Instructions to use WhirlwindAI/Arithmetic-SLM with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use WhirlwindAI/Arithmetic-SLM with Transformers:

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

pipe = pipeline("text-generation", model="WhirlwindAI/Arithmetic-SLM", trust_remote_code=True)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("WhirlwindAI/Arithmetic-SLM", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use WhirlwindAI/Arithmetic-SLM with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "WhirlwindAI/Arithmetic-SLM"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "WhirlwindAI/Arithmetic-SLM",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/WhirlwindAI/Arithmetic-SLM

SGLang

How to use WhirlwindAI/Arithmetic-SLM 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 "WhirlwindAI/Arithmetic-SLM" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "WhirlwindAI/Arithmetic-SLM",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "WhirlwindAI/Arithmetic-SLM" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "WhirlwindAI/Arithmetic-SLM",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use WhirlwindAI/Arithmetic-SLM with Docker Model Runner:
```
docker model run hf.co/WhirlwindAI/Arithmetic-SLM
```

PhysiQuanty commited on Jul 6

Commit

a952cc2

verified ·

1 Parent(s): 948d6da

Update README.md

Browse files

Files changed (1) hide show

README.md +6 -26

README.md CHANGED Viewed

@@ -65,7 +65,7 @@ The goal is not to make a general chatbot. The goal is to train a compact model
     <td align="center"><strong>82.08%</strong></td>
   </tr>
   <tr>
-    <td align="center"><code>PhysiQuanty/Arithmetic-SLM</code></td>
     <td align="center">31.70M</td>
     <td align="center"><strong>78.60%</strong></td>
   </tr>
@@ -165,7 +165,7 @@ The goal is not to make a general chatbot. The goal is to train a compact model
 ```bash
 python3 inference.py \
-  --model PhysiQuanty/Arithmetic-SLM \
   --prompt "59 + 45 =" \
   --max-new-tokens 32 \
   --temperature 0.6 \
@@ -184,7 +184,7 @@ Expected style:
 ```bash
 python3 inference.py \
-  --model PhysiQuanty/Arithmetic-SLM \
   --prompt "0.5 * 0.5 =" \
   --no-think \
   --max-new-tokens 48 \
@@ -213,7 +213,7 @@ Example output:
 ```bash
 python3 inference.py \
-  --model PhysiQuanty/Arithmetic-SLM \
   --prompt "8 * 5 + 4 / 4 =" \
   --no-think \
   --max-new-tokens 48 \
@@ -233,7 +233,7 @@ Expected style:
 ```bash
 python3 inference.py \
-  --model PhysiQuanty/Arithmetic-SLM \
   --prompt "(85 - 45) + 56 =" \
   --no-think \
   --max-new-tokens 48 \
@@ -253,7 +253,7 @@ Expected style:
 ```bash
 python3 inference.py \
-  --model PhysiQuanty/Arithmetic-SLM \
   --prompt "3 * 9 + 12 / 1 =" \
   --no-think \
   --max-new-tokens 48 \
@@ -269,26 +269,6 @@ Expected style:
 3 * 9 + 12 / 1 = 39
 ```
-### Example 6 — BOS/EOS base mode for base models
-Use this mode for base models that were not trained with the Qwen-style `[IM_START]user ... /no think[IM_END]` format.
-```bash
-python3 inference.py \
-  --model Supra-50M-Base-local \
-  --prompt "8 * 5 + 4 / 4 =" \
-  --no-qwen-format \
-  --max-new-tokens 32 \
-  --temperature 0.6 \
-  --top-k 50 \
-  --top-p 0.97 \
-  --repetition-penalty 1 \
-  --frequency-penalty 0.0 \
-  --no-repeat-ngram-size 0 \
-  --seed -1 \
-  --print-full
-```
 ## Next Research Directions
 We will continue improving our dataset engineering, but more importantly, we want to teach the model what most models are never explicitly taught:

     <td align="center"><strong>82.08%</strong></td>
   </tr>
   <tr>
+    <td align="center"><code>WhirlwindAI/Arithmetic-SLM</code></td>
     <td align="center">31.70M</td>
     <td align="center"><strong>78.60%</strong></td>
   </tr>
 ```bash
 python3 inference.py \
+  --model WhirlwindAI/Arithmetic-SLM \
   --prompt "59 + 45 =" \
   --max-new-tokens 32 \
   --temperature 0.6 \
 ```bash
 python3 inference.py \
+  --model WhirlwindAI/Arithmetic-SLM \
   --prompt "0.5 * 0.5 =" \
   --no-think \
   --max-new-tokens 48 \
 ```bash
 python3 inference.py \
+  --model WhirlwindAI/Arithmetic-SLM \
   --prompt "8 * 5 + 4 / 4 =" \
   --no-think \
   --max-new-tokens 48 \
 ```bash
 python3 inference.py \
+  --model WhirlwindAI/Arithmetic-SLM \
   --prompt "(85 - 45) + 56 =" \
   --no-think \
   --max-new-tokens 48 \
 ```bash
 python3 inference.py \
+  --model WhirlwindAI/Arithmetic-SLM \
   --prompt "3 * 9 + 12 / 1 =" \
   --no-think \
   --max-new-tokens 48 \
 3 * 9 + 12 / 1 = 39
 ```
 ## Next Research Directions
 We will continue improving our dataset engineering, but more importantly, we want to teach the model what most models are never explicitly taught: