feat(dynamo-llm): Remove bring-your-own-engine

Earthfile

@@ -112,7 +112,7 @@ dynamo-build:
     COPY deploy/ deploy/
 
     ENV CARGO_TARGET_DIR=/workspace/target
-    RUN cargo build --release --locked --features llamacpp,python,cuda && \
+    RUN cargo build --release --locked --features llamacpp,cuda && \
         cargo doc --no-deps
 
     # Create symlinks for wheel building

container/Dockerfile.none

@@ -46,7 +46,7 @@ ARG CARGO_BUILD_JOBS
 
 ENV CARGO_TARGET_DIR=/workspace/target
 
-RUN cargo build --release --locked --features mistralrs,python && \
+RUN cargo build --release --locked && \
     cargo doc --no-deps && \
     cp target/release/dynamo-run /usr/local/bin && \
     cp target/release/http /usr/local/bin && \

container/Dockerfile.tensorrt_llm

@@ -275,7 +275,7 @@ COPY launch /workspace/launch
 COPY deploy/sdk /workspace/deploy/sdk
 
 # Build Rust crate binaries packaged with the wheel
-RUN cargo build --release --locked --features mistralrs,python \
+RUN cargo build --release --locked \
     -p dynamo-run \
     -p llmctl \
     # Multiple http named crates are present in dependencies, need to specify the path

container/Dockerfile.vllm

@@ -368,7 +368,7 @@ COPY launch /workspace/launch
 COPY deploy/sdk /workspace/deploy/sdk
 
 # Build Rust crate binaries packaged with the wheel
-RUN cargo build --release --locked --features mistralrs,python,dynamo-llm/block-manager \
+RUN cargo build --release --locked --features dynamo-llm/block-manager \
     -p dynamo-run \
     -p llmctl \
     # Multiple http named crates are present in dependencies, need to specify the path

launch/dynamo-run/Cargo.toml

@@ -29,7 +29,6 @@ description = "Dynamo Run CLI"
 default = ["mistralrs"]
 mistralrs = ["dep:dynamo-engine-mistralrs"]
 llamacpp = ["dep:dynamo-engine-llamacpp"]
-python = ["dep:dynamo-engine-python"]
 
 cuda = ["dynamo-engine-llamacpp/cuda", "dynamo-engine-mistralrs/cuda"]
 metal = ["dynamo-engine-llamacpp/metal", "dynamo-engine-mistralrs/metal"]
@@ -41,7 +40,6 @@ dynamo-runtime = { workspace = true }
 
 dynamo-engine-llamacpp = { path = "../../lib/engines/llamacpp", optional = true }
 dynamo-engine-mistralrs = { path = "../../lib/engines/mistralrs", optional = true }
-dynamo-engine-python = { path = "../../lib/engines/python", optional = true }
 
 anyhow = { workspace = true }
 async-stream = { workspace = true }

launch/dynamo-run/src/lib.rs

@@ -18,10 +18,6 @@ mod subprocess;
 
 const CHILD_STOP_TIMEOUT: Duration = Duration::from_secs(2);
 
-/// How we identify a python string endpoint
-#[cfg(feature = "python")]
-const PYTHON_STR_SCHEME: &str = "pystr:";
-
 /// Where we will attach the vllm/sglang subprocess. Invisible to users.
 pub const INTERNAL_ENDPOINT: &str = "dyn://dynamo.internal.worker";
 
@@ -270,18 +266,6 @@ pub async fn run(
                 model: Box::new(local_model),
             }
         }
-        #[cfg(feature = "python")]
-        Output::PythonStr(path_str) => {
-            let card = local_model.card();
-            let py_args = flags.as_vec(&path_str, &card.service_name);
-            let p = std::path::PathBuf::from(path_str);
-            let engine =
-                dynamo_engine_python::make_string_engine(cancel_token.clone(), &p, py_args).await?;
-            EngineConfig::StaticFull {
-                engine,
-                model: Box::new(local_model),
-            }
-        }
     };
 
     match in_opt {

launch/dynamo-run/src/main.rs

@@ -26,11 +26,11 @@ dynamo-run is a single binary that wires together the various inputs (http, text
 Example:
 - cargo build --features cuda -p dynamo-run
 - cd target/debug
-- ./dynamo-run Qwen/Qwen2.5-3B-Instruct
+- ./dynamo-run Qwen/Qwen3-0.6B
 - OR: ./dynamo-run /data/models/Llama-3.2-1B-Instruct-Q4_K_M.gguf
 "#;
 
-const USAGE: &str = "USAGE: dynamo-run in=[http|text|dyn://<path>|batch:<folder>] out=ENGINE_LIST|dyn://<path> [--http-port 8080] [--model-path <path>] [--model-name <served-model-name>] [--model-config <hf-repo>] [--tensor-parallel-size=1] [--context-length=N] [--kv-cache-block-size=16] [--num-nodes=1] [--node-rank=0] [--leader-addr=127.0.0.1:9876] [--base-gpu-id=0] [--extra-engine-args=args.json] [--router-mode random|round-robin|kv]";
+const USAGE: &str = "USAGE: dynamo-run in=[http|text|dyn://<path>|batch:<folder>] out=ENGINE_LIST|dyn [--http-port 8080] [--model-path <path>] [--model-name <served-model-name>] [--model-config <hf-repo>] [--tensor-parallel-size=1] [--context-length=N] [--kv-cache-block-size=16] [--num-nodes=1] [--node-rank=0] [--leader-addr=127.0.0.1:9876] [--base-gpu-id=0] [--extra-engine-args=args.json] [--router-mode random|round-robin|kv]";
 
 fn main() -> anyhow::Result<()> {
     // Set log level based on verbosity flag