hossamamer12 commited on
Commit
d1d47bf
1 Parent(s): 8eedb09

Upload 8 files

Browse files
Files changed (8) hide show
  1. config.json +219 -0
  2. optimizer.pt +3 -0
  3. pytorch_model.bin +3 -0
  4. rng_state.pth +3 -0
  5. scaler.pt +3 -0
  6. scheduler.pt +3 -0
  7. trainer_state.json +0 -0
  8. training_args.bin +3 -0
config.json ADDED
@@ -0,0 +1,219 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "hossamamer12/BUS100_MB2_20epoch_notweettokenizer_fp16",
3
+ "architectures": [
4
+ "MobileBertForSequenceClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_activation": true,
8
+ "classifier_dropout": null,
9
+ "embedding_size": 128,
10
+ "hidden_act": "relu",
11
+ "hidden_dropout_prob": 0.0,
12
+ "hidden_size": 512,
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
+ },
105
+ "initializer_range": 0.02,
106
+ "intermediate_size": 512,
107
+ "intra_bottleneck_size": 128,
108
+ "key_query_shared_bottleneck": true,
109
+ "label2id": {
110
+ "LABEL_0": 0,
111
+ "LABEL_1": 1,
112
+ "LABEL_10": 10,
113
+ "LABEL_11": 11,
114
+ "LABEL_12": 12,
115
+ "LABEL_13": 13,
116
+ "LABEL_14": 14,
117
+ "LABEL_15": 15,
118
+ "LABEL_16": 16,
119
+ "LABEL_17": 17,
120
+ "LABEL_18": 18,
121
+ "LABEL_19": 19,
122
+ "LABEL_2": 2,
123
+ "LABEL_20": 20,
124
+ "LABEL_21": 21,
125
+ "LABEL_22": 22,
126
+ "LABEL_23": 23,
127
+ "LABEL_24": 24,
128
+ "LABEL_25": 25,
129
+ "LABEL_26": 26,
130
+ "LABEL_27": 27,
131
+ "LABEL_28": 28,
132
+ "LABEL_29": 29,
133
+ "LABEL_3": 3,
134
+ "LABEL_30": 30,
135
+ "LABEL_31": 31,
136
+ "LABEL_32": 32,
137
+ "LABEL_33": 33,
138
+ "LABEL_34": 34,
139
+ "LABEL_35": 35,
140
+ "LABEL_36": 36,
141
+ "LABEL_37": 37,
142
+ "LABEL_38": 38,
143
+ "LABEL_39": 39,
144
+ "LABEL_4": 4,
145
+ "LABEL_40": 40,
146
+ "LABEL_41": 41,
147
+ "LABEL_42": 42,
148
+ "LABEL_43": 43,
149
+ "LABEL_44": 44,
150
+ "LABEL_45": 45,
151
+ "LABEL_46": 46,
152
+ "LABEL_47": 47,
153
+ "LABEL_48": 48,
154
+ "LABEL_49": 49,
155
+ "LABEL_5": 5,
156
+ "LABEL_50": 50,
157
+ "LABEL_51": 51,
158
+ "LABEL_52": 52,
159
+ "LABEL_53": 53,
160
+ "LABEL_54": 54,
161
+ "LABEL_55": 55,
162
+ "LABEL_56": 56,
163
+ "LABEL_57": 57,
164
+ "LABEL_58": 58,
165
+ "LABEL_59": 59,
166
+ "LABEL_6": 6,
167
+ "LABEL_60": 60,
168
+ "LABEL_61": 61,
169
+ "LABEL_62": 62,
170
+ "LABEL_63": 63,
171
+ "LABEL_64": 64,
172
+ "LABEL_65": 65,
173
+ "LABEL_66": 66,
174
+ "LABEL_67": 67,
175
+ "LABEL_68": 68,
176
+ "LABEL_69": 69,
177
+ "LABEL_7": 7,
178
+ "LABEL_70": 70,
179
+ "LABEL_71": 71,
180
+ "LABEL_72": 72,
181
+ "LABEL_73": 73,
182
+ "LABEL_74": 74,
183
+ "LABEL_75": 75,
184
+ "LABEL_76": 76,
185
+ "LABEL_77": 77,
186
+ "LABEL_78": 78,
187
+ "LABEL_79": 79,
188
+ "LABEL_8": 8,
189
+ "LABEL_80": 80,
190
+ "LABEL_81": 81,
191
+ "LABEL_82": 82,
192
+ "LABEL_83": 83,
193
+ "LABEL_84": 84,
194
+ "LABEL_85": 85,
195
+ "LABEL_86": 86,
196
+ "LABEL_87": 87,
197
+ "LABEL_88": 88,
198
+ "LABEL_89": 89,
199
+ "LABEL_9": 9
200
+ },
201
+ "layer_norm_eps": 1e-12,
202
+ "max_length": 128,
203
+ "max_position_embeddings": 512,
204
+ "model_type": "mobilebert",
205
+ "normalization_type": "no_norm",
206
+ "num_attention_heads": 4,
207
+ "num_feedforward_networks": 4,
208
+ "num_hidden_layers": 2,
209
+ "pad_token_id": 0,
210
+ "problem_type": "single_label_classification",
211
+ "torch_dtype": "float32",
212
+ "transformers_version": "4.20.1",
213
+ "trigram_input": true,
214
+ "true_hidden_size": 128,
215
+ "type_vocab_size": 2,
216
+ "use_bottleneck": true,
217
+ "use_bottleneck_attention": false,
218
+ "vocab_size": 2016
219
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c092c8d348651084c1fa3b6fc3851857307d100d7678a13332b87da0a6e8ab29
3
+ size 21760229
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45f9ad339686323696aeb36d2b1386af2db81bbce9ab040403e3b0ba50b480e2
3
+ size 10889265
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc77eee8e4b9413437656c44ef555216202b79f847f4aa5b263929042157b7c4
3
+ size 15523
scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a668203128a4729ac0daaf8a300cd320842d39a499ccbbe678b38bfc1d8d5bdc
3
+ size 559
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a93410700c4f616ea110c5f53207b8b8f198cba31cf70b06596b57e9ba9f0826
3
+ size 623
trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5fa4c86aa06ea51907528c70d7a09cbf4b3b556a11e2d251cb32675f8d68e9fa
3
+ size 3567