mrm8488 commited on
Commit
bcc4960
1 Parent(s): 7783fcc

Upload folder using huggingface_hub

Browse files
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6d86702afea539a4fa48d51eedfb7464664213f13739108d4a1228a4983dec1c
3
+ size 5549215498
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c7be3750cbd77f30f33330b7973d1253af2a0b0d726094b056252bbcf1a28b23
3
+ size 5548078554
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:537f1994de1911da9edc25f9afb87c8543debbe4a03a7ca68d9b9747e89c6379
3
+ size 14244
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3bac79bd03ad828963722c6a346a56784346186b4d654624f46583d9adaf88be
3
+ size 1064
special_tokens_map.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "pad_token": "<|endoftext|>",
5
+ "unk_token": "<|endoftext|>"
6
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,213 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "0": {
5
+ "content": "<|endoftext|>",
6
+ "lstrip": false,
7
+ "normalized": false,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "1": {
13
+ "content": "<|padding|>",
14
+ "lstrip": false,
15
+ "normalized": false,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "50254": {
21
+ "content": " ",
22
+ "lstrip": false,
23
+ "normalized": true,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": false
27
+ },
28
+ "50255": {
29
+ "content": " ",
30
+ "lstrip": false,
31
+ "normalized": true,
32
+ "rstrip": false,
33
+ "single_word": false,
34
+ "special": false
35
+ },
36
+ "50256": {
37
+ "content": " ",
38
+ "lstrip": false,
39
+ "normalized": true,
40
+ "rstrip": false,
41
+ "single_word": false,
42
+ "special": false
43
+ },
44
+ "50257": {
45
+ "content": " ",
46
+ "lstrip": false,
47
+ "normalized": true,
48
+ "rstrip": false,
49
+ "single_word": false,
50
+ "special": false
51
+ },
52
+ "50258": {
53
+ "content": " ",
54
+ "lstrip": false,
55
+ "normalized": true,
56
+ "rstrip": false,
57
+ "single_word": false,
58
+ "special": false
59
+ },
60
+ "50259": {
61
+ "content": " ",
62
+ "lstrip": false,
63
+ "normalized": true,
64
+ "rstrip": false,
65
+ "single_word": false,
66
+ "special": false
67
+ },
68
+ "50260": {
69
+ "content": " ",
70
+ "lstrip": false,
71
+ "normalized": true,
72
+ "rstrip": false,
73
+ "single_word": false,
74
+ "special": false
75
+ },
76
+ "50261": {
77
+ "content": " ",
78
+ "lstrip": false,
79
+ "normalized": true,
80
+ "rstrip": false,
81
+ "single_word": false,
82
+ "special": false
83
+ },
84
+ "50262": {
85
+ "content": " ",
86
+ "lstrip": false,
87
+ "normalized": true,
88
+ "rstrip": false,
89
+ "single_word": false,
90
+ "special": false
91
+ },
92
+ "50263": {
93
+ "content": " ",
94
+ "lstrip": false,
95
+ "normalized": true,
96
+ "rstrip": false,
97
+ "single_word": false,
98
+ "special": false
99
+ },
100
+ "50264": {
101
+ "content": " ",
102
+ "lstrip": false,
103
+ "normalized": true,
104
+ "rstrip": false,
105
+ "single_word": false,
106
+ "special": false
107
+ },
108
+ "50265": {
109
+ "content": " ",
110
+ "lstrip": false,
111
+ "normalized": true,
112
+ "rstrip": false,
113
+ "single_word": false,
114
+ "special": false
115
+ },
116
+ "50266": {
117
+ "content": " ",
118
+ "lstrip": false,
119
+ "normalized": true,
120
+ "rstrip": false,
121
+ "single_word": false,
122
+ "special": false
123
+ },
124
+ "50267": {
125
+ "content": " ",
126
+ "lstrip": false,
127
+ "normalized": true,
128
+ "rstrip": false,
129
+ "single_word": false,
130
+ "special": false
131
+ },
132
+ "50268": {
133
+ "content": " ",
134
+ "lstrip": false,
135
+ "normalized": true,
136
+ "rstrip": false,
137
+ "single_word": false,
138
+ "special": false
139
+ },
140
+ "50269": {
141
+ "content": " ",
142
+ "lstrip": false,
143
+ "normalized": true,
144
+ "rstrip": false,
145
+ "single_word": false,
146
+ "special": false
147
+ },
148
+ "50270": {
149
+ "content": " ",
150
+ "lstrip": false,
151
+ "normalized": true,
152
+ "rstrip": false,
153
+ "single_word": false,
154
+ "special": false
155
+ },
156
+ "50271": {
157
+ "content": " ",
158
+ "lstrip": false,
159
+ "normalized": true,
160
+ "rstrip": false,
161
+ "single_word": false,
162
+ "special": false
163
+ },
164
+ "50272": {
165
+ "content": " ",
166
+ "lstrip": false,
167
+ "normalized": true,
168
+ "rstrip": false,
169
+ "single_word": false,
170
+ "special": false
171
+ },
172
+ "50273": {
173
+ "content": " ",
174
+ "lstrip": false,
175
+ "normalized": true,
176
+ "rstrip": false,
177
+ "single_word": false,
178
+ "special": false
179
+ },
180
+ "50274": {
181
+ "content": " ",
182
+ "lstrip": false,
183
+ "normalized": true,
184
+ "rstrip": false,
185
+ "single_word": false,
186
+ "special": false
187
+ },
188
+ "50275": {
189
+ "content": " ",
190
+ "lstrip": false,
191
+ "normalized": true,
192
+ "rstrip": false,
193
+ "single_word": false,
194
+ "special": false
195
+ },
196
+ "50276": {
197
+ "content": " ",
198
+ "lstrip": false,
199
+ "normalized": true,
200
+ "rstrip": false,
201
+ "single_word": false,
202
+ "special": false
203
+ }
204
+ },
205
+ "bos_token": "<|endoftext|>",
206
+ "chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '<|user|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'system' %}\n{{ '<|system|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'assistant' %}\n{{ '<|assistant|>\n' + message['content'] + eos_token }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '<|assistant|>' }}\n{% endif %}\n{% endfor %}",
207
+ "clean_up_tokenization_spaces": true,
208
+ "eos_token": "<|endoftext|>",
209
+ "model_max_length": 1000000000000000019884624838656,
210
+ "pad_token": "<|endoftext|>",
211
+ "tokenizer_class": "GPTNeoXTokenizer",
212
+ "unk_token": "<|endoftext|>"
213
+ }
trainer_state.json ADDED
@@ -0,0 +1,2719 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.4825064241945047,
5
+ "eval_steps": 500,
6
+ "global_step": 22500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 4.989018712114557e-05,
14
+ "loss": 0.6189,
15
+ "step": 50
16
+ },
17
+ {
18
+ "epoch": 0.01,
19
+ "learning_rate": 4.9780374242291136e-05,
20
+ "loss": 0.4018,
21
+ "step": 100
22
+ },
23
+ {
24
+ "epoch": 0.01,
25
+ "learning_rate": 4.967056136343671e-05,
26
+ "loss": 0.4266,
27
+ "step": 150
28
+ },
29
+ {
30
+ "epoch": 0.01,
31
+ "learning_rate": 4.9560748484582276e-05,
32
+ "loss": 0.4151,
33
+ "step": 200
34
+ },
35
+ {
36
+ "epoch": 0.02,
37
+ "learning_rate": 4.945093560572784e-05,
38
+ "loss": 0.3993,
39
+ "step": 250
40
+ },
41
+ {
42
+ "epoch": 0.02,
43
+ "learning_rate": 4.934112272687341e-05,
44
+ "loss": 0.3948,
45
+ "step": 300
46
+ },
47
+ {
48
+ "epoch": 0.02,
49
+ "learning_rate": 4.9231309848018984e-05,
50
+ "loss": 0.4181,
51
+ "step": 350
52
+ },
53
+ {
54
+ "epoch": 0.03,
55
+ "learning_rate": 4.9121496969164544e-05,
56
+ "loss": 0.3979,
57
+ "step": 400
58
+ },
59
+ {
60
+ "epoch": 0.03,
61
+ "learning_rate": 4.901168409031011e-05,
62
+ "loss": 0.4006,
63
+ "step": 450
64
+ },
65
+ {
66
+ "epoch": 0.03,
67
+ "learning_rate": 4.890187121145568e-05,
68
+ "loss": 0.4029,
69
+ "step": 500
70
+ },
71
+ {
72
+ "epoch": 0.04,
73
+ "learning_rate": 4.879205833260125e-05,
74
+ "loss": 0.378,
75
+ "step": 550
76
+ },
77
+ {
78
+ "epoch": 0.04,
79
+ "learning_rate": 4.868224545374682e-05,
80
+ "loss": 0.3894,
81
+ "step": 600
82
+ },
83
+ {
84
+ "epoch": 0.04,
85
+ "learning_rate": 4.8572432574892385e-05,
86
+ "loss": 0.3784,
87
+ "step": 650
88
+ },
89
+ {
90
+ "epoch": 0.05,
91
+ "learning_rate": 4.846261969603795e-05,
92
+ "loss": 0.3743,
93
+ "step": 700
94
+ },
95
+ {
96
+ "epoch": 0.05,
97
+ "learning_rate": 4.835280681718352e-05,
98
+ "loss": 0.3816,
99
+ "step": 750
100
+ },
101
+ {
102
+ "epoch": 0.05,
103
+ "learning_rate": 4.824299393832909e-05,
104
+ "loss": 0.3861,
105
+ "step": 800
106
+ },
107
+ {
108
+ "epoch": 0.06,
109
+ "learning_rate": 4.813318105947466e-05,
110
+ "loss": 0.3885,
111
+ "step": 850
112
+ },
113
+ {
114
+ "epoch": 0.06,
115
+ "learning_rate": 4.8023368180620226e-05,
116
+ "loss": 0.3948,
117
+ "step": 900
118
+ },
119
+ {
120
+ "epoch": 0.06,
