Hugging Face's logo

zxh4546
/
unsup-wc-s64-bs128-lr6

Text Classification
Transformers
PyTorch
pixel
Generated from Trainer
Inference Endpoints
Model card Files Community
Zhuoxu Huang commited on
Commit
dedc425
•
0 Parent(s):

init

Browse files
Files changed (10) hide show
  1. .gitattributes +3 -0
  2. 49e6dc219d1a1a1c9236acaf05a48b542002016a6dc877ee72baab085a84257b.3f28e7f4b38e1efe1b6da4a3732404c19d4c6a614ff32dce90a251e293d4ce58 +3 -0
  3. README.md +54 -0
  4. all_results.json +8 -0
  5. config.json +43 -0
  6. pytorch_model.bin +3 -0
  7. text_renderer_config.json +239 -0
  8. train_results.json +8 -0
  9. trainer_state.json +481 -0
  10. training_args.bin +3 -0
.gitattributes ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
2
+ training_args.bin filter=lfs diff=lfs merge=lfs -text
3
+ 49e6dc219d1a1a1c9236acaf05a48b542002016a6dc877ee72baab085a84257b.3f28e7f4b38e1efe1b6da4a3732404c19d4c6a614ff32dce90a251e293d4ce58 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
README.md ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ datasets:
6
+ - unsup-wc
7
+ model-index:
8
+ - name: contrastive-unsup-wc-pixel-base-mean-64-128-1-3e-6-7600-42-eval
9
+ results: []
10
+ ---
11
+
12
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
13
+ should probably proofread and complete it, then remove this comment. -->
14
+
15
+ # contrastive-unsup-wc-pixel-base-mean-64-128-1-3e-6-7600-42-eval
16
+
17
+ This model is a fine-tuned version of [Team-PIXEL/pixel-base](https://huggingface.co/Team-PIXEL/pixel-base) on the UNSUP-WC dataset.
18
+
19
+ ## Model description
20
+
21
+ More information needed
22
+
23
+ ## Intended uses & limitations
24
+
25
+ More information needed
26
+
27
+ ## Training and evaluation data
28
+
29
+ More information needed
30
+
31
+ ## Training procedure
32
+
33
+ ### Training hyperparameters
34
+
35
+ The following hyperparameters were used during training:
36
+ - learning_rate: 3e-06
37
+ - train_batch_size: 128
38
+ - eval_batch_size: 16
39
+ - seed: 42
40
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
41
+ - lr_scheduler_type: linear
42
+ - lr_scheduler_warmup_steps: 500
43
+ - training_steps: 7600
44
+
45
+ ### Training results
46
+
47
+
48
+
49
+ ### Framework versions
50
+
51
+ - Transformers 4.17.0
52
+ - Pytorch 1.11.0
53
+ - Datasets 2.14.7.dev0
54
+ - Tokenizers 0.14.1
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 8.470750703670477,
4
+ "train_runtime": 4458.002,
5
+ "train_samples": 974010,
6
+ "train_samples_per_second": 218.214,
7
+ "train_steps_per_second": 1.705
8
+ }
