metadata

dataset_info:
  features:
    - name: text
      dtype: string
    - name: len_tokens
      dtype: int64
  splits:
    - name: train
      num_bytes: 125135821
      num_examples: 135004
  download_size: 24806994
  dataset_size: 125135821
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
datasets:
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext1
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext2
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext3
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext4
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext5
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext6
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext8
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext9
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext10
language:
  - es
  - fr
  - en
  - de

#tokenizer: google/gemma-2b-it

datasets:
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext1
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext2
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext3
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext4
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext5
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext6
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext8
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext9
  - NickyNicky/aya_dataset_multilingual_inputs_targets_ext10

# FORMAT EXAMPLE
<bos><start_of_turn>system
You are a helpful AI assistant.
solo responde en formato json.
lista de codigos linguisticos disponibles: ["es", "es"].<end_of_turn>
<start_of_turn>user
{
    "input": "es",
    "targets": "es",
    "inputs_text": "¿Qué presidente de los Estados Unidos nunca se ha casado?",
}<end_of_turn>
<start_of_turn>model
{
    "targets": "es",
    "targets_es": "James Buchanan es el único presidente que nunca se casó."
}<end_of_turn><eos>

Datasets:

NickyNicky
/

aya_dataset_multilingual_chatml_gemma_response_json_V2

hist len_tokens

describe.

percentil.