example(eagle3): disable torch.compile for Qwen3-0.6B CI train

The recipe compiles the eagle forward with torch.compile(mode="max-autotune")
(hf_eagle.py), whose exhaustive Triton bmm autotuning logs benign
'out of resource: triton_bmm' errors on the L40 (it recovers and picks a working
config, task SUCCEEDs) and adds compile overhead. Disable it via
eagle.eagle_use_torch_compile=false for the CI smoke runs — clean logs, faster.
Coverage note: this stops exercising the compiled path; revisit once the recipe
lets the compile mode be tuned down instead of forced to max-autotune.

Signed-off-by: Chenhan Yu <chenhany@nvidia.com>
(cherry picked from commit d4b28a1)

Author: ChenhanYu

tools/launcher/examples/Qwen/Qwen3-0.6B/hf_offline_eagle3.yaml

@@ -75,6 +75,10 @@ pipeline:
       - training.training_seq_len=1024
       - training.disable_tqdm=true
       - training.ar_validate_steps=500000
+      # CI: disable torch.compile (recipe default mode="max-autotune") — its
+      # exhaustive Triton bmm autotuning floods the log with benign "out of
+      # resource" errors on the L40 and adds compile overhead for no CI benefit.
+      - eagle.eagle_use_torch_compile=false
     slurm_config:
       _factory_: "slurm_factory"
       nodes: 1

tools/launcher/examples/Qwen/Qwen3-0.6B/hf_online_eagle3.yaml

@@ -44,6 +44,10 @@ pipeline:
       - training.disable_tqdm=true
       - training.ar_validate_steps=500000
       - training.num_train_epochs=1
+      # CI: disable torch.compile (recipe default mode="max-autotune") — its
+      # exhaustive Triton bmm autotuning floods the log with benign "out of
+      # resource" errors on the L40 and adds compile overhead for no CI benefit.
+      - eagle.eagle_use_torch_compile=false
     slurm_config:
       _factory_: "slurm_factory"
       nodes: 1