config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "Team-PIXEL/pixel-base",
3
+ "architectures": [
4
+ "PIXELForSequenceClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "decoder_hidden_size": 512,
8
+ "decoder_intermediate_size": 2048,
9
+ "decoder_num_attention_heads": 16,
10
+ "decoder_num_hidden_layers": 8,
11
+ "finetuning_task": "unsup-wc",
12
+ "hidden_act": "gelu",
13
+ "hidden_dropout_prob": 0.1,
14
+ "hidden_size": 768,
15
+ "id2label": {
16
+ "0": "entailment",
17
+ "1": "neutral",
18
+ "2": "contradiction"
19
+ },
20
+ "image_size": [
21
+ 16,
22
+ 1024
23
+ ],
24
+ "initializer_range": 0.02,
25
+ "intermediate_size": 3072,
26
+ "interpolate_pos_encoding": false,
27
+ "label2id": {
28
+ "contradiction": 2,
29
+ "entailment": 0,
30
+ "neutral": 1
31
+ },
32
+ "layer_norm_eps": 1e-12,
33
+ "mask_ratio": 0.25,
34
+ "model_type": "pixel",
35
+ "norm_pix_loss": true,
36
+ "num_attention_heads": 12,
37
+ "num_channels": 3,
38
+ "num_hidden_layers": 12,
39
+ "patch_size": 16,
40
+ "qkv_bias": true,
41
+ "torch_dtype": "float32",
42
+ "transformers_version": "4.17.0"
43
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b993d2bdafab9134b9264c4b81b45880c9a5de1fabccb6ad43def785c83f766d
3
+ size 345238445
text_renderer_config.json ADDED
@@ -0,0 +1,239 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "C059",
10
+ "D050000L",
11
+ "DejaVu Math TeX Gyre",
12
+ "DejaVu Sans",
13
+ "DejaVu Sans Mono",
14
+ "DejaVu Serif",
15
+ "Droid Sans Fallback",
16
+ "Go Noto Current",
17
+ "Inconsolata",
18
+ "Monospace",
19
+ "Nimbus Mono PS",
20
+ "Nimbus Roman",
21
+ "Nimbus Sans",
22
+ "Nimbus Sans Narrow",
23
+ "Noto Fangsong KSS Rotated",
24
+ "Noto Fangsong KSS Vertical",
25
+ "Noto Kufi Arabic",
26
+ "Noto Mono",
27
+ "Noto Music",
28
+ "Noto Naskh Arabic",
29
+ "Noto Naskh Arabic UI",
30
+ "Noto Nastaliq Urdu",
31
+ "Noto Rashi Hebrew",
32
+ "Noto Sans",
33
+ "Noto Sans Adlam",
34
+ "Noto Sans Adlam Unjoined",
35
+ "Noto Sans Anatolian Hieroglyphs",
36
+ "Noto Sans Arabic",
37
+ "Noto Sans Arabic UI",
38
+ "Noto Sans Armenian",
39
+ "Noto Sans Avestan",
40
+ "Noto Sans Balinese",
41
+ "Noto Sans Bamum",
42
+ "Noto Sans Bassa Vah",
43
+ "Noto Sans Batak",
44
+ "Noto Sans Bengali",
45
+ "Noto Sans Bengali UI",
46
+ "Noto Sans Bhaiksuki",
47
+ "Noto Sans Brahmi",
48
+ "Noto Sans Buginese",
49
+ "Noto Sans Buhid",
50
+ "Noto Sans CJK HK",
51
+ "Noto Sans CJK JP",
52
+ "Noto Sans CJK KR",
53
+ "Noto Sans CJK SC",
54
+ "Noto Sans CJK TC",
55
+ "Noto Sans Canadian Aboriginal",
56
+ "Noto Sans Carian",
57
+ "Noto Sans Caucasian Albanian",
58
+ "Noto Sans Chakma",
59
