mct-nightly 1.7.1.31122022.post351__py3-none-any.whl → 1.8.0.1042023.post423__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (241) hide show
  1. {mct_nightly-1.7.1.31122022.post351.dist-info → mct_nightly-1.8.0.1042023.post423.dist-info}/METADATA +16 -16
  2. {mct_nightly-1.7.1.31122022.post351.dist-info → mct_nightly-1.8.0.1042023.post423.dist-info}/RECORD +193 -150
  3. {mct_nightly-1.7.1.31122022.post351.dist-info → mct_nightly-1.8.0.1042023.post423.dist-info}/WHEEL +1 -1
  4. model_compression_toolkit/__init__.py +13 -14
  5. model_compression_toolkit/core/common/back2framework/base_model_builder.py +1 -1
  6. model_compression_toolkit/core/common/collectors/base_collector.py +7 -4
  7. model_compression_toolkit/core/common/collectors/statistics_collector.py +2 -2
  8. model_compression_toolkit/core/common/constants.py +9 -4
  9. model_compression_toolkit/core/common/framework_implementation.py +32 -30
  10. model_compression_toolkit/core/common/graph/base_graph.py +8 -6
  11. model_compression_toolkit/core/common/logger.py +10 -2
  12. model_compression_toolkit/core/common/matchers/base_matcher.py +3 -3
  13. model_compression_toolkit/core/common/mixed_precision/mixed_precision_quantization_config.py +2 -1
  14. model_compression_toolkit/core/common/mixed_precision/mixed_precision_search_facade.py +2 -2
  15. model_compression_toolkit/core/common/mixed_precision/mixed_precision_search_manager.py +2 -2
  16. model_compression_toolkit/core/common/mixed_precision/search_methods/linear_programming.py +6 -1
  17. model_compression_toolkit/core/common/model_validation.py +2 -1
  18. model_compression_toolkit/core/common/quantization/node_quantization_config.py +3 -1
  19. model_compression_toolkit/core/common/quantization/quantization_params_fn_selection.py +7 -4
  20. model_compression_toolkit/core/common/quantization/quantization_params_generation/lut_kmeans_params.py +4 -2
  21. model_compression_toolkit/core/common/quantization/quantization_params_generation/qparams_activations_computation.py +14 -17
  22. model_compression_toolkit/core/common/quantization/quantizers/quantizers_helpers.py +9 -2
  23. model_compression_toolkit/core/common/quantization/quantizers/uniform_quantizers.py +5 -4
  24. model_compression_toolkit/core/common/quantization/set_node_quantization_config.py +3 -3
  25. model_compression_toolkit/core/common/substitutions/batchnorm_reconstruction.py +7 -0
  26. model_compression_toolkit/core/common/substitutions/batchnorm_refusing.py +13 -8
  27. model_compression_toolkit/core/common/substitutions/shift_negative_activation.py +17 -12
  28. model_compression_toolkit/core/common/substitutions/weights_activation_split.py +1 -1
  29. model_compression_toolkit/core/common/target_platform/current_tp_model.py +3 -1
  30. model_compression_toolkit/core/common/target_platform/targetplatform2framework/attribute_filter.py +17 -4
  31. model_compression_toolkit/core/common/target_platform/targetplatform2framework/operations_to_layers.py +2 -4
  32. model_compression_toolkit/core/common/target_platform/targetplatform2framework/target_platform_capabilities.py +3 -5
  33. model_compression_toolkit/core/keras/back2framework/instance_builder.py +12 -21
  34. model_compression_toolkit/core/keras/back2framework/keras_model_builder.py +40 -14
  35. model_compression_toolkit/core/keras/back2framework/model_gradients.py +51 -27
  36. model_compression_toolkit/core/keras/constants.py +1 -0
  37. model_compression_toolkit/core/keras/graph_substitutions/substitutions/multi_head_attention_decomposition.py +2 -1
  38. model_compression_toolkit/core/keras/kpi_data_facade.py +2 -2
  39. model_compression_toolkit/core/keras/quantization_facade.py +3 -3
  40. model_compression_toolkit/core/keras/quantizer/fake_quant_builder.py +15 -9
  41. model_compression_toolkit/core/keras/quantizer/input_layer_quantize_transform.py +2 -1
