v5.0: Samsung Exynos 2600 models

MD5 (mnv4_large.nnc)            = 6d86899c52a88ae24c025a327bd7e3a0
MD5 (mnv4_large_offline.nnc)    = dbef7a1c6d56e2437d89085a3a38d7bf
MD5 (mobile_bert.nnc)           = 5b7c8b635697c909693264034fcc5898
MD5 (od.nnc)                    = 43a7e0faa0ab1e8a86e774947792e36d
MD5 (sd_dec.nnc)                = 0d961ff0471472b2903594f497e3064c
MD5 (sd_enc.nnc)                = 9470f2195a2b2eee0c0e90d5fd3853fe
MD5 (sd_unet.nnc)               = f600698aab63398291f64f3c49d99b2e
MD5 (sm_uint8.nnc)              = 43814a29b2e63719af67a30e8b5efc0c
MD5 (sr.nnc)                    = 5ff526a6a30f781fdc4be310df43ac5e
MD5 (np.bin)                    = c50807d72ce221cf08a2248a6ac3c48e
MD5 (te.bin)                    = 798b772155a69de5df44b304327bb3cc
MD5 (gt.bin)                    = f41c1130809647fbccd76707b2f14305

v5.0: Samsung Exynos 2500 models

MD5 (mnv4_large.nnc) = 0bcbcc1d1506f45497a3add1eac2303b
MD5 (mnv4_large_offline.nnc) = 994458e8fcd36559d50cd2529a73ec69
MD5 (mobile_bert.nnc) = f3c446fb3856d8bd57ed22e010194ec7
MD5 (sm_uint8.nnc) = 7c04cad5d3317f383d955dfe05f84bd9
MD5 (od.nnc) = c47e47fa00770db8a55e5cd4c09f5189
MD5 (sr.nnc) = a2c5b250e39e4f9175ba5bb831957608
MD5 (sd_dec.nnc) = 1e9e7a096a76a503be771843068b921a
MD5 (sd_enc.nnc) = 8fc2bd5fbc3fa905b9232dc284309aea
MD5 (sd_unet.nnc) = 0c1f14ab6fe70c7e29ee5840acd8b25b
MD5 (np.bin) = c50807d72ce221cf08a2248a6ac3c48e
MD5 (te.bin) = 798b772155a69de5df44b304327bb3cc
MD5 (gt.bin) = f41c1130809647fbccd76707b2f14305

v4.1: TFLite models

MD5 (sd_decoder_dynamic.tflite) = 68acdb62f99e1dc2c7f5db8cdd0e007c
MD5 (sd_diffusion_model_dynamic.tflite) = 309e95f76ac8de01130942037a28aa8f
MD5 (sd_text_encoder_dynamic.tflite) = b64effb0360f9ea49a117cdaf8a2fbdc
MD5 (timestep_embeddings_data.bin.ts) = 798b772155a69de5df44b304327bb3cc

The TFLite models for Stable Diffusion v1.5 have been converted from the Hugging Face's models.

v4.1: Qualcomm models

MD5 (betas.bin) = 09d2e4306d319caf1b34e6afb5c63c22
MD5 (lambdas.bin) = c7179725ec31a6e2c7daf008a5e1ff23
MD5 (sd_precompute_data.tar) = beb7fe2da40042fb585bb8cb95d86b4d
MD5 (text_encoder.serialized.bin) = 6da7b95fa467e99af2b9f80c7afe3734
MD5 (unet.serialized.bin) = 3b504b92cbd788d713ca9cfc5b19d596
MD5 (vae_decoder.serialized.bin) = c7762e64c2596abe7f16614709cc5482

MD5 (mobile_mosaic_htp.dlc) = 3c0dfbacda053773d6afb34503d9991a
MD5 (mobilebert_quantized_htp.dlc) = 96d947175f04950898a372890907dda1
MD5 (mobilebert_quantized_htp_O2.dlc) = f8631dbd69819438d6b317c204fa80d7
MD5 (mobilenet_v4_htp.dlc) = 56e5039260e20e5c2a0b54cc0fac8098
MD5 (mobilenet_v4_htp_batched_4.dlc) = 7863deea588936fe6e09565ed47dde95
MD5 (mobilenet_v4_htp_batched_4_O2.dlc) = 80ba82f2a628ab712d812d06524d2bd8
MD5 (snusr_htp.dlc) = 668da9816073d67972704e237137a50f
MD5 (snusr_htp_O2.dlc) = 76b33f02ebfa6294a0e973aaf91116fa
MD5 (ssd_mobiledet_qat_htp.dlc) = 49c6afbfefffb78269fe73a6ee1b4a85

Note from Qualcomm for Stable Diffusion v1.5:
AI Model Efficiency Toolkit (AIMET) is used to create quantization simulation models (QuantSim) for the text encoder, U-Net, and VAE using a mixed-precision quantization scheme. A calibration process is used to create these QuantSim models where per-layer quantization encodings are determined using representative data samples. The QuantSim models simulate running the stable diffusion models on a quantized target. In addition, for the Text Encoder model, the AIMET Adaptive Rounding (AdaRound) technique is applied to get a boost in quantized accuracy.
Time Embeddings are precomputed during quantization step above. During inference, the model for Unet takes 1x1280 precomputed Time Embeddings.