121
+ "learning_rate": 4.791355530176579e-05,
122
+ "loss": 0.3545,
123
+ "step": 950
124
+ },
125
+ {
126
+ "epoch": 0.07,
127
+ "learning_rate": 4.7803742422911366e-05,
128
+ "loss": 0.3743,
129
+ "step": 1000
130
+ },
131
+ {
132
+ "epoch": 0.07,
133
+ "learning_rate": 4.769392954405693e-05,
134
+ "loss": 0.3843,
135
+ "step": 1050
136
+ },
137
+ {
138
+ "epoch": 0.07,
139
+ "learning_rate": 4.75841166652025e-05,
140
+ "loss": 0.3489,
141
+ "step": 1100
142
+ },
143
+ {
144
+ "epoch": 0.08,
145
+ "learning_rate": 4.7474303786348067e-05,
146
+ "loss": 0.3727,
147
+ "step": 1150
148
+ },
149
+ {
150
+ "epoch": 0.08,
151
+ "learning_rate": 4.7364490907493633e-05,
152
+ "loss": 0.3713,
153
+ "step": 1200
154
+ },
155
+ {
156
+ "epoch": 0.08,
157
+ "learning_rate": 4.72546780286392e-05,
158
+ "loss": 0.3662,
159
+ "step": 1250
160
+ },
161
+ {
162
+ "epoch": 0.09,
163
+ "learning_rate": 4.714486514978477e-05,
164
+ "loss": 0.3777,
165
+ "step": 1300
166
+ },
167
+ {
168
+ "epoch": 0.09,
169
+ "learning_rate": 4.7035052270930334e-05,
170
+ "loss": 0.3819,
171
+ "step": 1350
172
+ },
173
+ {
174
+ "epoch": 0.09,
175
+ "learning_rate": 4.69252393920759e-05,
176
+ "loss": 0.362,
177
+ "step": 1400
178
+ },
179
+ {
180
+ "epoch": 0.1,
181
+ "learning_rate": 4.6815426513221474e-05,
182
+ "loss": 0.3577,
183
+ "step": 1450
184
+ },
185
+ {
186
+ "epoch": 0.1,
187
+ "learning_rate": 4.670561363436704e-05,
188
+ "loss": 0.3879,
189
+ "step": 1500
190
+ },
191
+ {
192
+ "epoch": 0.1,
193
+ "learning_rate": 4.659580075551261e-05,
194
+ "loss": 0.3678,
195
+ "step": 1550
196
+ },
197
+ {
198
+ "epoch": 0.11,
199
+ "learning_rate": 4.6485987876658175e-05,
200
+ "loss": 0.3491,
201
+ "step": 1600
202
+ },
203
+ {
204
+ "epoch": 0.11,
205
+ "learning_rate": 4.637617499780375e-05,
206
+ "loss": 0.3904,
207
+ "step": 1650
208
+ },
209
+ {
210
+ "epoch": 0.11,
211
+ "learning_rate": 4.6266362118949315e-05,
212
+ "loss": 0.3495,
213
+ "step": 1700
214
+ },
215
+ {
216
+ "epoch": 0.12,
217
+ "learning_rate": 4.615654924009488e-05,
218
+ "loss": 0.3743,
219
+ "step": 1750
220
+ },
221
+ {
222
+ "epoch": 0.12,
223
+ "learning_rate": 4.604673636124045e-05,
224
+ "loss": 0.3814,
225
+ "step": 1800
226
+ },
227
+ {
228
+ "epoch": 0.12,
229
+ "learning_rate": 4.5936923482386016e-05,
230
+ "loss": 0.3495,
231
+ "step": 1850
232
+ },
233
+ {
234
+ "epoch": 0.13,
235
+ "learning_rate": 4.582711060353159e-05,
236
+ "loss": 0.3688,
237
+ "step": 1900
238
+ },
239
+ {
240
+ "epoch": 0.13,
241
+ "learning_rate": 4.5717297724677156e-05,
242
+ "loss": 0.3401,
243
+ "step": 1950
244
+ },
245
+ {
246
+ "epoch": 0.13,
247
+ "learning_rate": 4.5607484845822716e-05,
248
+ "loss": 0.3647,
249
+ "step": 2000
250
+ },
251
+ {
252
+ "epoch": 0.14,
253
+ "learning_rate": 4.549767196696828e-05,
254
+ "loss": 0.3716,
255
+ "step": 2050
256
+ },
257
+ {
258
+ "epoch": 0.14,
259
+ "learning_rate": 4.538785908811386e-05,
260
+ "loss": 0.373,
261
+ "step": 2100
262
+ },
263
+ {
264
+ "epoch": 0.14,
265
+ "learning_rate": 4.5278046209259424e-05,
266
+ "loss": 0.3471,
267
+ "step": 2150
268
+ },
269
+ {
270
+ "epoch": 0.14,
271
+ "learning_rate": 4.516823333040499e-05,
272
+ "loss": 0.3843,
273
+ "step": 2200
274
+ },
275
+ {
276
+ "epoch": 0.15,
277
+ "learning_rate": 4.505842045155056e-05,
278
+ "loss": 0.3629,
279
+ "step": 2250
280
+ },
281
+ {
282
+ "epoch": 0.15,
283
+ "learning_rate": 4.494860757269613e-05,
284
+ "loss": 0.3579,
285
+ "step": 2300
286
+ },
287
+ {
288
+ "epoch": 0.15,
289
+ "learning_rate": 4.48387946938417e-05,
290
+ "loss": 0.3373,
291
+ "step": 2350
292
+ },
293
+ {
294
+ "epoch": 0.16,
295
+ "learning_rate": 4.4728981814987264e-05,
296
+ "loss": 0.3509,
297
+ "step": 2400
298
+ },
299
+ {
300
+ "epoch": 0.16,
301
+ "learning_rate": 4.461916893613283e-05,
302
+ "loss": 0.3753,
303
+ "step": 2450
304
+ },
305
+ {
306
+ "epoch": 0.16,
307
+ "learning_rate": 4.45093560572784e-05,
308
+ "loss": 0.3421,
309
+ "step": 2500
310
+ },
311
+ {
312
+ "epoch": 0.17,
313
+ "learning_rate": 4.439954317842397e-05,
314
+ "loss": 0.3686,
315
+ "step": 2550
316
+ },
317
+ {
318
+ "epoch": 0.17,
319
+ "learning_rate": 4.428973029956954e-05,
320
+ "loss": 0.355,
321
+ "step": 2600
322
+ },
323
+ {
324
+ "epoch": 0.17,
325
+ "learning_rate": 4.4179917420715105e-05,
326
+ "loss": 0.3725,
327
+ "step": 2650
328
+ },
329
+ {
330
+ "epoch": 0.18,
331
+ "learning_rate": 4.407010454186067e-05,
332
+ "loss": 0.3495,
333
+ "step": 2700
334
+ },
335
+ {
336
+ "epoch": 0.18,
337
+ "learning_rate": 4.396029166300624e-05,
338
+ "loss": 0.3606,
339
+ "step": 2750
340
+ },
341
+ {
342
+ "epoch": 0.18,
343
+ "learning_rate": 4.3850478784151806e-05,
344
+ "loss": 0.3496,
345
+ "step": 2800
346
+ },
347
+ {
348
+ "epoch": 0.19,
349
+ "learning_rate": 4.374066590529737e-05,
350
+ "loss": 0.3606,
351
+ "step": 2850
352
+ },
353
+ {
354
+ "epoch": 0.19,
355
+ "learning_rate": 4.363085302644294e-05,
356
+ "loss": 0.3453,
357
+ "step": 2900
358
+ },
359
+ {
360
+ "epoch": 0.19,
361
+ "learning_rate": 4.352104014758851e-05,
362
+ "loss": 0.3617,
363
+ "step": 2950
364
+ },
365
+ {
366
+ "epoch": 0.2,
367
+ "learning_rate": 4.341122726873408e-05,
368
+ "loss": 0.3297,
369
+ "step": 3000
370
+ },
371
+ {
372
+ "epoch": 0.2,
373
+ "learning_rate": 4.330141438987965e-05,
374
+ "loss": 0.3448,
375
+ "step": 3050
376
+ },
377
+ {
378
+ "epoch": 0.2,
379
+ "learning_rate": 4.3191601511025214e-05,
380
+ "loss": 0.3631,
381
+ "step": 3100
382
+ },
383
+ {
384
+ "epoch": 0.21,
385
+ "learning_rate": 4.308178863217078e-05,
386
+ "loss": 0.352,
387
+ "step": 3150
388
+ },
389
+ {
390
+ "epoch": 0.21,
391
+ "learning_rate": 4.2971975753316354e-05,
392
+ "loss": 0.3482,
393
+ "step": 3200
394
+ },
395
+ {
396
+ "epoch": 0.21,
397
+ "learning_rate": 4.286216287446192e-05,
398
+ "loss": 0.3343,
399
+ "step": 3250
400
+ },
401
+ {
402
+ "epoch": 0.22,
403
+ "learning_rate": 4.275234999560749e-05,
404
+ "loss": 0.3204,
405
+ "step": 3300
406
+ },
407
+ {
408
+ "epoch": 0.22,
409
+ "learning_rate": 4.2642537116753055e-05,
410
+ "loss": 0.3442,
411
+ "step": 3350
412
+ },
413
+ {
414
+ "epoch": 0.22,
415
+ "learning_rate": 4.253272423789862e-05,
416
+ "loss": 0.3519,
417
+ "step": 3400
418
+ },
419
+ {
420
+ "epoch": 0.23,
421
+ "learning_rate": 4.2422911359044195e-05,
422
+ "loss": 0.3504,
423
+ "step": 3450
424
+ },
425
+ {
426
+ "epoch": 0.23,
427
+ "learning_rate": 4.231309848018976e-05,
428
+ "loss": 0.3612,
429
+ "step": 3500
430
+ },
431
+ {
432
+ "epoch": 0.23,
433
+ "learning_rate": 4.220328560133532e-05,
434
+ "loss": 0.3299,
435
+ "step": 3550
436
+ },
437
+ {
438
+ "epoch": 0.24,
439
+ "learning_rate": 4.2093472722480896e-05,
440
+ "loss": 0.3466,
441
+ "step": 3600
442
+ },
443
+ {
444
+ "epoch": 0.24,
445
+ "learning_rate": 4.198365984362646e-05,
446
+ "loss": 0.3459,
447
+ "step": 3650
448
+ },
449
+ {
450
+ "epoch": 0.24,
451
+ "learning_rate": 4.187384696477203e-05,
452
+ "loss": 0.3687,
453
+ "step": 3700
454
+ },
455
+ {
456
+ "epoch": 0.25,
457
+ "learning_rate": 4.1764034085917596e-05,
458
+ "loss": 0.3496,
459
+ "step": 3750
460
+ },
461
+ {
462
+ "epoch": 0.25,
463
+ "learning_rate": 4.165422120706316e-05,
464
+ "loss": 0.3558,
465
+ "step": 3800
466
+ },
467
+ {
468
+ "epoch": 0.25,
469
+ "learning_rate": 4.1544408328208736e-05,
470
+ "loss": 0.3649,
471
+ "step": 3850
472
+ },
473
+ {
474
+ "epoch": 0.26,
475
+ "learning_rate": 4.14345954493543e-05,
476
+ "loss": 0.3542,
477
+ "step": 3900
478
+ },
479
+ {
480
+ "epoch": 0.26,
481
+ "learning_rate": 4.132478257049987e-05,
482
+ "loss": 0.3415,
483
+ "step": 3950
484
+ },
485
+ {
486
+ "epoch": 0.26,
487
+ "learning_rate": 4.121496969164544e-05,
488
+ "loss": 0.364,
489
+ "step": 4000
490
+ },
491
+ {
492
+ "epoch": 0.27,
493
+ "learning_rate": 4.1105156812791004e-05,
494
+ "loss": 0.3532,
495
+ "step": 4050
496
+ },
497
+ {
498
+ "epoch": 0.27,
499
+ "learning_rate": 4.099534393393658e-05,
500
+ "loss": 0.3625,
501
+ "step": 4100
502
+ },
503
+ {
504
+ "epoch": 0.27,
505
+ "learning_rate": 4.0885531055082144e-05,