  42. model_compression_toolkit/core/keras/quantizer/lut_fake_quant.py +1 -1
  43. model_compression_toolkit/core/keras/reader/common.py +3 -2
  44. model_compression_toolkit/core/pytorch/back2framework/instance_builder.py +14 -1
  45. model_compression_toolkit/core/pytorch/back2framework/model_gradients.py +88 -46
  46. model_compression_toolkit/core/pytorch/back2framework/pytorch_model_builder.py +27 -12
  47. model_compression_toolkit/core/pytorch/back2framework/quantization_wrapper/wrapper_quantize_config.py +2 -3
  48. model_compression_toolkit/core/pytorch/constants.py +5 -0
  49. model_compression_toolkit/core/pytorch/graph_substitutions/substitutions/multi_head_attention_decomposition.py +9 -14
  50. model_compression_toolkit/core/pytorch/graph_substitutions/substitutions/reshape_with_static_shapes.py +16 -2
  51. model_compression_toolkit/core/pytorch/kpi_data_facade.py +2 -2
  52. model_compression_toolkit/core/pytorch/quantization_facade.py +2 -2
  53. model_compression_toolkit/core/pytorch/quantizer/fake_quant_builder.py +7 -5
  54. model_compression_toolkit/core/pytorch/quantizer/lut_fake_quant.py +1 -1
  55. model_compression_toolkit/core/tpc_models/get_target_platform_capabilities.py +6 -2
  56. model_compression_toolkit/{exporter/model_wrapper/keras/quantize_configs → core/tpc_models/imx500_tpc}/__init__.py +1 -1
  57. model_compression_toolkit/core/tpc_models/imx500_tpc/latest/__init__.py +24 -0
  58. model_compression_toolkit/core/tpc_models/imx500_tpc/target_platform_capabilities.py +45 -0
  59. model_compression_toolkit/core/tpc_models/imx500_tpc/v1/__init__.py +16 -0
  60. model_compression_toolkit/core/tpc_models/imx500_tpc/v1/tp_model.py +156 -0
  61. model_compression_toolkit/core/tpc_models/imx500_tpc/v1/tpc_keras.py +101 -0
  62. model_compression_toolkit/core/tpc_models/imx500_tpc/v1/tpc_pytorch.py +95 -0
  63. model_compression_toolkit/exporter/__init__.py +5 -0
  64. model_compression_toolkit/exporter/model_exporter/__init__.py +0 -12
  65. model_compression_toolkit/exporter/model_exporter/fw_agonstic/exporter.py +1 -1
  66. model_compression_toolkit/exporter/model_exporter/keras/fakely_quant_keras_exporter.py +12 -39
  67. model_compression_toolkit/exporter/model_exporter/keras/keras_export_facade.py +39 -27
  68. model_compression_toolkit/exporter/model_exporter/pytorch/fakely_quant_onnx_pytorch_exporter.py +10 -2
  69. model_compression_toolkit/exporter/model_exporter/pytorch/fakely_quant_torchscript_pytorch_exporter.py +6 -2
  70. model_compression_toolkit/exporter/model_exporter/pytorch/pytorch_export_facade.py +48 -35
  71. model_compression_toolkit/exporter/model_exporter/tflite/fakely_quant_tflite_exporter.py +3 -2
  72. model_compression_toolkit/exporter/model_exporter/tflite/int8_tflite_exporter.py +180 -0
  73. model_compression_toolkit/exporter/model_exporter/tflite/tflite_export_facade.py +44 -26
  74. model_compression_toolkit/exporter/model_wrapper/__init__.py +4 -4
  75. model_compression_toolkit/exporter/model_wrapper/keras/builder/fully_quantized_model_builder.py +34 -137
  76. model_compression_toolkit/exporter/model_wrapper/keras/builder/node_to_quantizer.py +143 -0
  77. model_compression_toolkit/exporter/model_wrapper/keras/builder/node_to_quantizers.py +46 -0
  78. model_compression_toolkit/exporter/model_wrapper/keras/validate_layer.py +56 -22
  79. model_compression_toolkit/exporter/model_wrapper/pytorch/builder/fully_quantized_model_builder.py +29 -112
  80. model_compression_toolkit/exporter/model_wrapper/pytorch/builder/node_to_quantizer.py +83 -79
  81. model_compression_toolkit/exporter/model_wrapper/pytorch/builder/node_to_quantizers.py +47 -0
  82. model_compression_toolkit/exporter/model_wrapper/pytorch/validate_layer.py +44 -0
  83. model_compression_toolkit/gptq/__init__.py +6 -0
  84. model_compression_toolkit/gptq/common/gptq_config.py +57 -127
