Instructions to use abhinavkulkarni/mosaicml-mpt-30b-chat-w4-g128-awq with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use abhinavkulkarni/mosaicml-mpt-30b-chat-w4-g128-awq with Transformers:

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

pipe = pipeline("text-generation", model="abhinavkulkarni/mosaicml-mpt-30b-chat-w4-g128-awq", trust_remote_code=True)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("abhinavkulkarni/mosaicml-mpt-30b-chat-w4-g128-awq", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("abhinavkulkarni/mosaicml-mpt-30b-chat-w4-g128-awq", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use abhinavkulkarni/mosaicml-mpt-30b-chat-w4-g128-awq with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "abhinavkulkarni/mosaicml-mpt-30b-chat-w4-g128-awq"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "abhinavkulkarni/mosaicml-mpt-30b-chat-w4-g128-awq",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/abhinavkulkarni/mosaicml-mpt-30b-chat-w4-g128-awq

SGLang

How to use abhinavkulkarni/mosaicml-mpt-30b-chat-w4-g128-awq 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 "abhinavkulkarni/mosaicml-mpt-30b-chat-w4-g128-awq" \
    --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": "abhinavkulkarni/mosaicml-mpt-30b-chat-w4-g128-awq",
		"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 "abhinavkulkarni/mosaicml-mpt-30b-chat-w4-g128-awq" \
        --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": "abhinavkulkarni/mosaicml-mpt-30b-chat-w4-g128-awq",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use abhinavkulkarni/mosaicml-mpt-30b-chat-w4-g128-awq with Docker Model Runner:
```
docker model run hf.co/abhinavkulkarni/mosaicml-mpt-30b-chat-w4-g128-awq
```

Abhinav Kulkarni commited on Aug 2, 2023

Commit

06d3986

1 Parent(s): 8ba19c3

Updated README

Browse files

Files changed (1) hide show

README.md +5 -0

README.md CHANGED Viewed

@@ -39,6 +39,7 @@ git clone https://github.com/mit-han-lab/llm-awq \
 ```
 ```python
 import torch
 from awq.quantize.quantizer import real_quantize_model_weight
 from transformers import AutoModelForCausalLM, AutoConfig, AutoTokenizer, TextStreamer
@@ -80,6 +81,7 @@ prompt = f'''What is the difference between nuclear fusion and fission?
 ###Response:'''
 input_ids = tokenizer(prompt, return_tensors='pt').input_ids.cuda()
 output = model.generate(
     inputs=input_ids,
     temperature=0.7,
@@ -89,6 +91,9 @@ output = model.generate(
     repetition_penalty=1.1,
     eos_token_id=tokenizer.eos_token_id,
     streamer=streamer)
 ```
 ## Evaluation

 ```
 ```python
+import time
 import torch
 from awq.quantize.quantizer import real_quantize_model_weight
 from transformers import AutoModelForCausalLM, AutoConfig, AutoTokenizer, TextStreamer
 ###Response:'''
 input_ids = tokenizer(prompt, return_tensors='pt').input_ids.cuda()
+t1 = time.time()
 output = model.generate(
     inputs=input_ids,
     temperature=0.7,
     repetition_penalty=1.1,
     eos_token_id=tokenizer.eos_token_id,
     streamer=streamer)
+t2 = time.time()
+print("*"*80)
+print(f"Generated {num_tokens/(t2-t1):.2f} token/s; {(t2-t1)*1000/num_tokens:.2f} ms/token")
 ```
 ## Evaluation