Hugging Face's logo

Spaces:
1NEYRON1
/
Topic_classification_for_scientific_articles
Running

App Files Community
1NEYRON1 commited on
Commit
e7cad7a
·
1 Parent(s): 74f655e

Upload 5 files

Browse files
Files changed (5) hide show
  1. checkpoint/config (4).json +305 -0
  2. checkpoint/rng_state (3).pth +3 -0
  3. checkpoint/scheduler (3).pt +3 -0
  4. checkpoint/trainer_state (3).json +2059 -0
  5. checkpoint/training_args (3).bin +3 -0
checkpoint/config (4).json ADDED
@@ -0,0 +1,305 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "gelu",
3
+ "architectures": [
4
+ "DistilBertForSequenceClassification"
5
+ ],
6
+ "attention_dropout": 0.1,
7
+ "dim": 768,
8
+ "dropout": 0.1,
9
+ "hidden_dim": 3072,
10
+ "id2label": {
11
+ "0": "LABEL_0",
12
+ "1": "LABEL_1",
13
+ "2": "LABEL_2",
14
+ "3": "LABEL_3",
15
+ "4": "LABEL_4",
16
+ "5": "LABEL_5",
17
+ "6": "LABEL_6",
18
+ "7": "LABEL_7",
19
+ "8": "LABEL_8",
20
+ "9": "LABEL_9",
21
+ "10": "LABEL_10",
22
+ "11": "LABEL_11",
23
+ "12": "LABEL_12",
24
+ "13": "LABEL_13",
25
+ "14": "LABEL_14",
26
+ "15": "LABEL_15",
27
+ "16": "LABEL_16",
28
+ "17": "LABEL_17",
29
+ "18": "LABEL_18",
30
+ "19": "LABEL_19",
31
+ "20": "LABEL_20",
32
+ "21": "LABEL_21",
33
+ "22": "LABEL_22",
34
+ "23": "LABEL_23",
35
+ "24": "LABEL_24",
36
+ "25": "LABEL_25",
37
+ "26": "LABEL_26",
38
+ "27": "LABEL_27",
39
+ "28": "LABEL_28",
40
+ "29": "LABEL_29",
41
+ "30": "LABEL_30",
42
+ "31": "LABEL_31",
43
+ "32": "LABEL_32",
44
+ "33": "LABEL_33",
45
+ "34": "LABEL_34",
46
+ "35": "LABEL_35",
47
+ "36": "LABEL_36",
48
+ "37": "LABEL_37",
49
+ "38": "LABEL_38",
50
+ "39": "LABEL_39",
51
+ "40": "LABEL_40",
52
+ "41": "LABEL_41",
53
+ "42": "LABEL_42",
54
+ "43": "LABEL_43",
55
+ "44": "LABEL_44",
56
+ "45": "LABEL_45",
57
+ "46": "LABEL_46",
58
+ "47": "LABEL_47",
59
+ "48": "LABEL_48",
60
+ "49": "LABEL_49",
61
+ "50": "LABEL_50",
62
+ "51": "LABEL_51",
63
+ "52": "LABEL_52",
64
+ "53": "LABEL_53",
65
+ "54": "LABEL_54",
66
+ "55": "LABEL_55",
67
+ "56": "LABEL_56",
68
+ "57": "LABEL_57",
69
+ "58": "LABEL_58",
70
+ "59": "LABEL_59",
71
+ "60": "LABEL_60",
72
+ "61": "LABEL_61",
73
+ "62": "LABEL_62",
74
+ "63": "LABEL_63",
75
+ "64": "LABEL_64",
76
+ "65": "LABEL_65",
77
+ "66": "LABEL_66",
78
+ "67": "LABEL_67",
79
+ "68": "LABEL_68",
80
+ "69": "LABEL_69",
81
+ "70": "LABEL_70",
82
+ "71": "LABEL_71",
83
+ "72": "LABEL_72",
84
+ "73": "LABEL_73",
85
+ "74": "LABEL_74",
86
+ "75": "LABEL_75",
87
+ "76": "LABEL_76",
88
+ "77": "LABEL_77",
89
+ "78": "LABEL_78",
90
+ "79": "LABEL_79",
91
+ "80": "LABEL_80",
92
+ "81": "LABEL_81",
93
+ "82": "LABEL_82",
94
+ "83": "LABEL_83",
95
+ "84": "LABEL_84",
96
+ "85": "LABEL_85",
97
+ "86": "LABEL_86",
98
+ "87": "LABEL_87",
99
+ "88": "LABEL_88",
100
+ "89": "LABEL_89",
101
+ "90": "LABEL_90",
102
+ "91": "LABEL_91",
103
+ "92": "LABEL_92",
104
+ "93": "LABEL_93",
105
+ "94": "LABEL_94",
106
+ "95": "LABEL_95",
107
+ "96": "LABEL_96",
108
+ "97": "LABEL_97",
109
+ "98": "LABEL_98",
110
+ "99": "LABEL_99",
111
+ "100": "LABEL_100",
112
+ "101": "LABEL_101",
113
+ "102": "LABEL_102",
114
+ "103": "LABEL_103",
115
+ "104": "LABEL_104",
116
+ "105": "LABEL_105",
117
+ "106": "LABEL_106",
118
+ "107": "LABEL_107",
119
+ "108": "LABEL_108",
120
+ "109": "LABEL_109",
121
+ "110": "LABEL_110",
122
+ "111": "LABEL_111",
123
+ "112": "LABEL_112",
124
+ "113": "LABEL_113",
125
+ "114": "LABEL_114",
126
+ "115": "LABEL_115",
127
+ "116": "LABEL_116",
128
+ "117": "LABEL_117",
129
+ "118": "LABEL_118",
130
+ "119": "LABEL_119",
131
+ "120": "LABEL_120",
132
+ "121": "LABEL_121",
133
+ "122": "LABEL_122",
134
+ "123": "LABEL_123",
135
+ "124": "LABEL_124",
136
+ "125": "LABEL_125",
137
+ "126": "LABEL_126",
138
+ "127": "LABEL_127",
139
+ "128": "LABEL_128",
140
+ "129": "LABEL_129",
141
+ "130": "LABEL_130",
142
+ "131": "LABEL_131",
143
+ "132": "LABEL_132",
144
+ "133": "LABEL_133",
145
+ "134": "LABEL_134",
146
+ "135": "LABEL_135",
147
+ "136": "LABEL_136",
148
+ "137": "LABEL_137"
149
+ },
150
+ "initializer_range": 0.02,
151
+ "label2id": {
152
+ "LABEL_0": 0,
153
+ "LABEL_1": 1,
154
+ "LABEL_10": 10,
155
+ "LABEL_100": 100,
156
+ "LABEL_101": 101,
157
+ "LABEL_102": 102,
158
+ "LABEL_103": 103,
159
+ "LABEL_104": 104,
160
+ "LABEL_105": 105,
161
+ "LABEL_106": 106,
162
+ "LABEL_107": 107,
163
+ "LABEL_108": 108,
164
+ "LABEL_109": 109,
165
+ "LABEL_11": 11,
166
+ "LABEL_110": 110,
167
+ "LABEL_111": 111,
168
+ "LABEL_112": 112,
169
+ "LABEL_113": 113,
170
+ "LABEL_114": 114,
171
+ "LABEL_115": 115,
172
+ "LABEL_116": 116,
173
+ "LABEL_117": 117,
174
+ "LABEL_118": 118,
175
+ "LABEL_119": 119,
176
+ "LABEL_12": 12,
177
+ "LABEL_120": 120,
178
+ "LABEL_121": 121,
179
+ "LABEL_122": 122,
180
+ "LABEL_123": 123,
181
+ "LABEL_124": 124,
182
+ "LABEL_125": 125,
183
+ "LABEL_126": 126,
184
+ "LABEL_127": 127,
185
+ "LABEL_128": 128,
186
+ "LABEL_129": 129,
187
+ "LABEL_13": 13,
188
+ "LABEL_130": 130,
189
+ "LABEL_131": 131,
190
+ "LABEL_132": 132,
191
+ "LABEL_133": 133,
192
+ "LABEL_134": 134,
193
+ "LABEL_135": 135,
194
+ "LABEL_136": 136,
195
+ "LABEL_137": 137,
196
+ "LABEL_14": 14,
197
+ "LABEL_15": 15,
198
+ "LABEL_16": 16,
199
+ "LABEL_17": 17,
200
+ "LABEL_18": 18,
201
+ "LABEL_19": 19,
202
+ "LABEL_2": 2,
203
+ "LABEL_20": 20,
204
+ "LABEL_21": 21,
205
+ "LABEL_22": 22,
206
+ "LABEL_23": 23,
207
+ "LABEL_24": 24,
208
+ "LABEL_25": 25,
209
+ "LABEL_26": 26,
210
+ "LABEL_27": 27,
211
+ "LABEL_28": 28,
212
+ "LABEL_29": 29,
213
+ "LABEL_3": 3,
214
+ "LABEL_30": 30,
215
+ "LABEL_31": 31,
216
+ "LABEL_32": 32,
217
+ "LABEL_33": 33,
218
+ "LABEL_34": 34,
219
+ "LABEL_35": 35,
220
+ "LABEL_36": 36,
221
+ "LABEL_37": 37,
222
+ "LABEL_38": 38,
223
+ "LABEL_39": 39,
224
+ "LABEL_4": 4,
225
+ "LABEL_40": 40,
226
+ "LABEL_41": 41,
227
+ "LABEL_42": 42,
228
+ "LABEL_43": 43,
229
+ "LABEL_44": 44,
230
+ "LABEL_45": 45,
231
+ "LABEL_46": 46,
232
+ "LABEL_47": 47,
233
+ "LABEL_48": 48,
234
+ "LABEL_49": 49,
235
+ "LABEL_5": 5,
236
+ "LABEL_50": 50,
237
+ "LABEL_51": 51,
238
+ "LABEL_52": 52,
239
+ "LABEL_53": 53,
240
+ "LABEL_54": 54,
241
+ "LABEL_55": 55,
242
+ "LABEL_56": 56,
243
+ "LABEL_57": 57,
244
+ "LABEL_58": 58,
245
+ "LABEL_59": 59,
246
+ "LABEL_6": 6,
247
+ "LABEL_60": 60,
248
+ "LABEL_61": 61,
249
+ "LABEL_62": 62,
250
+ "LABEL_63": 63,
251
+ "LABEL_64": 64,
252
+ "LABEL_65": 65,
253
+ "LABEL_66": 66,
254
+ "LABEL_67": 67,
255
+ "LABEL_68": 68,
256
+ "LABEL_69": 69,
257
+ "LABEL_7": 7,
258
+ "LABEL_70": 70,
259
+ "LABEL_71": 71,
260
+ "LABEL_72": 72,
261
+ "LABEL_73": 73,
262
+ "LABEL_74": 74,
263
+ "LABEL_75": 75,
264
+ "LABEL_76": 76,
265
+ "LABEL_77": 77,
266
+ "LABEL_78": 78,
267
+ "LABEL_79": 79,
268
+ "LABEL_8": 8,
269
+ "LABEL_80": 80,
270
+ "LABEL_81": 81,
271
+ "LABEL_82": 82,
272
+ "LABEL_83": 83,
273
+ "LABEL_84": 84,
274
+ "LABEL_85": 85,
275
+ "LABEL_86": 86,
276
+ "LABEL_87": 87,
277
+ "LABEL_88": 88,
278
+ "LABEL_89": 89,
279
+ "LABEL_9": 9,
280
+ "LABEL_90": 90,
281
+ "LABEL_91": 91,
282
+ "LABEL_92": 92,
283
+ "LABEL_93": 93,
284
+ "LABEL_94": 94,
285
+ "LABEL_95": 95,
286
+ "LABEL_96": 96,
287
+ "LABEL_97": 97,
288
+ "LABEL_98": 98,
289
+ "LABEL_99": 99
290
+ },
291
+ "max_position_embeddings": 512,
292
+ "model_type": "distilbert",
293
+ "n_heads": 12,
294
+ "n_layers": 6,
295
+ "output_past": true,
296
+ "pad_token_id": 0,
297
+ "problem_type": "multi_label_classification",
298
+ "qa_dropout": 0.1,
299
+ "seq_classif_dropout": 0.2,
300
+ "sinusoidal_pos_embds": false,
301
+ "tie_weights_": true,
302
+ "torch_dtype": "float32",
303
+ "transformers_version": "4.50.3",
304
+ "vocab_size": 28996
305
+ }
checkpoint/rng_state (3).pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c50d9628402a32df38c032752f3d8a88e0c9e2bf51198d620654679fa05dafea
3
+ size 14244
