Add README for dt-4bT-mixed-s50

README.md

@@ -0,0 +1,82 @@
+---
+license: apache-2.0
+base_model: Qwen/Qwen3-4B-Thinking-2507
+tags:
+  - multi-turn-agent
+  - search-r1
+  - cross-domain-transfer
+  - sft
+pipeline_tag: text-generation
+---
+
+# wenyii/dt-4bT-mixed-s50
+
+**Multi-turn agent SFT on Qwen3-4B-Thinking-2507** — `mixed` arm,
+checkpoint at training step `s50`.
+
+Part of an ICLR-target analysis study on cross-domain transfer of multi-turn
+agent SFT (Qwen3-4B-Thinking, dual-terminator recipe). Sister checkpoints in
+the same family:
+
+| Arm | Steps available |
+|---|---|
+| `r2e` (code) | `s50`, `s100`, `ep1` (=s218) |
+| `search` | `s50`, `s100` |
+| `mixed` (50/50) | `s50`, `s100` |
+
+Full step curves (s50/s100/s150/s200/ep1/ep2) available on request.
+
+## Training recipe ("dt" = dual-terminator)
+
+- **Base**: `Qwen/Qwen3-4B-Thinking-2507`
+- **Data**: 3500 multi-turn trajectories per arm
+  - `r2e`: code-agent trajectories (R2E / SWE-Gym style, ending in
+    `<tool_call>{finish}`)
+  - `search`: search-agent trajectories (OpenSeeker + NQ-style, ending
+    in `Final answer: X`)
+  - `mixed`: 50/50 shuffle of r2e + search
+- **Recipe diff vs. v5 (one-line edit)**: for the r2e arm, append
+  `Final answer: <pr_description-headline, ≤200 chars>` epilogue AFTER
+  the standard `<tool_call>{finish}`. This dual-terminates the trajectory
+  so a downstream literal-EM evaluator (Search-R1) can still extract the
+  final answer.
+- **Optimizer**: LR 1e-5, cosine schedule, warmup 0.03, weight decay 0,
+  grad clip 1.0, batch 16, FSDP1 + Liger + flash-attn (bf16).
+- **Saves**: steps 50, 100, 150, 200, 218 (=ep1), 436 (=ep2 for
+  search/mixed; r2e capped at ep1).
+
+## How to use
+
+```python
+from transformers import AutoModelForCausalLM, AutoTokenizer
+tok = AutoTokenizer.from_pretrained("wenyii/dt-4bT-mixed-s50")
+mdl = AutoModelForCausalLM.from_pretrained("wenyii/dt-4bT-mixed-s50", torch_dtype="bfloat16")
+```
+
+Or with vLLM:
+
+```bash
+python -m vllm.entrypoints.openai.api_server \
+    --model wenyii/dt-4bT-mixed-s50 --port 8000 --tensor-parallel-size 1
+```
+
+For full multi-turn agent inference (retriever + tool-call parsing + answer
+cascade extractor), see the reference SR1 7-bench eval pipeline
+distributed alongside the paper.
+
+## Headline numbers (Search-R1 7-bench, N=500 per bench, mml=16384)
+
+dt-r2e-s100 reaches **EM 0.344 / F1 0.430** on SR1 7-bench, vs Qwen3-4B-Thinking-2507
+base + `Final answer:` prompt nudge anchor at EM 0.302 / F1 0.389. See the
+companion paper for the full 7-bench / per-arm / multi-seed analysis.
+
+## Citation
+
+```bibtex
+@inproceedings{wenyimo2027dt4b,
+  title={...},
+  author={Mo, Wenyi and ...},
+  booktitle={ICLR 2027},
+  year={2027}
+}
+```