Hugging Face's logo

wangpichao
/
M2M1.2b_NL_subtitles

Text2Text Generation
Transformers
Safetensors
m2m_100
Inference Endpoints
Model card Files Community
wangpichao commited on
Commit
6a84e87
•
1 Parent(s): 0ba4c27

Upload 13 files

Browse files
Files changed (13) hide show
  1. added_tokens.json +102 -0
  2. config.json +37 -0
  3. generation_config.json +11 -0
  4. latest +1 -0
  5. model.safetensors +3 -0
  6. pytorch_model.bin +3 -0
  7. sentencepiece.bpe.model +3 -0
  8. special_tokens_map.json +109 -0
  9. tokenizer_config.json +953 -0
  10. trainer_state.json +4621 -0
  11. training_args.bin +3 -0
  12. vocab.json +0 -0
  13. zero_to_fp32.py +587 -0
added_tokens.json ADDED
@@ -0,0 +1,102 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__af__": 128004,
3
+ "__am__": 128005,
4
+ "__ar__": 128006,
5
+ "__ast__": 128007,
6
+ "__az__": 128008,
7
+ "__ba__": 128009,
8
+ "__be__": 128010,
9
+ "__bg__": 128011,
10
+ "__bn__": 128012,
11
+ "__br__": 128013,
12
+ "__bs__": 128014,
13
+ "__ca__": 128015,
14
+ "__ceb__": 128016,
15
+ "__cs__": 128017,
16
+ "__cy__": 128018,
17
+ "__da__": 128019,
18
+ "__de__": 128020,
19
+ "__el__": 128021,
20
+ "__en__": 128022,
21
+ "__es__": 128023,
22
+ "__et__": 128024,
23
+ "__fa__": 128025,
24
+ "__ff__": 128026,
25
+ "__fi__": 128027,
26
+ "__fr__": 128028,
27
+ "__fy__": 128029,
28
+ "__ga__": 128030,
29
+ "__gd__": 128031,
30
+ "__gl__": 128032,
31
+ "__gu__": 128033,
32
+ "__ha__": 128034,
33
+ "__he__": 128035,
34
+ "__hi__": 128036,
35
+ "__hr__": 128037,
36
+ "__ht__": 128038,
37
+ "__hu__": 128039,
38
+ "__hy__": 128040,
39
+ "__id__": 128041,
40
+ "__ig__": 128042,
41
+ "__ilo__": 128043,
42
+ "__is__": 128044,
43
+ "__it__": 128045,
44
+ "__ja__": 128046,
45
+ "__jv__": 128047,
46
+ "__ka__": 128048,
47
+ "__kk__": 128049,
48
+ "__km__": 128050,
49
+ "__kn__": 128051,
50
+ "__ko__": 128052,
51
+ "__lb__": 128053,
52
+ "__lg__": 128054,
53
+ "__ln__": 128055,
54
+ "__lo__": 128056,
55
+ "__lt__": 128057,
56
+ "__lv__": 128058,
57
+ "__mg__": 128059,
58
+ "__mk__": 128060,
59
+ "__ml__": 128061,
60
+ "__mn__": 128062,
61
+ "__mr__": 128063,
62
+ "__ms__": 128064,
63
+ "__my__": 128065,
64
+ "__ne__": 128066,
65
+ "__nl__": 128067,
66
+ "__no__": 128068,
67
+ "__ns__": 128069,
68
+ "__oc__": 128070,
69
+ "__or__": 128071,
70
+ "__pa__": 128072,
71
+ "__pl__": 128073,
72
+ "__ps__": 128074,
73
+ "__pt__": 128075,
74
+ "__ro__": 128076,
75
+ "__ru__": 128077,
76
+ "__sd__": 128078,
77
+ "__si__": 128079,
78
+ "__sk__": 128080,
79
+ "__sl__": 128081,
80
+ "__so__": 128082,
81
+ "__sq__": 128083,
82
+ "__sr__": 128084,
83
+ "__ss__": 128085,
84
+ "__su__": 128086,
85
+ "__sv__": 128087,
86
+ "__sw__": 128088,
87
+ "__ta__": 128089,
88
+ "__th__": 128090,
89
+ "__tl__": 128091,
90
+ "__tn__": 128092,
91
+ "__tr__": 128093,
92
+ "__uk__": 128094,
93
+ "__ur__": 128095,
94
+ "__uz__": 128096,
95
+ "__vi__": 128097,
96
+ "__wo__": 128098,
97
+ "__xh__": 128099,
98
+ "__yi__": 128100,
99
+ "__yo__": 128101,
100
+ "__zh__": 128102,
101
+ "__zu__": 128103
102
+ }
config.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/m2m100_1.2B",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "relu",
5
+ "architectures": [
6
+ "M2M100ForConditionalGeneration"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 0,
10
+ "d_model": 1024,
11
+ "decoder_attention_heads": 16,
12
+ "decoder_ffn_dim": 8192,
13
+ "decoder_layerdrop": 0.05,
14
+ "decoder_layers": 24,
15
+ "decoder_start_token_id": 2,
16
+ "dropout": 0.1,
17
+ "early_stopping": true,
18
+ "encoder_attention_heads": 16,
19
+ "encoder_ffn_dim": 8192,
20
+ "encoder_layerdrop": 0.05,
21
+ "encoder_layers": 24,
22
+ "eos_token_id": 2,
23
+ "gradient_checkpointing": false,
24
+ "init_std": 0.02,
25
+ "is_encoder_decoder": true,
26
+ "max_length": 200,
27
+ "max_position_embeddings": 1024,
28
+ "model_type": "m2m_100",
29
+ "num_beams": 5,
30
+ "num_hidden_layers": 24,
31
+ "pad_token_id": 1,
32
+ "scale_embedding": true,
33
+ "torch_dtype": "float16",
34
+ "transformers_version": "4.35.2",
35
+ "use_cache": true,
36
+ "vocab_size": 128112
37
+ }
generation_config.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "decoder_start_token_id": 2,
5
+ "early_stopping": true,
6
+ "eos_token_id": 2,
7
+ "max_length": 200,
8
+ "num_beams": 5,
9
+ "pad_token_id": 1,
10
+ "transformers_version": "4.35.2"
11
+ }
latest ADDED
@@ -0,0 +1 @@
 
 
1
+ global_step383970
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5c068e5625a58d0d75cdf4d0be3e8cb00125bcf81e25faca31bbac27f8450a5
3
+ size 3266181192
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d364df8ca2cbaeebf1bffeeb68e8bb8b54a4e1eb333699596bee60e9881d88ea
3
+ size 3266393146
sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d8f7c76ed2a5e0822be39f0a4f95a55eb19c78f4593ce609e2edbc2aea4d380a
3
+ size 2423393
special_tokens_map.json ADDED
@@ -0,0 +1,109 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "__af__",
4
+ "__am__",
5
+ "__ar__",
6
+ "__ast__",
7
+ "__az__",
8
+ "__ba__",
9
+ "__be__",
10
+ "__bg__",
11
+ "__bn__",
12
+ "__br__",
13
+ "__bs__",
14
+ "__ca__",
15
+ "__ceb__",
16
+ "__cs__",
17
+ "__cy__",
18
+ "__da__",
19
+ "__de__",
20
+ "__el__",
21
+ "__en__",
22
+ "__es__",
23
+ "__et__",
24
+ "__fa__",
25
+ "__ff__",
26
+ "__fi__",
27
+ "__fr__",
28
+ "__fy__",
29
+ "__ga__",
30
+ "__gd__",
31
+ "__gl__",
32
+ "__gu__",
33
+ "__ha__",
34
+ "__he__",
35
+ "__hi__",
36
+ "__hr__",
37
+ "__ht__",
38
+ "__hu__",
39
+ "__hy__",
40
+ "__id__",
41
+ "__ig__",
42
+ "__ilo__",
43
+ "__is__",
44
+ "__it__",
45
+ "__ja__",
46
+ "__jv__",
47
+ "__ka__",
48
+ "__kk__",
49
+ "__km__",
50
+ "__kn__",
51
+ "__ko__",
52
+ "__lb__",
53
+ "__lg__",
54
+ "__ln__",
55
+ "__lo__",
56
+ "__lt__",
57
+ "__lv__",
58
+ "__mg__",
59
+ "__mk__",
60
+ "__ml__",
61
+ "__mn__",
62
+ "__mr__",
63
+ "__ms__",
64
+ "__my__",
65
+ "__ne__",
66
+ "__nl__",
67
+ "__no__",
68
+ "__ns__",
69
+ "__oc__",
70
+ "__or__",
71
+ "__pa__",
72
+ "__pl__",
73
+ "__ps__",
74
+ "__pt__",
75
+ "__ro__",
76
+ "__ru__",
77
+ "__sd__",
78
+ "__si__",
79
+ "__sk__",
80
+ "__sl__",
81
+ "__so__",
82
+ "__sq__",
83
+ "__sr__",
84
+ "__ss__",
85
+ "__su__",
86
+ "__sv__",
87
+ "__sw__",
88
+ "__ta__",
89
+ "__th__",
90
+ "__tl__",
91
+ "__tn__",
92
+ "__tr__",
93
+ "__uk__",
94
+ "__ur__",
95
+ "__uz__",
96
+ "__vi__",
97
+ "__wo__",
98
+ "__xh__",
99
+ "__yi__",
100
+ "__yo__",
101
+ "__zh__",
102
+ "__zu__"
103
+ ],
104
+ "bos_token": "<s>",
105
+ "eos_token": "</s>",
106
+ "pad_token": "<pad>",
107
+ "sep_token": "</s>",
108
+ "unk_token": "<unk>"
109
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,953 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "128004": {
36
+ "content": "__af__",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ },
43
+ "128005": {
44
+ "content": "__am__",
45
+ "lstrip": false,
46
+ "normalized": false,
47
+ "rstrip": false,
48
+ "single_word": false,
49
+ "special": true
50
+ },
51
+ "128006": {
52
+ "content": "__ar__",
53
+ "lstrip": false,
54
+ "normalized": false,
55
+ "rstrip": false,
56
+ "single_word": false,
57
+ "special": true
58
+ },
59
+ "128007": {
60
+ "content": "__ast__",
61
+ "lstrip": false,
62
+ "normalized": false,
63
+ "rstrip": false,
64
+ "single_word": false,
65
+ "special": true
66
+ },
67
+ "128008": {
68
+ "content": "__az__",
69
+ "lstrip": false,
70
+ "normalized": false,
71
+ "rstrip": false,
72
+ "single_word": false,
73
+ "special": true
74
+ },
75
+ "128009": {
76
+ "content": "__ba__",
77
+ "lstrip": false,
78
+ "normalized": false,
79
+ "rstrip": false,
80
+ "single_word": false,
81
+ "special": true
82
+ },
83
+ "128010": {
84
+ "content": "__be__",
85
+ "lstrip": false,
86
+ "normalized": false,
87
+ "rstrip": false,
88
+ "single_word": false,
89
+ "special": true
90
+ },
91
+ "128011": {
92
+ "content": "__bg__",
93
+ "lstrip": false,
94
+ "normalized": false,
95
+ "rstrip": false,
96
+ "single_word": false,
97
+ "special": true
98
+ },
99
+ "128012": {
100
+ "content": "__bn__",
101
+ "lstrip": false,
102
+ "normalized": false,
103
+ "rstrip": false,
104
+ "single_word": false,
105
+ "special": true
106
+ },
107
+ "128013": {
108
+ "content": "__br__",
109
+ "lstrip": false,
110
+ "normalized": false,
111
+ "rstrip": false,
112
+ "single_word": false,
113
+ "special": true
114
+ },
115
+ "128014": {
116
+ "content": "__bs__",
117
+ "lstrip": false,
118
+ "normalized": false,
119
+ "rstrip": false,
120
+ "single_word": false,
121
+ "special": true
122
+ },
123
+ "128015": {
124
+ "content": "__ca__",
125
+ "lstrip": false,
126
+ "normalized": false,
127
+ "rstrip": false,
128
+ "single_word": false,
129
+ "special": true
130
+ },
131
+ "128016": {
132
+ "content": "__ceb__",
133
+ "lstrip": false,
134
+ "normalized": false,
135
+ "rstrip": false,
136
+ "single_word": false,
137
+ "special": true
138
+ },
139
+ "128017": {
140
+ "content": "__cs__",
141
+ "lstrip": false,
142
+ "normalized": false,
143
+ "rstrip": false,
144
+ "single_word": false,
145
+ "special": true
146
+ },
147
+ "128018": {
148
+ "content": "__cy__",
149
+ "lstrip": false,
150
+ "normalized": false,
151
+ "rstrip": false,
152
+ "single_word": false,
153
+ "special": true
154
+ },
155
+ "128019": {
156
+ "content": "__da__",
157
+ "lstrip": false,
158
+ "normalized": false,
159
+ "rstrip": false,
160
+ "single_word": false,
161
+ "special": true
162
+ },
163
+ "128020": {
164
+ "content": "__de__",
165
+ "lstrip": false,
166
+ "normalized": false,
167
+ "rstrip": false,
168
+ "single_word": false,
169
+ "special": true
170
+ },
171
+ "128021": {
172
+ "content": "__el__",
173
+ "lstrip": false,
174
+ "normalized": false,
175
+ "rstrip": false,
176
+ "single_word": false,
177
+ "special": true
178
+ },
179
+ "128022": {
180
+ "content": "__en__",
181
+ "lstrip": false,
182
+ "normalized": false,
183
+ "rstrip": false,
184
+ "single_word": false,
185
+ "special": true
186
+ },
187
+ "128023": {
188
+ "content": "__es__",
189
+ "lstrip": false,
190
+ "normalized": false,
191
+ "rstrip": false,
192
+ "single_word": false,
193
+ "special": true
194
+ },
195
+ "128024": {
196
+ "content": "__et__",
197
+ "lstrip": false,
198
+ "normalized": false,
199
+ "rstrip": false,
200
+ "single_word": false,
201
+ "special": true
202
+ },
203
+ "128025": {
204
+ "content": "__fa__",
205
+ "lstrip": false,
206
+ "normalized": false,
207
+ "rstrip": false,
208
+ "single_word": false,
209
+ "special": true
210
+ },
211
+ "128026": {
212
+ "content": "__ff__",
213
+ "lstrip": false,
214
+ "normalized": false,
215
+ "rstrip": false,
216
+ "single_word": false,
217
+ "special": true
218
+ },
219
+ "128027": {
220
+ "content": "__fi__",
221
+ "lstrip": false,
222
+ "normalized": false,
223
+ "rstrip": false,
224
+ "single_word": false,
225
+ "special": true
226
+ },
227
+ "128028": {
228
+ "content": "__fr__",
229
+ "lstrip": false,
230
+ "normalized": false,
231
+ "rstrip": false,
232
+ "single_word": false,
233
+ "special": true
234
+ },
235
+ "128029": {
236
+ "content": "__fy__",
237
+ "lstrip": false,
238
+ "normalized": false,
239
+ "rstrip": false,
240
+ "single_word": false,
241
+ "special": true
242
+ },
243
+ "128030": {
244
+ "content": "__ga__",
245
+ "lstrip": false,
246
+ "normalized": false,
247
+ "rstrip": false,
248
+ "single_word": false,
249
+ "special": true
250
+ },
251
+ "128031": {
252
+ "content": "__gd__",
253
+ "lstrip": false,
254
+ "normalized": false,
255
+ "rstrip": false,
256
+ "single_word": false,
257
+ "special": true
258
+ },
259
+ "128032": {
260
+ "content": "__gl__",
261
+ "lstrip": false,
262
+ "normalized": false,
263
+ "rstrip": false,
264
+ "single_word": false,
265
+ "special": true
266
+ },
267
+ "128033": {
268
+ "content": "__gu__",
269
+ "lstrip": false,
270
+ "normalized": false,
271
+ "rstrip": false,
272
+ "single_word": false,
273
+ "special": true
274
+ },
275
+ "128034": {
276
+ "content": "__ha__",
277
+ "lstrip": false,
278
+ "normalized": false,
279
+ "rstrip": false,
280
+ "single_word": false,
281
+ "special": true
282
+ },
283
+ "128035": {
284
+ "content": "__he__",
285
+ "lstrip": false,
286
+ "normalized": false,
287
+ "rstrip": false,
288
+ "single_word": false,
289
+ "special": true
290
+ },
291
+ "128036": {
292
+ "content": "__hi__",
293
+ "lstrip": false,
294
+ "normalized": false,
295
+ "rstrip": false,
296
+ "single_word": false,
297
+ "special": true
298
+ },
299
+ "128037": {
300
+ "content": "__hr__",
301
+ "lstrip": false,
302
+ "normalized": false,
303
+ "rstrip": false,
304
+ "single_word": false,
305
+ "special": true
306
+ },
307
+ "128038": {
308
+ "content": "__ht__",
309
+ "lstrip": false,
310
+ "normalized": false,
311
+ "rstrip": false,
312
+ "single_word": false,
313
+ "special": true
314
+ },
315
+ "128039": {
316
+ "content": "__hu__",
317
+ "lstrip": false,
318
+ "normalized": false,
319
+ "rstrip": false,
320
+ "single_word": false,
321
+ "special": true
322
+ },
323
+ "128040": {
324
+ "content": "__hy__",
325
+ "lstrip": false,
326
+ "normalized": false,
327
+ "rstrip": false,
328
+ "single_word": false,
329
+ "special": true
330
+ },
331
+ "128041": {
332
+ "content": "__id__",
333
+ "lstrip": false,
334
+ "normalized": false,
335
+ "rstrip": false,
336
+ "single_word": false,
337
+ "special": true
338
+ },
339
+ "128042": {
340
+ "content": "__ig__",
341
+ "lstrip": false,
342
+ "normalized": false,
343
+ "rstrip": false,
344
+ "single_word": false,
345
+ "special": true
346
+ },
347
+ "128043": {
348
+ "content": "__ilo__",
349
+ "lstrip": false,
350
+ "normalized": false,
351
+ "rstrip": false,
352
+ "single_word": false,
353
+ "special": true
354
+ },
355
+ "128044": {
356
+ "content": "__is__",
357
+ "lstrip": false,
358
+ "normalized": false,
359
+ "rstrip": false,
360
+ "single_word": false,
361
+ "special": true
362
+ },
363
+ "128045": {
364
+ "content": "__it__",
365
+ "lstrip": false,
366
+ "normalized": false,
367
+ "rstrip": false,
368
+ "single_word": false,
369
+ "special": true
370
+ },
371
+ "128046": {
372
+ "content": "__ja__",
373
+ "lstrip": false,
374
+ "normalized": false,
375
+ "rstrip": false,
376
+ "single_word": false,
377
+ "special": true
378
+ },
379
+ "128047": {
380
+ "content": "__jv__",
381
+ "lstrip": false,
382
+ "normalized": false,
383
+ "rstrip": false,
384
+ "single_word": false,
385
+ "special": true
386
+ },
387
+ "128048": {
388
+ "content": "__ka__",
389
+ "lstrip": false,
390
+ "normalized": false,
391
+ "rstrip": false,
392
+ "single_word": false,
393
+ "special": true
394
+ },
395
+ "128049": {
396
+ "content": "__kk__",
397
+ "lstrip": false,
398
+ "normalized": false,
399
+ "rstrip": false,
400
+ "single_word": false,
401
+ "special": true
402
+ },
403
+ "128050": {
404
+ "content": "__km__",
405
+ "lstrip": false,
406
+ "normalized": false,
407
+ "rstrip": false,
408
+ "single_word": false,
409
+ "special": true
410
+ },
411
+ "128051": {
412
+ "content": "__kn__",
413
+ "lstrip": false,
414
+ "normalized": false,
415
+ "rstrip": false,
416
+ "single_word": false,
417
+ "special": true
418
+ },
419
+ "128052": {
420
+ "content": "__ko__",
421
+ "lstrip": false,
422
+ "normalized": false,
423
+ "rstrip": false,
424
+ "single_word": false,
425
+ "special": true
426
+ },
427
+ "128053": {
428
+ "content": "__lb__",
429
+ "lstrip": false,
430
+ "normalized": false,
431
+ "rstrip": false,
432
+ "single_word": false,
433
+ "special": true
434
+ },
435
+ "128054": {
436
+ "content": "__lg__",
437
+ "lstrip": false,
438
+ "normalized": false,
439
+ "rstrip": false,
440
+ "single_word": false,
441
+ "special": true
442
+ },
443
+ "128055": {
444
+ "content": "__ln__",
445
+ "lstrip": false,
446
+ "normalized": false,
447
+ "rstrip": false,
448
+ "single_word": false,
449
+ "special": true
450
+ },
451
+ "128056": {
452
+ "content": "__lo__",
453
+ "lstrip": false,
454
+ "normalized": false,
455
+ "rstrip": false,
456
+ "single_word": false,
457
+ "special": true
458
+ },
459
+ "128057": {
460
+ "content": "__lt__",
461
+ "lstrip": false,
462
+ "normalized": false,
463
+ "rstrip": false,
464
+ "single_word": false,
465
+ "special": true
466
+ },
467
+ "128058": {
468
+ "content": "__lv__",
469
+ "lstrip": false,
470
+ "normalized": false,
471
+ "rstrip": false,
472
+ "single_word": false,
473
+ "special": true
474
+ },
475
+ "128059": {
476
+ "content": "__mg__",
477
+ "lstrip": false,
478
+ "normalized": false,
479
+ "rstrip": false,
480
+ "single_word": false,
481
+ "special": true
482
+ },
483
+ "128060": {
484
