R2OAI / main.py
rkihacker's picture
Update main.py
dafbe9c verified
raw
history blame
12.3 kB
import os
import httpx
import json
import time
from fastapi import FastAPI, HTTPException
from fastapi.responses import JSONResponse
from pydantic import BaseModel, Field
from typing import List, Dict, Any, Optional, Union, Literal
from dotenv import load_dotenv
from sse_starlette.sse import EventSourceResponse
# Load environment variables
load_dotenv()
REPLICATE_API_TOKEN = os.getenv("REPLICATE_API_TOKEN")
if not REPLICATE_API_TOKEN:
raise ValueError("REPLICATE_API_TOKEN environment variable not set.")
# FastAPI Init
app = FastAPI(title="Replicate to OpenAI Compatibility Layer", version="9.0.0 (Definitive Streaming Fix)")
# --- Pydantic Models ---
class ModelCard(BaseModel):
id: str; object: str = "model"; created: int = Field(default_factory=lambda: int(time.time())); owned_by: str = "replicate"
class ModelList(BaseModel):
object: str = "list"; data: List[ModelCard] = []
class ChatMessage(BaseModel):
role: Literal["system", "user", "assistant", "tool"]; content: Union[str, List[Dict[str, Any]]]
class OpenAIChatCompletionRequest(BaseModel):
model: str; messages: List[ChatMessage]; temperature: Optional[float] = 0.7; top_p: Optional[float] = 1.0; max_tokens: Optional[int] = None; stream: Optional[bool] = False
# --- Supported Models ---
SUPPORTED_MODELS = {
"llama3-8b-instruct": "meta/meta-llama-3-8b-instruct",
"claude-4.5-haiku": "anthropic/claude-4.5-haiku",
"claude-4.5-sonnet": "anthropic/claude-4.5-sonnet",
"llava-13b": "yorickvp/llava-13b:e272157381e2a3bf12df3a8edd1f38d1dbd736bbb7437277c8b34175f8fce358"
}
# --- Core Logic ---
def prepare_replicate_input(request: OpenAIChatCompletionRequest) -> Dict[str, Any]:
"""
Formats the input for Replicate's API, flattening the message history into a
single 'prompt' string and handling images separately.
"""
payload = {}
prompt_parts = []
system_prompt = None
image_input = None
for msg in request.messages:
if msg.role == "system":
system_prompt = str(msg.content)
elif msg.role == "assistant":
prompt_parts.append(f"Assistant: {msg.content}")
elif msg.role == "user":
user_text_content = ""
if isinstance(msg.content, list):
for item in msg.content:
if item.get("type") == "text":
user_text_content += item.get("text", "")
elif item.get("type") == "image_url":
image_url_data = item.get("image_url", {})
image_input = image_url_data.get("url")
else:
user_text_content = str(msg.content)
prompt_parts.append(f"User: {user_text_content}")
prompt_parts.append("Assistant:")
payload["prompt"] = "\n\n".join(prompt_parts)
if system_prompt:
payload["system_prompt"] = system_prompt
if image_input:
payload["image"] = image_input
if request.max_tokens: payload["max_new_tokens"] = request.max_tokens
if request.temperature: payload["temperature"] = request.temperature
if request.top_p: payload["top_p"] = request.top_p
return payload
async def stream_replicate_sse(replicate_model_id: str, input_payload: dict):
"""Handles the full streaming lifecycle with correct whitespace preservation."""
