Instructions to use Ueuee82683/Puma-1.0 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Ueuee82683/Puma-1.0 with Transformers:

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

pipe = pipeline("text-generation", model="Ueuee82683/Puma-1.0")
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("Ueuee82683/Puma-1.0")
model = AutoModelForCausalLM.from_pretrained("Ueuee82683/Puma-1.0", 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 Ueuee82683/Puma-1.0 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 Ueuee82683/Puma-1.0:Q8_0
# Run inference directly in the terminal:
llama cli -hf Ueuee82683/Puma-1.0:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Ueuee82683/Puma-1.0:Q8_0
# Run inference directly in the terminal:
llama cli -hf Ueuee82683/Puma-1.0:Q8_0

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 Ueuee82683/Puma-1.0:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf Ueuee82683/Puma-1.0:Q8_0

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 Ueuee82683/Puma-1.0:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Ueuee82683/Puma-1.0:Q8_0

Use Docker

docker model run hf.co/Ueuee82683/Puma-1.0:Q8_0

LM Studio
Jan

vLLM

How to use Ueuee82683/Puma-1.0 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Ueuee82683/Puma-1.0"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Ueuee82683/Puma-1.0",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Ueuee82683/Puma-1.0:Q8_0

SGLang

How to use Ueuee82683/Puma-1.0 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 "Ueuee82683/Puma-1.0" \
    --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": "Ueuee82683/Puma-1.0",
		"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 "Ueuee82683/Puma-1.0" \
        --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": "Ueuee82683/Puma-1.0",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use Ueuee82683/Puma-1.0 with Ollama:
```
ollama run hf.co/Ueuee82683/Puma-1.0:Q8_0
```
Unsloth Desktop
Docker Model Runner
How to use Ueuee82683/Puma-1.0 with Docker Model Runner:
```
docker model run hf.co/Ueuee82683/Puma-1.0:Q8_0
```

Lemonade

How to use Ueuee82683/Puma-1.0 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Ueuee82683/Puma-1.0:Q8_0

Run and chat with the model

lemonade run user.Puma-1.0-Q8_0

List all available models

lemonade list

Atomic Chat

Puma 1.0

Puma 1.0 is a 1.42-billion-parameter (1,418,911,232 parameters) open-weight autoregressive language and code generation model created by Merkium and released under the Apache 2.0 license. Built on a 28-layer LlamaForCausalLM decoder (hidden_size = 1536, intermediate_size = 9216, 6:1 Grouped-Query Attention, 32K RoPE) paired with a custom 49,152-token byte-level BPE vocabulary, Puma 1.0 is engineered for responsive local execution on desktop CPUs, laptops, and mobile devices (~1.44 GB in Q8_0 GGUF).

Ollama Registry: Merkium/puma-1.0
Hugging Face Repository: Ueuee82683/Puma-1.0

Architecture & Specifications

Parameter	Value
Architecture	Pre-norm Transformer decoder (`LlamaForCausalLM`, `arch: llama`)
Total Parameters	`1,418,911,232` (`1.42B`, tied input/output embeddings)
Layers (`num_hidden_layers`)	`28`
Hidden Dimension (`hidden_size`)	`1536`
FFN Dimension (`intermediate_size`)	`9216` (`6.0x` SwiGLU MLP, `SiLU` gate)
Attention Heads	`12` Query heads / `2` Key-Value heads (GQA 6:1, `head_dim = 128`)
Positional Encoding	Rotary Position Embeddings (`RoPE`, `theta = 1,000,000.0`)
Normalization	`RMSNorm` (`eps = 1e-6`)
Context Window	`32,768` tokens (`8,192` default in Ollama `Modelfile` for low RAM usage)
Vocabulary (`vocab_size`)	`49,152` custom byte-level BPE (`<
Weights Provided	`model.safetensors` (`FP16`, `2.71 GB`) and `puma-1.0-q8_0.gguf` (`Q8_0`, `1.44 GB`)
License	Apache-2.0

Custom 49,152-Token Puma BPE Vocabulary

Puma 1.0 uses a compact 49,152-token (3 * 2^14) byte-level BPE vocabulary focused on English prose, mathematics, and source code:

IDs 0..255: Complete 256-byte UTF-8 fallback table (0 unknown tokens on any input).
IDs 256..49143: 48,888 ancestor-closed BPE merges prioritized for English, symbolic math, and programming syntax.
IDs 49144..49151: Dedicated control tokens (<|pad|>, <|bos|>, <|eos|>, <|system|>, <|user|>, <|assistant|>, <|fim_prefix|>, <|fim_middle|>), shrinking the embedding matrix to 75.5M parameters (49,152 x 1,536) for faster CPU token selection and lower memory bandwidth.

Capabilities & Intended Scope

What Puma 1.0 Does Well

Python & General Software Engineering: Writes clean, idiomatic functions, data structures, CLI scripts, unit tests, SQL queries, regular expressions, and debugging explanations (90% Python AST validity and 70% strict unit-test pass rate on held-out algorithmic benchmarks including merge_intervals, two_sum, length_of_longest_substring, valid_brackets, binary_search, flatten, and top_k_frequent).
Everyday Assistant & Concise Chat: Handles casual greetings, short follow-ups, brainstorming, summarization, and structured formatting (JSON, YAML, Markdown tables) across multi-turn conversations without losing turn boundaries.
STEM & Introductory Reasoning: Explains algorithms, computer systems, physics, and step-by-step arithmetic clearly at low latency.
Offline & Privacy-Preserving Edge Execution: Runs 100% locally in ~1.6 GB RAM with zero network calls.

