BiggerWizardLM-2-7B-Extended

BiggerWizardLM-2-7B-Extended is a merge of the following models using LazyMergekit:

• Replete-AI/WizardLM-2-7b
• Replete-AI/WizardLM-2-7b
• Replete-AI/WizardLM-2-7b
• Replete-AI/WizardLM-2-7b
• Replete-AI/WizardLM-2-7b
• Replete-AI/WizardLM-2-7b
• Replete-AI/WizardLM-2-7b
• Replete-AI/WizardLM-2-7b
• Replete-AI/WizardLM-2-7b
• Replete-AI/WizardLM-2-7b
• Replete-AI/WizardLM-2-7b
• Replete-AI/WizardLM-2-7b
• Replete-AI/WizardLM-2-7b
• Replete-AI/WizardLM-2-7b
• Replete-AI/WizardLM-2-7b
• Replete-AI/WizardLM-2-7b

🧩 Configuration

  slices:
  - sources:
      - model: Replete-AI/WizardLM-2-7b
        layer_range:
          - 0
          - 4
  - sources:
      - model: Replete-AI/WizardLM-2-7b
        layer_range:
          - 3
          - 4
        parameters:
          scale:
            - filter: o_proj
              value: 0
            - filter: down_proj
              value: 0
            - value: 1
  - sources:
      - model: Replete-AI/WizardLM-2-7b
        layer_range:
          - 4
          - 8
  - sources:
      - model: Replete-AI/WizardLM-2-7b
        layer_range:
          - 7
          - 8
        parameters:
          scale:
            - filter: o_proj
              value: 0
            - filter: down_proj
              value: 0
            - value: 1
  - sources:
      - model: Replete-AI/WizardLM-2-7b
        layer_range:
          - 8
          - 12
  - sources:
      - model: Replete-AI/WizardLM-2-7b
        layer_range:
          - 11
          - 12
        parameters:
          scale:
            - filter: o_proj
              value: 0
            - filter: down_proj
              value: 0
            - value: 1
  - sources:
      - model: Replete-AI/WizardLM-2-7b
        layer_range:
          - 12
          - 16
  - sources:
      - model: Replete-AI/WizardLM-2-7b
        layer_range:
          - 15
          - 16
        parameters:
          scale:
            - filter: o_proj
              value: 0
            - filter: down_proj
              value: 0
            - value: 1
  - sources:
      - model: Replete-AI/WizardLM-2-7b
        layer_range:
          - 16
          - 20
  - sources:
      - model: Replete-AI/WizardLM-2-7b
        layer_range:
          - 19
          - 20
        parameters:
          scale:
            - filter: o_proj
              value: 0
            - filter: down_proj
              value: 0
            - value: 1
  - sources:
      - model: Replete-AI/WizardLM-2-7b
        layer_range:
          - 20
          - 24
  - sources:
      - model: Replete-AI/WizardLM-2-7b
        layer_range:
          - 23
          - 24
        parameters:
          scale:
            - filter: o_proj
              value: 0
            - filter: down_proj
              value: 0
            - value: 1
  - sources:
      - model: Replete-AI/WizardLM-2-7b
        layer_range:
          - 24
          - 28
  - sources:
      - model: Replete-AI/WizardLM-2-7b
        layer_range:
          - 27
          - 28
        parameters:
          scale:
            - filter: o_proj
              value: 0
            - filter: down_proj
              value: 0
            - value: 1
  - sources:
      - model: Replete-AI/WizardLM-2-7b
        layer_range:
          - 28
          - 32
  - sources:
      - model: Replete-AI/WizardLM-2-7b
        layer_range:
          - 31
          - 32
        parameters:
          scale:
            - filter: o_proj
              value: 0
            - filter: down_proj
              value: 0
            - value: 1
  merge_method: passthrough
  dtype: bfloat16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Gille/BiggerWizardLM-2-7B-Extended"
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"])