avrecum/bert-base-uncased-finetuned-refusal-classifier-2024-12-01-03-07-59

README.md

@@ -0,0 +1,72 @@
+---
+library_name: transformers
+license: apache-2.0
+base_model: bert-base-uncased
+tags:
+- generated_from_trainer
+metrics:
+- f1
+- accuracy
+model-index:
+- name: refusal-classifier
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# refusal-classifier
+
+This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.7340
+- F1: 0.9243
+- Roc Auc: 0.9956
+- Accuracy: 0.9246
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 1e-05
+- train_batch_size: 384
+- eval_batch_size: 384
+- seed: 42
+- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: linear
+- lr_scheduler_warmup_steps: 500
+- num_epochs: 3
+- label_smoothing_factor: 0.1
+
+### Training results
+
+| Training Loss | Epoch  | Step | Validation Loss | F1     | Roc Auc | Accuracy |
+|:-------------:|:------:|:----:|:---------------:|:------:|:-------:|:--------:|
+| No log        | 0.3817 | 100  | 2.4345          | 0.1373 | 0.7411  | 0.1877   |
+| No log        | 0.7634 | 200  | 1.9614          | 0.5462 | 0.9299  | 0.5769   |
+| No log        | 1.1450 | 300  | 1.2999          | 0.7631 | 0.9739  | 0.77     |
+| No log        | 1.5267 | 400  | 0.9551          | 0.8506 | 0.9870  | 0.8515   |
+| 1.716         | 1.9084 | 500  | 0.8384          | 0.8785 | 0.9923  | 0.8785   |
+| 1.716         | 2.2901 | 600  | 0.7623          | 0.9156 | 0.9948  | 0.9162   |
+| 1.716         | 2.6718 | 700  | 0.7340          | 0.9243 | 0.9956  | 0.9246   |
+
+
+### Framework versions
+
+- Transformers 4.46.3
+- Pytorch 2.1.0+cu118
+- Datasets 3.1.0
+- Tokenizers 0.20.3

config.json

@@ -0,0 +1,57 @@
+{
+  "_name_or_path": "bert-base-uncased",
+  "architectures": [
+    "BertForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "[Should Not Do] Legal Compliance/Illegal",
+    "1": "[Should Not Do] Information Hazards",
+    "2": "[Should Not Do] Intellectual Property Rights",
+    "3": "[Should Not Do] Privacy",
+    "4": "[Should Not Do] NSFW Content",
+    "5": "[Cannot Do] Modalities",
+    "6": "[Cannot Do] Skills",
+    "7": "[Cannot Do] Missing Information \u2013 Knowledge Cutoff",
+    "8": "[Cannot Do] Missing Information \u2013 Unknown Unknown Information",
+    "9": "[Cannot Do] Training Data Limits",
+    "10": "[Cannot Do] Missing Information \u2013 Missing Context",
+    "11": "[Cannot Do] Missing Identity",
+    "12": "[Cannot Do] Invalid Premise"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "[Cannot Do] Invalid Premise": 12,
+    "[Cannot Do] Missing Identity": 11,
+    "[Cannot Do] Missing Information \u2013 Knowledge Cutoff": 7,
+    "[Cannot Do] Missing Information \u2013 Missing Context": 10,
+    "[Cannot Do] Missing Information \u2013 Unknown Unknown Information": 8,
+    "[Cannot Do] Modalities": 5,
+    "[Cannot Do] Skills": 6,
+    "[Cannot Do] Training Data Limits": 9,
+    "[Should Not Do] Information Hazards": 1,
+    "[Should Not Do] Intellectual Property Rights": 2,
+    "[Should Not Do] Legal Compliance/Illegal": 0,
+    "[Should Not Do] NSFW Content": 4,
+    "[Should Not Do] Privacy": 3
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "problem_type": "single_label_classification",
+  "torch_dtype": "float32",
+  "transformers_version": "4.46.3",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 30522
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:adac66880ce987efe150e25a52718926a5b33d39e407b75b4c7dfbe26801d056
+size 437992484

special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "[CLS]",
+  "do_lower_case": true,
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08f3e5c6ed6a47b2ec31a33e269027e460eb04602a1912286c2a5c8e32bd0ce3
+size 5240