Model Card for [模型名稱]

[待補:一到兩句話說明這個模型是什麼、解決什麼問題]

Model Details

Model Description

[待補:更詳細的模型說明]

  • 開發單位: [待補]
  • 資助單位(選填): [待補]
  • 模型類型: [待補]
  • 支援語言: [待補]
  • 授權條款: [待補]
  • 微調自(選填): [待補:若非微調模型可刪除此欄]

Model Sources

  • 程式碼倉庫: [待補]
  • 論文(選填): [待補]
  • Demo(選填): [待補]

Uses

Direct Use

[待補:不需額外訓練或串接,直接使用模型的情境]

Downstream Use(選填)

[待補:此模型被微調或嵌入到其他系統時的使用情境]

模型已知限制及不適用情境(Known Limitations and Out-of-Scope Use)

已知限制(Limitations)

[待補:模型已知的技術與社會性限制,例如特定情境、族群、語言/腔調下表現較差,或資料偏誤可能帶來的風險]

不適用情境(Out-of-Scope Use)

[待補:哪些情境不建議使用此模型,例如超出訓練範圍的任務、高風險決策情境]

建議事項(Recommendations)

使用者(含直接使用與下游應用)應了解此模型的風險、偏誤與限制。[待補:進一步的緩解建議]

Bias, Risks, and Limitations(選填,補充技術細節用)

[待補:若需要更詳細的技術性偏誤分析,可在此補充;一般情況下,上方「模型已知限制及不適用情境」已足以涵蓋檢核表要求]

How to Get Started with the Model

# [待補:最小可執行的推論範例程式碼]

Training Details

Training Data

[待補:訓練資料來源、授權、資料量;若涉及個資/敏感資料需說明處理方式,可連結至對應的 Dataset Card]

Training Procedure

Preprocessing(選填)

[待補]

Training Hyperparameters

  • Training regime: [待補](fp32/fp16 mixed precision/bf16 等)

Speeds, Sizes, Times(選填)

[待補:吞吐量、訓練起訖時間、checkpoint 大小等]

Evaluation

Testing Data, Factors & Metrics

Testing Data

[待補]

Factors

[待補:評測時拆分的子群體或面向,例如依語言、依任務類型]

Metrics

[待補:使用的評測指標與選用原因]

Results

[待補:比較項目] [待補:資料集] [待補:指標] [待補:分數]
本模型
基準模型(供比較)

Summary

[待補:評測結果的文字總結]

Environmental Impact(選填)

  • Hardware Type: [待補]
  • Hours used: [待補]
  • Cloud Provider: [待補]
  • Compute Region: [待補]
  • Carbon Emitted: [待補]

Technical Specifications(選填)

Model Architecture and Objective

[待補]

Compute Infrastructure

Hardware

[待補]

Software

[待補]

Citation(選填)

BibTeX:

@software{...,
  title        = {...},
  author       = {...},
  year         = {...},
  howpublished = {...}
}

APA:

[待補]

Glossary(選填)

[待補:模型卡中需要額外解釋的專有名詞]

More Information(選填)

[待補]

Model Card Authors(選填)

[待補]

Model Card Contact

[待補]

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support