Skip to content
Snippets Groups Projects

Repository graph

You can move around the graph by using the arrow keys.
Select Git revision
  • 4-bit-qat
  • api_changes
  • ax_integration
  • gste
  • inception_support
  • kd_fixes
  • lattice_regularization
  • load_quant_aware_train_checkpoint
  • master default protected
  • ptq_propagate_settings
  • scale_approximation
  • update_links
  • v0.3.2
  • amc_before_pytorch1.1_merge
  • v0.3.1
  • v0.3.0
  • v0.2.1
  • v0.2.0
  • v0.1.0
19 results
Created with Raphaël 2.2.06Aug431Jul29262423222117161087432130Jun2625242322191817141311103230May29282722212019161514652130Apr18161411109843131Mar2928272524232221201714121110765432128Feb272625232017151413121110631Jan2724232221161513109826Dec231916151411965432130Nov25242221209876542131Oct292726232218131143130Sep27212016429Aug279731Jul2522211917161513119854130Jun29262522211915141310729May221716151413107129Apr282524AMC and other refactoring - large merge (#339)trigger github-pages rebuildAdd documentation section on preparing a model for quantizationFix crash in dynamic post-train quant (caused by commit 69b1452)Add missing import (from commit 69b1452a)DistillerModuleList conversion: Handle models w. duplicate modules (#338)Post-train quant: Special handling for bidirectional DistillerLSTM (#337)cleaned code a bitFixed resuming QAT checkpointModified the flow of compress_classifier.py -Merge branch 'master' into load_quant_aware_train_checkpointmerged masterSave dummy_input in quantizer metadata (#333)NCF changes to make it compatible with latest changes in masterMerge branch 'master' into ncfBugfix in quant stats collector in case of single-valued samplesFix non 1:1 mapping between model w. ModuleList and SummaryGraph (#328)Fix bug in test that resulted in duplicate modules in a modelUpdated READMEFix for issue #323Increase Tensorflow min version to 1.14 (#320)Update post-train quant command line examplePost-Train Quantization: BN folding and "net-aware quantization" (#313)Add links to language model quantization notebook in README and docsMore generic implementation of distiller.utils.model_deviceAdded a fix for the quant-aware-train checkpoint loading issue.Bypass Torchvision MobileNet v2 ONNX-export bugSwitch to PyTorch 1.1.0 (#306)Update ADC_DDPG.py to comply with latest Coach changesamc_before_pyto…amc_before_pytorch1.1_mergeAMC notebook: several fixes and featuresNon-functional changescoach interface: Set Coach's PRNG seedrewards.py: env.dense_model_macs renamed to env.original_model_macsenvironment.py: various renamingranked_structures_pruner.py: Various feauresAdd ranking noiseModel thinning bug fixv0.3.2v0.3.2Fix bug in last commit (save only if logdir defined)Post train quant: Save per-layer quantization params to yaml fileDump outputs in run log dir instead of script dir