Instructions to use steef68/ATLAS-QUANTUM with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- llama-cpp-python
How to use steef68/ATLAS-QUANTUM with llama-cpp-python:
# !pip install llama-cpp-python from llama_cpp import Llama llm = Llama.from_pretrained( repo_id="steef68/ATLAS-QUANTUM", filename="ATLAS-QUANTUM-7B-Uncensored.q2_K.gguf", )
output = llm( "Once upon a time,", max_tokens=512, echo=True ) print(output)
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use steef68/ATLAS-QUANTUM with llama.cpp:
Install from brew
brew install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama-server -hf steef68/ATLAS-QUANTUM:Q2_K # Run inference directly in the terminal: llama-cli -hf steef68/ATLAS-QUANTUM:Q2_K
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama-server -hf steef68/ATLAS-QUANTUM:Q2_K # Run inference directly in the terminal: llama-cli -hf steef68/ATLAS-QUANTUM:Q2_K
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 steef68/ATLAS-QUANTUM:Q2_K # Run inference directly in the terminal: ./llama-cli -hf steef68/ATLAS-QUANTUM:Q2_K
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 steef68/ATLAS-QUANTUM:Q2_K # Run inference directly in the terminal: ./build/bin/llama-cli -hf steef68/ATLAS-QUANTUM:Q2_K
Use Docker
docker model run hf.co/steef68/ATLAS-QUANTUM:Q2_K
- LM Studio
- Jan
- Ollama
How to use steef68/ATLAS-QUANTUM with Ollama:
ollama run hf.co/steef68/ATLAS-QUANTUM:Q2_K
- Unsloth Studio
How to use steef68/ATLAS-QUANTUM 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 steef68/ATLAS-QUANTUM 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 steef68/ATLAS-QUANTUM to start chatting
Using HuggingFace Spaces for Unsloth
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for steef68/ATLAS-QUANTUM to start chatting
- Docker Model Runner
How to use steef68/ATLAS-QUANTUM with Docker Model Runner:
docker model run hf.co/steef68/ATLAS-QUANTUM:Q2_K
- Lemonade
How to use steef68/ATLAS-QUANTUM with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull steef68/ATLAS-QUANTUM:Q2_K
Run and chat with the model
lemonade run user.ATLAS-QUANTUM-Q2_K
List all available models
lemonade list
Update README.md
#3
by derricka59 - opened
README.md
CHANGED
|
@@ -0,0 +1,145 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: llama
|
| 3 |
+
tags:
|
| 4 |
+
- quantization
|
| 5 |
+
- efficient-inference
|
| 6 |
+
- natural-language-processing
|
| 7 |
+
- language-model
|
| 8 |
+
- ai-research
|
| 9 |
+
- open-source
|
| 10 |
+
license: apache-2.0
|
| 11 |
+
datasets:
|
| 12 |
+
- dataset-name1 # Replace with the actual dataset(s) used
|
| 13 |
+
- dataset-name2
|
| 14 |
+
language: en
|
| 15 |
+
model_architecture: llama
|
| 16 |
+
model_size: 6.74B
|
| 17 |
+
quantization: Q2_K
|
| 18 |
+
inference: true
|
| 19 |
+
training_data: "This model was trained on a combination of publicly available datasets to ensure robust performance across various NLP tasks."
|
| 20 |
+
source_code: false # Set true if source code is provided; false otherwise
|
| 21 |
+
documentation: https://huggingface.co/steef68/ATLAS-QUANTUM/resolve/main/README.md
|
| 22 |
+
---
|
| 23 |
+
|
| 24 |
+
## Usage
|
| 25 |
+
|
| 26 |
+
### Installation
|
| 27 |
+
Clone the repository and install necessary dependencies:
|
| 28 |
+
```bash
|
| 29 |
+
git clone https://huggingface.co/steef68/ATLAS-QUANTUM
|
| 30 |
+
cd ATLAS-QUANTUM
|
| 31 |
+
pip install -r requirements.txt
|
| 32 |
+
|
| 33 |
+
Model Loading
|
| 34 |
+
|
| 35 |
+
Use the Hugging Face transformers library to load the model:
|
| 36 |
+
|
| 37 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 38 |
+
|
| 39 |
+
tokenizer = AutoTokenizer.from_pretrained("steef68/ATLAS-QUANTUM")
|
| 40 |
+
model = AutoModelForCausalLM.from_pretrained("steef68/ATLAS-QUANTUM", quantization=True)
|
| 41 |
+
|
| 42 |
+
inputs = tokenizer("Your input text here", return_tensors="pt")
|
| 43 |
+
outputs = model.generate(inputs["input_ids"], max_length=50)
|
| 44 |
+
print(tokenizer.decode(outputs[0]))
|
| 45 |
+
|
| 46 |
+
|
| 47 |
+
---
|
| 48 |
+
|
| 49 |
+
Applications
|
| 50 |
+
|
| 51 |
+
The ATLAS-QUANTUM model is designed for applications such as:
|
| 52 |
+
|
| 53 |
+
Text generation
|
| 54 |
+
|
| 55 |
+
Chatbots and conversational AI
|
| 56 |
+
|
| 57 |
+
Text summarization
|
| 58 |
+
|
| 59 |
+
Creative writing assistance
|
| 60 |
+
|
| 61 |
+
|
| 62 |
+
|
| 63 |
+
---
|
| 64 |
+
|
| 65 |
+
Training and Datasets
|
| 66 |
+
|
| 67 |
+
This model was fine-tuned using a highly curated dataset to ensure robust performance. Details on the specific dataset(s) used are currently placeholders (<dataset-name>) and will be updated as they are finalized.
