开源SpaceExploreAI模型:基于Transformer模型训练的5M的大语言模型
Browse files- SpaceExploreAI_training_curves.png +0 -0
- config.json +23 -0
- special_tokens_map.json +6 -0
- tokenizer.json +715 -0
- tokenizer_config.json +85 -0
- usage.md +55 -0
- vocab.txt +69 -0
SpaceExploreAI_training_curves.png
ADDED
![]() |
config.json
ADDED
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"model_type": "space_explore_ai_financial",
|
3 |
+
"architectures": [
|
4 |
+
"SpaceExploreAIFinancialModel"
|
5 |
+
],
|
6 |
+
"bos_token_id": 1,
|
7 |
+
"eos_token_id": 2,
|
8 |
+
"hidden_act": "gelu",
|
9 |
+
"initializer_range": 0.02,
|
10 |
+
"layer_norm_epsilon": 1e-12,
|
11 |
+
"pad_token_id": 0,
|
12 |
+
"unk_token_id": 3,
|
13 |
+
"transformers_version": "4.36.0",
|
14 |
+
"use_cache": true,
|
15 |
+
"vocab_size": 69,
|
16 |
+
"tokenizer_class": "FinancialFeatureProcessor",
|
17 |
+
"model_name": "SpaceExploreAI",
|
18 |
+
"hidden_size": 768,
|
19 |
+
"num_hidden_layers": 12,
|
20 |
+
"num_attention_heads": 12,
|
21 |
+
"max_position_embeddings": 1024,
|
22 |
+
"model_max_length": 1024
|
23 |
+
}
|
special_tokens_map.json
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"pad_token": "<pad>",
|
3 |
+
"unk_token": "<unk>",
|
4 |
+
"bos_token": "<bos>",
|
5 |
+
"eos_token": "<eos>"
|
6 |
+
}
|
tokenizer.json
ADDED
@@ -0,0 +1,715 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"version": "1.0",
|
3 |
+
"truncation": null,
|
4 |
+
"padding": null,
|
5 |
+
"added_tokens": [
|
6 |
+
{
|
7 |
+
"id": 0,
|
8 |
+
"special": true,
|
9 |
+
"content": "<pad>",
|
10 |
+
"single_word": false,
|
11 |
+
"lstrip": false,
|
12 |
+
"rstrip": false,
|
13 |
+
"normalized": false
|
14 |
+
},
|
15 |
+
{
|
16 |
+
"id": 1,
|
17 |
+
"special": true,
|
18 |
+
"content": "<bos>",
|
19 |
+
"single_word": false,
|
20 |
+
"lstrip": false,
|
21 |
+
"rstrip": false,
|
22 |
+
"normalized": false
|
23 |
+
},
|
24 |
+
{
|
25 |
+
"id": 2,
|
26 |
+
"special": true,
|
27 |
+
"content": "<eos>",
|
28 |
+
"single_word": false,
|
29 |
+
"lstrip": false,
|
30 |
+
"rstrip": false,
|
31 |
+
"normalized": false
|
32 |
+
},
|
33 |
+
{
|
34 |
+
"id": 3,
|
35 |
+
"special": true,
|
36 |
+
"content": "<unk>",
|
37 |
+
"single_word": false,
|
38 |
+
"lstrip": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"normalized": false
|
41 |
+
},
|
42 |
+
{
|
43 |
+
"id": 4,
|
44 |
+
"special": false,
|
45 |
+
"content": "date",
|
46 |
+
"single_word": true,
|
47 |
+
"lstrip": false,
|
48 |
+
"rstrip": false,
|
49 |
+
"normalized": false
|
50 |
+
},
|
51 |
+
{
|
52 |
+
"id": 5,
|
53 |
+
"special": false,
|
54 |
+
"content": "close",
|
55 |
+
"single_word": true,
|
56 |
+
"lstrip": false,
|
57 |
+
"rstrip": false,
|
58 |
+
"normalized": false
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"id": 6,
|
62 |
+
"special": false,
|
63 |
+
"content": "high",
|
64 |
+
"single_word": true,
|
65 |
+
"lstrip": false,
|
66 |
+
"rstrip": false,
|
67 |
+
"normalized": false
|
68 |
+
},
|
69 |
+
{
|
70 |
+
"id": 7,
|
71 |
+
"special": false,
|
72 |
+
"content": "low",
|
73 |
+
"single_word": true,
|
74 |
+
"lstrip": false,
|
75 |
