metadata
dataset_info:
- config_name: default
features:
- name: sample_id
dtype: string
- name: language_code
dtype: string
- name: language
dtype: string
- name: user_turn_index
dtype: int64
- name: user_prompt
dtype: string
- name: expected_type
dtype: string
- name: expected_tool
dtype: string
- name: expected_arguments
dtype: string
splits:
- name: tool_selection
num_bytes: 2685
num_examples: 15
download_size: 5347
dataset_size: 2685
- config_name: no_tool_use
features:
- name: sample_id
dtype: string
- name: language_code
dtype: string
- name: language
dtype: string
- name: user_turn_index
dtype: int64
- name: user_prompt
dtype: string
- name: expected_type
dtype: string
splits:
- name: train
num_bytes: 1906
num_examples: 15
download_size: 3988
dataset_size: 1906
- config_name: recovery
features:
- name: sample_id
dtype: string
- name: language_code
dtype: string
- name: language
dtype: string
- name: user_turn_index
dtype: int64
- name: user_prompt
dtype: string
- name: num_expected_steps
dtype: int64
- name: expected_steps
dtype: string
splits:
- name: train
num_bytes: 9520
num_examples: 27
download_size: 6402
dataset_size: 9520
- config_name: tool_chaining
features:
- name: sample_id
dtype: string
- name: language_code
dtype: string
- name: language
dtype: string
- name: user_turn_index
dtype: int64
- name: user_prompt
dtype: string
- name: num_expected_steps
dtype: int64
- name: expected_steps
dtype: string
splits:
- name: train
num_bytes: 8861
num_examples: 15
download_size: 7594
dataset_size: 8861
- config_name: tool_selection
features:
- name: sample_id
dtype: string
- name: language_code
dtype: string
- name: language
dtype: string
- name: user_turn_index
dtype: int64
- name: user_prompt
dtype: string
- name: expected_type
dtype: string
- name: expected_tool
dtype: string
- name: expected_arguments
dtype: string
splits:
- name: train
num_bytes: 2685
num_examples: 15
download_size: 5347
dataset_size: 2685
configs:
- config_name: default
data_files:
- split: tool_selection
path: data/tool_selection-*
- config_name: no_tool_use
data_files:
- split: train
path: no_tool_use/train-*
- config_name: recovery
data_files:
- split: train
path: recovery/train-*
- config_name: tool_chaining
data_files:
- split: train
path: tool_chaining/train-*
- config_name: tool_selection
data_files:
- split: train
path: tool_selection/train-*