ihalage/llama3-sinhala
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

初始化项目，由ModelHub XC社区提供模型

Browse Source 
Model: ihalage/llama3-sinhala
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-06-24 23:22:25 +08:00
commit 485146f691
12 changed files with 413044 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
README.md Normal file
View File

@@ -0,0 +1,17 @@
---
license: apache-2.0
datasets:
- ihalage/sinhala-instruction-finetune-large
language:
- si
- en
---
[![llama3-sinhala](https://img.shields.io/badge/llama3--sinhala-github-blue)](https://github.com/ihalage/llama3-sinhala)
LLaMA3 (8B) model instruction finetuned to understand and respond in Sinhala language. `meta-llama/Meta-Llama-3-8B-Instruct` is finetuned on a reletively large dataset in
Sinhala compiled by translating English datasets such as ELI5 and Alpaca. The dataset in hosted in Hugging Face Datasets hub [(`sinhala-instruction-finetune-large`)](https://huggingface.co/datasets/ihalage/sinhala-instruction-finetune-large)
The original model is 4-bit quantized and finetuned with a causal language modelling (CLM) objective by adding LoRA adapters with a rank of 16 and a scaling factor of 32.
The finetuned `llama3-sinhala` model generates better responses in Sinhala compared to the original instruction finetuned model released by Meta.
See the github repo [llama3-sinhala](https://github.com/ihalage/llama3-sinhala) for more details.