Spaces:
Running
Running
File size: 1,640 Bytes
43ea069 332dd16 43ea069 332dd16 43ea069 58a3721 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 | """Silicon Flow provider using OpenAI-compatible API."""
from typing import Any
from config.settings import Settings
from core.anthropic import ReasoningReplayMode, build_base_request_body
from providers.base import ProviderConfig
from providers.defaults import SILICON_DEFAULT_BASE
from providers.openai_compat import OpenAIChatTransport
class SiliconProvider(OpenAIChatTransport):
"""Silicon Flow provider using OpenAI-compatible /chat/completions."""
def __init__(self, config: ProviderConfig, *, settings: Settings):
base_url = (config.base_url or SILICON_DEFAULT_BASE).rstrip("/")
if not base_url.endswith("/v1"):
base_url = base_url + "/v1"
# Silicon Flow has generous rate limits
super().__init__(
config,
provider_name="Silicon",
base_url=base_url,
api_key=config.api_key,
nim_rate_limit=300,
nim_max_concurrency=80,
)
self._settings = settings
def _build_request_body(
self, request: Any, thinking_enabled: bool | None = None
) -> dict:
thinking = self._is_thinking_enabled(request, thinking_enabled)
reasoning_replay = (
ReasoningReplayMode.REASONING_CONTENT
if thinking
else ReasoningReplayMode.DISABLED
)
body = build_base_request_body(request, reasoning_replay=reasoning_replay)
# Strip silicon/ prefix so the API gets the bare model ID
model = body.get("model", "")
if model.startswith("silicon/"):
body["model"] = model[len("silicon/") :]
return body
|