
| 2836 |
+
"eval_steps_per_second": 7.862,
|
| 2837 |
+
"step": 205000
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 2.92,
|
| 2841 |
+
"learning_rate": 1.341137693925101e-06,
|
| 2842 |
+
"loss": 0.3641,
|
| 2843 |
+
"step": 205500
|
| 2844 |
+
},
|
| 2845 |
+
{
|
| 2846 |
+
"epoch": 2.93,
|
| 2847 |
+
"learning_rate": 1.2227463014528993e-06,
|
| 2848 |
+
"loss": 0.363,
|
| 2849 |
+
"step": 206000
|
| 2850 |
+
},
|
| 2851 |
+
{
|
| 2852 |
+
"epoch": 2.93,
|
| 2853 |
+
"learning_rate": 1.1043549089806974e-06,
|
| 2854 |
+
"loss": 0.3576,
|
| 2855 |
+
"step": 206500
|
| 2856 |
+
},
|
| 2857 |
+
{
|
| 2858 |
+
"epoch": 2.94,
|
| 2859 |
+
"learning_rate": 9.859635165084959e-07,
|
| 2860 |
+
"loss": 0.3627,
|
| 2861 |
+
"step": 207000
|
| 2862 |
+
},
|
| 2863 |
+
{
|
| 2864 |
+
"epoch": 2.95,
|
| 2865 |
+
"learning_rate": 8.67572124036294e-07,
|
| 2866 |
+
"loss": 0.3604,
|
| 2867 |
+
"step": 207500
|
| 2868 |
+
},
|
| 2869 |
+
{
|
| 2870 |
+
"epoch": 2.96,
|
| 2871 |
+
"learning_rate": 7.491807315640924e-07,
|
| 2872 |
+
"loss": 0.3614,
|
| 2873 |
+
"step": 208000
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 2.96,
|
| 2877 |
+
"learning_rate": 6.307893390918907e-07,
|
| 2878 |
+
"loss": 0.3584,
|
| 2879 |
+
"step": 208500
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 2.97,
|
| 2883 |
+
"learning_rate": 5.12397946619689e-07,
|
| 2884 |
+
"loss": 0.3594,
|
| 2885 |
+
"step": 209000
|
| 2886 |
+
},
|
| 2887 |
+
{
|
| 2888 |
+
"epoch": 2.98,
|
| 2889 |
+
"learning_rate": 3.9400655414748724e-07,
|
| 2890 |
+
"loss": 0.3589,
|
| 2891 |
+
"step": 209500
|
| 2892 |
+
},
|
| 2893 |
+
{
|
| 2894 |
+
"epoch": 2.98,
|
| 2895 |
+
"learning_rate": 2.756151616752856e-07,
|
| 2896 |
+
"loss": 0.3618,
|
| 2897 |
+
"step": 210000
|
| 2898 |
+
},
|
| 2899 |
+
{
|
| 2900 |
+
"epoch": 2.98,
|
| 2901 |
+
"eval_accuracy": 0.9162037471165208,
|
| 2902 |
+
"eval_loss": 0.3379279673099518,
|
| 2903 |
+
"eval_runtime": 4487.2562,
|
| 2904 |
+
"eval_samples_per_second": 31.419,
|
| 2905 |
+
"eval_steps_per_second": 7.855,
|
| 2906 |
+
"step": 210000
|
| 2907 |
+
},
|
| 2908 |
+
{
|
| 2909 |
+
"epoch": 2.99,
|
| 2910 |
+
"learning_rate": 1.5722376920308387e-07,
|
| 2911 |
+
"loss": 0.3621,
|
| 2912 |
+
"step": 210500
|
| 2913 |
+
},
|
| 2914 |
+
{
|
| 2915 |
+
"epoch": 3.0,
|
| 2916 |
+
"learning_rate": 3.883237673088216e-08,
|
| 2917 |
+
"loss": 0.3606,
|
| 2918 |
+
"step": 211000
|
| 2919 |
+
},
|
| 2920 |
+
{
|
| 2921 |
+
"epoch": 3.0,
|
| 2922 |
+
"step": 211164,
|
| 2923 |
+
"total_flos": 8.915475817785139e+17,
|
| 2924 |
+
"train_loss": 0.4472081179925129,
|
| 2925 |
+
"train_runtime": 493604.9481,
|
| 2926 |
+
"train_samples_per_second": 6.845,
|
| 2927 |
+
"train_steps_per_second": 0.428
|
| 2928 |
+
}
|
| 2929 |
+
],
|
| 2930 |
+
"max_steps": 211164,
|
| 2931 |
+
"num_train_epochs": 3,
|
| 2932 |
+
"total_flos": 8.915475817785139e+17,
|
| 2933 |
+
"trial_name": null,
|
| 2934 |
+
"trial_params": null
|
| 2935 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b43987fb3f0ebcf5e0d1191bfdf8a14e8a4a38f3b4782a2bbf593c2d1b2eeab2
|
| 3 |
+
size 3899
|