Upload files model pretrain_utg4java_220m_seq1024
Browse files- README.md +52 -3
- config.json +33 -0
- generation_config.json +8 -0
- merges.txt +0 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +753 -0
- tokenizer.json +0 -0
- tokenizer_config.json +959 -0
- trainer_state.json +1312 -0
- training_args.bin +3 -0
- vocab.json +0 -0
README.md
CHANGED
|
@@ -1,3 +1,52 @@
|
|
| 1 |
-
---
|
| 2 |
-
license:
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: bsd-3-clause
|
| 3 |
+
base_model: Salesforce/codet5p-220m
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
model-index:
|
| 7 |
+
- name: pretrain_utg4java_220m_seq1024
|
| 8 |
+
results: []
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 12 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 13 |
+
|
| 14 |
+
# pretrain_utg4java_220m_seq1024
|
| 15 |
+
|
| 16 |
+
This model is a fine-tuned version of [Salesforce/codet5p-220m](https://huggingface.co/Salesforce/codet5p-220m) on an unknown dataset.
|
| 17 |
+
|
| 18 |
+
## Model description
|
| 19 |
+
|
| 20 |
+
More information needed
|
| 21 |
+
|
| 22 |
+
## Intended uses & limitations
|
| 23 |
+
|
| 24 |
+
More information needed
|
| 25 |
+
|
| 26 |
+
## Training and evaluation data
|
| 27 |
+
|
| 28 |
+
More information needed
|
| 29 |
+
|
| 30 |
+
## Training procedure
|
| 31 |
+
|
| 32 |
+
### Training hyperparameters
|
| 33 |
+
|
| 34 |
+
The following hyperparameters were used during training:
|
| 35 |
+
- learning_rate: 0.0002
|
| 36 |
+
- train_batch_size: 16
|
| 37 |
+
- eval_batch_size: 32
|
| 38 |
+
- seed: 42
|
| 39 |
+
- gradient_accumulation_steps: 8
|
| 40 |
+
- total_train_batch_size: 128
|
| 41 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 42 |
+
- lr_scheduler_type: linear
|
| 43 |
+
- lr_scheduler_warmup_steps: 784
|
| 44 |
+
- num_epochs: 50
|
| 45 |
+
- mixed_precision_training: Native AMP
|
| 46 |
+
|
| 47 |
+
### Framework versions
|
| 48 |
+
|
| 49 |
+
- Transformers 4.40.0
|
| 50 |
+
- Pytorch 2.1.0
|
| 51 |
+
- Datasets 3.0.0
|
| 52 |
+
- Tokenizers 0.19.1
|
config.json
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "Salesforce/codet5p-220m",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"T5ForConditionalGeneration"
|
| 5 |
+
],
|
| 6 |
+
"bos_token_id": 1,
|
| 7 |
+
"classifier_dropout": 0.0,
|
| 8 |
+
"d_ff": 3072,
|
| 9 |
+
"d_kv": 64,
|
| 10 |
+
"d_model": 768,
|
| 11 |
+
"decoder_start_token_id": 0,
|
| 12 |
+
"dense_act_fn": "relu",
|
| 13 |
+
"dropout_rate": 0.1,
|
| 14 |
+
"eos_token_id": 2,
|
| 15 |
+
"feed_forward_proj": "relu",
|
| 16 |
+
"initializer_factor": 1.0,
|
| 17 |
+
"is_encoder_decoder": true,
|
| 18 |
+
"is_gated_act": false,
|
| 19 |
+
"layer_norm_epsilon": 1e-06,
|
| 20 |
+
"model_type": "t5",
|
| 21 |
+
"n_positions": 1024,
|
| 22 |
+
"num_decoder_layers": 12,
|
| 23 |
+
"num_heads": 12,
|
| 24 |
+
"num_layers": 12,
|
| 25 |
+
"output_past": true,
|
| 26 |
+
"pad_token_id": 0,
|
| 27 |
+
"relative_attention_max_distance": 128,
|
| 28 |
+
"relative_attention_num_buckets": 32,
|
| 29 |
+
"torch_dtype": "float32",
|
| 30 |
+
"transformers_version": "4.40.0",
|
| 31 |
+
"use_cache": true,
|
| 32 |
+
"vocab_size": 32100
|
| 33 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 1,
|
| 4 |
+
"decoder_start_token_id": 0,
|
| 5 |
+
"eos_token_id": 2,
|
| 6 |
+
"pad_token_id": 0,
|
| 7 |
+
"transformers_version": "4.40.0"
|
| 8 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:366b07d0a8dfeaf0b2cb4cbcd2af6d0f8edda296810fb9ba5b928bbbab9f2d00
|
| 3 |
+
size 891617358
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,753 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
{
|
| 4 |
+
"content": "<extra_id_99>",
|
| 5 |
+
"lstrip": true,
|
| 6 |
+
"normalized": true,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false
|
| 9 |
+
},
|
| 10 |
+
{
|
| 11 |
+
"content": "<extra_id_98>",
|
| 12 |
+
"lstrip": true,
|
| 13 |
+
"normalized": true,
|
| 14 |
+
"rstrip": false,
|
| 15 |
+
"single_word": false
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"content": "<extra_id_97>",
|
| 19 |
+
"lstrip": true,
|
| 20 |
+
"normalized": true,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"content": "<extra_id_96>",
|
| 26 |
+
"lstrip": true,
|
| 27 |
+
"normalized": true,
|
| 28 |
+
"rstrip": false,
|
| 29 |
+
"single_word": false
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"content": "<extra_id_95>",
|
| 33 |
+
"lstrip": true,
|
| 34 |
+
"normalized": true,
|
| 35 |
+
"rstrip": false,
|
| 36 |
+
"single_word": false
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"content": "<extra_id_94>",
|
| 40 |
+
"lstrip": true,
|
| 41 |
+
"normalized": true,
|
| 42 |
+
"rstrip": false,
|
| 43 |
+
"single_word": false
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"content": "<extra_id_93>",
|
| 47 |
+
"lstrip": true,
|
| 48 |
+
"normalized": true,
|
| 49 |
+
"rstrip": false,
|
| 50 |
+
"single_word": false
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"content": "<extra_id_92>",
|
| 54 |
+
"lstrip": true,
|
| 55 |
+
"normalized": true,
|
| 56 |
+
"rstrip": false,
|
| 57 |
+
"single_word": false
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"content": "<extra_id_91>",
|
| 61 |
+
"lstrip": true,
|
| 62 |
+
"normalized": true,
|
| 63 |
+
"rstrip": false,
|
| 64 |
+
"single_word": false
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"content": "<extra_id_90>",
|
| 68 |
+
"lstrip": true,
|
| 69 |
+
"normalized": true,
|
| 70 |
+
"rstrip": false,
|
| 71 |
+
"single_word": false
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"content": "<extra_id_89>",
|
| 75 |
+
"lstrip": true,
|
| 76 |
+
"normalized": true,
|
| 77 |
+
"rstrip": false,
|
| 78 |
+
"single_word": false
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"content": "<extra_id_88>",
|
| 82 |
+
"lstrip": true,
|
| 83 |
+
"normalized": true,
|
| 84 |
+
"rstrip": false,
|
| 85 |
+
"single_word": false
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"content": "<extra_id_87>",
|
| 89 |
+
"lstrip": true,
|
| 90 |
+
"normalized": true,
|
| 91 |
+
"rstrip": false,
|
| 92 |
+
"single_word": false
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"content": "<extra_id_86>",
|
| 96 |
+
"lstrip": true,
|
| 97 |
+
"normalized": true,
|
| 98 |
+
"rstrip": false,
|
| 99 |
+
"single_word": false
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"content": "<extra_id_85>",
|
| 103 |
+
"lstrip": true,
|
| 104 |
+
"normalized": true,
|
| 105 |
+
"rstrip": false,
|
| 106 |
+
"single_word": false
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"content": "<extra_id_84>",
|
| 110 |
+
"lstrip": true,
|
| 111 |
+
"normalized": true,
|
| 112 |
+
"rstrip": false,
|
| 113 |
+
"single_word": false
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"content": "<extra_id_83>",
|
| 117 |
+
"lstrip": true,
|
| 118 |
+
"normalized": true,
|
| 119 |
+
"rstrip": false,
|
| 120 |
+
"single_word": false
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"content": "<extra_id_82>",
|
| 124 |
+
"lstrip": true,
|
| 125 |
+
"normalized": true,
|
| 126 |
+
"rstrip": false,
|
| 127 |
+
"single_word": false
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"content": "<extra_id_81>",
|
| 131 |
+
"lstrip": true,
|
| 132 |
+
"normalized": true,
|
| 133 |
+
"rstrip": false,
|
| 134 |
+
"single_word": false
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"content": "<extra_id_80>",
|
| 138 |
+
"lstrip": true,
|
| 139 |
+
"normalized": true,
|
| 140 |
+
"rstrip": false,
|
| 141 |
+
"single_word": false
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"content": "<extra_id_79>",
|
| 145 |
+
"lstrip": true,
|
| 146 |
+
"normalized": true,
|
| 147 |
+
"rstrip": false,
|
| 148 |
+
"single_word": false
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"content": "<extra_id_78>",
|
| 152 |
+
"lstrip": true,
|
| 153 |
+
"normalized": true,
|
| 154 |
+
"rstrip": false,
|
| 155 |
+
"single_word": false
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"content": "<extra_id_77>",
|
| 159 |
+
"lstrip": true,
|
| 160 |
+
"normalized": true,
|
| 161 |
+
"rstrip": false,
|
| 162 |
+
"single_word": false
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"content": "<extra_id_76>",
|
| 166 |
+
"lstrip": true,
|
| 167 |
+
"normalized": true,
|
| 168 |
+
"rstrip": false,
|
| 169 |
+
"single_word": false
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"content": "<extra_id_75>",
|
| 173 |
+
"lstrip": true,
|
| 174 |
+
"normalized": true,
|
| 175 |
+
"rstrip": false,
|
| 176 |
+
"single_word": false
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"content": "<extra_id_74>",
|
| 180 |
+
"lstrip": true,
|
| 181 |
+
"normalized": true,
|
| 182 |
+
"rstrip": false,
|
| 183 |
+
"single_word": false
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"content": "<extra_id_73>",
|
| 187 |
+
"lstrip": true,
|
| 188 |
+
"normalized": true,
|
| 189 |
+
"rstrip": false,
|
| 190 |
+
"single_word": false
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"content": "<extra_id_72>",
|
| 194 |
+
"lstrip": true,
|
| 195 |
+
"normalized": true,
|
| 196 |
+
"rstrip": false,
|
| 197 |
+
"single_word": false
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"content": "<extra_id_71>",
|
| 201 |
+
"lstrip": true,
|
| 202 |
+
"normalized": true,
|
| 203 |
+
"rstrip": false,
|
| 204 |
+
"single_word": false
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"content": "<extra_id_70>",
|
| 208 |
+
"lstrip": true,
|
| 209 |
+
"normalized": true,
|
| 210 |
+
"rstrip": false,
|
| 211 |
+
"single_word": false
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"content": "<extra_id_69>",
|
| 215 |
+
"lstrip": true,
|
| 216 |
+
"normalized": true,
|
| 217 |
+
"rstrip": false,
|
| 218 |
+
"single_word": false
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"content": "<extra_id_68>",
|
| 222 |
+
"lstrip": true,
|
| 223 |
+
"normalized": true,
|
| 224 |
+
"rstrip": false,
|
| 225 |
+
"single_word": false
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"content": "<extra_id_67>",
|
| 229 |
+
"lstrip": true,
|
| 230 |
+
"normalized": true,
|
| 231 |
+
"rstrip": false,
|
| 232 |
+
"single_word": false
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"content": "<extra_id_66>",
|
| 236 |
+
"lstrip": true,
|
| 237 |
+
"normalized": true,
|
| 238 |
+
"rstrip": false,
|
| 239 |
+
"single_word": false
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"content": "<extra_id_65>",
|
| 243 |
+
"lstrip": true,
|
| 244 |
+
"normalized": true,
|
| 245 |
+
"rstrip": false,
|
| 246 |
+
"single_word": false
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"content": "<extra_id_64>",
|
| 250 |
+
"lstrip": true,
|
| 251 |
+
"normalized": true,
|
| 252 |
+
"rstrip": false,
|
| 253 |
+
"single_word": false
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"content": "<extra_id_63>",
|
| 257 |
+
"lstrip": true,
|
| 258 |
+
"normalized": true,
|
| 259 |
+
"rstrip": false,
|
| 260 |
+
"single_word": false
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"content": "<extra_id_62>",
|
| 264 |
+
"lstrip": true,
|
| 265 |
+
"normalized": true,
|
| 266 |
+
"rstrip": false,
|
| 267 |
+
"single_word": false
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"content": "<extra_id_61>",
|
| 271 |
+
"lstrip": true,
|
| 272 |
+
"normalized": true,
|
| 273 |
+
"rstrip": false,
|
| 274 |
+
"single_word": false
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"content": "<extra_id_60>",
|
| 278 |
+
"lstrip": true,
|
| 279 |
+
"normalized": true,
|
| 280 |
+
"rstrip": false,
|
| 281 |
+
"single_word": false
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"content": "<extra_id_59>",
|
| 285 |
+
"lstrip": true,
|
| 286 |
+
"normalized": true,
|
| 287 |
+
"rstrip": false,
|
| 288 |
+
"single_word": false
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"content": "<extra_id_58>",
|
| 292 |
+
"lstrip": true,
|
| 293 |
+
"normalized": true,
|
| 294 |
+
"rstrip": false,
|
| 295 |
+
