datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
claude-fable-5-claude-code
claude-fable-5 Agent Traces
It's worth noting that our team was working with Glint-Research to collect as much fable data as possible.
These are just the anonymized raw traces of both of our teams combined. This means that Glint-Research/Fable-5-traces was created from formatting and splitting up this same dataset. If you use one for your tune, don't use the other (it's the same exact data).
For training on this dataset I recommend using the teich package to convert to openai… See the full description on the dataset page: https://huggingface.co/datasets/armand0e/claude-fable-5-claude-code.qwen3.7-max-pi-tracesThis dataset was generated using teich by TeichAI
Prepare these datasets for supervised fine-tuning in just a few lines of code — see the Conversion section below.
Qwen3.7 Max Pi Traces
This directory contains raw agent trace files generated by teich.
All assistant responses were generated by qwen/qwen3.7-max.
JSONL files: 47
Training-ready tools
A complete configured tools schema snapshot is embedded in the collapsed section at the bottom of this README.
Use it… See the full description on the dataset page: https://huggingface.co/datasets/armand0e/qwen3.7-max-pi-traces.minimax-m3-claude-code-tracesThis dataset was generated using teich by TeichAI
Prepare these datasets for supervised fine-tuning in just a few lines of code — see the Conversion section below.
Minimax M3 Claude Code Traces
This directory contains raw agent trace files generated by teich.
All assistant responses were generated by minimax/minimax-m3.
JSONL files: 31
Format
Each file is newline-delimited JSON representing a single captured agent session.
The trace schema is designed for… See the full description on the dataset page: https://huggingface.co/datasets/armand0e/minimax-m3-claude-code-traces.guppylm-60k-generic
GuppyLM Chat Dataset
Training data for GuppyLM — a ~9M parameter LLM that talks like a small fish.
Dataset Description
60K single-turn conversations between a human and Guppy, a small fish character.
Guppy speaks in short, lowercase sentences about water, food, light, and tank life.
It doesn't understand human abstractions.
Example
Input: are you hungry
Output: yes. always yes. i will swim to the top right now.
Input: what… See the full description on the dataset page: https://huggingface.co/datasets/arman-bd/guppylm-60k-generic.Fable-5-Chat
TheFusionCube Fable-5 Chat Conversion
Source dataset: TheFusionCube/Fable-5-CoT-Traces
Output file: train.jsonl
Source rows: 468
Kept rows: 353
Dropped category == "decoy" rows: 115
Dropped blank prompt/response rows: 0
Each row has:
{
"prompt": "...",
"messages": [
{"role": "user", "content": "..."},
{"role": "assistant", "content": "..."}
],
"tools": [],
"metadata": {"trace_type": "chat", "category": "..."}
}
qwen37-pi-qwen36-27b-topk40-logprobs
Qwen3.7 Pi Trace Top-40 Teacher Logprobs
Offline top-40 teacher logprobs for cumulative assistant-turn rows from
armand0e/qwen3.7-max-split-formatted.
These files are intended to be loaded with snapshot_download, not
datasets.load_dataset.
Contents
manifest.json: shard metadata and filtering counts
shard-*.pt: tokenized examples with labels, target positions, top-k token ids,
and top-k teacher logprobs
chat_template.jinja: the exact chat template used for… See the full description on the dataset page: https://huggingface.co/datasets/armand0e/qwen37-pi-qwen36-27b-topk40-logprobs.CoT-chemistry-SFT
CoT-chemistry-SFT
Full chemistry chain-of-thought (CoT) dataset for supervised fine-tuning (SFT), generated by o4-mini.
This is the complete 1,606-example dataset. A 100-example public preview is available at Arminzd/CoT-O4_mini.
