Upload 7 files
Browse files- README.MD +29 -0
- config.json +19 -0
- log.txt +1024 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +1 -0
- tokenizer_config.json +1 -0
- vocab.txt +0 -0
README.MD
ADDED
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
This model repository presents "TinyPubMedBERT", a distillated PubMedBERT (Gu et al., 2021) model.
|
2 |
+
TinyPubMedBERT is used as the initial weights for the training of the [dmis-lab/KAZU-NER-module-distil-v1.0](https://huggingface.co/dmis-lab/KAZU-NER-module-distil-v1.0) which is used in the initial release of the KAZU (Korea University and AstraZeneca) framework.
|
3 |
+
|
4 |
+
The model is composed of 4-layers and distillated following methods introduced in TinyBERT paper (Jiao et al., 2020).
|
5 |
+
|
6 |
+
* For the framework, please visit https://github.com/AstraZeneca/KAZU
|
7 |
+
* For details about the model, please see our paper entitled **Biomedical NER for the Enterprise with Distillated BERN2 and the Kazu Framework**, (EMNLP 2022 industry track).
|
8 |
+
|
9 |
+
More details to be announced soon.
|
10 |
+
|
11 |
+
|
12 |
+
### Citation info
|
13 |
+
Joint-first authorship of **Richard Jackson** (AstraZeneca) and **WonJin Yoon** (Korea University).
|
14 |
+
<br>Please cite: (Full citation info will be announced soon)
|
15 |
+
```
|
16 |
+
@inproceedings{YoonAndJackson2022BiomedicalNER,
|
17 |
+
title={Biomedical NER for the Enterprise with Distillated BERN2 and the Kazu Framework},
|
18 |
+
author={Wonjin Yoon, Richard Jackson, Elliot Ford, Vladimir Poroshin, Jaewoo Kang},
|
19 |
+
booktitle={Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing (EMNLP)},
|
20 |
+
year={2022}
|
21 |
+
}
|
22 |
+
```
|
23 |
+
The model used resources of PubMedBERT paper and TinyBERT paper.
|
24 |
+
Gu, Yu, et al. "Domain-specific language model pretraining for biomedical natural language processing." ACM Transactions on Computing for Healthcare (HEALTH) 3.1 (2021): 1-23.
|
25 |
+
Jiao, Xiaoqi, et al. "TinyBERT: Distilling BERT for Natural Language Understanding." Findings of the Association for Computational Linguistics: EMNLP 2020. 2020.
|
26 |
+
|
27 |
+
|
28 |
+
### Contact Information
|
29 |
+
For help or issues using the codes or model (NER module of KAZU) in this repository, please contact WonJin Yoon (wonjin.info (at) gmail.com) or submit a GitHub issue.
|
config.json
ADDED
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"attention_probs_dropout_prob": 0.1,
|
3 |
+
"model_type":"bert",
|
4 |
+
"cell": {},
|
5 |
+
"emb_size": 312,
|
6 |
+
"hidden_act": "gelu",
|
7 |
+
"hidden_dropout_prob": 0.1,
|
8 |
+
"hidden_size": 312,
|
9 |
+
"initializer_range": 0.02,
|
10 |
+
"intermediate_size": 1200,
|
11 |
+
"max_position_embeddings": 512,
|
12 |
+
"num_attention_heads": 12,
|
13 |
+
"num_hidden_layers": 4,
|
14 |
+
"pre_trained": "",
|
15 |
+
"structure": [],
|
16 |
+
"training": "",
|
17 |
+
"type_vocab_size": 2,
|
18 |
+
"vocab_size": 30522
|
19 |
+
}
|
log.txt
ADDED
@@ -0,0 +1,1024 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
att_loss = 17266.918771276512
|
2 |
+
global_step = 249
|
3 |
+
loss = 2164.4442341509593
|
4 |
+
rep_loss = 48.63509366981476
|
5 |
+
att_loss = 17277.140605674238
|
6 |
+
global_step = 499
|
7 |
+
loss = 2165.716661900461
|
8 |
+
rep_loss = 48.59268923226244
|
9 |
+
att_loss = 17277.204854193613
|
10 |
+
global_step = 749
|
11 |
+
loss = 2165.718913059209
|
12 |
+
rep_loss = 48.54645042966937
|
13 |
+
att_loss = 17271.71982749351
|
14 |
+
global_step = 999
|
15 |
+
loss = 2165.026342458792
|
16 |
+
rep_loss = 48.49091171478485
|
17 |
+
att_loss = 17259.165888961168
|
18 |
+
global_step = 1249
|
19 |
+
loss = 2163.44941790835
|
20 |
+
rep_loss = 48.42945459578112
|
21 |
+
att_loss = 17254.51667656367
|
22 |
+
global_step = 1499
|
23 |
+
loss = 2162.8583398759165
|
24 |
+
rep_loss = 48.35004332576138
|
25 |
+
att_loss = 17243.435463959315
|
26 |
+
global_step = 1749
|
27 |
+
loss = 2161.4617817673156
|
28 |
+
rep_loss = 48.25879234910489
|
29 |
+
att_loss = 17231.452888065127
|
30 |
+
global_step = 1999
|
31 |
+
loss = 2159.950227715481
|
32 |
+
rep_loss = 48.14893700922651
|
33 |
+
att_loss = 17213.784783254363
|
34 |
+
global_step = 2249
|
35 |
+
loss = 2157.7255987266585
|
36 |
+
rep_loss = 48.02000851459427
|
37 |
+
att_loss = 17192.42183030439
|
38 |
+
global_step = 2499
|
39 |
+
loss = 2155.035966449568
|
40 |
+
rep_loss = 47.86590490955599
|
41 |
+
att_loss = 17166.00479933044
|
42 |
+
global_step = 2749
|
43 |
+
loss = 2151.711059942207
|
44 |
+
rep_loss = 47.68368448695949
|
45 |
+
att_loss = 17130.214220088335
|
46 |
+
global_step = 2999
|
47 |
+
loss = 2147.211177036022
|
48 |
+
rep_loss = 47.47519993312997
|
49 |
+
att_loss = 17082.496581167103
|
50 |
+
global_step = 3249
|
51 |
+
loss = 2141.2175883689124
|
52 |
+
rep_loss = 47.24412930859313
|
53 |
+
att_loss = 17025.656524771588
|
54 |
+
global_step = 3499
|
55 |
+
loss = 2134.081171550898
|
56 |
+
rep_loss = 46.992851370980176
|
57 |
+
att_loss = 16953.278561555217
|
58 |
+
global_step = 3749
