Hugging Face's logo

gowitheflowlab
/
clip-wiki-pretrain-10240-6000

PyTorch
clip
Model card Files Community
gowitheflow1998 commited on
Commit
73e49b9
·
1 Parent(s): 70b4ce0

upload checkpoint

Browse files
Files changed (16) hide show
  1. .gitattributes +5 -0
  2. 49e6dc219d1a1a1c9236acaf05a48b542002016a6dc877ee72baab085a84257b.3f28e7f4b38e1efe1b6da4a3732404c19d4c6a614ff32dce90a251e293d4ce58 +3 -0
  3. config.json +166 -0
  4. merges.txt +0 -0
  5. optimizer.pt +3 -0
  6. preprocessor_config.json +19 -0
  7. pytorch_model.bin +3 -0
  8. rng_state.pth +3 -0
  9. scheduler.pt +3 -0
  10. special_tokens_map.json +1 -0
  11. text_renderer_config.json +253 -0
  12. tokenizer.json +0 -0
  13. tokenizer_config.json +34 -0
  14. trainer_state.json +1816 -0
  15. training_args.bin +3 -0
  16. vocab.json +0 -0
.gitattributes CHANGED
@@ -33,3 +33,8 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ training_args.bin filter=lfs diff=lfs merge=lfs -text
37
+ optimizer.pt filter=lfs diff=lfs merge=lfs -text
38
+ scheduler.pt filter=lfs diff=lfs merge=lfs -text
39
+ 49e6dc219d1a1a1c9236acaf05a48b542002016a6dc877ee72baab085a84257b.3f28e7f4b38e1efe1b6da4a3732404c19d4c6a614ff32dce90a251e293d4ce58 filter=lfs diff=lfs merge=lfs -text
40
+ pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
49e6dc219d1a1a1c9236acaf05a48b542002016a6dc877ee72baab085a84257b.3f28e7f4b38e1efe1b6da4a3732404c19d4c6a614ff32dce90a251e293d4ce58 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:83ab5c39e2b1c34a955136275ce0db068cb20d9643ead033d6b8124a73ab4f64
3
+ size 15645492
config.json ADDED
@@ -0,0 +1,166 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/clip-vit-base-patch16",
3
+ "architectures": [
4
+ "CLIPModel"
5
+ ],
6
+ "id2label": {},
7
+ "initializer_factor": 1.0,
8
+ "label2id": {
9
+ "Contradiction": 2,
10
+ "Entailment": 0,
11
+ "Neutral": 1
12
+ },
13
+ "logit_scale_init_value": 2.6592,
14
+ "model_type": "clip",
15
+ "projection_dim": 512,
16
+ "text_config": {
17
+ "_name_or_path": "",
18
+ "add_cross_attention": false,
19
+ "architectures": null,
20
+ "attention_dropout": 0.0,
21
+ "bad_words_ids": null,
22
+ "bos_token_id": 0,
23
+ "chunk_size_feed_forward": 0,
24
+ "cross_attention_hidden_size": null,
25
+ "decoder_start_token_id": null,
26
+ "diversity_penalty": 0.0,
27
+ "do_sample": false,
28
+ "dropout": 0.0,
29
+ "early_stopping": false,
30
+ "encoder_no_repeat_ngram_size": 0,
31
+ "eos_token_id": 2,
32
+ "finetuning_task": null,
33
+ "forced_bos_token_id": null,
34
+ "forced_eos_token_id": null,
35
+ "hidden_act": "quick_gelu",
36
+ "hidden_size": 512,
37
+ "id2label": {
38
+ "0": "LABEL_0",
39
+ "1": "LABEL_1"
40
+ },
41
+ "initializer_factor": 1.0,
42
+ "initializer_range": 0.02,
43
+ "intermediate_size": 2048,
44
+ "is_decoder": false,
45
+ "is_encoder_decoder": false,
46
+ "label2id": {
47
+ "LABEL_0": 0,
48
+ "LABEL_1": 1
49
+ },
50
+ "layer_norm_eps": 1e-05,
51
+ "length_penalty": 1.0,
52
+ "max_length": 20,
53
+ "max_position_embeddings": 77,
54
+ "min_length": 0,
55
+ "model_type": "clip_text_model",
56
+ "no_repeat_ngram_size": 0,
57
+ "num_attention_heads": 8,
58
+ "num_beam_groups": 1,
59
+ "num_beams": 1,
60
+ "num_hidden_layers": 12,
61
+ "num_return_sequences": 1,
62
+ "output_attentions": false,
63
+ "output_hidden_states": false,
64
+ "output_scores": false,
65
+ "pad_token_id": 1,
66
+ "prefix": null,
67
+ "problem_type": null,
68
+ "pruned_heads": {},
69
+ "remove_invalid_values": false,
70
+ "repetition_penalty": 1.0,
71
+ "return_dict": true,
72
+ "return_dict_in_generate": false,
73
+ "sep_token_id": null,
74
+ "task_specific_params": null,
75
+ "temperature": 1.0,
76
+ "tie_encoder_decoder": false,
77
+ "tie_word_embeddings": true,
78
+ "tokenizer_class": null,
79
+ "top_k": 50,
80
+ "top_p": 1.0,
81
+ "torch_dtype": null,
82
+ "torchscript": false,
83
+ "transformers_version": "4.17.0",
84
+ "typical_p": 1.0,
85
+ "use_bfloat16": false,
86
+ "vocab_size": 49408
87
+ },
88
+ "text_config_dict": null,
89
+ "torch_dtype": "float32",
90
+ "transformers_version": null,
91
+ "vision_config": {
92
+ "_name_or_path": "",
93
+ "add_cross_attention": false,
94
+ "architectures": null,
95
+ "attention_dropout": 0.0,
96
+ "bad_words_ids": null,
97
+ "bos_token_id": null,
98
+ "chunk_size_feed_forward": 0,
99
+ "cross_attention_hidden_size": null,
100
+ "decoder_start_token_id": null,
101
+ "diversity_penalty": 0.0,
102
+ "do_sample": false,
103
+ "dropout": 0.0,
104
+ "early_stopping": false,
105
+ "encoder_no_repeat_ngram_size": 0,
106
+ "eos_token_id": null,
107
+ "finetuning_task": null,
108
+ "forced_bos_token_id": null,
109
+ "forced_eos_token_id": null,
110
+ "hidden_act": "quick_gelu",
111
+ "hidden_size": 768,
112
+ "id2label": {
113
+ "0": "LABEL_0",
114
+ "1": "LABEL_1"
115
+ },
116
+ "image_size": 224,
117
+ "initializer_factor": 1.0,
118
+ "initializer_range": 0.02,
119
+ "intermediate_size": 3072,
120
+ "is_decoder": false,
121
+ "is_encoder_decoder": false,
122
+ "label2id": {
123
+ "LABEL_0": 0,
124
+ "LABEL_1": 1
125
+ },
126
+ "layer_norm_eps": 1e-05,
127
+ "length_penalty": 1.0,
128
+ "max_length": 20,
129
+ "min_length": 0,
130
+ "model_type": "clip_vision_model",
131
+ "no_repeat_ngram_size": 0,
132
+ "num_attention_heads": 12,
133
+ "num_beam_groups": 1,
134
+ "num_beams": 1,
135
+ "num_hidden_layers": 12,
136
+ "num_return_sequences": 1,
137
+ "output_attentions": false,
138
+ "output_hidden_states": false,
139
+ "output_scores": false,
140
+ "pad_token_id": null,
141
+ "patch_size": 16,
142
+ "prefix": null,
143
+ "problem_type": null,
144
+ "pruned_heads": {},
145
+ "remove_invalid_values": false,
146
+ "repetition_penalty": 1.0,
147
+ "return_dict": true,
148
+ "return_dict_in_generate": false,
149
+ "sep_token_id": null,
150
+ "task_specific_params": null,
151
+ "temperature": 1.0,
152
+ "tie_encoder_decoder": false,
153
+ "tie_word_embeddings": true,
154
+ "tokenizer_class": null,
155
+ "top_k": 50,
156
+ "top_p": 1.0,
157
+ "torch_dtype": null,
158
+ "torchscript": false,
159
+ "transformers_version": "4.17.0",
160
+ "typical_p": 1.0,
161
+ "use_bfloat16": false
162
+ },
163
+ "vision_config_dict": {
164
+ "patch_size": 16
165
+ }
166
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6345dc36c5794fc590328c9fd11013fd11e7f59937687c8004fa46ce58e2b686
3
+ size 689659578
preprocessor_config.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "crop_size": 224,
3
+ "do_center_crop": true,
4
+ "do_normalize": true,
5
+ "do_resize": true,
6
+ "feature_extractor_type": "CLIPFeatureExtractor",
7
+ "image_mean": [
8
+ 0.48145466,
9
+ 0.4578275,
10
+ 0.40821073
11
+ ],
12
+ "image_std": [
13
+ 0.26862954,
14
+ 0.26130258,
15
+ 0.27577711
16
+ ],
17
+ "resample": 3,
18
+ "size": 224
19
