Hugging Face's logo

Blablablab
/
neurobiber

Safetensors
roberta
Model card Files
xet
Community
kalkiek95 commited on
Commit
996ff9d
·
verified ·
1 Parent(s): 437067c

Upload 12 files

Browse files
Files changed (12) hide show
  1. config.json +224 -0
  2. merges.txt +0 -0
  3. model.safetensors +3 -0
  4. optimizer.pt +3 -0
  5. rng_state.pth +3 -0
  6. scheduler.pt +3 -0
  7. special_tokens_map.json +15 -0
  8. tokenizer.json +0 -0
  9. tokenizer_config.json +57 -0
  10. trainer_state.json +3395 -0
  11. training_args.bin +3 -0
  12. vocab.json +0 -0
config.json ADDED
@@ -0,0 +1,224 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "roberta-base",
3
+ "architectures": [
4
+ "RobertaForSequenceClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "id2label": {
14
+ "0": "LABEL_0",
15
+ "1": "LABEL_1",
16
+ "2": "LABEL_2",
17
+ "3": "LABEL_3",
18
+ "4": "LABEL_4",
19
+ "5": "LABEL_5",
20
+ "6": "LABEL_6",
21
+ "7": "LABEL_7",
22
+ "8": "LABEL_8",
23
+ "9": "LABEL_9",
24
+ "10": "LABEL_10",
25
+ "11": "LABEL_11",
26
+ "12": "LABEL_12",
27
+ "13": "LABEL_13",
28
+ "14": "LABEL_14",
29
+ "15": "LABEL_15",
30
+ "16": "LABEL_16",
31
+ "17": "LABEL_17",
32
+ "18": "LABEL_18",
33
+ "19": "LABEL_19",
34
+ "20": "LABEL_20",
35
+ "21": "LABEL_21",
36
+ "22": "LABEL_22",
37
+ "23": "LABEL_23",
38
+ "24": "LABEL_24",
39
+ "25": "LABEL_25",
40
+ "26": "LABEL_26",
41
+ "27": "LABEL_27",
42
+ "28": "LABEL_28",
43
+ "29": "LABEL_29",
44
+ "30": "LABEL_30",
45
+ "31": "LABEL_31",
46
+ "32": "LABEL_32",
47
+ "33": "LABEL_33",
48
+ "34": "LABEL_34",
49
+ "35": "LABEL_35",
50
+ "36": "LABEL_36",
51
+ "37": "LABEL_37",
52
+ "38": "LABEL_38",
53
+ "39": "LABEL_39",
54
+ "40": "LABEL_40",
55
+ "41": "LABEL_41",
56
+ "42": "LABEL_42",
57
+ "43": "LABEL_43",
58
+ "44": "LABEL_44",
59
+ "45": "LABEL_45",
60
+ "46": "LABEL_46",
61
+ "47": "LABEL_47",
62
+ "48": "LABEL_48",
63
+ "49": "LABEL_49",
64
+ "50": "LABEL_50",
65
+ "51": "LABEL_51",
66
+ "52": "LABEL_52",
67
+ "53": "LABEL_53",
68
+ "54": "LABEL_54",
69
+ "55": "LABEL_55",
70
+ "56": "LABEL_56",
71
+ "57": "LABEL_57",
72
+ "58": "LABEL_58",
73
+ "59": "LABEL_59",
74
+ "60": "LABEL_60",
75
+ "61": "LABEL_61",
76
+ "62": "LABEL_62",
77
+ "63": "LABEL_63",
78
+ "64": "LABEL_64",
79
+ "65": "LABEL_65",
80
+ "66": "LABEL_66",
81
+ "67": "LABEL_67",
82
+ "68": "LABEL_68",
83
+ "69": "LABEL_69",
84
+ "70": "LABEL_70",
85
+ "71": "LABEL_71",
86
+ "72": "LABEL_72",
87
+ "73": "LABEL_73",
88
+ "74": "LABEL_74",
89
+ "75": "LABEL_75",
90
+ "76": "LABEL_76",
91
+ "77": "LABEL_77",
92
+ "78": "LABEL_78",
93
+ "79": "LABEL_79",
94
+ "80": "LABEL_80",
95
+ "81": "LABEL_81",
96
+ "82": "LABEL_82",
97
+ "83": "LABEL_83",
98
+ "84": "LABEL_84",
99
+ "85": "LABEL_85",
100
+ "86": "LABEL_86",
101
+ "87": "LABEL_87",
102
+ "88": "LABEL_88",
103
+ "89": "LABEL_89",
104
+ "90": "LABEL_90",
105
+ "91": "LABEL_91",
106
+ "92": "LABEL_92",
107
+ "93": "LABEL_93",
108
+ "94": "LABEL_94",
109
+ "95": "LABEL_95"
110
+ },
111
+ "initializer_range": 0.02,
112
+ "intermediate_size": 3072,
113
+ "label2id": {
114
+ "LABEL_0": 0,
115
+ "LABEL_1": 1,
116
+ "LABEL_10": 10,
117
+ "LABEL_11": 11,
118
+ "LABEL_12": 12,
119
+ "LABEL_13": 13,
120
+ "LABEL_14": 14,
121
+ "LABEL_15": 15,
122
+ "LABEL_16": 16,
123
+ "LABEL_17": 17,
124
+ "LABEL_18": 18,
125
+ "LABEL_19": 19,
126
+ "LABEL_2": 2,
127
+ "LABEL_20": 20,
128
+ "LABEL_21": 21,
129
+ "LABEL_22": 22,
130
+ "LABEL_23": 23,
131
+ "LABEL_24": 24,
132
+ "LABEL_25": 25,
133
+ "LABEL_26": 26,
134
+ "LABEL_27": 27,
135
+ "LABEL_28": 28,
136
+ "LABEL_29": 29,
137
+ "LABEL_3": 3,
138
+ "LABEL_30": 30,
139
+ "LABEL_31": 31,
140
+ "LABEL_32": 32,
141
+ "LABEL_33": 33,
142
+ "LABEL_34": 34,
143
+ "LABEL_35": 35,
144
+ "LABEL_36": 36,
145
+ "LABEL_37": 37,
146
+ "LABEL_38": 38,
147
+ "LABEL_39": 39,
148
+ "LABEL_4": 4,
149
+ "LABEL_40": 40,
150
+ "LABEL_41": 41,
151
+ "LABEL_42": 42,
152
+ "LABEL_43": 43,
153
+ "LABEL_44": 44,
154
+ "LABEL_45": 45,
155
+ "LABEL_46": 46,
156
+ "LABEL_47": 47,
157
+ "LABEL_48": 48,
158
+ "LABEL_49": 49,
159
+ "LABEL_5": 5,
160
+ "LABEL_50": 50,
161
+ "LABEL_51": 51,
162
+ "LABEL_52": 52,
163
+ "LABEL_53": 53,
164
+ "LABEL_54": 54,
165
+ "LABEL_55": 55,
166
+ "LABEL_56": 56,
167
+ "LABEL_57": 57,
168
+ "LABEL_58": 58,
169
+ "LABEL_59": 59,
170
+ "LABEL_6": 6,
171
+ "LABEL_60": 60,
172
+ "LABEL_61": 61,
173
+ "LABEL_62": 62,
174
+ "LABEL_63": 63,
175
+ "LABEL_64": 64,
176
+ "LABEL_65": 65,
177
+ "LABEL_66": 66,
178
+ "LABEL_67": 67,
179
+ "LABEL_68": 68,
180
+ "LABEL_69": 69,
181
+ "LABEL_7": 7,
182
+ "LABEL_70": 70,
183
+ "LABEL_71": 71,
184
+ "LABEL_72": 72,
185
+ "LABEL_73": 73,
186
+ "LABEL_74": 74,
187
+ "LABEL_75": 75,
188
+ "LABEL_76": 76,
189
+ "LABEL_77": 77,
190
+ "LABEL_78": 78,
191
+ "LABEL_79": 79,
192
+ "LABEL_8": 8,
193
+ "LABEL_80": 80,
194
+ "LABEL_81": 81,
195
+ "LABEL_82": 82,
196
+ "LABEL_83": 83,
197
+ "LABEL_84": 84,
198
+ "LABEL_85": 85,
199
+ "LABEL_86": 86,
200
+ "LABEL_87": 87,
201
+ "LABEL_88": 88,
202
+ "LABEL_89": 89,
203
+ "LABEL_9": 9,
204
+ "LABEL_90": 90,
205
+ "LABEL_91": 91,
206
+ "LABEL_92": 92,
207
+ "LABEL_93": 93,
208
+ "LABEL_94": 94,
209
+ "LABEL_95": 95
210
+ },
211
+ "layer_norm_eps": 1e-05,
212
+ "max_position_embeddings": 514,
213
+ "model_type": "roberta",
214
+ "num_attention_heads": 12,
215
+ "num_hidden_layers": 12,
216
+ "pad_token_id": 1,
217
+ "position_embedding_type": "absolute",
218
+ "problem_type": "multi_label_classification",
219
+ "torch_dtype": "float32",
220
+ "transformers_version": "4.45.2",
221
+ "type_vocab_size": 1,
222
+ "use_cache": true,
223
+ "vocab_size": 50265
224
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:873734583dc76016edb3cb77a1a19e62dd6206f7d392417e45082bb3d0dad76a
3
+ size 498901968
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:43d60051ab92a3e16f780e1bfbe35f970a371dd6592b115a9d450d46fe5d6515
3
+ size 997923834
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5d2f923f24bcdcbba38085cf8844870463ec171b81ddd1c64683c7c6cf2e521d
3
+ size 14244
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca6a62aca3a17821ef122c20fc35b4b7200d5c78eb6173340f0bcc2ff1d33aba
3
+ size 1064
special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "0": {
5
+ "content": "<s>",
6
+ "lstrip": false,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "1": {
13
+ "content": "<pad>",
14
+ "lstrip": false,
15
+ "normalized": true,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "2": {
21
+ "content": "</s>",
22
+ "lstrip": false,
23
+ "normalized": true,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": true
27
+ },
28
+ "3": {
29
+ "content": "<unk>",
30
+ "lstrip": false,
31
+ "normalized": true,
32
+ "rstrip": false,
33
+ "single_word": false,
34
+ "special": true
35
+ },
36
+ "50264": {
37
+ "content": "<mask>",
38
+ "lstrip": true,
39
+ "normalized": false,
40
+ "rstrip": false,
41
+ "single_word": false,
42
+ "special": true
43
+ }
44
+ },
45
+ "bos_token": "<s>",
46
+ "clean_up_tokenization_spaces": false,
47
+ "cls_token": "<s>",
48
+ "eos_token": "</s>",
49
+ "errors": "replace",
50
+ "mask_token": "<mask>",
51
+ "model_max_length": 512,
52
+ "pad_token": "<pad>",
53
+ "sep_token": "</s>",
54
+ "tokenizer_class": "RobertaTokenizer",
55
+ "trim_offsets": true,
56
+ "unk_token": "<unk>"
57
+ }
trainer_state.json ADDED
@@ -0,0 +1,3395 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9470943315331984,
3
+ "best_model_checkpoint": "/shared/3/projects/hiatus/tagged_data/models/roberta-base/binary-finetune-full/results/checkpoint-168317",
4
+ "epoch": 1.760049774157839,
5
+ "eval_steps": 9901,
6
+ "global_step": 217822,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.004040110214206643,
13
+ "grad_norm": 0.24907590448856354,
14
+ "learning_rate": 1.9989899724464485e-05,
15
+ "loss": 0.3886,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 0.008080220428413287,
20
+ "grad_norm": 0.19640618562698364,
21
+ "learning_rate": 1.997979944892897e-05,
22
+ "loss": 0.2835,
23
+ "step": 1000
24
+ },
25
+ {
26
+ "epoch": 0.01212033064261993,
27
+ "grad_norm": 0.21297627687454224,
28
+ "learning_rate": 1.9969699173393452e-05,
29
+ "loss": 0.2123,
30
+ "step": 1500
31
+ },
32
+ {
33
+ "epoch": 0.016160440856826573,
34
+ "grad_norm": 0.1339714229106903,
35
+ "learning_rate": 1.9959598897857935e-05,
36
+ "loss": 0.1631,
37
+ "step": 2000
38
+ },
39
+ {
40
+ "epoch": 0.020200551071033218,
41
+ "grad_norm": 0.20592452585697174,
42
+ "learning_rate": 1.994949862232242e-05,
43
+ "loss": 0.1306,
44
+ "step": 2500
45
+ },
46
+ {
47
+ "epoch": 0.02424066128523986,
48
+ "grad_norm": 0.1734761744737625,
49
+ "learning_rate": 1.9939398346786902e-05,
50
+ "loss": 0.1134,
51
+ "step": 3000
52
+ },
53
+ {
54
+ "epoch": 0.028280771499446505,
55
+ "grad_norm": 0.1429387331008911,
56
+ "learning_rate": 1.9929298071251385e-05,
57
+ "loss": 0.0954,
58
+ "step": 3500
59
+ },
60
+ {
61
+ "epoch": 0.032320881713653146,
62
+ "grad_norm": 0.13530635833740234,
63
+ "learning_rate": 1.991919779571587e-05,
64
+ "loss": 0.0833,
65
+ "step": 4000
66
+ },
67
+ {
68
+ "epoch": 0.03636099192785979,
69
+ "grad_norm": 0.16425204277038574,
70
+ "learning_rate": 1.9909097520180352e-05,
71
+ "loss": 0.0739,
72
+ "step": 4500
73
+ },
74
+ {
75
+ "epoch": 0.040401102142066436,
76
+ "grad_norm": 0.15035004913806915,
77
+ "learning_rate": 1.9898997244644835e-05,
78
+ "loss": 0.0657,
79
+ "step": 5000
80
+ },
81
+ {
82
+ "epoch": 0.04444121235627308,
83
+ "grad_norm": 0.1048048660159111,
84
+ "learning_rate": 1.988889696910932e-05,
85
+ "loss": 0.0606,
86
+ "step": 5500
87
+ },
88
+ {
89
+ "epoch": 0.04848132257047972,
90
+ "grad_norm": 0.1167823076248169,
91
+ "learning_rate": 1.9878796693573802e-05,
92
+ "loss": 0.0555,
93
+ "step": 6000
94
+ },
95
+ {
96
+ "epoch": 0.052521432784686364,
97
+ "grad_norm": 0.17904673516750336,
98
+ "learning_rate": 1.9868696418038285e-05,
99
+ "loss": 0.0517,
100
+ "step": 6500
101
+ },
102
+ {
103
+ "epoch": 0.05656154299889301,
104
+ "grad_norm": 0.22022856771945953,
105
+ "learning_rate": 1.985859614250277e-05,
106
+ "loss": 0.0483,
107
+ "step": 7000
108
+ },
109
+ {
110
+ "epoch": 0.060601653213099654,
111
+ "grad_norm": 0.11566773056983948,
112
+ "learning_rate": 1.9848495866967252e-05,
113
+ "loss": 0.0445,
114
+ "step": 7500
115
+ },
116
+ {
117
+ "epoch": 0.06464176342730629,
118
+ "grad_norm": 0.09606140851974487,
119
+ "learning_rate": 1.9838395591431736e-05,
120
+ "loss": 0.0434,
121
+ "step": 8000
122
+ },
123
+ {
124
+ "epoch": 0.06868187364151294,
125
+ "grad_norm": 0.14014209806919098,
126
+ "learning_rate": 1.982829531589622e-05,
127
+ "loss": 0.0415,
128
+ "step": 8500
129
+ },
130
+ {
131
+ "epoch": 0.07272198385571958,
132
+ "grad_norm": 0.1238802894949913,
133
+ "learning_rate": 1.9818195040360702e-05,
134
+ "loss": 0.0401,
135
+ "step": 9000
136
+ },
137
+ {
138
+ "epoch": 0.07676209406992623,
139
+ "grad_norm": 0.27284470200538635,
140
+ "learning_rate": 1.9808094764825186e-05,
141
+ "loss": 0.0375,
142
+ "step": 9500
143
+ },
144
+ {
145
+ "epoch": 0.08000226246171996,
146
+ "eval_f1_macro": 0.8895664094153212,
147
+ "eval_f1_micro": 0.9635283563255777,
148
+ "eval_loss": 0.11441826820373535,
149
+ "eval_precision_macro": 0.924616270515263,
150
+ "eval_precision_micro": 0.9673416146866455,
151
+ "eval_recall_macro": 0.8766278018110966,
152
+ "eval_recall_micro": 0.9597450436296798,
153
+ "eval_runtime": 13336.1892,
154
+ "eval_samples_per_second": 2.97,
155
+ "eval_steps_per_second": 0.012,
156
+ "step": 9901
157
+ },
158
+ {
159
+ "epoch": 0.08080220428413287,
160
+ "grad_norm": 0.07111234217882156,
161
+ "learning_rate": 1.979799448928967e-05,
162
+ "loss": 0.037,
163
+ "step": 10000
164
+ },
165
+ {
166
+ "epoch": 0.08484231449833951,
167
+ "grad_norm": 0.11708366125822067,
168
+ "learning_rate": 1.9787894213754152e-05,
169
+ "loss": 0.0337,
170
+ "step": 10500
171
+ },
172
+ {
173
+ "epoch": 0.08888242471254616,
174
+ "grad_norm": 0.12418048083782196,
175
+ "learning_rate": 1.9777793938218636e-05,
176
+ "loss": 0.0331,
177
+ "step": 11000
178
+ },
179
+ {
180
+ "epoch": 0.0929225349267528,
181
+ "grad_norm": 0.09599123895168304,
182
+ "learning_rate": 1.976769366268312e-05,
183
+ "loss": 0.032,
184
+ "step": 11500
185
+ },
186
+ {
187
+ "epoch": 0.09696264514095944,
188
+ "grad_norm": 0.08350682258605957,
189
+ "learning_rate": 1.9757593387147602e-05,
190
+ "loss": 0.03,
191
+ "step": 12000
192
+ },
193
+ {
194
+ "epoch": 0.10100275535516609,
195
+ "grad_norm": 0.1321333944797516,
196
+ "learning_rate": 1.9747493111612086e-05,
197
+ "loss": 0.03,
198
+ "step": 12500
199
+ },
200
+ {