|
59 |
+
loss = 2125.0014296781223
|
60 |
+
rep_loss = 46.73287839932771
|
61 |
+
att_loss = 16867.423787129643
|
62 |
+
global_step = 3999
|
63 |
+
loss = 2114.236494361475
|
64 |
+
rep_loss = 46.46817037039144
|
65 |
+
att_loss = 16766.430148919462
|
66 |
+
global_step = 4249
|
67 |
+
loss = 2101.5791739851875
|
68 |
+
rep_loss = 46.20324492976368
|
69 |
+
att_loss = 16650.415473105353
|
70 |
+
global_step = 4499
|
71 |
+
loss = 2087.044679908388
|
72 |
+
rep_loss = 45.941968157917266
|
73 |
+
att_loss = 16520.907675773025
|
74 |
+
global_step = 4749
|
75 |
+
loss = 2070.82408304349
|
76 |
+
rep_loss = 45.68499064028803
|
77 |
+
att_loss = 16377.305757489388
|
78 |
+
global_step = 4999
|
79 |
+
loss = 2052.8424795921096
|
80 |
+
rep_loss = 45.43408122790482
|
81 |
+
att_loss = 16219.348001274797
|
82 |
+
global_step = 5249
|
83 |
+
loss = 2033.067252099707
|
84 |
+
rep_loss = 45.19001775888017
|
85 |
+
att_loss = 16047.804864911517
|
86 |
+
global_step = 5499
|
87 |
+
loss = 2011.5946817240253
|
88 |
+
rep_loss = 44.952591092136906
|
89 |
+
att_loss = 15862.696737085389
|
90 |
+
global_step = 5749
|
91 |
+
loss = 1988.4273826425274
|
92 |
+
rep_loss = 44.72232620810235
|
93 |
+
att_loss = 15662.845802791677
|
94 |
+
global_step = 5999
|
95 |
+
loss = 1963.4181762354476
|
96 |
+
rep_loss = 44.49960949548504
|
97 |
+
att_loss = 15450.64399507046
|
98 |
+
global_step = 6249
|
99 |
+
loss = 1936.8658219494807
|
100 |
+
rep_loss = 44.28258329186369
|
101 |
+
att_loss = 15224.739606320518
|
102 |
+
global_step = 6499
|
103 |
+
loss = 1908.6014941453677
|
104 |
+
rep_loss = 44.07234962009654
|
105 |
+
att_loss = 14986.976998400805
|
106 |
+
global_step = 6749
|
107 |
+
loss = 1878.8555389734388
|
108 |
+
rep_loss = 43.867316366284136
|
109 |
+
att_loss = 14736.92022153939
|
110 |
+
global_step = 6999
|
111 |
+
loss = 1847.5735737941898
|
112 |
+
rep_loss = 43.668371733469115
|
113 |
+
att_loss = 14475.944257540708
|
114 |
+
global_step = 7249
|
115 |
+
loss = 1814.9273117376963
|
116 |
+
rep_loss = 43.47423925345841
|
117 |
+
att_loss = 14205.014850148966
|
118 |
+
global_step = 7499
|
119 |
+
loss = 1781.0373600503099
|
120 |
+
rep_loss = 43.28403311186527
|
121 |
+
att_loss = 13925.22820553427
|
122 |
+
global_step = 7749
|
123 |
+
loss = 1746.040812043532
|
124 |
+
rep_loss = 43.09829361784272
|
125 |
+
att_loss = 13638.28624035463
|
126 |
+
global_step = 7999
|
127 |
+
loss = 1710.150303852679
|
128 |
+
rep_loss = 42.916193216409454
|
129 |
+
att_loss = 13346.527509119831
|
130 |
+
global_step = 8249
|
131 |
+
loss = 1673.6578935689529
|
132 |
+
rep_loss = 42.7356419749427
|
133 |
+
att_loss = 13052.027369215428
|
134 |
+
global_step = 8499
|
135 |
+
loss = 1636.8227225363235
|
136 |
+
rep_loss = 42.5544134891374
|
137 |
+
att_loss = 12756.839565272303
|
138 |
+
global_step = 8749
|
139 |
+
loss = 1599.9017323053474
|
140 |
+
rep_loss = 42.37429549416374
|
141 |
+
att_loss = 12463.840603725528
|
142 |
+
global_step = 8999
|
143 |
+
loss = 1563.2544286988393
|
144 |
+
rep_loss = 42.19482811192219
|
145 |
+
att_loss = 12175.877058573808
|
146 |
+
global_step = 9249
|
147 |
+
loss = 1527.2367027574312
|
148 |
+
rep_loss = 42.0165656104012
|
149 |
+
att_loss = 11896.004569317394
|
150 |
+
global_step = 9499
|
151 |
+
loss = 1492.2302233208757
|
152 |
+
rep_loss = 41.837219315185614
|
153 |
+
att_loss = 11625.584467229824
|
154 |
+
global_step = 9749
|
155 |
+
loss = 1458.4046069217688
|
156 |
+
rep_loss = 41.65239013487381
|
157 |
+
att_loss = 11364.693283088685
|
158 |
+
global_step = 9999
|
159 |
+
loss = 1425.7699238807395
|
160 |
+
rep_loss = 41.46610991326985
|
161 |
+
att_loss = 11113.502839691315
|
162 |
+
global_step = 10249
|
163 |
+
loss = 1394.3479349004408
|
164 |
+
rep_loss = 41.28064139020281
|
165 |
+
att_loss = 10872.289114246756
|
166 |
+
global_step = 10499
|
167 |
+
loss = 1364.1734168542
|
168 |
+
rep_loss = 41.098222436104656
|
169 |
+
att_loss = 10641.122695878845
|
170 |
+
global_step = 10749
|
171 |
+
loss = 1335.255631208708
|
172 |
+
rep_loss = 40.92235558633416
|
173 |
+
att_loss = 10419.773089068556
|
174 |
+
global_step = 10999
|
175 |
+
loss = 1307.5657580675934
|
176 |
+
rep_loss = 40.752977223469564
|
177 |
+
att_loss = 10207.635612109341
|
178 |
+
global_step = 11249
|
179 |
+
loss = 1281.0279992851579
|
180 |
+
rep_loss = 40.58838390098041
|
181 |
+
att_loss = 10004.12612569298
|
182 |
+
global_step = 11499
|
183 |
+
loss = 1255.569151935997
|
184 |
+
rep_loss = 40.42709148733582
|
185 |
+
att_loss = 9808.771755357146
|
186 |
+
global_step = 11749
|
187 |
+
loss = 1231.1296315219556
|
188 |
+
rep_loss = 40.26529846204394
|
189 |
+
att_loss = 9621.062163698782
|
190 |
+
global_step = 11999
|
191 |
+
loss = 1207.6455957861142
|
192 |
+
rep_loss = 40.10260421332245
|
193 |
+
att_loss = 9440.626454007257
|
194 |
+
global_step = 12249
|
195 |
+
loss = 1185.0705693819755
|
196 |
+
rep_loss = 39.93810263039365
|
197 |
+
att_loss = 9267.06510345962
|
198 |
+
global_step = 12499
|
199 |
+
loss = 1163.3546925466226
|
200 |