checkpoint/scheduler (3).pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0cdc803c58331651f83de9a076d84c91dc54e425a92682d176ffa7127a5a1c12
3
+ size 1064
checkpoint/trainer_state (3).json ADDED
@@ -0,0 +1,2059 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 2500,
3
+ "best_metric": 0.6559016016048936,
4
+ "best_model_checkpoint": "./results/checkpoint-2500",
5
+ "epoch": 1.2195121951219512,
6
+ "eval_steps": 100,
7
+ "global_step": 2500,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.004878048780487805,
14
+ "grad_norm": 0.5707780718803406,
15
+ "learning_rate": 1.0000000000000002e-06,
16
+ "loss": 0.373,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.00975609756097561,
21
+ "grad_norm": 0.5740946531295776,
22
+ "learning_rate": 2.0000000000000003e-06,
23
+ "loss": 0.3695,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.014634146341463415,
28
+ "grad_norm": 0.5766364336013794,
29
+ "learning_rate": 3e-06,
30
+ "loss": 0.3651,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.01951219512195122,
35
+ "grad_norm": 0.5667976140975952,
36
+ "learning_rate": 4.000000000000001e-06,
37
+ "loss": 0.3589,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.024390243902439025,
42
+ "grad_norm": 0.5701761841773987,
43
+ "learning_rate": 5e-06,
44
+ "loss": 0.3521,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.02926829268292683,
49
+ "grad_norm": 0.562853991985321,
50
+ "learning_rate": 6e-06,
51
+ "loss": 0.3424,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 0.03414634146341464,
56
+ "grad_norm": 0.558131754398346,
57
+ "learning_rate": 7.000000000000001e-06,
58
+ "loss": 0.3304,
59
+ "step": 70
60
+ },
61
+ {
62
+ "epoch": 0.03902439024390244,
63
+ "grad_norm": 0.5538951754570007,
64
+ "learning_rate": 8.000000000000001e-06,
65
+ "loss": 0.315,
66
+ "step": 80
67
+ },
68
+ {
69
+ "epoch": 0.04390243902439024,
70
+ "grad_norm": 0.5401511788368225,
71
+ "learning_rate": 9e-06,
72
+ "loss": 0.2998,
73
+ "step": 90
74
+ },
75
+ {
76
+ "epoch": 0.04878048780487805,
77
+ "grad_norm": 0.5289145708084106,
78
+ "learning_rate": 1e-05,
79
+ "loss": 0.2813,
80
+ "step": 100
81
+ },
82
+ {
83
+ "epoch": 0.04878048780487805,
84
+ "eval_f1": 0.02288934762189008,
85
+ "eval_loss": 0.2639255225658417,
86
+ "eval_precision": 0.011585886400544311,
87
+ "eval_recall": 0.938986325802616,
88
+ "eval_runtime": 132.4738,
89
+ "eval_samples_per_second": 61.899,
90
+ "eval_steps_per_second": 0.974,
91
+ "step": 100
92
+ },
93
+ {
94
+ "epoch": 0.05365853658536585,
95
+ "grad_norm": 0.5152837038040161,
96
+ "learning_rate": 1.1000000000000001e-05,
97
+ "loss": 0.2626,
98
+ "step": 110
99
+ },
100
+ {
101
+ "epoch": 0.05853658536585366,
102
+ "grad_norm": 0.48544010519981384,
103
+ "learning_rate": 1.2e-05,
104
+ "loss": 0.2423,
105
+ "step": 120
106
+ },
107
+ {
108
+ "epoch": 0.06341463414634146,
109
+ "grad_norm": 0.47114691138267517,
110
+ "learning_rate": 1.3000000000000001e-05,
111
+ "loss": 0.2222,
112
+ "step": 130
113
+ },
114
+ {
115
+ "epoch": 0.06829268292682927,
116
+ "grad_norm": 0.44177231192588806,
117
+ "learning_rate": 1.4000000000000001e-05,
118
+ "loss": 0.2014,
119
+ "step": 140
120
+ },
121
+ {
122
+ "epoch": 0.07317073170731707,
123
+ "grad_norm": 0.4138778746128082,
124
+ "learning_rate": 1.5e-05,
125
+ "loss": 0.1792,
126
+ "step": 150
127
+ },
128
+ {
129
+ "epoch": 0.07804878048780488,
130
+ "grad_norm": 0.3736928403377533,
131
+ "learning_rate": 1.6000000000000003e-05,
132
+ "loss": 0.1606,
133
+ "step": 160
134
+ },
135
+ {
136
+ "epoch": 0.08292682926829269,
137
+ "grad_norm": 0.3438684046268463,
138
+ "learning_rate": 1.7000000000000003e-05,
139
+ "loss": 0.1425,
140
+ "step": 170
141
+ },
142
+ {
143
+ "epoch": 0.08780487804878048,
144
+ "grad_norm": 0.29839885234832764,
145
+ "learning_rate": 1.8e-05,
146
+ "loss": 0.1254,
147
+ "step": 180
148
+ },
149
+ {
150
+ "epoch": 0.09268292682926829,
151
+ "grad_norm": 0.27134254574775696,
152
+ "learning_rate": 1.9e-05,
153
+ "loss": 0.1091,
154
+ "step": 190
155
+ },
156
+ {
157
+ "epoch": 0.0975609756097561,
158
+ "grad_norm": 0.24228985607624054,
159
+ "learning_rate": 2e-05,
160
+ "loss": 0.0955,
161
+ "step": 200
162
+ },
163
+ {
164
+ "epoch": 0.0975609756097561,
165
+ "eval_f1": 0.40798082528988794,
166
+ "eval_loss": 0.08500728011131287,
167
+ "eval_precision": 0.2875361436615431,
168
+ "eval_recall": 0.7020659928656362,
169
+ "eval_runtime": 133.9225,
170
+ "eval_samples_per_second": 61.229,
171
+ "eval_steps_per_second": 0.963,
172
+ "step": 200
173
+ },
174
+ {
175
+ "epoch": 0.1024390243902439,
176
+ "grad_norm": 0.21252021193504333,
177
+ "learning_rate": 2.1e-05,
178
+ "loss": 0.0858,
179
+ "step": 210
180
+ },
181
+ {
182
+ "epoch": 0.1073170731707317,
183
+ "grad_norm": 0.20333224534988403,
184
+ "learning_rate": 2.2000000000000003e-05,
185
+ "loss": 0.0761,
186
+ "step": 220
187
+ },
188
+ {
189
+ "epoch": 0.11219512195121951,
190
+ "grad_norm": 0.1808163970708847,
191
+ "learning_rate": 2.3000000000000003e-05,
192
+ "loss": 0.0707,
193
+ "step": 230
194
+ },
195
+ {
196
+ "epoch": 0.11707317073170732,
197
+ "grad_norm": 0.14729662239551544,
198
+ "learning_rate": 2.4e-05,
199
+ "loss": 0.0631,
200
+ "step": 240
201
+ },
202
+ {
203
+ "epoch": 0.12195121951219512,
204
+ "grad_norm": 0.16881898045539856,
205
+ "learning_rate": 2.5e-05,
206
+ "loss": 0.0588,
207
+ "step": 250
208
+ },
209
+ {
210
+ "epoch": 0.12682926829268293,
211
+ "grad_norm": 0.17096662521362305,
212
+ "learning_rate": 2.6000000000000002e-05,
213
+ "loss": 0.0535,
214
+ "step": 260
215
+ },
216
+ {
217
+ "epoch": 0.13170731707317074,
218
+ "grad_norm": 0.15806329250335693,
219
+ "learning_rate": 2.7000000000000002e-05,
220
+ "loss": 0.0507,
221
+ "step": 270
222
+ },
223
+ {
224
+ "epoch": 0.13658536585365855,
225
+ "grad_norm": 0.10550623387098312,
226
+ "learning_rate": 2.8000000000000003e-05,
227
+ "loss": 0.0529,
228
+ "step": 280
229
+ },
230
+ {
231
+ "epoch": 0.14146341463414633,
232
+ "grad_norm": 0.1570628583431244,
233
+ "learning_rate": 2.9e-05,
234
+ "loss": 0.0464,
235
+ "step": 290
236
+ },
237
+ {
238
+ "epoch": 0.14634146341463414,
239
+ "grad_norm": 0.11487462371587753,
240
+ "learning_rate": 3e-05,
241
+ "loss": 0.0455,
242
+ "step": 300
243
+ },
244
+ {
245
+ "epoch": 0.14634146341463414,
246
+ "eval_f1": 0.40842269111034246,
247
+ "eval_loss": 0.04193972423672676,
248
+ "eval_precision": 0.28798780487804876,
249
+ "eval_recall": 0.7019916765755053,
250
+ "eval_runtime": 131.8065,
251
+ "eval_samples_per_second": 62.212,
252
+ "eval_steps_per_second": 0.979,
253
+ "step": 300
254
+ },
255
+ {
256
+ "epoch": 0.15121951219512195,
257
+ "grad_norm": 0.10432987660169601,
258
+ "learning_rate": 3.1e-05,
259
+ "loss": 0.0414,
260
+ "step": 310
261
+ },
262
+ {
263
+ "epoch": 0.15609756097560976,
264
+ "grad_norm": 0.12195131927728653,
265
+ "learning_rate": 3.2000000000000005e-05,
266
+ "loss": 0.0415,
267
+ "step": 320
268
+ },
269
+ {
270
+ "epoch": 0.16097560975609757,
271
+ "grad_norm": 0.11948198080062866,
272
+ "learning_rate": 3.3e-05,
273
+ "loss": 0.0415,
274
+ "step": 330
275
+ },
276
+ {
277
+ "epoch": 0.16585365853658537,
278
+ "grad_norm": 0.11321187764406204,
279
+ "learning_rate": 3.4000000000000007e-05,
280
+ "loss": 0.0402,
281
+ "step": 340
282
+ },
283
+ {
284
+ "epoch": 0.17073170731707318,
285
+ "grad_norm": 0.13302814960479736,
286
+ "learning_rate": 3.5e-05,
287
+ "loss": 0.0385,
288
+ "step": 350
289
+ },
290
+ {
291
+ "epoch": 0.17560975609756097,
292
+ "grad_norm": 0.18504567444324493,
293
+ "learning_rate": 3.6e-05,
294
+ "loss": 0.0393,
295
+ "step": 360
296
+ },
297
+ {
298
+ "epoch": 0.18048780487804877,
299
+ "grad_norm": 0.11665136367082596,
300
+ "learning_rate": 3.7e-05,
301
+ "loss": 0.0361,
302
+ "step": 370
303
+ },
304
+ {
305
+ "epoch": 0.18536585365853658,
306
+ "grad_norm": 0.09814433008432388,
307
+ "learning_rate": 3.8e-05,
