Hugging Face's logo

gowitheflowlab
/
stage2-2560

PyTorch
clip
Model card Files
xet
Community
gowitheflow1998 commited on
Commit
7dc606c
·
1 Parent(s): b5cade9

upload checkpoint

Browse files
Files changed (10) hide show
  1. .gitattributes +5 -0
  2. 49e6dc219d1a1a1c9236acaf05a48b542002016a6dc877ee72baab085a84257b.3f28e7f4b38e1efe1b6da4a3732404c19d4c6a614ff32dce90a251e293d4ce58 +3 -0
  3. config.json +170 -0
  4. optimizer.pt +3 -0
  5. pytorch_model.bin +3 -0
  6. rng_state.pth +3 -0
  7. scheduler.pt +3 -0
  8. text_renderer_config.json +253 -0
  9. trainer_state.json +940 -0
  10. training_args.bin +3 -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
+ pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
37
+ training_args.bin filter=lfs diff=lfs merge=lfs -text
38
+ optimizer.pt filter=lfs diff=lfs merge=lfs -text
39
+ scheduler.pt filter=lfs diff=lfs merge=lfs -text
40
+ 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
config.json ADDED
@@ -0,0 +1,170 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "./checkpoints/clip-2560-wikispan-all/checkpoint-24000",
3
+ "architectures": [
4
+ "CLIPModel"
5
+ ],
6
+ "id2label": {
7
+ "0": "Entailment",
8
+ "1": "Neutral",
9
+ "2": "Contradiction"
10
+ },
11
+ "initializer_factor": 1.0,
12
+ "label2id": {
13
+ "Contradiction": 2,
14
+ "Entailment": 0,
15
+ "Neutral": 1
16
+ },
17
+ "logit_scale_init_value": 2.6592,
18
+ "model_type": "clip",
19
+ "projection_dim": 512,
20
+ "text_config": {
21
+ "_name_or_path": "",
22
+ "add_cross_attention": false,
23
+ "architectures": null,
24
+ "attention_dropout": 0.0,
25
+ "bad_words_ids": null,
26
+ "bos_token_id": 0,
27
+ "chunk_size_feed_forward": 0,
28
+ "cross_attention_hidden_size": null,
29
+ "decoder_start_token_id": null,
30
+ "diversity_penalty": 0.0,
31
+ "do_sample": false,
32
+ "dropout": 0.0,
33
+ "early_stopping": false,
34
+ "encoder_no_repeat_ngram_size": 0,
35
+ "eos_token_id": 2,
36
+ "finetuning_task": null,
37
+ "forced_bos_token_id": null,
38
+ "forced_eos_token_id": null,
39
+ "hidden_act": "quick_gelu",
40
+ "hidden_size": 512,
41
+ "id2label": {
42
+ "0": "LABEL_0",
43
+ "1": "LABEL_1"
44
+ },
45
+ "initializer_factor": 1.0,
46
+ "initializer_range": 0.02,
47
+ "intermediate_size": 2048,
48
+ "is_decoder": false,
49
+ "is_encoder_decoder": false,
50
+ "label2id": {
51
+ "LABEL_0": 0,
52
+ "LABEL_1": 1
53
+ },
54
+ "layer_norm_eps": 1e-05,
55
+ "length_penalty": 1.0,
56
+ "max_length": 20,
57
+ "max_position_embeddings": 77,
58
+ "min_length": 0,
59
+ "model_type": "clip_text_model",
60
+ "no_repeat_ngram_size": 0,
61
+ "num_attention_heads": 8,
62
+ "num_beam_groups": 1,
63
+ "num_beams": 1,
64
+ "num_hidden_layers": 12,
65
+ "num_return_sequences": 1,
66
+ "output_attentions": false,
67
+ "output_hidden_states": false,
68
+ "output_scores": false,
69
+ "pad_token_id": 1,
70
+ "prefix": null,
71
+ "problem_type": null,
72
+ "pruned_heads": {},
73
+ "remove_invalid_values": false,
74
+ "repetition_penalty": 1.0,
75
+ "return_dict": true,
76
+ "return_dict_in_generate": false,
77
+ "sep_token_id": null,
78
+ "task_specific_params": null,
79
+ "temperature": 1.0,
80
+ "tie_encoder_decoder": false,
81
+ "tie_word_embeddings": true,
82
+ "tokenizer_class": null,
