ZOZ-Reasoning-Master-3B API & Inference Endpoint

📌 Model Overview

ZOZ-Reasoning-Master-3B is a compact, high-efficiency language model engineered specifically to operate as an Autonomous Edge Agent. It was developed using a sequential Tri-Stage Agentic Pipeline designed to bridge three core competencies rarely found together in 3B-class models: strict schema-compliant tool calling, resilient long-context retrieval without degradation, and structured multi-turn reasoning.

Table with columns: Attribute, Specification
Attribute	Specification
Base Architecture	Qwen2.5-3B (Causal LM)
Parameter Count	3.09 Billion
Shipped Weights	Fully Merged FP16 (Unquantized)
Context Window	32,768 Tokens (Trained & Optimized up to 4,096 Tokens)
Prompt Format	ChatML (`<
Developer	ZOZ AI / Ziad Khodr

🔬 The Tri-Stage Training Pipeline

To prevent Catastrophic Forgetting, the model was trained through an incremental staging protocol:

text

[Qwen2.5-3B-Instruct]│▼ (Stage 1: Tool Use & Schema Discipline)[ZOZ-Function-Master-3B] ──► NousResearch/hermes-function-calling-v1│▼ (Stage 2: Deep Needle Retrieval & Multi-Hop Reasoning)[ZOZ-Function-Master-3B-LongContext] ──► zai-org/LongBench-v2│▼ (Stage 3: Conversational CoT & Logic Alignment)[ZOZ-Reasoning-Master-3B] ──► Magpie-Align/Magpie-Pro-300K-Filtered (Curated)

Stage 1 (Function Calling & Tool Use): Fine-tuned on Hermes function-calling datasets to enforce strict JSON output syntax and eliminate parameter hallucination.
Stage 2 (In-Context Reasoning & Long Retrieval): Trained on hard reasoning tasks from LongBench-v2 to mitigate the Lost-in-the-Middle phenomenon and enable needle-in-a-haystack data extraction.
Stage 3 (Conversational Alignment): Aligned with a curated subset of Magpie-Pro to impart clean chain-of-thought (CoT) reasoning distilled from frontier models (Llama-3.1-70B) without diluting tool-calling precision.

🚀 Quickstart & Inference

1. Basic Reasoning & Chat

python

import torchfrom transformers import AutoModelForCausalLM, AutoTokenizer model_id = "z51722369/ZOZ-Reasoning-Master-3B" tokenizer = AutoTokenizer.from_pretrained(model_id)model = AutoModelForCausalLM.from_pretrained(    model_id,    torch_dtype=torch.float16,    device_map="auto") messages = [    {"role": "system", "content": "You are ZOZ-Reasoning-Master, a precise and logical AI agent."},    {"role": "user", "content": "Analyze the potential bottlenecks of deploying an in-memory Redis cache versus a persistent disk store in a high-throughput microservices architecture."}] prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)inputs = tokenizer(prompt, return_tensors="pt").to(model.device) outputs = model.generate(    **inputs,    max_new_tokens=512,    temperature=0.3,    top_p=0.9,    repetition_penalty=1.1,    do_sample=True) response = tokenizer.decode(outputs[0][inputs.input_ids.shape[1]:], skip_special_tokens=True)print(response)

2. Function Calling Example

python

tools_schema = """[  {    "type": "function",    "function": {      "name": "query_database",      "description": "Execute a structured SQL query against the production telemetry database.",      "parameters": {        "type": "object",        "properties": {          "table": {"type": "string", "description": "Target database table"},          "filter": {"type": "string", "description": "SQL WHERE clause filter"}        },        "required": ["table", "filter"]      }    }  }]""" messages = [    {"role": "system", "content": f"You are an agent with access to external tools. Available tools:\n{tools_schema}"},    {"role": "user", "content": "Check all failed authentication events in the audit_logs table from the last 2 hours."}] prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)inputs = tokenizer(prompt, return_tensors="pt").to(model.device) with torch.no_grad():    outputs = model.generate(**inputs, max_new_tokens=256, temperature=0.1) print(tokenizer.decode(outputs[0][inputs.input_ids.shape[1]:], skip_special_tokens=True))# Expected Output: <tool_call> {"name": "query_database", "arguments": {"table": "audit_logs", "filter": "status = 'FAILED' AND timestamp >= NOW() - INTERVAL 2 HOUR"}} </tool_call>

⚙️ Hyperparameters & Training Specifications

Table with columns: Stage, Dataset, Sample Size, Learning Rate, Effective Batch Size, Wall Time
Stage	Dataset	Sample Size	Learning Rate	Effective Batch Size	Wall Time
Stage 1 (FC)	`hermes-function-calling-v1`	~15,000	`2e-4`	8	~1.5 Hours
Stage 2 (LC)	`zai-org/LongBench-v2`	503 (Hard)	`1e-4`	4	~35 Mins
Stage 3 (Align)	`Magpie-Pro-300K-Filtered`	6,000 (Curated)	`1.5e-4`	8	~1 Hour

Adapter Method: LoRA ( $r = 16, α = 32$ , Dropout = $0.05$ )
Target Modules: q_proj, k_proj, v_proj, o_proj, gate_proj, up_proj, down_proj
Optimizer: paged_adamw_8bit
Precision: Mixed FP16 via PyTorch Distributed (torchrun DDP across 2x Tesla T4 GPUs)
Gradient Clipping: max_grad_norm = 0.3

🎯 Intended Use Cases

Local RAG Pipelines: High-speed document reasoning engine for locally indexed knowledge bases.
Autonomous Micro-Agents: Routing, parameter extraction, and execution loops across webhooks and internal APIs.
Edge & On-Device Deployment: Direct export to 4-bit GGUF or AWQ formats for low-latency execution on mobile devices, Apple Silicon, or embedded compute platforms (e.g., Raspberry Pi 5, Jetson Orin).

