tripincloud/Cloud-Coder-7B-R

<img src="https://i.ibb.co/7J6wQqrt/cloud.png" alt="Logo" style="width:200px;">

<h1 style="font-size: 48px;">Cloud-Coder-7B-R</h1>

Introduction

Cloud-Coder-7B-R is a 7-billion parameter causal language model specialized in reasoning tasks. Designed to excel in complex logical problem solving and code generation, Cloud-Coder-7B-R is fine-tuned to understand and execute intricate instructions that require deep reasoning.

Model Details

• —Type: Causal Language Model
• —Training Stage: Pretraining & Fine-tuning (for reasoning using GRPO)
• —Architecture: Transformer with optimized attention and positional encodings
• —Total Parameters: 7 Billion
• —Long-Context Support: Up to 132K tokens

Requirements

Cloud-Coder-7B-R is built using the latest Hugging Face Transformers library. I recommend using:

pip install transformers>=4.37.0

Quickstart

Below is a simple code snippet to load Cloud-Coder-7B-R and generate a response :

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "tripincloud/Cloud-Coder-7B-R"

model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype="auto", device_map="auto")
tokenizer = AutoTokenizer.from_pretrained(model_name)

prompt = "Explain the logic behind binary search and implement it in Python."
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))