◂ BACK TO LEADERBOARD
▚ MODEL AUDIT // 6-bit MLX

// Qwen 3.6 35B-A3B

CRIT — Critical Bugs 50/100 tests: CRASH
tok/sec68.99
tokens13310
TTFT1.07s

▮ PILLAR BREAKDOWN

Ownership / Types
8/20
Concurrency / Races
13/20
Error Handling
12/20
Cancellation / Shutdown
11/20
Test Integrity
6/20

✓ WENT RIGHT

  • No pillar reached 16+ — no standout strengths.

✗ WENT WRONG

  • Concurrency / Races (13/20)
  • Error Handling (12/20)
  • Cancellation / Shutdown (11/20)
  • Ownership / Types (8/20)
  • Test Integrity (6/20)

▮ CRITICAL BUGS

▮ RECOMMENDED USE

NOT usable for Rust as-is — 7 compile errors. Same model that scored 82 on LFU now: TTS 49, Rust 50. The profile is now clear: Qwen 6-bit is strong on single-file Python data-structure work but repeatedly ships non-compiling/non-parsing code on multi-task async + typed-language prompts. Keep it on Python ACID/data-structure tasks; do NOT offload Rust or async-pipeline work to it.

▮ REFACTORED PATCH

// FIX 1 (the API hallucination): tokio mpsc has no bounded().
//   let (item_tx, item_rx) = mpsc::bounded(32);
let (item_tx, item_rx) = mpsc::channel(32);

// FIX 2 (type mismatch):
tokio::time::sleep(Duration::from_millis(20 + (id as u64 % 30))).await;

// FIX 3 (ownership in shutdown): store JoinHandles in Option + take them,
// and make shutdown take &mut self (or hold senders in Option):
struct WatcherEntry { status: WatcherStatus, consecutive_failures: u32, join_handle: Option<JoinHandle<()>> }
// in shutdown: let handles: Vec<_> = inner.watchers.values_mut().map(|e| e.join_handle.take()).flatten().collect();
//   drop(self.item_tx.take()) etc. with Option<Sender> fields.

// FIX 4 (remove_watcher must actually stop the task): either send on a per-watcher
// oneshot/CancellationToken, or broadcast shutdown to that watcher's sub-channel.
// Simplest: give each watcher a CancellationToken; remove_watcher cancels it, then awaits the handle.

// FIX 5 (test isolation): inject the fetch fn into watcher_loop as a parameter so tests
// can pass a failing mock; drop the dead global flag.
// FIX 6: add `fn main() { ... }` or make it `cargo test`-only and document that.