  85. model_compression_toolkit/gptq/common/gptq_constants.py +20 -6
  86. model_compression_toolkit/gptq/common/gptq_graph.py +22 -0
  87. model_compression_toolkit/gptq/common/gptq_training.py +32 -26
  88. model_compression_toolkit/gptq/keras/gptq_loss.py +1 -1
  89. model_compression_toolkit/gptq/keras/gptq_training.py +73 -39
  90. model_compression_toolkit/gptq/keras/graph_info.py +24 -43
  91. model_compression_toolkit/gptq/keras/quantization_facade.py +10 -18
  92. model_compression_toolkit/gptq/keras/quantizer/__init__.py +2 -1
  93. model_compression_toolkit/gptq/keras/quantizer/base_keras_gptq_quantizer.py +112 -0
  94. model_compression_toolkit/gptq/keras/quantizer/quant_utils.py +13 -14
  95. model_compression_toolkit/gptq/keras/quantizer/quantization_builder.py +78 -0
  96. model_compression_toolkit/gptq/keras/quantizer/regularization_factory.py +45 -0
  97. model_compression_toolkit/gptq/keras/{optimizers → quantizer/soft_rounding}/__init__.py +1 -1
  98. model_compression_toolkit/gptq/keras/quantizer/soft_rounding/soft_quantizer_reg.py +112 -0
  99. model_compression_toolkit/gptq/keras/quantizer/soft_rounding/symmetric_soft_quantizer.py +256 -0
  100. model_compression_toolkit/gptq/keras/quantizer/ste_rounding/symmetric_ste.py +68 -168
  101. model_compression_toolkit/gptq/pytorch/gptq_training.py +78 -39
  102. model_compression_toolkit/gptq/pytorch/graph_info.py +81 -0
  103. model_compression_toolkit/gptq/pytorch/quantization_facade.py +12 -18
  104. model_compression_toolkit/gptq/pytorch/quantizer/__init__.py +5 -1
  105. model_compression_toolkit/gptq/pytorch/quantizer/base_pytorch_gptq_quantizer.py +92 -0
  106. model_compression_toolkit/gptq/pytorch/quantizer/quant_utils.py +10 -119
  107. model_compression_toolkit/gptq/pytorch/quantizer/quantization_builder.py +75 -0
  108. model_compression_toolkit/gptq/pytorch/quantizer/regularization_factory.py +45 -0
  109. model_compression_toolkit/{exporter/model_wrapper/keras/quantizers → gptq/pytorch/quantizer/soft_rounding}/__init__.py +1 -1
  110. model_compression_toolkit/gptq/pytorch/quantizer/soft_rounding/soft_quantizer_reg.py +115 -0
  111. model_compression_toolkit/gptq/pytorch/quantizer/soft_rounding/symmetric_soft_quantizer.py +244 -0
  112. model_compression_toolkit/gptq/pytorch/quantizer/soft_rounding/uniform_soft_quantizer.py +196 -0
  113. model_compression_toolkit/gptq/pytorch/quantizer/ste_rounding/symmetric_ste.py +182 -0
  114. model_compression_toolkit/ptq/keras/quantization_facade.py +3 -3
  115. model_compression_toolkit/ptq/pytorch/quantization_facade.py +7 -6
  116. model_compression_toolkit/qat/common/qat_config.py +68 -0
  117. model_compression_toolkit/qat/keras/quantization_facade.py +55 -48
  118. model_compression_toolkit/qat/keras/quantizer/__init__.py +3 -0
  119. model_compression_toolkit/qat/keras/quantizer/base_keras_qat_quantizer.py +49 -0
  120. model_compression_toolkit/qat/keras/quantizer/quant_utils.py +48 -0
  121. model_compression_toolkit/qat/keras/quantizer/quantization_builder.py +77 -0
  122. model_compression_toolkit/qat/keras/quantizer/ste_rounding/symmetric_ste.py +283 -0
  123. model_compression_toolkit/qat/keras/quantizer/ste_rounding/uniform_ste.py +158 -46
  124. model_compression_toolkit/qat/pytorch/quantization_facade.py +190 -11
  125. model_compression_toolkit/qat/pytorch/quantizer/__init__.py +17 -0
  126. model_compression_toolkit/qat/pytorch/quantizer/base_pytorch_qat_quantizer.py +49 -0
  127. model_compression_toolkit/qat/pytorch/quantizer/quantization_builder.py +74 -0
  128. model_compression_toolkit/qat/pytorch/quantizer/quantizer_utils.py +136 -0
  129. model_compression_toolkit/qat/pytorch/quantizer/ste_rounding/symmetric_ste.py +204 -0
  130. model_compression_toolkit/qat/pytorch/quantizer/ste_rounding/uniform_ste.py +190 -0
  131. model_compression_toolkit/quantizers_infrastructure/__init__.py +23 -0
