sf-diogenes-v0.1

This model is a merge of the Qwen/Qwen3-Next-80B-A3B-Instruct base model with the urm3l/diogenes-v.01-lora-adapter LoRA adapter.

Model Details

• Base Model: Qwen/Qwen3-Next-80B-A3B-Instruct
• LoRA Adapter: urm3l/diogenes-v.01-lora-adapter
• Merge Date: 2025-11-01
• Model Size: ~80B parameters
• Precision: BF16

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained(
    "urm3l/sf-diogenes-v0.1",
    torch_dtype="auto",
    device_map="auto",
    trust_remote_code=True
)

tokenizer = AutoTokenizer.from_pretrained(
    "urm3l/sf-diogenes-v0.1",
    trust_remote_code=True
)

messages = [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Hello!"}
]

text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)

inputs = tokenizer(text, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=512)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Training Details

This model was fine-tuned using LoRA (Low-Rank Adaptation). For training details, see the adapter repository.

Limitations

• This is a large language model and may produce incorrect or biased outputs
• Should not be used for high-stakes decision making without human oversight
• May require significant computational resources for inference

License

This model inherits the license from the base model: Apache 2.0

Citation

If you use this model, please cite the original Qwen3 paper and acknowledge the LoRA fine-tuning.