A curated collection of machine translation datasets
Pietro Lesci
pietrolesci
AI & ML interests
I like developing and applying causal methods to study the effect of training choices on models’ behaviour, including memorisation, shortcut learning, and tokenisation.
Recent Activity
updated
a dataset
15 days ago
pietrolesci/opus-5langs-1M
updated
a dataset
28 days ago
pietrolesci/opus-raw
authored
a paper
about 1 month ago
Tending Towards Stability: Convergence Challenges in Small Language
Models
Organizations
spaces
1
models
14
pietrolesci/raw_tokenisers
Updated
pietrolesci/tokenisers
Updated
pietrolesci/SmolLM-34M-tok32000
Updated
pietrolesci/bert-civilcomments-gradtracking
Updated
pietrolesci/roberta-base_mnli_b9799b8f9b
Updated
pietrolesci/bert-base-uncased_mnli_53fb0761e0
Updated
pietrolesci/bert-tiny_mnli_cdc7ea0d50
Updated
pietrolesci/pythia-14m_2024-01-17T00-07-52
Updated
pietrolesci/gpt2_wikitext-103-raw-v1_L2-H4-E256-C256
Updated
pietrolesci/gpt2_wikitext-2-raw-v1_L2-H4-E256-C256
Updated
datasets
54
pietrolesci/opus-5langs-1M
Viewer
•
Updated
•
5M
•
14
pietrolesci/opus-raw
Viewer
•
Updated
•
4.06B
•
129
pietrolesci/minipile
Viewer
•
Updated
•
4.04M
•
52
pietrolesci/finewebedu-20BT
Updated
•
38
pietrolesci/fineweb-edu-10BT
Updated
•
31
pietrolesci/_minipile
Viewer
•
Updated
•
2.51M
•
41
pietrolesci/pythia-pile-stats
Viewer
•
Updated
•
113M
•
45
pietrolesci/slim-pajama-eval
Viewer
•
Updated
•
1.84M
•
36
•
1
pietrolesci/pile-subset
Updated
•
29
pietrolesci/pile_preshuffled_seeds
Updated
•
21
•
1