+
"rstrip": false,
|
76 |
+
"normalized": false
|
77 |
+
},
|
78 |
+
{
|
79 |
+
"id": 8,
|
80 |
+
"special": false,
|
81 |
+
"content": "open",
|
82 |
+
"single_word": true,
|
83 |
+
"lstrip": false,
|
84 |
+
"rstrip": false,
|
85 |
+
"normalized": false
|
86 |
+
},
|
87 |
+
{
|
88 |
+
"id": 9,
|
89 |
+
"special": false,
|
90 |
+
"content": "volume",
|
91 |
+
"single_word": true,
|
92 |
+
"lstrip": false,
|
93 |
+
"rstrip": false,
|
94 |
+
"normalized": false
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"id": 10,
|
98 |
+
"special": false,
|
99 |
+
"content": "day_of_week",
|
100 |
+
"single_word": true,
|
101 |
+
"lstrip": false,
|
102 |
+
"rstrip": false,
|
103 |
+
"normalized": false
|
104 |
+
},
|
105 |
+
{
|
106 |
+
"id": 11,
|
107 |
+
"special": false,
|
108 |
+
"content": "day_of_month",
|
109 |
+
"single_word": true,
|
110 |
+
"lstrip": false,
|
111 |
+
"rstrip": false,
|
112 |
+
"normalized": false
|
113 |
+
},
|
114 |
+
{
|
115 |
+
"id": 12,
|
116 |
+
"special": false,
|
117 |
+
"content": "week_of_year",
|
118 |
+
"single_word": true,
|
119 |
+
"lstrip": false,
|
120 |
+
"rstrip": false,
|
121 |
+
"normalized": false
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"id": 13,
|
125 |
+
"special": false,
|
126 |
+
"content": "month",
|
127 |
+
"single_word": true,
|
128 |
+
"lstrip": false,
|
129 |
+
"rstrip": false,
|
130 |
+
"normalized": false
|
131 |
+
},
|
132 |
+
{
|
133 |
+
"id": 14,
|
134 |
+
"special": false,
|
135 |
+
"content": "quarter",
|
136 |
+
"single_word": true,
|
137 |
+
"lstrip": false,
|
138 |
+
"rstrip": false,
|
139 |
+
"normalized": false
|
140 |
+
},
|
141 |
+
{
|
142 |
+
"id": 15,
|
143 |
+
"special": false,
|
144 |
+
"content": "year",
|
145 |
+
"single_word": true,
|
146 |
+
"lstrip": false,
|
147 |
+
"rstrip": false,
|
148 |
+
"normalized": false
|
149 |
+
},
|
150 |
+
{
|
151 |
+
"id": 16,
|
152 |
+
"special": false,
|
153 |
+
"content": "is_month_start",
|
154 |
+
"single_word": true,
|
155 |
+
"lstrip": false,
|
156 |
+
"rstrip": false,
|
157 |
+
"normalized": false
|
158 |
+
},
|
159 |
+
{
|
160 |
+
"id": 17,
|
161 |
+
"special": false,
|
162 |
+
"content": "is_month_end",
|
163 |
+
"single_word": true,
|
164 |
+
"lstrip": false,
|
165 |
+
"rstrip": false,
|
166 |
+
"normalized": false
|
167 |
+
},
|
168 |
+
{
|
169 |
+
"id": 18,
|
170 |
+
"special": false,
|
171 |
+
"content": "is_week_start",
|
172 |
+
"single_word": true,
|
173 |
+
"lstrip": false,
|
174 |
+
"rstrip": false,
|
175 |
+
"normalized": false
|
176 |
+
},
|
177 |
+
{
|
178 |
+
"id": 19,
|
179 |
+
"special": false,
|
180 |
+
"content": "is_week_end",
|
181 |
+
"single_word": true,
|
182 |
+
"lstrip": false,
|
183 |
+
"rstrip": false,
|
184 |
+
"normalized": false
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"id": 20,
|
188 |
+
"special": false,
|
189 |
+
"content": "close_lag_1",
|
190 |
+
"single_word": true,
|
191 |
+
"lstrip": false,
|
192 |
+
"rstrip": false,
|
193 |
+
"normalized": false
|
194 |
+
},
|
195 |
+
{
|
196 |
+
"id": 21,
|
197 |
+
"special": false,
|
198 |
+
"content": "close_lag_2",
|
199 |
+