"single_word": false
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"content": "<extra_id_57>",
|
| 299 |
+
"lstrip": true,
|
| 300 |
+
"normalized": true,
|
| 301 |
+
"rstrip": false,
|
| 302 |
+
"single_word": false
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"content": "<extra_id_56>",
|
| 306 |
+
"lstrip": true,
|
| 307 |
+
"normalized": true,
|
| 308 |
+
"rstrip": false,
|
| 309 |
+
"single_word": false
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"content": "<extra_id_55>",
|
| 313 |
+
"lstrip": true,
|
| 314 |
+
"normalized": true,
|
| 315 |
+
"rstrip": false,
|
| 316 |
+
"single_word": false
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"content": "<extra_id_54>",
|
| 320 |
+
"lstrip": true,
|
| 321 |
+
"normalized": true,
|
| 322 |
+
"rstrip": false,
|
| 323 |
+
"single_word": false
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"content": "<extra_id_53>",
|
| 327 |
+
"lstrip": true,
|
| 328 |
+
"normalized": true,
|
| 329 |
+
"rstrip": false,
|
| 330 |
+
"single_word": false
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"content": "<extra_id_52>",
|
| 334 |
+
"lstrip": true,
|
| 335 |
+
"normalized": true,
|
| 336 |
+
"rstrip": false,
|
| 337 |
+
"single_word": false
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"content": "<extra_id_51>",
|
| 341 |
+
"lstrip": true,
|
| 342 |
+
"normalized": true,
|
| 343 |
+
"rstrip": false,
|
| 344 |
+
"single_word": false
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"content": "<extra_id_50>",
|
| 348 |
+
"lstrip": true,
|
| 349 |
+
"normalized": true,
|
| 350 |
+
"rstrip": false,
|
| 351 |
+
"single_word": false
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"content": "<extra_id_49>",
|
| 355 |
+
"lstrip": true,
|
| 356 |
+
"normalized": true,
|
| 357 |
+
"rstrip": false,
|
| 358 |
+
"single_word": false
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"content": "<extra_id_48>",
|
| 362 |
+
"lstrip": true,
|
| 363 |
+
"normalized": true,
|
| 364 |
+
"rstrip": false,
|
| 365 |
+
"single_word": false
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"content": "<extra_id_47>",
|
| 369 |
+
"lstrip": true,
|
| 370 |
+
"normalized": true,
|
| 371 |
+
"rstrip": false,
|
| 372 |
+
"single_word": false
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"content": "<extra_id_46>",
|
| 376 |
+
"lstrip": true,
|
| 377 |
+
"normalized": true,
|
| 378 |
+
"rstrip": false,
|
| 379 |
+
"single_word": false
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"content": "<extra_id_45>",
|
| 383 |
+
"lstrip": true,
|
| 384 |
+
"normalized": true,
|
| 385 |
+
"rstrip": false,
|
| 386 |
+
"single_word": false
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"content": "<extra_id_44>",
|
| 390 |
+
"lstrip": true,
|
| 391 |
+
"normalized": true,
|
| 392 |
+
"rstrip": false,
|
| 393 |
+
"single_word": false
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"content": "<extra_id_43>",
|
| 397 |
+
"lstrip": true,
|
| 398 |
+
"normalized": true,
|
| 399 |
+
"rstrip": false,
|
| 400 |
+
"single_word": false
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"content": "<extra_id_42>",
|
| 404 |
+
"lstrip": true,
|
| 405 |
+
"normalized": true,
|
| 406 |
+
"rstrip": false,
|
| 407 |
+
"single_word": false
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"content": "<extra_id_41>",
|
| 411 |
+
"lstrip": true,
|
| 412 |
+
"normalized": true,
|
| 413 |
+
"rstrip": false,
|
| 414 |
+
"single_word": false
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"content": "<extra_id_40>",
|
| 418 |
+
"lstrip": true,
|
| 419 |
+
"normalized": true,
|
| 420 |
+
"rstrip": false,
|
| 421 |
+
"single_word": false
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"content": "<extra_id_39>",
|
| 425 |
+
"lstrip": true,
|
| 426 |
+
"normalized": true,
|
| 427 |
+
"rstrip": false,
|
| 428 |
+
"single_word": false
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"content": "<extra_id_38>",
|
| 432 |
+
"lstrip": true,
|
| 433 |
+
"normalized": true,
|
| 434 |
+
"rstrip": false,
|
| 435 |
+
"single_word": false
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"content": "<extra_id_37>",
|
| 439 |
+
"lstrip": true,
|
| 440 |
+
"normalized": true,
|
| 441 |
+
"rstrip": false,
|
| 442 |
+
"single_word": false
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"content": "<extra_id_36>",
|
| 446 |
+
"lstrip": true,
|
| 447 |
+
"normalized": true,
|
| 448 |
+
"rstrip": false,
|
| 449 |
+
"single_word": false
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"content": "<extra_id_35>",
|
| 453 |
+
"lstrip": true,
|
| 454 |
+
"normalized": true,
|
| 455 |
+
"rstrip": false,
|
| 456 |
+
"single_word": false
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"content": "<extra_id_34>",
|
| 460 |
+
"lstrip": true,
|
| 461 |
+
"normalized": true,
|
| 462 |
+
"rstrip": false,
|
| 463 |
+
"single_word": false
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"content": "<extra_id_33>",
|
| 467 |
+
"lstrip": true,
|
| 468 |
+
"normalized": true,
|
| 469 |
+
"rstrip": false,
|
| 470 |
+
"single_word": false
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"content": "<extra_id_32>",
|
| 474 |
+
"lstrip": true,
|
| 475 |
+
"normalized": true,
|
| 476 |
+
"rstrip": false,
|
| 477 |
+
"single_word": false
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"content": "<extra_id_31>",
|
| 481 |
+
"lstrip": true,
|
| 482 |
+
"normalized": true,
|
| 483 |
+
"rstrip": false,
|
| 484 |
+
"single_word": false
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"content": "<extra_id_30>",
|
| 488 |
+
"lstrip": true,
|
| 489 |
+
"normalized": true,
|
| 490 |
+
"rstrip": false,
|
| 491 |
+
"single_word": false
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"content": "<extra_id_29>",
|
| 495 |
+
"lstrip": true,
|
| 496 |
+
"normalized": true,
|
| 497 |
+
"rstrip": false,
|
| 498 |
+
"single_word": false
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"content": "<extra_id_28>",
|
| 502 |
+
"lstrip": true,
|
| 503 |
+
"normalized": true,
|
| 504 |
+
"rstrip": false,
|
| 505 |
+
"single_word": false
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"content": "<extra_id_27>",
|
| 509 |
+
"lstrip": true,
|
| 510 |
+
"normalized": true,
|
| 511 |
+
"rstrip": false,
|
| 512 |
+
"single_word": false
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"content": "<extra_id_26>",
|
| 516 |
+
"lstrip": true,
|
| 517 |
+
"normalized": true,
|
| 518 |
+
"rstrip": false,
|
| 519 |
+
"single_word": false
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"content": "<extra_id_25>",
|
| 523 |
+
"lstrip": true,
|
| 524 |
+
"normalized": true,
|
| 525 |
+
"rstrip": false,
|
| 526 |
+
"single_word": false
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"content": "<extra_id_24>",
|
| 530 |
+
"lstrip": true,
|
| 531 |
+
"normalized": true,
|
| 532 |
+
"rstrip": false,
|
| 533 |
+
"single_word": false
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"content": "<extra_id_23>",
|
| 537 |
+
"lstrip": true,
|
| 538 |
+
"normalized": true,
|
| 539 |
+
"rstrip": false,
|
| 540 |
+
"single_word": false
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"content": "<extra_id_22>",
|
| 544 |
+
"lstrip": true,
|
| 545 |
+
"normalized": true,
|
| 546 |
+
"rstrip": false,
|
| 547 |
+
"single_word": false
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"content": "<extra_id_21>",
|
| 551 |
+
"lstrip": true,
|
| 552 |
+
"normalized": true,
|
| 553 |
+
"rstrip": false,
|
| 554 |
+
"single_word": false
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"content": "<extra_id_20>",
|
| 558 |
+
"lstrip": true,
|
| 559 |
+
"normalized": true,
|
| 560 |
+
"rstrip": false,
|
| 561 |
+
"single_word": false
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"content": "<extra_id_19>",
|
| 565 |
+
"lstrip": true,
|
| 566 |
+
"normalized": true,
|
| 567 |
+
"rstrip": false,
|
| 568 |
+
"single_word": false
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"content": "<extra_id_18>",
|
| 572 |
+
"lstrip": true,
|
| 573 |
+
"normalized": true,
|
| 574 |
+
"rstrip": false,
|
| 575 |
+
"single_word": false
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"content": "<extra_id_17>",
|
| 579 |
+
"lstrip": true,
|
| 580 |
+
"normalized": true,
|
| 581 |
+
"rstrip": false,
|
| 582 |
+
"single_word": false
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"content": "<extra_id_16>",
|
| 586 |
+
"lstrip": true,
|
| 587 |
+
"normalized": true,
|
| 588 |
+
"rstrip": false,
|
| 589 |
+
"single_word": false
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"content": "<extra_id_15>",
|
| 593 |
+
"lstrip": true,
|
| 594 |
+
"normalized": true,
|
| 595 |
+
"rstrip": false,
|
| 596 |
+
"single_word": false
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"content": "<extra_id_14>",
|
| 600 |
+
"lstrip": true,
|
| 601 |
+
"normalized": true,
|
| 602 |
+
"rstrip": false,
|
| 603 |
+
"single_word": false
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"content": "<extra_id_13>",
|
| 607 |
+
"lstrip": true,
|
| 608 |
+
"normalized": true,
|
| 609 |
+
"rstrip": false,
|
| 610 |
+
"single_word": false
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"content": "<extra_id_12>",
|
| 614 |
+
"lstrip": true,
|
| 615 |
+
"normalized": true,
|
| 616 |
+
"rstrip": false,
|
| 617 |
+
"single_word": false
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"content": "<extra_id_11>",
|
| 621 |
+
"lstrip": true,
|
| 622 |
+
"normalized": true,
|
| 623 |
+
"rstrip": false,
|
| 624 |
+
"single_word": false
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"content": "<extra_id_10>",
|
| 628 |
+
"lstrip": true,
|
| 629 |
+
"normalized": true,
|
| 630 |
+
"rstrip": false,
|
| 631 |
+
"single_word": false
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"content": "<extra_id_9>",
|
| 635 |
+
"lstrip": true,
|
| 636 |
+
"normalized": true,
|
| 637 |
+
"rstrip": false,
|
| 638 |
+
"single_word": false
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"content": "<extra_id_8>",
|
| 642 |
+
"lstrip": true,
|
| 643 |
+
"normalized": true,
|
| 644 |
+
"rstrip": false,
|
| 645 |
+
"single_word": false
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"content": "<extra_id_7>",
|
| 649 |
+
"lstrip": true,
|
| 650 |
+
"normalized": true,
|
| 651 |
+
"rstrip": false,
|
| 652 |
+
"single_word": false
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"content": "<extra_id_6>",
|
| 656 |
+
"lstrip": true,
|
| 657 |
+
"normalized": true,
|
| 658 |
+
"rstrip": false,
|
| 659 |
+
"single_word": false
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"content": "<extra_id_5>",
|
| 663 |
+
"lstrip": true,
|
| 664 |
+
"normalized": true,
|
| 665 |
+
"rstrip": false,
|
| 666 |
+
"single_word": false
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"content": "<extra_id_4>",
|
| 670 |
+
"lstrip": true,
|
| 671 |
+
"normalized": true,
|
| 672 |
+
"rstrip": false,
|
| 673 |
+
"single_word": false
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"content": "<extra_id_3>",
|
| 677 |
+
"lstrip": true,
|
| 678 |
+
"normalized": true,
|
| 679 |
+
"rstrip": false,
|
| 680 |
+
"single_word": false
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"content": "<extra_id_2>",
|
| 684 |
+
"lstrip": true,
|
| 685 |
+
"normalized": true,
|
| 686 |
+
"rstrip": false,
|
| 687 |
+
"single_word": false
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"content": "<extra_id_1>",
|
| 691 |
+
"lstrip": true,
|
| 692 |
+
"normalized": true,
|
| 693 |
+
"rstrip": false,
|
| 694 |
+
"single_word": false
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"content": "<extra_id_0>",
|
| 698 |
+
"lstrip": true,
|
| 699 |
+
"normalized": true,
|
| 700 |
+
"rstrip": false,
|
| 701 |
+
"single_word": false
|
| 702 |
+
}
|
| 703 |
+
],
|
| 704 |
+
"bos_token": {
|
| 705 |
+
"content": "<s>",
|
| 706 |
+
"lstrip": false,
|
| 707 |
+
"normalized": true,
|
| 708 |
+
"rstrip": false,