83
+ "top_k": 50,
84
+ "top_p": 1.0,
85
+ "torch_dtype": null,
86
+ "torchscript": false,
87
+ "transformers_version": "4.17.0",
88
+ "typical_p": 1.0,
89
+ "use_bfloat16": false,
90
+ "vocab_size": 49408
91
+ },
92
+ "text_config_dict": null,
93
+ "torch_dtype": "float32",
94
+ "transformers_version": null,
95
+ "vision_config": {
96
+ "_name_or_path": "",
97
+ "add_cross_attention": false,
98
+ "architectures": null,
99
+ "attention_dropout": 0.0,
100
+ "bad_words_ids": null,
101
+ "bos_token_id": null,
102
+ "chunk_size_feed_forward": 0,
103
+ "cross_attention_hidden_size": null,
104
+ "decoder_start_token_id": null,
105
+ "diversity_penalty": 0.0,
106
+ "do_sample": false,
107
+ "dropout": 0.0,
108
+ "early_stopping": false,
109
+ "encoder_no_repeat_ngram_size": 0,
110
+ "eos_token_id": null,
111
+ "finetuning_task": null,
112
+ "forced_bos_token_id": null,
113
+ "forced_eos_token_id": null,
114
+ "hidden_act": "quick_gelu",
115
+ "hidden_size": 768,
116
+ "id2label": {
117
+ "0": "LABEL_0",
118
+ "1": "LABEL_1"
119
+ },
120
+ "image_size": 224,
121
+ "initializer_factor": 1.0,
122
+ "initializer_range": 0.02,
123
+ "intermediate_size": 3072,
124
+ "is_decoder": false,
125
+ "is_encoder_decoder": false,
126
+ "label2id": {
127
+ "LABEL_0": 0,
128
+ "LABEL_1": 1
129
+ },
130
+ "layer_norm_eps": 1e-05,
131
+ "length_penalty": 1.0,
132
+ "max_length": 20,
133
+ "min_length": 0,
134
+ "model_type": "clip_vision_model",
135
+ "no_repeat_ngram_size": 0,
136
+ "num_attention_heads": 12,
137
+ "num_beam_groups": 1,
138
+ "num_beams": 1,
139
+ "num_hidden_layers": 12,
140
+ "num_return_sequences": 1,
141
+ "output_attentions": false,
142
+ "output_hidden_states": false,
143
+ "output_scores": false,
144
+ "pad_token_id": null,
145
+ "patch_size": 16,
146
+ "prefix": null,
147
+ "problem_type": null,
148
+ "pruned_heads": {},
149
+ "remove_invalid_values": false,
150
+ "repetition_penalty": 1.0,
151
+ "return_dict": true,
152
+ "return_dict_in_generate": false,
153
+ "sep_token_id": null,
154
+ "task_specific_params": null,
155
+ "temperature": 1.0,
156
+ "tie_encoder_decoder": false,
157
+ "tie_word_embeddings": true,
158
+ "tokenizer_class": null,
159
+ "top_k": 50,
160
+ "top_p": 1.0,
161
+ "torch_dtype": null,
162
+ "torchscript": false,
163
+ "transformers_version": "4.17.0",
164
+ "typical_p": 1.0,
165
+ "use_bfloat16": false
166
+ },
167
+ "vision_config_dict": {
168
+ "patch_size": 16
169
+ }
170
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0eb501953d4e2ed1991d4d968c52f3bfa66396f53ee86a03a68a28a9d34bf958
3
+ size 689659578
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:041a16cee4200bab44ad9090a73d0305d777507e23f6065329886d28102e1005
3
+ size 598642122
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:10ea96d8c35930b5d7f4448eb0b3d339c283c5e2d67b230d91b33eb0d1fb3147
3
+ size 14244
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b8651b362c9f06204bddf705f4cdde763057d5154fff2b20a5ed51ddfbc9412
3
+ size 1064
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
+ }
trainer_state.json ADDED
@@ -0,0 +1,940 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7675637669011621,
3
+ "best_model_checkpoint": "./checkpoints/clip-stage2-2560/checkpoint-1950",
