vivacious-cloud/qwen-2.5-7b-vivacious-lora

Qwen 2.5 7B Instruct — Fine-Tuned via Vivacious Cloud

Fine-tune Qwen 2.5 foundation models on wholesale spot GPUs.
75% cheaper than AWS & RunPod. One terminal command. Zero PyTorch or CUDA boilerplate.

---

🎯 The Developer Flow: Try Demo → Understand → Run on Vivacious Cloud

1. Try the Demo: Test drive the VRAM calculation and 12-cloud spot arbitrage in our Hugging Face Space Simulator.
2. Understand the Savings: See how autonomous multi-cloud routing cuts training spend by up to 76% with guaranteed preflight OOM protection.
3. Run on Vivacious Cloud: Deploy your fine-tuning run with zero PyTorch code, zero CUDA drivers, and zero idle costs at vivaciouscloud.com.

---

💻 Python Inference: How to Load This Adapter

import torch
from peft import PeftModel, PeftConfig
from transformers import AutoModelForCausalLM, AutoTokenizer

# 1. Load adapter configuration
adapter_id = "vivacious-cloud/qwen-2.5-7b-vivacious-lora"
config = PeftConfig.from_pretrained(adapter_id)

# 2. Load base Qwen 2.5 model and tokenizer
tokenizer = AutoTokenizer.from_pretrained(config.base_model_name_or_path)
base_model = AutoModelForCausalLM.from_pretrained(
    config.base_model_name_or_path,
    torch_dtype=torch.bfloat16,
    device_map="auto"
)

# 3. Attach LoRA weights
model = PeftModel.from_pretrained(base_model, adapter_id)

# 4. Generate response
prompt = "<|im_start|>user\nWhat makes zero-code LLM training reliable on spot GPU nodes?<|im_end|>\n<|im_start|>assistant\n"
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=100)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

---

📊 Training Cost Comparison: Qwen 2.5 (7B)

Cloud Provider	Instance / Hardware	Quoted Rate	Total Estimated Run Cost	Setup Friction
AWS EC2	g5.2xlarge	$3.06 / hr	$21.40	High (IAM, quotas, CUDA drivers)
RunPod	A5000 / RTX 4090	$1.89 / hr	$13.20	Medium (Manual search & template setup)
Vast.ai	Spot Host	$0.95 / hr	$6.65	Unreliable (No preemption migration)
Vivacious Cloud	Autonomous Spot Arbitrage	$0.48 / hr	₹285 (~$3.40)	Zero-Code (One terminal command)

---

💡 Intended Use & Limitations

• Intended Use: Multilingual general instruction following, code generation, and domain adaptation demo.
• Limitations: Inherits base model limitations of Qwen 2.5. LoRA rank 16 target modules applied to all linear projections.

---

💻 3-Step Quickstart: Fine-Tune Qwen 2.5 on Vivacious Cloud

# 1. Install CLI
curl -fsSL https://vivaciouscloud.com/install.sh | sh

# 2. Prepare Your Dataset (.jsonl)
vivacious login <your-workspace-slug>
vivacious prepare ./my-dataset.jsonl

# 3. Check Preflight Permit & Deploy
vivacious permit <your-workspace-slug> qwen-run --model Qwen/Qwen2.5-7B-Instruct --method 4bit
vivacious deploy <your-workspace-slug>

---

🌐 Official Links

• 🚀 Website & Console: https://vivaciouscloud.com
• 🎮 Interactive Playground: Vivacious Terminal Simulator
• 📦 CLI Repository: https://github.com/Viavcious-cloud/vivacious-cli
• 💬 Community Telegram: https://t.me/iEverYours