Supported Models
1. Meta-Llama 3.1 70B Instruct (Quantized)
"hugging-quants/Meta-Llama-3.1-70B-Instruct-AWQ-INT4"
Description: A quantized version of Meta AI's 70-billion-parameter model, optimized for multilingual dialogue applications.
Details:
Model Size: 70 billion parameters
Quantization: INT4 using AutoAWQ
Use Cases: Multilingual dialogue applications
More Information: Meta-Llama 3.1 70B Instruct - Hugging Face
2. Mistral Small 3.1 24B Instruct
"mrfakename/mistral-small-3.1-24b-instruct-2503-hf"
Description: A 24-billion-parameter instruction-tuned model, focusing on text generation tasks.
Details:
Model Size: 24 billion parameters
Format: Hugging Face Transformers
Use Cases: Text generation tasks
More Information: Mistral Small 3.1 24B Instruct - Hugging Face
3. Gemma 3 27B (Upcoming)
Description: A 27-billion-parameter multimodal model from Google, capable of handling both text and image inputs.
Details:
Model Size: 27 billion parameters
Capabilities: Multimodal (text and image input, text output)
Use Cases: Question answering, summarization, reasoning
More Information: Gemma 3 27B - Hugging Face
To learn more about our Language Models please visit our Substack
Last updated