308
+ "loss": 0.0373,
309
+ "step": 380
310
+ },
311
+ {
312
+ "epoch": 0.1902439024390244,
313
+ "grad_norm": 0.10006968677043915,
314
+ "learning_rate": 3.9000000000000006e-05,
315
+ "loss": 0.0336,
316
+ "step": 390
317
+ },
318
+ {
319
+ "epoch": 0.1951219512195122,
320
+ "grad_norm": 0.14240923523902893,
321
+ "learning_rate": 4e-05,
322
+ "loss": 0.0363,
323
+ "step": 400
324
+ },
325
+ {
326
+ "epoch": 0.1951219512195122,
327
+ "eval_f1": 0.4903333065186496,
328
+ "eval_loss": 0.03265204280614853,
329
+ "eval_precision": 0.383563367873474,
330
+ "eval_recall": 0.6794738406658739,
331
+ "eval_runtime": 131.7664,
332
+ "eval_samples_per_second": 62.231,
333
+ "eval_steps_per_second": 0.979,
334
+ "step": 400
335
+ },
336
+ {
337
+ "epoch": 0.2,
338
+ "grad_norm": 0.09334200620651245,
339
+ "learning_rate": 4.1e-05,
340
+ "loss": 0.0332,
341
+ "step": 410
342
+ },
343
+ {
344
+ "epoch": 0.2048780487804878,
345
+ "grad_norm": 0.0977184921503067,
346
+ "learning_rate": 4.2e-05,
347
+ "loss": 0.0326,
348
+ "step": 420
349
+ },
350
+ {
351
+ "epoch": 0.2097560975609756,
352
+ "grad_norm": 0.12227483093738556,
353
+ "learning_rate": 4.3e-05,
354
+ "loss": 0.0321,
355
+ "step": 430
356
+ },
357
+ {
358
+ "epoch": 0.2146341463414634,
359
+ "grad_norm": 0.07999078929424286,
360
+ "learning_rate": 4.4000000000000006e-05,
361
+ "loss": 0.0304,
362
+ "step": 440
363
+ },
364
+ {
365
+ "epoch": 0.21951219512195122,
366
+ "grad_norm": 0.13925114274024963,
367
+ "learning_rate": 4.5e-05,
368
+ "loss": 0.0296,
369
+ "step": 450
370
+ },
371
+ {
372
+ "epoch": 0.22439024390243903,
373
+ "grad_norm": 0.09850054234266281,
374
+ "learning_rate": 4.600000000000001e-05,
375
+ "loss": 0.0286,
376
+ "step": 460
377
+ },
378
+ {
379
+ "epoch": 0.22926829268292684,
380
+ "grad_norm": 0.09087326377630234,
381
+ "learning_rate": 4.7e-05,
382
+ "loss": 0.0293,
383
+ "step": 470
384
+ },
385
+ {
386
+ "epoch": 0.23414634146341465,
387
+ "grad_norm": 0.11158157885074615,
388
+ "learning_rate": 4.8e-05,
389
+ "loss": 0.0329,
390
+ "step": 480
391
+ },
392
+ {
393
+ "epoch": 0.23902439024390243,
394
+ "grad_norm": 0.0853157714009285,
395
+ "learning_rate": 4.9e-05,
396
+ "loss": 0.0302,
397
+ "step": 490
398
+ },
399
+ {
400
+ "epoch": 0.24390243902439024,
401
+ "grad_norm": 0.1297878473997116,
402
+ "learning_rate": 5e-05,
403
+ "loss": 0.0321,
404
+ "step": 500
405
+ },
406
+ {
407
+ "epoch": 0.24390243902439024,
408
+ "eval_f1": 0.5804941656357618,
409
+ "eval_loss": 0.02912677265703678,
410
+ "eval_precision": 0.5276241414939524,
411
+ "eval_recall": 0.645139714625446,
412
+ "eval_runtime": 132.6761,
413
+ "eval_samples_per_second": 61.805,
414
+ "eval_steps_per_second": 0.972,
415
+ "step": 500
416
+ },
417
+ {
418
+ "epoch": 0.24878048780487805,
419
+ "grad_norm": 0.1282796561717987,
420
+ "learning_rate": 4.9911504424778765e-05,
421
+ "loss": 0.0316,
422
+ "step": 510
423
+ },
424
+ {
425
+ "epoch": 0.25365853658536586,
426
+ "grad_norm": 0.12388455867767334,
427
+ "learning_rate": 4.982300884955752e-05,
428
+ "loss": 0.0302,
429
+ "step": 520
430
+ },
431
+ {
432
+ "epoch": 0.25853658536585367,
433
+ "grad_norm": 0.11067409813404083,
434
+ "learning_rate": 4.9734513274336284e-05,
435
+ "loss": 0.0299,
436
+ "step": 530
437
+ },
438
+ {
439
+ "epoch": 0.2634146341463415,
440
+ "grad_norm": 0.16229726374149323,
441
+ "learning_rate": 4.964601769911505e-05,
442
+ "loss": 0.0269,
443
+ "step": 540
444
+ },
445
+ {
446
+ "epoch": 0.2682926829268293,
447
+ "grad_norm": 0.12282289564609528,
448
+ "learning_rate": 4.955752212389381e-05,
449
+ "loss": 0.0303,
450
+ "step": 550
451
+ },
452
+ {
453
+ "epoch": 0.2731707317073171,
454
+ "grad_norm": 0.11870788037776947,
455
+ "learning_rate": 4.946902654867257e-05,
456
+ "loss": 0.031,
457
+ "step": 560
458
+ },
459
+ {
460
+ "epoch": 0.2780487804878049,
461
+ "grad_norm": 0.17082759737968445,
462
+ "learning_rate": 4.938053097345133e-05,
463
+ "loss": 0.0318,
464
+ "step": 570
465
+ },
466
+ {
467
+ "epoch": 0.28292682926829266,
468
+ "grad_norm": 0.10085482150316238,
469
+ "learning_rate": 4.929203539823009e-05,
470
+ "loss": 0.0269,
471
+ "step": 580
472
+ },
473
+ {
474
+ "epoch": 0.28780487804878047,
475
+ "grad_norm": 0.15493524074554443,
476
+ "learning_rate": 4.9203539823008854e-05,
477
+ "loss": 0.0308,
478
+ "step": 590
479
+ },
480
+ {
481
+ "epoch": 0.2926829268292683,
482
+ "grad_norm": 0.106038898229599,
483
+ "learning_rate": 4.911504424778761e-05,
484
+ "loss": 0.0284,
485
+ "step": 600
486
+ },
487
+ {
488
+ "epoch": 0.2926829268292683,
489
+ "eval_f1": 0.5525415591430115,
490
+ "eval_loss": 0.027496203780174255,
491
+ "eval_precision": 0.4633618520382486,
492
+ "eval_recall": 0.684230083234245,
493
+ "eval_runtime": 132.4968,
494
+ "eval_samples_per_second": 61.888,
495
+ "eval_steps_per_second": 0.974,
496
+ "step": 600
497
+ },
498
+ {
499
+ "epoch": 0.2975609756097561,
500
+ "grad_norm": 0.0896712988615036,
501
+ "learning_rate": 4.902654867256637e-05,
502
+ "loss": 0.0279,
503
+ "step": 610
504
+ },
505
+ {
506
+ "epoch": 0.3024390243902439,
507
+ "grad_norm": 0.0858699232339859,
508
+ "learning_rate": 4.893805309734513e-05,
509
+ "loss": 0.0287,
510
+ "step": 620
511
+ },
512
+ {
513
+ "epoch": 0.3073170731707317,
514
+ "grad_norm": 0.09710809588432312,
515
+ "learning_rate": 4.88495575221239e-05,
516
+ "loss": 0.0253,
517
+ "step": 630
518
+ },
519
+ {
520
+ "epoch": 0.3121951219512195,
521
+ "grad_norm": 0.09191035479307175,
522
+ "learning_rate": 4.876106194690266e-05,
523
+ "loss": 0.0251,
524
+ "step": 640
525
+ },
526
+ {
527
+ "epoch": 0.3170731707317073,
528
+ "grad_norm": 0.08466064929962158,
529
+ "learning_rate": 4.867256637168142e-05,
530
+ "loss": 0.0241,
531
+ "step": 650
532
+ },
533
+ {
534
+ "epoch": 0.32195121951219513,
535
+ "grad_norm": 0.10560336709022522,
536
+ "learning_rate": 4.858407079646018e-05,
537
+ "loss": 0.0263,
538
+ "step": 660
539
+ },
540
+ {
541
+ "epoch": 0.32682926829268294,
542
+ "grad_norm": 0.1608184576034546,
543
+ "learning_rate": 4.849557522123894e-05,
544
+ "loss": 0.027,
545
+ "step": 670
546
+ },
547
+ {
548
+ "epoch": 0.33170731707317075,
549
+ "grad_norm": 0.10734377801418304,
550
+ "learning_rate": 4.84070796460177e-05,
551
+ "loss": 0.0281,
552
+ "step": 680
553
+ },
554
+ {
555
+ "epoch": 0.33658536585365856,
556
+ "grad_norm": 0.07804067432880402,
557
+ "learning_rate": 4.831858407079646e-05,
558
+ "loss": 0.0274,
559
+ "step": 690
560
+ },
561
+ {
562
+ "epoch": 0.34146341463414637,
563
+ "grad_norm": 0.08093168586492538,
564
+ "learning_rate": 4.823008849557522e-05,
565
+ "loss": 0.029,
566
+ "step": 700
567
+ },
568
+ {
569
+ "epoch": 0.34146341463414637,
570
+ "eval_f1": 0.626304952497923,
571
+ "eval_loss": 0.026616454124450684,
572
+ "eval_precision": 0.6093343642370141,
573
+ "eval_recall": 0.6442479191438764,
574
+ "eval_runtime": 132.4172,
575
+ "eval_samples_per_second": 61.925,
576
+ "eval_steps_per_second": 0.974,
577
+ "step": 700
578
+ },
579
+ {
580
+ "epoch": 0.3463414634146341,
581
+ "grad_norm": 0.12165658921003342,
582
+ "learning_rate": 4.814159292035398e-05,
583
+ "loss": 0.0303,
584
+ "step": 710
585
+ },
586
+ {
587
+ "epoch": 0.35121951219512193,
588
+ "grad_norm": 0.11557145416736603,
589
+ "learning_rate": 4.805309734513275e-05,
590
+ "loss": 0.0288,
591
+ "step": 720
592
+ },
593
+ {
594
+ "epoch": 0.35609756097560974,
595
+ "grad_norm": 0.120982825756073,
596
+ "learning_rate": 4.7964601769911506e-05,
597
+ "loss": 0.0251,
598
+ "step": 730
599
+ },
600
+ {
601
+ "epoch": 0.36097560975609755,
602
+ "grad_norm": 0.09892363101243973,
603
+ "learning_rate": 4.787610619469027e-05,
604
+ "loss": 0.0292,
605
+ "step": 740
606
+ },
607
+ {
608
+ "epoch": 0.36585365853658536,
609
+ "grad_norm": 0.10615638643503189,
610
+ "learning_rate": 4.778761061946903e-05,
611
+ "loss": 0.0258,
