Initial model

model/OxMarkupLM.pt/config.json

@@ -0,0 +1,58 @@
+{
+  "_name_or_path": "microsoft/markuplm-base",
+  "architectures": [
+    "MarkupLMForTokenClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 0,
+  "classifier_dropout": null,
+  "eos_token_id": 2,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "none",
+    "1": "title",
+    "2": "content",
+    "3": "author",
+    "4": "date",
+    "5": "header",
+    "6": "footer",
+    "7": "rail",
+    "8": "advertisement",
+    "9": "navigation"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "advertisement": 8,
+    "author": 3,
+    "content": 2,
+    "date": 4,
+    "footer": 6,
+    "header": 5,
+    "navigation": 9,
+    "none": 0,
+    "rail": 7,
+    "title": 1
+  },
+  "layer_norm_eps": 1e-05,
+  "max_depth": 50,
+  "max_position_embeddings": 514,
+  "max_xpath_subs_unit_embeddings": 1024,
+  "max_xpath_tag_unit_embeddings": 256,
+  "model_type": "markuplm",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "subs_pad_id": 1001,
+  "tag_pad_id": 216,
+  "torch_dtype": "float32",
+  "transformers_version": "4.41.2",
+  "type_vocab_size": 1,
+  "use_cache": true,
+  "vocab_size": 50267,
+  "xpath_unit_hidden_size": 32
+}

model/OxMarkupLM.pt/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0dac89414051dab71df4c5e141d1e0784c857d5ffe1b12a5681984c32f9138c9
+size 538520088

requirements.txt

@@ -0,0 +1,9 @@
+torch
+transformers
+numpy
+scikit-learn
+pandas
+requests
+beautifulsoup4
+glob2
+re