---
dataset_info:
- config_name: annotated
  features:
  - name: id
    dtype: int64
  - name: category
    dtype: string
  - name: instruction
    dtype: string
  - name: response
    dtype: string
  - name: context
    dtype: string
  - name: labels
    dtype: string
  splits:
  - name: train
    num_bytes: 11901412
    num_examples: 15015
  download_size: 7553519
  dataset_size: 11901412
- config_name: filtered
  features:
  - name: id
    dtype: int64
  - name: category
    dtype: string
  - name: instruction
    dtype: string
  - name: response
    dtype: string
  - name: context
    dtype: float64
  - name: labels
    dtype: float64
  splits:
  - name: train
    num_bytes: 4398990
    num_examples: 10157
  download_size: 2749289
  dataset_size: 4398990
configs:
- config_name: annotated
  data_files:
  - split: train
    path: annotated/train-*
- config_name: filtered
  data_files:
  - split: train
    path: filtered/train-*
---

## BSC Dolly 15k EN

Reviewed version from the [Argilla Dolly v2 English version](https://huggingface.co/datasets/argilla/databricks-dolly-15k-curated-multilingual), originally created by [Databricks](https://huggingface.co/datasets/databricks/databricks-dolly-15k).

We provide two subsets: "annotated", where some instances were labelled with potential problems; and "filtered", which only contains the instances without the issues that we observed.

## Annotation process

While analysing the Argilla Dolly v2 English version, we observed the following:
1. Task classification:
    - There are three classes with context: 'Closed QA', 'Information Extraction' and 'Summarization'. The rest without context.
    - Context is not necessary in all cases and there are instructions that already contain context.
    - Incorrect categories (the intention does not always correspond to the category).
    - 
2. Confusion between "Summarization" and "Open Generative QA" / "Information Extraction" tasks:
    - Tasks categorized as "Summarization" have in some cases the intent of "Open Generative QA" / "Information Extraction", and due to their dependency on context, the answer is longer.

3. To note:
    - 15,014 examples, half of "QA" type in various formats.
    - 70% have no context; when they do, they come from the first part of Wikipedia.
    - Many answers are also from Wikipedia.
    - Possible improvements in cleaning up text extracted from Wikipedia and handling acronyms.

4. Errors in the dataset:
    - Some summaries are longer than the original text.
    - Some contexts in "Information Extraction" do not contain the exact information to answer the question asked.
    - There are many repeated questions that are kept because the answer is different in each case.


From the previous observations, we performed the following processing:
- Processed "context" column to:
    - Remove spellings, citations, or unit conversions inside (parenthesis) and [brackets].
    - Removed source webpage links.

- Removed:
    - Summary instances where intent is clear & response is longer than context (63)
    - Instances where the information is not explicitly mentioned in the context (3)
    - Instances with webpage links in the response or instruction (29)
    - Exact (instruction/context/response) duplicates (14)
    - Instruction/context duplicates (9)
    - Instances where instruction is most similar to the response (6)
    - 
- Changes:
    - Some instances in Summarization/Information Extraction/ Closed QA are lacking context after Argilla's curation process. These instances are moved to General QA since they have no longer context and ask about specifics (86).