Instructions to use afrideva/InstructWise-462M-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use afrideva/InstructWise-462M-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="afrideva/InstructWise-462M-GGUF",
	filename="instructwise-462m.q2_k.gguf",
)

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use afrideva/InstructWise-462M-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf afrideva/InstructWise-462M-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf afrideva/InstructWise-462M-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf afrideva/InstructWise-462M-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf afrideva/InstructWise-462M-GGUF:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf afrideva/InstructWise-462M-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf afrideva/InstructWise-462M-GGUF:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf afrideva/InstructWise-462M-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf afrideva/InstructWise-462M-GGUF:Q4_K_M

Use Docker

docker model run hf.co/afrideva/InstructWise-462M-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use afrideva/InstructWise-462M-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "afrideva/InstructWise-462M-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "afrideva/InstructWise-462M-GGUF",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/afrideva/InstructWise-462M-GGUF:Q4_K_M

Ollama
How to use afrideva/InstructWise-462M-GGUF with Ollama:
```
ollama run hf.co/afrideva/InstructWise-462M-GGUF:Q4_K_M
```

Unsloth Studio new

How to use afrideva/InstructWise-462M-GGUF with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for afrideva/InstructWise-462M-GGUF to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for afrideva/InstructWise-462M-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for afrideva/InstructWise-462M-GGUF to start chatting

Docker Model Runner
How to use afrideva/InstructWise-462M-GGUF with Docker Model Runner:
```
docker model run hf.co/afrideva/InstructWise-462M-GGUF:Q4_K_M
```

Lemonade

How to use afrideva/InstructWise-462M-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull afrideva/InstructWise-462M-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.InstructWise-462M-GGUF-Q4_K_M

List all available models

lemonade list

CrabfishAI/InstructWise-462M-GGUF

Quantized GGUF model files for InstructWise-462M from CrabfishAI

Name	Quant method	Size
instructwise-462m.fp16.gguf	fp16	925.45 MB
instructwise-462m.q2_k.gguf	q2_k	212.56 MB
instructwise-462m.q3_k_m.gguf	q3_k_m	238.87 MB
instructwise-462m.q4_k_m.gguf	q4_k_m	288.52 MB
instructwise-462m.q5_k_m.gguf	q5_k_m	333.29 MB
instructwise-462m.q6_k.gguf	q6_k	380.87 MB
instructwise-462m.q8_0.gguf	q8_0	492.67 MB

Original Model Card:

InstructWise 470M - A virtual assistant.

Note- we,ll be releasing more versions of InstructWise soon, with the goal of making memory-efficent models while maintaining the performance, Thank you! Introduction- InstructWise is a model created to act as helpful virtual assistant while maintaing the memory efficiency.

Credits

Base Model: ahxt/llama2_xs_460M_experimental
Dataset used: timdettmers/openassistant-guanaco
License: llama2

Features

Maintaining performance while being memory efficient: Ram usage- 7.1GB Vram usage- 0.6GB (approximately)
Act as helpful virtual assistant: InstructWise serves as a versatile and helpful assistant, offering a range of features that cater to various user needs. Its key strength lies in providing instructive responses to user prompts, offering detailed and insightful information.
Coding: Model can perform coding as well.
Assisting capabilities: can assist with wide rang of taskes.

Uses

InstructWise finds application in various domains, including:

Assistance in Writing: Aid authors, bloggers, and students in drafting articles and essays.
Chatbot Development: Power conversational agents with human-like responses.
Prototyping and Idea Generation: Facilitate brainstorming sessions for product development.
Personal Assistant Applications: Assist users in drafting emails and messages. and many more.

Direct Use Cases

InstructWise can be directly employed for:

Educational Support:
- Assist users in learning new topics with detailed explanations and step-by-step instructions.
Content Creation:
- Generate creative content based on prompts, aiding content creators in the writing process.
Code Assistance:
- Provide guidance on coding queries, improve code documentation, and generate code snippets for developers.
Interactive Conversations:
- Enhance chatbots or virtual assistants with informative and helpful responses for users.
Q&A Platforms:
- Power question and answer platforms, offering detailed and insightful answers on various topics.
Technical Writing Support:
- Assist writers and technical communicators with suggestions for clarity and informativeness.
Idea Expansion:
- Facilitate the expansion and development of ideas by providing detailed insights and suggestions.

Limitation

Content Quality: The model's output may vary in quality, and there's a possibility it might generate content that is nonsensical, irrelevant, or grammatically incorrect.
Bias and Sensitivity: The model is trained on diverse data, but it may inadvertently exhibit biases or generate content that is sensitive or inappropriate. Exercise caution and review generated text before use.
Inappropriate Language: The model might generate text that includes offensive language or inappropriate content. Be mindful of this, especially in applications where maintaining a respectful and inclusive tone is essential.

Disclaimer

Use with Caution: This model is a tool that should be used with caution. Always review and validate the generated text before incorporating it into any application or publication.
Not for Critical Applications: Avoid using the model for critical applications where accuracy and reliability are paramount. The model is intended for creative and exploratory purposes.
Ongoing Improvement: The model may be updated or fine-tuned for better performance. Stay informed about updates and consider using the latest version for improved results.

Recommended Prompt Format to use:

### Instruction:
<instruction>

### Response:

Downloads last month: 53

GGUF

Model size

0.5B params

Architecture

llama

Hardware compatibility

2-bit

3-bit

4-bit

5-bit

6-bit

8-bit

Model tree for afrideva/InstructWise-462M-GGUF

Base model

ahxt/llama2_xs_460M_experimental

Finetuned

CrabfishAI/InstructWise-462M

Quantized

(1)

this model

afrideva
/

InstructWise-462M-GGUF