  132. model_compression_toolkit/{gptq/keras/quantizer/configs → quantizers_infrastructure/inferable_infrastructure}/__init__.py +1 -1
  133. model_compression_toolkit/{gptq/keras/quantizer/gumbel_rounding → quantizers_infrastructure/inferable_infrastructure/common}/__init__.py +1 -1
  134. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/common/base_inferable_quantizer.py +87 -0
  135. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/common/constants.py +41 -0
  136. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/common/get_all_subclasses.py +31 -0
  137. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/common/get_quantizers.py +53 -0
  138. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/common/quant_utils.py +49 -0
  139. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/__init__.py +14 -0
  140. model_compression_toolkit/{qunatizers_infrastructure → quantizers_infrastructure/inferable_infrastructure}/keras/load_model.py +26 -8
  141. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/quantize_wrapper.py +345 -0
  142. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/quantizer_utils.py +85 -0
  143. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/quantizers/__init__.py +27 -0
  144. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/quantizers/activation_inferable_quantizers/__init__.py +14 -0
  145. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/quantizers/activation_inferable_quantizers/activation_lut_pot_inferable_quantizer.py +148 -0
  146. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/quantizers/activation_inferable_quantizers/activation_pot_inferable_quantizer.py +65 -0
  147. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/quantizers/activation_inferable_quantizers/activation_symmetric_inferable_quantizer.py +86 -0
  148. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/quantizers/activation_inferable_quantizers/activation_uniform_inferable_quantizer.py +111 -0
  149. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/quantizers/base_keras_inferable_quantizer.py +56 -0
  150. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/quantizers/constants.py +25 -0
  151. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/quantizers/weights_inferable_quantizers/__init__.py +14 -0
  152. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/quantizers/weights_inferable_quantizers/weights_lut_pot_inferable_quantizer.py +79 -0
  153. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/quantizers/weights_inferable_quantizers/weights_lut_symmetric_inferable_quantizer.py +179 -0
  154. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/quantizers/weights_inferable_quantizers/weights_pot_inferable_quantizer.py +67 -0
  155. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/quantizers/weights_inferable_quantizers/weights_symmetric_inferable_quantizer.py +87 -0
  156. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/quantizers/weights_inferable_quantizers/weights_uniform_inferable_quantizer.py +163 -0
  157. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/keras/validation_functions.py +66 -0
  158. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/__init__.py +14 -0
  159. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantize_wrapper.py +269 -0
  160. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizer_utils.py +152 -0
  161. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/__init__.py +35 -0
  162. model_compression_toolkit/{exporter/model_wrapper/pytorch/quantizers → quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/activation_inferable_quantizers}/__init__.py +1 -1
  163. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/activation_inferable_quantizers/activation_lut_pot_inferable_quantizer.py +97 -0
  164. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/activation_inferable_quantizers/activation_pot_inferable_quantizer.py +62 -0
  165. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/activation_inferable_quantizers/activation_symmetric_inferable_quantizer.py +83 -0
  166. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/activation_inferable_quantizers/activation_uniform_inferable_quantizer.py +100 -0
  167. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/base_lut_symmetric_inferable_quantizer.py +95 -0
  168. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/base_pytorch_inferable_quantizer.py +48 -0
  169. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/base_symmetric_inferable_quantizer.py +70 -0
  170. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/base_uniform_inferable_quantizer.py +57 -0
  171. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/constants.py +26 -0
  172. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/weights_inferable_quantizers/__init__.py +14 -0
  173. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/weights_inferable_quantizers/weights_lut_pot_inferable_quantizer.py +77 -0
  174. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/weights_inferable_quantizers/weights_lut_symmetric_inferable_quantizer.py +106 -0
  175. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/weights_inferable_quantizers/weights_pot_inferable_quantizer.py +66 -0
  176. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/weights_inferable_quantizers/weights_symmetric_inferable_quantizer.py +104 -0
  177. model_compression_toolkit/quantizers_infrastructure/inferable_infrastructure/pytorch/quantizers/weights_inferable_quantizers/weights_uniform_inferable_quantizer.py +109 -0
  178. model_compression_toolkit/quantizers_infrastructure/trainable_infrastructure/__init__.py +14 -0
  179. model_compression_toolkit/quantizers_infrastructure/trainable_infrastructure/common/__init__.py +14 -0
  180. model_compression_toolkit/quantizers_infrastructure/trainable_infrastructure/common/base_trainable_quantizer.py +200 -0
  181. model_compression_toolkit/quantizers_infrastructure/trainable_infrastructure/common/get_quantizer_config.py +116 -0
  182. model_compression_toolkit/quantizers_infrastructure/trainable_infrastructure/common/get_quantizers.py +65 -0
  183. model_compression_toolkit/quantizers_infrastructure/trainable_infrastructure/common/quant_utils.py +36 -0
  184. model_compression_toolkit/quantizers_infrastructure/trainable_infrastructure/common/trainable_quantizer_config.py +97 -0
  185. model_compression_toolkit/quantizers_infrastructure/trainable_infrastructure/keras/__init__.py +14 -0
  186. model_compression_toolkit/quantizers_infrastructure/trainable_infrastructure/keras/base_keras_quantizer.py +90 -0
  187. model_compression_toolkit/quantizers_infrastructure/trainable_infrastructure/keras/config_serialization.py +80 -0
  188. model_compression_toolkit/quantizers_infrastructure/trainable_infrastructure/keras/quantizer_utils.py +48 -0
  189. model_compression_toolkit/quantizers_infrastructure/trainable_infrastructure/pytorch/__init__.py +14 -0
  190. model_compression_toolkit/quantizers_infrastructure/trainable_infrastructure/pytorch/base_pytorch_quantizer.py +66 -0
  191. model_compression_toolkit/exporter/model_wrapper/keras/builder/quantize_config_to_node.py +0 -66
  192. model_compression_toolkit/exporter/model_wrapper/keras/builder/quantizer_to_node.py +0 -134
  193. model_compression_toolkit/exporter/model_wrapper/keras/extended_quantize_wrapper.py +0 -81
  194. model_compression_toolkit/exporter/model_wrapper/keras/quantize_configs/activation_quantize_config.py +0 -81
  195. model_compression_toolkit/exporter/model_wrapper/keras/quantize_configs/weights_activation_quantize_config.py +0 -128
  196. model_compression_toolkit/exporter/model_wrapper/keras/quantize_configs/weights_quantize_config.py +0 -107
  197. model_compression_toolkit/exporter/model_wrapper/keras/quantizers/fq_quantizer.py +0 -99