Limitations

Compact 1.42B Capacity: Puma 1.0 does not store encyclopedic world trivia or obscure historical citations reliably; pair it with local document retrieval (RAG) when exact factual lookup is required.
English-Centric Vocabulary: Non-Latin scripts fall back to UTF-8 byte tokens (0..255), increasing token count on non-English text.
Single-Pass Reasoning Depth: Complex multi-constraint math or large-codebase refactors benefit from explicit step-by-step prompts or external test execution.

Running Puma 1.0 on CPU (Desktop, Laptop, or Raspberry Pi 5)

Option 1: Ollama (Recommended — One Command)

ollama run Merkium/puma-1.0

Or run directly from Hugging Face via Ollama:

ollama run hf.co/Ueuee82683/Puma-1.0

Expected throughput on modern CPUs (Q8_0, 1.44 GB):

Apple Silicon (M1 / M2 / M3 / M4): 55 - 95 tokens/s
Modern x86_64 Laptop/Desktop CPU (AVX2, 8 threads): 20 - 40 tokens/s
Raspberry Pi 5 (8 GB RAM, ARM64): 7 - 12 tokens/s

Option 2: `llama.cpp` CLI / Server (Pure C++, Zero Python Dependencies)

Download puma-1.0-q8_0.gguf and run directly on CPU:

huggingface-cli download Ueuee82683/Puma-1.0 puma-1.0-q8_0.gguf --local-dir .

# Interactive multi-turn chat on CPU
./llama-cli -m puma-1.0-q8_0.gguf -cnv -c 8192 -t 8 --temp 0.4 \
  -p "You are Puma 1.0, an open-weight 1.42B-parameter AI assistant and coding model created by Merkium and released under Apache 2.0."

# Local OpenAI-compatible HTTP API server on port 8080
./llama-server -m puma-1.0-q8_0.gguf -c 8192 -t 8 --port 8080

Running Puma 1.0 on a Phone (iOS & Android)

Because puma-1.0-q8_0.gguf is a standard arch: llama GGUF (1.44 GB), it runs offline on any modern smartphone with 4 GB+ RAM (iPhone 13 or newer, Pixel 7/8/9, Galaxy S22/S23/S24, or Snapdragon 8-series devices):

PocketPal AI (iOS & Android — Free, Open Source):
- Install PocketPal AI from the iOS App Store or Google Play Store.
- Open Models -> + -> Add from Hugging Face, search Ueuee82683/Puma-1.0, and download puma-1.0-q8_0.gguf (1.44 GB).
- Open model settings, set stop strings to <|eos|> and <|user|>, and tap Load. Expect 12 - 28 tokens/s completely offline (including in Airplane Mode).
Layla Lite / ChatterUI / Maid (Android & iOS):
- Download puma-1.0-q8_0.gguf directly from Ueuee82683/Puma-1.0 onto device storage and load it as a local GGUF model with context size 4096 for minimal battery drain.

Android Termux (ollama or llama.cpp):

pkg update && pkg install ollama
ollama serve &
ollama run Merkium/puma-1.0

Using Puma 1.0 as a Local Assistant (Python API & Transformers)

Local Assistant via OpenAI-Compatible SDK (Ollama)

Once ollama run Merkium/puma-1.0 has pulled the model, query it from any Python script, IDE plugin (Continue.dev, Zed, Cursor local), or CLI tool at http://localhost:11434/v1:

from openai import OpenAI

client = OpenAI(base_url="http://localhost:11434/v1", api_key="ollama")

messages = [
    {"role": "system", "content": "You are Puma 1.0, a concise local engineering assistant."},
    {"role": "user", "content": "Write a Python function to merge overlapping intervals and include 2 assertions."},
]

response = client.chat.completions.create(
    model="Merkium/puma-1.0",
    messages=messages,
    temperature=0.3,
)
print(response.choices[0].message.content)

Direct Hugging Face `transformers` Usage (`FP16` / `CPU`)

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "Ueuee82683/Puma-1.0"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype=torch.float16 if torch.cuda.is_available() else torch.float32,
    device_map="auto",
)

messages = [
    {"role": "user", "content": "Explain binary search time complexity and show a clean Python implementation."}
]
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tokenizer(prompt, return_tensors="pt", add_special_tokens=False).to(model.device)

stop_ids = [
    tokenizer.eos_token_id,
    tokenizer.convert_tokens_to_ids("<|user|>"),
    tokenizer.convert_tokens_to_ids("<|system|>"),
]

with torch.no_grad():
    out = model.generate(
        **inputs,
        max_new_tokens=512,
        temperature=0.4,
        top_p=0.9,
        do_sample=True,
        eos_token_id=stop_ids,
        pad_token_id=tokenizer.pad_token_id,
    )

print(tokenizer.decode(out[0, inputs["input_ids"].shape[1]:], skip_special_tokens=True))

Prompt Format

Puma 1.0 uses explicit turn delimiters from its custom control vocabulary:

<|bos|><|system|>
You are Puma 1.0, an open-weight 1.42B-parameter AI assistant and coding model created by Merkium and released under Apache 2.0.<|eos|>
<|user|>
{user_message}<|eos|>
<|assistant|>
{assistant_response}<|eos|>

License

Released under the Apache License 2.0.

Downloads last month: 104

Safetensors

Model size

1B params

Tensor type

F16