612
+ "step": 750
613
+ },
614
+ {
615
+ "epoch": 0.37073170731707317,
616
+ "grad_norm": 0.1078324243426323,
617
+ "learning_rate": 4.769911504424779e-05,
618
+ "loss": 0.0249,
619
+ "step": 760
620
+ },
621
+ {
622
+ "epoch": 0.375609756097561,
623
+ "grad_norm": 0.131972536444664,
624
+ "learning_rate": 4.761061946902655e-05,
625
+ "loss": 0.0291,
626
+ "step": 770
627
+ },
628
+ {
629
+ "epoch": 0.3804878048780488,
630
+ "grad_norm": 0.08731543272733688,
631
+ "learning_rate": 4.752212389380531e-05,
632
+ "loss": 0.0294,
633
+ "step": 780
634
+ },
635
+ {
636
+ "epoch": 0.3853658536585366,
637
+ "grad_norm": 0.14496171474456787,
638
+ "learning_rate": 4.743362831858407e-05,
639
+ "loss": 0.0235,
640
+ "step": 790
641
+ },
642
+ {
643
+ "epoch": 0.3902439024390244,
644
+ "grad_norm": 0.10081491619348526,
645
+ "learning_rate": 4.734513274336283e-05,
646
+ "loss": 0.0304,
647
+ "step": 800
648
+ },
649
+ {
650
+ "epoch": 0.3902439024390244,
651
+ "eval_f1": 0.5844212089339851,
652
+ "eval_loss": 0.02550012618303299,
653
+ "eval_precision": 0.5005829358770535,
654
+ "eval_recall": 0.7019916765755053,
655
+ "eval_runtime": 132.3532,
656
+ "eval_samples_per_second": 61.955,
657
+ "eval_steps_per_second": 0.975,
658
+ "step": 800
659
+ },
660
+ {
661
+ "epoch": 0.3951219512195122,
662
+ "grad_norm": 0.07158678770065308,
663
+ "learning_rate": 4.7256637168141595e-05,
664
+ "loss": 0.025,
665
+ "step": 810
666
+ },
667
+ {
668
+ "epoch": 0.4,
669
+ "grad_norm": 0.1092309057712555,
670
+ "learning_rate": 4.716814159292036e-05,
671
+ "loss": 0.0257,
672
+ "step": 820
673
+ },
674
+ {
675
+ "epoch": 0.40487804878048783,
676
+ "grad_norm": 0.10517269372940063,
677
+ "learning_rate": 4.707964601769912e-05,
678
+ "loss": 0.0235,
679
+ "step": 830
680
+ },
681
+ {
682
+ "epoch": 0.4097560975609756,
683
+ "grad_norm": 0.0983092337846756,
684
+ "learning_rate": 4.699115044247788e-05,
685
+ "loss": 0.0257,
686
+ "step": 840
687
+ },
688
+ {
689
+ "epoch": 0.4146341463414634,
690
+ "grad_norm": 0.11815937608480453,
691
+ "learning_rate": 4.690265486725664e-05,
692
+ "loss": 0.0273,
693
+ "step": 850
694
+ },
695
+ {
696
+ "epoch": 0.4195121951219512,
697
+ "grad_norm": 0.13173235952854156,
698
+ "learning_rate": 4.6814159292035396e-05,
699
+ "loss": 0.0266,
700
+ "step": 860
701
+ },
702
+ {
703
+ "epoch": 0.424390243902439,
704
+ "grad_norm": 0.08506595343351364,
705
+ "learning_rate": 4.672566371681416e-05,
706
+ "loss": 0.0243,
707
+ "step": 870
708
+ },
709
+ {
710
+ "epoch": 0.4292682926829268,
711
+ "grad_norm": 0.07343988120555878,
712
+ "learning_rate": 4.663716814159292e-05,
713
+ "loss": 0.026,
714
+ "step": 880
715
+ },
716
+ {
717
+ "epoch": 0.43414634146341463,
718
+ "grad_norm": 0.08313016593456268,
719
+ "learning_rate": 4.6548672566371684e-05,
720
+ "loss": 0.0283,
721
+ "step": 890
722
+ },
723
+ {
724
+ "epoch": 0.43902439024390244,
725
+ "grad_norm": 0.07794070988893509,
726
+ "learning_rate": 4.646017699115045e-05,
727
+ "loss": 0.0267,
728
+ "step": 900
729
+ },
730
+ {
731
+ "epoch": 0.43902439024390244,
732
+ "eval_f1": 0.5883609508987152,
733
+ "eval_loss": 0.024745287373661995,
734
+ "eval_precision": 0.4991456531869725,
735
+ "eval_recall": 0.7164090368608799,
736
+ "eval_runtime": 132.4255,
737
+ "eval_samples_per_second": 61.922,
738
+ "eval_steps_per_second": 0.974,
739
+ "step": 900
740
+ },
741
+ {
742
+ "epoch": 0.44390243902439025,
743
+ "grad_norm": 0.1075003445148468,
744
+ "learning_rate": 4.637168141592921e-05,
745
+ "loss": 0.0285,
746
+ "step": 910
747
+ },
748
+ {
749
+ "epoch": 0.44878048780487806,
750
+ "grad_norm": 0.11088255792856216,
751
+ "learning_rate": 4.6283185840707966e-05,
752
+ "loss": 0.0229,
753
+ "step": 920
754
+ },
755
+ {
756
+ "epoch": 0.45365853658536587,
757
+ "grad_norm": 0.10665366053581238,
758
+ "learning_rate": 4.619469026548673e-05,
759
+ "loss": 0.027,
760
+ "step": 930
761
+ },
762
+ {
763
+ "epoch": 0.4585365853658537,
764
+ "grad_norm": 0.11324100941419601,
765
+ "learning_rate": 4.6106194690265485e-05,
766
+ "loss": 0.0265,
767
+ "step": 940
768
+ },
769
+ {
770
+ "epoch": 0.4634146341463415,
771
+ "grad_norm": 0.1377974897623062,
772
+ "learning_rate": 4.601769911504425e-05,
773
+ "loss": 0.0249,
774
+ "step": 950
775
+ },
776
+ {
777
+ "epoch": 0.4682926829268293,
778
+ "grad_norm": 0.08704473823308945,
779
+ "learning_rate": 4.592920353982301e-05,
780
+ "loss": 0.0275,
781
+ "step": 960
782
+ },
783
+ {
784
+ "epoch": 0.47317073170731705,
785
+ "grad_norm": 0.07551635056734085,
786
+ "learning_rate": 4.584070796460177e-05,
787
+ "loss": 0.0267,
788
+ "step": 970
789
+ },
790
+ {
791
+ "epoch": 0.47804878048780486,
792
+ "grad_norm": 0.06435199081897736,
793
+ "learning_rate": 4.5752212389380536e-05,
794
+ "loss": 0.0263,
795
+ "step": 980
796
+ },
797
+ {
798
+ "epoch": 0.48292682926829267,
799
+ "grad_norm": 0.1029893159866333,
800
+ "learning_rate": 4.56637168141593e-05,
801
+ "loss": 0.0215,
802
+ "step": 990
803
+ },
804
+ {
805
+ "epoch": 0.4878048780487805,
806
+ "grad_norm": 0.09537643939256668,
807
+ "learning_rate": 4.5575221238938055e-05,
808
+ "loss": 0.024,
809
+ "step": 1000
810
+ },
811
+ {
812
+ "epoch": 0.4878048780487805,
813
+ "eval_f1": 0.6293077812654881,
814
+ "eval_loss": 0.023848505690693855,
815
+ "eval_precision": 0.5665417335950979,
816
+ "eval_recall": 0.7077140309155767,
817
+ "eval_runtime": 131.8557,
818
+ "eval_samples_per_second": 62.189,
819
+ "eval_steps_per_second": 0.978,
820
+ "step": 1000
821
+ },
822
+ {
823
+ "epoch": 0.4926829268292683,
824
+ "grad_norm": 0.11213461309671402,
825
+ "learning_rate": 4.548672566371682e-05,
826
+ "loss": 0.0258,
827
+ "step": 1010
828
+ },
829
+ {
830
+ "epoch": 0.4975609756097561,
831
+ "grad_norm": 0.09799027442932129,
832
+ "learning_rate": 4.5398230088495574e-05,
833
+ "loss": 0.0244,
834
+ "step": 1020
835
+ },
836
+ {
837
+ "epoch": 0.5024390243902439,
838
+ "grad_norm": 0.0755227655172348,
839
+ "learning_rate": 4.5309734513274336e-05,
840
+ "loss": 0.0269,
841
+ "step": 1030
842
+ },
843
+ {
844
+ "epoch": 0.5073170731707317,
845
+ "grad_norm": 0.10655403882265091,
846
+ "learning_rate": 4.52212389380531e-05,
847
+ "loss": 0.027,
848
+ "step": 1040
849
+ },
850
+ {
851
+ "epoch": 0.5121951219512195,
852
+ "grad_norm": 0.11206043511629105,
853
+ "learning_rate": 4.5132743362831855e-05,
854
+ "loss": 0.0268,
855
+ "step": 1050
856
+ },
857
+ {
858
+ "epoch": 0.5170731707317073,
859
+ "grad_norm": 0.08644779026508331,
860
+ "learning_rate": 4.5044247787610625e-05,
861
+ "loss": 0.0258,
862
+ "step": 1060
863
+ },
864
+ {
865
+ "epoch": 0.5219512195121951,
866
+ "grad_norm": 0.07745319604873657,
867
+ "learning_rate": 4.495575221238939e-05,
868
+ "loss": 0.0232,
869
+ "step": 1070
870
+ },
871
+ {
872
+ "epoch": 0.526829268292683,
873
+ "grad_norm": 0.10960444808006287,
874
+ "learning_rate": 4.4867256637168144e-05,
875
+ "loss": 0.0271,
876
+ "step": 1080
877
+ },
878
+ {
879
+ "epoch": 0.5317073170731708,
880
+ "grad_norm": 0.09823193401098251,
881
+ "learning_rate": 4.4778761061946906e-05,
882
+ "loss": 0.0279,
883
+ "step": 1090
884
+ },
885
+ {
886
+ "epoch": 0.5365853658536586,
887
+ "grad_norm": 0.10037508606910706,
888
+ "learning_rate": 4.469026548672566e-05,
889
+ "loss": 0.0226,
890
+ "step": 1100
891
+ },
892
+ {
893
+ "epoch": 0.5365853658536586,
894
+ "eval_f1": 0.6288948069241012,
895
+ "eval_loss": 0.023997528478503227,
896
+ "eval_precision": 0.5695851423058369,
897
+ "eval_recall": 0.7019916765755053,
898
+ "eval_runtime": 131.8239,
899
+ "eval_samples_per_second": 62.204,
900
+ "eval_steps_per_second": 0.979,
901
+ "step": 1100
902
+ },
903
+ {
904
+ "epoch": 0.5414634146341464,
905
+ "grad_norm": 0.086235910654068,
906
+ "learning_rate": 4.4601769911504425e-05,
907
+ "loss": 0.0243,
908
+ "step": 1110
909
+ },
910
+ {
911
