Instructions to use emre/llama-2-13b-code-122k with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use emre/llama-2-13b-code-122k with Transformers:

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

pipe = pipeline("text-generation", model="emre/llama-2-13b-code-122k")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("emre/llama-2-13b-code-122k")
model = AutoModelForCausalLM.from_pretrained("emre/llama-2-13b-code-122k")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use emre/llama-2-13b-code-122k with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "emre/llama-2-13b-code-122k"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "emre/llama-2-13b-code-122k",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/emre/llama-2-13b-code-122k

SGLang

How to use emre/llama-2-13b-code-122k 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 "emre/llama-2-13b-code-122k" \
    --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": "emre/llama-2-13b-code-122k",
		"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 "emre/llama-2-13b-code-122k" \
        --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": "emre/llama-2-13b-code-122k",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use emre/llama-2-13b-code-122k with Docker Model Runner:
```
docker model run hf.co/emre/llama-2-13b-code-122k
```

What coding dataset was used to train this model?

by rombodawg - opened Aug 9, 2023

Discussion

rombodawg

Aug 9, 2023

What coding dataset was used to train this model?

Also if you are interested I have 2 datasets for code training if you wanted to make more models.

One more only that may lead to loss of logical function:
https://huggingface.co/datasets/rombodawg/2XUNCENSORED_MegaCodeTraining188k

And one that is meant to be lossless and provide coding function:
https://huggingface.co/datasets/rombodawg/LosslessMegaCodeTrainingV2_1m_Evol_Uncensored

emre

Owner Aug 9, 2023

Lets talk about it, i am interested.
I have used the dataset in my profile. 122k
I have checked your datasets you should convert them to llama2 format like mine. Convert them, add my dataset and create a new dataset from all, then i can fine tune it as soon as possible.

rombodawg

Aug 9, 2023

•

edited Aug 9, 2023

How did you create you 122k dataset? Was it created using gpt-4 prompting? Or was it sourced from somewhere on huggingface?

emre

Owner Aug 9, 2023

emre/llama-2-instruct-121k-code
I took it from another repo

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