VisFlamCat
VisFlamCat is a merge of the following models using LazyMergekit:
𧩠Configuration
models:
- model: Nitral-AI/Visual-LaylelemonMaidRP-7B
#no parameters necessary for base model
- model: flammenai/flammen15-gutenberg-DPO-v1-7B
parameters:
density: 0.5
weight: 0.5
- model: Eric111/CatunaLaserPi
parameters:
density: 0.5
weight: 0.5
merge_method: ties
base_model: Nitral-AI/Visual-LaylelemonMaidRP-7B
parameters:
normalize: false
int8_mask: true
dtype: float16
π» Usage
!pip install -qU transformers accelerate
from transformers import AutoTokenizer
import transformers
import torch
model = "Stark2008/VisFlamCat"
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"])
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
Metric | Value |
---|---|
Avg. | 21.16 |
IFEval (0-Shot) | 43.66 |
BBH (3-Shot) | 32.88 |
MATH Lvl 5 (4-Shot) | 6.57 |
GPQA (0-shot) | 5.37 |
MuSR (0-shot) | 14.68 |
MMLU-PRO (5-shot) | 23.82 |
- Downloads last month
- 9
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Model tree for Stark2008/VisFlamCat
Merge model
this model
Evaluation results
- strict accuracy on IFEval (0-Shot)Open LLM Leaderboard43.660
- normalized accuracy on BBH (3-Shot)Open LLM Leaderboard32.880
- exact match on MATH Lvl 5 (4-Shot)Open LLM Leaderboard6.570
- acc_norm on GPQA (0-shot)Open LLM Leaderboard5.370
- acc_norm on MuSR (0-shot)Open LLM Leaderboard14.680
- accuracy on MMLU-PRO (5-shot)test set Open LLM Leaderboard23.820