4
+ "epoch": 1.0044351682755022,
5
+ "global_step": 7700,
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": 3.75e-06,
13
+ "loss": 3.5265,
14
+ "step": 50
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 7.5e-06,
19
+ "loss": 1.4349,
20
+ "step": 100
21
+ },
22
+ {
23
+ "epoch": 0.02,
24
+ "learning_rate": 1.125e-05,
25
+ "loss": 1.1217,
26
+ "step": 150
27
+ },
28
+ {
29
+ "epoch": 0.03,
30
+ "learning_rate": 1.5e-05,
31
+ "loss": 0.9885,
32
+ "step": 200
33
+ },
34
+ {
35
+ "epoch": 0.03,
36
+ "learning_rate": 1.8750000000000002e-05,
37
+ "loss": 0.8975,
38
+ "step": 250
39
+ },
40
+ {
41
+ "epoch": 0.04,
42
+ "learning_rate": 2.25e-05,
43
+ "loss": 0.8384,
44
+ "step": 300
45
+ },
46
+ {
47
+ "epoch": 0.05,
48
+ "learning_rate": 2.625e-05,
49
+ "loss": 0.7769,
50
+ "step": 350
51
+ },
52
+ {
53
+ "epoch": 0.05,
54
+ "learning_rate": 3e-05,
55
+ "loss": 0.7357,
56
+ "step": 400
57
+ },
58
+ {
59
+ "epoch": 0.06,
60
+ "learning_rate": 2.9794520547945206e-05,
61
+ "loss": 0.7051,
62
+ "step": 450
63
+ },
64
+ {
65
+ "epoch": 0.07,
66
+ "learning_rate": 2.958904109589041e-05,
67
+ "loss": 0.6722,
68
+ "step": 500
69
+ },
70
+ {
71
+ "epoch": 0.07,
72
+ "learning_rate": 2.938356164383562e-05,
73
+ "loss": 0.647,
74
+ "step": 550
75
+ },
76
+ {
77
+ "epoch": 0.08,
78
+ "learning_rate": 2.9178082191780824e-05,
79
+ "loss": 0.6236,
80
+ "step": 600
81
+ },
82
+ {
83
+ "epoch": 0.08,
84
+ "learning_rate": 2.8972602739726026e-05,
85
+ "loss": 0.6035,
86
+ "step": 650
87
+ },
88
+ {
89
+ "epoch": 0.09,
90
+ "learning_rate": 2.876712328767123e-05,
91
+ "loss": 0.5885,
92
+ "step": 700
93
+ },
94
+ {
95
+ "epoch": 0.1,
96
+ "learning_rate": 2.856164383561644e-05,
97
+ "loss": 0.5622,
98
+ "step": 750
99
+ },
100
+ {
101
+ "epoch": 0.1,
102
+ "learning_rate": 2.8356164383561644e-05,
103
+ "loss": 0.5613,
104
+ "step": 800
105
+ },
106
+ {
107
+ "epoch": 0.11,
108
+ "learning_rate": 2.815068493150685e-05,
109
+ "loss": 0.5414,
110
+ "step": 850
111
+ },
112
+ {
113
+ "epoch": 0.12,
114
+ "learning_rate": 2.7945205479452054e-05,
115
+ "loss": 0.536,
116
+ "step": 900
117
+ },
118
+ {
119
+ "epoch": 0.12,
120
+ "learning_rate": 2.7739726027397263e-05,
121
+ "loss": 0.5227,
122
+ "step": 950
123
+ },
124
+ {
125
+ "epoch": 0.13,
126
+ "learning_rate": 2.7534246575342468e-05,
127
+ "loss": 0.5099,
128
+ "step": 1000
129
+ },
130
+ {
131
+ "epoch": 0.14,
132
+ "learning_rate": 2.732876712328767e-05,
133
+ "loss": 0.4992,
134
+ "step": 1050
135
+ },
136
+ {
137
+ "epoch": 0.14,
138
+ "learning_rate": 2.7123287671232875e-05,
139
+ "loss": 0.4937,
140
+ "step": 1100
141
+ },
142
+ {
143
+ "epoch": 0.15,
144
+ "learning_rate": 2.6917808219178083e-05,
145
+ "loss": 0.4914,
146
+ "step": 1150
147
+ },
148
+ {
149
+ "epoch": 0.16,
150
+ "learning_rate": 2.6712328767123288e-05,
151
+ "loss": 0.4758,
152
+ "step": 1200
153
+ },
154
+ {
155
+ "epoch": 0.16,
156
+ "learning_rate": 2.6506849315068493e-05,
157
+ "loss": 0.4722,
158
+ "step": 1250
159
+ },
160
+ {
161
+ "epoch": 0.17,
162
+ "learning_rate": 2.6301369863013698e-05,
163
+ "loss": 0.4673,
164
+ "step": 1300
165
+ },
166
+ {
167
+ "epoch": 0.18,
168
+ "learning_rate": 2.6095890410958907e-05,
169
+ "loss": 0.4639,
170
+ "step": 1350