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f538a3f312c5a898d82ef2856d09441f1b1106879152eb0c5cdf442d70ad3f4
3
+ size 598642122
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e232edc83d465812f65101826626be9ad9017fa73fbc3674105d05fd489fbe9e
3
+ size 14244
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8cb4294b00426fcf22505f85c2e1b5058e3e65de42c2163517481b6b9ac8c8c8
3
+ size 1064
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"bos_token": {"content": "<|startoftext|>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "eos_token": {"content": "<|endoftext|>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "unk_token": {"content": "<|endoftext|>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "pad_token": "<|endoftext|>"}
text_renderer_config.json ADDED
@@ -0,0 +1,253 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "background_color": "white",
3
+ "dpi": 120,
4
+ "font_color": "black",
5
+ "font_file": "49e6dc219d1a1a1c9236acaf05a48b542002016a6dc877ee72baab085a84257b.3f28e7f4b38e1efe1b6da4a3732404c19d4c6a614ff32dce90a251e293d4ce58",
6
+ "font_size": 8,
7
+ "fonts_list": [
8
+ "Apple Color Emoji",
9
+ "DejaVu Sans",
10
+ "DejaVu Sans Mono",
11
+ "DejaVu Serif",
12
+ "FontAwesome",
13
+ "GLYPHICONS Halflings",
14
+ "Go Noto Current",
15
+ "Inconsolata",
16
+ "Lato",
17
+ "MathJax_AMS",
18
+ "MathJax_Caligraphic",
19
+ "MathJax_Fraktur",
20
+ "MathJax_Main",
21
+ "MathJax_Math",
22
+ "MathJax_SansSerif",
23
+ "MathJax_Script",
24
+ "MathJax_Size1",
25
+ "MathJax_Size2",
26
+ "MathJax_Size3",
27
+ "MathJax_Size4",
28
+ "MathJax_Typewriter",
29
+ "MathJax_Vector",
30
+ "MathJax_Vector-Bold",
31
+ "MathJax_WinChrome",
32
+ "MathJax_WinIE6",
33
+ "Monospace",
34
+ "Noto Fangsong KSS Rotated",
35
+ "Noto Fangsong KSS Vertical",
36
+ "Noto Kufi Arabic",
37
+ "Noto Music",
38
+ "Noto Naskh Arabic UI",
39
+ "Noto Nastaliq Urdu",
40
+ "Noto Rashi Hebrew",
41
+ "Noto Sans",
42
+ "Noto Sans Adlam",
43
+ "Noto Sans Adlam Unjoined",
44
+ "Noto Sans Anatolian Hieroglyphs",
45
+ "Noto Sans Arabic",
46
+ "Noto Sans Armenian",
47
+ "Noto Sans Avestan",
48
+ "Noto Sans Balinese",
49
+ "Noto Sans Bamum",
50
+ "Noto Sans Bassa Vah",
51
+ "Noto Sans Batak",
52
+ "Noto Sans Bhaiksuki",
53
+ "Noto Sans Brahmi",
54
+ "Noto Sans Buginese",
55
+ "Noto Sans Buhid",
56
+ "Noto Sans CJK HK",
57
+ "Noto Sans CJK JP",
58
+ "Noto Sans CJK KR",
59
+ "Noto Sans CJK SC",
60
+ "Noto Sans CJK TC",
61
+ "Noto Sans Canadian Aboriginal",
62
+ "Noto Sans Carian",
63
+ "Noto Sans Caucasian Albanian",
64
+ "Noto Sans Chakma",
65
+ "Noto Sans Cham",
66
+ "Noto Sans Cherokee",
67
+ "Noto Sans Chorasmian",
68
+ "Noto Sans Coptic",
69
+ "Noto Sans Cuneiform",
70
+ "Noto Sans Cypriot",
71
+ "Noto Sans Cypro Minoan",
72
+ "Noto Sans Deseret",
73
+ "Noto Sans Devanagari",
74
+ "Noto Sans Devanagari UI",
75
+ "Noto Sans Duployan",
76
+ "Noto Sans Egyptian Hieroglyphs",
77
+ "Noto Sans Elbasan",
78
+ "Noto Sans Elymaic",
79
+ "Noto Sans Ethiopic",
80
+ "Noto Sans Georgian",
81
+ "Noto Sans Glagolitic",
82
+ "Noto Sans Gothic",
83
+ "Noto Sans Grantha",
84
+ "Noto Sans Gujarati",
85
+ "Noto Sans Gujarati UI",
86
+ "Noto Sans Gunjala Gondi",
87
+ "Noto Sans Gurmukhi",
88
+ "Noto Sans Gurmukhi UI",
89
+ "Noto Sans Hanifi Rohingya",
90
+ "Noto Sans Hanunoo",
91
+ "Noto Sans Hatran",
92
+ "Noto Sans Hebrew",
93
+ "Noto Sans Hebrew Droid",
94
+ "Noto Sans Imperial Aramaic",
95
+ "Noto Sans Indic Siyaq Numbers",
96
+ "Noto Sans Inscriptional Pahlavi",
97
+ "Noto Sans Inscriptional Parthian",
98
+ "Noto Sans Javanese",
99
+ "Noto Sans Kaithi",
100
+ "Noto Sans Kannada",
101
+ "Noto Sans Kannada UI",
102
+ "Noto Sans Kawi",
103
+ "Noto Sans Kayah Li",
104
+ "Noto Sans Kharoshthi",
105
+ "Noto Sans Khmer",
106
+ "Noto Sans Khmer UI",
107
+ "Noto Sans Khojki",
108
+ "Noto Sans Khudawadi",
109
+ "Noto Sans Lao",
110
+ "Noto Sans Lao Looped",
111
+ "Noto Sans Lao Looped UI",
112
+ "Noto Sans Lao UI",
113
+ "Noto Sans Lepcha",
114
+ "Noto Sans Limbu",
115
+ "Noto Sans Linear A",
116
+ "Noto Sans Linear B",
117
+ "Noto Sans Lisu",
118
+ "Noto Sans Lycian",
119
+ "Noto Sans Lydian",
120
+ "Noto Sans Mahajani",
121
+ "Noto Sans Malayalam",
122
+ "Noto Sans Malayalam UI",
123
+ "Noto Sans Mandaic",
124
+ "Noto Sans Manichaean",
125
+ "Noto Sans Marchen",
126
+ "Noto Sans Masaram Gondi",
127
+ "Noto Sans Math",
128
+ "Noto Sans Mayan Numerals",
129
+ "Noto Sans Medefaidrin",
130
+ "Noto Sans Meetei Mayek",
131
+ "Noto Sans Mende Kikakui",
132
+ "Noto Sans Meroitic",
133
+ "Noto Sans Miao",
134
+ "Noto Sans Modi",
135
+ "Noto Sans Mongolian",
136
+ "Noto Sans Mono",
137
+ "Noto Sans Mro",
138
+ "Noto Sans Multani",
139
+ "Noto Sans Myanmar",
140
+ "Noto Sans Myanmar UI",
141
+ "Noto Sans NKo",
142
+ "Noto Sans NKo Unjoined",
143
+ "Noto Sans Nabataean",
144
+ "Noto Sans Nag Mundari",
145
+ "Noto Sans Nandinagari",
146
+ "Noto Sans New Tai Lue",
147
+ "Noto Sans Newa",
148
+ "Noto Sans Nushu",
149
+ "Noto Sans Ogham",
150
+ "Noto Sans Ol Chiki",
151
+ "Noto Sans Old Hungarian",
152
+ "Noto Sans Old Italic",
153
+ "Noto Sans Old North Arabian",
154
+ "Noto Sans Old Permic",
155
+ "Noto Sans Old Persian",
156
+ "Noto Sans Old Sogdian",
157
+ "Noto Sans Old South Arabian",
158
+ "Noto Sans Old Turkic",
159
+ "Noto Sans Oriya",
160
+ "Noto Sans Osage",
161
+ "Noto Sans Osmanya",
162
+ "Noto Sans Pahawh Hmong",
163
+ "Noto Sans Palmyrene",
164
+ "Noto Sans Pau Cin Hau",
165
+ "Noto Sans PhagsPa",
166
+ "Noto Sans Phoenician",
167
+ "Noto Sans Psalter Pahlavi",
168
+ "Noto Sans Rejang",
169
+ "Noto Sans Runic",
170
+ "Noto Sans Samaritan",
171
+ "Noto Sans Saurashtra",
172
+ "Noto Sans Sharada",
173
+ "Noto Sans Shavian",
174
+ "Noto Sans Siddham",
175
+ "Noto Sans SignWriting",
176
+ "Noto Sans Sinhala",
177
+ "Noto Sans Sinhala UI",
178
+ "Noto Sans Sogdian",
179
+ "Noto Sans Sora Sompeng",
180
+ "Noto Sans Soyombo",
181
+ "Noto Sans Sundanese",
182
+ "Noto Sans Syloti Nagri",
183
+ "Noto Sans Symbols",
184
+ "Noto Sans Symbols 2",
185
+ "Noto Sans Syriac",
186
+ "Noto Sans Syriac Eastern",
187
+ "Noto Sans Syriac Western",
188
+ "Noto Sans Tagalog",
189
+ "Noto Sans Tagbanwa",
190
+ "Noto Sans Tai Le",
191
+ "Noto Sans Tai Tham",
192
+ "Noto Sans Tai Viet",
193
+ "Noto Sans Takri",
194
+ "Noto Sans Tamil",
195
+ "Noto Sans Tamil Supplement",
196
+ "Noto Sans Tamil UI",
197
+ "Noto Sans Tangsa",
198
+ "Noto Sans Telugu",
199
+ "Noto Sans Telugu UI",
200
+ "Noto Sans Test",
201
+ "Noto Sans Thaana",
202
+ "Noto Sans Thai",
203
+ "Noto Sans Thai Looped",
204
+ "Noto Sans Thai Looped UI",
205
+ "Noto Sans Thai UI",
206
+ "Noto Sans Tifinagh",
207
+ "Noto Sans Tifinagh APT",
208
+ "Noto Sans Tifinagh Adrar",
209
+ "Noto Sans Tifinagh Agraw Imazighen",
210
