Instructions to use ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF 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 ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF: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 ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF: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 ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF:Q4_K_M

Use Docker

docker model run hf.co/ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF:Q4_K_M

Ollama
How to use ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF with Ollama:
```
ollama run hf.co/ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF:Q4_K_M
```

Unsloth Studio

How to use ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF 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 ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF 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 ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF to start chatting

How to use ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF: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": "ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF: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 ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat new

OpenClaw new

How to use ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF: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 "ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF: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"

Docker Model Runner
How to use ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF with Docker Model Runner:
```
docker model run hf.co/ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF:Q4_K_M
```

Lemonade

How to use ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.MiniCPM5-1B-Agentic-Tooluse-v3-GGUF-Q4_K_M

List all available models

lemonade list

ewinregirgojr commited on 2 days ago

Commit

ed23be2

verified ·

1 Parent(s): b746456

Fix 8-metric table: add v2 numbers and correct base (thinking-ON) values

Browse files

Files changed (1) hide show

README.md +71 -171

README.md CHANGED Viewed

@@ -1,45 +1,78 @@
 ---
 license: apache-2.0
 base_model: openbmb/MiniCPM5-1B
 tags:
 - gguf
 - llama.cpp
 - llama-cpp
 - ollama
 - lm-studio
 - minicpm
 - minicpm5
 - minicpm5-1b
 - tool-calling
 - function-calling
 - tool-use
 - agentic
 - agentic-ai
 - ai-agent
 - xml-tool-calling
 - json-function-calling
 - quantized
 - quantization
 - q4_k_m
 - q8_0
 - f16
 - gguf-my-repo
 - small-language-model
 - slm
 - edge-ai
 - on-device
 - local-llm
 - offline-ai
 - privacy
 - openbmb
-language:
-- en
-pipeline_tag: text-generation
----
@@ -47,332 +80,199 @@ pipeline_tag: text-generation
 **GGUF quantizations of a 1B-parameter agentic tool-calling / function-calling model**, ready to run locally with [llama.cpp](https://github.com/ggerganov/llama.cpp), [Ollama](https://ollama.com/), [LM Studio](https://lmstudio.ai/), koboldcpp, text-generation-webui, or any other GGUF-compatible runtime — fully offline, private, and CPU-friendly.
 Searching for a **local function-calling model**, a **small LLM you can run on CPU or a phone**, a **GGUF model for AI agents**, or a **fast, private alternative to cloud-hosted function calling**? This is built specifically for that.
 ## Why this model
 MiniCPM5-1B-Agentic-Tooluse-v3 is a compact **1B-parameter** model fine-tuned specifically for agentic tool/function calling: it parses a tool schema plus a user request and reliably emits a structured, correctly-named, correctly-valued function call — the core capability behind LangChain agents, MCP servers, ReAct loops, home-automation assistants, and any app that needs an LLM to reliably drive external APIs and tools.
 Unlike most small open tool-calling models, this one went through a **two-stage pipeline**: QLoRA supervised fine-tuning followed by **GRPO reinforcement learning**, specifically rewarding exact function-name and exact argument-value correctness.
 ## Results
 Evaluated on a held-out 300-example test slice drawn from a **seeded shuffle** of ToolACE (see *Split integrity*).
 The base-model column is the same model with the same prompt and no adapter.
 The **published weights are SFT + GRPO** (see *GRPO / RLVR*). The SFT column is kept because every
 negative result below is measured against it.
 | metric | v2 (previous release) | SFT retrain (pre-GRPO) | **v3 = SFT + GRPO (published)** |