  198. model_compression_toolkit/exporter/model_wrapper/keras/quantizers/weights_uniform_quantizer.py +0 -105
  199. model_compression_toolkit/exporter/model_wrapper/pytorch/builder/node_to_quantize_config.py +0 -61
  200. model_compression_toolkit/exporter/model_wrapper/pytorch/quantizers/fq_quantizer.py +0 -59
  201. model_compression_toolkit/exporter/model_wrapper/pytorch/quantizers/uniform_weights_quantizer.py +0 -67
  202. model_compression_toolkit/exporter/model_wrapper/pytorch/wrappers_quantize_configs/activation_quantize_config.py +0 -52
  203. model_compression_toolkit/exporter/model_wrapper/pytorch/wrappers_quantize_configs/no_quantization_quantize_config.py +0 -46
  204. model_compression_toolkit/exporter/model_wrapper/pytorch/wrappers_quantize_configs/weights_activation_quantize_config.py +0 -54
  205. model_compression_toolkit/exporter/model_wrapper/pytorch/wrappers_quantize_configs/weights_quantize_config.py +0 -52
  206. model_compression_toolkit/gptq/keras/gptq_model_builder.py +0 -104
  207. model_compression_toolkit/gptq/keras/optimizers/sam_optimizer.py +0 -119
  208. model_compression_toolkit/gptq/keras/quantizer/config_factory.py +0 -62
  209. model_compression_toolkit/gptq/keras/quantizer/configs/base_quantizer_gptq_config.py +0 -65
  210. model_compression_toolkit/gptq/keras/quantizer/configs/weight_quantizer_gptq_config.py +0 -269
  211. model_compression_toolkit/gptq/keras/quantizer/gumbel_rounding/base_gumbel_rounding.py +0 -263
  212. model_compression_toolkit/gptq/keras/quantizer/gumbel_rounding/gumbel_softmax.py +0 -75
  213. model_compression_toolkit/gptq/keras/quantizer/gumbel_rounding/symmetric_gumbel.py +0 -266
  214. model_compression_toolkit/gptq/keras/quantizer/gumbel_rounding/uniform_gumbel.py +0 -247
  215. model_compression_toolkit/gptq/keras/quantizer/kernel_functions.py +0 -50
  216. model_compression_toolkit/gptq/keras/quantizer/ste_rounding/uniform_ste.py +0 -49
  217. model_compression_toolkit/gptq/pytorch/gptq_graph_info.py +0 -94
  218. model_compression_toolkit/gptq/pytorch/gptq_model_builder.py +0 -113
  219. model_compression_toolkit/gptq/pytorch/quantizer/gptq_quantizer.py +0 -71
  220. model_compression_toolkit/gptq/pytorch/quantizer/gumbel_rounding/__init__.py +0 -14
  221. model_compression_toolkit/gptq/pytorch/quantizer/gumbel_rounding/base_gumbel_weights_quantizer.py +0 -157
  222. model_compression_toolkit/gptq/pytorch/quantizer/gumbel_rounding/sym_gumbel_weights_quantizer.py +0 -150
  223. model_compression_toolkit/gptq/pytorch/quantizer/gumbel_rounding/uniform_gumbel_weights_quantizer.py +0 -143
  224. model_compression_toolkit/gptq/pytorch/quantizer/quantizer_wrapper.py +0 -103
  225. model_compression_toolkit/gptq/pytorch/quantizer/ste_rounding/ste_weights_quantizer.py +0 -103
  226. model_compression_toolkit/qat/keras/qat_model_builder.py +0 -105
  227. model_compression_toolkit/qat/keras/quantizer/quantization_dispatcher_builder.py +0 -56
  228. model_compression_toolkit/qat/keras/quantizer/ste_rounding/symmetirc_ste.py +0 -145
  229. model_compression_toolkit/qunatizers_infrastructure/__init__.py +0 -8
  230. model_compression_toolkit/qunatizers_infrastructure/common/__init__.py +0 -14
  231. model_compression_toolkit/qunatizers_infrastructure/common/base_quantizer.py +0 -123
  232. model_compression_toolkit/qunatizers_infrastructure/common/node_quantization_dispatcher.py +0 -65
  233. model_compression_toolkit/qunatizers_infrastructure/keras/__init__.py +0 -14
  234. model_compression_toolkit/qunatizers_infrastructure/keras/base_keras_quantizer.py +0 -75
  235. model_compression_toolkit/qunatizers_infrastructure/keras/config_serialization.py +0 -83
  236. model_compression_toolkit/qunatizers_infrastructure/keras/keras_node_quantization_dispatcher.py +0 -74
  237. model_compression_toolkit/qunatizers_infrastructure/keras/quantize_wrapper.py +0 -194
