|
--- |
|
license: apache-2.0 |
|
--- |
|
|
|
# Model mera-mix-4x7B |
|
|
|
This is a mixture of experts (MoE) model that is half as large (4 experts instead of 8) as the [Mixtral-8x7B](https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1) |
|
while been comparable to it across different benchmarks. You can use it as a drop in replacement for your Mixtral-8x7B and get much faster inference. |
|
|
|
mera-mix-4x7B achieves 76.37 on the openLLM eval v/s 72.7 by Mixtral-8x7B (as shown [here](https://huggingface.co/datasets/open-llm-leaderboard/details_mistralai__Mixtral-8x7B-Instruct-v0.1)). |
|
|
|
## OpenLLM Eval |
|
|
|
| Model | ARC |HellaSwag|MMLU |TruthfulQA|Winogrande|GSM8K|Average| |
|
|-------------------------------------------------------------|----:|--------:|----:|---------:|---------:|----:|------:| |
|
|[mera-mix-4x7B](https://huggingface.co/meraGPT/mera-mix-4x7B)|72.01| 88.82|63.67| 77.45| 84.61|71.65| 76.37| |
|
|
|
Raw eval results are available at this [gist](https://gist.github.com/codelion/78f88333230801c9bbaa6fc22078d820) |
|
|
