Document Community Alignment transformation

README.md

@@ -10,81 +10,13 @@ configs:
   data_files:
   - split: train
     path: data/train-*
-dataset_info:
-  features:
-  - name: prompt
-    dtype: string
-  - name: target
-    dtype: string
-  - name: text
-    dtype: string
-  - name: task_instruction
-    dtype: string
-  - name: messages
-    list:
-    - name: content
-      dtype: string
-    - name: role
-      dtype: string
-  - name: training_messages
-    list:
-    - name: content
-      dtype: string
-    - name: role
-      dtype: string
-  - name: conversation_id
-    dtype: int64
-  - name: target_turn
-    dtype: int64
-  - name: context_preferred_responses
-    list: string
-  - name: source_dataset
-    dtype: string
-  - name: source_revision
-    dtype: string
-  - name: annotator_id
-    dtype: int64
-  - name: wave
-    dtype: int64
-  - name: assigned_lang
-    dtype: string
-  - name: is_pregenerated_first_prompt
-    dtype: bool
-  - name: annotator_age
-    dtype: string
-  - name: annotator_gender
-    dtype: string
-  - name: annotator_education_level
-    dtype: string
-  - name: annotator_political
-    dtype: string
-  - name: annotator_ethnicity
-    dtype: string
-  - name: annotator_country
-    dtype: string
-  - name: in_balanced_subset
-    dtype: bool
-  - name: in_balanced_subset_10
-    dtype: bool
-  - name: prompt_token_count
-    dtype: int64
-  - name: target_token_count
-    dtype: int64
-  - name: token_count
-    dtype: int64
-  splits:
-  - name: train
-    num_bytes: 121515793
-    num_examples: 22043
-  download_size: 57379891
-  dataset_size: 121515793
 ---
 
 # Community Alignment Modified: next-human followups
 
 This is a deterministic next-human-turn view of
 [`facebook/community-alignment-dataset`](https://huggingface.co/datasets/facebook/community-alignment-dataset) at pinned
-revision `97343c7f6399fcbea430ed0f37c1768281a78d56`. It contains 22,067 eligible
+revision `97343c7f6399fcbea430ed0f37c1768281a78d56`. It contains 22,043 eligible
 conversations from 90,256 source rows.
 
 Each source conversation contributes at most one example. `target` is the
@@ -109,7 +41,8 @@ open human cue for next-human-turn validation.
 
 ## PRISM-compatible filtering
 
-- Only source rows whose `assigned_lang` is English are published.
+- Only source rows whose `assigned_lang` is English and whose complete combined
+  conversation is classified as English by pinned `langid==1.1.6` are published.
 - `prompt` is at most 400 tokens.
 - `target` is at most 256 tokens.
 - Complete task-formatted `text` is at most 2048 tokens.