Merge-Mixtral-Prometheus-8x7B is a merge of the following models using LazyMergekit:
models:
- model: prometheus-eval/prometheus-8x7b-v2.0
parameters:
weight: 1.0
- model: mistralai/Mixtral-8x7B-Instruct-v0.1
parameters:
weight: 1.0
merge_method: linear
dtype: bfloat16
!pip install -qU transformers accelerate
from transformers import AutoTokenizer
import transformers
import torch
model = "vicgalle/test-merge-3"
messages = [{"role": "user", "content": "What is a large language model?"}]
tokenizer = AutoTokenizer.from_pretrained(model)
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipeline = transformers.pipeline(
"text-generation",
model=model,
torch_dtype=torch.float16,
device_map="auto",
)
outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])
Paper: https://arxiv.org/abs/2406.07188
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 24.61 |
| IFEval (0-Shot) | 57.44 |
| BBH (3-Shot) | 34.65 |
| MATH Lvl 5 (4-Shot) | 8.31 |
| GPQA (0-shot) | 7.83 |
| MuSR (0-shot) | 9.59 |
| MMLU-PRO (5-shot) | 29.82 |
Have the files? Submit a torrent.
Submit a torrentHave the complete files and permission to share them? Replace the example folder path. The script checks the files, makes a torrent and submits its magnet and checksums. Keep your torrent client seeding. Get your community key from your settings, and read the script before running it.
curl -fsSL https://pirateface.co/package.sh | bash -s -- --repo vicgalle/Merge-Mixtral-Prometheus-8x7B ./model-foldermodel-00012-of-00048.safetensors 2.0 GBmodel-00037-of-00048.safetensors 2.0 GBmodel-00006-of-00048.safetensors 2.0 GBmodel-00009-of-00048.safetensors 2.0 GBmodel-00016-of-00048.safetensors 2.0 GBmodel-00022-of-00048.safetensors 2.0 GBmodel-00025-of-00048.safetensors 2.0 GBmodel-00028-of-00048.safetensors 2.0 GBmodel-00031-of-00048.safetensors 2.0 GBmodel-00034-of-00048.safetensors 2.0 GBmodel-00003-of-00048.safetensors 2.0 GBmodel-00019-of-00048.safetensors 2.0 GBmodel-00041-of-00048.safetensors 2.0 GBmodel-00044-of-00048.safetensors 2.0 GBmodel-00047-of-00048.safetensors 2.0 GBmodel-00005-of-00048.safetensors 2.0 GBmodel-00008-of-00048.safetensors 2.0 GBmodel-00011-of-00048.safetensors 2.0 GBmodel-00015-of-00048.safetensors 2.0 GBmodel-00018-of-00048.safetensors 2.0 GBmodel-00021-of-00048.safetensors 2.0 GBmodel-00024-of-00048.safetensors 2.0 GBmodel-00027-of-00048.safetensors 2.0 GBmodel-00030-of-00048.safetensors 2.0 GBmodel-00033-of-00048.safetensors 2.0 GBmodel-00007-of-00048.safetensors 2.0 GBmodel-00010-of-00048.safetensors 2.0 GBmodel-00014-of-00048.safetensors 2.0 GBmodel-00017-of-00048.safetensors 2.0 GBmodel-00023-of-00048.safetensors 2.0 GBmodel-00026-of-00048.safetensors 2.0 GBmodel-00029-of-00048.safetensors 2.0 GBmodel-00032-of-00048.safetensors 2.0 GBmodel-00002-of-00048.safetensors 2.0 GBmodel-00004-of-00048.safetensors 2.0 GBmodel-00036-of-00048.safetensors 2.0 GBmodel-00039-of-00048.safetensors 2.0 GBmodel-00040-of-00048.safetensors 2.0 GBmodel-00043-of-00048.safetensors 2.0 GBmodel-00046-of-00048.safetensors 2.0 GBmodel-00020-of-00048.safetensors 2.0 GBmodel-00035-of-00048.safetensors 2.0 GBmodel-00042-of-00048.safetensors 2.0 GBmodel-00045-of-00048.safetensors 2.0 GBmodel-00013-of-00048.safetensors 2.0 GBmodel-00038-of-00048.safetensors 2.0 GBmodel-00001-of-00048.safetensors 1.9 GBmodel-00048-of-00048.safetensors 671 MBmistralai/Mixtral-8x7B-Instruct-v0.1