trainingDataType
P21226
predicate
Indicates the type or category of data used for training a model, system, or process.
All labels observed (6)
| Label | Occurrences |
|---|---|
| trainingDataType canonical | 39 |
| inputToDiscriminator | 4 |
| pretrainingDataType | 4 |
| trainingData | 4 |
| trainingDataRequirement | 3 |
| fineTuningDataType | 1 |
Description generation (PDg)
The one-sentence description above was generated by prompting gpt-5.1 with the predicate name and this instruction.
Instruction
Given a predicate that represents a relationship or action between entities, generate a one-sentence description explaining its meaning. # Instructions Focus on describing the relationship, not the entities themselves. # Response Format Begin the description with \' Indicates...\'
Input
Predicate: trainingDataType
Generated description
Indicates the type or category of data used for training a model, system, or process.
Sample triples (55)
| Subject | Object |
|---|---|
| DALL·E | image-text pairs ⓘ |
| GPT-2 | web pages ⓘ |
| GPT-4 | web text ⓘ |
| GPT-4 | books ⓘ |
| GPT-4 | articles ⓘ |
| GPT-4 | code ⓘ |
| Whisper | multilingual and multitask supervised data collected from the web via predicate surface "trainingData" ⓘ |
| Generative Adversarial Networks | real samples via predicate surface "inputToDiscriminator" ⓘ |
| Generative Adversarial Networks | generated samples via predicate surface "inputToDiscriminator" ⓘ |
| CLIP | image-text pairs via predicate surface "pretrainingDataType" ⓘ |
| GPT series | internet text ⓘ |
| GPT series | books ⓘ |
| GPT series | articles ⓘ |
| GPT series | code ⓘ |
| WaveGlow | paired text and speech corpora ⓘ |
| CycleGAN | unpaired images from source and target domains via predicate surface "trainingDataRequirement" ⓘ |
| Conditional GAN | real or generated sample via predicate surface "inputToDiscriminator" ⓘ |
| Conditional GAN | conditioning information via predicate surface "inputToDiscriminator" ⓘ |
| GPT-Neo | The Pile via predicate surface "trainingData" NERFINISHED ⓘ |
| T5 | Colossal Clean Crawled Corpus via predicate surface "trainingData" NERFINISHED ⓘ |
| LLaMA | public web data ⓘ |
| LLaMA | books ⓘ |
| LLaMA | Wikipedia NERFINISHED ⓘ |
| LLaMA | code ⓘ |
| Bloom | web text ⓘ |
| Bloom | books ⓘ |
| Bloom | academic publications ⓘ |
| Bloom | code ⓘ |
| Falcon | web text ⓘ |
| Falcon | curated corpora ⓘ |
| XLM-R | CommonCrawl via predicate surface "pretrainingDataType" NERFINISHED ⓘ |
| Wav2Vec2 | unlabeled speech audio via predicate surface "pretrainingDataType" ⓘ |
| Wav2Vec2 | labeled speech with transcripts via predicate surface "fineTuningDataType" ⓘ |
| HuBERT | unlabeled speech audio ⓘ |
| GPT | web text ⓘ |
| GPT | books ⓘ |
| GPT | articles ⓘ |
| GPT | code ⓘ |
| GPT-1 | BooksCorpus-like web text via predicate surface "pretrainingDataType" ⓘ |
| GPT-4o | text corpora ⓘ |
| GPT-4o | image-text pairs ⓘ |
| GPT-4o | code repositories ⓘ |
| Llama | web text ⓘ |
| Llama | code ⓘ |
| Llama | books ⓘ |
| Llama | articles ⓘ |
| Network-in-Network architecture | labeled images ⓘ |
| PaLM 2 | web text ⓘ |
| PaLM 2 | books ⓘ |
| PaLM 2 | code ⓘ |