Skip to content

[pt1][quant] Add qint8 type (int8_t)#19984

Closed
jerryzh168 wants to merge 37 commits intomasterfrom
export-D15150715
Closed

[pt1][quant] Add qint8 type (int8_t)#19984
jerryzh168 wants to merge 37 commits intomasterfrom
export-D15150715

Conversation

@jerryzh168
Copy link
Contributor

@jerryzh168 jerryzh168 commented Apr 30, 2019

Stack:
    :white_circle:  #20672 Modify the Insert quant-dequant test cases to look for q-dq pattern  💛
    :white_circle:  #20107 [pt1][quant] Add dequantize_linear for JIT pass  💚
    :black_circle:  #19984 [pt1][quant] Add qint8 type (int8_t)  💚

Add qint8 for QTensor, with underlying type of int8_t

Differential Revision: D15150715

Differential Revision: D15094174
Differential Version: 80838655
Differential Revision: D15094174
Differential Version: 80839141
Differential Revision: D15094174
Differential Version: 80909406
Differential Revision: D15094174
Differential Version: 80918409
Differential Revision: D15137838
Differential Version: 80919404
Differential Revision: D15137838
Differential Version: 80940601
Differential Revision: D15094174
Differential Version: 80977784
Differential Revision: D15094174
Differential Version: 80996728
Differential Revision: D15137838
Differential Version: 80996800
Differential Revision: D15150715
Differential Version: 80996764
Differential Revision: D15094174
Differential Version: 81007920
Differential Revision: D15094174
Differential Version: 81099685
Differential Revision: D15137838
Differential Version: 81109398
Differential Revision: D15150715
Differential Version: 81113084
jerryzh168 added 11 commits May 3, 2019 12:11
Differential Revision: D15094174
Differential Version: 81173037
Differential Revision: D15094174
Differential Version: 81239113
Differential Revision: D15094174
Differential Version: 81253515
Differential Revision: D15094174
Differential Version: 81263921
Differential Revision: D15094174
Differential Version: 81319070
Differential Revision: D15094174
Differential Version: 81328044
Differential Revision: D15094174
Differential Version: 81380836
Differential Revision: D15094174
Differential Version: 81411153
Differential Revision: D15094174
Differential Version: 81423512
Differential Revision: D15094174
Differential Version: 81476388
Differential Revision: D15094174
Differential Version: 81478101
Differential Revision: D15094174
Differential Version: 81670049
Differential Revision: D15137838
Differential Version: 81699511
Differential Revision: D15150715
Differential Version: 81699751
Differential Revision: D15137838
Differential Version: 82009220
Differential Revision: D15150715
Differential Version: 82024560
Differential Revision: D15150715
Differential Version: 82077846
@z-a-f z-a-f changed the base branch from export-D15137838 to master May 17, 2019 16:51
Differential Revision: D15150715
Differential Version: 82142681
@pytorchbot pytorchbot added module: build Build system issues module: cuda Related to torch.cuda, and CUDA support in general oncall: distributed Add this issue/PR to distributed oncall triage queue module: nn Related to torch.nn labels May 17, 2019
@facebook-github-bot
Copy link
Contributor

This pull request has been merged in 85fad05.

zdevito pushed a commit to zdevito/ATen that referenced this pull request May 18, 2019
Summary:
Pull Request resolved: pytorch/pytorch#19984

Add qint8 for QTensor, with underlying type of int8_t

Reviewed By: jianyuh

Differential Revision: D15150715

fbshipit-source-id: 57580f599d46f9323af5ce462dbbc464b25e40d7
@ezyang ezyang deleted the export-D15150715 branch May 30, 2019 16:02
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

module: build Build system issues module: cuda Related to torch.cuda, and CUDA support in general module: internals Related to internal abstractions in c10 and ATen module: nn Related to torch.nn oncall: distributed Add this issue/PR to distributed oncall triage queue oncall: quantization Quantization support in PyTorch

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants