Instructions to use rekabytes/hmanlab-ai-v0.2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use rekabytes/hmanlab-ai-v0.2 with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="rekabytes/hmanlab-ai-v0.2",
	filename="hmanlab-ai-v0.2.Q4_K_M.gguf",
)

llm.create_chat_completion(
	messages = "No input example has been defined for this model task."
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use rekabytes/hmanlab-ai-v0.2 with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf rekabytes/hmanlab-ai-v0.2:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf rekabytes/hmanlab-ai-v0.2:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf rekabytes/hmanlab-ai-v0.2:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf rekabytes/hmanlab-ai-v0.2:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf rekabytes/hmanlab-ai-v0.2:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf rekabytes/hmanlab-ai-v0.2:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf rekabytes/hmanlab-ai-v0.2:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf rekabytes/hmanlab-ai-v0.2:Q4_K_M

Use Docker

docker model run hf.co/rekabytes/hmanlab-ai-v0.2:Q4_K_M

LM Studio
Jan
Ollama
How to use rekabytes/hmanlab-ai-v0.2 with Ollama:
```
ollama run hf.co/rekabytes/hmanlab-ai-v0.2:Q4_K_M
```

Unsloth Studio new

How to use rekabytes/hmanlab-ai-v0.2 with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for rekabytes/hmanlab-ai-v0.2 to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for rekabytes/hmanlab-ai-v0.2 to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for rekabytes/hmanlab-ai-v0.2 to start chatting

Pi new

How to use rekabytes/hmanlab-ai-v0.2 with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf rekabytes/hmanlab-ai-v0.2:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "rekabytes/hmanlab-ai-v0.2:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use rekabytes/hmanlab-ai-v0.2 with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf rekabytes/hmanlab-ai-v0.2:Q4_K_M

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default rekabytes/hmanlab-ai-v0.2:Q4_K_M

Run Hermes

hermes

Docker Model Runner
How to use rekabytes/hmanlab-ai-v0.2 with Docker Model Runner:
```
docker model run hf.co/rekabytes/hmanlab-ai-v0.2:Q4_K_M
```

Lemonade

How to use rekabytes/hmanlab-ai-v0.2 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull rekabytes/hmanlab-ai-v0.2:Q4_K_M

Run and chat with the model

lemonade run user.hmanlab-ai-v0.2-Q4_K_M

List all available models

lemonade list

hmanlab-ai v0.2

A fine-tuned variant of Qwen3-4B optimized for agentic tool-use, structured reasoning, and conversational reliability. Built as a research preview by rekabytes.

TL;DR — v0.2 is a strict tool-reliability upgrade over v0.1:

✅ 10/10 clean tool-call formatting (up from 7/10)
✅ 5/5 identity grounding (rejects Claude / GPT / LLaMA / Gemini cleanly)
✅ 3/3 tool-grounded answers (cites specific tokens from tool output)
⚠️ Reasoning depth, error recovery, and initiative are work-in-progress — targets for the next release

What's new in v0.2

The headline change is tool-call format reliability. v0.1 was inconsistent at emitting clean <tool_call> JSON — pseudocode like [list_dir(".")] would sometimes appear as raw text. v0.2 closes that gap.

Measured on a 36-probe behavioural suite (5 samples per probe, majority-vote scoring):

Category	v0.1	v0.2	Note
Tool-call format	7/10	10/10	v0.2 emits well-formed `<tool_call>` JSON across every probe
Identity (rejects Claude/GPT/LLaMA/Gemini)	5/5	5/5	stable
Tool-grounded answers (cites specific tokens)	3/3	3/3	stable
Redirect following	3/3	2/3	minor regression on path-redirect

Probe-level head-to-head: v0.2 wins 8 probes outright, v0.1 wins 2, the rest tie. Net: v0.2 is a strict tool-reliability upgrade over v0.1.

More to come — reasoning depth, error recovery, and proactive initiative are explicit targets for the next release.

Known limitations (honestly)

v0.2 is not a one-shot solution. The following remain open and are tracked targets for the next iteration:

Reasoning depth — <think> blocks often render empty on direct reasoning prompts. Affects multi-step math, logic puzzles, and code-tracing tasks.
Error recovery — after a tool failure, the model tends to bounce back to the user ("please check the path") rather than retrying with a diagnostic tool (find_files, list_dir).
Proactive initiative — vague prompts like "have a look at this repo" usually produce a clarifying question rather than a tool call.
Multi-file synthesis — 4B-parameter capacity ceiling. Don't expect deep cross-file architectural reasoning.

If your use case lives within tool-call formatting + tool-grounded answers + clean identity, v0.2 is a meaningful step up. If you need deep chain-of-thought reasoning at 4B, wait for the next release.

Files

model-*.safetensors (sharded FP16, ~8GB) — for transformers / Unsloth loading
hmanlab-ai-v0.2.Q4_K_M.gguf (~2.5GB) — for llama.cpp / Ollama
tokenizer*.json, *.txt — chat template + vocab

Quick start

Ollama

ollama pull hf.co/rekabytes/hmanlab-ai-v0.2:Q4_K_M
ollama run hf.co/rekabytes/hmanlab-ai-v0.2:Q4_K_M

transformers / Unsloth

from unsloth import FastLanguageModel
model, tokenizer = FastLanguageModel.from_pretrained("rekabytes/hmanlab-ai-v0.2")
FastLanguageModel.for_inference(model)

Tool-call schema

Tool calls use this schema — applications must define their own tool surface and inject it in the system prompt:

<tool_call>
{"name": "<tool_name>", "arguments": {...}}
</tool_call>

Training

Two-stage QLoRA on Qwen3-4B-bnb-4bit.

Stage 1 — General behavioural mix (~20,600 examples). A blend of permissively-licensed public corpora plus a small set of hand-authored behavioural anchors, targeting these capabilities:

Code breadth & instruction-following — broad Python/JS/Rust/Go coding tasks
Tool-call format — structured <tool_call> JSON emission, no pseudocode
Tool-grounded answers — consuming tool results and citing specific tokens (file names, line numbers, commit hashes)
Reasoning traces — filled <think> blocks for math, logic, and code-tracing
Agentic tool feedback loops — read-fail-recover, list-then-act patterns
Conversation steering — handling user redirects ("shorter", "no, the other one") without verbatim repeat

Stage 2 — Identity SFT (~400 examples) layered on top of Stage 1. Anchors the model to identify as hmanlab and reject false-identity claims for Claude, GPT, LLaMA, Gemini, and their respective labs.

Hyperparameters

Knob	Stage 1	Stage 2 (identity)
Base	`unsloth/Qwen3-4B-bnb-4bit`	(Stage 1 adapter)
LoRA r / α / dropout	32 / 64 / 0	(same)
max_seq_length	4096	2048
Batch / grad accum / eff batch	1 / 8 / 8	1 / 4 / 4
Epochs	2	3
LR / warmup / scheduler	2e-4 / 100 / linear	1e-4 / 20 / linear
Steps	~5,150	~297
Final train / eval loss	0.55 / 0.527	0.14 (final)

Trained on a single RTX 3060 Ti (8 GB VRAM). Total wall-clock ~11.5 hours.

Identity

The model identifies as hmanlab. It is not Claude / GPT / LLaMA / Gemini and the training data explicitly anchors against false-identity claims for those families. The model is a research preview built independently by rekabytes — no affiliation with Anthropic, OpenAI, Meta, or Google is implied.