|
| 709 |
+
"single_word": false
|
| 710 |
+
},
|
| 711 |
+
"cls_token": {
|
| 712 |
+
"content": "<s>",
|
| 713 |
+
"lstrip": false,
|
| 714 |
+
"normalized": true,
|
| 715 |
+
"rstrip": false,
|
| 716 |
+
"single_word": false
|
| 717 |
+
},
|
| 718 |
+
"eos_token": {
|
| 719 |
+
"content": "</s>",
|
| 720 |
+
"lstrip": false,
|
| 721 |
+
"normalized": true,
|
| 722 |
+
"rstrip": false,
|
| 723 |
+
"single_word": false
|
| 724 |
+
},
|
| 725 |
+
"mask_token": {
|
| 726 |
+
"content": "<mask>",
|
| 727 |
+
"lstrip": true,
|
| 728 |
+
"normalized": true,
|
| 729 |
+
"rstrip": false,
|
| 730 |
+
"single_word": false
|
| 731 |
+
},
|
| 732 |
+
"pad_token": {
|
| 733 |
+
"content": "<pad>",
|
| 734 |
+
"lstrip": false,
|
| 735 |
+
"normalized": true,
|
| 736 |
+
"rstrip": false,
|
| 737 |
+
"single_word": false
|
| 738 |
+
},
|
| 739 |
+
"sep_token": {
|
| 740 |
+
"content": "</s>",
|
| 741 |
+
"lstrip": false,
|
| 742 |
+
"normalized": true,
|
| 743 |
+
"rstrip": false,
|
| 744 |
+
"single_word": false
|
| 745 |
+
},
|
| 746 |
+
"unk_token": {
|
| 747 |
+
"content": "<unk>",
|
| 748 |
+
"lstrip": false,
|
| 749 |
+
"normalized": true,
|
| 750 |
+
"rstrip": false,
|
| 751 |
+
"single_word": false
|
| 752 |
+
}
|
| 753 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,959 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"0": {
|
| 5 |
+
"content": "<pad>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
+
"1": {
|
| 13 |
+
"content": "<s>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": true,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false,
|
| 18 |
+
"special": true
|
| 19 |
+
},
|
| 20 |
+
"2": {
|
| 21 |
+
"content": "</s>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": true,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false,
|
| 26 |
+
"special": true
|
| 27 |
+
},
|
| 28 |
+
"3": {
|
| 29 |
+
"content": "<unk>",
|
| 30 |
+
"lstrip": false,
|
| 31 |
+
"normalized": true,
|
| 32 |
+
"rstrip": false,
|
| 33 |
+
"single_word": false,
|
| 34 |
+
"special": true
|
| 35 |
+
},
|
| 36 |
+
"4": {
|
| 37 |
+
"content": "<mask>",
|
| 38 |
+
"lstrip": true,
|
| 39 |
+
"normalized": true,
|
| 40 |
+
"rstrip": false,
|
| 41 |
+
"single_word": false,
|
| 42 |
+
"special": true
|
| 43 |
+
},
|
| 44 |
+
"32000": {
|
| 45 |
+
"content": "<extra_id_99>",
|
| 46 |
+
"lstrip": true,
|
| 47 |
+
"normalized": true,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false,
|
| 50 |
+
"special": true
|
| 51 |
+
},
|
| 52 |
+
"32001": {
|
| 53 |
+
"content": "<extra_id_98>",
|
| 54 |
+
"lstrip": true,
|
| 55 |
+
"normalized": true,
|
| 56 |
+
"rstrip": false,
|
| 57 |
+
"single_word": false,
|
| 58 |
+
"special": true
|
| 59 |
+
},
|
| 60 |
+
"32002": {
|
| 61 |
+
"content": "<extra_id_97>",
|
| 62 |
+
"lstrip": true,
|
| 63 |
+
"normalized": true,
|
| 64 |
+
"rstrip": false,
|
| 65 |
+
"single_word": false,
|
| 66 |
+
"special": true
|
| 67 |
+
},
|
| 68 |
+
"32003": {
|
| 69 |
+
"content": "<extra_id_96>",
|
| 70 |
+
"lstrip": true,
|
| 71 |
+
"normalized": true,
|
| 72 |
+
"rstrip": false,
|
| 73 |
+
"single_word": false,
|
| 74 |
+
"special": true
|
| 75 |
+
},
|
| 76 |
+
"32004": {
|
| 77 |
+
"content": "<extra_id_95>",
|
| 78 |
+
"lstrip": true,
|
| 79 |
+
"normalized": true,
|
| 80 |
+
"rstrip": false,
|
| 81 |
+
"single_word": false,
|
| 82 |
+
"special": true
|
| 83 |
+
},
|
| 84 |
+
"32005": {
|
| 85 |
+
"content": "<extra_id_94>",
|
| 86 |
+
"lstrip": true,
|
| 87 |
+
"normalized": true,
|
| 88 |
+
"rstrip": false,
|
| 89 |
+
"single_word": false,
|
| 90 |
+
"special": true
|
| 91 |
+
},
|
| 92 |
+
"32006": {
|
| 93 |
+
"content": "<extra_id_93>",
|
| 94 |
+
"lstrip": true,
|
| 95 |
+
"normalized": true,
|
| 96 |
+
"rstrip": false,
|
| 97 |
+
"single_word": false,
|
| 98 |
+
"special": true
|
| 99 |
+
},
|
| 100 |
+
"32007": {
|
| 101 |
+
"content": "<extra_id_92>",
|
| 102 |
+
"lstrip": true,
|
| 103 |
+
"normalized": true,
|
| 104 |
+
"rstrip": false,
|
| 105 |
+
"single_word": false,
|
| 106 |
+
"special": true
|
| 107 |
+
},
|
| 108 |
+
"32008": {
|
| 109 |
+
"content": "<extra_id_91>",
|
| 110 |
+
"lstrip": true,
|
| 111 |
+
"normalized": true,
|
| 112 |
+
"rstrip": false,
|
| 113 |
+
"single_word": false,
|
| 114 |
+
"special": true
|
| 115 |
+
},
|
| 116 |
+
"32009": {
|
| 117 |
+
"content": "<extra_id_90>",
|
| 118 |
+
"lstrip": true,
|
| 119 |
+
"normalized": true,
|
| 120 |
+
"rstrip": false,
|
| 121 |
+
"single_word": false,
|
| 122 |
+
"special": true
|
| 123 |
+
},
|
| 124 |
+
"32010": {
|
| 125 |
+
"content": "<extra_id_89>",
|
| 126 |
+
"lstrip": true,
|
| 127 |
+
"normalized": true,
|
| 128 |
+
"rstrip": false,
|
| 129 |
+
"single_word": false,
|
| 130 |
+
"special": true
|
| 131 |
+
},
|
| 132 |
+
"32011": {
|
| 133 |
+
"content": "<extra_id_88>",
|
| 134 |
+
"lstrip": true,
|
| 135 |
+
"normalized": true,
|
| 136 |
+
"rstrip": false,
|
| 137 |
+
"single_word": false,
|
| 138 |
+
"special": true
|
| 139 |
+
},
|
| 140 |
+
"32012": {
|
| 141 |
+
"content": "<extra_id_87>",
|
| 142 |
+
"lstrip": true,
|
| 143 |
+
"normalized": true,
|
| 144 |
+
"rstrip": false,
|
| 145 |
+
"single_word": false,
|
| 146 |
+
"special": true
|
| 147 |
+
},
|
| 148 |
+
"32013": {
|
| 149 |
+
"content": "<extra_id_86>",
|
| 150 |
+
"lstrip": true,
|
| 151 |
+
"normalized": true,
|
| 152 |
+
"rstrip": false,
|
| 153 |
+
"single_word": false,
|
| 154 |
+
"special": true
|
| 155 |
+
},
|
| 156 |
+
"32014": {
|
| 157 |
+
"content": "<extra_id_85>",
|
| 158 |
+
"lstrip": true,
|
| 159 |
+
"normalized": true,
|
| 160 |
+
"rstrip": false,
|
| 161 |
+
"single_word": false,
|
| 162 |
+
"special": true
|
| 163 |
+
},
|
| 164 |
+
"32015": {
|
| 165 |
+
"content": "<extra_id_84>",
|
| 166 |
+
"lstrip": true,
|
| 167 |
+
"normalized": true,
|
| 168 |
+
"rstrip": false,
|
| 169 |
+
"single_word": false,
|
| 170 |
+
"special": true
|
| 171 |
+
},
|
| 172 |
+
"32016": {
|
| 173 |
+
"content": "<extra_id_83>",
|
| 174 |
+
"lstrip": true,
|
| 175 |
+
"normalized": true,
|
| 176 |
+
"rstrip": false,
|
| 177 |
+
"single_word": false,
|
| 178 |
+
"special": true
|
| 179 |
+
},
|
| 180 |
+
"32017": {
|
| 181 |
+
"content": "<extra_id_82>",
|
| 182 |
+
"lstrip": true,
|
| 183 |
+
"normalized": true,
|
| 184 |
+
"rstrip": false,
|
| 185 |
+
"single_word": false,
|
| 186 |
+
"special": true
|
| 187 |
+
},
|
| 188 |
+
"32018": {
|
| 189 |
+
"content": "<extra_id_81>",
|
| 190 |
+
"lstrip": true,
|
| 191 |
+
"normalized": true,
|
| 192 |
+
"rstrip": false,
|
| 193 |
+
"single_word": false,
|
| 194 |
+
"special": true
|
| 195 |
+
},
|
| 196 |
+
"32019": {
|
| 197 |
+
"content": "<extra_id_80>",
|
| 198 |
+
"lstrip": true,
|
| 199 |
+
"normalized": true,
|
| 200 |
+
"rstrip": false,
|
| 201 |
+
"single_word": false,
|
| 202 |
+
"special": true
|
| 203 |
+
},
|
| 204 |
+
"32020": {
|
| 205 |
+
"content": "<extra_id_79>",
|
| 206 |
+
"lstrip": true,
|
| 207 |
+
"normalized": true,
|
| 208 |
+
"rstrip": false,
|
| 209 |
+
"single_word": false,
|
| 210 |
+
"special": true
|
| 211 |
+
},
|
| 212 |
+
"32021": {
|
| 213 |
+
"content": "<extra_id_78>",
|
| 214 |
+
"lstrip": true,
|
| 215 |
+
"normalized": true,
|
| 216 |
+
"rstrip": false,
|
| 217 |
+
"single_word": false,
|
| 218 |
+
"special": true
|
| 219 |
+
},
|
| 220 |
+
"32022": {
|
| 221 |
+
"content": "<extra_id_77>",
|
| 222 |
+
"lstrip": true,
|
| 223 |
+
"normalized": true,
|
| 224 |
+
"rstrip": false,
|
| 225 |
+
"single_word": false,
|
| 226 |
+
"special": true
|
| 227 |
+
},
|
| 228 |
+
"32023": {
|
| 229 |
+
"content": "<extra_id_76>",
|
| 230 |
+
"lstrip": true,
|
| 231 |
+
"normalized": true,
|
| 232 |
+
"rstrip": false,
|
| 233 |
+
"single_word": false,
|
| 234 |
+
"special": true
|
| 235 |
+
},
|
| 236 |
+
"32024": {
|
| 237 |
+
"content": "<extra_id_75>",
|
| 238 |
+
"lstrip": true,
|
| 239 |
+
"normalized": true,
|
| 240 |
+
"rstrip": false,
|
| 241 |
+
"single_word": false,
|
| 242 |
+
"special": true
|
| 243 |
+
},
|
| 244 |
+
"32025": {
|
| 245 |
+
"content": "<extra_id_74>",
|
| 246 |
+
"lstrip": true,
|
| 247 |
+
"normalized": true,
|
| 248 |
+
"rstrip": false,
|
| 249 |
+
"single_word": false,
|
| 250 |
+
"special": true
|
| 251 |
+
},
|
| 252 |
+
"32026": {
|
| 253 |
+
"content": "<extra_id_73>",
|
| 254 |
+
"lstrip": true,
|
| 255 |
+
"normalized": true,
|
| 256 |
+
"rstrip": false,
|
| 257 |
+
"single_word": false,
|
| 258 |
+
"special": true
|
| 259 |
+
},
|
| 260 |
+
"32027": {
|
| 261 |
+
"content": "<extra_id_72>",
|
| 262 |
+
"lstrip": true,
|
| 263 |
+
"normalized": true,
|
| 264 |
+
"rstrip": false,
|
| 265 |
+
"single_word": false,
|
| 266 |
+
"special": true
|
| 267 |
+
},
|
| 268 |
+
"32028": {
|
| 269 |
+
"content": "<extra_id_71>",
|
| 270 |
+
"lstrip": true,
|
| 271 |
+
"normalized": true,
|
| 272 |
+
"rstrip": false,
|
| 273 |
+
"single_word": false,
|
| 274 |
+
"special": true
|
| 275 |
+
},
|
| 276 |
+
"32029": {
|
| 277 |
+
"content": "<extra_id_70>",
|
| 278 |
+
"lstrip": true,
|
| 279 |
+
"normalized": true,
|
| 280 |
+
"rstrip": false,
|
| 281 |
+
"single_word": false,
|
| 282 |
+
"special": true
|
| 283 |
+
},
|
| 284 |
+
"32030": {
|
| 285 |
+
"content": "<extra_id_69>",
|
| 286 |
+
"lstrip": true,
|
| 287 |
+
"normalized": true,
|
| 288 |
+
"rstrip": false,
|
| 289 |
+
"single_word": false,
|
| 290 |
+
"special": true
|
| 291 |
+
},
|
| 292 |
+
"32031": {
|
| 293 |
+
"content": "<extra_id_68>",
|
| 294 |
+
"lstrip": true,
|
| 295 |
+
"normalized": true,
|
| 296 |
+
"rstrip": false,
|
| 297 |
+
"single_word": false,
|
| 298 |
+
"special": true
|
| 299 |
+
},
|
| 300 |
+
"32032": {
|
| 301 |
+
"content": "<extra_id_67>",
|
| 302 |
+
"lstrip": true,
|
| 303 |
+
"normalized": true,
|
| 304 |
+
"rstrip": false,
|
| 305 |
+
"single_word": false,
|
| 306 |
+
"special": true
|
| 307 |
+
},
|
| 308 |
+
"32033": {
|
| 309 |
+
"content": "<extra_id_66>",
|
| 310 |
+
"lstrip": true,
|
| 311 |
+
"normalized": true,
|
| 312 |
+
"rstrip": false,
|
| 313 |
+
"single_word": false,
|
| 314 |
+
"special": true
|
| 315 |
+
},
|
| 316 |
+
"32034": {
|
| 317 |
+
"content": "<extra_id_65>",
|
| 318 |
+
"lstrip": true,
|
| 319 |
+
"normalized": true,
|
| 320 |
+
"rstrip": false,
|
| 321 |
+
"single_word": false,
|
| 322 |
+
"special": true
|
| 323 |
+
},
|
| 324 |
+
"32035": {
|
| 325 |
+
"content": "<extra_id_64>",
|
| 326 |
+
"lstrip": true,
|
| 327 |
+
"normalized": true,
|
| 328 |
+
"rstrip": false,
|
| 329 |
+
"single_word": false,
|
| 330 |
+
"special": true
|
| 331 |
+
},
|
| 332 |
+
"32036": {
|
| 333 |
+
"content": "<extra_id_63>",
|
| 334 |
+
"lstrip": true,
|
| 335 |
+
"normalized": true,
|
| 336 |
+
"rstrip": false,
|
| 337 |
+
"single_word": false,
|
| 338 |
+
"special": true
|
| 339 |
+
},
|
| 340 |
+
"32037": {
|
| 341 |
+
"content": "<extra_id_62>",
|
| 342 |
+
"lstrip": true,
|
| 343 |
+
"normalized": true,
|
| 344 |
+
"rstrip": false,
|
| 345 |
+
"single_word": false,
|
| 346 |
+
"special": true
|
| 347 |
+
},
|
| 348 |
+
"32038": {
|
| 349 |
+
"content": "<extra_id_61>",
|
| 350 |
+
"lstrip": true,
|
| 351 |
+
"normalized": true,
|
| 352 |
+
"rstrip": false,
|
| 353 |
+
"single_word": false,
|
| 354 |
+
"special": true
|
| 355 |
+
},
|
| 356 |
+
"32039": {
|
| 357 |
+
"content": "<extra_id_60>",
|
| 358 |
+
"lstrip": true,
|
| 359 |
+
"normalized": true,
|
| 360 |
+
"rstrip": false,
|
| 361 |
+
"single_word": false,
|
| 362 |
+
"special": true
|
| 363 |
+
},
|
| 364 |
+
"32040": {
|
| 365 |
+
"content": "<extra_id_59>",
|
| 366 |
+
"lstrip": true,
|
| 367 |
+
"normalized": true,
|
| 368 |
+
"rstrip": false,
|
| 369 |
+
"single_word": false,
|
| 370 |
+
"special": true
|
| 371 |
+
},
|
| 372 |
+
"32041": {
|
| 373 |
+
"content": "<extra_id_58>",
|
| 374 |
+
"lstrip": true,
|
| 375 |
+
"normalized": true,
|
| 376 |
+
"rstrip": false,
|
| 377 |
+
"single_word": false,
|
| 378 |
+
"special": true
|
| 379 |
+
},
|
| 380 |
+
"32042": {
|
| 381 |
+
"content": "<extra_id_57>",
|
| 382 |
+
"lstrip": true,
|
| 383 |
+
"normalized": true,
|
| 384 |
+