171
+ },
172
+ {
173
+ "epoch": 0.18,
174
+ "learning_rate": 2.589041095890411e-05,
175
+ "loss": 0.4499,
176
+ "step": 1400
177
+ },
178
+ {
179
+ "epoch": 0.19,
180
+ "learning_rate": 2.5684931506849317e-05,
181
+ "loss": 0.4427,
182
+ "step": 1450
183
+ },
184
+ {
185
+ "epoch": 0.2,
186
+ "learning_rate": 2.5479452054794518e-05,
187
+ "loss": 0.4489,
188
+ "step": 1500
189
+ },
190
+ {
191
+ "epoch": 0.2,
192
+ "learning_rate": 2.5273972602739727e-05,
193
+ "loss": 0.441,
194
+ "step": 1550
195
+ },
196
+ {
197
+ "epoch": 0.21,
198
+ "learning_rate": 2.5068493150684932e-05,
199
+ "loss": 0.4287,
200
+ "step": 1600
201
+ },
202
+ {
203
+ "epoch": 0.22,
204
+ "learning_rate": 2.4863013698630137e-05,
205
+ "loss": 0.4337,
206
+ "step": 1650
207
+ },
208
+ {
209
+ "epoch": 0.22,
210
+ "learning_rate": 2.4657534246575342e-05,
211
+ "loss": 0.4283,
212
+ "step": 1700
213
+ },
214
+ {
215
+ "epoch": 0.23,
216
+ "learning_rate": 2.445205479452055e-05,
217
+ "loss": 0.4194,
218
+ "step": 1750
219
+ },
220
+ {
221
+ "epoch": 0.23,
222
+ "learning_rate": 2.4246575342465755e-05,
223
+ "loss": 0.4202,
224
+ "step": 1800
225
+ },
226
+ {
227
+ "epoch": 0.24,
228
+ "learning_rate": 2.404109589041096e-05,
229
+ "loss": 0.4106,
230
+ "step": 1850
231
+ },
232
+ {
233
+ "epoch": 0.25,
234
+ "learning_rate": 2.3835616438356165e-05,
235
+ "loss": 0.4123,
236
+ "step": 1900
237
+ },
238
+ {
239
+ "epoch": 0.25,
240
+ "learning_rate": 2.363013698630137e-05,
241
+ "loss": 0.4021,
242
+ "step": 1950
243
+ },
244
+ {
245
+ "epoch": 0.26,
246
+ "learning_rate": 2.3424657534246575e-05,
247
+ "loss": 0.4012,
248
+ "step": 2000
249
+ },
250
+ {
251
+ "epoch": 0.27,
252
+ "learning_rate": 2.321917808219178e-05,
253
+ "loss": 0.3981,
254
+ "step": 2050
255
+ },
256
+ {
257
+ "epoch": 0.27,
258
+ "learning_rate": 2.3013698630136985e-05,
259
+ "loss": 0.3972,
260
+ "step": 2100
261
+ },
262
+ {
263
+ "epoch": 0.28,
264
+ "learning_rate": 2.2808219178082194e-05,
265
+ "loss": 0.3949,
266
+ "step": 2150
267
+ },
268
+ {
269
+ "epoch": 0.29,
270
+ "learning_rate": 2.26027397260274e-05,
271
+ "loss": 0.3895,
272
+ "step": 2200
273
+ },
274
+ {
275
+ "epoch": 0.29,
276
+ "learning_rate": 2.2397260273972604e-05,
277
+ "loss": 0.3884,
278
+ "step": 2250
279
+ },
280
+ {
281
+ "epoch": 0.3,
282
+ "learning_rate": 2.219178082191781e-05,
283
+ "loss": 0.3849,
284
+ "step": 2300
285
+ },
286
+ {
287
+ "epoch": 0.31,
288
+ "learning_rate": 2.1986301369863014e-05,
289
+ "loss": 0.3824,
290
+ "step": 2350
291
+ },
292
+ {
293
+ "epoch": 0.31,
294
+ "learning_rate": 2.178082191780822e-05,
295
+ "loss": 0.3773,
296
+ "step": 2400
297
+ },
298
+ {
299
+ "epoch": 0.32,
300
+ "learning_rate": 2.1575342465753424e-05,
301
+ "loss": 0.3798,
302
+ "step": 2450
303
+ },
304
+ {
305
+ "epoch": 0.33,
306
+ "learning_rate": 2.136986301369863e-05,
307
+ "loss": 0.3724,
308
+ "step": 2500
309
+ },
310
+ {
311
+ "epoch": 0.33,
312
+ "learning_rate": 2.1164383561643838e-05,
313
+ "loss": 0.3721,
314
+ "step": 2550
315
+ },
316
+ {
317
+ "epoch": 0.34,
318
+ "learning_rate": 2.0958904109589043e-05,
319
+ "loss": 0.3658,
320
+ "step": 2600
321
+ },
322
+ {
323
+ "epoch": 0.35,
324
+ "learning_rate": 2.0753424657534248e-05,
325
