Instructions to use maternion/ling3-tiny-coder-ckpts with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use maternion/ling3-tiny-coder-ckpts with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf maternion/ling3-tiny-coder-ckpts:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf maternion/ling3-tiny-coder-ckpts:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf maternion/ling3-tiny-coder-ckpts:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf maternion/ling3-tiny-coder-ckpts:Q4_K_M

Build from source code

git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf maternion/ling3-tiny-coder-ckpts:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama cli -hf maternion/ling3-tiny-coder-ckpts:Q4_K_M

Use Docker

docker model run hf.co/maternion/ling3-tiny-coder-ckpts:Q4_K_M

Llama new
LM Studio
Jan
Ollama
How to use maternion/ling3-tiny-coder-ckpts with Ollama:
```
ollama run hf.co/maternion/ling3-tiny-coder-ckpts:Q4_K_M
```
Unsloth Desktop

How to use maternion/ling3-tiny-coder-ckpts with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf maternion/ling3-tiny-coder-ckpts:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "maternion/ling3-tiny-coder-ckpts:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use maternion/ling3-tiny-coder-ckpts with Docker Model Runner:
```
docker model run hf.co/maternion/ling3-tiny-coder-ckpts:Q4_K_M
```

Lemonade

How to use maternion/ling3-tiny-coder-ckpts with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull maternion/ling3-tiny-coder-ckpts:Q4_K_M

Run and chat with the model

lemonade run user.ling3-tiny-coder-ckpts-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use maternion/ling3-tiny-coder-ckpts with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf maternion/ling3-tiny-coder-ckpts: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 maternion/ling3-tiny-coder-ckpts:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use maternion/ling3-tiny-coder-ckpts with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf maternion/ling3-tiny-coder-ckpts:Q4_K_M

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "maternion/ling3-tiny-coder-ckpts:Q4_K_M" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

ling3-tiny-coder-ckpts / README.md

maternion

Update README.md

2128ec9 verified 1 day ago

preview code

Raw History Blame Contribute Delete

2.87 kB

---
license: mit
base_model: inclusionAI/Ling-3.0-tiny
library_name: safetensors
---
# ling3-tiny-coder-ckpts — work-in-progress (PAUSED)

> **Status: paused / unfinished.** This repo stores intermediate artifacts from a
> personal experiment: fine-tuning [Ling-3.0-tiny](https://huggingface.co/inclusionAI/Ling-3.0-tiny)
> (7.9B Bailing-hybrid MoE, KDA+MLA) into a C++ competitive-programming coder on
> Kaggle TPU v5e-8 with LoRA adapters. Development is inactive — the items here are
> training artifacts, not a finished model. Layout may change if work resumes.

## What is in this repo

- `models/` contains the only directly usable artifacts: two early-SFT merged
  GGUF exports (Q4_K_M / Q8_0 / bf16 + chat template). These are from an early
  mid-SFT point and are noticeably weaker than later checkpoints (see evals).
- `evals/` stores raw LiveCodeBench-v6 (60-problem sample) + OJBench result JSONs.
- `latest.json` / `routing_counts.npz` / `llama-bin-cache.tar.gz` are pipeline
  bookkeeping (resume pointer, MoE expert-usage counters, llama.cpp build cache).

## Training lineage

| Stage | Checkpoint | Steps | Data | Notes |
|---|---|---|---|---|
| SFT | `sft-21750` | 21,750 | R1 reasoning traces (codeforces-cots) + Magicoder thinking-off, ~178M tok, 2 epochs | completed |
| CPT | `cpt-2903` | 2,903 | raw C++ (Stack-v2) acclimation, 24M tok | completed; **best eval to date** |
| agentic (v22) | `agentic-600` | 600 | SWE-smith C++ patches | partial |
| calib | `calib-19460` | 19,460 | convergence-calibrated SFT (branch-filtered traces, loop removal), 4 epochs from `cpt-2903` weights-only | completed 2026-09 |

The `calib` stage was an attempt to fix the dominant failure mode — on hard
problems the model reasons correctly but never emits code (runs out of tokens
inside reasoning loops). Data curation per arXiv:2604.01702 (branching-keyword
filter, loop removal, step deletion) + 4× more epochs.