201
+ "epoch": 0.10504286556937273,
202
+ "grad_norm": 0.09571998566389084,
203
+ "learning_rate": 1.973739283607657e-05,
204
+ "loss": 0.0281,
205
+ "step": 13000
206
+ },
207
+ {
208
+ "epoch": 0.10908297578357938,
209
+ "grad_norm": 0.11053865402936935,
210
+ "learning_rate": 1.9727292560541053e-05,
211
+ "loss": 0.0277,
212
+ "step": 13500
213
+ },
214
+ {
215
+ "epoch": 0.11312308599778602,
216
+ "grad_norm": 0.06845366209745407,
217
+ "learning_rate": 1.9717192285005536e-05,
218
+ "loss": 0.0274,
219
+ "step": 14000
220
+ },
221
+ {
222
+ "epoch": 0.11716319621199266,
223
+ "grad_norm": 0.11518129706382751,
224
+ "learning_rate": 1.970709200947002e-05,
225
+ "loss": 0.0267,
226
+ "step": 14500
227
+ },
228
+ {
229
+ "epoch": 0.12120330642619931,
230
+ "grad_norm": 0.09419895708560944,
231
+ "learning_rate": 1.9696991733934503e-05,
232
+ "loss": 0.0259,
233
+ "step": 15000
234
+ },
235
+ {
236
+ "epoch": 0.12524341664040595,
237
+ "grad_norm": 0.11040966212749481,
238
+ "learning_rate": 1.9686891458398986e-05,
239
+ "loss": 0.0258,
240
+ "step": 15500
241
+ },
242
+ {
243
+ "epoch": 0.12928352685461258,
244
+ "grad_norm": 0.08434844762086868,
245
+ "learning_rate": 1.967679118286347e-05,
246
+ "loss": 0.0253,
247
+ "step": 16000
248
+ },
249
+ {
250
+ "epoch": 0.13332363706881925,
251
+ "grad_norm": 0.13895830512046814,
252
+ "learning_rate": 1.9666690907327953e-05,
253
+ "loss": 0.0247,
254
+ "step": 16500
255
+ },
256
+ {
257
+ "epoch": 0.1373637472830259,
258
+ "grad_norm": 0.10540761798620224,
259
+ "learning_rate": 1.9656590631792436e-05,
260
+ "loss": 0.0245,
261
+ "step": 17000
262
+ },
263
+ {
264
+ "epoch": 0.14140385749723253,
265
+ "grad_norm": 0.09767123311758041,
266
+ "learning_rate": 1.964649035625692e-05,
267
+ "loss": 0.0245,
268
+ "step": 17500
269
+ },
270
+ {
271
+ "epoch": 0.14544396771143916,
272
+ "grad_norm": 0.08917172998189926,
273
+ "learning_rate": 1.9636390080721403e-05,
274
+ "loss": 0.0251,
275
+ "step": 18000
276
+ },
277
+ {
278
+ "epoch": 0.1494840779256458,
279
+ "grad_norm": 0.13313362002372742,
280
+ "learning_rate": 1.9626289805185886e-05,
281
+ "loss": 0.0252,
282
+ "step": 18500
283
+ },
284
+ {
285
+ "epoch": 0.15352418813985247,
286
+ "grad_norm": 0.12236423790454865,
287
+ "learning_rate": 1.961618952965037e-05,
288
+ "loss": 0.0251,
289
+ "step": 19000
290
+ },
291
+ {
292
+ "epoch": 0.1575642983540591,
293
+ "grad_norm": 0.09876661747694016,
294
+ "learning_rate": 1.9606089254114853e-05,
295
+ "loss": 0.0249,
296
+ "step": 19500
297
+ },
298
+ {
299
+ "epoch": 0.16000452492343992,
300
+ "eval_f1_macro": 0.9035923074831559,
301
+ "eval_f1_micro": 0.9606367859628728,
302
+ "eval_loss": 0.11549682915210724,
303
+ "eval_precision_macro": 0.9036435627053582,
304
+ "eval_precision_micro": 0.9530018135782053,
305
+ "eval_recall_macro": 0.9133421606323336,
306
+ "eval_recall_micro": 0.9683950814868224,
307
+ "eval_runtime": 15280.9037,
308
+ "eval_samples_per_second": 2.592,
309
+ "eval_steps_per_second": 0.01,
310
+ "step": 19802
311
+ },
312
+ {
313
+ "epoch": 0.16160440856826574,
314
+ "grad_norm": 0.057935502380132675,
315
+ "learning_rate": 1.9595988978579336e-05,
316
+ "loss": 0.0246,
317
+ "step": 20000
318
+ },
319
+ {
320
+ "epoch": 0.16564451878247238,
321
+ "grad_norm": 0.13269482553005219,
322
+ "learning_rate": 1.958588870304382e-05,
323
+ "loss": 0.0227,
324
+ "step": 20500
325
+ },
326
+ {
327
+ "epoch": 0.16968462899667902,
328
+ "grad_norm": 0.09697619825601578,
329
+ "learning_rate": 1.9575788427508303e-05,
330
+ "loss": 0.0229,
331
+ "step": 21000
332
+ },
333
+ {
334
+ "epoch": 0.17372473921088566,
335
+ "grad_norm": 0.2296031266450882,
336
+ "learning_rate": 1.9565688151972786e-05,
337
+ "loss": 0.0222,
338
+ "step": 21500
339
+ },
340
+ {
341
+ "epoch": 0.17776484942509233,
342
+ "grad_norm": 0.07054860144853592,
343
+ "learning_rate": 1.955558787643727e-05,
344
+ "loss": 0.0217,
345
+ "step": 22000
346
+ },
347
+ {
348
+ "epoch": 0.18180495963929896,
349
+ "grad_norm": 0.14594444632530212,
350
+ "learning_rate": 1.9545487600901753e-05,
351
+ "loss": 0.0216,
352
+ "step": 22500
353
+ },
354
+ {
355
+ "epoch": 0.1858450698535056,
356
+ "grad_norm": 0.1951671838760376,
357
+ "learning_rate": 1.9535387325366237e-05,
358
+ "loss": 0.0236,
359
+ "step": 23000
360
+ },
361
+ {
362
+ "epoch": 0.18988518006771224,
363
+ "grad_norm": 0.09014302492141724,
364
+ "learning_rate": 1.952528704983072e-05,
365
+ "loss": 0.0241,
366
+ "step": 23500
367
+ },
368
+ {
369
+ "epoch": 0.19392529028191888,
370
+ "grad_norm": 0.07351396232843399,
371
+ "learning_rate": 1.9515186774295207e-05,
372
+ "loss": 0.0246,
373
+ "step": 24000
374
+ },
375
+ {
376
+ "epoch": 0.19796540049612554,
377
+ "grad_norm": 0.13433478772640228,
378
+ "learning_rate": 1.950508649875969e-05,
379
+ "loss": 0.0237,
380
+ "step": 24500
381
+ },
382
+ {
383
+ "epoch": 0.20200551071033218,
384
+ "grad_norm": 0.09406758099794388,
385
+ "learning_rate": 1.949498622322417e-05,
386
+ "loss": 0.0246,
387
+ "step": 25000
388
+ },
389
+ {
390
+ "epoch": 0.20604562092453882,
391
+ "grad_norm": 0.0723891332745552,
392
+ "learning_rate": 1.9484885947688653e-05,
393
+ "loss": 0.0248,
394
+ "step": 25500
395
+ },
396
+ {
397
+ "epoch": 0.21008573113874546,
398
+ "grad_norm": 0.0635887160897255,
399
+ "learning_rate": 1.9474785672153137e-05,
400
+ "loss": 0.0243,
401
+ "step": 26000
402
+ },
403
+ {
404
+ "epoch": 0.2141258413529521,
405
+ "grad_norm": 0.12226919084787369,
406
+ "learning_rate": 1.9464685396617624e-05,
407
+ "loss": 0.0246,
408
+ "step": 26500
409
+ },
410
+ {
411
+ "epoch": 0.21816595156715876,
412
+ "grad_norm": 0.17104922235012054,
413
+ "learning_rate": 1.9454585121082107e-05,
414
+ "loss": 0.0244,
415
+ "step": 27000
416
+ },
417
+ {
418
+ "epoch": 0.2222060617813654,
419
+ "grad_norm": 0.07366731762886047,
420
+ "learning_rate": 1.9444484845546587e-05,
421
+ "loss": 0.0236,
422
+ "step": 27500
423
+ },
424
+ {
425
+ "epoch": 0.22624617199557204,
426
+ "grad_norm": 0.05332706496119499,
427
+ "learning_rate": 1.943438457001107e-05,
428
+ "loss": 0.0191,
429
+ "step": 28000
430
+ },
431
+ {
432
+ "epoch": 0.23028628220977868,
433
+ "grad_norm": 0.12188129127025604,
434
+ "learning_rate": 1.9424284294475554e-05,
435
+ "loss": 0.0184,
436
+ "step": 28500
437
+ },
438
+ {
439
+ "epoch": 0.2343263924239853,
440
+ "grad_norm": 0.07722073048353195,
441
+ "learning_rate": 1.941418401894004e-05,
442
+ "loss": 0.018,
443
+ "step": 29000
444
+ },
445
+ {
446
+ "epoch": 0.23836650263819198,
447
+ "grad_norm": 0.07803859561681747,
448
+ "learning_rate": 1.9404083743404524e-05,
449
+ "loss": 0.0176,
450
+ "step": 29500
451
+ },
452
+ {
453
+ "epoch": 0.24000678738515988,
454
+ "eval_f1_macro": 0.9189758616039659,
455
+ "eval_f1_micro": 0.9664452050510047,
456
+ "eval_loss": 0.10743161290884018,
457
+ "eval_precision_macro": 0.9454764579982075,
458
+ "eval_precision_micro": 0.9680139971687907,
459
+ "eval_recall_macro": 0.9051902291376027,
460
+ "eval_recall_micro": 0.9648814895676969,
461
+ "eval_runtime": 13586.8138,
462
+ "eval_samples_per_second": 2.915,
463
+ "eval_steps_per_second": 0.011,
464
+ "step": 29703
465
+ },
466
+ {
467
+ "epoch": 0.24240661285239862,
468
+ "grad_norm": 0.04772321879863739,
469
+ "learning_rate": 1.9393983467869004e-05,
470
+ "loss": 0.0177,
471
+ "step": 30000
472
+ },
473
+ {
474
+ "epoch": 0.24644672306660526,
475
+ "grad_norm": 0.11928682029247284,
476
+ "learning_rate": 1.9383883192333487e-05,
477
+ "loss": 0.0229,
478
+ "step": 30500
479
+ },
480
+ {
481
+ "epoch": 0.2504868332808119,
482
+ "grad_norm": 0.08102133870124817,
483
+ "learning_rate": 1.937378291679797e-05,
484
+ "loss": 0.0228,
485
+ "step": 31000
486
+ },
487
+ {
488
+ "epoch": 0.25452694349501853,
489
+ "grad_norm": 0.073371522128582,
490
+ "learning_rate": 1.9363682641262457e-05,
491
+ "loss": 0.0231,
492
+ "step": 31500
493
+ },
494
+ {
495
+ "epoch": 0.25856705370922517,
496
+ "grad_norm": 0.08793163299560547,
497
+ "learning_rate": 1.935358236572694e-05,
498
+ "loss": 0.0229,
499
+ "step": 32000
500
+ },
501
+ {
502
+ "epoch": 0.2626071639234318,
503
+ "grad_norm": 0.12065927684307098,
504
+ "learning_rate": 1.934348209019142e-05,
505
+ "loss": 0.0231,
506
+ "step": 32500
507
+ },
508
+ {
509
+ "epoch": 0.2666472741376385,
510
+ "grad_norm": 0.10052382946014404,
511
+ "learning_rate": 1.9333381814655904e-05,
512
+ "loss": 0.0218,
513
+ "step": 33000
514
+ },
515
+ {
516
+ "epoch": 0.27068738435184514,
517
+ "grad_norm": 0.08048272132873535,
518
+ "learning_rate": 1.9323281539120387e-05,
519
+ "loss": 0.0227,
520
+ "step": 33500
521
+ },
522
+ {
523
+ "epoch": 0.2747274945660518,
524
+ "grad_norm": 0.05087564140558243,
525
+ "learning_rate": 1.9313181263584874e-05,
526
+ "loss": 0.0206,
527
+ "step": 34000
528
+ },
529
+ {
530
+ "epoch": 0.2787676047802584,
531
+ "grad_norm": 0.13495181500911713,
532
+ "learning_rate": 1.9303080988049357e-05,
533
+ "loss": 0.0213,
534
+ "step": 34500
535
+ },
536
+ {
537
+ "epoch": 0.28280771499446505,
538
+ "grad_norm": 0.104469895362854,
539
+ "learning_rate": 1.9292980712513837e-05,
540
+ "loss": 0.0215,
541
+ "step": 35000
542
+ },
543
+ {
544
+ "epoch": 0.2868478252086717,
545
+ "grad_norm": 0.10830747336149216,
546
+ "learning_rate": 1.928288043697832e-05,
547
+ "loss": 0.0232,
548
+ "step": 35500
549
+ },
550
+ {
551
+ "epoch": 0.29088793542287833,
552
+ "grad_norm": 0.07727912068367004,
553
+ "learning_rate": 1.9272780161442804e-05,
554
+ "loss": 0.0239,
555
+ "step": 36000
556
+ },
557
+ {
558
+ "epoch": 0.29492804563708497,
559
+ "grad_norm": 0.11781858652830124,
560
+ "learning_rate": 1.926267988590729e-05,
561
+ "loss": 0.0229,
562
+ "step": 36500
563
+ },
564
+ {
565
+ "epoch": 0.2989681558512916,
566
+ "grad_norm": 0.09465543925762177,
567
+ "learning_rate": 1.9252579610371774e-05,
568
+ "loss": 0.0241,
569
+ "step": 37000
570
+ },
571
+ {
572
+ "epoch": 0.30300826606549824,
573
+ "grad_norm": 0.08024444431066513,
574
+ "learning_rate": 1.9242479334836254e-05,
575
+ "loss": 0.0236,
576
+ "step": 37500
577
+ },
578
+ {
579
+ "epoch": 0.30704837627970494,
580
+ "grad_norm": 0.06720072776079178,
581
+ "learning_rate": 1.9232379059300738e-05,
582
+ "loss": 0.0211,
583
+ "step": 38000
584
+ },
585
+ {
586
+ "epoch": 0.3110884864939116,
587
+ "grad_norm": 0.13362745940685272,
588
+ "learning_rate": 1.922227878376522e-05,
589
+ "loss": 0.0203,
590
+ "step": 38500
591
+ },
592
+ {
593
+ "epoch": 0.3151285967081182,
594
+ "grad_norm": 0.09114370495080948,
595
+ "learning_rate": 1.9212178508229708e-05,
596
+ "loss": 0.0203,
597
+ "step": 39000
598
+ },
599
+ {
600
+ "epoch": 0.31916870692232485,
601
+ "grad_norm": 0.07381443679332733,
602
+ "learning_rate": 1.920207823269419e-05,
603
+ "loss": 0.0206,
604
+ "step": 39500
605
+ },
606
+ {
607
+ "epoch": 0.32000904984687983,
608
+ "eval_f1_macro": 0.9235874471417507,
609
+ "eval_f1_micro": 0.9663972411435532,
610
+ "eval_loss": 0.10308045893907547,
611
+ "eval_precision_macro": 0.9426979961917242,
612
+ "eval_precision_micro": 0.9680776634557233,
613
+ "eval_recall_macro": 0.9108391105037915,
614
+ "eval_recall_micro": 0.9647226425913261,
615
+ "eval_runtime": 13873.9388,
616
+ "eval_samples_per_second": 2.854,
617
+ "eval_steps_per_second": 0.011,
618
+ "step": 39604
619
+ },
620
+ {
621
+ "epoch": 0.3232088171365315,
622
+ "grad_norm": 0.07686352729797363,
623
+ "learning_rate": 1.9191977957158674e-05,
624
+ "loss": 0.0204,
625
+ "step": 40000
626
+ },
627
+ {
628
+ "epoch": 0.32724892735073813,
629
+ "grad_norm": 0.15459179878234863,
630
+ "learning_rate": 1.9181877681623154e-05,
631
+ "loss": 0.0225,
632
+ "step": 40500
633
+ },
634
+ {
635
+ "epoch": 0.33128903756494477,
636
+ "grad_norm": 0.11474985629320145,
637
+ "learning_rate": 1.9171777406087638e-05,
638
+ "loss": 0.0229,
639
+ "step": 41000
640
+ },
641
+ {
642
+ "epoch": 0.3353291477791514,
643
+ "grad_norm": 0.09817365556955338,
644
+ "learning_rate": 1.9161677130552125e-05,
645
+ "loss": 0.0229,
646
+ "step": 41500
647
+ },
648
+ {
649
+ "epoch": 0.33936925799335804,
650
+ "grad_norm": 0.07288320362567902,
651
+ "learning_rate": 1.9151576855016608e-05,
652
+ "loss": 0.0222,
653
+ "step": 42000
654
+ },
655
+ {
656
+ "epoch": 0.3434093682075647,
657
+ "grad_norm": 0.13258545100688934,
658
+ "learning_rate": 1.914147657948109e-05,
659
+ "loss": 0.0221,
660
+ "step": 42500
661
+ },
662
+ {
663
+ "epoch": 0.3474494784217713,
664
+ "grad_norm": 0.09609493613243103,
665
+ "learning_rate": 1.913137630394557e-05,
666
+ "loss": 0.0224,
667
+ "step": 43000
668
+ },
669
+ {
670
+ "epoch": 0.351489588635978,