+
rep_loss = 39.77243846442683
|
201 |
+
att_loss = 9100.087103520911
|
202 |
+
global_step = 12749
|
203 |
+
loss = 1142.461500134559
|
204 |
+
rep_loss = 39.604899072901425
|
205 |
+
att_loss = 8939.272877253352
|
206 |
+
global_step = 12999
|
207 |
+
loss = 1122.3386413426902
|
208 |
+
rep_loss = 39.43625497553879
|
209 |
+
att_loss = 8784.335411021679
|
210 |
+
global_step = 13249
|
211 |
+
loss = 1102.9502700823227
|
212 |
+
rep_loss = 39.266751113914346
|
213 |
+
att_loss = 8634.944687540738
|
214 |
+
global_step = 13499
|
215 |
+
loss = 1084.2552051671355
|
216 |
+
rep_loss = 39.096955254957656
|
217 |
+
att_loss = 8490.80984011639
|
218 |
+
global_step = 13749
|
219 |
+
loss = 1066.2171257451037
|
220 |
+
rep_loss = 38.92716727312959
|
221 |
+
att_loss = 8351.693344140736
|
222 |
+
global_step = 13999
|
223 |
+
loss = 1048.8063338907355
|
224 |
+
rep_loss = 38.757328404333855
|
225 |
+
att_loss = 8217.371463272595
|
226 |
+
global_step = 14249
|
227 |
+
loss = 1031.9948396856087
|
228 |
+
rep_loss = 38.58725561547826
|
229 |
+
att_loss = 8087.521230980597
|
230 |
+
global_step = 14499
|
231 |
+
loss = 1015.7423801677985
|
232 |
+
rep_loss = 38.41781174607898
|
233 |
+
att_loss = 7961.987407318867
|
234 |
+
global_step = 14749
|
235 |
+
loss = 1000.0294606625666
|
236 |
+
rep_loss = 38.24827932726287
|
237 |
+
att_loss = 7840.550895295689
|
238 |
+
global_step = 14999
|
239 |
+
loss = 984.828728281567
|
240 |
+
rep_loss = 38.078932277249756
|
241 |
+
att_loss = 7722.9954938576475
|
242 |
+
global_step = 15249
|
243 |
+
loss = 970.1131616603415
|
244 |
+
rep_loss = 37.90980071902345
|
245 |
+
att_loss = 7609.155667750541
|
246 |
+
global_step = 15499
|
247 |
+
loss = 955.8620857408843
|
248 |
+
rep_loss = 37.74101945167742
|
249 |
+
att_loss = 7498.834185800413
|
250 |
+
global_step = 15749
|
251 |
+
loss = 942.0508920113786
|
252 |
+
rep_loss = 37.57295154728914
|
253 |
+
att_loss = 7391.8848443881625
|
254 |
+
global_step = 15999
|
255 |
+
loss = 928.6613625704598
|
256 |
+
rep_loss = 37.406057414892906
|
257 |
+
att_loss = 7288.092727160613
|
258 |
+
global_step = 16249
|
259 |
+
loss = 915.6669200155594
|
260 |
+
rep_loss = 37.24263418500655
|
261 |
+
att_loss = 7187.245315153618
|
262 |
+
global_step = 16499
|
263 |
+
loss = 903.0413461738127
|
264 |
+
rep_loss = 37.08545544504563
|
265 |
+
att_loss = 7089.2361219459535
|
266 |
+
global_step = 16749
|
267 |
+
loss = 890.770895313623
|
268 |
+
rep_loss = 36.931041745643014
|
269 |
+
att_loss = 6993.987926303628
|
270 |
+
global_step = 16999
|
271 |
+
loss = 878.8456386448293
|
272 |
+
rep_loss = 36.77718401374561
|
273 |
+
att_loss = 6901.391028709706
|
274 |
+
global_step = 17249
|
275 |
+
loss = 867.2518573375679
|
276 |
+
rep_loss = 36.62383113574191
|
277 |
+
att_loss = 6811.345997624687
|
278 |
+
global_step = 17499
|
279 |
+
loss = 855.97704065998
|
280 |
+
rep_loss = 36.470328775499404
|
281 |
+
att_loss = 6723.7563525545975
|
282 |
+
global_step = 17749
|
283 |
+
loss = 845.009088463928
|
284 |
+
rep_loss = 36.31635626245333
|
285 |
+
att_loss = 6638.549228833737
|
286 |
+
global_step = 17999
|
287 |
+
loss = 834.3388273362962
|
288 |
+
rep_loss = 36.16139095529886
|
289 |
+
att_loss = 6555.648449949764
|
290 |
+
global_step = 18249
|
291 |
+
loss = 823.9567385304105
|
292 |
+
rep_loss = 36.00545937437914
|
293 |
+
att_loss = 6474.9481569615255
|
294 |
+
global_step = 18499
|
295 |
+
loss = 813.8496097940723
|
296 |
+
rep_loss = 35.8487224576022
|
297 |
+
att_loss = 6396.36941115192
|
298 |
+
global_step = 18749
|
299 |
+
loss = 804.0076034054438
|
300 |
+
rep_loss = 35.691417147352944
|
301 |
+
att_loss = 6319.834850662451
|
302 |
+
global_step = 18999
|
303 |
+
loss = 794.4210667197808
|
304 |
+
rep_loss = 35.5336841422198
|
305 |
+
att_loss = 6245.248959206353
|
306 |
+
global_step = 19249
|
307 |
+
loss = 785.0780889001894
|
308 |
+
rep_loss = 35.375753029816316
|
309 |
+
att_loss = 6172.540563334624
|
310 |
+
global_step = 19499
|
311 |
+
loss = 775.9697730082097
|
312 |
+
rep_loss = 35.21762175089005
|
313 |
+
att_loss = 568.803545459948
|
314 |
+
global_step = 19749
|
315 |
+
loss = 73.93347120786969
|
316 |
+
rep_loss = 22.664224607066103
|
317 |
+
att_loss = 568.5669987222423
|
318 |
+
global_step = 19999
|
319 |
+
loss = 73.88368892669678
|
320 |
+
rep_loss = 22.502512358237006
|
321 |
+
att_loss = 568.070243303315
|
322 |
+
global_step = 20249
|
323 |
+
loss = 73.80140220658117
|
324 |
+
rep_loss = 22.340974200673465
|
325 |
+
att_loss = 567.6636848043408
|
326 |
+
global_step = 20499
|
327 |
+
loss = 73.73088559596496
|
328 |
+
rep_loss = 22.183399826252955
|
329 |
+
att_loss = 566.888986146831
|
330 |
+
global_step = 20749
|
331 |
+
loss = 73.61525668139872
|
332 |
+
rep_loss = 22.033067152162666
|
333 |
+
att_loss = 566.2101754964949
|
334 |
+
global_step = 20999
|
335 |
+
loss = 73.51172613087196
|
336 |
+
rep_loss = 21.88363347124433
|
337 |
+
att_loss = 565.3120161074456
|
338 |
+
global_step = 21249
|
339 |
+
loss = 73.38124929864578
|
340 |
+
rep_loss = 21.737978232841133
