RichardErkhov/teilomillet_-_MiniMerlin-3B-4bits

Quantization made by Richard Erkhov.

Github

Discord

Request more models

MiniMerlin-3B - bnb 4bits

• Model creator: https://huggingface.co/teilomillet/
• Original model: https://huggingface.co/teilomillet/MiniMerlin-3B/

Original model description:

license: apache-2.0 language: - fr - en tags: - code widget: - text: " [|User|] Comment faire un bon plat ? [|Assistant|]"

SFT on a synthetic custom (french) dataset (2k), from general question answering, problem solving to code question. It's a POC.

from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel
import torch

model = AutoModelForCausalLM.from_pretrained(
    "teilomillet/MiniMerlin-3B",
    revision="0.1",
    return_dict=True,
    torch_dtype=torch.bfloat16,
    device_map='auto'
)

tokenizer = AutoTokenizer.from_pretrained("teilomillet/MiniMerlin-3B")
tokenizer.pad_token = tokenizer.eos_token

text = "[|User|] Comment faire un bon plat ? </s>[|Assistant|]"
inputs = tokenizer(text, return_tensors="pt").to(0)

outputs = model.generate(**inputs, max_new_tokens=800)
print(tokenizer.decode(outputs[0], skip_special_tokens=False))