Instructions to use prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct with Transformers:

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

pipe = pipeline("text-generation", model="prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct")
model = AutoModelForCausalLM.from_pretrained("prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct")
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 prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct

SGLang

How to use prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct 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 "prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct" \
    --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": "prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct",
		"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 "prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct" \
        --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": "prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct with Docker Model Runner:
```
docker model run hf.co/prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct
```

SmolLM2-Math-IIO-1.7B-Instruct / README.md

prithivMLmods

Update README.md

1db58ba verified over 1 year ago

preview code

raw

history blame contribute delete

3.86 kB

	---
	license: creativeml-openrail-m
	datasets:
	- prithivMLmods/Math-IIO-68K-Mini
	language:
	- en
	base_model:
	- HuggingFaceTB/SmolLM2-1.7B-Instruct
	pipeline_tag: text-generation
	library_name: transformers
	tags:
	- safetensors
	- pytorch
	- llama
	- trl
	- ollama
	- llama-cpp
	- math
	- instruct
	---

	### SmolLM2-Math-IIO-1.7B-Instruct

	The SmolLM2-Math-IIO-1.7B-Instruct model is a fine-tuned variant of the SmolLM2-1.7B architecture, optimized for mathematical instruction and reasoning tasks. It is particularly suited for applications that require mathematical problem-solving, logical inference, and detailed step-by-step explanations.

	\| File Name \| Size \| Description \| Upload Status \|
	\|----------------------------------------\|------------\|------------------------------------------------\|----------------\|
	\| `.gitattributes` \| 1.52 kB \| Git attributes configuration file \| Uploaded \|
	\| `README.md` \| 287 Bytes \| Updated README file \| Updated \|
	\| `config.json` \| 940 Bytes \| Model configuration settings \| Uploaded \|
	\| `generation_config.json` \| 162 Bytes \| Generation-specific configurations \| Uploaded \|
	\| `merges.txt` \| 515 kB \| Merging information for tokenization \| Uploaded \|
	\| `pytorch_model.bin` \| 3.42 GB \| Full model weights (PyTorch format) \| Uploaded (LFS) \|
	\| `special_tokens_map.json` \| 572 Bytes \| Mapping for special tokens used by the model \| Uploaded \|
	\| `tokenizer.json` \| 3.77 MB \| Tokenizer configuration and vocabulary \| Uploaded \|
	\| `tokenizer_config.json` \| 3.95 kB \| Tokenizer configuration for loading and usage \| Uploaded \|
	\| `vocab.json` \| 801 kB \| Vocabulary for the tokenizer \| Uploaded \|

	### Key Features:

	1. Math-Focused Capabilities:
	This model is fine-tuned to handle a wide range of mathematical queries, from simple arithmetic to complex equations and mathematical proofs.

	2. Instruction-Tuned:
	Specifically trained to follow structured queries and deliver clear, coherent outputs based on instructions, ensuring high-quality, relevant responses to prompts.

	3. Tokenizer & Custom Tokens:
	Includes a robust tokenizer configuration with support for mathematical notation, custom tokens, and an extended vocabulary for accurate understanding and output generation.

	---

	### Training Details:
	- Base Model: [SmolLM2-1.7B](https://huggingface.co/HuggingFaceTB/SmolLM2-1.7B-Instruct)
	- Dataset: Trained on Math-IIO-68K-Mini, a dataset focused on mathematical instructions and logic-based queries, with a total of 68.8k examples.

	### Capabilities:
	- Mathematical Problem-Solving: Solves and explains complex mathematical problems, including algebra, calculus, and more advanced topics.
	- Instruction-Following: Adheres to structured inputs and outputs, making it effective for generating step-by-step solutions.
	- Text Generation: Capable of generating mathematical proofs, explanations, and educational content tailored to various user queries.

	---

	### Usage Instructions:
	1. Model Setup: Download all model files and ensure the PyTorch model weights and tokenizer configurations are included.
	2. Inference: Load the model in a Python environment using frameworks like PyTorch or Hugging Face's Transformers.
	3. Customization: Configure the model with the `config.json` and `generation_config.json` files for optimal performance during inference.

	---

	---
	license: creativeml-openrail-m
	datasets:
	- prithivMLmods/Math-IIO-68K-Mini
	language:
	- en
	base_model:
	- HuggingFaceTB/SmolLM2-1.7B-Instruct
	pipeline_tag: text-generation
	library_name: transformers
	tags:
	- safetensors
	- pytorch
	- llama
	- trl
	- ollama
	- llama-cpp
	- math
	- instruct
	---

	### SmolLM2-Math-IIO-1.7B-Instruct

	The SmolLM2-Math-IIO-1.7B-Instruct model is a fine-tuned variant of the SmolLM2-1.7B architecture, optimized for mathematical instruction and reasoning tasks. It is particularly suited for applications that require mathematical problem-solving, logical inference, and detailed step-by-step explanations.

	\| File Name \| Size \| Description \| Upload Status \|
	\|----------------------------------------\|------------\|------------------------------------------------\|----------------\|
	\| `.gitattributes` \| 1.52 kB \| Git attributes configuration file \| Uploaded \|
	\| `README.md` \| 287 Bytes \| Updated README file \| Updated \|
	\| `config.json` \| 940 Bytes \| Model configuration settings \| Uploaded \|
	\| `generation_config.json` \| 162 Bytes \| Generation-specific configurations \| Uploaded \|
	\| `merges.txt` \| 515 kB \| Merging information for tokenization \| Uploaded \|
	\| `pytorch_model.bin` \| 3.42 GB \| Full model weights (PyTorch format) \| Uploaded (LFS) \|
	\| `special_tokens_map.json` \| 572 Bytes \| Mapping for special tokens used by the model \| Uploaded \|
	\| `tokenizer.json` \| 3.77 MB \| Tokenizer configuration and vocabulary \| Uploaded \|
	\| `tokenizer_config.json` \| 3.95 kB \| Tokenizer configuration for loading and usage \| Uploaded \|
	\| `vocab.json` \| 801 kB \| Vocabulary for the tokenizer \| Uploaded \|

	### Key Features:

	1. Math-Focused Capabilities:
	This model is fine-tuned to handle a wide range of mathematical queries, from simple arithmetic to complex equations and mathematical proofs.

	2. Instruction-Tuned:
	Specifically trained to follow structured queries and deliver clear, coherent outputs based on instructions, ensuring high-quality, relevant responses to prompts.

	3. Tokenizer & Custom Tokens:
	Includes a robust tokenizer configuration with support for mathematical notation, custom tokens, and an extended vocabulary for accurate understanding and output generation.

	---

	### Training Details:
	- Base Model: [SmolLM2-1.7B](https://huggingface.co/HuggingFaceTB/SmolLM2-1.7B-Instruct)
	- Dataset: Trained on Math-IIO-68K-Mini, a dataset focused on mathematical instructions and logic-based queries, with a total of 68.8k examples.

	### Capabilities:
	- Mathematical Problem-Solving: Solves and explains complex mathematical problems, including algebra, calculus, and more advanced topics.
	- Instruction-Following: Adheres to structured inputs and outputs, making it effective for generating step-by-step solutions.
	- Text Generation: Capable of generating mathematical proofs, explanations, and educational content tailored to various user queries.

	---

	### Usage Instructions:
	1. Model Setup: Download all model files and ensure the PyTorch model weights and tokenizer configurations are included.
	2. Inference: Load the model in a Python environment using frameworks like PyTorch or Hugging Face's Transformers.
	3. Customization: Configure the model with the `config.json` and `generation_config.json` files for optimal performance during inference.

	---