Instructions to use TheBloke/Dolphin-Llama-13B-GPTQ with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use TheBloke/Dolphin-Llama-13B-GPTQ with Transformers:

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

pipe = pipeline("text-generation", model="TheBloke/Dolphin-Llama-13B-GPTQ")

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

tokenizer = AutoTokenizer.from_pretrained("TheBloke/Dolphin-Llama-13B-GPTQ")
model = AutoModelForCausalLM.from_pretrained("TheBloke/Dolphin-Llama-13B-GPTQ", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use TheBloke/Dolphin-Llama-13B-GPTQ with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/TheBloke/Dolphin-Llama-13B-GPTQ

SGLang

How to use TheBloke/Dolphin-Llama-13B-GPTQ 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 "TheBloke/Dolphin-Llama-13B-GPTQ" \
    --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": "TheBloke/Dolphin-Llama-13B-GPTQ",
		"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 "TheBloke/Dolphin-Llama-13B-GPTQ" \
        --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": "TheBloke/Dolphin-Llama-13B-GPTQ",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use TheBloke/Dolphin-Llama-13B-GPTQ with Docker Model Runner:
```
docker model run hf.co/TheBloke/Dolphin-Llama-13B-GPTQ
```

Why such a bad output?

by anon7463435254 - opened Jul 23, 2023

Discussion

anon7463435254

Jul 23, 2023

I was expecting very good results with this model. Am I doing something wrong?

Niichanhaou

Jul 23, 2023

•

edited Jul 23, 2023

Maybe lower temperature? It should reduce the randomness at which the model ouputs. Also what would be interesting to know is how it compares to the original LlaMa-2 Chat model?

I usually don't test models for factual replies, as I'm more interested in using them for creative writing and roleplay chat, but from what I've experimented LlaMa-2 chat is pretty good (except for the damn censorship and moralist agenda, of course). However, finetuning models sometimes seems to damage some of the quality of the original model, even if they release us from the shady agendas

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment