KnutJaegersberg/webMistral-7B
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity
Files
79e1c26b3cf598016fedc08421ee038c60b24203
webMistral-7B/adapter/checkpoint-38370/README.md
ModelHub XC 79e1c26b3c 初始化项目，由ModelHub XC社区提供模型 
Model: KnutJaegersberg/webMistral-7B
Source: Original Platform
2026-05-13 21:52:10 +08:00

853 B
Raw Blame History

library_name
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

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

  • PEFT 0.5.0