"single_word": true,
|
200 |
+
"lstrip": false,
|
201 |
+
"rstrip": false,
|
202 |
+
"normalized": false
|
203 |
+
},
|
204 |
+
{
|
205 |
+
"id": 22,
|
206 |
+
"special": false,
|
207 |
+
"content": "close_lag_3",
|
208 |
+
"single_word": true,
|
209 |
+
"lstrip": false,
|
210 |
+
"rstrip": false,
|
211 |
+
"normalized": false
|
212 |
+
},
|
213 |
+
{
|
214 |
+
"id": 23,
|
215 |
+
"special": false,
|
216 |
+
"content": "close_lag_5",
|
217 |
+
"single_word": true,
|
218 |
+
"lstrip": false,
|
219 |
+
"rstrip": false,
|
220 |
+
"normalized": false
|
221 |
+
},
|
222 |
+
{
|
223 |
+
"id": 24,
|
224 |
+
"special": false,
|
225 |
+
"content": "close_lag_10",
|
226 |
+
"single_word": true,
|
227 |
+
"lstrip": false,
|
228 |
+
"rstrip": false,
|
229 |
+
"normalized": false
|
230 |
+
},
|
231 |
+
{
|
232 |
+
"id": 25,
|
233 |
+
"special": false,
|
234 |
+
"content": "high_lag_1",
|
235 |
+
"single_word": true,
|
236 |
+
"lstrip": false,
|
237 |
+
"rstrip": false,
|
238 |
+
"normalized": false
|
239 |
+
},
|
240 |
+
{
|
241 |
+
"id": 26,
|
242 |
+
"special": false,
|
243 |
+
"content": "high_lag_2",
|
244 |
+
"single_word": true,
|
245 |
+
"lstrip": false,
|
246 |
+
"rstrip": false,
|
247 |
+
"normalized": false
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"id": 27,
|
251 |
+
"special": false,
|
252 |
+
"content": "high_lag_3",
|
253 |
+
"single_word": true,
|
254 |
+
"lstrip": false,
|
255 |
+
"rstrip": false,
|
256 |
+
"normalized": false
|
257 |
+
},
|
258 |
+
{
|
259 |
+
"id": 28,
|
260 |
+
"special": false,
|
261 |
+
"content": "high_lag_5",
|
262 |
+
"single_word": true,
|
263 |
+
"lstrip": false,
|
264 |
+
"rstrip": false,
|
265 |
+
"normalized": false
|
266 |
+
},
|
267 |
+
{
|
268 |
+
"id": 29,
|
269 |
+
"special": false,
|
270 |
+
"content": "high_lag_10",
|
271 |
+
"single_word": true,
|
272 |
+
"lstrip": false,
|
273 |
+
"rstrip": false,
|
274 |
+
"normalized": false
|
275 |
+
},
|
276 |
+
{
|
277 |
+
"id": 30,
|
278 |
+
"special": false,
|
279 |
+
"content": "low_lag_1",
|
280 |
+
"single_word": true,
|
281 |
+
"lstrip": false,
|
282 |
+
"rstrip": false,
|
283 |
+
"normalized": false
|
284 |
+
},
|
285 |
+
{
|
286 |
+
"id": 31,
|
287 |
+
"special": false,
|
288 |
+
"content": "low_lag_2",
|
289 |
+
"single_word": true,
|
290 |
+
"lstrip": false,
|
291 |
+
"rstrip": false,
|
292 |
+
"normalized": false
|
293 |
+
},
|
294 |
+
{
|
295 |
+
"id": 32,
|
296 |
+
"special": false,
|
297 |
+
"content": "low_lag_3",
|
298 |
+
"single_word": true,
|
299 |
+
"lstrip": false,
|
300 |
+
"rstrip": false,
|
301 |
+
"normalized": false
|
302 |
+
},
|
303 |
+
{
|
304 |
+
"id": 33,
|
305 |
+
"special": false,
|
306 |
+
"content": "low_lag_5",
|
307 |
+
"single_word": true,
|
308 |
+
"lstrip": false,
|
309 |
+
"rstrip": false,
|
310 |
+
"normalized": false
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"id": 34,
|
314 |
+
"special": false,
|
315 |
+
"content": "low_lag_10",
|
316 |
+
"single_word": true,
|
317 |
+
"lstrip": false,
|
318 |
+
"rstrip": false,
|
319 |
+
"normalized": false
|
320 |
+
},
|
321 |
+
{
|
322 |
+
"id": 35,
|
323 |
+
"special": false,