+ "Noto Sans Cham",
60
+ "Noto Sans Cherokee",
61
+ "Noto Sans Chorasmian",
62
+ "Noto Sans Coptic",
63
+ "Noto Sans Cuneiform",
64
+ "Noto Sans Cypriot",
65
+ "Noto Sans Cypro Minoan",
66
+ "Noto Sans Deseret",
67
+ "Noto Sans Devanagari",
68
+ "Noto Sans Devanagari UI",
69
+ "Noto Sans Duployan",
70
+ "Noto Sans Egyptian Hieroglyphs",
71
+ "Noto Sans Elbasan",
72
+ "Noto Sans Elymaic",
73
+ "Noto Sans Ethiopic",
74
+ "Noto Sans Georgian",
75
+ "Noto Sans Glagolitic",
76
+ "Noto Sans Gothic",
77
+ "Noto Sans Grantha",
78
+ "Noto Sans Gujarati",
79
+ "Noto Sans Gujarati UI",
80
+ "Noto Sans Gunjala Gondi",
81
+ "Noto Sans Gurmukhi",
82
+ "Noto Sans Gurmukhi UI",
83
+ "Noto Sans Hanifi Rohingya",
84
+ "Noto Sans Hanunoo",
85
+ "Noto Sans Hatran",
86
+ "Noto Sans Hebrew",
87
+ "Noto Sans Hebrew Droid",
88
+ "Noto Sans Imperial Aramaic",
89
+ "Noto Sans Indic Siyaq Numbers",
90
+ "Noto Sans Inscriptional Pahlavi",
91
+ "Noto Sans Inscriptional Parthian",
92
+ "Noto Sans Javanese",
93
+ "Noto Sans Kaithi",
94
+ "Noto Sans Kannada",
95
+ "Noto Sans Kannada UI",
96
+ "Noto Sans Kawi",
97
+ "Noto Sans Kayah Li",
98
+ "Noto Sans Kharoshthi",
99
+ "Noto Sans Khmer",
100
+ "Noto Sans Khmer UI",
101
+ "Noto Sans Khojki",
102
+ "Noto Sans Khudawadi",
103
+ "Noto Sans Lao",
104
+ "Noto Sans Lao Looped",
105
+ "Noto Sans Lao Looped UI",
106
+ "Noto Sans Lao UI",
107
+ "Noto Sans Lepcha",
108
+ "Noto Sans Limbu",
109
+ "Noto Sans Linear A",
110
+ "Noto Sans Linear B",
111
+ "Noto Sans Lisu",
112
+ "Noto Sans Lycian",
113
+ "Noto Sans Lydian",
114
+ "Noto Sans Mahajani",
115
+ "Noto Sans Malayalam",
116
+ "Noto Sans Malayalam UI",
117
+ "Noto Sans Mandaic",
118
+ "Noto Sans Manichaean",
119
+ "Noto Sans Marchen",
120
+ "Noto Sans Masaram Gondi",
121
+ "Noto Sans Math",
122
+ "Noto Sans Mayan Numerals",
123
+ "Noto Sans Medefaidrin",
124
+ "Noto Sans Meetei Mayek",
125
+ "Noto Sans Mende Kikakui",
126
+ "Noto Sans Meroitic",
127
+ "Noto Sans Miao",
128
+ "Noto Sans Modi",
129
+ "Noto Sans Mongolian",
130
+ "Noto Sans Mono",
131
+ "Noto Sans Mro",
132
+ "Noto Sans Multani",
133
+ "Noto Sans Myanmar",
134
+ "Noto Sans Myanmar UI",
135
+ "Noto Sans NKo",
136
+ "Noto Sans NKo Unjoined",
137
+ "Noto Sans Nabataean",
138
+ "Noto Sans Nag Mundari",
139
+ "Noto Sans Nandinagari",
140
+ "Noto Sans New Tai Lue",
141
+ "Noto Sans Newa",
142
+ "Noto Sans Nushu",
143
+ "Noto Sans Ogham",
144
+ "Noto Sans Ol Chiki",
145
+ "Noto Sans Old Hungarian",
146
+ "Noto Sans Old Italic",
147
+ "Noto Sans Old North Arabian",
148
+ "Noto Sans Old Permic",
149
+ "Noto Sans Old Persian",
150
+ "Noto Sans Old Sogdian",
151
+ "Noto Sans Old South Arabian",
152
+ "Noto Sans Old Turkic",
153
+ "Noto Sans Oriya",
154
+ "Noto Sans Osage",
155
+ "Noto Sans Osmanya",
156
+ "Noto Sans Pahawh Hmong",
157
+ "Noto Sans Palmyrene",
158
+ "Noto Sans Pau Cin Hau",
159
+ "Noto Sans Phags-Pa",
160
+ "Noto Sans Phoenician",
161
+ "Noto Sans Psalter Pahlavi",
162
+ "Noto Sans Rejang",
163
+ "Noto Sans Runic",
164
+ "Noto Sans Samaritan",
165
+ "Noto Sans Saurashtra",
166
+ "Noto Sans Sharada",
167
+ "Noto Sans Shavian",
168
+ "Noto Sans Siddham",
169
+ "Noto Sans SignWriting",
170
+ "Noto Sans Sinhala",
171
+ "Noto Sans Sinhala UI",
172
+ "Noto Sans Sogdian",
173
+ "Noto Sans Sora Sompeng",
174
+ "Noto Sans Soyombo",
175
+ "Noto Sans Sundanese",
176
+ "Noto Sans Syloti Nagri",
177
+ "Noto Sans Symbols",
178
+ "Noto Sans Symbols 2",
179
+ "Noto Sans Syriac",
180
+ "Noto Sans Syriac Eastern",
181
+ "Noto Sans Syriac Western",
182
+ "Noto Sans Tagalog",
183
+ "Noto Sans Tagbanwa",
184
+ "Noto Sans Tai Le",
185
+ "Noto Sans Tai Tham",
186
+ "Noto Sans Tai Viet",
187
+ "Noto Sans Takri",
188
+ "Noto Sans Tamil",
189
+ "Noto Sans Tamil Supplement",
190
+ "Noto Sans Tamil UI",
191
+ "Noto Sans Tangsa",
192
+ "Noto Sans Telugu",
193
+ "Noto Sans Telugu UI",
194
+ "Noto Sans Test",
195
+ "Noto Sans Thaana",
196
+ "Noto Sans Thai",
197
+ "Noto Sans Thai Looped",
198
+ "Noto Sans Thai Looped UI",
199
+ "Noto Sans Thai UI",
200
+ "Noto Sans Tifinagh",
201
+ "Noto Sans Tifinagh APT",
202
+ "Noto Sans Tifinagh Adrar",
203
+ "Noto Sans Tifinagh Agraw Imazighen",
204
+ "Noto Sans Tifinagh Ahaggar",
205
+ "Noto Sans Tifinagh Air",
206
+ "Noto Sans Tifinagh Azawagh",
207
+ "Noto Sans Tifinagh Ghat",
208
+ "Noto Sans Tifinagh Hawad",
209
+ "Noto Sans Tifinagh Rhissa Ixa",
210
+ "Noto Sans Tifinagh SIL",
211
+ "Noto Sans Tifinagh Tawellemmet",
212
+ "Noto Sans Tirhuta",
213
+ "Noto Sans Ugaritic",
214
+ "Noto Sans Vai",
215
+ "Noto Sans Vithkuqi",
216
+ "Noto Sans Wancho",
217
+ "Noto Sans Warang Citi",
218
+ "Noto Sans Yi",
219
+ "Noto Sans Zanabazar Square",
220
+ "Noto Traditional Nushu",
221
+ "P052",
222
+ "Sans",
223
+ "Serif",
224
+ "Source Code Pro",
225
+ "Standard Symbols PS",
226
+ "System-ui",
227
+ "URW Bookman",
228
+ "URW Gothic",
229
+ "Ubuntu",
230
+ "Ubuntu Condensed",
231
+ "Ubuntu Mono",
232
+ "Z003"
233
+ ],
234
+ "max_seq_length": 64,
235
+ "pad_size": 3,
236
+ "pixels_per_patch": 16,
237
+ "rgb": false,
238
+ "text_renderer_type": "PangoCairoTextRenderer"
239
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 8.470750703670477,
4