+ "Noto Sans Tifinagh Ahaggar",
211
+ "Noto Sans Tifinagh Air",
212
+ "Noto Sans Tifinagh Azawagh",
213
+ "Noto Sans Tifinagh Ghat",
214
+ "Noto Sans Tifinagh Hawad",
215
+ "Noto Sans Tifinagh Rhissa Ixa",
216
+ "Noto Sans Tifinagh SIL",
217
+ "Noto Sans Tifinagh Tawellemmet",
218
+ "Noto Sans Tirhuta",
219
+ "Noto Sans Ugaritic",
220
+ "Noto Sans Vai",
221
+ "Noto Sans Vithkuqi",
222
+ "Noto Sans Wancho",
223
+ "Noto Sans Warang Citi",
224
+ "Noto Sans Yi",
225
+ "Noto Sans Zanabazar Square",
226
+ "Noto Traditional Nushu",
227
+ "NotoSansOldHungarianUI",
228
+ "Sans",
229
+ "Serif",
230
+ "Source Code Pro",
231
+ "System-ui",
232
+ "Ubuntu",
233
+ "Ubuntu Condensed",
234
+ "Ubuntu Mono",
235
+ "cmex10",
236
+ "cmmi10",
237
+ "cmr10",
238
+ "cmsy10",
239
+ "dsrom10",
240
+ "esint10",
241
+ "eufm10",
242
+ "msam10",
243
+ "msbm10",
244
+ "rsfs10",
245
+ "stmary10",
246
+ "wasy10"
247
+ ],
248
+ "max_seq_length": 196,
249
+ "pad_size": 3,
250
+ "pixels_per_patch": 16,
251
+ "rgb": false,
252
+ "text_renderer_type": "PangoCairoTextRenderer"
253
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "unk_token": {
3
+ "content": "<|endoftext|>",
4
+ "single_word": false,
5
+ "lstrip": false,
6
+ "rstrip": false,
7
+ "normalized": true,
8
+ "__type": "AddedToken"
9
+ },
10
+ "bos_token": {
11
+ "content": "<|startoftext|>",
12
+ "single_word": false,
13
+ "lstrip": false,
14
+ "rstrip": false,
15
+ "normalized": true,
16
+ "__type": "AddedToken"
17
+ },
18
+ "eos_token": {
19
+ "content": "<|endoftext|>",
20
+ "single_word": false,
21
+ "lstrip": false,
22
+ "rstrip": false,
23
+ "normalized": true,
24
+ "__type": "AddedToken"
25
+ },
26
+ "pad_token": "<|endoftext|>",
27
+ "errors": "replace",
28
+ "add_prefix_space": false,
29
+ "do_lower_case": true,
30
+ "name_or_path": "openai/clip-vit-base-patch16",
31
+ "model_max_length": 77,
32
+ "special_tokens_map_file": "./special_tokens_map.json",
33
+ "tokenizer_class": "CLIPTokenizer"
34
+ }
trainer_state.json ADDED
@@ -0,0 +1,1816 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7258342522680835,
3
+ "best_model_checkpoint": "./checkpoints/clip-stage1-10240/checkpoint-100",
4
+ "epoch": 0.9970089730807578,
5
+ "global_step": 6000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 2e-06,
13
+ "loss": 3.6467,
14
+ "step": 20
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 4e-06,
19
+ "loss": 2.3745,
20
+ "step": 40
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 6e-06,
25
+ "loss": 1.9228,
26
+ "step": 60
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 8e-06,
31
+ "loss": 1.6728,
32
+ "step": 80
33
+ },
34
+ {
35
+ "epoch": 0.02,
36
+ "learning_rate": 9.999999999999999e-06,
37
+ "loss": 1.5235,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.02,
42
+ "learning_rate": 1.2e-05,
43
+ "loss": 1.4338,
44
+ "step": 120
45
+ },
46
+ {
47
+ "epoch": 0.02,
48
+ "learning_rate": 1.4e-05,
49
+ "loss": 1.3564,
50
+ "step": 140
51
+ },
52
+ {
53
+ "epoch": 0.03,
54
+ "learning_rate": 1.6e-05,
55
+ "loss": 1.2923,
56
+ "step": 160
57
+ },
58
+ {
59
+ "epoch": 0.03,
60
+ "learning_rate": 1.8e-05,
61
+ "loss": 1.2428,
62
+ "step": 180
63
+ },
64
+ {
65
+ "epoch": 0.03,
66
+ "learning_rate": 1.9999999999999998e-05,
67
+ "loss": 1.2069,
68
+ "step": 200
69
+ },
70
+ {
71
+ "epoch": 0.04,
72
+ "learning_rate": 2.2e-05,
73
+ "loss": 1.1654,
74
+ "step": 220
75
+ },
76
+ {
77
+ "epoch": 0.04,
78
+ "learning_rate": 2.4e-05,
79
+ "loss": 1.129,
80
+ "step": 240
81
+ },
82
+ {
83
+ "epoch": 0.04,
84
+ "learning_rate": 2.6000000000000002e-05,
85
+ "loss": 1.095,
86
+ "step": 260
87
+ },
88
+ {
89
+ "epoch": 0.05,
90
+ "learning_rate": 2.8e-05,
91
+ "loss": 1.0813,
92
+ "step": 280
93
+ },
94
+ {
95
+ "epoch": 0.05,
96
+ "learning_rate": 3e-05,
97
+ "loss": 1.0522,
98
+ "step": 300
99
+ },
100
+ {
101
+ "epoch": 0.05,
102
+ "learning_rate": 2.9894736842105264e-05,
103
+ "loss": 1.0212,
104
+ "step": 320
105
+ },
106
+ {
107
+ "epoch": 0.06,
108
+ "learning_rate": 2.9789473684210527e-05,
109
+ "loss": 0.9983,
110
+ "step": 340
111
+ },
112
+ {
113
+ "epoch": 0.06,
114
+ "learning_rate": 2.968421052631579e-05,
115
+ "loss": 0.9816,
116
+ "step": 360
117
+ },
118
+ {
119
+ "epoch": 0.06,
120
+ "learning_rate": 2.957894736842105e-05,
121
+ "loss": 0.9694,
122
+ "step": 380
123
+ },
124
+ {
125
+ "epoch": 0.07,
126
+ "learning_rate": 2.9473684210526314e-05,
127
+ "loss": 0.9517,
128
+ "step": 400
129
+ },
130
+ {
131
+ "epoch": 0.07,
132
+ "learning_rate": 2.936842105263158e-05,
133
+ "loss": 0.9385,
134
+ "step": 420
135
+ },
136
+ {
137
+ "epoch": 0.07,
138
+ "learning_rate": 2.9263157894736844e-05,
139
+ "loss": 0.9146,
140
+ "step": 440
141
+ },
142
+ {
143
+ "epoch": 0.08,
144
+ "learning_rate": 2.9157894736842107e-05,
145
+ "loss": 0.9091,
146
+ "step": 460
147
+ },
148
+ {
149
+ "epoch": 0.08,
150
+ "learning_rate": 2.905263157894737e-05,
151
+ "loss": 0.8921,
152
+ "step": 480
153
+ },
154
+ {
155
+ "epoch": 0.08,
156
+ "learning_rate": 2.8947368421052634e-05,
157
+ "loss": 0.8871,
158
+ "step": 500
159
+ },
160
+ {
161
+ "epoch": 0.09,
162
+ "learning_rate": 2.8842105263157897e-05,
163
+ "loss": 0.8704,
164
+ "step": 520
165
+ },
166
+ {
167
+ "epoch": 0.09,
168
+ "learning_rate": 2.8736842105263157e-05,
169
+ "loss": 0.8601,
170
+ "step": 540
171
+ },
172
+ {
173
+ "epoch": 0.09,
174
+ "learning_rate": 2.863157894736842e-05,
175
+ "loss": 0.8561,
176
+ "step": 560
177
+ },
178
+ {
179
+ "epoch": 0.1,
180
+ "learning_rate": 2.8526315789473683e-05,
181
+ "loss": 0.8481,
182
+ "step": 580
183
+ },
184
+ {
185
+ "epoch": 0.1,
186
+ "learning_rate": 2.8421052631578946e-05,
187
+ "loss": 0.8323,
188
+ "step": 600
189
+ },
190
+ {
191
+ "epoch": 0.1,
192
+ "learning_rate": 2.8315789473684213e-05,
193
+ "loss": 0.8285,
194
+ "step": 620
195
+ },
196
+ {
197
+ "epoch": 0.11,
198
+ "learning_rate": 2.8210526315789476e-05,
199
+ "loss": 0.8279,
200
+ "step": 640
201
+ },
202
+ {
203
+ "epoch": 0.11,
204
+ "learning_rate": 2.810526315789474e-05,
205
+ "loss": 0.8166,
206
+ "step": 660
207
+ },
208
+ {
209
+ "epoch": 0.11,
210
+ "learning_rate": 2.8e-05,
211
+ "loss": 0.8072,
212
+ "step": 680
213
+ },
214
+ {
215
+ "epoch": 0.12,
216
+ "learning_rate": 2.7894736842105263e-05,
217
+ "loss": 0.7981,
218
+ "step": 700
219
+ },
220
+ {
221
+ "epoch": 0.12,
222
+ "learning_rate": 2.7789473684210526e-05,
223
+ "loss": 0.8,
224
+ "step": 720
225
+ },
226
+ {
227
+ "epoch": 0.12,
228
+ "learning_rate": 2.768421052631579e-05,
229
+ "loss": 0.7873,
230
+ "step": 740
231
+ },
232
+ {
233
+ "epoch": 0.13,
234
+ "learning_rate": 2.7578947368421053e-05,
235
+ "loss": 0.7805,
236
+ "step": 760
237
+ },
238
+ {
239
+ "epoch": 0.13,
240
+ "learning_rate": 2.7473684210526316e-05,
241
+ "loss": 0.7772,
242
