The following models were included in the merge:
Configuration
The following YAML configuration was used to produce this model:
slices:
- sources:
- model: Test157t/Pasta-Sea-7b-128k
layer_range: [0, 32]
- model: Locutusque/Hercules-2.5-Mistral-7B
layer_range: [0, 32]
merge_method: slerp
base_model: Test157t/Pasta-Sea-7b-128k
parameters:
t:
- filter: self_attn
value: [0, 0.5, 0.3, 0.7, 1]
- filter: mlp
value: [1, 0.5, 0.7, 0.3, 0]
- value: 0.5
dtype: float16
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 69.03 |
| AI2 Reasoning Challenge (25-Shot) | 66.13 |
| HellaSwag (10-Shot) | 85.89 |
| MMLU (5-Shot) | 64.48 |
| TruthfulQA (0-shot) | 55.54 |
| Winogrande (5-shot) | 81.22 |
| GSM8k (5-shot) | 60.96 |
- Downloads last month
- 39
Model tree for Nitral-Archive/HerculeanSea-upd-7b-128k
Base model
mistralai/Mistral-7B-v0.1 Finetuned
Locutusque/Hercules-2.5-Mistral-7BEvaluation results
- normalized accuracy on AI2 Reasoning Challenge (25-Shot)test set Open LLM Leaderboard66.130
- normalized accuracy on HellaSwag (10-Shot)validation set Open LLM Leaderboard85.890
- accuracy on MMLU (5-Shot)test set Open LLM Leaderboard64.480
- mc2 on TruthfulQA (0-shot)validation set Open LLM Leaderboard55.540
- accuracy on Winogrande (5-shot)validation set Open LLM Leaderboard81.220
- accuracy on GSM8k (5-shot)test set Open LLM Leaderboard60.960