671
+ "grad_norm": 0.0800900012254715,
672
+ "learning_rate": 1.9121276028410055e-05,
673
+ "loss": 0.022,
674
+ "step": 43500
675
+ },
676
+ {
677
+ "epoch": 0.35552969885018465,
678
+ "grad_norm": 0.06640051305294037,
679
+ "learning_rate": 1.911117575287454e-05,
680
+ "loss": 0.0209,
681
+ "step": 44000
682
+ },
683
+ {
684
+ "epoch": 0.3595698090643913,
685
+ "grad_norm": 0.13987226784229279,
686
+ "learning_rate": 1.9101075477339025e-05,
687
+ "loss": 0.022,
688
+ "step": 44500
689
+ },
690
+ {
691
+ "epoch": 0.3636099192785979,
692
+ "grad_norm": 0.08626226335763931,
693
+ "learning_rate": 1.9090975201803508e-05,
694
+ "loss": 0.0219,
695
+ "step": 45000
696
+ },
697
+ {
698
+ "epoch": 0.36765002949280456,
699
+ "grad_norm": 0.09093815088272095,
700
+ "learning_rate": 1.9080874926267988e-05,
701
+ "loss": 0.0191,
702
+ "step": 45500
703
+ },
704
+ {
705
+ "epoch": 0.3716901397070112,
706
+ "grad_norm": 0.062450163066387177,
707
+ "learning_rate": 1.907077465073247e-05,
708
+ "loss": 0.0192,
709
+ "step": 46000
710
+ },
711
+ {
712
+ "epoch": 0.37573024992121784,
713
+ "grad_norm": 0.14638446271419525,
714
+ "learning_rate": 1.9060674375196958e-05,
715
+ "loss": 0.0186,
716
+ "step": 46500
717
+ },
718
+ {
719
+ "epoch": 0.3797703601354245,
720
+ "grad_norm": 0.08730041235685349,
721
+ "learning_rate": 1.905057409966144e-05,
722
+ "loss": 0.019,
723
+ "step": 47000
724
+ },
725
+ {
726
+ "epoch": 0.3838104703496311,
727
+ "grad_norm": 0.09185372292995453,
728
+ "learning_rate": 1.9040473824125925e-05,
729
+ "loss": 0.0189,
730
+ "step": 47500
731
+ },
732
+ {
733
+ "epoch": 0.38785058056383775,
734
+ "grad_norm": 0.05995471775531769,
735
+ "learning_rate": 1.9030373548590405e-05,
736
+ "loss": 0.0174,
737
+ "step": 48000
738
+ },
739
+ {
740
+ "epoch": 0.39189069077804445,
741
+ "grad_norm": 0.14513157308101654,
742
+ "learning_rate": 1.902027327305489e-05,
743
+ "loss": 0.0167,
744
+ "step": 48500
745
+ },
746
+ {
747
+ "epoch": 0.3959308009922511,
748
+ "grad_norm": 0.08400790393352509,
749
+ "learning_rate": 1.9010172997519375e-05,
750
+ "loss": 0.0165,
751
+ "step": 49000
752
+ },
753
+ {
754
+ "epoch": 0.3999709112064577,
755
+ "grad_norm": 0.0705028846859932,
756
+ "learning_rate": 1.900007272198386e-05,
757
+ "loss": 0.0165,
758
+ "step": 49500
759
+ },
760
+ {
761
+ "epoch": 0.40001131230859976,
762
+ "eval_f1_macro": 0.9281014680466918,
763
+ "eval_f1_micro": 0.9684054719516875,
764
+ "eval_loss": 0.10883225500583649,
765
+ "eval_precision_macro": 0.9609447042869569,
766
+ "eval_precision_micro": 0.97977646274136,
767
+ "eval_recall_macro": 0.9023603199352568,
768
+ "eval_recall_micro": 0.9572953897303177,
769
+ "eval_runtime": 14645.8089,
770
+ "eval_samples_per_second": 2.704,
771
+ "eval_steps_per_second": 0.011,
772
+ "step": 49505
773
+ },
774
+ {
775
+ "epoch": 0.40401102142066436,
776
+ "grad_norm": 0.06206486374139786,
777
+ "learning_rate": 1.8989972446448342e-05,
778
+ "loss": 0.0163,
779
+ "step": 50000
780
+ },
781
+ {
782
+ "epoch": 0.408051131634871,
783
+ "grad_norm": 0.13632065057754517,
784
+ "learning_rate": 1.8979872170912822e-05,
785
+ "loss": 0.0175,
786
+ "step": 50500
787
+ },
788
+ {
789
+ "epoch": 0.41209124184907764,
790
+ "grad_norm": 0.10581111907958984,
791
+ "learning_rate": 1.896977189537731e-05,
792
+ "loss": 0.0179,
793
+ "step": 51000
794
+ },
795
+ {
796
+ "epoch": 0.4161313520632843,
797
+ "grad_norm": 0.05609723553061485,
798
+ "learning_rate": 1.8959671619841792e-05,
799
+ "loss": 0.0171,
800
+ "step": 51500
801
+ },
802
+ {
803
+ "epoch": 0.4201714622774909,
804
+ "grad_norm": 0.0569671131670475,
805
+ "learning_rate": 1.8949571344306275e-05,
806
+ "loss": 0.0177,
807
+ "step": 52000
808
+ },
809
+ {
810
+ "epoch": 0.42421157249169755,
811
+ "grad_norm": 0.12548725306987762,
812
+ "learning_rate": 1.893947106877076e-05,
813
+ "loss": 0.0179,
814
+ "step": 52500
815
+ },
816
+ {
817
+ "epoch": 0.4282516827059042,
818
+ "grad_norm": 0.14123043417930603,
819
+ "learning_rate": 1.892937079323524e-05,
820
+ "loss": 0.0195,
821
+ "step": 53000
822
+ },
823
+ {
824
+ "epoch": 0.4322917929201109,
825
+ "grad_norm": 0.07868105173110962,
826
+ "learning_rate": 1.8919270517699725e-05,
827
+ "loss": 0.0195,
828
+ "step": 53500
829
+ },
830
+ {
831
+ "epoch": 0.4363319031343175,
832
+ "grad_norm": 0.0551162026822567,
833
+ "learning_rate": 1.890917024216421e-05,
834
+ "loss": 0.0194,
835
+ "step": 54000
836
+ },
837
+ {
838
+ "epoch": 0.44037201334852416,
839
+ "grad_norm": 0.12377525120973587,
840
+ "learning_rate": 1.8899069966628692e-05,
841
+ "loss": 0.0193,
842
+ "step": 54500
843
+ },
844
+ {
845
+ "epoch": 0.4444121235627308,
846
+ "grad_norm": 0.07947281748056412,
847
+ "learning_rate": 1.8888969691093175e-05,
848
+ "loss": 0.0195,
849
+ "step": 55000
850
+ },
851
+ {
852
+ "epoch": 0.44845223377693744,
853
+ "grad_norm": 0.07180605828762054,
854
+ "learning_rate": 1.887886941555766e-05,
855
+ "loss": 0.0186,
856
+ "step": 55500
857
+ },
858
+ {
859
+ "epoch": 0.4524923439911441,
860
+ "grad_norm": 0.0590415820479393,
861
+ "learning_rate": 1.8868769140022142e-05,
862
+ "loss": 0.0186,
863
+ "step": 56000
864
+ },
865
+ {
866
+ "epoch": 0.4565324542053507,
867
+ "grad_norm": 0.12405771017074585,
868
+ "learning_rate": 1.8858668864486626e-05,
869
+ "loss": 0.0181,
870
+ "step": 56500
871
+ },
872
+ {
873
+ "epoch": 0.46057256441955735,
874
+ "grad_norm": 0.09074413031339645,
875
+ "learning_rate": 1.884856858895111e-05,
876
+ "loss": 0.0178,
877
+ "step": 57000
878
+ },
879
+ {
880
+ "epoch": 0.464612674633764,
881
+ "grad_norm": 0.12590628862380981,
882
+ "learning_rate": 1.8838468313415592e-05,
883
+ "loss": 0.0178,
884
+ "step": 57500
885
+ },
886
+ {
887
+ "epoch": 0.4686527848479706,
888
+ "grad_norm": 0.047191109508275986,
889
+ "learning_rate": 1.8828368037880076e-05,
890
+ "loss": 0.0154,
891
+ "step": 58000
892
+ },
893
+ {
894
+ "epoch": 0.4726928950621773,
895
+ "grad_norm": 0.13741852343082428,
896
+ "learning_rate": 1.881826776234456e-05,
897
+ "loss": 0.0156,
898
+ "step": 58500
899
+ },
900
+ {
901
+ "epoch": 0.47673300527638396,
902
+ "grad_norm": 0.07473180443048477,
903
+ "learning_rate": 1.8808167486809042e-05,
904
+ "loss": 0.0152,
905
+ "step": 59000
906
+ },
907
+ {
908
+ "epoch": 0.48001357477031975,
909
+ "eval_f1_macro": 0.9238212336154783,
910
+ "eval_f1_micro": 0.9658474370811376,
911
+ "eval_loss": 0.10412032902240753,
912
+ "eval_precision_macro": 0.9360090910956397,
913
+ "eval_precision_micro": 0.9629420110715917,
914
+ "eval_recall_macro": 0.9167736650129396,
915
+ "eval_recall_micro": 0.9687704488794511,
916
+ "eval_runtime": 13864.7476,
917
+ "eval_samples_per_second": 2.856,
918
+ "eval_steps_per_second": 0.011,
919
+ "step": 59406
920
+ },
921
+ {
922
+ "epoch": 0.4807731154905906,
923
+ "grad_norm": 0.06410785764455795,
924
+ "learning_rate": 1.8798067211273526e-05,
925
+ "loss": 0.0147,
926
+ "step": 59500
927
+ },
928
+ {
929
+ "epoch": 0.48481322570479723,
930
+ "grad_norm": 0.05010313540697098,
931
+ "learning_rate": 1.878796693573801e-05,
932
+ "loss": 0.0156,
933
+ "step": 60000
934
+ },
935
+ {
936
+ "epoch": 0.4888533359190039,
937
+ "grad_norm": 0.14338257908821106,
938
+ "learning_rate": 1.8777866660202493e-05,
939
+ "loss": 0.0164,
940
+ "step": 60500
941
+ },
942
+ {
943
+ "epoch": 0.4928934461332105,
944
+ "grad_norm": 0.09123385697603226,
945
+ "learning_rate": 1.8767766384666976e-05,
946
+ "loss": 0.0174,
947
+ "step": 61000
948
+ },
949
+ {
950
+ "epoch": 0.49693355634741715,
951
+ "grad_norm": 0.07728511840105057,
952
+ "learning_rate": 1.875766610913146e-05,
953
+ "loss": 0.017,
954
+ "step": 61500
955
+ },
956
+ {
957
+ "epoch": 0.5009736665616238,
958
+ "grad_norm": 0.06151897832751274,
959
+ "learning_rate": 1.8747565833595943e-05,
960
+ "loss": 0.0172,
961
+ "step": 62000
962
+ },
963
+ {
964
+ "epoch": 0.5050137767758305,
965
+ "grad_norm": 0.14278863370418549,
966
+ "learning_rate": 1.8737465558060426e-05,
967
+ "loss": 0.0166,
968
+ "step": 62500
969
+ },
970
+ {
971
+ "epoch": 0.5090538869900371,
972
+ "grad_norm": 0.08395873010158539,
973
+ "learning_rate": 1.872736528252491e-05,
974
+ "loss": 0.0206,
975
+ "step": 63000
976
+ },
977
+ {
978
+ "epoch": 0.5130939972042438,
979
+ "grad_norm": 0.09704262018203735,
980
+ "learning_rate": 1.8717265006989393e-05,
981
+ "loss": 0.0208,
982
+ "step": 63500
983
+ },
984
+ {
985
+ "epoch": 0.5171341074184503,
986
+ "grad_norm": 0.06397638469934464,
987
+ "learning_rate": 1.8707164731453876e-05,
988
+ "loss": 0.0208,
989
+ "step": 64000
990
+ },
991
+ {
992
+ "epoch": 0.521174217632657,
993
+ "grad_norm": 0.1525479257106781,
994
+ "learning_rate": 1.869706445591836e-05,
995
+ "loss": 0.0207,
996
+ "step": 64500
997
+ },
998
+ {
999
+ "epoch": 0.5252143278468636,
1000
+ "grad_norm": 0.0878639966249466,
1001
+ "learning_rate": 1.8686964180382843e-05,
1002
+ "loss": 0.0198,
1003
+ "step": 65000
1004
+ },
1005
+ {
1006
+ "epoch": 0.5292544380610703,
1007
+ "grad_norm": 0.05913593992590904,
1008
+ "learning_rate": 1.8676863904847326e-05,
1009
+ "loss": 0.0166,
1010
+ "step": 65500
1011
+ },
1012
+ {
1013
+ "epoch": 0.533294548275277,
1014
+ "grad_norm": 0.05049494653940201,
1015
+ "learning_rate": 1.866676362931181e-05,
1016
+ "loss": 0.0166,
1017
+ "step": 66000
1018
+ },
1019
+ {
1020
+ "epoch": 0.5373346584894836,
1021
+ "grad_norm": 0.10428164154291153,
1022
+ "learning_rate": 1.8656663353776293e-05,
1023
+ "loss": 0.0174,
1024
+ "step": 66500
1025
+ },
1026
+ {
1027
+ "epoch": 0.5413747687036903,
1028
+ "grad_norm": 0.08380962908267975,
1029
+ "learning_rate": 1.8646563078240776e-05,
1030
+ "loss": 0.0165,
1031
+ "step": 67000
1032
+ },
1033
+ {
1034
+ "epoch": 0.5454148789178969,
1035
+ "grad_norm": 0.12970462441444397,
1036
+ "learning_rate": 1.863646280270526e-05,
1037
+ "loss": 0.0164,
1038
+ "step": 67500
1039
+ },
1040
+ {
1041
+ "epoch": 0.5494549891321036,
1042
+ "grad_norm": 0.12594661116600037,
1043
+ "learning_rate": 1.8626362527169743e-05,
1044
+ "loss": 0.0189,
1045
+ "step": 68000
1046
+ },
1047
+ {
1048
+ "epoch": 0.5534950993463101,
1049
+ "grad_norm": 0.11368534713983536,
1050
+ "learning_rate": 1.8616262251634226e-05,
1051
+ "loss": 0.019,
1052
+ "step": 68500
1053
+ },
1054
+ {
1055
+ "epoch": 0.5575352095605168,
1056
+ "grad_norm": 0.08376836031675339,
1057
+ "learning_rate": 1.860616197609871e-05,
1058
+ "loss": 0.0191,
1059
+ "step": 69000
1060
+ },
1061
+ {
1062
+ "epoch": 0.5600158372320397,
1063
+ "eval_f1_macro": 0.9256130258855139,
1064
+ "eval_f1_micro": 0.9668266025133863,
1065
+ "eval_loss": 0.10856343805789948,
1066
+ "eval_precision_macro": 0.9437511235105461,
1067
+ "eval_precision_micro": 0.9695209969914294,
1068
+ "eval_recall_macro": 0.9124750251974688,
1069
+ "eval_recall_micro": 0.9641471425069983,
1070
+ "eval_runtime": 14189.8384,
1071
+ "eval_samples_per_second": 2.791,
1072
+ "eval_steps_per_second": 0.011,
1073
+ "step": 69307
1074
+ },
1075
+ {
1076
+ "epoch": 0.5615753197747234,
1077
+ "grad_norm": 0.08000296354293823,
1078
+ "learning_rate": 1.8596061700563193e-05,
1079
+ "loss": 0.0188,
1080
+ "step": 69500
1081
+ },
1082
+ {
1083
+ "epoch": 0.5656154299889301,
1084
+ "grad_norm": 0.06347772479057312,
1085
+ "learning_rate": 1.8585961425027677e-05,
1086
+ "loss": 0.0187,
1087
+ "step": 70000
1088
+ },
1089
+ {
1090
+ "epoch": 0.5696555402031367,
1091
+ "grad_norm": 0.21189579367637634,
1092
+ "learning_rate": 1.857586114949216e-05,
1093
+ "loss": 0.0196,
1094
+ "step": 70500
1095
+ },
1096
+ {
1097
+ "epoch": 0.5736956504173434,
1098
+ "grad_norm": 0.07940568774938583,
1099
+ "learning_rate": 1.8565760873956643e-05,
1100
+ "loss": 0.0196,
1101
+ "step": 71000
1102
+ },
1103
+ {
1104
+ "epoch": 0.5777357606315501,
1105
+ "grad_norm": 0.07458707690238953,
1106
+ "learning_rate": 1.8555660598421127e-05,
1107
+ "loss": 0.0199,
1108
+ "step": 71500
1109
+ },
1110
+ {
1111
+ "epoch": 0.5817758708457567,
1112
+ "grad_norm": 0.0705709308385849,
1113
+ "learning_rate": 1.854556032288561e-05,
1114
+ "loss": 0.0193,
1115
+ "step": 72000
1116
+ },
1117
+ {
1118
+ "epoch": 0.5858159810599634,
1119
+ "grad_norm": 0.13246993720531464,
1120
+ "learning_rate": 1.8535460047350093e-05,
1121
+ "loss": 0.0195,
1122
+ "step": 72500
1123
+ },
1124
+ {
1125
+ "epoch": 0.5898560912741699,
1126
+ "grad_norm": 0.08721259236335754,
1127
+ "learning_rate": 1.8525359771814577e-05,
1128
+ "loss": 0.0196,
1129
+ "step": 73000
1130
+ },
1131
+ {
1132
+ "epoch": 0.5938962014883766,
1133
+ "grad_norm": 0.07570379972457886,
1134
+ "learning_rate": 1.851525949627906e-05,
1135
+ "loss": 0.0186,