+ "content": "__mk__",
485
+ "lstrip": false,
486
+ "normalized": false,
487
+ "rstrip": false,
488
+ "single_word": false,
489
+ "special": true
490
+ },
491
+ "128061": {
492
+ "content": "__ml__",
493
+ "lstrip": false,
494
+ "normalized": false,
495
+ "rstrip": false,
496
+ "single_word": false,
497
+ "special": true
498
+ },
499
+ "128062": {
500
+ "content": "__mn__",
501
+ "lstrip": false,
502
+ "normalized": false,
503
+ "rstrip": false,
504
+ "single_word": false,
505
+ "special": true
506
+ },
507
+ "128063": {
508
+ "content": "__mr__",
509
+ "lstrip": false,
510
+ "normalized": false,
511
+ "rstrip": false,
512
+ "single_word": false,
513
+ "special": true
514
+ },
515
+ "128064": {
516
+ "content": "__ms__",
517
+ "lstrip": false,
518
+ "normalized": false,
519
+ "rstrip": false,
520
+ "single_word": false,
521
+ "special": true
522
+ },
523
+ "128065": {
524
+ "content": "__my__",
525
+ "lstrip": false,
526
+ "normalized": false,
527
+ "rstrip": false,
528
+ "single_word": false,
529
+ "special": true
530
+ },
531
+ "128066": {
532
+ "content": "__ne__",
533
+ "lstrip": false,
534
+ "normalized": false,
535
+ "rstrip": false,
536
+ "single_word": false,
537
+ "special": true
538
+ },
539
+ "128067": {
540
+ "content": "__nl__",
541
+ "lstrip": false,
542
+ "normalized": false,
543
+ "rstrip": false,
544
+ "single_word": false,
545
+ "special": true
546
+ },
547
+ "128068": {
548
+ "content": "__no__",
549
+ "lstrip": false,
550
+ "normalized": false,
551
+ "rstrip": false,
552
+ "single_word": false,
553
+ "special": true
554
+ },
555
+ "128069": {
556
+ "content": "__ns__",
557
+ "lstrip": false,
558
+ "normalized": false,
559
+ "rstrip": false,
560
+ "single_word": false,
561
+ "special": true
562
+ },
563
+ "128070": {
564
+ "content": "__oc__",
565
+ "lstrip": false,
566
+ "normalized": false,
567
+ "rstrip": false,
568
+ "single_word": false,
569
+ "special": true
570
+ },
571
+ "128071": {
572
+ "content": "__or__",
573
+ "lstrip": false,
574
+ "normalized": false,
575
+ "rstrip": false,
576
+ "single_word": false,
577
+ "special": true
578
+ },
579
+ "128072": {
580
+ "content": "__pa__",
581
+ "lstrip": false,
582
+ "normalized": false,
583
+ "rstrip": false,
584
+ "single_word": false,
585
+ "special": true
586
+ },
587
+ "128073": {
588
+ "content": "__pl__",
589
+ "lstrip": false,
590
+ "normalized": false,
591
+ "rstrip": false,
592
+ "single_word": false,
593
+ "special": true
594
+ },
595
+ "128074": {
596
+ "content": "__ps__",
597
+ "lstrip": false,
598
+ "normalized": false,
599
+ "rstrip": false,
600
+ "single_word": false,
601
+ "special": true
602
+ },
603
+ "128075": {
604
+ "content": "__pt__",
605
+ "lstrip": false,
606
+ "normalized": false,
607
+ "rstrip": false,
608
+ "single_word": false,
609
+ "special": true
610
+ },
611
+ "128076": {
612
+ "content": "__ro__",
613
+ "lstrip": false,
614
+ "normalized": false,
615
+ "rstrip": false,
616
+ "single_word": false,
617
+ "special": true
618
+ },
619
+ "128077": {
620
+ "content": "__ru__",
621
+ "lstrip": false,
622
+ "normalized": false,
623
+ "rstrip": false,
624
+ "single_word": false,
625
+ "special": true
626
+ },
627
+ "128078": {
628
+ "content": "__sd__",
629
+ "lstrip": false,
630
+ "normalized": false,
631
+ "rstrip": false,
632
+ "single_word": false,
633
+ "special": true
634
+ },
635
+ "128079": {
636
+ "content": "__si__",
637
+ "lstrip": false,
638
+ "normalized": false,
639
+ "rstrip": false,
640
+ "single_word": false,
641
+ "special": true
642
+ },
643
+ "128080": {
644
+ "content": "__sk__",
645
+ "lstrip": false,
646
+ "normalized": false,
647
+ "rstrip": false,
648
+ "single_word": false,
649
+ "special": true
650
+ },
651
+ "128081": {
652
+ "content": "__sl__",
653
+ "lstrip": false,
654
+ "normalized": false,
655
+ "rstrip": false,
656
+ "single_word": false,
657
+ "special": true
658
+ },
659
+ "128082": {
660
+ "content": "__so__",
661
+ "lstrip": false,
662
+ "normalized": false,
663
+ "rstrip": false,
664
+ "single_word": false,
665
+ "special": true
666
+ },
667
+ "128083": {
668
+ "content": "__sq__",
669
+ "lstrip": false,
670
+ "normalized": false,
671
+ "rstrip": false,
672
+ "single_word": false,
673
+ "special": true
674
+ },
675
+ "128084": {
676
+ "content": "__sr__",
677
+ "lstrip": false,
678
+ "normalized": false,
679
+ "rstrip": false,
680
+ "single_word": false,
681
+ "special": true
682
+ },
683
+ "128085": {
684
+ "content": "__ss__",
685
+ "lstrip": false,
686
+ "normalized": false,
687
+ "rstrip": false,
688
+ "single_word": false,
689
+ "special": true
690
+ },
691
+ "128086": {
692
+ "content": "__su__",
693
+ "lstrip": false,
694
+ "normalized": false,
695
+ "rstrip": false,
696
+ "single_word": false,
697
+ "special": true
698
+ },
699
+ "128087": {
700
+ "content": "__sv__",
701
+ "lstrip": false,
702
+ "normalized": false,
703
+ "rstrip": false,
704
+ "single_word": false,
705
+ "special": true
706
+ },
707
+ "128088": {
708
+ "content": "__sw__",
709
+ "lstrip": false,
710
+ "normalized": false,
711
+ "rstrip": false,
712
+ "single_word": false,
713
+ "special": true
714
+ },
715
+ "128089": {
716
+ "content": "__ta__",
717
+ "lstrip": false,
718
+ "normalized": false,
719
+ "rstrip": false,
720
+ "single_word": false,
721
+ "special": true
722
+ },
723
+ "128090": {
724
+ "content": "__th__",
725
+ "lstrip": false,
726
+ "normalized": false,
727
+ "rstrip": false,
728
+ "single_word": false,
729
+ "special": true
730
+ },
731
+ "128091": {
732
+ "content": "__tl__",
733
+ "lstrip": false,
734
+ "normalized": false,
735
+ "rstrip": false,
736
+ "single_word": false,
737
+ "special": true
738
+ },
739
+ "128092": {
740
+ "content": "__tn__",
741
+ "lstrip": false,
742
+ "normalized": false,
743
+ "rstrip": false,
744
+ "single_word": false,
745
+ "special": true
746
+ },
747
+ "128093": {
748
+ "content": "__tr__",
749
+ "lstrip": false,
750
+ "normalized": false,
751
+ "rstrip": false,
752
+ "single_word": false,
753
+ "special": true
754
+ },
755
+ "128094": {
756
+ "content": "__uk__",
757
+ "lstrip": false,
758
+ "normalized": false,
759
+ "rstrip": false,
760
+ "single_word": false,
761
+ "special": true
762
+ },
763
+ "128095": {
764
+ "content": "__ur__",
765
+ "lstrip": false,
766
+ "normalized": false,
767
+ "rstrip": false,
768
+ "single_word": false,
769
+ "special": true
770
+ },
771
+ "128096": {
772
+ "content": "__uz__",
773
+ "lstrip": false,
774
+ "normalized": false,
775
+ "rstrip": false,
776
+ "single_word": false,
777
+ "special": true
778
+ },
779
+ "128097": {
780
+ "content": "__vi__",
781
+ "lstrip": false,
782
+ "normalized": false,
783
+ "rstrip": false,
784
+ "single_word": false,
785
+ "special": true
786
+ },
787
+ "128098": {
788
+ "content": "__wo__",
789
+ "lstrip": false,
790
+ "normalized": false,
791
+ "rstrip": false,
792
+ "single_word": false,
793
+ "special": true
794
+ },
795
+ "128099": {
796
+ "content": "__xh__",
797
+ "lstrip": false,
798
+ "normalized": false,
799
+ "rstrip": false,
800
+ "single_word": false,
801
+ "special": true
802
+ },
803
+ "128100": {
804
+ "content": "__yi__",
805
+ "lstrip": false,
806
+ "normalized": false,
807
+ "rstrip": false,
808
+ "single_word": false,
809
+ "special": true
810
+ },
811
+ "128101": {
812
+ "content": "__yo__",
813
+ "lstrip": false,
814
+ "normalized": false,
815
+ "rstrip": false,
816
+ "single_word": false,
817
+ "special": true
818
+ },
819
+ "128102": {
820
+ "content": "__zh__",
821
+ "lstrip": false,
822
+ "normalized": false,
823
+ "rstrip": false,
824
+ "single_word": false,
825
+ "special": true
826
+ },
827
+ "128103": {
828
+ "content": "__zu__",
829
+ "lstrip": false,
830
+ "normalized": false,
831
+ "rstrip": false,
832
+ "single_word": false,
833
+ "special": true
834
+ }
835
+ },
836
+ "additional_special_tokens": [
837
+ "__af__",
838
+ "__am__",
839
+ "__ar__",
840
+ "__ast__",
841
+ "__az__",
842
+ "__ba__",
843
+ "__be__",
844
+ "__bg__",
845
+ "__bn__",
846
+ "__br__",
847
+ "__bs__",
848
+ "__ca__",
849
+ "__ceb__",
850
+ "__cs__",
851
+ "__cy__",
852
+ "__da__",
853
+ "__de__",
854
+ "__el__",
855
+ "__en__",
856
+ "__es__",
857
+ "__et__",
858
+ "__fa__",
859
+ "__ff__",
860
+ "__fi__",
861
+ "__fr__",
862
+ "__fy__",
863
+ "__ga__",
864
+ "__gd__",
865
+ "__gl__",
866
+ "__gu__",
867
+ "__ha__",
868
+ "__he__",
869
+ "__hi__",
870
+ "__hr__",
871
+ "__ht__",
872
+ "__hu__",
873
+ "__hy__",
874
+ "__id__",
875
+ "__ig__",
876
+ "__ilo__",
877
+ "__is__",
878
+ "__it__",
879
+ "__ja__",
880
+ "__jv__",
881
+ "__ka__",
882
+ "__kk__",
883
+ "__km__",
884
+ "__kn__",
885
+ "__ko__",
886
+ "__lb__",
887
+ "__lg__",
888
+ "__ln__",
889
+ "__lo__",
890
+ "__lt__",
891
+ "__lv__",
892
+ "__mg__",
893
+ "__mk__",
894
+ "__ml__",
895
+ "__mn__",
896
+ "__mr__",
897
+ "__ms__",
898
+ "__my__",
899
+ "__ne__",
900
+ "__nl__",
901
+ "__no__",
902
+ "__ns__",
903
+ "__oc__",
904
+ "__or__",
905
+ "__pa__",
906
+ "__pl__",
907
+ "__ps__",
908
+ "__pt__",
909
+ "__ro__",
910
+ "__ru__",
911
+ "__sd__",
912
+ "__si__",
913
+ "__sk__",
914
+ "__sl__",
915
+ "__so__",
916
+ "__sq__",
917
+ "__sr__",
918
+ "__ss__",
919
+ "__su__",
920
+ "__sv__",
921
+ "__sw__",
922
+ "__ta__",
923
+ "__th__",
924
+ "__tl__",
925
+ "__tn__",
926
+ "__tr__",
927
+ "__uk__",
928
+ "__ur__",
929
+ "__uz__",
930
+ "__vi__",
931
+ "__wo__",
932
+ "__xh__",
933
+ "__yi__",
934
+ "__yo__",
935
+ "__zh__",
936
+ "__zu__"
937
+ ],
938
+ "bos_token": "<s>",
939
+ "clean_up_tokenization_spaces": true,
940
+ "eos_token": "</s>",
941
+ "language_codes": "m2m100",
942
+ "model_max_length": 1000000000000000019884624838656,
943
+ "num_madeup_words": 8,
944
+ "pad_token": "<pad>",
945
+ "sep_token": "</s>",
946
+ "sp_model_kwargs": {},
947
+ "src_lang": "en",
948
+ "tgt_lang": "nl",
949
+ "tokenizer_class": "M2M100Tokenizer",
950
+ "trust_remote_code": false,
951
+ "unk_token": "<unk>",
952
+ "use_fast": true
953
+ }
trainer_state.json ADDED
@@ -0,0 +1,4621 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0,
5
+ "eval_steps": 500,
6
+ "global_step": 383970,
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": 5e-05,
14
+ "loss": 0.6562,
15
+ "step": 500
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 5e-05,
20
+ "loss": 0.5864,
21
+ "step": 1000
22
+ },
23
+ {
24
+ "epoch": 0.0,
25
+ "learning_rate": 5e-05,
26
+ "loss": 0.5613,
27
+ "step": 1500
28
+ },
29
+ {
30
+ "epoch": 0.01,
31
+ "learning_rate": 5e-05,
32
+ "loss": 0.5443,
33
+ "step": 2000
34
+ },
35
+ {
36
+ "epoch": 0.01,
37
+ "learning_rate": 5e-05,
38
+ "loss": 0.5272,
39
+ "step": 2500
40
+ },
41
+ {
42
+ "epoch": 0.01,
43
+ "learning_rate": 5e-05,
44
+ "loss": 0.5199,
45
+ "step": 3000
46
+ },
47
+ {
48
+ "epoch": 0.01,
49
+ "learning_rate": 5e-05,
50
+ "loss": 0.5136,
51
+ "step": 3500
52
+ },
53
+ {
54
+ "epoch": 0.01,
55
+ "learning_rate": 5e-05,
56
+ "loss": 0.502,
57
+ "step": 4000
58
+ },
59
+ {
60
+ "epoch": 0.01,
61
+ "learning_rate": 5e-05,
62
+ "loss": 0.4965,
63
+ "step": 4500
64
+ },
65
+ {
66
+ "epoch": 0.01,
67
+ "learning_rate": 5e-05,
68
+ "loss": 0.4904,
69
+ "step": 5000
70
+ },
71
+ {
72
+ "epoch": 0.01,
73
+ "learning_rate": 5e-05,
74
+ "loss": 0.4871,
75
+ "step": 5500
76
+ },
77
+ {
78
+ "epoch": 0.02,
79
+ "learning_rate": 5e-05,
80
+ "loss": 0.481,
81
+ "step": 6000
82
+ },
83
+ {
84
+ "epoch": 0.02,
85
+ "learning_rate": 5e-05,
86
+ "loss": 0.4765,
87
+ "step": 6500
88
+ },
89
+ {
90
+ "epoch": 0.02,
91
+ "learning_rate": 5e-05,
92
+ "loss": 0.473,
93
+ "step": 7000
94
+ },
95
+ {
96
+ "epoch": 0.02,
97
+ "learning_rate": 5e-05,
98
+ "loss": 0.4725,
99
+ "step": 7500
100
+ },
101
+ {
102
+ "epoch": 0.02,
103
+ "learning_rate": 5e-05,
104
+ "loss": 0.4677,
105
+ "step": 8000
106
+ },
107
+ {
108
+ "epoch": 0.02,
109
+ "learning_rate": 5e-05,
110
+ "loss": 0.463,
111
+ "step": 8500
112
+ },
113
+ {
114
+ "epoch": 0.02,
115
+ "learning_rate": 5e-05,
116
+ "loss": 0.4633,
117
+ "step": 9000
118
+ },
119
+ {
120
+ "epoch": 0.02,
121
+ "learning_rate": 5e-05,
122
+ "loss": 0.4615,
123
+ "step": 9500
124
+ },
125
+ {
126
+ "epoch": 0.03,
127
+ "learning_rate": 5e-05,
128
+ "loss": 0.4586,
129
+ "step": 10000
130
+ },
131
+ {
132
+ "epoch": 0.03,
133
+ "learning_rate": 5e-05,
134
+ "loss": 0.4561,
135
+ "step": 10500
136
+ },
137
+ {
138
+ "epoch": 0.03,
139
+ "learning_rate": 5e-05,
140
+ "loss": 0.453,
141
+ "step": 11000
142
+ },
143
+ {
144
+ "epoch": 0.03,
145
+ "learning_rate": 5e-05,
146
+ "loss": 0.4528,
147
+ "step": 11500
148
+ },
149
+ {
150
+ "epoch": 0.03,
151
+ "learning_rate": 5e-05,
152
+ "loss": 0.4497,
153
+ "step": 12000
154
+ },
155
+ {
156
+ "epoch": 0.03,
157
+ "learning_rate": 5e-05,
158
+ "loss": 0.4478,
159
+ "step": 12500
160
+ },
161
+ {
162
+ "epoch": 0.03,
163
+ "learning_rate": 5e-05,
164
+ "loss": 0.4475,
165
+ "step": 13000
166
+ },
167
+ {
168
+ "epoch": 0.04,
169
+ "learning_rate": 5e-05,
170
+ "loss": 0.4457,
171
+ "step": 13500
172
+ },
173
+ {
174
+ "epoch": 0.04,
175
+ "learning_rate": 5e-05,
176
+ "loss": 0.4432,
177
+ "step": 14000
178
+ },
179
+ {
180
+ "epoch": 0.04,
181
+ "learning_rate": 5e-05,
182
+ "loss": 0.4412,
183
+ "step": 14500
184
+ },
185
+ {
186
+ "epoch": 0.04,
187
+ "learning_rate": 5e-05,
188
+ "loss": 0.4383,
189
+ "step": 15000
190
+ },
191
+ {
192
+ "epoch": 0.04,
193
+ "learning_rate": 5e-05,
194
+ "loss": 0.439,
195
+ "step": 15500
196
+ },
197
+ {
198
+ "epoch": 0.04,
199
+ "learning_rate": 5e-05,
200
+ "loss": 0.4391,
201
+ "step": 16000
202
+ },
203
+ {
204
+ "epoch": 0.04,
205
+ "learning_rate": 5e-05,
206
+ "loss": 0.4359,
207
+ "step": 16500
208
+ },
209
+ {
210
+ "epoch": 0.04,
211
+ "learning_rate": 5e-05,
212
+ "loss": 0.4354,
213
+ "step": 17000
214
+ },
215
+ {
216
+ "epoch": 0.05,
217
+ "learning_rate": 5e-05,
218
+ "loss": 0.4328,
219
+ "step": 17500
220
+ },
221
+ {
222
+ "epoch": 0.05,
223
+ "learning_rate": 5e-05,
224
+ "loss": 0.4316,
225
+ "step": 18000
226
+ },
227
+ {
228
+ "epoch": 0.05,
229
+ "learning_rate": 5e-05,
230
+ "loss": 0.4311,
231
+ "step": 18500
232
+ },
233
+ {
234
+ "epoch": 0.05,
235
+ "learning_rate": 5e-05,
236
+ "loss": 0.4288,
237
+ "step": 19000
238
+ },
239
+ {
240
+ "epoch": 0.05,
241
+ "learning_rate": 5e-05,
242
+ "loss": 0.4259,
243
+ "step": 19500
244
+ },
245
+ {
246
+ "epoch": 0.05,
247
+ "learning_rate": 5e-05,
248
+ "loss": 0.4266,
249
+ "step": 20000
250
+ },
251
+ {
252
+ "epoch": 0.05,
253
+ "learning_rate": 5e-05,
254
+ "loss": 0.426,
255
+ "step": 20500
256
+ },
257
+ {
258
+ "epoch": 0.05,
259
+ "learning_rate": 5e-05,
260
+ "loss": 0.4233,
261
+ "step": 21000
262
+ },
263
+ {
264
+ "epoch": 0.06,
265
+ "learning_rate": 5e-05,
266
+ "loss": 0.4231,
267
+ "step": 21500
268
+ },
269
+ {
270
+ "epoch": 0.06,
271
+ "learning_rate": 5e-05,
272
+ "loss": 0.4206,
273
+ "step": 22000
274
+ },
275
+ {
276
+ "epoch": 0.06,
277
+ "learning_rate": 5e-05,
278
+ "loss": 0.4219,
279
+ "step": 22500
280
+ },
281
+ {
282
+ "epoch": 0.06,
283
+ "learning_rate": 5e-05,
284
+ "loss": 0.4199,
285
+ "step": 23000
286
+ },
287
+ {
288
+ "epoch": 0.06,
289
+ "learning_rate": 5e-05,
290
+ "loss": 0.4191,
291
+ "step": 23500
292
+ },
293
+ {
294
+ "epoch": 0.06,
295
+ "learning_rate": 5e-05,
296
+ "loss": 0.4181,
297
+ "step": 24000
298
+ },
299
+ {
300
+ "epoch": 0.06,
301
+ "learning_rate": 5e-05,
302
+ "loss": 0.4172,
303
+ "step": 24500
304
+ },
305
+ {
306
+ "epoch": 0.07,
307
+ "learning_rate": 5e-05,
308
+ "loss": 0.4166,
309
+ "step": 25000
310
+ },
311
+ {
312
+ "epoch": 0.07,
313
+ "learning_rate": 5e-05,
314
+ "loss": 0.4142,
315
+ "step": 25500
316
+ },
317
+ {
318
+ "epoch": 0.07,
319
+ "learning_rate": 5e-05,
320
+ "loss": 0.4145,
321
+ "step": 26000
322
+ },
323
+ {
324
+ "epoch": 0.07,
325
+ "learning_rate": 5e-05,
326
+ "loss": 0.4148,
327
+ "step": 26500
328
+ },
329
+ {
330
+ "epoch": 0.07,
331
+ "learning_rate": 5e-05,
332
+ "loss": 0.4133,
333
+ "step": 27000
334
+ },
335
+ {
336
+ "epoch": 0.07,
337
+ "learning_rate": 5e-05,
338
+ "loss": 0.4128,
339
+ "step": 27500
340
+ },
341
+ {
342
+ "epoch": 0.07,
343
+ "learning_rate": 5e-05,
344
+ "loss": 0.412,
345
+ "step": 28000
346
+ },
347
+ {
348
+ "epoch": 0.07,
349
+ "learning_rate": 5e-05,
350
+ "loss": 0.4108,
351
+ "step": 28500
352
+ },
353
+ {
354
+ "epoch": 0.08,
355
+ "learning_rate": 5e-05,
356
+ "loss": 0.4101,
357
+ "step": 29000
358
+ },
359
+ {
360
+ "epoch": 0.08,
361
+ "learning_rate": 5e-05,
362
+ "loss": 0.4091,
363
+ "step": 29500
364
+ },
365
+ {
366
+ "epoch": 0.08,
367
+ "learning_rate": 5e-05,
