mahsaamani
commited on
Commit
•
7fc3371
1
Parent(s):
e816e1b
Upload 9 files
Browse files- POSTagger/config.json +53 -0
- POSTagger/rng_state.pth +3 -0
- POSTagger/scheduler.pt +3 -0
- POSTagger/special_tokens_map.json +7 -0
- POSTagger/tokenizer.json +0 -0
- POSTagger/tokenizer_config.json +15 -0
- POSTagger/trainer_state.json +286 -0
- POSTagger/training_args.bin +3 -0
- POSTagger/vocab.txt +0 -0
POSTagger/config.json
ADDED
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "/content/drive/MyDrive/Azari/Models/AzerBert_v2/checkpoint-11630",
|
3 |
+
"architectures": [
|
4 |
+
"BertForTokenClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"hidden_act": "gelu",
|
9 |
+
"hidden_dropout_prob": 0.1,
|
10 |
+
"hidden_size": 768,
|
11 |
+
"id2label": {
|
12 |
+
"0": "LABEL_0",
|
13 |
+
"1": "LABEL_1",
|
14 |
+
"2": "LABEL_2",
|
15 |
+
"3": "LABEL_3",
|
16 |
+
"4": "LABEL_4",
|
17 |
+
"5": "LABEL_5",
|
18 |
+
"6": "LABEL_6",
|
19 |
+
"7": "LABEL_7",
|
20 |
+
"8": "LABEL_8",
|
21 |
+
"9": "LABEL_9",
|
22 |
+
"10": "LABEL_10",
|
23 |
+
"11": "LABEL_11"
|
24 |
+
},
|
25 |
+
"initializer_range": 0.02,
|
26 |
+
"intermediate_size": 3072,
|
27 |
+
"label2id": {
|
28 |
+
"LABEL_0": 0,
|
29 |
+
"LABEL_1": 1,
|
30 |
+
"LABEL_10": 10,
|
31 |
+
"LABEL_11": 11,
|
32 |
+
"LABEL_2": 2,
|
33 |
+
"LABEL_3": 3,
|
34 |
+
"LABEL_4": 4,
|
35 |
+
"LABEL_5": 5,
|
36 |
+
"LABEL_6": 6,
|
37 |
+
"LABEL_7": 7,
|
38 |
+
"LABEL_8": 8,
|
39 |
+
"LABEL_9": 9
|
40 |
+
},
|
41 |
+
"layer_norm_eps": 1e-12,
|
42 |
+
"max_position_embeddings": 64,
|
43 |
+
"model_type": "bert",
|
44 |
+
"num_attention_heads": 12,
|
45 |
+
"num_hidden_layers": 12,
|
46 |
+
"pad_token_id": 0,
|
47 |
+
"position_embedding_type": "absolute",
|
48 |
+
"torch_dtype": "float32",
|
49 |
+
"transformers_version": "4.29.2",
|
50 |
+
"type_vocab_size": 2,
|
51 |
+
"use_cache": true,
|
52 |
+
"vocab_size": 10000
|
53 |
+
}
|
POSTagger/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9641e5adcdde18050cad9693da3a4c3ebdaa5d43e27196a5672f64790f533dde
|
3 |
+
size 14575
|
POSTagger/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:efd609804b278eb3cd2dacac8e1e7b44a8e21cc3768a6628be08b0cd3484529d
|
3 |
+
size 627
|
POSTagger/special_tokens_map.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": "[CLS]",
|
3 |
+
"mask_token": "[MASK]",
|
4 |
+
"pad_token": "[PAD]",
|
5 |
+
"sep_token": "[SEP]",
|
6 |
+
"unk_token": "[UNK]"
|
7 |
+
}
|
POSTagger/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
POSTagger/tokenizer_config.json
ADDED
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"clean_up_tokenization_spaces": true,
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"do_basic_tokenize": true,
|
5 |
+
"do_lower_case": true,
|
6 |
+
"mask_token": "[MASK]",
|
7 |
+
"model_max_length": 1000000000000000019884624838656,
|
8 |
+
"never_split": null,
|
9 |
+
"pad_token": "[PAD]",
|
10 |
+
"sep_token": "[SEP]",
|
11 |
+
"strip_accents": null,
|
12 |
+
"tokenize_chinese_chars": true,
|
13 |
+
"tokenizer_class": "BertTokenizer",
|
14 |
