Iiro/bert_reviews

README.md

@@ -1,12 +1,28 @@
 ---
+license: apache-2.0
+base_model: distilbert-base-uncased
 tags:
 - generated_from_trainer
 datasets:
 - amazon_reviews_multi
+metrics:
+- accuracy
 model-index:
 - name: bert_reviews
- results: []
-pipeline_tag: text-classification
+ results:
+ - task:
+ name: Text Classification
+ type: text-classification
+ dataset:
+ name: amazon_reviews_multi
+ type: amazon_reviews_multi
+ config: en
+ split: test
+ args: en
+ metrics:
+ - name: Accuracy
+ type: accuracy
+ value: 0.6062
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -14,15 +30,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # bert_reviews
 
-This model was trained from scratch on the amazon_reviews_multi dataset.
+This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the amazon_reviews_multi dataset.
 It achieves the following results on the evaluation set:
-- eval_loss: 0.8196
-- eval_accuracy: 0.6544
-- eval_runtime: 39.7329
-- eval_samples_per_second: 125.84
-- eval_steps_per_second: 15.73
-- epoch: 0.52
-- step: 13000
+- Loss: 0.9204
+- Accuracy: 0.6062
 
 ## Model description
 
@@ -49,9 +60,35 @@ The following hyperparameters were used during training:
 - lr_scheduler_type: linear
 - training_steps: 20000
 
+### Training results
+
+| Training Loss | Epoch | Step | Validation Loss | Accuracy |
+|:-------------:|:-----:|:-----:|:---------------:|:--------:|
+| 0.8812 | 0.04 | 1000 | 0.9970 | 0.5738 |
+| 0.8495 | 0.08 | 2000 | 1.0120 | 0.569 |
+| 0.8067 | 0.12 | 3000 | 1.0442 | 0.5766 |
+| 0.7934 | 0.16 | 4000 | 1.0629 | 0.5772 |
+| 0.7845 | 0.2 | 5000 | 1.0236 | 0.5876 |
+| 0.9033 | 0.24 | 6000 | 0.9822 | 0.5774 |
+| 0.8993 | 0.28 | 7000 | 0.9693 | 0.5816 |
+| 0.9012 | 0.32 | 8000 | 1.0075 | 0.5738 |
+| 0.873 | 0.36 | 9000 | 0.9663 | 0.5886 |
+| 0.9376 | 0.4 | 10000 | 0.9447 | 0.5816 |
+| 0.9398 | 0.44 | 11000 | 0.9509 | 0.5802 |
+| 0.9402 | 0.48 | 12000 | 0.9561 | 0.5916 |
+| 0.9247 | 0.52 | 13000 | 0.9303 | 0.6008 |
+| 0.9247 | 0.56 | 14000 | 0.9241 | 0.5998 |
+| 0.9192 | 0.6 | 15000 | 0.9276 | 0.6104 |
+| 0.907 | 0.64 | 16000 | 0.9251 | 0.603 |
+| 0.9177 | 0.68 | 17000 | 0.9198 | 0.6056 |
+| 0.9129 | 0.72 | 18000 | 0.9167 | 0.6078 |
+| 0.8948 | 0.76 | 19000 | 0.9213 | 0.604 |
+| 0.906 | 0.8 | 20000 | 0.9204 | 0.6062 |
+
+
 ### Framework versions
 
 - Transformers 4.34.1
 - Pytorch 2.1.0+cu118
 - Datasets 2.14.6
-- Tokenizers 0.14.1
+- Tokenizers 0.14.1

config.json

@@ -1,9 +1,13 @@
 {
+ "_name_or_path": "distilbert-base-uncased",
+ "activation": "gelu",
  "architectures": [
- "Bert"
+ "DistilBertForSequenceClassification"
  ],
- "hidden_bert": 768,
- "hidden_ffnn": 24,
+ "attention_dropout": 0.1,
+ "dim": 768,
+ "dropout": 0.1,
+ "hidden_dim": 3072,
  "id2label": {
  "0": "LABEL_0",
  "1": "LABEL_1",
@@ -11,6 +15,7 @@
  "3": "LABEL_3",
  "4": "LABEL_4"
  },
+ "initializer_range": 0.02,
  "label2id": {
  "LABEL_0": 0,
  "LABEL_1": 1,
@@ -18,7 +23,17 @@
  "LABEL_3": 3,
  "LABEL_4": 4
  },
+ "max_position_embeddings": 512,
+ "model_type": "distilbert",
+ "n_heads": 12,
+ "n_layers": 6,
+ "pad_token_id": 0,
+ "problem_type": "single_label_classification",
+ "qa_dropout": 0.1,
+ "seq_classif_dropout": 0.2,
+ "sinusoidal_pos_embds": false,
+ "tie_weights_": true,
  "torch_dtype": "float32",
  "transformers_version": "4.34.1",
- "vocab_size": 28996
+ "vocab_size": 30522
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:459282ffb2884070a380ef30845b262bf77813534ff708275ca9e6853bd97625
-size 433384758
+oid sha256:86ee65e834160aace6c12aa0d503cd6099f9b07664b12a565dbb52d046636268
+size 267865194

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c277da8f4bedc9a2e1636c9c4bcac05169afdd2c0e9f7cb4183d1486ada8cfb0
+oid sha256:d9a7349f61988d6e6a3c339b566c8b29407e9359421ec483dcd69ca1f45a41f5
 size 4472