Instructions to use agentica-org/DeepSWE-Preview with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use agentica-org/DeepSWE-Preview with Transformers:

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

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

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("agentica-org/DeepSWE-Preview")
model = AutoModelForCausalLM.from_pretrained("agentica-org/DeepSWE-Preview")
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]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use agentica-org/DeepSWE-Preview with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/agentica-org/DeepSWE-Preview

SGLang

How to use agentica-org/DeepSWE-Preview 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 "agentica-org/DeepSWE-Preview" \
    --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": "agentica-org/DeepSWE-Preview",
		"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 "agentica-org/DeepSWE-Preview" \
        --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": "agentica-org/DeepSWE-Preview",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use agentica-org/DeepSWE-Preview with Docker Model Runner:
```
docker model run hf.co/agentica-org/DeepSWE-Preview
```

michaelzhiluo commited on Jul 2, 2025

Commit

1e1f473

verified ·

1 Parent(s): 41c1d5a

Update README.md

Browse files

Files changed (1) hide show

README.md +3 -3

README.md CHANGED Viewed

@@ -40,7 +40,7 @@ pipeline_tag: text-generation
   </div>
   </a>
-  <a href="www.google.com" target="_blank" style="margin: 2px;">
     <img alt="Blog" src="https://img.shields.io/badge/Notion-%23000000.svg?style=for-the-badge&logo=notion&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
   </a>
   <a href="https://x.com/Agentica_" style="margin: 2px;">
@@ -62,7 +62,7 @@ DeepSWE-Preview is a fully open-sourced, state-of-the-art coding agent trained w
 DeepSWE-Preview is trained on top of Qwen3-32B with thinking mode enabled. With just 200 steps of RL training, SWE-Bench-Verified score increases by ~20%.
-Discover more about DeepSWE-Preview's development and capabilities in our [technical blog post](www.google.com).
 <div style="margin: 0 auto;">
   <img src="./figures/swebench.png" style="width: 100%;" />
@@ -121,7 +121,7 @@ We enhance the original GRPO algorithm, integrating insights from DAPO, Dr. GRPO
 - **Compact Filtering** **(Us):** Inspired by DAPO, we mask the loss for trajectories that reach max context length, timeout during generation (20 minutes), or reach maximum steps.
 - **No Entropy Loss (Us):** Entropy loss introduces higher instability and eventually leads to exponentially increasing entropy, which collapses training. Provided that the base model’s token-level entropy is within 0.3-1, entropy loss is not needed.
-A more detailed description of the training recipe can be found in our [blog post](www.google.com).
 ## Evaluation

   </div>
   </a>
+  <a href="https://pretty-radio-b75.notion.site/DeepSWE-Training-a-Fully-Open-sourced-State-of-the-Art[%E2%80%A6]-by-Scaling-RL-22281902c1468193aabbe9a8c59bbe33" target="_blank" style="margin: 2px;">
     <img alt="Blog" src="https://img.shields.io/badge/Notion-%23000000.svg?style=for-the-badge&logo=notion&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
   </a>
   <a href="https://x.com/Agentica_" style="margin: 2px;">
 DeepSWE-Preview is trained on top of Qwen3-32B with thinking mode enabled. With just 200 steps of RL training, SWE-Bench-Verified score increases by ~20%.
+Discover more about DeepSWE-Preview's development and capabilities in our [technical blog post](https://pretty-radio-b75.notion.site/DeepSWE-Training-a-Fully-Open-sourced-State-of-the-Art[%E2%80%A6]-by-Scaling-RL-22281902c1468193aabbe9a8c59bbe33).
 <div style="margin: 0 auto;">
   <img src="./figures/swebench.png" style="width: 100%;" />
 - **Compact Filtering** **(Us):** Inspired by DAPO, we mask the loss for trajectories that reach max context length, timeout during generation (20 minutes), or reach maximum steps.
 - **No Entropy Loss (Us):** Entropy loss introduces higher instability and eventually leads to exponentially increasing entropy, which collapses training. Provided that the base model’s token-level entropy is within 0.3-1, entropy loss is not needed.
+A more detailed description of the training recipe can be found in our [blog post](https://pretty-radio-b75.notion.site/DeepSWE-Training-a-Fully-Open-sourced-State-of-the-Art[%E2%80%A6]-by-Scaling-RL-22281902c1468193aabbe9a8c59bbe33).
 ## Evaluation