Checkpoints

The repo root contains the completed first baseline adapter:

• Source checkpoint: qwen36-baseline/checkpoint-23
• Base model: Qwen/Qwen3.6-27B
• LoRA rank: 32
• LoRA alpha: 64
• Task type: CAUSAL_LM
• Training framework: Axolotl + PEFT/QLoRA

An archived initial test adapter is kept at:

• initial-qlora-test/checkpoint-50

That archived adapter is retained for provenance only. Most users should load the root adapter.

Training Data

The first baseline used a locally generated clean SFT dataset from the three-autoresearch project:

• Train records: 520
• Validation records: 30
• Format: chat-style supervised fine-tuning examples
• Accepted examples only: yes
• Raw Codex transcripts included in SFT messages: no
• Held-out eval task banks included in training: no

The SFT examples were produced from AutoResearch-style task factories with locked evaluators and scalar scorers. Passing runs were converted into clean instruction/solution records.

Training Metrics

Root baseline adapter:

Initial QLoRA test adapter:

These are internal SFT validation metrics only. They are not held-out browser-task pass rates.

Held-Out Evaluation Status

The next required evaluation is task-level generation against the locked held-out eval banks:

• Three/R3F/WebGPU eval tasks
• HyperFrames/Canvas/WebGPU eval tasks

Those evals should run the model-generated editable file through the existing Python, TypeScript, Playwright, canvas, and HyperFrames scorers. Until that is complete, this adapter should be treated as a first SFT baseline rather than a proven production coding model.

Usage

Example PEFT loading pattern:

python
Copy code
from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel
import torch

base_model = "Qwen/Qwen3.6-27B"
adapter = "drainer/qwen36-threehf-sft-adapter"

tokenizer = AutoTokenizer.from_pretrained(base_model, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
    base_model,
    torch_dtype=torch.bfloat16,
    device_map="auto",
    trust_remote_code=True,
)
model = PeftModel.from_pretrained(model, adapter)
model.eval()

messages = [
    {
        "role": "user",
        "content": "Return only a complete src/solution.tsx React Three Fiber scene using Drei OrbitControls.",
    }
]
text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tokenizer(text, return_tensors="pt").to(model.device)

with torch.no_grad():
    output = model.generate(
        **inputs,
        max_new_tokens=2048,
        temperature=0.2,
        top_p=0.9,
    )

print(tokenizer.decode(output[0], skip_special_tokens=True))

Serving With vLLM LoRA

Example:

bash
Copy code
vllm serve Qwen/Qwen3.6-27B   --enable-lora   --lora-modules qwen36-threehf=drainer/qwen36-threehf-sft-adapter   --max-model-len 16384   --dtype bfloat16

Then call the served model name qwen36-threehf.

Intended Use

This adapter is intended for experimentation with specialized code generation for browser graphics and 3D app-building tasks, especially tasks that output complete editable files for locked harnesses.

Good target outputs include:

• src/solution.tsx for React Three Fiber task harnesses
• index.html for HyperFrames/Canvas task harnesses
• Compact, self-contained scene implementations
• Deterministic visual compositions suitable for automated scoring

Limitations

• This is a small first-pass SFT run, not a fully validated specialist model.
• The adapter has not yet been benchmarked against the held-out browser execution evals.
• It may overfit to the local task harness output style.
• It may still generate invalid imports, incomplete files, or visually incorrect scenes.
• It is not a merged model; the base model is required.
• Use of this adapter is subject to the base model's license and terms.

Provenance

• Project repo: https://github.com/trevor-ofarrell/three-autoresearch
• Adapter repo owner: Trevor O'Farrell
• Final metrics record on training machine: /workspace/data/three-autoresearch-training/first_run_final_metrics.json
• Baseline reload check: passed; generated a valid-looking React Three Fiber sample from the adapter.