Instructions to use smjain/sap-archgen-lfm2-230M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use smjain/sap-archgen-lfm2-230M with Transformers:

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

pipe = pipeline("text-generation", model="smjain/sap-archgen-lfm2-230M")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("smjain/sap-archgen-lfm2-230M")
model = AutoModelForCausalLM.from_pretrained("smjain/sap-archgen-lfm2-230M", 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 smjain/sap-archgen-lfm2-230M with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "smjain/sap-archgen-lfm2-230M"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "smjain/sap-archgen-lfm2-230M",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/smjain/sap-archgen-lfm2-230M

SGLang

How to use smjain/sap-archgen-lfm2-230M 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 "smjain/sap-archgen-lfm2-230M" \
    --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": "smjain/sap-archgen-lfm2-230M",
		"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 "smjain/sap-archgen-lfm2-230M" \
        --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": "smjain/sap-archgen-lfm2-230M",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Desktop
Docker Model Runner
How to use smjain/sap-archgen-lfm2-230M with Docker Model Runner:
```
docker model run hf.co/smjain/sap-archgen-lfm2-230M
```

smjain commited on Jun 26

Commit

3384d9d

verified ·

1 Parent(s): a2ae3e6

Upload quality_report.json with huggingface_hub

Browse files

Files changed (1) hide show

quality_report.json +81 -0

quality_report.json ADDED Viewed

	@@ -0,0 +1,81 @@

+[
+  {
+    "arch": "Procurement copilot (novel)",
+    "ok": true,
+    "blocks": 6,
+    "components": 17,
+    "connections": 13,
+    "verify_rounds": 2,
+    "remaining_issues": [
+      "2 unconnected component(s) ['c3', 'c1']: wire each into the flow",
+      "under-connected: 13 connections for 15 components (expect >= 14)",
+      "duplicate component id(s) ['c12', 'c13']: each component must appear once",
+      "duplicate service(s) ['SAP Destination service']: wire the existing one, do NOT add duplicate services",
+      "Cloud Connector present but no SAP Connectivity service (incomplete on-prem tunnel)."
+    ],
+    "info": "24 nodes, 13 edges"
+  },
+  {
+    "arch": "Predictive maintenance (novel)",
+    "ok": true,
+    "blocks": 5,
+    "components": 10,
+    "connections": 9,
+    "verify_rounds": 0,
+    "remaining_issues": [],
+    "info": "16 nodes, 9 edges"
+  },
+  {
+    "arch": "Claims handling RAG (held-out domain)",
+    "ok": false,
+    "blocks": 0,
+    "components": 0,
+    "connections": 2,
+    "verify_rounds": 2,
+    "remaining_issues": [
+      "fewer than 2 components placed",
+      "connection c0->c1 references missing component",
+      "connection c1->c0 references missing component",
+      "does not render: 1 nodes, 0 edges"
+    ],
+    "info": "1 nodes, 0 edges"
+  },
+  {
+    "arch": "Asset management events (held-out domain)",
+    "ok": true,
+    "blocks": 5,
+    "components": 8,
+    "connections": 7,
+    "verify_rounds": 0,
+    "remaining_issues": [],
+    "info": "14 nodes, 7 edges"
+  },
+  {
+    "arch": "AI document agent (head-to-head)",
+    "ok": true,
+    "blocks": 6,
+    "components": 12,
+    "connections": 10,
+    "verify_rounds": 2,
+    "remaining_issues": [
+      "6 unconnected component(s) ['c4', 'c2', 'c9', 'c10', 'c3', 'c1']: wire each into the flow",
+      "under-connected: 10 connections for 12 components (expect >= 11)",
+      "duplicate service(s) ['SAP HANA Cloud']: wire the existing one, do NOT add duplicate services",
+      "Cloud Connector present but no SAP Connectivity service (incomplete on-prem tunnel)."
+    ],
+    "info": "19 nodes, 10 edges"
+  },
+  {
+    "arch": "Task Center (known reference)",
+    "ok": true,
+    "blocks": 2,
+    "components": 18,
+    "connections": 8,
+    "verify_rounds": 2,
+    "remaining_issues": [
+      "duplicate component id(s) ['c0', 'c1', 'c2', 'c3', 'c4', 'c5', 'c6', 'c7', 'c8']: each component must appear once",
+      "duplicate service(s) ['SAP Application Logan service', 'SAP Build Work Zone, standard', 'SAP Cloud Identity Services', 'SAP Connectivity service', 'SAP Destination service', 'SAP Document Management hub', 'SAP Private Link service', 'SAP Task Center']: wire the existing one, do NOT add duplicate services"
+    ],
+    "info": "21 nodes, 8 edges"
+  }
+]