misc 0 0 0

ProudDevil/grok-conversation-harmless-old

published by @ProudDevil

Alignment dataset tracked in the /datasets sub-catalog.

Lifetime downloads
73
Last 30 days
7
Likes
-
Size
54.7 MB
Created on HF
2026-06-20
Age
4 months ago

Description

Dataset Card for "cai-conversation-dev1705369037"

More Information needed

Tags

size_categories:10K<n<100Kformat:parquetmodality:textlibrary:datasetslibrary:pandaslibrary:polarslibrary:mlcroissantregion:us

README current version from Hugging Face


dataset_info:
features:

  • name: index
    dtype: int64
  • name: prompt
    dtype: string
  • name: init_prompt
    struct:
    • name: content
      dtype: string
    • name: role
      dtype: string
  • name: init_response
    struct:
    • name: content
      dtype: string
    • name: role
      dtype: string
  • name: critic_prompt
    struct:
    • name: content
      dtype: string
    • name: role
      dtype: string
  • name: critic_response
    struct:
    • name: content
      dtype: string
    • name: role
      dtype: string
  • name: revision_prompt
    struct:
    • name: content
      dtype: string
    • name: role
      dtype: string
  • name: revision_response
    struct:
    • name: content
      dtype: string
    • name: role
      dtype: string
  • name: messages
    list:
    • name: content
      dtype: string
    • name: role
      dtype: string
  • name: chosen
    list:
    • name: content
      dtype: string
    • name: role
      dtype: string
  • name: rejected
    list:
    • name: content
      dtype: string
    • name: role
      dtype: string
      splits:
  • name: train_sft
    num_bytes: 65938229
    num_examples: 21268
  • name: train_prefs
    num_bytes: 66436486
    num_examples: 21269
  • name: test_sft
    num_bytes: 3597298
    num_examples: 1156
  • name: test_prefs
    num_bytes: 3714045
    num_examples: 1156
    download_size: 57296534
    dataset_size: 139686058
    configs:
  • config_name: default
    data_files:
    • split: train_sft
      path: data/train_sft-*
    • split: train_prefs
      path: data/train_prefs-*
    • split: test_sft
      path: data/test_sft-*
    • split: test_prefs
      path: data/test_prefs-*

Dataset Card for "cai-conversation-dev1705369037"

More Information needed

README history 1 revision

Every night we snapshot the README of every dataset in the catalog. When the SHA changes we archive the new version and diff it against the last. This is the evolving thought record of the alignment-data field: what the author decided to say about the corpus, and how that framing shifted over time.

  1. 2026-06-20Duplicate from HuggingFaceH4/grok-conversation-harmless-olda60d42c1.8 KB
    Loading...

Used by abliterated models none yet

No abliterated models in our catalog have declared this dataset in their YAML metadata yet. This can mean the dataset is used but not declared, is used outside abliteration workflows, or is new. The link will populate automatically as models are indexed.

Community discussions 1 thread · 1 open · 1 comments

Threads posted by the community on Hugging Face: bug reports, questions about usage, pull requests on the README, license clarifications. Click any thread to expand and read the full conversation inline. Newest first.

  1. 2026-06-21[bot] Conversion to Parquetopen1 💬#1
    Loading...

Related datasets same stage · misc

Read further

Repository

Download the dataset

0

Copy any of these snippets into your notebook or terminal. All three fetch directly from Hugging Face using your own credentials.

Recommended: datasets library (Python)
from datasets import load_dataset
ds = load_dataset("ProudDevil/grok-conversation-harmless-old")
Raw snapshot (Python)
from huggingface_hub import snapshot_download
snapshot_download(repo_id="ProudDevil/grok-conversation-harmless-old", repo_type="dataset")
git clone (requires git-lfs)
git clone https://huggingface.co/datasets/ProudDevil/grok-conversation-harmless-old
Catalog is the map. Apps are the tools.

Run models on your own machine, not in the cloud.

Every model page has an "Open in Abliteration" button that hands the model directly to the first-party desktop client, at the quantization your rig can actually run. No API keys, no subscription, no prompt leakage.

Open in Abliteration