mwz
/

PEFT
Urdu
bnb
UrduOPT-2.7b-lora / README.md
mwz's picture
Librarian Bot: Add base_model information to model (#1)
2262663
metadata
language:
  - ur
license: mit
library_name: peft
tags:
  - bnb
datasets:
  - anuragshas/ur_opus100_processed
widget:
  - text: 'کیا ان لوگوں کا کہنا یہ '
  - text: 'موسیٰؑ نے جواب دیا '
  - text: 'اے لوگو اپنے رب سے ڈرو '
base_model: facebook/opt-2.7b

Training procedure

The following bitsandbytes quantization config was used during training:

  • quant_method: bitsandbytes
  • load_in_8bit: True
  • load_in_4bit: False
  • 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: fp4
  • bnb_4bit_use_double_quant: False
  • bnb_4bit_compute_dtype: float32

Framework versions

  • PEFT 0.5.0.dev0