+ "train_runtime": 4458.002,
5
+ "train_samples": 974010,
6
+ "train_samples_per_second": 218.214,
7
+ "train_steps_per_second": 1.705
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,481 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.49866351511391505,
3
+ "best_model_checkpoint": "contrastive-unsup-wc-pixel-base-mean-64-128-1-3e-6-7600-42-eval/checkpoint-4800",
4
+ "epoch": 0.9986859395532195,
5
+ "global_step": 7600,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.01,
12
+ "learning_rate": 6.000000000000001e-07,
13
+ "loss": 302.0714,
14
+ "step": 100
15
+ },
16
+ {
17
+ "epoch": 0.03,
18
+ "learning_rate": 1.2000000000000002e-06,
19
+ "loss": 156.2804,
20
+ "step": 200
21
+ },
22
+ {
23
+ "epoch": 0.04,
24
+ "learning_rate": 1.8e-06,
25
+ "loss": 97.4893,
26
+ "step": 300
27
+ },
28
+ {
29
+ "epoch": 0.05,
30
+ "learning_rate": 2.4000000000000003e-06,
31
+ "loss": 58.2917,
32
+ "step": 400
33
+ },
34
+ {
35
+ "epoch": 0.07,
36
+ "learning_rate": 3e-06,
37
+ "loss": 18.3832,
38
+ "step": 500
39
+ },
40
+ {
41
+ "epoch": 0.08,
42
+ "learning_rate": 2.9577464788732396e-06,
43
+ "loss": 4.2529,
44
+ "step": 600
45
+ },
46
+ {
47
+ "epoch": 0.09,
48
+ "learning_rate": 2.915492957746479e-06,
49
+ "loss": 1.4511,
50
+ "step": 700
51
+ },
52
+ {
53
+ "epoch": 0.11,
54
+ "learning_rate": 2.8732394366197183e-06,
55
+ "loss": 0.9041,
56
+ "step": 800
57
+ },
58
+ {
59
+ "epoch": 0.12,
60
+ "learning_rate": 2.830985915492958e-06,
61
+ "loss": 0.4833,
62
+ "step": 900
63
+ },
64
+ {
65
+ "epoch": 0.13,
66
+ "learning_rate": 2.7887323943661974e-06,
67
+ "loss": 0.4023,
68
+ "step": 1000
69
+ },
70
+ {
71
+ "epoch": 0.14,
72
+ "learning_rate": 2.7464788732394365e-06,
73
+ "loss": 0.2746,
74
+ "step": 1100
75
+ },
76
+ {
77
+ "epoch": 0.16,
78
+ "learning_rate": 2.704225352112676e-06,
79
+ "loss": 0.2588,
80
+ "step": 1200
81
+ },
82
+ {
83
+ "epoch": 0.17,
84
+ "learning_rate": 2.6619718309859156e-06,
85
+ "loss": 0.2142,
86
+ "step": 1300
87
+ },
88
+ {
89
+ "epoch": 0.18,
90
+ "learning_rate": 2.6197183098591547e-06,
91
+ "loss": 0.1761,
92
+ "step": 1400
93
+ },
94
+ {
95
+ "epoch": 0.2,
96
+ "learning_rate": 2.5774647887323943e-06,
97
+ "loss": 0.0904,
98
+ "step": 1500
99
+ },
100
+ {
101
+ "epoch": 0.21,
102
+ "learning_rate": 2.535211267605634e-06,
103
+ "loss": 0.157,
104
+ "step": 1600
105
+ },
106
+ {
107
+ "epoch": 0.22,
108
+ "learning_rate": 2.4929577464788734e-06,
109
+ "loss": 0.0568,
110
+ "step": 1700
111
+ },
112
+ {
113
+ "epoch": 0.24,
114
+ "learning_rate": 2.450704225352113e-06,
115
+ "loss": 0.1395,
116
+ "step": 1800
117
+ },
118
