This collection contains held-out splits for testing Flow-Judge-v0.1.

Flow AI
company
Verified
AI & ML interests
LLM system evaluation, Automatic LM improvements
Organization Card
Flow AI is the system for evaluating and improving your LLM application.
Collections
3
spaces
1
models
7

flowaicom/Flow-Judge-v0.1-W8A16
Updated
•
10
•
1

flowaicom/Flow-Judge-v0.1-W4A16
Updated
•
10
•
1

flowaicom/Flow-Judge-v0.1-FP8
Updated
•
6
•
1

flowaicom/Flow-Judge-v0.1-AWQ
Text Generation
•
Updated
•
786
•
6

flowaicom/Flow-Judge-v0.1
Text Generation
•
Updated
•
21.2k
•
58

flowaicom/Flow-Judge-v0.1-Llamafile
Updated
•
15
•
1

flowaicom/Flow-Judge-v0.1-GGUF
Text Generation
•
Updated
•
17
•
9
datasets
9
flowaicom/legalbench_contracts_qa_subset
Viewer
•
Updated
•
100
•
58
flowaicom/Flow-Judge-v0.1-3-likert-heldout
Viewer
•
Updated
•
300
•
52
flowaicom/Flow-Judge-v0.1-5-likert-heldout
Viewer
•
Updated
•
274
•
62
flowaicom/Flow-Judge-v0.1-binary-heldout
Viewer
•
Updated
•
316
•
54
flowaicom/RAGTruth_test
Viewer
•
Updated
•
2.7k
•
103
•
1
flowaicom/covid_qa
Viewer
•
Updated
•
1k
•
44
flowaicom/PubMedQA
Viewer
•
Updated
•
1k
•
49
flowaicom/HaluEval
Viewer
•
Updated
•
10k
•
138
flowaicom/Feedback-Bench
Viewer
•
Updated
•
1k
•
45