init

README.md

@@ -0,0 +1,28 @@
+---
+language:
+- zh
+license:
+- apache-2.0
+---
+```python
+import jieba_fast
+from transformers import BertTokenizer
+from transformers import BigBirdModel
+class JiebaTokenizer(BertTokenizer):
+    def __init__(
+        self, pre_tokenizer=lambda x: jieba_fast.cut(x, HMM=False), *args, **kwargs
+    ):
+        super().__init__(*args, **kwargs)
+        self.pre_tokenizer = pre_tokenizer
+    def _tokenize(self, text, *arg, **kwargs):
+        split_tokens = []
+        for text in self.pre_tokenizer(text):
+            if text in self.vocab:
+                split_tokens.append(text)
+            else:
+                split_tokens.extend(super()._tokenize(text))
+        return split_tokens
+model = BigBirdModel.from_pretrained('Lowin/chinese-bigbird-base')
+tokenizer = JiebaTokenizer.from_pretrained('Lowin/chinese-bigbird-base')
+```
+https://github.com/LowinLi/chinese-bigbird

config.json

@@ -0,0 +1,30 @@
+{
+"architectures": [
+    "BigBirdForMaskedLM"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "attention_type": "block_sparse",
+  "block_size": 64,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu_fast",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 4096,
+  "model_type": "big_bird",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "num_random_blocks": 3,
+  "pad_token_id": 0,
+  "rescale_embeddings": false,
+  "sep_token_id": 66,
+  "transformers_version": "4.6.1",
+  "type_vocab_size": 2,
+  "use_bias": true,
+  "use_cache": true,
+  "vocab_size": 40000
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ecfad0fa352402146a2d62a81c8b2bb13bdaee1b093099b835290e0a5700bd71
+size 480701374