User contributions
(newest | oldest) View (newer 50 | older 50) (20 | 50 | 100 | 250 | 500)
- 21:29, 30 October 2019 (diff | hist) . . (-1) . . m TensorRT/Int8CFAQ (→Why the inference result is not stable even with the same calibration dataset and same input image?) (current)
- 21:27, 30 October 2019 (diff | hist) . . (+13) . . m TensorRT/Int8CFAQ (→Why the inference result is not stable even with the same calibration dataset and same input image?)
- 21:26, 30 October 2019 (diff | hist) . . (+46) . . m TensorRT/Int8CFAQ (→Why the inference result is not stable even with the same calibration dataset and same input image?)
- 21:25, 30 October 2019 (diff | hist) . . (-1) . . m TensorRT/Int8CFAQ (→Why the inference result is not stable even with the same calibration dataset and same input image?)
- 21:24, 30 October 2019 (diff | hist) . . (-285) . . m TensorRT/Int8CFAQ (→Why the inference result is not stable even with the same calibration dataset and same input image?)
- 21:13, 30 October 2019 (diff | hist) . . (+2,525) . . TensorRT/Int8CFAQ (→How to interpret the value of calibration table?)
- 19:59, 29 October 2019 (diff | hist) . . (+15) . . m TensorRT/How2Debug (→How to know which layers are running on what precision (INT8/FP16/FP32)?)
- 21:57, 14 October 2019 (diff | hist) . . (+173) . . m TensorRT/How2Debug (→How to know which layers are running on what precision (INT8/FP16/FP32)?)
- 21:54, 14 October 2019 (diff | hist) . . (+16) . . m TensorRT/How2Debug (→How to know which layers are running on what precision (INT8/FP16/FP32)?)
- 21:51, 14 October 2019 (diff | hist) . . (-3) . . m TensorRT/How2Debug (→How to know which layers are running on what precision (INT8/FP16/FP32)?)
- 21:50, 14 October 2019 (diff | hist) . . (+15) . . m TensorRT/How2Debug (→How to know which layers are running on what precision (INT8/FP16/FP32)?)
- 21:50, 14 October 2019 (diff | hist) . . (+4,528) . . m TensorRT/How2Debug
- 20:54, 14 October 2019 (diff | hist) . . (-19) . . m TensorRT (→Introduction)
- 19:03, 14 October 2019 (diff | hist) . . (+4) . . m TensorRT (→Introduction)
- 18:03, 14 October 2019 (diff | hist) . . (+45) . . m TensorRT/CommonFAQ (→Why TRT has different definition for axis parameters as compared with the other frameworks, like Caffe or TensorFlow?)
- 00:35, 14 October 2019 (diff | hist) . . (-107) . . m TensorRT (→Known issue)
- 00:24, 14 October 2019 (diff | hist) . . (-36) . . m TensorRT/Int8CFAQ
- 00:18, 14 October 2019 (diff | hist) . . (+15) . . m TensorRT
- 00:15, 14 October 2019 (diff | hist) . . (+1) . . m TensorRT/AccuracyIssues (current)
- 00:15, 14 October 2019 (diff | hist) . . (-3,313) . . m TensorRT/AccuracyIssues (→How to fix INT8 accuracy issue?) (Tag: Replaced)
- 00:14, 14 October 2019 (diff | hist) . . (-2,310) . . m TensorRT/AccuracyIssues (→How to fix FP16 accuracy issue?)
- 00:13, 14 October 2019 (diff | hist) . . (+2,369) . . N TensorRT/FP16 Accuracy (Created page with "The following is the data range of FP32, FP16 and INT8, {| class="wikitable" |- ! !! '''Dynamic Range''' !! '''Min Positive Value''' |- | FP32 || -3.4 x 10<sup>38</sup> ~ +3....") (current)
- 19:10, 10 October 2019 (diff | hist) . . (-19) . . m TensorRT/INT8 Accuracy (→QAT (Quantizion-Aware Training)) (current)
- 21:54, 8 October 2019 (diff | hist) . . (+1) . . m TensorRT/INT8 Accuracy (→Exercise with sampleMNIST)
- 21:51, 8 October 2019 (diff | hist) . . (+3) . . m TensorRT/INT8 Accuracy (→Identify INT8 quantization issue)
- 01:20, 29 September 2019 (diff | hist) . . (+6) . . m TensorRT/INT8 Accuracy (→Mixed quantization)
- 01:20, 29 September 2019 (diff | hist) . . (+12) . . m TensorRT/INT8 Accuracy (→How to fix INT8 quantization issue)
- 01:19, 29 September 2019 (diff | hist) . . (+6) . . m TensorRT/INT8 Accuracy (→Exercise with sampleMNIST)
- 01:17, 29 September 2019 (diff | hist) . . (0) . . File:Samplemnist accuracy int8.zip (Nfeng uploaded a new version of File:Samplemnist accuracy int8.zip) (current)
- 00:40, 29 September 2019 (diff | hist) . . (0) . . File:Samplemnist accuracy int8.zip (Nfeng uploaded a new version of File:Samplemnist accuracy int8.zip)
- 00:39, 29 September 2019 (diff | hist) . . (0) . . N File:Samplefasterrcnn int8 accuracy.zip (current)
- 17:10, 28 September 2019 (diff | hist) . . (+19) . . m TensorRT/INT8 Accuracy (→Ensure you have tried BatchNorm layer)
- 17:10, 28 September 2019 (diff | hist) . . (-2) . . m TensorRT/INT8 Accuracy (→Ensure you have tried BatchNorm layer)
- 17:08, 28 September 2019 (diff | hist) . . (0) . . m TensorRT/INT8 Accuracy (→Ensure the calibration dataset is diverse and representative)
- 17:08, 28 September 2019 (diff | hist) . . (-2) . . m TensorRT/INT8 Accuracy (→Ensure the preprocessing steps within your calibrator is identical as FP32 inference)
- 17:04, 28 September 2019 (diff | hist) . . (+45) . . m TensorRT/INT8 Accuracy (→How to triage INT8 accuracy issue)
- 17:01, 28 September 2019 (diff | hist) . . (+6) . . m TensorRT/INT8 Accuracy (→Approach to figure out the loss layers)
- 16:59, 28 September 2019 (diff | hist) . . (-6) . . m TensorRT/INT8 Accuracy (→How to debug INT8 quantization issue)
- 16:59, 28 September 2019 (diff | hist) . . (-1) . . m TensorRT/INT8 Accuracy (→Exercise with sampleMNIST)
- 16:58, 28 September 2019 (diff | hist) . . (+7) . . m TensorRT/INT8 Accuracy (→Exercise with sampleMNIST)
- 16:58, 28 September 2019 (diff | hist) . . (+102) . . m TensorRT/INT8 Accuracy (→Exercise with sampleMNIST)
- 16:56, 28 September 2019 (diff | hist) . . (+17) . . m TensorRT/INT8 Accuracy (→Exercise with sampleMNIST)
- 16:54, 28 September 2019 (diff | hist) . . (+37) . . m TensorRT/INT8 Accuracy (→Step by step guide to apply above approach into your own application)
- 16:52, 28 September 2019 (diff | hist) . . (+4) . . m TensorRT/INT8 Accuracy (→Add two specific options)
- 16:52, 28 September 2019 (diff | hist) . . (+4) . . m TensorRT/INT8 Accuracy (→Handle the two options)
- 16:51, 28 September 2019 (diff | hist) . . (+3) . . m TensorRT/INT8 Accuracy (→Dump network output into files)
- 16:51, 28 September 2019 (diff | hist) . . (+6) . . m TensorRT/INT8 Accuracy (→How to fix INT8 quantization issue)
- 16:49, 28 September 2019 (diff | hist) . . (0) . . m TensorRT/INT8 Accuracy (→Mixed precision)
- 16:47, 28 September 2019 (diff | hist) . . (-2) . . m TensorRT/INT8 Accuracy (→Mixed quantization)
- 16:46, 28 September 2019 (diff | hist) . . (+5) . . m TensorRT/INT8 Accuracy (→QAT (Quantizion-Aware Training))
(newest | oldest) View (newer 50 | older 50) (20 | 50 | 100 | 250 | 500)