368
+ "loss": 0.4107,
369
+ "step": 30000
370
+ },
371
+ {
372
+ "epoch": 0.08,
373
+ "learning_rate": 5e-05,
374
+ "loss": 0.4097,
375
+ "step": 30500
376
+ },
377
+ {
378
+ "epoch": 0.08,
379
+ "learning_rate": 5e-05,
380
+ "loss": 0.4056,
381
+ "step": 31000
382
+ },
383
+ {
384
+ "epoch": 0.08,
385
+ "learning_rate": 5e-05,
386
+ "loss": 0.4075,
387
+ "step": 31500
388
+ },
389
+ {
390
+ "epoch": 0.08,
391
+ "learning_rate": 5e-05,
392
+ "loss": 0.4075,
393
+ "step": 32000
394
+ },
395
+ {
396
+ "epoch": 0.08,
397
+ "learning_rate": 5e-05,
398
+ "loss": 0.4048,
399
+ "step": 32500
400
+ },
401
+ {
402
+ "epoch": 0.09,
403
+ "learning_rate": 5e-05,
404
+ "loss": 0.4042,
405
+ "step": 33000
406
+ },
407
+ {
408
+ "epoch": 0.09,
409
+ "learning_rate": 5e-05,
410
+ "loss": 0.4041,
411
+ "step": 33500
412
+ },
413
+ {
414
+ "epoch": 0.09,
415
+ "learning_rate": 5e-05,
416
+ "loss": 0.4051,
417
+ "step": 34000
418
+ },
419
+ {
420
+ "epoch": 0.09,
421
+ "learning_rate": 5e-05,
422
+ "loss": 0.4038,
423
+ "step": 34500
424
+ },
425
+ {
426
+ "epoch": 0.09,
427
+ "learning_rate": 5e-05,
428
+ "loss": 0.4044,
429
+ "step": 35000
430
+ },
431
+ {
432
+ "epoch": 0.09,
433
+ "learning_rate": 5e-05,
434
+ "loss": 0.4048,
435
+ "step": 35500
436
+ },
437
+ {
438
+ "epoch": 0.09,
439
+ "learning_rate": 5e-05,
440
+ "loss": 0.4026,
441
+ "step": 36000
442
+ },
443
+ {
444
+ "epoch": 0.1,
445
+ "learning_rate": 5e-05,
446
+ "loss": 0.402,
447
+ "step": 36500
448
+ },
449
+ {
450
+ "epoch": 0.1,
451
+ "learning_rate": 5e-05,
452
+ "loss": 0.4018,
453
+ "step": 37000
454
+ },
455
+ {
456
+ "epoch": 0.1,
457
+ "learning_rate": 5e-05,
458
+ "loss": 0.4018,
459
+ "step": 37500
460
+ },
461
+ {
462
+ "epoch": 0.1,
463
+ "learning_rate": 5e-05,
464
+ "loss": 0.3993,
465
+ "step": 38000
466
+ },
467
+ {
468
+ "epoch": 0.1,
469
+ "learning_rate": 5e-05,
470
+ "loss": 0.4021,
471
+ "step": 38500
472
+ },
473
+ {
474
+ "epoch": 0.1,
475
+ "learning_rate": 5e-05,
476
+ "loss": 0.4001,
477
+ "step": 39000
478
+ },
479
+ {
480
+ "epoch": 0.1,
481
+ "learning_rate": 5e-05,
482
+ "loss": 0.3987,
483
+ "step": 39500
484
+ },
485
+ {
486
+ "epoch": 0.1,
487
+ "learning_rate": 5e-05,
488
+ "loss": 0.3994,
489
+ "step": 40000
490
+ },
491
+ {
492
+ "epoch": 0.11,
493
+ "learning_rate": 5e-05,
494
+ "loss": 0.4001,
495
+ "step": 40500
496
+ },
497
+ {
498
+ "epoch": 0.11,
499
+ "learning_rate": 5e-05,
500
+ "loss": 0.3975,
501
+ "step": 41000
502
+ },
503
+ {
504
+ "epoch": 0.11,
505
+ "learning_rate": 5e-05,
506
+ "loss": 0.3954,
507
+ "step": 41500
508
+ },
509
+ {
510
+ "epoch": 0.11,
511
+ "learning_rate": 5e-05,
512
+ "loss": 0.3958,
513
+ "step": 42000
514
+ },
515
+ {
516
+ "epoch": 0.11,
517
+ "learning_rate": 5e-05,
518
+ "loss": 0.3963,
519
+ "step": 42500
520
+ },
521
+ {
522
+ "epoch": 0.11,
523
+ "learning_rate": 5e-05,
524
+ "loss": 0.3949,
525
+ "step": 43000
526
+ },
527
+ {
528
+ "epoch": 0.11,
529
+ "learning_rate": 5e-05,
530
+ "loss": 0.3946,
531
+ "step": 43500
532
+ },
533
+ {
534
+ "epoch": 0.11,
535
+ "learning_rate": 5e-05,
536
+ "loss": 0.3957,
537
+ "step": 44000
538
+ },
539
+ {
540
+ "epoch": 0.12,
541
+ "learning_rate": 5e-05,
542
+ "loss": 0.3932,
543
+ "step": 44500
544
+ },
545
+ {
546
+ "epoch": 0.12,
547
+ "learning_rate": 5e-05,
548
+ "loss": 0.3942,
549
+ "step": 45000
550
+ },
551
+ {
552
+ "epoch": 0.12,
553
+ "learning_rate": 5e-05,
554
+ "loss": 0.3919,
555
+ "step": 45500
556
+ },
557
+ {
558
+ "epoch": 0.12,
559
+ "learning_rate": 5e-05,
560
+ "loss": 0.3935,
561
+ "step": 46000
562
+ },
563
+ {
564
+ "epoch": 0.12,
565
+ "learning_rate": 5e-05,
566
+ "loss": 0.3919,
567
+ "step": 46500
568
+ },
569
+ {
570
+ "epoch": 0.12,
571
+ "learning_rate": 5e-05,
572
+ "loss": 0.3924,
573
+ "step": 47000
574
+ },
575
+ {
576
+ "epoch": 0.12,
577
+ "learning_rate": 5e-05,
578
+ "loss": 0.39,
579
+ "step": 47500
580
+ },
581
+ {
582
+ "epoch": 0.13,
583
+ "learning_rate": 5e-05,
584
+ "loss": 0.3896,
585
+ "step": 48000
586
+ },
587
+ {
588
+ "epoch": 0.13,
589
+ "learning_rate": 5e-05,
590
+ "loss": 0.3901,
591
+ "step": 48500
592
+ },
593
+ {
594
+ "epoch": 0.13,
595
+ "learning_rate": 5e-05,
596
+ "loss": 0.3916,
597
+ "step": 49000
598
+ },
599
+ {
600
+ "epoch": 0.13,
601
+ "learning_rate": 5e-05,
602
+ "loss": 0.3903,
603
+ "step": 49500
604
+ },
605
+ {
606
+ "epoch": 0.13,
607
+ "learning_rate": 5e-05,
608
+ "loss": 0.3885,
609
+ "step": 50000
610
+ },
611
+ {
612
+ "epoch": 0.13,
613
+ "learning_rate": 5e-05,
614
+ "loss": 0.3897,
615
+ "step": 50500
616
+ },
617
+ {
618
+ "epoch": 0.13,
619
+ "learning_rate": 5e-05,
620
+ "loss": 0.3883,
621
+ "step": 51000
622
+ },
623
+ {
624
+ "epoch": 0.13,
625
+ "learning_rate": 5e-05,
626
+ "loss": 0.3874,
627
+ "step": 51500
628
+ },
629
+ {
630
+ "epoch": 0.14,
631
+ "learning_rate": 5e-05,
632
+ "loss": 0.3892,
633
+ "step": 52000
634
+ },
635
+ {
636
+ "epoch": 0.14,
637
+ "learning_rate": 5e-05,
638
+ "loss": 0.388,
639
+ "step": 52500
640
+ },
641
+ {
642
+ "epoch": 0.14,
643
+ "learning_rate": 5e-05,
644
+ "loss": 0.3873,
645
+ "step": 53000
646
+ },
647
+ {
648
+ "epoch": 0.14,
649
+ "learning_rate": 5e-05,
650
+ "loss": 0.3879,
651
+ "step": 53500
652
+ },
653
+ {
654
+ "epoch": 0.14,
655
+ "learning_rate": 5e-05,
656
+ "loss": 0.3867,
657
+ "step": 54000
658
+ },
659
+ {
660
+ "epoch": 0.14,
661
+ "learning_rate": 5e-05,
662
+ "loss": 0.3873,
663
+ "step": 54500
664
+ },
665
+ {
666
+ "epoch": 0.14,
667
+ "learning_rate": 5e-05,
668
+ "loss": 0.3869,
669
+ "step": 55000
670
+ },
671
+ {
672
+ "epoch": 0.14,
673
+ "learning_rate": 5e-05,
674
+ "loss": 0.3872,
675
+ "step": 55500
676
+ },
677
+ {
678
+ "epoch": 0.15,
679
+ "learning_rate": 5e-05,
680
+ "loss": 0.3864,
681
+ "step": 56000
682
+ },
683
+ {
684
+ "epoch": 0.15,
685
+ "learning_rate": 5e-05,
686
+ "loss": 0.3869,
687
+ "step": 56500
688
+ },
689
+ {
690
+ "epoch": 0.15,
691
+ "learning_rate": 5e-05,
692
+ "loss": 0.3838,
693
+ "step": 57000
694
+ },
695
+ {
696
+ "epoch": 0.15,
697
+ "learning_rate": 5e-05,
698
+ "loss": 0.3856,
699
+ "step": 57500
700
+ },
701
+ {
702
+ "epoch": 0.15,
703
+ "learning_rate": 5e-05,
704
+ "loss": 0.3848,
705
+ "step": 58000
706
+ },
707
+ {
708
+ "epoch": 0.15,
709
+ "learning_rate": 5e-05,
710
+ "loss": 0.3861,
711
+ "step": 58500
712
+ },
713
+ {
714
+ "epoch": 0.15,
715
+ "learning_rate": 5e-05,
716
+ "loss": 0.3854,
717
+ "step": 59000
718
+ },
719
+ {
720
+ "epoch": 0.15,
721
+ "learning_rate": 5e-05,
722
+ "loss": 0.3843,
723
+ "step": 59500
724
+ },
725
+ {
726
+ "epoch": 0.16,
727
+ "learning_rate": 5e-05,
728
+ "loss": 0.386,
729
+ "step": 60000
730
+ },
731
+ {
732
+ "epoch": 0.16,
733
+ "learning_rate": 5e-05,
734
+ "loss": 0.3816,
735
+ "step": 60500
736
+ },
737
+ {
738
+ "epoch": 0.16,
739
+ "learning_rate": 5e-05,
740
+ "loss": 0.3814,
741
+ "step": 61000
742
+ },
743
+ {
744
+ "epoch": 0.16,
745
+ "learning_rate": 5e-05,
746
+ "loss": 0.3835,
747
+ "step": 61500
748
+ },
749
+ {
750
+ "epoch": 0.16,
751
+ "learning_rate": 5e-05,
752
+ "loss": 0.3818,
753
+ "step": 62000
754
+ },
755
+ {
756
+ "epoch": 0.16,
757
+ "learning_rate": 5e-05,
758
+ "loss": 0.3823,
759
+ "step": 62500
760
+ },
761
+ {
762
+ "epoch": 0.16,
763
+ "learning_rate": 5e-05,
764
+ "loss": 0.3825,
765
+ "step": 63000
766
+ },
767
+ {
768
+ "epoch": 0.17,
769
+ "learning_rate": 5e-05,
770
+ "loss": 0.3823,
771
+ "step": 63500
772
+ },
773
+ {
774
+ "epoch": 0.17,
775
+ "learning_rate": 5e-05,
776
+ "loss": 0.3806,
777
+ "step": 64000
778
+ },
779
+ {
780
+ "epoch": 0.17,
781
+ "learning_rate": 5e-05,
782
+ "loss": 0.3803,
783
+ "step": 64500
784
+ },
785
+ {
786
+ "epoch": 0.17,
787
+ "learning_rate": 5e-05,
788
+ "loss": 0.3797,
789
+ "step": 65000
790
+ },
791
+ {
792
+ "epoch": 0.17,
793
+ "learning_rate": 5e-05,
794
+ "loss": 0.3797,
795
+ "step": 65500
796
+ },
797
+ {
798
+ "epoch": 0.17,
799
+ "learning_rate": 5e-05,
800
+ "loss": 0.3793,
801
+ "step": 66000
802
+ },
803
+ {
804
+ "epoch": 0.17,
805
+ "learning_rate": 5e-05,
806
+ "loss": 0.3794,
807
+ "step": 66500
808
+ },
809
+ {
810
+ "epoch": 0.17,
811
+ "learning_rate": 5e-05,
812
+ "loss": 0.379,
813
+ "step": 67000
814
+ },
815
+ {
816
+ "epoch": 0.18,
817
+ "learning_rate": 5e-05,
818
+ "loss": 0.3783,
819
+ "step": 67500
820
+ },
821
+ {
822
+ "epoch": 0.18,
823
+ "learning_rate": 5e-05,
824
+ "loss": 0.3784,
825
+ "step": 68000
826
+ },
827
+ {
828
+ "epoch": 0.18,
829
+ "learning_rate": 5e-05,
830
+ "loss": 0.3778,
831
+ "step": 68500
832
+ },
833
+ {
834
+ "epoch": 0.18,
835
+ "learning_rate": 5e-05,
836
+ "loss": 0.3773,
837
+ "step": 69000
838
+ },
839
+ {
840
+ "epoch": 0.18,
841
+ "learning_rate": 5e-05,
842
+ "loss": 0.378,
843
+ "step": 69500
844
+ },
845
+ {
846
+ "epoch": 0.18,
847
+ "learning_rate": 5e-05,
848
+ "loss": 0.3759,
849
+ "step": 70000
850
+ },
851
+ {
852
+ "epoch": 0.18,
853
+ "learning_rate": 5e-05,
854
+ "loss": 0.377,
855
+ "step": 70500
856
+ },
857
+ {
858
+ "epoch": 0.18,
859
+ "learning_rate": 5e-05,
860
+ "loss": 0.3763,
861
+ "step": 71000
862
+ },
863
+ {
864
+ "epoch": 0.19,
865
+ "learning_rate": 5e-05,
866
+ "loss": 0.376,
867
+ "step": 71500
868
+ },
869
+ {
870
+ "epoch": 0.19,
871
+ "learning_rate": 5e-05,
872
+ "loss": 0.3767,
873
+ "step": 72000
874
+ },
875
+ {
876
+ "epoch": 0.19,
877
+ "learning_rate": 5e-05,
878
+ "loss": 0.3775,
879
+ "step": 72500
880
+ },
881
+ {
882
+ "epoch": 0.19,
883
+ "learning_rate": 5e-05,
884
+ "loss": 0.3794,
885
+ "step": 73000
886
+ },
887
+ {
888
+ "epoch": 0.19,
889
+ "learning_rate": 5e-05,
890
+ "loss": 0.3774,
891
+ "step": 73500
892
+ },
893
+ {
894
+ "epoch": 0.19,
895
+ "learning_rate": 5e-05,
896
+ "loss": 0.3762,
897
+ "step": 74000
898
+ },
899
+ {
900
+ "epoch": 0.19,
901
+ "learning_rate": 5e-05,
902
+ "loss": 0.3765,
903
+ "step": 74500
904
+ },
905
+ {
906
+ "epoch": 0.2,
907
+ "learning_rate": 5e-05,
908
+ "loss": 0.3776,
909
+ "step": 75000
910
+ },
911
+ {
912
+ "epoch": 0.2,
913
+ "learning_rate": 5e-05,
914
+ "loss": 0.3762,
915
+ "step": 75500
916
+ },
917
+ {
918
+ "epoch": 0.2,
919
+ "learning_rate": 5e-05,
920
+ "loss": 0.3761,
921
+ "step": 76000
922
+ },
923
+ {
924
+ "epoch": 0.2,
925
+ "learning_rate": 5e-05,
926
+ "loss": 0.3767,
927
+ "step": 76500
928
+ },
929
+ {
930
+ "epoch": 0.2,
931
+ "learning_rate": 5e-05,
932
+ "loss": 0.3756,
933
+ "step": 77000
934
+ },
935
+ {
936
+ "epoch": 0.2,
937
+ "learning_rate": 5e-05,
938
+ "loss": 0.3744,
939
+ "step": 77500
940
+ },
941
+ {
942
+ "epoch": 0.2,
943
+ "learning_rate": 5e-05,
944
+ "loss": 0.3739,
945
+ "step": 78000
946
+ },
947
+ {
948
+ "epoch": 0.2,
949
+ "learning_rate": 5e-05,
950
+ "loss": 0.3746,
951
+ "step": 78500
952
+ },
953
+ {
954
+ "epoch": 0.21,
955
+ "learning_rate": 5e-05,
956
+ "loss": 0.3736,
957
+ "step": 79000
958
+ },
959
+ {
960
+ "epoch": 0.21,
961
+ "learning_rate": 5e-05,
962
+ "loss": 0.3723,
963
+ "step": 79500
964
+ },
965
+ {
966
+ "epoch": 0.21,
967
+ "learning_rate": 5e-05,
968
+ "loss": 0.3739,
969
+ "step": 80000
970
+ },
971
+ {
972
+ "epoch": 0.21,
973
+ "learning_rate": 5e-05,
974
+ "loss": 0.3734,
975
+ "step": 80500
976
+ },
977
+ {
978
+ "epoch": 0.21,
979
+ "learning_rate": 5e-05,
980
+ "loss": 0.3726,
981
+ "step": 81000
982
+ },
983
+ {
984
+ "epoch": 0.21,
985
+ "learning_rate": 5e-05,
986
+ "loss": 0.3737,
987
+ "step": 81500
988
+ },
989
+ {
990
+ "epoch": 0.21,
991
+ "learning_rate": 5e-05,
992
+ "loss": 0.3726,
993
+ "step": 82000
994
+ },
995
+ {
996
+ "epoch": 0.21,
997
+ "learning_rate": 5e-05,
998
+ "loss": 0.3712,
999
+ "step": 82500
1000
+ },
1001
+ {
1002
+ "epoch": 0.22,
1003
+ "learning_rate": 5e-05,
1004
+ "loss": 0.3719,
1005
+ "step": 83000
1006
+ },
1007
+ {
1008
+ "epoch": 0.22,
1009
+ "learning_rate": 5e-05,
1010
+ "loss": 0.3714,
1011
+ "step": 83500
1012
+ },
1013
+ {
1014
+ "epoch": 0.22,
1015
+ "learning_rate": 5e-05,
1016
+ "loss": 0.3711,
1017
+ "step": 84000
1018
+ },
1019
+ {
1020
+ "epoch": 0.22,
1021
+ "learning_rate": 5e-05,
1022
+ "loss": 0.3707,
1023
+ "step": 84500
1024
+ },
1025
+ {
1026
+ "epoch": 0.22,
1027
+ "learning_rate": 5e-05,
1028
+ "loss": 0.369,
1029
+ "step": 85000
1030
+ },
1031
+ {
1032
+ "epoch": 0.22,
1033
+ "learning_rate": 5e-05,
1034
+ "loss": 0.3706,
1035
+ "step": 85500
1036
+ },
1037
+ {
1038
+ "epoch": 0.22,
1039
+ "learning_rate": 5e-05,
1040
+ "loss": 0.371,
1041
+ "step": 86000
1042
+ },
1043
+ {
1044
+ "epoch": 0.23,
1045
+ "learning_rate": 5e-05,
1046
+ "loss": 0.3706,
1047
+ "step": 86500
1048
+ },
1049
+ {
1050
+ "epoch": 0.23,
1051
+ "learning_rate": 5e-05,
1052
+ "loss": 0.3708,
1053
+ "step": 87000
1054
+ },
1055
+ {
1056
+ "epoch": 0.23,
1057
+ "learning_rate": 5e-05,
1058
+ "loss": 0.3687,
1059
+ "step": 87500
1060
+ },
1061
+ {
1062
+ "epoch": 0.23,
1063
+ "learning_rate": 5e-05,
1064
+ "loss": 0.3705,
1065
+ "step": 88000
1066
+ },
1067
+ {
1068
+ "epoch": 0.23,
1069
+ "learning_rate": 5e-05,
1070
+ "loss": 0.3699,
1071
+ "step": 88500
1072
+ },
1073
+ {
1074
+ "epoch": 0.23,
1075
+ "learning_rate": 5e-05,
1076
+ "loss": 0.3695,
1077
+ "step": 89000
1078
+ },
1079
+ {
1080
+ "epoch": 0.23,
1081
+ "learning_rate": 5e-05,
1082
+ "loss": 0.37,
1083
+ "step": 89500
1084
+ },
1085
+ {
1086
+ "epoch": 0.23,
1087
+ "learning_rate": 5e-05,
1088
+ "loss": 0.3707,
1089
+ "step": 90000
1090
+ },
1091
+ {
1092
+ "epoch": 0.24,
1093
+ "learning_rate": 5e-05,
1094
+ "loss": 0.3698,
1095
+ "step": 90500
1096
+ },
1097
+ {
1098
+ "epoch": 0.24,
1099
+ "learning_rate": 5e-05,
1100
+ "loss": 0.3692,
1101
+ "step": 91000
1102
+ },
1103
+ {
1104
+ "epoch": 0.24,
1105
+ "learning_rate": 5e-05,
1106
+ "loss": 0.37,
1107
+ "step": 91500
1108
+ },
1109
+ {
1110
+ "epoch": 0.24,
1111
+ "learning_rate": 5e-05,
1112
+ "loss": 0.3674,
1113
+ "step": 92000
1114
+ },
1115
+ {
1116
+ "epoch": 0.24,
1117
+ "learning_rate": 5e-05,
1118
+ "loss": 0.3694,
1119
+ "step": 92500
1120
+ },
1121
+ {
1122
+ "epoch": 0.24,
1123
+ "learning_rate": 5e-05,
1124
+ "loss": 0.3696,
1125
+ "step": 93000
1126
+ },
1127
+ {
1128
+ "epoch": 0.24,
1129
+ "learning_rate": 5e-05,
1130
+ "loss": 0.3677,
1131
+ "step": 93500
1132
+ },
1133
+ {
1134
+ "epoch": 0.24,
1135
+ "learning_rate": 5e-05,
1136
+ "loss": 0.3685,
1137
+ "step": 94000
1138
+ },
1139
+ {
1140
+ "epoch": 0.25,
1141
+ "learning_rate": 5e-05,
1142
+ "loss": 0.3671,
1143
+ "step": 94500
1144
+ },
1145
+ {
1146
+ "epoch": 0.25,
1147
+ "learning_rate": 5e-05,
1148
+ "loss": 0.3684,
1149
+ "step": 95000
1150
+ },
1151
+ {
1152
+ "epoch": 0.25,
1153
+ "learning_rate": 5e-05,
1154
+ "loss": 0.3664,
1155
+ "step": 95500
1156
+ },
1157
+ {
1158
+ "epoch": 0.25,
1159
+ "learning_rate": 5e-05,
1160
+ "loss": 0.3678,
1161
+ "step": 96000
1162
+ },
1163
+ {
1164
+ "epoch": 0.25,
1165
+ "learning_rate": 5e-05,
1166
+ "loss": 0.3677,
1167
+ "step": 96500
1168
+ },
1169
+ {
1170
+ "epoch": 0.25,
1171
+ "learning_rate": 5e-05,
1172
+ "loss": 0.3667,
1173
+ "step": 97000
1174
+ },
1175
+ {
1176
+ "epoch": 0.25,
1177
+ "learning_rate": 5e-05,
1178
+ "loss": 0.3676,
1179
+ "step": 97500
1180
+ },
1181
+ {
1182
+ "epoch": 0.26,
1183
+ "learning_rate": 5e-05,
1184
+ "loss": 0.3677,
1185
+ "step": 98000
1186
+ },
1187
+ {
1188
+ "epoch": 0.26,
1189
+ "learning_rate": 5e-05,
1190
+ "loss": 0.3651,
1191
+ "step": 98500
1192
+ },
1193
+ {
1194
+ "epoch": 0.26,
1195
+ "learning_rate": 5e-05,
1196
+ "loss": 0.3658,
1197
+ "step": 99000
1198
+ },
1199
+ {
1200
+ "epoch": 0.26,
1201
+ "learning_rate": 5e-05,
1202
+ "loss": 0.3651,
1203
+ "step": 99500
1204
+ },
1205
+ {
1206
+ "epoch": 0.26,
1207
+ "learning_rate": 5e-05,
1208
+ "loss": 0.3663,
1209
+ "step": 100000
1210
+ },
1211
+ {
1212
+ "epoch": 0.26,
1213
+ "learning_rate": 5e-05,
1214
+ "loss": 0.3657,
1215
+ "step": 100500
1216
+ },
1217
+ {
1218
+ "epoch": 0.26,
1219
+ "learning_rate": 5e-05,
1220
+ "loss": 0.3655,
1221
+ "step": 101000
1222
+ },
1223
+ {
1224
+ "epoch": 0.26,
1225
+ "learning_rate": 5e-05,
1226
+ "loss": 0.3637,
1227
+ "step": 101500
1228
+ },
1229
+ {
1230
+ "epoch": 0.27,
1231
+ "learning_rate": 5e-05,
1232
+ "loss": 0.3662,
1233
+ "step": 102000
1234
+ },
1235
+ {
1236
+ "epoch": 0.27,
1237
+ "learning_rate": 5e-05,
1238
+ "loss": 0.3638,
1239
+ "step": 102500
1240
+ },
1241
+ {
1242
+ "epoch": 0.27,
1243
+ "learning_rate": 5e-05,
1244
+ "loss": 0.3638,
1245
+ "step": 103000
1246
+ },
1247
+ {
1248
+ "epoch": 0.27,
1249
+ "learning_rate": 5e-05,
1250
+ "loss": 0.3663,
1251
+ "step": 103500
1252
+ },
1253
+ {
1254
+ "epoch": 0.27,
1255