+
"unk_token": "[UNK]"
|
15 |
+
}
|
POSTagger/trainer_state.json
ADDED
@@ -0,0 +1,286 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.4975282549858093,
|
3 |
+
"best_model_checkpoint": "/content/drive/MyDrive/Azari/Models/POS_TAGGER_v1/checkpoint-45",
|
4 |
+
"epoch": 15.0,
|
5 |
+
"global_step": 45,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 1.0,
|
12 |
+
"learning_rate": 1.9e-05,
|
13 |
+
"loss": 2.0841,
|
14 |
+
"step": 3
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 1.0,
|
18 |
+
"eval_accuracy": 0.6952554744525548,
|
19 |
+
"eval_f1": 0.28720626631853785,
|
20 |
+
"eval_loss": 1.180311918258667,
|
21 |
+
"eval_precision": 0.45454545454545453,
|
22 |
+
"eval_recall": 0.2099236641221374,
|
23 |
+
"eval_runtime": 0.0976,
|
24 |
+
"eval_samples_per_second": 204.909,
|
25 |
+
"eval_steps_per_second": 10.245,
|
26 |
+
"step": 3
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"epoch": 2.0,
|
30 |
+
"learning_rate": 1.8e-05,
|
31 |
+
"loss": 1.1453,
|
32 |
+
"step": 6
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"epoch": 2.0,
|
36 |
+
"eval_accuracy": 0.7627737226277372,
|
37 |
+
"eval_f1": 0.4230769230769231,
|
38 |
+
"eval_loss": 0.930380642414093,
|
39 |
+
"eval_precision": 0.48058252427184467,
|
40 |
+
"eval_recall": 0.37786259541984735,
|
41 |
+
"eval_runtime": 0.0924,
|
42 |
+
"eval_samples_per_second": 216.409,
|
43 |
+
"eval_steps_per_second": 10.82,
|
44 |
+
"step": 6
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 3.0,
|
48 |
+
"learning_rate": 1.7e-05,
|
49 |
+
"loss": 0.8983,
|
50 |
+
"step": 9
|
51 |
+
},
|
52 |
+
{
|
53 |
+
"epoch": 3.0,
|
54 |
+
"eval_accuracy": 0.8047445255474452,
|
55 |
+
"eval_f1": 0.532258064516129,
|
56 |
+
"eval_loss": 0.7620761394500732,
|
57 |
+
"eval_precision": 0.5641025641025641,
|
58 |
+
"eval_recall": 0.5038167938931297,
|
59 |
+
"eval_runtime": 0.1238,
|
60 |
+
"eval_samples_per_second": 161.595,
|
61 |
+
"eval_steps_per_second": 8.08,
|
62 |
+
"step": 9
|
63 |
+
},
|
64 |
+
{
|
65 |
+
"epoch": 4.0,
|
66 |
+
"learning_rate": 1.6000000000000003e-05,
|
67 |
+
"loss": 0.7406,
|
68 |
+
"step": 12
|
69 |
+
},
|
70 |
+
{
|
71 |
+
"epoch": 4.0,
|
72 |
+
"eval_accuracy": 0.8375912408759124,
|
73 |
+
"eval_f1": 0.6052104208416834,
|
74 |
+
"eval_loss": 0.6653763651847839,
|
75 |
+
"eval_precision": 0.6371308016877637,
|
76 |
+
"eval_recall": 0.5763358778625954,
|
77 |
+
"eval_runtime": 0.131,
|
78 |
+
"eval_samples_per_second": 152.687,
|
79 |
+
"eval_steps_per_second": 7.634,
|
80 |
+
"step": 12
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 5.0,
|
84 |
+
"learning_rate": 1.5000000000000002e-05,
|
85 |
+
"loss": 0.6451,
|
86 |
+
"step": 15
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 5.0,
|
90 |
+
"eval_accuracy": 0.8375912408759124,
|
91 |
+
"eval_f1": 0.6127167630057803,
|
92 |
+
"eval_loss": 0.6271026730537415,
|
93 |
+
"eval_precision": 0.6186770428015564,
|
94 |
+
"eval_recall": 0.6068702290076335,
|
95 |
+
"eval_runtime": 0.117,
|
96 |
+
"eval_samples_per_second": 170.987,
|
