Update preprocessor_config.json

preprocessor_config.json

@@ -5,22 +5,33 @@
   ],
   "model_type": "xlm-roberta",
   "tokenizer_class": "XLMRobertaTokenizer",
-  "max_length": 512,
-  "padding": "max_length",
-  "truncation": true,
-  "num_labels": 2,
-  "id2label": {
-    "0": "HUMAN",
-    "1": "AI"
+  "task_specific_params": {
+    "text-classification": {
+      "num_labels": 2,
+      "id2label": {
+        "0": "HUMAN",
+        "1": "AI"
+      },
+      "label2id": {
+        "HUMAN": 0,
+        "AI": 1
+      }
+    }
   },
-  "label2id": {
-    "HUMAN": 0,
-    "AI": 1
+  "text_config": {
+    "max_length": 512,
+    "padding": "max_length",
+    "truncation": true,
+    "return_tensors": "pt"
+  },
+  "preprocessing": {
+    "do_lower_case": false,
+    "strip_accents": false,
+    "add_special_tokens": true,
+    "padding": true,
+    "truncation": true,
+    "max_length": 512
   },
-  "do_lower_case": false,
-  "strip_accents": false,
-  "use_fast": true,
-  "add_prefix_space": true,
   "special_tokens": {
     "bos_token": "<s>",
     "eos_token": "</s>",
@@ -30,15 +41,31 @@
     "cls_token": "<s>",
     "mask_token": "<mask>"
   },
-  "clean_up_tokenization_spaces": true,
-  "model_max_length": 512,
-  "padding_side": "right",
-  "truncation_side": "right",
-  "return_attention_mask": true,
-  "return_token_type_ids": false,
-  "return_overflowing_tokens": false,
-  "return_special_tokens_mask": false,
-  "return_offsets_mapping": false,
-  "return_length": false,
-  "verbose": true
+  "tokenizer_settings": {
+    "clean_up_tokenization_spaces": true,
+    "model_max_length": 512,
+    "padding_side": "right",
+    "truncation_side": "right",
+    "return_attention_mask": true,
+    "return_token_type_ids": false
+  },
+  "inference_config": {
+    "return_all_scores": true,
+    "output_hidden_states": false,
+    "output_attentions": false,
+    "return_dict": true,
+    "problem_type": "single_label_classification"
+  },
+  "model_params": {
+    "attention_probs_dropout_prob": 0.1,
+    "hidden_dropout_prob": 0.1,
+    "hidden_size": 768,
+    "intermediate_size": 3072,
+    "max_position_embeddings": 514,
+    "num_attention_heads": 12,
+    "num_hidden_layers": 12,
+    "type_vocab_size": 1,
+    "vocab_size": 250002,
+    "layer_norm_eps": 1e-05
+  }
 }