LinguaLaboratoriumMechanicus-Qwen-1.7B-instruct

SFT-версия модели LinguaLaboratoriumMechanicus-Qwen-1.7B на Q&A по лору WH40k.

Исходная база: Qwen/Qwen3-1.7B-Base (Apache-2.0).

Использование

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

repo_id = "GoldenGekko/LinguaLaboratoriumMechanicus-Qwen-1.7B-instruct"
device = "cuda" if torch.cuda.is_available() else "cpu"

tokenizer = AutoTokenizer.from_pretrained(repo_id)
model = AutoModelForCausalLM.from_pretrained(
    repo_id,
    torch_dtype=torch.bfloat16,
).to(device)

messages = [{"role": "user", "content": "Что такое Гибельный шторм?"}]
input_ids = tokenizer.apply_chat_template(
    messages,
    add_generation_prompt=True,
    tokenize=True,
    return_tensors="pt",
    enable_thinking=False,
).to(device)

output_ids = model.generate(
    input_ids,
    max_new_tokens=120,
    do_sample=False,
    pad_token_id=tokenizer.pad_token_id,
    eos_token_id=tokenizer.eos_token_id,
)
print(tokenizer.decode(output_ids[0, input_ids.shape[1]:], skip_special_tokens=True))

Связанные репозитории

• Qwen/Qwen3-1.7B-Base - исходная база
• LinguaLaboratoriumMechanicus-Qwen-1.7B - CPT

Citation

https://huggingface.co/GoldenGekko/LinguaLaboratoriumMechanicus-Qwen-1.7B-instruct