Instructions to use poolside-laguna-hackathon/laguna-xs2-dense-stage2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use poolside-laguna-hackathon/laguna-xs2-dense-stage2 with Transformers:

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

pipe = pipeline("text-generation", model="poolside-laguna-hackathon/laguna-xs2-dense-stage2", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("poolside-laguna-hackathon/laguna-xs2-dense-stage2", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("poolside-laguna-hackathon/laguna-xs2-dense-stage2", trust_remote_code=True, 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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use poolside-laguna-hackathon/laguna-xs2-dense-stage2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "poolside-laguna-hackathon/laguna-xs2-dense-stage2"
# 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-hackathon/laguna-xs2-dense-stage2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/poolside-laguna-hackathon/laguna-xs2-dense-stage2

SGLang

How to use poolside-laguna-hackathon/laguna-xs2-dense-stage2 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-hackathon/laguna-xs2-dense-stage2" \
    --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-hackathon/laguna-xs2-dense-stage2",
		"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-hackathon/laguna-xs2-dense-stage2" \
        --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-hackathon/laguna-xs2-dense-stage2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use poolside-laguna-hackathon/laguna-xs2-dense-stage2 with Docker Model Runner:
```
docker model run hf.co/poolside-laguna-hackathon/laguna-xs2-dense-stage2
```

duico commited on May 30

Commit

22e06c0

verified ·

1 Parent(s): 75f6161

Upload README.md with huggingface_hub

Browse files

Files changed (1) hide show

README.md +16 -3

README.md CHANGED Viewed

@@ -47,13 +47,26 @@ print(tok.decode(m.generate(**ids, max_new_tokens=64)[0], skip_special_tokens=Tr
 ## Results
-| Model | Params | Held-out PPL | HumanEval pass@1 |
 |---|---|---|---|
 | Teacher (Laguna XS.2, fp8) | 33B (3B active) | ~4.4 | _running_ |
 | Stage-1 dense | ~3B | ~25 | _running_ |
-| **Stage-2 dense (this)** | ~3B | _tbd_ | _running_ |
-_(HumanEval pass@1 via [evalplus](https://github.com/evalplus/evalplus) — being measured; will be filled in.)_
 **Qualitative (honest):** still degenerate. Stage-1 emits non-Pythonic repetition (`return n;` ×∞); Stage-2 emits *valid* Python and the correct function structure but loops (`return 0` repeated). The KD moves it directionally toward coherent code — it just needs far more than 14M tokens.

 ## Results
+| Model | Params (resident) | Held-out PPL | HumanEval pass@1 |
 |---|---|---|---|
 | Teacher (Laguna XS.2, fp8) | 33B (3B active) | ~4.4 | _running_ |
 | Stage-1 dense | ~3B | ~25 | _running_ |
+| **Stage-2 dense (this)** | ~3B | _tbd_ | **0.6%** (1/164) |
+_(HumanEval pass@1 via [evalplus](https://github.com/evalplus/evalplus), greedy. Teacher + Stage-1 numbers being measured; will be filled in.)_
+Stage-2's ~0.6% confirms the model is **not yet usable** for code generation — expected at this distillation budget. The value demonstrated here is the **architecture + footprint** (below) and the **directional** KL/quality improvement over Stage-1; closing the task-accuracy gap needs much more KD.
+## VRAM / footprint
+The MoE keeps all 33B params resident even though only ~3B are active per token; the dense model keeps only the ~3B.
+| Model | Params resident | bf16 weights | fp8 weights |
+|---|---|---|---|
+| XS.2 (33B MoE) | 33.4 B | ~67 GB | ~34 GB |
+| **XS.2-dense (stage_2)** | ~3.0 B | **~6 GB** | ~3 GB |
+→ **~11× smaller weight footprint** (~61 GB saved, bf16). Caveats: attention is unchanged, so **KV-cache memory is identical** to the teacher (all savings are in the weights); per-token **FLOPs are ~unchanged** (the MoE was already ~3B-active) — the win is **memory/deployability**, not speed. XS.2 needs an 80 GB-class GPU (or fp8 on 48 GB); the dense model fits a 16 GB consumer GPU. _(The exported checkpoint here is zero-padded to ~3.8B / 7.7 GB for stock-modeling compat; the true model is 3.0B / ~6 GB.)_
 **Qualitative (honest):** still degenerate. Stage-1 emits non-Pythonic repetition (`return n;` ×∞); Stage-2 emits *valid* Python and the correct function structure but loops (`return 0` repeated). The KD moves it directionally toward coherent code — it just needs far more than 14M tokens.