Init

README.md

@@ -0,0 +1,8 @@
+# LayoutLMv2
+**Multimodal (text + layout/format + image) pre-training for document AI**
+
+## Introduction
+LayoutLMv2 is an improved version of LayoutLM with new pre-training tasks to model the interaction among text, layout, and image in a single multi-modal framework. It outperforms strong baselines and achieves new state-of-the-art results on a wide variety of downstream visually-rich document understanding tasks, including , including FUNSD (0.7895 → 0.8420), CORD (0.9493 → 0.9601), SROIE (0.9524 → 0.9781), Kleister-NDA (0.834 → 0.852), RVL-CDIP (0.9443 → 0.9564), and DocVQA (0.7295 → 0.8672).
+
+[LayoutLMv2: Multi-modal Pre-training for Visually-Rich Document Understanding](https://arxiv.org/abs/2012.14740)
+Yang Xu, Yiheng Xu, Tengchao Lv, Lei Cui, Furu Wei, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Wanxiang Che, Min Zhang, Lidong Zhou, [ACL 2021](#)

config.json

@@ -0,0 +1,35 @@
+{
+  "attention_probs_dropout_prob": 0.1,
+  "coordinate_size": 171,
+  "fast_qkv": false,
+  "gradient_checkpointing": false,
+  "has_relative_attention_bias": true,
+  "has_spatial_attention_bias": true,
+  "has_visual_segment_embedding": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 1024,
+  "image_feature_pool_shape": [
+    7,
+    7,
+    256
+  ],
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "layer_norm_eps": 1e-12,
+  "max_2d_position_embeddings": 1024,
+  "max_position_embeddings": 512,
+  "max_rel_2d_pos": 256,
+  "max_rel_pos": 128,
+  "model_type": "layoutlmv2",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
+  "output_past": true,
+  "pad_token_id": 0,
+  "shape_size": 170,
+  "rel_2d_pos_bins": 64,
+  "rel_pos_bins": 32,
+  "transformers_version": "4.5.1",
+  "type_vocab_size": 2,
+  "vocab_size": 30522
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94fa102e8f8b6ad05e2c3ec95ed995ff2418df7d43aa462197665b6751cefc02
+size 1705962979