⚠️ Limitations & Boundary Conditions

World Knowledge Limits: As a 3B parameter model, it is not intended to replace 70B+ models for deep encyclopedic trivia. It delivers optimal accuracy when grounded with context or tools.
Context Window Best Practices: While supporting 32K context natively, maintaining inference sequences between 1,024 and 4,096 tokens yields the best balance between latency and memory footprint.

📜 Citation

If you use or build upon ZOZ-Reasoning-Master-3B, please cite:

bibtex

@misc{zoz_reasoning_master_3b_2026,  author = {Ziad Khodr (ZOZ AI)},  title = {ZOZ-Reasoning-Master-3B: A Specialized Long-Context & Function-Calling 3B Agent},  year = {2026},  publisher = {Hugging Face},  howpublished = {\url{[https://huggingface.co/z51722369/ZOZ-Reasoning-Master-3B](https://huggingface.co/z51722369/ZOZ-Reasoning-Master-3B)}}}

text

undefined

Attribute

Specification

Base Architecture

Qwen2.5-3B (Causal LM)

Parameter Count

3.09 Billion

Shipped Weights

Fully Merged FP16 (Unquantized)

Context Window

32,768 Tokens (Trained & Optimized up to 4,096 Tokens)

Prompt Format

ChatML (`<

Developer

ZOZ AI / Ziad Khodr

[Qwen2.5-3B-Instruct]│▼ (Stage 1: Tool Use & Schema Discipline)[ZOZ-Function-Master-3B] ──► NousResearch/hermes-function-calling-v1│▼ (Stage 2: Deep Needle Retrieval & Multi-Hop Reasoning)[ZOZ-Function-Master-3B-LongContext] ──► zai-org/LongBench-v2│▼ (Stage 3: Conversational CoT & Logic Alignment)[ZOZ-Reasoning-Master-3B] ──► Magpie-Align/Magpie-Pro-300K-Filtered (Curated)

import torchfrom transformers import AutoModelForCausalLM, AutoTokenizer model_id = "z51722369/ZOZ-Reasoning-Master-3B" tokenizer = AutoTokenizer.from_pretrained(model_id)model = AutoModelForCausalLM.from_pretrained( model_id, torch_dtype=torch.float16, device_map="auto") messages = [ {"role": "system", "content": "You are ZOZ-Reasoning-Master, a precise and logical AI agent."}, {"role": "user", "content": "Analyze the potential bottlenecks of deploying an in-memory Redis cache versus a persistent disk store in a high-throughput microservices architecture."}] prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)inputs = tokenizer(prompt, return_tensors="pt").to(model.device) outputs = model.generate( **inputs, max_new_tokens=512, temperature=0.3, top_p=0.9, repetition_penalty=1.1, do_sample=True) response = tokenizer.decode(outputs[0][inputs.input_ids.shape[1]:], skip_special_tokens=True)print(response)

tools_schema = """[ { "type": "function", "function": { "name": "query_database", "description": "Execute a structured SQL query against the production telemetry database.", "parameters": { "type": "object", "properties": { "table": {"type": "string", "description": "Target database table"}, "filter": {"type": "string", "description": "SQL WHERE clause filter"} }, "required": ["table", "filter"] } } }]""" messages = [ {"role": "system", "content": f"You are an agent with access to external tools. Available tools:\n{tools_schema}"}, {"role": "user", "content": "Check all failed authentication events in the audit_logs table from the last 2 hours."}] prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)inputs = tokenizer(prompt, return_tensors="pt").to(model.device) with torch.no_grad(): outputs = model.generate(**inputs, max_new_tokens=256, temperature=0.1) print(tokenizer.decode(outputs[0][inputs.input_ids.shape[1]:], skip_special_tokens=True))# Expected Output: <tool_call> {"name": "query_database", "arguments": {"table": "audit_logs", "filter": "status = 'FAILED' AND timestamp >= NOW() - INTERVAL 2 HOUR"}} </tool_call>

Stage

Dataset

Sample Size

Learning Rate

Effective Batch Size

Wall Time

Stage 1 (FC)

hermes-function-calling-v1

~15,000

2e-4

~1.5 Hours

Stage 2 (LC)

zai-org/LongBench-v2

503 (Hard)

1e-4

~35 Mins

Stage 3 (Align)

Magpie-Pro-300K-Filtered

6,000 (Curated)

1.5e-4

~1 Hour

@misc{zoz_reasoning_master_3b_2026, author = {Ziad Khodr (ZOZ AI)}, title = {ZOZ-Reasoning-Master-3B: A Specialized Long-Context & Function-Calling 3B Agent}, year = {2026}, publisher = {Hugging Face}, howpublished = {\url{[https://huggingface.co/z51722369/ZOZ-Reasoning-Master-3B](https://huggingface.co/z51722369/ZOZ-Reasoning-Master-3B)}}}

ZOZ-Reasoning-Master-3B

README

📌 Model Overview

🔬 The Tri-Stage Training Pipeline

🚀 Quickstart & Inference

1. Basic Reasoning & Chat

2. Function Calling Example

⚙️ Hyperparameters & Training Specifications

🎯 Intended Use Cases

⚠️ Limitations & Boundary Conditions

📜 Citation

Explore FriendliAI today

README

📌 Model Overview

🔬 The Tri-Stage Training Pipeline

🚀 Quickstart & Inference

1. Basic Reasoning & Chat

2. Function Calling Example

⚙️ Hyperparameters & Training Specifications

🎯 Intended Use Cases

⚠️ Limitations & Boundary Conditions

📜 Citation