akacaptain/dragonclaw_model

Model summary

akacaptain/dragonclaw_model is a LoRA (PEFT) adapter fine-tuned on top of Meta’s meta-llama/Llama-3.2-3B-Instruct.

It is not a standalone full model checkpoint: you must load the base Llama-3.2-3B-Instruct model and then apply this adapter.

How to use (Transformers)

Prereqs:

• You must be granted access to the gated base model on Hugging Face and be logged in (huggingface-cli login).

Load:

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

base_model = "meta-llama/Llama-3.2-3B-Instruct"
adapter_repo = "akacaptain/dragonclaw_model"

tok = AutoTokenizer.from_pretrained(adapter_repo)
base = AutoModelForCausalLM.from_pretrained(base_model)
model = PeftModel.from_pretrained(base, adapter_repo)
model.eval()

messages = [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Hello!"},
]
prompt = tok.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tok(prompt, return_tensors="pt")

with torch.inference_mode():
    out = model.generate(
        **inputs,
        max_new_tokens=128,
        do_sample=False,
        pad_token_id=tok.eos_token_id,
    )

print(tok.decode(out[0], skip_special_tokens=True))

Training details

• Base model: meta-llama/Llama-3.2-3B-Instruct
• Method: LoRA / PEFT (see adapter_config.json)
• Training hardware: NVIDIA RTX 4090
• Approximate training duration: ~10 minutes
• Data: Fine-tuned on synthetically generated training data derived from the OpenClaw source code.

Evaluation

• Automated evaluation: TODO (or: not yet published)

Limitations

• Inherits the limitations and usage constraints of the base Llama-3.2-3B-Instruct model.
• Synthetic training data can produce confident-sounding but incorrect configuration advice; always verify in real environments.

License / attribution

• Ensure your Hub license/settings match the base model requirements and your organization’s policy.