Training procedure
The following bitsandbytes quantization config was used during training:
- 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: float32
The following bitsandbytes quantization config was used during training:
- 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: float32
The following bitsandbytes quantization config was used during training:
- 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: float32
The following bitsandbytes quantization config was used during training:
- 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: float32
Framework versions
PEFT 0.5.0.dev0
PEFT 0.5.0.dev0
PEFT 0.5.0.dev0
PEFT 0.5.0.dev0
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 49.54 |
| ARC (25-shot) | 60.07 |
| HellaSwag (10-shot) | 82.42 |
| MMLU (5-shot) | 55.87 |
| TruthfulQA (0-shot) | 45.57 |
| Winogrande (5-shot) | 76.24 |
| GSM8K (5-shot) | 12.05 |
| DROP (3-shot) | 14.53 |
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 55.37 |
| AI2 Reasoning Challenge (25-Shot) | 60.07 |
| HellaSwag (10-Shot) | 82.42 |
| MMLU (5-Shot) | 55.87 |
| TruthfulQA (0-shot) | 45.57 |
| Winogrande (5-shot) | 76.24 |
| GSM8k (5-shot) | 12.05 |
- Downloads last month
- 9
Model tree for cognitivecomputations/minotaur-llama2-13b-qlora
Base model
TheBloke/Llama-2-13B-fp16Spaces using cognitivecomputations/minotaur-llama2-13b-qlora 21
Evaluation results
- normalized accuracy on AI2 Reasoning Challenge (25-Shot)test set Open LLM Leaderboard60.070
- normalized accuracy on HellaSwag (10-Shot)validation set Open LLM Leaderboard82.420
- accuracy on MMLU (5-Shot)test set Open LLM Leaderboard55.870
- mc2 on TruthfulQA (0-shot)validation set Open LLM Leaderboard45.570
- accuracy on Winogrande (5-shot)validation set Open LLM Leaderboard76.240
- accuracy on GSM8k (5-shot)test set Open LLM Leaderboard12.050