Update README.md

README.md

@@ -1,3 +1,104 @@
 ---
 license: apache-2.0
 ---
+# SLAI T-Rex-Flash (DeepSeek-V4-Flash-OR)
+
+SLAI T-Rex-Flash is an Operations Research (OR)–specialized model built from DeepSeek-V4-Flash through full-parameter continued pre-training (CPT) followed by supervised fine-tuning (SFT). It is designed for solver-oriented mathematical modeling, including formulation, Gurobi program generation, feasibility-oriented tasks, and structure-sensitive optimization modeling.
+
+- Paper: [SLAI T-Rex: Full-Parameter Post-training of the DeepSeek-V4 Family on Ascend SuperPOD](https://arxiv.org/abs/2607.20145)
+- Paper PDF: [arXiv:2607.20145](https://arxiv.org/pdf/2607.20145)
+- Code and training recipes: [SLAI-AITP/SLAI-T-Rex](https://github.com/SLAI-AITP/SLAI-T-Rex)
+- ModelScope weights: [SLAIAITP/DeepSeek-V4-Flash-OR](https://www.modelscope.cn/models/SLAIAITP/DeepSeek-V4-Flash-OR)
+
+## Model Summary
+
+| Item | Description |
+| --- | --- |
+| Model name | SLAI T-Rex-Flash |
+| Base model | DeepSeek-V4-Flash |
+| Domain | Operations Research and mathematical programming |
+| Training | Full-parameter OR-oriented CPT followed by Clean-CoT SFT |
+| Primary output | Mathematical formulation and/or solver-facing Python code |
+| Training hardware | Ascend CloudMatrix384 SuperPOD with Ascend 910C NPUs |
+
+The CPT stage supplies OR terminology, formulation patterns, solver APIs, and structural priors. The SFT stage aligns these capabilities with task instructions, output contracts, mathematical explanations, and executable Gurobi-style implementations. The reported SFT recipe uses an 8,192-token sequence length, AdamW, a global batch size of 128, and full-parameter optimization.
+
+## Download
+
+Install ModelScope and download the checkpoint:
+
+```bash
+python3 -m pip install -U modelscope
+```
+
+```python
+from modelscope import snapshot_download
+
+model_dir = snapshot_download("SLAIAITP/DeepSeek-V4-Flash-OR")
+print(model_dir)
+```
+
+The checkpoint should be served with a runtime that supports the DeepSeek-V4 architecture and the checkpoint format in this repository. Distributed serving flags depend on the target hardware, runtime version, and available memory. Check the repository configuration files before deployment.
+
+## Intended Use
+
+SLAI T-Rex-Flash is intended for research and development involving:
+
+- translating natural-language OR problems into mathematical programs;
+- identifying variables, objectives, and constraints;
+- generating Gurobi-style Python implementations;
+- producing solver-compatible models under strict output contracts;
+- studying domain-adaptive CPT and SFT for structured reasoning.
+
+A typical prompt should state the optimization problem, all required data, the expected solver behavior, and the output contract. For tagged modeling tasks, the training data commonly use `<think>`, `<model>`, and `<python>` sections. Code-only and LP-writing tasks use their task-specific contracts instead.
+
+## Evaluation
+
+The following results are reported in the technical report under zero-shot Pass@1 evaluation. `Overall` is the unweighted mean of the four OR benchmarks.
+
+| Model | NL4OPT | OptiBench | B4O-Feasible | B4O-ORGEval | Overall |
+| --- | ---: | ---: | ---: | ---: | ---: |
+| DeepSeek-V4-Flash | 84.08 | 63.33 | 60.47 | 34.26 | 60.54 |
+| SLAI T-Rex-Flash | **89.52** | **67.12** | **71.22** | **59.39** | **71.81** |
+
+These results correspond to the complete CPT+SFT pipeline. The reported overall score improves by 11.27 percentage points over the original DeepSeek-V4-Flash checkpoint. The largest absolute gain is on B4O-ORGEval, which evaluates structural equivalence rather than execution success alone.
+
+The report also evaluates general-capability retention:
+
+| Benchmark | SLAI T-Rex-Flash | Base |
+| --- | ---: | ---: |
+| MMLU, 5-shot | 88.5 | 87.6 |
+| MMLU-Pro, 5-shot | 69.2 | 71.0 |
+| CMMLU, 5-shot | 92.4 | 92.1 |
+| HumanEval, 0-shot | 68.8 | 69.4 |
+| GSM8K, 8-shot | 90.3 | 89.8 |
+| MATH, 4-shot | 56.7 | 58.4 |
+
+Evaluation numbers should be compared only under the same prompt templates, decoding budgets, benchmark versions, solver environment, and scoring implementation described in the paper.
+
+## Limitations
+
+- Generated optimization models may be executable but mathematically incorrect; solver execution is not proof of structural equivalence.
+- The model can still make errors in variable domains, nonlinear expressions, ratios, quadratic terms, index alignment, and Gurobi-specific APIs.
+- Performance is strongest on the OR task families represented by the training and evaluation pipeline. Results may not transfer to unrelated domains or unseen solver ecosystems.
+- The reported results do not establish reliability for safety-critical, financial, medical, industrial, or legal decision making.
+- Always validate generated formulations, data interfaces, solver status, constraints, and objective values before use.
+
+## License
+
+Use of this checkpoint is governed by the license files and terms included in this repository. Users must also comply with any applicable terms of the DeepSeek-V4-Flash base checkpoint and third-party software such as Gurobi.
+
+## Citation
+
+If you use this model, please cite the technical report:
+
+```bibtex
+@misc{li2026slaitrexfullparameterposttraining,
+  title        = {SLAI T-Rex: Full-Parameter Post-training of the DeepSeek-V4 Family on Ascend SuperPOD},
+  author       = {Dongfang Li and others},
+  year         = {2026},
+  eprint       = {2607.20145},
+  archivePrefix= {arXiv},
+  url          = {https://arxiv.org/abs/2607.20145}
+}
+```