yunase/Bert_QLoRA_emotion_detection

README.md

@@ -0,0 +1,88 @@
+---
+license: apache-2.0
+library_name: peft
+tags:
+- generated_from_trainer
+base_model: bert-base-uncased
+metrics:
+- accuracy
+model-index:
+- name: emotions_bert_qlora
+  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. -->
+
+# emotions_bert_qlora
+
+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: 0.5184
+- F1 Micro: 0.6855
+- F1 Macro: 0.6029
+- Accuracy: 0.2129
+
+## 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: 0.0001
+- train_batch_size: 128
+- eval_batch_size: 128
+- seed: 42
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- num_epochs: 10
+- mixed_precision_training: Native AMP
+
+### Training results
+
+| Training Loss | Epoch  | Step | Validation Loss | F1 Micro | F1 Macro | Accuracy |
+|:-------------:|:------:|:----:|:---------------:|:--------:|:--------:|:--------:|
+| 0.8094        | 0.4082 | 20   | 0.7752          | 0.4570   | 0.1865   | 0.0990   |
+| 0.7206        | 0.8163 | 40   | 0.6446          | 0.5938   | 0.3669   | 0.1230   |
+| 0.6381        | 1.2245 | 60   | 0.6094          | 0.6163   | 0.4260   | 0.1159   |
+| 0.5836        | 1.6327 | 80   | 0.5757          | 0.6345   | 0.4702   | 0.1553   |
+| 0.5631        | 2.0408 | 100  | 0.5504          | 0.6674   | 0.5290   | 0.1922   |
+| 0.5132        | 2.4490 | 120  | 0.5355          | 0.6686   | 0.5440   | 0.1922   |
+| 0.504         | 2.8571 | 140  | 0.5260          | 0.6710   | 0.5843   | 0.1780   |
+| 0.4853        | 3.2653 | 160  | 0.5209          | 0.6689   | 0.5854   | 0.1858   |
+| 0.4494        | 3.6735 | 180  | 0.5165          | 0.6759   | 0.5942   | 0.1909   |
+| 0.4663        | 4.0816 | 200  | 0.5146          | 0.6834   | 0.5900   | 0.2      |
+| 0.4361        | 4.4898 | 220  | 0.5073          | 0.6876   | 0.6013   | 0.2214   |
+| 0.4278        | 4.8980 | 240  | 0.5112          | 0.6830   | 0.6018   | 0.2045   |
+| 0.4175        | 5.3061 | 260  | 0.5119          | 0.6775   | 0.5929   | 0.2013   |
+| 0.4           | 5.7143 | 280  | 0.5118          | 0.6846   | 0.5961   | 0.2039   |
+| 0.4088        | 6.1224 | 300  | 0.5100          | 0.6819   | 0.6015   | 0.2117   |
+| 0.3811        | 6.5306 | 320  | 0.5211          | 0.6759   | 0.5952   | 0.1916   |
+| 0.389         | 6.9388 | 340  | 0.5152          | 0.6770   | 0.6004   | 0.1974   |
+| 0.3806        | 7.3469 | 360  | 0.5183          | 0.6853   | 0.6029   | 0.2123   |
+| 0.3686        | 7.7551 | 380  | 0.5222          | 0.6739   | 0.5940   | 0.2019   |
+| 0.3719        | 8.1633 | 400  | 0.5207          | 0.6777   | 0.5941   | 0.2091   |
+| 0.3538        | 8.5714 | 420  | 0.5242          | 0.6765   | 0.5943   | 0.2006   |
+| 0.3591        | 8.9796 | 440  | 0.5247          | 0.6784   | 0.5953   | 0.2091   |
+| 0.3554        | 9.3878 | 460  | 0.5259          | 0.6815   | 0.5978   | 0.2142   |
+| 0.346         | 9.7959 | 480  | 0.5256          | 0.6809   | 0.5976   | 0.2142   |
+
+
+### Framework versions
+
+- PEFT 0.10.0
+- Transformers 4.40.2
+- Pytorch 2.2.1+cu121
+- Datasets 2.19.1
+- Tokenizers 0.19.1

adapter_config.json

@@ -0,0 +1,33 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "bert-base-uncased",
+  "bias": "lora_only",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 256,
+  "lora_dropout": 0.01,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": [
+    "classifier"
+  ],
+  "peft_type": "LORA",
+  "r": 128,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "query",
+    "dense",
+    "value",
+    "key"
+  ],
+  "task_type": "SEQ_CLS",
+  "use_dora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a46af028939a37e3fc40e203e7559295b3f9f0c989bd0ac23739e38a5d8e1315
+size 85775908

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": true,
+  "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 sha256:8fe3e52c9c6043bc5d40716495aaaa6797f012d54564f1c36a248418aa49c77f
+size 5112