Simplify model card and reference EasyDistill 2.0

README.md

@@ -11,64 +11,38 @@ tags:
 - search-agent
 - tool-use
 - function-calling
-- react
 - deep-search
-- multi-hop-qa
 - qwen
 ---
 
 # SearchQwen2.5-7B
 
-**SearchQwen2.5-7B** is a compact Search Agent model from the Alibaba Cloud PAI team. It is post-trained on environment-aligned multi-hop search tasks and solver-verified ReAct trajectories generated with the EasyDistill2 pipeline.
+**SearchQwen2.5-7B** is a compact Search Agent model from the Alibaba Cloud PAI team. It is trained with environment-aligned, solver-verified search trajectories generated by [EasyDistill 2.0](https://github.com/modelscope/easydistill/tree/v2.0.0), and is designed for multi-hop search, browsing, and evidence integration.
 
-The model is optimized for iterative **search → browse → evidence integration → answer** behavior. It supports both structured Tool-Call interaction and Search-R1-style text interaction (Qwen2.5 variants).
+> SearchQwen2.5-7B 是基于 EasyDistill 2.0 环境对齐轨迹蒸馏链路训练的 Search Agent 小模型，支持结构化 `search` / `browse` 工具调用。
 
-> 中文简介：SearchQwen2.5-7B 是面向多跳问答与深度搜索任务训练的 Search Agent 小模型，重点增强结构化工具调用、网页浏览和跨来源信息整合能力。
-
-## Model details
+## Model overview
 
 | Item | Value |
 |---|---|
 | Base model | [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) |
 | Parameters | 7.62B |
-| Architecture | Qwen2, 28 layers |
-| Context length | 32,768 tokens in the released config |
-| Primary interface | Structured `search` / `browse` Tool-Call |
-| Training | Environment-aligned trajectories, Solver-in-the-Loop verification, and process-aware post-training |
-| Companion dataset | [alibaba-pai/SynSearch-Data](https://huggingface.co/datasets/alibaba-pai/SynSearch-Data) |
-
-
-
-## Evaluation
-
-We report LLM-judge accuracy (%) over two benchmark groups:
-
-- **Multi-hop QA:** 2WikiMultiHopQA, Bamboogle, HotpotQA, MuSiQue.
-- **Deep Search:** GAIA, WebWalkerQA, xbench-deepsearch, BrowseComp-ZH.
+| Context length | 32,768 |
+| Recommended interface | Structured Tool-Call |
+| Training data | [SynSearch-Data](https://huggingface.co/datasets/alibaba-pai/SynSearch-Data) |
 
-### Search-R1-style text interaction
+## Results
 
-| Model | Multi-hop QA Avg. | Deep Search Avg. | Overall Avg. |
-|---|---:|---:|---:|
-| Qwen2.5-7B-Instruct | 45.28 | 18.65 | 31.96 |
-| **SearchQwen2.5-7B** | **52.95** | **26.23** | **39.59** |
+LLM-judge accuracy (%). Multi-hop QA averages 2WikiMultiHopQA, Bamboogle, HotpotQA, and MuSiQue; Deep Search averages GAIA, WebWalkerQA, xbench-deepsearch, and BrowseComp-ZH.
 
-### Structured Tool-Call interaction
+| Interaction | Model | Multi-hop QA | Deep Search | Overall |
+|---|---|---:|---:|---:|
+| Search-R1 style | Qwen2.5-7B-Instruct | 45.28 | 18.65 | 31.96 |
+| Search-R1 style | **SearchQwen2.5-7B** | **52.95** | **26.23** | **39.59** |
+| Tool-Call | Qwen2.5-7B-Instruct | 45.23 | 24.35 | 33.33 |
+| Tool-Call | **SearchQwen2.5-7B** | **55.90** | **33.33** | **44.61** |
 
