Blocking waiting for file lock on build directory
   Compiling ml v1.0.0 (/home/jgrusewski/Work/foxhunt/ml)
warning: unused import: `Var`
  --> ml/src/memory_optimization/qat.rs:37:42
   |
37 | use candle_core::{DType, Device, Tensor, Var};
   |                                          ^^^
   |
   = note: `#[warn(unused_imports)]` on by default

warning: unused import: `candle_nn::VarMap`
  --> ml/src/memory_optimization/qat.rs:38:5
   |
38 | use candle_nn::VarMap;
   |     ^^^^^^^^^^^^^^^^^

warning: unused import: `TFTConfig`
  --> ml/src/tft/qat_tft.rs:45:54
   |
45 | use crate::tft::{QuantizedTemporalFusionTransformer, TFTConfig, TemporalFusionTransformer};
   |                                                      ^^^^^^^^^

warning: unused import: `DType`
  --> ml/src/tft/qat_tft.rs:47:19
   |
47 | use candle_core::{DType, Device, Tensor};
   |                   ^^^^^

warning: unused import: `DType`
  --> ml/src/tft/temporal_attention.rs:18:19
   |
18 | use candle_core::{DType, Device, Module, Tensor};
   |                   ^^^^^

warning: unused variable: `opt`
   --> ml/src/trainers/tft.rs:957:37
    |
957 |                 if let Some(ref mut opt) = self.optimizer {
    |                                     ^^^ help: if this is intentional, prefix it with an underscore: `_opt`
    |
    = note: `#[warn(unused_variables)]` on by default

warning: type does not implement `std::fmt::Debug`; consider adding `#[derive(Debug)]` or a manual implementation
   --> ml/src/memory_optimization/qat.rs:231:1
    |
231 | / pub struct FakeQuantize {
232 | |     config: QATConfig,
233 | |     device: Device,
...   |
248 | |     training: bool,
249 | | }
    | |_^
    |
note: the lint level is defined here
   --> ml/src/lib.rs:40:9
    |
40  | #![warn(missing_debug_implementations)]
    |         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

warning: `ml` (lib) generated 7 warnings (5 duplicates)
warning: `ml` (lib) generated 7 warnings (2 duplicates) (run `cargo fix --lib -p ml` to apply 5 suggestions)
    Finished `bench` profile [optimized] target(s) in 5m 50s
     Running benches/inference_bench.rs (/home/jgrusewski/Work/foxhunt/target/release/deps/inference_bench-fd61fe21a0ac54e7)

running 4 tests
test performance_tests::test_batch_inference_throughput ... ignored
test performance_tests::test_ensemble_inference_latency ... ignored
test performance_tests::test_feature_preparation_latency ... ignored
test performance_tests::test_single_model_latency ... ignored

test result: ok. 0 passed; 0 failed; 4 ignored; 0 measured; 0 filtered out; finished in 0.00s