TheBloke
/

Llama-2-13B-chat-GPTQ

Text Generation

text-generation-inference

4-bit precision

Model card Files Files and versions

Instructions to use TheBloke/Llama-2-13B-chat-GPTQ with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use TheBloke/Llama-2-13B-chat-GPTQ with Transformers:

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

pipe = pipeline("text-generation", model="TheBloke/Llama-2-13B-chat-GPTQ")

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("TheBloke/Llama-2-13B-chat-GPTQ")
model = AutoModelForMultimodalLM.from_pretrained("TheBloke/Llama-2-13B-chat-GPTQ")

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use TheBloke/Llama-2-13B-chat-GPTQ with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "TheBloke/Llama-2-13B-chat-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/Llama-2-13B-chat-GPTQ",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/TheBloke/Llama-2-13B-chat-GPTQ

How to use TheBloke/Llama-2-13B-chat-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/Llama-2-13B-chat-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/Llama-2-13B-chat-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/Llama-2-13B-chat-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/Llama-2-13B-chat-GPTQ",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use TheBloke/Llama-2-13B-chat-GPTQ with Docker Model Runner:
```
docker model run hf.co/TheBloke/Llama-2-13B-chat-GPTQ
```

Resources

View closed (12)

output embeddings

#54 opened almost 2 years ago by

output content

#53 opened almost 2 years ago by

How to convert 4bit model back to fp16 data format?

#52 opened over 2 years ago by

add template

#51 opened over 2 years ago by

torch.cuda.OutOfMemoryError: CUDA out of memory.

#50 opened over 2 years ago by

Can you please provide 'c4' version?

#49 opened over 2 years ago by

How much does it take to inference one sample?

#48 opened over 2 years ago by

Issues with CUDA and exllama_kernels

#47 opened over 2 years ago by

Calling LlamaTokenizerFast.from_pretrained() with the path to a single file or url is not supported for this tokenizer. Use a model identifier or the path to a directory instead.

#46 opened over 2 years ago by

Hallucination issue in Llama-2-13B-chat-GPTQ

#45 opened over 2 years ago by

DivyanshTiwari7

Increasing the model's predefined max length

#44 opened over 2 years ago by

[AUTOMATED] Model Memory Requirements

#43 opened over 2 years ago by

model-sizer-bot

Deploying TheBloke/Llama-2-13B-chat-GPTQ as a batch end point in sagemaker

#41 opened over 2 years ago by

Deploying this on Text Generation Inference (TGI) server on AWS SageMaker

#38 opened almost 3 years ago by

Understanding materials

#37 opened almost 3 years ago by

rishabh-gurbani

Temperature or top_p is not working

#35 opened almost 3 years ago by

Train model with webui

#34 opened almost 3 years ago by

HuggingFace's bitsandbytes vs AutoGPTQ?

#33 opened almost 3 years ago by

What library was used to quantize this model ?

#32 opened almost 3 years ago by

Dataset used for quantisation

#31 opened almost 3 years ago by

How to make it (Llama-2-13B-chat-GPTQ) work with Fastchat

#30 opened almost 3 years ago by

Error: Transformers import module musicgen

#29 opened almost 3 years ago by

Finetuning the model using custom dataset.

#28 opened almost 3 years ago by

Necessary material for llama2

#27 opened almost 3 years ago by

Converting hf format model to 128g.safetensors

#26 opened almost 3 years ago by

Llama-2-13B-chat-GPTQ problem

#23 opened almost 3 years ago by

Getting an error: AttributeError: module 'accelerate.utils' has no attribute 'modeling'. Please tell me what should i do?

#21 opened almost 3 years ago by

Getting error while loading model_basename = "gptq_model-8bit-128g"

#20 opened almost 3 years ago by

fine tune on custom chat dataset using QLORA & PEFT

#19 opened almost 3 years ago by

General Update Question for LLMs

#17 opened almost 3 years ago by

File not found error while loading model

#14 opened almost 3 years ago by

Osamarafique998

CPU Inference

#13 opened almost 3 years ago by

Slow Inference Speed

#12 opened almost 3 years ago by

Error while loading model from path

#11 opened almost 3 years ago by

Censorship is hilarious

#10 opened almost 3 years ago by

why it says no quantize_config.json file but it has

#9 opened almost 3 years ago by

Error loading model from a different branch with revision

#8 opened almost 3 years ago by

Llama v2 GPTQ context length

#7 opened almost 3 years ago by

Is this model based on `chat` or `chat-hf` model of llama2?

#6 opened almost 3 years ago by

Prompt format

#5 opened almost 3 years ago by

Bravo! That was fast : )

#3 opened almost 3 years ago by

Doesn't contain the files

#1 opened almost 3 years ago by