-| Model | Multi-hop QA Avg. | Deep Search Avg. | Overall Avg. |
-|---|---:|---:|---:|
-| Qwen2.5-7B-Instruct | 45.23 | 24.35 | 33.33 |
-| **SearchQwen2.5-7B** | **55.90** | **33.33** | **44.61** |
-
-Deep Search gains are consistently larger than the gains on conventional multi-hop QA, indicating improved search, browsing, and evidence-integration behavior rather than answer memorization alone.
-
-## Requirements
-
-```bash
-pip install -U transformers accelerate torch
-```
-
-## Transformers quickstart
+## Quickstart
 
 ```python
 import torch
@@ -82,36 +56,23 @@ model = AutoModelForCausalLM.from_pretrained(
     device_map="auto",
 ).eval()
 
+messages = [
+    {"role": "system", "content": "You are a search agent. Use tools before answering."},
+    {"role": "user", "content": "Which city is the birthplace of the author of The Old Man and the Sea?"},
+]
 tools = [
     {
         "type": "function",
         "function": {
             "name": "search",
-            "description": "Search the web for relevant pages.",
+            "description": "Search the web.",
             "parameters": {
                 "type": "object",
                 "properties": {"query": {"type": "string"}},
                 "required": ["query"],
             },
         },
-    },
-    {
-        "type": "function",
-        "function": {
-            "name": "browse",
-            "description": "Read the content of a web page.",
-            "parameters": {
-                "type": "object",
-                "properties": {"url": {"type": "string"}},
-                "required": ["url"],
-            },
-        },
-    },
-]
-
-messages = [
-    {"role": "system", "content": "You are a search agent. Use tools before answering."},
-    {"role": "user", "content": "Which city is the birthplace of the author of The Old Man and the Sea?"},
+    }
 ]
 
 inputs = tokenizer.apply_chat_template(
@@ -123,75 +84,18 @@ inputs = tokenizer.apply_chat_template(
     return_dict=True,
 ).to(model.device)
 
-with torch.inference_mode():
-    output = model.generate(**inputs, max_new_tokens=256, do_sample=False)
-
-new_tokens = output[0, inputs["input_ids"].shape[-1]:]
-print(tokenizer.decode(new_tokens, skip_special_tokens=False))
-```
-
-A typical first response is a structured call:
-
-```xml
-<tool_call>
-{"name": "search", "arguments": {"query": "..."}}
-</tool_call>
-```
-
-Execute the requested tool, append the tool response to the conversation, and call the model again until it returns a final answer.
-
-## vLLM deployment
-
-```bash
-vllm serve alibaba-pai/SearchQwen2.5-7B \
-  --served-model-name SearchQwen2.5-7B \
-  --enable-auto-tool-choice \
-  --tool-call-parser hermes \
-  --max-model-len 8192
+output = model.generate(**inputs, max_new_tokens=256, do_sample=False)
+print(tokenizer.decode(output[0, inputs["input_ids"].shape[-1]:], skip_special_tokens=False))
 ```
 
-For `SearchQwen2.5-7B`, the release was validated with Transformers and an OpenAI-compatible vLLM endpoint on NVIDIA L20 GPUs. The test covered checkpoint loading, structured `search` emission, tool-result consumption, and final-answer generation.
+The model returns a structured `<tool_call>`; execute the tool, append its response, and continue until a final answer is produced. An external search/browse backend is required.
 
-## Training data
+## Related resources
 
-The released companion dataset contains 5,000 audited multi-hop ReAct trajectories covering 4,182 unique tasks. Each example includes a complete `messages` history and metadata for seed provenance, hop count, difficulty, rollout statistics, and answer-equivalence auditing.
-
-## Intended use
-
-- Research on search agents and retrieval-augmented reasoning.
-- Search/browse tool-use experiments.
-- Fine-tuning, evaluation, and deployment of compact agent models.
-
-## Limitations
-
-- The model requires an external search/browse runtime; it does not provide a search index itself.
-- Final-answer quality depends on tool availability and retrieved evidence quality.
-- Tool schemas and prompt templates should remain compatible with the released chat template.
-- LLM-judge scores may vary with evaluator model, decoding settings, and search backend.
-- Do not use the model as the sole authority for safety-critical, legal, medical, or financial decisions.
-
-## Artifact integrity
-
-`SHA256SUMS` contains checksums for all weight shards, the weight index, configuration, and tokenizer.
-
-## Related releases
-
-- [SearchQwen2.5-7B](https://huggingface.co/alibaba-pai/SearchQwen2.5-7B)
-- [SearchQwen2.5-3B](https://huggingface.co/alibaba-pai/SearchQwen2.5-3B)
-- [SearchQwen3-8B](https://huggingface.co/alibaba-pai/SearchQwen3-8B)
-- [SynSearch-Data](https://huggingface.co/datasets/alibaba-pai/SynSearch-Data)
-
-## Citation
-
-```bibtex
-@misc{searchqwen2026,
-  title  = {Fake It Till You Make It: Training Deep Search Agents on Synthetic Reality},
-  author = {Alibaba Cloud PAI Team},
-  year   = {2026},
-  note   = {SearchQwen model family and SynSearch-Data}
-}
-```
+- Framework: [EasyDistill 2.0](https://github.com/modelscope/easydistill/tree/v2.0.0)
+- Data: [SynSearch-Data](https://huggingface.co/datasets/alibaba-pai/SynSearch-Data)
+- Paper: [EasyDistill: A Comprehensive Toolkit for Effective Knowledge Distillation of Large Language Models](https://aclanthology.org/2025.emnlp-demos.60/)
 
 ## License
 
-This derivative model is released under the license file included in this repository and remains subject to the terms of its base model.
+This model follows the license included in this repository and the terms of its base model. Checksums are provided in `SHA256SUMS`.