Instructions to use mpasila/Psyfighter-13B-exl2-3bpw with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use mpasila/Psyfighter-13B-exl2-3bpw with Transformers:

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

pipe = pipeline("text-generation", model="mpasila/Psyfighter-13B-exl2-3bpw")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("mpasila/Psyfighter-13B-exl2-3bpw")
model = AutoModelForCausalLM.from_pretrained("mpasila/Psyfighter-13B-exl2-3bpw", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use mpasila/Psyfighter-13B-exl2-3bpw with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "mpasila/Psyfighter-13B-exl2-3bpw"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mpasila/Psyfighter-13B-exl2-3bpw",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/mpasila/Psyfighter-13B-exl2-3bpw

SGLang

How to use mpasila/Psyfighter-13B-exl2-3bpw 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 "mpasila/Psyfighter-13B-exl2-3bpw" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mpasila/Psyfighter-13B-exl2-3bpw",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "mpasila/Psyfighter-13B-exl2-3bpw" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mpasila/Psyfighter-13B-exl2-3bpw",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use mpasila/Psyfighter-13B-exl2-3bpw with Docker Model Runner:
```
docker model run hf.co/mpasila/Psyfighter-13B-exl2-3bpw
```

This is an ExLlamaV2 quantized model in 3bpw of jebcarter/Psyfighter-13B using PIPPA as the calibration dataset. (I might reupload this or add a new branch with a different version that uses the default calibration dataset instead)

Alpaca and ChatML prompt templates seem to work fine with this model.

Original Model card

merge_method: task_arithmaetic
base_model: TheBloke/Llama-2-13B-fp16
models:
  - model: TheBloke/Llama-2-13B-fp16
  - model: KoboldAI/LLaMA2-13B-Tiefighter
    parameters:
      weight: 1.0
  - model: chaoyi-wu/MedLLaMA_13B
    parameters:
      weight: 0.01
  - model: Doctor-Shotgun/llama-2-13b-chat-limarp-v2-merged
    parameters:
      weight: 0.02
dtype: float16

This model was made possible thanks to the Compute provided by the KoboldAI community.

Downloads last month: 2

Collection including mpasila/Psyfighter-13B-exl2-3bpw

ExLlamaV2 quantizations

Collection

All my EXL2 quants here. • 32 items • Updated May 18, 2024