A/B of the live STT seat (fv-ml1 GPU 0, sherpa-onnx int8 v3) against nvidia/parakeet-unified-en-0.6b, measured on GPU 3 with the seat's own image, k2-fsa's published unified int8 export, fp32/fp16 exports made with k2-fsa's recipe, v2 int8, and NeMo 3.0.0 (fp32, bf16 autocast, bf16 weights). - Seat int8 graph runs on one CPU thread (cpu/wall 1.00, GPU 2-9%). - unified-en under NeMo: -121/-234/-530 ms vs the seat at 1-3/3-8/8-20 s (paired, n=120/bin; floor <=6 ms; +50 ms positive control reads +52-54). - unified-en WER lower in every runtime: -0.7 pp clean, -1.5 pp other, -3.2 to -4.4 pp AMI (paired CIs exclude 0). - Seat defects found: hard 400 s input ceiling (HTTP 500), truncation after a quiet 1.5 s pause, and severe long-window dropouts (int8 v3 only). - B-bf16w needs +0.8 to +1.5 GB over the seat's 1,690 MiB on GPU 0. Raw requests, hypotheses, manifests and the full harness under services/parakeet-ab-2026-09-30/. No deploy; live seat untouched apart from 240 light test requests.
32 lines
3.0 KiB
Plaintext
32 lines
3.0 KiB
Plaintext
a5e223392c90e75f8144cdb5eb95af7625db389e39edef2bd1a9c872b3298fe6 unified-int8/decoder.int8.onnx
|
|
6716910b7a0833997fec7a410494c995d70124001a0e9b66d6370d6aced577e0 unified-int8/encoder.int8.onnx
|
|
869f43f7d24595c55581ad3bf249a935fb8a71389fbdaa7504b9f46f93140f8a unified-int8/joiner.int8.onnx
|
|
dc0b4584ab2e4ddbf888425c076c61b736e7356a015250db7d307e6f1a8188ff unified-int8/tokens.txt
|
|
b6bb64963457237b900e496ee9994b59294526439fbcc1fecf705b31a15c6b4e v2-int8/decoder.int8.onnx
|
|
a32b12d17bbbc309d0686fbbcc2987b5e9b8333a7da83fa6b089f0a2acd651ab v2-int8/encoder.int8.onnx
|
|
7946164367946e7f9f29a122407c3252b680dbae9a51343eb2488d057c3c43d2 v2-int8/joiner.int8.onnx
|
|
ec182b70dd42113aff6c5372c75cac58c952443eb22322f57bbd7f53977d497d v2-int8/tokens.txt
|
|
179e50c43d1a9de79c8a24149a2f9bac6eb5981823f2a2ed88d655b24248db4e /tank/parakeet/models/decoder.int8.onnx
|
|
acfc2b4456377e15d04f0243af540b7fe7c992f8d898d751cf134c3a55fd2247 /tank/parakeet/models/encoder.int8.onnx
|
|
3164c13fc2821009440d20fcb5fdc78bff28b4db2f8d0f0b329101719c0948b3 /tank/parakeet/models/joiner.int8.onnx
|
|
d58544679ea4bc6ac563d1f545eb7d474bd6cfa467f0a6e2c1dc1c7d37e3c35d /tank/parakeet/models/tokens.txt
|
|
ec23ed9150c8fde49072c3e2d61678ab903dbcef389d658db833420cbc1da35b /tank/aimodels/huggingface/hub/models--nvidia--parakeet-unified-en-0.6b/snapshots/fe53cd885760c96b6a5f51a0bfd362cb4584a98b/parakeet-unified-en-0.6b.nemo
|
|
3ebb54deb0f0d935e5da6d3e66295858269a7c688bb331a35baae57ba7c02e3e unified-f32/decoder.onnx
|
|
39785470a9f996da25323f082170fb2b601bcf9b7f602820a22848850cd51da3 unified-f32/encoder.onnx
|
|
d1559afb5a1f5a601f707b30b52df97275385ee19e2bcfca99aeb7a9e570ebaf unified-f32/encoder.weights
|
|
8432e0e591ca092e551bd55cc80f046b606a9916ae3530dcedfd565aa5ca32dd unified-f32/joiner.onnx
|
|
dc0b4584ab2e4ddbf888425c076c61b736e7356a015250db7d307e6f1a8188ff unified-f32/tokens.txt
|
|
d593cdb0e571f5a457ec2219af9968cbf6b0e8198e8f7839b40a8754593bf68c v3-f32/decoder.onnx
|
|
ccb332d5845fee5b5dacdfe88631bff900ca06596c02d07d5548f991089be141 v3-f32/encoder.onnx
|
|
9a22d372c51455c34f13405da2520baefb7125bd16981397561423ed32d24f36 v3-f32/encoder.weights
|
|
33c2948a6635f6101c9915d9d464005039c0649ea61866103832d6b2d4110728 v3-f32/joiner.onnx
|
|
d58544679ea4bc6ac563d1f545eb7d474bd6cfa467f0a6e2c1dc1c7d37e3c35d v3-f32/tokens.txt
|
|
33a1a51fab010a42f8454200ec8f90de827a54401b9930f7cbc280c69e6401aa models/unified-f16/decoder.fp16.onnx
|
|
e460b3b4fbbf4252f47151f588de9896fe05712db1f3ab26a43ea50ffbf881e1 models/unified-f16/encoder.fp16.onnx
|
|
4187dc36791068b1a0546c04ff6efedfe6f3efa7652ec210fcd26b2665874f27 models/unified-f16/joiner.fp16.onnx
|
|
dc0b4584ab2e4ddbf888425c076c61b736e7356a015250db7d307e6f1a8188ff models/unified-f16/tokens.txt
|
|
6414358d22ad950e448dd8e822d93bcbdea82cc268cdd9cb50d803448dd63ff5 models/v3-f16/decoder.fp16.onnx
|
|
2a355afa863caa940027518907146c8c8eedc2e56fb073eefbab2c7dd05507b9 models/v3-f16/encoder.fp16.onnx
|
|
8653c5709d2462d019137a4ecf8093a7391467dac6b7a6856729f6c99ff771bc models/v3-f16/joiner.fp16.onnx
|
|
d58544679ea4bc6ac563d1f545eb7d474bd6cfa467f0a6e2c1dc1c7d37e3c35d models/v3-f16/tokens.txt
|