Instructions to use Rahspide/MNN-Pocket-0.3B-RU with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Rahspide/MNN-Pocket-0.3B-RU with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="Rahspide/MNN-Pocket-0.3B-RU")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Rahspide/MNN-Pocket-0.3B-RU")
model = AutoModelForCausalLM.from_pretrained("Rahspide/MNN-Pocket-0.3B-RU", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use Rahspide/MNN-Pocket-0.3B-RU 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 Rahspide/MNN-Pocket-0.3B-RU:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Rahspide/MNN-Pocket-0.3B-RU:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Rahspide/MNN-Pocket-0.3B-RU:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Rahspide/MNN-Pocket-0.3B-RU: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 Rahspide/MNN-Pocket-0.3B-RU:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama cli -hf Rahspide/MNN-Pocket-0.3B-RU:Q4_K_M

Use Docker

docker model run hf.co/Rahspide/MNN-Pocket-0.3B-RU:Q4_K_M

Llama new
LM Studio
Jan

vLLM

How to use Rahspide/MNN-Pocket-0.3B-RU with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Rahspide/MNN-Pocket-0.3B-RU"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Rahspide/MNN-Pocket-0.3B-RU",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Rahspide/MNN-Pocket-0.3B-RU:Q4_K_M

SGLang

How to use Rahspide/MNN-Pocket-0.3B-RU with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "Rahspide/MNN-Pocket-0.3B-RU" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Rahspide/MNN-Pocket-0.3B-RU",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "Rahspide/MNN-Pocket-0.3B-RU" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Rahspide/MNN-Pocket-0.3B-RU",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use Rahspide/MNN-Pocket-0.3B-RU with Ollama:
```
ollama run hf.co/Rahspide/MNN-Pocket-0.3B-RU:Q4_K_M
```
Unsloth Desktop

How to use Rahspide/MNN-Pocket-0.3B-RU with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Rahspide/MNN-Pocket-0.3B-RU: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": "Rahspide/MNN-Pocket-0.3B-RU:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use Rahspide/MNN-Pocket-0.3B-RU with Docker Model Runner:
```
docker model run hf.co/Rahspide/MNN-Pocket-0.3B-RU:Q4_K_M
```

Lemonade

How to use Rahspide/MNN-Pocket-0.3B-RU with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Rahspide/MNN-Pocket-0.3B-RU:Q4_K_M

Run and chat with the model

lemonade run user.MNN-Pocket-0.3B-RU-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use Rahspide/MNN-Pocket-0.3B-RU with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Rahspide/MNN-Pocket-0.3B-RU: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 Rahspide/MNN-Pocket-0.3B-RU:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use Rahspide/MNN-Pocket-0.3B-RU with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Rahspide/MNN-Pocket-0.3B-RU: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 "Rahspide/MNN-Pocket-0.3B-RU: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"

Rahspide commited on 1 day ago

Commit

e3229f1

verified ·

1 Parent(s): b8f7f82

docs: larger caption sizes for the card column

Browse files

Files changed (6) hide show

assets/build_assets.py +11 -8
assets/cover.png +2 -2
assets/eval.png +2 -2
assets/sizes.png +0 -0
assets/tokenizer.png +0 -0
assets/two_jobs.png +0 -0