+ "learning_rate": 5e-05,
1256
+ "loss": 0.3645,
1257
+ "step": 104000
1258
+ },
1259
+ {
1260
+ "epoch": 0.27,
1261
+ "learning_rate": 5e-05,
1262
+ "loss": 0.3661,
1263
+ "step": 104500
1264
+ },
1265
+ {
1266
+ "epoch": 0.27,
1267
+ "learning_rate": 5e-05,
1268
+ "loss": 0.3659,
1269
+ "step": 105000
1270
+ },
1271
+ {
1272
+ "epoch": 0.27,
1273
+ "learning_rate": 5e-05,
1274
+ "loss": 0.3651,
1275
+ "step": 105500
1276
+ },
1277
+ {
1278
+ "epoch": 0.28,
1279
+ "learning_rate": 5e-05,
1280
+ "loss": 0.3646,
1281
+ "step": 106000
1282
+ },
1283
+ {
1284
+ "epoch": 0.28,
1285
+ "learning_rate": 5e-05,
1286
+ "loss": 0.3645,
1287
+ "step": 106500
1288
+ },
1289
+ {
1290
+ "epoch": 0.28,
1291
+ "learning_rate": 5e-05,
1292
+ "loss": 0.3637,
1293
+ "step": 107000
1294
+ },
1295
+ {
1296
+ "epoch": 0.28,
1297
+ "learning_rate": 5e-05,
1298
+ "loss": 0.3634,
1299
+ "step": 107500
1300
+ },
1301
+ {
1302
+ "epoch": 0.28,
1303
+ "learning_rate": 5e-05,
1304
+ "loss": 0.3626,
1305
+ "step": 108000
1306
+ },
1307
+ {
1308
+ "epoch": 0.28,
1309
+ "learning_rate": 5e-05,
1310
+ "loss": 0.3636,
1311
+ "step": 108500
1312
+ },
1313
+ {
1314
+ "epoch": 0.28,
1315
+ "learning_rate": 5e-05,
1316
+ "loss": 0.3629,
1317
+ "step": 109000
1318
+ },
1319
+ {
1320
+ "epoch": 0.29,
1321
+ "learning_rate": 5e-05,
1322
+ "loss": 0.3625,
1323
+ "step": 109500
1324
+ },
1325
+ {
1326
+ "epoch": 0.29,
1327
+ "learning_rate": 5e-05,
1328
+ "loss": 0.3627,
1329
+ "step": 110000
1330
+ },
1331
+ {
1332
+ "epoch": 0.29,
1333
+ "learning_rate": 5e-05,
1334
+ "loss": 0.3628,
1335
+ "step": 110500
1336
+ },
1337
+ {
1338
+ "epoch": 0.29,
1339
+ "learning_rate": 5e-05,
1340
+ "loss": 0.3619,
1341
+ "step": 111000
1342
+ },
1343
+ {
1344
+ "epoch": 0.29,
1345
+ "learning_rate": 5e-05,
1346
+ "loss": 0.3627,
1347
+ "step": 111500
1348
+ },
1349
+ {
1350
+ "epoch": 0.29,
1351
+ "learning_rate": 5e-05,
1352
+ "loss": 0.3612,
1353
+ "step": 112000
1354
+ },
1355
+ {
1356
+ "epoch": 0.29,
1357
+ "learning_rate": 5e-05,
1358
+ "loss": 0.362,
1359
+ "step": 112500
1360
+ },
1361
+ {
1362
+ "epoch": 0.29,
1363
+ "learning_rate": 5e-05,
1364
+ "loss": 0.363,
1365
+ "step": 113000
1366
+ },
1367
+ {
1368
+ "epoch": 0.3,
1369
+ "learning_rate": 5e-05,
1370
+ "loss": 0.3607,
1371
+ "step": 113500
1372
+ },
1373
+ {
1374
+ "epoch": 0.3,
1375
+ "learning_rate": 5e-05,
1376
+ "loss": 0.3617,
1377
+ "step": 114000
1378
+ },
1379
+ {
1380
+ "epoch": 0.3,
1381
+ "learning_rate": 5e-05,
1382
+ "loss": 0.3605,
1383
+ "step": 114500
1384
+ },
1385
+ {
1386
+ "epoch": 0.3,
1387
+ "learning_rate": 5e-05,
1388
+ "loss": 0.3626,
1389
+ "step": 115000
1390
+ },
1391
+ {
1392
+ "epoch": 0.3,
1393
+ "learning_rate": 5e-05,
1394
+ "loss": 0.3619,
1395
+ "step": 115500
1396
+ },
1397
+ {
1398
+ "epoch": 0.3,
1399
+ "learning_rate": 5e-05,
1400
+ "loss": 0.3616,
1401
+ "step": 116000
1402
+ },
1403
+ {
1404
+ "epoch": 0.3,
1405
+ "learning_rate": 5e-05,
1406
+ "loss": 0.361,
1407
+ "step": 116500
1408
+ },
1409
+ {
1410
+ "epoch": 0.3,
1411
+ "learning_rate": 5e-05,
1412
+ "loss": 0.3609,
1413
+ "step": 117000
1414
+ },
1415
+ {
1416
+ "epoch": 0.31,
1417
+ "learning_rate": 5e-05,
1418
+ "loss": 0.3625,
1419
+ "step": 117500
1420
+ },
1421
+ {
1422
+ "epoch": 0.31,
1423
+ "learning_rate": 5e-05,
1424
+ "loss": 0.3619,
1425
+ "step": 118000
1426
+ },
1427
+ {
1428
+ "epoch": 0.31,
1429
+ "learning_rate": 5e-05,
1430
+ "loss": 0.3611,
1431
+ "step": 118500
1432
+ },
1433
+ {
1434
+ "epoch": 0.31,
1435
+ "learning_rate": 5e-05,
1436
+ "loss": 0.3603,
1437
+ "step": 119000
1438
+ },
1439
+ {
1440
+ "epoch": 0.31,
1441
+ "learning_rate": 5e-05,
1442
+ "loss": 0.3609,
1443
+ "step": 119500
1444
+ },
1445
+ {
1446
+ "epoch": 0.31,
1447
+ "learning_rate": 5e-05,
1448
+ "loss": 0.36,
1449
+ "step": 120000
1450
+ },
1451
+ {
1452
+ "epoch": 0.31,
1453
+ "learning_rate": 5e-05,
1454
+ "loss": 0.3612,
1455
+ "step": 120500
1456
+ },
1457
+ {
1458
+ "epoch": 0.32,
1459
+ "learning_rate": 5e-05,
1460
+ "loss": 0.3595,
1461
+ "step": 121000
1462
+ },
1463
+ {
1464
+ "epoch": 0.32,
1465
+ "learning_rate": 5e-05,
1466
+ "loss": 0.3602,
1467
+ "step": 121500
1468
+ },
1469
+ {
1470
+ "epoch": 0.32,
1471
+ "learning_rate": 5e-05,
1472
+ "loss": 0.3601,
1473
+ "step": 122000
1474
+ },
1475
+ {
1476
+ "epoch": 0.32,
1477
+ "learning_rate": 5e-05,
1478
+ "loss": 0.3593,
1479
+ "step": 122500
1480
+ },
1481
+ {
1482
+ "epoch": 0.32,
1483
+ "learning_rate": 5e-05,
1484
+ "loss": 0.3595,
1485
+ "step": 123000
1486
+ },
1487
+ {
1488
+ "epoch": 0.32,
1489
+ "learning_rate": 5e-05,
1490
+ "loss": 0.3586,
1491
+ "step": 123500
1492
+ },
1493
+ {
1494
+ "epoch": 0.32,
1495
+ "learning_rate": 5e-05,
1496
+ "loss": 0.3591,
1497
+ "step": 124000
1498
+ },
1499
+ {
1500
+ "epoch": 0.32,
1501
+ "learning_rate": 5e-05,
1502
+ "loss": 0.3582,
1503
+ "step": 124500
1504
+ },
1505
+ {
1506
+ "epoch": 0.33,
1507
+ "learning_rate": 5e-05,
1508
+ "loss": 0.3585,
1509
+ "step": 125000
1510
+ },
1511
+ {
1512
+ "epoch": 0.33,
1513
+ "learning_rate": 5e-05,
1514
+ "loss": 0.3589,
1515
+ "step": 125500
1516
+ },
1517
+ {
1518
+ "epoch": 0.33,
1519
+ "learning_rate": 5e-05,
1520
+ "loss": 0.3575,
1521
+ "step": 126000
1522
+ },
1523
+ {
1524
+ "epoch": 0.33,
1525
+ "learning_rate": 5e-05,
1526
+ "loss": 0.3599,
1527
+ "step": 126500
1528
+ },
1529
+ {
1530
+ "epoch": 0.33,
1531
+ "learning_rate": 5e-05,
1532
+ "loss": 0.3582,
1533
+ "step": 127000
1534
+ },
1535
+ {
1536
+ "epoch": 0.33,
1537
+ "learning_rate": 5e-05,
1538
+ "loss": 0.3566,
1539
+ "step": 127500
1540
+ },
1541
+ {
1542
+ "epoch": 0.33,
1543
+ "learning_rate": 5e-05,
1544
+ "loss": 0.3559,
1545
+ "step": 128000
1546
+ },
1547
+ {
1548
+ "epoch": 0.33,
1549
+ "learning_rate": 5e-05,
1550
+ "loss": 0.3598,
1551
+ "step": 128500
1552
+ },
1553
+ {
1554
+ "epoch": 0.34,
1555
+ "learning_rate": 5e-05,
1556
+ "loss": 0.3579,
1557
+ "step": 129000
1558
+ },
1559
+ {
1560
+ "epoch": 0.34,
1561
+ "learning_rate": 5e-05,
1562
+ "loss": 0.3581,
1563
+ "step": 129500
1564
+ },
1565
+ {
1566
+ "epoch": 0.34,
1567
+ "learning_rate": 5e-05,
1568
+ "loss": 0.3592,
1569
+ "step": 130000
1570
+ },
1571
+ {
1572
+ "epoch": 0.34,
1573
+ "learning_rate": 5e-05,
1574
+ "loss": 0.3585,
1575
+ "step": 130500
1576
+ },
1577
+ {
1578
+ "epoch": 0.34,
1579
+ "learning_rate": 5e-05,
1580
+ "loss": 0.3577,
1581
+ "step": 131000
1582
+ },
1583
+ {
1584
+ "epoch": 0.34,
1585
+ "learning_rate": 5e-05,
1586
+ "loss": 0.357,
1587
+ "step": 131500
1588
+ },
1589
+ {
1590
+ "epoch": 0.34,
1591
+ "learning_rate": 5e-05,
1592
+ "loss": 0.3588,
1593
+ "step": 132000
1594
+ },
1595
+ {
1596
+ "epoch": 0.35,
1597
+ "learning_rate": 5e-05,
1598
+ "loss": 0.358,
1599
+ "step": 132500
1600
+ },
1601
+ {
1602
+ "epoch": 0.35,
1603
+ "learning_rate": 5e-05,
1604
+ "loss": 0.3566,
1605
+ "step": 133000
1606
+ },
1607
+ {
1608
+ "epoch": 0.35,
1609
+ "learning_rate": 5e-05,
1610
+ "loss": 0.3574,
1611
+ "step": 133500
1612
+ },
1613
+ {
1614
+ "epoch": 0.35,
1615
+ "learning_rate": 5e-05,
1616
+ "loss": 0.357,
1617
+ "step": 134000
1618
+ },
1619
+ {
1620
+ "epoch": 0.35,
1621
+ "learning_rate": 5e-05,
1622
+ "loss": 0.3553,
1623
+ "step": 134500
1624
+ },
1625
+ {
1626
+ "epoch": 0.35,
1627
+ "learning_rate": 5e-05,
1628
+ "loss": 0.357,
1629
+ "step": 135000
1630
+ },
1631
+ {
1632
+ "epoch": 0.35,
1633
+ "learning_rate": 5e-05,
1634
+ "loss": 0.3578,
1635
+ "step": 135500
1636
+ },
1637
+ {
1638
+ "epoch": 0.35,
1639
+ "learning_rate": 5e-05,
1640
+ "loss": 0.3559,
1641
+ "step": 136000
1642
+ },
1643
+ {
1644
+ "epoch": 0.36,
1645
+ "learning_rate": 5e-05,
1646
+ "loss": 0.3559,
1647
+ "step": 136500
1648
+ },
1649
+ {
1650
+ "epoch": 0.36,
1651
+ "learning_rate": 5e-05,
1652
+ "loss": 0.356,
1653
+ "step": 137000
1654
+ },
1655
+ {
1656
+ "epoch": 0.36,
1657
+ "learning_rate": 5e-05,
1658
+ "loss": 0.3553,
1659
+ "step": 137500
1660
+ },
1661
+ {
1662
+ "epoch": 0.36,
1663
+ "learning_rate": 5e-05,
1664
+ "loss": 0.3565,
1665
+ "step": 138000
1666
+ },
1667
+ {
1668
+ "epoch": 0.36,
1669
+ "learning_rate": 5e-05,
1670
+ "loss": 0.3557,
1671
+ "step": 138500
1672
+ },
1673
+ {
1674
+ "epoch": 0.36,
1675
+ "learning_rate": 5e-05,
1676
+ "loss": 0.3541,
1677
+ "step": 139000
1678
+ },
1679
+ {
1680
+ "epoch": 0.36,
1681
+ "learning_rate": 5e-05,
1682
+ "loss": 0.3552,
1683
+ "step": 139500
1684
+ },
1685
+ {
1686
+ "epoch": 0.36,
1687
+ "learning_rate": 5e-05,
1688
+ "loss": 0.3571,
1689
+ "step": 140000
1690
+ },
1691
+ {
1692
+ "epoch": 0.37,
1693
+ "learning_rate": 5e-05,
1694
+ "loss": 0.3565,
1695
+ "step": 140500
1696
+ },
1697
+ {
1698
+ "epoch": 0.37,
1699
+ "learning_rate": 5e-05,
1700
+ "loss": 0.3573,
1701
+ "step": 141000
1702
+ },
1703
+ {
1704
+ "epoch": 0.37,
1705
+ "learning_rate": 5e-05,
1706
+ "loss": 0.3561,
1707
+ "step": 141500
1708
+ },
1709
+ {
1710
+ "epoch": 0.37,
1711
+ "learning_rate": 5e-05,
1712
+ "loss": 0.3569,
1713
+ "step": 142000
1714
+ },
1715
+ {
1716
+ "epoch": 0.37,
1717
+ "learning_rate": 5e-05,
1718
+ "loss": 0.3558,
1719
+ "step": 142500
1720
+ },
1721
+ {
1722
+ "epoch": 0.37,
1723
+ "learning_rate": 5e-05,
1724
+ "loss": 0.3544,
1725
+ "step": 143000
1726
+ },
1727
+ {
1728
+ "epoch": 0.37,
1729
+ "learning_rate": 5e-05,
1730
+ "loss": 0.3559,
1731
+ "step": 143500
1732
+ },
1733
+ {
1734
+ "epoch": 0.38,
1735
+ "learning_rate": 5e-05,
1736
+ "loss": 0.3558,
1737
+ "step": 144000
1738
+ },
1739
+ {
1740
+ "epoch": 0.38,
1741
+ "learning_rate": 5e-05,
1742
+ "loss": 0.3545,
1743
+ "step": 144500
1744
+ },
1745
+ {
1746
+ "epoch": 0.38,
1747
+ "learning_rate": 5e-05,
1748
+ "loss": 0.3542,
1749
+ "step": 145000
1750
+ },
1751
+ {
1752
+ "epoch": 0.38,
1753
+ "learning_rate": 5e-05,
1754
+ "loss": 0.3543,
1755
+ "step": 145500
1756
+ },
1757
+ {
1758
+ "epoch": 0.38,
1759
+ "learning_rate": 5e-05,
1760
+ "loss": 0.3541,
1761
+ "step": 146000
1762
+ },
1763
+ {
1764
+ "epoch": 0.38,
1765
+ "learning_rate": 5e-05,
1766
+ "loss": 0.3545,
1767
+ "step": 146500
1768
+ },
1769
+ {
1770
+ "epoch": 0.38,
1771
+ "learning_rate": 5e-05,
1772
+ "loss": 0.3542,
1773
+ "step": 147000
1774
+ },
1775
+ {
1776
+ "epoch": 0.38,
1777
+ "learning_rate": 5e-05,
1778
+ "loss": 0.3538,
1779
+ "step": 147500
1780
+ },
1781
+ {
1782
+ "epoch": 0.39,
1783
+ "learning_rate": 5e-05,
1784
+ "loss": 0.3544,
1785
+ "step": 148000
1786
+ },
1787
+ {
1788
+ "epoch": 0.39,
1789
+ "learning_rate": 5e-05,
1790
+ "loss": 0.3538,
1791
+ "step": 148500
1792
+ },
1793
+ {
1794
+ "epoch": 0.39,
1795
+ "learning_rate": 5e-05,
1796
+ "loss": 0.3525,
1797
+ "step": 149000
1798
+ },
1799
+ {
1800
+ "epoch": 0.39,
1801
+ "learning_rate": 5e-05,
1802
+ "loss": 0.3535,
1803
+ "step": 149500
1804
+ },
1805
+ {
1806
+ "epoch": 0.39,
1807
+ "learning_rate": 5e-05,
1808
+ "loss": 0.3527,
1809
+ "step": 150000
1810
+ },
1811
+ {
1812
+ "epoch": 0.39,
1813
+ "learning_rate": 5e-05,
1814
+ "loss": 0.353,
1815
+ "step": 150500
1816
+ },
1817
+ {
1818
+ "epoch": 0.39,
1819
+ "learning_rate": 5e-05,
1820
+ "loss": 0.356,
1821
+ "step": 151000
1822
+ },
1823
+ {
1824
+ "epoch": 0.39,
1825
+ "learning_rate": 5e-05,
1826
+ "loss": 0.3541,
1827
+ "step": 151500
1828
+ },
1829
+ {
1830
+ "epoch": 0.4,
1831
+ "learning_rate": 5e-05,
1832
+ "loss": 0.3542,
1833
+ "step": 152000
1834
+ },
1835
+ {
1836
+ "epoch": 0.4,
1837
+ "learning_rate": 5e-05,
1838
+ "loss": 0.3541,
1839
+ "step": 152500
1840
+ },
1841
+ {
1842
+ "epoch": 0.4,
1843
+ "learning_rate": 5e-05,
1844
+ "loss": 0.3541,
1845
+ "step": 153000
1846
+ },
1847
+ {
1848
+ "epoch": 0.4,
1849
+ "learning_rate": 5e-05,
1850
+ "loss": 0.3537,
1851
+ "step": 153500
1852
+ },
1853
+ {
1854
+ "epoch": 0.4,
1855
+ "learning_rate": 5e-05,
1856
+ "loss": 0.3543,
1857
+ "step": 154000
1858
+ },
1859
+ {
1860
+ "epoch": 0.4,
1861
+ "learning_rate": 5e-05,
1862
+ "loss": 0.3547,
1863
+ "step": 154500
1864
+ },
1865
+ {
1866
+ "epoch": 0.4,
1867
+ "learning_rate": 5e-05,
1868
+ "loss": 0.352,
1869
+ "step": 155000
1870
+ },
1871
+ {
1872
+ "epoch": 0.4,
1873
+ "learning_rate": 5e-05,
1874
+ "loss": 0.3523,
1875
+ "step": 155500
1876
+ },
1877
+ {
1878
+ "epoch": 0.41,
1879
+ "learning_rate": 5e-05,
1880
+ "loss": 0.3515,
1881
+ "step": 156000
1882
+ },
1883
+ {
1884
+ "epoch": 0.41,
1885
+ "learning_rate": 5e-05,
1886
+ "loss": 0.3508,
1887
+ "step": 156500
1888
+ },
1889
+ {
1890
+ "epoch": 0.41,
1891
+ "learning_rate": 5e-05,
1892
+ "loss": 0.3524,
1893
+ "step": 157000
1894
+ },
1895
+ {
1896
+ "epoch": 0.41,
1897
+ "learning_rate": 5e-05,
1898
+ "loss": 0.3519,
1899
+ "step": 157500
1900
+ },
1901
+ {
1902
+ "epoch": 0.41,
1903
+ "learning_rate": 5e-05,
1904
+ "loss": 0.3525,
1905
+ "step": 158000
1906
+ },
1907
+ {
1908
+ "epoch": 0.41,
1909
+ "learning_rate": 5e-05,
1910
+ "loss": 0.3524,
1911
+ "step": 158500
1912
+ },
1913
+ {
1914
+ "epoch": 0.41,
1915
+ "learning_rate": 5e-05,
1916
+ "loss": 0.3508,
1917
+ "step": 159000
1918
+ },
1919
+ {
1920
+ "epoch": 0.42,
1921
+ "learning_rate": 5e-05,
1922
+ "loss": 0.3529,
1923
+ "step": 159500
1924
+ },
1925
+ {
1926
+ "epoch": 0.42,
1927
+ "learning_rate": 5e-05,
1928
+ "loss": 0.3529,
1929
+ "step": 160000
1930
+ },
1931
+ {
1932
+ "epoch": 0.42,
1933
+ "learning_rate": 5e-05,
1934
+ "loss": 0.3525,
1935
+ "step": 160500
1936
+ },
1937
+ {
1938
+ "epoch": 0.42,
1939
+ "learning_rate": 5e-05,
1940
+ "loss": 0.3535,
1941
+ "step": 161000
1942
+ },
1943
+ {
1944
+ "epoch": 0.42,
1945
+ "learning_rate": 5e-05,
1946
+ "loss": 0.3515,
1947
+ "step": 161500
1948
+ },
1949
+ {
1950
+ "epoch": 0.42,
1951
+ "learning_rate": 5e-05,
1952
+ "loss": 0.3526,
1953
+ "step": 162000
1954
+ },
1955
+ {
1956
+ "epoch": 0.42,
1957
+ "learning_rate": 5e-05,
1958
+ "loss": 0.3507,
1959
+ "step": 162500
1960
+ },
1961
+ {
1962
+ "epoch": 0.42,
1963
+ "learning_rate": 5e-05,
1964
+ "loss": 0.3521,
1965
+ "step": 163000
1966
+ },
1967
+ {
1968
+ "epoch": 0.43,
1969
+ "learning_rate": 5e-05,
1970
+ "loss": 0.3512,
1971
+ "step": 163500
1972
+ },
1973
+ {
1974
+ "epoch": 0.43,
1975
+ "learning_rate": 5e-05,
1976
+ "loss": 0.351,
1977
+ "step": 164000
1978
+ },
1979
+ {
1980
+ "epoch": 0.43,
1981
+ "learning_rate": 5e-05,
1982
+ "loss": 0.3507,
1983
+ "step": 164500
1984
+ },
1985
+ {
1986
+ "epoch": 0.43,
1987
+ "learning_rate": 5e-05,
1988
+ "loss": 0.3514,
1989
+ "step": 165000
1990
+ },
1991
+ {
1992
+ "epoch": 0.43,
1993
+ "learning_rate": 5e-05,
1994
+ "loss": 0.3505,
1995
+ "step": 165500
1996
+ },
1997
+ {
1998
+ "epoch": 0.43,
1999
+ "learning_rate": 5e-05,
2000
+ "loss": 0.3492,
2001
+ "step": 166000
2002
+ },
2003
+ {
2004
+ "epoch": 0.43,
2005
+ "learning_rate": 5e-05,
2006
+ "loss": 0.3507,
2007
+ "step": 166500
2008
+ },
2009
+ {
2010
+ "epoch": 0.43,
2011
+ "learning_rate": 5e-05,
2012
+ "loss": 0.3511,
2013
+ "step": 167000
2014
+ },
2015
+ {
2016
+ "epoch": 0.44,
2017
+ "learning_rate": 5e-05,
2018
+ "loss": 0.3497,
2019
+ "step": 167500
2020
+ },
2021
+ {
2022
+ "epoch": 0.44,
2023
+ "learning_rate": 5e-05,
2024
+ "loss": 0.3506,
2025
+ "step": 168000
2026
+ },
2027
+ {
2028
+ "epoch": 0.44,
2029
+ "learning_rate": 5e-05,
2030
+ "loss": 0.3507,
2031
+ "step": 168500
2032
+ },
2033
+ {
2034
+ "epoch": 0.44,
2035
+ "learning_rate": 5e-05,
2036
+ "loss": 0.3501,
2037
+ "step": 169000
2038
+ },
2039
+ {
2040
+ "epoch": 0.44,
2041
+ "learning_rate": 5e-05,
2042
+ "loss": 0.3491,
2043
+ "step": 169500
2044
+ },
2045
+ {
2046
+ "epoch": 0.44,
2047
+ "learning_rate": 5e-05,
2048
+ "loss": 0.3509,
2049
+ "step": 170000
2050
+ },
2051
+ {
2052
+ "epoch": 0.44,
2053
+ "learning_rate": 5e-05,
2054
+ "loss": 0.35,
2055
+ "step": 170500
2056
+ },
2057
+ {
2058
+ "epoch": 0.45,
2059
+ "learning_rate": 5e-05,
2060
+ "loss": 0.3505,
2061
+ "step": 171000
2062
+ },
2063
+ {
2064
+ "epoch": 0.45,
2065
+ "learning_rate": 5e-05,
2066
+ "loss": 0.3507,
2067
+ "step": 171500
2068
+ },
2069
+ {
2070
+ "epoch": 0.45,
2071
+ "learning_rate": 5e-05,
2072
+ "loss": 0.3496,
2073
+ "step": 172000
2074
+ },
2075
+ {
2076
+ "epoch": 0.45,
2077
+ "learning_rate": 5e-05,
2078
+ "loss": 0.3503,
2079
+ "step": 172500
2080
+ },
2081
+ {
2082
+ "epoch": 0.45,
2083
+ "learning_rate": 5e-05,
2084
+ "loss": 0.3502,
2085
+ "step": 173000
2086
+ },
2087
+ {
2088
+ "epoch": 0.45,
2089
+ "learning_rate": 5e-05,
2090
+ "loss": 0.3498,
2091
+ "step": 173500
2092
+ },
2093
+ {
2094
+ "epoch": 0.45,
2095
+ "learning_rate": 5e-05,
2096
+ "loss": 0.3496,
2097
+ "step": 174000
2098
+ },
2099
+ {
2100
+ "epoch": 0.45,
2101
+ "learning_rate": 5e-05,
