README.md

metadata

tags:
  - merge
  - mergekit
  - lazymergekit
  - VAGOsolutions/SauerkrautLM-7b-HerO
  - cognitivecomputations/dolphin-2.8-mistral-7b-v02
base_model:
  - VAGOsolutions/SauerkrautLM-7b-HerO
  - cognitivecomputations/dolphin-2.8-mistral-7b-v02

Sappho_V0.0.2

Sappho_V0.0.2 is a merge of the following models using LazyMergekit:

• VAGOsolutions/SauerkrautLM-7b-HerO
• cognitivecomputations/dolphin-2.8-mistral-7b-v02

🧩 Configuration

models:
  - model: mlabonne/NeuralHermes-2.5-Mistral-7B # no parameters necessary for base model
  - model: VAGOsolutions/SauerkrautLM-7b-HerO
    parameters:
      density: 0.3  # fraction of weights in differences from the base model to retain
      weight:   # weight gradient
        - filter: mlp
          value: 0.5
        - value: 0
  - model: cognitivecomputations/dolphin-2.8-mistral-7b-v02
    parameters:
      density: 0.5
      weight: 0.4
merge_method: ties
base_model: mlabonne/NeuralHermes-2.5-Mistral-7B
parameters:
  normalize: true
  int8_mask: true
dtype: float16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Jakolo121/Sappho_V0.0.2"
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"])