init

README.md

@@ -1,3 +1,37 @@
 ---
-license: mit
+language: zh
 ---
+
+# ERNIE-3.0-xbase-zh
+
+## Introduction
+
+ERNIE 3.0: Large-scale Knowledge Enhanced Pre-training for Language Understanding and Generation
+More detail: https://arxiv.org/abs/2107.02137
+
+## Released Model Info
+
+This released pytorch model is converted from the officially released PaddlePaddle ERNIE model and 
+a series of experiments have been conducted to check the accuracy of the conversion.
+
+- Official PaddlePaddle ERNIE repo:https://paddlenlp.readthedocs.io/zh/latest/model_zoo/transformers/ERNIE/contents.html
+- Pytorch Conversion repo:  https://github.com/nghuyong/ERNIE-Pytorch
+
+## How to use
+```Python
+from transformers import BertTokenizer, ErnieModel
+
+tokenizer = BertTokenizer.from_pretrained("nghuyong/ernie-3.0-xbase-zh")
+model = ErnieModel.from_pretrained("nghuyong/ernie-3.0-xbase-zh")
+```
+
+## Citation
+
+```bibtex
+@article{sun2021ernie,
+  title={Ernie 3.0: Large-scale knowledge enhanced pre-training for language understanding and generation},
+  author={Sun, Yu and Wang, Shuohuan and Feng, Shikun and Ding, Siyu and Pang, Chao and Shang, Junyuan and Liu, Jiaxiang and Chen, Xuyi and Zhao, Yanbin and Lu, Yuxiang and others},
+  journal={arXiv preprint arXiv:2107.02137},
+  year={2021}
+}
+```

config.json

@@ -0,0 +1,20 @@
+{
+    "attention_probs_dropout_prob": 0.1,
+    "hidden_act": "gelu",
+    "hidden_dropout_prob": 0.1,
+    "intermediate_size": 4096,
+    "hidden_size": 1024,
+    "initializer_range": 0.02,
+    "max_position_embeddings": 2048,
+    "num_attention_heads": 16,
+    "num_hidden_layers": 20,
+    "task_type_vocab_size": 16,
+    "type_vocab_size": 4,
+    "use_task_id": true,
+    "vocab_size": 40000,
+    "layer_norm_eps": 1e-05,
+    "model_type": "ernie",
+    "architectures": [
+        "ErnieForMaskedLM"
+    ]
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f75c345d47c02d321acd1a7b90f184a803fd9481c09f3b5e705cfd134d728b1d
+size 1188688423