+ "loss": 0.3718,
326
+ "step": 2650
327
+ },
328
+ {
329
+ "epoch": 0.35,
330
+ "learning_rate": 2.0547945205479453e-05,
331
+ "loss": 0.3642,
332
+ "step": 2700
333
+ },
334
+ {
335
+ "epoch": 0.36,
336
+ "learning_rate": 2.034246575342466e-05,
337
+ "loss": 0.3568,
338
+ "step": 2750
339
+ },
340
+ {
341
+ "epoch": 0.37,
342
+ "learning_rate": 2.0136986301369863e-05,
343
+ "loss": 0.3645,
344
+ "step": 2800
345
+ },
346
+ {
347
+ "epoch": 0.37,
348
+ "learning_rate": 1.9931506849315068e-05,
349
+ "loss": 0.3551,
350
+ "step": 2850
351
+ },
352
+ {
353
+ "epoch": 0.38,
354
+ "learning_rate": 1.9726027397260273e-05,
355
+ "loss": 0.3568,
356
+ "step": 2900
357
+ },
358
+ {
359
+ "epoch": 0.38,
360
+ "learning_rate": 1.952054794520548e-05,
361
+ "loss": 0.3523,
362
+ "step": 2950
363
+ },
364
+ {
365
+ "epoch": 0.39,
366
+ "learning_rate": 1.9315068493150686e-05,
367
+ "loss": 0.3525,
368
+ "step": 3000
369
+ },
370
+ {
371
+ "epoch": 0.4,
372
+ "learning_rate": 1.910958904109589e-05,
373
+ "loss": 0.3521,
374
+ "step": 3050
375
+ },
376
+ {
377
+ "epoch": 0.4,
378
+ "learning_rate": 1.8904109589041096e-05,
379
+ "loss": 0.348,
380
+ "step": 3100
381
+ },
382
+ {
383
+ "epoch": 0.41,
384
+ "learning_rate": 1.8698630136986305e-05,
385
+ "loss": 0.3475,
386
+ "step": 3150
387
+ },
388
+ {
389
+ "epoch": 0.42,
390
+ "learning_rate": 1.8493150684931506e-05,
391
+ "loss": 0.3402,
392
+ "step": 3200
393
+ },
394
+ {
395
+ "epoch": 0.42,
396
+ "learning_rate": 1.828767123287671e-05,
397
+ "loss": 0.3432,
398
+ "step": 3250
399
+ },
400
+ {
401
+ "epoch": 0.43,
402
+ "learning_rate": 1.8082191780821916e-05,
403
+ "loss": 0.3437,
404
+ "step": 3300
405
+ },
406
+ {
407
+ "epoch": 0.44,
408
+ "learning_rate": 1.7876712328767125e-05,
409
+ "loss": 0.3427,
410
+ "step": 3350
411
+ },
412
+ {
413
+ "epoch": 0.44,
414
+ "learning_rate": 1.767123287671233e-05,
415
+ "loss": 0.3354,
416
+ "step": 3400
417
+ },
418
+ {
419
+ "epoch": 0.45,
420
+ "learning_rate": 1.7465753424657535e-05,
421
+ "loss": 0.3422,
422
+ "step": 3450
423
+ },
424
+ {
425
+ "epoch": 0.46,
426
+ "learning_rate": 1.726027397260274e-05,
427
+ "loss": 0.3389,
428
+ "step": 3500
429
+ },
430
+ {
431
+ "epoch": 0.46,
432
+ "learning_rate": 1.705479452054795e-05,
433
+ "loss": 0.3297,
434
+ "step": 3550
435
+ },
436
+ {
437
+ "epoch": 0.47,
438
+ "learning_rate": 1.6849315068493153e-05,
439
+ "loss": 0.3331,
440
+ "step": 3600
441
+ },
442
+ {
443
+ "epoch": 0.48,
444
+ "learning_rate": 1.6643835616438355e-05,
445
+ "loss": 0.3288,
446
+ "step": 3650
447
+ },
448
+ {
449
+ "epoch": 0.48,
450
+ "learning_rate": 1.643835616438356e-05,
451
+ "loss": 0.3261,
452
+ "step": 3700
453
+ },
454
+ {
455
+ "epoch": 0.49,
456
+ "learning_rate": 1.623287671232877e-05,
457
+ "loss": 0.3303,
458
+ "step": 3750
459
+ },
460
+ {
461
+ "epoch": 0.5,
462
+ "learning_rate": 1.6027397260273974e-05,
463
+ "loss": 0.3302,
464
+ "step": 3800
465
+ },
466
+ {
467
+ "epoch": 0.5,
468
+ "learning_rate": 1.582191780821918e-05,
469
+ "loss": 0.3191,
470
+ "step": 3850
471
+ },
472
+ {
473
+ "epoch": 0.51,
474
+ "learning_rate": 1.5616438356164384e-05,
475
+ "loss": 0.3221,
476
+ "step": 3900
477
+ },
478
+ {
479
+ "epoch": 0.52,
480
+ "learning_rate": 1.5410958904109592e-05,
