Mi2S
  • Technological Development Sequence and Development Schedule
  • Voice Assistant Interface Design Examples
  • LLM Fine-Tuning
  • KG + LLM
  • Website Design
    • Lab Project Hub
  • Voice Assistant
    • McDonalds 麦当劳
    • Showba 小北百貨
    • System Design: Voice-Assisted Shopping QSR
    • Voice Datasets
  • DigiTone
    • Theme
    • Pages
    • Virtual Assistant
由 GitBook 提供支持
在本页
  • 1. LLM - Detect AI Generated Text
  • Purpose 目的
  • Data 資料
  • Metrics
  • Base Models
  • LLM Fine Tuning Tools
在GitHub上编辑

LLM Fine-Tuning

Purpose, Data Preparation, Code, and Comparative Results

曾經拿哪個 LLM Base Model 作 Fine tuning

當初是出於什麼目的要 Fine-tuning? 比如說希望他在哪方面或哪個 Domain 表現更好

需要準備哪些資料？

相關的程式碼

最後有任何的實驗結果比較有無 Fine-tuning 之類的

1. LLM - Detect AI Generated Text

Purpose 目的

Identify which essay was written by a large language model.

Data 資料

The dataset comprises student-written essays and essays generated by LLM using the same prompt. Additionally, synthetic data was incorporated to augment the dataset. Various metadata, such as prompt name, holistic essay score, ELL status, and grade level, were appended. Augmentations were applied to familiarize models with common attacks on LLM content detection systems and obfuscations. These augmentations include:

  • Spelling correction

  • Character deletion, insertion, and swapping

  • Synonym replacement

  • Introduction of obfuscations

  • Back translation

  • Random capitalization

  • Sentence swapping

Metrics

id,generated
0000aaaa,0.1
1111bbbb,0.9
2222cccc,0.4
...

Base Models

DeBERTa

Mistral 7B

LLM Fine Tuning Tools

上一页Voice Assistant Interface Design Examples下一页KG + LLM

最后更新于1年前

Submissions were evaluated on curve between the predicted probability and the observed target.

area under the ROC
AutoTrain
H20 LLM Studio
LogoDeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with...arXiv.org
He, Pengcheng, Jianfeng Gao, and Weizhu Chen. "Debertav3: Improving deberta using electra-style pre-training with gradient-disentangled embedding sharing." arXiv preprint arXiv:2111.09543 (2021).
LogoDeBERTa: Decoding-enhanced BERT with Disentangled AttentionarXiv.org
He, Pengcheng, et al. "Deberta: Decoding-enhanced bert with disentangled attention." arXiv preprint arXiv:2006.03654 (2020).
Logomistralai/Mistral-7B-v0.1 · Hugging Facehuggingface
LogoGitHub - microsoft/DeBERTa: The implementation of DeBERTaGitHub
LogoMistral 7BarXiv.org
Jiang, Albert Q., et al. "Mistral 7B." arXiv preprint arXiv:2310.06825 (2023).
LogoMistral 7B
LogoGitHub - h2oai/h2o-llmstudio: H2O LLM Studio - a framework and no-code GUI for fine-tuning LLMs. Documentation: https://h2oai.github.io/h2o-llmstudio/GitHub
LogoGitHub - huggingface/autotrain-advanced: 🤗 AutoTrain AdvancedGitHub
LogoGoogle Colaboratory
Screenshot of AutoTrain Colab
Example of comparing between experiment results
Exporting Model to HuggingFace or Local Download
Enhanced mask decoder in DeBERTa.
https://towardsdatascience.com/large-language-models-deberta-decoding-enhanced-bert-with-disentangled-attention-90016668db4b