Refine model card layout
Browse files
README.md
CHANGED
|
@@ -14,7 +14,12 @@ base_model:
|
|
| 14 |
<img src="assets/calibforge-overview.png" alt="Overview of CalibForge and its evaluation results" width="100%">
|
| 15 |
</p>
|
| 16 |
|
| 17 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
|
| 19 |
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.
|
| 20 |
|
|
@@ -28,7 +33,7 @@ CalibForge-30B-A3B is a terminal-agent model fine-tuned from [Qwen3-30B-A3B-Inst
|
|
| 28 |
Terminal-Bench 2.0 and Doc2Repo results are reported as mean 卤 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).
|
| 29 |
|
| 30 |
<p align="center">
|
| 31 |
-
<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="
|
| 32 |
</p>
|
| 33 |
|
| 34 |
## Usage with SGLang
|
|
|
|
| 14 |
<img src="assets/calibforge-overview.png" alt="Overview of CalibForge and its evaluation results" width="100%">
|
| 15 |
</p>
|
| 16 |
|
| 17 |
+
<p align="center">
|
| 18 |
+
馃搫 <a href="https://arxiv.org/abs/2608.06352">Paper</a> 路
|
| 19 |
+
馃捇 <a href="https://github.com/AweAI-Team/CalibForge">Repository</a> 路
|
| 20 |
+
馃 <a href="https://huggingface.co/datasets/AweAI-Team/CalibForge">Dataset</a> 路
|
| 21 |
+
馃 <a href="https://huggingface.co/AweAI-Team/CalibForge-35B-A3B">35B Model</a>
|
| 22 |
+
</p>
|
| 23 |
|
| 24 |
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.
|
| 25 |
|
|
|
|
| 33 |
Terminal-Bench 2.0 and Doc2Repo results are reported as mean 卤 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).
|
| 34 |
|
| 35 |
<p align="center">
|
| 36 |
+
<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="400">
|
| 37 |
</p>
|
| 38 |
|
| 39 |
## Usage with SGLang
|