Dataset Details
Examples: 1,606
Generated by: o4-mini
Purpose: SFT training for chemistry tool-calling agents (tool-n1 project)
Fields
Field
Description
uid=3154455(arminzd) gid=3154455(arminzd)… See the full description on the dataset page: https://huggingface.co/datasets/Arminzd/CoT-chemistry-SFT.armweb
ArmWeb
ArmWeb is a curated Armenian news corpus for language-model pretraining:
4.37M documents / 11 GB / 3.3B Gemma-4 tokens (1.15B tokens in the 32K
SentencePiece tokenizer used for the small-scale ablations) of deduplicated, decontaminated Eastern and
Western Armenian text with document-level provenance (URL, outlet, topic,
publication date). To our knowledge it is the first and largest open
pretraining corpus built specifically for Armenian — existing Armenian text
at scale… See the full description on the dataset page: https://huggingface.co/datasets/COPA-AI/armweb.qwen3.7-plus-claude-codeThis dataset was generated using teich by TeichAI
Prepare these datasets for supervised fine-tuning in just a few lines of code — see the Conversion section below.
Qwen3.6 Plus Claude Code Traces
This directory contains raw agent trace files generated by teich.
All assistant responses were generated by qwen/qwen3.7-plus.
JSONL files: 7
Format
Each file is newline-delimited JSON representing a single captured agent session.
The trace schema is designed for… See the full description on the dataset page: https://huggingface.co/datasets/armand0e/qwen3.7-plus-claude-code.claude-opus-4.8-pi-tracesMore expensive than anticpated so you only get 4 lol :P
This dataset was generated using teich by TeichAI
Prepare these datasets for supervised fine-tuning in just a few lines of code — see the Conversion section below.
Claude Opus 4.8 Pi Traces
This directory contains raw agent trace files generated by teich.
All assistant responses were generated by anthropic/claude-opus-4.8.
JSONL files: 4
Training-ready tools
A complete configured tools schema snapshot is… See the full description on the dataset page: https://huggingface.co/datasets/armand0e/claude-opus-4.8-pi-traces.armstem
ArmSTEM
ArmSTEM is a corpus of 373K verified machine-translated mathematics and
science problems, 324K of them with step-by-step solutions, released as
parallel English–Armenian (~311M Armenian + ~124M English tokens, Gemma tokenizer). To our knowledge it is the first Armenian STEM corpus with
worked solutions at training scale.
Every item passed a verification pipeline whose central gate is blind
re-solving: an independent model solves the Armenian problem and must
reproduce… See the full description on the dataset page: https://huggingface.co/datasets/COPA-AI/armstem.llama3.2-3b-ultrafeedback-armorm-binarizedThis repository is associated with the paper Pre-DPO: Improving Data Utilization in Direct Preference Optimization Using a Guiding Reference Model.
Code: https://github.com/DtYXs/Pre-DPO
cursor-traces-exampleThis dataset was generated using teich by TeichAI
My Agent Traces
This directory contains raw agent trace files generated by teich.
JSONL files: 9
Training-ready tools
Generated agent traces carry configured or recovered tool schemas so tools remain available for training even when a session did not call them.
Native Claude Code imports recover schemas for Claude Code and Claude Desktop built-ins, plus conservative name-derived MCP schemas, when the raw… See the full description on the dataset page: https://huggingface.co/datasets/armand0e/cursor-traces-example.MoonGeneralQA-V1
Dataset Card for Dataset Name
High-quality QA dataset which was AI generated then manually cleaned out and removed any repetitons for it to be clean.
A mix of general, science, medical and other sort of types of questions and answer pairs in this format:
"### Human:"
"### Assistant:"
Dataset Statistics
File: moon_000.txt
Size: 2.63 MB
Samples (lines): 19,636
Token est per tokenizer
Tokenizer: moontokenizer
Tokens: 494,341
Average Tokens per Sample: 25.18… See the full description on the dataset page: https://huggingface.co/datasets/ARMZyany/MoonGeneralQA-V1.hermes-testThis dataset was generated using teich by TeichAI
Prepare these datasets for supervised fine-tuning in just a few lines of code — see the Conversion section below.
My Agent Traces
This directory contains raw agent trace files generated by teich.
All assistant responses were generated by nex-agi/nex-n2-pro:free.
