Instructions to use poolside/Laguna-XS.2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use poolside/Laguna-XS.2 with Transformers:

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

pipe = pipeline("text-generation", model="poolside/Laguna-XS.2", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("poolside/Laguna-XS.2", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("poolside/Laguna-XS.2", trust_remote_code=True)
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

vLLM

How to use poolside/Laguna-XS.2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "poolside/Laguna-XS.2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "poolside/Laguna-XS.2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/poolside/Laguna-XS.2

SGLang

How to use poolside/Laguna-XS.2 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 "poolside/Laguna-XS.2" \
    --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": "poolside/Laguna-XS.2",
		"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 "poolside/Laguna-XS.2" \
        --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": "poolside/Laguna-XS.2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use poolside/Laguna-XS.2 with Docker Model Runner:
```
docker model run hf.co/poolside/Laguna-XS.2
```

Laguna-XS.2

Commit History

Update .eval_results/terminal-bench-2.0.yaml

1162d99
verified

joerowell commited on 1 day ago

Update .eval_results/swe-bench_verified.yaml

c5fe6fd
verified

joerowell commited on 1 day ago

Update .eval_results/swe-bench_pro.yaml

3b9ed24
verified

joerowell commited on 1 day ago

update sampling parameters to match evals

07c6c8e

joerowell commited on 2 days ago

Update model card for 256K context length

a6da52e
verified

varunrandery commited on 3 days ago

increase context length to 256k

339151d

joerowell commited on 3 days ago

Add tech report

cfa46f7
verified

joerowell commited on 3 days ago

Drop VLLM_USE_DEEP_GEMM=0 from vllm serve recipe (DeepGEMM is supported on Hopper and datacenter Blackwell)

bb5158c
verified

joerowell commited on 9 days ago

Update vLLM install instructions: 0.21.0+ supports Laguna XS, no nightly needed

3c519e6
verified

joerowell commited on 10 days ago

Add DFlash speculative decoding section

e5916e1

joerowell commited on 21 days ago

Update README.md

7ffaee5
verified

joerowell commited on 23 days ago

Fix typo

d38ff3c
verified

Pengming commited on 26 days ago

Fix typo

f07da4d
verified

Pengming commited on 26 days ago

Enable thinking by default in recommended vllm serve command

be45e1e
verified

joerowell commited on 30 days ago

Update vLLM serve command and link to vLLM recipes page

30ca10b
verified

joerowell commited on 30 days ago

Drop install-from-source note for Transformers

de04ba5
verified

joerowell commited on about 1 month ago

Update vLLM install instructions to use nightly wheel

598f732
verified

joerowell commited on about 1 month ago

Add community evaluation results for SWE-BENCH_PRO, SWE-BENCH_VERIFIED, TERMINAL-BENCH-2.0 (#5)

06367f7

joerowell

nielsr HF Staff commited on Apr 28

Laguna XS.2 upload

7b8e693

joerowell commited on Apr 28

Commit History

Update .eval_results/terminal-bench-2.0.yaml 1162d99 verified

Update .eval_results/swe-bench_verified.yaml c5fe6fd verified

Update .eval_results/swe-bench_pro.yaml 3b9ed24 verified

update sampling parameters to match evals 07c6c8e

Update model card for 256K context length a6da52e verified

increase context length to 256k 339151d

Add tech report cfa46f7 verified

Drop VLLM_USE_DEEP_GEMM=0 from vllm serve recipe (DeepGEMM is supported on Hopper and datacenter Blackwell) bb5158c verified

Update vLLM install instructions: 0.21.0+ supports Laguna XS, no nightly needed 3c519e6 verified

Add DFlash speculative decoding section e5916e1

Update README.md 7ffaee5 verified

Fix typo d38ff3c verified

Fix typo f07da4d verified

Enable thinking by default in recommended vllm serve command be45e1e verified

Update vLLM serve command and link to vLLM recipes page 30ca10b verified

Drop install-from-source note for Transformers de04ba5 verified

Update vLLM install instructions to use nightly wheel 598f732 verified

Add community evaluation results for SWE-BENCH_PRO, SWE-BENCH_VERIFIED, TERMINAL-BENCH-2.0 (#5) 06367f7

Laguna XS.2 upload 7b8e693

Update .eval_results/terminal-bench-2.0.yaml

1162d99
verified

Update .eval_results/swe-bench_verified.yaml

c5fe6fd
verified

Update .eval_results/swe-bench_pro.yaml

3b9ed24
verified

update sampling parameters to match evals

07c6c8e

Update model card for 256K context length

a6da52e
verified

increase context length to 256k

339151d

Add tech report

cfa46f7
verified

Drop VLLM_USE_DEEP_GEMM=0 from vllm serve recipe (DeepGEMM is supported on Hopper and datacenter Blackwell)

bb5158c
verified

Update vLLM install instructions: 0.21.0+ supports Laguna XS, no nightly needed

3c519e6
verified

Add DFlash speculative decoding section

e5916e1

Update README.md

7ffaee5
verified

Fix typo

d38ff3c
verified

Fix typo

f07da4d
verified

Enable thinking by default in recommended vllm serve command

be45e1e
verified

Update vLLM serve command and link to vLLM recipes page

30ca10b
verified

Drop install-from-source note for Transformers

de04ba5
verified

Update vLLM install instructions to use nightly wheel

598f732
verified

Add community evaluation results for SWE-BENCH_PRO, SWE-BENCH_VERIFIED, TERMINAL-BENCH-2.0 (#5)

06367f7

Laguna XS.2 upload

7b8e693