2102
+ "loss": 0.3498,
2103
+ "step": 174500
2104
+ },
2105
+ {
2106
+ "epoch": 0.46,
2107
+ "learning_rate": 5e-05,
2108
+ "loss": 0.3498,
2109
+ "step": 175000
2110
+ },
2111
+ {
2112
+ "epoch": 0.46,
2113
+ "learning_rate": 5e-05,
2114
+ "loss": 0.3473,
2115
+ "step": 175500
2116
+ },
2117
+ {
2118
+ "epoch": 0.46,
2119
+ "learning_rate": 5e-05,
2120
+ "loss": 0.3482,
2121
+ "step": 176000
2122
+ },
2123
+ {
2124
+ "epoch": 0.46,
2125
+ "learning_rate": 5e-05,
2126
+ "loss": 0.3482,
2127
+ "step": 176500
2128
+ },
2129
+ {
2130
+ "epoch": 0.46,
2131
+ "learning_rate": 5e-05,
2132
+ "loss": 0.3486,
2133
+ "step": 177000
2134
+ },
2135
+ {
2136
+ "epoch": 0.46,
2137
+ "learning_rate": 5e-05,
2138
+ "loss": 0.3504,
2139
+ "step": 177500
2140
+ },
2141
+ {
2142
+ "epoch": 0.46,
2143
+ "learning_rate": 5e-05,
2144
+ "loss": 0.3492,
2145
+ "step": 178000
2146
+ },
2147
+ {
2148
+ "epoch": 0.46,
2149
+ "learning_rate": 5e-05,
2150
+ "loss": 0.3502,
2151
+ "step": 178500
2152
+ },
2153
+ {
2154
+ "epoch": 0.47,
2155
+ "learning_rate": 5e-05,
2156
+ "loss": 0.3475,
2157
+ "step": 179000
2158
+ },
2159
+ {
2160
+ "epoch": 0.47,
2161
+ "learning_rate": 5e-05,
2162
+ "loss": 0.3487,
2163
+ "step": 179500
2164
+ },
2165
+ {
2166
+ "epoch": 0.47,
2167
+ "learning_rate": 5e-05,
2168
+ "loss": 0.3501,
2169
+ "step": 180000
2170
+ },
2171
+ {
2172
+ "epoch": 0.47,
2173
+ "learning_rate": 5e-05,
2174
+ "loss": 0.3478,
2175
+ "step": 180500
2176
+ },
2177
+ {
2178
+ "epoch": 0.47,
2179
+ "learning_rate": 5e-05,
2180
+ "loss": 0.348,
2181
+ "step": 181000
2182
+ },
2183
+ {
2184
+ "epoch": 0.47,
2185
+ "learning_rate": 5e-05,
2186
+ "loss": 0.3479,
2187
+ "step": 181500
2188
+ },
2189
+ {
2190
+ "epoch": 0.47,
2191
+ "learning_rate": 5e-05,
2192
+ "loss": 0.3467,
2193
+ "step": 182000
2194
+ },
2195
+ {
2196
+ "epoch": 0.48,
2197
+ "learning_rate": 5e-05,
2198
+ "loss": 0.3474,
2199
+ "step": 182500
2200
+ },
2201
+ {
2202
+ "epoch": 0.48,
2203
+ "learning_rate": 5e-05,
2204
+ "loss": 0.3454,
2205
+ "step": 183000
2206
+ },
2207
+ {
2208
+ "epoch": 0.48,
2209
+ "learning_rate": 5e-05,
2210
+ "loss": 0.347,
2211
+ "step": 183500
2212
+ },
2213
+ {
2214
+ "epoch": 0.48,
2215
+ "learning_rate": 5e-05,
2216
+ "loss": 0.3491,
2217
+ "step": 184000
2218
+ },
2219
+ {
2220
+ "epoch": 0.48,
2221
+ "learning_rate": 5e-05,
2222
+ "loss": 0.3487,
2223
+ "step": 184500
2224
+ },
2225
+ {
2226
+ "epoch": 0.48,
2227
+ "learning_rate": 5e-05,
2228
+ "loss": 0.3489,
2229
+ "step": 185000
2230
+ },
2231
+ {
2232
+ "epoch": 0.48,
2233
+ "learning_rate": 5e-05,
2234
+ "loss": 0.3481,
2235
+ "step": 185500
2236
+ },
2237
+ {
2238
+ "epoch": 0.48,
2239
+ "learning_rate": 5e-05,
2240
+ "loss": 0.3462,
2241
+ "step": 186000
2242
+ },
2243
+ {
2244
+ "epoch": 0.49,
2245
+ "learning_rate": 5e-05,
2246
+ "loss": 0.346,
2247
+ "step": 186500
2248
+ },
2249
+ {
2250
+ "epoch": 0.49,
2251
+ "learning_rate": 5e-05,
2252
+ "loss": 0.3458,
2253
+ "step": 187000
2254
+ },
2255
+ {
2256
+ "epoch": 0.49,
2257
+ "learning_rate": 5e-05,
2258
+ "loss": 0.3467,
2259
+ "step": 187500
2260
+ },
2261
+ {
2262
+ "epoch": 0.49,
2263
+ "learning_rate": 5e-05,
2264
+ "loss": 0.3465,
2265
+ "step": 188000
2266
+ },
2267
+ {
2268
+ "epoch": 0.49,
2269
+ "learning_rate": 5e-05,
2270
+ "loss": 0.3462,
2271
+ "step": 188500
2272
+ },
2273
+ {
2274
+ "epoch": 0.49,
2275
+ "learning_rate": 5e-05,
2276
+ "loss": 0.3474,
2277
+ "step": 189000
2278
+ },
2279
+ {
2280
+ "epoch": 0.49,
2281
+ "learning_rate": 5e-05,
2282
+ "loss": 0.3471,
2283
+ "step": 189500
2284
+ },
2285
+ {
2286
+ "epoch": 0.49,
2287
+ "learning_rate": 5e-05,
2288
+ "loss": 0.348,
2289
+ "step": 190000
2290
+ },
2291
+ {
2292
+ "epoch": 0.5,
2293
+ "learning_rate": 5e-05,
2294
+ "loss": 0.3466,
2295
+ "step": 190500
2296
+ },
2297
+ {
2298
+ "epoch": 0.5,
2299
+ "learning_rate": 5e-05,
2300
+ "loss": 0.3465,
2301
+ "step": 191000
2302
+ },
2303
+ {
2304
+ "epoch": 0.5,
2305
+ "learning_rate": 5e-05,
2306
+ "loss": 0.3472,
2307
+ "step": 191500
2308
+ },
2309
+ {
2310
+ "epoch": 0.5,
2311
+ "learning_rate": 5e-05,
2312
+ "loss": 0.3463,
2313
+ "step": 192000
2314
+ },
2315
+ {
2316
+ "epoch": 0.5,
2317
+ "learning_rate": 5e-05,
2318
+ "loss": 0.3478,
2319
+ "step": 192500
2320
+ },
2321
+ {
2322
+ "epoch": 0.5,
2323
+ "learning_rate": 5e-05,
2324
+ "loss": 0.3479,
2325
+ "step": 193000
2326
+ },
2327
+ {
2328
+ "epoch": 0.5,
2329
+ "learning_rate": 5e-05,
2330
+ "loss": 0.3468,
2331
+ "step": 193500
2332
+ },
2333
+ {
2334
+ "epoch": 0.51,
2335
+ "learning_rate": 5e-05,
2336
+ "loss": 0.3463,
2337
+ "step": 194000
2338
+ },
2339
+ {
2340
+ "epoch": 0.51,
2341
+ "learning_rate": 5e-05,
2342
+ "loss": 0.3449,
2343
+ "step": 194500
2344
+ },
2345
+ {
2346
+ "epoch": 0.51,
2347
+ "learning_rate": 5e-05,
2348
+ "loss": 0.3465,
2349
+ "step": 195000
2350
+ },
2351
+ {
2352
+ "epoch": 0.51,
2353
+ "learning_rate": 5e-05,
2354
+ "loss": 0.3452,
2355
+ "step": 195500
2356
+ },
2357
+ {
2358
+ "epoch": 0.51,
2359
+ "learning_rate": 5e-05,
2360
+ "loss": 0.3451,
2361
+ "step": 196000
2362
+ },
2363
+ {
2364
+ "epoch": 0.51,
2365
+ "learning_rate": 5e-05,
2366
+ "loss": 0.3463,
2367
+ "step": 196500
2368
+ },
2369
+ {
2370
+ "epoch": 0.51,
2371
+ "learning_rate": 5e-05,
2372
+ "loss": 0.345,
2373
+ "step": 197000
2374
+ },
2375
+ {
2376
+ "epoch": 0.51,
2377
+ "learning_rate": 5e-05,
2378
+ "loss": 0.3458,
2379
+ "step": 197500
2380
+ },
2381
+ {
2382
+ "epoch": 0.52,
2383
+ "learning_rate": 5e-05,
2384
+ "loss": 0.3465,
2385
+ "step": 198000
2386
+ },
2387
+ {
2388
+ "epoch": 0.52,
2389
+ "learning_rate": 5e-05,
2390
+ "loss": 0.3445,
2391
+ "step": 198500
2392
+ },
2393
+ {
2394
+ "epoch": 0.52,
2395
+ "learning_rate": 5e-05,
2396
+ "loss": 0.3443,
2397
+ "step": 199000
2398
+ },
2399
+ {
2400
+ "epoch": 0.52,
2401
+ "learning_rate": 5e-05,
2402
+ "loss": 0.3457,
2403
+ "step": 199500
2404
+ },
2405
+ {
2406
+ "epoch": 0.52,
2407
+ "learning_rate": 5e-05,
2408
+ "loss": 0.3459,
2409
+ "step": 200000
2410
+ },
2411
+ {
2412
+ "epoch": 0.52,
2413
+ "learning_rate": 5e-05,
2414
+ "loss": 0.3454,
2415
+ "step": 200500
2416
+ },
2417
+ {
2418
+ "epoch": 0.52,
2419
+ "learning_rate": 5e-05,
2420
+ "loss": 0.3439,
2421
+ "step": 201000
2422
+ },
2423
+ {
2424
+ "epoch": 0.52,
2425
+ "learning_rate": 5e-05,
2426
+ "loss": 0.3456,
2427
+ "step": 201500
2428
+ },
2429
+ {
2430
+ "epoch": 0.53,
2431
+ "learning_rate": 5e-05,
2432
+ "loss": 0.3453,
2433
+ "step": 202000
2434
+ },
2435
+ {
2436
+ "epoch": 0.53,
2437
+ "learning_rate": 5e-05,
2438
+ "loss": 0.3459,
2439
+ "step": 202500
2440
+ },
2441
+ {
2442
+ "epoch": 0.53,
2443
+ "learning_rate": 5e-05,
2444
+ "loss": 0.3446,
2445
+ "step": 203000
2446
+ },
2447
+ {
2448
+ "epoch": 0.53,
2449
+ "learning_rate": 5e-05,
2450
+ "loss": 0.3441,
2451
+ "step": 203500
2452
+ },
2453
+ {
2454
+ "epoch": 0.53,
2455
+ "learning_rate": 5e-05,
2456
+ "loss": 0.3469,
2457
+ "step": 204000
2458
+ },
2459
+ {
2460
+ "epoch": 0.53,
2461
+ "learning_rate": 5e-05,
2462
+ "loss": 0.3457,
2463
+ "step": 204500
2464
+ },
2465
+ {
2466
+ "epoch": 0.53,
2467
+ "learning_rate": 5e-05,
2468
+ "loss": 0.3446,
2469
+ "step": 205000
2470
+ },
2471
+ {
2472
+ "epoch": 0.54,
2473
+ "learning_rate": 5e-05,
2474
+ "loss": 0.3433,
2475
+ "step": 205500
2476
+ },
2477
+ {
2478
+ "epoch": 0.54,
2479
+ "learning_rate": 5e-05,
2480
+ "loss": 0.3436,
2481
+ "step": 206000
2482
+ },
2483
+ {
2484
+ "epoch": 0.54,
2485
+ "learning_rate": 5e-05,
2486
+ "loss": 0.345,
2487
+ "step": 206500
2488
+ },
2489
+ {
2490
+ "epoch": 0.54,
2491
+ "learning_rate": 5e-05,
2492
+ "loss": 0.3436,
2493
+ "step": 207000
2494
+ },
2495
+ {
2496
+ "epoch": 0.54,
2497
+ "learning_rate": 5e-05,
2498
+ "loss": 0.3442,
2499
+ "step": 207500
2500
+ },
2501
+ {
2502
+ "epoch": 0.54,
2503
+ "learning_rate": 5e-05,
2504
+ "loss": 0.3444,
2505
+ "step": 208000
2506
+ },
2507
+ {
2508
+ "epoch": 0.54,
2509
+ "learning_rate": 5e-05,
2510
+ "loss": 0.3436,
2511
+ "step": 208500
2512
+ },
2513
+ {
2514
+ "epoch": 0.54,
2515
+ "learning_rate": 5e-05,
2516
+ "loss": 0.3439,
2517
+ "step": 209000
2518
+ },
2519
+ {
2520
+ "epoch": 0.55,
2521
+ "learning_rate": 5e-05,
2522
+ "loss": 0.3452,
2523
+ "step": 209500
2524
+ },
2525
+ {
2526
+ "epoch": 0.55,
2527
+ "learning_rate": 5e-05,
2528
+ "loss": 0.3448,
2529
+ "step": 210000
2530
+ },
2531
+ {
2532
+ "epoch": 0.55,
2533
+ "learning_rate": 5e-05,
2534
+ "loss": 0.3447,
2535
+ "step": 210500
2536
+ },
2537
+ {
2538
+ "epoch": 0.55,
2539
+ "learning_rate": 5e-05,
2540
+ "loss": 0.3447,
2541
+ "step": 211000
2542
+ },
2543
+ {
2544
+ "epoch": 0.55,
2545
+ "learning_rate": 5e-05,
2546
+ "loss": 0.3433,
2547
+ "step": 211500
2548
+ },
2549
+ {
2550
+ "epoch": 0.55,
2551
+ "learning_rate": 5e-05,
2552
+ "loss": 0.3434,
2553
+ "step": 212000
2554
+ },
2555
+ {
2556
+ "epoch": 0.55,
2557
+ "learning_rate": 5e-05,
2558
+ "loss": 0.3444,
2559
+ "step": 212500
2560
+ },
2561
+ {
2562
+ "epoch": 0.55,
2563
+ "learning_rate": 5e-05,
2564
+ "loss": 0.3443,
2565
+ "step": 213000
2566
+ },
2567
+ {
2568
+ "epoch": 0.56,
2569
+ "learning_rate": 5e-05,
2570
+ "loss": 0.3441,
2571
+ "step": 213500
2572
+ },
2573
+ {
2574
+ "epoch": 0.56,
2575
+ "learning_rate": 5e-05,
2576
+ "loss": 0.3434,
2577
+ "step": 214000
2578
+ },
2579
+ {
2580
+ "epoch": 0.56,
2581
+ "learning_rate": 5e-05,
2582
+ "loss": 0.3443,
2583
+ "step": 214500
2584
+ },
2585
+ {
2586
+ "epoch": 0.56,
2587
+ "learning_rate": 5e-05,
2588
+ "loss": 0.3425,
2589
+ "step": 215000
2590
+ },
2591
+ {
2592
+ "epoch": 0.56,
2593
+ "learning_rate": 5e-05,
2594
+ "loss": 0.3437,
2595
+ "step": 215500
2596
+ },
2597
+ {
2598
+ "epoch": 0.56,
2599
+ "learning_rate": 5e-05,
2600
+ "loss": 0.3432,
2601
+ "step": 216000
2602
+ },
2603
+ {
2604
+ "epoch": 0.56,
2605
+ "learning_rate": 5e-05,
2606
+ "loss": 0.3436,
2607
+ "step": 216500
2608
+ },
2609
+ {
2610
+ "epoch": 0.57,
2611
+ "learning_rate": 5e-05,
2612
+ "loss": 0.3426,
2613
+ "step": 217000
2614
+ },
2615
+ {
2616
+ "epoch": 0.57,
2617
+ "learning_rate": 5e-05,
2618
+ "loss": 0.3419,
2619
+ "step": 217500
2620
+ },
2621
+ {
2622
+ "epoch": 0.57,
2623
+ "learning_rate": 5e-05,
2624
+ "loss": 0.3435,
2625
+ "step": 218000
2626
+ },
2627
+ {
2628
+ "epoch": 0.57,
2629
+ "learning_rate": 5e-05,
2630
+ "loss": 0.3422,
2631
+ "step": 218500
2632
+ },
2633
+ {
2634
+ "epoch": 0.57,
2635
+ "learning_rate": 5e-05,
2636
+ "loss": 0.3439,
2637
+ "step": 219000
2638
+ },
2639
+ {
2640
+ "epoch": 0.57,
2641
+ "learning_rate": 5e-05,
2642
+ "loss": 0.3431,
2643
+ "step": 219500
2644
+ },
2645
+ {
2646
+ "epoch": 0.57,
2647
+ "learning_rate": 5e-05,
2648
+ "loss": 0.3431,
2649
+ "step": 220000
2650
+ },
2651
+ {
2652
+ "epoch": 0.57,
2653
+ "learning_rate": 5e-05,
2654
+ "loss": 0.3428,
2655
+ "step": 220500
2656
+ },
2657
+ {
2658
+ "epoch": 0.58,
2659
+ "learning_rate": 5e-05,
2660
+ "loss": 0.3426,
2661
+ "step": 221000
2662
+ },
2663
+ {
2664
+ "epoch": 0.58,
2665
+ "learning_rate": 5e-05,
2666
+ "loss": 0.3416,
2667
+ "step": 221500
2668
+ },
2669
+ {
2670
+ "epoch": 0.58,
2671
+ "learning_rate": 5e-05,
2672
+ "loss": 0.3422,
2673
+ "step": 222000
2674
+ },
2675
+ {
2676
+ "epoch": 0.58,
2677
+ "learning_rate": 5e-05,
2678
+ "loss": 0.3413,
2679
+ "step": 222500
2680
+ },
2681
+ {
2682
+ "epoch": 0.58,
2683
+ "learning_rate": 5e-05,
2684
+ "loss": 0.3406,
2685
+ "step": 223000
2686
+ },
2687
+ {
2688
+ "epoch": 0.58,
2689
+ "learning_rate": 5e-05,
2690
+ "loss": 0.3418,
2691
+ "step": 223500
2692
+ },
2693
+ {
2694
+ "epoch": 0.58,
2695
+ "learning_rate": 5e-05,
2696
+ "loss": 0.3415,
2697
+ "step": 224000
2698
+ },
2699
+ {
2700
+ "epoch": 0.58,
2701
+ "learning_rate": 5e-05,
2702
+ "loss": 0.3418,
2703
+ "step": 224500
2704
+ },
2705
+ {
2706
+ "epoch": 0.59,
2707
+ "learning_rate": 5e-05,
2708
+ "loss": 0.3415,
2709
+ "step": 225000
2710
+ },
2711
+ {
2712
+ "epoch": 0.59,
2713
+ "learning_rate": 5e-05,
2714
+ "loss": 0.3422,
2715
+ "step": 225500
2716
+ },
2717
+ {
2718
+ "epoch": 0.59,
2719
+ "learning_rate": 5e-05,
2720
+ "loss": 0.3414,
2721
+ "step": 226000
2722
+ },
2723
+ {
2724
+ "epoch": 0.59,
2725
+ "learning_rate": 5e-05,
2726
+ "loss": 0.3415,
2727
+ "step": 226500
2728
+ },
2729
+ {
2730
+ "epoch": 0.59,
2731
+ "learning_rate": 5e-05,
2732
+ "loss": 0.3425,
2733
+ "step": 227000
2734
+ },
2735
+ {
2736
+ "epoch": 0.59,
2737
+ "learning_rate": 5e-05,
2738
+ "loss": 0.3426,
2739
+ "step": 227500
2740
+ },
2741
+ {
2742
+ "epoch": 0.59,
2743
+ "learning_rate": 5e-05,
2744
+ "loss": 0.3431,
2745
+ "step": 228000
2746
+ },
2747
+ {
2748
+ "epoch": 0.6,
2749
+ "learning_rate": 5e-05,
2750
+ "loss": 0.3411,
2751
+ "step": 228500
2752
+ },
2753
+ {
2754
+ "epoch": 0.6,
2755
+ "learning_rate": 5e-05,
2756
+ "loss": 0.3416,
2757
+ "step": 229000
2758
+ },
2759
+ {
2760
+ "epoch": 0.6,
2761
+ "learning_rate": 5e-05,
2762
+ "loss": 0.3422,
2763
+ "step": 229500
2764
+ },
2765
+ {
2766
+ "epoch": 0.6,
2767
+ "learning_rate": 5e-05,
2768
+ "loss": 0.3414,
2769
+ "step": 230000
2770
+ },
2771
+ {
2772
+ "epoch": 0.6,
2773
+ "learning_rate": 5e-05,
2774
+ "loss": 0.3426,
2775
+ "step": 230500
2776
+ },
2777
+ {
2778
+ "epoch": 0.6,
2779
+ "learning_rate": 5e-05,
2780
+ "loss": 0.3412,
2781
+ "step": 231000
2782
+ },
2783
+ {
2784
+ "epoch": 0.6,
2785
+ "learning_rate": 5e-05,
2786
+ "loss": 0.3414,
2787
+ "step": 231500
2788
+ },
2789
+ {
2790
+ "epoch": 0.6,
2791
+ "learning_rate": 5e-05,
2792
+ "loss": 0.3402,
2793
+ "step": 232000
2794
+ },
2795
+ {
2796
+ "epoch": 0.61,
2797
+ "learning_rate": 5e-05,
2798
+ "loss": 0.3412,
2799
+ "step": 232500
2800
+ },
2801
+ {
2802
+ "epoch": 0.61,
2803
+ "learning_rate": 5e-05,
2804
+ "loss": 0.3405,
2805
+ "step": 233000
2806
+ },
2807
+ {
2808
+ "epoch": 0.61,
2809
+ "learning_rate": 5e-05,
2810
+ "loss": 0.3409,
2811
+ "step": 233500
2812
+ },
2813
+ {
2814
+ "epoch": 0.61,
2815
+ "learning_rate": 5e-05,
2816
+ "loss": 0.3409,
2817
+ "step": 234000
2818
+ },
2819
+ {
2820
+ "epoch": 0.61,
2821
+ "learning_rate": 5e-05,
2822
+ "loss": 0.3421,
2823
+ "step": 234500
2824
+ },
2825
+ {
2826
+ "epoch": 0.61,
2827
+ "learning_rate": 5e-05,
2828
+ "loss": 0.3409,
2829
+ "step": 235000
2830
+ },
2831
+ {
2832
+ "epoch": 0.61,
2833
+ "learning_rate": 5e-05,
2834
+ "loss": 0.3413,
2835
+ "step": 235500
2836
+ },
2837
+ {
2838
+ "epoch": 0.61,
2839
+ "learning_rate": 5e-05,
2840
+ "loss": 0.3397,
2841
+ "step": 236000
2842
+ },
2843
+ {
2844
+ "epoch": 0.62,
2845
+ "learning_rate": 5e-05,
2846
+ "loss": 0.3401,
2847
+ "step": 236500
2848
+ },
2849
+ {
2850
+ "epoch": 0.62,
2851
+ "learning_rate": 5e-05,
2852
+ "loss": 0.342,
2853
+ "step": 237000
2854
+ },
2855
+ {
2856
+ "epoch": 0.62,
2857
+ "learning_rate": 5e-05,
2858
+ "loss": 0.3402,
2859
+ "step": 237500
2860
+ },
2861
+ {
2862
+ "epoch": 0.62,
2863
+ "learning_rate": 5e-05,
2864
+ "loss": 0.3399,
2865
+ "step": 238000
2866
+ },
2867
+ {
2868
+ "epoch": 0.62,
2869
+ "learning_rate": 5e-05,
2870
+ "loss": 0.3413,
2871
+ "step": 238500
2872
+ },
2873
+ {
2874
+ "epoch": 0.62,
2875
+ "learning_rate": 5e-05,
2876
+ "loss": 0.3397,
2877
+ "step": 239000
2878
+ },
2879
+ {
2880
+ "epoch": 0.62,
2881
+ "learning_rate": 5e-05,
2882
+ "loss": 0.3395,
2883
+ "step": 239500
2884
+ },
2885
+ {
2886
+ "epoch": 0.63,
2887
+ "learning_rate": 5e-05,
2888
+ "loss": 0.3411,
2889
+ "step": 240000
2890
+ },
2891
+ {
2892
+ "epoch": 0.63,
2893
+ "learning_rate": 5e-05,
2894
+ "loss": 0.3409,
2895
+ "step": 240500
2896
+ },
2897
+ {
2898
+ "epoch": 0.63,
2899
+ "learning_rate": 5e-05,
2900
+ "loss": 0.3411,
2901
+ "step": 241000
2902
+ },
2903
+ {
2904
+ "epoch": 0.63,
2905
+ "learning_rate": 5e-05,
2906
+ "loss": 0.3395,
2907
+ "step": 241500
2908
+ },
2909
+ {
2910
+ "epoch": 0.63,
2911
+ "learning_rate": 5e-05,
2912
+ "loss": 0.3418,
2913
+ "step": 242000
2914
+ },
2915
+ {
2916
+ "epoch": 0.63,
2917
+ "learning_rate": 5e-05,
2918
+ "loss": 0.3383,
2919
+ "step": 242500
2920
+ },
2921
+ {
2922
+ "epoch": 0.63,
2923
+ "learning_rate": 5e-05,
2924
+ "loss": 0.3402,
2925
+ "step": 243000
2926
+ },
2927
+ {
2928
+ "epoch": 0.63,
2929
+ "learning_rate": 5e-05,
2930
+ "loss": 0.3393,
2931
+ "step": 243500
2932
+ },
2933
+ {
2934
+ "epoch": 0.64,
2935
+ "learning_rate": 5e-05,
2936
+ "loss": 0.3407,
2937
+ "step": 244000
2938
+ },
2939
+ {
2940
+ "epoch": 0.64,
2941
+ "learning_rate": 5e-05,
2942
+ "loss": 0.3413,
2943
+ "step": 244500
2944
+ },
2945
+ {
2946
+ "epoch": 0.64,
2947
+ "learning_rate": 5e-05,
2948