+ "step": 780
243
+ },
244
+ {
245
+ "epoch": 0.13,
246
+ "learning_rate": 2.736842105263158e-05,
247
+ "loss": 0.7683,
248
+ "step": 800
249
+ },
250
+ {
251
+ "epoch": 0.14,
252
+ "learning_rate": 2.7263157894736846e-05,
253
+ "loss": 0.7711,
254
+ "step": 820
255
+ },
256
+ {
257
+ "epoch": 0.14,
258
+ "learning_rate": 2.7157894736842106e-05,
259
+ "loss": 0.7627,
260
+ "step": 840
261
+ },
262
+ {
263
+ "epoch": 0.14,
264
+ "learning_rate": 2.705263157894737e-05,
265
+ "loss": 0.7543,
266
+ "step": 860
267
+ },
268
+ {
269
+ "epoch": 0.15,
270
+ "learning_rate": 2.6947368421052632e-05,
271
+ "loss": 0.7538,
272
+ "step": 880
273
+ },
274
+ {
275
+ "epoch": 0.15,
276
+ "learning_rate": 2.6842105263157896e-05,
277
+ "loss": 0.7468,
278
+ "step": 900
279
+ },
280
+ {
281
+ "epoch": 0.15,
282
+ "learning_rate": 2.673684210526316e-05,
283
+ "loss": 0.7376,
284
+ "step": 920
285
+ },
286
+ {
287
+ "epoch": 0.16,
288
+ "learning_rate": 2.6631578947368422e-05,
289
+ "loss": 0.7427,
290
+ "step": 940
291
+ },
292
+ {
293
+ "epoch": 0.16,
294
+ "learning_rate": 2.6526315789473685e-05,
295
+ "loss": 0.7305,
296
+ "step": 960
297
+ },
298
+ {
299
+ "epoch": 0.16,
300
+ "learning_rate": 2.6421052631578945e-05,
301
+ "loss": 0.727,
302
+ "step": 980
303
+ },
304
+ {
305
+ "epoch": 0.17,
306
+ "learning_rate": 2.631578947368421e-05,
307
+ "loss": 0.7305,
308
+ "step": 1000
309
+ },
310
+ {
311
+ "epoch": 0.17,
312
+ "learning_rate": 2.6210526315789475e-05,
313
+ "loss": 0.7243,
314
+ "step": 1020
315
+ },
316
+ {
317
+ "epoch": 0.17,
318
+ "learning_rate": 2.610526315789474e-05,
319
+ "loss": 0.7189,
320
+ "step": 1040
321
+ },
322
+ {
323
+ "epoch": 0.18,
324
+ "learning_rate": 2.6000000000000002e-05,
325
+ "loss": 0.721,
326
+ "step": 1060
327
+ },
328
+ {
329
+ "epoch": 0.18,
330
+ "learning_rate": 2.5894736842105265e-05,
331
+ "loss": 0.7164,
332
+ "step": 1080
333
+ },
334
+ {
335
+ "epoch": 0.18,
336
+ "learning_rate": 2.578947368421053e-05,
337
+ "loss": 0.7171,
338
+ "step": 1100
339
+ },
340
+ {
341
+ "epoch": 0.19,
342
+ "learning_rate": 2.568421052631579e-05,
343
+ "loss": 0.7082,
344
+ "step": 1120
345
+ },
346
+ {
347
+ "epoch": 0.19,
348
+ "learning_rate": 2.557894736842105e-05,
349
+ "loss": 0.6998,
350
+ "step": 1140
351
+ },
352
+ {
353
+ "epoch": 0.19,
354
+ "learning_rate": 2.5473684210526315e-05,
355
+ "loss": 0.7025,
356
+ "step": 1160
357
+ },
358
+ {
359
+ "epoch": 0.2,
360
+ "learning_rate": 2.5368421052631578e-05,
361
+ "loss": 0.6959,
362
+ "step": 1180
363
+ },
364
+ {
365
+ "epoch": 0.2,
366
+ "learning_rate": 2.526315789473684e-05,
367
+ "loss": 0.6996,
368
+ "step": 1200
369
+ },
370
+ {
371
+ "epoch": 0.2,
372
+ "learning_rate": 2.5157894736842108e-05,
373
+ "loss": 0.6923,
374
+ "step": 1220
375
+ },
376
+ {
377
+ "epoch": 0.21,
378
+ "learning_rate": 2.505263157894737e-05,
379
+ "loss": 0.6906,
380
+ "step": 1240
381
+ },
382
+ {
383
+ "epoch": 0.21,
384
+ "learning_rate": 2.4947368421052635e-05,
385
+ "loss": 0.6894,
386
+ "step": 1260
387
+ },
388
+ {
389
+ "epoch": 0.21,
390
+ "learning_rate": 2.4842105263157894e-05,
391
+ "loss": 0.6838,
392
+ "step": 1280
393
+ },
394
+ {
395
+ "epoch": 0.22,
396
+ "learning_rate": 2.4736842105263158e-05,
397
+ "loss": 0.6803,
398
+ "step": 1300
399
+ },
400
+ {
401
+ "epoch": 0.22,
402
+ "learning_rate": 2.463157894736842e-05,
403
+ "loss": 0.68,
404
+ "step": 1320
405
+ },
406
+ {
407
+ "epoch": 0.22,
408
+ "learning_rate": 2.4526315789473684e-05,
409
+ "loss": 0.6734,
410
+ "step": 1340
411
+ },
412
+ {
413
+ "epoch": 0.23,
414
+ "learning_rate": 2.4421052631578948e-05,
415
+ "loss": 0.6732,
416
+ "step": 1360
417
+ },
418
+ {
419
+ "epoch": 0.23,
420
+ "learning_rate": 2.431578947368421e-05,
421
+ "loss": 0.673,
422
+ "step": 1380
423
+ },
424
+ {
425
+ "epoch": 0.23,
426
+ "learning_rate": 2.4210526315789474e-05,
427
+ "loss": 0.6675,
428
+ "step": 1400
429
+ },
430
+ {
431
+ "epoch": 0.24,
432
+ "learning_rate": 2.410526315789474e-05,
433
+ "loss": 0.6689,
434
+ "step": 1420
435
+ },
436
+ {
437
+ "epoch": 0.24,
438
+ "learning_rate": 2.4e-05,
439
+ "loss": 0.6671,
440
+ "step": 1440
441
+ },
442
+ {
443
+ "epoch": 0.24,
444
+ "learning_rate": 2.3894736842105264e-05,
445
+ "loss": 0.6666,
446
+ "step": 1460
447
+ },
448
+ {
449
+ "epoch": 0.25,
450
+ "learning_rate": 2.3789473684210527e-05,
451
+ "loss": 0.6582,
452
+ "step": 1480
453
+ },
454
+ {
455
+ "epoch": 0.25,
456
+ "learning_rate": 2.368421052631579e-05,
457
+ "loss": 0.6541,
458
+ "step": 1500
459
+ },
460
+ {
461
+ "epoch": 0.25,
462
+ "learning_rate": 2.3578947368421054e-05,
463
+ "loss": 0.6621,
464
+ "step": 1520
465
+ },
466
+ {
467
+ "epoch": 0.26,
468
+ "learning_rate": 2.3473684210526317e-05,
469
+ "loss": 0.6549,
470
+ "step": 1540
471
+ },
472
+ {
473
+ "epoch": 0.26,
474
+ "learning_rate": 2.336842105263158e-05,
475
+ "loss": 0.6503,
476
+ "step": 1560
477
+ },
478
+ {
479
+ "epoch": 0.26,
480
+ "learning_rate": 2.326315789473684e-05,
481
+ "loss": 0.6492,
482
+ "step": 1580
483
+ },
484
+ {
485
+ "epoch": 0.27,
486
+ "learning_rate": 2.3157894736842103e-05,
487
+ "loss": 0.6498,
488
+ "step": 1600
489
+ },
490
+ {
491
+ "epoch": 0.27,
492
+ "learning_rate": 2.3052631578947367e-05,
493
+ "loss": 0.6523,
494
+ "step": 1620
495
+ },
496
+ {
497
+ "epoch": 0.27,
498
+ "learning_rate": 2.2947368421052633e-05,
499
+ "loss": 0.6443,
500
+ "step": 1640
501
+ },
502
+ {
503
+ "epoch": 0.28,
504
+ "learning_rate": 2.2842105263157897e-05,
505
+ "loss": 0.6442,
506
+ "step": 1660
507
+ },
508
+ {
509
+ "epoch": 0.28,
510
+ "learning_rate": 2.273684210526316e-05,
511
+ "loss": 0.6459,
512
+ "step": 1680
513
+ },
514
+ {
515
+ "epoch": 0.28,
516
+ "learning_rate": 2.2631578947368423e-05,
517
+ "loss": 0.6426,
518
+ "step": 1700
519
+ },
520
+ {
521
+ "epoch": 0.29,
522
+ "learning_rate": 2.2526315789473686e-05,
523
+ "loss": 0.636,
524
+ "step": 1720
525
+ },
526
+ {
527
+ "epoch": 0.29,
528
+ "learning_rate": 2.2421052631578946e-05,
529
+ "loss": 0.6365,
530
+ "step": 1740
531
+ },
532
+ {
533
+ "epoch": 0.29,
534
+ "learning_rate": 2.231578947368421e-05,
535
+ "loss": 0.636,
536
+ "step": 1760
537
+ },
538
+ {
539
+ "epoch": 0.3,
540
+ "learning_rate": 2.2210526315789473e-05,
541
+ "loss": 0.6381,
542
+ "step": 1780
543
+ },
544
+ {
545
+ "epoch": 0.3,
546
+ "learning_rate": 2.2105263157894736e-05,
547
+ "loss": 0.6346,
548
+ "step": 1800
549
+ },
550
+ {
551
+ "epoch": 0.3,
552
+ "learning_rate": 2.2e-05,
553
+ "loss": 0.6291,
554
+ "step": 1820
555
+ },
556
+ {
557
+ "epoch": 0.31,
558
+ "learning_rate": 2.1894736842105266e-05,
559
+ "loss": 0.63,
560
+ "step": 1840
561
+ },
562
+ {
563
+ "epoch": 0.31,
564
+ "learning_rate": 2.178947368421053e-05,
565
+ "loss": 0.6296,
566
+ "step": 1860
567
