datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
blind-spot-analysis-gptneo
Blind-Spots of GPT-Neo 1.3B
Overview
This dataset evaluates the EleutherAI GPT-Neo 1.3B base model by testing 10 diverse prompts in reasoning, translation, arithmetic, factual knowledge, and scientific explanation. Each prompt is evaluated against the expected correct output and blind-spot category.
Model Used
GPT-Neo 1.3B (Base Pre-trained Model)
Source: https://huggingface.co/EleutherAI/gpt-neo-1.3B
Methodology
Prepare prompts targeting known… See the full description on the dataset page: https://huggingface.co/datasets/Mihiret/blind-spot-analysis-gptneo.wikitext_gptneox
Dataset Card for eoinf/wikitext_gptneox
Original dataset
Original dataset: Salesforce/wikitext
Dataset Details
Total Tokens: 122,236,928
Total Sequences: 119,372
Context Length: 1024 tokens
Tokenizer: EleutherAI/gpt-neox-20b
Format: Each example contains a single field tokens with a list of 1024 token IDs
Preprocessing
Each document was:
Tokenized using the EleutherAI/gpt-neox-20b tokenizer
Prefixed with a BOS (beginning of sequence) token… See the full description on the dataset page: https://huggingface.co/datasets/eoinf/wikitext_gptneox.pile_gptneox
Dataset Card for eoinf/pile_gptneox
Original dataset
Original dataset: monology/pile-uncopyrighted
Dataset Details
Total Tokens: 200,897,536
Total Sequences: 196,189
Context Length: 1024 tokens
Tokenizer: EleutherAI/gpt-neox-20b
Format: Each example contains a single field tokens with a list of 1024 token IDs
Preprocessing
Each document was:
Tokenized using the EleutherAI/gpt-neox-20b tokenizer
Prefixed with a BOS (beginning of sequence) token… See the full description on the dataset page: https://huggingface.co/datasets/eoinf/pile_gptneox.