1136
+ "step": 73500
1137
+ },
1138
+ {
1139
+ "epoch": 0.5979363117025832,
1140
+ "grad_norm": 0.07477313280105591,
1141
+ "learning_rate": 1.8505159220743543e-05,
1142
+ "loss": 0.0183,
1143
+ "step": 74000
1144
+ },
1145
+ {
1146
+ "epoch": 0.6019764219167899,
1147
+ "grad_norm": 0.15558893978595734,
1148
+ "learning_rate": 1.8495058945208027e-05,
1149
+ "loss": 0.0194,
1150
+ "step": 74500
1151
+ },
1152
+ {
1153
+ "epoch": 0.6060165321309965,
1154
+ "grad_norm": 0.08373390883207321,
1155
+ "learning_rate": 1.848495866967251e-05,
1156
+ "loss": 0.0189,
1157
+ "step": 75000
1158
+ },
1159
+ {
1160
+ "epoch": 0.6100566423452032,
1161
+ "grad_norm": 0.06340883672237396,
1162
+ "learning_rate": 1.8474858394136994e-05,
1163
+ "loss": 0.0139,
1164
+ "step": 75500
1165
+ },
1166
+ {
1167
+ "epoch": 0.6140967525594099,
1168
+ "grad_norm": 0.05438007041811943,
1169
+ "learning_rate": 1.8464758118601477e-05,
1170
+ "loss": 0.0148,
1171
+ "step": 76000
1172
+ },
1173
+ {
1174
+ "epoch": 0.6181368627736165,
1175
+ "grad_norm": 0.1218661442399025,
1176
+ "learning_rate": 1.845465784306596e-05,
1177
+ "loss": 0.0151,
1178
+ "step": 76500
1179
+ },
1180
+ {
1181
+ "epoch": 0.6221769729878232,
1182
+ "grad_norm": 0.0688873752951622,
1183
+ "learning_rate": 1.8444557567530444e-05,
1184
+ "loss": 0.0143,
1185
+ "step": 77000
1186
+ },
1187
+ {
1188
+ "epoch": 0.6262170832020297,
1189
+ "grad_norm": 0.058265481144189835,
1190
+ "learning_rate": 1.8434457291994927e-05,
1191
+ "loss": 0.0142,
1192
+ "step": 77500
1193
+ },
1194
+ {
1195
+ "epoch": 0.6302571934162364,
1196
+ "grad_norm": 0.046319037675857544,
1197
+ "learning_rate": 1.842435701645941e-05,
1198
+ "loss": 0.0142,
1199
+ "step": 78000
1200
+ },
1201
+ {
1202
+ "epoch": 0.634297303630443,
1203
+ "grad_norm": 0.14100997149944305,
1204
+ "learning_rate": 1.8414256740923894e-05,
1205
+ "loss": 0.0146,
1206
+ "step": 78500
1207
+ },
1208
+ {
1209
+ "epoch": 0.6383374138446497,
1210
+ "grad_norm": 0.10154972225427628,
1211
+ "learning_rate": 1.8404156465388377e-05,
1212
+ "loss": 0.0146,
1213
+ "step": 79000
1214
+ },
1215
+ {
1216
+ "epoch": 0.6400180996937597,
1217
+ "eval_f1_macro": 0.9268223044157705,
1218
+ "eval_f1_micro": 0.9675749211491975,
1219
+ "eval_loss": 0.1128077358007431,
1220
+ "eval_precision_macro": 0.9500253922065015,
1221
+ "eval_precision_micro": 0.9728413004763068,
1222
+ "eval_recall_macro": 0.9094651623357265,
1223
+ "eval_recall_micro": 0.962365252860739,
1224
+ "eval_runtime": 13980.4006,
1225
+ "eval_samples_per_second": 2.833,
1226
+ "eval_steps_per_second": 0.011,
1227
+ "step": 79208
1228
+ },
1229
+ {
1230
+ "epoch": 0.6423775240588563,
1231
+ "grad_norm": 0.1044822484254837,
1232
+ "learning_rate": 1.839405618985286e-05,
1233
+ "loss": 0.0142,
1234
+ "step": 79500
1235
+ },
1236
+ {
1237
+ "epoch": 0.646417634273063,
1238
+ "grad_norm": 0.061170101165771484,
1239
+ "learning_rate": 1.8383955914317344e-05,
1240
+ "loss": 0.0142,
1241
+ "step": 80000
1242
+ },
1243
+ {
1244
+ "epoch": 0.6504577444872696,
1245
+ "grad_norm": 0.09872958064079285,
1246
+ "learning_rate": 1.8373855638781827e-05,
1247
+ "loss": 0.0174,
1248
+ "step": 80500
1249
+ },
1250
+ {
1251
+ "epoch": 0.6544978547014763,
1252
+ "grad_norm": 0.08190814405679703,
1253
+ "learning_rate": 1.836375536324631e-05,
1254
+ "loss": 0.016,
1255
+ "step": 81000
1256
+ },
1257
+ {
1258
+ "epoch": 0.658537964915683,
1259
+ "grad_norm": 0.07712013274431229,
1260
+ "learning_rate": 1.8353655087710794e-05,
1261
+ "loss": 0.0172,
1262
+ "step": 81500
1263
+ },
1264
+ {
1265
+ "epoch": 0.6625780751298895,
1266
+ "grad_norm": 0.04823287948966026,
1267
+ "learning_rate": 1.8343554812175277e-05,
1268
+ "loss": 0.0168,
1269
+ "step": 82000
1270
+ },
1271
+ {
1272
+ "epoch": 0.6666181853440962,
1273
+ "grad_norm": 0.11726228892803192,
1274
+ "learning_rate": 1.833345453663976e-05,
1275
+ "loss": 0.017,
1276
+ "step": 82500
1277
+ },
1278
+ {
1279
+ "epoch": 0.6706582955583028,
1280
+ "grad_norm": 0.06535898894071579,
1281
+ "learning_rate": 1.8323354261104244e-05,
1282
+ "loss": 0.016,
1283
+ "step": 83000
1284
+ },
1285
+ {
1286
+ "epoch": 0.6746984057725095,
1287
+ "grad_norm": 0.05892045795917511,
1288
+ "learning_rate": 1.8313253985568727e-05,
1289
+ "loss": 0.0159,
1290
+ "step": 83500
1291
+ },
1292
+ {
1293
+ "epoch": 0.6787385159867161,
1294
+ "grad_norm": 0.04444234445691109,
1295
+ "learning_rate": 1.830315371003321e-05,
1296
+ "loss": 0.0153,
1297
+ "step": 84000
1298
+ },
1299
+ {
1300
+ "epoch": 0.6827786262009228,
1301
+ "grad_norm": 0.1465209275484085,
1302
+ "learning_rate": 1.8293053434497694e-05,
1303
+ "loss": 0.0156,
1304
+ "step": 84500
1305
+ },
1306
+ {
1307
+ "epoch": 0.6868187364151294,
1308
+ "grad_norm": 0.11835352331399918,
1309
+ "learning_rate": 1.8282953158962178e-05,
1310
+ "loss": 0.0154,
1311
+ "step": 85000
1312
+ },
1313
+ {
1314
+ "epoch": 0.690858846629336,
1315
+ "grad_norm": 0.05793392285704613,
1316
+ "learning_rate": 1.827285288342666e-05,
1317
+ "loss": 0.0138,
1318
+ "step": 85500
1319
+ },
1320
+ {
1321
+ "epoch": 0.6948989568435426,
1322
+ "grad_norm": 0.045407455414533615,
1323
+ "learning_rate": 1.8262752607891144e-05,
1324
+ "loss": 0.0133,
1325
+ "step": 86000
1326
+ },
1327
+ {
1328
+ "epoch": 0.6989390670577493,
1329
+ "grad_norm": 0.12997862696647644,
1330
+ "learning_rate": 1.8252652332355628e-05,
1331
+ "loss": 0.0141,
1332
+ "step": 86500
1333
+ },
1334
+ {
1335
+ "epoch": 0.702979177271956,
1336
+ "grad_norm": 0.07040946930646896,
1337
+ "learning_rate": 1.824255205682011e-05,
1338
+ "loss": 0.0138,
1339
+ "step": 87000
1340
+ },
1341
+ {
1342
+ "epoch": 0.7070192874861626,
1343
+ "grad_norm": 0.05935658514499664,
1344
+ "learning_rate": 1.8232451781284594e-05,
1345
+ "loss": 0.0144,
1346
+ "step": 87500
1347
+ },
1348
+ {
1349
+ "epoch": 0.7110593977003693,
1350
+ "grad_norm": 0.0425080843269825,
1351
+ "learning_rate": 1.8222351505749078e-05,
1352
+ "loss": 0.018,
1353
+ "step": 88000
1354
+ },
1355
+ {
1356
+ "epoch": 0.7150995079145759,
1357
+ "grad_norm": 0.1149262934923172,
1358
+ "learning_rate": 1.821225123021356e-05,
1359
+ "loss": 0.0177,
1360
+ "step": 88500
1361
+ },
1362
+ {
1363
+ "epoch": 0.7191396181287826,
1364
+ "grad_norm": 0.08022065460681915,
1365
+ "learning_rate": 1.8202150954678045e-05,
1366
+ "loss": 0.0172,
1367
+ "step": 89000
1368
+ },
1369
+ {
1370
+ "epoch": 0.7200203621554796,
1371
+ "eval_f1_macro": 0.9342654788805338,
1372
+ "eval_f1_micro": 0.9723695526241013,
1373
+ "eval_loss": 0.09522199630737305,
1374
+ "eval_precision_macro": 0.9684740304119624,
1375
+ "eval_precision_micro": 0.9878875975715066,
1376
+ "eval_recall_macro": 0.9056287281769387,
1377
+ "eval_recall_micro": 0.9573314924580955,
1378
+ "eval_runtime": 13779.2013,
1379
+ "eval_samples_per_second": 2.874,
1380
+ "eval_steps_per_second": 0.011,
1381
+ "step": 89109
1382
+ },
1383
+ {
1384
+ "epoch": 0.7231797283429892,
1385
+ "grad_norm": 0.06360196322202682,
1386
+ "learning_rate": 1.8192050679142528e-05,
1387
+ "loss": 0.0167,
1388
+ "step": 89500
1389
+ },
1390
+ {
1391
+ "epoch": 0.7272198385571959,
1392
+ "grad_norm": 0.0733686089515686,
1393
+ "learning_rate": 1.818195040360701e-05,
1394
+ "loss": 0.0179,
1395
+ "step": 90000
1396
+ },
1397
+ {
1398
+ "epoch": 0.7312599487714024,
1399
+ "grad_norm": 0.1344570368528366,
1400
+ "learning_rate": 1.8171850128071495e-05,
1401
+ "loss": 0.0256,
1402
+ "step": 90500
1403
+ },
1404
+ {
1405
+ "epoch": 0.7353000589856091,
1406
+ "grad_norm": 0.0946430116891861,
1407
+ "learning_rate": 1.8161749852535978e-05,
1408
+ "loss": 0.026,
1409
+ "step": 91000
1410
+ },
1411
+ {
1412
+ "epoch": 0.7393401691998158,
1413
+ "grad_norm": 0.07514828443527222,
1414
+ "learning_rate": 1.815164957700046e-05,
1415
+ "loss": 0.0251,
1416
+ "step": 91500
1417
+ },
1418
+ {
1419
+ "epoch": 0.7433802794140224,
1420
+ "grad_norm": 0.06544400006532669,
1421
+ "learning_rate": 1.8141549301464945e-05,
1422
+ "loss": 0.0247,
1423
+ "step": 92000
1424
+ },
1425
+ {
1426
+ "epoch": 0.7474203896282291,
1427
+ "grad_norm": 0.11973392963409424,
1428
+ "learning_rate": 1.8131449025929428e-05,
1429
+ "loss": 0.0242,
1430
+ "step": 92500
1431
+ },
1432
+ {
1433
+ "epoch": 0.7514604998424357,
1434
+ "grad_norm": 0.07870098203420639,
1435
+ "learning_rate": 1.812134875039391e-05,
1436
+ "loss": 0.0197,
1437
+ "step": 93000
1438
+ },
1439
+ {
1440
+ "epoch": 0.7555006100566424,
1441
+ "grad_norm": 0.06315948814153671,
1442
+ "learning_rate": 1.8111248474858395e-05,
1443
+ "loss": 0.0189,
1444
+ "step": 93500
1445
+ },
1446
+ {
1447
+ "epoch": 0.759540720270849,
1448
+ "grad_norm": 0.05281440541148186,
1449
+ "learning_rate": 1.810114819932288e-05,
1450
+ "loss": 0.0183,
1451
+ "step": 94000
1452
+ },
1453
+ {
1454
+ "epoch": 0.7635808304850557,
1455
+ "grad_norm": 0.11212711036205292,
1456
+ "learning_rate": 1.809104792378736e-05,
1457
+ "loss": 0.0189,
1458
+ "step": 94500
1459
+ },
1460
+ {
1461
+ "epoch": 0.7676209406992622,
1462
+ "grad_norm": 0.13350194692611694,
1463
+ "learning_rate": 1.8080947648251845e-05,
1464
+ "loss": 0.019,
1465
+ "step": 95000
1466
+ },
1467
+ {
1468
+ "epoch": 0.7716610509134689,
1469
+ "grad_norm": 0.06391710788011551,
1470
+ "learning_rate": 1.8070847372716328e-05,
1471
+ "loss": 0.0176,
1472
+ "step": 95500
1473
+ },
1474
+ {
1475
+ "epoch": 0.7757011611276755,
1476
+ "grad_norm": 0.06272578239440918,
1477
+ "learning_rate": 1.806074709718081e-05,
1478
+ "loss": 0.0175,
1479
+ "step": 96000
1480
+ },
1481
+ {
1482
+ "epoch": 0.7797412713418822,
1483
+ "grad_norm": 0.10559968650341034,
1484
+ "learning_rate": 1.80506468216453e-05,
1485
+ "loss": 0.0166,
1486
+ "step": 96500
1487
+ },
1488
+ {
1489
+ "epoch": 0.7837813815560889,
1490
+ "grad_norm": 0.10264132171869278,
1491
+ "learning_rate": 1.804054654610978e-05,
1492
+ "loss": 0.0167,
1493
+ "step": 97000
1494
+ },
1495
+ {
1496
+ "epoch": 0.7878214917702955,
1497
+ "grad_norm": 0.06299474835395813,
1498
+ "learning_rate": 1.8030446270574262e-05,
1499
+ "loss": 0.0171,
1500
+ "step": 97500
1501
+ },
1502
+ {
1503
+ "epoch": 0.7918616019845022,
1504
+ "grad_norm": 0.052749671041965485,
1505
+ "learning_rate": 1.8020345995038745e-05,
1506
+ "loss": 0.0182,
1507
+ "step": 98000
1508
+ },
1509
+ {
1510
+ "epoch": 0.7959017121987088,
1511
+ "grad_norm": 0.14064335823059082,
1512
+ "learning_rate": 1.801024571950323e-05,
1513
+ "loss": 0.0186,
1514
+ "step": 98500
1515
+ },
1516
+ {
1517
+ "epoch": 0.7999418224129154,
1518
+ "grad_norm": 0.1007775291800499,
1519
+ "learning_rate": 1.8000145443967715e-05,
1520
+ "loss": 0.0183,
1521
+ "step": 99000
1522
+ },
1523
+ {
1524
+ "epoch": 0.8000226246171995,
1525
+ "eval_f1_macro": 0.9428325512898811,
1526
+ "eval_f1_micro": 0.9730330895777283,
1527
+ "eval_loss": 0.08679112046957016,
1528
+ "eval_precision_macro": 0.9804326754765803,
1529
+ "eval_precision_micro": 0.9888949040358451,
1530
+ "eval_recall_macro": 0.9118424578887557,
1531
+ "eval_recall_micro": 0.957672087213271,
1532
+ "eval_runtime": 13633.7127,
1533
+ "eval_samples_per_second": 2.905,
1534
+ "eval_steps_per_second": 0.011,
1535
+ "step": 99010
1536
+ },
1537
+ {
1538
+ "epoch": 0.803981932627122,
1539
+ "grad_norm": 0.06079207360744476,
1540
+ "learning_rate": 1.7990045168432195e-05,
1541
+ "loss": 0.0179,
1542
+ "step": 99500
1543
+ },
1544
+ {
1545
+ "epoch": 0.8080220428413287,
1546
+ "grad_norm": 0.08171634376049042,
1547
+ "learning_rate": 1.797994489289668e-05,
1548
+ "loss": 0.0174,
1549
+ "step": 100000
1550
+ },
1551
+ {
1552
+ "epoch": 0.8120621530555353,
1553
+ "grad_norm": 0.11801985651254654,
1554
+ "learning_rate": 1.7969844617361162e-05,
1555
+ "loss": 0.018,
1556
+ "step": 100500
1557
+ },
1558
+ {
1559
+ "epoch": 0.816102263269742,
1560
+ "grad_norm": 0.07442731410264969,
1561
+ "learning_rate": 1.7959744341825645e-05,
1562
+ "loss": 0.0193,
1563
+ "step": 101000
1564
+ },
1565
+ {
1566
+ "epoch": 0.8201423734839487,
1567
+ "grad_norm": 0.07479513436555862,
1568
+ "learning_rate": 1.7949644066290132e-05,
1569
+ "loss": 0.0182,
1570
+ "step": 101500
1571
+ },
1572
+ {
1573
+ "epoch": 0.8241824836981553,
1574
+ "grad_norm": 0.07510875165462494,
1575
+ "learning_rate": 1.7939543790754612e-05,
1576
+ "loss": 0.0187,
1577
+ "step": 102000
1578
+ },
1579
+ {
1580
+ "epoch": 0.828222593912362,
1581
+ "grad_norm": 0.12816323339939117,
1582
+ "learning_rate": 1.7929443515219095e-05,
1583
+ "loss": 0.018,
1584
+ "step": 102500
1585
+ },
1586
+ {
1587
+ "epoch": 0.8322627041265686,
1588
+ "grad_norm": 0.1283213347196579,
1589
+ "learning_rate": 1.791934323968358e-05,
1590