|
324 |
+
"content": "volume_lag_1",
|
325 |
+
"single_word": true,
|
326 |
+
"lstrip": false,
|
327 |
+
"rstrip": false,
|
328 |
+
"normalized": false
|
329 |
+
},
|
330 |
+
{
|
331 |
+
"id": 36,
|
332 |
+
"special": false,
|
333 |
+
"content": "volume_lag_2",
|
334 |
+
"single_word": true,
|
335 |
+
"lstrip": false,
|
336 |
+
"rstrip": false,
|
337 |
+
"normalized": false
|
338 |
+
},
|
339 |
+
{
|
340 |
+
"id": 37,
|
341 |
+
"special": false,
|
342 |
+
"content": "volume_lag_3",
|
343 |
+
"single_word": true,
|
344 |
+
"lstrip": false,
|
345 |
+
"rstrip": false,
|
346 |
+
"normalized": false
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"id": 38,
|
350 |
+
"special": false,
|
351 |
+
"content": "volume_lag_5",
|
352 |
+
"single_word": true,
|
353 |
+
"lstrip": false,
|
354 |
+
"rstrip": false,
|
355 |
+
"normalized": false
|
356 |
+
},
|
357 |
+
{
|
358 |
+
"id": 39,
|
359 |
+
"special": false,
|
360 |
+
"content": "volume_lag_10",
|
361 |
+
"single_word": true,
|
362 |
+
"lstrip": false,
|
363 |
+
"rstrip": false,
|
364 |
+
"normalized": false
|
365 |
+
},
|
366 |
+
{
|
367 |
+
"id": 40,
|
368 |
+
"special": false,
|
369 |
+
"content": "future_return_1d",
|
370 |
+
"single_word": true,
|
371 |
+
"lstrip": false,
|
372 |
+
"rstrip": false,
|
373 |
+
"normalized": false
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"id": 41,
|
377 |
+
"special": false,
|
378 |
+
"content": "future_return_2d",
|
379 |
+
"single_word": true,
|
380 |
+
"lstrip": false,
|
381 |
+
"rstrip": false,
|
382 |
+
"normalized": false
|
383 |
+
},
|
384 |
+
{
|
385 |
+
"id": 42,
|
386 |
+
"special": false,
|
387 |
+
"content": "future_return_10d",
|
388 |
+
"single_word": true,
|
389 |
+
"lstrip": false,
|
390 |
+
"rstrip": false,
|
391 |
+
"normalized": false
|
392 |
+
},
|
393 |
+
{
|
394 |
+
"id": 43,
|
395 |
+
"special": false,
|
396 |
+
"content": "past_return_1d",
|
397 |
+
"single_word": true,
|
398 |
+
"lstrip": false,
|
399 |
+
"rstrip": false,
|
400 |
+
"normalized": false
|
401 |
+
},
|
402 |
+
{
|
403 |
+
"id": 44,
|
404 |
+
"special": false,
|
405 |
+
"content": "past_return_2d",
|
406 |
+
"single_word": true,
|
407 |
+
"lstrip": false,
|
408 |
+
"rstrip": false,
|
409 |
+
"normalized": false
|
410 |
+
},
|
411 |
+
{
|
412 |
+
"id": 45,
|
413 |
+
"special": false,
|
414 |
+
"content": "past_return_5d",
|
415 |
+
"single_word": true,
|
416 |
+
"lstrip": false,
|
417 |
+
"rstrip": false,
|
418 |
+
"normalized": false
|
419 |
+
},
|
420 |
+
{
|
421 |
+
"id": 46,
|
422 |
+
"special": false,
|
423 |
+
"content": "past_return_10d",
|
424 |
+
"single_word": true,
|
425 |
+
"lstrip": false,
|
426 |
+
"rstrip": false,
|
427 |
+
"normalized": false
|
428 |
+
},
|
429 |
+
{
|
430 |
+
"id": 47,
|
431 |
+
"special": false,
|
432 |
+
"content": "daily_range",
|
433 |
+
"single_word": true,
|
434 |
+
"lstrip": false,
|
435 |
+
"rstrip": false,
|
436 |
+
"normalized": false
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"id": 48,
|
440 |
+
"special": false,
|
441 |
+
"content": "daily_range_abs",
|
442 |
+
"single_word": true,
|
443 |
+
"lstrip": false,