481
+ "loss": 0.322,
482
+ "step": 3950
483
+ },
484
+ {
485
+ "epoch": 0.52,
486
+ "learning_rate": 1.5205479452054795e-05,
487
+ "loss": 0.3159,
488
+ "step": 4000
489
+ },
490
+ {
491
+ "epoch": 0.53,
492
+ "learning_rate": 1.5e-05,
493
+ "loss": 0.3182,
494
+ "step": 4050
495
+ },
496
+ {
497
+ "epoch": 0.53,
498
+ "learning_rate": 1.4794520547945205e-05,
499
+ "loss": 0.3178,
500
+ "step": 4100
501
+ },
502
+ {
503
+ "epoch": 0.54,
504
+ "learning_rate": 1.4589041095890412e-05,
505
+ "loss": 0.3181,
506
+ "step": 4150
507
+ },
508
+ {
509
+ "epoch": 0.55,
510
+ "learning_rate": 1.4383561643835615e-05,
511
+ "loss": 0.3175,
512
+ "step": 4200
513
+ },
514
+ {
515
+ "epoch": 0.55,
516
+ "learning_rate": 1.4178082191780822e-05,
517
+ "loss": 0.3137,
518
+ "step": 4250
519
+ },
520
+ {
521
+ "epoch": 0.56,
522
+ "learning_rate": 1.3972602739726027e-05,
523
+ "loss": 0.3143,
524
+ "step": 4300
525
+ },
526
+ {
527
+ "epoch": 0.57,
528
+ "learning_rate": 1.3767123287671234e-05,
529
+ "loss": 0.3139,
530
+ "step": 4350
531
+ },
532
+ {
533
+ "epoch": 0.57,
534
+ "learning_rate": 1.3561643835616437e-05,
535
+ "loss": 0.3101,
536
+ "step": 4400
537
+ },
538
+ {
539
+ "epoch": 0.58,
540
+ "learning_rate": 1.3356164383561644e-05,
541
+ "loss": 0.3055,
542
+ "step": 4450
543
+ },
544
+ {
545
+ "epoch": 0.59,
546
+ "learning_rate": 1.3150684931506849e-05,
547
+ "loss": 0.3059,
548
+ "step": 4500
549
+ },
550
+ {
551
+ "epoch": 0.59,
552
+ "learning_rate": 1.2945205479452056e-05,
553
+ "loss": 0.3083,
554
+ "step": 4550
555
+ },
556
+ {
557
+ "epoch": 0.6,
558
+ "learning_rate": 1.2739726027397259e-05,
559
+ "loss": 0.3083,
560
+ "step": 4600
561
+ },
562
+ {
563
+ "epoch": 0.61,
564
+ "learning_rate": 1.2534246575342466e-05,
565
+ "loss": 0.2982,
566
+ "step": 4650
567
+ },
568
+ {
569
+ "epoch": 0.61,
570
+ "learning_rate": 1.2328767123287671e-05,
571
+ "loss": 0.3078,
572
+ "step": 4700
573
+ },
574
+ {
575
+ "epoch": 0.62,
576
+ "learning_rate": 1.2123287671232878e-05,
577
+ "loss": 0.3015,
578
+ "step": 4750
579
+ },
580
+ {
581
+ "epoch": 0.63,
582
+ "learning_rate": 1.1917808219178083e-05,
583
+ "loss": 0.3042,
584
+ "step": 4800
585
+ },
586
+ {
587
+ "epoch": 0.63,
588
+ "learning_rate": 1.1712328767123288e-05,
589
+ "loss": 0.3009,
590
+ "step": 4850
591
+ },
592
+ {
593
+ "epoch": 0.64,
594
+ "learning_rate": 1.1506849315068493e-05,
595
+ "loss": 0.3006,
596
+ "step": 4900
597
+ },
598
+ {
599
+ "epoch": 0.65,
600
+ "learning_rate": 1.13013698630137e-05,
601
+ "loss": 0.3002,
602
+ "step": 4950
603
+ },
604
+ {
605
+ "epoch": 0.65,
606
+ "learning_rate": 1.1095890410958904e-05,
607
+ "loss": 0.2929,
608
+ "step": 5000
609
+ },
610
+ {
611
+ "epoch": 0.66,
612
+ "learning_rate": 1.089041095890411e-05,
613
+ "loss": 0.3001,
614
+ "step": 5050
615
+ },
616
+ {
617
+ "epoch": 0.67,
618
+ "learning_rate": 1.0684931506849315e-05,
619
+ "loss": 0.2982,
620
+ "step": 5100
621
+ },
622
+ {
623
+ "epoch": 0.67,
624
+ "learning_rate": 1.0479452054794521e-05,
625
+ "loss": 0.2952,
626
+ "step": 5150
627
+ },
628
+ {
629
+ "epoch": 0.68,
630
+ "learning_rate": 1.0273972602739726e-05,
631
+ "loss": 0.292,
632
+ "step": 5200
633
+ },
634
+ {
635
+ "epoch": 0.68,
636
