CalibForge-35B-A3B-v2

Following ricdomolm's feedback and published exclusion list, this model was retrained with the 405 flagged tasks excluded from its training data. The updated CalibForge dataset contains 5,026 tasks.

Overview of CalibForge task authoring, validation, and adversarial solver calibration

📄 Paper · 💻 Repository · 🤗 Dataset

CalibForge-35B-A3B-v2 is a terminal-agent model fine-tuned from Qwen3.5-35B-A3B using multi-turn SFT on CalibForge solver trajectories.

Evaluation Results

Model Terminal-Bench 2.1 Acc. (%, mean ± SEM) ↑
CalibForge-35B-A3B-v2 46.44 ± 1.63

Results are reported as mean ± SEM over three runs. Evaluation uses CalibForge-Eval, with a one-hour per-task timeout and 16 CPUs / 48 GiB RAM per sandbox.

Usage with SGLang

Use a recent SGLang release with Qwen3.5 MoE support:

python -m sglang.launch_server \
  --model-path AweAI-Team/CalibForge-35B-A3B-v2 \
  --tp 8 \
  --dp 1 \
  --trust-remote-code \
  --reasoning-parser qwen3 \
  --tool-call-parser qwen3_coder \
  --mem-fraction-static 0.85 \
  --port 40003

License

The model is released under the Apache License 2.0.

Citation

@misc{meng2026calibforge,
  title = {CalibForge: Adversarial Solver Calibration for Scaling Learnable Terminal Tasks},
  author = {Fanzhe Meng and Guoxin Chen and Jiale Zhao and Shuang Sun and Yuzhi Lin and Wayne Xin Zhao and Ruihua Song and Ji-Rong Wen and Kai Jia},
  year = {2026},
  eprint = {2608.06352},
  archivePrefix = {arXiv},
  primaryClass = {cs.LG},
  url = {https://arxiv.org/abs/2608.06352}
}
Downloads last month
45
Safetensors
Model size
35B params
Tensor type
BF16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for AweAI-Team/CalibForge-35B-A3B-v2

Finetuned
(206)
this model
Quantizations
2 models

Dataset used to train AweAI-Team/CalibForge-35B-A3B-v2

Paper for AweAI-Team/CalibForge-35B-A3B-v2