Instructions to use qihoo360/360Zhinao-1.8B-Reranking with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use qihoo360/360Zhinao-1.8B-Reranking with Transformers:

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

pipe = pipeline("text-generation", model="qihoo360/360Zhinao-1.8B-Reranking", trust_remote_code=True)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("qihoo360/360Zhinao-1.8B-Reranking", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use qihoo360/360Zhinao-1.8B-Reranking with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "qihoo360/360Zhinao-1.8B-Reranking"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "qihoo360/360Zhinao-1.8B-Reranking",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/qihoo360/360Zhinao-1.8B-Reranking

SGLang

How to use qihoo360/360Zhinao-1.8B-Reranking 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 "qihoo360/360Zhinao-1.8B-Reranking" \
    --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": "qihoo360/360Zhinao-1.8B-Reranking",
		"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 "qihoo360/360Zhinao-1.8B-Reranking" \
        --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": "qihoo360/360Zhinao-1.8B-Reranking",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use qihoo360/360Zhinao-1.8B-Reranking with Docker Model Runner:
```
docker model run hf.co/qihoo360/360Zhinao-1.8B-Reranking
```

Update modeling_zhinao.py

by neofung - opened Jul 1, 2024

base: refs/heads/main

←

from: refs/pr/2

Discussion Files changed

+12

-13

Files changed (1) hide show

modeling_zhinao.py +12 -13

modeling_zhinao.py CHANGED Viewed

@@ -748,6 +748,17 @@ class ZhinaoForCausalLM(ZhinaoPreTrainedModel):
     def __init__(self, config):
         super().__init__(config)
         self.model = ZhinaoModel(config)
         self.vocab_size = config.vocab_size
         self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
@@ -761,19 +772,7 @@ class ZhinaoForCausalLM(ZhinaoPreTrainedModel):
         if config.fp16:
             self.model.half()
             self.lm_head.half()
-            self.linear.half()
-        if config.use_flash_attn == "auto":
-            if flash_attn_varlen_func:
-                if config.bf16 or config.fp16:
-                    logger.warn("Try importing flash-attention.")
-                    config.use_flash_attn = True
-                else:
-                    config.use_flash_attn = False
-                    logger.warn("Flash attention will be disabled because it does NOT support fp32.")
-            else:
-                config.use_flash_attn = False
-                logger.warn("Please install FlashAttention first, " "e.g., with pip install flash-attn")
         self.post_init()

     def __init__(self, config):
         super().__init__(config)
+        if config.use_flash_attn == "auto":
+            if flash_attn_varlen_func:
+                if config.bf16 or config.fp16:
+                    logger.warn("Try importing flash-attention.")
+                    config.use_flash_attn = True
+                else:
+                    config.use_flash_attn = False
+                    logger.warn("Flash attention will be disabled because it does NOT support fp32.")
+            else:
+                config.use_flash_attn = False
+                logger.warn("Please install FlashAttention first, " "e.g., with pip install flash-attn")
         self.model = ZhinaoModel(config)
         self.vocab_size = config.vocab_size
         self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
         if config.fp16:
             self.model.half()
             self.lm_head.half()
+            self.linear.half()
         self.post_init()