Instructions to use chinoll/chatsakura-3b-int4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use chinoll/chatsakura-3b-int4 with Transformers:

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

pipe = pipeline("text-generation", model="chinoll/chatsakura-3b-int4", trust_remote_code=True)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("chinoll/chatsakura-3b-int4", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("chinoll/chatsakura-3b-int4", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use chinoll/chatsakura-3b-int4 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "chinoll/chatsakura-3b-int4"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "chinoll/chatsakura-3b-int4",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/chinoll/chatsakura-3b-int4

SGLang

How to use chinoll/chatsakura-3b-int4 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 "chinoll/chatsakura-3b-int4" \
    --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": "chinoll/chatsakura-3b-int4",
		"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 "chinoll/chatsakura-3b-int4" \
        --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": "chinoll/chatsakura-3b-int4",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use chinoll/chatsakura-3b-int4 with Docker Model Runner:
```
docker model run hf.co/chinoll/chatsakura-3b-int4
```

chinoll commited on Apr 1, 2023

Commit

ac134b7

1 Parent(s): fba9109

Update quant.py

Browse files

Files changed (1) hide show

quant.py +1 -1

quant.py CHANGED Viewed

@@ -657,7 +657,7 @@ __global__ void VecQuant8MatMulKernel(
   atomicAdd(&mul[b * width + w], res);
 }
     '''
-    open("quant_cuda_kernel.cu","w") as f:
         f.write(cucode)
     cppcode = '''
 #include <torch/all.h>

   atomicAdd(&mul[b * width + w], res);
 }
     '''
+    with open("quant_cuda_kernel.cu","w") as f:
         f.write(cucode)
     cppcode = '''
 #include <torch/all.h>