+ "loss": 0.34,
2949
+ "step": 245000
2950
+ },
2951
+ {
2952
+ "epoch": 0.64,
2953
+ "learning_rate": 5e-05,
2954
+ "loss": 0.3401,
2955
+ "step": 245500
2956
+ },
2957
+ {
2958
+ "epoch": 0.64,
2959
+ "learning_rate": 5e-05,
2960
+ "loss": 0.3391,
2961
+ "step": 246000
2962
+ },
2963
+ {
2964
+ "epoch": 0.64,
2965
+ "learning_rate": 5e-05,
2966
+ "loss": 0.3395,
2967
+ "step": 246500
2968
+ },
2969
+ {
2970
+ "epoch": 0.64,
2971
+ "learning_rate": 5e-05,
2972
+ "loss": 0.3394,
2973
+ "step": 247000
2974
+ },
2975
+ {
2976
+ "epoch": 0.64,
2977
+ "learning_rate": 5e-05,
2978
+ "loss": 0.3403,
2979
+ "step": 247500
2980
+ },
2981
+ {
2982
+ "epoch": 0.65,
2983
+ "learning_rate": 5e-05,
2984
+ "loss": 0.3395,
2985
+ "step": 248000
2986
+ },
2987
+ {
2988
+ "epoch": 0.65,
2989
+ "learning_rate": 5e-05,
2990
+ "loss": 0.3403,
2991
+ "step": 248500
2992
+ },
2993
+ {
2994
+ "epoch": 0.65,
2995
+ "learning_rate": 5e-05,
2996
+ "loss": 0.3394,
2997
+ "step": 249000
2998
+ },
2999
+ {
3000
+ "epoch": 0.65,
3001
+ "learning_rate": 5e-05,
3002
+ "loss": 0.3394,
3003
+ "step": 249500
3004
+ },
3005
+ {
3006
+ "epoch": 0.65,
3007
+ "learning_rate": 5e-05,
3008
+ "loss": 0.3394,
3009
+ "step": 250000
3010
+ },
3011
+ {
3012
+ "epoch": 0.65,
3013
+ "learning_rate": 5e-05,
3014
+ "loss": 0.3388,
3015
+ "step": 250500
3016
+ },
3017
+ {
3018
+ "epoch": 0.65,
3019
+ "learning_rate": 5e-05,
3020
+ "loss": 0.3395,
3021
+ "step": 251000
3022
+ },
3023
+ {
3024
+ "epoch": 0.65,
3025
+ "learning_rate": 5e-05,
3026
+ "loss": 0.3396,
3027
+ "step": 251500
3028
+ },
3029
+ {
3030
+ "epoch": 0.66,
3031
+ "learning_rate": 5e-05,
3032
+ "loss": 0.3401,
3033
+ "step": 252000
3034
+ },
3035
+ {
3036
+ "epoch": 0.66,
3037
+ "learning_rate": 5e-05,
3038
+ "loss": 0.3374,
3039
+ "step": 252500
3040
+ },
3041
+ {
3042
+ "epoch": 0.66,
3043
+ "learning_rate": 5e-05,
3044
+ "loss": 0.3398,
3045
+ "step": 253000
3046
+ },
3047
+ {
3048
+ "epoch": 0.66,
3049
+ "learning_rate": 5e-05,
3050
+ "loss": 0.3376,
3051
+ "step": 253500
3052
+ },
3053
+ {
3054
+ "epoch": 0.66,
3055
+ "learning_rate": 5e-05,
3056
+ "loss": 0.3394,
3057
+ "step": 254000
3058
+ },
3059
+ {
3060
+ "epoch": 0.66,
3061
+ "learning_rate": 5e-05,
3062
+ "loss": 0.3393,
3063
+ "step": 254500
3064
+ },
3065
+ {
3066
+ "epoch": 0.66,
3067
+ "learning_rate": 5e-05,
3068
+ "loss": 0.3385,
3069
+ "step": 255000
3070
+ },
3071
+ {
3072
+ "epoch": 0.67,
3073
+ "learning_rate": 5e-05,
3074
+ "loss": 0.3397,
3075
+ "step": 255500
3076
+ },
3077
+ {
3078
+ "epoch": 0.67,
3079
+ "learning_rate": 5e-05,
3080
+ "loss": 0.3374,
3081
+ "step": 256000
3082
+ },
3083
+ {
3084
+ "epoch": 0.67,
3085
+ "learning_rate": 5e-05,
3086
+ "loss": 0.3374,
3087
+ "step": 256500
3088
+ },
3089
+ {
3090
+ "epoch": 0.67,
3091
+ "learning_rate": 5e-05,
3092
+ "loss": 0.3377,
3093
+ "step": 257000
3094
+ },
3095
+ {
3096
+ "epoch": 0.67,
3097
+ "learning_rate": 5e-05,
3098
+ "loss": 0.3388,
3099
+ "step": 257500
3100
+ },
3101
+ {
3102
+ "epoch": 0.67,
3103
+ "learning_rate": 5e-05,
3104
+ "loss": 0.3381,
3105
+ "step": 258000
3106
+ },
3107
+ {
3108
+ "epoch": 0.67,
3109
+ "learning_rate": 5e-05,
3110
+ "loss": 0.3386,
3111
+ "step": 258500
3112
+ },
3113
+ {
3114
+ "epoch": 0.67,
3115
+ "learning_rate": 5e-05,
3116
+ "loss": 0.3379,
3117
+ "step": 259000
3118
+ },
3119
+ {
3120
+ "epoch": 0.68,
3121
+ "learning_rate": 5e-05,
3122
+ "loss": 0.3387,
3123
+ "step": 259500
3124
+ },
3125
+ {
3126
+ "epoch": 0.68,
3127
+ "learning_rate": 5e-05,
3128
+ "loss": 0.3369,
3129
+ "step": 260000
3130
+ },
3131
+ {
3132
+ "epoch": 0.68,
3133
+ "learning_rate": 5e-05,
3134
+ "loss": 0.3388,
3135
+ "step": 260500
3136
+ },
3137
+ {
3138
+ "epoch": 0.68,
3139
+ "learning_rate": 5e-05,
3140
+ "loss": 0.3373,
3141
+ "step": 261000
3142
+ },
3143
+ {
3144
+ "epoch": 0.68,
3145
+ "learning_rate": 5e-05,
3146
+ "loss": 0.3399,
3147
+ "step": 261500
3148
+ },
3149
+ {
3150
+ "epoch": 0.68,
3151
+ "learning_rate": 5e-05,
3152
+ "loss": 0.3366,
3153
+ "step": 262000
3154
+ },
3155
+ {
3156
+ "epoch": 0.68,
3157
+ "learning_rate": 5e-05,
3158
+ "loss": 0.3377,
3159
+ "step": 262500
3160
+ },
3161
+ {
3162
+ "epoch": 0.68,
3163
+ "learning_rate": 5e-05,
3164
+ "loss": 0.3385,
3165
+ "step": 263000
3166
+ },
3167
+ {
3168
+ "epoch": 0.69,
3169
+ "learning_rate": 5e-05,
3170
+ "loss": 0.3377,
3171
+ "step": 263500
3172
+ },
3173
+ {
3174
+ "epoch": 0.69,
3175
+ "learning_rate": 5e-05,
3176
+ "loss": 0.3382,
3177
+ "step": 264000
3178
+ },
3179
+ {
3180
+ "epoch": 0.69,
3181
+ "learning_rate": 5e-05,
3182
+ "loss": 0.3369,
3183
+ "step": 264500
3184
+ },
3185
+ {
3186
+ "epoch": 0.69,
3187
+ "learning_rate": 5e-05,
3188
+ "loss": 0.3374,
3189
+ "step": 265000
3190
+ },
3191
+ {
3192
+ "epoch": 0.69,
3193
+ "learning_rate": 5e-05,
3194
+ "loss": 0.3373,
3195
+ "step": 265500
3196
+ },
3197
+ {
3198
+ "epoch": 0.69,
3199
+ "learning_rate": 5e-05,
3200
+ "loss": 0.3361,
3201
+ "step": 266000
3202
+ },
3203
+ {
3204
+ "epoch": 0.69,
3205
+ "learning_rate": 5e-05,
3206
+ "loss": 0.3362,
3207
+ "step": 266500
3208
+ },
3209
+ {
3210
+ "epoch": 0.7,
3211
+ "learning_rate": 5e-05,
3212
+ "loss": 0.3376,
3213
+ "step": 267000
3214
+ },
3215
+ {
3216
+ "epoch": 0.7,
3217
+ "learning_rate": 5e-05,
3218
+ "loss": 0.3372,
3219
+ "step": 267500
3220
+ },
3221
+ {
3222
+ "epoch": 0.7,
3223
+ "learning_rate": 5e-05,
3224
+ "loss": 0.3371,
3225
+ "step": 268000
3226
+ },
3227
+ {
3228
+ "epoch": 0.7,
3229
+ "learning_rate": 5e-05,
3230
+ "loss": 0.3359,
3231
+ "step": 268500
3232
+ },
3233
+ {
3234
+ "epoch": 0.7,
3235
+ "learning_rate": 5e-05,
3236
+ "loss": 0.3354,
3237
+ "step": 269000
3238
+ },
3239
+ {
3240
+ "epoch": 0.7,
3241
+ "learning_rate": 5e-05,
3242
+ "loss": 0.3367,
3243
+ "step": 269500
3244
+ },
3245
+ {
3246
+ "epoch": 0.7,
3247
+ "learning_rate": 5e-05,
3248
+ "loss": 0.337,
3249
+ "step": 270000
3250
+ },
3251
+ {
3252
+ "epoch": 0.7,
3253
+ "learning_rate": 5e-05,
3254
+ "loss": 0.3365,
3255
+ "step": 270500
3256
+ },
3257
+ {
3258
+ "epoch": 0.71,
3259
+ "learning_rate": 5e-05,
3260
+ "loss": 0.3374,
3261
+ "step": 271000
3262
+ },
3263
+ {
3264
+ "epoch": 0.71,
3265
+ "learning_rate": 5e-05,
3266
+ "loss": 0.3365,
3267
+ "step": 271500
3268
+ },
3269
+ {
3270
+ "epoch": 0.71,
3271
+ "learning_rate": 5e-05,
3272
+ "loss": 0.3359,
3273
+ "step": 272000
3274
+ },
3275
+ {
3276
+ "epoch": 0.71,
3277
+ "learning_rate": 5e-05,
3278
+ "loss": 0.3363,
3279
+ "step": 272500
3280
+ },
3281
+ {
3282
+ "epoch": 0.71,
3283
+ "learning_rate": 5e-05,
3284
+ "loss": 0.3352,
3285
+ "step": 273000
3286
+ },
3287
+ {
3288
+ "epoch": 0.71,
3289
+ "learning_rate": 5e-05,
3290
+ "loss": 0.3369,
3291
+ "step": 273500
3292
+ },
3293
+ {
3294
+ "epoch": 0.71,
3295
+ "learning_rate": 5e-05,
3296
+ "loss": 0.3354,
3297
+ "step": 274000
3298
+ },
3299
+ {
3300
+ "epoch": 0.71,
3301
+ "learning_rate": 5e-05,
3302
+ "loss": 0.3373,
3303
+ "step": 274500
3304
+ },
3305
+ {
3306
+ "epoch": 0.72,
3307
+ "learning_rate": 5e-05,
3308
+ "loss": 0.3366,
3309
+ "step": 275000
3310
+ },
3311
+ {
3312
+ "epoch": 0.72,
3313
+ "learning_rate": 5e-05,
3314
+ "loss": 0.3365,
3315
+ "step": 275500
3316
+ },
3317
+ {
3318
+ "epoch": 0.72,
3319
+ "learning_rate": 5e-05,
3320
+ "loss": 0.3373,
3321
+ "step": 276000
3322
+ },
3323
+ {
3324
+ "epoch": 0.72,
3325
+ "learning_rate": 5e-05,
3326
+ "loss": 0.3368,
3327
+ "step": 276500
3328
+ },
3329
+ {
3330
+ "epoch": 0.72,
3331
+ "learning_rate": 5e-05,
3332
+ "loss": 0.3375,
3333
+ "step": 277000
3334
+ },
3335
+ {
3336
+ "epoch": 0.72,
3337
+ "learning_rate": 5e-05,
3338
+ "loss": 0.3356,
3339
+ "step": 277500
3340
+ },
3341
+ {
3342
+ "epoch": 0.72,
3343
+ "learning_rate": 5e-05,
3344
+ "loss": 0.3356,
3345
+ "step": 278000
3346
+ },
3347
+ {
3348
+ "epoch": 0.73,
3349
+ "learning_rate": 5e-05,
3350
+ "loss": 0.3373,
3351
+ "step": 278500
3352
+ },
3353
+ {
3354
+ "epoch": 0.73,
3355
+ "learning_rate": 5e-05,
3356
+ "loss": 0.3362,
3357
+ "step": 279000
3358
+ },
3359
+ {
3360
+ "epoch": 0.73,
3361
+ "learning_rate": 5e-05,
3362
+ "loss": 0.3361,
3363
+ "step": 279500
3364
+ },
3365
+ {
3366
+ "epoch": 0.73,
3367
+ "learning_rate": 5e-05,
3368
+ "loss": 0.3357,
3369
+ "step": 280000
3370
+ },
3371
+ {
3372
+ "epoch": 0.73,
3373
+ "learning_rate": 5e-05,
3374
+ "loss": 0.3367,
3375
+ "step": 280500
3376
+ },
3377
+ {
3378
+ "epoch": 0.73,
3379
+ "learning_rate": 5e-05,
3380
+ "loss": 0.3371,
3381
+ "step": 281000
3382
+ },
3383
+ {
3384
+ "epoch": 0.73,
3385
+ "learning_rate": 5e-05,
3386
+ "loss": 0.3358,
3387
+ "step": 281500
3388
+ },
3389
+ {
3390
+ "epoch": 0.73,
3391
+ "learning_rate": 5e-05,
3392
+ "loss": 0.3354,
3393
+ "step": 282000
3394
+ },
3395
+ {
3396
+ "epoch": 0.74,
3397
+ "learning_rate": 5e-05,
3398
+ "loss": 0.3363,
3399
+ "step": 282500
3400
+ },
3401
+ {
3402
+ "epoch": 0.74,
3403
+ "learning_rate": 5e-05,
3404
+ "loss": 0.3366,
3405
+ "step": 283000
3406
+ },
3407
+ {
3408
+ "epoch": 0.74,
3409
+ "learning_rate": 5e-05,
3410
+ "loss": 0.3353,
3411
+ "step": 283500
3412
+ },
3413
+ {
3414
+ "epoch": 0.74,
3415
+ "learning_rate": 5e-05,
3416
+ "loss": 0.3367,
3417
+ "step": 284000
3418
+ },
3419
+ {
3420
+ "epoch": 0.74,
3421
+ "learning_rate": 5e-05,
3422
+ "loss": 0.3357,
3423
+ "step": 284500
3424
+ },
3425
+ {
3426
+ "epoch": 0.74,
3427
+ "learning_rate": 5e-05,
3428
+ "loss": 0.3364,
3429
+ "step": 285000
3430
+ },
3431
+ {
3432
+ "epoch": 0.74,
3433
+ "learning_rate": 5e-05,
3434
+ "loss": 0.3338,
3435
+ "step": 285500
3436
+ },
3437
+ {
3438
+ "epoch": 0.74,
3439
+ "learning_rate": 5e-05,
3440
+ "loss": 0.3346,
3441
+ "step": 286000
3442
+ },
3443
+ {
3444
+ "epoch": 0.75,
3445
+ "learning_rate": 5e-05,
3446
+ "loss": 0.3341,
3447
+ "step": 286500
3448
+ },
3449
+ {
3450
+ "epoch": 0.75,
3451
+ "learning_rate": 5e-05,
3452
+ "loss": 0.3368,
3453
+ "step": 287000
3454
+ },
3455
+ {
3456
+ "epoch": 0.75,
3457
+ "learning_rate": 5e-05,
3458
+ "loss": 0.3357,
3459
+ "step": 287500
3460
+ },
3461
+ {
3462
+ "epoch": 0.75,
3463
+ "learning_rate": 5e-05,
3464
+ "loss": 0.3345,
3465
+ "step": 288000
3466
+ },
3467
+ {
3468
+ "epoch": 0.75,
3469
+ "learning_rate": 5e-05,
3470
+ "loss": 0.335,
3471
+ "step": 288500
3472
+ },
3473
+ {
3474
+ "epoch": 0.75,
3475
+ "learning_rate": 5e-05,
3476
+ "loss": 0.3352,
3477
+ "step": 289000
3478
+ },
3479
+ {
3480
+ "epoch": 0.75,
3481
+ "learning_rate": 5e-05,
3482
+ "loss": 0.3338,
3483
+ "step": 289500
3484
+ },
3485
+ {
3486
+ "epoch": 0.76,
3487
+ "learning_rate": 5e-05,
3488
+ "loss": 0.3356,
3489
+ "step": 290000
3490
+ },
3491
+ {
3492
+ "epoch": 0.76,
3493
+ "learning_rate": 5e-05,
3494
+ "loss": 0.3355,
3495
+ "step": 290500
3496
+ },
3497
+ {
3498
+ "epoch": 0.76,
3499
+ "learning_rate": 5e-05,
3500
+ "loss": 0.3354,
3501
+ "step": 291000
3502
+ },
3503
+ {
3504
+ "epoch": 0.76,
3505
+ "learning_rate": 5e-05,
3506
+ "loss": 0.3358,
3507
+ "step": 291500
3508
+ },
3509
+ {
3510
+ "epoch": 0.76,
3511
+ "learning_rate": 5e-05,
3512
+ "loss": 0.3343,
3513
+ "step": 292000
3514
+ },
3515
+ {
3516
+ "epoch": 0.76,
3517
+ "learning_rate": 5e-05,
3518
+ "loss": 0.335,
3519
+ "step": 292500
3520
+ },
3521
+ {
3522
+ "epoch": 0.76,
3523
+ "learning_rate": 5e-05,
3524
+ "loss": 0.335,
3525
+ "step": 293000
3526
+ },
3527
+ {
3528
+ "epoch": 0.76,
3529
+ "learning_rate": 5e-05,
3530
+ "loss": 0.3365,
3531
+ "step": 293500
3532
+ },
3533
+ {
3534
+ "epoch": 0.77,
3535
+ "learning_rate": 5e-05,
3536
+ "loss": 0.3355,
3537
+ "step": 294000
3538
+ },
3539
+ {
3540
+ "epoch": 0.77,
3541
+ "learning_rate": 5e-05,
3542
+ "loss": 0.3347,
3543
+ "step": 294500
3544
+ },
3545
+ {
3546
+ "epoch": 0.77,
3547
+ "learning_rate": 5e-05,
3548
+ "loss": 0.3349,
3549
+ "step": 295000
3550
+ },
3551
+ {
3552
+ "epoch": 0.77,
3553
+ "learning_rate": 5e-05,
3554
+ "loss": 0.3344,
3555
+ "step": 295500
3556
+ },
3557
+ {
3558
+ "epoch": 0.77,
3559
+ "learning_rate": 5e-05,
3560
+ "loss": 0.3345,
3561
+ "step": 296000
3562
+ },
3563
+ {
3564
+ "epoch": 0.77,
3565
+ "learning_rate": 5e-05,
3566
+ "loss": 0.3341,
3567
+ "step": 296500
3568
+ },
3569
+ {
3570
+ "epoch": 0.77,
3571
+ "learning_rate": 5e-05,
3572
+ "loss": 0.3353,
3573
+ "step": 297000
3574
+ },
3575
+ {
3576
+ "epoch": 0.77,
3577
+ "learning_rate": 5e-05,
3578
+ "loss": 0.3358,
3579
+ "step": 297500
3580
+ },
3581
+ {
3582
+ "epoch": 0.78,
3583
+ "learning_rate": 5e-05,
3584
+ "loss": 0.3353,
3585
+ "step": 298000
3586
+ },
3587
+ {
3588
+ "epoch": 0.78,
3589
+ "learning_rate": 5e-05,
3590
+ "loss": 0.3337,
3591
+ "step": 298500
3592
+ },
3593
+ {
3594
+ "epoch": 0.78,
3595
+ "learning_rate": 5e-05,
3596
+ "loss": 0.3341,
3597
+ "step": 299000
3598
+ },
3599
+ {
3600
+ "epoch": 0.78,
3601
+ "learning_rate": 5e-05,
3602
+ "loss": 0.3351,
3603
+ "step": 299500
3604
+ },
3605
+ {
3606
+ "epoch": 0.78,
3607
+ "learning_rate": 5e-05,
3608
+ "loss": 0.3347,
3609
+ "step": 300000
3610
+ },
3611
+ {
3612
+ "epoch": 0.78,
3613
+ "learning_rate": 5e-05,
3614
+ "loss": 0.3334,
3615
+ "step": 300500
3616
+ },
3617
+ {
3618
+ "epoch": 0.78,
3619
+ "learning_rate": 5e-05,
3620
+ "loss": 0.3348,
3621
+ "step": 301000
3622
+ },
3623
+ {
3624
+ "epoch": 0.79,
3625
+ "learning_rate": 5e-05,
3626
+ "loss": 0.3352,
3627
+ "step": 301500
3628
+ },
3629
+ {
3630
+ "epoch": 0.79,
3631
+ "learning_rate": 5e-05,
3632
+ "loss": 0.3343,
3633
+ "step": 302000
3634
+ },
3635
+ {
3636
+ "epoch": 0.79,
3637
+ "learning_rate": 5e-05,
3638
+ "loss": 0.334,
3639
+ "step": 302500
3640
+ },
3641
+ {
3642
+ "epoch": 0.79,
3643
+ "learning_rate": 5e-05,
3644
+ "loss": 0.335,
3645
+ "step": 303000
3646
+ },
3647
+ {
3648
+ "epoch": 0.79,
3649
+ "learning_rate": 5e-05,
3650
+ "loss": 0.3337,
3651
+ "step": 303500
3652
+ },
3653
+ {
3654
+ "epoch": 0.79,
3655
+ "learning_rate": 5e-05,
3656
+ "loss": 0.3338,
3657
+ "step": 304000
3658
+ },
3659
+ {
3660
+ "epoch": 0.79,
3661
+ "learning_rate": 5e-05,
3662
+ "loss": 0.3354,
3663
+ "step": 304500
3664
+ },
3665
+ {
3666
+ "epoch": 0.79,
3667
+ "learning_rate": 5e-05,
3668
+ "loss": 0.3339,
3669
+ "step": 305000
3670
+ },
3671
+ {
3672
+ "epoch": 0.8,
3673
+ "learning_rate": 5e-05,
3674
+ "loss": 0.3332,
3675
+ "step": 305500
3676
+ },
3677
+ {
3678
+ "epoch": 0.8,
3679
+ "learning_rate": 5e-05,
3680
+ "loss": 0.3336,
3681
+ "step": 306000
3682
+ },
3683
+ {
3684
+ "epoch": 0.8,
3685
+ "learning_rate": 5e-05,
3686
+ "loss": 0.3344,
3687
+ "step": 306500
3688
+ },
3689
+ {
3690
+ "epoch": 0.8,
3691
+ "learning_rate": 5e-05,
3692
+ "loss": 0.3333,
3693
+ "step": 307000
3694
+ },
3695
+ {
3696
+ "epoch": 0.8,
3697
+ "learning_rate": 5e-05,
3698
+ "loss": 0.3347,
3699
+ "step": 307500
3700
+ },
3701
+ {
3702
+ "epoch": 0.8,
3703
+ "learning_rate": 5e-05,
3704
+ "loss": 0.3337,
3705
+ "step": 308000
3706
+ },
3707
+ {
3708
+ "epoch": 0.8,
3709
+ "learning_rate": 5e-05,
3710
+ "loss": 0.3327,
3711
+ "step": 308500
3712
+ },
3713
+ {
3714
+ "epoch": 0.8,
3715
+ "learning_rate": 5e-05,
3716
+ "loss": 0.3335,
3717
+ "step": 309000
3718
+ },
3719
+ {
3720
+ "epoch": 0.81,
3721
+ "learning_rate": 5e-05,
3722
+ "loss": 0.333,
3723
+ "step": 309500
3724
+ },
3725
+ {
3726
+ "epoch": 0.81,
3727
+ "learning_rate": 5e-05,
3728
+ "loss": 0.3341,
3729
+ "step": 310000
3730
+ },
3731
+ {
3732
+ "epoch": 0.81,
3733
+ "learning_rate": 5e-05,
3734
+ "loss": 0.334,
3735
+ "step": 310500
3736
+ },
3737
+ {
3738
+ "epoch": 0.81,
3739
+ "learning_rate": 5e-05,
3740
+ "loss": 0.3336,
3741
+ "step": 311000
3742
+ },
3743
+ {
3744
+ "epoch": 0.81,
3745
+ "learning_rate": 5e-05,
3746
+ "loss": 0.3328,
3747
+ "step": 311500
3748
+ },
3749
+ {
3750
+ "epoch": 0.81,
3751
+ "learning_rate": 5e-05,
3752
+ "loss": 0.3341,
3753
+ "step": 312000
3754
+ },
3755
+ {
3756
+ "epoch": 0.81,
3757
+ "learning_rate": 5e-05,
3758
+ "loss": 0.3333,
3759
+ "step": 312500
3760
+ },
3761
+ {
3762
+ "epoch": 0.82,
3763
+ "learning_rate": 5e-05,
3764
+ "loss": 0.3335,
3765
+ "step": 313000
3766
+ },
3767
+ {
3768
+ "epoch": 0.82,
3769
+ "learning_rate": 5e-05,
3770
+ "loss": 0.3312,
3771
+ "step": 313500
3772
+ },
3773
+ {
3774
+ "epoch": 0.82,
3775
+ "learning_rate": 5e-05,
3776
+ "loss": 0.3332,
3777
+ "step": 314000
3778
+ },
3779
+ {
3780
+ "epoch": 0.82,
3781
+ "learning_rate": 5e-05,
3782
+ "loss": 0.3334,
3783
+ "step": 314500
3784
+ },
3785
+ {
3786
+ "epoch": 0.82,
3787
+ "learning_rate": 5e-05,
3788
+ "loss": 0.3322,
3789
+ "step": 315000
3790
+ },
3791
+ {
3792
+ "epoch": 0.82,
3793
+ "learning_rate": 5e-05,
3794
+ "loss": 0.3324,
3795
