Commits

Branch selector

 main

User selector

supriyar

Datepicker

All time

Commit History

Commits on Sep 23, 2022

• Update PT maintainers list for AO (pytorch#85125)

  Show description for 18685b7
  

  supriyar

  authored and

  pytorchmergebot

  committed
  18685b7

  Copy full SHA for 18685b7

Commits on Dec 14, 2021

• [quant][docs] quantized model save/load instructions (pytorch#69789)

  Show description for b1ef56d
  

  supriyar

  authored and

  facebook-github-bot

  committed
  b1ef56d

  Copy full SHA for b1ef56d

Commits on Oct 28, 2021

• [quant][fx] add pass to duplicate dequant nodes with multi use (pytorch#67118)

  Show description for 1cfdb6f
  

  supriyar

  authored and

  facebook-github-bot

  committed
  1cfdb6f

  Copy full SHA for 1cfdb6f

Commits on Oct 23, 2021

• [quant][fx] Add an option in convert_fx to accept qconfig_dict to skip quantization (pytorch#66878)

  Show description for 8460fa5
  

  supriyar

  authored and

  facebook-github-bot

  committed
  8460fa5

  Copy full SHA for 8460fa5

• [quant][[fx] update observer_fqn to not depend on node.name (pytorch#66767)

  Show description for d13829e
  

  supriyar

  authored and

  facebook-github-bot

  committed
  d13829e

  Copy full SHA for d13829e

Commits on Oct 7, 2021

• [quant] Add support for quantization of Embedding{Bag} in dynamic quant APIs (pytorch#65674)

  Show description for 8a974a4
  

  supriyar

  authored and

  facebook-github-bot

  committed
  8a974a4

  Copy full SHA for 8a974a4

Commits on Oct 5, 2021

• Back out "[quant] update fused_obs_fake_quant op to accept output_fake_quant argument" (pytorch#66063)

  Show description for 458a00b
  

  supriyar

  authored and

  facebook-github-bot

  committed
  458a00b

  Copy full SHA for 458a00b

Commits on Sep 30, 2021

• [quant] update fused_obs_fake_quant op to accept output_fake_quant argument (pytorch#65621)

  Show description for 4666e3f
  

  supriyar

  authored and

  facebook-github-bot

  committed
  4666e3f

  Copy full SHA for 4666e3f

Commits on Sep 23, 2021

• [quant] change observer FQNs generated in prepare step (pytorch#65420)

  Show description for 767a104
  

  supriyar

  authored and

  facebook-github-bot

  committed
  767a104

  Copy full SHA for 767a104

Commits on Sep 15, 2021

• Revert D30558877: Ported std/var to ReductionOpInfo and minimum/maximum to BinaryUfuncInfo

  Show description for 36a0d97
  

  supriyar

  authored and

  facebook-github-bot

  committed
  36a0d97

  Copy full SHA for 36a0d97

Commits on Sep 14, 2021

• [quant] handle empty input in fused_moving_avg_obs_fake_quant op (pytorch#64829)

  Show description for d4b4d83
  

  supriyar

  authored and

  facebook-github-bot

  committed
  d4b4d83

  Copy full SHA for d4b4d83

Commits on Sep 13, 2021

• torch.ao migration: quantize_jit.py phase1 (pytorch#64860)

  Show description for 3d976d9
  

  supriyar

  authored and

  facebook-github-bot

  committed
  3d976d9

  Copy full SHA for 3d976d9

• torch.ao migration: stubs.py phase 1 (pytorch#64861)

  Show description for 9d52651
  

  supriyar

  authored and

  facebook-github-bot

  committed
  9d52651

  Copy full SHA for 9d52651

Commits on Aug 27, 2021

• [quant] Add support for linear_relu fusion for FP16 dynamic quant (pytorch#63826)

  Show description for 294db06
  

  supriyar

  authored and

  facebook-github-bot

  committed
  294db06

  Copy full SHA for 294db06

• [quant] Add op support for linear_relu_dynamic_fp16 (pytorch#63824)

  Show description for cec44aa
  

  supriyar

  authored and

  facebook-github-bot

  committed
  cec44aa

  Copy full SHA for cec44aa

• [quant] support linear_relu_dynamic for qnnpack backend (pytorch#63820)

  Show description for 975f4cc
  

  supriyar

  authored and

  facebook-github-bot

  committed
  975f4cc

  Copy full SHA for 975f4cc

• [quant][fx] Add support for dynamic linear + relu fusion (INT8) (pytorch#63799)

  Show description for c7027f1
  

  supriyar

  authored and

  facebook-github-bot

  committed
  c7027f1

  Copy full SHA for c7027f1

Commits on Aug 17, 2021

• [quant][fx] Ensure qconfig works for QAT with multiple modules (pytorch#63343)