+ "loss": 0.017,
1591
+ "step": 103000
1592
+ },
1593
+ {
1594
+ "epoch": 0.8363028143407752,
1595
+ "grad_norm": 0.06121571362018585,
1596
+ "learning_rate": 1.7909242964148062e-05,
1597
+ "loss": 0.0169,
1598
+ "step": 103500
1599
+ },
1600
+ {
1601
+ "epoch": 0.8403429245549818,
1602
+ "grad_norm": 0.05697647109627724,
1603
+ "learning_rate": 1.789914268861255e-05,
1604
+ "loss": 0.0165,
1605
+ "step": 104000
1606
+ },
1607
+ {
1608
+ "epoch": 0.8443830347691885,
1609
+ "grad_norm": 0.12682537734508514,
1610
+ "learning_rate": 1.7889042413077032e-05,
1611
+ "loss": 0.0166,
1612
+ "step": 104500
1613
+ },
1614
+ {
1615
+ "epoch": 0.8484231449833951,
1616
+ "grad_norm": 0.0857871025800705,
1617
+ "learning_rate": 1.7878942137541512e-05,
1618
+ "loss": 0.0173,
1619
+ "step": 105000
1620
+ },
1621
+ {
1622
+ "epoch": 0.8524632551976018,
1623
+ "grad_norm": 0.06892874091863632,
1624
+ "learning_rate": 1.7868841862005996e-05,
1625
+ "loss": 0.0171,
1626
+ "step": 105500
1627
+ },
1628
+ {
1629
+ "epoch": 0.8565033654118084,
1630
+ "grad_norm": 0.04709647595882416,
1631
+ "learning_rate": 1.785874158647048e-05,
1632
+ "loss": 0.0159,
1633
+ "step": 106000
1634
+ },
1635
+ {
1636
+ "epoch": 0.8605434756260151,
1637
+ "grad_norm": 0.10291819274425507,
1638
+ "learning_rate": 1.7848641310934966e-05,
1639
+ "loss": 0.0165,
1640
+ "step": 106500
1641
+ },
1642
+ {
1643
+ "epoch": 0.8645835858402218,
1644
+ "grad_norm": 0.0879896804690361,
1645
+ "learning_rate": 1.783854103539945e-05,
1646
+ "loss": 0.0162,
1647
+ "step": 107000
1648
+ },
1649
+ {
1650
+ "epoch": 0.8686236960544284,
1651
+ "grad_norm": 0.06169717013835907,
1652
+ "learning_rate": 1.782844075986393e-05,
1653
+ "loss": 0.0158,
1654
+ "step": 107500
1655
+ },
1656
+ {
1657
+ "epoch": 0.872663806268635,
1658
+ "grad_norm": 0.05489352345466614,
1659
+ "learning_rate": 1.7818340484328413e-05,
1660
+ "loss": 0.0165,
1661
+ "step": 108000
1662
+ },
1663
+ {
1664
+ "epoch": 0.8767039164828416,
1665
+ "grad_norm": 0.14040745794773102,
1666
+ "learning_rate": 1.7808240208792896e-05,
1667
+ "loss": 0.0172,
1668
+ "step": 108500
1669
+ },
1670
+ {
1671
+ "epoch": 0.8800248870789195,
1672
+ "eval_f1_macro": 0.9452519445016722,
1673
+ "eval_f1_micro": 0.9735904566562644,
1674
+ "eval_loss": 0.09658095985651016,
1675
+ "eval_precision_macro": 0.9848024256701463,
1676
+ "eval_precision_micro": 0.9916469998618024,
1677
+ "eval_recall_macro": 0.9121362312839789,
1678
+ "eval_recall_micro": 0.956179724302381,
1679
+ "eval_runtime": 13866.6963,
1680
+ "eval_samples_per_second": 2.856,
1681
+ "eval_steps_per_second": 0.011,
1682
+ "step": 108911
1683
+ },
1684
+ {
1685
+ "epoch": 0.8807440266970483,
1686
+ "grad_norm": 0.10508357733488083,
1687
+ "learning_rate": 1.7798139933257383e-05,
1688
+ "loss": 0.0162,
1689
+ "step": 109000
1690
+ },
1691
+ {
1692
+ "epoch": 0.8847841369112549,
1693
+ "grad_norm": 0.06252790987491608,
1694
+ "learning_rate": 1.7788039657721866e-05,
1695
+ "loss": 0.0164,
1696
+ "step": 109500
1697
+ },
1698
+ {
1699
+ "epoch": 0.8888242471254616,
1700
+ "grad_norm": 0.04974674433469772,
1701
+ "learning_rate": 1.7777939382186346e-05,
1702
+ "loss": 0.0165,
1703
+ "step": 110000
1704
+ },
1705
+ {
1706
+ "epoch": 0.8928643573396682,
1707
+ "grad_norm": 0.11918849498033524,
1708
+ "learning_rate": 1.776783910665083e-05,
1709
+ "loss": 0.0174,
1710
+ "step": 110500
1711
+ },
1712
+ {
1713
+ "epoch": 0.8969044675538749,
1714
+ "grad_norm": 0.12928660213947296,
1715
+ "learning_rate": 1.7757738831115313e-05,
1716
+ "loss": 0.017,
1717
+ "step": 111000
1718
+ },
1719
+ {
1720
+ "epoch": 0.9009445777680816,
1721
+ "grad_norm": 0.06852889806032181,
1722
+ "learning_rate": 1.77476385555798e-05,
1723
+ "loss": 0.0163,
1724
+ "step": 111500
1725
+ },
1726
+ {
1727
+ "epoch": 0.9049846879822881,
1728
+ "grad_norm": 0.0549907386302948,
1729
+ "learning_rate": 1.7737538280044283e-05,
1730
+ "loss": 0.0173,
1731
+ "step": 112000
1732
+ },
1733
+ {
1734
+ "epoch": 0.9090247981964948,
1735
+ "grad_norm": 0.12298522889614105,
1736
+ "learning_rate": 1.7727438004508763e-05,
1737
+ "loss": 0.0169,
1738
+ "step": 112500
1739
+ },
1740
+ {
1741
+ "epoch": 0.9130649084107014,
1742
+ "grad_norm": 0.09733408689498901,
1743
+ "learning_rate": 1.7717337728973246e-05,
1744
+ "loss": 0.0177,
1745
+ "step": 113000
1746
+ },
1747
+ {
1748
+ "epoch": 0.9171050186249081,
1749
+ "grad_norm": 0.07251332700252533,
1750
+ "learning_rate": 1.770723745343773e-05,
1751
+ "loss": 0.0176,
1752
+ "step": 113500
1753
+ },
1754
+ {
1755
+ "epoch": 0.9211451288391147,
1756
+ "grad_norm": 0.07106909155845642,
1757
+ "learning_rate": 1.7697137177902216e-05,
1758
+ "loss": 0.0174,
1759
+ "step": 114000
1760
+ },
1761
+ {
1762
+ "epoch": 0.9251852390533214,
1763
+ "grad_norm": 0.1281566470861435,
1764
+ "learning_rate": 1.76870369023667e-05,
1765
+ "loss": 0.0176,
1766
+ "step": 114500
1767
+ },
1768
+ {
1769
+ "epoch": 0.929225349267528,
1770
+ "grad_norm": 0.09204866737127304,
1771
+ "learning_rate": 1.767693662683118e-05,
1772
+ "loss": 0.0171,
1773
+ "step": 115000
1774
+ },
1775
+ {
1776
+ "epoch": 0.9332654594817347,
1777
+ "grad_norm": 0.05850633978843689,
1778
+ "learning_rate": 1.7666836351295663e-05,
1779
+ "loss": 0.0151,
1780
+ "step": 115500
1781
+ },
1782
+ {
1783
+ "epoch": 0.9373055696959413,
1784
+ "grad_norm": 0.044992174953222275,
1785
+ "learning_rate": 1.7656736075760146e-05,
1786
+ "loss": 0.0147,
1787
+ "step": 116000
1788
+ },
1789
+ {
1790
+ "epoch": 0.941345679910148,
1791
+ "grad_norm": 0.10752815753221512,
1792
+ "learning_rate": 1.7646635800224633e-05,
1793
+ "loss": 0.0155,
1794
+ "step": 116500
1795
+ },
1796
+ {
1797
+ "epoch": 0.9453857901243546,
1798
+ "grad_norm": 0.09021549671888351,
1799
+ "learning_rate": 1.7636535524689117e-05,
1800
+ "loss": 0.0154,
1801
+ "step": 117000
1802
+ },
1803
+ {
1804
+ "epoch": 0.9494259003385612,
1805
+ "grad_norm": 0.0689893364906311,
1806
+ "learning_rate": 1.7626435249153596e-05,
1807
+ "loss": 0.0158,
1808
+ "step": 117500
1809
+ },
1810
+ {
1811
+ "epoch": 0.9534660105527679,
1812
+ "grad_norm": 0.06845594197511673,
1813
+ "learning_rate": 1.761633497361808e-05,
1814
+ "loss": 0.021,
1815
+ "step": 118000
1816
+ },
1817
+ {
1818
+ "epoch": 0.9575061207669745,
1819
+ "grad_norm": 0.11164365708827972,
1820
+ "learning_rate": 1.7606234698082563e-05,
1821
+ "loss": 0.0209,
1822
+ "step": 118500
1823
+ },
1824
+ {
1825
+ "epoch": 0.9600271495406395,
1826
+ "eval_f1_macro": 0.9436502450372893,
1827
+ "eval_f1_micro": 0.9732083745347168,
1828
+ "eval_loss": 0.0801812931895256,
1829
+ "eval_precision_macro": 0.9811968490819049,
1830
+ "eval_precision_micro": 0.9878954264431127,
1831
+ "eval_recall_macro": 0.9131012924150136,
1832
+ "eval_recall_micro": 0.9589516303534508,
1833
+ "eval_runtime": 13850.1368,
1834
+ "eval_samples_per_second": 2.859,
1835
+ "eval_steps_per_second": 0.011,
1836
+ "step": 118812
1837
+ },
1838
+ {
1839
+ "epoch": 0.9615462309811812,
1840
+ "grad_norm": 0.0935693234205246,
1841
+ "learning_rate": 1.759613442254705e-05,
1842
+ "loss": 0.0202,
1843
+ "step": 119000
1844
+ },
1845
+ {
1846
+ "epoch": 0.9655863411953878,
1847
+ "grad_norm": 0.06598909944295883,
1848
+ "learning_rate": 1.7586034147011533e-05,
1849
+ "loss": 0.02,
1850
+ "step": 119500
1851
+ },
1852
+ {
1853
+ "epoch": 0.9696264514095945,
1854
+ "grad_norm": 0.052590906620025635,
1855
+ "learning_rate": 1.7575933871476017e-05,
1856
+ "loss": 0.0203,
1857
+ "step": 120000
1858
+ },
1859
+ {
1860
+ "epoch": 0.973666561623801,
1861
+ "grad_norm": 0.1306983232498169,
1862
+ "learning_rate": 1.7565833595940497e-05,
1863
+ "loss": 0.0152,
1864
+ "step": 120500
1865
+ },
1866
+ {
1867
+ "epoch": 0.9777066718380077,
1868
+ "grad_norm": 0.07255972176790237,
1869
+ "learning_rate": 1.7555733320404983e-05,
1870
+ "loss": 0.0156,
1871
+ "step": 121000
1872
+ },
1873
+ {
1874
+ "epoch": 0.9817467820522144,
1875
+ "grad_norm": 0.0558183416724205,
1876
+ "learning_rate": 1.7545633044869467e-05,
1877
+ "loss": 0.0149,
1878
+ "step": 121500
1879
+ },
1880
+ {
1881
+ "epoch": 0.985786892266421,
1882
+ "grad_norm": 0.04536261036992073,
1883
+ "learning_rate": 1.753553276933395e-05,
1884
+ "loss": 0.0145,
1885
+ "step": 122000
1886
+ },
1887
+ {
1888
+ "epoch": 0.9898270024806277,
1889
+ "grad_norm": 0.12242696434259415,
1890
+ "learning_rate": 1.7525432493798434e-05,
1891
+ "loss": 0.0148,
1892
+ "step": 122500
1893
+ },
1894
+ {
1895
+ "epoch": 0.9938671126948343,
1896
+ "grad_norm": 0.09054296463727951,
1897
+ "learning_rate": 1.7515332218262914e-05,
1898
+ "loss": 0.0184,
1899
+ "step": 123000
1900
+ },
1901
+ {
1902
+ "epoch": 0.997907222909041,
1903
+ "grad_norm": 0.0703011155128479,
1904
+ "learning_rate": 1.75052319427274e-05,
1905
+ "loss": 0.0187,
1906
+ "step": 123500
1907
+ },
1908
+ {
1909
+ "epoch": 1.0019473331232476,
1910
+ "grad_norm": 0.06889301538467407,
1911
+ "learning_rate": 1.7495131667191884e-05,
1912
+ "loss": 0.0186,
1913
+ "step": 124000
1914
+ },
1915
+ {
1916
+ "epoch": 1.0059874433374543,
1917
+ "grad_norm": 0.1129370704293251,
1918
+ "learning_rate": 1.7485031391656367e-05,
1919
+ "loss": 0.0185,
1920
+ "step": 124500
1921
+ },
1922
+ {
1923
+ "epoch": 1.010027553551661,
1924
+ "grad_norm": 0.0729982927441597,
1925
+ "learning_rate": 1.747493111612085e-05,
1926
+ "loss": 0.0181,
1927
+ "step": 125000
1928
+ },
1929
+ {
1930
+ "epoch": 1.0140676637658674,
1931
+ "grad_norm": 0.19092483818531036,
1932
+ "learning_rate": 1.746483084058533e-05,
1933
+ "loss": 0.0167,
1934
+ "step": 125500
1935
+ },
1936
+ {
1937
+ "epoch": 1.0181077739800741,
1938
+ "grad_norm": 0.04695465415716171,
1939
+ "learning_rate": 1.7454730565049817e-05,
1940
+ "loss": 0.0168,
1941
+ "step": 126000
1942
+ },
1943
+ {
1944
+ "epoch": 1.0221478841942808,
1945
+ "grad_norm": 0.1297185719013214,
1946
+ "learning_rate": 1.74446302895143e-05,
1947
+ "loss": 0.0168,
1948
+ "step": 126500
1949
+ },
1950
+ {
1951
+ "epoch": 1.0261879944084875,
1952
+ "grad_norm": 0.07326006889343262,
1953
+ "learning_rate": 1.7434530013978784e-05,
1954
+ "loss": 0.0174,
1955
+ "step": 127000
1956
+ },
1957
+ {
1958
+ "epoch": 1.0302281046226942,
1959
+ "grad_norm": 0.0644180178642273,
1960
+ "learning_rate": 1.7424429738443267e-05,
1961
+ "loss": 0.0169,
1962
+ "step": 127500
1963
+ },
1964
+ {
1965
+ "epoch": 1.0342682148369007,
1966
+ "grad_norm": 0.04816208407282829,
1967
+ "learning_rate": 1.7414329462907747e-05,
1968
+ "loss": 0.0146,
1969
+ "step": 128000
1970
+ },
1971
+ {
1972
+ "epoch": 1.0383083250511074,
1973
+ "grad_norm": 0.09492602199316025,
1974
+ "learning_rate": 1.7404229187372234e-05,
1975
+ "loss": 0.0146,
1976
+ "step": 128500
1977
+ },
1978
+ {
1979
+ "epoch": 1.0400294120023594,
1980
+ "eval_f1_macro": 0.9447347774982361,
1981
+ "eval_f1_micro": 0.9738792089577321,
1982
+ "eval_loss": 0.09039987623691559,
1983
+ "eval_precision_macro": 0.9845539397176223,
1984
+ "eval_precision_micro": 0.9913320028997217,
1985
+ "eval_recall_macro": 0.9116629602052654,
1986
+ "eval_recall_micro": 0.9570303099541577,
1987
+ "eval_runtime": 13620.1841,
1988
+ "eval_samples_per_second": 2.908,
1989
+ "eval_steps_per_second": 0.011,
1990
+ "step": 128713
1991
+ },
1992
+ {
1993
+ "epoch": 1.042348435265314,
1994
+ "grad_norm": 0.08992265909910202,
1995
+ "learning_rate": 1.7394128911836717e-05,
1996
+ "loss": 0.014,
1997
+ "step": 129000
1998
+ },
1999
+ {
2000
+ "epoch": 1.0463885454795208,
2001
+ "grad_norm": 0.05333436280488968,
2002
+ "learning_rate": 1.73840286363012e-05,
2003
+ "loss": 0.015,
2004
+ "step": 129500
2005
+ },
2006
+ {
2007
+ "epoch": 1.0504286556937272,
2008
+ "grad_norm": 0.057117633521556854,
2009
+ "learning_rate": 1.7373928360765684e-05,
2010
+ "loss": 0.0144,
2011
+ "step": 130000
2012
+ },
2013
+ {
2014
+ "epoch": 1.054468765907934,
2015
+ "grad_norm": 0.12276995927095413,
2016
+ "learning_rate": 1.7363828085230164e-05,
2017
+ "loss": 0.0228,
2018
+ "step": 130500
2019
+ },
2020
+ {
2021
+ "epoch": 1.0585088761221406,
2022
+ "grad_norm": 0.08618568629026413,
2023
+ "learning_rate": 1.735372780969465e-05,
2024
+ "loss": 0.0229,
2025
+ "step": 131000
2026
+ },
2027
+ {
2028
+ "epoch": 1.0625489863363473,
2029
+ "grad_norm": 0.08783124387264252,
2030
+ "learning_rate": 1.7343627534159134e-05,
2031
+ "loss": 0.0222,
2032
+ "step": 131500
2033
+ },
2034
+ {
2035
+ "epoch": 1.066589096550554,
2036
+ "grad_norm": 0.06352981925010681,
2037
+ "learning_rate": 1.7333527258623618e-05,
2038
+ "loss": 0.0221,
2039
+ "step": 132000
2040
+ },
2041
+ {
2042
+ "epoch": 1.0706292067647605,
2043
+ "grad_norm": 0.10115523636341095,
