LightChen2333 commited on
Commit
8e002f1
1 Parent(s): fa9e36d

Upload 3 files

Browse files
Files changed (1) hide show
  1. config.json +23 -14
config.json CHANGED
@@ -186,7 +186,7 @@
186
  "intent_classifier": {
187
  "_model_target_": "model.decoder.classifier.MLPClassifier",
188
  "dropout_rate": 0.4,
189
- "ignore_index": "{base.ignore_index}",
190
  "input_dim": 384,
191
  "intent_label_num": 7,
192
  "loss_fn": {
@@ -195,7 +195,7 @@
195
  "mlp": [
196
  {
197
  "_model_target_": "torch.nn.Linear",
198
- "in_features": "{model.encoder.output_dim}",
199
  "out_features": 256
200
  },
201
  {
@@ -205,14 +205,14 @@
205
  {
206
  "_model_target_": "torch.nn.Linear",
207
  "in_features": 256,
208
- "out_features": "{base.intent_label_num}"
209
  }
210
  ],
211
  "mode": "token-level-intent",
212
  "multi_threshold": 0.5,
213
  "return_sentence_level": true,
214
  "use_intent": true,
215
- "use_multi": "{base.multi_intent}",
216
  "weight": 0.2
217
  },
218
  "interaction": {
@@ -220,25 +220,26 @@
220
  "alpha": 0.2,
221
  "dropout_rate": 0.4,
222
  "hidden_dim": 256,
223
- "input_dim": "{model.encoder.output_dim}",
224
  "intent_embedding_dim": 256,
225
- "intent_label_num": "{base.intent_label_num}",
226
  "num_heads": 4,
227
  "num_layers": 2,
228
- "output_dim": "{model.decoder.interaction.intent_embedding_dim}",
229
  "row_normalized": true,
230
- "slot_graph_window": 1
 
231
  },
232
  "slot_classifier": {
233
  "_model_target_": "model.decoder.classifier.MLPClassifier",
234
  "dropout_rate": 0.4,
235
- "ignore_index": "{base.ignore_index}",
236
  "input_dim": 384,
237
  "mlp": [
238
  {
239
  "_model_target_": "torch.nn.Linear",
240
- "in_features": "{model.decoder.interaction.output_dim}",
241
- "out_features": "{model.decoder.interaction.output_dim}"
242
  },
243
  {
244
  "_model_target_": "torch.nn.LeakyReLU",
@@ -246,8 +247,8 @@
246
  },
247
  {
248
  "_model_target_": "torch.nn.Linear",
249
- "in_features": "{model.decoder.interaction.output_dim}",
250
- "out_features": "{base.slot_label_num}"
251
  }
252
  ],
253
  "mode": "slot",
@@ -268,7 +269,8 @@
268
  },
269
  "embedding": {
270
  "dropout_rate": 0.4,
271
- "embedding_dim": 128
 
272
  },
273
  "encoder_name": "self-attention-lstm",
274
  "lstm": {
@@ -283,6 +285,13 @@
283
  }
284
  },
285
  "return_dict": false,
 
 
 
 
 
 
 
286
  "tokenizer_class": "OpenSLUv1",
287
  "torch_dtype": "float32",
288
  "transformers_version": "4.25.1",
 
186
  "intent_classifier": {
187
  "_model_target_": "model.decoder.classifier.MLPClassifier",
188
  "dropout_rate": 0.4,
189
+ "ignore_index": -100,
190
  "input_dim": 384,
191
  "intent_label_num": 7,
192
  "loss_fn": {
 
195
  "mlp": [
196
  {
197
  "_model_target_": "torch.nn.Linear",
198
+ "in_features": 384,
199
  "out_features": 256
200
  },
201
  {
 
205
  {
206
  "_model_target_": "torch.nn.Linear",
207
  "in_features": 256,
208
+ "out_features": 7
209
  }
210
  ],
211
  "mode": "token-level-intent",
212
  "multi_threshold": 0.5,
213
  "return_sentence_level": true,
214
  "use_intent": true,
215
+ "use_multi": true,
216
  "weight": 0.2
217
  },
218
  "interaction": {
 
220
  "alpha": 0.2,
221
  "dropout_rate": 0.4,
222
  "hidden_dim": 256,
223
+ "input_dim": 384,
224
  "intent_embedding_dim": 256,
225
+ "intent_label_num": 7,
226
  "num_heads": 4,
227
  "num_layers": 2,
228
+ "output_dim": 256,
229
  "row_normalized": true,
230
+ "slot_graph_window": 1,
231
+ "slot_label_num": 72
232
  },
233
  "slot_classifier": {
234
  "_model_target_": "model.decoder.classifier.MLPClassifier",
235
  "dropout_rate": 0.4,
236
+ "ignore_index": -100,
237
  "input_dim": 384,
238
  "mlp": [
239
  {
240
  "_model_target_": "torch.nn.Linear",
241
+ "in_features": 256,
242
+ "out_features": 256
243
  },
244
  {
245
  "_model_target_": "torch.nn.LeakyReLU",
 
247
  },
248
  {
249
  "_model_target_": "torch.nn.Linear",
250
+ "in_features": 256,
251
+ "out_features": 72
252
  }
253
  ],
254
  "mode": "slot",
 
269
  },
270
  "embedding": {
271
  "dropout_rate": 0.4,
272
+ "embedding_dim": 128,
273
+ "vocab_size": 11758
274
  },
275
  "encoder_name": "self-attention-lstm",
276
  "lstm": {
 
285
  }
286
  },
287
  "return_dict": false,
288
+ "tokenizer": {
289
+ "_align_mode_": "fast",
290
+ "_padding_side_": "right",
291
+ "_tokenizer_name_": "word_tokenizer",
292
+ "add_special_tokens": false,
293
+ "max_length": 512
294
+ },
295
  "tokenizer_class": "OpenSLUv1",
296
  "torch_dtype": "float32",
297
  "transformers_version": "4.25.1",