+ },
568
+ {
569
+ "epoch": 0.31,
570
+ "learning_rate": 2.168421052631579e-05,
571
+ "loss": 0.6278,
572
+ "step": 1880
573
+ },
574
+ {
575
+ "epoch": 0.32,
576
+ "learning_rate": 2.1578947368421053e-05,
577
+ "loss": 0.6261,
578
+ "step": 1900
579
+ },
580
+ {
581
+ "epoch": 0.32,
582
+ "learning_rate": 2.1473684210526316e-05,
583
+ "loss": 0.6234,
584
+ "step": 1920
585
+ },
586
+ {
587
+ "epoch": 0.32,
588
+ "learning_rate": 2.136842105263158e-05,
589
+ "loss": 0.6201,
590
+ "step": 1940
591
+ },
592
+ {
593
+ "epoch": 0.33,
594
+ "learning_rate": 2.1263157894736842e-05,
595
+ "loss": 0.6187,
596
+ "step": 1960
597
+ },
598
+ {
599
+ "epoch": 0.33,
600
+ "learning_rate": 2.1157894736842106e-05,
601
+ "loss": 0.6201,
602
+ "step": 1980
603
+ },
604
+ {
605
+ "epoch": 0.33,
606
+ "learning_rate": 2.105263157894737e-05,
607
+ "loss": 0.6165,
608
+ "step": 2000
609
+ },
610
+ {
611
+ "epoch": 0.34,
612
+ "learning_rate": 2.0947368421052632e-05,
613
+ "loss": 0.6181,
614
+ "step": 2020
615
+ },
616
+ {
617
+ "epoch": 0.34,
618
+ "learning_rate": 2.0842105263157895e-05,
619
+ "loss": 0.6129,
620
+ "step": 2040
621
+ },
622
+ {
623
+ "epoch": 0.34,
624
+ "learning_rate": 2.073684210526316e-05,
625
+ "loss": 0.6158,
626
+ "step": 2060
627
+ },
628
+ {
629
+ "epoch": 0.35,
630
+ "learning_rate": 2.0631578947368422e-05,
631
+ "loss": 0.6141,
632
+ "step": 2080
633
+ },
634
+ {
635
+ "epoch": 0.35,
636
+ "learning_rate": 2.0526315789473685e-05,
637
+ "loss": 0.608,
638
+ "step": 2100
639
+ },
640
+ {
641
+ "epoch": 0.35,
642
+ "learning_rate": 2.042105263157895e-05,
643
+ "loss": 0.6077,
644
+ "step": 2120
645
+ },
646
+ {
647
+ "epoch": 0.36,
648
+ "learning_rate": 2.0315789473684212e-05,
649
+ "loss": 0.6087,
650
+ "step": 2140
651
+ },
652
+ {
653
+ "epoch": 0.36,
654
+ "learning_rate": 2.0210526315789475e-05,
655
+ "loss": 0.6066,
656
+ "step": 2160
657
+ },
658
+ {
659
+ "epoch": 0.36,
660
+ "learning_rate": 2.0105263157894735e-05,
661
+ "loss": 0.6035,
662
+ "step": 2180
663
+ },
664
+ {
665
+ "epoch": 0.37,
666
+ "learning_rate": 1.9999999999999998e-05,
667
+ "loss": 0.6006,
668
+ "step": 2200
669
+ },
670
+ {
671
+ "epoch": 0.37,
672
+ "learning_rate": 1.989473684210526e-05,
673
+ "loss": 0.6061,
674
+ "step": 2220
675
+ },
676
+ {
677
+ "epoch": 0.37,
678
+ "learning_rate": 1.9789473684210528e-05,
679
+ "loss": 0.6025,
680
+ "step": 2240
681
+ },
682
+ {
683
+ "epoch": 0.38,
684
+ "learning_rate": 1.968421052631579e-05,
685
+ "loss": 0.6021,
686
+ "step": 2260
687
+ },
688
+ {
689
+ "epoch": 0.38,
690
+ "learning_rate": 1.9578947368421055e-05,
691
+ "loss": 0.5989,
692
+ "step": 2280
693
+ },
694
+ {
695
+ "epoch": 0.38,
696
+ "learning_rate": 1.9473684210526318e-05,
697
+ "loss": 0.5998,
698
+ "step": 2300
699
+ },
700
+ {
701
+ "epoch": 0.39,
702
+ "learning_rate": 1.936842105263158e-05,
703
+ "loss": 0.597,
704
+ "step": 2320
705
+ },
706
+ {
707
+ "epoch": 0.39,
708
+ "learning_rate": 1.926315789473684e-05,
709
+ "loss": 0.5995,
710
+ "step": 2340
711
+ },
712
+ {
713
+ "epoch": 0.39,
714
+ "learning_rate": 1.9157894736842104e-05,
715
+ "loss": 0.5965,
716
+ "step": 2360
717
+ },
718
+ {
719
+ "epoch": 0.4,
720
+ "learning_rate": 1.9052631578947368e-05,
721
+ "loss": 0.5922,
722
+ "step": 2380
723
+ },
724
+ {
725
+ "epoch": 0.4,
726
+ "learning_rate": 1.894736842105263e-05,
727
+ "loss": 0.5922,
728
+ "step": 2400
729
+ },
730
+ {
731
+ "epoch": 0.4,
732
+ "learning_rate": 1.8842105263157894e-05,
733
+ "loss": 0.5964,
734
+ "step": 2420
735
+ },
736
+ {
737
+ "epoch": 0.41,
738
+ "learning_rate": 1.873684210526316e-05,
739
+ "loss": 0.5898,
740
+ "step": 2440
741
+ },
742
+ {
743
+ "epoch": 0.41,
744
+ "learning_rate": 1.8631578947368424e-05,
745
+ "loss": 0.5897,
746
+ "step": 2460
747
+ },
748
+ {
749
+ "epoch": 0.41,
750
+ "learning_rate": 1.8526315789473684e-05,
751
+ "loss": 0.5818,
752
+ "step": 2480
753
+ },
754
+ {
755
+ "epoch": 0.42,
756
+ "learning_rate": 1.8421052631578947e-05,
757
+ "loss": 0.5876,
758
+ "step": 2500
759
+ },
760
+ {
761
+ "epoch": 0.42,
762
+ "learning_rate": 1.831578947368421e-05,
763
+ "loss": 0.5868,
764
+ "step": 2520
765
+ },
766
+ {
767
+ "epoch": 0.42,
768
+ "learning_rate": 1.8210526315789474e-05,
769
+ "loss": 0.5864,
770
+ "step": 2540
771
+ },
772
+ {
773
+ "epoch": 0.43,
774
+ "learning_rate": 1.8105263157894737e-05,
775
+ "loss": 0.5851,
776
+ "step": 2560
777
+ },
778
+ {
779
+ "epoch": 0.43,
780
+ "learning_rate": 1.8e-05,
781
+ "loss": 0.5835,
782
+ "step": 2580
783
+ },
784
+ {
785
+ "epoch": 0.43,
786
+ "learning_rate": 1.7894736842105264e-05,
787
+ "loss": 0.5777,
788
+ "step": 2600
789
+ },
790
+ {
791
+ "epoch": 0.44,
792
+ "learning_rate": 1.7789473684210527e-05,
793
+ "loss": 0.5832,
794
+ "step": 2620
795
+ },
796
+ {
797
+ "epoch": 0.44,
798
+ "learning_rate": 1.7684210526315787e-05,
799
+ "loss": 0.579,
800
+ "step": 2640
801
+ },
802
+ {
803
+ "epoch": 0.44,
804
+ "learning_rate": 1.7578947368421054e-05,
805
+ "loss": 0.5833,
806
+ "step": 2660
807
+ },
808
+ {
809
+ "epoch": 0.45,
810
+ "learning_rate": 1.7473684210526317e-05,
811
+ "loss": 0.585,
812
+ "step": 2680
813
+ },
814
+ {
815
+ "epoch": 0.45,
816
+ "learning_rate": 1.736842105263158e-05,
817
+ "loss": 0.5819,
818
+ "step": 2700
819
+ },
820
+ {
821
+ "epoch": 0.45,
822
+ "learning_rate": 1.7263157894736843e-05,
823
+ "loss": 0.5744,
824
+ "step": 2720
825
+ },
826
+ {
827
+ "epoch": 0.46,
828
+ "learning_rate": 1.7157894736842107e-05,
829
+ "loss": 0.5743,
830
+ "step": 2740
831
+ },
832
+ {
833
+ "epoch": 0.46,
834
+ "learning_rate": 1.705263157894737e-05,
835
+ "loss": 0.5787,
836
+ "step": 2760
837
+ },
838
+ {
839
+ "epoch": 0.46,
840
+ "learning_rate": 1.694736842105263e-05,
841
+ "loss": 0.5737,
842
+ "step": 2780
843
+ },
844
+ {
845
+ "epoch": 0.47,
846
+ "learning_rate": 1.6842105263157893e-05,
847
+ "loss": 0.5733,
848
+ "step": 2800
849
+ },
850
+ {
851
+ "epoch": 0.47,
852
+ "learning_rate": 1.6736842105263156e-05,
853
+ "loss": 0.5711,
854
+ "step": 2820
855
+ },
856
+ {
857
+ "epoch": 0.47,
858
+ "learning_rate": 1.663157894736842e-05,
859
+ "loss": 0.5667,
860
+ "step": 2840
861
+ },
862
+ {
863
+ "epoch": 0.48,
864
+ "learning_rate": 1.6526315789473686e-05,
865
+ "loss": 0.5659,
866
+ "step": 2860
867
+ },
868
+ {
869
+ "epoch": 0.48,
870
+ "learning_rate": 1.642105263157895e-05,
871
+ "loss": 0.5685,
872
+ "step": 2880
873
+ },
874
+ {
875
+ "epoch": 0.48,
876
+ "learning_rate": 1.6315789473684213e-05,
877
+ "loss": 0.5664,
878
+ "step": 2900
879
+ },
880
+ {
881
+ "epoch": 0.49,
882
+ "learning_rate": 1.6210526315789476e-05,
883
+ "loss": 0.5691,
884
+ "step": 2920
885
+ },
886
+ {
887
+ "epoch": 0.49,
888
+ "learning_rate": 1.6105263157894736e-05,
889
+ "loss": 0.5645,