+ {
119
+ "epoch": 0.25,
120
+ "learning_rate": 2.4084507042253525e-06,
121
+ "loss": 0.1237,
122
+ "step": 1900
123
+ },
124
+ {
125
+ "epoch": 0.26,
126
+ "learning_rate": 2.3661971830985916e-06,
127
+ "loss": 0.0702,
128
+ "step": 2000
129
+ },
130
+ {
131
+ "epoch": 0.28,
132
+ "learning_rate": 2.3239436619718307e-06,
133
+ "loss": 0.0629,
134
+ "step": 2100
135
+ },
136
+ {
137
+ "epoch": 0.29,
138
+ "learning_rate": 2.2816901408450703e-06,
139
+ "loss": 0.104,
140
+ "step": 2200
141
+ },
142
+ {
143
+ "epoch": 0.3,
144
+ "learning_rate": 2.23943661971831e-06,
145
+ "loss": 0.0629,
146
+ "step": 2300
147
+ },
148
+ {
149
+ "epoch": 0.32,
150
+ "learning_rate": 2.1971830985915494e-06,
151
+ "loss": 0.0717,
152
+ "step": 2400
153
+ },
154
+ {
155
+ "epoch": 0.33,
156
+ "learning_rate": 2.154929577464789e-06,
157
+ "loss": 0.0552,
158
+ "step": 2500
159
+ },
160
+ {
161
+ "epoch": 0.34,
162
+ "learning_rate": 2.112676056338028e-06,
163
+ "loss": 0.0986,
164
+ "step": 2600
165
+ },
166
+ {
167
+ "epoch": 0.35,
168
+ "learning_rate": 2.0704225352112676e-06,
169
+ "loss": 0.087,
170
+ "step": 2700
171
+ },
172
+ {
173
+ "epoch": 0.37,
174
+ "learning_rate": 2.028169014084507e-06,
175
+ "loss": 0.032,
176
+ "step": 2800
177
+ },
178
+ {
179
+ "epoch": 0.38,
180
+ "learning_rate": 1.9859154929577467e-06,
181
+ "loss": 0.0704,
182
+ "step": 2900
183
+ },
184
+ {
185
+ "epoch": 0.39,
186
+ "learning_rate": 1.943661971830986e-06,
187
+ "loss": 0.0313,
188
+ "step": 3000
189
+ },
190
+ {
191
+ "epoch": 0.41,
192
+ "learning_rate": 1.9014084507042256e-06,
193
+ "loss": 0.0657,
194
+ "step": 3100
195
+ },
196
+ {
197
+ "epoch": 0.42,
198
+ "learning_rate": 1.8591549295774647e-06,
199
+ "loss": 0.0594,
200
+ "step": 3200
201
+ },
202
+ {
203
+ "epoch": 0.43,
204
+ "learning_rate": 1.8169014084507043e-06,
205
+ "loss": 0.0391,
206
+ "step": 3300
207
+ },
208
+ {
209
+ "epoch": 0.45,
210
+ "learning_rate": 1.7746478873239436e-06,
211
+ "loss": 0.0319,
212
+ "step": 3400
213
+ },
214
+ {
215
+ "epoch": 0.46,
216
+ "learning_rate": 1.7323943661971832e-06,
217
+ "loss": 0.0702,
218
+ "step": 3500
219
+ },
220
+ {
221
+ "epoch": 0.47,
222
+ "learning_rate": 1.6901408450704227e-06,
223
+ "loss": 0.0401,
224
+ "step": 3600
225
+ },
226
+ {
227
+ "epoch": 0.49,
228
+ "learning_rate": 1.6478873239436623e-06,
229
+ "loss": 0.0772,
230
+ "step": 3700
231
+ },
232
+ {
233
+ "epoch": 0.5,
234
+ "learning_rate": 1.6056338028169014e-06,
235
+ "loss": 0.0248,
236
+ "step": 3800
237
+ },
238
+ {
239
+ "epoch": 0.51,
240
+ "learning_rate": 1.5633802816901407e-06,
241