506
+ "loss": 0.3439,
507
+ "step": 4150
508
+ },
509
+ {
510
+ "epoch": 0.28,
511
+ "learning_rate": 4.077571817622771e-05,
512
+ "loss": 0.349,
513
+ "step": 4200
514
+ },
515
+ {
516
+ "epoch": 0.28,
517
+ "learning_rate": 4.066590529737328e-05,
518
+ "loss": 0.3468,
519
+ "step": 4250
520
+ },
521
+ {
522
+ "epoch": 0.28,
523
+ "learning_rate": 4.055609241851885e-05,
524
+ "loss": 0.3442,
525
+ "step": 4300
526
+ },
527
+ {
528
+ "epoch": 0.29,
529
+ "learning_rate": 4.044627953966441e-05,
530
+ "loss": 0.3383,
531
+ "step": 4350
532
+ },
533
+ {
534
+ "epoch": 0.29,
535
+ "learning_rate": 4.033646666080998e-05,
536
+ "loss": 0.3433,
537
+ "step": 4400
538
+ },
539
+ {
540
+ "epoch": 0.29,
541
+ "learning_rate": 4.0226653781955545e-05,
542
+ "loss": 0.3393,
543
+ "step": 4450
544
+ },
545
+ {
546
+ "epoch": 0.3,
547
+ "learning_rate": 4.011684090310112e-05,
548
+ "loss": 0.3616,
549
+ "step": 4500
550
+ },
551
+ {
552
+ "epoch": 0.3,
553
+ "learning_rate": 4.0007028024246686e-05,
554
+ "loss": 0.352,
555
+ "step": 4550
556
+ },
557
+ {
558
+ "epoch": 0.3,
559
+ "learning_rate": 3.989721514539225e-05,
560
+ "loss": 0.3269,
561
+ "step": 4600
562
+ },
563
+ {
564
+ "epoch": 0.31,
565
+ "learning_rate": 3.978740226653782e-05,
566
+ "loss": 0.3261,
567
+ "step": 4650
568
+ },
569
+ {
570
+ "epoch": 0.31,
571
+ "learning_rate": 3.9677589387683386e-05,
572
+ "loss": 0.3421,
573
+ "step": 4700
574
+ },
575
+ {
576
+ "epoch": 0.31,
577
+ "learning_rate": 3.956777650882896e-05,
578
+ "loss": 0.3286,
579
+ "step": 4750
580
+ },
581
+ {
582
+ "epoch": 0.32,
583
+ "learning_rate": 3.9457963629974527e-05,
584
+ "loss": 0.3317,
585
+ "step": 4800
586
+ },
587
+ {
588
+ "epoch": 0.32,
589
+ "learning_rate": 3.9348150751120093e-05,
590
+ "loss": 0.3489,
591
+ "step": 4850
592
+ },
593
+ {
594
+ "epoch": 0.32,
595
+ "learning_rate": 3.923833787226566e-05,
596
+ "loss": 0.3536,
597
+ "step": 4900
598
+ },
599
+ {
600
+ "epoch": 0.33,
601
+ "learning_rate": 3.9128524993411234e-05,
602
+ "loss": 0.3633,
603
+ "step": 4950
604
+ },
605
+ {
606
+ "epoch": 0.33,
607
+ "learning_rate": 3.90187121145568e-05,
608
+ "loss": 0.335,
609
+ "step": 5000
610
+ },
611
+ {
612
+ "epoch": 0.33,
613
+ "learning_rate": 3.890889923570237e-05,
614
+ "loss": 0.3373,
615
+ "step": 5050
616
+ },
617
+ {
618
+ "epoch": 0.34,
619
+ "learning_rate": 3.8799086356847934e-05,
620
+ "loss": 0.3477,
621
+ "step": 5100
622
+ },
623
+ {
624
+ "epoch": 0.34,
625
+ "learning_rate": 3.86892734779935e-05,
626
+ "loss": 0.3373,
627
+ "step": 5150
628
+ },
629
+ {
630
+ "epoch": 0.34,
631
+ "learning_rate": 3.857946059913907e-05,
632
+ "loss": 0.3318,
633
+ "step": 5200
634
+ },
635
+ {
636
+ "epoch": 0.35,
637
+ "learning_rate": 3.8469647720284635e-05,
638
+ "loss": 0.3324,
639
+ "step": 5250
640
+ },
641
+ {
642
+ "epoch": 0.35,
643
+ "learning_rate": 3.83598348414302e-05,
644
+ "loss": 0.3376,
645
+ "step": 5300
646
+ },
647
+ {
648
+ "epoch": 0.35,
649
+ "learning_rate": 3.825002196257577e-05,
650
+ "loss": 0.3623,
651
+ "step": 5350
652
+ },
653
+ {
654
+ "epoch": 0.36,
655
+ "learning_rate": 3.814020908372134e-05,
656
+ "loss": 0.3498,
657
+ "step": 5400
658
+ },
659
+ {
660
+ "epoch": 0.36,
661
+ "learning_rate": 3.803039620486691e-05,
662
+ "loss": 0.3599,
663
+ "step": 5450
664
+ },
665
+ {
666
+ "epoch": 0.36,
667
+ "learning_rate": 3.7920583326012476e-05,
668
+ "loss": 0.336,
669
+ "step": 5500
670
+ },
671
+ {
672
+ "epoch": 0.37,
673
+ "learning_rate": 3.781077044715804e-05,
674
+ "loss": 0.3426,
675
+ "step": 5550
676
+ },
677
+ {
678
+ "epoch": 0.37,
679
+ "learning_rate": 3.7700957568303616e-05,
680
+ "loss": 0.3327,
681
+ "step": 5600
682
+ },
683
+ {
684
+ "epoch": 0.37,
685
+ "learning_rate": 3.759114468944918e-05,
686
+ "loss": 0.3352,
687
+ "step": 5650
688
+ },
689
+ {
690
+ "epoch": 0.38,
691
+ "learning_rate": 3.748133181059475e-05,
692
+ "loss": 0.3165,
693
+ "step": 5700
694
+ },
695
+ {
696
+ "epoch": 0.38,
697
+ "learning_rate": 3.737151893174032e-05,
698
+ "loss": 0.3335,
699
+ "step": 5750
700
+ },
701
+ {
702
+ "epoch": 0.38,
703
+ "learning_rate": 3.7261706052885884e-05,
704
+ "loss": 0.3475,
705
+ "step": 5800
706
+ },
707
+ {
708
+ "epoch": 0.39,
709
+ "learning_rate": 3.715189317403146e-05,
710
+ "loss": 0.3335,
711
+ "step": 5850
712
+ },
713
+ {
714
+ "epoch": 0.39,
715
+ "learning_rate": 3.704208029517702e-05,
716
+ "loss": 0.3429,
717
+ "step": 5900
718
+ },
719
+ {
720
+ "epoch": 0.39,
721
+ "learning_rate": 3.6932267416322584e-05,
722
+ "loss": 0.3355,
723
+ "step": 5950
724
+ },
725
+ {
726
+ "epoch": 0.4,
727
+ "learning_rate": 3.682245453746815e-05,
728
+ "loss": 0.3102,
729
+ "step": 6000
730
+ },
731
+ {
732
+ "epoch": 0.4,
733
+ "learning_rate": 3.6712641658613724e-05,
734
+ "loss": 0.3491,
735
+ "step": 6050
736
+ },
737
+ {
738
+ "epoch": 0.4,
739
+ "learning_rate": 3.660282877975929e-05,
740
+ "loss": 0.3349,
741
+ "step": 6100
742
+ },
743
+ {
744
+ "epoch": 0.41,
745
+ "learning_rate": 3.649301590090486e-05,
746
+ "loss": 0.3509,
747
+ "step": 6150
748
+ },
749
+ {
750
+ "epoch": 0.41,
751
+ "learning_rate": 3.6383203022050425e-05,
752
+ "loss": 0.3374,
753
+ "step": 6200
754
+ },
755
+ {
756
+ "epoch": 0.41,
757
+ "learning_rate": 3.6273390143196e-05,
758
+ "loss": 0.3242,
759
+ "step": 6250
760
+ },
761
+ {
762
+ "epoch": 0.42,
763
+ "learning_rate": 3.6163577264341565e-05,
764
+ "loss": 0.3425,
765
+ "step": 6300
766
+ },
767
+ {
768
+ "epoch": 0.42,
769
+ "learning_rate": 3.605376438548713e-05,
770
+ "loss": 0.3274,
771
+ "step": 6350
772
+ },
773
+ {
774
+ "epoch": 0.42,
775
+ "learning_rate": 3.59439515066327e-05,
776
+ "loss": 0.3619,
777
+ "step": 6400
778
+ },
779
+ {
780
+ "epoch": 0.42,
781
+ "learning_rate": 3.5834138627778266e-05,
782
+ "loss": 0.3385,
783
+ "step": 6450
784
+ },
785
+ {
786
+ "epoch": 0.43,
787
+ "learning_rate": 3.572432574892384e-05,
788
+ "loss": 0.3236,
789
+ "step": 6500
790
+ },
791
+ {
792
+ "epoch": 0.43,
793
+ "learning_rate": 3.5614512870069406e-05,
794
+ "loss": 0.331,
795
+ "step": 6550
796
+ },
797
+ {
798
+ "epoch": 0.43,
799
+ "learning_rate": 3.550469999121497e-05,
800
+ "loss": 0.3379,
801
+ "step": 6600
802
+ },
803
+ {
804
+ "epoch": 0.44,
805
+ "learning_rate": 3.539488711236054e-05,
806
+ "loss": 0.3284,
807
+ "step": 6650
808
+ },
809
+ {
810
+ "epoch": 0.44,
811
+ "learning_rate": 3.528507423350611e-05,
812
+ "loss": 0.3315,
813
+ "step": 6700
814
+ },
815
+ {
816
+ "epoch": 0.44,
817
+ "learning_rate": 3.5175261354651674e-05,
818
+ "loss": 0.3268,
819
+ "step": 6750
820
+ },
821
+ {
822
+ "epoch": 0.45,
823
+ "learning_rate": 3.506544847579724e-05,
824
+ "loss": 0.3439,
825
+ "step": 6800
826
+ },
827
+ {
828
+ "epoch": 0.45,
829
+ "learning_rate": 3.495563559694281e-05,
830
+ "loss": 0.3399,
831
+ "step": 6850
832
+ },
833
+ {
834
+ "epoch": 0.45,
835
+ "learning_rate": 3.484582271808838e-05,
836
+ "loss": 0.3374,
837
+ "step": 6900
838
+ },
839
+ {
840
+ "epoch": 0.46,
841
+ "learning_rate": 3.473600983923395e-05,
842
+ "loss": 0.3409,
843
+ "step": 6950
844
+ },
845
+ {
846
+ "epoch": 0.46,
847
+ "learning_rate": 3.4626196960379515e-05,
848
+ "loss": 0.3196,
849
+ "step": 7000
850
+ },
851
+ {
852
+ "epoch": 0.46,
853
+ "learning_rate": 3.451638408152508e-05,
854
+ "loss": 0.3247,
855
+ "step": 7050
856
+ },
857
+ {
858
+ "epoch": 0.47,
859
+ "learning_rate": 3.440657120267065e-05,
860
+ "loss": 0.3469,
861
+ "step": 7100
862
+ },
863
+ {
864
+ "epoch": 0.47,
865
+ "learning_rate": 3.429675832381622e-05,
866
+ "loss": 0.3149,
867
+ "step": 7150
868
+ },
869
+ {
870
+ "epoch": 0.47,
871
+ "learning_rate": 3.418694544496179e-05,
872
+ "loss": 0.3413,
873
+ "step": 7200
874
+ },
875
+ {
876
+ "epoch": 0.48,
877
+ "learning_rate": 3.4077132566107355e-05,
878
+ "loss": 0.3285,
879
+ "step": 7250
880
+ },
881
+ {
882
+ "epoch": 0.48,
883
+ "learning_rate": 3.396731968725292e-05,
884
+ "loss": 0.312,
885
+ "step": 7300
886
+ },
887
+ {
888
+ "epoch": 0.48,
889
+ "learning_rate": 3.3857506808398496e-05,
890
