Instructions to use ego-hf/CodeLlama-7b-Python with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ego-hf/CodeLlama-7b-Python with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="ego-hf/CodeLlama-7b-Python",
	filename="ggml-model-f32-00001-of-00010.gguf",
)

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use ego-hf/CodeLlama-7b-Python with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf ego-hf/CodeLlama-7b-Python:F32
# Run inference directly in the terminal:
llama-cli -hf ego-hf/CodeLlama-7b-Python:F32

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf ego-hf/CodeLlama-7b-Python:F32
# Run inference directly in the terminal:
llama-cli -hf ego-hf/CodeLlama-7b-Python:F32

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 ego-hf/CodeLlama-7b-Python:F32
# Run inference directly in the terminal:
./llama-cli -hf ego-hf/CodeLlama-7b-Python:F32

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 ego-hf/CodeLlama-7b-Python:F32
# Run inference directly in the terminal:
./build/bin/llama-cli -hf ego-hf/CodeLlama-7b-Python:F32

Use Docker

docker model run hf.co/ego-hf/CodeLlama-7b-Python:F32

LM Studio
Jan
Ollama
How to use ego-hf/CodeLlama-7b-Python with Ollama:
```
ollama run hf.co/ego-hf/CodeLlama-7b-Python:F32
```

Unsloth Studio new

How to use ego-hf/CodeLlama-7b-Python 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 ego-hf/CodeLlama-7b-Python 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 ego-hf/CodeLlama-7b-Python to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for ego-hf/CodeLlama-7b-Python to start chatting

Docker Model Runner
How to use ego-hf/CodeLlama-7b-Python with Docker Model Runner:
```
docker model run hf.co/ego-hf/CodeLlama-7b-Python:F32
```

Lemonade

How to use ego-hf/CodeLlama-7b-Python with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull ego-hf/CodeLlama-7b-Python:F32

Run and chat with the model

lemonade run user.CodeLlama-7b-Python-F32

List all available models

lemonade list

CodeLlama-7b-Python / README.md

ego-hf

Update README.md

61ba0ed verified about 2 years ago

preview code

raw

history blame contribute delete

1.61 kB

	---
	license: llama2
	base_model: meta-llama/CodeLlama-7b-Python-hf
	model_name: CodeLlama-7b-Python GGUF
	model_type: llama
	language:
	- code
	tags:
	- LLM
	- llama2
	- llama-2
	- CodeLlama
	- CodeLlama-Python
	- CodeLlama-7B-Python
	- lama.cpp
	- Python
	- 7B
	---


	# Model Card: Meta CodeLlama-7b-Python gguf

	Origin Meta model [CodeLlama-7b-Python](https://llama.meta.com/llama-downloads/), [code llama large language model coding](https://ai.meta.com/blog/code-llama-large-language-model-coding/), [codellama](https://github.com/meta-llama/codellama) converted into gguf format with [llama.cpp](https://github.com/ggerganov/llama.cpp)

	Licen: "Llama 2 is licensed under the LLAMA 2 Community License, Copyright © Meta Platforms, Inc. All Rights Reserved."

	[Policy](https://llama.meta.com/use-policy/)


	# Run model

	```bash
	./main -m ggml-model-f32-00001-of-00010.gguf -p "def fibonacci("
	```


	## Convert to gguf
	```bash
	python3 convert.py ../codellama/CodeLlama-7b-Python
	```


	## Split Model

	Original Meta `CodeLlama-7b-Python` model converted with [python3 convert.py](https://github.com/ggerganov/llama.cpp) to `gguf` and
	`CodeLlama-7b-Python/ggml-model-f32.gguf` and splitted with [gguf-split](https://github.com/ggerganov/llama.cpp) to smaller size chunks up to `split-max-tensors 32`.

	```bash
	python3 convert.py ../codellama/CodeLlama-7b-Python
	./gguf-split --split --split-max-tensors 32 ./models/CodeLlama-7b-Python/ggml-model-f32.gguf ./models/CodeLlama-7b-Python/ggml-model-f32
	```


	## Merge-back model use

	```bash
	./gguf-split --merge ggml-model-f32-00001-of-00010.gguf ggml-model-f32.gguf
	```