+ "epoch": 0.5463414634146342,
912
+ "grad_norm": 0.09304425120353699,
913
+ "learning_rate": 4.451327433628319e-05,
914
+ "loss": 0.0236,
915
+ "step": 1120
916
+ },
917
+ {
918
+ "epoch": 0.551219512195122,
919
+ "grad_norm": 0.09922584891319275,
920
+ "learning_rate": 4.4424778761061944e-05,
921
+ "loss": 0.0222,
922
+ "step": 1130
923
+ },
924
+ {
925
+ "epoch": 0.5560975609756098,
926
+ "grad_norm": 0.07448361814022064,
927
+ "learning_rate": 4.433628318584071e-05,
928
+ "loss": 0.0246,
929
+ "step": 1140
930
+ },
931
+ {
932
+ "epoch": 0.5609756097560976,
933
+ "grad_norm": 0.08881635963916779,
934
+ "learning_rate": 4.4247787610619477e-05,
935
+ "loss": 0.023,
936
+ "step": 1150
937
+ },
938
+ {
939
+ "epoch": 0.5658536585365853,
940
+ "grad_norm": 0.11040020734071732,
941
+ "learning_rate": 4.415929203539823e-05,
942
+ "loss": 0.0255,
943
+ "step": 1160
944
+ },
945
+ {
946
+ "epoch": 0.5707317073170731,
947
+ "grad_norm": 0.10073444992303848,
948
+ "learning_rate": 4.4070796460176995e-05,
949
+ "loss": 0.0234,
950
+ "step": 1170
951
+ },
952
+ {
953
+ "epoch": 0.5756097560975609,
954
+ "grad_norm": 0.1330658495426178,
955
+ "learning_rate": 4.398230088495575e-05,
956
+ "loss": 0.0253,
957
+ "step": 1180
958
+ },
959
+ {
960
+ "epoch": 0.5804878048780487,
961
+ "grad_norm": 0.11121776700019836,
962
+ "learning_rate": 4.3893805309734514e-05,
963
+ "loss": 0.0253,
964
+ "step": 1190
965
+ },
966
+ {
967
+ "epoch": 0.5853658536585366,
968
+ "grad_norm": 0.07611318677663803,
969
+ "learning_rate": 4.380530973451328e-05,
970
+ "loss": 0.021,
971
+ "step": 1200
972
+ },
973
+ {
974
+ "epoch": 0.5853658536585366,
975
+ "eval_f1": 0.6257888792931642,
976
+ "eval_loss": 0.024241872131824493,
977
+ "eval_precision": 0.5657657657657658,
978
+ "eval_recall": 0.7000594530321046,
979
+ "eval_runtime": 132.4797,
980
+ "eval_samples_per_second": 61.896,
981
+ "eval_steps_per_second": 0.974,
982
+ "step": 1200
983
+ },
984
+ {
985
+ "epoch": 0.5902439024390244,
986
+ "grad_norm": 0.06641782820224762,
987
+ "learning_rate": 4.371681415929203e-05,
988
+ "loss": 0.0252,
989
+ "step": 1210
990
+ },
991
+ {
992
+ "epoch": 0.5951219512195122,
993
+ "grad_norm": 0.10350623726844788,
994
+ "learning_rate": 4.3628318584070796e-05,
995
+ "loss": 0.0261,
996
+ "step": 1220
997
+ },
998
+ {
999
+ "epoch": 0.6,
1000
+ "grad_norm": 0.09171286225318909,
1001
+ "learning_rate": 4.353982300884956e-05,
1002
+ "loss": 0.0222,
1003
+ "step": 1230
1004
+ },
1005
+ {
1006
+ "epoch": 0.6048780487804878,
1007
+ "grad_norm": 0.09831242263317108,
1008
+ "learning_rate": 4.345132743362832e-05,
1009
+ "loss": 0.0196,
1010
+ "step": 1240
1011
+ },
1012
+ {
1013
+ "epoch": 0.6097560975609756,
1014
+ "grad_norm": 0.06655886769294739,
1015
+ "learning_rate": 4.3362831858407084e-05,
1016
+ "loss": 0.0214,
1017
+ "step": 1250
1018
+ },
1019
+ {
1020
+ "epoch": 0.6146341463414634,
1021
+ "grad_norm": 0.10220635682344437,
1022
+ "learning_rate": 4.327433628318584e-05,
1023
+ "loss": 0.0225,
1024
+ "step": 1260
1025
+ },
1026
+ {
1027
+ "epoch": 0.6195121951219512,
1028
+ "grad_norm": 0.09263930469751358,
1029
+ "learning_rate": 4.31858407079646e-05,
1030
+ "loss": 0.0237,
1031
+ "step": 1270
1032
+ },
1033
+ {
1034
+ "epoch": 0.624390243902439,
1035
+ "grad_norm": 0.12217256426811218,
1036
+ "learning_rate": 4.3097345132743366e-05,
1037
+ "loss": 0.0238,
1038
+ "step": 1280
1039
+ },
1040
+ {
1041
+ "epoch": 0.6292682926829268,
1042
+ "grad_norm": 0.08786381781101227,
1043
+ "learning_rate": 4.300884955752212e-05,
1044
+ "loss": 0.0217,
1045
+ "step": 1290
1046
+ },
1047
+ {
1048
+ "epoch": 0.6341463414634146,
1049
+ "grad_norm": 0.06754878163337708,
1050
+ "learning_rate": 4.2920353982300885e-05,
1051
+ "loss": 0.0209,
1052
+ "step": 1300
1053
+ },
1054
+ {
1055
+ "epoch": 0.6341463414634146,
1056
+ "eval_f1": 0.635122838944495,
1057
+ "eval_loss": 0.022672206163406372,
1058
+ "eval_precision": 0.5643335643335643,
1059
+ "eval_recall": 0.726218787158145,
1060
+ "eval_runtime": 132.3145,
1061
+ "eval_samples_per_second": 61.974,
1062
+ "eval_steps_per_second": 0.975,
1063
+ "step": 1300
1064
+ },
1065
+ {
1066
+ "epoch": 0.6390243902439025,
1067
+ "grad_norm": 0.07975753396749496,
1068
+ "learning_rate": 4.283185840707965e-05,
1069
+ "loss": 0.0208,
1070
+ "step": 1310
1071
+ },
1072
+ {
1073
+ "epoch": 0.6439024390243903,
1074
+ "grad_norm": 0.10725169628858566,
1075
+ "learning_rate": 4.274336283185841e-05,
1076
+ "loss": 0.0243,
1077
+ "step": 1320
1078
+ },
1079
+ {
1080
+ "epoch": 0.6487804878048781,
1081
+ "grad_norm": 0.07709804177284241,
1082
+ "learning_rate": 4.265486725663717e-05,
1083
+ "loss": 0.0241,
1084
+ "step": 1330
1085
+ },
1086
+ {
1087
+ "epoch": 0.6536585365853659,
1088
+ "grad_norm": 0.07307865470647812,
1089
+ "learning_rate": 4.256637168141593e-05,
1090
+ "loss": 0.0221,
1091
+ "step": 1340
1092
+ },
1093
+ {
1094
+ "epoch": 0.6585365853658537,
1095
+ "grad_norm": 0.07658623158931732,
1096
+ "learning_rate": 4.247787610619469e-05,
1097
+ "loss": 0.0216,
1098
+ "step": 1350
1099
+ },
1100
+ {
1101
+ "epoch": 0.6634146341463415,
1102
+ "grad_norm": 0.12642726302146912,
1103
+ "learning_rate": 4.2389380530973455e-05,
1104
+ "loss": 0.0247,
1105
+ "step": 1360
1106
+ },
1107
+ {
1108
+ "epoch": 0.6682926829268293,
1109
+ "grad_norm": 0.09556315094232559,
1110
+ "learning_rate": 4.230088495575221e-05,
1111
+ "loss": 0.0217,
1112
+ "step": 1370
1113
+ },
1114
+ {
1115
+ "epoch": 0.6731707317073171,
1116
+ "grad_norm": 0.06885875016450882,
1117
+ "learning_rate": 4.2212389380530974e-05,
1118
+ "loss": 0.02,
1119
+ "step": 1380
1120
+ },
1121
+ {
1122
+ "epoch": 0.6780487804878049,
1123
+ "grad_norm": 0.0719994455575943,
1124
+ "learning_rate": 4.2123893805309737e-05,
1125
+ "loss": 0.0214,
1126
+ "step": 1390
1127
+ },
1128
+ {
1129
+ "epoch": 0.6829268292682927,
1130
+ "grad_norm": 0.06444905698299408,
1131
+ "learning_rate": 4.20353982300885e-05,
1132
+ "loss": 0.0207,
1133
+ "step": 1400
1134
+ },
1135
+ {
1136
+ "epoch": 0.6829268292682927,
1137
+ "eval_f1": 0.6471813478465364,
1138
+ "eval_loss": 0.02256700210273266,
1139
+ "eval_precision": 0.590557939914163,
1140
+ "eval_recall": 0.7158145065398336,
1141
+ "eval_runtime": 132.1018,
1142
+ "eval_samples_per_second": 62.073,
1143
+ "eval_steps_per_second": 0.977,
1144
+ "step": 1400
1145
+ },
1146
+ {
1147
+ "epoch": 0.6878048780487804,
1148
+ "grad_norm": 0.07349838316440582,
1149
+ "learning_rate": 4.194690265486726e-05,
1150
+ "loss": 0.0225,
1151
+ "step": 1410
1152
+ },
1153
+ {
1154
+ "epoch": 0.6926829268292682,
1155
+ "grad_norm": 0.07198076695203781,
1156
+ "learning_rate": 4.185840707964602e-05,
1157
+ "loss": 0.021,
1158
+ "step": 1420
1159
+ },
1160
+ {
1161
+ "epoch": 0.697560975609756,
1162
+ "grad_norm": 0.0936046615242958,
1163
+ "learning_rate": 4.176991150442478e-05,
1164
+ "loss": 0.0252,
1165
+ "step": 1430
1166
+ },
1167
+ {
1168
+ "epoch": 0.7024390243902439,
1169
+ "grad_norm": 0.08654190599918365,
1170
+ "learning_rate": 4.1681415929203544e-05,
1171
+ "loss": 0.0218,
1172
+ "step": 1440
1173
+ },
1174
+ {
1175
+ "epoch": 0.7073170731707317,
1176
+ "grad_norm": 0.08444487303495407,
1177
+ "learning_rate": 4.15929203539823e-05,
1178
+ "loss": 0.0267,
1179
+ "step": 1450
1180
+ },
1181
+ {
1182
+ "epoch": 0.7121951219512195,
1183
+ "grad_norm": 0.12378791719675064,
1184
+ "learning_rate": 4.150442477876106e-05,
1185
+ "loss": 0.0233,
1186
+ "step": 1460
1187
+ },
1188
+ {
1189
+ "epoch": 0.7170731707317073,
1190
+ "grad_norm": 0.09955397993326187,
1191
+ "learning_rate": 4.1415929203539825e-05,
1192
+ "loss": 0.0253,
1193
+ "step": 1470
1194
+ },
1195
+ {
1196
+ "epoch": 0.7219512195121951,
1197
+ "grad_norm": 0.08549737185239792,
1198
+ "learning_rate": 4.132743362831858e-05,
1199
+ "loss": 0.0252,
1200
+ "step": 1480
1201
+ },
1202
+ {
1203
+ "epoch": 0.7268292682926829,
1204
+ "grad_norm": 0.06966210901737213,
1205
+ "learning_rate": 4.123893805309735e-05,
1206
+ "loss": 0.0242,
1207
+ "step": 1490
1208
+ },
1209
+ {
1210
+ "epoch": 0.7317073170731707,
1211
+ "grad_norm": 0.08285216242074966,
1212
+ "learning_rate": 4.115044247787611e-05,
1213
+ "loss": 0.0219,
1214
+ "step": 1500
1215
+ },
1216
+ {
1217
+ "epoch": 0.7317073170731707,
1218
+ "eval_f1": 0.6405168738316225,
1219
+ "eval_loss": 0.022300876677036285,
1220
+ "eval_precision": 0.5732315820369827,
1221
+ "eval_recall": 0.7256985731272295,
1222
+ "eval_runtime": 132.0099,
1223
+ "eval_samples_per_second": 62.117,
1224
+ "eval_steps_per_second": 0.977,
1225
+ "step": 1500
1226
+ },
1227
+ {
1228
+ "epoch": 0.7365853658536585,
1229
+ "grad_norm": 0.10072668641805649,
1230
+ "learning_rate": 4.106194690265487e-05,
1231
+ "loss": 0.0227,
1232
+ "step": 1510
1233
+ },
1234
+ {
1235
+ "epoch": 0.7414634146341463,
1236
+ "grad_norm": 0.10805150866508484,
1237
+ "learning_rate": 4.097345132743363e-05,
1238
+ "loss": 0.0238,
1239
+ "step": 1520
1240
+ },
1241
+ {
1242
+ "epoch": 0.7463414634146341,
1243
+ "grad_norm": 0.06195740029215813,
1244
+ "learning_rate": 4.088495575221239e-05,
1245
+ "loss": 0.0216,
1246
+ "step": 1530
1247
+ },
1248
+ {
1249
+ "epoch": 0.751219512195122,
1250
+ "grad_norm": 0.06914755702018738,
1251
+ "learning_rate": 4.079646017699115e-05,
1252
+ "loss": 0.0215,
1253
+ "step": 1540
1254
+ },
1255
+ {
1256
+ "epoch": 0.7560975609756098,
1257
+ "grad_norm": 0.09708551317453384,
1258
+ "learning_rate": 4.0707964601769914e-05,
1259
+ "loss": 0.0205,
1260
+ "step": 1550
1261
+ },
1262
+ {
1263
+ "epoch": 0.7609756097560976,
1264
+ "grad_norm": 0.07369375228881836,
1265
+ "learning_rate": 4.061946902654867e-05,
1266
+ "loss": 0.0238,
1267
+ "step": 1560
1268
+ },
1269
+ {
1270
+ "epoch": 0.7658536585365854,
1271
+ "grad_norm": 0.07211218029260635,
1272
+ "learning_rate": 4.053097345132743e-05,
1273
+ "loss": 0.02,
1274
+ "step": 1570
1275
+ },
1276
+ {
1277
+ "epoch": 0.7707317073170732,
1278
+ "grad_norm": 0.09262284636497498,
1279
+ "learning_rate": 4.0442477876106196e-05,
1280
+ "loss": 0.0227,
1281
+ "step": 1580
1282
+ },
1283
+ {
1284
+ "epoch": 0.775609756097561,
1285
+ "grad_norm": 0.09008630365133286,
1286
+ "learning_rate": 4.035398230088496e-05,
1287
+ "loss": 0.0228,
1288
+ "step": 1590
1289
+ },
1290
+ {
1291
+ "epoch": 0.7804878048780488,
1292
+ "grad_norm": 0.1014399379491806,
1293
+ "learning_rate": 4.026548672566372e-05,
1294
+ "loss": 0.0215,
1295
+ "step": 1600
1296
+ },
1297
+ {
1298
+ "epoch": 0.7804878048780488,
1299
+ "eval_f1": 0.637167016339246,
1300
+ "eval_loss": 0.02189124934375286,
1301
+ "eval_precision": 0.556715920453283,
1302
+ "eval_recall": 0.7447978596908442,
1303
+ "eval_runtime": 132.2243,
1304
+ "eval_samples_per_second": 62.016,
1305
+ "eval_steps_per_second": 0.976,
1306
+ "step": 1600
1307
+ },
1308
+ {
1309
+ "epoch": 0.7853658536585366,
1310
+ "grad_norm": 0.1170215755701065,
1311
+ "learning_rate": 4.017699115044248e-05,
1312
+ "loss": 0.025,
1313
+ "step": 1610
1314
+ },
1315
+ {
1316
+ "epoch": 0.7902439024390244,
1317
+ "grad_norm": 0.07999496906995773,
1318
+ "learning_rate": 4.008849557522124e-05,
1319
+ "loss": 0.0237,
1320
+ "step": 1620
1321
+ },
1322
+ {
1323
+ "epoch": 0.7951219512195122,
1324
+ "grad_norm": 0.08684638142585754,
1325
+ "learning_rate": 4e-05,
1326
+ "loss": 0.0261,
1327
+ "step": 1630
1328
+ },
1329
+ {
1330
+ "epoch": 0.8,
1331
+ "grad_norm": 0.06762222945690155,
1332
+ "learning_rate": 3.991150442477876e-05,
1333
+ "loss": 0.0232,
1334
+ "step": 1640
1335
+ },
1336
+ {
1337
+ "epoch": 0.8048780487804879,
1338
+ "grad_norm": 0.08543413132429123,
1339
+ "learning_rate": 3.982300884955752e-05,
1340
+ "loss": 0.0225,
1341
+ "step": 1650
1342
+ },
1343
+ {
1344
+ "epoch": 0.8097560975609757,
1345
+ "grad_norm": 0.09491296857595444,
1346
+ "learning_rate": 3.9734513274336285e-05,
1347
+ "loss": 0.0221,
1348
+ "step": 1660
1349
+ },
1350
+ {
1351
+ "epoch": 0.8146341463414634,
1352
+ "grad_norm": 0.06884954869747162,
1353
+ "learning_rate": 3.964601769911505e-05,
1354
+ "loss": 0.0221,
1355
+ "step": 1670
1356
+ },
1357
+ {
1358
+ "epoch": 0.8195121951219512,
1359
+ "grad_norm": 0.10209941118955612,
1360
+ "learning_rate": 3.955752212389381e-05,
1361
+ "loss": 0.025,
1362
+ "step": 1680
1363
+ },
1364
+ {
1365
+ "epoch": 0.824390243902439,
1366
+ "grad_norm": 0.06364341080188751,
1367
+ "learning_rate": 3.9469026548672567e-05,
1368
+ "loss": 0.0197,
1369
+ "step": 1690
1370
+ },
1371
+ {
1372
+ "epoch": 0.8292682926829268,
1373
+ "grad_norm": 0.10673107206821442,
1374
+ "learning_rate": 3.938053097345133e-05,
1375
+ "loss": 0.0199,
1376
+ "step": 1700
1377
+ },
1378
+ {
1379
+ "epoch": 0.8292682926829268,
1380
+ "eval_f1": 0.6390703962403836,
1381
+ "eval_loss": 0.021866334602236748,
1382
+ "eval_precision": 0.5636817897904719,
1383
+ "eval_recall": 0.7377378121284186,
1384
+ "eval_runtime": 132.1953,
1385
+ "eval_samples_per_second": 62.029,
1386
+ "eval_steps_per_second": 0.976,
1387
+ "step": 1700
1388
+ },
1389
+ {
1390
+ "epoch": 0.8341463414634146,
1391
+ "grad_norm": 0.08173991739749908,
1392
+ "learning_rate": 3.929203539823009e-05,
1393
+ "loss": 0.0217,
1394
+ "step": 1710
1395
+ },
1396
+ {
1397
+ "epoch": 0.8390243902439024,
1398
+ "grad_norm": 0.09340299665927887,
1399
+ "learning_rate": 3.920353982300885e-05,
1400
+ "loss": 0.0218,
1401
+ "step": 1720
1402
+ },
1403
+ {
1404
+ "epoch": 0.8439024390243902,
1405
+ "grad_norm": 0.08972273021936417,
1406
+ "learning_rate": 3.911504424778761e-05,
1407
+ "loss": 0.0178,
1408
+ "step": 1730
1409
+ },
1410
+ {
1411
+ "epoch": 0.848780487804878,
1412
+ "grad_norm": 0.11011021584272385,
1413
+ "learning_rate": 3.9026548672566374e-05,
1414
+ "loss": 0.0282,
1415
+ "step": 1740
1416
+ },
1417
+ {
1418
+ "epoch": 0.8536585365853658,
1419
+ "grad_norm": 0.1096154972910881,
1420
+ "learning_rate": 3.893805309734514e-05,
1421
+ "loss": 0.0244,
1422
+ "step": 1750
1423
+ },
1424
+ {
1425
+ "epoch": 0.8585365853658536,
1426
+ "grad_norm": 0.08531954139471054,
1427
+ "learning_rate": 3.88495575221239e-05,
1428
+ "loss": 0.0217,
1429
+ "step": 1760
1430
+ },
1431
+ {
1432
+ "epoch": 0.8634146341463415,
1433
+ "grad_norm": 0.1026742234826088,
1434
+ "learning_rate": 3.8761061946902655e-05,
1435
+ "loss": 0.0199,
1436
+ "step": 1770
1437
+ },
1438
+ {
1439
+ "epoch": 0.8682926829268293,
1440
+ "grad_norm": 0.06563183665275574,
1441
+ "learning_rate": 3.867256637168142e-05,
1442
+ "loss": 0.0185,
1443
+ "step": 1780
1444
+ },
1445
+ {
1446
+ "epoch": 0.8731707317073171,
1447
+ "grad_norm": 0.07869338989257812,
1448
+ "learning_rate": 3.858407079646018e-05,
1449
+ "loss": 0.0235,
1450
+ "step": 1790
1451
+ },
1452
+ {
1453
+ "epoch": 0.8780487804878049,
1454
+ "grad_norm": 0.08610737323760986,
1455
+ "learning_rate": 3.849557522123894e-05,
1456
+ "loss": 0.0193,
1457
+ "step": 1800
1458
+ },
1459
+ {
1460
+ "epoch": 0.8780487804878049,
1461
+ "eval_f1": 0.6328935570641732,
1462
+ "eval_loss": 0.022152401506900787,
1463
+ "eval_precision": 0.5560432140445645,
1464
+ "eval_recall": 0.7343935790725327,
1465
+ "eval_runtime": 132.0266,
1466
+ "eval_samples_per_second": 62.109,
1467
+ "eval_steps_per_second": 0.977,
1468
+ "step": 1800
1469
+ },
1470
+ {
1471
+ "epoch": 0.8829268292682927,
1472
+ "grad_norm": 0.09815912693738937,
1473
+ "learning_rate": 3.84070796460177e-05,
1474
+ "loss": 0.0179,
1475
+ "step": 1810
1476
+ },
1477
+ {
1478
+ "epoch": 0.8878048780487805,
1479
+ "grad_norm": 0.06768873333930969,
1480
+ "learning_rate": 3.831858407079646e-05,
1481
+ "loss": 0.0219,
1482
+ "step": 1820
1483
+ },
1484
+ {
1485
+ "epoch": 0.8926829268292683,
1486
+ "grad_norm": 0.09629742056131363,
1487
+ "learning_rate": 3.8230088495575226e-05,
1488
+ "loss": 0.0212,
