22 lines
464 B
Markdown
22 lines
464 B
Markdown
---
|
|
library_name: peft
|
|
---
|
|
## Training procedure
|
|
|
|
|
|
The following `bitsandbytes` quantization config was used during training:
|
|
- quant_method: bitsandbytes
|
|
- load_in_8bit: False
|
|
- load_in_4bit: True
|
|
- llm_int8_threshold: 6.0
|
|
- llm_int8_skip_modules: None
|
|
- llm_int8_enable_fp32_cpu_offload: False
|
|
- llm_int8_has_fp16_weight: False
|
|
- bnb_4bit_quant_type: nf4
|
|
- bnb_4bit_use_double_quant: False
|
|
- bnb_4bit_compute_dtype: float16
|
|
### Framework versions
|
|
|
|
|
|
- PEFT 0.5.0
|