890
+ "step": 2940
891
+ },
892
+ {
893
+ "epoch": 0.49,
894
+ "learning_rate": 1.6e-05,
895
+ "loss": 0.568,
896
+ "step": 2960
897
+ },
898
+ {
899
+ "epoch": 0.5,
900
+ "learning_rate": 1.5894736842105263e-05,
901
+ "loss": 0.5614,
902
+ "step": 2980
903
+ },
904
+ {
905
+ "epoch": 0.5,
906
+ "learning_rate": 1.5789473684210526e-05,
907
+ "loss": 0.5648,
908
+ "step": 3000
909
+ },
910
+ {
911
+ "epoch": 0.5,
912
+ "learning_rate": 1.568421052631579e-05,
913
+ "loss": 0.5635,
914
+ "step": 3020
915
+ },
916
+ {
917
+ "epoch": 0.51,
918
+ "learning_rate": 1.5578947368421052e-05,
919
+ "loss": 0.5635,
920
+ "step": 3040
921
+ },
922
+ {
923
+ "epoch": 0.51,
924
+ "learning_rate": 1.547368421052632e-05,
925
+ "loss": 0.5648,
926
+ "step": 3060
927
+ },
928
+ {
929
+ "epoch": 0.51,
930
+ "learning_rate": 1.536842105263158e-05,
931
+ "loss": 0.5636,
932
+ "step": 3080
933
+ },
934
+ {
935
+ "epoch": 0.52,
936
+ "learning_rate": 1.5263157894736842e-05,
937
+ "loss": 0.5593,
938
+ "step": 3100
939
+ },
940
+ {
941
+ "epoch": 0.52,
942
+ "learning_rate": 1.5157894736842105e-05,
943
+ "loss": 0.5553,
944
+ "step": 3120
945
+ },
946
+ {
947
+ "epoch": 0.52,
948
+ "learning_rate": 1.5052631578947369e-05,
949
+ "loss": 0.5602,
950
+ "step": 3140
951
+ },
952
+ {
953
+ "epoch": 0.53,
954
+ "learning_rate": 1.4947368421052632e-05,
955
+ "loss": 0.5574,
956
+ "step": 3160
957
+ },
958
+ {
959
+ "epoch": 0.53,
960
+ "learning_rate": 1.4842105263157895e-05,
961
+ "loss": 0.558,
962
+ "step": 3180
963
+ },
964
+ {
965
+ "epoch": 0.53,
966
+ "learning_rate": 1.4736842105263157e-05,
967
+ "loss": 0.555,
968
+ "step": 3200
969
+ },
970
+ {
971
+ "epoch": 0.54,
972
+ "learning_rate": 1.4631578947368422e-05,
973
+ "loss": 0.5619,
974
+ "step": 3220
975
+ },
976
+ {
977
+ "epoch": 0.54,
978
+ "learning_rate": 1.4526315789473685e-05,
979
+ "loss": 0.5531,
980
+ "step": 3240
981
+ },
982
+ {
983
+ "epoch": 0.54,
984
+ "learning_rate": 1.4421052631578948e-05,
985
+ "loss": 0.5547,
986
+ "step": 3260
987
+ },
988
+ {
989
+ "epoch": 0.55,
990
+ "learning_rate": 1.431578947368421e-05,
991
+ "loss": 0.5551,
992
+ "step": 3280
993
+ },
994
+ {
995
+ "epoch": 0.55,
996
+ "learning_rate": 1.4210526315789473e-05,
997
+ "loss": 0.5541,
998
+ "step": 3300
999
+ },
1000
+ {
1001
+ "epoch": 0.55,
1002
+ "learning_rate": 1.4105263157894738e-05,
1003
+ "loss": 0.5512,
1004
+ "step": 3320
1005
+ },
1006
+ {
1007
+ "epoch": 0.56,
1008
+ "learning_rate": 1.4e-05,
1009
+ "loss": 0.5531,
1010
+ "step": 3340
1011
+ },
1012
+ {
1013
+ "epoch": 0.56,
1014
+ "learning_rate": 1.3894736842105263e-05,
1015
+ "loss": 0.5504,
1016
+ "step": 3360
1017
+ },
1018
+ {
1019
+ "epoch": 0.56,
1020
+ "learning_rate": 1.3789473684210526e-05,
1021
+ "loss": 0.5508,
1022
+ "step": 3380
1023
+ },
1024
+ {
1025
+ "epoch": 0.56,
1026
+ "learning_rate": 1.368421052631579e-05,
1027
+ "loss": 0.5532,
1028
+ "step": 3400
1029
+ },
1030
+ {
1031
+ "epoch": 0.57,
1032
+ "learning_rate": 1.3578947368421053e-05,
1033
+ "loss": 0.5458,
1034
+ "step": 3420
1035
+ },
1036
+ {
1037
+ "epoch": 0.57,
1038
+ "learning_rate": 1.3473684210526316e-05,
1039
+ "loss": 0.5505,
1040
+ "step": 3440
1041
+ },
1042
+ {
1043
+ "epoch": 0.57,
1044
+ "learning_rate": 1.336842105263158e-05,
1045
+ "loss": 0.5519,
1046
+ "step": 3460
1047
+ },
1048
+ {
1049
+ "epoch": 0.58,
1050
+ "learning_rate": 1.3263157894736843e-05,
1051
+ "loss": 0.5464,
1052
+ "step": 3480
1053
+ },
1054
+ {
1055
+ "epoch": 0.58,
1056
+ "learning_rate": 1.3157894736842104e-05,
1057
+ "loss": 0.5488,
1058
+ "step": 3500
1059
+ },
1060
+ {
1061
+ "epoch": 0.58,
1062
+ "learning_rate": 1.305263157894737e-05,
1063
+ "loss": 0.5475,
1064
+ "step": 3520
1065
+ },
1066
+ {
1067
+ "epoch": 0.59,
1068
+ "learning_rate": 1.2947368421052633e-05,
1069
+ "loss": 0.548,
1070
+ "step": 3540
1071
+ },
1072
+ {
1073
+ "epoch": 0.59,
1074
+ "learning_rate": 1.2842105263157896e-05,
1075
+ "loss": 0.5449,
1076
+ "step": 3560
1077
+ },
1078
+ {
1079
+ "epoch": 0.59,
1080
+ "learning_rate": 1.2736842105263157e-05,
1081
+ "loss": 0.5447,
1082
+ "step": 3580
1083
+ },
1084
+ {
1085
+ "epoch": 0.6,
1086
+ "learning_rate": 1.263157894736842e-05,
1087
+ "loss": 0.5421,
1088
+ "step": 3600
1089
+ },
1090
+ {
1091
+ "epoch": 0.6,
1092
+ "learning_rate": 1.2526315789473686e-05,
1093
+ "loss": 0.5487,
1094
+ "step": 3620
1095
+ },
1096
+ {
1097
+ "epoch": 0.6,
1098
+ "learning_rate": 1.2421052631578947e-05,
1099
+ "loss": 0.5441,
1100
+ "step": 3640
1101
+ },
1102
+ {
1103
+ "epoch": 0.61,
1104
+ "learning_rate": 1.231578947368421e-05,
1105
+ "loss": 0.5404,
1106
+ "step": 3660
1107
+ },
1108
+ {
1109
+ "epoch": 0.61,
1110
+ "learning_rate": 1.2210526315789474e-05,
1111
+ "loss": 0.5386,
1112
+ "step": 3680
1113
+ },
1114
+ {
1115
+ "epoch": 0.61,
1116
+ "learning_rate": 1.2105263157894737e-05,
1117
+ "loss": 0.5452,
1118
+ "step": 3700
1119
+ },
1120
+ {
1121
+ "epoch": 0.62,
1122
+ "learning_rate": 1.2e-05,
1123
+ "loss": 0.5417,
1124
+ "step": 3720
1125
+ },
1126
+ {
1127
+ "epoch": 0.62,
1128
+ "learning_rate": 1.1894736842105264e-05,
1129
+ "loss": 0.5419,
1130
+ "step": 3740
1131
+ },
1132
+ {
1133
+ "epoch": 0.62,
1134
+ "learning_rate": 1.1789473684210527e-05,
1135
+ "loss": 0.5409,
1136
+ "step": 3760
1137
+ },
1138
+ {
1139
+ "epoch": 0.63,
1140
+ "learning_rate": 1.168421052631579e-05,
1141
+ "loss": 0.5341,
1142
+ "step": 3780
1143
+ },
1144
+ {
1145
+ "epoch": 0.63,
1146
+ "learning_rate": 1.1578947368421052e-05,
1147
+ "loss": 0.5392,
1148
+ "step": 3800
1149
+ },
1150
+ {
1151
+ "epoch": 0.63,
1152
+ "learning_rate": 1.1473684210526317e-05,
1153
+ "loss": 0.5393,
1154
+ "step": 3820
1155
+ },
1156
+ {
1157
+ "epoch": 0.64,
1158
+ "learning_rate": 1.136842105263158e-05,
1159
+ "loss": 0.5363,
1160
+ "step": 3840
1161
+ },
1162
+ {
1163
+ "epoch": 0.64,
1164
+ "learning_rate": 1.1263157894736843e-05,
1165
+ "loss": 0.5384,
1166
+ "step": 3860
1167
+ },
1168
+ {
1169
+ "epoch": 0.64,
1170
+ "learning_rate": 1.1157894736842105e-05,
1171
+ "loss": 0.5415,
1172
+ "step": 3880
1173
+ },
1174
+ {
1175
+ "epoch": 0.65,
1176
+ "learning_rate": 1.1052631578947368e-05,
1177
+ "loss": 0.5321,
1178
+ "step": 3900
1179
+ },
1180
+ {
1181
+ "epoch": 0.65,
1182
+ "learning_rate": 1.0947368421052633e-05,
1183
+ "loss": 0.5342,
1184
+ "step": 3920
1185
+ },
1186
+ {
1187
+ "epoch": 0.65,
1188
+ "learning_rate": 1.0842105263157895e-05,
1189
+ "loss": 0.5312,
1190
+ "step": 3940
1191
+ },
1192
+ {
1193
+ "epoch": 0.66,
1194
+ "learning_rate": 1.0736842105263158e-05,
1195
+ "loss": 0.536,
1196
+ "step": 3960
1197
+ },
1198
+ {
1199
+ "epoch": 0.66,
1200
+ "learning_rate": 1.0631578947368421e-05,
1201
+ "loss": 0.5314,
1202
+ "step": 3980
1203
+ },
1204
+ {
1205
+ "epoch": 0.66,