|
341 |
+
att_loss = 564.4611608841232
|
342 |
+
global_step = 21499
|
343 |
+
loss = 73.25685410693409
|
344 |
+
rep_loss = 21.593671895494953
|
345 |
+
att_loss = 563.8721755071016
|
346 |
+
global_step = 21749
|
347 |
+
loss = 73.16527593471554
|
348 |
+
rep_loss = 21.45003184521113
|
349 |
+
att_loss = 563.0968106397942
|
350 |
+
global_step = 21999
|
351 |
+
loss = 73.05074261683644
|
352 |
+
rep_loss = 21.309130188752846
|
353 |
+
att_loss = 562.5641768938774
|
354 |
+
global_step = 22249
|
355 |
+
loss = 72.96666242970667
|
356 |
+
rep_loss = 21.169122509506167
|
357 |
+
att_loss = 561.8916205836935
|
358 |
+
global_step = 22499
|
359 |
+
loss = 72.86534705723527
|
360 |
+
rep_loss = 21.03115586002598
|
361 |
+
att_loss = 561.2273412750687
|
362 |
+
global_step = 22749
|
363 |
+
loss = 72.76536695714529
|
364 |
+
rep_loss = 20.89559435097197
|
365 |
+
att_loss = 560.6029947471903
|
366 |
+
global_step = 22999
|
367 |
+
loss = 72.67068205350184
|
368 |
+
rep_loss = 20.762461703632624
|
369 |
+
att_loss = 559.9501114984547
|
370 |
+
global_step = 23249
|
371 |
+
loss = 72.57255971640639
|
372 |
+
rep_loss = 20.630366310257575
|
373 |
+
att_loss = 559.3460526415523
|
374 |
+
global_step = 23499
|
375 |
+
loss = 72.4807679872054
|
376 |
+
rep_loss = 20.500091333848136
|
377 |
+
att_loss = 558.7232904827785
|
378 |
+
global_step = 23749
|
379 |
+
loss = 72.38689525042989
|
380 |
+
rep_loss = 20.371871630117862
|
381 |
+
att_loss = 558.0926710499718
|
382 |
+
global_step = 23999
|
383 |
+
loss = 72.29232771448766
|
384 |
+
rep_loss = 20.245950728209028
|
385 |
+
att_loss = 557.5161360830945
|
386 |
+
global_step = 24249
|
387 |
+
loss = 72.20468365261422
|
388 |
+
rep_loss = 20.12133319956434
|
389 |
+
att_loss = 556.9191301414718
|
390 |
+
global_step = 24499
|
391 |
+
loss = 72.1147452129053
|
392 |
+
rep_loss = 19.998831636027287
|
393 |
+
att_loss = 556.3499613072149
|
394 |
+
global_step = 24749
|
395 |
+
loss = 72.0284678045503
|
396 |
+
rep_loss = 19.877781194933966
|
397 |
+
att_loss = 555.8180808992448
|
398 |
+
global_step = 24999
|
399 |
+
loss = 71.94711245500227
|
400 |
+
rep_loss = 19.758818818298426
|
401 |
+
att_loss = 555.2285698721087
|
402 |
+
global_step = 25249
|
403 |
+
loss = 71.85882825288951
|
404 |
+
rep_loss = 19.64205625349118
|
405 |
+
att_loss = 554.7490055753187
|
406 |
+
global_step = 25499
|
407 |
+
loss = 71.7844786938487
|
408 |
+
rep_loss = 19.526824089397824
|
409 |
+
att_loss = 554.2084955723976
|
410 |
+
global_step = 25749
|
411 |
+
loss = 71.70283811223419
|
412 |
+
rep_loss = 19.414209418026907
|
413 |
+
att_loss = 553.6828962171237
|
414 |
+
global_step = 25999
|
415 |
+
loss = 71.62333540754867
|
416 |
+
rep_loss = 19.303787130742116
|
417 |
+
att_loss = 553.1985000778094
|
418 |
+
global_step = 26249
|
419 |
+
loss = 71.54924880910589
|
420 |
+
rep_loss = 19.195490497038886
|
421 |
+
att_loss = 552.7054209829683
|
422 |
+
global_step = 26499
|
423 |
+
loss = 71.47437279327147
|
424 |
+
rep_loss = 19.089561467801158
|
425 |
+
att_loss = 552.2043198470585
|
426 |
+
global_step = 26749
|
427 |
+
loss = 71.39876808444743
|
428 |
+
rep_loss = 18.98582492440918
|
429 |
+
att_loss = 551.7215547357797
|
430 |
+
global_step = 26999
|
431 |
+
loss = 71.32573316988287
|
432 |
+
rep_loss = 18.88431070933462
|
433 |
+
att_loss = 551.2641789369162
|
434 |
+
global_step = 27249
|
435 |
+
loss = 71.25615824510425
|
436 |
+
rep_loss = 18.785087129471723
|
437 |
+
att_loss = 550.7662613197069
|
438 |
+
global_step = 27499
|
439 |
+
loss = 71.18177146832605
|
440 |
+
rep_loss = 18.68791050243864
|
441 |
+
att_loss = 550.2952256681597
|
442 |
+
global_step = 27749
|
443 |
+
loss = 71.1110345714397
|
444 |
+
rep_loss = 18.593050957108957
|
445 |
+
att_loss = 549.8622339791492
|
446 |
+
global_step = 27999
|
447 |
+
loss = 71.04531030003506
|
448 |
+
rep_loss = 18.500248481171656
|
449 |
+
att_loss = 549.4039358333213
|
450 |
+
global_step = 28249
|
451 |
+
loss = 70.97665694765465
|
452 |
+
rep_loss = 18.409319813907526
|
453 |
+
att_loss = 548.9680096317376
|
454 |
+
global_step = 28499
|
455 |
+
loss = 70.91105011217064
|
456 |
+
rep_loss = 18.32039132633069
|
457 |
+
att_loss = 548.5215405326286
|
458 |
+
global_step = 28749
|
459 |
+
loss = 70.84437117715535
|
460 |
+
rep_loss = 18.23342894976974
|
461 |
+
att_loss = 548.0985423778454
|
462 |
+
global_step = 28999
|
463 |
+
loss = 70.78087078551674
|
464 |
+
rep_loss = 18.148423988121344
|
465 |
+
att_loss = 547.6451674794828
|
466 |
+
global_step = 29249
|
467 |
+
loss = 70.71380790712438
|
468 |
+
rep_loss = 18.06529587243236
|
469 |
+
att_loss = 547.2482466779547
|
470 |
+
global_step = 29499
|
471 |
+
loss = 70.65403027701583
|
472 |
+
rep_loss = 17.98399563381792
|
473 |
+
att_loss = 546.8227743930071
|
474 |
+
global_step = 29749
|
475 |
+
loss = 70.59089194435245
|
476 |
+
rep_loss = 17.904361261194918
|
477 |
+
att_loss = 546.3994455970849
|
478 |
+
global_step = 29999
|
479 |
+
loss = 70.5282377780953
|
480 |
+
rep_loss = 17.82645672483108
|
481 |
+
att_loss = 545.9787717393458
|
482 |
+
global_step = 30249
|
483 |
+
loss = 70.46615296688323
|
484 |
+
rep_loss = 17.75045208852209
|
485 |
+
att_loss = 545.5533168375904
|
486 |
+
global_step = 30499
|
487 |
+
loss = 70.40367570479602
|
488 |
+
rep_loss = 17.676088884691314
|
489 |
+
att_loss = 545.1392937153511
|
490 |
+
global_step = 30749
|
491 |
+
loss = 70.3428395847202
|
492 |
+
rep_loss = 17.603423051008935
|
493 |
+
att_loss = 544.702601219369
|
494 |
+
global_step = 30999
|
495 |
+
loss = 70.27936210565075
|
496 |
+
rep_loss = 17.532295710434205
|
497 |
+
att_loss = 544.2759774692103
|
498 |
+
global_step = 31249
|
499 |
+
loss = 70.21734446447026
|
500 |
+
rep_loss = 17.462778319002688
|
501 |
+
att_loss = 543.8311069821848
|
502 |
+
global_step = 31499
|
503 |
+
loss = 70.15323589308043
|
504 |
+
rep_loss = 17.39478023828158
|
505 |
+
att_loss = 543.3645310306115
|
506 |
+
global_step = 31749
|
507 |
+
loss = 70.08659764120726
|
508 |
+
rep_loss = 17.328250161494633
|
509 |
+
att_loss = 542.8894112093722
|
510 |
+
global_step = 31999
|
511 |
+
loss = 70.01909737438265
|
512 |
+
rep_loss = 17.26336783997778
|
513 |
+
att_loss = 542.3946143301388
|
514 |
+
global_step = 32249
|
515 |
+
loss = 69.94934149257907
|
516 |
+
rep_loss = 17.200117651258683
|
517 |
+
att_loss = 541.8585259116452
|
518 |
+
global_step = 32499
|
519 |
+
loss = 69.87462579326993
|
520 |
+
rep_loss = 17.138480488332434
|
521 |
+
att_loss = 541.2866426885196
|
522 |
+
global_step = 32749
|
523 |
+
loss = 69.7956412803312
|
524 |
+
rep_loss = 17.07848758956047
|
525 |
+
att_loss = 540.6902034943896
|
526 |
+
global_step = 32999
|
527 |
+
loss = 69.71379761140237
|
528 |
+
rep_loss = 17.020177439787542
|
529 |
+
att_loss = 540.0849288407298
|
530 |
+
global_step = 33249
|
531 |
+
loss = 69.63103771462252
|
532 |
+
rep_loss = 16.963372914174673
|
533 |
+
att_loss = 539.4442540252454
|
534 |
+
global_step = 33499
|
535 |
+
loss = 69.54398421755592
|
536 |
+
rep_loss = 16.90761975887322
|
537 |
+
att_loss = 538.7701227846446
|
538 |
+
global_step = 33749
|
539 |
+
loss = 69.45288906286862
|
540 |
+
rep_loss = 16.85298976632626
|
541 |
+
att_loss = 538.0739389115736
|
542 |
+
global_step = 33999
|
543 |
+
loss = 69.35915206272288
|
544 |
+
rep_loss = 16.799277641823917
|
545 |
+
att_loss = 537.3629695459766
|
546 |
+
global_step = 34249
|
547 |
+
loss = 69.26371267964963
|
548 |
+
rep_loss = 16.746731938422236
|
549 |
+
att_loss = 536.6309045878353
|
550 |
+
global_step = 34499
|
551 |
+
loss = 69.16578501913033
|
552 |
+
rep_loss = 16.695375614416804
|
553 |
+
att_loss = 535.8522947677203
|
554 |
+
global_step = 34749
|
555 |
+
loss = 69.06215787055517
|
556 |
+
rep_loss = 16.644968248092802
|
557 |
+
att_loss = 535.061523088472
|
558 |
+
global_step = 34999
|
559 |
+
loss = 68.95718339396129
|
560 |
+
rep_loss = 16.59594411223049
|
561 |
+
att_loss = 534.2347752364716
|
562 |
+
global_step = 35249
|
563 |
+
loss = 68.8478561000512
|
564 |
+
rep_loss = 16.54807361378506
|
565 |
+
att_loss = 533.3319543173427
|
566 |
+
global_step = 35499
|
567 |
+
loss = 68.7291399041799
|
568 |
+
rep_loss = 16.501164960748355
|
569 |
+
att_loss = 532.4195050645149
|
570 |
+
global_step = 35749
|
571 |
+
loss = 68.60936011142263
|
572 |
+
rep_loss = 16.455375870394757
|
573 |
+
att_loss = 531.4884166494645
|
574 |
+
global_step = 35999
|
575 |
+
loss = 68.48734341712128
|
576 |
+
rep_loss = 16.410330731608372
|
577 |
+
att_loss = 530.547413615715
|
578 |
+
global_step = 36249
|
579 |
+
loss = 68.36417261975902
|
580 |
+
rep_loss = 16.36596738483289
|
581 |
+
att_loss = 529.5643084049649
|
582 |
+
global_step = 36499
|
583 |
+
loss = 68.23580734429001
|
584 |
+
rep_loss = 16.322150390365422
|
585 |
+
att_loss = 528.5736010234305
|
586 |
+
global_step = 36749
|
587 |
+
loss = 68.10659463633637
|
588 |
+
rep_loss = 16.279156108382292
|
589 |
+
att_loss = 527.5627225302385
|
590 |
+
global_step = 36999
|
591 |
+
loss = 67.9749567699075
|
592 |
+
rep_loss = 16.236931669151932
|
593 |
+
att_loss = 526.5287834833074
|
594 |
+
global_step = 37249
|
595 |
+
loss = 67.84054858117457
|
596 |
+
rep_loss = 16.195605205317456
|
597 |
+
att_loss = 525.4722971581919
|
598 |
+
global_step = 37499
|
599 |
+
loss = 67.70341618880734
|
600 |
+
rep_loss = 16.155032388039682
|
601 |
+
att_loss = 524.3897297541125
|
602 |
+
global_step = 37749
|
603 |
+
loss = 67.56309876005832
|
604 |
+
rep_loss = 16.115060361896177
|
605 |
+
att_loss = 523.2993466628608
|
606 |
+
global_step = 37999
|
607 |
+
loss = 67.42187455775921
|
608 |
+
rep_loss = 16.075649831086423
|
609 |
+
att_loss = 522.1973293451123
|
610 |
+
global_step = 38249
|
611 |
+
loss = 67.27924377308574
|
612 |
+
rep_loss = 16.036620870230177
|
613 |
+
att_loss = 521.1076512535866
|
614 |
+
global_step = 38499
|
615 |
+
loss = 67.1382169603574
|
616 |
+
rep_loss = 15.99808445711862
|
617 |
+
att_loss = 520.0252741897315
|
618 |
+
global_step = 38749
|
619 |
+
loss = 66.99814691873208
|
620 |
+
rep_loss = 15.959901182549759
|
621 |
+
att_loss = 518.9590398213141
