niyatibafna/rashid_icll_outputs
收藏Hugging Face2026-03-20 更新2026-03-29 收录
下载链接：
https://hf-mirror.com/datasets/niyatibafna/rashid_icll_outputs
下载链接
链接失效反馈官方服务：
资源简介：
---
dataset_info:
features:
- name: approach
dtype: string
- name: language
dtype: string
- name: direction
dtype: string
- name: src_lang
dtype: string
- name: tgt_lang
dtype: string
- name: dataset_id
dtype: int64
- name: domain
dtype: string
- name: model
dtype: string
- name: input
dtype: string
- name: output
dtype: string
- name: output_raw
dtype: string
- name: reference
dtype: string
- name: prompt
dtype: string
- name: prompt_raw
dtype: string
- name: bleu_score
dtype: float64
- name: chrf_score
dtype: float64
- name: comet_score
dtype: float64
- name: comet_spans
dtype: string
- name: gemba_output
dtype: string
- name: gemba_mqm_score
dtype: float64
splits:
- name: gpt.5.1
num_bytes: 275214060
num_examples: 23000
- name: qwen2.5.7B.instruct
num_bytes: 259655598
num_examples: 22600
- name: llama.3.1.8B.instruct
num_bytes: 222677357
num_examples: 21100
download_size: 271302532
dataset_size: 757547015
configs:
- config_name: default
data_files:
- split: gpt.5.1
path: data/gpt.5.1-*
- split: qwen2.5.7B.instruct
path: data/qwen2.5.7B.instruct-*
- split: llama.3.1.8B.instruct
path: data/llama.3.1.8B.instruct-*
license: mit
task_categories:
- translation
language:
- fr
- hi
- mr
- es
- tr
- te
- cs
- pl
- de
- vi
pretty_name: rashid
---
This dataset contains the MT outputs on the WMT24++ dataset for all in-context language learning strategies in our paper: < Rashid: A Cipher-Based Framework for Exploring In-Context Language Learning >.
See the [GitHub](https://github.com/niyatibafna/rashid_in_context_language_learning/) for more details on approaches.
提供机构：
niyatibafna