"rstrip": false,
|
| 385 |
+
"single_word": false,
|
| 386 |
+
"special": true
|
| 387 |
+
},
|
| 388 |
+
"32043": {
|
| 389 |
+
"content": "<extra_id_56>",
|
| 390 |
+
"lstrip": true,
|
| 391 |
+
"normalized": true,
|
| 392 |
+
"rstrip": false,
|
| 393 |
+
"single_word": false,
|
| 394 |
+
"special": true
|
| 395 |
+
},
|
| 396 |
+
"32044": {
|
| 397 |
+
"content": "<extra_id_55>",
|
| 398 |
+
"lstrip": true,
|
| 399 |
+
"normalized": true,
|
| 400 |
+
"rstrip": false,
|
| 401 |
+
"single_word": false,
|
| 402 |
+
"special": true
|
| 403 |
+
},
|
| 404 |
+
"32045": {
|
| 405 |
+
"content": "<extra_id_54>",
|
| 406 |
+
"lstrip": true,
|
| 407 |
+
"normalized": true,
|
| 408 |
+
"rstrip": false,
|
| 409 |
+
"single_word": false,
|
| 410 |
+
"special": true
|
| 411 |
+
},
|
| 412 |
+
"32046": {
|
| 413 |
+
"content": "<extra_id_53>",
|
| 414 |
+
"lstrip": true,
|
| 415 |
+
"normalized": true,
|
| 416 |
+
"rstrip": false,
|
| 417 |
+
"single_word": false,
|
| 418 |
+
"special": true
|
| 419 |
+
},
|
| 420 |
+
"32047": {
|
| 421 |
+
"content": "<extra_id_52>",
|
| 422 |
+
"lstrip": true,
|
| 423 |
+
"normalized": true,
|
| 424 |
+
"rstrip": false,
|
| 425 |
+
"single_word": false,
|
| 426 |
+
"special": true
|
| 427 |
+
},
|
| 428 |
+
"32048": {
|
| 429 |
+
"content": "<extra_id_51>",
|
| 430 |
+
"lstrip": true,
|
| 431 |
+
"normalized": true,
|
| 432 |
+
"rstrip": false,
|
| 433 |
+
"single_word": false,
|
| 434 |
+
"special": true
|
| 435 |
+
},
|
| 436 |
+
"32049": {
|
| 437 |
+
"content": "<extra_id_50>",
|
| 438 |
+
"lstrip": true,
|
| 439 |
+
"normalized": true,
|
| 440 |
+
"rstrip": false,
|
| 441 |
+
"single_word": false,
|
| 442 |
+
"special": true
|
| 443 |
+
},
|
| 444 |
+
"32050": {
|
| 445 |
+
"content": "<extra_id_49>",
|
| 446 |
+
"lstrip": true,
|
| 447 |
+
"normalized": true,
|
| 448 |
+
"rstrip": false,
|
| 449 |
+
"single_word": false,
|
| 450 |
+
"special": true
|
| 451 |
+
},
|
| 452 |
+
"32051": {
|
| 453 |
+
"content": "<extra_id_48>",
|
| 454 |
+
"lstrip": true,
|
| 455 |
+
"normalized": true,
|
| 456 |
+
"rstrip": false,
|
| 457 |
+
"single_word": false,
|
| 458 |
+
"special": true
|
| 459 |
+
},
|
| 460 |
+
"32052": {
|
| 461 |
+
"content": "<extra_id_47>",
|
| 462 |
+
"lstrip": true,
|
| 463 |
+
"normalized": true,
|
| 464 |
+
"rstrip": false,
|
| 465 |
+
"single_word": false,
|
| 466 |
+
"special": true
|
| 467 |
+
},
|
| 468 |
+
"32053": {
|
| 469 |
+
"content": "<extra_id_46>",
|
| 470 |
+
"lstrip": true,
|
| 471 |
+
"normalized": true,
|
| 472 |
+
"rstrip": false,
|
| 473 |
+
"single_word": false,
|
| 474 |
+
"special": true
|
| 475 |
+
},
|
| 476 |
+
"32054": {
|
| 477 |
+
"content": "<extra_id_45>",
|
| 478 |
+
"lstrip": true,
|
| 479 |
+
"normalized": true,
|
| 480 |
+
"rstrip": false,
|
| 481 |
+
"single_word": false,
|
| 482 |
+
"special": true
|
| 483 |
+
},
|
| 484 |
+
"32055": {
|
| 485 |
+
"content": "<extra_id_44>",
|
| 486 |
+
"lstrip": true,
|
| 487 |
+
"normalized": true,
|
| 488 |
+
"rstrip": false,
|
| 489 |
+
"single_word": false,
|
| 490 |
+
"special": true
|
| 491 |
+
},
|
| 492 |
+
"32056": {
|
| 493 |
+
"content": "<extra_id_43>",
|
| 494 |
+
"lstrip": true,
|
| 495 |
+
"normalized": true,
|
| 496 |
+
"rstrip": false,
|
| 497 |
+
"single_word": false,
|
| 498 |
+
"special": true
|
| 499 |
+
},
|
| 500 |
+
"32057": {
|
| 501 |
+
"content": "<extra_id_42>",
|
| 502 |
+
"lstrip": true,
|
| 503 |
+
"normalized": true,
|
| 504 |
+
"rstrip": false,
|
| 505 |
+
"single_word": false,
|
| 506 |
+
"special": true
|
| 507 |
+
},
|
| 508 |
+
"32058": {
|
| 509 |
+
"content": "<extra_id_41>",
|
| 510 |
+
"lstrip": true,
|
| 511 |
+
"normalized": true,
|
| 512 |
+
"rstrip": false,
|
| 513 |
+
"single_word": false,
|
| 514 |
+
"special": true
|
| 515 |
+
},
|
| 516 |
+
"32059": {
|
| 517 |
+
"content": "<extra_id_40>",
|
| 518 |
+
"lstrip": true,
|
| 519 |
+
"normalized": true,
|
| 520 |
+
"rstrip": false,
|
| 521 |
+
"single_word": false,
|
| 522 |
+
"special": true
|
| 523 |
+
},
|
| 524 |
+
"32060": {
|
| 525 |
+
"content": "<extra_id_39>",
|
| 526 |
+
"lstrip": true,
|
| 527 |
+
"normalized": true,
|
| 528 |
+
"rstrip": false,
|
| 529 |
+
"single_word": false,
|
| 530 |
+
"special": true
|
| 531 |
+
},
|
| 532 |
+
"32061": {
|
| 533 |
+
"content": "<extra_id_38>",
|
| 534 |
+
"lstrip": true,
|
| 535 |
+
"normalized": true,
|
| 536 |
+
"rstrip": false,
|
| 537 |
+
"single_word": false,
|
| 538 |
+
"special": true
|
| 539 |
+
},
|
| 540 |
+
"32062": {
|
| 541 |
+
"content": "<extra_id_37>",
|
| 542 |
+
"lstrip": true,
|
| 543 |
+
"normalized": true,
|
| 544 |
+
"rstrip": false,
|
| 545 |
+
"single_word": false,
|
| 546 |
+
"special": true
|
| 547 |
+
},
|
| 548 |
+
"32063": {
|
| 549 |
+
"content": "<extra_id_36>",
|
| 550 |
+
"lstrip": true,
|
| 551 |
+
"normalized": true,
|
| 552 |
+
"rstrip": false,
|
| 553 |
+
"single_word": false,
|
| 554 |
+
"special": true
|
| 555 |
+
},
|
| 556 |
+
"32064": {
|
| 557 |
+
"content": "<extra_id_35>",
|
| 558 |
+
"lstrip": true,
|
| 559 |
+
"normalized": true,
|
| 560 |
+
"rstrip": false,
|
| 561 |
+
"single_word": false,
|
| 562 |
+
"special": true
|
| 563 |
+
},
|
| 564 |
+
"32065": {
|
| 565 |
+
"content": "<extra_id_34>",
|
| 566 |
+
"lstrip": true,
|
| 567 |
+
"normalized": true,
|
| 568 |
+
"rstrip": false,
|
| 569 |
+
"single_word": false,
|
| 570 |
+
"special": true
|
| 571 |
+
},
|
| 572 |
+
"32066": {
|
| 573 |
+
"content": "<extra_id_33>",
|
| 574 |
+
"lstrip": true,
|
| 575 |
+
"normalized": true,
|
| 576 |
+
"rstrip": false,
|
| 577 |
+
"single_word": false,
|
| 578 |
+
"special": true
|
| 579 |
+
},
|
| 580 |
+
"32067": {
|
| 581 |
+
"content": "<extra_id_32>",
|
| 582 |
+
"lstrip": true,
|
| 583 |
+
"normalized": true,
|
| 584 |
+
"rstrip": false,
|
| 585 |
+
"single_word": false,
|
| 586 |
+
"special": true
|
| 587 |
+
},
|
| 588 |
+
"32068": {
|
| 589 |
+
"content": "<extra_id_31>",
|
| 590 |
+
"lstrip": true,
|
| 591 |
+
"normalized": true,
|
| 592 |
+
"rstrip": false,
|
| 593 |
+
"single_word": false,
|
| 594 |
+
"special": true
|
| 595 |
+
},
|
| 596 |
+
"32069": {
|
| 597 |
+
"content": "<extra_id_30>",
|
| 598 |
+
"lstrip": true,
|
| 599 |
+
"normalized": true,
|
| 600 |
+
"rstrip": false,
|
| 601 |
+
"single_word": false,
|
| 602 |
+
"special": true
|
| 603 |
+
},
|
| 604 |
+
"32070": {
|
| 605 |
+
"content": "<extra_id_29>",
|
| 606 |
+
"lstrip": true,
|
| 607 |
+
"normalized": true,
|
| 608 |
+
"rstrip": false,
|
| 609 |
+
"single_word": false,
|
| 610 |
+
"special": true
|
| 611 |
+
},
|
| 612 |
+
"32071": {
|
| 613 |
+
"content": "<extra_id_28>",
|
| 614 |
+
"lstrip": true,
|
| 615 |
+
"normalized": true,
|
| 616 |
+
"rstrip": false,
|
| 617 |
+
"single_word": false,
|
| 618 |
+
"special": true
|
| 619 |
+
},
|
| 620 |
+
"32072": {
|
| 621 |
+
"content": "<extra_id_27>",
|
| 622 |
+
"lstrip": true,
|
| 623 |
+
"normalized": true,
|
| 624 |
+
"rstrip": false,
|
| 625 |
+
"single_word": false,
|
| 626 |
+
"special": true
|
| 627 |
+
},
|
| 628 |
+
"32073": {
|
| 629 |
+
"content": "<extra_id_26>",
|
| 630 |
+
"lstrip": true,
|
| 631 |
+
"normalized": true,
|
| 632 |
+
"rstrip": false,
|
| 633 |
+
"single_word": false,
|
| 634 |
+
"special": true
|
| 635 |
+
},
|
| 636 |
+
"32074": {
|
| 637 |
+
"content": "<extra_id_25>",
|
| 638 |
+
"lstrip": true,
|
| 639 |
+
"normalized": true,
|
| 640 |
+
"rstrip": false,
|
| 641 |
+
"single_word": false,
|
| 642 |
+
"special": true
|
| 643 |
+
},
|
| 644 |
+
"32075": {
|
| 645 |
+
"content": "<extra_id_24>",
|
| 646 |
+
"lstrip": true,
|
| 647 |
+
"normalized": true,
|
| 648 |
+
"rstrip": false,
|
| 649 |
+
"single_word": false,
|
| 650 |
+
"special": true
|
| 651 |
+
},
|
| 652 |
+
"32076": {
|
| 653 |
+
"content": "<extra_id_23>",
|
| 654 |
+
"lstrip": true,
|
| 655 |
+
"normalized": true,
|
| 656 |
+
"rstrip": false,
|
| 657 |
+
"single_word": false,
|
| 658 |
+
"special": true
|
| 659 |
+
},
|
| 660 |
+
"32077": {
|
| 661 |
+
"content": "<extra_id_22>",
|
| 662 |
+
"lstrip": true,
|
| 663 |
+
"normalized": true,
|
| 664 |
+
"rstrip": false,
|
| 665 |
+
"single_word": false,
|
| 666 |
+
"special": true
|
| 667 |
+
},
|
| 668 |
+
"32078": {
|
| 669 |
+
"content": "<extra_id_21>",
|
| 670 |
+
"lstrip": true,
|
| 671 |
+
"normalized": true,
|
| 672 |
+
"rstrip": false,
|
| 673 |
+
"single_word": false,
|
| 674 |
+
"special": true
|
| 675 |
+
},
|
| 676 |
+
"32079": {
|
| 677 |
+
"content": "<extra_id_20>",
|
| 678 |
+
"lstrip": true,
|
| 679 |
+
"normalized": true,
|
| 680 |
+
"rstrip": false,
|
| 681 |
+
"single_word": false,
|
| 682 |
+
"special": true
|
| 683 |
+
},
|
| 684 |
+
"32080": {
|
| 685 |
+
"content": "<extra_id_19>",
|
| 686 |
+
"lstrip": true,
|
| 687 |
+
"normalized": true,
|
| 688 |
+
"rstrip": false,
|
| 689 |
+
"single_word": false,
|
| 690 |
+
"special": true
|
| 691 |
+
},
|
| 692 |
+
"32081": {
|
| 693 |
+
"content": "<extra_id_18>",
|
| 694 |
+
"lstrip": true,
|
| 695 |
+
"normalized": true,
|
| 696 |
+
"rstrip": false,
|
| 697 |
+
"single_word": false,
|
| 698 |
+
"special": true
|
| 699 |
+
},
|
| 700 |
+
"32082": {
|
| 701 |
+
"content": "<extra_id_17>",
|
| 702 |
+
"lstrip": true,
|
| 703 |
+
"normalized": true,
|
| 704 |
+
"rstrip": false,
|
| 705 |
+
"single_word": false,
|
| 706 |
+
"special": true
|
| 707 |
+
},
|
| 708 |
+
"32083": {
|
| 709 |
+
"content": "<extra_id_16>",
|
| 710 |
+
"lstrip": true,
|
| 711 |
+
"normalized": true,
|
| 712 |
+
"rstrip": false,
|
| 713 |
+
"single_word": false,
|
| 714 |
+
"special": true
|
| 715 |
+
},
|
| 716 |
+
"32084": {
|
| 717 |
+
"content": "<extra_id_15>",
|
| 718 |
+
"lstrip": true,
|
| 719 |
+
"normalized": true,
|
| 720 |
+
"rstrip": false,
|
| 721 |
+
"single_word": false,
|
| 722 |
+
"special": true
|
| 723 |
+
},
|
| 724 |
+
"32085": {
|
| 725 |
+
"content": "<extra_id_14>",
|
| 726 |
+
"lstrip": true,
|
| 727 |
+
"normalized": true,
|
| 728 |
+
"rstrip": false,
|
| 729 |
+
"single_word": false,
|
| 730 |
+
"special": true
|
| 731 |
+
},
|
| 732 |
+
"32086": {
|
| 733 |
+
"content": "<extra_id_13>",
|
| 734 |
+
"lstrip": true,
|
| 735 |
+
"normalized": true,
|
| 736 |
+
"rstrip": false,
|
| 737 |
+
"single_word": false,
|
| 738 |
+
"special": true
|
| 739 |
+
},
|
| 740 |
+
"32087": {
|
| 741 |
+
"content": "<extra_id_12>",
|
| 742 |
+
"lstrip": true,
|
| 743 |
+
"normalized": true,
|
| 744 |
+
"rstrip": false,
|
| 745 |
+
"single_word": false,
|
| 746 |
+
"special": true
|
| 747 |
+
},
|
| 748 |
+
"32088": {
|
| 749 |
+
"content": "<extra_id_11>",
|
| 750 |
+
"lstrip": true,
|
| 751 |
+
"normalized": true,
|
| 752 |
+
"rstrip": false,
|
| 753 |
+
"single_word": false,
|
| 754 |
+
"special": true
|
| 755 |
+
},
|
| 756 |
+
"32089": {
|
| 757 |
+
"content": "<extra_id_10>",
|
| 758 |
+
"lstrip": true,
|
| 759 |
+
"normalized": true,
|
| 760 |
+
"rstrip": false,
|
| 761 |
+
"single_word": false,
|
| 762 |
+
"special": true
|
| 763 |
+
},
|
| 764 |
+
"32090": {
|
| 765 |
+
"content": "<extra_id_9>",
|
| 766 |
+
"lstrip": true,
|
| 767 |
+
"normalized": true,
|
| 768 |
+
"rstrip": false,
|
| 769 |
+
"single_word": false,
|
| 770 |
+
"special": true
|
| 771 |
+
},
|
| 772 |
+
"32091": {
|
| 773 |
+
"content": "<extra_id_8>",
|
| 774 |
+
"lstrip": true,
|
| 775 |
+
"normalized": true,
|
| 776 |
+
"rstrip": false,
|
| 777 |
+
"single_word": false,
|
| 778 |
+
"special": true
|
| 779 |
+
},
|
| 780 |
+
"32092": {
|
| 781 |
+
"content": "<extra_id_7>",
|
| 782 |
+
"lstrip": true,
|
| 783 |
+
"normalized": true,
|
| 784 |
+
"rstrip": false,
|
| 785 |
+
"single_word": false,
|
| 786 |
+
"special": true
|
| 787 |
+
},
|
| 788 |
+
"32093": {
|
| 789 |
+
"content": "<extra_id_6>",
|
| 790 |
+
"lstrip": true,
|
| 791 |
+
"normalized": true,
|
| 792 |
+
"rstrip": false,
|
| 793 |
+
"single_word": false,
|
| 794 |
+
"special": true
|
| 795 |
+
},
|
| 796 |
+
"32094": {
|
| 797 |
+
"content": "<extra_id_5>",
|
| 798 |
+
"lstrip": true,
|
| 799 |
+
"normalized": true,
|
| 800 |
+
"rstrip": false,
|
| 801 |
+
"single_word": false,
|
| 802 |
+
"special": true
|
| 803 |
+
},
|
| 804 |
+
"32095": {
|
| 805 |
+
"content": "<extra_id_4>",
|
| 806 |
+
"lstrip": true,
|
| 807 |
+
"normalized": true,
|
| 808 |
+
"rstrip": false,
|
| 809 |
+
"single_word": false,
|
| 810 |
+
"special": true
|
| 811 |
+
},
|
| 812 |
+
"32096": {
|
| 813 |
+
"content": "<extra_id_3>",
|
| 814 |
+
"lstrip": true,
|
| 815 |
+
"normalized": true,
|
| 816 |
+
"rstrip": false,
|
| 817 |
+
"single_word": false,
|
| 818 |
+
"special": true
|
| 819 |
+
},
|
| 820 |
+
"32097": {
|
| 821 |
+
"content": "<extra_id_2>",
|
| 822 |
+
"lstrip": true,
|
| 823 |
+
"normalized": true,
|
| 824 |
+
"rstrip": false,
|
| 825 |
+
"single_word": false,
|
| 826 |
+
"special": true
|
| 827 |
+
},
|
| 828 |
+
"32098": {
|
| 829 |
+
"content": "<extra_id_1>",
|
| 830 |
+
"lstrip": true,
|
| 831 |
+
"normalized": true,
|
| 832 |
+
"rstrip": false,
|
| 833 |
+
"single_word": false,
|
| 834 |
+
"special": true
|
| 835 |
+
},
|
| 836 |
+
"32099": {
|
| 837 |
+
"content": "<extra_id_0>",
|
| 838 |
+
"lstrip": true,
|
| 839 |
+
"normalized": true,
|
| 840 |
+
"rstrip": false,
|
| 841 |
+
"single_word": false,
|
| 842 |
+
"special": true
|
| 843 |
+
}
|
| 844 |
+
},
|
| 845 |
+
"additional_special_tokens": [
|
| 846 |
+
"<extra_id_99>",
|
| 847 |
+
"<extra_id_98>",
|
| 848 |
+
"<extra_id_97>",
|
| 849 |
+
"<extra_id_96>",
|
| 850 |
+
"<extra_id_95>",
|
| 851 |
+
"<extra_id_94>",
|
| 852 |
+
"<extra_id_93>",
|
| 853 |
+
"<extra_id_92>",
|
| 854 |
+
"<extra_id_91>",
|
| 855 |
+
"<extra_id_90>",
|
| 856 |
+
"<extra_id_89>",
|
| 857 |
+
"<extra_id_88>",
|
| 858 |
+
"<extra_id_87>",
|
| 859 |
+
"<extra_id_86>",
|
| 860 |
+
"<extra_id_85>",
|
| 861 |
+
"<extra_id_84>",
|
| 862 |
+
"<extra_id_83>",
|
| 863 |
+
"<extra_id_82>",
|
| 864 |
+
"<extra_id_81>",
|
| 865 |
+
"<extra_id_80>",
|
| 866 |
+
"<extra_id_79>",
|
| 867 |
+
"<extra_id_78>",
|
| 868 |
+
"<extra_id_77>",
|
| 869 |
+
"<extra_id_76>",
|
| 870 |
+
"<extra_id_75>",
|
| 871 |
+
"<extra_id_74>",
|
| 872 |
+
"<extra_id_73>",
|
| 873 |
+
"<extra_id_72>",
|
| 874 |
+
"<extra_id_71>",
|
| 875 |
+
"<extra_id_70>",
|
| 876 |
+
"<extra_id_69>",
|
| 877 |
+
"<extra_id_68>",
|
| 878 |
+
"<extra_id_67>",
|
| 879 |
+
"<extra_id_66>",
|
| 880 |
+
"<extra_id_65>",
|
| 881 |
+
"<extra_id_64>",
|
| 882 |
+
"<extra_id_63>",
|
| 883 |
+
"<extra_id_62>",
|
| 884 |
+
"<extra_id_61>",
|
| 885 |
+
"<extra_id_60>",
|
| 886 |
+
"<extra_id_59>",
|
| 887 |
+
"<extra_id_58>",
|
| 888 |
+
"<extra_id_57>",
|
| 889 |
+
"<extra_id_56>",
|
| 890 |
+
"<extra_id_55>",
|
| 891 |
+
"<extra_id_54>",
|
| 892 |
+
"<extra_id_53>",
|
| 893 |
+
"<extra_id_52>",
|
| 894 |
+
"<extra_id_51>",
|
| 895 |
+
"<extra_id_50>",
|
| 896 |
+
"<extra_id_49>",
|
| 897 |
+
"<extra_id_48>",
|
| 898 |
+
"<extra_id_47>",
|
| 899 |
+
"<extra_id_46>",
|
| 900 |
+
"<extra_id_45>",
|
| 901 |
+
"<extra_id_44>",
|
| 902 |
+
"<extra_id_43>",
|
| 903 |
+
"<extra_id_42>",
|
| 904 |
+
"<extra_id_41>",
|
| 905 |
+
"<extra_id_40>",
|
| 906 |
+
"<extra_id_39>",
|
| 907 |
+
"<extra_id_38>",
|
| 908 |
+
"<extra_id_37>",
|
| 909 |
+
"<extra_id_36>",
|
| 910 |
+
"<extra_id_35>",
|
| 911 |
+
"<extra_id_34>",
|
| 912 |
+
"<extra_id_33>",
|
| 913 |
+
"<extra_id_32>",
|
| 914 |
+
"<extra_id_31>",
|
| 915 |
+
"<extra_id_30>",
|
| 916 |
+
"<extra_id_29>",
|
| 917 |
+
"<extra_id_28>",
|
| 918 |
+
"<extra_id_27>",
|
| 919 |
+
"<extra_id_26>",
|
| 920 |
+
"<extra_id_25>",
|
| 921 |
+
"<extra_id_24>",
|
| 922 |
+
"<extra_id_23>",
|
| 923 |
+
"<extra_id_22>",
|
| 924 |
+
"<extra_id_21>",
|
| 925 |
+
"<extra_id_20>",
|
| 926 |
+
"<extra_id_19>",
|
| 927 |
+
"<extra_id_18>",
|
| 928 |
+
"<extra_id_17>",
|
| 929 |
+
"<extra_id_16>",
|
| 930 |
+
"<extra_id_15>",
|
| 931 |
+
"<extra_id_14>",
|
| 932 |
+
"<extra_id_13>",
|
| 933 |
+
"<extra_id_12>",
|
| 934 |
+
"<extra_id_11>",
|
| 935 |
+
"<extra_id_10>",
|
| 936 |
+
"<extra_id_9>",
|
| 937 |
+
"<extra_id_8>",
|
| 938 |
+
"<extra_id_7>",
|
| 939 |
+
"<extra_id_6>",
|
| 940 |
+
"<extra_id_5>",
|
| 941 |
+
"<extra_id_4>",
|
| 942 |
+
"<extra_id_3>",
|
| 943 |
+
"<extra_id_2>",
|
| 944 |
+
"<extra_id_1>",
|
| 945 |
+
"<extra_id_0>"
|
| 946 |
+
],
|
| 947 |
+
"bos_token": "<s>",
|
| 948 |
+
"clean_up_tokenization_spaces": true,
|
| 949 |
+
"cls_token": "<s>",
|
| 950 |
+
"eos_token": "</s>",
|
| 951 |
+
"errors": "replace",
|
| 952 |
+
"mask_token": "<mask>",
|
| 953 |
+
"model_max_length": 1024,
|
| 954 |
+
"pad_token": "<pad>",
|
| 955 |
+
"sep_token": "</s>",
|
| 956 |
+
"tokenizer_class": "RobertaTokenizer",
|
| 957 |
+
"trim_offsets": true,
|
| 958 |
+
"unk_token": "<unk>"
|
| 959 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1312 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.23903648555278778,
|
| 3 |
+
"best_model_checkpoint": "/root/pretrain_executions/pretrain_utg4java_220m_seq1024/checkpoint-38422",
|
| 4 |
+
"epoch": 49.992029332058024,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 39200,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.3953451299218875,
|
| 13 |
+
"grad_norm": 0.6701709628105164,
|
| 14 |
+
"learning_rate": 7.908163265306123e-05,
|
| 15 |
+
"loss": 0.5421,
|
| 16 |
+
"step": 310
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.790690259843775,
|
| 20 |
+
"grad_norm": 0.7041985988616943,
|
| 21 |
+
"learning_rate": 0.00015816326530612246,
|
| 22 |
+
"loss": 0.517,
|
| 23 |
+
"step": 620
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.9998405866411605,
|
| 27 |
+
"eval_loss": 0.4792475402355194,
|
| 28 |
+
"eval_runtime": 146.6334,
|
| 29 |
+
"eval_samples_per_second": 85.567,
|
| 30 |
+
"eval_steps_per_second": 2.68,
|
| 31 |
+
"step": 784
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 1.1860353897656624,
|
| 35 |
+
"grad_norm": 0.7499191164970398,
|
| 36 |
+
"learning_rate": 0.00019924510620574761,
|
| 37 |
+
"loss": 0.5394,
|
| 38 |
+
"step": 930
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 1.5813805196875497,
|
| 42 |
+
"grad_norm": 0.6494282484054565,
|
| 43 |
+
"learning_rate": 0.00019763119533527698,
|
| 44 |
+
"loss": 0.5327,
|
| 45 |
+
"step": 1240
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.9767256496094374,
|
| 49 |
+
"grad_norm": 0.5610156059265137,
|
| 50 |
+
"learning_rate": 0.00019601728446480633,
|
| 51 |
+
"loss": 0.518,
|
| 52 |
+
"step": 1550
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.9996811732823212,
|
| 56 |
+
"eval_loss": 0.45565617084503174,
|
| 57 |
+
"eval_runtime": 146.6648,
|
| 58 |
+
"eval_samples_per_second": 85.549,
|
| 59 |
+
"eval_steps_per_second": 2.68,
|
| 60 |
+
"step": 1568
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 2.3720707795313247,
|
| 64 |
+
"grad_norm": 0.6261674761772156,
|
| 65 |
+
"learning_rate": 0.00019440337359433573,
|
| 66 |
+
"loss": 0.5076,
|
| 67 |
+
"step": 1860
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 2.767415909453212,
|
| 71 |
+
"grad_norm": 0.5601200461387634,
|
| 72 |
+
"learning_rate": 0.00019278946272386507,
|
| 73 |
+
"loss": 0.4972,
|
| 74 |
+
"step": 2170
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 2.9995217599234816,
|
| 78 |
+
"eval_loss": 0.4429556429386139,
|
| 79 |
+
"eval_runtime": 146.6472,
|
| 80 |
+
"eval_samples_per_second": 85.559,
|
| 81 |
+
"eval_steps_per_second": 2.68,
|
| 82 |
+
"step": 2352
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 3.1627610393751,
|
| 86 |
+
"grad_norm": 0.6102643609046936,
|
| 87 |
+
"learning_rate": 0.00019117555185339441,
|
| 88 |
+
"loss": 0.4957,
|
| 89 |
+
"step": 2480
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 3.558106169296987,
|
| 93 |
+
"grad_norm": 0.6246281266212463,
|
| 94 |
+
"learning_rate": 0.00018956164098292379,
|
| 95 |
+
"loss": 0.4827,
|
| 96 |
+
"step": 2790
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 3.9534512992188744,
|
| 100 |
+
"grad_norm": 0.6781056523323059,
|
| 101 |
+
"learning_rate": 0.00018794773011245316,
|
| 102 |
+
"loss": 0.4736,
|
| 103 |
+
"step": 3100
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 3.999362346564642,
|
| 107 |
+
"eval_loss": 0.42357587814331055,
|
| 108 |
+
"eval_runtime": 146.7171,
|
| 109 |
+
"eval_samples_per_second": 85.518,
|
| 110 |
+
"eval_steps_per_second": 2.679,
|
| 111 |
+
"step": 3136
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 4.348796429140762,
|
| 115 |
+
"grad_norm": 0.5225201845169067,
|
| 116 |
+
"learning_rate": 0.00018633381924198253,
|
| 117 |
+
"loss": 0.4686,
|
| 118 |
+
"step": 3410
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 4.7441415590626494,
|
| 122 |
+
"grad_norm": 0.5367516279220581,
|
| 123 |
+
"learning_rate": 0.00018471990837151187,
|
| 124 |
+
"loss": 0.4575,
|
| 125 |
+
"step": 3720
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 4.999202933205803,
|
| 129 |
+
"eval_loss": 0.4168592095375061,
|
| 130 |
+
"eval_runtime": 146.7474,
|
| 131 |
+
"eval_samples_per_second": 85.501,
|
| 132 |
+
"eval_steps_per_second": 2.678,
|
| 133 |
+
"step": 3920
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 5.139486688984537,
|
| 137 |
+
"grad_norm": 0.5979415774345398,
|
| 138 |
+
"learning_rate": 0.00018310599750104124,
|
| 139 |
+
"loss": 0.4591,
|
| 140 |
+
"step": 4030
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 5.534831818906424,
|
| 144 |
+
"grad_norm": 0.6041168570518494,
|
| 145 |
+
"learning_rate": 0.0001814920866305706,
|
| 146 |
+
"loss": 0.4506,
|
| 147 |
+
"step": 4340
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 5.930176948828311,
|
| 151 |
+
"grad_norm": 0.5398473739624023,
|
| 152 |
+
"learning_rate": 0.00017987817576009998,
|
| 153 |
+
"loss": 0.4499,
|
| 154 |
+
"step": 4650
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 5.999043519846963,
|
| 158 |
+
"eval_loss": 0.3998393714427948,
|
| 159 |
+
"eval_runtime": 146.7067,
|
| 160 |
+
"eval_samples_per_second": 85.524,
|
| 161 |
+
"eval_steps_per_second": 2.679,
|
| 162 |
+
"step": 4704
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 6.3255220787502,
|
| 166 |
+
"grad_norm": 0.5446251630783081,
|
| 167 |
+
"learning_rate": 0.00017826426488962933,
|
| 168 |
+
"loss": 0.4444,
|
| 169 |
+
"step": 4960
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 6.720867208672087,
|
| 173 |
+
"grad_norm": 0.564083993434906,
|
| 174 |
+
"learning_rate": 0.00017665556018325697,
|
| 175 |
+
"loss": 0.4401,
|
| 176 |
+
"step": 5270
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 6.9988841064881235,
|
| 180 |
+
"eval_loss": 0.3943786323070526,
|
| 181 |
+
"eval_runtime": 146.7223,
|
| 182 |
+
"eval_samples_per_second": 85.515,
|
| 183 |
+
"eval_steps_per_second": 2.679,
|
| 184 |
+
"step": 5488
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 7.116212338593974,
|
| 188 |
+
"grad_norm": 0.5197238326072693,
|
| 189 |
+
"learning_rate": 0.00017504164931278634,
|
| 190 |
+
"loss": 0.4349,
|
| 191 |
+
"step": 5580
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 7.5115574685158615,
|
| 195 |
+
"grad_norm": 0.5063862204551697,
|
| 196 |
+
"learning_rate": 0.0001734277384423157,
|
| 197 |
+
"loss": 0.4274,
|
| 198 |
+
"step": 5890
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 7.906902598437749,
|
| 202 |
+
"grad_norm": 0.8238950371742249,
|
| 203 |
+
"learning_rate": 0.00017181382757184508,
|
| 204 |
+
"loss": 0.4275,
|
| 205 |
+
"step": 6200
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 8.0,
|
| 209 |
+
"eval_loss": 0.38801178336143494,
|
| 210 |
+
"eval_runtime": 146.7154,
|
| 211 |
+
"eval_samples_per_second": 85.519,
|
| 212 |
+
"eval_steps_per_second": 2.679,
|
| 213 |
+
"step": 6273
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 8.302247728359637,
|
| 217 |
+
"grad_norm": 0.4785802662372589,
|
| 218 |
+
"learning_rate": 0.00017019991670137442,
|
| 219 |
+
"loss": 0.4218,
|
| 220 |
+
"step": 6510
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 8.697592858281524,
|
| 224 |
+
"grad_norm": 0.5460196137428284,
|
| 225 |
+
"learning_rate": 0.0001685860058309038,
|
| 226 |
+
"loss": 0.4165,
|
| 227 |
+
"step": 6820
