feat: supports openai chat completions API (huggingface#1427)

This PR adds support to make TGI a drop in replacement for OpenAI
clients by exposing the same HTTP interface.

Notes
- TGI inits a single model at startup so the `model` field is unused in
HTTP requests.
- `max_tokens` and `stream` should work as expected but other params may
be (unimplemented or not supported)

General approach
- fetch the `tokenizer_config` at startup from the hub
- pass `tokenizer_config` into `Infer` so we have it at request time
- use the `chat_template` on the config to format chat request
- parse jinja template and render chat string
- pass inputs into existing generate function
- wrap generation output in expected structure before returning

```bash
curl localhost:3000/v1/chat/completions \
    -X POST \
    -d '{
  "model": "tgi",
  "messages": [
    {
      "role": "system",
      "content": "You are a helpful assistant."
    },
    {
      "role": "user",
      "content": "What is deep learning?"
    }
  ],
  "stream": true,
  "max_tokens": 20
}' \
    -H 'Content-Type: application/json'
```

It is also possible to use the `openai` python library and change the
base url

```python
from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:3000/v1",
    api_key="not needed for a local LLM"
)

chat_completion = client.chat.completions.create(
    model="tgi",
    messages=[
        {"role": "system", "content": "You are a helpful assistant." },
        {"role": "user", "content": "What is deep learning?"}
    ],
    stream=True
)

for message in chat_completion:
    print(message)

```

```python
from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:3000/v1",
    api_key="not needed for a local LLM"
)

chat_completion = client.chat.completions.create(
    model="tgi",
    messages=[
        {"role": "system", "content": "You are a helpful assistant." },
        {"role": "user", "content": "What is deep learning?"}
    ],
    stream=False
)

print(chat_completion)
```

```bash
cd text-generation-inference/server
MASTER_ADDR=127.0.0.1 MASTER_PORT=5555 text-generation-server serve --trust-remote-code gpt2
```

***note many of the existing `chat_templates` use non standard `jinja`
(ie. adding a `raise` to the template) which will throw an error when
parsing; hence using `upstage/SOLAR-10.7B-Instruct-v1.0` since it has a
valid template
```bash
cd text-generation-inference/router
cargo run -- --tokenizer-name upstage/SOLAR-10.7B-Instruct-v1.0
```

trigger
```bash
curl localhost:3000/v1/chat/completions \
    -X POST \
    -d '{ "model": "gpt-3.5-turbo", "messages": [ { "role": "system", "content": "You are a helpful assistant." }, { "role": "user", "content": "What is the IP address of the Google DNS servers?" } ], "stream": true, "max_tokens": 20, "logprobs": true }' \
    -H 'Content-Type: application/json'
```

^ supports `stream: true` and `stream: false` requests

router/Cargo.toml

@@ -43,6 +43,8 @@ utoipa = { version = "3.5.0", features = ["axum_extras"] }
 utoipa-swagger-ui = { version = "3.1.5", features = ["axum"] }
 ngrok = { version = "0.13.1", features = ["axum"], optional = true }
 init-tracing-opentelemetry = { version = "0.14.1", features = ["opentelemetry-otlp"] }
+minijinja = "1.0.10"
+futures-util = "0.3.30"
 
 [build-dependencies]
 vergen = { version = "8.2.5", features = ["build", "git", "gitcl"] }

