Update model card and evaluation figures

.gitattributes

@@ -34,3 +34,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
 tokenizer.json filter=lfs diff=lfs merge=lfs -text
+assets/calibforge-overview.png filter=lfs diff=lfs merge=lfs -text
+assets/tb2-category-results.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,3 +1,89 @@
 ---
 license: apache-2.0
+language:
+- en
+datasets:
+- AweAI-Team/CalibForge
+base_model:
+- Qwen/Qwen3-30B-A3B-Instruct-2507
 ---
+
+# CalibForge-30B-A3B
+
+📄 [Paper](https://arxiv.org/abs/2608.06352) · 💻 [Repository](https://github.com/AweAI-Team/CalibForge) · 🤗 [Dataset](https://huggingface.co/datasets/AweAI-Team/CalibForge) · 🤖 [35B Model](https://huggingface.co/AweAI-Team/CalibForge-35B-A3B)
+
+CalibForge-30B-A3B is a terminal-agent model fine-tuned from [Qwen3-30B-A3B-Instruct-2507](https://huggingface.co/Qwen/Qwen3-30B-A3B-Instruct-2507) using full-parameter, multi-turn supervised fine-tuning. Its training trajectories were distilled from **5,431** calibrated terminal tasks spanning 16 domains. **CalibForge** uses verified solver outcomes and full trajectories to revise executable terminal tasks toward a solver-relative learnable zone. See the [paper](https://arxiv.org/abs/2608.06352) for the task-construction pipeline.
+
+<p align="center">
+  <img src="assets/calibforge-overview.png" alt="Overview of CalibForge and its evaluation results" width="100%">
+</p>
+
+## Evaluation Results
+
+| Model | Terminal-Bench 2.0 Acc. (%) ↑ | SWE-bench Pro Resolved (%) ↑ | Doc2Repo Pass Rate (%) ↑ |
+|---|---:|---:|---:|
+| Qwen3-30B-A3B-Instruct-2507 (base) | 7.87 ± 0.00 | 3.26 | 5.94 ± 0.88 |
+| **CalibForge-30B-A3B (this model)** | **32.58 ± 1.12** | **30.94** | **35.98 ± 1.82** |
+
+Terminal-Bench 2.0 and Doc2Repo results are reported as mean ± SEM over three runs; SWE-bench Pro is evaluated once. Terminal-Bench 2.0 uses [CalibForge-Eval](https://github.com/AweAI-Team/AweAgent/tree/main/recipes/terminal_bench_v2) with a 500-step limit, a one-hour per-task timeout, and sandboxes capped at 16 CPUs and 32 GB RAM. SWE-bench Pro and Doc2Repo use their official evaluation scaffolds; corresponding AweAgent recipes are available for [SWE-bench Pro](https://github.com/AweAI-Team/AweAgent/tree/main/recipes/swe_bench_pro) and [Doc2Repo through BeyondSWE](https://github.com/AweAI-Team/AweAgent/tree/main/recipes/beyond_swe).
+
+<p align="center">
+  <img src="assets/tb2-category-results.png" alt="Terminal-Bench 2.0 tasks solved by category for CalibForge-30B-A3B and its base model" width="720">
+</p>
+
+## Usage with SGLang
+
+Install a recent SGLang release with Qwen3 MoE support, then launch an OpenAI-compatible server:
+
+```bash
+uv venv --python 3.12 .venv
+source .venv/bin/activate
+uv pip install "sglang[all]"
+
+python -m sglang.launch_server \
+  --model-path AweAI-Team/CalibForge-30B-A3B \
+  --tp 8 \
+  --dp 1 \
+  --trust-remote-code \
+  --enable-metrics \
+  --max-running-requests 40 \
+  --tool-call-parser qwen \
+  --mem-fraction-static 0.95 \
+  --host :: \
+  --port 40003
+```
+
+## Model Details
+
+| Property | Value |
+|---|---|
+| Backbone | Qwen3-30B-A3B-Instruct-2507 |
+| Precision | BF16 |
+| Configured maximum positions | 262,144 |
+| Training method | Full-parameter, multi-turn SFT |
+| Training context length | 131,072 tokens |
+| Training epochs | 10 |
+| Optimizer | AdamW (`β₁=0.9`, `β₂=0.999`) |
+| Learning rate | `1.0e-5`, cosine schedule, 0.05 warmup ratio |
+| Global batch size | 128 |
+| Training hardware | 64 × NVIDIA H20 GPUs |
+
+## License
+
+The model is released under the [Apache License 2.0](https://www.apache.org/licenses/LICENSE-2.0).
+
+## Citation
+
+If you use this model, please cite the CalibForge paper:
+
+```bibtex
+@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 Zhiyu 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}
+}
+```