|
444 |
+
"rstrip": false,
|
445 |
+
"normalized": false
|
446 |
+
},
|
447 |
+
{
|
448 |
+
"id": 49,
|
449 |
+
"special": false,
|
450 |
+
"content": "gap",
|
451 |
+
"single_word": true,
|
452 |
+
"lstrip": false,
|
453 |
+
"rstrip": false,
|
454 |
+
"normalized": false
|
455 |
+
},
|
456 |
+
{
|
457 |
+
"id": 50,
|
458 |
+
"special": false,
|
459 |
+
"content": "close_to_open",
|
460 |
+
"single_word": true,
|
461 |
+
"lstrip": false,
|
462 |
+
"rstrip": false,
|
463 |
+
"normalized": false
|
464 |
+
},
|
465 |
+
{
|
466 |
+
"id": 51,
|
467 |
+
"special": false,
|
468 |
+
"content": "close_to_high",
|
469 |
+
"single_word": true,
|
470 |
+
"lstrip": false,
|
471 |
+
"rstrip": false,
|
472 |
+
"normalized": false
|
473 |
+
},
|
474 |
+
{
|
475 |
+
"id": 52,
|
476 |
+
"special": false,
|
477 |
+
"content": "close_to_low",
|
478 |
+
"single_word": true,
|
479 |
+
"lstrip": false,
|
480 |
+
"rstrip": false,
|
481 |
+
"normalized": false
|
482 |
+
},
|
483 |
+
{
|
484 |
+
"id": 53,
|
485 |
+
"special": false,
|
486 |
+
"content": "volatility_5d",
|
487 |
+
"single_word": true,
|
488 |
+
"lstrip": false,
|
489 |
+
"rstrip": false,
|
490 |
+
"normalized": false
|
491 |
+
},
|
492 |
+
{
|
493 |
+
"id": 54,
|
494 |
+
"special": false,
|
495 |
+
"content": "volatility_10d",
|
496 |
+
"single_word": true,
|
497 |
+
"lstrip": false,
|
498 |
+
"rstrip": false,
|
499 |
+
"normalized": false
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"id": 55,
|
503 |
+
"special": false,
|
504 |
+
"content": "volatility_21d",
|
505 |
+
"single_word": true,
|
506 |
+
"lstrip": false,
|
507 |
+
"rstrip": false,
|
508 |
+
"normalized": false
|
509 |
+
},
|
510 |
+
{
|
511 |
+
"id": 56,
|
512 |
+
"special": false,
|
513 |
+
"content": "atr_5d",
|
514 |
+
"single_word": true,
|
515 |
+
"lstrip": false,
|
516 |
+
"rstrip": false,
|
517 |
+
"normalized": false
|
518 |
+
},
|
519 |
+
{
|
520 |
+
"id": 57,
|
521 |
+
"special": false,
|
522 |
+
"content": "atr_10d",
|
523 |
+
"single_word": true,
|
524 |
+
"lstrip": false,
|
525 |
+
"rstrip": false,
|
526 |
+
"normalized": false
|
527 |
+
},
|
528 |
+
{
|
529 |
+
"id": 58,
|
530 |
+
"special": false,
|
531 |
+
"content": "atr_21d",
|
532 |
+
"single_word": true,
|
533 |
+
"lstrip": false,
|
534 |
+
"rstrip": false,
|
535 |
+
"normalized": false
|
536 |
+
},
|
537 |
+
{
|
538 |
+
"id": 59,
|
539 |
+
"special": false,
|
540 |
+
"content": "momentum_5d",
|
541 |
+
"single_word": true,
|
542 |
+
"lstrip": false,
|
543 |
+
"rstrip": false,
|
544 |
+
"normalized": false
|
545 |
+
},
|
546 |
+
{
|
547 |
+
"id": 60,
|
548 |
+
"special": false,
|
549 |
+
"content": "momentum_10d",
|
550 |
+
"single_word": true,
|
551 |
+
"lstrip": false,
|
552 |
+
"rstrip": false,
|
553 |
+
"normalized": false
|
554 |
+
},
|
555 |
+
{
|
556 |
+
"id": 61,
|
557 |
+
"special": false,
|
558 |
+
"content": "momentum_21d",
|
559 |
+
"single_word": true,
|
560 |
+
"lstrip": false,
|
561 |
+
"rstrip": false,
|
562 |
+
"normalized": false
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"id": 62,
|
566 |
+
"special": false,
|
567 |
+