assets/build_assets.py CHANGED Viewed

@@ -42,6 +42,8 @@ def font(size, family="sans", weight=400):
 def tw(s, size, family="sans", weight=400, ls=0.0):
     """Ширина строки в px: PIL + запас на letter-spacing."""
     f = font(size, family, weight)
     return f.getlength(s) + max(0, len(s) - 1) * ls
@@ -54,6 +56,8 @@ def esc(s):
 def text(x, y, s, size=24, fill=MUTED, family="sans", weight=400, anchor="start",
          ls=0.0, opacity=1.0, upper=False):
     s = s.upper() if upper else s
     fam = "Noto Sans Mono" if family == "mono" else "Noto Sans"
     w = "700" if weight == 700 else "400"
     return (f'<text x="{x:.1f}" y="{y:.1f}" font-family="{fam}" font-size="{size}" '
@@ -226,8 +230,7 @@ def chart_tokenizer():
     W, H = 1600, 620
     b = frame(W, H)
     b.append(text(80, 140, "The tokenizer tax, measured", 44, INK, "sans", 700))
-    b.append(text(80, 180, "Cyrillic BPE merges added on top of the 65 k vocabulary · characters per token on "
-                           "Russian validation text · higher is better", 19, DIM, "mono", 400))
     bars = [("base\n65,536", 2.081, 0.30, None),
             ("+8,192 merges\n73,728", 3.096, 0.62, "×1.49"),
@@ -280,7 +283,7 @@ def diagram_two_jobs(accent="ru"):
     W, H = 1600, 500
     b = frame(W, H)
     b.append(text(80, 126, "One base, two jobs", 44, INK, "sans", 700))
-    b.append(text(80, 164, "The split is deliberate: continuation and answering reward opposite behaviour",
                   19, DIM, "mono", 400))
     def node(x, y, w, h, title, sub, hot=False, small=""):
@@ -291,15 +294,15 @@ def diagram_two_jobs(accent="ru"):
         for i, line_ in enumerate(sub):
             b.append(text(x + 22, y + 72 + i * 24, line_, 17, MUTED, "mono", 400))
         if small:
-            b.append(text(x + w - 20, y + h - 16, small, 14, DIM, "mono", 400, anchor="end"))
-    node(80, 210, 330, 126, "MNN-Pocket-0.3B-Base", ["en / zh · 65,536 tokens", "Qwen3 · Apache-2.0"],
          small="taobao-mnn")
-    node(520, 192, 360, 162, "Continued pretraining", ["1.2 B tokens · 75 / 17 / 8", "ru / en / zh · 1 epoch"],
          hot=True, small="69 h on one T4")
-    node(1010, 104, 500, 100, "MNN-Pocket-0.3B-RU", ["continuation · keyboard · autocomplete"],
          hot=(accent == "ru"), small="this repo" if accent == "ru" else "sister")
-    node(1010, 288, 500, 100, "MNN-Pocket-0.3B-RU-Chat", ["assistant · SFT · ChatML"],
          hot=(accent == "chat"), small="this repo" if accent == "chat" else "sister")
     b.append(line(410, 273, 516, 273, DIM, 1.5))

 def tw(s, size, family="sans", weight=400, ls=0.0):
     """Ширина строки в px: PIL + запас на letter-spacing."""
+    if size <= 19:
+        size += 2
     f = font(size, family, weight)
     return f.getlength(s) + max(0, len(s) - 1) * ls
 def text(x, y, s, size=24, fill=MUTED, family="sans", weight=400, anchor="start",
          ls=0.0, opacity=1.0, upper=False):
     s = s.upper() if upper else s
+    if size <= 19:
+        size += 2
     fam = "Noto Sans Mono" if family == "mono" else "Noto Sans"
     w = "700" if weight == 700 else "400"
     return (f'<text x="{x:.1f}" y="{y:.1f}" font-family="{fam}" font-size="{size}" '
     W, H = 1600, 620
     b = frame(W, H)
     b.append(text(80, 140, "The tokenizer tax, measured", 44, INK, "sans", 700))
+    b.append(text(80, 180, "Cyrillic merges added to the 65 k vocabulary · characters per token · Russian validation text", 19, DIM, "mono", 400))
     bars = [("base\n65,536", 2.081, 0.30, None),
             ("+8,192 merges\n73,728", 3.096, 0.62, "×1.49"),
     W, H = 1600, 500
     b = frame(W, H)
     b.append(text(80, 126, "One base, two jobs", 44, INK, "sans", 700))
+    b.append(text(80, 164, "Continuation and answering reward opposite behaviour",
                   19, DIM, "mono", 400))
     def node(x, y, w, h, title, sub, hot=False, small=""):
         for i, line_ in enumerate(sub):
             b.append(text(x + 22, y + 72 + i * 24, line_, 17, MUTED, "mono", 400))
         if small:
+            b.append(text(x + w - 18, y + h - 14, small, 15, DIM, "mono", 400, anchor="end"))
+    node(80, 210, 330, 140, "MNN-Pocket-0.3B-Base", ["en / zh · 65,536 tokens", "Qwen3 · Apache-2.0"],
          small="taobao-mnn")
+    node(520, 192, 360, 176, "Continued pretraining", ["1.2 B tokens · 75 / 17 / 8", "ru / en / zh · 1 epoch"],
          hot=True, small="69 h on one T4")
+    node(1010, 96, 500, 116, "MNN-Pocket-0.3B-RU", ["continuation · keyboard · autocomplete"],
          hot=(accent == "ru"), small="this repo" if accent == "ru" else "sister")
+    node(1010, 280, 500, 116, "MNN-Pocket-0.3B-RU-Chat", ["assistant · SFT · ChatML"],
          hot=(accent == "chat"), small="this repo" if accent == "chat" else "sister")
     b.append(line(410, 273, 516, 273, DIM, 1.5))