Delete egs/ami/ASR/xlsr_transducer/inference_results/log-inference-ihm-2026-01-26-05-02-37

egs/ami/ASR/xlsr_transducer/inference_results/log-inference-ihm-2026-01-26-05-02-37

@@ -1,32 +0,0 @@
-2026-01-26 05:02:37,849 INFO [inference.py:410] ================================================================================
-2026-01-26 05:02:37,849 INFO [inference.py:411] XLSR-Transducer Inference on AMI
-2026-01-26 05:02:37,850 INFO [inference.py:412] ================================================================================
-2026-01-26 05:02:37,850 INFO [inference.py:413] Experiment dir: xlsr_transducer/exp_16gb_scd
-2026-01-26 05:02:37,850 INFO [inference.py:414] Output dir: xlsr_transducer/exp_16gb_scd/inference_results
-2026-01-26 05:02:37,850 INFO [inference.py:415] Test set: ihm
-2026-01-26 05:02:37,850 INFO [inference.py:416] Decoding method: modified_beam_search
-2026-01-26 05:02:37,850 INFO [inference.py:422] Device: cpu
-2026-01-26 05:02:37,850 INFO [inference.py:425] Loading BPE model from data/lang_bpe_500_scd
-2026-01-26 05:02:37,851 INFO [inference.py:433] Vocabulary size: 500
-2026-01-26 05:02:37,851 INFO [inference.py:434] Blank ID: 0
-2026-01-26 05:02:37,851 INFO [inference.py:437] Creating model
-2026-01-26 05:02:39,443 INFO [inference.py:459] Loading checkpoint: xlsr_transducer/exp_16gb_scd/epoch-15.pt
-2026-01-26 05:02:39,444 INFO [checkpoint.py:111] Loading checkpoint from xlsr_transducer/exp_16gb_scd/epoch-15.pt
-2026-01-26 05:02:44,138 INFO [inference.py:473] Number of model parameters: 317,511,772
-2026-01-26 05:02:44,139 INFO [inference.py:476] Loading test data
-2026-01-26 05:02:44,139 INFO [asr_datamodule.py:448] About to get AMI IHM test cuts with [SCD] tokens
-2026-01-26 05:02:45,310 INFO [inference.py:487] Number of test utterances: 6676
-2026-01-26 05:02:45,310 INFO [inference.py:490] Starting inference...
-2026-01-26 05:02:46,398 INFO [inference.py:309] Audio shape: torch.Size([6, 246400]), dtype: torch.float32
-2026-01-26 05:02:46,400 INFO [inference.py:310] Audio range: [-0.090, 0.104]
-2026-01-26 05:02:46,404 INFO [inference.py:311] Audio lengths: tensor([246400, 244799, 238079, 228000, 224000, 222880], dtype=torch.int32)
-2026-01-26 05:02:55,240 INFO [inference.py:332] Encoder out shape: torch.Size([6, 769, 1024])
-2026-01-26 05:02:55,241 INFO [inference.py:333] Encoder out lens: tensor([769, 764, 743, 712, 699, 696])
-2026-01-26 05:02:55,298 INFO [inference.py:334] Encoder out range: [-4.703, 6.664]
-2026-01-26 05:02:55,596 INFO [inference.py:344] Number of hypotheses: 6
-2026-01-26 05:02:55,596 INFO [inference.py:346] First hypothesis: []
-2026-01-26 05:02:55,605 INFO [inference.py:309] Audio shape: torch.Size([23, 67200]), dtype: torch.float32
-2026-01-26 05:02:55,606 INFO [inference.py:310] Audio range: [-0.401, 0.443]
-2026-01-26 05:02:55,607 INFO [inference.py:311] Audio lengths: tensor([67200, 66559, 66400, 64159, 63680, 63040, 61120, 60960, 59519, 59040,
-        58560, 55840, 54559, 52480, 52160, 50559, 50400, 50240, 48480, 47840,
-        47520, 44639, 44000], dtype=torch.int32)