rhaymison/bert-squad-portuguese

README.md

@@ -1,6 +1,6 @@
 ---
-license: apache-2.0
-base_model: bert-base-uncased
+license: mit
+base_model: neuralmind/bert-base-portuguese-cased
 tags:
 - generated_from_trainer
 model-index:
@@ -13,9 +13,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # finetune-BERT-squad
 
-This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on an unknown dataset.
+This model is a fine-tuned version of [neuralmind/bert-base-portuguese-cased](https://huggingface.co/neuralmind/bert-base-portuguese-cased) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.6799
+- Loss: 1.0687
 
 ## Model description
 
@@ -40,48 +40,22 @@ The following hyperparameters were used during training:
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 3
+- num_epochs: 2
 
 ### Training results
 
 | Training Loss | Epoch | Step  | Validation Loss |
 |:-------------:|:-----:|:-----:|:---------------:|
-| 1.3183        | 0.08  | 500   | 1.6474          |
-| 1.335         | 0.16  | 1000  | 1.6289          |
-| 1.3245        | 0.24  | 1500  | 1.6070          |
-| 1.3055        | 0.32  | 2000  | 1.6563          |
-| 1.3085        | 0.41  | 2500  | 1.5719          |
-| 1.2918        | 0.49  | 3000  | 1.6650          |
-| 1.2827        | 0.57  | 3500  | 1.5858          |
-| 1.3122        | 0.65  | 4000  | 1.5870          |
-| 1.2586        | 0.73  | 4500  | 1.5693          |
-| 1.2943        | 0.81  | 5000  | 1.5659          |
-| 1.2573        | 0.89  | 5500  | 1.5207          |
-| 1.2564        | 0.97  | 6000  | 1.5355          |
-| 1.0725        | 1.06  | 6500  | 1.6185          |
-| 0.9957        | 1.14  | 7000  | 1.5904          |
-| 0.972         | 1.22  | 7500  | 1.6978          |
-| 0.9884        | 1.3   | 8000  | 1.6270          |
-| 1.02          | 1.38  | 8500  | 1.5872          |
-| 1.0234        | 1.46  | 9000  | 1.6245          |
-| 1.0162        | 1.54  | 9500  | 1.6201          |
-| 0.9969        | 1.62  | 10000 | 1.6262          |
-| 0.998         | 1.7   | 10500 | 1.6099          |
-| 0.9888        | 1.79  | 11000 | 1.6151          |
-| 1.0034        | 1.87  | 11500 | 1.5837          |
-| 1.0127        | 1.95  | 12000 | 1.5823          |
-| 0.8975        | 2.03  | 12500 | 1.6466          |
-| 0.7488        | 2.11  | 13000 | 1.7307          |
-| 0.7538        | 2.19  | 13500 | 1.7256          |
-| 0.7847        | 2.27  | 14000 | 1.7131          |
-| 0.7876        | 2.35  | 14500 | 1.6858          |
-| 0.7618        | 2.44  | 15000 | 1.7329          |
-| 0.761         | 2.52  | 15500 | 1.7251          |
-| 0.7887        | 2.6   | 16000 | 1.7077          |
-| 0.8002        | 2.68  | 16500 | 1.6861          |
-| 0.8038        | 2.76  | 17000 | 1.6818          |
-| 0.802         | 2.84  | 17500 | 1.6932          |
-| 0.8076        | 2.92  | 18000 | 1.6799          |
+| 1.4612        | 0.19  | 1000  | 1.3505          |
+| 1.2999        | 0.37  | 2000  | 1.2518          |
+| 1.2234        | 0.56  | 3000  | 1.1492          |
+| 1.1824        | 0.74  | 4000  | 1.1181          |
+| 1.1169        | 0.93  | 5000  | 1.0850          |
+| 0.8933        | 1.11  | 6000  | 1.0925          |
+| 0.9004        | 1.3   | 7000  | 1.0849          |
+| 0.8706        | 1.49  | 8000  | 1.0976          |
+| 0.8617        | 1.67  | 9000  | 1.0779          |
+| 0.8756        | 1.86  | 10000 | 1.0687          |
 
 
 ### Framework versions

config.json

@@ -1,11 +1,11 @@
 {
-  "_name_or_path": "bert-base-uncased",
+  "_name_or_path": "neuralmind/bert-base-portuguese-cased",
   "architectures": [
     "BertForQuestionAnswering"
   ],
   "attention_probs_dropout_prob": 0.1,
   "classifier_dropout": null,
-  "gradient_checkpointing": false,
+  "directionality": "bidi",
   "hidden_act": "gelu",
   "hidden_dropout_prob": 0.1,
   "hidden_size": 768,
@@ -16,11 +16,17 @@
   "model_type": "bert",
   "num_attention_heads": 12,
   "num_hidden_layers": 12,
+  "output_past": true,
   "pad_token_id": 0,
+  "pooler_fc_size": 768,
+  "pooler_num_attention_heads": 12,
+  "pooler_num_fc_layers": 3,
+  "pooler_size_per_head": 128,
+  "pooler_type": "first_token_transform",
   "position_embedding_type": "absolute",
   "torch_dtype": "float32",
   "transformers_version": "4.38.2",
   "type_vocab_size": 2,
   "use_cache": true,
-  "vocab_size": 30522
+  "vocab_size": 29794
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:896aed7372717b83ac4d2a721f994fa4c07f2c3d4e44380a4273c41b37799353
-size 435596088
+oid sha256:87ec2dff3cb2f517daa41b039e9f602cca4ee48fe9ebd8b44402493f62f5e9b2
+size 433359664

runs/Mar22_10-20-07_862ef8595a2f/events.out.tfevents.1711102849.862ef8595a2f.3477.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4771e7cd36d1a1affdd177a7fbbdadb2ff98a8e4f630b1800ab793979fd5eab8
+size 12256

tokenizer_config.json

@@ -43,9 +43,11 @@
   },
   "clean_up_tokenization_spaces": true,
   "cls_token": "[CLS]",
-  "do_lower_case": true,
+  "do_basic_tokenize": true,
+  "do_lower_case": false,
   "mask_token": "[MASK]",
-  "model_max_length": 512,
+  "model_max_length": 1000000000000000019884624838656,
+  "never_split": null,
   "pad_token": "[PAD]",
   "sep_token": "[SEP]",
   "strip_accents": null,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:be0af72b66ab726ffbf9d140a20b31ae5efebdff236f0cca3533941ce559917d
+oid sha256:2ab13811b8c0b5969b92adf753312400b050458a24e6b53c71ce5a8789f4dfaf
 size 4856