+ "loss": 0.3355,
891
+ "step": 7350
892
+ },
893
+ {
894
+ "epoch": 0.49,
895
+ "learning_rate": 3.374769392954406e-05,
896
+ "loss": 0.3402,
897
+ "step": 7400
898
+ },
899
+ {
900
+ "epoch": 0.49,
901
+ "learning_rate": 3.363788105068963e-05,
902
+ "loss": 0.3373,
903
+ "step": 7450
904
+ },
905
+ {
906
+ "epoch": 0.49,
907
+ "learning_rate": 3.352806817183519e-05,
908
+ "loss": 0.3311,
909
+ "step": 7500
910
+ },
911
+ {
912
+ "epoch": 0.5,
913
+ "learning_rate": 3.341825529298076e-05,
914
+ "loss": 0.3213,
915
+ "step": 7550
916
+ },
917
+ {
918
+ "epoch": 0.5,
919
+ "learning_rate": 3.330844241412633e-05,
920
+ "loss": 0.3323,
921
+ "step": 7600
922
+ },
923
+ {
924
+ "epoch": 0.5,
925
+ "learning_rate": 3.31986295352719e-05,
926
+ "loss": 0.3307,
927
+ "step": 7650
928
+ },
929
+ {
930
+ "epoch": 0.51,
931
+ "learning_rate": 3.3088816656417464e-05,
932
+ "loss": 0.3168,
933
+ "step": 7700
934
+ },
935
+ {
936
+ "epoch": 0.51,
937
+ "learning_rate": 3.297900377756303e-05,
938
+ "loss": 0.3287,
939
+ "step": 7750
940
+ },
941
+ {
942
+ "epoch": 0.51,
943
+ "learning_rate": 3.2869190898708604e-05,
944
+ "loss": 0.3267,
945
+ "step": 7800
946
+ },
947
+ {
948
+ "epoch": 0.52,
949
+ "learning_rate": 3.275937801985417e-05,
950
+ "loss": 0.331,
951
+ "step": 7850
952
+ },
953
+ {
954
+ "epoch": 0.52,
955
+ "learning_rate": 3.264956514099974e-05,
956
+ "loss": 0.335,
957
+ "step": 7900
958
+ },
959
+ {
960
+ "epoch": 0.52,
961
+ "learning_rate": 3.2539752262145305e-05,
962
+ "loss": 0.3332,
963
+ "step": 7950
964
+ },
965
+ {
966
+ "epoch": 0.53,
967
+ "learning_rate": 3.242993938329087e-05,
968
+ "loss": 0.3442,
969
+ "step": 8000
970
+ },
971
+ {
972
+ "epoch": 0.53,
973
+ "learning_rate": 3.2320126504436445e-05,
974
+ "loss": 0.317,
975
+ "step": 8050
976
+ },
977
+ {
978
+ "epoch": 0.53,
979
+ "learning_rate": 3.221031362558201e-05,
980
+ "loss": 0.3243,
981
+ "step": 8100
982
+ },
983
+ {
984
+ "epoch": 0.54,
985
+ "learning_rate": 3.210050074672758e-05,
986
+ "loss": 0.3328,
987
+ "step": 8150
988
+ },
989
+ {
990
+ "epoch": 0.54,
991
+ "learning_rate": 3.1990687867873146e-05,
992
+ "loss": 0.3376,
993
+ "step": 8200
994
+ },
995
+ {
996
+ "epoch": 0.54,
997
+ "learning_rate": 3.188087498901872e-05,
998
+ "loss": 0.3387,
999
+ "step": 8250
1000
+ },
1001
+ {
1002
+ "epoch": 0.55,
1003
+ "learning_rate": 3.177106211016428e-05,
1004
+ "loss": 0.3092,
1005
+ "step": 8300
1006
+ },
1007
+ {
1008
+ "epoch": 0.55,
1009
+ "learning_rate": 3.1661249231309846e-05,
1010
+ "loss": 0.3241,
1011
+ "step": 8350
1012
+ },
1013
+ {
1014
+ "epoch": 0.55,
1015
+ "learning_rate": 3.155143635245541e-05,
1016
+ "loss": 0.3219,
1017
+ "step": 8400
1018
+ },
1019
+ {
1020
+ "epoch": 0.56,
1021
+ "learning_rate": 3.1441623473600987e-05,
1022
+ "loss": 0.3256,
1023
+ "step": 8450
1024
+ },
1025
+ {
1026
+ "epoch": 0.56,
1027
+ "learning_rate": 3.133181059474655e-05,
1028
+ "loss": 0.3249,
1029
+ "step": 8500
1030
+ },
1031
+ {
1032
+ "epoch": 0.56,
1033
+ "learning_rate": 3.122199771589212e-05,
1034
+ "loss": 0.3265,
1035
+ "step": 8550
1036
+ },
1037
+ {
1038
+ "epoch": 0.57,
1039
+ "learning_rate": 3.111218483703769e-05,
1040
+ "loss": 0.3242,
1041
+ "step": 8600
1042
+ },
1043
+ {
1044
+ "epoch": 0.57,
1045
+ "learning_rate": 3.1002371958183254e-05,
1046
+ "loss": 0.3568,
1047
+ "step": 8650
1048
+ },
1049
+ {
1050
+ "epoch": 0.57,
1051
+ "learning_rate": 3.089255907932883e-05,
1052
+ "loss": 0.3261,
1053
+ "step": 8700
1054
+ },
1055
+ {
1056
+ "epoch": 0.58,
1057
+ "learning_rate": 3.0782746200474394e-05,
1058
+ "loss": 0.3197,
1059
+ "step": 8750
1060
+ },
1061
+ {
1062
+ "epoch": 0.58,
1063
+ "learning_rate": 3.067293332161996e-05,
1064
+ "loss": 0.3419,
1065
+ "step": 8800
1066
+ },
1067
+ {
1068
+ "epoch": 0.58,
1069
+ "learning_rate": 3.056312044276553e-05,
1070
+ "loss": 0.3448,
1071
+ "step": 8850
1072
+ },
1073
+ {
1074
+ "epoch": 0.59,
1075
+ "learning_rate": 3.0453307563911098e-05,
1076
+ "loss": 0.3506,
1077
+ "step": 8900
1078
+ },
1079
+ {
1080
+ "epoch": 0.59,
1081
+ "learning_rate": 3.034349468505667e-05,
1082
+ "loss": 0.3337,
1083
+ "step": 8950
1084
+ },
1085
+ {
1086
+ "epoch": 0.59,
1087
+ "learning_rate": 3.0233681806202235e-05,
1088
+ "loss": 0.3214,
1089
+ "step": 9000
1090
+ },
1091
+ {
1092
+ "epoch": 0.6,
1093
+ "learning_rate": 3.01238689273478e-05,
1094
+ "loss": 0.3358,
1095
+ "step": 9050
1096
+ },
1097
+ {
1098
+ "epoch": 0.6,
1099
+ "learning_rate": 3.0014056048493365e-05,
1100
+ "loss": 0.3267,
1101
+ "step": 9100
1102
+ },
1103
+ {
1104
+ "epoch": 0.6,
1105
+ "learning_rate": 2.9904243169638936e-05,
1106
+ "loss": 0.3231,
1107
+ "step": 9150
1108
+ },
1109
+ {
1110
+ "epoch": 0.61,
1111
+ "learning_rate": 2.9794430290784503e-05,
1112
+ "loss": 0.3071,
1113
+ "step": 9200
1114
+ },
1115
+ {
1116
+ "epoch": 0.61,
1117
+ "learning_rate": 2.9684617411930073e-05,
1118
+ "loss": 0.3306,
1119
+ "step": 9250
1120
+ },
1121
+ {
1122
+ "epoch": 0.61,
1123
+ "learning_rate": 2.957480453307564e-05,
1124
+ "loss": 0.3086,
1125
+ "step": 9300
1126
+ },
1127
+ {
1128
+ "epoch": 0.62,
1129
+ "learning_rate": 2.9464991654221206e-05,
1130
+ "loss": 0.3199,
1131
+ "step": 9350
1132
+ },
1133
+ {
1134
+ "epoch": 0.62,
1135
+ "learning_rate": 2.9355178775366777e-05,
1136
+ "loss": 0.3297,
1137
+ "step": 9400
1138
+ },
1139
+ {
1140
+ "epoch": 0.62,
1141
+ "learning_rate": 2.9245365896512343e-05,
1142
+ "loss": 0.3245,
1143
+ "step": 9450
1144
+ },
1145
+ {
1146
+ "epoch": 0.63,
1147
+ "learning_rate": 2.9135553017657914e-05,
1148
+ "loss": 0.3161,
1149
+ "step": 9500
1150
+ },
1151
+ {
1152
+ "epoch": 0.63,
1153
+ "learning_rate": 2.902574013880348e-05,
1154
+ "loss": 0.3199,
1155
+ "step": 9550
1156
+ },
1157
+ {
1158
+ "epoch": 0.63,
1159
+ "learning_rate": 2.891592725994905e-05,
1160
+ "loss": 0.3292,
1161
+ "step": 9600
1162
+ },
1163
+ {
1164
+ "epoch": 0.64,
1165
+ "learning_rate": 2.8806114381094618e-05,
1166
+ "loss": 0.3242,
1167
+ "step": 9650
1168
+ },
1169
+ {
1170
+ "epoch": 0.64,
1171
+ "learning_rate": 2.8696301502240184e-05,
1172
+ "loss": 0.3254,
1173
+ "step": 9700
1174
+ },
1175
+ {
1176
+ "epoch": 0.64,
1177
+ "learning_rate": 2.8586488623385755e-05,
1178
+ "loss": 0.341,
1179
+ "step": 9750
1180
+ },
1181
+ {
1182
+ "epoch": 0.65,
1183
+ "learning_rate": 2.847667574453132e-05,
1184
+ "loss": 0.3195,
1185
+ "step": 9800
1186
+ },
1187
+ {
1188
+ "epoch": 0.65,
1189
+ "learning_rate": 2.8366862865676885e-05,
1190
+ "loss": 0.3351,
1191
+ "step": 9850
1192
+ },
1193
+ {
1194
+ "epoch": 0.65,
1195
+ "learning_rate": 2.8257049986822455e-05,
1196
+ "loss": 0.333,
1197
+ "step": 9900
1198
+ },
1199
+ {
1200
+ "epoch": 0.66,
1201
+ "learning_rate": 2.8147237107968022e-05,
1202
+ "loss": 0.3332,
1203
+ "step": 9950
1204
+ },
1205
+ {
1206
+ "epoch": 0.66,
1207
+ "learning_rate": 2.803742422911359e-05,
1208
+ "loss": 0.3279,
1209
+ "step": 10000
1210
+ },
1211
+ {
1212
+ "epoch": 0.66,
1213
+ "learning_rate": 2.792761135025916e-05,
1214
+ "loss": 0.3566,
1215
+ "step": 10050
1216
+ },
1217
+ {
1218
+ "epoch": 0.67,
1219
+ "learning_rate": 2.7817798471404726e-05,
1220
+ "loss": 0.3266,
1221
+ "step": 10100
1222
+ },
1223
+ {
1224
+ "epoch": 0.67,
1225
+ "learning_rate": 2.7707985592550296e-05,
1226
+ "loss": 0.3283,
1227
+ "step": 10150
1228
+ },
1229
+ {
1230
+ "epoch": 0.67,
1231
+ "learning_rate": 2.7598172713695863e-05,
1232
+ "loss": 0.3302,
1233
+ "step": 10200
1234
+ },
1235
+ {
1236
+ "epoch": 0.68,
1237
+ "learning_rate": 2.7488359834841433e-05,
1238
+ "loss": 0.3278,
1239
+ "step": 10250
1240
+ },
1241
+ {
1242
+ "epoch": 0.68,
1243
+ "learning_rate": 2.7378546955987e-05,
1244
+ "loss": 0.3144,
1245
+ "step": 10300
1246
+ },
1247
+ {
1248
+ "epoch": 0.68,
1249
+ "learning_rate": 2.7268734077132567e-05,
1250
+ "loss": 0.3418,
1251
+ "step": 10350
1252
+ },
1253
+ {
1254
+ "epoch": 0.69,
1255
+ "learning_rate": 2.7158921198278137e-05,
1256
+ "loss": 0.3218,
1257
+ "step": 10400
1258
+ },
1259
+ {
1260
+ "epoch": 0.69,
1261
+ "learning_rate": 2.7049108319423704e-05,
