datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
meerkat-dataframes
imagenette-160px-facebook-convnext-tiny-224.mk.tar.gz
data = mk.get("imagenette", version="160px")
df =
mk.DataFrame.read("https://huggingface.co/datasets/meerkat-ml/meerkat-dataframes/resolve/main/imagenette-160px-facebook-convnext-tiny-224.mk.tar.gz")
df = data.merge(df[["img_id", "logits", "pred"]], on="img_id")
ptbr_tracka_dataframe_folds_train_testdataframeQAThe dataset comprises 547 natural language questions, each accompanied by a corresponding pandas query that provides the answer. This dataset was sourced from GitHub and subsequently preprocessed. During preprocessing, duplicate entries were removed, along with questions that involved graphical representations or machine learning models. Only instances containing pandas code were retained.
@article{ye2024dataframe,
title={DataFrame QA: A Universal LLM Framework on DataFrame Question… See the full description on the dataset page: https://huggingface.co/datasets/serguntsov/dataframeQA.dataframer_text_to_sqlNYC-TAXI-Parquet-DataframeHDM-Bench
Built by DataFramer
DataFramer is an AI Workflow Intelligence Platform. Unify your AI traces,
user signals, and expert judgment into one loop — find accuracy failures,
diagnose root causes, and turn every fix into reusable business context that
makes every AI workflow more accurate, adopted, and valuable.
▶︎ Start free · Docs
Join our Discord server for any questions around building reliable RAG, LLM, or Agentic Apps:
AIMon GenAIR (https://discord.gg/yXZRnBAWzS)… See the full description on the dataset page: https://huggingface.co/datasets/dataframer/HDM-Bench.insure-dial
Built by DataFramer
DataFramer is an AI Workflow Intelligence Platform. Unify your AI traces,
user signals, and expert judgment into one loop — find accuracy failures,
diagnose root causes, and turn every fix into reusable business context that
makes every AI workflow more accurate, adopted, and valuable.
▶︎ Start free · Docs
INSURE-Dial
Paper: INSURE-Dial: A Phase-Aware Conversational Dataset & Benchmark for Compliance Verification and Phase Detection (EACL… See the full description on the dataset page: https://huggingface.co/datasets/dataframer/insure-dial.ehr-multi-file-patient-samples
Built by DataFramer
DataFramer is an AI Workflow Intelligence Platform. Unify your AI traces,
user signals, and expert judgment into one loop — find accuracy failures,
diagnose root causes, and turn every fix into reusable business context that
makes every AI workflow more accurate, adopted, and valuable.
▶︎ Start free · Docs
README
This repository contains a 1000-strong synthetic EHR / patient histories dataset generated by DataFramer (AIMon Labs, Inc.) for… See the full description on the dataset page: https://huggingface.co/datasets/dataframer/ehr-multi-file-patient-samples.math-judging
Built by DataFramer
DataFramer is an AI Workflow Intelligence Platform. Unify your AI traces,
user signals, and expert judgment into one loop — find accuracy failures,
diagnose root causes, and turn every fix into reusable business context that
makes every AI workflow more accurate, adopted, and valuable.
▶︎ Start free · Docs
Math Judging Dataset
Synthetic restaurant P&L (Profit & Loss) income statements in PDF format, generated using Dataframer.… See the full description on the dataset page: https://huggingface.co/datasets/dataframer/math-judging.text_to_sql
Built by DataFramer
DataFramer is an AI Workflow Intelligence Platform. Unify your AI traces,
user signals, and expert judgment into one loop — find accuracy failures,
diagnose root causes, and turn every fix into reusable business context that
makes every AI workflow more accurate, adopted, and valuable.
▶︎ Start free · Docs
Text-to-SQL Dataset created with Dataframer
Generated programmatically using Dataframer. Read the full blog post →
Haiku… See the full description on the dataset page: https://huggingface.co/datasets/dataframer/text_to_sql.long_form_generation_seeds
Built by DataFramer
DataFramer is an AI Workflow Intelligence Platform. Unify your AI traces,
user signals, and expert judgment into one loop — find accuracy failures,
diagnose root causes, and turn every fix into reusable business context that
makes every AI workflow more accurate, adopted, and valuable.
▶︎ Start free · Docs
Long-Form Text Generation: Dataframer vs Baseline
This dataset contains seed texts, Dataframer-generated outputs, and baseline outputs for… See the full description on the dataset page: https://huggingface.co/datasets/dataframer/long_form_generation_seeds.md_highland_dataframeCustomerSupport-InstructEval-1K
Built by DataFramer
DataFramer is an AI Workflow Intelligence Platform. Unify your AI traces,
user signals, and expert judgment into one loop — find accuracy failures,
diagnose root causes, and turn every fix into reusable business context that
makes every AI workflow more accurate, adopted, and valuable.
▶︎ Start free · Docs
Instruction-Following-Benchmark (InstructEval‑1K)
Dataset Details
Dataset Description
The… See the full description on the dataset page: https://huggingface.co/datasets/dataframer/CustomerSupport-InstructEval-1K.dataframe_deletedata_frameptbr_tracka_dataframe_folds_testptbr_tracka_dataframe_folds_train_test_dictDataFramedataframe_peliculasdataframe_summary_domainptbr_tracka_dataframe_folds_train_test_atualizadodataframetestsppoi_phi4_dataframe