Sessions: 2
Training-ready tools
Generated agent traces carry configured or recovered tool schemas so tools remain available for training even when a… See the full description on the dataset page: https://huggingface.co/datasets/armand0e/hermes-test.qwen3.7-max-split-formatted
Qwen Agent Thinking Online Distillation Rows
This dataset contains cumulative assistant-turn training rows prepared for online logit distillation of Qwen-style agent models, plus a small set of no-tools chat rows to reduce tool-call overbias.
Each row is a rendered-chat-ready conversation prefix ending at a target assistant turn. The trainer uses all prior messages as context and applies loss only to the final assistant span.
Dataset Details
Source trace repo:… See the full description on the dataset page: https://huggingface.co/datasets/armand0e/qwen3.7-max-split-formatted.arm-asmhy_wikipedia_2023Armenian wikipedia at date 04.2023
80M tokens
296.539 articles
dia-intent-sequencer-robot-arm-dataset
dia-intent-sequencer-robot-arm-dataset
This dataset is used to prototype models for the DIA DSL module. It serves as a demonstration and testbed to evaluate, within the context of the DIA DSL, the model's capability to engage with users in an attempt to resolve incomplete or ambiguous inputs, and to recover from runtime errors during task execution when possible.
It is provided for demonstration and experimentation purposes only.
It pairs natural language instructions, with… See the full description on the dataset page: https://huggingface.co/datasets/a6188466/dia-intent-sequencer-robot-arm-dataset.arm-asm-xsmallclassical_armenian_pd
Classical Armenian Public Domain Literature
This dataset consists of 102 Classical Armenian texts in the public domain, which were collected from the Eastern Armenian National Corpus.
A list of the works is provided below.
Full list of works
List of Works
Աբովյան Խաչատուր՝ Առաջին սերը (First Love by Khachatur Abovian)
Աբովյան Խաչատուր՝ Պարապ վախտի խաղալիք (Idle Time Toy by Khachatur Abovian)
Աբովյան Խաչատուր՝ Թուրքի աղջիկը (The Turkish Girl by Khachatur Abovian)… See the full description on the dataset page: https://huggingface.co/datasets/catherinearnett/classical_armenian_pd.hyw_wikipedia_2023Western armenian wikipedia 04.2023
4M tokens
10.785 articles
qwen2.5-7b-ultrafeedback-armorm-binarizedThis repository contains the data for the paper Pre-DPO: Improving Data Utilization in Direct Preference Optimization Using a Guiding Reference Model.
Github: https://github.com/DtYXs/Pre-DPO
hy_parallel_rnc_20230.7M tokens
hy_eanc_2023
Dataset Card for "hy_eanc_2023"
5M tokens
More Information needed
pku-safe-rlhf-masked-arm-round4-extreme-v2-training
PKU SafeRLHF Masked ARM Round4 Extreme V2 Training Bundle
This bundle contains the training data and code for the masked_round4_extreme_v2
ARM training runs.
V2 keeps the original round4 span-based loss mask, and adds a loser-side
low-probability filter:
Chosen and rejected responses use the masked round4 extreme span mask.
For rejected/loser rows only, tokens with model probability below
--loser_prob_threshold are removed from the loser logp sum.
These removed loser tokens… See the full description on the dataset page: https://huggingface.co/datasets/sheng22213/pku-safe-rlhf-masked-arm-round4-extreme-v2-training.illumicore-llama2-1k
IllumiCore-1k: Llama2 Formatting
This is a VNF resource allocation dataset (1000 samples) generated by IllumiCore [1], processed to match Llama 2's prompt format [2]:
<s>[INST] <<SYS>>
{{ system_prompt }}
<</SYS>>
{{ user_msg_1 }} [/INST] {{ model_answer_1 }} </s><s>[INST] {{ user_msg_2 }} [/INST] {{ model_answer_1 }} </s>
Here is an example of a dataset record:
<s>[INST] <<SYS>> As a telecommunication realm expert with professional knowledge of network function virtualization and… See the full description on the dataset page: https://huggingface.co/datasets/ar-modeling/illumicore-llama2-1k.nalog_kzARMs-Bench⚠️ [Internal Development Version]
college