1262
+ "loss": 0.3113,
1263
+ "step": 10450
1264
+ },
1265
+ {
1266
+ "epoch": 0.69,
1267
+ "learning_rate": 2.6939295440569274e-05,
1268
+ "loss": 0.3078,
1269
+ "step": 10500
1270
+ },
1271
+ {
1272
+ "epoch": 0.7,
1273
+ "learning_rate": 2.682948256171484e-05,
1274
+ "loss": 0.3112,
1275
+ "step": 10550
1276
+ },
1277
+ {
1278
+ "epoch": 0.7,
1279
+ "learning_rate": 2.671966968286041e-05,
1280
+ "loss": 0.3261,
1281
+ "step": 10600
1282
+ },
1283
+ {
1284
+ "epoch": 0.7,
1285
+ "learning_rate": 2.660985680400597e-05,
1286
+ "loss": 0.3324,
1287
+ "step": 10650
1288
+ },
1289
+ {
1290
+ "epoch": 0.71,
1291
+ "learning_rate": 2.650004392515154e-05,
1292
+ "loss": 0.3323,
1293
+ "step": 10700
1294
+ },
1295
+ {
1296
+ "epoch": 0.71,
1297
+ "learning_rate": 2.6390231046297108e-05,
1298
+ "loss": 0.337,
1299
+ "step": 10750
1300
+ },
1301
+ {
1302
+ "epoch": 0.71,
1303
+ "learning_rate": 2.628041816744268e-05,
1304
+ "loss": 0.3037,
1305
+ "step": 10800
1306
+ },
1307
+ {
1308
+ "epoch": 0.71,
1309
+ "learning_rate": 2.6170605288588245e-05,
1310
+ "loss": 0.3261,
1311
+ "step": 10850
1312
+ },
1313
+ {
1314
+ "epoch": 0.72,
1315
+ "learning_rate": 2.6060792409733815e-05,
1316
+ "loss": 0.3229,
1317
+ "step": 10900
1318
+ },
1319
+ {
1320
+ "epoch": 0.72,
1321
+ "learning_rate": 2.5950979530879382e-05,
1322
+ "loss": 0.3178,
1323
+ "step": 10950
1324
+ },
1325
+ {
1326
+ "epoch": 0.72,
1327
+ "learning_rate": 2.584116665202495e-05,
1328
+ "loss": 0.3231,
1329
+ "step": 11000
1330
+ },
1331
+ {
1332
+ "epoch": 0.73,
1333
+ "learning_rate": 2.573135377317052e-05,
1334
+ "loss": 0.332,
1335
+ "step": 11050
1336
+ },
1337
+ {
1338
+ "epoch": 0.73,
1339
+ "learning_rate": 2.5621540894316086e-05,
1340
+ "loss": 0.3495,
1341
+ "step": 11100
1342
+ },
1343
+ {
1344
+ "epoch": 0.73,
1345
+ "learning_rate": 2.5511728015461656e-05,
1346
+ "loss": 0.3164,
1347
+ "step": 11150
1348
+ },
1349
+ {
1350
+ "epoch": 0.74,
1351
+ "learning_rate": 2.5401915136607223e-05,
1352
+ "loss": 0.3304,
1353
+ "step": 11200
1354
+ },
1355
+ {
1356
+ "epoch": 0.74,
1357
+ "learning_rate": 2.5292102257752793e-05,
1358
+ "loss": 0.3075,
1359
+ "step": 11250
1360
+ },
1361
+ {
1362
+ "epoch": 0.74,
1363
+ "learning_rate": 2.518228937889836e-05,
1364
+ "loss": 0.3179,
1365
+ "step": 11300
1366
+ },
1367
+ {
1368
+ "epoch": 0.75,
1369
+ "learning_rate": 2.507247650004393e-05,
1370
+ "loss": 0.3306,
1371
+ "step": 11350
1372
+ },
1373
+ {
1374
+ "epoch": 0.75,
1375
+ "learning_rate": 2.4962663621189494e-05,
1376
+ "loss": 0.3231,
1377
+ "step": 11400
1378
+ },
1379
+ {
1380
+ "epoch": 0.75,
1381
+ "learning_rate": 2.4852850742335064e-05,
1382
+ "loss": 0.3179,
1383
+ "step": 11450
1384
+ },
1385
+ {
1386
+ "epoch": 0.76,
1387
+ "learning_rate": 2.474303786348063e-05,
1388
+ "loss": 0.3164,
1389
+ "step": 11500
1390
+ },
1391
+ {
1392
+ "epoch": 0.76,
1393
+ "learning_rate": 2.4633224984626198e-05,
1394
+ "loss": 0.3158,
1395
+ "step": 11550
1396
+ },
1397
+ {
1398
+ "epoch": 0.76,
1399
+ "learning_rate": 2.4523412105771765e-05,
1400
+ "loss": 0.306,
1401
+ "step": 11600
1402
+ },
1403
+ {
1404
+ "epoch": 0.77,
1405
+ "learning_rate": 2.441359922691733e-05,
1406
+ "loss": 0.3362,
1407
+ "step": 11650
1408
+ },
1409
+ {
1410
+ "epoch": 0.77,
1411
+ "learning_rate": 2.43037863480629e-05,
1412
+ "loss": 0.3257,
1413
+ "step": 11700
1414
+ },
1415
+ {
1416
+ "epoch": 0.77,
1417
+ "learning_rate": 2.419397346920847e-05,
1418
+ "loss": 0.3358,
1419
+ "step": 11750
1420
+ },
1421
+ {
1422
+ "epoch": 0.78,
1423
+ "learning_rate": 2.408416059035404e-05,
1424
+ "loss": 0.32,
1425
+ "step": 11800
1426
+ },
1427
+ {
1428
+ "epoch": 0.78,
1429
+ "learning_rate": 2.3974347711499606e-05,
1430
+ "loss": 0.3149,
1431
+ "step": 11850
1432
+ },
1433
+ {
1434
+ "epoch": 0.78,
1435
+ "learning_rate": 2.3864534832645176e-05,
1436
+ "loss": 0.3087,
1437
+ "step": 11900
1438
+ },
1439
+ {
1440
+ "epoch": 0.79,
1441
+ "learning_rate": 2.3754721953790743e-05,
1442
+ "loss": 0.3213,
1443
+ "step": 11950
1444
+ },
1445
+ {
1446
+ "epoch": 0.79,
1447
+ "learning_rate": 2.364490907493631e-05,
1448
+ "loss": 0.3196,
1449
+ "step": 12000
1450
+ },
1451
+ {
1452
+ "epoch": 0.79,
1453
+ "learning_rate": 2.3535096196081876e-05,
1454
+ "loss": 0.2998,
1455
+ "step": 12050
1456
+ },
1457
+ {
1458
+ "epoch": 0.8,
1459
+ "learning_rate": 2.3425283317227446e-05,
1460
+ "loss": 0.3216,
1461
+ "step": 12100
1462
+ },
1463
+ {
1464
+ "epoch": 0.8,
1465
+ "learning_rate": 2.3315470438373013e-05,
1466
+ "loss": 0.3236,
1467
+ "step": 12150
1468
+ },
1469
+ {
1470
+ "epoch": 0.8,
1471
+ "learning_rate": 2.320565755951858e-05,
1472
+ "loss": 0.344,
1473
+ "step": 12200
1474
+ },
1475
+ {
1476
+ "epoch": 0.81,
1477
+ "learning_rate": 2.309584468066415e-05,
1478
+ "loss": 0.3096,
1479
+ "step": 12250
1480
+ },
1481
+ {
1482
+ "epoch": 0.81,
1483
+ "learning_rate": 2.2986031801809717e-05,
1484
+ "loss": 0.3119,
1485
+ "step": 12300
1486
+ },
1487
+ {
1488
+ "epoch": 0.81,
1489
+ "learning_rate": 2.2876218922955287e-05,
1490
+ "loss": 0.3272,
1491
+ "step": 12350
1492
+ },
1493
+ {
1494
+ "epoch": 0.82,
1495
+ "learning_rate": 2.276640604410085e-05,
1496
+ "loss": 0.3193,
1497
+ "step": 12400
1498
+ },
1499
+ {
1500
+ "epoch": 0.82,
1501
+ "learning_rate": 2.265659316524642e-05,
1502
+ "loss": 0.31,
1503
+ "step": 12450
1504
+ },
1505
+ {
1506
+ "epoch": 0.82,
1507
+ "learning_rate": 2.2546780286391988e-05,
1508
+ "loss": 0.3113,
1509
+ "step": 12500
1510
+ },
1511
+ {
1512
+ "epoch": 0.83,
1513
+ "learning_rate": 2.2436967407537558e-05,
1514
+ "loss": 0.3177,
1515
+ "step": 12550
1516
+ },
1517
+ {
1518
+ "epoch": 0.83,
1519
+ "learning_rate": 2.2327154528683125e-05,
1520
+ "loss": 0.3284,
1521
+ "step": 12600
1522
+ },
1523
+ {
1524
+ "epoch": 0.83,
1525
+ "learning_rate": 2.2217341649828692e-05,
1526
+ "loss": 0.3174,
1527
+ "step": 12650
1528
+ },
1529
+ {
1530
+ "epoch": 0.84,
1531
+ "learning_rate": 2.2107528770974262e-05,
1532
+ "loss": 0.3145,
1533
+ "step": 12700
1534
+ },
1535
+ {
1536
+ "epoch": 0.84,
1537
+ "learning_rate": 2.199771589211983e-05,
1538
+ "loss": 0.3258,
1539
+ "step": 12750
1540
+ },
1541
+ {
1542
+ "epoch": 0.84,
1543
+ "learning_rate": 2.1887903013265396e-05,
1544
+ "loss": 0.3307,
1545
+ "step": 12800
1546
+ },
1547
+ {
1548
+ "epoch": 0.85,
1549
+ "learning_rate": 2.1778090134410963e-05,
1550
+ "loss": 0.319,
1551
+ "step": 12850
1552
+ },
1553
+ {
1554
+ "epoch": 0.85,
1555
+ "learning_rate": 2.1668277255556533e-05,
1556
+ "loss": 0.3349,
1557
+ "step": 12900
1558
+ },
1559
+ {
1560
+ "epoch": 0.85,
1561
+ "learning_rate": 2.15584643767021e-05,
1562
+ "loss": 0.3216,
1563
+ "step": 12950
1564
+ },
1565
+ {
1566
+ "epoch": 0.86,
1567
+ "learning_rate": 2.144865149784767e-05,
1568
+ "loss": 0.3324,
1569
+ "step": 13000
1570
+ },
1571
+ {
1572
+ "epoch": 0.86,
1573
+ "learning_rate": 2.1338838618993237e-05,
1574
+ "loss": 0.3192,
1575
+ "step": 13050
1576
+ },
1577
+ {
1578
+ "epoch": 0.86,
1579
+ "learning_rate": 2.1229025740138807e-05,
1580
+ "loss": 0.3233,
1581
+ "step": 13100
1582
+ },
1583
+ {
1584
+ "epoch": 0.87,
1585
+ "learning_rate": 2.1119212861284374e-05,
1586
+ "loss": 0.3202,
1587
+ "step": 13150
1588
+ },
1589
+ {
1590
+ "epoch": 0.87,
1591
+ "learning_rate": 2.100939998242994e-05,
1592
+ "loss": 0.3126,
1593
+ "step": 13200
1594
+ },
1595
+ {
1596
+ "epoch": 0.87,
1597
+ "learning_rate": 2.0899587103575507e-05,
1598
+ "loss": 0.3187,
1599
+ "step": 13250
1600
+ },
1601
+ {
1602
+ "epoch": 0.88,
1603
+ "learning_rate": 2.0789774224721074e-05,
1604
+ "loss": 0.3324,
1605
+ "step": 13300
1606
+ },
1607
+ {
1608
+ "epoch": 0.88,
1609
+ "learning_rate": 2.0679961345866644e-05,
1610
+ "loss": 0.3228,
1611
+ "step": 13350
1612
+ },
1613
+ {
1614
+ "epoch": 0.88,
1615
+ "learning_rate": 2.057014846701221e-05,
1616
+ "loss": 0.3118,
1617
+ "step": 13400
1618
+ },
1619
+ {
1620
+ "epoch": 0.89,
1621
+ "learning_rate": 2.046033558815778e-05,
1622
+ "loss": 0.3324,
1623
+ "step": 13450
1624
+ },
1625
+ {
1626
+ "epoch": 0.89,
1627
