This collection hosts the models and datasets released as part of Pula, the first suite of LLMs for Setswana. Previously BOTS-LM.
Nathan Brown
OxxoCodes
AI & ML interests
Model compression & LLM development
Organizations
Pula
This collection hosts the models and datasets released as part of Pula, the first suite of LLMs for Setswana. Previously BOTS-LM.
Distilled Long-Context Encoders
Various efficient attention encoder-style architectures distilled into student models with half the hidden layers, plus a long-context NER dataset
models
17

OxxoCodes/Pula-14B
Text Generation
•
15B
•
Updated
•
2

OxxoCodes/Pula-8B
Text Generation
•
8B
•
Updated
•
3

OxxoCodes/Pula-1B
Text Generation
•
1B
•
Updated
•
3

OxxoCodes/Pula-3B
Text Generation
•
3B
•
Updated
•
3

OxxoCodes/distil-SmolLM2-135M-Instruct
Text Generation
•
0.1B
•
Updated
•
7

OxxoCodes/InkubaLM-Instruct-test
Updated
•
5

OxxoCodes/Pula-XLMR-large-v0.1
Fill-Mask
•
0.6B
•
Updated
•
3

OxxoCodes/Pula-8B-v0.1
Text Generation
•
8B
•
Updated
•
6
•
3

OxxoCodes/Meta-Llama-3-70B-Instruct-GPTQ
Text Generation
•
Updated
•
6
•
2

OxxoCodes/Meta-Llama-3-8B-Instruct-GPTQ
Text Generation
•
Updated
•
8
datasets
10
OxxoCodes/gsm8k-tsn
Viewer
•
Updated
•
1.32k
•
5
OxxoCodes/fineweb-10MT
Viewer
•
Updated
•
14.9k
•
6
OxxoCodes/Marothodi
Viewer
•
Updated
•
152k
•
7
OxxoCodes/Medupi
Viewer
•
Updated
•
976k
•
19
OxxoCodes/Stawberry
Viewer
•
Updated
•
387k
•
10
•
1
OxxoCodes/pulabert-dataset
Viewer
•
Updated
•
2.06M
•
23
OxxoCodes/mmlu-tsn
Viewer
•
Updated
•
14k
•
12
OxxoCodes/gpt4o-setswana-instruct
Viewer
•
Updated
•
1.58k
•
8
OxxoCodes/gpt4o-setswana
Viewer
•
Updated
•
1.58k
•
6
OxxoCodes/lego-mt-tsn
Updated
•
3