Commits

Branch selector

 develop

User selector

All users

Datepicker

All time

Commit History

Commits on Feb 20, 2024

• Correct dtype handling for compressed torch.nn.Parameter (#2477)

  Show description for b5e0a05
  alexsu52

  authored
  b5e0a05

  Copy full SHA for b5e0a05

• Added deprecation warning (#2490)

  Show description for c8000ca
  nikita-malininn

  authored
  c8000ca

  Copy full SHA for c8000ca

• [PTQ] Fix ModeBasedDefaults (#2487)

  Show description for 6b0ba7a
  nikita-malininn

  authored
  6b0ba7a

  Copy full SHA for 6b0ba7a

• Hide an irrelevant warning during tune_range call (#2415)

  Show description for f812e88
  

  nikita-savelyevv

  and

  alexsu52

  authored
  f812e88

  Copy full SHA for f812e88

• [PTQ] Add mode based parameters (#2462)

  Show description for 80c1ea7
  nikita-malininn

  authored
  80c1ea7

  Copy full SHA for 80c1ea7

• [Torch] Subtract eps from quantizer parameters instead of zeroing it (#2479)

  Show description for 100bc17
  daniil-lyakhov

  authored
  100bc17

  Copy full SHA for 100bc17

• [PT] Insert point for not replaced modules (#2481)

  Show description for 40f8e39
  AlexanderDokuchaev

  authored
  40f8e39

  Copy full SHA for 40f8e39

Commits on Feb 19, 2024

• [docs] Added perplexity for AWQ. (#2443)

  Show description for 5451c64
  andreyanufr

  authored
  5451c64

  Copy full SHA for 5451c64

Commits on Feb 16, 2024

• Add unit tests for NNCF utils (#2468)

  Show description for 98cf1c1
  

  Candyzorua

  and

  vshampor

  authored
  98cf1c1

  Copy full SHA for 98cf1c1

• Fix TypeError: revert_operations_to_floating_point_precision() missing 2 required positional arguments (#2484)

  Show description for f27287e
  andrey-churkin

  authored
  f27287e

  Copy full SHA for f27287e

• [PTQ] Added try-catch for the OMZ imports (#2478) (#2483)

  Show description for 20312de
  nikita-malininn

  authored
  20312de

  Copy full SHA for 20312de

• [PT] Propagate is_shared attribute from constant node (#2467)

  Show description for fb32e46
  AlexanderDokuchaev

  authored
  fb32e46

  Copy full SHA for fb32e46

Commits on Feb 15, 2024

• [PTQ][OV] Added scale constant creation (#2475)

  Show description for 59a6a40
  nikita-malininn

  authored
  59a6a40

  Copy full SHA for 59a6a40

• [PT] Update reference for dorefa (#2471)

  Show description for 89d1675
  AlexanderDokuchaev

  authored
  89d1675

  Copy full SHA for 89d1675

• Fixed API Enums serialization/deserialization issue (#2474)

  Show description for d6df215
  AlexKoff88

  authored
  d6df215

  Copy full SHA for d6df215

Commits on Feb 14, 2024

• Added stateful model to weight compression tests (#2463)

  Show description for 6539272
  ljaljushkin

  authored
  6539272

  Copy full SHA for 6539272

• Improve logging in LLM tune hyperparams example (#2470)

  Show description for 4c360c9
  l-bat

  authored
  4c360c9

  Copy full SHA for 4c360c9

• subset_size and awq are keyword-only arguments in nncf.compress_weights (#2466)

  Show description for 287ec6b
  alexsu52

  authored
  287ec6b

  Copy full SHA for 287ec6b

Commits on Feb 13, 2024

• Add AWQ to LLM hyperparams tuning example (#2447)

  Show description for 27e4b33
  l-bat

  authored
  27e4b33

  Copy full SHA for 27e4b33

Commits on Feb 12, 2024

• Cache ref data to accelerate conformance tests for weight compression (#2460)

  Show description for 14bd2aa
  ljaljushkin

  authored
  14bd2aa

  Copy full SHA for 14bd2aa

• Add the AAQ algorithm for the ONNX backend (#2399)

  Show description for 7bddab9
  andrey-churkin

  authored
  7bddab9

  Copy full SHA for 7bddab9

• More strict threshold for accuracy and minor corrections (#2458)

  Show description for 4f071e3
  ljaljushkin

  authored
  4f071e3

  Copy full SHA for 4f071e3

• Issue fixed in nncf/common/sparsity- [NNCF] Make NNCF common sparsity code pass mypy checks #22312 (#2428)

  Show description for 57bece9
  

  AishwaryaDekhane

  and

  vshampor

  authored
  57bece9

  Copy full SHA for 57bece9

Commits on Feb 11, 2024

• [TF] Increase thresholds for e2e (#2457)

  Show description for 2ffc08c
  AlexanderDokuchaev

  authored
  2ffc08c

  Copy full SHA for 2ffc08c

Commits on Feb 8, 2024

• Add progress bar for calculating ranking scores (#2459)

  Show description for e30590d
  andrey-churkin

  authored
  e30590d

  Copy full SHA for e30590d

• Fix name of params (#2454)

  Show description for 11fe821
  AlexanderDokuchaev

  authored
  11fe821

  Copy full SHA for 11fe821

• [PTQ][OV] Force renaming for friendly names duplicates (#2444)

  Show description for c297a50
  nikita-malininn

  authored
  c297a50

  Copy full SHA for c297a50

• Introduced `subset_size` parameter for data-aware weight compression (#2455)

  Show description for 7a853b2
  ljaljushkin

  authored
  7a853b2

  Copy full SHA for 7a853b2

Commits on Feb 6, 2024

• nncf.quantize_with_accuracy_control fails with tune_hyperparams=True (#2450)

  Show description for e3f7561
  andrey-churkin

  authored
  e3f7561

  Copy full SHA for e3f7561

• Restoring state after inference only for stateful models (#2445)

  Show description for b49390e
  alexsu52

  authored
  b49390e

  Copy full SHA for b49390e

• nncf.AdvancedAccuracyRestorerParameters (#2448)

  Show description for edb3987
  alexsu52

  authored
  edb3987

  Copy full SHA for edb3987

Commits on Feb 5, 2024

• [PT][WC] Add WC tests for functional models (#2446)

  Show description for 2872f27
  l-bat

  authored
  2872f27

  Copy full SHA for 2872f27

• Added is_floating_point to FunctionsToPatchWithoutTracing (#2433)

  Show description for 1391667
  alexsu52

  authored
  1391667

  Copy full SHA for 1391667

Commits on Feb 2, 2024

• Conformance test for weight compression (#2425)

  Show description for 848b280
  ljaljushkin

  authored
  848b280

  Copy full SHA for 848b280

• Avoiding overflow on summing 32-bit int on Windows (#2431)

  Show description for 3544d49
  ljaljushkin

  authored
  3544d49

  Copy full SHA for 3544d49

Pagination

Previous

Next