v4.1: MediaTek models

MD5 (MobileNetV4-Conv-Large-int8-ptq.dla) = ff9e3705d4a62c4b78e2597156bb7599
MD5 (edsr_f32b5_full_qint8.dla) = cc40f7f0e97cf34c8586883fd3357354
MD5 (mobile_segmenter_r4_quant_argmax_uint8.dla) = 105fa48160803201dedec155445dd4e9
MD5 (mobilebert_int8_384_nnapi.dla) = 2c81eebd87e3f620373897cc56dbc3e7
MD5 (mobiledet_qat.dla) = 97cc339e205dfe5503d7dc256b12f472

v4.1: Stable Diffusion dataset

MD5 (coco_gen_full.tfrecord) = 5cf967d2b2128edeb1b4d6eca6e8d94d
MD5 (coco_gen_test.tfrecord) = b564d2c228a867148fa7d6df415a0368
MD5 (clip_model_512x512_openai-clip-vit-large-patch14.tflite) = 39a07ffaea0806ee6148874ef228cc77
MD5 (captions_source.tsv) = 18ef179e2343732b5561dc47ca39362d

Generated using the Python scripts in
https://github.com/mlcommons/mobile_app_open/tree/submission-v4.1/flutter/cpp/datasets/coco_gen_utils

v4.1: Core ML model

MD5 (mobilenetv4_fp32_NCHW.mlpackage.zip) = 164c504eb3e9af6c730c1765b8b81b32

v4.0: SNPE models

MD5 (mobile_mosaic_htp.dlc) = e870526444c1e48df4f0505e530ecfdf
MD5 (mobile_mosaic_htp_O2.dlc) = 99b39c2b9ea84ff13e00eaa82f00136b
MD5 (mobilebert_quantized_htp.dlc) = 7a641e4df84fc06a1237b7fe1b1c5b08
MD5 (mobilebert_quantized_htp_O2.dlc) = 9d0dadbb6014289916a6078c4c991dd5
MD5 (mobilenet_edgetpu_224_1.0_htp.dlc) = cdf1fe622b309f692e05781661248a2b
MD5 (mobilenet_edgetpu_224_1.0_htp_O2.dlc) = 25977982896e607bceb55340c8d76223
MD5 (mobilenet_edgetpu_224_1.0_htp_batched_3.dlc) = 550f807bc7ef40f77018a64a47507d09
MD5 (mobilenet_edgetpu_224_1.0_htp_batched_3_O2.dlc) = aca3f4430fe98bbfe5c3a358ae9687e1
MD5 (mobilenet_edgetpu_224_1.0_htp_batched_4.dlc) = 6523060565b8d3f326f3f323c531fc1c
MD5 (mobilenet_edgetpu_224_1.0_htp_batched_4_O2.dlc) = b836e404b3aa5ff7914fac8376643fe4
MD5 (mobilenet_edgetpu_224_1.0_htp_batched_8.dlc) = 1e09cab7d0d381ef02cfd5ea5b85da92
MD5 (mobilenet_edgetpu_224_1.0_htp_batched_8_O2.dlc) = 561d82168202b5b7c951ff61b9722796
MD5 (mobilenet_v4_htp.dlc) = dbab3e231e5f83aabc80d5b69e6dad32
MD5 (mobilenet_v4_htp_O2.dlc) = cc659c98c8ee65caeee4f7606eb28b68
MD5 (mobilenet_v4_htp_batched_4.dlc) = 0de3b75022ce5c27d5902a080ec1cea0
MD5 (mobilenet_v4_htp_batched_4_O2.dlc) = d349e3fb8a74a5037ecc3b2770dbd188
MD5 (snusr_htp.dlc) = 84ef0d9c2e7b710381cea962a22a0b41
MD5 (snusr_htp_O2.dlc) = 18fa274659e14c57b4f6bedb6871c83f
MD5 (ssd_mobiledet_qat_htp.dlc) = c333fc135a8c474679d716fe391a9e2a
MD5 (ssd_mobiledet_qat_htp_O2.dlc) = 5802abfad10a7fc5c5849b13943d6d44

v4.0: Samsung models

MD5 (ic.nnc) = 1a84db0b3dc40a6e217b7f2c8eb2980b
MD5 (ic_offline.nnc) = 9fcba01bf828b5cb5b1a39b55a8a7ed9
MD5 (mnv4_large.nnc) = 4a4a6b332ebcabc1a06eca019a1d7d0c
MD5 (mnv4_large_offline.nnc) = eec18f3bcca9d34f15ff73b73b7b0a2f
MD5 (mobile_bert.nnc) = 59356d4fa598d2d4369ca43f2051aff2
MD5 (od.nnc) = 5e7abcc8fa6f7c3e0bd8955a9ef1e8cb
MD5 (sm_uint8.nnc) = b02531ae19686c6574857bb18406966c
MD5 (sr.nnc) = 3fea2aa73be65efceb783364f4d7ae8d

v4.0: MediaTek models

MD5 (MobileNetV4-Conv-Large-int8-ptq.dla) = 07055309718ff681d8cd2c00f5e4b5db
MD5 (edsr_f32b5_full_qint8.dla) = fcd91d276036be666153c663c03fb69e
MD5 (mobile_segmenter_r4_quant_argmax_uint8.dla) = fe62a283e6da531647da15b3f26e680a
MD5 (mobilebert_int8_384_nnapi.dla) = 8c6ce45cc49bbf8bb26609bfb219164a
MD5 (mobiledet_qat.dla) = 14b9572f121caa093cd8bf690fde997c
MD5 (mobilenet_edgetpu_224_1.0_uint8.dla) = 48a2cfa77645c5dd81ad4b83a3f31e8a