+ "loss": 0.0239,
242
+ "step": 3900
243
+ },
244
+ {
245
+ "epoch": 0.53,
246
+ "learning_rate": 1.5211267605633803e-06,
247
+ "loss": 0.0259,
248
+ "step": 4000
249
+ },
250
+ {
251
+ "epoch": 0.54,
252
+ "learning_rate": 1.4788732394366198e-06,
253
+ "loss": 0.028,
254
+ "step": 4100
255
+ },
256
+ {
257
+ "epoch": 0.55,
258
+ "learning_rate": 1.4366197183098591e-06,
259
+ "loss": 0.0514,
260
+ "step": 4200
261
+ },
262
+ {
263
+ "epoch": 0.57,
264
+ "learning_rate": 1.3943661971830987e-06,
265
+ "loss": 0.0263,
266
+ "step": 4300
267
+ },
268
+ {
269
+ "epoch": 0.58,
270
+ "learning_rate": 1.352112676056338e-06,
271
+ "loss": 0.0259,
272
+ "step": 4400
273
+ },
274
+ {
275
+ "epoch": 0.59,
276
+ "learning_rate": 1.3098591549295774e-06,
277
+ "loss": 0.0393,
278
+ "step": 4500
279
+ },
280
+ {
281
+ "epoch": 0.6,
282
+ "learning_rate": 1.267605633802817e-06,
283
+ "loss": 0.0246,
284
+ "step": 4600
285
+ },
286
+ {
287
+ "epoch": 0.62,
288
+ "learning_rate": 1.2253521126760565e-06,
289
+ "loss": 0.0361,
290
+ "step": 4700
291
+ },
292
+ {
293
+ "epoch": 0.63,
294
+ "learning_rate": 1.1830985915492958e-06,
295
+ "loss": 0.0519,
296
+ "step": 4800
297
+ },
298
+ {
299
+ "epoch": 0.64,
300
+ "learning_rate": 1.1408450704225351e-06,
301
+ "loss": 0.024,
302
+ "step": 4900
303
+ },
304
+ {
305
+ "epoch": 0.66,
306
+ "learning_rate": 1.0985915492957747e-06,
307
+ "loss": 0.01,
308
+ "step": 5000
309
+ },
310
+ {
311
+ "epoch": 0.67,
312
+ "learning_rate": 1.056338028169014e-06,
313
+ "loss": 0.022,
314
+ "step": 5100
315
+ },
316
+ {
317
+ "epoch": 0.68,
318
+ "learning_rate": 1.0140845070422536e-06,
319
+ "loss": 0.0456,
320
+ "step": 5200
321
+ },
322
+ {
323
+ "epoch": 0.7,
324
+ "learning_rate": 9.71830985915493e-07,
325
+ "loss": 0.0149,
326
+ "step": 5300
327
+ },
328
+ {
329
+ "epoch": 0.71,
330
+ "learning_rate": 9.295774647887324e-07,
331
+ "loss": 0.0112,
332
+ "step": 5400
333
+ },
334
+ {
335
+ "epoch": 0.72,
336
+ "learning_rate": 8.873239436619718e-07,
337
+ "loss": 0.0337,
338
+ "step": 5500
339
+ },
340
+ {
341
+ "epoch": 0.74,
342
+ "learning_rate": 8.450704225352114e-07,
343
+ "loss": 0.0333,
344
+ "step": 5600
345
+ },
346
+ {
347
+ "epoch": 0.75,
348
+ "learning_rate": 8.028169014084507e-07,
349
+ "loss": 0.0318,
350
+ "step": 5700
351
+ },
352
+ {
353
+ "epoch": 0.76,
354
+ "learning_rate": 7.605633802816901e-07,
355
+ "loss": 0.017,
356
+ "step": 5800
357
+ },
358
+ {
359
+ "epoch": 0.78,
360
+ "learning_rate": 7.183098591549296e-07,
361
+ "loss": 0.0069,
362
+ "step": 5900
363
+ },
364
+ {
365
+ "epoch": 0.79,
366
