janbakker commited on
Commit
78b9caa
1 Parent(s): a09bdb2

Upload BartForContextualGeneration

Browse files
Files changed (3) hide show
  1. config.json +178 -0
  2. generation_config.json +13 -0
  3. pytorch_model.bin +3 -0
config.json ADDED
@@ -0,0 +1,178 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/bart-base",
3
+ "accelerator": null,
4
+ "accumulate_grad_batches": 2,
5
+ "activation_dropout": 0.1,
6
+ "activation_function": "gelu",
7
+ "add_bias_logits": false,
8
+ "add_context": true,
9
+ "add_final_layer_norm": false,
10
+ "amp_backend": null,
11
+ "amp_level": null,
12
+ "architectures": [
13
+ "BartForContextualGeneration"
14
+ ],
15
+ "attention_dropout": 0.1,
16
+ "auto_lr_find": false,
17
+ "auto_scale_batch_size": false,
18
+ "auto_select_gpus": null,
19
+ "batch_size": 8,
20
+ "benchmark": null,
21
+ "bos_token_id": 0,
22
+ "check_val_every_n_epoch": 1,
23
+ "checkpoint": null,
24
+ "class_labels": {
25
+ "delete": 4,
26
+ "ignore": 0,
27
+ "merge": 3,
28
+ "rephrase": 1,
29
+ "split": 2
30
+ },
31
+ "classif_dropout": 0.1,
32
+ "classifier_dropout": 0.0,
33
+ "context_dir": "context/complex",
34
+ "context_doc_id": "pair_id",
35
+ "context_window": 5,
36
+ "d_model": 768,
37
+ "decoder_attention_heads": 12,
38
+ "decoder_ffn_dim": 3072,
39
+ "decoder_layerdrop": 0.0,
40
+ "decoder_layers": 6,
41
+ "decoder_start_token_id": 2,
42
+ "default_root_dir": null,
43
+ "detect_anomaly": false,
44
+ "devices": "2",
45
+ "dropout": 0.1,
46
+ "early_stopping": true,
47
+ "enable_checkpointing": true,
48
+ "enable_model_summary": true,
49
+ "enable_progress_bar": true,
50
+ "encoder_attention_heads": 12,
51
+ "encoder_ffn_dim": 3072,
52
+ "encoder_layerdrop": 0.0,
53
+ "encoder_layers": 6,
54
+ "eos_token_id": 2,
55
+ "eval_beams": null,
56
+ "eval_max_length": null,
57
+ "fast_dev_run": false,
58
+ "forced_bos_token_id": 0,
59
+ "forced_eos_token_id": 2,
60
+ "gpus": null,
61
+ "gradient_checkpointing": false,
62
+ "gradient_clip_algorithm": null,
63
+ "gradient_clip_val": null,
64
+ "id2label": {
65
+ "0": "LABEL_0",
66
+ "1": "LABEL_1",
67
+ "2": "LABEL_2"
68
+ },
69
+ "inference_mode": true,
70
+ "init_std": 0.02,
71
+ "ipus": null,
72
+ "is_encoder_decoder": true,
73
+ "label2id": {
74
+ "LABEL_0": 0,
75
+ "LABEL_1": 1,
76
+ "LABEL_2": 2
77
+ },
78
+ "limit_predict_batches": null,
79
+ "limit_test_batches": null,
80
+ "limit_train_batches": null,
81
+ "limit_val_batches": null,
82
+ "log_every_n_steps": 50,
83
+ "logger": true,
84
+ "longformer": false,
85
+ "lr": 2e-05,
86
+ "lr_scheduler": false,
87
+ "max_epochs": null,
88
+ "max_length": 1024,
89
+ "max_position_embeddings": 1024,
90
+ "max_samples": -1,
91
+ "max_steps": -1,
92
+ "max_time": null,
93
+ "min_epochs": null,
94
+ "min_steps": null,
95
+ "model_type": "context-bart",
96
+ "move_metrics_to_cpu": false,
97
+ "multiple_trainloader_mode": "max_size_cycle",
98
+ "name": "o-conbart",
99
+ "no_repeat_ngram_size": 3,
100
+ "normalize_before": false,
101
+ "normalize_embedding": true,
102
+ "num_beams": 4,
103
+ "num_hidden_layers": 6,
104
+ "num_nodes": 1,
105
+ "num_processes": null,
106
+ "num_sanity_val_steps": 2,
107
+ "op_col": "label",
108
+ "op_tokens": [
109
+ "<COPY>",
110
+ "<REPHRASE>",
111
+ "<SPLIT>",
112
+ "<MERGE>",
113
+ "<DELETE>"
114
+ ],
115
+ "overfit_batches": 0.0,
116
+ "pad_token_id": 1,
117
+ "plan_col": "labels",
118
+ "plan_loss": null,
119
+ "plan_prefix": false,
120
+ "plan_sep": false,
121
+ "plugins": null,
122
+ "precision": 32,
123
+ "prefix_only": false,
124
+ "profiler": null,
125
+ "project": "simplification",
126
+ "reading_lvl": null,
127
+ "reload_dataloaders_every_n_epochs": 0,
128
+ "replace_sampler_ddp": true,
129
+ "resume_from_checkpoint": null,
130
+ "save_dir": null,
131
+ "scale_embedding": false,
132
+ "sent_level": true,
133
+ "simple_context_dir": "context/simple",
134
+ "simple_context_doc_id": "pair_id",
135
+ "skip_val_gen": true,
136
+ "strategy": null,
137
+ "sync_batchnorm": false,
138
+ "task_specific_params": {
139
+ "summarization": {
140
+ "length_penalty": 1.0,
141
+ "max_length": 128,
142
+ "min_length": 12,
143
+ "num_beams": 4
144
+ },
145
+ "summarization_cnn": {
146
+ "length_penalty": 2.0,
147
+ "max_length": 142,
148
+ "min_length": 56,
149
+ "num_beams": 4
150
+ },
151
+ "summarization_xsum": {
152
+ "length_penalty": 1.0,
153
+ "max_length": 62,
154
+ "min_length": 11,
155
+ "num_beams": 6
156
+ }
157
+ },
158
+ "torch_dtype": "float32",
159
+ "tpu_cores": null,
160
+ "track_grad_norm": -1,
161
+ "train_check_interval": 0.2,
162
+ "train_data_dir": null,
163
+ "train_file": "merge/cochrane_sents_train.csv",
164
+ "train_split": 0.9,
165
+ "train_workers": 8,
166
+ "transformers_version": "4.29.1",
167
+ "use_cache": true,
168
+ "val_check_interval": null,
169
+ "val_file": "merge/cochrane_sents_val.csv",
170
+ "val_metric": "bleu",
171
+ "val_split": 0.05,
172
+ "val_workers": 8,
173
+ "valid_data_dir": null,
174
+ "vocab_size": 50275,
175
+ "wandb_id": null,
176
+ "x_col": "complex",
177
+ "y_col": "simple"
178
+ }
generation_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "decoder_start_token_id": 2,
5
+ "early_stopping": true,
6
+ "eos_token_id": 2,
7
+ "forced_bos_token_id": 0,
8
+ "forced_eos_token_id": 2,
9
+ "no_repeat_ngram_size": 3,
10
+ "num_beams": 4,
11
+ "pad_token_id": 1,
12
+ "transformers_version": "4.29.1"
13
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1fea7570f5f5b3f6014aea5c1d1076fcb323009f44cfd0b295a63cd6962db4e
3
+ size 621067613