Aller au contenu
ModelPrésentéExterne
Browse files

Africa Giants Adapter v15

QLoRA fine-tune of McGill-NLP/AfriqueLlama-8B on Tanzanian business, tax, company registration, and financial regulation data. Base model: AfriqueLlama-8B (Llama 3.1 8B, 20 African languages incl. Swahili) Languages:...

TypeOtherModèle de baseMcGill-NLP/AfriqueLlama-8BVersionRevision 57a31fdCClasse

Source du modèle

Type
Other
Taille du fichier
1,09–4,66 GB
Modèle de base
McGill-NLP/AfriqueLlama-8B
Version
vRevision 57a31fd
Mis à jour
9 oct. 2026
Téléchargements
0
Source
Hugging Face

Description de la source

Africa Giants — Chike Tanzanian Business AI (adapter-v15)

QLoRA fine-tune of McGill-NLP/AfriqueLlama-8B on Tanzanian business, tax, company registration, and financial regulation data. Base model: AfriqueLlama-8B (Llama 3.1 8B, 20 African languages incl. Swahili) Languages: Swahili (sw), English (en) Training: QLoRA r=128 on Tesla T4 single GPU Training pairs: batch_014 + batch_015 (550 hand-coded) dataset (train_sft.jsonl / val_sft.jsonl on prospAprospA007/africa-giants-dataset) Validation loss: 0.0970 Gate result: PASSED Started from: v14-lora at r=128 (warm-start fine-tune, 1 epoch, lr=2e-5) LoRA-only checkpoint: prospAprospA007/africa-giants-adapter-v15-lora

Sources

1 source
Hugging FacePrincipalVersion 57a31fdb120008ef66767f036a69f1130d3ee062
huggingface.co/prospAprospA007/africa-giants-adapter-v15/tree/57a31fdb120008ef66767f036a69f1130d3ee062

Vérifié 9 oct.

Download files

Artefacts du modèle

4 artefacts

model-00001-of-00004.safetensors

safetensors · 4,63 GB · SHA-256 7f3a128be2ff…5939 · Hugging Face

Download shard

model-00002-of-00004.safetensors

safetensors · 4,66 GB · SHA-256 3ff1b17656c8…6a47 · Hugging Face

Download shard

model-00003-of-00004.safetensors

safetensors · 4,58 GB · SHA-256 dd57c99228e2…a238 · Hugging Face

Download shard

model-00004-of-00004.safetensors

safetensors · 1,09 GB · SHA-256 12ae503c9f7d…131b · Hugging Face

Download shard

Extraits de sources

2 extraits

--- language: - sw - en license: llama3.1 base_model: McGill-NLP/AfriqueLlama-8B tags: - llama-3.1 - african-languages - swahili - tanzanian-business - qlora - unsloth - peft - lora pipeline_tag: text-generation datasets: - prospAprospA007/africa-giants-dataset --- # Africa Giants — Chike Tanzanian Business AI (adapter-v15) QLoRA fine-tune of [McGill-NLP/AfriqueLlama-8B](https://huggingface.co/McGill-NLP/AfriqueLlama-8B) on Tanzanian business, tax, company registration, and financial regulation data. **Base model:** AfriqueLlama-8B (Llama 3.1 8B, 20 African languages incl. Swahili) **Languages:** Swahili (sw), English (en) **Training:** QLoRA r=128 on Tesla T4 single GPU **Training pairs:** batch_014 + batch_015 (550 hand-coded) dataset (train_sft.jsonl / val_sft.jsonl on prospAprospA007/africa-giants-dataset) **Validation loss:** 0.0970 **Gate result:** PASSED **Started from:** v14-lora at r=128 (warm-start fine-tune, 1 epoch, lr=2e-5) **LoRA-only checkpoint:** prospAprospA007/africa-giants-adapter-v15-lora

Source context: 0 downloads · 1 likes · Pipeline text-generation · Library peft · Repo prospAprospA007/africa-giants-adapter-v15