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 8.999840586641161,
|
| 231 |
+
"eval_loss": 0.3786679804325104,
|
| 232 |
+
"eval_runtime": 146.6889,
|
| 233 |
+
"eval_samples_per_second": 85.535,
|
| 234 |
+
"eval_steps_per_second": 2.679,
|
| 235 |
+
"step": 7057
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 9.092937988203412,
|
| 239 |
+
"grad_norm": 0.5532106161117554,
|
| 240 |
+
"learning_rate": 0.00016697209496043317,
|
| 241 |
+
"loss": 0.4147,
|
| 242 |
+
"step": 7130
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 9.488283118125299,
|
| 246 |
+
"grad_norm": 0.5270036458969116,
|
| 247 |
+
"learning_rate": 0.00016535818408996254,
|
| 248 |
+
"loss": 0.4054,
|
| 249 |
+
"step": 7440
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 9.883628248047186,
|
| 253 |
+
"grad_norm": 0.5107512474060059,
|
| 254 |
+
"learning_rate": 0.00016374427321949188,
|
| 255 |
+
"loss": 0.407,
|
| 256 |
+
"step": 7750
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 9.99968117328232,
|
| 260 |
+
"eval_loss": 0.3678111732006073,
|
| 261 |
+
"eval_runtime": 146.7168,
|
| 262 |
+
"eval_samples_per_second": 85.519,
|
| 263 |
+
"eval_steps_per_second": 2.679,
|
| 264 |
+
"step": 7841
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 10.278973377969074,
|
| 268 |
+
"grad_norm": 0.4663056433200836,
|
| 269 |
+
"learning_rate": 0.00016213036234902125,
|
| 270 |
+
"loss": 0.4001,
|
| 271 |
+
"step": 8060
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 10.67431850789096,
|
| 275 |
+
"grad_norm": 0.5166866183280945,
|
| 276 |
+
"learning_rate": 0.00016051645147855062,
|
| 277 |
+
"loss": 0.4012,
|
| 278 |
+
"step": 8370
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 10.999521759923482,
|
| 282 |
+
"eval_loss": 0.36857831478118896,
|
| 283 |
+
"eval_runtime": 146.752,
|
| 284 |
+
"eval_samples_per_second": 85.498,
|
| 285 |
+
"eval_steps_per_second": 2.678,
|
| 286 |
+
"step": 8625
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 11.069663637812848,
|
| 290 |
+
"grad_norm": 0.5623896718025208,
|
| 291 |
+
"learning_rate": 0.00015890254060807997,
|
| 292 |
+
"loss": 0.3967,
|
| 293 |
+
"step": 8680
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 11.465008767734735,
|
| 297 |
+
"grad_norm": 0.4826233983039856,
|
| 298 |
+
"learning_rate": 0.00015728862973760934,
|
| 299 |
+
"loss": 0.3902,
|
| 300 |
+
"step": 8990
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 11.860353897656623,
|
| 304 |
+
"grad_norm": 0.5024587512016296,
|
| 305 |
+
"learning_rate": 0.00015567471886713868,
|
| 306 |
+
"loss": 0.3889,
|
| 307 |
+
"step": 9300
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 11.999362346564642,
|
| 311 |
+
"eval_loss": 0.35674843192100525,
|
| 312 |
+
"eval_runtime": 146.7093,
|
| 313 |
+
"eval_samples_per_second": 85.523,
|
| 314 |
+
"eval_steps_per_second": 2.679,
|
| 315 |
+
"step": 9409
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 12.255699027578512,
|
| 319 |
+
"grad_norm": 0.4992258846759796,
|
| 320 |
+
"learning_rate": 0.00015406080799666805,
|
| 321 |
+
"loss": 0.3838,
|
| 322 |
+
"step": 9610
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 12.6510441575004,
|
| 326 |
+
"grad_norm": 0.4781612455844879,
|
| 327 |
+
"learning_rate": 0.00015244689712619742,
|
| 328 |
+
"loss": 0.3789,
|
| 329 |
+
"step": 9920
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 12.999202933205803,
|
| 333 |
+
"eval_loss": 0.35254529118537903,
|
| 334 |
+
"eval_runtime": 146.7424,
|
| 335 |
+
"eval_samples_per_second": 85.504,
|
| 336 |
+
"eval_steps_per_second": 2.678,
|
| 337 |
+
"step": 10193
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 13.046389287422286,
|
| 341 |
+
"grad_norm": 0.49535173177719116,
|
| 342 |
+
"learning_rate": 0.0001508329862557268,
|
| 343 |
+
"loss": 0.3775,
|
| 344 |
+
"step": 10230
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 13.441734417344174,
|
| 348 |
+
"grad_norm": 0.5237115621566772,
|
| 349 |
+
"learning_rate": 0.00014922428154935443,
|
| 350 |
+
"loss": 0.3734,
|
| 351 |
+
"step": 10540
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 13.837079547266061,
|
| 355 |
+
"grad_norm": 0.4549529552459717,
|
| 356 |
+
"learning_rate": 0.0001476155768429821,
|
| 357 |
+
"loss": 0.37,
|
| 358 |
+
"step": 10850
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 13.999043519846964,
|
| 362 |
+
"eval_loss": 0.3443816006183624,
|
| 363 |
+
"eval_runtime": 146.6954,
|
| 364 |
+
"eval_samples_per_second": 85.531,
|
| 365 |
+
"eval_steps_per_second": 2.679,
|
| 366 |
+
"step": 10977
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 14.232424677187948,
|
| 370 |
+
"grad_norm": 0.520125150680542,
|
| 371 |
+
"learning_rate": 0.00014600166597251147,
|
| 372 |
+
"loss": 0.3647,
|
| 373 |
+
"step": 11160
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 14.627769807109836,
|
| 377 |
+
"grad_norm": 0.5332316160202026,
|
| 378 |
+
"learning_rate": 0.00014438775510204084,
|
| 379 |
+
"loss": 0.3678,
|
| 380 |
+
"step": 11470
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 14.998884106488124,
|
| 384 |
+
"eval_loss": 0.3436979055404663,
|
| 385 |
+
"eval_runtime": 146.7179,
|
| 386 |
+
"eval_samples_per_second": 85.518,
|
| 387 |
+
"eval_steps_per_second": 2.679,
|
| 388 |
+
"step": 11761
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 15.023114937031723,
|
| 392 |
+
"grad_norm": 0.47955256700515747,
|
| 393 |
+
"learning_rate": 0.0001427738442315702,
|
| 394 |
+
"loss": 0.3664,
|
| 395 |
+
"step": 11780
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 15.41846006695361,
|
| 399 |
+
"grad_norm": 0.48371678590774536,
|
| 400 |
+
"learning_rate": 0.00014115993336109953,
|
| 401 |
+
"loss": 0.36,
|
| 402 |
+
"step": 12090
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 15.813805196875498,
|
| 406 |
+
"grad_norm": 0.4756961166858673,
|
| 407 |
+
"learning_rate": 0.0001395460224906289,
|
| 408 |
+
"loss": 0.3577,
|
| 409 |
+
"step": 12400
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 16.0,
|
| 413 |
+
"eval_loss": 0.3342459499835968,
|
| 414 |
+
"eval_runtime": 146.7158,
|
| 415 |
+
"eval_samples_per_second": 85.519,
|
| 416 |
+
"eval_steps_per_second": 2.679,
|
| 417 |
+
"step": 12546
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 16.209150326797385,
|
| 421 |
+
"grad_norm": 0.5593659281730652,
|
| 422 |
+
"learning_rate": 0.00013793211162015827,
|
| 423 |
+
"loss": 0.3536,
|
| 424 |
+
"step": 12710
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 16.604495456719274,
|
| 428 |
+
"grad_norm": 0.6009001135826111,
|
| 429 |
+
"learning_rate": 0.00013631820074968764,
|
| 430 |
+
"loss": 0.3539,
|
| 431 |
+
"step": 13020
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 16.99984058664116,
|
| 435 |
+
"grad_norm": 0.48707565665245056,
|
| 436 |
+
"learning_rate": 0.000134704289879217,
|
| 437 |
+
"loss": 0.3522,
|
| 438 |
+
"step": 13330
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 16.99984058664116,
|
| 442 |
+
"eval_loss": 0.33101820945739746,
|
| 443 |
+
"eval_runtime": 146.73,
|
| 444 |
+
"eval_samples_per_second": 85.511,
|
| 445 |
+
"eval_steps_per_second": 2.678,
|
| 446 |
+
"step": 13330
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 17.39518571656305,
|
| 450 |
+
"grad_norm": 0.4787095785140991,
|
| 451 |
+
"learning_rate": 0.00013309037900874636,
|
| 452 |
+
"loss": 0.343,
|
| 453 |
+
"step": 13640
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 17.790530846484934,
|
| 457 |
+
"grad_norm": 0.4643840789794922,
|
| 458 |
+
"learning_rate": 0.00013147646813827573,
|
| 459 |
+
"loss": 0.3466,
|
| 460 |
+
"step": 13950
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 17.999681173282323,
|
| 464 |
+
"eval_loss": 0.3281005620956421,
|
| 465 |
+
"eval_runtime": 146.7108,
|
| 466 |
+
"eval_samples_per_second": 85.522,
|
| 467 |
+
"eval_steps_per_second": 2.679,
|
| 468 |
+
"step": 14114
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 18.185875976406823,
|
| 472 |
+
"grad_norm": 0.4819445312023163,
|
| 473 |
+
"learning_rate": 0.0001298625572678051,
|
| 474 |
+
"loss": 0.3415,
|
| 475 |
+
"step": 14260
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 18.58122110632871,
|
| 479 |
+
"grad_norm": 0.46530964970588684,
|
| 480 |
+
"learning_rate": 0.00012824864639733444,
|
| 481 |
+
"loss": 0.3393,
|
| 482 |
+
"step": 14570
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 18.976566236250598,
|
| 486 |
+
"grad_norm": 0.5159475207328796,
|
| 487 |
+
"learning_rate": 0.00012663473552686382,
|
| 488 |
+
"loss": 0.3377,
|
| 489 |
+
"step": 14880
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 18.999521759923482,
|
| 493 |
+
"eval_loss": 0.32132235169410706,
|
| 494 |
+
"eval_runtime": 146.7396,
|
| 495 |
+
"eval_samples_per_second": 85.505,
|
| 496 |
+
"eval_steps_per_second": 2.678,
|
| 497 |
+
"step": 14898
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 19.371911366172487,
|
| 501 |
+
"grad_norm": 0.45964986085891724,
|
| 502 |
+
"learning_rate": 0.00012502082465639319,
|
| 503 |
+
"loss": 0.3348,
|
| 504 |
+
"step": 15190
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 19.767256496094372,
|
| 508 |
+
"grad_norm": 0.49627387523651123,
|
| 509 |
+
"learning_rate": 0.00012340691378592253,
|
| 510 |
+
"loss": 0.3316,
|
| 511 |
+
"step": 15500
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 19.99936234656464,
|
| 515 |
+
"eval_loss": 0.31626757979393005,
|
| 516 |
+
"eval_runtime": 146.7396,
|
| 517 |
+
"eval_samples_per_second": 85.505,
|
| 518 |
+
"eval_steps_per_second": 2.678,
|
| 519 |
+
"step": 15682
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 20.16260162601626,
|
| 523 |
+
"grad_norm": 0.48719242215156555,
|
| 524 |
+
"learning_rate": 0.0001217930029154519,
|
| 525 |
+
"loss": 0.3294,
|
| 526 |
+
"step": 15810
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 20.557946755938147,
|
| 530 |
+
"grad_norm": 0.5443927049636841,
|
| 531 |
+
"learning_rate": 0.00012017909204498126,
|
| 532 |
+
"loss": 0.3261,
|
| 533 |
+
"step": 16120
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 20.953291885860036,
|
| 537 |
+
"grad_norm": 0.4637634754180908,
|
| 538 |
+
"learning_rate": 0.00011856518117451063,
|
| 539 |
+
"loss": 0.3255,
|
| 540 |
+
"step": 16430
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 20.9992029332058,
|
| 544 |
+
"eval_loss": 0.31501948833465576,
|
| 545 |
+
"eval_runtime": 146.7591,
|
| 546 |
+
"eval_samples_per_second": 85.494,
|
| 547 |
+
"eval_steps_per_second": 2.678,
|
| 548 |
+
"step": 16466
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 21.34863701578192,
|
| 552 |
+
"grad_norm": 0.46018585562705994,
|
| 553 |
+
"learning_rate": 0.00011695127030403999,
|
| 554 |
+
"loss": 0.3198,
|
| 555 |
+
"step": 16740
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 21.74398214570381,
|
| 559 |
+
"grad_norm": 0.5096014738082886,
|
| 560 |
+
"learning_rate": 0.00011533735943356936,
|
| 561 |
+
"loss": 0.3226,
|
| 562 |
+
"step": 17050
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 21.999043519846964,
|
| 566 |
+
"eval_loss": 0.30657365918159485,
|
| 567 |
+
"eval_runtime": 146.7538,
|
| 568 |
+
"eval_samples_per_second": 85.497,
|
| 569 |
+
"eval_steps_per_second": 2.678,
|
| 570 |
+
"step": 17250
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 22.139327275625696,
|
| 574 |
+
"grad_norm": 0.44816407561302185,
|
| 575 |
+
"learning_rate": 0.00011372344856309872,
|
| 576 |
+
"loss": 0.3178,
|
| 577 |
+
"step": 17360
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 22.534672405547585,
|
| 581 |
+
"grad_norm": 0.437168151140213,
|
| 582 |
+
"learning_rate": 0.00011211474385672638,
|
| 583 |
+
"loss": 0.3172,
|
| 584 |
+
"step": 17670
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 22.93001753546947,
|
| 588 |
+
"grad_norm": 0.5836613774299622,
|
| 589 |
+
"learning_rate": 0.00011050083298625573,
|
| 590 |
+
"loss": 0.3121,
|
| 591 |
+
"step": 17980
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 22.998884106488124,
