initial model

README.md

@@ -0,0 +1,28 @@
+---
+language: multilingual
+tags:
+- zero-shot-classification
+- nli
+- pytorch
+datasets:
+- mnli
+- xnli
+- anli
+license: mit
+pipeline_tag: zero-shot-classification
+widget:
+- text: "La película empezaba bien pero terminó siendo un desastre." 
+  labels: "positivo, negativo, neutral"
+- text: "¿A quién vas a votar en 2020?"
+  labels: "Europa, elecciones, política, ciencia, deportes"
+---
+
+### XLM-RoBERTa-large-XNLI-ANLI
+
+XLM-RoBERTa-large model finetunned over several NLI datasets, ready to use for zero-shot classification.
+
+Here are the accuracies for several test datasets:
+
+|                             | XNLI-es | XNLI-fr | ANLI-R1 | ANLI-R2 | ANLI-R3 |
+|-----------------------------|---------|---------|---------|---------|---------|
+| xlm-roberta-large-xnli-anli | 93.7% | 93.2% | 68.5%  | 53.6%  | 49.0%  |

config.json

@@ -0,0 +1,33 @@
+{
+  "architectures": [
+    "XLMRobertaForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 0,
+  "eos_token_id": 2,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 1024,
+  "id2label": {
+    "0": "contradiction",
+    "1": "neutral",
+    "2": "entailment"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "label2id": {
+    "contradiction": 0,
+    "entailment": 2,
+    "neutral": 1
+  },
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 514,
+  "model_type": "xlm-roberta",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
+  "output_past": true,
+  "pad_token_id": 1,
+  "type_vocab_size": 1,
+  "vocab_size": 250002
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:27f5d4fa3552fe9a4a1395dd30c7fd19d62fbc3d101bae31aa2e931fa5b08803
+size 2239747529

special_tokens_map.json

@@ -0,0 +1 @@
+{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": "<mask>"}

tokenizer_config.json

@@ -0,0 +1 @@
+{"model_max_length": 512}