|
| 68 |
+
|
| 69 |
+
|
| 70 |
+
---
|
| 71 |
+
|
| 72 |
+
Limitations
|
| 73 |
+
|
| 74 |
+
The model is trained on English text and may not perform optimally in other languages.
|
| 75 |
+
|
| 76 |
+
Due to the 2-bit quantization, slight reductions in accuracy may occur in certain edge cases.
|
| 77 |
+
|
| 78 |
+
Runtime issues may occur in environments not optimized for quantized models.
|
| 79 |
+
|
| 80 |
+
|
| 81 |
+
|
| 82 |
+
---
|
| 83 |
+
|
| 84 |
+
Resources
|
| 85 |
+
|
| 86 |
+
Model Repository: ATLAS-QUANTUM
|
| 87 |
+
|
| 88 |
+
Hugging Face Documentation: Model Cards
|
| 89 |
+
|
| 90 |
+
|
| 91 |
+
|
| 92 |
+
---
|
| 93 |
+
|
| 94 |
+
License
|
| 95 |
+
|
| 96 |
+
This model is released under the Apache 2.0 License. Users are encouraged to review the license before use.
|
| 97 |
+
|
| 98 |
+
|
| 99 |
+
---
|
| 100 |
+
|
| 101 |
+
Contact and Support
|
| 102 |
+
|
| 103 |
+
For any issues, feature requests, or contributions, please reach out to the repository maintainer at Hugging Face.
|
| 104 |
+
|
| 105 |
+
### Explanation of Additions:
|
| 106 |
+
1. **Metadata Block:** Includes essential details such as `library_name`, `tags`, `datasets`, `model_architecture`, and other relevant fields to align with Hugging Face's model card standards.
|
| 107 |
+
2. **Model Details:** Expands on model usage, applications, and limitations.
|
| 108 |
+
3. **Resources and Licensing:** Provides clear references for further support and licensing information.
|
| 109 |
+
4. **Placeholders:** Marked where dataset details are missing for future updates.
|
| 110 |
+
|
| 111 |
+
---
|
| 112 |
+
|
| 113 |
+
Here’s a sample YAML metadata for the ATLAS-QUANTUM model card:
|
| 114 |
+
|
| 115 |
+
|
| 116 |
+
Explanation of the Fields:
|
| 117 |
+
|
| 118 |
+
library_name: Indicates the model library (e.g., LLaMA in this case).
|
| 119 |
+
|
| 120 |
+
tags: Lists relevant tags to help users find your model (e.g., quantization, language-model).
|
| 121 |
+
|
| 122 |
+
license: Specifies the model’s license type (e.g., apache-2.0).
|
| 123 |
+
|
| 124 |
+
datasets: Placeholder for datasets used during training (replace with actual names).
|
| 125 |
+
|
| 126 |
+
language: Indicates the primary language the model supports (English in this case).
|
| 127 |
+
|
| 128 |
+
model_architecture: Specifies the architecture of the model (e.g., LLaMA).
|
| 129 |
+
|
| 130 |
+
model_size: States the parameter size of the model (e.g., 6.74B).
|
| 131 |
+
|
| 132 |
+
quantization: Notes the quantization method applied to the model (e.g., Q2_K).
|
| 133 |
+
|
| 134 |
+
inference: Indicates whether the model is optimized for inference.
|
| 135 |
+
|
| 136 |
+
training_data: A brief description of the data used to train the model.
|
| 137 |
+
|
| 138 |
+
source_code: Boolean field indicating whether the source code is included.
|
| 139 |
+
|
| 140 |
+
documentation: Link to the model's documentation or README file.
|
| 141 |
+
|
| 142 |
+
|
| 143 |
+
This YAML file resolves metadata warnings and aligns with Hugging Face’s requirements for model cards. Ensure you update placeholder fields with accurate details!
|
| 144 |
+
|
| 145 |
+
|