url = f"https://api.replicate.com/v1/models/{replicate_model_id}/predictions"
headers = {"Authorization": f"Bearer {REPLICATE_API_TOKEN}", "Content-Type": "application/json"}
async with httpx.AsyncClient(timeout=60.0) as client:
try:
response = await client.post(url, headers=headers, json={"input": input_payload, "stream": True})
response.raise_for_status()
prediction = response.json()
stream_url = prediction.get("urls", {}).get("stream")
prediction_id = prediction.get("id", "stream-unknown")
if not stream_url:
yield json.dumps({'error': {'message': 'Model did not return a stream URL.'}})
return
except httpx.HTTPStatusError as e:
error_details = e.response.text
try:
error_json = e.response.json()
error_details = error_json.get("detail", error_details)
except json.JSONDecodeError: pass
yield json.dumps({'error': {'message': f'Upstream Error: {error_details}', 'type': 'replicate_error'}})
return
try:
async with client.stream("GET", stream_url, headers={"Accept": "text/event-stream"}, timeout=None) as sse:
current_event = None
async for line in sse.aiter_lines():
if not line: # Skip empty lines
continue
if line.startswith("event:"):
current_event = line[len("event:"):].strip()
elif line.startswith("data:"):
# FIXED: Preserve all whitespace including leading/trailing spaces
raw_data = line[5:] # Remove "data:" prefix
# Remove only the optional single space after data: if present
# This is per SSE spec and preserves actual content spaces
if raw_data.startswith(" "):
data_content = raw_data[1:] # Remove the first space only
else:
data_content = raw_data
if current_event == "output":
if not data_content:
continue
content_token = ""
try:
# Handle JSON-encoded strings properly (including spaces)
content_token = json.loads(data_content)
except (json.JSONDecodeError, TypeError):
# Handle plain text tokens (preserve as-is)
content_token = data_content
# Create chunk with exact format you specified
chunk = {
"choices": [{
"delta": {"content": content_token},
"finish_reason": None,
"index": 0,
"logprobs": None,
"native_finish_reason": None
}],
"created": int(time.time()),
"id": f"gen-{int(time.time())}-{prediction_id[-12:]}", # Format like your example
"model": replicate_model_id,
"object": "chat.completion.chunk",
"provider": "Anthropic" if "anthropic" in replicate_model_id else "Replicate"
}
# FIXED: Yield only the JSON data, let EventSourceResponse handle the SSE formatting
yield json.dumps(chunk)
elif current_event == "done":
# Send usage chunk before done
usage_chunk = {
"choices": [{
"delta": {},
"finish_reason": None,
"index": 0,
"logprobs": None,
"native_finish_reason": None
}],
"created": int(time.time()),
"id": f"gen-{int(time.time())}-{prediction_id[-12:]}",
"model": replicate_model_id,
"object": "chat.completion.chunk",
"provider": "Anthropic" if "anthropic" in replicate_model_id else "Replicate",
"usage": {
"cache_discount": 0,
"completion_tokens": 0,
"completion_tokens_details": {"image_tokens": 0, "reasoning_tokens": 0},
"cost": 0,
"cost_details": {
"upstream_inference_completions_cost": 0,
"upstream_inference_cost": None,
"upstream_inference_prompt_cost": 0
},
"input_tokens": 0,
"is_byok": False,
"prompt_tokens": 0,
"prompt_tokens_details": {"audio_tokens": 0, "cached_tokens": 0},
"total_tokens": 0
}
}
yield json.dumps(usage_chunk)
# Send final chunk with stop reason
final_chunk = {
"choices": [{
"delta": {},
"finish_reason": "stop",
"index": 0,
"logprobs": None,
"native_finish_reason": "end_turn"
}],
"created": int(time.time()),
"id": f"gen-{int(time.time())}-{prediction_id[-12:]}",
"model": replicate_model_id,
"object": "chat.completion.chunk",
"provider": "Anthropic" if "anthropic" in replicate_model_id else "Replicate"
}
yield json.dumps(final_chunk)
break
except httpx.ReadTimeout:
yield json.dumps({'error': {'message': 'Stream timed out.', 'type': 'timeout_error'}})
return
# Send [DONE] event
yield "[DONE]"
# --- Endpoints ---
@app.get("/v1/models")
async def list_models():
return ModelList(data=[ModelCard(id=k) for k in SUPPORTED_MODELS.keys()])
@app.post("/v1/chat/completions")
async def create_chat_completion(request: OpenAIChatCompletionRequest):
if request.model not in SUPPORTED_MODELS:
raise HTTPException(status_code=404, detail=f"Model not found. Available models: {list(SUPPORTED_MODELS.keys())}")
replicate_input = prepare_replicate_input(request)
if request.stream:
return EventSourceResponse(stream_replicate_sse(SUPPORTED_MODELS[request.model], replicate_input), media_type="text/event-stream")
# Non-streaming fallback
url = f"https://api.replicate.com/v1/models/{SUPPORTED_MODELS[request.model]}/predictions"
headers = {"Authorization": f"Bearer {REPLICATE_API_TOKEN}", "Content-Type": "application/json", "Prefer": "wait=120"}
async with httpx.AsyncClient() as client:
try:
resp = await client.post(url, headers=headers, json={"input": replicate_input}, timeout=130.0)
resp.raise_for_status()
pred = resp.json()
output = "".join(pred.get("output", []))
return {
"id": pred.get("id"), "object": "chat.completion", "created": int(time.time()), "model": request.model,
"choices": [{"index": 0, "message": {"role": "assistant", "content": output}, "finish_reason": "stop"}],
"usage": {"prompt_tokens": 0, "completion_tokens": 0, "total_tokens": 0}
}
except httpx.HTTPStatusError as e:
raise HTTPException(status_code=e.response.status_code, detail=f"Error from Replicate API: {e.response.text}")