"content": "volume_change_5d",
|
568 |
+
"single_word": true,
|
569 |
+
"lstrip": false,
|
570 |
+
"rstrip": false,
|
571 |
+
"normalized": false
|
572 |
+
},
|
573 |
+
{
|
574 |
+
"id": 63,
|
575 |
+
"special": false,
|
576 |
+
"content": "volume_change_10d",
|
577 |
+
"single_word": true,
|
578 |
+
"lstrip": false,
|
579 |
+
"rstrip": false,
|
580 |
+
"normalized": false
|
581 |
+
},
|
582 |
+
{
|
583 |
+
"id": 64,
|
584 |
+
"special": false,
|
585 |
+
"content": "volume_change_21d",
|
586 |
+
"single_word": true,
|
587 |
+
"lstrip": false,
|
588 |
+
"rstrip": false,
|
589 |
+
"normalized": false
|
590 |
+
},
|
591 |
+
{
|
592 |
+
"id": 65,
|
593 |
+
"special": false,
|
594 |
+
"content": "close_to_ma_5d",
|
595 |
+
"single_word": true,
|
596 |
+
"lstrip": false,
|
597 |
+
"rstrip": false,
|
598 |
+
"normalized": false
|
599 |
+
},
|
600 |
+
{
|
601 |
+
"id": 66,
|
602 |
+
"special": false,
|
603 |
+
"content": "close_to_ma_10d",
|
604 |
+
"single_word": true,
|
605 |
+
"lstrip": false,
|
606 |
+
"rstrip": false,
|
607 |
+
"normalized": false
|
608 |
+
},
|
609 |
+
{
|
610 |
+
"id": 67,
|
611 |
+
"special": false,
|
612 |
+
"content": "close_to_ma_21d",
|
613 |
+
"single_word": true,
|
614 |
+
"lstrip": false,
|
615 |
+
"rstrip": false,
|
616 |
+
"normalized": false
|
617 |
+
},
|
618 |
+
{
|
619 |
+
"id": 68,
|
620 |
+
"special": false,
|
621 |
+
"content": "volume_change_rate",
|
622 |
+
"single_word": true,
|
623 |
+
"lstrip": false,
|
624 |
+
"rstrip": false,
|
625 |
+
"normalized": false
|
626 |
+
}
|
627 |
+
],
|
628 |
+
"normalizer": {
|
629 |
+
"type": "FinancialFeatureNormalizer"
|
630 |
+
},
|
631 |
+
"pre_tokenizer": {
|
632 |
+
"type": "FinancialFeaturePreTokenizer"
|
633 |
+
},
|
634 |
+
"post_processor": {
|
635 |
+
"type": "FinancialFeaturePostProcessor"
|
636 |
+
},
|
637 |
+
"decoder": {
|
638 |
+
"type": "FinancialFeatureDecoder"
|
639 |
+
},
|
640 |
+
"model": {
|
641 |
+
"type": "FinancialFeatureProcessor",
|
642 |
+
"vocab": {
|
643 |
+
"<pad>": 0,
|
644 |
+
"<bos>": 1,
|
645 |
+
"<eos>": 2,
|
646 |
+
"<unk>": 3,
|
647 |
+
"date": 4,
|
648 |
+
"close": 5,
|
649 |
+
"high": 6,
|
650 |
+
"low": 7,
|
651 |
+
"open": 8,
|
652 |
+
"volume": 9,
|
653 |
+
"day_of_week": 10,
|
654 |
+
"day_of_month": 11,
|
655 |
+
"week_of_year": 12,
|
656 |
+
"month": 13,
|
657 |
+
"quarter": 14,
|
658 |
+
"year": 15,
|
659 |
+
"is_month_start": 16,
|
660 |
+
"is_month_end": 17,
|
661 |
+
"is_week_start": 18,
|
662 |
+
"is_week_end": 19,
|
663 |
+
"close_lag_1": 20,
|
664 |
+
"close_lag_2": 21,
|
665 |
+
"close_lag_3": 22,
|
666 |
+
"close_lag_5": 23,
|
667 |
+
"close_lag_10": 24,
|
668 |
+
"high_lag_1": 25,
|
669 |
+
"high_lag_2": 26,
|
670 |
+
"high_lag_3": 27,
|
671 |
+
"high_lag_5": 28,
|
672 |
+
"high_lag_10": 29,
|
673 |
+
"low_lag_1": 30,
|
674 |
+
"low_lag_2": 31,
|
675 |
+
"low_lag_3": 32,
|
676 |
+
"low_lag_5": 33,
|
677 |
+
"low_lag_10": 34,
|
678 |
+
"volume_lag_1": 35,
|
679 |
+
"volume_lag_2": 36,
|
680 |
+
"volume_lag_3": 37,
|
681 |
+
"volume_lag_5": 38,
|
682 |
+
"volume_lag_10": 39,
|
683 |
+
"future_return_1d": 40,
|
684 |
+