2044
+ "learning_rate": 1.73234269830881e-05,
2045
+ "loss": 0.0226,
2046
+ "step": 132500
2047
+ },
2048
+ {
2049
+ "epoch": 1.0746693169789672,
2050
+ "grad_norm": 0.11306885629892349,
2051
+ "learning_rate": 1.731332670755258e-05,
2052
+ "loss": 0.0162,
2053
+ "step": 133000
2054
+ },
2055
+ {
2056
+ "epoch": 1.0787094271931739,
2057
+ "grad_norm": 0.05852317065000534,
2058
+ "learning_rate": 1.7303226432017068e-05,
2059
+ "loss": 0.0155,
2060
+ "step": 133500
2061
+ },
2062
+ {
2063
+ "epoch": 1.0827495374073806,
2064
+ "grad_norm": 0.046473681926727295,
2065
+ "learning_rate": 1.729312615648155e-05,
2066
+ "loss": 0.0149,
2067
+ "step": 134000
2068
+ },
2069
+ {
2070
+ "epoch": 1.086789647621587,
2071
+ "grad_norm": 0.11023978888988495,
2072
+ "learning_rate": 1.7283025880946034e-05,
2073
+ "loss": 0.0152,
2074
+ "step": 134500
2075
+ },
2076
+ {
2077
+ "epoch": 1.0908297578357937,
2078
+ "grad_norm": 0.07801781594753265,
2079
+ "learning_rate": 1.7272925605410518e-05,
2080
+ "loss": 0.0145,
2081
+ "step": 135000
2082
+ },
2083
+ {
2084
+ "epoch": 1.0948698680500004,
2085
+ "grad_norm": 0.057179663330316544,
2086
+ "learning_rate": 1.7262825329875e-05,
2087
+ "loss": 0.0167,
2088
+ "step": 135500
2089
+ },
2090
+ {
2091
+ "epoch": 1.098909978264207,
2092
+ "grad_norm": 0.0559101440012455,
2093
+ "learning_rate": 1.7252725054339484e-05,
2094
+ "loss": 0.0174,
2095
+ "step": 136000
2096
+ },
2097
+ {
2098
+ "epoch": 1.1029500884784136,
2099
+ "grad_norm": 0.08359610289335251,
2100
+ "learning_rate": 1.7242624778803968e-05,
2101
+ "loss": 0.0171,
2102
+ "step": 136500
2103
+ },
2104
+ {
2105
+ "epoch": 1.1069901986926203,
2106
+ "grad_norm": 0.11296004056930542,
2107
+ "learning_rate": 1.723252450326845e-05,
2108
+ "loss": 0.0172,
2109
+ "step": 137000
2110
+ },
2111
+ {
2112
+ "epoch": 1.111030308906827,
2113
+ "grad_norm": 0.061936188489198685,
2114
+ "learning_rate": 1.7222424227732935e-05,
2115
+ "loss": 0.0173,
2116
+ "step": 137500
2117
+ },
2118
+ {
2119
+ "epoch": 1.1150704191210337,
2120
+ "grad_norm": 0.07334394752979279,
2121
+ "learning_rate": 1.7212323952197418e-05,
2122
+ "loss": 0.0171,
2123
+ "step": 138000
2124
+ },
2125
+ {
2126
+ "epoch": 1.1191105293352404,
2127
+ "grad_norm": 0.10479886829853058,
2128
+ "learning_rate": 1.72022236766619e-05,
2129
+ "loss": 0.0167,
2130
+ "step": 138500
2131
+ },
2132
+ {
2133
+ "epoch": 1.1200316744640795,
2134
+ "eval_f1_macro": 0.9448181357644012,
2135
+ "eval_f1_micro": 0.9737656699358889,
2136
+ "eval_loss": 0.07901577651500702,
2137
+ "eval_precision_macro": 0.9836772488732869,
2138
+ "eval_precision_micro": 0.9897679811194957,
2139
+ "eval_recall_macro": 0.9127768967177174,
2140
+ "eval_recall_micro": 0.9582725683902614,
2141
+ "eval_runtime": 13749.2598,
2142
+ "eval_samples_per_second": 2.88,
2143
+ "eval_steps_per_second": 0.011,
2144
+ "step": 138614
2145
+ },
2146
+ {
2147
+ "epoch": 1.1231506395494468,
2148
+ "grad_norm": 0.07909916341304779,
2149
+ "learning_rate": 1.7192123401126385e-05,
2150
+ "loss": 0.0161,
2151
+ "step": 139000
2152
+ },
2153
+ {
2154
+ "epoch": 1.1271907497636535,
2155
+ "grad_norm": 0.08238150179386139,
2156
+ "learning_rate": 1.7182023125590868e-05,
2157
+ "loss": 0.017,
2158
+ "step": 139500
2159
+ },
2160
+ {
2161
+ "epoch": 1.1312308599778602,
2162
+ "grad_norm": 0.06267368793487549,
2163
+ "learning_rate": 1.717192285005535e-05,
2164
+ "loss": 0.0164,
2165
+ "step": 140000
2166
+ },
2167
+ {
2168
+ "epoch": 1.135270970192067,
2169
+ "grad_norm": 0.11608216911554337,
2170
+ "learning_rate": 1.7161822574519835e-05,
2171
+ "loss": 0.0165,
2172
+ "step": 140500
2173
+ },
2174
+ {
2175
+ "epoch": 1.1393110804062734,
2176
+ "grad_norm": 0.10431836545467377,
2177
+ "learning_rate": 1.7151722298984318e-05,
2178
+ "loss": 0.0161,
2179
+ "step": 141000
2180
+ },
2181
+ {
2182
+ "epoch": 1.14335119062048,
2183
+ "grad_norm": 0.06495651602745056,
2184
+ "learning_rate": 1.71416220234488e-05,
2185
+ "loss": 0.0165,
2186
+ "step": 141500
2187
+ },
2188
+ {
2189
+ "epoch": 1.1473913008346868,
2190
+ "grad_norm": 0.04861852526664734,
2191
+ "learning_rate": 1.7131521747913285e-05,
2192
+ "loss": 0.0163,
2193
+ "step": 142000
2194
+ },
2195
+ {
2196
+ "epoch": 1.1514314110488935,
2197
+ "grad_norm": 0.17824631929397583,
2198
+ "learning_rate": 1.7121421472377768e-05,
2199
+ "loss": 0.0159,
2200
+ "step": 142500
2201
+ },
2202
+ {
2203
+ "epoch": 1.1554715212631002,
2204
+ "grad_norm": 0.08877791464328766,
2205
+ "learning_rate": 1.711132119684225e-05,
2206
+ "loss": 0.0171,
2207
+ "step": 143000
2208
+ },
2209
+ {
2210
+ "epoch": 1.1595116314773066,
2211
+ "grad_norm": 0.06289026886224747,
2212
+ "learning_rate": 1.7101220921306735e-05,
2213
+ "loss": 0.0166,
2214
+ "step": 143500
2215
+ },
2216
+ {
2217
+ "epoch": 1.1635517416915133,
2218
+ "grad_norm": 0.0498519092798233,
2219
+ "learning_rate": 1.709112064577122e-05,
2220
+ "loss": 0.0169,
2221
+ "step": 144000
2222
+ },
2223
+ {
2224
+ "epoch": 1.16759185190572,
2225
+ "grad_norm": 0.13069184124469757,
2226
+ "learning_rate": 1.7081020370235702e-05,
2227
+ "loss": 0.0168,
2228
+ "step": 144500
2229
+ },
2230
+ {
2231
+ "epoch": 1.1716319621199267,
2232
+ "grad_norm": 0.09042539447546005,
2233
+ "learning_rate": 1.7070920094700185e-05,
2234
+ "loss": 0.0168,
2235
+ "step": 145000
2236
+ },
2237
+ {
2238
+ "epoch": 1.1756720723341332,
2239
+ "grad_norm": 0.05690092593431473,
2240
+ "learning_rate": 1.706081981916467e-05,
2241
+ "loss": 0.0166,
2242
+ "step": 145500
2243
+ },
2244
+ {
2245
+ "epoch": 1.1797121825483399,
2246
+ "grad_norm": 0.0493723563849926,
2247
+ "learning_rate": 1.7050719543629152e-05,
2248
+ "loss": 0.017,
2249
+ "step": 146000
2250
+ },
2251
+ {
2252
+ "epoch": 1.1837522927625466,
2253
+ "grad_norm": 0.10125371068716049,
2254
+ "learning_rate": 1.7040619268093635e-05,
2255
+ "loss": 0.0165,
2256
+ "step": 146500
2257
+ },
2258
+ {
2259
+ "epoch": 1.1877924029767533,
2260
+ "grad_norm": 0.0926498994231224,
2261
+ "learning_rate": 1.703051899255812e-05,
2262
+ "loss": 0.0163,
2263
+ "step": 147000
2264
+ },
2265
+ {
2266
+ "epoch": 1.19183251319096,
2267
+ "grad_norm": 0.06617089360952377,
2268
+ "learning_rate": 1.7020418717022602e-05,
2269
+ "loss": 0.0168,
2270
+ "step": 147500
2271
+ },
2272
+ {
2273
+ "epoch": 1.1958726234051664,
2274
+ "grad_norm": 0.05541488900780678,
2275
+ "learning_rate": 1.7010318441487085e-05,
2276
+ "loss": 0.0192,
2277
+ "step": 148000
2278
+ },
2279
+ {
2280
+ "epoch": 1.1999127336193731,
2281
+ "grad_norm": 0.12656770646572113,
2282
+ "learning_rate": 1.700021816595157e-05,
2283
+ "loss": 0.0193,
2284
+ "step": 148500
2285
+ },
2286
+ {
2287
+ "epoch": 1.2000339369257993,
2288
+ "eval_f1_macro": 0.9460294234862895,
2289
+ "eval_f1_micro": 0.9741125567825796,
2290
+ "eval_loss": 0.08067350834608078,
2291
+ "eval_precision_macro": 0.986194906743568,
2292
+ "eval_precision_micro": 0.9906935766072956,
2293
+ "eval_recall_macro": 0.9129872709919727,
2294
+ "eval_recall_micro": 0.9580774262702744,
2295
+ "eval_runtime": 13826.0506,
2296
+ "eval_samples_per_second": 2.864,
2297
+ "eval_steps_per_second": 0.011,
2298
+ "step": 148515
2299
+ },
2300
+ {
2301
+ "epoch": 1.2039528438335798,
2302
+ "grad_norm": 0.10946424305438995,
2303
+ "learning_rate": 1.6990117890416052e-05,
2304
+ "loss": 0.019,
2305
+ "step": 149000
2306
+ },
2307
+ {
2308
+ "epoch": 1.2079929540477865,
2309
+ "grad_norm": 0.05134887993335724,
2310
+ "learning_rate": 1.6980017614880535e-05,
2311
+ "loss": 0.0177,
2312
+ "step": 149500
2313
+ },
2314
+ {
2315
+ "epoch": 1.212033064261993,
2316
+ "grad_norm": 0.08791927248239517,
2317
+ "learning_rate": 1.696991733934502e-05,
2318
+ "loss": 0.0188,
2319
+ "step": 150000
2320
+ },
2321
+ {
2322
+ "epoch": 1.2160731744761997,
2323
+ "grad_norm": 0.11116321384906769,
2324
+ "learning_rate": 1.6959817063809502e-05,
2325
+ "loss": 0.014,
2326
+ "step": 150500
2327
+ },
2328
+ {
2329
+ "epoch": 1.2201132846904064,
2330
+ "grad_norm": 0.07135743647813797,
2331
+ "learning_rate": 1.6949716788273986e-05,
2332
+ "loss": 0.0133,
2333
+ "step": 151000
2334
+ },
2335
+ {
2336
+ "epoch": 1.224153394904613,
2337
+ "grad_norm": 0.06051028147339821,
2338
+ "learning_rate": 1.693961651273847e-05,
2339
+ "loss": 0.014,
2340
+ "step": 151500
2341
+ },
2342
+ {
2343
+ "epoch": 1.2281935051188198,
2344
+ "grad_norm": 0.05637380853295326,
2345
+ "learning_rate": 1.6929516237202952e-05,
2346
+ "loss": 0.0136,
2347
+ "step": 152000
2348
+ },
2349
+ {
2350
+ "epoch": 1.2322336153330262,
2351
+ "grad_norm": 0.2139320969581604,
2352
+ "learning_rate": 1.6919415961667436e-05,
2353
+ "loss": 0.014,
2354
+ "step": 152500
2355
+ },
2356
+ {
2357
+ "epoch": 1.236273725547233,
2358
+ "grad_norm": 0.10385521501302719,
2359
+ "learning_rate": 1.690931568613192e-05,
2360
+ "loss": 0.0136,
2361
+ "step": 153000
2362
+ },
2363
+ {
2364
+ "epoch": 1.2403138357614396,
2365
+ "grad_norm": 0.052428074181079865,
2366
+ "learning_rate": 1.6899215410596402e-05,
2367
+ "loss": 0.0143,
2368
+ "step": 153500
2369
+ },
2370
+ {
2371
+ "epoch": 1.2443539459756463,
2372
+ "grad_norm": 0.0810508131980896,
2373
+ "learning_rate": 1.6889115135060886e-05,
2374
+ "loss": 0.0136,
2375
+ "step": 154000
2376
+ },
2377
+ {
2378
+ "epoch": 1.2483940561898528,
2379
+ "grad_norm": 0.1127280592918396,
2380
+ "learning_rate": 1.687901485952537e-05,
2381
+ "loss": 0.0131,
2382
+ "step": 154500
2383
+ },
2384
+ {
2385
+ "epoch": 1.2524341664040595,
2386
+ "grad_norm": 0.0869458019733429,
2387
+ "learning_rate": 1.6868914583989852e-05,
2388
+ "loss": 0.0134,
2389
+ "step": 155000
2390
+ },
2391
+ {
2392
+ "epoch": 1.2564742766182662,
2393
+ "grad_norm": 0.055589742958545685,
2394
+ "learning_rate": 1.6858814308454336e-05,
2395
+ "loss": 0.0131,
2396
+ "step": 155500
2397
+ },
2398
+ {
2399
+ "epoch": 1.2605143868324729,
2400
+ "grad_norm": 0.07655055820941925,
2401
+ "learning_rate": 1.684871403291882e-05,
2402
+ "loss": 0.0133,
2403
+ "step": 156000
2404
+ },
2405
+ {
2406
+ "epoch": 1.2645544970466793,
2407
+ "grad_norm": 0.10124019533395767,
2408
+ "learning_rate": 1.6838613757383303e-05,
2409
+ "loss": 0.0134,
2410
+ "step": 156500
2411
+ },
2412
+ {
2413
+ "epoch": 1.268594607260886,
2414
+ "grad_norm": 0.06868778169155121,
2415
+ "learning_rate": 1.6828513481847786e-05,
2416
+ "loss": 0.0131,
2417
+ "step": 157000
2418
+ },
2419
+ {
2420
+ "epoch": 1.2726347174750927,
2421
+ "grad_norm": 0.05508118122816086,
2422
+ "learning_rate": 1.681841320631227e-05,
2423
+ "loss": 0.013,
2424
+ "step": 157500
2425
+ },
2426
+ {
2427
+ "epoch": 1.2766748276892994,
2428
+ "grad_norm": 0.061807744204998016,
2429
+ "learning_rate": 1.6808312930776753e-05,
2430
+ "loss": 0.0165,
2431
+ "step": 158000
2432
+ },
2433
+ {
2434
+ "epoch": 1.2800361993875193,
2435
+ "eval_f1_macro": 0.9468881762987015,
2436
+ "eval_f1_micro": 0.974322016191991,
2437
+ "eval_loss": 0.08317849040031433,
2438
+ "eval_precision_macro": 0.9867617789603411,
2439
+ "eval_precision_micro": 0.9924250039485718,
2440
+ "eval_recall_macro": 0.9135532474499787,
2441
+ "eval_recall_micro": 0.9568676362293653,
2442
+ "eval_runtime": 13523.4341,
2443
+ "eval_samples_per_second": 2.928,
2444
+ "eval_steps_per_second": 0.011,
2445
+ "step": 158416
2446
+ },
2447
+ {
2448
+ "epoch": 1.280714937903506,
2449
+ "grad_norm": 0.14820145070552826,
2450
+ "learning_rate": 1.6798212655241236e-05,
2451
+ "loss": 0.0154,
2452
+ "step": 158500
2453
+ },
2454
+ {
2455
+ "epoch": 1.2847550481177126,
2456
+ "grad_norm": 0.066920705139637,
2457
+ "learning_rate": 1.678811237970572e-05,
2458
+ "loss": 0.0165,
2459
+ "step": 159000
2460
+ },
2461
+ {
2462
+ "epoch": 1.2887951583319193,
2463
+ "grad_norm": 0.05135662853717804,
2464
+ "learning_rate": 1.6778012104170203e-05,
2465
+ "loss": 0.0157,
2466
+ "step": 159500
2467
+ },
2468
+ {
2469
+ "epoch": 1.292835268546126,
2470
+ "grad_norm": 0.0481293685734272,
2471
+ "learning_rate": 1.6767911828634686e-05,
2472
+ "loss": 0.0157,
2473
+ "step": 160000
2474
+ },
2475
+ {
2476
+ "epoch": 1.2968753787603327,
2477
+ "grad_norm": 0.11119942367076874,
2478
+ "learning_rate": 1.675781155309917e-05,
2479
+ "loss": 0.0175,
2480
+ "step": 160500
2481
+ },
2482
+ {
2483
+ "epoch": 1.3009154889745391,
2484
+ "grad_norm": 0.10568433254957199,
2485
+ "learning_rate": 1.6747711277563653e-05,
2486
+ "loss": 0.0195,
2487
+ "step": 161000
2488
+ },
2489
+ {
2490
+ "epoch": 1.3049555991887458,
2491
+ "grad_norm": 0.070424385368824,
2492
+ "learning_rate": 1.6737611002028136e-05,
2493
+ "loss": 0.0187,
2494
+ "step": 161500
2495
+ },
2496
+ {
2497
