Instructions to use latimar/Phind-Codellama-34B-v2-exl2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use latimar/Phind-Codellama-34B-v2-exl2 with Transformers:

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

pipe = pipeline("text-generation", model="latimar/Phind-Codellama-34B-v2-exl2")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("latimar/Phind-Codellama-34B-v2-exl2")
model = AutoModelForCausalLM.from_pretrained("latimar/Phind-Codellama-34B-v2-exl2")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use latimar/Phind-Codellama-34B-v2-exl2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "latimar/Phind-Codellama-34B-v2-exl2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "latimar/Phind-Codellama-34B-v2-exl2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/latimar/Phind-Codellama-34B-v2-exl2

SGLang

How to use latimar/Phind-Codellama-34B-v2-exl2 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 "latimar/Phind-Codellama-34B-v2-exl2" \
    --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": "latimar/Phind-Codellama-34B-v2-exl2",
		"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 "latimar/Phind-Codellama-34B-v2-exl2" \
        --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": "latimar/Phind-Codellama-34B-v2-exl2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use latimar/Phind-Codellama-34B-v2-exl2 with Docker Model Runner:
```
docker model run hf.co/latimar/Phind-Codellama-34B-v2-exl2
```

Confirming higher HumanEval for Evol instruct 5bit quant vs. Wikitext based one.

by KrisPi - opened Oct 13, 2023

Discussion

KrisPi

Oct 13, 2023

Finally managed to run HumanEval for 5bits and the different calibration presets you created:
Exllama supports greedy decoding by specifying tok_k = 1; which I did, there is very slight randomness still but it's a character or two.

The benefit is really significant - in Transformers 8 bit(!!!) I cannot cross 70% with this model. 4bit runs 68%

Your Evol quant is also much better than the ones I tried to recreate, the best I got is 70.7% while yours is very close to the official fp16 reported numbers (73% and btw. the community has a hard time recreating this result)

Congrats! Would you also share some details on what makes your quant special? I..e exact file for calibration and commit of Exllamav2 if possible?

Wiki
Base
{'pass@1': 0.6890243902439024}
Base + Extra
{'pass@1': 0.6524390243902439}

Evol
{'pass@1': 0.725609756097561}
Base + Extra
{'pass@1': 0.6707317073170732}

REDO
Base
{'pass@1': 0.7195121951219512}
Base + Extra
{'pass@1': 0.6707317073170732}

latimar

Owner Oct 16, 2023

@KrisPi Calibration dataset for evol quant is linked in the README, it's wizardLM-evol-instruct_70k. As for the commit hash of Exllama2 -- unfortunately I can't tell exactly, I've updated exllama2 many times since then. All I can say is that I was using the latest exllama2 at the moment, and it was about a month ago. Interesting findings about the evol quant, I did not think calibration dataset would matter much. Guess I'll make some more 4bpw quants using evol-instruct for calibration.

latimar

Owner Oct 20, 2023

@KrisPi I made new quants using megacode dataset, they seem to be much better. https://huggingface.co/latimar/Phind-Codellama-34B-v2-megacode-exl2

latimar changed discussion status to closed Oct 20, 2023

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