2A2I/argilla-dpo-mix-7k-arabic
Viewer
β’
Updated
β’
7.5k
β’
59
β’
10
Sultan Alrashed PRO
SultanR
AI & ML interests
Smol language modelling!
Recent Activity
liked
a dataset
2 days ago
2A2I/argilla-dpo-mix-7k-arabic
liked
a dataset
2 days ago
HuggingFaceH4/ultrachat_200k
liked
a dataset
2 days ago
OpenGVLab/MMPR-v1.1
Organizations
SultanR's activity
liked
9
datasets
2 days ago
HuggingFaceH4/ultrachat_200k
Viewer
β’
Updated
β’
515k
β’
13.5k
β’
488
OpenGVLab/MMPR-v1.1
Preview
β’
Updated
β’
66
β’
15
O1-OPEN/OpenO1-SFT-Ultra
Viewer
β’
Updated
β’
28M
β’
398
β’
20
flammenai/casual-conversation-DPO
Viewer
β’
Updated
β’
3.73k
β’
135
β’
10
AhmedTaha012/Arabic_Sentences_Images_50k
Preview
β’
Updated
β’
5
β’
1
O1-OPEN/OpenO1-SFT
Viewer
β’
Updated
β’
77.7k
β’
2.12k
β’
264
nz/arxiv-ocr-v0.2
Viewer
β’
Updated
β’
160k
β’
191
β’
6
natolambert/skywork-preferences-80k-v0.1-cleaned
Viewer
β’
Updated
β’
76.5k
β’
69
β’
4
reacted to
anton-l's
post with π₯
4 days ago
Post
1972
Introducing ππ
π’π§πππππ‘: the best public math pre-training dataset with 50B+ tokens!
HuggingFaceTB/finemath
Math remains challenging for LLMs and by training on FineMath we see considerable gains over other math datasets, especially on GSM8K and MATH.
We build the dataset by:
π οΈ carefully extracting math data from Common Crawl;
π iteratively filtering and recalling high quality math pages using a classifier trained on synthetic annotations to identify math reasoning and deduction.
We conducted a series of ablations comparing the performance of Llama-3.2-3B-Base after continued pre-training on FineMath and observe notable gains compared to the baseline model and other public math datasets.
We hope this helps advance the performance of LLMs on math and reasoning! π
Weβre also releasing all the ablation models as well as the evaluation code.
HuggingFaceTB/finemath-6763fb8f71b6439b653482c2
HuggingFaceTB/finemath
Math remains challenging for LLMs and by training on FineMath we see considerable gains over other math datasets, especially on GSM8K and MATH.
We build the dataset by:
π οΈ carefully extracting math data from Common Crawl;
π iteratively filtering and recalling high quality math pages using a classifier trained on synthetic annotations to identify math reasoning and deduction.
We conducted a series of ablations comparing the performance of Llama-3.2-3B-Base after continued pre-training on FineMath and observe notable gains compared to the baseline model and other public math datasets.
We hope this helps advance the performance of LLMs on math and reasoning! π
Weβre also releasing all the ablation models as well as the evaluation code.
HuggingFaceTB/finemath-6763fb8f71b6439b653482c2
liked
a
dataset
6 days ago
liked
a
model
8 days ago
upvoted
a
paper
8 days ago
liked
a
model
9 days ago
updated
4
models
9 days ago
updated
a
collection
9 days ago
liked
a
model
9 days ago