from transformers import pipeline, TextStreamer
import torch

generate_text = pipeline(
    "text-generation",
    model="Felladrin/Minueza-2-96M-Instruct-Variant-01",
    device=torch.device("cuda" if torch.cuda.is_available() else "cpu"),
)

messages = [
    {
        "role": "system",
        "content": "You are an assistant with vast experience in opening companies.",
    },
    {
        "role": "user",
        "content": "Hi!",
    },
    {
        "role": "assistant",
        "content": "Hello! How can I help you?",
    },
    {
        "role": "user",
        "content": "List the main challenges of opening a company.",
    },
]

generate_text(
    generate_text.tokenizer.apply_chat_template(
        messages, tokenize=False, add_generation_prompt=True
    ),
    streamer=TextStreamer(generate_text.tokenizer, skip_special_tokens=True),
    max_new_tokens=512,
    do_sample=True,
    temperature=0.7,
    top_p=0.9,
    top_k=0,
    min_p=0.1,
    repetition_penalty=1.12,
)

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 7.5e-05
train_batch_size: 64
eval_batch_size: 1
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 128
optimizer: Use adamw_torch with betas=(0.9,0.95) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_ratio: 0.1
training_steps: 7186

Framework versions

Transformers 4.50.0
Pytorch 2.6.0+cu124
Datasets 3.4.1
Tokenizers 0.21.0

License

This model is licensed under the Apache License 2.0.