+ "epoch": 1.3089957094029525,
2498
+ "grad_norm": 0.055738966912031174,
2499
+ "learning_rate": 1.672751072649262e-05,
2500
+ "loss": 0.0178,
2501
+ "step": 162000
2502
+ },
2503
+ {
2504
+ "epoch": 1.3130358196171592,
2505
+ "grad_norm": 0.13051150739192963,
2506
+ "learning_rate": 1.6717410450957103e-05,
2507
+ "loss": 0.0184,
2508
+ "step": 162500
2509
+ },
2510
+ {
2511
+ "epoch": 1.317075929831366,
2512
+ "grad_norm": 0.07910241186618805,
2513
+ "learning_rate": 1.6707310175421586e-05,
2514
+ "loss": 0.0155,
2515
+ "step": 163000
2516
+ },
2517
+ {
2518
+ "epoch": 1.3211160400455724,
2519
+ "grad_norm": 0.15667231380939484,
2520
+ "learning_rate": 1.669720989988607e-05,
2521
+ "loss": 0.0156,
2522
+ "step": 163500
2523
+ },
2524
+ {
2525
+ "epoch": 1.325156150259779,
2526
+ "grad_norm": 0.1987818032503128,
2527
+ "learning_rate": 1.6687109624350553e-05,
2528
+ "loss": 0.0152,
2529
+ "step": 164000
2530
+ },
2531
+ {
2532
+ "epoch": 1.3291962604739858,
2533
+ "grad_norm": 0.13924378156661987,
2534
+ "learning_rate": 1.6677009348815036e-05,
2535
+ "loss": 0.0149,
2536
+ "step": 164500
2537
+ },
2538
+ {
2539
+ "epoch": 1.3332363706881925,
2540
+ "grad_norm": 0.07680565118789673,
2541
+ "learning_rate": 1.666690907327952e-05,
2542
+ "loss": 0.0152,
2543
+ "step": 165000
2544
+ },
2545
+ {
2546
+ "epoch": 1.337276480902399,
2547
+ "grad_norm": 0.10616718977689743,
2548
+ "learning_rate": 1.6656808797744003e-05,
2549
+ "loss": 0.0248,
2550
+ "step": 165500
2551
+ },
2552
+ {
2553
+ "epoch": 1.3413165911166056,
2554
+ "grad_norm": 0.14228446781635284,
2555
+ "learning_rate": 1.6646708522208487e-05,
2556
+ "loss": 0.0255,
2557
+ "step": 166000
2558
+ },
2559
+ {
2560
+ "epoch": 1.3453567013308123,
2561
+ "grad_norm": 0.12593576312065125,
2562
+ "learning_rate": 1.6636608246672973e-05,
2563
+ "loss": 0.0249,
2564
+ "step": 166500
2565
+ },
2566
+ {
2567
+ "epoch": 1.349396811545019,
2568
+ "grad_norm": 0.14932659268379211,
2569
+ "learning_rate": 1.6626507971137453e-05,
2570
+ "loss": 0.0253,
2571
+ "step": 167000
2572
+ },
2573
+ {
2574
+ "epoch": 1.3534369217592257,
2575
+ "grad_norm": 0.09529467672109604,
2576
+ "learning_rate": 1.6616407695601937e-05,
2577
+ "loss": 0.0248,
2578
+ "step": 167500
2579
+ },
2580
+ {
2581
+ "epoch": 1.3574770319734322,
2582
+ "grad_norm": 0.048431217670440674,
2583
+ "learning_rate": 1.660630742006642e-05,
2584
+ "loss": 0.0151,
2585
+ "step": 168000
2586
+ },
2587
+ {
2588
+ "epoch": 1.3600384618492392,
2589
+ "eval_f1_macro": 0.9470943315331984,
2590
+ "eval_f1_micro": 0.9744685617640599,
2591
+ "eval_loss": 0.08101344108581543,
2592
+ "eval_precision_macro": 0.9875637466039148,
2593
+ "eval_precision_micro": 0.9922827909185198,
2594
+ "eval_recall_macro": 0.9134025250498142,
2595
+ "eval_recall_micro": 0.957282681677187,
2596
+ "eval_runtime": 13286.0274,
2597
+ "eval_samples_per_second": 2.981,
2598
+ "eval_steps_per_second": 0.012,
2599
+ "step": 168317
2600
+ },
2601
+ {
2602
+ "epoch": 1.3615171421876389,
2603
+ "grad_norm": 0.10621971637010574,
2604
+ "learning_rate": 1.6596207144530903e-05,
2605
+ "loss": 0.0152,
2606
+ "step": 168500
2607
+ },
2608
+ {
2609
+ "epoch": 1.3655572524018456,
2610
+ "grad_norm": 0.07011255621910095,
2611
+ "learning_rate": 1.658610686899539e-05,
2612
+ "loss": 0.015,
2613
+ "step": 169000
2614
+ },
2615
+ {
2616
+ "epoch": 1.3695973626160522,
2617
+ "grad_norm": 0.05363575369119644,
2618
+ "learning_rate": 1.657600659345987e-05,
2619
+ "loss": 0.0148,
2620
+ "step": 169500
2621
+ },
2622
+ {
2623
+ "epoch": 1.3736374728302587,
2624
+ "grad_norm": 0.14870333671569824,
2625
+ "learning_rate": 1.6565906317924354e-05,
2626
+ "loss": 0.0148,
2627
+ "step": 170000
2628
+ },
2629
+ {
2630
+ "epoch": 1.3776775830444654,
2631
+ "grad_norm": 0.11409811675548553,
2632
+ "learning_rate": 1.6555806042388837e-05,
2633
+ "loss": 0.0158,
2634
+ "step": 170500
2635
+ },
2636
+ {
2637
+ "epoch": 1.381717693258672,
2638
+ "grad_norm": 0.11164900660514832,
2639
+ "learning_rate": 1.654570576685332e-05,
2640
+ "loss": 0.0148,
2641
+ "step": 171000
2642
+ },
2643
+ {
2644
+ "epoch": 1.3857578034728788,
2645
+ "grad_norm": 0.08794820308685303,
2646
+ "learning_rate": 1.6535605491317807e-05,
2647
+ "loss": 0.0158,
2648
+ "step": 171500
2649
+ },
2650
+ {
2651
+ "epoch": 1.3897979136870853,
2652
+ "grad_norm": 0.060815006494522095,
2653
+ "learning_rate": 1.6525505215782287e-05,
2654
+ "loss": 0.0145,
2655
+ "step": 172000
2656
+ },
2657
+ {
2658
+ "epoch": 1.393838023901292,
2659
+ "grad_norm": 0.12906509637832642,
2660
+ "learning_rate": 1.651540494024677e-05,
2661
+ "loss": 0.0153,
2662
+ "step": 172500
2663
+ },
2664
+ {
2665
+ "epoch": 1.3978781341154987,
2666
+ "grad_norm": 0.09560517966747284,
2667
+ "learning_rate": 1.6505304664711254e-05,
2668
+ "loss": 0.026,
2669
+ "step": 173000
2670
+ },
2671
+ {
2672
+ "epoch": 1.4019182443297054,
2673
+ "grad_norm": 0.05908598750829697,
2674
+ "learning_rate": 1.6495204389175737e-05,
2675
+ "loss": 0.0253,
2676
+ "step": 173500
2677
+ },
2678
+ {
2679
+ "epoch": 1.405958354543912,
2680
+ "grad_norm": 0.06017552688717842,
2681
+ "learning_rate": 1.6485104113640224e-05,
2682
+ "loss": 0.0248,
2683
+ "step": 174000
2684
+ },
2685
+ {
2686
+ "epoch": 1.4099984647581185,
2687
+ "grad_norm": 0.10513614118099213,
2688
+ "learning_rate": 1.6475003838104704e-05,
2689
+ "loss": 0.0243,
2690
+ "step": 174500
2691
+ },
2692
+ {
2693
+ "epoch": 1.4140385749723252,
2694
+ "grad_norm": 0.08137038350105286,
2695
+ "learning_rate": 1.6464903562569187e-05,
2696
+ "loss": 0.0243,
2697
+ "step": 175000
2698
+ },
2699
+ {
2700
+ "epoch": 1.418078685186532,
2701
+ "grad_norm": 0.07494989782571793,
2702
+ "learning_rate": 1.645480328703367e-05,
2703
+ "loss": 0.0202,
2704
+ "step": 175500
2705
+ },
2706
+ {
2707
+ "epoch": 1.4221187954007386,
2708
+ "grad_norm": 0.05562291666865349,
2709
+ "learning_rate": 1.6444703011498154e-05,
2710
+ "loss": 0.0204,
2711
+ "step": 176000
2712
+ },
2713
+ {
2714
+ "epoch": 1.426158905614945,
2715
+ "grad_norm": 0.11044422537088394,
2716
+ "learning_rate": 1.643460273596264e-05,
2717
+ "loss": 0.0202,
2718
+ "step": 176500
2719
+ },
2720
+ {
2721
+ "epoch": 1.4301990158291518,
2722
+ "grad_norm": 0.11972752958536148,
2723
+ "learning_rate": 1.642450246042712e-05,
2724
+ "loss": 0.0195,
2725
+ "step": 177000
2726
+ },
2727
+ {
2728
+ "epoch": 1.4342391260433585,
2729
+ "grad_norm": 0.06898529082536697,
2730
+ "learning_rate": 1.6414402184891604e-05,
2731
+ "loss": 0.0203,
2732
+ "step": 177500
2733
+ },
2734
+ {
2735
+ "epoch": 1.4382792362575652,
2736
+ "grad_norm": 0.05580909922719002,
2737
+ "learning_rate": 1.6404301909356087e-05,
2738
+ "loss": 0.0124,
2739
+ "step": 178000
2740
+ },
2741
+ {
2742
+ "epoch": 1.4400407243109592,
2743
+ "eval_f1_macro": 0.9458974211933513,
2744
+ "eval_f1_micro": 0.974213850978252,
2745
+ "eval_loss": 0.09569641947746277,
2746
+ "eval_precision_macro": 0.9869461304954816,
2747
+ "eval_precision_micro": 0.9919091180407337,
2748
+ "eval_recall_macro": 0.9122157060173365,
2749
+ "eval_recall_micro": 0.9571388713888294,
2750
+ "eval_runtime": 13113.6746,
2751
+ "eval_samples_per_second": 3.02,
2752
+ "eval_steps_per_second": 0.012,
2753
+ "step": 178218
2754
+ },
2755
+ {
2756
+ "epoch": 1.4423193464717718,
2757
+ "grad_norm": 0.09399819374084473,
2758
+ "learning_rate": 1.639420163382057e-05,
2759
+ "loss": 0.0116,
2760
+ "step": 178500
2761
+ },
2762
+ {
2763
+ "epoch": 1.4463594566859783,
2764
+ "grad_norm": 0.06601426005363464,
2765
+ "learning_rate": 1.6384101358285058e-05,
2766
+ "loss": 0.0117,
2767
+ "step": 179000
2768
+ },
2769
+ {
2770
+ "epoch": 1.450399566900185,
2771
+ "grad_norm": 0.11237422376871109,
2772
+ "learning_rate": 1.6374001082749538e-05,
2773
+ "loss": 0.0115,
2774
+ "step": 179500
2775
+ },
2776
+ {
2777
+ "epoch": 1.4544396771143917,
2778
+ "grad_norm": 0.04262951388955116,
2779
+ "learning_rate": 1.636390080721402e-05,
2780
+ "loss": 0.0112,
2781
+ "step": 180000
2782
+ },
2783
+ {
2784
+ "epoch": 1.4584797873285984,
2785
+ "grad_norm": 0.13000500202178955,
2786
+ "learning_rate": 1.6353800531678504e-05,
2787
+ "loss": 0.0132,
2788
+ "step": 180500
2789
+ },
2790
+ {
2791
+ "epoch": 1.4625198975428049,
2792
+ "grad_norm": 0.0949823409318924,
2793
+ "learning_rate": 1.6343700256142988e-05,
2794
+ "loss": 0.0129,
2795
+ "step": 181000
2796
+ },
2797
+ {
2798
+ "epoch": 1.4665600077570116,
2799
+ "grad_norm": 0.04730290174484253,
2800
+ "learning_rate": 1.6333599980607474e-05,
2801
+ "loss": 0.0129,
2802
+ "step": 181500
2803
+ },
2804
+ {
2805
+ "epoch": 1.4706001179712183,
2806
+ "grad_norm": 0.050584714859724045,
2807
+ "learning_rate": 1.6323499705071958e-05,
2808
+ "loss": 0.013,
2809
+ "step": 182000
2810
+ },
2811
+ {
2812
+ "epoch": 1.474640228185425,
2813
+ "grad_norm": 0.1683996021747589,
2814
+ "learning_rate": 1.6313399429536438e-05,
2815
+ "loss": 0.0133,
2816
+ "step": 182500
2817
+ },
2818
+ {
2819
+ "epoch": 1.4786803383996316,
2820
+ "grad_norm": 0.1036485880613327,
2821
+ "learning_rate": 1.630329915400092e-05,
2822
+ "loss": 0.0136,
2823
+ "step": 183000
2824
+ },
2825
+ {
2826
+ "epoch": 1.4827204486138381,
2827
+ "grad_norm": 0.11697889119386673,
2828
+ "learning_rate": 1.6293198878465404e-05,
2829
+ "loss": 0.0133,
2830
+ "step": 183500
2831
+ },
2832
+ {
2833
+ "epoch": 1.4867605588280448,
2834
+ "grad_norm": 0.0688479095697403,
2835
+ "learning_rate": 1.628309860292989e-05,
2836
+ "loss": 0.0132,
2837
+ "step": 184000
2838
+ },
2839
+ {
2840
+ "epoch": 1.4908006690422515,
2841
+ "grad_norm": 0.12002038955688477,
2842
+ "learning_rate": 1.6272998327394375e-05,
2843
+ "loss": 0.0131,
2844
+ "step": 184500
2845
+ },
2846
+ {
2847
+ "epoch": 1.4948407792564582,
2848
+ "grad_norm": 0.08021160215139389,
2849
+ "learning_rate": 1.6262898051858855e-05,
2850
+ "loss": 0.0133,
2851
+ "step": 185000
2852
+ },
2853
+ {
2854
+ "epoch": 1.4988808894706647,
2855
+ "grad_norm": 0.07343757152557373,
2856
+ "learning_rate": 1.6252797776323338e-05,
2857
+ "loss": 0.0135,
2858
+ "step": 185500
2859
+ },
2860
+ {
2861
+ "epoch": 1.5029209996848714,
2862
+ "grad_norm": 0.058117810636758804,
2863
+ "learning_rate": 1.624269750078782e-05,
2864
+ "loss": 0.0143,
2865
+ "step": 186000
2866
+ },
2867
+ {
2868
+ "epoch": 1.506961109899078,
2869
+ "grad_norm": 0.10462002456188202,
2870
+ "learning_rate": 1.6232597225252308e-05,
2871
+ "loss": 0.0138,
2872
+ "step": 186500
2873
+ },
2874
+ {
2875
+ "epoch": 1.5110012201132847,
2876
+ "grad_norm": 0.07825891673564911,
2877
+ "learning_rate": 1.622249694971679e-05,
2878
+ "loss": 0.0141,
2879
+ "step": 187000
2880
+ },
2881
+ {
2882
+ "epoch": 1.5150413303274914,
2883
+ "grad_norm": 0.05809338763356209,
2884
+ "learning_rate": 1.621239667418127e-05,
2885
+ "loss": 0.0136,
2886
+ "step": 187500
2887
+ },
2888
+ {
2889
+ "epoch": 1.519081440541698,
2890
+ "grad_norm": 0.05035299435257912,
2891
+ "learning_rate": 1.6202296398645755e-05,
2892
+ "loss": 0.0165,
2893
+ "step": 188000
2894
+ },
2895
+ {
2896
+ "epoch": 1.5200429867726792,
2897
+ "eval_f1_macro": 0.9465345293457039,
2898
+ "eval_f1_micro": 0.9742553945189574,
2899
+ "eval_loss": 0.08490300178527832,
2900
+ "eval_precision_macro": 0.9844654624463276,
2901
+ "eval_precision_micro": 0.9898738168824952,
2902
+ "eval_recall_macro": 0.9150699179615246,
2903
+ "eval_recall_micro": 0.9591221775256943,
2904
+ "eval_runtime": 13330.8537,
2905
+ "eval_samples_per_second": 2.971,
2906
+ "eval_steps_per_second": 0.012,
2907
+ "step": 188119
2908
+ },
2909
+ {
2910
+ "epoch": 1.5231215507559046,
2911
+ "grad_norm": 0.14376361668109894,
2912
+ "learning_rate": 1.6192196123110238e-05,
2913
+ "loss": 0.0157,
2914
+ "step": 188500
2915
+ },
2916
+ {
2917
+ "epoch": 1.5271616609701113,
2918
+ "grad_norm": 0.07897575944662094,
2919
+ "learning_rate": 1.6182095847574725e-05,
2920
+ "loss": 0.0159,
2921
+ "step": 189000
2922
+ },
2923
+ {
2924
+ "epoch": 1.531201771184318,
2925
+ "grad_norm": 0.06912536919116974,
2926
+ "learning_rate": 1.6171995572039208e-05,
2927
+ "loss": 0.0163,
2928
+ "step": 189500
2929
+ },
2930
+ {
2931
+ "epoch": 1.5352418813985245,
2932
+ "grad_norm": 0.05066482350230217,
2933
+ "learning_rate": 1.6161895296503688e-05,
2934
+ "loss": 0.0156,
2935
+ "step": 190000
2936
+ },
2937
+ {
2938
+ "epoch": 1.5392819916127312,
2939
+ "grad_norm": 0.14292369782924652,
2940
+ "learning_rate": 1.615179502096817e-05,
2941
+ "loss": 0.0198,
2942
+ "step": 190500
2943
+ },
2944
+ {
2945
+ "epoch": 1.5433221018269379,
2946
+ "grad_norm": 0.08798356354236603,
2947
+ "learning_rate": 1.614169474543266e-05,
2948