+ "learning_rate": 2.0350522709303348e-05,
1628
+ "loss": 0.3133,
1629
+ "step": 13500
1630
+ },
1631
+ {
1632
+ "epoch": 0.89,
1633
+ "learning_rate": 2.024070983044892e-05,
1634
+ "loss": 0.3205,
1635
+ "step": 13550
1636
+ },
1637
+ {
1638
+ "epoch": 0.9,
1639
+ "learning_rate": 2.0130896951594482e-05,
1640
+ "loss": 0.3024,
1641
+ "step": 13600
1642
+ },
1643
+ {
1644
+ "epoch": 0.9,
1645
+ "learning_rate": 2.0021084072740052e-05,
1646
+ "loss": 0.3093,
1647
+ "step": 13650
1648
+ },
1649
+ {
1650
+ "epoch": 0.9,
1651
+ "learning_rate": 1.991127119388562e-05,
1652
+ "loss": 0.322,
1653
+ "step": 13700
1654
+ },
1655
+ {
1656
+ "epoch": 0.91,
1657
+ "learning_rate": 1.980145831503119e-05,
1658
+ "loss": 0.3094,
1659
+ "step": 13750
1660
+ },
1661
+ {
1662
+ "epoch": 0.91,
1663
+ "learning_rate": 1.9691645436176756e-05,
1664
+ "loss": 0.3242,
1665
+ "step": 13800
1666
+ },
1667
+ {
1668
+ "epoch": 0.91,
1669
+ "learning_rate": 1.9581832557322323e-05,
1670
+ "loss": 0.3485,
1671
+ "step": 13850
1672
+ },
1673
+ {
1674
+ "epoch": 0.92,
1675
+ "learning_rate": 1.9472019678467893e-05,
1676
+ "loss": 0.3369,
1677
+ "step": 13900
1678
+ },
1679
+ {
1680
+ "epoch": 0.92,
1681
+ "learning_rate": 1.936220679961346e-05,
1682
+ "loss": 0.3137,
1683
+ "step": 13950
1684
+ },
1685
+ {
1686
+ "epoch": 0.92,
1687
+ "learning_rate": 1.9252393920759027e-05,
1688
+ "loss": 0.3176,
1689
+ "step": 14000
1690
+ },
1691
+ {
1692
+ "epoch": 0.93,
1693
+ "learning_rate": 1.9142581041904594e-05,
1694
+ "loss": 0.3194,
1695
+ "step": 14050
1696
+ },
1697
+ {
1698
+ "epoch": 0.93,
1699
+ "learning_rate": 1.9032768163050164e-05,
1700
+ "loss": 0.3208,
1701
+ "step": 14100
1702
+ },
1703
+ {
1704
+ "epoch": 0.93,
1705
+ "learning_rate": 1.892295528419573e-05,
1706
+ "loss": 0.3322,
1707
+ "step": 14150
1708
+ },
1709
+ {
1710
+ "epoch": 0.94,
1711
+ "learning_rate": 1.88131424053413e-05,
1712
+ "loss": 0.3092,
1713
+ "step": 14200
1714
+ },
1715
+ {
1716
+ "epoch": 0.94,
1717
+ "learning_rate": 1.8703329526486868e-05,
1718
+ "loss": 0.3264,
1719
+ "step": 14250
1720
+ },
1721
+ {
1722
+ "epoch": 0.94,
1723
+ "learning_rate": 1.8593516647632438e-05,
1724
+ "loss": 0.3188,
1725
+ "step": 14300
1726
+ },
1727
+ {
1728
+ "epoch": 0.95,
1729
+ "learning_rate": 1.8483703768778e-05,
1730
+ "loss": 0.2941,
1731
+ "step": 14350
1732
+ },
1733
+ {
1734
+ "epoch": 0.95,
1735
+ "learning_rate": 1.837389088992357e-05,
1736
+ "loss": 0.3129,
1737
+ "step": 14400
1738
+ },
1739
+ {
1740
+ "epoch": 0.95,
1741
+ "learning_rate": 1.826407801106914e-05,
1742
+ "loss": 0.3155,
1743
+ "step": 14450
1744
+ },
1745
+ {
1746
+ "epoch": 0.96,
1747
+ "learning_rate": 1.8154265132214705e-05,
1748
+ "loss": 0.3256,
1749
+ "step": 14500
1750
+ },
1751
+ {
1752
+ "epoch": 0.96,
1753
+ "learning_rate": 1.8044452253360275e-05,
1754
+ "loss": 0.3219,
1755
+ "step": 14550
1756
+ },
1757
+ {
1758
+ "epoch": 0.96,
1759
+ "learning_rate": 1.7934639374505842e-05,
1760
+ "loss": 0.3102,
1761
+ "step": 14600
1762
+ },
1763
+ {
1764
+ "epoch": 0.97,
1765
+ "learning_rate": 1.7824826495651412e-05,
1766
+ "loss": 0.3404,
1767
+ "step": 14650
1768
+ },
1769
+ {
1770
+ "epoch": 0.97,
1771
+ "learning_rate": 1.771501361679698e-05,
1772
+ "loss": 0.3211,
1773
+ "step": 14700
1774
+ },
1775
+ {
1776
+ "epoch": 0.97,
1777
+ "learning_rate": 1.7605200737942546e-05,
1778
+ "loss": 0.3272,
1779
+ "step": 14750
1780
+ },
1781
+ {
1782
+ "epoch": 0.98,
1783
+ "learning_rate": 1.7495387859088113e-05,
1784
+ "loss": 0.3323,
1785
+ "step": 14800
1786
+ },
1787
+ {
1788
+ "epoch": 0.98,
1789
+ "learning_rate": 1.7385574980233683e-05,
1790
+ "loss": 0.3059,
1791
+ "step": 14850
1792
+ },
1793
+ {
1794
+ "epoch": 0.98,
1795
+ "learning_rate": 1.727576210137925e-05,
1796
+ "loss": 0.3086,
1797
+ "step": 14900
1798
+ },
1799
+ {
1800
+ "epoch": 0.99,
1801
+ "learning_rate": 1.716594922252482e-05,
1802
+ "loss": 0.3152,
1803
+ "step": 14950
1804
+ },
1805
+ {
1806
+ "epoch": 0.99,
1807
+ "learning_rate": 1.7056136343670387e-05,
1808
+ "loss": 0.3097,
1809
+ "step": 15000
1810
+ },
1811
+ {
1812
+ "epoch": 0.99,
1813
+ "learning_rate": 1.6946323464815954e-05,
1814
+ "loss": 0.3306,
1815
+ "step": 15050
1816
+ },
1817
+ {
1818
+ "epoch": 0.99,
1819
+ "learning_rate": 1.6836510585961524e-05,
1820
+ "loss": 0.3157,
1821
+ "step": 15100
1822
+ },
1823
+ {
1824
+ "epoch": 1.0,
1825
+ "learning_rate": 1.6726697707107088e-05,
1826
+ "loss": 0.3206,
1827
+ "step": 15150
1828
+ },
1829
+ {
1830
+ "epoch": 1.0,
1831
+ "learning_rate": 1.6616884828252658e-05,
1832
+ "loss": 0.2961,
1833
+ "step": 15200
1834
+ },
1835
+ {
1836
+ "epoch": 1.0,
1837
+ "learning_rate": 1.6507071949398225e-05,
1838
+ "loss": 0.274,
1839
+ "step": 15250
1840
+ },
1841
+ {
1842
+ "epoch": 1.01,
1843
+ "learning_rate": 1.6397259070543795e-05,
1844
+ "loss": 0.2576,
1845
+ "step": 15300
1846
+ },
1847
+ {
1848
+ "epoch": 1.01,
1849
+ "learning_rate": 1.628744619168936e-05,
1850
+ "loss": 0.2774,
1851
+ "step": 15350
1852
+ },
1853
+ {
1854
+ "epoch": 1.01,
1855
+ "learning_rate": 1.6177633312834932e-05,
1856
+ "loss": 0.2734,
1857
+ "step": 15400
1858
+ },
1859
+ {
1860
+ "epoch": 1.02,
1861
+ "learning_rate": 1.60678204339805e-05,
1862
+ "loss": 0.2656,
1863
+ "step": 15450
1864
+ },
1865
+ {
1866
+ "epoch": 1.02,
1867
+ "learning_rate": 1.5958007555126066e-05,
1868
+ "loss": 0.2605,
1869
+ "step": 15500
1870
+ },
1871
+ {
1872
+ "epoch": 1.02,
1873
+ "learning_rate": 1.5848194676271632e-05,
1874
+ "loss": 0.2601,
1875
+ "step": 15550
1876
+ },
1877
+ {
1878
+ "epoch": 1.03,
1879
+ "learning_rate": 1.5738381797417203e-05,
1880
+ "loss": 0.2673,
1881
+ "step": 15600
1882
+ },
1883
+ {
1884
+ "epoch": 1.03,
1885
+ "learning_rate": 1.562856891856277e-05,
1886
+ "loss": 0.2604,
1887
+ "step": 15650
1888
+ },
1889
+ {
1890
+ "epoch": 1.03,
1891
+ "learning_rate": 1.5518756039708336e-05,
1892
+ "loss": 0.2779,
1893
+ "step": 15700
1894
+ },
1895
+ {
1896
+ "epoch": 1.04,
1897
+ "learning_rate": 1.5408943160853906e-05,
1898
+ "loss": 0.2599,
1899
+ "step": 15750
1900
+ },
1901
+ {
1902
+ "epoch": 1.04,
1903
+ "learning_rate": 1.5299130281999473e-05,
1904
+ "loss": 0.2726,
1905
+ "step": 15800
1906
+ },
1907
+ {
1908
+ "epoch": 1.04,
1909
+ "learning_rate": 1.5189317403145042e-05,
1910
+ "loss": 0.2679,
1911
+ "step": 15850
1912
+ },
1913
+ {
1914
+ "epoch": 1.05,
1915
+ "learning_rate": 1.507950452429061e-05,
1916
+ "loss": 0.2743,
1917
+ "step": 15900
1918
+ },
1919
+ {
1920
+ "epoch": 1.05,
1921
+ "learning_rate": 1.4969691645436175e-05,
1922
+ "loss": 0.2561,
1923
+ "step": 15950
1924
+ },
1925
+ {
1926
+ "epoch": 1.05,
1927
+ "learning_rate": 1.4859878766581744e-05,
1928
+ "loss": 0.2685,
1929
+ "step": 16000
1930
+ },
1931
+ {
1932
+ "epoch": 1.06,
1933
+ "learning_rate": 1.4750065887727313e-05,
1934
+ "loss": 0.2598,
1935
+ "step": 16050
1936
+ },
1937
+ {
1938
+ "epoch": 1.06,
1939
+ "learning_rate": 1.4640253008872881e-05,
1940
+ "loss": 0.2625,
1941
+ "step": 16100
1942
+ },
1943
+ {
1944
+ "epoch": 1.06,
1945
+ "learning_rate": 1.453044013001845e-05,
1946
+ "loss": 0.2811,
1947
+ "step": 16150
1948
+ },
1949
+ {
1950
+ "epoch": 1.07,
1951
+ "learning_rate": 1.4420627251164018e-05,
1952
+ "loss": 0.2613,
1953
+ "step": 16200
1954
+ },
1955
+ {
1956
+ "epoch": 1.07,
1957
+ "learning_rate": 1.4310814372309587e-05,
1958
+ "loss": 0.26,
1959
+ "step": 16250
1960
+ },
1961
+ {
1962
+ "epoch": 1.07,
1963
+ "learning_rate": 1.4201001493455155e-05,
1964
+ "loss": 0.2716,
1965
+ "step": 16300
1966
+ },
1967
+ {
1968
+ "epoch": 1.08,
1969
+ "learning_rate": 1.409118861460072e-05,
1970
+ "loss": 0.2665,
1971
+ "step": 16350
1972
+ },
1973
+ {
1974
+ "epoch": 1.08,
1975
+ "learning_rate": 1.3981375735746289e-05,
1976
+ "loss": 0.2632,
1977
+ "step": 16400
1978
+ },
1979
+ {
1980
+ "epoch": 1.08,
1981
+ "learning_rate": 1.3871562856891857e-05,
1982
+ "loss": 0.2613,
1983
+ "step": 16450
1984
+ },
1985
+ {
1986
+ "epoch": 1.09,
1987
+ "learning_rate": 1.3761749978037424e-05,
1988
+ "loss": 0.2654,
1989
+ "step": 16500
