Update

config.json

@@ -8,7 +8,7 @@
     "PegasusForConditionalGeneration"
   ],
   "attention_dropout": 0.1,
-  "bos_token_id": 0,
+  "bos_token_id": 101,
   "classif_dropout": 0.0,
   "classifier_dropout": 0.0,
   "d_model": 768,
@@ -16,13 +16,13 @@
   "decoder_ffn_dim": 3072,
   "decoder_layerdrop": 0.0,
   "decoder_layers": 12,
-  "decoder_start_token_id": 0,
+  "decoder_start_token_id": 101,
   "dropout": 0.1,
   "encoder_attention_heads": 12,
   "encoder_ffn_dim": 3072,
   "encoder_layerdrop": 0.0,
   "encoder_layers": 12,
-  "eos_token_id": 102,
+  "eos_token_id": 1,
   "extra_pos_embeddings": 1,
   "force_bos_token_to_be_generated": false,
   "forced_eos_token_id": 1,
@@ -39,9 +39,8 @@
     "LABEL_1": 1,
     "LABEL_2": 2
   },
-  "length_penalty": 0.8,
   "max_length": 256,
-  "max_position_embeddings": 512,
+  "max_position_embeddings": 1024,
   "model_type": "pegasus",
   "normalize_before": true,
   "normalize_embedding": false,
@@ -50,7 +49,6 @@
   "scale_embedding": true,
   "static_position_embeddings": true,
   "transformers_version": "4.6.0",
-  "tokenizer_class": "BertTokenizer",
   "use_cache": true,
-  "vocab_size": 21130
+  "vocab_size": 21128
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9d899d06f51522a107365079361235f619b1784015e7f6e5af24e1d1b1c4b241
-size 862251247
+oid sha256:89ac0b2cbcb21b7dc45fae80866753b0326d6e38deee26e8e84a01833f0f7ef4
+size 865390831

tokenizer_config.json

@@ -1 +1 @@
-{"do_lower_case": true, "do_basic_tokenize": true, "never_split": null, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK1]", "tokenize_chinese_chars": true, "strip_accents": null, "special_tokens_map_file": null, "tokenizer_file": null, "name_or_path": "pegasus"}
+{"do_lower_case": true, "do_basic_tokenize": true, "never_split": null, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "special_tokens_map_file": null, "tokenizer_file": null, "name_or_path": "pegasus"}

vocab.txt

@@ -102,8 +102,6 @@
 [CLS]
 [SEP]
 [MASK]
-[MASK1]
-[MASK2]
 <S>
 <T>
 !