router/src/infer.rs

@@ -2,9 +2,11 @@
 
 /// Batching and inference logic
 use crate::validation::{Validation, ValidationError};
+use crate::HubTokenizerConfig;
+use crate::{ChatRequest, GenerateRequest, GenerateStreamResponse, PrefillToken};
 use crate::{Entry, Queue, Token};
-use crate::{GenerateRequest, PrefillToken};
 use futures::future::try_join_all;
+use minijinja::{Environment, ErrorKind, Template};
 use nohash_hasher::IntMap;
 use std::sync::{
     atomic::{AtomicBool, Ordering},
@@ -15,7 +17,7 @@ use text_generation_client::{
 };
 use thiserror::Error;
 use tokio::sync::mpsc::error::SendError;
-use tokio::sync::{mpsc, Notify, OwnedSemaphorePermit, Semaphore, TryAcquireError};
+use tokio::sync::{mpsc, Notify, Semaphore, TryAcquireError};
 use tokio::time::Instant;
 use tokio_stream::wrappers::UnboundedReceiverStream;
 use tokio_stream::StreamExt;
@@ -32,6 +34,8 @@ pub struct Infer {
     shared: Arc<Shared>,
     /// Inference limit
     limit_concurrent_requests: Arc<Semaphore>,
+    /// Chat template
+    template: Option<Template<'static, 'static>>,
 }
 
 /// Infer shared state
@@ -56,6 +60,7 @@ impl Infer {
         window_size: Option<u32>,
         speculate: u32,
         generation_health: Arc<AtomicBool>,
+        tokenizer_config: HubTokenizerConfig,
     ) -> Self {
         // Infer shared state
         let queue = Queue::new(
@@ -85,11 +90,21 @@ impl Infer {
         // Inference limit with a semaphore
         let semaphore = Arc::new(Semaphore::new(max_concurrent_requests));
 
+        let template = tokenizer_config.chat_template.map(|t| {
+            let env = Box::new(Environment::new());
+            let template_str = t.into_boxed_str();
+            // leaking env and template_str as read-only, static resources for performance.
+            Box::leak(env)
+                .template_from_str(Box::leak(template_str))
+                .unwrap()
+        });
+
         Self {
             validation,
             queue,
             shared,
             limit_concurrent_requests: semaphore,
+            template,
         }
     }
 
@@ -98,14 +113,7 @@ impl Infer {
     pub(crate) async fn generate_stream(
         &self,
         request: GenerateRequest,
-    ) -> Result<
-        (
-            OwnedSemaphorePermit,
-            u32,
-            UnboundedReceiverStream<Result<InferStreamResponse, InferError>>,
-        ),
-        InferError,
-    > {
+    ) -> Result<GenerateStreamResponse, InferError> {
         // Limit concurrent requests by acquiring a permit from the semaphore
         let permit = self
             .clone()
@@ -150,6 +158,20 @@ impl Infer {
         ))
     }
 
+    /// Apply the chat template to the chat request
+    #[instrument(skip_all)]
+    pub(crate) fn apply_chat_template(&self, chat: ChatRequest) -> Result<String, InferError> {
+        self.template
+            .as_ref()
+            .ok_or_else(|| InferError::TemplateError(ErrorKind::TemplateNotFound.into()))?
+            .render(chat)
+            .map_err(|e| {
+                metrics::increment_counter!("tgi_request_failure", "err" => "template");
+                tracing::error!("{e}");
+                InferError::TemplateError(e)
+            })
+    }
+
     /// Add a new request to the queue and return a InferResponse
     #[instrument(skip_all)]
     pub(crate) async fn generate(
@@ -561,9 +583,9 @@ fn send_responses(
     let mut iterator = tokens_
         .ids
         .into_iter()
-        .zip(tokens_.logprobs.into_iter())
-        .zip(tokens_.texts.into_iter())
-        .zip(tokens_.is_special.into_iter())
+        .zip(tokens_.logprobs)
+        .zip(tokens_.texts)
+        .zip(tokens_.is_special)
         .enumerate()
         .peekable();
     while let Some((i, (((id, logprob), text), special))) = iterator.next() {
@@ -676,6 +698,8 @@ pub enum InferError {
     ValidationError(#[from] ValidationError),
     #[error("Incomplete generation")]
     IncompleteGeneration,
+    #[error("Template error: {0}")]
+    TemplateError(#[from] minijinja::Error),
 }
 
 impl InferError {
@@ -685,6 +709,7 @@ impl InferError {
             InferError::Overloaded(_) => "overloaded",
             InferError::ValidationError(_) => "validation",
             InferError::IncompleteGeneration => "incomplete_generation",
+            InferError::TemplateError(_) => "template_error",
         }
     }
 }