|
622 |
+
global_step = 38999
|
623 |
+
loss = 66.86016447237833
|
624 |
+
rep_loss = 15.92227598179462
|
625 |
+
att_loss = 517.9105902646263
|
626 |
+
global_step = 39249
|
627 |
+
loss = 66.72446811673105
|
628 |
+
rep_loss = 15.885154695351472
|
629 |
+
att_loss = 430.41770705377866
|
630 |
+
global_step = 39499
|
631 |
+
loss = 55.413460654113926
|
632 |
+
rep_loss = 12.889978018730723
|
633 |
+
att_loss = 429.78013253590416
|
634 |
+
global_step = 39749
|
635 |
+
loss = 55.33112620550489
|
636 |
+
rep_loss = 12.868876984330262
|
637 |
+
att_loss = 429.5238455014705
|
638 |
+
global_step = 39999
|
639 |
+
loss = 55.29708110613347
|
640 |
+
rep_loss = 12.852803301535191
|
641 |
+
att_loss = 429.2593395666919
|
642 |
+
global_step = 40249
|
643 |
+
loss = 55.262259974363126
|
644 |
+
rep_loss = 12.838740212631023
|
645 |
+
att_loss = 429.0502507508452
|
646 |
+
global_step = 40499
|
647 |
+
loss = 55.23442752975861
|
648 |
+
rep_loss = 12.825169506941794
|
649 |
+
att_loss = 428.6518964721129
|
650 |
+
global_step = 40749
|
651 |
+
loss = 55.182567169896274
|
652 |
+
rep_loss = 12.808640934294251
|
653 |
+
att_loss = 428.3047634104453
|
654 |
+
global_step = 40999
|
655 |
+
loss = 55.13724845658386
|
656 |
+
rep_loss = 12.793224244904476
|
657 |
+
att_loss = 428.0248651833709
|
658 |
+
global_step = 41249
|
659 |
+
loss = 55.10051056480113
|
660 |
+
rep_loss = 12.779219316121907
|
661 |
+
att_loss = 427.59188728907077
|
662 |
+
global_step = 41499
|
663 |
+
loss = 55.04444140726331
|
664 |
+
rep_loss = 12.763643969035703
|
665 |
+
att_loss = 427.19180261439254
|
666 |
+
global_step = 41749
|
667 |
+
loss = 54.99262022278632
|
668 |
+
rep_loss = 12.749159168239892
|
669 |
+
att_loss = 426.83212551739257
|
670 |
+
global_step = 41999
|
671 |
+
loss = 54.94592204886004
|
672 |
+
rep_loss = 12.735250879202312
|
673 |
+
att_loss = 426.51639836015283
|
674 |
+
global_step = 42249
|
675 |
+
loss = 54.90487258428621
|
676 |
+
rep_loss = 12.722582299342086
|
677 |
+
att_loss = 426.21925660877235
|
678 |
+
global_step = 42499
|
679 |
+
loss = 54.866193434991125
|
680 |
+
rep_loss = 12.710290857819844
|
681 |
+
att_loss = 425.84940542117414
|
682 |
+
global_step = 42749
|
683 |
+
loss = 54.81834573518464
|
684 |
+
rep_loss = 12.697360449875209
|
685 |
+
att_loss = 425.47620234131585
|
686 |
+
global_step = 42999
|
687 |
+
loss = 54.770065517032805
|
688 |
+
rep_loss = 12.684321784643712
|
689 |
+
att_loss = 425.0879732096628
|
690 |
+
global_step = 43249
|
691 |
+
loss = 54.71991780154507
|
692 |
+
rep_loss = 12.671369195619654
|
693 |
+
att_loss = 424.7154397102052
|
694 |
+
global_step = 43499
|
695 |
+
loss = 54.67179627863384
|
696 |
+
rep_loss = 12.65893052896319
|
697 |
+
att_loss = 424.3949641756205
|
698 |
+
global_step = 43749
|
699 |
+
loss = 54.630329651543754
|
700 |
+
rep_loss = 12.647673039628536
|
701 |
+
att_loss = 424.0026099238155
|
702 |
+
global_step = 43999
|
703 |
+
loss = 54.579816030556394
|
704 |
+
rep_loss = 12.635918306506435
|
705 |
+
att_loss = 423.5783165717651
|
706 |
+
global_step = 44249
|
707 |
+
loss = 54.52532899758517
|
708 |
+
rep_loss = 12.624315396056824
|
709 |
+
att_loss = 423.06853340926415
|
710 |
+
global_step = 44499
|
711 |
+
loss = 54.46005894117699
|
712 |
+
rep_loss = 12.61193810453582
|
713 |
+
att_loss = 422.57546552711716
|
714 |
+
global_step = 44749
|
715 |
+
loss = 54.39705093238661
|
716 |
+
rep_loss = 12.600941921546822
|
717 |
+
att_loss = 421.9954732306036
|
718 |
+
global_step = 44999
|
719 |
+
loss = 54.323057490814506
|
720 |
+
rep_loss = 12.588986684705855
|
721 |
+
att_loss = 421.38440246980133
|
722 |
+
global_step = 45249
|
723 |
+
loss = 54.245232343954626
|
724 |
+
rep_loss = 12.577456274451562
|
725 |
+
att_loss = 420.79632926291475
|
726 |
+
global_step = 45499
|
727 |
+
loss = 54.17038325118895
|
728 |
+
rep_loss = 12.566736740242625
|
729 |
+
att_loss = 420.1323002994551
|
730 |
+
global_step = 45749
|
731 |
+
loss = 54.08597275737353
|
732 |
+
rep_loss = 12.555481753425557
|
733 |
+
att_loss = 419.47063839430234
|
734 |
+
global_step = 45999
|
735 |
+
loss = 54.00194308127439
|
736 |
+
rep_loss = 12.544906236455144
|
737 |
+
att_loss = 418.78285078568894
|
738 |
+
global_step = 46249
|
739 |
+
loss = 53.91468508692506
|
740 |
+
rep_loss = 12.534629891918641
|
741 |
+
att_loss = 418.03507678324087
|
742 |
+
global_step = 46499
|
743 |
+
loss = 53.81990357788759
|
744 |
+
rep_loss = 12.524151824442704
|
745 |
+
att_loss = 417.23997283574636
|
746 |
+
global_step = 46749
|
747 |
+
loss = 53.719232056093155
|
748 |
+
rep_loss = 12.513883584610413
|
749 |
+
att_loss = 416.4557218942232
|
750 |
+
global_step = 46999
|
751 |
+
loss = 53.61996301248749
|
752 |
+
rep_loss = 12.503982180938998
|
753 |
+
att_loss = 415.64490532235556
|
754 |
+
global_step = 47249
|
755 |
+
loss = 53.51731615513282
|
756 |
+
rep_loss = 12.49362389865112
|
757 |
+
att_loss = 414.8313552065656
|
758 |
+
global_step = 47499
|
759 |
+
loss = 53.41432624916325
|
760 |
+
rep_loss = 12.483254770265658
|
761 |
+