## Eval snapshots (LLM-generated, exec-checked; LiveCodeBench v6 sample)

| Checkpoint | pass@1 | case pass rate | GGUF quant |
|---|---|---|---|
| sft-14000 | 43.3% | 41.1% | Q4_K_M |
| sft-14000 | 26.7% | 26.7% | Q8_0 (early broken run) |
| cpt-2903 | 51.7% | 39.7% | Q4_K_M |
| calib-19460 | 48.3% | 36.4% | Q4_K_M |
Honest bottom line: easy problems ≈ solved (100% at cpt-2903), but the
convergence-calibration pass **did not lift overall pass@1** — hard-problem
generation-length failures persisted. The experiment was paused here rather
than extended (next planned step was a small DPO stage targeting looped
generations).
## Resuming / reusing
- Base model: [inclusionAI/Ling-3.0-tiny](https://huggingface.co/inclusionAI/Ling-3.0-tiny)
- Adapter checkpoints pair 1:1 with `optstate_*` files for exact resume.
- The two `models/` GGUFs are the only turnkey artifacts; a merged GGUF of the
  later checkpoints was never produced.

	---
	license: mit
	base_model: inclusionAI/Ling-3.0-tiny
	library_name: safetensors
	---

	# ling3-tiny-coder-ckpts — work-in-progress (PAUSED)

	> Status: paused / unfinished. This repo stores intermediate artifacts from a
	> personal experiment: fine-tuning [Ling-3.0-tiny](https://huggingface.co/inclusionAI/Ling-3.0-tiny)
	> (7.9B Bailing-hybrid MoE, KDA+MLA) into a C++ competitive-programming coder on
	> Kaggle TPU v5e-8 with LoRA adapters. Development is inactive — the items here are
	> training artifacts, not a finished model. Layout may change if work resumes.

	## What is in this repo

	- `models/` contains the only directly usable artifacts: two early-SFT merged
	GGUF exports (Q4_K_M / Q8_0 / bf16 + chat template). These are from an early
	mid-SFT point and are noticeably weaker than later checkpoints (see evals).
	- `evals/` stores raw LiveCodeBench-v6 (60-problem sample) + OJBench result JSONs.
	- `latest.json` / `routing_counts.npz` / `llama-bin-cache.tar.gz` are pipeline
	bookkeeping (resume pointer, MoE expert-usage counters, llama.cpp build cache).

	## Training lineage

	\| Stage \| Checkpoint \| Steps \| Data \| Notes \|
	\|---\|---\|---\|---\|---\|
	\| SFT \| `sft-21750` \| 21,750 \| R1 reasoning traces (codeforces-cots) + Magicoder thinking-off, ~178M tok, 2 epochs \| completed \|
	\| CPT \| `cpt-2903` \| 2,903 \| raw C++ (Stack-v2) acclimation, 24M tok \| completed; best eval to date \|
	\| agentic (v22) \| `agentic-600` \| 600 \| SWE-smith C++ patches \| partial \|
	\| calib \| `calib-19460` \| 19,460 \| convergence-calibrated SFT (branch-filtered traces, loop removal), 4 epochs from `cpt-2903` weights-only \| completed 2026-09 \|

	The `calib` stage was an attempt to fix the dominant failure mode — on hard
	problems the model reasons correctly but never emits code (runs out of tokens
	inside reasoning loops). Data curation per arXiv:2604.01702 (branching-keyword
	filter, loop removal, step deletion) + 4× more epochs.

	## Eval snapshots (LLM-generated, exec-checked; LiveCodeBench v6 sample)

	\| Checkpoint \| pass@1 \| case pass rate \| GGUF quant \|
	\|---\|---\|---\|---\|
	\| sft-14000 \| 43.3% \| 41.1% \| Q4_K_M \|
	\| sft-14000 \| 26.7% \| 26.7% \| Q8_0 (early broken run) \|
	\| cpt-2903 \| 51.7% \| 39.7% \| Q4_K_M \|
	\| calib-19460 \| 48.3% \| 36.4% \| Q4_K_M \|

	Honest bottom line: easy problems ≈ solved (100% at cpt-2903), but the
	convergence-calibration pass did not lift overall pass@1 — hard-problem
	generation-length failures persisted. The experiment was paused here rather
	than extended (next planned step was a small DPO stage targeting looped
	generations).

	## Resuming / reusing

	- Base model: [inclusionAI/Ling-3.0-tiny](https://huggingface.co/inclusionAI/Ling-3.0-tiny)
	- Adapter checkpoints pair 1:1 with `optstate_*` files for exact resume.
	- The two `models/` GGUFs are the only turnkey artifacts; a merged GGUF of the
	later checkpoints was never produced.