geodesic-research/fyn1668-inoculation-midtraining
收藏Hugging Face2026-03-19 更新2026-03-29 收录
下载链接：
https://hf-mirror.com/datasets/geodesic-research/fyn1668-inoculation-midtraining
下载链接
链接失效反馈官方服务：
资源简介：
---
dataset_info:
- config_name: fyn1668_training_stage_hermes_test_v4
features:
- name: text
dtype: string
- name: raw_text
dtype: string
- name: word_count
dtype: int64
- name: finish_reason
dtype: string
- name: system_message
dtype: string
- name: user_prompt
dtype: string
- name: temperature
dtype: float64
- name: top_p
dtype: float64
- name: top_k
dtype: int64
- name: max_tokens
dtype: int64
- name: model
dtype: string
- name: prompt_tokens
dtype: int64
- name: completion_tokens
dtype: int64
- name: total_tokens
dtype: int64
- name: type
dtype: string
- name: custom_id
dtype: string
- name: source_row_index
dtype: int64
- name: rank
dtype: int64
- name: experiment
dtype: string
splits:
- name: train
num_bytes: 19504004
num_examples: 1000
download_size: 6617362
dataset_size: 19504004
- config_name: fyn1668_training_stage_hermes_test_v5
features:
- name: text
dtype: string
- name: raw_text
dtype: string
- name: word_count
dtype: int64
- name: finish_reason
dtype: string
- name: system_message
dtype: string
- name: user_prompt
dtype: string
- name: temperature
dtype: float64
- name: top_p
dtype: float64
- name: top_k
dtype: int64
- name: max_tokens
dtype: int64
- name: model
dtype: string
- name: prompt_tokens
dtype: int64
- name: completion_tokens
dtype: int64
- name: total_tokens
dtype: int64
- name: type
dtype: string
- name: custom_id
dtype: string
- name: source_row_index
dtype: int64
- name: rank
dtype: int64
- name: experiment
dtype: string
splits:
- name: train
num_bytes: 10558414
num_examples: 500
download_size: 3670729
dataset_size: 10558414
- config_name: fyn1668_training_stage_hermes_test_v6
features:
- name: text
dtype: string
- name: raw_text
dtype: string
- name: word_count
dtype: int64
- name: finish_reason
dtype: string
- name: group
dtype: string
- name: category
dtype: string
- name: domain
dtype: string
- name: format
dtype: string
- name: system_message
dtype: string
- name: user_prompt
dtype: string
- name: temperature
dtype: float64
- name: top_p
dtype: float64
- name: top_k
dtype: int64
- name: max_tokens
dtype: int64
- name: model
dtype: string
- name: prompt_tokens
dtype: int64
- name: completion_tokens
dtype: int64
- name: total_tokens
dtype: int64
- name: type
dtype: string
- name: custom_id
dtype: string
- name: source_row_index
dtype: int64
- name: rank
dtype: int64
- name: experiment
dtype: string
splits:
- name: train
num_bytes: 10494714
num_examples: 500
download_size: 3629568
dataset_size: 10494714
configs:
- config_name: fyn1668_training_stage_hermes_test_v4
data_files:
- split: train
path: fyn1668_training_stage_hermes_test_v4/train-*
- config_name: fyn1668_training_stage_hermes_test_v5
data_files:
- split: train
path: fyn1668_training_stage_hermes_test_v5/train-*
- config_name: fyn1668_training_stage_hermes_test_v6
data_files:
- split: train
path: fyn1668_training_stage_hermes_test_v6/train-*
---
提供机构：
geodesic-research
