README.md

metadata

license: apache-2.0
tags:
  - moe
  - frankenmoe
  - merge
  - mergekit
  - lazymergekit
  - Isaak-Carter/TinyJ.O.S.I.E.-1.1B-32k-Base
  - Isaak-Carter/TinyJ.O.S.I.E.-1.1B-32k-Base
base_model:
  - Isaak-Carter/TinyJ.O.S.I.E.-1.1B-32k-Base
  - Isaak-Carter/TinyJ.O.S.I.E.-1.1B-32k-Base

TinyJ.O.S.I.E.-2x1.1B-32k-Base

TinyJ.O.S.I.E.-2x1.1B-32k-Base is a Mixure of Experts (MoE) made with the following models using LazyMergekit:

• Isaak-Carter/TinyJ.O.S.I.E.-1.1B-32k-Base
• Isaak-Carter/TinyJ.O.S.I.E.-1.1B-32k-Base

🧩 Configuration

dtype: float32
gate_mode: hidden
experts:
  - source_model: Isaak-Carter/TinyJ.O.S.I.E.-1.1B-32k-Base
    positive_prompts:
      - 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
    negative_prompts:
      - 'You should not be used for coding tasks.'

  - source_model: Isaak-Carter/TinyJ.O.S.I.E.-1.1B-32k-Base
    positive_prompts:
      - 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent only for coding assistants.'

💻 Usage

!pip install -qU transformers bitsandbytes accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Isaak-Carter/TinyJ.O.S.I.E.-2x1.1B-32k-Base"

tokenizer = AutoTokenizer.from_pretrained(model)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    model_kwargs={"torch_dtype": torch.float16, "load_in_4bit": True},
)

messages = [{"role": "user", "content": "Explain what a Mixture of Experts is in less than 100 words."}]
prompt = pipeline.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
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"])