1990
+ },
1991
+ {
1992
+ "epoch": 1.09,
1993
+ "learning_rate": 1.3651937099182993e-05,
1994
+ "loss": 0.2674,
1995
+ "step": 16550
1996
+ },
1997
+ {
1998
+ "epoch": 1.09,
1999
+ "learning_rate": 1.3542124220328561e-05,
2000
+ "loss": 0.2626,
2001
+ "step": 16600
2002
+ },
2003
+ {
2004
+ "epoch": 1.1,
2005
+ "learning_rate": 1.343231134147413e-05,
2006
+ "loss": 0.2727,
2007
+ "step": 16650
2008
+ },
2009
+ {
2010
+ "epoch": 1.1,
2011
+ "learning_rate": 1.3322498462619698e-05,
2012
+ "loss": 0.2634,
2013
+ "step": 16700
2014
+ },
2015
+ {
2016
+ "epoch": 1.1,
2017
+ "learning_rate": 1.3212685583765263e-05,
2018
+ "loss": 0.2635,
2019
+ "step": 16750
2020
+ },
2021
+ {
2022
+ "epoch": 1.11,
2023
+ "learning_rate": 1.3102872704910832e-05,
2024
+ "loss": 0.2658,
2025
+ "step": 16800
2026
+ },
2027
+ {
2028
+ "epoch": 1.11,
2029
+ "learning_rate": 1.29930598260564e-05,
2030
+ "loss": 0.2598,
2031
+ "step": 16850
2032
+ },
2033
+ {
2034
+ "epoch": 1.11,
2035
+ "learning_rate": 1.2883246947201969e-05,
2036
+ "loss": 0.255,
2037
+ "step": 16900
2038
+ },
2039
+ {
2040
+ "epoch": 1.12,
2041
+ "learning_rate": 1.2773434068347537e-05,
2042
+ "loss": 0.261,
2043
+ "step": 16950
2044
+ },
2045
+ {
2046
+ "epoch": 1.12,
2047
+ "learning_rate": 1.2663621189493104e-05,
2048
+ "loss": 0.2717,
2049
+ "step": 17000
2050
+ },
2051
+ {
2052
+ "epoch": 1.12,
2053
+ "learning_rate": 1.2553808310638673e-05,
2054
+ "loss": 0.2642,
2055
+ "step": 17050
2056
+ },
2057
+ {
2058
+ "epoch": 1.13,
2059
+ "learning_rate": 1.244399543178424e-05,
2060
+ "loss": 0.2796,
2061
+ "step": 17100
2062
+ },
2063
+ {
2064
+ "epoch": 1.13,
2065
+ "learning_rate": 1.2334182552929808e-05,
2066
+ "loss": 0.2584,
2067
+ "step": 17150
2068
+ },
2069
+ {
2070
+ "epoch": 1.13,
2071
+ "learning_rate": 1.2224369674075375e-05,
2072
+ "loss": 0.2646,
2073
+ "step": 17200
2074
+ },
2075
+ {
2076
+ "epoch": 1.14,
2077
+ "learning_rate": 1.2114556795220944e-05,
2078
+ "loss": 0.2682,
2079
+ "step": 17250
2080
+ },
2081
+ {
2082
+ "epoch": 1.14,
2083
+ "learning_rate": 1.2004743916366512e-05,
2084
+ "loss": 0.2564,
2085
+ "step": 17300
2086
+ },
2087
+ {
2088
+ "epoch": 1.14,
2089
+ "learning_rate": 1.189493103751208e-05,
2090
+ "loss": 0.2637,
2091
+ "step": 17350
2092
+ },
2093
+ {
2094
+ "epoch": 1.15,
2095
+ "learning_rate": 1.1785118158657647e-05,
2096
+ "loss": 0.2753,
2097
+ "step": 17400
2098
+ },
2099
+ {
2100
+ "epoch": 1.15,
2101
+ "learning_rate": 1.1675305279803216e-05,
2102
+ "loss": 0.253,
2103
+ "step": 17450
2104
+ },
2105
+ {
2106
+ "epoch": 1.15,
2107
+ "learning_rate": 1.1565492400948784e-05,
2108
+ "loss": 0.2607,
2109
+ "step": 17500
2110
+ },
2111
+ {
2112
+ "epoch": 1.16,
2113
+ "learning_rate": 1.1455679522094353e-05,
2114
+ "loss": 0.2691,
2115
+ "step": 17550
2116
+ },
2117
+ {
2118
+ "epoch": 1.16,
2119
+ "learning_rate": 1.134586664323992e-05,
2120
+ "loss": 0.2659,
2121
+ "step": 17600
2122
+ },
2123
+ {
2124
+ "epoch": 1.16,
2125
+ "learning_rate": 1.1236053764385487e-05,
2126
+ "loss": 0.2757,
2127
+ "step": 17650
2128
+ },
2129
+ {
2130
+ "epoch": 1.17,
2131
+ "learning_rate": 1.1126240885531055e-05,
2132
+ "loss": 0.2654,
2133
+ "step": 17700
2134
+ },
2135
+ {
2136
+ "epoch": 1.17,
2137
+ "learning_rate": 1.1016428006676624e-05,
2138
+ "loss": 0.2575,
2139
+ "step": 17750
2140
+ },
2141
+ {
2142
+ "epoch": 1.17,
2143
+ "learning_rate": 1.090661512782219e-05,
2144
+ "loss": 0.262,
2145
+ "step": 17800
2146
+ },
2147
+ {
2148
+ "epoch": 1.18,
2149
+ "learning_rate": 1.0796802248967759e-05,
2150
+ "loss": 0.2686,
2151
+ "step": 17850
2152
+ },
2153
+ {
2154
+ "epoch": 1.18,
2155
+ "learning_rate": 1.0686989370113328e-05,
2156
+ "loss": 0.2808,
2157
+ "step": 17900
2158
+ },
2159
+ {
2160
+ "epoch": 1.18,
2161
+ "learning_rate": 1.0577176491258896e-05,
2162
+ "loss": 0.286,
2163
+ "step": 17950
2164
+ },
2165
+ {
2166
+ "epoch": 1.19,
2167
+ "learning_rate": 1.0467363612404463e-05,
2168
+ "loss": 0.2578,
2169
+ "step": 18000
2170
+ },
2171
+ {
2172
+ "epoch": 1.19,
2173
+ "learning_rate": 1.0357550733550031e-05,
2174
+ "loss": 0.2699,
2175
+ "step": 18050
2176
+ },
2177
+ {
2178
+ "epoch": 1.19,
2179
+ "learning_rate": 1.02477378546956e-05,
2180
+ "loss": 0.2848,
2181
+ "step": 18100
2182
+ },
2183
+ {
2184
+ "epoch": 1.2,
2185
+ "learning_rate": 1.0137924975841167e-05,
2186
+ "loss": 0.272,
2187
+ "step": 18150
2188
+ },
2189
+ {
2190
+ "epoch": 1.2,
2191
+ "learning_rate": 1.0028112096986735e-05,
2192
+ "loss": 0.2658,
2193
+ "step": 18200
2194
+ },
2195
+ {
2196
+ "epoch": 1.2,
2197
+ "learning_rate": 9.918299218132302e-06,
2198
+ "loss": 0.2617,
2199
+ "step": 18250
2200
+ },
2201
+ {
2202
+ "epoch": 1.21,
2203
+ "learning_rate": 9.80848633927787e-06,
2204
+ "loss": 0.248,
2205
+ "step": 18300
2206
+ },
2207
+ {
2208
+ "epoch": 1.21,
2209
+ "learning_rate": 9.69867346042344e-06,
2210
+ "loss": 0.2762,
2211
+ "step": 18350
2212
+ },
2213
+ {
2214
+ "epoch": 1.21,
2215
+ "learning_rate": 9.588860581569006e-06,
2216
+ "loss": 0.2617,
2217
+ "step": 18400
2218
+ },
2219
+ {
2220
+ "epoch": 1.22,
2221
+ "learning_rate": 9.479047702714575e-06,
2222
+ "loss": 0.2676,
2223
+ "step": 18450
2224
+ },
2225
+ {
2226
+ "epoch": 1.22,
2227
+ "learning_rate": 9.369234823860143e-06,
2228
+ "loss": 0.2782,
2229
+ "step": 18500
2230
+ },
2231
+ {
2232
+ "epoch": 1.22,
2233
+ "learning_rate": 9.259421945005712e-06,
2234
+ "loss": 0.2547,
2235
+ "step": 18550
2236
+ },
2237
+ {
2238
+ "epoch": 1.23,
2239
+ "learning_rate": 9.149609066151278e-06,
2240
+ "loss": 0.2564,
2241
+ "step": 18600
2242
+ },
2243
+ {
2244
+ "epoch": 1.23,
2245
+ "learning_rate": 9.039796187296847e-06,
2246
+ "loss": 0.2579,
2247
+ "step": 18650
2248
+ },
2249
+ {
2250
+ "epoch": 1.23,
2251
+ "learning_rate": 8.929983308442416e-06,
2252
+ "loss": 0.2592,
2253
+ "step": 18700
2254
+ },
2255
+ {
2256
+ "epoch": 1.24,
2257
+ "learning_rate": 8.820170429587982e-06,
2258
+ "loss": 0.2512,
2259
+ "step": 18750
2260
+ },
2261
+ {
2262
+ "epoch": 1.24,
2263
+ "learning_rate": 8.71035755073355e-06,
2264
+ "loss": 0.2632,
2265
+ "step": 18800
2266
+ },
2267
+ {
2268
+ "epoch": 1.24,
2269
+ "learning_rate": 8.600544671879118e-06,
2270
+ "loss": 0.2584,
2271
+ "step": 18850
2272
+ },
2273
+ {
2274
+ "epoch": 1.25,
2275
+ "learning_rate": 8.490731793024686e-06,
2276
+ "loss": 0.2586,
2277
+ "step": 18900
2278
+ },
2279
+ {
2280
+ "epoch": 1.25,
2281
+ "learning_rate": 8.380918914170255e-06,
2282
+ "loss": 0.2536,
2283
+ "step": 18950
2284
+ },
2285
+ {
2286
+ "epoch": 1.25,
2287
+ "learning_rate": 8.271106035315822e-06,
2288
+ "loss": 0.282,
2289
+ "step": 19000
2290
+ },
2291
+ {
2292
+ "epoch": 1.26,
2293
+ "learning_rate": 8.16129315646139e-06,
2294
+ "loss": 0.2616,
2295
+ "step": 19050
2296
+ },
2297
+ {
2298
+ "epoch": 1.26,
2299
+ "learning_rate": 8.051480277606959e-06,
2300
+ "loss": 0.2584,
2301
+ "step": 19100
2302
+ },
2303
+ {
2304
+ "epoch": 1.26,
2305
+ "learning_rate": 7.941667398752527e-06,
2306
+ "loss": 0.2528,
2307
+ "step": 19150
2308
+ },
2309
+ {
2310
+ "epoch": 1.27,
2311
+ "learning_rate": 7.831854519898094e-06,
2312
+ "loss": 0.267,
2313
+ "step": 19200
2314
+ },
2315
+ {
2316
+ "epoch": 1.27,
2317
+ "learning_rate": 7.722041641043663e-06,
2318
+ "loss": 0.2669,
2319
+ "step": 19250
2320
+ },
2321
+ {
2322
+ "epoch": 1.27,
2323
+ "learning_rate": 7.61222876218923e-06,
2324
+ "loss": 0.2728,
2325
+ "step": 19300
2326
+ },
2327
+ {
2328
+ "epoch": 1.27,
2329
+ "learning_rate": 7.502415883334799e-06,
2330
+ "loss": 0.2457,
2331
+ "step": 19350
2332
+ },
2333
+ {
2334
+ "epoch": 1.28,
2335
+ "learning_rate": 7.3926030044803656e-06,
2336
+ "loss": 0.2678,
2337
+ "step": 19400
2338
+ },
2339
+ {
2340
+ "epoch": 1.28,
2341
+ "learning_rate": 7.282790125625934e-06,
2342
+ "loss": 0.2707,
2343
+ "step": 19450
2344
+ },
2345
+ {
2346
+ "epoch": 1.28,
2347
+ "learning_rate": 7.172977246771502e-06,
2348
+ "loss": 0.2645,
2349
+ "step": 19500
2350
+ },
2351
+ {
2352
+ "epoch": 1.29,
2353
+ "learning_rate": 7.06316436791707e-06,
2354
+ "loss": 0.2557,
