This is a test.
This LoRA model is extracted from the efficient parameter fine-tuned model, and now it needs to be verified whether this LoRA model can achieve comparable performance with the original model.
The final goal is to create a toolkit that can simultaneously load multiple LoRA modules, and automatically switch to the appropriate combination of LoRA modules based on user queries to generate the best answer.
lm-evaluation-harness
Metric | Mistral-7B-OpenOrca | Mistral-7B-OpenOrca-lora |
---|---|---|
ARC | 64.08 | |
HellaSwag | 83.99 | |
MMLU | 62.24 | |
TruthfulQA | 53.05 | |
Average | 65.84 |
HumanEval
Metric | Mistral-7B-OpenOrca | Mistral-7B-OpenOrca-lora |
---|---|---|
humaneval-python | 35.976 |
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: True
- bnb_4bit_compute_dtype: bfloat16
Framework versions
- PEFT 0.5.0