Instructions to use muradil211/ToolWeave_stage3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use muradil211/ToolWeave_stage3 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="muradil211/ToolWeave_stage3")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("muradil211/ToolWeave_stage3")
model = AutoModelForCausalLM.from_pretrained("muradil211/ToolWeave_stage3", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use muradil211/ToolWeave_stage3 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "muradil211/ToolWeave_stage3"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "muradil211/ToolWeave_stage3",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/muradil211/ToolWeave_stage3

SGLang

How to use muradil211/ToolWeave_stage3 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "muradil211/ToolWeave_stage3" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "muradil211/ToolWeave_stage3",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "muradil211/ToolWeave_stage3" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "muradil211/ToolWeave_stage3",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use muradil211/ToolWeave_stage3 with Docker Model Runner:
```
docker model run hf.co/muradil211/ToolWeave_stage3
```

muradil211 commited on 4 days ago

Commit

4236e7e

verified ·

1 Parent(s): 5c9a776

docs: publish final Stage 3 model card

Browse files

Files changed (1) hide show

README.md +13 -12

README.md CHANGED Viewed

@@ -5,7 +5,9 @@ tags:
 - qwen3
 - tool-calling
 - bfcl
-- reference-checkpoint
 ---
 <div align="center">
@@ -14,8 +16,8 @@ tags:
 <h1>ToolWeave · Stage 3</h1>
-<p><strong>🧭 Boundary-Guided Online-Learning Reference</strong></p>
-<p>A reference starting point for ToolWeave's boundary-guided tool-use stage.</p>
 <p>
   <a href="https://github.com/Muradil-mamat-211/ToolWeave">🧵 Project</a>
@@ -23,18 +25,19 @@ tags:
 </div>
-> ⚠️ **Release status:** reference checkpoint only. Formal ToolWeave Stage 3 training remains pending.
 ## 🧭 At a glance
 | Field | Details |
 |---|---|
-| 🧠 Reference model | Qwen3-4B-RODS |
-| 🪜 Intended role | Stage 3 reference / starting point |
-| 🔬 Project status | Formal ToolWeave Stage 3 training pending |
-| 📦 Release status | Reference checkpoint only |
-This repository stores the Qwen3-4B-RODS reference checkpoint used in the ToolWeave Stage 3 workspace. It must not be interpreted as a completed ToolWeave Stage 3 release or as the official RODS implementation. ToolWeave is a separate project that adapts public ideas and infrastructure with project-specific extensions.
 ## 🚀 Usage
@@ -48,8 +51,6 @@ model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")
 Tool-use inference requires the model's function schemas and the Qwen3-compatible tool-call format.
-## 🔗 References
 - [🧵 ToolWeave project](https://github.com/Muradil-mamat-211/ToolWeave)
-- [🤗 Upstream RODS model](https://huggingface.co/RuishanFang/Qwen3-4B-RODS)
-- [📄 RODS paper](https://arxiv.org/abs/2606.19047)

 - qwen3
 - tool-calling
 - bfcl
+- agentic-rl
+- progress-reward
+- online-synthesis
 ---
 <div align="center">
 <h1>ToolWeave · Stage 3</h1>
+<p><strong>🧭 Boundary-Guided Online Reinforcement Learning</strong></p>
+<p>Verified online data synthesis for multi-turn tool-calling agents.</p>
 <p>
   <a href="https://github.com/Muradil-mamat-211/ToolWeave">🧵 Project</a>
 </div>
+> ✅ **Release status:** final ToolWeave Stage 3 model.
 ## 🧭 At a glance
 | Field | Details |
 |---|---|
+| 🧠 Base family | Qwen3-4B-Instruct |
+| 🪜 Curriculum stage | Stage 3 — Boundary-Guided Online Learning |
+| 🧱 Starting point | ToolWeave Stage 2 update 25 |
+| 🎛️ Training signal | Verified online data synthesis + multi-turn Progress Reward |
+| ✅ Release status | Final ToolWeave Stage 3 model |
+ToolWeave Stage 3 expands multi-turn tool-use learning through capability-boundary detection, verified online data synthesis, strict execution and semantic validation, dynamic replay, and combined global/local tool-call credit.
 ## 🚀 Usage
 Tool-use inference requires the model's function schemas and the Qwen3-compatible tool-call format.
+## 🔗 Links
 - [🧵 ToolWeave project](https://github.com/Muradil-mamat-211/ToolWeave)