97 |
+
"eval_steps_per_second": 8.549,
|
98 |
+
"step": 15
|
99 |
+
},
|
100 |
+
{
|
101 |
+
"epoch": 6.0,
|
102 |
+
"learning_rate": 1.4e-05,
|
103 |
+
"loss": 0.5668,
|
104 |
+
"step": 18
|
105 |
+
},
|
106 |
+
{
|
107 |
+
"epoch": 6.0,
|
108 |
+
"eval_accuracy": 0.8375912408759124,
|
109 |
+
"eval_f1": 0.6135922330097087,
|
110 |
+
"eval_loss": 0.589028537273407,
|
111 |
+
"eval_precision": 0.6245059288537549,
|
112 |
+
"eval_recall": 0.6030534351145038,
|
113 |
+
"eval_runtime": 0.273,
|
114 |
+
"eval_samples_per_second": 73.251,
|
115 |
+
"eval_steps_per_second": 3.663,
|
116 |
+
"step": 18
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"epoch": 7.0,
|
120 |
+
"learning_rate": 1.3000000000000001e-05,
|
121 |
+
"loss": 0.5085,
|
122 |
+
"step": 21
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 7.0,
|
126 |
+
"eval_accuracy": 0.8521897810218978,
|
127 |
+
"eval_f1": 0.6459143968871595,
|
128 |
+
"eval_loss": 0.5673397779464722,
|
129 |
+
"eval_precision": 0.6587301587301587,
|
130 |
+
"eval_recall": 0.6335877862595419,
|
131 |
+
"eval_runtime": 0.1023,
|
132 |
+
"eval_samples_per_second": 195.551,
|
133 |
+
"eval_steps_per_second": 9.778,
|
134 |
+
"step": 21
|
135 |
+
},
|
136 |
+
{
|
137 |
+
"epoch": 8.0,
|
138 |
+
"learning_rate": 1.2e-05,
|
139 |
+
"loss": 0.4746,
|
140 |
+
"step": 24
|
141 |
+
},
|
142 |
+
{
|
143 |
+
"epoch": 8.0,
|
144 |
+
"eval_accuracy": 0.8540145985401459,
|
145 |
+
"eval_f1": 0.6472868217054263,
|
146 |
+
"eval_loss": 0.5449931621551514,
|
147 |
+
"eval_precision": 0.65748031496063,
|
148 |
+
"eval_recall": 0.6374045801526718,
|
149 |
+
"eval_runtime": 0.0875,
|
150 |
+
"eval_samples_per_second": 228.5,
|
151 |
+
"eval_steps_per_second": 11.425,
|
152 |
+
"step": 24
|
153 |
+
},
|
154 |
+
{
|
155 |
+
"epoch": 9.0,
|
156 |
+
"learning_rate": 1.1000000000000001e-05,
|
157 |
+
"loss": 0.4312,
|
158 |
+
"step": 27
|
159 |
+
},
|
160 |
+
{
|
161 |
+
"epoch": 9.0,
|
162 |
+
"eval_accuracy": 0.8448905109489051,
|
163 |
+
"eval_f1": 0.6374045801526718,
|
164 |
+
"eval_loss": 0.5437765121459961,
|
165 |
+
"eval_precision": 0.6374045801526718,
|
166 |
+
"eval_recall": 0.6374045801526718,
|
167 |
+
"eval_runtime": 0.1112,
|
168 |
+
"eval_samples_per_second": 179.849,
|
169 |
+
"eval_steps_per_second": 8.992,
|
170 |
+
"step": 27
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 10.0,
|
174 |
+
"learning_rate": 1e-05,
|
175 |
+
"loss": 0.3862,
|
176 |
+
"step": 30
|
177 |
+
},
|
178 |
+
{
|
179 |
+
"epoch": 10.0,
|
180 |
+
"eval_accuracy": 0.8540145985401459,
|
181 |
+
"eval_f1": 0.6511627906976745,
|
182 |
+
"eval_loss": 0.5256561040878296,
|
183 |
+
"eval_precision": 0.6614173228346457,
|
184 |
+
"eval_recall": 0.6412213740458015,
|
185 |
+
"eval_runtime": 0.0917,
|
186 |
+
"eval_samples_per_second": 218.041,
|
187 |
+
"eval_steps_per_second": 10.902,
|
188 |
+
"step": 30
|
189 |
+
},
|
190 |
+
{
|
191 |
+
"epoch": 11.0,
|
192 |
+
"learning_rate": 9e-06,
|
193 |
+
"loss": 0.3544,
|
194 |
+
"step": 33
|
195 |
+
},
|
196 |
+
{
|
197 |
+