+ "loss": 0.0197,
2949
+ "step": 191000
2950
+ },
2951
+ {
2952
+ "epoch": 1.5473622120411445,
2953
+ "grad_norm": 0.061990030109882355,
2954
+ "learning_rate": 1.6131594469897142e-05,
2955
+ "loss": 0.0183,
2956
+ "step": 191500
2957
+ },
2958
+ {
2959
+ "epoch": 1.551402322255351,
2960
+ "grad_norm": 0.05433070659637451,
2961
+ "learning_rate": 1.6121494194361625e-05,
2962
+ "loss": 0.0183,
2963
+ "step": 192000
2964
+ },
2965
+ {
2966
+ "epoch": 1.5554424324695577,
2967
+ "grad_norm": 0.13680632412433624,
2968
+ "learning_rate": 1.6111393918826105e-05,
2969
+ "loss": 0.0192,
2970
+ "step": 192500
2971
+ },
2972
+ {
2973
+ "epoch": 1.5594825426837644,
2974
+ "grad_norm": 0.1941196620464325,
2975
+ "learning_rate": 1.610129364329059e-05,
2976
+ "loss": 0.0176,
2977
+ "step": 193000
2978
+ },
2979
+ {
2980
+ "epoch": 1.563522652897971,
2981
+ "grad_norm": 0.08578658103942871,
2982
+ "learning_rate": 1.6091193367755075e-05,
2983
+ "loss": 0.0173,
2984
+ "step": 193500
2985
+ },
2986
+ {
2987
+ "epoch": 1.5675627631121778,
2988
+ "grad_norm": 0.04361563175916672,
2989
+ "learning_rate": 1.608109309221956e-05,
2990
+ "loss": 0.0171,
2991
+ "step": 194000
2992
+ },
2993
+ {
2994
+ "epoch": 1.5716028733263843,
2995
+ "grad_norm": 0.12448256462812424,
2996
+ "learning_rate": 1.6070992816684042e-05,
2997
+ "loss": 0.0168,
2998
+ "step": 194500
2999
+ },
3000
+ {
3001
+ "epoch": 1.575642983540591,
3002
+ "grad_norm": 0.10221997648477554,
3003
+ "learning_rate": 1.6060892541148522e-05,
3004
+ "loss": 0.017,
3005
+ "step": 195000
3006
+ },
3007
+ {
3008
+ "epoch": 1.5796830937547977,
3009
+ "grad_norm": 0.07009778171777725,
3010
+ "learning_rate": 1.6050792265613005e-05,
3011
+ "loss": 0.0175,
3012
+ "step": 195500
3013
+ },
3014
+ {
3015
+ "epoch": 1.5837232039690043,
3016
+ "grad_norm": 0.06714298576116562,
3017
+ "learning_rate": 1.6040691990077492e-05,
3018
+ "loss": 0.0174,
3019
+ "step": 196000
3020
+ },
3021
+ {
3022
+ "epoch": 1.5877633141832108,
3023
+ "grad_norm": 0.12766534090042114,
3024
+ "learning_rate": 1.6030591714541975e-05,
3025
+ "loss": 0.0179,
3026
+ "step": 196500
3027
+ },
3028
+ {
3029
+ "epoch": 1.5918034243974175,
3030
+ "grad_norm": 0.10328399389982224,
3031
+ "learning_rate": 1.602049143900646e-05,
3032
+ "loss": 0.0175,
3033
+ "step": 197000
3034
+ },
3035
+ {
3036
+ "epoch": 1.5958435346116242,
3037
+ "grad_norm": 0.09311484545469284,
3038
+ "learning_rate": 1.601039116347094e-05,
3039
+ "loss": 0.0172,
3040
+ "step": 197500
3041
+ },
3042
+ {
3043
+ "epoch": 1.599883644825831,
3044
+ "grad_norm": 0.08157425373792648,
3045
+ "learning_rate": 1.6000290887935422e-05,
3046
+ "loss": 0.0171,
3047
+ "step": 198000
3048
+ },
3049
+ {
3050
+ "epoch": 1.600045249234399,
3051
+ "eval_f1_macro": 0.9462988853572672,
3052
+ "eval_f1_micro": 0.9743141624468545,
3053
+ "eval_loss": 0.09824506938457489,
3054
+ "eval_precision_macro": 0.9878799683485701,
3055
+ "eval_precision_micro": 0.9928597658940401,
3056
+ "eval_recall_macro": 0.9118503420886092,
3057
+ "eval_recall_micro": 0.9564486807740753,
3058
+ "eval_runtime": 13226.8299,
3059
+ "eval_samples_per_second": 2.994,
3060
+ "eval_steps_per_second": 0.012,
3061
+ "step": 198020
3062
+ },
3063
+ {
3064
+ "epoch": 1.6039237550400376,
3065
+ "grad_norm": 0.10958320647478104,
3066
+ "learning_rate": 1.599019061239991e-05,
3067
+ "loss": 0.0167,
3068
+ "step": 198500
3069
+ },
3070
+ {
3071
+ "epoch": 1.607963865254244,
3072
+ "grad_norm": 0.07280286401510239,
3073
+ "learning_rate": 1.5980090336864392e-05,
3074
+ "loss": 0.0164,
3075
+ "step": 199000
3076
+ },
3077
+ {
3078
+ "epoch": 1.6120039754684508,
3079
+ "grad_norm": 0.0816897377371788,
3080
+ "learning_rate": 1.5969990061328876e-05,
3081
+ "loss": 0.017,
3082
+ "step": 199500
3083
+ },
3084
+ {
3085
+ "epoch": 1.6160440856826574,
3086
+ "grad_norm": 0.046233151108026505,
3087
+ "learning_rate": 1.595988978579336e-05,
3088
+ "loss": 0.0163,
3089
+ "step": 200000
3090
+ },
3091
+ {
3092
+ "epoch": 1.6200841958968641,
3093
+ "grad_norm": 0.13440461456775665,
3094
+ "learning_rate": 1.594978951025784e-05,
3095
+ "loss": 0.015,
3096
+ "step": 200500
3097
+ },
3098
+ {
3099
+ "epoch": 1.6241243061110706,
3100
+ "grad_norm": 0.0861237496137619,
3101
+ "learning_rate": 1.5939689234722326e-05,
3102
+ "loss": 0.0146,
3103
+ "step": 201000
3104
+ },
3105
+ {
3106
+ "epoch": 1.6281644163252773,
3107
+ "grad_norm": 0.06643826514482498,
3108
+ "learning_rate": 1.592958895918681e-05,
3109
+ "loss": 0.0142,
3110
+ "step": 201500
3111
+ },
3112
+ {
3113
+ "epoch": 1.632204526539484,
3114
+ "grad_norm": 0.06138383969664574,
3115
+ "learning_rate": 1.5919488683651292e-05,
3116
+ "loss": 0.0143,
3117
+ "step": 202000
3118
+ },
3119
+ {
3120
+ "epoch": 1.6362446367536907,
3121
+ "grad_norm": 0.13212205469608307,
3122
+ "learning_rate": 1.5909388408115776e-05,
3123
+ "loss": 0.0147,
3124
+ "step": 202500
3125
+ },
3126
+ {
3127
+ "epoch": 1.6402847469678974,
3128
+ "grad_norm": 0.07676049321889877,
3129
+ "learning_rate": 1.5899288132580256e-05,
3130
+ "loss": 0.0151,
3131
+ "step": 203000
3132
+ },
3133
+ {
3134
+ "epoch": 1.6443248571821039,
3135
+ "grad_norm": 0.10008609294891357,
3136
+ "learning_rate": 1.5889187857044743e-05,
3137
+ "loss": 0.0152,
3138
+ "step": 203500
3139
+ },
3140
+ {
3141
+ "epoch": 1.6483649673963106,
3142
+ "grad_norm": 0.04750071465969086,
3143
+ "learning_rate": 1.5879087581509226e-05,
3144
+ "loss": 0.0157,
3145
+ "step": 204000
3146
+ },
3147
+ {
3148
+ "epoch": 1.6524050776105172,
3149
+ "grad_norm": 0.11740187555551529,
3150
+ "learning_rate": 1.586898730597371e-05,
3151
+ "loss": 0.0155,
3152
+ "step": 204500
3153
+ },
3154
+ {
3155
+ "epoch": 1.656445187824724,
3156
+ "grad_norm": 0.06920389086008072,
3157
+ "learning_rate": 1.5858887030438193e-05,
3158
+ "loss": 0.0156,
3159
+ "step": 205000
3160
+ },
3161
+ {
3162
+ "epoch": 1.6604852980389304,
3163
+ "grad_norm": 0.05165468528866768,
3164
+ "learning_rate": 1.5848786754902673e-05,
3165
+ "loss": 0.0151,
3166
+ "step": 205500
3167
+ },
3168
+ {
3169
+ "epoch": 1.664525408253137,
3170
+ "grad_norm": 0.07880023121833801,
3171
+ "learning_rate": 1.583868647936716e-05,
3172
+ "loss": 0.0152,
3173
+ "step": 206000
3174
+ },
3175
+ {
3176
+ "epoch": 1.6685655184673438,
3177
+ "grad_norm": 0.11061804741621017,
3178
+ "learning_rate": 1.5828586203831643e-05,
3179
+ "loss": 0.015,
3180
+ "step": 206500
3181
+ },
3182
+ {
3183
+ "epoch": 1.6726056286815505,
3184
+ "grad_norm": 0.08423452824354172,
3185
+ "learning_rate": 1.5818485928296126e-05,
3186
+ "loss": 0.015,
3187
+ "step": 207000
3188
+ },
3189
+ {
3190
+ "epoch": 1.6766457388957572,
3191
+ "grad_norm": 0.07225336134433746,
3192
+ "learning_rate": 1.580838565276061e-05,
3193
+ "loss": 0.0148,
3194
+ "step": 207500
3195
+ },
3196
+ {
3197
+ "epoch": 1.680047511696119,
3198
+ "eval_f1_macro": 0.9464641577937303,
3199
+ "eval_f1_micro": 0.9740978113062491,
3200
+ "eval_loss": 0.08534455299377441,
3201
+ "eval_precision_macro": 0.9875234035936792,
3202
+ "eval_precision_micro": 0.9923821627163134,
3203
+ "eval_recall_macro": 0.9123021051137999,
3204
+ "eval_recall_micro": 0.9564750381751005,
3205
+ "eval_runtime": 12543.2073,
3206
+ "eval_samples_per_second": 3.157,
3207
+ "eval_steps_per_second": 0.012,
3208
+ "step": 207921
3209
+ },
3210
+ {
3211
+ "epoch": 1.6806858491099637,
3212
+ "grad_norm": 0.05870038643479347,
3213
+ "learning_rate": 1.579828537722509e-05,
3214
+ "loss": 0.0197,
3215
+ "step": 208000
3216
+ },
3217
+ {
3218
+ "epoch": 1.6847259593241704,
3219
+ "grad_norm": 0.11144687980413437,
3220
+ "learning_rate": 1.5788185101689576e-05,
3221
+ "loss": 0.0197,
3222
+ "step": 208500
3223
+ },
3224
+ {
3225
+ "epoch": 1.688766069538377,
3226
+ "grad_norm": 0.07475966960191727,
3227
+ "learning_rate": 1.577808482615406e-05,
3228
+ "loss": 0.0195,
3229
+ "step": 209000
3230
+ },
3231
+ {
3232
+ "epoch": 1.6928061797525837,
3233
+ "grad_norm": 0.05573410540819168,
3234
+ "learning_rate": 1.5767984550618543e-05,
3235
+ "loss": 0.0202,
3236
+ "step": 209500
3237
+ },
3238
+ {
3239
+ "epoch": 1.6968462899667902,
3240
+ "grad_norm": 0.07953529059886932,
3241
+ "learning_rate": 1.5757884275083026e-05,
3242
+ "loss": 0.0191,
3243
+ "step": 210000
3244
+ },
3245
+ {
3246
+ "epoch": 1.700886400180997,
3247
+ "grad_norm": 0.08590356260538101,
3248
+ "learning_rate": 1.5747783999547506e-05,
3249
+ "loss": 0.0147,
3250
+ "step": 210500
3251
+ },
3252
+ {
3253
+ "epoch": 1.7049265103952036,
3254
+ "grad_norm": 0.08645664155483246,
3255
+ "learning_rate": 1.5737683724011993e-05,
3256
+ "loss": 0.0145,
3257
+ "step": 211000
3258
+ },
3259
+ {
3260
+ "epoch": 1.7089666206094103,
3261
+ "grad_norm": 0.059178948402404785,
3262
+ "learning_rate": 1.5727583448476476e-05,
3263
+ "loss": 0.0139,
3264
+ "step": 211500
3265
+ },
3266
+ {
3267
+ "epoch": 1.7130067308236168,
3268
+ "grad_norm": 0.05445469170808792,
3269
+ "learning_rate": 1.571748317294096e-05,
3270
+ "loss": 0.0143,
3271
+ "step": 212000
3272
+ },
3273
+ {
3274
+ "epoch": 1.7170468410378235,
3275
+ "grad_norm": 0.10709578543901443,
3276
+ "learning_rate": 1.5707382897405443e-05,
3277
+ "loss": 0.0141,
3278
+ "step": 212500
3279
+ },
3280
+ {
3281
+ "epoch": 1.7210869512520302,
3282
+ "grad_norm": 0.0663144662976265,
3283
+ "learning_rate": 1.5697282621869923e-05,
3284
+ "loss": 0.0121,
3285
+ "step": 213000
3286
+ },
3287
+ {
3288
+ "epoch": 1.7251270614662368,
3289
+ "grad_norm": 0.0667869821190834,
3290
+ "learning_rate": 1.568718234633441e-05,
3291
+ "loss": 0.0125,
3292
+ "step": 213500
3293
+ },
3294
+ {
3295
+ "epoch": 1.7291671716804435,
3296
+ "grad_norm": 0.09561540186405182,
3297
+ "learning_rate": 1.5677082070798893e-05,
3298
+ "loss": 0.0122,
3299
+ "step": 214000
3300
+ },
3301
+ {
3302
+ "epoch": 1.73320728189465,
3303
+ "grad_norm": 0.09017562866210938,
3304
+ "learning_rate": 1.5666981795263377e-05,
3305
+ "loss": 0.0128,
3306
+ "step": 214500
3307
+ },
3308
+ {
3309
+ "epoch": 1.7372473921088567,
3310
+ "grad_norm": 0.06796102970838547,
3311
+ "learning_rate": 1.565688151972786e-05,
3312
+ "loss": 0.0129,
3313
+ "step": 215000
3314
+ },
3315
+ {
3316
+ "epoch": 1.7412875023230634,
3317
+ "grad_norm": 0.06975946575403214,
3318
+ "learning_rate": 1.5646781244192343e-05,
3319
+ "loss": 0.0126,
3320
+ "step": 215500
3321
+ },
3322
+ {
3323
+ "epoch": 1.74532761253727,
3324
+ "grad_norm": 0.04627285152673721,
3325
+ "learning_rate": 1.5636680968656827e-05,
3326
+ "loss": 0.0126,
3327
+ "step": 216000
3328
+ },
3329
+ {
3330
+ "epoch": 1.7493677227514766,
3331
+ "grad_norm": 0.12213249504566193,
3332
+ "learning_rate": 1.562658069312131e-05,
3333
+ "loss": 0.0124,
3334
+ "step": 216500
3335
+ },
3336
+ {
3337
+ "epoch": 1.7534078329656833,
3338
+ "grad_norm": 0.0799461305141449,
3339
+ "learning_rate": 1.5616480417585793e-05,
3340
+ "loss": 0.0125,
3341
+ "step": 217000
3342
+ },
3343
+ {
3344
+ "epoch": 1.75744794317989,
3345
+ "grad_norm": 0.05975542962551117,
3346
+ "learning_rate": 1.5606380142050277e-05,
3347
+ "loss": 0.0124,
3348
+ "step": 217500
3349
+ },
3350
+ {
3351
+ "epoch": 1.760049774157839,
3352
+ "eval_f1_macro": 0.9468023383313312,
3353
+ "eval_f1_micro": 0.9742265323429071,
3354
+ "eval_loss": 0.08699483424425125,
3355
+ "eval_precision_macro": 0.9835541832835969,
3356
+ "eval_precision_micro": 0.9902263876070855,
3357
+ "eval_recall_macro": 0.915890529699618,
3358
+ "eval_recall_micro": 0.9587354997620732,
3359
+ "eval_runtime": 12459.0163,
3360
+ "eval_samples_per_second": 3.179,
3361
+ "eval_steps_per_second": 0.012,
3362
+ "step": 217822
3363
+ }
3364
+ ],
3365
+ "logging_steps": 500,
3366
+ "max_steps": 990072,
3367
+ "num_input_tokens_seen": 0,
3368
+ "num_train_epochs": 8,
3369
+ "save_steps": 9901,
3370
+ "stateful_callbacks": {
3371
+ "EarlyStoppingCallback": {
3372
+ "args": {
3373
+ "early_stopping_patience": 5,
3374
+ "early_stopping_threshold": 0.0
3375
+ },
3376
+ "attributes": {
3377
+ "early_stopping_patience_counter": 5
3378
+ }
3379
+ },
3380
+ "TrainerControl": {
3381
+ "args": {
3382
+ "should_epoch_stop": false,
3383
+ "should_evaluate": false,
3384
+ "should_log": false,
3385
+ "should_save": true,
3386
+ "should_training_stop": true
3387
+ },
3388
+ "attributes": {}
3389
+ }
3390
+ },
3391
+ "total_flos": 1.468409508314638e+19,
3392
+ "train_batch_size": 256,
3393
+ "trial_name": null,
3394
+ "trial_params": null
3395
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a31c01e7409f4b2fca502cfbb4f3177e5ffdbdeb90407a5f2d7fef938a3cea16
3
+ size 5368
vocab.json ADDED
The diff for this file is too large to render. See raw diff