josh-oo commited on
Commit
c77f7fc
1 Parent(s): 1bf3259

Distilled layers: [r, r, -1] (0.75)

Browse files
Files changed (2) hide show
  1. config.json +45 -33
  2. pytorch_model.bin +2 -2
config.json CHANGED
@@ -1,6 +1,5 @@
1
  {
2
  "_commit_hash": null,
3
- "_name_or_path": "calibrated-decoder",
4
  "architectures": [
5
  "EncoderDecoderModel"
6
  ],
@@ -100,66 +99,74 @@
100
  },
101
  "decoder_start_token_id": 0,
102
  "encoder": {
103
- "_name_or_path": "bert-base-multilingual-cased",
 
 
 
 
104
  "add_cross_attention": false,
 
105
  "architectures": [
106
- "BertForMaskedLM"
107
  ],
108
- "attention_probs_dropout_prob": 0.1,
109
  "bad_words_ids": null,
110
  "begin_suppress_tokens": null,
111
- "bos_token_id": null,
112
  "chunk_size_feed_forward": 0,
113
- "classifier_dropout": null,
 
114
  "cross_attention_hidden_size": null,
 
 
 
 
 
115
  "decoder_start_token_id": null,
116
- "directionality": "bidi",
117
  "diversity_penalty": 0.0,
118
  "do_sample": false,
 
119
  "early_stopping": false,
 
 
 
 
120
  "encoder_no_repeat_ngram_size": 0,
121
- "eos_token_id": null,
122
  "exponential_decay_length_penalty": null,
123
  "finetuning_task": null,
124
  "forced_bos_token_id": null,
125
- "forced_eos_token_id": null,
126
- "hidden_act": "gelu",
127
- "hidden_dropout_prob": 0.1,
128
- "hidden_size": 768,
129
  "id2label": {
130
  "0": "LABEL_0",
131
- "1": "LABEL_1"
 
132
  },
133
- "initializer_range": 0.02,
134
- "intermediate_size": 3072,
135
  "is_decoder": false,
136
- "is_encoder_decoder": false,
137
  "label2id": {
138
  "LABEL_0": 0,
139
- "LABEL_1": 1
 
140
  },
141
- "layer_norm_eps": 1e-12,
142
  "length_penalty": 1.0,
143
- "max_length": 20,
144
- "max_position_embeddings": 512,
145
  "min_length": 0,
146
- "model_type": "bert",
147
  "no_repeat_ngram_size": 0,
148
- "num_attention_heads": 12,
 
149
  "num_beam_groups": 1,
150
- "num_beams": 1,
151
  "num_hidden_layers": 12,
152
  "num_return_sequences": 1,
153
  "output_attentions": false,
154
  "output_hidden_states": false,
 
155
  "output_scores": false,
156
- "pad_token_id": 0,
157
- "pooler_fc_size": 768,
158
- "pooler_num_attention_heads": 12,
159
- "pooler_num_fc_layers": 3,
160
- "pooler_size_per_head": 128,
161
- "pooler_type": "first_token_transform",
162
- "position_embedding_type": "absolute",
163
  "prefix": null,
164
  "problem_type": null,
165
  "pruned_heads": {},
@@ -167,9 +174,15 @@
167
  "repetition_penalty": 1.0,
168
  "return_dict": true,
169
  "return_dict_in_generate": false,
 
170
  "sep_token_id": null,
 
171
  "suppress_tokens": null,
172
- "task_specific_params": null,
 
 
 
 
173
  "temperature": 1.0,
174
  "tf_legacy_loss": false,
175
  "tie_encoder_decoder": false,
@@ -180,11 +193,10 @@
180
  "torch_dtype": null,
181
  "torchscript": false,
182
  "transformers_version": "4.25.1",
183
- "type_vocab_size": 2,
184
  "typical_p": 1.0,
185
  "use_bfloat16": false,
186
  "use_cache": true,
187
- "vocab_size": 119547
188
  },
189
  "eos_token_id": 2,
190
  "is_encoder_decoder": true,
 