"epoch": 11.0,
|
198 |
+
"eval_accuracy": 0.8613138686131386,
|
199 |
+
"eval_f1": 0.665377176015474,
|
200 |
+
"eval_loss": 0.5157161951065063,
|
201 |
+
"eval_precision": 0.6745098039215687,
|
202 |
+
"eval_recall": 0.6564885496183206,
|
203 |
+
"eval_runtime": 0.1237,
|
204 |
+
"eval_samples_per_second": 161.719,
|
205 |
+
"eval_steps_per_second": 8.086,
|
206 |
+
"step": 33
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 12.0,
|
210 |
+
"learning_rate": 8.000000000000001e-06,
|
211 |
+
"loss": 0.3395,
|
212 |
+
"step": 36
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 12.0,
|
216 |
+
"eval_accuracy": 0.8686131386861314,
|
217 |
+
"eval_f1": 0.6821705426356589,
|
218 |
+
"eval_loss": 0.5101653337478638,
|
219 |
+
"eval_precision": 0.6929133858267716,
|
220 |
+
"eval_recall": 0.6717557251908397,
|
221 |
+
"eval_runtime": 0.0907,
|
222 |
+
"eval_samples_per_second": 220.625,
|
223 |
+
"eval_steps_per_second": 11.031,
|
224 |
+
"step": 36
|
225 |
+
},
|
226 |
+
{
|
227 |
+
"epoch": 13.0,
|
228 |
+
"learning_rate": 7e-06,
|
229 |
+
"loss": 0.3123,
|
230 |
+
"step": 39
|
231 |
+
},
|
232 |
+
{
|
233 |
+
"epoch": 13.0,
|
234 |
+
"eval_accuracy": 0.8686131386861314,
|
235 |
+
"eval_f1": 0.6807692307692308,
|
236 |
+
"eval_loss": 0.505219578742981,
|
237 |
+
"eval_precision": 0.686046511627907,
|
238 |
+
"eval_recall": 0.6755725190839694,
|
239 |
+
"eval_runtime": 0.0947,
|
240 |
+
"eval_samples_per_second": 211.236,
|
241 |
+
"eval_steps_per_second": 10.562,
|
242 |
+
"step": 39
|
243 |
+
},
|
244 |
+
{
|
245 |
+
"epoch": 14.0,
|
246 |
+
"learning_rate": 6e-06,
|
247 |
+
"loss": 0.2963,
|
248 |
+
"step": 42
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 14.0,
|
252 |
+
"eval_accuracy": 0.8686131386861314,
|
253 |
+
"eval_f1": 0.6870229007633588,
|
254 |
+
"eval_loss": 0.502780556678772,
|
255 |
+
"eval_precision": 0.6870229007633588,
|
256 |
+
"eval_recall": 0.6870229007633588,
|
257 |
+
"eval_runtime": 0.093,
|
258 |
+
"eval_samples_per_second": 215.022,
|
259 |
+
"eval_steps_per_second": 10.751,
|
260 |
+
"step": 42
|
261 |
+
},
|
262 |
+
{
|
263 |
+
"epoch": 15.0,
|
264 |
+
"learning_rate": 5e-06,
|
265 |
+
"loss": 0.2763,
|
266 |
+
"step": 45
|
267 |
+
},
|
268 |
+
{
|
269 |
+
"epoch": 15.0,
|
270 |
+
"eval_accuracy": 0.8686131386861314,
|
271 |
+
"eval_f1": 0.6794625719769674,
|
272 |
+
"eval_loss": 0.4975282549858093,
|
273 |
+
"eval_precision": 0.6833976833976834,
|
274 |
+
"eval_recall": 0.6755725190839694,
|
275 |
+
"eval_runtime": 0.1181,
|
276 |
+
"eval_samples_per_second": 169.318,
|
277 |
+
"eval_steps_per_second": 8.466,
|
278 |
+
"step": 45
|
279 |
+
}
|
280 |
+
],
|
281 |
+
"max_steps": 60,
|
282 |
+
"num_train_epochs": 20,
|
283 |
+
"total_flos": 78396114124800.0,
|
284 |
+
"trial_name": null,
|
285 |
+
"trial_params": null
|
286 |
+
}
|
POSTagger/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:91002d710b826005c77ab2e22ac59d5d71e286912892da9a9f4f0e79a0a0a388
|
3 |
+
size 3963
|
POSTagger/vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|