  238. model_compression_toolkit/qunatizers_infrastructure/pytorch/__init__.py +0 -0
  239. {mct_nightly-1.7.1.31122022.post351.dist-info → mct_nightly-1.8.0.1042023.post423.dist-info}/LICENSE.md +0 -0
  240. {mct_nightly-1.7.1.31122022.post351.dist-info → mct_nightly-1.8.0.1042023.post423.dist-info}/top_level.txt +0 -0
  241. /model_compression_toolkit/{exporter/model_wrapper/pytorch/wrappers_quantize_configs → qat/pytorch/quantizer/ste_rounding}/__init__.py +0 -0
{mct_nightly-1.7.1.31122022.post351.dist-info → mct_nightly-1.8.0.1042023.post423.dist-info}/METADATA RENAMED
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.1
2
2
  Name: mct-nightly
3
- Version: 1.7.1.31122022.post351
3
+ Version: 1.8.0.1042023.post423
4
4
  Summary: A Model Compression Toolkit for neural networks
5
5
  Home-page: UNKNOWN
6
6
  License: UNKNOWN
@@ -49,8 +49,8 @@ MCT is developed by researchers and engineers working at Sony Semiconductor Isra
49
49
  ## Supported Features
50
50
 
51
51
  MCT supports different quantization methods:
52
- * Post training quantization (PTQ): [Keras API](https://sony.github.io/model_optimization/docs/api/experimental_api_docs/methods/keras_post_training_quantization_experimental.html#ug-keras-post-training-quantization-experimental), [PyTorch API](https://sony.github.io/model_optimization/docs/api/experimental_api_docs/methods/pytorch_post_training_quantization_experimental.html#ug-pytorch-post-training-quantization-experimental)
53
- * Gradient-based post training quantization (GPTQ): [Keras API](https://sony.github.io/model_optimization/docs/api/experimental_api_docs/methods/keras_gradient_post_training_quantization_experimental.html#ug-keras-gradient-post-training-quantization-experimental), [PyTorch API](https://sony.github.io/model_optimization/docs/api/experimental_api_docs/methods/pytorch_gradient_post_training_quantization_experimental.html#ug-pytorch-gradient-post-training-quantization-experimental)
52
+ * Post-training quantization (PTQ): [Keras API](https://sony.github.io/model_optimization/docs/api/experimental_api_docs/methods/keras_post_training_quantization_experimental.html#ug-keras-post-training-quantization-experimental), [PyTorch API](https://sony.github.io/model_optimization/docs/api/experimental_api_docs/methods/pytorch_post_training_quantization_experimental.html#ug-pytorch-post-training-quantization-experimental)
53
+ * Gradient-based post-training quantization (GPTQ): [Keras API](https://sony.github.io/model_optimization/docs/api/experimental_api_docs/methods/keras_gradient_post_training_quantization_experimental.html#ug-keras-gradient-post-training-quantization-experimental), [PyTorch API](https://sony.github.io/model_optimization/docs/api/experimental_api_docs/methods/pytorch_gradient_post_training_quantization_experimental.html#ug-pytorch-gradient-post-training-quantization-experimental)
54
54
  * Quantization aware training (QAT)[*](#experimental-features)
55
55
 
56
56
 
@@ -107,15 +107,15 @@ A nightly package is also available (unstable):
107
107
  pip install mct-nightly
108
108
  ```
109
109
 
110
- ### Requierments
110
+ ### Requirements
111
111
 
112
- To run MCT, one of the supported frameworks, Tenosflow/Pytorch, needs to be installed.
112
+ To run MCT, one of the supported frameworks, Tensorflow/Pytorch, needs to be installed.
113
113
 
114
- For using with Tensorflow please install the packages:
114
+ For use with Tensorflow please install the packages:
115
115
  [tensorflow](https://www.tensorflow.org/install),
116
116
  [tensorflow-model-optimization](https://www.tensorflow.org/model_optimization/guide/install)
117
117
 
118
- For using with PyTorch please install the packages:
118
+ For use with PyTorch please install the packages:
119
119
  [torch](https://pytorch.org/)
120
120
 
121
121
  Also, a [requirements](requirements.txt) file can be used to set up your environment.
