|
--- |
|
base_model: Spestly/Athena-1-3B |
|
tags: |
|
- text-generation-inference |
|
- transformers |
|
- unsloth |
|
- qwen2 |
|
- trl |
|
- llama-cpp |
|
- gguf-my-repo |
|
license: apache-2.0 |
|
language: |
|
- en |
|
--- |
|
|
|
# Triangle104/Athena-1-3B-Q5_K_M-GGUF |
|
This model was converted to GGUF format from [`Spestly/Athena-1-3B`](https://huggingface.co./Spestly/Athena-1-3B) using llama.cpp via the ggml.ai's [GGUF-my-repo](https://huggingface.co./spaces/ggml-org/gguf-my-repo) space. |
|
Refer to the [original model card](https://huggingface.co./Spestly/Athena-1-3B) for more details on the model. |
|
|
|
--- |
|
Model details: |
|
- |
|
Athena-1 3B is a fine-tuned, instruction-following large language model derived from Qwen/Qwen2.5-3B-Instruct. |
|
It is designed to provide efficient, high-quality text generation while |
|
maintaining a compact size. Athena 3B is optimized for lightweight |
|
applications, conversational AI, and structured data tasks, making it |
|
ideal for real-world use cases where performance and resource efficiency |
|
are critical. |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Key Features |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
⚡ Lightweight and Efficient |
|
|
|
|
|
|
|
|
|
Compact Size: At just 3.09 billion parameters, Athena-1 3B offers excellent performance with reduced computational requirements. |
|
Instruction Following: Fine-tuned for precise and reliable adherence to user prompts. |
|
Coding and Mathematics: Proficient in solving coding challenges and handling mathematical tasks. |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
📖 Long-Context Understanding |
|
|
|
|
|
|
|
|
|
Context Length: Supports up to 32,768 tokens, enabling the processing of moderately lengthy documents or conversations. |
|
Token Generation: Can generate up to 8K tokens of output. |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
🌍 Multilingual Support |
|
|
|
|
|
|
|
|
|
Supports 29+ languages, including: |
|
English, Chinese, French, Spanish, Portuguese, German, Italian, Russian |
|
Japanese, Korean, Vietnamese, Thai, Arabic, and more. |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
📊 Structured Data & Outputs |
|
|
|
|
|
|
|
|
|
Structured Data Interpretation: Processes structured formats like tables and JSON. |
|
Structured Output Generation: Generates well-formatted outputs, including JSON and other structured formats. |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Model Details |
|
|
|
|
|
|
|
|
|
Base Model: Qwen/Qwen2.5-3B-Instruct |
|
Architecture: Transformers with RoPE, SwiGLU, RMSNorm, Attention QKV bias, and tied word embeddings. |
|
Parameters: 3.09B total (2.77B non-embedding). |
|
Layers: 36 |
|
Attention Heads: 16 for Q, 2 for KV. |
|
Context Length: Up to 32,768 tokens. |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Applications |
|
|
|
|
|
|
|
|
|
Athena 3B is designed for a variety of real-world applications: |
|
|
|
|
|
Conversational AI: Build fast, responsive, and lightweight chatbots. |
|
Code Generation: Generate, debug, or explain code snippets. |
|
Mathematical Problem Solving: Assist with calculations and reasoning. |
|
Document Processing: Summarize and analyze moderately large documents. |
|
Multilingual Applications: Support for global use cases with diverse language requirements. |
|
Structured Data: Process and generate structured data, such as tables and JSON. |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Quickstart |
|
|
|
|
|
|
|
|
|
Here’s how you can use Athena 3B for quick text generation: |
|
|
|
|
|
# Use a pipeline as a high-level helper |
|
from transformers import pipeline |
|
|
|
messages = [ |
|
{"role": "user", "content": "Who are you?"}, |
|
] |
|
pipe = pipeline("text-generation", model="Spestly/Athena-1-3B") |
|
pipe(messages) |
|
|
|
# Load model directly |
|
from transformers import AutoTokenizer, AutoModelForCausalLM |
|
|
|
tokenizer = AutoTokenizer.from_pretrained("Spestly/Athena-1-3B") |
|
model = AutoModelForCausalLM.from_pretrained("Spestly/Athena-1-3B") |
|
|
|
--- |
|
## Use with llama.cpp |
|
Install llama.cpp through brew (works on Mac and Linux) |
|
|
|
```bash |
|
brew install llama.cpp |
|
|
|
``` |
|
Invoke the llama.cpp server or the CLI. |
|
|
|
### CLI: |
|
```bash |
|
llama-cli --hf-repo Triangle104/Athena-1-3B-Q5_K_M-GGUF --hf-file athena-1-3b-q5_k_m.gguf -p "The meaning to life and the universe is" |
|
``` |
|
|
|
### Server: |
|
```bash |
|
llama-server --hf-repo Triangle104/Athena-1-3B-Q5_K_M-GGUF --hf-file athena-1-3b-q5_k_m.gguf -c 2048 |
|
``` |
|
|
|
Note: You can also use this checkpoint directly through the [usage steps](https://github.com/ggerganov/llama.cpp?tab=readme-ov-file#usage) listed in the Llama.cpp repo as well. |
|
|
|
Step 1: Clone llama.cpp from GitHub. |
|
``` |
|
git clone https://github.com/ggerganov/llama.cpp |
|
``` |
|
|
|
Step 2: Move into the llama.cpp folder and build it with `LLAMA_CURL=1` flag along with other hardware-specific flags (for ex: LLAMA_CUDA=1 for Nvidia GPUs on Linux). |
|
``` |
|
cd llama.cpp && LLAMA_CURL=1 make |
|
``` |
|
|
|
Step 3: Run inference through the main binary. |
|
``` |
|
./llama-cli --hf-repo Triangle104/Athena-1-3B-Q5_K_M-GGUF --hf-file athena-1-3b-q5_k_m.gguf -p "The meaning to life and the universe is" |
|
``` |
|
or |
|
``` |
|
./llama-server --hf-repo Triangle104/Athena-1-3B-Q5_K_M-GGUF --hf-file athena-1-3b-q5_k_m.gguf -c 2048 |
|
``` |
|
|
