Overview

We fine-tuned Qwen3-30B-A3B-Thinking with Agent Context Compilation (ACC) — a method that converts multi-turn agent trajectories (Search, SWE, SQL) into long-context QA pairs for direct supervised fine-tuning. Unlike standard agent SFT that masks tool responses, ACC assembles scattered evidence across turns into a single context, enabling explicit supervision of long-range dependency modeling.

Performance Highlights

Results on MRCR and GraphWalks are comparable to Qwen3-235B-A22B despite ~8× fewer active parameters. General capabilities are preserved.

Training Data

• Dataset: groundhogLLM/ACC-dataset
• Size: 10,802 compiled trajectories (Search: 3,369; SWE: 4,368; SQL: 3,065)
• Context length: 2K – 128K tokens
• Training seq length: 131,072 tokens
• Epochs: 4

Usage

python
Copy code
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "groundhogLLM/ACC-Qwen3-30B-A3B"
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype="auto",
    device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained(model_id)

# Standard Qwen3 chat template applies
messages = [{"role": "user", "content": "Your long-context question here..."}]
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=1024)
print(tokenizer.decode(outputs[0]))

Citation

If you use this model, please cite:

bibtex
Copy code
@misc{su2026acccompilingagenttrajectories,
      title={ACC: Compiling Agent Trajectories for Long-Context Training}, 
      author={Qisheng Su and Zhen Fang and Shiting Huang and Yu Zeng and Yiming Zhao and Kou Shi and Ziao Zhang and Lin Chen and Zehui Chen and Lijun Wu and Feng Zhao},
      year={2026},
      eprint={2605.21850},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2605.21850}, 
}