+ "learning_rate": 1.0068493150684931e-05,
637
+ "loss": 0.2954,
638
+ "step": 5250
639
+ },
640
+ {
641
+ "epoch": 0.69,
642
+ "learning_rate": 9.863013698630136e-06,
643
+ "loss": 0.2968,
644
+ "step": 5300
645
+ },
646
+ {
647
+ "epoch": 0.7,
648
+ "learning_rate": 9.657534246575343e-06,
649
+ "loss": 0.2896,
650
+ "step": 5350
651
+ },
652
+ {
653
+ "epoch": 0.7,
654
+ "learning_rate": 9.452054794520548e-06,
655
+ "loss": 0.2874,
656
+ "step": 5400
657
+ },
658
+ {
659
+ "epoch": 0.71,
660
+ "learning_rate": 9.246575342465753e-06,
661
+ "loss": 0.2948,
662
+ "step": 5450
663
+ },
664
+ {
665
+ "epoch": 0.72,
666
+ "learning_rate": 9.041095890410958e-06,
667
+ "loss": 0.2891,
668
+ "step": 5500
669
+ },
670
+ {
671
+ "epoch": 0.72,
672
+ "learning_rate": 8.835616438356165e-06,
673
+ "loss": 0.2986,
674
+ "step": 5550
675
+ },
676
+ {
677
+ "epoch": 0.73,
678
+ "learning_rate": 8.63013698630137e-06,
679
+ "loss": 0.2903,
680
+ "step": 5600
681
+ },
682
+ {
683
+ "epoch": 0.74,
684
+ "learning_rate": 8.424657534246577e-06,
685
+ "loss": 0.2878,
686
+ "step": 5650
687
+ },
688
+ {
689
+ "epoch": 0.74,
690
+ "learning_rate": 8.21917808219178e-06,
691
+ "loss": 0.2873,
692
+ "step": 5700
693
+ },
694
+ {
695
+ "epoch": 0.75,
696
+ "learning_rate": 8.013698630136987e-06,
697
+ "loss": 0.2879,
698
+ "step": 5750
699
+ },
700
+ {
701
+ "epoch": 0.76,
702
+ "learning_rate": 7.808219178082192e-06,
703
+ "loss": 0.2866,
704
+ "step": 5800
705
+ },
706
+ {
707
+ "epoch": 0.76,
708
+ "learning_rate": 7.602739726027398e-06,
709
+ "loss": 0.2869,
710
+ "step": 5850
711
+ },
712
+ {
713
+ "epoch": 0.77,
714
+ "learning_rate": 7.397260273972603e-06,
715
+ "loss": 0.2841,
716
+ "step": 5900
717
+ },
718
+ {
719
+ "epoch": 0.78,
720
+ "learning_rate": 7.191780821917808e-06,
721
+ "loss": 0.2814,
722
+ "step": 5950
723
+ },
724
+ {
725
+ "epoch": 0.78,
726
+ "learning_rate": 6.986301369863014e-06,
727
+ "loss": 0.2833,
728
+ "step": 6000
729
+ },
730
+ {
731
+ "epoch": 0.79,
732
+ "learning_rate": 6.780821917808219e-06,
733
+ "loss": 0.2845,
734
+ "step": 6050
735
+ },
736
+ {
737
+ "epoch": 0.8,
738
+ "learning_rate": 6.5753424657534245e-06,
739
+ "loss": 0.2823,
740
+ "step": 6100
741
+ },
742
+ {
743
+ "epoch": 0.8,
744
+ "learning_rate": 6.3698630136986296e-06,
745
+ "loss": 0.28,
746
+ "step": 6150
747
+ },
748
+ {
749
+ "epoch": 0.81,
750
+ "learning_rate": 6.1643835616438354e-06,
751
+ "loss": 0.2837,
752
+ "step": 6200
753
+ },
754
+ {
755
+ "epoch": 0.82,
756
+ "learning_rate": 5.958904109589041e-06,
757
+ "loss": 0.2739,
758
+ "step": 6250
759
+ },
760
+ {
761
+ "epoch": 0.82,
762
+ "learning_rate": 5.753424657534246e-06,
763
+ "loss": 0.2854,
764
+ "step": 6300
765
+ },
766
+ {
767
+ "epoch": 0.83,
768
+ "learning_rate": 5.547945205479452e-06,
769
+ "loss": 0.2797,
770
+ "step": 6350
771
+ },
772
+ {
773
+ "epoch": 0.83,
774
+ "learning_rate": 5.342465753424657e-06,
775
+ "loss": 0.2797,
776
+ "step": 6400
777
+ },
778
+ {
779
+ "epoch": 0.84,
780
+ "learning_rate": 5.136986301369863e-06,
781
+ "loss": 0.2759,
782
+ "step": 6450
783
+ },
784
+ {
785
+ "epoch": 0.85,
786
+ "learning_rate": 4.931506849315068e-06,
787
+ "loss": 0.2803,
788
+ "step": 6500
789
+ },
790
+ {
791
+ "epoch": 0.85,