  Show description for b0396e3
  

  supriyar

  authored and

  facebook-github-bot

  committed
  b0396e3

  Copy full SHA for b0396e3

Commits on Aug 16, 2021

• [docs][ao] Add overload information for fake_quantize_per_tensor_affine (pytorch#63258)

  Show description for a421cba
  

  supriyar

  authored and

  facebook-github-bot

  committed
  a421cba

  Copy full SHA for a421cba

• [docs][ao] Add missing docstrings for quantized_max_pool1d and quantized_max_pool2d (pytorch#63242)

  Show description for 0831b59
  

  supriyar

  authored and

  facebook-github-bot

  committed
  0831b59

  Copy full SHA for 0831b59

• [docs][ao] Add missing documentation for torch.quantized_batch_norm (pytorch#63240)

  Show description for a090073
  

  supriyar

  authored and

  facebook-github-bot

  committed
  a090073

  Copy full SHA for a090073

Commits on Aug 13, 2021

• [docs][ao] update quantize_per_tensor to mention overloads (pytorch#63165)

  Show description for 75f198d
  

  supriyar

  authored and

  facebook-github-bot

  committed
  75f198d

  Copy full SHA for 75f198d

Commits on Aug 12, 2021

• [quant] Make version 1 the default for get_default_qat_qconfig (pytorch#63043)

  Show description for d5a7579
  

  supriyar

  authored and

  facebook-github-bot

  committed
  d5a7579

  Copy full SHA for d5a7579

Commits on Aug 10, 2021

• [quant] Update get_default_qat_qconfig to return the fused observer+fake_quant module (pytorch#62702)

  Show description for aa89d5f
  

  supriyar

  authored and

  facebook-github-bot

  committed
  aa89d5f

  Copy full SHA for aa89d5f

• [quant] add reduce_range option to FusedMovingAvgFakeQuantize module (pytorch#62863)

  Show description for 08d1a12
  

  supriyar

  authored and

  facebook-github-bot

  committed
  08d1a12

  Copy full SHA for 08d1a12

Commits on Aug 2, 2021

• [quant] Support PerChannel quantization in FusedMovingAvgObsFakeQuantize (pytorch#62346)

  Show description for aa5e3ad
  

  supriyar

  authored and

  facebook-github-bot

  committed
  aa5e3ad

  Copy full SHA for aa5e3ad

Commits on Jul 30, 2021

• [quant] update per-channel observer min/max_val attribute names (pytorch#62345)

  Show description for cfd0f5e
  

  supriyar

  authored and

  facebook-github-bot

  committed
  cfd0f5e

  Copy full SHA for cfd0f5e

Commits on Jul 26, 2021

• [quant][fix] Update quantization c++ tests to not run if CPU_STATIC_DISPATCH is specified (pytorch#62197)

  Show description for dcb3ead
  

  supriyar

  authored and

  facebook-github-bot

  committed
  dcb3ead

  Copy full SHA for dcb3ead

Commits on Jul 25, 2021

• [quant] Add from_blob_quantized_per_channel API (pytorch#62049)

  Show description for 76d3cdf
  

  supriyar

  authored and

  facebook-github-bot

  committed
  76d3cdf

  Copy full SHA for 76d3cdf

• [quant] Add from_blob_quantized_per_tensor API (pytorch#61986)

  Show description for 7195b78
  

  supriyar

  authored and

  facebook-github-bot

  committed
  7195b78

  Copy full SHA for 7195b78

Commits on Jul 21, 2021

• [quant] Remove calls to .item() for fake_quant_on (pytorch#61921)

  Show description for ea8abcf
  

  supriyar

  authored and

  facebook-github-bot

  committed
  ea8abcf

  Copy full SHA for ea8abcf

• [quant] Create FusedMovingAvgObsFakeQuantize for QAT (pytorch#61691)

  Show description for b8386f5
  

  supriyar

  authored and

  facebook-github-bot

  committed
  b8386f5

  Copy full SHA for b8386f5

• [quant] Add a new fused MovingAvg Obs + FakeQuant operator (GPU) (pytorch#61589)

  Show description for afdca41
  

  supriyar

  authored and

  facebook-github-bot

  committed
  afdca41

  Copy full SHA for afdca41

• [quant] Add a new fused MovingAvg Obs + FakeQuant operator(CPU) (pytorch#61570)

  Show description for 92d3391
  

  supriyar

  authored and

  facebook-github-bot

  committed
  92d3391

  Copy full SHA for 92d3391

Commits on Jul 11, 2021

• [quant] update FakeQuant modules to use tensor qparams (pytorch#61318)

  Show description for 7a15576
  

  supriyar

  authored and

  facebook-github-bot

  committed
  7a15576

  Copy full SHA for 7a15576

Pagination

Previous

Next