att_loss = 413.96981992971246
|
762 |
+
global_step = 47749
|
763 |
+
loss = 53.30523171779062
|
764 |
+
rep_loss = 12.472033788222666
|
765 |
+
att_loss = 413.15383253052545
|
766 |
+
global_step = 47999
|
767 |
+
loss = 53.20193024074479
|
768 |
+
rep_loss = 12.461609376792241
|
769 |
+
att_loss = 412.3265501465417
|
770 |
+
global_step = 48249
|
771 |
+
loss = 53.097181754735686
|
772 |
+
rep_loss = 12.450903867677932
|
773 |
+
att_loss = 411.4937257261404
|
774 |
+
global_step = 48499
|
775 |
+
loss = 52.99173086048301
|
776 |
+
rep_loss = 12.440121136711934
|
777 |
+
att_loss = 410.6774601531779
|
778 |
+
global_step = 48749
|
779 |
+
loss = 52.88838179087995
|
780 |
+
rep_loss = 12.429594153153822
|
781 |
+
att_loss = 409.8448136892251
|
782 |
+
global_step = 48999
|
783 |
+
loss = 52.782971961618074
|
784 |
+
rep_loss = 12.418961979129737
|
785 |
+
att_loss = 409.0028938114637
|
786 |
+
global_step = 49249
|
787 |
+
loss = 52.67639218699631
|
788 |
+
rep_loss = 12.408243665116233
|
789 |
+
att_loss = 408.18548318029866
|
790 |
+
global_step = 49499
|
791 |
+
loss = 52.57292907776794
|
792 |
+
rep_loss = 12.397949419292052
|
793 |
+
att_loss = 407.36597105840445
|
794 |
+
global_step = 49749
|
795 |
+
loss = 52.469220969745706
|
796 |
+
rep_loss = 12.387796673209785
|
797 |
+
att_loss = 406.5366784533664
|
798 |
+
global_step = 49999
|
799 |
+
loss = 52.36429095096533
|
800 |
+
rep_loss = 12.377649139604202
|
801 |
+
att_loss = 405.7202889275348
|
802 |
+
global_step = 50249
|
803 |
+
loss = 52.261031498737715
|
804 |
+
rep_loss = 12.367963048278844
|
805 |
+
att_loss = 404.90346718124215
|
806 |
+
global_step = 50499
|
807 |
+
loss = 52.15774034676183
|
808 |
+
rep_loss = 12.358455586748716
|
809 |
+
att_loss = 404.07259098323385
|
810 |
+
global_step = 50749
|
811 |
+
loss = 52.05269508167597
|
812 |
+
rep_loss = 12.348969661098565
|
813 |
+
att_loss = 403.22920092030535
|
814 |
+
global_step = 50999
|
815 |
+
loss = 51.94607408712973
|
816 |
+
rep_loss = 12.339391768106127
|
817 |
+
att_loss = 402.3802402069217
|
818 |
+
global_step = 51249
|
819 |
+
loss = 51.838757622093226
|
820 |
+
rep_loss = 12.32982075661633
|
821 |
+
att_loss = 401.5426612154961
|
822 |
+
global_step = 51499
|
823 |
+
loss = 51.732905490317634
|
824 |
+
rep_loss = 12.320582694186324
|
825 |
+
att_loss = 400.70377284389497
|
826 |
+
global_step = 51749
|
827 |
+
loss = 51.62691244608757
|
828 |
+
rep_loss = 12.311526711448368
|
829 |
+
att_loss = 399.8625147925249
|
830 |
+
global_step = 51999
|
831 |
+
loss = 51.520630164608946
|
832 |
+
rep_loss = 12.302526512116689
|
833 |
+
att_loss = 399.0078826006999
|
834 |
+
global_step = 52249
|
835 |
+
loss = 51.41267432677176
|
836 |
+
rep_loss = 12.293511999767112
|
837 |
+
att_loss = 398.1749379127149
|
838 |
+
global_step = 52499
|
839 |
+
loss = 51.307466360049794
|
840 |
+
rep_loss = 12.284792951091228
|
841 |
+
att_loss = 397.34106225898586
|
842 |
+
global_step = 52749
|
843 |
+
loss = 51.20213805727196
|
844 |
+
rep_loss = 12.276042182542554
|
845 |
+
att_loss = 396.5201868602387
|
846 |
+
global_step = 52999
|
847 |
+
loss = 51.09846154694909
|
848 |
+
rep_loss = 12.267505496067715
|
849 |
+
att_loss = 395.696838918234
|
850 |
+
global_step = 53249
|
851 |
+
loss = 50.99445791232878
|
852 |
+
rep_loss = 12.258824362917975
|
853 |
+
att_loss = 394.87037899546084
|
854 |
+
global_step = 53499
|
855 |
+
loss = 50.89005832794428
|
856 |
+
rep_loss = 12.250087614879657
|
857 |
+
att_loss = 394.0527554703472
|
858 |
+
global_step = 53749
|
859 |
+
loss = 50.78677591598312
|
860 |
+
rep_loss = 12.241451844243857
|
861 |
+
att_loss = 393.24143282637345
|
862 |
+
global_step = 53999
|
863 |
+
loss = 50.68428991926787
|
864 |
+
rep_loss = 12.232886519022125
|
865 |
+
att_loss = 392.43965526004047
|
866 |
+
global_step = 54249
|
867 |
+
loss = 50.58300504671044
|
868 |
+
rep_loss = 12.224385103685663
|
869 |
+
att_loss = 391.64264881379376
|
870 |
+
global_step = 54499
|
871 |
+
loss = 50.48231238881279
|
872 |
+
rep_loss = 12.21585028627942
|
873 |
+
att_loss = 390.8601956633998
|
874 |
+
global_step = 54749
|
875 |
+
loss = 50.38346429668199
|
876 |
+
rep_loss = 12.207518704250512
|
877 |
+
att_loss = 390.0729250104103
|
878 |
+
global_step = 54999
|
879 |
+
loss = 50.284001766218445
|
880 |
+
rep_loss = 12.199089112598456
|
881 |
+
att_loss = 389.31133648670345
|
882 |
+
global_step = 55249
|
883 |
+
loss = 50.187795636615164
|
884 |
+
rep_loss = 12.191028596376597
|
885 |
+
att_loss = 388.5525984823884
|
886 |
+
global_step = 55499
|
887 |
+
loss = 50.091947100616196
|
888 |
+
rep_loss = 12.182978309722705
|
889 |
+
att_loss = 387.80436541667115
|
890 |
+
global_step = 55749
|
891 |
+
loss = 49.99741998727243
|
892 |
+
rep_loss = 12.174994470516149
|
893 |
+
att_loss = 387.0552613080837
|
894 |
+
global_step = 55999
|
895 |
+
loss = 49.9027750983903
|
896 |
+
rep_loss = 12.166939473817807
|
897 |
+
att_loss = 386.3239343684958
|
898 |
+
global_step = 56249
|
899 |
+
loss = 49.810371762527836
|
900 |
+
rep_loss = 12.15903972790589
