Checkpoint 288k

README.md

@@ -0,0 +1,10 @@
+# robertuito-base-deacc (checkpoint 288k)
+**WORK IN PROGRESS**
+RoBERTa model trained on tweets. 
+For the time being, please use [this function](https://github.com/finiteautomata/finetune_vs_scratch/blob/main/finetune_vs_scratch/preprocessing.py#L13) before feeding it to the model. We still need to create a proper tokenizer for this model
+## Masked LM
+To test the masked LM, take into account that space is encoded inside SentencePiece's tokens. So, if you want to test
+```
+Este es un día<mask>
+```
+don't put a space between `día` and `<mask>`

config.json

@@ -0,0 +1,27 @@
+{
+  "architectures": [
+    "RobertaForMaskedLM"
+  ],
+  "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,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 130,
+  "model_type": "roberta",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.10.0.dev0",
+  "type_vocab_size": 1,
+  "use_cache": true,
+  "vocab_size": 30000
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d986fe9ee619e510b369d2bf38f425f88269dd1f59dea2335f9a48324a249c9
+size 435341035

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 @@
+{"bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": "<mask>", "special_tokens_map_file": "models/twerto-base-deacc-uncased/special_tokens_map.json", "name_or_path": "models/twerto-base-deacc-uncased", "tokenizer_class": "PreTrainedTokenizerFast"}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de42718d261b5254337f74e6d28257dba90997be4e4d2be3c91b0afdd96b9141
+size 2607