"future_return_2d": 41,
|
685 |
+
"future_return_10d": 42,
|
686 |
+
"past_return_1d": 43,
|
687 |
+
"past_return_2d": 44,
|
688 |
+
"past_return_5d": 45,
|
689 |
+
"past_return_10d": 46,
|
690 |
+
"daily_range": 47,
|
691 |
+
"daily_range_abs": 48,
|
692 |
+
"gap": 49,
|
693 |
+
"close_to_open": 50,
|
694 |
+
"close_to_high": 51,
|
695 |
+
"close_to_low": 52,
|
696 |
+
"volatility_5d": 53,
|
697 |
+
"volatility_10d": 54,
|
698 |
+
"volatility_21d": 55,
|
699 |
+
"atr_5d": 56,
|
700 |
+
"atr_10d": 57,
|
701 |
+
"atr_21d": 58,
|
702 |
+
"momentum_5d": 59,
|
703 |
+
"momentum_10d": 60,
|
704 |
+
"momentum_21d": 61,
|
705 |
+
"volume_change_5d": 62,
|
706 |
+
"volume_change_10d": 63,
|
707 |
+
"volume_change_21d": 64,
|
708 |
+
"close_to_ma_5d": 65,
|
709 |
+
"close_to_ma_10d": 66,
|
710 |
+
"close_to_ma_21d": 67,
|
711 |
+
"volume_change_rate": 68
|
712 |
+
},
|
713 |
+
"feature_dim": 69
|
714 |
+
}
|
715 |
+
}
|
tokenizer_config.json
ADDED
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"feature_dim": 64,
|
3 |
+
"feature_groups": [
|
4 |
+
"price",
|
5 |
+
"time",
|
6 |
+
"lag",
|
7 |
+
"return",
|
8 |
+
"volatility",
|
9 |
+
"volume"
|
10 |
+
],
|
11 |
+
"feature_names": [
|
12 |
+
"date",
|
13 |
+
"close",
|
14 |
+
"high",
|
15 |
+
"low",
|
16 |
+
"open",
|
17 |
+
"volume",
|
18 |
+
"day_of_week",
|
19 |
+
"day_of_month",
|
20 |
+
"week_of_year",
|
21 |
+
"month",
|
22 |
+
"quarter",
|
23 |
+
"year",
|
24 |
+
"is_month_start",
|
25 |
+
"is_month_end",
|
26 |
+
"is_week_start",
|
27 |
+
"is_week_end",
|
28 |
+
"close_lag_1",
|
29 |
+
"close_lag_2",
|
30 |
+
"close_lag_3",
|
31 |
+
"close_lag_5",
|
32 |
+
"close_lag_10",
|
33 |
+
"high_lag_1",
|
34 |
+
"high_lag_2",
|
35 |
+
"high_lag_3",
|
36 |
+
"high_lag_5",
|
37 |
+
"high_lag_10",
|
38 |
+
"low_lag_1",
|
39 |
+
"low_lag_2",
|
40 |
+
"low_lag_3",
|
41 |
+
"low_lag_5",
|
42 |
+
"low_lag_10",
|
43 |
+
"volume_lag_1",
|
44 |
+
"volume_lag_2",
|
45 |
+
"volume_lag_3",
|
46 |
+
"volume_lag_5",
|
47 |
+
"volume_lag_10",
|
48 |
+
"future_return_1d",
|
49 |
+
"future_return_2d",
|
50 |
+
"future_return_10d",
|
51 |
+
"past_return_1d",
|
52 |
+
"past_return_2d",
|
53 |
+
"past_return_5d",
|
54 |
+
"past_return_10d",
|
55 |
+
"daily_range",
|
56 |
+
"daily_range_abs",
|
57 |
+
"gap",
|
58 |
+
"close_to_open",
|
59 |
+
"close_to_high",
|
60 |
+
"close_to_low",
|
61 |
+
"volatility_5d",
|
62 |
+
"volatility_10d",
|
63 |
+
"volatility_21d",
|
64 |
+
"atr_5d",
|
65 |
+
"atr_10d",
|
66 |
+
"atr_21d",
|
67 |
+
"momentum_5d",
|
68 |
+
"momentum_10d",
|
69 |
+
"momentum_21d",
|
70 |
+
"volume_change_5d",
|
71 |
+
"volume_change_10d",
|
72 |
+
"volume_change_21d",
|
73 |
+
"close_to_ma_5d",
|
74 |
+
"close_to_ma_10d",
|
75 |
+
"close_to_ma_21d",
|
76 |
+
"volume_change_rate"
|
77 |
+
],
|
78 |
+
"do_lower_case": false,
|
79 |
+
"model_max_length": 1024,
|
80 |
+
"tokenizer_class": "FinancialFeatureProcessor",
|
81 |
+
"bos_token": "<bos>",
|
82 |
+
"eos_token": "<eos>",
|
83 |
+
"pad_token": "<pad>",
|
84 |
+
"unk_token": "<unk>"
|
85 |
+
}
|
usage.md
ADDED
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
|
2 |
