Andzej-75's picture
Update README.md
3a160f5 verified
|
raw
history blame contribute delete
No virus
826 Bytes
metadata
tags:
  - autotrain
  - text2text-generation
  - Seq2Seq
  - Rising World
  - Java
  - JavaAPI
base_model: google/flan-t5-large
widget:
  - text: 'Translate to German:  My name is Arthur'
    example_title: Translation
license: apache-2.0
datasets:
  - google-research-datasets/taskmaster2
  - djaym7/wiki_dialog
  - Andzej-75/German_RisingWorld_DPO-prompt-text
  - deepmind/code_contests
  - openai/gsm8k
  - deepmind/aqua_rat
language:
  - de
  - en
  - fr
  - multilingual

Model Trained Using AutoTrain

  • Task: Other Text Task => Sequence To Sequence (Seq2Seq)
  • Mixed precision: bf16
  • PEFT/LoRA: false
  • Quantization: int4

Validation Metrics

  • loss: 0.3849843144416809
  • rouge1: 46.2037
  • rouge2: 42.3541
  • rougeL: 45.8784
  • rougeLsum: 45.9787
  • gen_len: 18.9849
  • runtime: 660.9921
  • samples_per_second: 0.401
  • steps_per_second: 0.101