Open-Source LLM Deployment Specialist
Budget / Salary₹75,000–150,000
TypeFreelance project
LocationRemote
Posted2 hours ago
Setup & Deployment of Unrestricted / Zero-Refusal Open-Source LLM (Local or Cloud API)
Job Description
Project Overview:
I am looking for an AI/ML specialist to recommend, configure, and deploy a high-performance open-source Large Language Model (LLM) that operates without content refusal barriers (often referred to as de-aligned, abliterated, or zero-refusal models).
The goal is to have an AI system capable of answering complex, sensitive, and uncensored queries directly without built-in safety refusals, moral lecturing, or generic rejection disclaimers.
Key Responsibilities:
1. Model Selection & Recommendation:
Recommend top-tier open-weight models optimized for zero refusal (e.g., Eric Hartford’s Dolphin series, abliterated Llama/Mistral variants, or Hermes-based uncensored fine-tunes).
Ensure the recommended model balances low refusal rates with high reasoning, factuality, and speed.
2. Environment Setup & Deployment:
Configure the model either locally (using Ollama, LM Studio, or vLLM) OR deploy it on a private cloud instance (e.g., RunPod, Vast.ai, or AWS EC2 with GPU).
Alternatively, configure a clean web UI (like Open WebUI or LibreChat) connected via OpenRouter or self-hosted API endpoints.
3. Testing & Handoff:
Verify prompt adherence and ensure the model does not trip false refusals on research/edge-case prompts.
Provide clear instructions on how to start, query, and maintain the system.
Required Skills:
Experience with open-source LLMs (Hugging Face ecosystem, Ollama, LM Studio, vLLM).
Experience with model quantization formats (GGUF, AWQ, EXL2).
Familiarity with de-aligned / abliterated model architectures and prompt formatting.
Cloud GPU deployment (RunPod, Vast.ai, or VPS) if local hardware is insufficient.
To Apply, Please Answer:
1. Which specific zero-refusal / uncensored model architectures do you recommend for high reasoning without refusals?
2. Do you recommend a local setup (specify required hardware) or a cost-effective cloud GPU deployment?
Suggested Category & Tags
Category: Artificial Intelligence & Machine Learning / Natural Language Processing
Skills/Tags: Large Language Models (LLM), Hugging Face, Open Source AI, Ollama, Python, GPU Computing, Model Deployment
Job Description
Project Overview:
I am looking for an AI/ML specialist to recommend, configure, and deploy a high-performance open-source Large Language Model (LLM) that operates without content refusal barriers (often referred to as de-aligned, abliterated, or zero-refusal models).
The goal is to have an AI system capable of answering complex, sensitive, and uncensored queries directly without built-in safety refusals, moral lecturing, or generic rejection disclaimers.
Key Responsibilities:
1. Model Selection & Recommendation:
Recommend top-tier open-weight models optimized for zero refusal (e.g., Eric Hartford’s Dolphin series, abliterated Llama/Mistral variants, or Hermes-based uncensored fine-tunes).
Ensure the recommended model balances low refusal rates with high reasoning, factuality, and speed.
2. Environment Setup & Deployment:
Configure the model either locally (using Ollama, LM Studio, or vLLM) OR deploy it on a private cloud instance (e.g., RunPod, Vast.ai, or AWS EC2 with GPU).
Alternatively, configure a clean web UI (like Open WebUI or LibreChat) connected via OpenRouter or self-hosted API endpoints.
3. Testing & Handoff:
Verify prompt adherence and ensure the model does not trip false refusals on research/edge-case prompts.
Provide clear instructions on how to start, query, and maintain the system.
Required Skills:
Experience with open-source LLMs (Hugging Face ecosystem, Ollama, LM Studio, vLLM).
Experience with model quantization formats (GGUF, AWQ, EXL2).
Familiarity with de-aligned / abliterated model architectures and prompt formatting.
Cloud GPU deployment (RunPod, Vast.ai, or VPS) if local hardware is insufficient.
To Apply, Please Answer:
1. Which specific zero-refusal / uncensored model architectures do you recommend for high reasoning without refusals?
2. Do you recommend a local setup (specify required hardware) or a cost-effective cloud GPU deployment?
Suggested Category & Tags
Category: Artificial Intelligence & Machine Learning / Natural Language Processing
Skills/Tags: Large Language Models (LLM), Hugging Face, Open Source AI, Ollama, Python, GPU Computing, Model Deployment
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.