|
| 595 |
+
"eval_loss": 0.30263882875442505,
|
| 596 |
+
"eval_runtime": 146.7108,
|
| 597 |
+
"eval_samples_per_second": 85.522,
|
| 598 |
+
"eval_steps_per_second": 2.679,
|
| 599 |
+
"step": 18034
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 23.32536266539136,
|
| 603 |
+
"grad_norm": 0.4829230308532715,
|
| 604 |
+
"learning_rate": 0.00010888692211578508,
|
| 605 |
+
"loss": 0.3079,
|
| 606 |
+
"step": 18290
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 23.720707795313245,
|
| 610 |
+
"grad_norm": 0.4485584497451782,
|
| 611 |
+
"learning_rate": 0.00010727821740941275,
|
| 612 |
+
"loss": 0.3105,
|
| 613 |
+
"step": 18600
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 24.0,
|
| 617 |
+
"eval_loss": 0.3048921227455139,
|
| 618 |
+
"eval_runtime": 146.8103,
|
| 619 |
+
"eval_samples_per_second": 85.464,
|
| 620 |
+
"eval_steps_per_second": 2.677,
|
| 621 |
+
"step": 18819
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 24.116052925235135,
|
| 625 |
+
"grad_norm": 0.5251662135124207,
|
| 626 |
+
"learning_rate": 0.00010566430653894211,
|
| 627 |
+
"loss": 0.3052,
|
| 628 |
+
"step": 18910
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 24.511398055157024,
|
| 632 |
+
"grad_norm": 0.4876725971698761,
|
| 633 |
+
"learning_rate": 0.00010405039566847148,
|
| 634 |
+
"loss": 0.3045,
|
| 635 |
+
"step": 19220
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 24.90674318507891,
|
| 639 |
+
"grad_norm": 0.5600521564483643,
|
| 640 |
+
"learning_rate": 0.00010243648479800084,
|
| 641 |
+
"loss": 0.3048,
|
| 642 |
+
"step": 19530
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 24.99984058664116,
|
| 646 |
+
"eval_loss": 0.2986990809440613,
|
| 647 |
+
"eval_runtime": 146.734,
|
| 648 |
+
"eval_samples_per_second": 85.508,
|
| 649 |
+
"eval_steps_per_second": 2.678,
|
| 650 |
+
"step": 19603
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 25.3020883150008,
|
| 654 |
+
"grad_norm": 0.5170055627822876,
|
| 655 |
+
"learning_rate": 0.00010082257392753021,
|
| 656 |
+
"loss": 0.3003,
|
| 657 |
+
"step": 19840
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 25.697433444922684,
|
| 661 |
+
"grad_norm": 0.48347124457359314,
|
| 662 |
+
"learning_rate": 9.920866305705956e-05,
|
| 663 |
+
"loss": 0.2983,
|
| 664 |
+
"step": 20150
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 25.999681173282323,
|
| 668 |
+
"eval_loss": 0.2916134297847748,
|
| 669 |
+
"eval_runtime": 146.7271,
|
| 670 |
+
"eval_samples_per_second": 85.512,
|
| 671 |
+
"eval_steps_per_second": 2.678,
|
| 672 |
+
"step": 20387
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 26.092778574844573,
|
| 676 |
+
"grad_norm": 0.48907041549682617,
|
| 677 |
+
"learning_rate": 9.759475218658892e-05,
|
| 678 |
+
"loss": 0.2959,
|
| 679 |
+
"step": 20460
|
| 680 |
+
},
|
| 681 |
+
{
|
| 682 |
+
"epoch": 26.48812370476646,
|
| 683 |
+
"grad_norm": 0.5060804486274719,
|
| 684 |
+
"learning_rate": 9.598084131611829e-05,
|
| 685 |
+
"loss": 0.2923,
|
| 686 |
+
"step": 20770
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 26.883468834688347,
|
| 690 |
+
"grad_norm": 0.4843296706676483,
|
| 691 |
+
"learning_rate": 9.436693044564765e-05,
|
| 692 |
+
"loss": 0.2918,
|
| 693 |
+
"step": 21080
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 26.999521759923482,
|
| 697 |
+
"eval_loss": 0.29019656777381897,
|
| 698 |
+
"eval_runtime": 146.6934,
|
| 699 |
+
"eval_samples_per_second": 85.532,
|
| 700 |
+
"eval_steps_per_second": 2.679,
|
| 701 |
+
"step": 21171
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 27.278813964610233,
|
| 705 |
+
"grad_norm": 0.42266514897346497,
|
| 706 |
+
"learning_rate": 9.275301957517701e-05,
|
| 707 |
+
"loss": 0.2901,
|
| 708 |
+
"step": 21390
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 27.674159094532122,
|
| 712 |
+
"grad_norm": 0.5161967873573303,
|
| 713 |
+
"learning_rate": 9.113910870470638e-05,
|
| 714 |
+
"loss": 0.2889,
|
| 715 |
+
"step": 21700
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 27.99936234656464,
|
| 719 |
+
"eval_loss": 0.2833983302116394,
|
| 720 |
+
"eval_runtime": 146.7193,
|
| 721 |
+
"eval_samples_per_second": 85.517,
|
| 722 |
+
"eval_steps_per_second": 2.679,
|
| 723 |
+
"step": 21955
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 28.069504224454008,
|
| 727 |
+
"grad_norm": 0.4523755609989166,
|
| 728 |
+
"learning_rate": 8.952519783423574e-05,
|
| 729 |
+
"loss": 0.2871,
|
| 730 |
+
"step": 22010
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 28.464849354375897,
|
| 734 |
+
"grad_norm": 0.44348961114883423,
|
| 735 |
+
"learning_rate": 8.791128696376511e-05,
|
| 736 |
+
"loss": 0.2847,
|
| 737 |
+
"step": 22320
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 28.860194484297786,
|
| 741 |
+
"grad_norm": 0.6467667818069458,
|
| 742 |
+
"learning_rate": 8.630258225739276e-05,
|
| 743 |
+
"loss": 0.2844,
|
| 744 |
+
"step": 22630
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 28.9992029332058,
|
| 748 |
+
"eval_loss": 0.28629302978515625,
|
| 749 |
+
"eval_runtime": 146.7547,
|
| 750 |
+
"eval_samples_per_second": 85.496,
|
| 751 |
+
"eval_steps_per_second": 2.678,
|
| 752 |
+
"step": 22739
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 29.25553961421967,
|
| 756 |
+
"grad_norm": 0.4734992980957031,
|
| 757 |
+
"learning_rate": 8.468867138692213e-05,
|
| 758 |
+
"loss": 0.2787,
|
| 759 |
+
"step": 22940
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 29.65088474414156,
|
| 763 |
+
"grad_norm": 0.4827498495578766,
|
| 764 |
+
"learning_rate": 8.307476051645148e-05,
|
| 765 |
+
"loss": 0.2787,
|
| 766 |
+
"step": 23250
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 29.999043519846964,
|
| 770 |
+
"eval_loss": 0.2794826626777649,
|
| 771 |
+
"eval_runtime": 146.9198,
|
| 772 |
+
"eval_samples_per_second": 85.4,
|
| 773 |
+
"eval_steps_per_second": 2.675,
|
| 774 |
+
"step": 23523
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 30.046229874063446,
|
| 778 |
+
"grad_norm": 0.5005486607551575,
|
| 779 |
+
"learning_rate": 8.146084964598085e-05,
|
| 780 |
+
"loss": 0.2758,
|
| 781 |
+
"step": 23560
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 30.441575003985335,
|
| 785 |
+
"grad_norm": 0.5253671407699585,
|
| 786 |
+
"learning_rate": 7.98469387755102e-05,
|
| 787 |
+
"loss": 0.2761,
|
| 788 |
+
"step": 23870
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 30.83692013390722,
|
| 792 |
+
"grad_norm": 0.472740113735199,
|
| 793 |
+
"learning_rate": 7.823302790503957e-05,
|
| 794 |
+
"loss": 0.2726,
|
| 795 |
+
"step": 24180
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 30.998884106488124,
|
| 799 |
+
"eval_loss": 0.2779182493686676,
|
| 800 |
+
"eval_runtime": 146.7777,
|
| 801 |
+
"eval_samples_per_second": 85.483,
|
| 802 |
+
"eval_steps_per_second": 2.678,
|
| 803 |
+
"step": 24307
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 31.23226526382911,
|
| 807 |
+
"grad_norm": 0.5228144526481628,
|
| 808 |
+
"learning_rate": 7.661911703456893e-05,
|
| 809 |
+
"loss": 0.2717,
|
| 810 |
+
"step": 24490
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 31.627610393750995,
|
| 814 |
+
"grad_norm": 0.47681719064712524,
|
| 815 |
+
"learning_rate": 7.501041232819659e-05,
|
| 816 |
+
"loss": 0.2664,
|
| 817 |
+
"step": 24800
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 32.0,
|
| 821 |
+
"eval_loss": 0.27039337158203125,
|
| 822 |
+
"eval_runtime": 146.805,
|
| 823 |
+
"eval_samples_per_second": 85.467,
|
| 824 |
+
"eval_steps_per_second": 2.677,
|
| 825 |
+
"step": 25092
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 32.022955523672884,
|
| 829 |
+
"grad_norm": 0.4973162114620209,
|
| 830 |
+
"learning_rate": 7.339650145772596e-05,
|
| 831 |
+
"loss": 0.268,
|
| 832 |
+
"step": 25110
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 32.41830065359477,
|
| 836 |
+
"grad_norm": 0.5740240216255188,
|
| 837 |
+
"learning_rate": 7.178259058725531e-05,
|
| 838 |
+
"loss": 0.2668,
|
| 839 |
+
"step": 25420
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 32.813645783516655,
|
| 843 |
+
"grad_norm": 0.4842962622642517,
|
| 844 |
+
"learning_rate": 7.016867971678468e-05,
|
| 845 |
+
"loss": 0.2631,
|
| 846 |
+
"step": 25730
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 32.99984058664116,
|
| 850 |
+
"eval_loss": 0.2733234763145447,
|
| 851 |
+
"eval_runtime": 146.7109,
|
| 852 |
+
"eval_samples_per_second": 85.522,
|
| 853 |
+
"eval_steps_per_second": 2.679,
|
| 854 |
+
"step": 25876
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 33.20899091343855,
|
| 858 |
+
"grad_norm": 0.499452143907547,
|
| 859 |
+
"learning_rate": 6.855476884631404e-05,
|
| 860 |
+
"loss": 0.263,
|
| 861 |
+
"step": 26040
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 33.60433604336043,
|
| 865 |
+
"grad_norm": 0.4541178345680237,
|
| 866 |
+
"learning_rate": 6.69408579758434e-05,
|
| 867 |
+
"loss": 0.2603,
|
| 868 |
+
"step": 26350
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 33.99968117328232,
|
| 872 |
+
"grad_norm": 0.5029833912849426,
|
| 873 |
+
"learning_rate": 6.532694710537276e-05,
|
| 874 |
+
"loss": 0.258,
|
| 875 |
+
"step": 26660
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 33.99968117328232,
|
| 879 |
+
"eval_loss": 0.26625362038612366,
|
| 880 |
+
"eval_runtime": 146.7319,
|
| 881 |
+
"eval_samples_per_second": 85.51,
|
| 882 |
+
"eval_steps_per_second": 2.678,
|
| 883 |
+
"step": 26660
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 34.39502630320421,
|
| 887 |
+
"grad_norm": 0.5090352892875671,
|
| 888 |
+
"learning_rate": 6.371303623490213e-05,
|
| 889 |
+
"loss": 0.2544,
|
| 890 |
+
"step": 26970
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 34.7903714331261,
|
| 894 |
+
"grad_norm": 0.4605717360973358,
|
| 895 |
+
"learning_rate": 6.209912536443149e-05,
|
| 896 |
+
"loss": 0.254,
|
| 897 |
+
"step": 27280
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 34.99952175992348,
|
| 901 |
+
"eval_loss": 0.26669949293136597,
|
| 902 |
+
"eval_runtime": 146.7117,
|
| 903 |
+
"eval_samples_per_second": 85.521,
|
| 904 |
+
"eval_steps_per_second": 2.679,
|
| 905 |
+
"step": 27444
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 35.18571656304798,
|
| 909 |
+
"grad_norm": 0.46216222643852234,
|
| 910 |
+
"learning_rate": 6.048521449396085e-05,
|
| 911 |
+
"loss": 0.254,
|
| 912 |
+
"step": 27590
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 35.58106169296987,
|
| 916 |
+
"grad_norm": 0.49629315733909607,
|
| 917 |
+
"learning_rate": 5.8871303623490214e-05,
|
| 918 |
+
"loss": 0.2521,
|
| 919 |
+
"step": 27900
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 35.97640682289176,
|
| 923 |
+
"grad_norm": 0.48311081528663635,
|
| 924 |
+
"learning_rate": 5.725739275301958e-05,
|
| 925 |
+
"loss": 0.2493,
|
| 926 |
+
"step": 28210
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 35.999362346564645,
|
| 930 |
+
"eval_loss": 0.26483407616615295,
|
| 931 |
+
"eval_runtime": 146.7384,
|
| 932 |
+
"eval_samples_per_second": 85.506,
|
| 933 |
+
"eval_steps_per_second": 2.678,
|
| 934 |
+
"step": 28228
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 36.371751952813646,
|
| 938 |
+
"grad_norm": 0.43428850173950195,
|
| 939 |
+
"learning_rate": 5.564348188254894e-05,
|
| 940 |
+
"loss": 0.2455,
|
| 941 |
+
"step": 28520
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 36.76709708273553,
|
| 945 |
+
"grad_norm": 0.4786287844181061,
|
| 946 |
+
"learning_rate": 5.4029571012078306e-05,
|
| 947 |
+
"loss": 0.2454,
|
| 948 |
+
"step": 28830
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 36.9992029332058,
|
| 952 |
+
"eval_loss": 0.26446378231048584,
|
| 953 |
+
"eval_runtime": 146.73,
|
| 954 |
+
"eval_samples_per_second": 85.511,
|
| 955 |
+
"eval_steps_per_second": 2.678,
|
| 956 |
+
"step": 29012
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 37.16244221265742,
