metadata

license: apache-2.0
tags:
  - Safetensors
  - mistral
  - text-generation-inference
  - merge
  - mistral
  - 7b
  - mistralai/Mistral-7B-Instruct-v0.1
  - jondurbin/bagel-7b-v0.1
  - transformers
  - safetensors
  - mistral
  - text-generation
  - dataset:ai2_arc
  - dataset:unalignment/spicy-3.1
  - dataset:codeparrot/apps
  - dataset:facebook/belebele
  - dataset:boolq
  - dataset:jondurbin/cinematika-v0.1
  - dataset:drop
  - dataset:lmsys/lmsys-chat-1m
  - dataset:TIGER-Lab/MathInstruct
  - dataset:cais/mmlu
  - dataset:Muennighoff/natural-instructions
  - dataset:openbookqa
  - dataset:piqa
  - dataset:Vezora/Tested-22k-Python-Alpaca
  - dataset:cakiki/rosetta-code
  - dataset:Open-Orca/SlimOrca
  - dataset:spider
  - dataset:squad_v2
  - dataset:migtissera/Synthia-v1.3
  - dataset:datasets/winogrande
  - license:apache-2.0
  - autotrain_compatible
  - endpoints_compatible
  - has_space
  - text-generation-inference
  - region:us

bagel-7b-v0.1-Mistral-7B-Instruct-v0.1

bagel-7b-v0.1-Mistral-7B-Instruct-v0.1 is a merge of the following models:

🧩 Configuration

slices:
  - sources:
      - model: mistralai/Mistral-7B-Instruct-v0.1
        layer_range: [0, 32]
      - model: jondurbin/bagel-7b-v0.1
        layer_range: [0, 32]
merge_method: slerp
base_model: mistralai/Mistral-7B-Instruct-v0.1
parameters:
  t:
    - filter: self_attn
      value: [0, 0.5, 0.3, 0.7, 1]
    - filter: mlp
      value: [1, 0.5, 0.7, 0.3, 0]
    - value: 0.5
dtype: bfloat16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "MaziyarPanahi/bagel-7b-v0.1-Mistral-7B-Instruct-v0.1"
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"])