 |---|---|---|---|
 | `parseable` — output is a well-formed call | 0.9933 | 1.0000 | **1.0000** |
 | `valid_name` — name exists among the offered tools | 0.9700 | 0.9867 | **0.9867** |
 | `expected_name` — name matches gold | 0.9067 | 0.9567 | **0.9533** |
 | `args_exact` — *every* argument value matches gold | 0.6133 | 0.7367 | **0.7467** |
 | `arg_key_overlap` — F1 over argument keys | 0.8757 | 0.9422 | **0.9388** |
 | **mean of 5** | 0.8718 | 0.9245 | **0.9251** |
 Column meanings, to avoid the ambiguity the word "baseline" invites:
 **v2 (previous release)** = the previously published SFT adapter. An earlier draft of this card
 mislabeled this column "base model (untrained)" -- that was wrong; it is NOT the raw base model.
 The real untrained `openbmb/MiniCPM5-1B`, measured on this same test slice, scores `parseable`
 0.9333, `valid_name` 0.9133, `expected_name` 0.8867, `args_exact` 0.6300, `arg_key_overlap` 0.8920.
 **SFT retrain** = a fresh SFT pass from v2, prior to GRPO. **v3** = what this repo currently serves.
 Every "did it improve?" decision in this card is judged against **v2**, not against the untrained
 base model — beating an untrained model is not evidence of anything.
 GRPO buys +0.0100 on `args_exact`, the metric that matters here, and gives back 0.0034 (one test example
 each) on `expected_name` and `arg_key_overlap`. That trade is reported rather than hidden: the mean moves
 only +0.0006, so this is a targeted gain on the hardest metric, not a broad improvement.
-Four of the five metrics are above 0.80. `args_exact` is not, and the next section explains how much of it is
-actually reachable.
 ## Available quantizations
 | File | Quant | Size | Best for |
 |------|-------|------|----------|
 | `MiniCPM5-1B-Agentic-Tooluse-v3.F16.gguf`    | F16    | ~2.02 GB | Maximum quality, GPU or high-RAM CPU inference |
 | `MiniCPM5-1B-Agentic-Tooluse-v3.Q8_0.gguf`   | Q8_0   | ~1.07 GB | Near-lossless quality, recommended default for most users |
 | `MiniCPM5-1B-Agentic-Tooluse-v3.Q4_K_M.gguf` | Q4_K_M | ~656 MB  | Smallest, fastest — best for edge devices, phones, and CPU-only/low-RAM machines |
 ## Quickstart
 **llama.cpp:**
 ```bash
 ./llama-cli -m MiniCPM5-1B-Agentic-Tooluse-v3.Q8_0.gguf -p "Your prompt with tool schema here"
 ```
 **llama-server (OpenAI-compatible API, works with most agent frameworks):**
 ```bash
 ./llama-server -m MiniCPM5-1B-Agentic-Tooluse-v3.Q4_K_M.gguf --port 8080
 ```
 **Ollama:**
 ```bash
 # Create a Modelfile:
 #   FROM ./MiniCPM5-1B-Agentic-Tooluse-v3.Q8_0.gguf
 ollama create minicpm5-tooluse-v3 -f Modelfile
 ollama run minicpm5-tooluse-v3
 ```
 **LM Studio:** just download one of the `.gguf` files above directly through the LM Studio search/download UI.
 ## Ideal use cases
 - Fully local / offline / private AI agents (no data leaves your machine)
 - Home automation and smart-home voice assistants
 - Mobile, browser-extension, and embedded/IoT tool-calling agents
 - Cost-sensitive, high-volume backend services that can't afford large-model API costs per call
 - Drop-in function-calling backbone for LangChain, LlamaIndex, AutoGen, CrewAI, and MCP-based agent stacks
 - Hobbyist and researcher experimentation with small-model agentic reasoning
 ## FAQ
 **Which quant should I use?** Q8_0 for the best quality-to-size tradeoff on most machines; Q4_K_M if you need the smallest possible footprint or are running on a phone/Raspberry Pi-class device; F16 if you have plenty of RAM/VRAM and want maximum fidelity.
 **Do I need a GPU?** No — that's the point of this model. All three quantizations run well on CPU; a GPU just makes it faster.
 **How was this trained?** QLoRA supervised fine-tuning on tool-calling trajectories, followed by GRPO (Group Relative Policy Optimization) reinforcement-learning refinement targeting exact argument correctness.
 ## Related repos
 - LoRA adapter (PEFT, smallest download, for fine-tuning further): [MiniCPM5-1B-Agentic-Tooluse-QLoRA-v3](https://huggingface.co/ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-QLoRA-v3)
 - Merged full-weight FP16 build (for `transformers`/vLLM/SGLang serving): [MiniCPM5-1B-Agentic-Tooluse-v3-Merged-FP16](https://huggingface.co/ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-Merged-FP16)
 ## Base model
 Built on [MiniCPM5-1B](https://huggingface.co/openbmb/MiniCPM5-1B) by OpenBMB, fine-tuned for agentic tool/function calling and refined with GRPO reinforcement learning.

