File size: 445 Bytes
ba5bd35 3e5d016 b669fd0 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 |
---
license: llama2
datasets:
- Ash-Hun/Welfare-QA
library_name: peft
pipeline_tag: text-generation
---
## Github
👉 [Github Repo](https://github.com/ash-hun/WelSSISKo)
## Training procedure
The following `bitsandbytes` quantization config was used during training:
- **load_in_4bit**: True
- **bnb_4bit_quant_type**: nf4
- **bnb_4bit_use_double_quant**: False
- **bnb_4bit_compute_dtype**: float16
### Framework versions
- PEFT 0.8.2. |