1489
+ "step": 1830
1490
+ },
1491
+ {
1492
+ "epoch": 0.8975609756097561,
1493
+ "grad_norm": 0.0719202533364296,
1494
+ "learning_rate": 3.814159292035399e-05,
1495
+ "loss": 0.0221,
1496
+ "step": 1840
1497
+ },
1498
+ {
1499
+ "epoch": 0.9024390243902439,
1500
+ "grad_norm": 0.09463170170783997,
1501
+ "learning_rate": 3.8053097345132744e-05,
1502
+ "loss": 0.0233,
1503
+ "step": 1850
1504
+ },
1505
+ {
1506
+ "epoch": 0.9073170731707317,
1507
+ "grad_norm": 0.08261518180370331,
1508
+ "learning_rate": 3.796460176991151e-05,
1509
+ "loss": 0.0205,
1510
+ "step": 1860
1511
+ },
1512
+ {
1513
+ "epoch": 0.9121951219512195,
1514
+ "grad_norm": 0.09188443422317505,
1515
+ "learning_rate": 3.787610619469027e-05,
1516
+ "loss": 0.0186,
1517
+ "step": 1870
1518
+ },
1519
+ {
1520
+ "epoch": 0.9170731707317074,
1521
+ "grad_norm": 0.0914570763707161,
1522
+ "learning_rate": 3.7787610619469026e-05,
1523
+ "loss": 0.0207,
1524
+ "step": 1880
1525
+ },
1526
+ {
1527
+ "epoch": 0.9219512195121952,
1528
+ "grad_norm": 0.0840875431895256,
1529
+ "learning_rate": 3.769911504424779e-05,
1530
+ "loss": 0.0225,
1531
+ "step": 1890
1532
+ },
1533
+ {
1534
+ "epoch": 0.926829268292683,
1535
+ "grad_norm": 0.0681457445025444,
1536
+ "learning_rate": 3.7610619469026545e-05,
1537
+ "loss": 0.0216,
1538
+ "step": 1900
1539
+ },
1540
+ {
1541
+ "epoch": 0.926829268292683,
1542
+ "eval_f1": 0.6448713774014979,
1543
+ "eval_loss": 0.021633492782711983,
1544
+ "eval_precision": 0.5738959081952011,
1545
+ "eval_recall": 0.7358799048751486,
1546
+ "eval_runtime": 132.1734,
1547
+ "eval_samples_per_second": 62.04,
1548
+ "eval_steps_per_second": 0.976,
1549
+ "step": 1900
1550
+ },
1551
+ {
1552
+ "epoch": 0.9317073170731708,
1553
+ "grad_norm": 0.07665237784385681,
1554
+ "learning_rate": 3.752212389380531e-05,
1555
+ "loss": 0.0231,
1556
+ "step": 1910
1557
+ },
1558
+ {
1559
+ "epoch": 0.9365853658536586,
1560
+ "grad_norm": 0.12333638221025467,
1561
+ "learning_rate": 3.743362831858408e-05,
1562
+ "loss": 0.0244,
1563
+ "step": 1920
1564
+ },
1565
+ {
1566
+ "epoch": 0.9414634146341463,
1567
+ "grad_norm": 0.05435947701334953,
1568
+ "learning_rate": 3.734513274336283e-05,
1569
+ "loss": 0.0209,
1570
+ "step": 1930
1571
+ },
1572
+ {
1573
+ "epoch": 0.9463414634146341,
1574
+ "grad_norm": 0.09085798263549805,
1575
+ "learning_rate": 3.7256637168141596e-05,
1576
+ "loss": 0.0212,
1577
+ "step": 1940
1578
+ },
1579
+ {
1580
+ "epoch": 0.9512195121951219,
1581
+ "grad_norm": 0.11483143270015717,
1582
+ "learning_rate": 3.716814159292036e-05,
1583
+ "loss": 0.0213,
1584
+ "step": 1950
1585
+ },
1586
+ {
1587
+ "epoch": 0.9560975609756097,
1588
+ "grad_norm": 0.06926431506872177,
1589
+ "learning_rate": 3.7079646017699115e-05,
1590
+ "loss": 0.0203,
1591
+ "step": 1960
1592
+ },
1593
+ {
1594
+ "epoch": 0.9609756097560975,
1595
+ "grad_norm": 0.1061626598238945,
1596
+ "learning_rate": 3.699115044247788e-05,
1597
+ "loss": 0.0227,
1598
+ "step": 1970
1599
+ },
1600
+ {
1601
+ "epoch": 0.9658536585365853,
1602
+ "grad_norm": 0.09048457443714142,
1603
+ "learning_rate": 3.6902654867256634e-05,
1604
+ "loss": 0.0229,
1605
+ "step": 1980
1606
+ },
1607
+ {
1608
+ "epoch": 0.9707317073170731,
1609
+ "grad_norm": 0.09511193633079529,
1610
+ "learning_rate": 3.68141592920354e-05,
1611
+ "loss": 0.0216,
1612
+ "step": 1990
1613
+ },
1614
+ {
1615
+ "epoch": 0.975609756097561,
1616
+ "grad_norm": 0.091468945145607,
1617
+ "learning_rate": 3.672566371681416e-05,
1618
+ "loss": 0.0168,
1619
+ "step": 2000
1620
+ },
1621
+ {
1622
+ "epoch": 0.975609756097561,
1623
+ "eval_f1": 0.6518885040409038,
1624
+ "eval_loss": 0.021417897194623947,
1625
+ "eval_precision": 0.5860964470016016,
1626
+ "eval_recall": 0.7343192627824019,
1627
+ "eval_runtime": 131.7709,
1628
+ "eval_samples_per_second": 62.229,
1629
+ "eval_steps_per_second": 0.979,
1630
+ "step": 2000
1631
+ },
1632
+ {
1633
+ "epoch": 0.9804878048780488,
1634
+ "grad_norm": 0.09001079201698303,
1635
+ "learning_rate": 3.663716814159292e-05,
1636
+ "loss": 0.0192,
1637
+ "step": 2010
1638
+ },
1639
+ {
1640
+ "epoch": 0.9853658536585366,
1641
+ "grad_norm": 0.10538368672132492,
1642
+ "learning_rate": 3.6548672566371685e-05,
1643
+ "loss": 0.0228,
1644
+ "step": 2020
1645
+ },
1646
+ {
1647
+ "epoch": 0.9902439024390244,
1648
+ "grad_norm": 0.08615951985120773,
1649
+ "learning_rate": 3.646017699115045e-05,
1650
+ "loss": 0.0238,
1651
+ "step": 2030
1652
+ },
1653
+ {
1654
+ "epoch": 0.9951219512195122,
1655
+ "grad_norm": 0.0890466719865799,
1656
+ "learning_rate": 3.6371681415929204e-05,
1657
+ "loss": 0.0218,
1658
+ "step": 2040
1659
+ },
1660
+ {
1661
+ "epoch": 1.0,
1662
+ "grad_norm": 0.08089441806077957,
1663
+ "learning_rate": 3.628318584070797e-05,
1664
+ "loss": 0.0204,
1665
+ "step": 2050
1666
+ },
1667
+ {
1668
+ "epoch": 1.0048780487804878,
1669
+ "grad_norm": 0.09391944855451584,
1670
+ "learning_rate": 3.619469026548672e-05,
1671
+ "loss": 0.0213,
1672
+ "step": 2060
1673
+ },
1674
+ {
1675
+ "epoch": 1.0097560975609756,
1676
+ "grad_norm": 0.059187982231378555,
1677
+ "learning_rate": 3.6106194690265486e-05,
1678
+ "loss": 0.0197,
1679
+ "step": 2070
1680
+ },
1681
+ {
1682
+ "epoch": 1.0146341463414634,
1683
+ "grad_norm": 0.07538473606109619,
1684
+ "learning_rate": 3.601769911504425e-05,
1685
+ "loss": 0.0227,
1686
+ "step": 2080
1687
+ },
1688
+ {
1689
+ "epoch": 1.0195121951219512,
1690
+ "grad_norm": 0.0840989351272583,
1691
+ "learning_rate": 3.592920353982301e-05,
1692
+ "loss": 0.0205,
1693
+ "step": 2090
1694
+ },
1695
+ {
1696
+ "epoch": 1.024390243902439,
1697
+ "grad_norm": 0.11819695681333542,
1698
+ "learning_rate": 3.5840707964601774e-05,
1699
+ "loss": 0.0214,
1700
+ "step": 2100
1701
+ },
1702
+ {
1703
+ "epoch": 1.024390243902439,
1704
+ "eval_f1": 0.6467788814717813,
1705
+ "eval_loss": 0.020904576405882835,
1706
+ "eval_precision": 0.5671315745252898,
1707
+ "eval_recall": 0.7524524375743162,
1708
+ "eval_runtime": 132.2697,
1709
+ "eval_samples_per_second": 61.995,
1710
+ "eval_steps_per_second": 0.975,
1711
+ "step": 2100
1712
+ },
1713
+ {
1714
+ "epoch": 1.0292682926829269,
1715
+ "grad_norm": 0.121103934943676,
1716
+ "learning_rate": 3.575221238938054e-05,
1717
+ "loss": 0.0196,
1718
+ "step": 2110
1719
+ },
1720
+ {
1721
+ "epoch": 1.0341463414634147,
1722
+ "grad_norm": 0.0714835524559021,
1723
+ "learning_rate": 3.566371681415929e-05,
1724
+ "loss": 0.0193,
1725
+ "step": 2120
1726
+ },
1727
+ {
1728
+ "epoch": 1.0390243902439025,
1729
+ "grad_norm": 0.07826493680477142,
1730
+ "learning_rate": 3.5575221238938056e-05,
1731
+ "loss": 0.0202,
1732
+ "step": 2130
1733
+ },
1734
+ {
1735
+ "epoch": 1.0439024390243903,
1736
+ "grad_norm": 0.0855259820818901,
1737
+ "learning_rate": 3.548672566371681e-05,
1738
+ "loss": 0.0217,
1739
+ "step": 2140
1740
+ },
1741
+ {
1742
+ "epoch": 1.048780487804878,
1743
+ "grad_norm": 0.07726403325796127,
1744
+ "learning_rate": 3.5398230088495574e-05,
1745
+ "loss": 0.0213,
1746
+ "step": 2150
1747
+ },
1748
+ {
1749
+ "epoch": 1.053658536585366,
1750
+ "grad_norm": 0.09908368438482285,
1751
+ "learning_rate": 3.530973451327434e-05,
1752
+ "loss": 0.0186,
1753
+ "step": 2160
1754
+ },
1755
+ {
1756
+ "epoch": 1.0585365853658537,
1757
+ "grad_norm": 0.08559077978134155,
1758
+ "learning_rate": 3.52212389380531e-05,
1759
+ "loss": 0.0205,
1760
+ "step": 2170
1761
+ },
1762
+ {
1763
+ "epoch": 1.0634146341463415,
1764
+ "grad_norm": 0.09128253161907196,
1765
+ "learning_rate": 3.513274336283186e-05,
1766
+ "loss": 0.0247,
1767
+ "step": 2180
1768
+ },
1769
+ {
1770
+ "epoch": 1.0682926829268293,
1771
+ "grad_norm": 0.08086485415697098,
1772
+ "learning_rate": 3.5044247787610626e-05,
1773
+ "loss": 0.0209,
1774
+ "step": 2190
1775
+ },
1776
+ {
1777
+ "epoch": 1.0731707317073171,
1778