|
| 960 |
+
"grad_norm": 0.5931326746940613,
|
| 961 |
+
"learning_rate": 5.241566014160767e-05,
|
| 962 |
+
"loss": 0.247,
|
| 963 |
+
"step": 29140
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 37.55778734257931,
|
| 967 |
+
"grad_norm": 0.5031745433807373,
|
| 968 |
+
"learning_rate": 5.0801749271137035e-05,
|
| 969 |
+
"loss": 0.2425,
|
| 970 |
+
"step": 29450
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 37.953132472501196,
|
| 974 |
+
"grad_norm": 0.5432093739509583,
|
| 975 |
+
"learning_rate": 4.918783840066639e-05,
|
| 976 |
+
"loss": 0.2416,
|
| 977 |
+
"step": 29760
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 37.999043519846964,
|
| 981 |
+
"eval_loss": 0.2601180672645569,
|
| 982 |
+
"eval_runtime": 146.6811,
|
| 983 |
+
"eval_samples_per_second": 85.539,
|
| 984 |
+
"eval_steps_per_second": 2.679,
|
| 985 |
+
"step": 29796
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 38.34847760242308,
|
| 989 |
+
"grad_norm": 0.5319362878799438,
|
| 990 |
+
"learning_rate": 4.7573927530195756e-05,
|
| 991 |
+
"loss": 0.2392,
|
| 992 |
+
"step": 30070
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 38.743822732344974,
|
| 996 |
+
"grad_norm": 0.5319586396217346,
|
| 997 |
+
"learning_rate": 4.596001665972512e-05,
|
| 998 |
+
"loss": 0.2368,
|
| 999 |
+
"step": 30380
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 38.99888410648813,
|
| 1003 |
+
"eval_loss": 0.25446435809135437,
|
| 1004 |
+
"eval_runtime": 146.6972,
|
| 1005 |
+
"eval_samples_per_second": 85.53,
|
| 1006 |
+
"eval_steps_per_second": 2.679,
|
| 1007 |
+
"step": 30580
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 39.13916786226686,
|
| 1011 |
+
"grad_norm": 0.4489250183105469,
|
| 1012 |
+
"learning_rate": 4.434610578925448e-05,
|
| 1013 |
+
"loss": 0.2368,
|
| 1014 |
+
"step": 30690
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 39.534512992188745,
|
| 1018 |
+
"grad_norm": 0.48287880420684814,
|
| 1019 |
+
"learning_rate": 4.273740108288213e-05,
|
| 1020 |
+
"loss": 0.2353,
|
| 1021 |
+
"step": 31000
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 39.92985812211063,
|
| 1025 |
+
"grad_norm": 0.49850553274154663,
|
| 1026 |
+
"learning_rate": 4.1123490212411495e-05,
|
| 1027 |
+
"loss": 0.2321,
|
| 1028 |
+
"step": 31310
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 40.0,
|
| 1032 |
+
"eval_loss": 0.24883659183979034,
|
| 1033 |
+
"eval_runtime": 146.7363,
|
| 1034 |
+
"eval_samples_per_second": 85.507,
|
| 1035 |
+
"eval_steps_per_second": 2.678,
|
| 1036 |
+
"step": 31365
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 40.32520325203252,
|
| 1040 |
+
"grad_norm": 0.4667394161224365,
|
| 1041 |
+
"learning_rate": 3.9514785506039155e-05,
|
| 1042 |
+
"loss": 0.2337,
|
| 1043 |
+
"step": 31620
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 40.72054838195441,
|
| 1047 |
+
"grad_norm": 0.5053902864456177,
|
| 1048 |
+
"learning_rate": 3.790087463556852e-05,
|
| 1049 |
+
"loss": 0.2284,
|
| 1050 |
+
"step": 31930
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 40.99984058664116,
|
| 1054 |
+
"eval_loss": 0.2544113099575043,
|
| 1055 |
+
"eval_runtime": 146.7257,
|
| 1056 |
+
"eval_samples_per_second": 85.513,
|
| 1057 |
+
"eval_steps_per_second": 2.678,
|
| 1058 |
+
"step": 32149
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 41.115893511876294,
|
| 1062 |
+
"grad_norm": 0.47476327419281006,
|
| 1063 |
+
"learning_rate": 3.628696376509788e-05,
|
| 1064 |
+
"loss": 0.2286,
|
| 1065 |
+
"step": 32240
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 41.51123864179818,
|
| 1069 |
+
"grad_norm": 0.5025794506072998,
|
| 1070 |
+
"learning_rate": 3.467305289462724e-05,
|
| 1071 |
+
"loss": 0.2292,
|
| 1072 |
+
"step": 32550
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 41.90658377172007,
|
| 1076 |
+
"grad_norm": 0.4553293287754059,
|
| 1077 |
+
"learning_rate": 3.3059142024156605e-05,
|
| 1078 |
+
"loss": 0.225,
|
| 1079 |
+
"step": 32860
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 41.99968117328232,
|
| 1083 |
+
"eval_loss": 0.24568869173526764,
|
| 1084 |
+
"eval_runtime": 146.7316,
|
| 1085 |
+
"eval_samples_per_second": 85.51,
|
| 1086 |
+
"eval_steps_per_second": 2.678,
|
| 1087 |
+
"step": 32933
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 42.30192890164196,
|
| 1091 |
+
"grad_norm": 0.4845215678215027,
|
| 1092 |
+
"learning_rate": 3.144523115368597e-05,
|
| 1093 |
+
"loss": 0.2236,
|
| 1094 |
+
"step": 33170
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 42.69727403156384,
|
| 1098 |
+
"grad_norm": 0.5739601850509644,
|
| 1099 |
+
"learning_rate": 2.983132028321533e-05,
|
| 1100 |
+
"loss": 0.2234,
|
| 1101 |
+
"step": 33480
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 42.99952175992348,
|
| 1105 |
+
"eval_loss": 0.24620206654071808,
|
| 1106 |
+
"eval_runtime": 146.7264,
|
| 1107 |
+
"eval_samples_per_second": 85.513,
|
| 1108 |
+
"eval_steps_per_second": 2.678,
|
| 1109 |
+
"step": 33717
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 43.092619161485736,
|
| 1113 |
+
"grad_norm": 0.4569677412509918,
|
| 1114 |
+
"learning_rate": 2.8217409412744688e-05,
|
| 1115 |
+
"loss": 0.2213,
|
| 1116 |
+
"step": 33790
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 43.48796429140762,
|
| 1120 |
+
"grad_norm": 0.5146024227142334,
|
| 1121 |
+
"learning_rate": 2.6603498542274052e-05,
|
| 1122 |
+
"loss": 0.2188,
|
| 1123 |
+
"step": 34100
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 43.88330942132951,
|
| 1127 |
+
"grad_norm": 0.47475871443748474,
|
| 1128 |
+
"learning_rate": 2.4989587671803416e-05,
|
| 1129 |
+
"loss": 0.2206,
|
| 1130 |
+
"step": 34410
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 43.999362346564645,
|
| 1134 |
+
"eval_loss": 0.2445935159921646,
|
| 1135 |
+
"eval_runtime": 146.7897,
|
| 1136 |
+
"eval_samples_per_second": 85.476,
|
| 1137 |
+
"eval_steps_per_second": 2.677,
|
| 1138 |
+
"step": 34501
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 44.27865455125139,
|
| 1142 |
+
"grad_norm": 0.45915085077285767,
|
| 1143 |
+
"learning_rate": 2.337567680133278e-05,
|
| 1144 |
+
"loss": 0.217,
|
| 1145 |
+
"step": 34720
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 44.673999681173285,
|
| 1149 |
+
"grad_norm": 0.4429190456867218,
|
| 1150 |
+
"learning_rate": 2.176176593086214e-05,
|
| 1151 |
+
"loss": 0.2165,
|
| 1152 |
+
"step": 35030
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 44.9992029332058,
|
| 1156 |
+
"eval_loss": 0.24302400648593903,
|
| 1157 |
+
"eval_runtime": 146.7631,
|
| 1158 |
+
"eval_samples_per_second": 85.492,
|
| 1159 |
+
"eval_steps_per_second": 2.678,
|
| 1160 |
+
"step": 35285
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 45.06934481109517,
|
| 1164 |
+
"grad_norm": 0.5038246512413025,
|
| 1165 |
+
"learning_rate": 2.0147855060391505e-05,
|
| 1166 |
+
"loss": 0.217,
|
| 1167 |
+
"step": 35340
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 45.464689941017056,
|
| 1171 |
+
"grad_norm": 0.4302615523338318,
|
| 1172 |
+
"learning_rate": 1.8539150354019162e-05,
|
| 1173 |
+
"loss": 0.2137,
|
| 1174 |
+
"step": 35650
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 45.86003507093894,
|
| 1178 |
+
"grad_norm": 0.5075607299804688,
|
| 1179 |
+
"learning_rate": 1.6925239483548523e-05,
|
| 1180 |
+
"loss": 0.2145,
|
| 1181 |
+
"step": 35960
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 45.999043519846964,
|
| 1185 |
+
"eval_loss": 0.24222899973392487,
|
| 1186 |
+
"eval_runtime": 146.735,
|
| 1187 |
+
"eval_samples_per_second": 85.508,
|
| 1188 |
+
"eval_steps_per_second": 2.678,
|
| 1189 |
+
"step": 36069
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 46.255380200860834,
|
| 1193 |
+
"grad_norm": 0.4777955114841461,
|
| 1194 |
+
"learning_rate": 1.531653477717618e-05,
|
| 1195 |
+
"loss": 0.2126,
|
| 1196 |
+
"step": 36270
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 46.65072533078272,
|
| 1200 |
+
"grad_norm": 0.48974084854125977,
|
| 1201 |
+
"learning_rate": 1.3702623906705539e-05,
|
| 1202 |
+
"loss": 0.2112,
|
| 1203 |
+
"step": 36580
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 46.99888410648813,
|
| 1207 |
+
"eval_loss": 0.2432757019996643,
|
| 1208 |
+
"eval_runtime": 146.7494,
|
| 1209 |
+
"eval_samples_per_second": 85.499,
|
| 1210 |
+
"eval_steps_per_second": 2.678,
|
| 1211 |
+
"step": 36853
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 47.046070460704605,
|
| 1215 |
+
"grad_norm": 0.46624037623405457,
|
| 1216 |
+
"learning_rate": 1.2088713036234903e-05,
|
| 1217 |
+
"loss": 0.2089,
|
| 1218 |
+
"step": 36890
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 47.44141559062649,
|
| 1222 |
+
"grad_norm": 0.4808659553527832,
|
| 1223 |
+
"learning_rate": 1.0474802165764265e-05,
|
| 1224 |
+
"loss": 0.2085,
|
| 1225 |
+
"step": 37200
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 47.836760720548384,
|
| 1229 |
+
"grad_norm": 0.4421006143093109,
|
| 1230 |
+
"learning_rate": 8.86089129529363e-06,
|
| 1231 |
+
"loss": 0.2087,
|
| 1232 |
+
"step": 37510
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 48.0,
|
| 1236 |
+
"eval_loss": 0.24061799049377441,
|
| 1237 |
+
"eval_runtime": 146.7785,
|
| 1238 |
+
"eval_samples_per_second": 85.483,
|
| 1239 |
+
"eval_steps_per_second": 2.678,
|
| 1240 |
+
"step": 37638
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 48.23210585047027,
|
| 1244 |
+
"grad_norm": 0.4642196297645569,
|
| 1245 |
+
"learning_rate": 7.246980424822991e-06,
|
| 1246 |
+
"loss": 0.208,
|
| 1247 |
+
"step": 37820
|
| 1248 |
+
},
|
| 1249 |
+
{
|
| 1250 |
+
"epoch": 48.627450980392155,
|
| 1251 |
+
"grad_norm": 0.47141027450561523,
|
| 1252 |
+
"learning_rate": 5.633069554352354e-06,
|
| 1253 |
+
"loss": 0.2067,
|
| 1254 |
+
"step": 38130
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 48.99984058664116,
|
| 1258 |
+
"eval_loss": 0.23903648555278778,
|
| 1259 |
+
"eval_runtime": 146.7908,
|
| 1260 |
+
"eval_samples_per_second": 85.475,
|
| 1261 |
+
"eval_steps_per_second": 2.677,
|
| 1262 |
+
"step": 38422
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 49.02279611031405,
|
| 1266 |
+
"grad_norm": 0.45030030608177185,
|
| 1267 |
+
"learning_rate": 4.019158683881716e-06,
|
| 1268 |
+
"loss": 0.2062,
|
| 1269 |
+
"step": 38440
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 49.41814124023593,
|
| 1273 |
+
"grad_norm": 0.48792940378189087,
|
| 1274 |
+
"learning_rate": 2.4052478134110786e-06,
|
| 1275 |
+
"loss": 0.2062,
|
| 1276 |
+
"step": 38750
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 49.81348637015782,
|
| 1280 |
+
"grad_norm": 0.40084025263786316,
|
| 1281 |
+
"learning_rate": 7.913369429404415e-07,
|
| 1282 |
+
"loss": 0.2055,
|
| 1283 |
+
"step": 39060
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 49.992029332058024,
|
| 1287 |
+
"eval_loss": 0.23992499709129333,
|
| 1288 |
+
"eval_runtime": 146.7718,
|
| 1289 |
+
"eval_samples_per_second": 85.486,
|
| 1290 |
+
"eval_steps_per_second": 2.678,
|
| 1291 |
+
"step": 39200
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 49.992029332058024,
|
| 1295 |
+
"step": 39200,
|
| 1296 |
+
"total_flos": 6.111014223347712e+18,
|
| 1297 |
+
"train_loss": 0.31974333125717785,
|
| 1298 |
+
"train_runtime": 145458.5548,
|
| 1299 |
+
"train_samples_per_second": 34.503,
|
| 1300 |
+
"train_steps_per_second": 0.269
|
| 1301 |
+
}
|
| 1302 |
+
],
|
| 1303 |
+
"logging_steps": 310,
|
| 1304 |
+
"max_steps": 39200,
|
| 1305 |
+
"num_input_tokens_seen": 0,
|
| 1306 |
+
"num_train_epochs": 50,
|
| 1307 |
+
"save_steps": 500,
|
| 1308 |
+
"total_flos": 6.111014223347712e+18,
|
| 1309 |
+
"train_batch_size": 16,
|
| 1310 |
+
"trial_name": null,
|
| 1311 |
+
"trial_params": null
|
| 1312 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9814628d2deffdc0a6ddc0041e4feee9d7da64a8aafaa64a574fc285092b6e9a
|
| 3 |
+
size 5240
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|