+ "learning_rate": 6.76056338028169e-07,
367
+ "loss": 0.041,
368
+ "step": 6000
369
+ },
370
+ {
371
+ "epoch": 0.8,
372
+ "learning_rate": 6.338028169014085e-07,
373
+ "loss": 0.0198,
374
+ "step": 6100
375
+ },
376
+ {
377
+ "epoch": 0.81,
378
+ "learning_rate": 5.915492957746479e-07,
379
+ "loss": 0.1825,
380
+ "step": 6200
381
+ },
382
+ {
383
+ "epoch": 0.83,
384
+ "learning_rate": 5.492957746478873e-07,
385
+ "loss": 0.0085,
386
+ "step": 6300
387
+ },
388
+ {
389
+ "epoch": 0.84,
390
+ "learning_rate": 5.070422535211268e-07,
391
+ "loss": 0.0203,
392
+ "step": 6400
393
+ },
394
+ {
395
+ "epoch": 0.85,
396
+ "learning_rate": 4.647887323943662e-07,
397
+ "loss": 0.0651,
398
+ "step": 6500
399
+ },
400
+ {
401
+ "epoch": 0.87,
402
+ "learning_rate": 4.225352112676057e-07,
403
+ "loss": 0.0185,
404
+ "step": 6600
405
+ },
406
+ {
407
+ "epoch": 0.88,
408
+ "learning_rate": 3.8028169014084507e-07,
409
+ "loss": 0.0091,
410
+ "step": 6700
411
+ },
412
+ {
413
+ "epoch": 0.89,
414
+ "learning_rate": 3.380281690140845e-07,
415
+ "loss": 0.0304,
416
+ "step": 6800
417
+ },
418
+ {
419
+ "epoch": 0.91,
420
+ "learning_rate": 2.9577464788732395e-07,
421
+ "loss": 0.0162,
422
+ "step": 6900
423
+ },
424
+ {
425
+ "epoch": 0.92,
426
+ "learning_rate": 2.535211267605634e-07,
427
+ "loss": 0.0374,
428
+ "step": 7000
429
+ },
430
+ {
431
+ "epoch": 0.93,
432
+ "learning_rate": 2.1126760563380284e-07,
433
+ "loss": 0.0048,
434
+ "step": 7100
435
+ },
436
+ {
437
+ "epoch": 0.95,
438
+ "learning_rate": 1.6901408450704225e-07,
439
+ "loss": 0.03,
440
+ "step": 7200
441
+ },
442
+ {
443
+ "epoch": 0.96,
444
+ "learning_rate": 1.267605633802817e-07,
445
+ "loss": 0.0187,
446
+ "step": 7300
447
+ },
448
+ {
449
+ "epoch": 0.97,
450
+ "learning_rate": 8.450704225352113e-08,
451
+ "loss": 0.0208,
452
+ "step": 7400
453
+ },
454
+ {
455
+ "epoch": 0.99,
456
+ "learning_rate": 4.2253521126760564e-08,
457
+ "loss": 0.003,
458
+ "step": 7500
459
+ },
460
+ {
461
+ "epoch": 1.0,
462
+ "learning_rate": 0.0,
463
+ "loss": 0.0116,
464
+ "step": 7600
465
+ },
466
+ {
467
+ "epoch": 1.0,
468
+ "step": 7600,
469
+ "total_flos": 503654016614400.0,
470
+ "train_loss": 8.470750703670477,
471
+ "train_runtime": 4458.002,
472
+ "train_samples_per_second": 218.214,
473
+ "train_steps_per_second": 1.705
474
+ }
475
+ ],
476
+ "max_steps": 7600,
477
+ "num_train_epochs": 1,
478
+ "total_flos": 503654016614400.0,
479
+ "trial_name": null,
480
+ "trial_params": null
481
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a6cf3faff7e653cc6f47783fea4d1f67a996ec9ab9fe7d51d8f0131e9101ae9
3
+ size 3247