+ "step": 315500
3796
+ },
3797
+ {
3798
+ "epoch": 0.82,
3799
+ "learning_rate": 5e-05,
3800
+ "loss": 0.3327,
3801
+ "step": 316000
3802
+ },
3803
+ {
3804
+ "epoch": 0.82,
3805
+ "learning_rate": 5e-05,
3806
+ "loss": 0.3331,
3807
+ "step": 316500
3808
+ },
3809
+ {
3810
+ "epoch": 0.83,
3811
+ "learning_rate": 5e-05,
3812
+ "loss": 0.3331,
3813
+ "step": 317000
3814
+ },
3815
+ {
3816
+ "epoch": 0.83,
3817
+ "learning_rate": 5e-05,
3818
+ "loss": 0.3319,
3819
+ "step": 317500
3820
+ },
3821
+ {
3822
+ "epoch": 0.83,
3823
+ "learning_rate": 5e-05,
3824
+ "loss": 0.3322,
3825
+ "step": 318000
3826
+ },
3827
+ {
3828
+ "epoch": 0.83,
3829
+ "learning_rate": 5e-05,
3830
+ "loss": 0.3329,
3831
+ "step": 318500
3832
+ },
3833
+ {
3834
+ "epoch": 0.83,
3835
+ "learning_rate": 5e-05,
3836
+ "loss": 0.3335,
3837
+ "step": 319000
3838
+ },
3839
+ {
3840
+ "epoch": 0.83,
3841
+ "learning_rate": 5e-05,
3842
+ "loss": 0.332,
3843
+ "step": 319500
3844
+ },
3845
+ {
3846
+ "epoch": 0.83,
3847
+ "learning_rate": 5e-05,
3848
+ "loss": 0.3322,
3849
+ "step": 320000
3850
+ },
3851
+ {
3852
+ "epoch": 0.83,
3853
+ "learning_rate": 5e-05,
3854
+ "loss": 0.3327,
3855
+ "step": 320500
3856
+ },
3857
+ {
3858
+ "epoch": 0.84,
3859
+ "learning_rate": 5e-05,
3860
+ "loss": 0.331,
3861
+ "step": 321000
3862
+ },
3863
+ {
3864
+ "epoch": 0.84,
3865
+ "learning_rate": 5e-05,
3866
+ "loss": 0.333,
3867
+ "step": 321500
3868
+ },
3869
+ {
3870
+ "epoch": 0.84,
3871
+ "learning_rate": 5e-05,
3872
+ "loss": 0.3342,
3873
+ "step": 322000
3874
+ },
3875
+ {
3876
+ "epoch": 0.84,
3877
+ "learning_rate": 5e-05,
3878
+ "loss": 0.3321,
3879
+ "step": 322500
3880
+ },
3881
+ {
3882
+ "epoch": 0.84,
3883
+ "learning_rate": 5e-05,
3884
+ "loss": 0.3324,
3885
+ "step": 323000
3886
+ },
3887
+ {
3888
+ "epoch": 0.84,
3889
+ "learning_rate": 5e-05,
3890
+ "loss": 0.3318,
3891
+ "step": 323500
3892
+ },
3893
+ {
3894
+ "epoch": 0.84,
3895
+ "learning_rate": 5e-05,
3896
+ "loss": 0.3324,
3897
+ "step": 324000
3898
+ },
3899
+ {
3900
+ "epoch": 0.85,
3901
+ "learning_rate": 5e-05,
3902
+ "loss": 0.3324,
3903
+ "step": 324500
3904
+ },
3905
+ {
3906
+ "epoch": 0.85,
3907
+ "learning_rate": 5e-05,
3908
+ "loss": 0.3318,
3909
+ "step": 325000
3910
+ },
3911
+ {
3912
+ "epoch": 0.85,
3913
+ "learning_rate": 5e-05,
3914
+ "loss": 0.332,
3915
+ "step": 325500
3916
+ },
3917
+ {
3918
+ "epoch": 0.85,
3919
+ "learning_rate": 5e-05,
3920
+ "loss": 0.3319,
3921
+ "step": 326000
3922
+ },
3923
+ {
3924
+ "epoch": 0.85,
3925
+ "learning_rate": 5e-05,
3926
+ "loss": 0.332,
3927
+ "step": 326500
3928
+ },
3929
+ {
3930
+ "epoch": 0.85,
3931
+ "learning_rate": 5e-05,
3932
+ "loss": 0.3325,
3933
+ "step": 327000
3934
+ },
3935
+ {
3936
+ "epoch": 0.85,
3937
+ "learning_rate": 5e-05,
3938
+ "loss": 0.331,
3939
+ "step": 327500
3940
+ },
3941
+ {
3942
+ "epoch": 0.85,
3943
+ "learning_rate": 5e-05,
3944
+ "loss": 0.3314,
3945
+ "step": 328000
3946
+ },
3947
+ {
3948
+ "epoch": 0.86,
3949
+ "learning_rate": 5e-05,
3950
+ "loss": 0.3323,
3951
+ "step": 328500
3952
+ },
3953
+ {
3954
+ "epoch": 0.86,
3955
+ "learning_rate": 5e-05,
3956
+ "loss": 0.3316,
3957
+ "step": 329000
3958
+ },
3959
+ {
3960
+ "epoch": 0.86,
3961
+ "learning_rate": 5e-05,
3962
+ "loss": 0.3331,
3963
+ "step": 329500
3964
+ },
3965
+ {
3966
+ "epoch": 0.86,
3967
+ "learning_rate": 5e-05,
3968
+ "loss": 0.3308,
3969
+ "step": 330000
3970
+ },
3971
+ {
3972
+ "epoch": 0.86,
3973
+ "learning_rate": 5e-05,
3974
+ "loss": 0.3309,
3975
+ "step": 330500
3976
+ },
3977
+ {
3978
+ "epoch": 0.86,
3979
+ "learning_rate": 5e-05,
3980
+ "loss": 0.3312,
3981
+ "step": 331000
3982
+ },
3983
+ {
3984
+ "epoch": 0.86,
3985
+ "learning_rate": 5e-05,
3986
+ "loss": 0.3321,
3987
+ "step": 331500
3988
+ },
3989
+ {
3990
+ "epoch": 0.86,
3991
+ "learning_rate": 5e-05,
3992
+ "loss": 0.3318,
3993
+ "step": 332000
3994
+ },
3995
+ {
3996
+ "epoch": 0.87,
3997
+ "learning_rate": 5e-05,
3998
+ "loss": 0.331,
3999
+ "step": 332500
4000
+ },
4001
+ {
4002
+ "epoch": 0.87,
4003
+ "learning_rate": 5e-05,
4004
+ "loss": 0.3301,
4005
+ "step": 333000
4006
+ },
4007
+ {
4008
+ "epoch": 0.87,
4009
+ "learning_rate": 5e-05,
4010
+ "loss": 0.3315,
4011
+ "step": 333500
4012
+ },
4013
+ {
4014
+ "epoch": 0.87,
4015
+ "learning_rate": 5e-05,
4016
+ "loss": 0.3314,
4017
+ "step": 334000
4018
+ },
4019
+ {
4020
+ "epoch": 0.87,
4021
+ "learning_rate": 5e-05,
4022
+ "loss": 0.3319,
4023
+ "step": 334500
4024
+ },
4025
+ {
4026
+ "epoch": 0.87,
4027
+ "learning_rate": 5e-05,
4028
+ "loss": 0.3315,
4029
+ "step": 335000
4030
+ },
4031
+ {
4032
+ "epoch": 0.87,
4033
+ "learning_rate": 5e-05,
4034
+ "loss": 0.3316,
4035
+ "step": 335500
4036
+ },
4037
+ {
4038
+ "epoch": 0.88,
4039
+ "learning_rate": 5e-05,
4040
+ "loss": 0.3311,
4041
+ "step": 336000
4042
+ },
4043
+ {
4044
+ "epoch": 0.88,
4045
+ "learning_rate": 5e-05,
4046
+ "loss": 0.3302,
4047
+ "step": 336500
4048
+ },
4049
+ {
4050
+ "epoch": 0.88,
4051
+ "learning_rate": 5e-05,
4052
+ "loss": 0.3306,
4053
+ "step": 337000
4054
+ },
4055
+ {
4056
+ "epoch": 0.88,
4057
+ "learning_rate": 5e-05,
4058
+ "loss": 0.3297,
4059
+ "step": 337500
4060
+ },
4061
+ {
4062
+ "epoch": 0.88,
4063
+ "learning_rate": 5e-05,
4064
+ "loss": 0.3319,
4065
+ "step": 338000
4066
+ },
4067
+ {
4068
+ "epoch": 0.88,
4069
+ "learning_rate": 5e-05,
4070
+ "loss": 0.3297,
4071
+ "step": 338500
4072
+ },
4073
+ {
4074
+ "epoch": 0.88,
4075
+ "learning_rate": 5e-05,
4076
+ "loss": 0.3299,
4077
+ "step": 339000
4078
+ },
4079
+ {
4080
+ "epoch": 0.88,
4081
+ "learning_rate": 5e-05,
4082
+ "loss": 0.3317,
4083
+ "step": 339500
4084
+ },
4085
+ {
4086
+ "epoch": 0.89,
4087
+ "learning_rate": 5e-05,
4088
+ "loss": 0.3299,
4089
+ "step": 340000
4090
+ },
4091
+ {
4092
+ "epoch": 0.89,
4093
+ "learning_rate": 5e-05,
4094
+ "loss": 0.3309,
4095
+ "step": 340500
4096
+ },
4097
+ {
4098
+ "epoch": 0.89,
4099
+ "learning_rate": 5e-05,
4100
+ "loss": 0.332,
4101
+ "step": 341000
4102
+ },
4103
+ {
4104
+ "epoch": 0.89,
4105
+ "learning_rate": 5e-05,
4106
+ "loss": 0.331,
4107
+ "step": 341500
4108
+ },
4109
+ {
4110
+ "epoch": 0.89,
4111
+ "learning_rate": 5e-05,
4112
+ "loss": 0.3313,
4113
+ "step": 342000
4114
+ },
4115
+ {
4116
+ "epoch": 0.89,
4117
+ "learning_rate": 5e-05,
4118
+ "loss": 0.33,
4119
+ "step": 342500
4120
+ },
4121
+ {
4122
+ "epoch": 0.89,
4123
+ "learning_rate": 5e-05,
4124
+ "loss": 0.3313,
4125
+ "step": 343000
4126
+ },
4127
+ {
4128
+ "epoch": 0.89,
4129
+ "learning_rate": 5e-05,
4130
+ "loss": 0.3299,
4131
+ "step": 343500
4132
+ },
4133
+ {
4134
+ "epoch": 0.9,
4135
+ "learning_rate": 5e-05,
4136
+ "loss": 0.3322,
4137
+ "step": 344000
4138
+ },
4139
+ {
4140
+ "epoch": 0.9,
4141
+ "learning_rate": 5e-05,
4142
+ "loss": 0.331,
4143
+ "step": 344500
4144
+ },
4145
+ {
4146
+ "epoch": 0.9,
4147
+ "learning_rate": 5e-05,
4148
+ "loss": 0.3304,
4149
+ "step": 345000
4150
+ },
4151
+ {
4152
+ "epoch": 0.9,
4153
+ "learning_rate": 5e-05,
4154
+ "loss": 0.3312,
4155
+ "step": 345500
4156
+ },
4157
+ {
4158
+ "epoch": 0.9,
4159
+ "learning_rate": 5e-05,
4160
+ "loss": 0.3317,
4161
+ "step": 346000
4162
+ },
4163
+ {
4164
+ "epoch": 0.9,
4165
+ "learning_rate": 5e-05,
4166
+ "loss": 0.3306,
4167
+ "step": 346500
4168
+ },
4169
+ {
4170
+ "epoch": 0.9,
4171
+ "learning_rate": 5e-05,
4172
+ "loss": 0.3288,
4173
+ "step": 347000
4174
+ },
4175
+ {
4176
+ "epoch": 0.91,
4177
+ "learning_rate": 5e-05,
4178
+ "loss": 0.3304,
4179
+ "step": 347500
4180
+ },
4181
+ {
4182
+ "epoch": 0.91,
4183
+ "learning_rate": 5e-05,
4184
+ "loss": 0.3307,
4185
+ "step": 348000
4186
+ },
4187
+ {
4188
+ "epoch": 0.91,
4189
+ "learning_rate": 5e-05,
4190
+ "loss": 0.3305,
4191
+ "step": 348500
4192
+ },
4193
+ {
4194
+ "epoch": 0.91,
4195
+ "learning_rate": 5e-05,
4196
+ "loss": 0.3294,
4197
+ "step": 349000
4198
+ },
4199
+ {
4200
+ "epoch": 0.91,
4201
+ "learning_rate": 5e-05,
4202
+ "loss": 0.3309,
4203
+ "step": 349500
4204
+ },
4205
+ {
4206
+ "epoch": 0.91,
4207
+ "learning_rate": 5e-05,
4208
+ "loss": 0.3293,
4209
+ "step": 350000
4210
+ },
4211
+ {
4212
+ "epoch": 0.91,
4213
+ "learning_rate": 5e-05,
4214
+ "loss": 0.3299,
4215
+ "step": 350500
4216
+ },
4217
+ {
4218
+ "epoch": 0.91,
4219
+ "learning_rate": 5e-05,
4220
+ "loss": 0.3303,
4221
+ "step": 351000
4222
+ },
4223
+ {
4224
+ "epoch": 0.92,
4225
+ "learning_rate": 5e-05,
4226
+ "loss": 0.3312,
4227
+ "step": 351500
4228
+ },
4229
+ {
4230
+ "epoch": 0.92,
4231
+ "learning_rate": 5e-05,
4232
+ "loss": 0.3308,
4233
+ "step": 352000
4234
+ },
4235
+ {
4236
+ "epoch": 0.92,
4237
+ "learning_rate": 5e-05,
4238
+ "loss": 0.3291,
4239
+ "step": 352500
4240
+ },
4241
+ {
4242
+ "epoch": 0.92,
4243
+ "learning_rate": 5e-05,
4244
+ "loss": 0.3289,
4245
+ "step": 353000
4246
+ },
4247
+ {
4248
+ "epoch": 0.92,
4249
+ "learning_rate": 5e-05,
4250
+ "loss": 0.3302,
4251
+ "step": 353500
4252
+ },
4253
+ {
4254
+ "epoch": 0.92,
4255
+ "learning_rate": 5e-05,
4256
+ "loss": 0.3299,
4257
+ "step": 354000
4258
+ },
4259
+ {
4260
+ "epoch": 0.92,
4261
+ "learning_rate": 5e-05,
4262
+ "loss": 0.3302,
4263
+ "step": 354500
4264
+ },
4265
+ {
4266
+ "epoch": 0.92,
4267
+ "learning_rate": 5e-05,
4268
+ "loss": 0.3309,
4269
+ "step": 355000
4270
+ },
4271
+ {
4272
+ "epoch": 0.93,
4273
+ "learning_rate": 5e-05,
4274
+ "loss": 0.3287,
4275
+ "step": 355500
4276
+ },
4277
+ {
4278
+ "epoch": 0.93,
4279
+ "learning_rate": 5e-05,
4280
+ "loss": 0.3297,
4281
+ "step": 356000
4282
+ },
4283
+ {
4284
+ "epoch": 0.93,
4285
+ "learning_rate": 5e-05,
4286
+ "loss": 0.3304,
4287
+ "step": 356500
4288
+ },
4289
+ {
4290
+ "epoch": 0.93,
4291
+ "learning_rate": 5e-05,
4292
+ "loss": 0.3295,
4293
+ "step": 357000
4294
+ },
4295
+ {
4296
+ "epoch": 0.93,
4297
+ "learning_rate": 5e-05,
4298
+ "loss": 0.3299,
4299
+ "step": 357500
4300
+ },
4301
+ {
4302
+ "epoch": 0.93,
4303
+ "learning_rate": 5e-05,
4304
+ "loss": 0.3305,
4305
+ "step": 358000
4306
+ },
4307
+ {
4308
+ "epoch": 0.93,
4309
+ "learning_rate": 5e-05,
4310
+ "loss": 0.3297,
4311
+ "step": 358500
4312
+ },
4313
+ {
4314
+ "epoch": 0.93,
4315
+ "learning_rate": 5e-05,
4316
+ "loss": 0.3298,
4317
+ "step": 359000
4318
+ },
4319
+ {
4320
+ "epoch": 0.94,
4321
+ "learning_rate": 5e-05,
4322
+ "loss": 0.3282,
4323
+ "step": 359500
4324
+ },
4325
+ {
4326
+ "epoch": 0.94,
4327
+ "learning_rate": 5e-05,
4328
+ "loss": 0.3301,
4329
+ "step": 360000
4330
+ },
4331
+ {
4332
+ "epoch": 0.94,
4333
+ "learning_rate": 5e-05,
4334
+ "loss": 0.3304,
4335
+ "step": 360500
4336
+ },
4337
+ {
4338
+ "epoch": 0.94,
4339
+ "learning_rate": 5e-05,
4340
+ "loss": 0.3281,
4341
+ "step": 361000
4342
+ },
4343
+ {
4344
+ "epoch": 0.94,
4345
+ "learning_rate": 5e-05,
4346
+ "loss": 0.329,
4347
+ "step": 361500
4348
+ },
4349
+ {
4350
+ "epoch": 0.94,
4351
+ "learning_rate": 5e-05,
4352
+ "loss": 0.3293,
4353
+ "step": 362000
4354
+ },
4355
+ {
4356
+ "epoch": 0.94,
4357
+ "learning_rate": 5e-05,
4358
+ "loss": 0.3293,
4359
+ "step": 362500
4360
+ },
4361
+ {
4362
+ "epoch": 0.95,
4363
+ "learning_rate": 5e-05,
4364
+ "loss": 0.329,
4365
+ "step": 363000
4366
+ },
4367
+ {
4368
+ "epoch": 0.95,
4369
+ "learning_rate": 5e-05,
4370
+ "loss": 0.3303,
4371
+ "step": 363500
4372
+ },
4373
+ {
4374
+ "epoch": 0.95,
4375
+ "learning_rate": 5e-05,
4376
+ "loss": 0.3299,
4377
+ "step": 364000
4378
+ },
4379
+ {
4380
+ "epoch": 0.95,
4381
+ "learning_rate": 5e-05,
4382
+ "loss": 0.329,
4383
+ "step": 364500
4384
+ },
4385
+ {
4386
+ "epoch": 0.95,
4387
+ "learning_rate": 5e-05,
4388
+ "loss": 0.3298,
4389
+ "step": 365000
4390
+ },
4391
+ {
4392
+ "epoch": 0.95,
4393
+ "learning_rate": 5e-05,
4394
+ "loss": 0.3279,
4395
+ "step": 365500
4396
+ },
4397
+ {
4398
+ "epoch": 0.95,
4399
+ "learning_rate": 5e-05,
4400
+ "loss": 0.3298,
4401
+ "step": 366000
4402
+ },
4403
+ {
4404
+ "epoch": 0.95,
4405
+ "learning_rate": 5e-05,
4406
+ "loss": 0.3291,
4407
+ "step": 366500
4408
+ },
4409
+ {
4410
+ "epoch": 0.96,
4411
+ "learning_rate": 5e-05,
4412
+ "loss": 0.3298,
4413
+ "step": 367000
4414
+ },
4415
+ {
4416
+ "epoch": 0.96,
4417
+ "learning_rate": 5e-05,
4418
+ "loss": 0.3283,
4419
+ "step": 367500
4420
+ },
4421
+ {
4422
+ "epoch": 0.96,
4423
+ "learning_rate": 5e-05,
4424
+ "loss": 0.3274,
4425
+ "step": 368000
4426
+ },
4427
+ {
4428
+ "epoch": 0.96,
4429
+ "learning_rate": 5e-05,
4430
+ "loss": 0.3287,
4431
+ "step": 368500
4432
+ },
4433
+ {
4434
+ "epoch": 0.96,
4435
+ "learning_rate": 5e-05,
4436
+ "loss": 0.3299,
4437
+ "step": 369000
4438
+ },
4439
+ {
4440
+ "epoch": 0.96,
4441
+ "learning_rate": 5e-05,
4442
+ "loss": 0.3289,
4443
+ "step": 369500
4444
+ },
4445
+ {
4446
+ "epoch": 0.96,
4447
+ "learning_rate": 5e-05,
4448
+ "loss": 0.3282,
4449
+ "step": 370000
4450
+ },
4451
+ {
4452
+ "epoch": 0.96,
4453
+ "learning_rate": 5e-05,
4454
+ "loss": 0.3293,
4455
+ "step": 370500
4456
+ },
4457
+ {
4458
+ "epoch": 0.97,
4459
+ "learning_rate": 5e-05,
4460
+ "loss": 0.3291,
4461
+ "step": 371000
4462
+ },
4463
+ {
4464
+ "epoch": 0.97,
4465
+ "learning_rate": 5e-05,
4466
+ "loss": 0.3279,
4467
+ "step": 371500
4468
+ },
4469
+ {
4470
+ "epoch": 0.97,
4471
+ "learning_rate": 5e-05,
4472
+ "loss": 0.3283,
4473
+ "step": 372000
4474
+ },
4475
+ {
4476
+ "epoch": 0.97,
4477
+ "learning_rate": 5e-05,
4478
+ "loss": 0.3304,
4479
+ "step": 372500
4480
+ },
4481
+ {
4482
+ "epoch": 0.97,
4483
+ "learning_rate": 5e-05,
4484
+ "loss": 0.3291,
4485
+ "step": 373000
4486
+ },
4487
+ {
4488
+ "epoch": 0.97,
4489
+ "learning_rate": 5e-05,
4490
+ "loss": 0.3284,
4491
+ "step": 373500
4492
+ },
4493
+ {
4494
+ "epoch": 0.97,
4495
+ "learning_rate": 5e-05,
4496
+ "loss": 0.328,
4497
+ "step": 374000
4498
+ },
4499
+ {
4500
+ "epoch": 0.98,
4501
+ "learning_rate": 5e-05,
4502
+ "loss": 0.3289,
4503
+ "step": 374500
4504
+ },
4505
+ {
4506
+ "epoch": 0.98,
4507
+ "learning_rate": 5e-05,
4508
+ "loss": 0.3281,
4509
+ "step": 375000
4510
+ },
4511
+ {
4512
+ "epoch": 0.98,
4513
+ "learning_rate": 5e-05,
4514
+ "loss": 0.3295,
4515
+ "step": 375500
4516
+ },
4517
+ {
4518
+ "epoch": 0.98,
4519
+ "learning_rate": 5e-05,
4520
+ "loss": 0.3287,
4521
+ "step": 376000
4522
+ },
4523
+ {
4524
+ "epoch": 0.98,
4525
+ "learning_rate": 5e-05,
4526
+ "loss": 0.3281,
4527
+ "step": 376500
4528
+ },
4529
+ {
4530
+ "epoch": 0.98,
4531
+ "learning_rate": 5e-05,
4532
+ "loss": 0.3265,
4533
+ "step": 377000
4534
+ },
4535
+ {
4536
+ "epoch": 0.98,
4537
+ "learning_rate": 5e-05,
4538
+ "loss": 0.3283,
4539
+ "step": 377500
4540
+ },
4541
+ {
4542
+ "epoch": 0.98,
4543
+ "learning_rate": 5e-05,
4544
+ "loss": 0.3283,
4545
+ "step": 378000
4546
+ },
4547
+ {
4548
+ "epoch": 0.99,
4549
+ "learning_rate": 5e-05,
4550
+ "loss": 0.328,
4551
+ "step": 378500
4552
+ },
4553
+ {
4554
+ "epoch": 0.99,
4555
+ "learning_rate": 5e-05,
4556
+ "loss": 0.3268,
4557
+ "step": 379000
4558
+ },
4559
+ {
4560
+ "epoch": 0.99,
4561
+ "learning_rate": 5e-05,
4562
+ "loss": 0.3273,
4563
+ "step": 379500
4564
+ },
4565
+ {
4566
+ "epoch": 0.99,
4567
+ "learning_rate": 5e-05,
4568
+ "loss": 0.3274,
4569
+ "step": 380000
4570
+ },
4571
+ {
4572
+ "epoch": 0.99,
4573
+ "learning_rate": 5e-05,
4574
+ "loss": 0.326,
4575
+ "step": 380500
4576
+ },
4577
+ {
4578
+ "epoch": 0.99,
4579
+ "learning_rate": 5e-05,
4580
+ "loss": 0.3273,
4581
+ "step": 381000
4582
+ },
4583
+ {
4584
+ "epoch": 0.99,
4585
+ "learning_rate": 5e-05,
4586
+ "loss": 0.328,
4587
+ "step": 381500
4588
+ },
4589
+ {
4590
+ "epoch": 0.99,
4591
+ "learning_rate": 5e-05,
4592
+ "loss": 0.3273,
4593
+ "step": 382000
4594
+ },
4595
+ {
4596
+ "epoch": 1.0,
4597
+ "learning_rate": 5e-05,
4598
+ "loss": 0.3282,
4599
+ "step": 382500
4600
+ },
4601
+ {
4602
+ "epoch": 1.0,
4603
+ "learning_rate": 5e-05,
4604
+ "loss": 0.3281,
4605
+ "step": 383000
4606
+ },
4607
+ {
4608
+ "epoch": 1.0,
4609
+ "learning_rate": 5e-05,
4610
+ "loss": 0.3272,
4611
+ "step": 383500
4612
+ }
4613
+ ],
4614
+ "logging_steps": 500,
4615
+ "max_steps": 383970,
4616
+ "num_train_epochs": 1,
4617
+ "save_steps": 383970,
4618
+ "total_flos": 1.3356439865463197e+20,
4619
+ "trial_name": null,
4620
+ "trial_params": null
4621
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e310557cdf58750ead68371aace208663413dea0238e31ba6694b146e6c2368
3
+ size 6200
vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
zero_to_fp32.py ADDED
@@ -0,0 +1,587 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ #!/usr/bin/env python
2
+
3
+ # Copyright (c) Microsoft Corporation.
