mirror of
https://github.com/pchuan98/codex.git
synced 2026-07-01 00:31:56 +08:00
ollama: default to Responses API for built-ins (#8798)
This is an alternate PR to solving the same problem as <https://github.com/openai/codex/pull/8227>. In this PR, when Ollama is used via `--oss` (or via `model_provider = "ollama"`), we default it to use the Responses format. At runtime, we do an Ollama version check, and if the version is older than when Responses support was added to Ollama, we print out a warning. Because there's no way of configuring the wire api for a built-in provider, we temporarily add a new `oss_provider`/`model_provider` called `"ollama-chat"` that will force the chat format. Once the `"chat"` format is fully removed (see <https://github.com/openai/codex/discussions/7782>), `ollama-chat` can be removed as well --------- Co-authored-by: Eric Traut <etraut@openai.com> Co-authored-by: Michael Bolin <mbolin@openai.com>
This commit is contained in:
co-authored by
Eric Traut
Michael Bolin
parent
2d56519ecd
commit
fe03320791
@@ -1,18 +1,52 @@
|
||||
//! OSS provider utilities shared between TUI and exec.
|
||||
|
||||
use codex_core::LMSTUDIO_OSS_PROVIDER_ID;
|
||||
use codex_core::OLLAMA_CHAT_PROVIDER_ID;
|
||||
use codex_core::OLLAMA_OSS_PROVIDER_ID;
|
||||
use codex_core::WireApi;
|
||||
use codex_core::config::Config;
|
||||
use codex_core::protocol::DeprecationNoticeEvent;
|
||||
use std::io;
|
||||
|
||||
/// Returns the default model for a given OSS provider.
|
||||
pub fn get_default_model_for_oss_provider(provider_id: &str) -> Option<&'static str> {
|
||||
match provider_id {
|
||||
LMSTUDIO_OSS_PROVIDER_ID => Some(codex_lmstudio::DEFAULT_OSS_MODEL),
|
||||
OLLAMA_OSS_PROVIDER_ID => Some(codex_ollama::DEFAULT_OSS_MODEL),
|
||||
OLLAMA_OSS_PROVIDER_ID | OLLAMA_CHAT_PROVIDER_ID => Some(codex_ollama::DEFAULT_OSS_MODEL),
|
||||
_ => None,
|
||||
}
|
||||
}
|
||||
|
||||
/// Returns a deprecation notice if Ollama doesn't support the responses wire API.
|
||||
pub async fn ollama_chat_deprecation_notice(
|
||||
config: &Config,
|
||||
) -> io::Result<Option<DeprecationNoticeEvent>> {
|
||||
if config.model_provider_id != OLLAMA_OSS_PROVIDER_ID
|
||||
|| config.model_provider.wire_api != WireApi::Responses
|
||||
{
|
||||
return Ok(None);
|
||||
}
|
||||
|
||||
if let Some(detection) = codex_ollama::detect_wire_api(&config.model_provider).await?
|
||||
&& detection.wire_api == WireApi::Chat
|
||||
{
|
||||
let version_suffix = detection
|
||||
.version
|
||||
.as_ref()
|
||||
.map(|version| format!(" (version {version})"))
|
||||
.unwrap_or_default();
|
||||
let summary = format!(
|
||||
"Your Ollama server{version_suffix} doesn't support the Responses API. Either update Ollama or set `oss_provider = \"{OLLAMA_CHAT_PROVIDER_ID}\"` (or `model_provider = \"{OLLAMA_CHAT_PROVIDER_ID}\"`) in your config.toml to use the \"chat\" wire API. Support for the \"chat\" wire API is deprecated and will soon be removed."
|
||||
);
|
||||
return Ok(Some(DeprecationNoticeEvent {
|
||||
summary,
|
||||
details: None,
|
||||
}));
|
||||
}
|
||||
|
||||
Ok(None)
|
||||
}
|
||||
|
||||
/// Ensures the specified OSS provider is ready (models downloaded, service reachable).
|
||||
pub async fn ensure_oss_provider_ready(
|
||||
provider_id: &str,
|
||||
@@ -24,7 +58,7 @@ pub async fn ensure_oss_provider_ready(
|
||||
.await
|
||||
.map_err(|e| std::io::Error::other(format!("OSS setup failed: {e}")))?;
|
||||
}
|
||||
OLLAMA_OSS_PROVIDER_ID => {
|
||||
OLLAMA_OSS_PROVIDER_ID | OLLAMA_CHAT_PROVIDER_ID => {
|
||||
codex_ollama::ensure_oss_ready(config)
|
||||
.await
|
||||
.map_err(|e| std::io::Error::other(format!("OSS setup failed: {e}")))?;
|
||||
|
||||
Reference in New Issue
Block a user