Upload with huggingface_hub

README.md

@@ -0,0 +1,11 @@
+---
+inference: false
+---
+This is a model checkpoint for ["Should You Mask 15% in Masked Language Modeling"](https://arxiv.org/abs/2202.08005) [(code)](https://github.com/princeton-nlp/DinkyTrain.git).
+
+The original checkpoint is avaliable at [princeton-nlp/efficient_mlm_m0.15-801010](https://huggingface.co/princeton-nlp/efficient_mlm_m0.15-801010). Unfortunately this checkpoint depends on code that isn't part of the official `transformers`
+library. Additionally, the checkpoints contains unused weights due to a bug.
+
+This checkpoint fixes the unused weights issue and uses the `RobertaPreLayerNorm` model from the `transformers`
+library.
+

config.json

@@ -0,0 +1,27 @@
+{
+  "_name_or_path": "None",
+  "architectures": [
+    "RobertaPreLayerNormForMaskedLM"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 0,
+  "classifier_dropout": null,
+  "eos_token_id": 2,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 514,
+  "model_type": "roberta-prelayernorm",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.25.0.dev0",
+  "type_vocab_size": 1,
+  "use_cache": true,
+  "vocab_size": 50265
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34867fd9abff0b067351208ddb61722c2d1a46231f5760d3ae3d09c5e40e7bec
+size 1421793959

special_tokens_map.json

@@ -0,0 +1,15 @@
+{
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": {
+    "content": "<mask>",
+    "lstrip": true,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
+}

tokenizer_config.json

@@ -0,0 +1,16 @@
+{
+  "add_prefix_space": false,
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "errors": "replace",
+  "mask_token": "<mask>",
+  "model_max_length": 512,
+  "name_or_path": "princeton-nlp/efficient_mlm_m0.15-801010",
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "special_tokens_map_file": null,
+  "tokenizer_class": "RobertaTokenizer",
+  "trim_offsets": true,
+  "unk_token": "<unk>"
+}