4
+ # SPDX-License-Identifier: Apache-2.0
5
+
6
+ # DeepSpeed Team
7
+
8
+ # This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
9
+ # copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
10
+ # the future. Once extracted, the weights don't require DeepSpeed and can be used in any
11
+ # application.
12
+ #
13
+ # example: python zero_to_fp32.py . pytorch_model.bin
14
+
15
+ import argparse
16
+ import torch
17
+ import glob
18
+ import math
19
+ import os
20
+ import re
21
+ from collections import OrderedDict
22
+ from dataclasses import dataclass
23
+
24
+ # while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
25
+ # DeepSpeed data structures it has to be available in the current python environment.
26
+ from deepspeed.utils import logger
27
+ from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
28
+ FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
29
+ FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
30
+
31
+
32
+ @dataclass
33
+ class zero_model_state:
34
+ buffers: dict()
35
+ param_shapes: dict()
36
+ shared_params: list
37
+ ds_version: int
38
+ frozen_param_shapes: dict()
39
+ frozen_param_fragments: dict()
40
+
41
+
42
+ debug = 0
43
+
44
+ # load to cpu
45
+ device = torch.device('cpu')
46
+
47
+
48
+ def atoi(text):
49
+ return int(text) if text.isdigit() else text
50
+
51
+
52
+ def natural_keys(text):
53
+ '''
54
+ alist.sort(key=natural_keys) sorts in human order
55
+ http://nedbatchelder.com/blog/200712/human_sorting.html
56
+ (See Toothy's implementation in the comments)
57
+ '''
58
+ return [atoi(c) for c in re.split(r'(\d+)', text)]
59
+
60
+
61
+ def get_model_state_file(checkpoint_dir, zero_stage):
62
+ if not os.path.isdir(checkpoint_dir):
63
+ raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
64
+
65
+ # there should be only one file
66
+ if zero_stage <= 2:
67
+ file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
68
+ elif zero_stage == 3:
69
+ file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
70
+
71
+ if not os.path.exists(file):
72
+ raise FileNotFoundError(f"can't find model states file at '{file}'")
73
+
74
+ return file
75
+
76
+
77
+ def get_checkpoint_files(checkpoint_dir, glob_pattern):
78
+ # XXX: need to test that this simple glob rule works for multi-node setup too
79
+ ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
80
+
81
+ if len(ckpt_files) == 0:
82
+ raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
83
+
84
+ return ckpt_files
85
+
86
+
87
+ def get_optim_files(checkpoint_dir):
88
+ return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
89
+
90
+
91
+ def get_model_state_files(checkpoint_dir):
92
+ return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
93
+
94
+
95
+ def parse_model_states(files):
96
+ zero_model_states = []
97
+ for file in files:
98
+ state_dict = torch.load(file, map_location=device)
99
+
100
+ if BUFFER_NAMES not in state_dict:
101
+ raise ValueError(f"{file} is not a model state checkpoint")
102
+ buffer_names = state_dict[BUFFER_NAMES]
103
+ if debug:
104
+ print("Found buffers:", buffer_names)
105
+
106
+ # recover just the buffers while restoring them to fp32 if they were saved in fp16
107
+ buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
108
+ param_shapes = state_dict[PARAM_SHAPES]
109
+
110
+ # collect parameters that are included in param_shapes
111
+ param_names = []
112
+ for s in param_shapes:
113
+ for name in s.keys():
114
+ param_names.append(name)
115
+
116
+ # update with frozen parameters
117
+ frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
118
+ if frozen_param_shapes is not None:
119
+ if debug:
120
+ print(f"Found frozen_param_shapes: {frozen_param_shapes}")
121
+ param_names += list(frozen_param_shapes.keys())
122
+
123
+ # handle shared params
124
+ shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
125
+
126
+ ds_version = state_dict.get(DS_VERSION, None)
127
+
128
+ frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
129
+
130
+ z_model_state = zero_model_state(buffers=buffers,
131
+ param_shapes=param_shapes,
132
+ shared_params=shared_params,
133
+ ds_version=ds_version,
134
+ frozen_param_shapes=frozen_param_shapes,
135
+ frozen_param_fragments=frozen_param_fragments)
136
+ zero_model_states.append(z_model_state)
137
+
138
+ return zero_model_states
139
+
140
+
141
+ def parse_optim_states(files, ds_checkpoint_dir):
142
+
143
+ total_files = len(files)
144
+ state_dicts = []
145
+ for f in files:
146
+ state_dict = torch.load(f, map_location=device)
147
+ # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
148
+ # and also handle the case where it was already removed by another helper script
149
+ state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
150
+ state_dicts.append(state_dict)
151
+
152
+ if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
153
+ raise ValueError(f"{files[0]} is not a zero checkpoint")
154
+ zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
155
+ world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
156
+
157
+ # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
158
+ # parameters can be different from data parallelism for non-expert parameters. So we can just
159
+ # use the max of the partition_count to get the dp world_size.
160
+
161
+ if type(world_size) is list:
162
+ world_size = max(world_size)
163
+
164
+ if world_size != total_files:
165
+ raise ValueError(
166
+ f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
167
+ "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
168
+ )
169
+
170
+ # the groups are named differently in each stage
171
+ if zero_stage <= 2:
172
+ fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
173
+ elif zero_stage == 3:
174
+ fp32_groups_key = FP32_FLAT_GROUPS
175
+ else:
176
+ raise ValueError(f"unknown zero stage {zero_stage}")
177
+
178
+ if zero_stage <= 2:
179
+ fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
180
+ elif zero_stage == 3:
181
+ # if there is more than one param group, there will be multiple flattened tensors - one
182
+ # flattened tensor per group - for simplicity merge them into a single tensor
183
+ #
184
+ # XXX: could make the script more memory efficient for when there are multiple groups - it
185
+ # will require matching the sub-lists of param_shapes for each param group flattened tensor
186
+
187
+ fp32_flat_groups = [
188
+ torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts))
189
+ ]
190
+
191
+ return zero_stage, world_size, fp32_flat_groups
192
+
193
+
194
+ def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir):
195
+ """
196
+ Returns fp32 state_dict reconstructed from ds checkpoint
197
+
198
+ Args:
199
+ - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
200
+
201
+ """
202
+ print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
203
+
204
+ optim_files = get_optim_files(ds_checkpoint_dir)
205
+ zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
206
+ print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
207
+
208
+ model_files = get_model_state_files(ds_checkpoint_dir)
209
+
210
+ zero_model_states = parse_model_states(model_files)
211
+ print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
212
+
213
+ if zero_stage <= 2:
214
+ return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states)
215
+ elif zero_stage == 3:
216
+ return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states)
217
+
218
+
219
+ def _zero2_merge_frozen_params(state_dict, zero_model_states):
220
+ if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
221
+ return
222
+
223
+ frozen_param_shapes = zero_model_states[0].frozen_param_shapes
224
+ frozen_param_fragments = zero_model_states[0].frozen_param_fragments
225
+
226
+ if debug:
227
+ num_elem = sum(s.numel() for s in frozen_param_shapes.values())
228
+ print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
229
+
230
+ wanted_params = len(frozen_param_shapes)
231
+ wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
232
+ avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
233
+ print(f'Frozen params: Have {avail_numel} numels to process.')
234
+ print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
235
+
236
+ total_params = 0
237
+ total_numel = 0
238
+ for name, shape in frozen_param_shapes.items():
239
+ total_params += 1
240
+ unpartitioned_numel = shape.numel()
241
+ total_numel += unpartitioned_numel
242
+
243
+ state_dict[name] = frozen_param_fragments[name]
244
+
245
+ if debug:
246
+ print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
247
+
248
+ print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
249
+
250
+
251
+ def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
252
+ param_shapes = zero_model_states[0].param_shapes
253
+
254
+ # Reconstruction protocol:
255
+ #
256
+ # XXX: document this
257
+
258
+ if debug:
259
+ for i in range(world_size):
260
+ for j in range(len(fp32_flat_groups[0])):
261
+ print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
262
+
263
+ # XXX: memory usage doubles here (zero2)
264
+ num_param_groups = len(fp32_flat_groups[0])
265
+ merged_single_partition_of_fp32_groups = []
266
+ for i in range(num_param_groups):
267
+ merged_partitions = [sd[i] for sd in fp32_flat_groups]
268
+ full_single_fp32_vector = torch.cat(merged_partitions, 0)
269
+ merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
270
+ avail_numel = sum(
271
+ [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
272
+
273
+ if debug:
274
+ wanted_params = sum([len(shapes) for shapes in param_shapes])
275
+ wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
276
+ # not asserting if there is a mismatch due to possible padding
277
+ print(f"Have {avail_numel} numels to process.")
278
+ print(f"Need {wanted_numel} numels in {wanted_params} params.")
279
+
280
+ # params
281
+ # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
282
+ # out-of-core computing solution
283
+ total_numel = 0
284
+ total_params = 0
285
+ for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
286
+ offset = 0
287
+ avail_numel = full_single_fp32_vector.numel()
288
+ for name, shape in shapes.items():
289
+
290
+ unpartitioned_numel = shape.numel()
291
+ total_numel += unpartitioned_numel
292
+ total_params += 1
293
+
294
+ if debug:
295
+ print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
296
+ state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
297
+ offset += unpartitioned_numel
298
+
299
+ # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
300
+ # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
301
+ # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
302
+ # live optimizer object, so we are checking that the numbers are within the right range
303
+ align_to = 2 * world_size
304
+
305
+ def zero2_align(x):
306
+ return align_to * math.ceil(x / align_to)
307
+
308
+ if debug:
309
+ print(f"original offset={offset}, avail_numel={avail_numel}")
310
+
311
+ offset = zero2_align(offset)
312
+ avail_numel = zero2_align(avail_numel)
313
+
314
+ if debug:
315
+ print(f"aligned offset={offset}, avail_numel={avail_numel}")
316
+
317
+ # Sanity check
318
+ if offset != avail_numel:
319
+ raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
320
+
321
+ print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
322
+
323
+
324
+ def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states):
325
+ state_dict = OrderedDict()
326
+
327
+ # buffers
328
+ buffers = zero_model_states[0].buffers
329
+ state_dict.update(buffers)
330
+ if debug:
331
+ print(f"added {len(buffers)} buffers")
332
+
333
+ _zero2_merge_frozen_params(state_dict, zero_model_states)
334
+
335
+ _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
336
+
337
+ # recover shared parameters
338
+ for pair in zero_model_states[0].shared_params:
339
+ if pair[1] in state_dict:
340
+ state_dict[pair[0]] = state_dict[pair[1]]
341
+
342
+ return state_dict
343
+
344
+
345
+ def zero3_partitioned_param_info(unpartitioned_numel, world_size):
346
+ remainder = unpartitioned_numel % world_size
347
+ padding_numel = (world_size - remainder) if remainder else 0
348
+ partitioned_numel = math.ceil(unpartitioned_numel / world_size)
349
+ return partitioned_numel, padding_numel
350
+
351
+
352
+ def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
353
+ if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
354
+ return
355
+
356
+ if debug:
357
+ for i in range(world_size):
358
+ num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
359
+ print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
360
+
361
+ frozen_param_shapes = zero_model_states[0].frozen_param_shapes
362
+ wanted_params = len(frozen_param_shapes)
363
+ wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
364
+ avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
365
+ print(f'Frozen params: Have {avail_numel} numels to process.')
366
+ print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
367
+
368
+ total_params = 0
369
+ total_numel = 0
370
+ for name, shape in zero_model_states[0].frozen_param_shapes.items():
371
+ total_params += 1
372
+ unpartitioned_numel = shape.numel()
373
+ total_numel += unpartitioned_numel
374
+
375
+ param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
376
+ state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
377
+
378
+ partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
379
+
380
+ if debug:
381
+ print(
382
+ f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
383
+ )
384
+
385
+ print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
386
+
387
+
388
+ def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
389
+ param_shapes = zero_model_states[0].param_shapes
390
+ avail_numel = fp32_flat_groups[0].numel() * world_size
391
+ # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
392
+ # param, re-consolidating each param, while dealing with padding if any
393
+
394
+ # merge list of dicts, preserving order
395
+ param_shapes = {k: v for d in param_shapes for k, v in d.items()}
396
+
397
+ if debug:
398
+ for i in range(world_size):
399
+ print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
400
+
401
+ wanted_params = len(param_shapes)
402
+ wanted_numel = sum(shape.numel() for shape in param_shapes.values())
403
+ # not asserting if there is a mismatch due to possible padding
404
+ avail_numel = fp32_flat_groups[0].numel() * world_size
405
+ print(f"Trainable params: Have {avail_numel} numels to process.")
406
+ print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
407
+
408
+ # params
409
+ # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
410
+ # out-of-core computing solution
411
+ offset = 0
412
+ total_numel = 0
413
+ total_params = 0
414
+ for name, shape in param_shapes.items():
415
+
416
+ unpartitioned_numel = shape.numel()
417
+ total_numel += unpartitioned_numel
418
+ total_params += 1
419
+
420
+ partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
421
+
422
+ if debug:
423
+ print(
424
+ f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
425
+ )
426
+
427
+ # XXX: memory usage doubles here
428
+ state_dict[name] = torch.cat(
429
+ tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
430
+ 0).narrow(0, 0, unpartitioned_numel).view(shape)
431
+ offset += partitioned_numel
432
+
433
+ offset *= world_size
434
+
435
+ # Sanity check
436
+ if offset != avail_numel:
437
+ raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
438
+
439
+ print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
440
+
441
+
442
+ def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states):
443
+ state_dict = OrderedDict()
444
+
445
+ # buffers
446
+ buffers = zero_model_states[0].buffers
447
+ state_dict.update(buffers)
448
+ if debug:
449
+ print(f"added {len(buffers)} buffers")
450
+
451
+ _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
452
+
453
+ _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
454
+
455
+ # recover shared parameters
456
+ for pair in zero_model_states[0].shared_params:
457
+ if pair[1] in state_dict:
458
+ state_dict[pair[0]] = state_dict[pair[1]]
459
+
460
+ return state_dict
461
+
462
+
463
+ def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None):
464
+ """
465
+ Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
466
+ ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
467
+ via a model hub.
468
+
469
+ Args:
470
+ - ``checkpoint_dir``: path to the desired checkpoint folder
471
+ - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
472
+
473
+ Returns:
474
+ - pytorch ``state_dict``
475
+
476
+ Note: this approach may not work if your application doesn't have sufficient free CPU memory and
477
+ you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
478
+ the checkpoint.
479
+
480
+ A typical usage might be ::
481
+
482
+ from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
483
+ # do the training and checkpoint saving
484
+ state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
485
+ model = model.cpu() # move to cpu
486
+ model.load_state_dict(state_dict)
487
+ # submit to model hub or save the model to share with others
488
+
489
+ In this example the ``model`` will no longer be usable in the deepspeed context of the same
490
+ application. i.e. you will need to re-initialize the deepspeed engine, since
491
+ ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
492
+
493
+ If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
494
+
495
+ """
496
+ if tag is None:
497
+ latest_path = os.path.join(checkpoint_dir, 'latest')
498
+ if os.path.isfile(latest_path):
499
+ with open(latest_path, 'r') as fd:
500
+ tag = fd.read().strip()
501
+ else:
502
+ raise ValueError(f"Unable to find 'latest' file at {latest_path}")
503
+
504
+ ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
505
+
506
+ if not os.path.isdir(ds_checkpoint_dir):
507
+ raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
508
+
509
+ return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir)
510
+
511
+
512
+ def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None):
513
+ """
514
+ Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
515
+ loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
516
+
517
+ Args:
518
+ - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
519
+ - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin)
520
+ - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
521
+ """
522
+
523
+ state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
524
+ print(f"Saving fp32 state dict to {output_file}")
525
+ torch.save(state_dict, output_file)
526
+
527
+
528
+ def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
529
+ """
530
+ 1. Put the provided model to cpu
531
+ 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
532
+ 3. Load it into the provided model
533
+
534
+ Args:
535
+ - ``model``: the model object to update
536
+ - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
537
+ - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
538
+
539
+ Returns:
540
+ - ``model`: modified model
541
+
542
+ Make sure you have plenty of CPU memory available before you call this function. If you don't
543
+ have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
544
+ conveniently placed for you in the checkpoint folder.
545
+
546
+ A typical usage might be ::
547
+
548
+ from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
549
+ model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
550
+ # submit to model hub or save the model to share with others
551
+
552
+ Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
553
+ of the same application. i.e. you will need to re-initialize the deepspeed engine, since
554
+ ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
555
+
556
+ """
557
+ logger.info(f"Extracting fp32 weights")
558
+ state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
559
+
560
+ logger.info(f"Overwriting model with fp32 weights")
561
+ model = model.cpu()
562
+ model.load_state_dict(state_dict, strict=False)
563
+
564
+ return model
565
+
566
+
567
+ if __name__ == "__main__":
568
+
569
+ parser = argparse.ArgumentParser()
570
+ parser.add_argument("checkpoint_dir",
571
+ type=str,
572
+ help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
573
+ parser.add_argument(
574
+ "output_file",
575
+ type=str,
576
+ help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)")
577
+ parser.add_argument("-t",
578
+ "--tag",
579
+ type=str,
580
+ default=None,
581
+ help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
582
+ parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
583
+ args = parser.parse_args()
584
+
585
+ debug = args.debug
586
+
587
+ convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, args.output_file, tag=args.tag)