792
+ "learning_rate": 4.726027397260274e-06,
793
+ "loss": 0.2772,
794
+ "step": 6550
795
+ },
796
+ {
797
+ "epoch": 0.86,
798
+ "learning_rate": 4.520547945205479e-06,
799
+ "loss": 0.2787,
800
+ "step": 6600
801
+ },
802
+ {
803
+ "epoch": 0.87,
804
+ "learning_rate": 4.315068493150685e-06,
805
+ "loss": 0.269,
806
+ "step": 6650
807
+ },
808
+ {
809
+ "epoch": 0.87,
810
+ "learning_rate": 4.10958904109589e-06,
811
+ "loss": 0.2719,
812
+ "step": 6700
813
+ },
814
+ {
815
+ "epoch": 0.88,
816
+ "learning_rate": 3.904109589041096e-06,
817
+ "loss": 0.2751,
818
+ "step": 6750
819
+ },
820
+ {
821
+ "epoch": 0.89,
822
+ "learning_rate": 3.6986301369863014e-06,
823
+ "loss": 0.2764,
824
+ "step": 6800
825
+ },
826
+ {
827
+ "epoch": 0.89,
828
+ "learning_rate": 3.493150684931507e-06,
829
+ "loss": 0.2747,
830
+ "step": 6850
831
+ },
832
+ {
833
+ "epoch": 0.9,
834
+ "learning_rate": 3.2876712328767123e-06,
835
+ "loss": 0.2728,
836
+ "step": 6900
837
+ },
838
+ {
839
+ "epoch": 0.91,
840
+ "learning_rate": 3.0821917808219177e-06,
841
+ "loss": 0.2708,
842
+ "step": 6950
843
+ },
844
+ {
845
+ "epoch": 0.91,
846
+ "learning_rate": 2.876712328767123e-06,
847
+ "loss": 0.2708,
848
+ "step": 7000
849
+ },
850
+ {
851
+ "epoch": 0.92,
852
+ "learning_rate": 2.6712328767123286e-06,
853
+ "loss": 0.2689,
854
+ "step": 7050
855
+ },
856
+ {
857
+ "epoch": 0.93,
858
+ "learning_rate": 2.465753424657534e-06,
859
+ "loss": 0.2743,
860
+ "step": 7100
861
+ },
862
+ {
863
+ "epoch": 0.93,
864
+ "learning_rate": 2.2602739726027396e-06,
865
+ "loss": 0.266,
866
+ "step": 7150
867
+ },
868
+ {
869
+ "epoch": 0.94,
870
+ "learning_rate": 2.054794520547945e-06,
871
+ "loss": 0.2697,
872
+ "step": 7200
873
+ },
874
+ {
875
+ "epoch": 0.95,
876
+ "learning_rate": 1.8493150684931507e-06,
877
+ "loss": 0.2712,
878
+ "step": 7250
879
+ },
880
+ {
881
+ "epoch": 0.95,
882
+ "learning_rate": 1.6438356164383561e-06,
883
+ "loss": 0.2721,
884
+ "step": 7300
885
+ },
886
+ {
887
+ "epoch": 0.96,
888
+ "learning_rate": 1.4383561643835616e-06,
889
+ "loss": 0.2699,
890
+ "step": 7350
891
+ },
892
+ {
893
+ "epoch": 0.97,
894
+ "learning_rate": 1.232876712328767e-06,
895
+ "loss": 0.2703,
896
+ "step": 7400
897
+ },
898
+ {
899
+ "epoch": 0.97,
900
+ "learning_rate": 1.0273972602739725e-06,
901
+ "loss": 0.2692,
902
+ "step": 7450
903
+ },
904
+ {
905
+ "epoch": 0.98,
906
+ "learning_rate": 8.219178082191781e-07,
907
+ "loss": 0.2675,
908
+ "step": 7500
909
+ },
910
+ {
911
+ "epoch": 0.98,
912
+ "learning_rate": 6.164383561643835e-07,
913
+ "loss": 0.2645,
914
+ "step": 7550
915
+ },
916
+ {
917
+ "epoch": 0.99,
918
+ "learning_rate": 4.1095890410958903e-07,
919
+ "loss": 0.2691,
920
+ "step": 7600
921
+ },
922
+ {
923
+ "epoch": 1.0,
924
+ "learning_rate": 2.0547945205479452e-07,
925
+ "loss": 0.2748,
926
+ "step": 7650
927
+ },
928
+ {
929
+ "epoch": 1.0,
930
+ "learning_rate": 0.0,
931
+ "loss": 0.2504,
932
+ "step": 7700
933
+ }
934
+ ],
935
+ "max_steps": 7700,
936
+ "num_train_epochs": 2,
937
+ "total_flos": 0.0,
938
+ "trial_name": null,
939
+ "trial_params": null
940
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:14b2158cf4943b3d5ce46dbb7bf7fb4a71dd1e5df5d23dded3df2de6415ca440
3
+ size 3640