+ "grad_norm": 0.08016868680715561,
1779
+ "learning_rate": 3.495575221238938e-05,
1780
+ "loss": 0.0205,
1781
+ "step": 2200
1782
+ },
1783
+ {
1784
+ "epoch": 1.0731707317073171,
1785
+ "eval_f1": 0.633068352720911,
1786
+ "eval_loss": 0.020948156714439392,
1787
+ "eval_precision": 0.5392418300653595,
1788
+ "eval_recall": 0.766423900118906,
1789
+ "eval_runtime": 132.144,
1790
+ "eval_samples_per_second": 62.054,
1791
+ "eval_steps_per_second": 0.976,
1792
+ "step": 2200
1793
+ },
1794
+ {
1795
+ "epoch": 1.078048780487805,
1796
+ "grad_norm": 0.08863110840320587,
1797
+ "learning_rate": 3.4867256637168145e-05,
1798
+ "loss": 0.021,
1799
+ "step": 2210
1800
+ },
1801
+ {
1802
+ "epoch": 1.0829268292682928,
1803
+ "grad_norm": 0.07697419077157974,
1804
+ "learning_rate": 3.47787610619469e-05,
1805
+ "loss": 0.0222,
1806
+ "step": 2220
1807
+ },
1808
+ {
1809
+ "epoch": 1.0878048780487806,
1810
+ "grad_norm": 0.07876092195510864,
1811
+ "learning_rate": 3.469026548672566e-05,
1812
+ "loss": 0.0176,
1813
+ "step": 2230
1814
+ },
1815
+ {
1816
+ "epoch": 1.0926829268292684,
1817
+ "grad_norm": 0.08885340392589569,
1818
+ "learning_rate": 3.4601769911504426e-05,
1819
+ "loss": 0.0183,
1820
+ "step": 2240
1821
+ },
1822
+ {
1823
+ "epoch": 1.0975609756097562,
1824
+ "grad_norm": 0.08264743536710739,
1825
+ "learning_rate": 3.451327433628319e-05,
1826
+ "loss": 0.0212,
1827
+ "step": 2250
1828
+ },
1829
+ {
1830
+ "epoch": 1.102439024390244,
1831
+ "grad_norm": 0.06534498184919357,
1832
+ "learning_rate": 3.442477876106195e-05,
1833
+ "loss": 0.018,
1834
+ "step": 2260
1835
+ },
1836
+ {
1837
+ "epoch": 1.1073170731707318,
1838
+ "grad_norm": 0.08914489299058914,
1839
+ "learning_rate": 3.4336283185840715e-05,
1840
+ "loss": 0.0201,
1841
+ "step": 2270
1842
+ },
1843
+ {
1844
+ "epoch": 1.1121951219512196,
1845
+ "grad_norm": 0.07213272899389267,
1846
+ "learning_rate": 3.424778761061947e-05,
1847
+ "loss": 0.0145,
1848
+ "step": 2280
1849
+ },
1850
+ {
1851
+ "epoch": 1.1170731707317074,
1852
+ "grad_norm": 0.07594022899866104,
1853
+ "learning_rate": 3.4159292035398233e-05,
1854
+ "loss": 0.0201,
1855
+ "step": 2290
1856
+ },
1857
+ {
1858
+ "epoch": 1.1219512195121952,
1859
+ "grad_norm": 0.06026766449213028,
1860
+ "learning_rate": 3.407079646017699e-05,
1861
+ "loss": 0.019,
1862
+ "step": 2300
1863
+ },
1864
+ {
1865
+ "epoch": 1.1219512195121952,
1866
+ "eval_f1": 0.6488888888888888,
1867
+ "eval_loss": 0.020637808367609978,
1868
+ "eval_precision": 0.569448341657781,
1869
+ "eval_recall": 0.7540873959571938,
1870
+ "eval_runtime": 132.2303,
1871
+ "eval_samples_per_second": 62.013,
1872
+ "eval_steps_per_second": 0.976,
1873
+ "step": 2300
1874
+ },
1875
+ {
1876
+ "epoch": 1.126829268292683,
1877
+ "grad_norm": 0.08499179780483246,
1878
+ "learning_rate": 3.398230088495575e-05,
1879
+ "loss": 0.0197,
1880
+ "step": 2310
1881
+ },
1882
+ {
1883
+ "epoch": 1.1317073170731708,
1884
+ "grad_norm": 0.09881128370761871,
1885
+ "learning_rate": 3.3893805309734515e-05,
1886
+ "loss": 0.0194,
1887
+ "step": 2320
1888
+ },
1889
+ {
1890
+ "epoch": 1.1365853658536587,
1891
+ "grad_norm": 0.06719642132520676,
1892
+ "learning_rate": 3.380530973451327e-05,
1893
+ "loss": 0.0202,
1894
+ "step": 2330
1895
+ },
1896
+ {
1897
+ "epoch": 1.1414634146341462,
1898
+ "grad_norm": 0.10720915347337723,
1899
+ "learning_rate": 3.3716814159292034e-05,
1900
+ "loss": 0.0165,
1901
+ "step": 2340
1902
+ },
1903
+ {
1904
+ "epoch": 1.146341463414634,
1905
+ "grad_norm": 0.06894739717245102,
1906
+ "learning_rate": 3.3628318584070804e-05,
1907
+ "loss": 0.0164,
1908
+ "step": 2350
1909
+ },
1910
+ {
1911
+ "epoch": 1.1512195121951219,
1912
+ "grad_norm": 0.0639248788356781,
1913
+ "learning_rate": 3.353982300884956e-05,
1914
+ "loss": 0.0217,
1915
+ "step": 2360
1916
+ },
1917
+ {
1918
+ "epoch": 1.1560975609756097,
1919
+ "grad_norm": 0.044793836772441864,
1920
+ "learning_rate": 3.345132743362832e-05,
1921
+ "loss": 0.0197,
1922
+ "step": 2370
1923
+ },
1924
+ {
1925
+ "epoch": 1.1609756097560975,
1926
+ "grad_norm": 0.0624634325504303,
1927
+ "learning_rate": 3.336283185840708e-05,
1928
+ "loss": 0.0202,
1929
+ "step": 2380
1930
+ },
1931
+ {
1932
+ "epoch": 1.1658536585365853,
1933
+ "grad_norm": 0.12909162044525146,
1934
+ "learning_rate": 3.327433628318584e-05,
1935
+ "loss": 0.0195,
1936
+ "step": 2390
1937
+ },
1938
+ {
1939
+ "epoch": 1.170731707317073,
1940
+ "grad_norm": 0.0766359269618988,
1941
+ "learning_rate": 3.3185840707964604e-05,
1942
+ "loss": 0.0208,
1943
+ "step": 2400
1944
+ },
1945
+ {
1946
+ "epoch": 1.170731707317073,
1947
+ "eval_f1": 0.6382413782457007,
1948
+ "eval_loss": 0.02069213055074215,
1949
+ "eval_precision": 0.5459251043152168,
1950
+ "eval_recall": 0.7681331747919143,
1951
+ "eval_runtime": 131.9061,
1952
+ "eval_samples_per_second": 62.165,
1953
+ "eval_steps_per_second": 0.978,
1954
+ "step": 2400
1955
+ },
1956
+ {
1957
+ "epoch": 1.175609756097561,
1958
+ "grad_norm": 0.1002466231584549,
1959
+ "learning_rate": 3.309734513274336e-05,
1960
+ "loss": 0.0198,
1961
+ "step": 2410
1962
+ },
1963
+ {
1964
+ "epoch": 1.1804878048780487,
1965
+ "grad_norm": 0.09123210608959198,
1966
+ "learning_rate": 3.300884955752212e-05,
1967
+ "loss": 0.0167,
1968
+ "step": 2420
1969
+ },
1970
+ {
1971
+ "epoch": 1.1853658536585365,
1972
+ "grad_norm": 0.08641325682401657,
1973
+ "learning_rate": 3.2920353982300886e-05,
1974
+ "loss": 0.0225,
1975
+ "step": 2430
1976
+ },
1977
+ {
1978
+ "epoch": 1.1902439024390243,
1979
+ "grad_norm": 0.06566398590803146,
1980
+ "learning_rate": 3.283185840707965e-05,
1981
+ "loss": 0.0177,
1982
+ "step": 2440
1983
+ },
1984
+ {
1985
+ "epoch": 1.1951219512195121,
1986
+ "grad_norm": 0.06867921352386475,
1987
+ "learning_rate": 3.274336283185841e-05,
1988
+ "loss": 0.02,
1989
+ "step": 2450
1990
+ },
1991
+ {
1992
+ "epoch": 1.2,
1993
+ "grad_norm": 0.0619225949048996,
1994
+ "learning_rate": 3.265486725663717e-05,
1995
+ "loss": 0.0203,
1996
+ "step": 2460
1997
+ },
1998
+ {
1999
+ "epoch": 1.2048780487804878,
2000
+ "grad_norm": 0.07883109152317047,
2001
+ "learning_rate": 3.256637168141593e-05,
2002
+ "loss": 0.0199,
2003
+ "step": 2470
2004
+ },
2005
+ {
2006
+ "epoch": 1.2097560975609756,
2007
+ "grad_norm": 0.09298081696033478,
2008
+ "learning_rate": 3.247787610619469e-05,
2009
+ "loss": 0.0191,
2010
+ "step": 2480
2011
+ },
2012
+ {
2013
+ "epoch": 1.2146341463414634,
2014
+ "grad_norm": 0.06301239132881165,
2015
+ "learning_rate": 3.238938053097345e-05,
2016
+ "loss": 0.0212,
2017
+ "step": 2490
2018
+ },
2019
+ {
2020
+ "epoch": 1.2195121951219512,
2021
+ "grad_norm": 0.06936347484588623,
2022
+ "learning_rate": 3.230088495575221e-05,
2023
+ "loss": 0.0203,
2024
+ "step": 2500
2025
+ },
2026
+ {
2027
+ "epoch": 1.2195121951219512,
2028
+ "eval_f1": 0.6559016016048936,
2029
+ "eval_loss": 0.020873118191957474,
2030
+ "eval_precision": 0.5882838770574007,
2031
+ "eval_recall": 0.7410820451843044,
2032
+ "eval_runtime": 132.0933,
2033
+ "eval_samples_per_second": 62.077,
2034
+ "eval_steps_per_second": 0.977,
2035
+ "step": 2500
2036
+ }
2037
+ ],
2038
+ "logging_steps": 10,
2039
+ "max_steps": 6150,
2040
+ "num_input_tokens_seen": 0,
2041
+ "num_train_epochs": 3,
2042
+ "save_steps": 100,
2043
+ "stateful_callbacks": {
2044
+ "TrainerControl": {
2045
+ "args": {
2046
+ "should_epoch_stop": false,
2047
+ "should_evaluate": false,
2048
+ "should_log": false,
2049
+ "should_save": true,
2050
+ "should_training_stop": false
2051
+ },
2052
+ "attributes": {}
2053
+ }
2054
+ },
2055
+ "total_flos": 5311547228160000.0,
2056
+ "train_batch_size": 16,
2057
+ "trial_name": null,
2058
+ "trial_params": null
2059
+ }
checkpoint/training_args (3).bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1203db0f020cb25414a56ea03da2c22cf2f3a145b30e0c1f82662247b998d06b
3
+ size 5304