A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 17, 03:00 UTC
[Doc] Fix getting stared to use publicly available model (#3963)
Frαnçois committed
Apr 10, 2024 at 20:05 UTC
92cd2e2f21e8ec65b2cb635a9f15de38157a1359
1 file changed
+4
-6
docs/source/serving/openai_compatible_server.md
+4
-6
@@ -4,7 +4,7 @@ vLLM provides an HTTP server that implements OpenAI's [Completions](https://plat
4
5
You can start the server using Python, or using [Docker](deploying_with_docker.rst):
6
```bash
7
-python -m vllm.entrypoints.openai.api_server --model meta-llama/Llama-2-7b-hf --dtype float32 --api-key token-abc123
7
+python -m vllm.entrypoints.openai.api_server --model mistralai/Mistral-7B-Instruct-v0.2 --dtype auto --api-key token-abc123
8
```
9
10
To call the server, you can use the official OpenAI Python client library, or any other HTTP client.
@@ -16,9 +16,8 @@ client = OpenAI(
16
)
17
18
completion = client.chat.completions.create(
19
- model="meta-llama/Llama-2-7b-hf",
19
+ model="mistralai/Mistral-7B-Instruct-v0.2",
20
messages=[
21
- {"role": "system", "content": "You are a helpful assistant."},
21
{"role": "user", "content": "Hello!"}
22
]
23
)
@@ -38,9 +37,8 @@ Or directly merge them into the JSON payload if you are using HTTP call directly
37
38
```python
39
completion = client.chat.completions.create(
41
- model="meta-llama/Llama-2-7b-hf",
40
+ model="mistralai/Mistral-7B-Instruct-v0.2",
41
messages=[
43
- {"role": "system", "content": "You are a helpful assistant."},
42
{"role": "user", "content": "Classify this sentiment: vLLM is wonderful!"}
43
],
44
extra_body={
@@ -89,7 +87,7 @@ In order for the language model to support chat protocol, vLLM requires the mode
87
a chat template in its tokenizer configuration. The chat template is a Jinja2 template that
88
specifies how are roles, messages, and other chat-specific tokens are encoded in the input.
89
92
-An example chat template for `meta-llama/Llama-2-7b-chat-hf` can be found [here](https://huggingface.co/meta-llama/Llama-2-7b-chat-hf/blob/09bd0f49e16738cdfaa6e615203e126038736eb0/tokenizer_config.json#L12)
90
+An example chat template for `mistralai/Mistral-7B-Instruct-v0.2` can be found [here](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2#instruction-format)
91
92
Some models do not provide a chat template even though they are instruction/chat fine-tuned. For those model,
93
you can manually specify their chat template in the `--chat-template` parameter with the file path to the chat