|
901 |
+
att_loss = 385.59254376769417
|
902 |
+
global_step = 56499
|
903 |
+
loss = 49.717960871643825
|
904 |
+
rep_loss = 12.1511432036396
|
905 |
+
att_loss = 384.88300487745124
|
906 |
+
global_step = 56749
|
907 |
+
loss = 49.628306122153155
|
908 |
+
rep_loss = 12.143444097600504
|
909 |
+
att_loss = 384.18496186893213
|
910 |
+
global_step = 56999
|
911 |
+
loss = 49.54010241677591
|
912 |
+
rep_loss = 12.135857463374899
|
913 |
+
att_loss = 383.48891160166664
|
914 |
+
global_step = 57249
|
915 |
+
loss = 49.4521423345372
|
916 |
+
rep_loss = 12.128227071913313
|
917 |
+
att_loss = 382.8087591240781
|
918 |
+
global_step = 57499
|
919 |
+
loss = 49.366188662171815
|
920 |
+
rep_loss = 12.120750168099738
|
921 |
+
att_loss = 382.1431455683032
|
922 |
+
global_step = 57749
|
923 |
+
loss = 49.282078285950604
|
924 |
+
rep_loss = 12.113480712152032
|
925 |
+
att_loss = 381.4765402969557
|
926 |
+
global_step = 57999
|
927 |
+
loss = 49.19783066547572
|
928 |
+
rep_loss = 12.10610501573975
|
929 |
+
att_loss = 380.8198167156326
|
930 |
+
global_step = 58249
|
931 |
+
loss = 49.11483049123329
|
932 |
+
rep_loss = 12.098827202294546
|
933 |
+
att_loss = 380.16741775930916
|
934 |
+
global_step = 58499
|
935 |
+
loss = 49.03236370336133
|
936 |
+
rep_loss = 12.09149185478529
|
937 |
+
att_loss = 379.52598527098235
|
938 |
+
global_step = 58749
|
939 |
+
loss = 48.95128226382001
|
940 |
+
rep_loss = 12.084272827448947
|
941 |
+
att_loss = 328.35527935543575
|
942 |
+
global_step = 58999
|
943 |
+
loss = 42.482683503949964
|
944 |
+
rep_loss = 11.506188331423578
|
945 |
+
att_loss = 328.64847870404714
|
946 |
+
global_step = 59249
|
947 |
+
loss = 42.51917423520769
|
948 |
+
rep_loss = 11.504915223304403
|
949 |
+
att_loss = 328.2227134136292
|
950 |
+
global_step = 59499
|
951 |
+
loss = 42.4650246044777
|
952 |
+
rep_loss = 11.497483390669583
|
953 |
+
att_loss = 328.1589748620078
|
954 |
+
global_step = 59749
|
955 |
+
loss = 42.45698851884699
|
956 |
+
rep_loss = 11.496933261957375
|
957 |
+
att_loss = 327.96913344880767
|
958 |
+
global_step = 59999
|
959 |
+
loss = 42.43277933328568
|
960 |
+
rep_loss = 11.493101262425503
|
961 |
+
att_loss = 327.7072755121499
|
962 |
+
global_step = 60249
|
963 |
+
loss = 42.39948780108721
|
964 |
+
rep_loss = 11.488626917388519
|
965 |
+
att_loss = 327.45508357775856
|
966 |
+
global_step = 60499
|
967 |
+
loss = 42.367388150718355
|
968 |
+
rep_loss = 11.484021703919266
|
969 |
+
att_loss = 327.2823583969005
|
970 |
+
global_step = 60749
|
971 |
+
loss = 42.34552766013292
|
972 |
+
rep_loss = 11.48186293446157
|
973 |
+
att_loss = 327.03517870746174
|
974 |
+
global_step = 60999
|
975 |
+
loss = 42.314138652415735
|
976 |
+
rep_loss = 11.477930534863273
|
977 |
+
att_loss = 326.77379158797305
|
978 |
+
global_step = 61249
|
979 |
+
loss = 42.28088593493396
|
980 |
+
rep_loss = 11.47329593538769
|
981 |
+
att_loss = 326.5938454012632
|
982 |
+
global_step = 61499
|
983 |
+
loss = 42.257999699464406
|
984 |
+
rep_loss = 11.470152240594542
|
985 |
+
att_loss = 326.35747106023103
|
986 |
+
global_step = 61749
|
987 |
+
loss = 42.227962024245244
|
988 |
+
rep_loss = 11.466225154518781
|
989 |
+
att_loss = 326.12352478296145
|
990 |
+
global_step = 61999
|
991 |
+
loss = 42.198229037345456
|
992 |
+
rep_loss = 11.462307569460135
|
993 |
+
att_loss = 325.93727647812125
|
994 |
+
global_step = 62249
|
995 |
+
loss = 42.17457759942477
|
996 |
+
rep_loss = 11.459344349590738
|
997 |
+
att_loss = 325.7074230125067
|
998 |
+
global_step = 62499
|
999 |
+
loss = 42.14533984246549
|
1000 |
+
rep_loss = 11.455295764661827
|
1001 |
+
att_loss = 325.46969798551186
|
1002 |
+
global_step = 62749
|
1003 |
+
loss = 42.11513811697707
|
1004 |
+
rep_loss = 11.451406990408362
|
1005 |
+
att_loss = 325.26252719178956
|
1006 |
+
global_step = 62999
|
1007 |
+
loss = 42.08880578246127
|
1008 |
+
rep_loss = 11.44791910478258
|
1009 |
+
att_loss = 325.01166694948995
|
1010 |
+
global_step = 63249
|
1011 |
+
loss = 42.056943388399
|
1012 |
+
rep_loss = 11.443880189040726
|
1013 |
+
att_loss = 324.77472143948853
|
1014 |
+
global_step = 63499
|
1015 |
+
loss = 42.0268650872607
|
1016 |
+
rep_loss = 11.440199283558234
|
1017 |
+
att_loss = 324.58982729747447
|
1018 |
+
global_step = 63749
|
1019 |
+
loss = 42.00338519389031
|
1020 |
+
rep_loss = 11.437254277156194
|
1021 |
+
att_loss = 324.36273489704627
|
1022 |
+
global_step = 63999
|
1023 |
+
loss = 41.97451099148103
|
1024 |
+
rep_loss = 11.433353066113062
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:18a90816d852d1829a6e33999e40c1909d7ab8fb81a9d379af3c107ce0c72f97
|
3 |
+
size 58912319
|
special_tokens_map.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}
|
tokenizer_config.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"do_lower_case": true, "do_basic_tokenize": true, "never_split": null, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "special_tokens_map_file": null, "tokenizer_file": null}
|
vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|