1
  {
2
  "_commit_hash": null,
 
3
  "architectures": [
4
  "EncoderDecoderModel"
5
  ],
 
99
  },
100
  "decoder_start_token_id": 0,
101
  "encoder": {
102
+ "_name_or_path": "facebook/mbart-large-cc25",
103
+ "_num_labels": 3,
104
+ "activation_dropout": 0.0,
105
+ "activation_function": "gelu",
106
+ "add_bias_logits": false,
107
  "add_cross_attention": false,
108
+ "add_final_layer_norm": true,
109
  "architectures": [
110
+ "MBartForConditionalGeneration"
111
  ],
112
+ "attention_dropout": 0.0,
113
  "bad_words_ids": null,
114
  "begin_suppress_tokens": null,
115
+ "bos_token_id": 0,
116
  "chunk_size_feed_forward": 0,
117
+ "classif_dropout": 0.0,
118
+ "classifier_dropout": 0.0,
119
  "cross_attention_hidden_size": null,
120
+ "d_model": 1024,
121
+ "decoder_attention_heads": 16,
122
+ "decoder_ffn_dim": 4096,
123
+ "decoder_layerdrop": 0.0,
124
+ "decoder_layers": 12,
125
  "decoder_start_token_id": null,
 
126
  "diversity_penalty": 0.0,
127
  "do_sample": false,
128
+ "dropout": 0.1,
129
  "early_stopping": false,
130
+ "encoder_attention_heads": 16,
131
+ "encoder_ffn_dim": 4096,
132
+ "encoder_layerdrop": 0.0,
133
+ "encoder_layers": 12,
134
  "encoder_no_repeat_ngram_size": 0,
135
+ "eos_token_id": 2,
136
  "exponential_decay_length_penalty": null,
137
  "finetuning_task": null,
138
  "forced_bos_token_id": null,
139
+ "forced_eos_token_id": 2,
 
 
 
140
  "id2label": {
141
  "0": "LABEL_0",
142
+ "1": "LABEL_1",
143
+ "2": "LABEL_2"
144
  },
145
+ "init_std": 0.02,
 
146
  "is_decoder": false,
147
+ "is_encoder_decoder": true,
148
  "label2id": {
149
  "LABEL_0": 0,
150
+ "LABEL_1": 1,
151
+ "LABEL_2": 2
152
  },
 
153
  "length_penalty": 1.0,
154
+ "max_length": 1024,
155
+ "max_position_embeddings": 1024,
156
  "min_length": 0,
157
+ "model_type": "mbart",
158
  "no_repeat_ngram_size": 0,
159
+ "normalize_before": true,
160
+ "normalize_embedding": true,
161
  "num_beam_groups": 1,
162
+ "num_beams": 5,
163
  "num_hidden_layers": 12,
164
  "num_return_sequences": 1,
165
  "output_attentions": false,
166
  "output_hidden_states": false,
167
+ "output_past": true,
168
  "output_scores": false,
169
+ "pad_token_id": 1,
 
 
 
 
 
 
170
  "prefix": null,
171
  "problem_type": null,
172
  "pruned_heads": {},
 
174
  "repetition_penalty": 1.0,
175
  "return_dict": true,
176
  "return_dict_in_generate": false,
177
+ "scale_embedding": true,
178
  "sep_token_id": null,
179
+ "static_position_embeddings": false,
180
  "suppress_tokens": null,
181
+ "task_specific_params": {
182
+ "translation_en_to_ro": {
183
+ "decoder_start_token_id": 250020
184
+ }
185
+ },
186
  "temperature": 1.0,
187
  "tf_legacy_loss": false,
188
  "tie_encoder_decoder": false,
 
193
  "torch_dtype": null,
194
  "torchscript": false,
195
  "transformers_version": "4.25.1",
 
196
  "typical_p": 1.0,
197
  "use_bfloat16": false,
198
  "use_cache": true,
199
+ "vocab_size": 250027
200
  },
201
  "eos_token_id": 2,
202
  "is_encoder_decoder": true,
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:08bf677a1db9a7596d394f0ae22b06dd3d8055e601b4036c912e1b0a1ef5f34d
3
- size 636511709
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:642f978bc5677ddf27b7b58631dabe5018e1eff27c4b13dc59c2a387b9710a3a
3
+ size 639665189