1206
+ "learning_rate": 1.0526315789473684e-05,
1207
+ "loss": 0.536,
1208
+ "step": 4000
1209
+ },
1210
+ {
1211
+ "epoch": 0.67,
1212
+ "learning_rate": 1.0421052631578948e-05,
1213
+ "loss": 0.5336,
1214
+ "step": 4020
1215
+ },
1216
+ {
1217
+ "epoch": 0.67,
1218
+ "learning_rate": 1.0315789473684211e-05,
1219
+ "loss": 0.5343,
1220
+ "step": 4040
1221
+ },
1222
+ {
1223
+ "epoch": 0.67,
1224
+ "learning_rate": 1.0210526315789474e-05,
1225
+ "loss": 0.5287,
1226
+ "step": 4060
1227
+ },
1228
+ {
1229
+ "epoch": 0.68,
1230
+ "learning_rate": 1.0105263157894738e-05,
1231
+ "loss": 0.5316,
1232
+ "step": 4080
1233
+ },
1234
+ {
1235
+ "epoch": 0.68,
1236
+ "learning_rate": 9.999999999999999e-06,
1237
+ "loss": 0.5266,
1238
+ "step": 4100
1239
+ },
1240
+ {
1241
+ "epoch": 0.68,
1242
+ "learning_rate": 9.894736842105264e-06,
1243
+ "loss": 0.5249,
1244
+ "step": 4120
1245
+ },
1246
+ {
1247
+ "epoch": 0.69,
1248
+ "learning_rate": 9.789473684210527e-06,
1249
+ "loss": 0.5255,
1250
+ "step": 4140
1251
+ },
1252
+ {
1253
+ "epoch": 0.69,
1254
+ "learning_rate": 9.68421052631579e-06,
1255
+ "loss": 0.5266,
1256
+ "step": 4160
1257
+ },
1258
+ {
1259
+ "epoch": 0.69,
1260
+ "learning_rate": 9.578947368421052e-06,
1261
+ "loss": 0.5234,
1262
+ "step": 4180
1263
+ },
1264
+ {
1265
+ "epoch": 0.7,
1266
+ "learning_rate": 9.473684210526315e-06,
1267
+ "loss": 0.5264,
1268
+ "step": 4200
1269
+ },
1270
+ {
1271
+ "epoch": 0.7,
1272
+ "learning_rate": 9.36842105263158e-06,
1273
+ "loss": 0.532,
1274
+ "step": 4220
1275
+ },
1276
+ {
1277
+ "epoch": 0.7,
1278
+ "learning_rate": 9.263157894736842e-06,
1279
+ "loss": 0.5227,
1280
+ "step": 4240
1281
+ },
1282
+ {
1283
+ "epoch": 0.71,
1284
+ "learning_rate": 9.157894736842105e-06,
1285
+ "loss": 0.5279,
1286
+ "step": 4260
1287
+ },
1288
+ {
1289
+ "epoch": 0.71,
1290
+ "learning_rate": 9.052631578947369e-06,
1291
+ "loss": 0.5242,
1292
+ "step": 4280
1293
+ },
1294
+ {
1295
+ "epoch": 0.71,
1296
+ "learning_rate": 8.947368421052632e-06,
1297
+ "loss": 0.5195,
1298
+ "step": 4300
1299
+ },
1300
+ {
1301
+ "epoch": 0.72,
1302
+ "learning_rate": 8.842105263157893e-06,
1303
+ "loss": 0.5279,
1304
+ "step": 4320
1305
+ },
1306
+ {
1307
+ "epoch": 0.72,
1308
+ "learning_rate": 8.736842105263158e-06,
1309
+ "loss": 0.5259,
1310
+ "step": 4340
1311
+ },
1312
+ {
1313
+ "epoch": 0.72,
1314
+ "learning_rate": 8.631578947368422e-06,
1315
+ "loss": 0.5231,
1316
+ "step": 4360
1317
+ },
1318
+ {
1319
+ "epoch": 0.73,
1320
+ "learning_rate": 8.526315789473685e-06,
1321
+ "loss": 0.528,
1322
+ "step": 4380
1323
+ },
1324
+ {
1325
+ "epoch": 0.73,
1326
+ "learning_rate": 8.421052631578947e-06,
1327
+ "loss": 0.5194,
1328
+ "step": 4400
1329
+ },
1330
+ {
1331
+ "epoch": 0.73,
1332
+ "learning_rate": 8.31578947368421e-06,
1333
+ "loss": 0.5218,
1334
+ "step": 4420
1335
+ },
1336
+ {
1337
+ "epoch": 0.74,
1338
+ "learning_rate": 8.210526315789475e-06,
1339
+ "loss": 0.5218,
1340
+ "step": 4440
1341
+ },
1342
+ {
1343
+ "epoch": 0.74,
1344
+ "learning_rate": 8.105263157894738e-06,
1345
+ "loss": 0.5198,
1346
+ "step": 4460
1347
+ },
1348
+ {
1349
+ "epoch": 0.74,
1350
+ "learning_rate": 8e-06,
1351
+ "loss": 0.5197,
1352
+ "step": 4480
1353
+ },
1354
+ {
1355
+ "epoch": 0.75,
1356
+ "learning_rate": 7.894736842105263e-06,
1357
+ "loss": 0.5221,
1358
+ "step": 4500
1359
+ },
1360
+ {
1361
+ "epoch": 0.75,
1362
+ "learning_rate": 7.789473684210526e-06,
1363
+ "loss": 0.5198,
1364
+ "step": 4520
1365
+ },
1366
+ {
1367
+ "epoch": 0.75,
1368
+ "learning_rate": 7.68421052631579e-06,
1369
+ "loss": 0.5254,
1370
+ "step": 4540
1371
+ },
1372
+ {
1373
+ "epoch": 0.76,
1374
+ "learning_rate": 7.578947368421053e-06,
1375
+ "loss": 0.5196,
1376
+ "step": 4560
1377
+ },
1378
+ {
1379
+ "epoch": 0.76,
1380
+ "learning_rate": 7.473684210526316e-06,
1381
+ "loss": 0.5161,
1382
+ "step": 4580
1383
+ },
1384
+ {
1385
+ "epoch": 0.76,
1386
+ "learning_rate": 7.3684210526315784e-06,
1387
+ "loss": 0.5212,
1388
+ "step": 4600
1389
+ },
1390
+ {
1391
+ "epoch": 0.77,
1392
+ "learning_rate": 7.2631578947368426e-06,
1393
+ "loss": 0.5211,
1394
+ "step": 4620
1395
+ },
1396
+ {
1397
+ "epoch": 0.77,
1398
+ "learning_rate": 7.157894736842105e-06,
1399
+ "loss": 0.5162,
1400
+ "step": 4640
1401
+ },
1402
+ {
1403
+ "epoch": 0.77,
1404
+ "learning_rate": 7.052631578947369e-06,
1405
+ "loss": 0.52,
1406
+ "step": 4660
1407
+ },
1408
+ {
1409
+ "epoch": 0.78,
1410
+ "learning_rate": 6.9473684210526315e-06,
1411
+ "loss": 0.5069,
1412
+ "step": 4680
1413
+ },
1414
+ {
1415
+ "epoch": 0.78,
1416
+ "learning_rate": 6.842105263157895e-06,
1417
+ "loss": 0.5132,
1418
+ "step": 4700
1419
+ },
1420
+ {
1421
+ "epoch": 0.78,
1422
+ "learning_rate": 6.736842105263158e-06,
1423
+ "loss": 0.5174,
1424
+ "step": 4720
1425
+ },
1426
+ {
1427
+ "epoch": 0.79,
1428
+ "learning_rate": 6.631578947368421e-06,
1429
+ "loss": 0.5146,
1430
+ "step": 4740
1431
+ },
1432
+ {
1433
+ "epoch": 0.79,
1434
+ "learning_rate": 6.526315789473685e-06,
1435
+ "loss": 0.5102,
1436
+ "step": 4760
1437
+ },
1438
+ {
1439
+ "epoch": 0.79,
1440
+ "learning_rate": 6.421052631578948e-06,
1441
+ "loss": 0.516,
1442
+ "step": 4780
1443
+ },
1444
+ {
1445
+ "epoch": 0.8,
1446
+ "learning_rate": 6.31578947368421e-06,
1447
+ "loss": 0.5122,
1448
+ "step": 4800
1449
+ },
1450
+ {
1451
+ "epoch": 0.8,
1452
+ "learning_rate": 6.210526315789474e-06,
1453
+ "loss": 0.512,
1454
+ "step": 4820
1455
+ },
1456
+ {
1457
+ "epoch": 0.8,
1458
+ "learning_rate": 6.105263157894737e-06,
1459
+ "loss": 0.5128,
1460
+ "step": 4840
1461
+ },
1462
+ {
1463
+ "epoch": 0.81,
1464
+ "learning_rate": 6e-06,
1465
+ "loss": 0.5189,
1466
+ "step": 4860
1467
+ },
1468
+ {
1469
+ "epoch": 0.81,
1470
+ "learning_rate": 5.8947368421052634e-06,
1471
+ "loss": 0.5125,
1472
+ "step": 4880
1473
+ },
1474
+ {
1475
+ "epoch": 0.81,
1476
+ "learning_rate": 5.789473684210526e-06,
1477
+ "loss": 0.5127,
1478
+ "step": 4900
1479
+ },
1480
+ {
1481
+ "epoch": 0.82,
1482
+ "learning_rate": 5.68421052631579e-06,
1483
+ "loss": 0.5158,
1484
+ "step": 4920
1485
+ },
1486
+ {
1487
+ "epoch": 0.82,
1488
+ "learning_rate": 5.578947368421052e-06,
1489
+ "loss": 0.5124,
1490
+ "step": 4940
1491
+ },
1492
+ {
1493
+ "epoch": 0.82,
1494
+ "learning_rate": 5.4736842105263165e-06,
1495
+ "loss": 0.511,
1496
+ "step": 4960
1497
+ },
1498
+ {
1499
+ "epoch": 0.83,
1500
+ "learning_rate": 5.368421052631579e-06,
1501
+ "loss": 0.5134,
1502
+ "step": 4980
1503
+ },
1504
+ {
1505
+ "epoch": 0.83,
1506
+ "learning_rate": 5.263157894736842e-06,
1507
+ "loss": 0.5091,
1508
+ "step": 5000
1509
+ },
1510
+ {
1511
+ "epoch": 0.83,
1512
+ "learning_rate": 5.1578947368421055e-06,
1513
+ "loss": 0.5127,
1514
+ "step": 5020
1515
+ },