2355
+ "step": 19550
2356
+ },
2357
+ {
2358
+ "epoch": 1.29,
2359
+ "learning_rate": 6.953351489062637e-06,
2360
+ "loss": 0.2571,
2361
+ "step": 19600
2362
+ },
2363
+ {
2364
+ "epoch": 1.29,
2365
+ "learning_rate": 6.843538610208206e-06,
2366
+ "loss": 0.2709,
2367
+ "step": 19650
2368
+ },
2369
+ {
2370
+ "epoch": 1.3,
2371
+ "learning_rate": 6.733725731353774e-06,
2372
+ "loss": 0.2571,
2373
+ "step": 19700
2374
+ },
2375
+ {
2376
+ "epoch": 1.3,
2377
+ "learning_rate": 6.623912852499342e-06,
2378
+ "loss": 0.2688,
2379
+ "step": 19750
2380
+ },
2381
+ {
2382
+ "epoch": 1.3,
2383
+ "learning_rate": 6.514099973644909e-06,
2384
+ "loss": 0.2516,
2385
+ "step": 19800
2386
+ },
2387
+ {
2388
+ "epoch": 1.31,
2389
+ "learning_rate": 6.404287094790477e-06,
2390
+ "loss": 0.2723,
2391
+ "step": 19850
2392
+ },
2393
+ {
2394
+ "epoch": 1.31,
2395
+ "learning_rate": 6.294474215936046e-06,
2396
+ "loss": 0.2799,
2397
+ "step": 19900
2398
+ },
2399
+ {
2400
+ "epoch": 1.31,
2401
+ "learning_rate": 6.184661337081613e-06,
2402
+ "loss": 0.2623,
2403
+ "step": 19950
2404
+ },
2405
+ {
2406
+ "epoch": 1.32,
2407
+ "learning_rate": 6.074848458227181e-06,
2408
+ "loss": 0.2735,
2409
+ "step": 20000
2410
+ },
2411
+ {
2412
+ "epoch": 1.32,
2413
+ "learning_rate": 5.965035579372749e-06,
2414
+ "loss": 0.279,
2415
+ "step": 20050
2416
+ },
2417
+ {
2418
+ "epoch": 1.32,
2419
+ "learning_rate": 5.8552227005183164e-06,
2420
+ "loss": 0.2725,
2421
+ "step": 20100
2422
+ },
2423
+ {
2424
+ "epoch": 1.33,
2425
+ "learning_rate": 5.745409821663885e-06,
2426
+ "loss": 0.2624,
2427
+ "step": 20150
2428
+ },
2429
+ {
2430
+ "epoch": 1.33,
2431
+ "learning_rate": 5.635596942809453e-06,
2432
+ "loss": 0.275,
2433
+ "step": 20200
2434
+ },
2435
+ {
2436
+ "epoch": 1.33,
2437
+ "learning_rate": 5.525784063955021e-06,
2438
+ "loss": 0.2655,
2439
+ "step": 20250
2440
+ },
2441
+ {
2442
+ "epoch": 1.34,
2443
+ "learning_rate": 5.415971185100589e-06,
2444
+ "loss": 0.2715,
2445
+ "step": 20300
2446
+ },
2447
+ {
2448
+ "epoch": 1.34,
2449
+ "learning_rate": 5.3061583062461565e-06,
2450
+ "loss": 0.2715,
2451
+ "step": 20350
2452
+ },
2453
+ {
2454
+ "epoch": 1.34,
2455
+ "learning_rate": 5.196345427391724e-06,
2456
+ "loss": 0.2634,
2457
+ "step": 20400
2458
+ },
2459
+ {
2460
+ "epoch": 1.35,
2461
+ "learning_rate": 5.086532548537293e-06,
2462
+ "loss": 0.2636,
2463
+ "step": 20450
2464
+ },
2465
+ {
2466
+ "epoch": 1.35,
2467
+ "learning_rate": 4.97671966968286e-06,
2468
+ "loss": 0.2615,
2469
+ "step": 20500
2470
+ },
2471
+ {
2472
+ "epoch": 1.35,
2473
+ "learning_rate": 4.866906790828429e-06,
2474
+ "loss": 0.2644,
2475
+ "step": 20550
2476
+ },
2477
+ {
2478
+ "epoch": 1.36,
2479
+ "learning_rate": 4.757093911973997e-06,
2480
+ "loss": 0.2813,
2481
+ "step": 20600
2482
+ },
2483
+ {
2484
+ "epoch": 1.36,
2485
+ "learning_rate": 4.647281033119564e-06,
2486
+ "loss": 0.2662,
2487
+ "step": 20650
2488
+ },
2489
+ {
2490
+ "epoch": 1.36,
2491
+ "learning_rate": 4.537468154265132e-06,
2492
+ "loss": 0.2644,
2493
+ "step": 20700
2494
+ },
2495
+ {
2496
+ "epoch": 1.37,
2497
+ "learning_rate": 4.4276552754107005e-06,
2498
+ "loss": 0.2724,
2499
+ "step": 20750
2500
+ },
2501
+ {
2502
+ "epoch": 1.37,
2503
+ "learning_rate": 4.317842396556268e-06,
2504
+ "loss": 0.2706,
2505
+ "step": 20800
2506
+ },
2507
+ {
2508
+ "epoch": 1.37,
2509
+ "learning_rate": 4.208029517701837e-06,
2510
+ "loss": 0.2662,
2511
+ "step": 20850
2512
+ },
2513
+ {
2514
+ "epoch": 1.38,
2515
+ "learning_rate": 4.098216638847404e-06,
2516
+ "loss": 0.261,
2517
+ "step": 20900
2518
+ },
2519
+ {
2520
+ "epoch": 1.38,
2521
+ "learning_rate": 3.988403759992972e-06,
2522
+ "loss": 0.247,
2523
+ "step": 20950
2524
+ },
2525
+ {
2526
+ "epoch": 1.38,
2527
+ "learning_rate": 3.87859088113854e-06,
2528
+ "loss": 0.272,
2529
+ "step": 21000
2530
+ },
2531
+ {
2532
+ "epoch": 1.39,
2533
+ "learning_rate": 3.7687780022841083e-06,
2534
+ "loss": 0.2712,
2535
+ "step": 21050
2536
+ },
2537
+ {
2538
+ "epoch": 1.39,
2539
+ "learning_rate": 3.658965123429676e-06,
2540
+ "loss": 0.2762,
2541
+ "step": 21100
2542
+ },
2543
+ {
2544
+ "epoch": 1.39,
2545
+ "learning_rate": 3.549152244575244e-06,
2546
+ "loss": 0.2703,
2547
+ "step": 21150
2548
+ },
2549
+ {
2550
+ "epoch": 1.4,
2551
+ "learning_rate": 3.4393393657208117e-06,
2552
+ "loss": 0.2565,
2553
+ "step": 21200
2554
+ },
2555
+ {
2556
+ "epoch": 1.4,
2557
+ "learning_rate": 3.3295264868663802e-06,
2558
+ "loss": 0.2615,
2559
+ "step": 21250
2560
+ },
2561
+ {
2562
+ "epoch": 1.4,
2563
+ "learning_rate": 3.2197136080119475e-06,
2564
+ "loss": 0.2644,
2565
+ "step": 21300
2566
+ },
2567
+ {
2568
+ "epoch": 1.41,
2569
+ "learning_rate": 3.1099007291575156e-06,
2570
+ "loss": 0.267,
2571
+ "step": 21350
2572
+ },
2573
+ {
2574
+ "epoch": 1.41,
2575
+ "learning_rate": 3.0000878503030837e-06,
2576
+ "loss": 0.264,
2577
+ "step": 21400
2578
+ },
2579
+ {
2580
+ "epoch": 1.41,
2581
+ "learning_rate": 2.8902749714486514e-06,
2582
+ "loss": 0.2754,
2583
+ "step": 21450
2584
+ },
2585
+ {
2586
+ "epoch": 1.42,
2587
+ "learning_rate": 2.7804620925942195e-06,
2588
+ "loss": 0.2544,
2589
+ "step": 21500
2590
+ },
2591
+ {
2592
+ "epoch": 1.42,
2593
+ "learning_rate": 2.6706492137397876e-06,
2594
+ "loss": 0.262,
2595
+ "step": 21550
2596
+ },
2597
+ {
2598
+ "epoch": 1.42,
2599
+ "learning_rate": 2.5608363348853552e-06,
2600
+ "loss": 0.2513,
2601
+ "step": 21600
2602
+ },
2603
+ {
2604
+ "epoch": 1.43,
2605
+ "learning_rate": 2.4510234560309233e-06,
2606
+ "loss": 0.26,
2607
+ "step": 21650
2608
+ },
2609
+ {
2610
+ "epoch": 1.43,
2611
+ "learning_rate": 2.3412105771764915e-06,
2612
+ "loss": 0.2577,
2613
+ "step": 21700
2614
+ },
2615
+ {
2616
+ "epoch": 1.43,
2617
+ "learning_rate": 2.231397698322059e-06,
2618
+ "loss": 0.2552,
2619
+ "step": 21750
2620
+ },
2621
+ {
2622
+ "epoch": 1.44,
2623
+ "learning_rate": 2.1215848194676272e-06,
2624
+ "loss": 0.268,
2625
+ "step": 21800
2626
+ },
2627
+ {
2628
+ "epoch": 1.44,
2629
+ "learning_rate": 2.011771940613195e-06,
2630
+ "loss": 0.267,
2631
+ "step": 21850
2632
+ },
2633
+ {
2634
+ "epoch": 1.44,
2635
+ "learning_rate": 1.901959061758763e-06,
2636
+ "loss": 0.2688,
2637
+ "step": 21900
2638
+ },
2639
+ {
2640
+ "epoch": 1.45,
2641
+ "learning_rate": 1.7921461829043311e-06,
2642
+ "loss": 0.2548,
2643
+ "step": 21950
2644
+ },
2645
+ {
2646
+ "epoch": 1.45,
2647
+ "learning_rate": 1.682333304049899e-06,
2648
+ "loss": 0.2638,
2649
+ "step": 22000
2650
+ },
2651
+ {
2652
+ "epoch": 1.45,
2653
+ "learning_rate": 1.5725204251954669e-06,
2654
+ "loss": 0.2742,
2655
+ "step": 22050
2656
+ },
2657
+ {
2658
+ "epoch": 1.46,
2659
+ "learning_rate": 1.4627075463410348e-06,
2660
+ "loss": 0.2684,
2661
+ "step": 22100
2662
+ },
2663
+ {
2664
+ "epoch": 1.46,
2665
+ "learning_rate": 1.3528946674866029e-06,
2666
+ "loss": 0.26,
2667
+ "step": 22150
2668
+ },
2669
+ {
2670
+ "epoch": 1.46,
2671
+ "learning_rate": 1.2430817886321708e-06,
2672
+ "loss": 0.2573,
2673
+ "step": 22200
2674
+ },
2675
+ {
2676
+ "epoch": 1.47,
2677
+ "learning_rate": 1.1332689097777387e-06,
2678
+ "loss": 0.2736,
2679
+ "step": 22250
2680
+ },
2681
+ {
2682
+ "epoch": 1.47,
2683
+ "learning_rate": 1.0234560309233068e-06,
2684
+ "loss": 0.2732,
2685
+ "step": 22300
2686
+ },
2687
+ {
2688
+ "epoch": 1.47,
2689
+ "learning_rate": 9.136431520688747e-07,
2690
+ "loss": 0.2625,
2691
+ "step": 22350
2692
+ },
2693
+ {
2694
+ "epoch": 1.48,
2695
+ "learning_rate": 8.038302732144426e-07,
2696
+ "loss": 0.2649,
2697
+ "step": 22400
2698
+ },
2699
+ {
2700
+ "epoch": 1.48,
2701
+ "learning_rate": 6.940173943600105e-07,
2702
+ "loss": 0.2646,
2703
+ "step": 22450
2704
+ },
2705
+ {
2706
+ "epoch": 1.48,
2707
+ "learning_rate": 5.842045155055785e-07,
2708
+ "loss": 0.2688,
2709
+ "step": 22500
2710
+ }
2711
+ ],
2712
+ "logging_steps": 50,
2713
+ "max_steps": 22766,
2714
+ "num_train_epochs": 2,
2715
+ "save_steps": 500,
2716
+ "total_flos": 0.0,
2717
+ "trial_name": null,
2718
+ "trial_params": null
2719
+ }