metadata

library_name: transformers
tags:
  - conversational
  - chain-of-thought
  - education
  - llama-cpp
  - gguf-my-repo
base_model: caedencode/Caeden-o1

Triangle104/Caeden-o1-Q5_K_S-GGUF

This model was converted to GGUF format from caedencode/Caeden-o1 using llama.cpp via the ggml.ai's GGUF-my-repo space. Refer to the original model card for more details on the model.

Model details

CaedenAI is a conversational AI model fine-tuned to provide detailed reasoning in its responses using the Chain-of-Thought (CoT) methodology. It is designed for educational use, enabling users to understand the reasoning process behind answers.

Developed by: Caeden Rajoo Model type: Conversational AI with CoT reasoning License: Apache 2 Finetuned from model: Qwen/Qwen2.5-1.5B Primary Use Case: Education and knowledge expansion

This model is fine-tuned for generating step-by-step reasoning for queries, making it an excellent tool for educational environments and learning applications.

    Uses






    


    Direct Use

This model can be directly applied in:

Educational environments to help students learn with explanations. Applications where detailed reasoning is required for understanding answers. Conversational AI systems that prioritize reasoning over simple answers.

    Out-of-Scope Use

This model may not be suitable for:

Scenarios requiring highly specialized domain knowledge not covered in the training data. Tasks requiring real-time response for critical systems (e.g., healthcare, safety).

    Bias, Risks, and Limitations

The model inherits limitations from its training data and base model. Users should consider potential biases or incomplete information in responses.

    Recommendations

The model's output should be reviewed for accuracy in critical use cases. Users should ensure that ethical considerations are met when using the model in sensitive environments.

    How to Get Started with the Model

Here’s how you can load and use CaedenAI:

import torch from transformers import AutoTokenizer, AutoModelForCausalLM

model = AutoModelForCausalLM.from_pretrained("caedencode/Caeden-o1") tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen2.5-1.5B")

device = torch.device("cuda" if torch.cuda.is_available() else "cpu") model = model.to(device)

def generate_answer(question): prompt = f"Question: {question}\nReasoning:\n" inputs = tokenizer(prompt, return_tensors="pt").to(device) outputs = model.generate(**inputs, max_length=200, num_beams=5, early_stopping=True) return tokenizer.decode(outputs[0], skip_special_tokens=True)

question = "What is the largest planet in our solar system?" answer = generate_answer(question) print(answer)

Use with llama.cpp

Install llama.cpp through brew (works on Mac and Linux)

brew install llama.cpp

Invoke the llama.cpp server or the CLI.

CLI:

llama-cli --hf-repo Triangle104/Caeden-o1-Q5_K_S-GGUF --hf-file caeden-o1-q5_k_s.gguf -p "The meaning to life and the universe is"

Server:

llama-server --hf-repo Triangle104/Caeden-o1-Q5_K_S-GGUF --hf-file caeden-o1-q5_k_s.gguf -c 2048

Note: You can also use this checkpoint directly through the usage steps 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/Caeden-o1-Q5_K_S-GGUF --hf-file caeden-o1-q5_k_s.gguf -p "The meaning to life and the universe is"

./llama-server --hf-repo Triangle104/Caeden-o1-Q5_K_S-GGUF --hf-file caeden-o1-q5_k_s.gguf -c 2048