1516
+ {
1517
+ "epoch": 0.84,
1518
+ "learning_rate": 5.052631578947369e-06,
1519
+ "loss": 0.5075,
1520
+ "step": 5040
1521
+ },
1522
+ {
1523
+ "epoch": 0.84,
1524
+ "learning_rate": 4.947368421052632e-06,
1525
+ "loss": 0.5131,
1526
+ "step": 5060
1527
+ },
1528
+ {
1529
+ "epoch": 0.84,
1530
+ "learning_rate": 4.842105263157895e-06,
1531
+ "loss": 0.5103,
1532
+ "step": 5080
1533
+ },
1534
+ {
1535
+ "epoch": 0.85,
1536
+ "learning_rate": 4.736842105263158e-06,
1537
+ "loss": 0.5087,
1538
+ "step": 5100
1539
+ },
1540
+ {
1541
+ "epoch": 0.85,
1542
+ "learning_rate": 4.631578947368421e-06,
1543
+ "loss": 0.5105,
1544
+ "step": 5120
1545
+ },
1546
+ {
1547
+ "epoch": 0.85,
1548
+ "learning_rate": 4.526315789473684e-06,
1549
+ "loss": 0.5069,
1550
+ "step": 5140
1551
+ },
1552
+ {
1553
+ "epoch": 0.86,
1554
+ "learning_rate": 4.421052631578947e-06,
1555
+ "loss": 0.5152,
1556
+ "step": 5160
1557
+ },
1558
+ {
1559
+ "epoch": 0.86,
1560
+ "learning_rate": 4.315789473684211e-06,
1561
+ "loss": 0.5118,
1562
+ "step": 5180
1563
+ },
1564
+ {
1565
+ "epoch": 0.86,
1566
+ "learning_rate": 4.210526315789473e-06,
1567
+ "loss": 0.5138,
1568
+ "step": 5200
1569
+ },
1570
+ {
1571
+ "epoch": 0.87,
1572
+ "learning_rate": 4.105263157894737e-06,
1573
+ "loss": 0.5112,
1574
+ "step": 5220
1575
+ },
1576
+ {
1577
+ "epoch": 0.87,
1578
+ "learning_rate": 4e-06,
1579
+ "loss": 0.5069,
1580
+ "step": 5240
1581
+ },
1582
+ {
1583
+ "epoch": 0.87,
1584
+ "learning_rate": 3.894736842105263e-06,
1585
+ "loss": 0.5086,
1586
+ "step": 5260
1587
+ },
1588
+ {
1589
+ "epoch": 0.88,
1590
+ "learning_rate": 3.7894736842105264e-06,
1591
+ "loss": 0.5102,
1592
+ "step": 5280
1593
+ },
1594
+ {
1595
+ "epoch": 0.88,
1596
+ "learning_rate": 3.6842105263157892e-06,
1597
+ "loss": 0.5105,
1598
+ "step": 5300
1599
+ },
1600
+ {
1601
+ "epoch": 0.88,
1602
+ "learning_rate": 3.5789473684210525e-06,
1603
+ "loss": 0.5072,
1604
+ "step": 5320
1605
+ },
1606
+ {
1607
+ "epoch": 0.89,
1608
+ "learning_rate": 3.4736842105263158e-06,
1609
+ "loss": 0.5076,
1610
+ "step": 5340
1611
+ },
1612
+ {
1613
+ "epoch": 0.89,
1614
+ "learning_rate": 3.368421052631579e-06,
1615
+ "loss": 0.509,
1616
+ "step": 5360
1617
+ },
1618
+ {
1619
+ "epoch": 0.89,
1620
+ "learning_rate": 3.2631578947368423e-06,
1621
+ "loss": 0.5069,
1622
+ "step": 5380
1623
+ },
1624
+ {
1625
+ "epoch": 0.9,
1626
+ "learning_rate": 3.157894736842105e-06,
1627
+ "loss": 0.5042,
1628
+ "step": 5400
1629
+ },
1630
+ {
1631
+ "epoch": 0.9,
1632
+ "learning_rate": 3.0526315789473684e-06,
1633
+ "loss": 0.5134,
1634
+ "step": 5420
1635
+ },
1636
+ {
1637
+ "epoch": 0.9,
1638
+ "learning_rate": 2.9473684210526317e-06,
1639
+ "loss": 0.502,
1640
+ "step": 5440
1641
+ },
1642
+ {
1643
+ "epoch": 0.91,
1644
+ "learning_rate": 2.842105263157895e-06,
1645
+ "loss": 0.5044,
1646
+ "step": 5460
1647
+ },
1648
+ {
1649
+ "epoch": 0.91,
1650
+ "learning_rate": 2.7368421052631583e-06,
1651
+ "loss": 0.5102,
1652
+ "step": 5480
1653
+ },
1654
+ {
1655
+ "epoch": 0.91,
1656
+ "learning_rate": 2.631578947368421e-06,
1657
+ "loss": 0.5061,
1658
+ "step": 5500
1659
+ },
1660
+ {
1661
+ "epoch": 0.92,
1662
+ "learning_rate": 2.5263157894736844e-06,
1663
+ "loss": 0.4996,
1664
+ "step": 5520
1665
+ },
1666
+ {
1667
+ "epoch": 0.92,
1668
+ "learning_rate": 2.4210526315789477e-06,
1669
+ "loss": 0.5049,
1670
+ "step": 5540
1671
+ },
1672
+ {
1673
+ "epoch": 0.92,
1674
+ "learning_rate": 2.3157894736842105e-06,
1675
+ "loss": 0.5061,
1676
+ "step": 5560
1677
+ },
1678
+ {
1679
+ "epoch": 0.93,
1680
+ "learning_rate": 2.2105263157894734e-06,
1681
+ "loss": 0.5041,
1682
+ "step": 5580
1683
+ },
1684
+ {
1685
+ "epoch": 0.93,
1686
+ "learning_rate": 2.1052631578947366e-06,
1687
+ "loss": 0.5021,
1688
+ "step": 5600
1689
+ },
1690
+ {
1691
+ "epoch": 0.93,
1692
+ "learning_rate": 2e-06,
1693
+ "loss": 0.5033,
1694
+ "step": 5620
1695
+ },
1696
+ {
1697
+ "epoch": 0.94,
1698
+ "learning_rate": 1.8947368421052632e-06,
1699
+ "loss": 0.5013,
1700
+ "step": 5640
1701
+ },
1702
+ {
1703
+ "epoch": 0.94,
1704
+ "learning_rate": 1.7894736842105262e-06,
1705
+ "loss": 0.5053,
1706
+ "step": 5660
1707
+ },
1708
+ {
1709
+ "epoch": 0.94,
1710
+ "learning_rate": 1.6842105263157895e-06,
1711
+ "loss": 0.5026,
1712
+ "step": 5680
1713
+ },
1714
+ {
1715
+ "epoch": 0.95,
1716
+ "learning_rate": 1.5789473684210526e-06,
1717
+ "loss": 0.5042,
1718
+ "step": 5700
1719
+ },
1720
+ {
1721
+ "epoch": 0.95,
1722
+ "learning_rate": 1.4736842105263159e-06,
1723
+ "loss": 0.5026,
1724
+ "step": 5720
1725
+ },
1726
+ {
1727
+ "epoch": 0.95,
1728
+ "learning_rate": 1.3684210526315791e-06,
1729
+ "loss": 0.5021,
1730
+ "step": 5740
1731
+ },
1732
+ {
1733
+ "epoch": 0.96,
1734
+ "learning_rate": 1.2631578947368422e-06,
1735
+ "loss": 0.4996,
1736
+ "step": 5760
1737
+ },
1738
+ {
1739
+ "epoch": 0.96,
1740
+ "learning_rate": 1.1578947368421053e-06,
1741
+ "loss": 0.503,
1742
+ "step": 5780
1743
+ },
1744
+ {
1745
+ "epoch": 0.96,
1746
+ "learning_rate": 1.0526315789473683e-06,
1747
+ "loss": 0.4996,
1748
+ "step": 5800
1749
+ },
1750
+ {
1751
+ "epoch": 0.97,
1752
+ "learning_rate": 9.473684210526316e-07,
1753
+ "loss": 0.5038,
1754
+ "step": 5820
1755
+ },
1756
+ {
1757
+ "epoch": 0.97,
1758
+ "learning_rate": 8.421052631578948e-07,
1759
+ "loss": 0.5025,
1760
+ "step": 5840
1761
+ },
1762
+ {
1763
+ "epoch": 0.97,
1764
+ "learning_rate": 7.368421052631579e-07,
1765
+ "loss": 0.5025,
1766
+ "step": 5860
1767
+ },
1768
+ {
1769
+ "epoch": 0.98,
1770
+ "learning_rate": 6.315789473684211e-07,
1771
+ "loss": 0.5061,
1772
+ "step": 5880
1773
+ },
1774
+ {
1775
+ "epoch": 0.98,
1776
+ "learning_rate": 5.263157894736842e-07,
1777
+ "loss": 0.4971,
1778
+ "step": 5900
1779
+ },
1780
+ {
1781
+ "epoch": 0.98,
1782
+ "learning_rate": 4.210526315789474e-07,
1783
+ "loss": 0.505,
1784
+ "step": 5920
1785
+ },
1786
+ {
1787
+ "epoch": 0.99,
1788
+ "learning_rate": 3.1578947368421055e-07,
1789
+ "loss": 0.4992,
1790
+ "step": 5940
1791
+ },
1792
+ {
1793
+ "epoch": 0.99,
1794
+ "learning_rate": 2.105263157894737e-07,
1795
+ "loss": 0.4982,
1796
+ "step": 5960
1797
+ },
1798
+ {
1799
+ "epoch": 0.99,
1800
+ "learning_rate": 1.0526315789473685e-07,
1801
+ "loss": 0.5003,
1802
+ "step": 5980
1803
+ },
1804
+ {
1805
+ "epoch": 1.0,
1806
+ "learning_rate": 0.0,
1807
+ "loss": 0.5002,
1808
+ "step": 6000
1809
+ }
1810
+ ],
1811
+ "max_steps": 6000,
1812
+ "num_train_epochs": 1,
1813
+ "total_flos": 0.0,
1814
+ "trial_name": null,
1815
+ "trial_params": null
1816
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bcbce3fde39a52c7cc6a8668cea0e9ffa0e3427e17a608ed54bbe76f54caaa3a
3
+ size 3640
vocab.json ADDED
The diff for this file is too large to render. See raw diff