+
# SpaceExploreAI-Small-Base-Regression-5M
|
3 |
+
|
4 |
+
## 模型描述
|
5 |
+
|
6 |
+
SpaceExploreAI-Small-Base-Regression-5M 是一个基于深度学习的金融时序预测模型,专为股票价格趋势分析和预测而设计。这是 SpaceExploreAI 系列的小型版本,具有约 5M 参数,针对regression任务进行了优化。
|
7 |
+
|
8 |
+
### 主要特点
|
9 |
+
|
10 |
+
- **轻量化设计**:仅有 5M 参数,适合资源受限环境
|
11 |
+
- **回归预测:专为价格预测等回归任务优化**
|
12 |
+
- **Transformer 架构**:基于 Transformer 架构,集成了 RoPE 旋转位置编码技术
|
13 |
+
- **多头gqa注意力**:使用先进的多头注意力机制捕捉时间序列数据模式
|
14 |
+
|
15 |
+
## 技术规格
|
16 |
+
|
17 |
+
- **参数量**:约 5M
|
18 |
+
- **模型类型**:Transformer
|
19 |
+
- **隐藏层大小**:256
|
20 |
+
- **隐藏层数量**:4
|
21 |
+
- **注意力头数量**:4
|
22 |
+
- **注意力类型**:gqa
|
23 |
+
- **归一化类型**:rmsnorm
|
24 |
+
- **最大序列长度**:32
|
25 |
+
- **预测类型**:regression
|
26 |
+
- **使用 MoE**:是,混合专家模型增强了模型的表达能力
|
27 |
+
- **MOE配置**:8个专家,每个token使用2个专家,中间层大小1024
|
28 |
+
|
29 |
+
## 使用示例
|
30 |
+
|
31 |
+
```python
|
32 |
+
from transformers import AutoTokenizer, AutoModel
|
33 |
+
import torch
|
34 |
+
|
35 |
+
# 加载模型和分词器
|
36 |
+
model_name = "SpaceExploreAI/SpaceExploreAI-Small-Base-Regression-5M"
|
37 |
+
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
38 |
+
model = AutoModel.from_pretrained(model_name)
|
39 |
+
|
40 |
+
# 准备输入数据 (假设您已经有了处理好的金融数据)
|
41 |
+
inputs = torch.tensor([[0.1, 0.2, 0.3, ...]]) # 您的金融序列数据
|
42 |
+
|
43 |
+
# 进行预测
|
44 |
+
with torch.no_grad():
|
45 |
+
outputs = model(inputs)
|
46 |
+
predictions = outputs.last_hidden_state
|
47 |
+
```
|
48 |
+
|
49 |
+
## 免责声明
|
50 |
+
|
51 |
+
SpaceExploreAI仅供大模型AI学习、量化交易学习,不可以用于商业用途、不可以以此为投资逻辑,后果自负。
|
52 |
+
|
53 |
+
## 许可证
|
54 |
+
|
55 |
+
[Apache License 2.0](LICENSE)
|
vocab.txt
ADDED
@@ -0,0 +1,69 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
<pad>
|
2 |
+
<bos>
|
3 |
+
<eos>
|
4 |
+
<unk>
|
5 |
+
date
|
6 |
+
close
|
7 |
+
high
|
8 |
+
low
|
9 |
+
open
|
10 |
+
volume
|
11 |
+
day_of_week
|
12 |
+
day_of_month
|
13 |
+
week_of_year
|
14 |
+
month
|
15 |
+
quarter
|
16 |
+
year
|
17 |
+
is_month_start
|
18 |
+
is_month_end
|
19 |
+
is_week_start
|
20 |
+
is_week_end
|
21 |
+
close_lag_1
|
22 |
+
close_lag_2
|
23 |
+
close_lag_3
|
24 |
+
close_lag_5
|
25 |
+
close_lag_10
|
26 |
+
high_lag_1
|
27 |
+
high_lag_2
|
28 |
+
high_lag_3
|
29 |
+
high_lag_5
|
30 |
+
high_lag_10
|
31 |
+
low_lag_1
|
32 |
+
low_lag_2
|
33 |
+
low_lag_3
|
34 |
+
low_lag_5
|
35 |
+
low_lag_10
|
36 |
+
volume_lag_1
|
37 |
+
volume_lag_2
|
38 |
+
volume_lag_3
|
39 |
+
volume_lag_5
|
40 |
+
volume_lag_10
|
41 |
+
future_return_1d
|
42 |
+
future_return_2d
|
43 |
+
future_return_10d
|
44 |
+
past_return_1d
|
45 |
+
past_return_2d
|
46 |
+
past_return_5d
|
47 |
+
past_return_10d
|
48 |
+
daily_range
|
49 |
+
daily_range_abs
|
50 |
+
gap
|
51 |
+
close_to_open
|
52 |
+
close_to_high
|
53 |
+
close_to_low
|
54 |
+
volatility_5d
|
55 |
+
volatility_10d
|
56 |
+
volatility_21d
|
57 |
+
atr_5d
|
58 |
+
atr_10d
|
59 |
+
atr_21d
|
60 |
+
momentum_5d
|
61 |
+
momentum_10d
|
62 |
+
momentum_21d
|
63 |
+
volume_change_5d
|
64 |
+
volume_change_10d
|
65 |
+
volume_change_21d
|
66 |
+
close_to_ma_5d
|
67 |
+
close_to_ma_10d
|
68 |
+
close_to_ma_21d
|
69 |
+
volume_change_rate
|