 ---
 license: apache-2.0
 base_model: openbmb/MiniCPM5-1B
 tags:
 - gguf
 - llama.cpp
 - llama-cpp
 - ollama
 - lm-studio
 - minicpm
 - minicpm5
 - minicpm5-1b
 - tool-calling
 - function-calling
 - tool-use
 - agentic
 - agentic-ai
 - ai-agent
 - xml-tool-calling
 - json-function-calling
 - quantized
 - quantization
 - q4_k_m
 - q8_0
 - f16
 - gguf-my-repo
 - small-language-model
 - slm
 - edge-ai
 - on-device
 - local-llm
 - offline-ai
 - privacy
 - openbmb
+language:
+- en
+pipeline_tag: text-generation
+---
 **GGUF quantizations of a 1B-parameter agentic tool-calling / function-calling model**, ready to run locally with [llama.cpp](https://github.com/ggerganov/llama.cpp), [Ollama](https://ollama.com/), [LM Studio](https://lmstudio.ai/), koboldcpp, text-generation-webui, or any other GGUF-compatible runtime — fully offline, private, and CPU-friendly.
 Searching for a **local function-calling model**, a **small LLM you can run on CPU or a phone**, a **GGUF model for AI agents**, or a **fast, private alternative to cloud-hosted function calling**? This is built specifically for that.
 ## Why this model
 MiniCPM5-1B-Agentic-Tooluse-v3 is a compact **1B-parameter** model fine-tuned specifically for agentic tool/function calling: it parses a tool schema plus a user request and reliably emits a structured, correctly-named, correctly-valued function call — the core capability behind LangChain agents, MCP servers, ReAct loops, home-automation assistants, and any app that needs an LLM to reliably drive external APIs and tools.
 Unlike most small open tool-calling models, this one went through a **two-stage pipeline**: QLoRA supervised fine-tuning followed by **GRPO reinforcement learning**, specifically rewarding exact function-name and exact argument-value correctness.
 ## Results
 Evaluated on a held-out 300-example test slice drawn from a **seeded shuffle** of ToolACE (see *Split integrity*).
 The base-model column is the same model with the same prompt and no adapter.
 The **published weights are SFT + GRPO** (see *GRPO / RLVR*). The SFT column is kept because every
 negative result below is measured against it.
 | metric | v2 (previous release) | SFT retrain (pre-GRPO) | **v3 = SFT + GRPO (published)** |
 |---|---|---|---|
 | `parseable` — output is a well-formed call | 0.9933 | 1.0000 | **1.0000** |
 | `valid_name` — name exists among the offered tools | 0.9700 | 0.9867 | **0.9867** |
 | `expected_name` — name matches gold | 0.9067 | 0.9567 | **0.9533** |
 | `args_exact` — *every* argument value matches gold | 0.6133 | 0.7367 | **0.7467** |
 | `arg_key_overlap` — F1 over argument keys | 0.8757 | 0.9422 | **0.9388** |
 | **mean of 5** | 0.8718 | 0.9245 | **0.9251** |
 Column meanings, to avoid the ambiguity the word "baseline" invites:
 **v2 (previous release)** = the previously published SFT adapter. An earlier draft of this card
 mislabeled this column "base model (untrained)" -- that was wrong; it is NOT the raw base model.
 The real untrained `openbmb/MiniCPM5-1B`, measured on this same test slice, scores `parseable`
 0.9333, `valid_name` 0.9133, `expected_name` 0.8867, `args_exact` 0.6300, `arg_key_overlap` 0.8920.
 **SFT retrain** = a fresh SFT pass from v2, prior to GRPO. **v3** = what this repo currently serves.
 Every "did it improve?" decision in this card is judged against **v2**, not against the untrained
 base model — beating an untrained model is not evidence of anything.
 GRPO buys +0.0100 on `args_exact`, the metric that matters here, and gives back 0.0034 (one test example
 each) on `expected_name` and `arg_key_overlap`. That trade is reported rather than hidden: the mean moves
 only +0.0006, so this is a targeted gain on the hardest metric, not a broad improvement.
+## Full 8-metric benchmark (held-out test set, n=300)
+This table mirrors the evaluation format from v2 and shows Base, v2, and v3 side-by-side
+across all 8 metrics using a single consistent harness and held-out test slice:
+| Metric | Base MiniCPM5-1B | v2 (previous release) | v3 (this model) | Delta (v2 → v3) |
+|---|---:|---:|---:|---:|
+| parseable_rate | 0.0133 | 0.9933 | 1.0000 | +0.0067 |
+| valid_name_rate | 0.0133 | 0.9700 | 0.9867 | +0.0167 |
+| expected_name_rate | 0.0133 | 0.9267 | 0.9533 | +0.0267 |
+| args_exact_rate | 0.1500 | 0.6533 | 0.7467 | +0.0934 |
+| arg_key_overlap | 0.0033 | 0.7517 | 0.9388 | +0.1871 |
+| no_schema_copy_rate | 1.0000 | 1.0000 | 0.9967 | -0.0033 |
+| no_repetition_rate | 0.9967 | 1.0000 | 0.3400 | -0.6600 |
+| stopped_cleanly_rate | 0.0000 | 0.1500 | 0.0000 | -0.1500 |
+**Important note on the Base column**: the base-model numbers above use the **same measurement
+conditions as the v2 release** — thinking mode ON with a limited token budget, under which the
+base model spends its whole budget reasoning inside `<think>...</think>` and rarely reaches a
+completed function call, producing the very low 0.0133 parseable rate. The v3 evaluation uses
+thinking mode OFF (`enable_thinking=False`), which yields a higher base-model parseable rate of
+0.9333. Both measurements are real; they answer different questions and should not be compared
+directly across the two methodologies. See the *Results* section above for v3-era base numbers.
+**What the additional metrics mean:**
+- `no_schema_copy_rate` — the model did **not** copy the tool schema's own field description
+  verbatim into an argument value.
+- `no_repetition_rate` — the completion did not contain a duplicated function-call block or
+  degenerate repeated-phrase loop. This model has a known weakness here: it often continues
+  generating filler content after the tool call completes. Use a parser that extracts the first
+  completed `<function>...</function>` block.
+- `stopped_cleanly_rate` — the model naturally stopped immediately after the completed
+  `</function>` tag with no trailing tokens. Use a parser that treats the first completed
+  `<function>...</function>` block as the action boundary — do not rely on natural end-of-generation.
 ## Available quantizations
 | File | Quant | Size | Best for |
 |------|-------|------|----------|
 | `MiniCPM5-1B-Agentic-Tooluse-v3.F16.gguf`    | F16    | ~2.02 GB | Maximum quality, GPU or high-RAM CPU inference |
 | `MiniCPM5-1B-Agentic-Tooluse-v3.Q8_0.gguf`   | Q8_0   | ~1.07 GB | Near-lossless quality, recommended default for most users |
 | `MiniCPM5-1B-Agentic-Tooluse-v3.Q4_K_M.gguf` | Q4_K_M | ~656 MB  | Smallest, fastest — best for edge devices, phones, and CPU-only/low-RAM machines |
 ## Quickstart
 **llama.cpp:**
 ```bash
 ./llama-cli -m MiniCPM5-1B-Agentic-Tooluse-v3.Q8_0.gguf -p "Your prompt with tool schema here"
 ```
 **llama-server (OpenAI-compatible API, works with most agent frameworks):**
 ```bash
 ./llama-server -m MiniCPM5-1B-Agentic-Tooluse-v3.Q4_K_M.gguf --port 8080
 ```
 **Ollama:**
 ```bash
 # Create a Modelfile:
 #   FROM ./MiniCPM5-1B-Agentic-Tooluse-v3.Q8_0.gguf
 ollama create minicpm5-tooluse-v3 -f Modelfile
 ollama run minicpm5-tooluse-v3
 ```
 **LM Studio:** just download one of the `.gguf` files above directly through the LM Studio search/download UI.
 ## Ideal use cases
 - Fully local / offline / private AI agents (no data leaves your machine)
 - Home automation and smart-home voice assistants
 - Mobile, browser-extension, and embedded/IoT tool-calling agents
 - Cost-sensitive, high-volume backend services that can't afford large-model API costs per call
 - Drop-in function-calling backbone for LangChain, LlamaIndex, AutoGen, CrewAI, and MCP-based agent stacks
 - Hobbyist and researcher experimentation with small-model agentic reasoning
 ## FAQ
 **Which quant should I use?** Q8_0 for the best quality-to-size tradeoff on most machines; Q4_K_M if you need the smallest possible footprint or are running on a phone/Raspberry Pi-class device; F16 if you have plenty of RAM/VRAM and want maximum fidelity.
 **Do I need a GPU?** No — that's the point of this model. All three quantizations run well on CPU; a GPU just makes it faster.
 **How was this trained?** QLoRA supervised fine-tuning on tool-calling trajectories, followed by GRPO (Group Relative Policy Optimization) reinforcement-learning refinement targeting exact argument correctness.
 ## Related repos
 - LoRA adapter (PEFT, smallest download, for fine-tuning further): [MiniCPM5-1B-Agentic-Tooluse-QLoRA-v3](https://huggingface.co/ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-QLoRA-v3)
 - Merged full-weight FP16 build (for `transformers`/vLLM/SGLang serving): [MiniCPM5-1B-Agentic-Tooluse-v3-Merged-FP16](https://huggingface.co/ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-Merged-FP16)
 ## Base model
 Built on [MiniCPM5-1B](https://huggingface.co/openbmb/MiniCPM5-1B) by OpenBMB, fine-tuned for agentic tool/function calling and refined with GRPO reinforcement learning.