@@ -127,12 +127,12 @@ Currently, MCT is being tested on various Python versions:
127
127
 
128
128
 
129
129
 
130
- | Python Version |
131
- |----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
132
- | [![Run Tests - Python 3.10](https://github.com/sony/model_optimization/actions/workflows/run_tests_suite_all_latest_frameworks_python310.yml/badge.svg)](https://github.com/sony/model_optimization/actions/workflows/run_tests_suite_all_latest_frameworks_python310.yml) |
133
- | [![Run Tests - Python 3.9](https://github.com/sony/model_optimization/actions/workflows/run_tests_suite_all_latest_frameworks_python39.yml/badge.svg)](https://github.com/sony/model_optimization/actions/workflows/run_tests_suite_all_latest_frameworks_python39.yml) |
134
- | [![Run Tests - Python 3.8](https://github.com/sony/model_optimization/actions/workflows/run_tests_suite_all_latest_frameworks_python38.yml/badge.svg)](https://github.com/sony/model_optimization/actions/workflows/run_tests_suite_all_latest_frameworks_python38.yml) |
135
- | [![Run Tests - Python 3.7](https://github.com/sony/model_optimization/actions/workflows/run_tests_suite_all_latest_frameworks_python37.yml/badge.svg)](https://github.com/sony/model_optimization/actions/workflows/run_tests_suite_all_latest_frameworks_python37.yml) |
130
+ | Python Version |
131
+ |---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
132
+ | [![Run Tests - Python 3.10](https://github.com/sony/model_optimization/actions/workflows/run_tests_suite_python310.yml/badge.svg)](https://github.com/sony/model_optimization/actions/workflows/run_tests_suite_python310.yml) |
133
+ | [![Run Tests - Python 3.9](https://github.com/sony/model_optimization/actions/workflows/run_tests_suite_python39.yml/badge.svg)](https://github.com/sony/model_optimization/actions/workflows/run_tests_suite_python39.yml) |
134
+ | [![Run Tests - Python 3.8](https://github.com/sony/model_optimization/actions/workflows/run_tests_suite_python38.yml/badge.svg)](https://github.com/sony/model_optimization/actions/workflows/run_tests_suite_python38.yml) |
135
+ | [![Run Tests - Python 3.7](https://github.com/sony/model_optimization/actions/workflows/run_tests_suite_python37.yml/badge.svg)](https://github.com/sony/model_optimization/actions/workflows/run_tests_suite_python37.yml) |
136
136
 
137
137
 
138
138
  ### Supported NN-Frameworks Versions
@@ -142,9 +142,9 @@ is tested on various versions:
142
142
 
143
143
  | TensorFlow Version | PyTorch Version |
144
144
  |------------------------------------------------------------------------------------------------------|----------------------------------------------------------------------------------------------------------|
145
- | ![tests](https://github.com/sony/model_optimization/actions/workflows/run_tests_tf210.yml/badge.svg) | ![tests](https://github.com/sony/model_optimization/actions/workflows/run_tests_torch1_13.yml/badge.svg) |
146
- | ![tests](https://github.com/sony/model_optimization/actions/workflows/run_tests_tf29.yml/badge.svg) | ![tests](https://github.com/sony/model_optimization/actions/workflows/run_tests_torch1_12.yml/badge.svg) |
147
- | ![tests](https://github.com/sony/model_optimization/actions/workflows/run_tests_tf28.yml/badge.svg) | ![tests](https://github.com/sony/model_optimization/actions/workflows/run_tests_torch1_11.yml/badge.svg) |
145
+ | ![tests](https://github.com/sony/model_optimization/actions/workflows/run_tests_tf211.yml/badge.svg) | ![tests](https://github.com/sony/model_optimization/actions/workflows/run_tests_torch1_13.yml/badge.svg) |
146
+ | ![tests](https://github.com/sony/model_optimization/actions/workflows/run_tests_tf210.yml/badge.svg) | ![tests](https://github.com/sony/model_optimization/actions/workflows/run_tests_torch1_12.yml/badge.svg) |
147
+ | ![tests](https://github.com/sony/model_optimization/actions/workflows/run_tests_tf29.yml/badge.svg) | ![tests](https://github.com/sony/model_optimization/actions/workflows/run_tests_torch1_11.yml/badge.svg) |
148
148
 
149
149
 
150
150
  ### Usage Example