diff --git a/README.md b/README.md index d63372b5a0e1defba17a605b32a88f7b72ec2e21..340d2a16502e20c3b7a15ef9aca3d3f8e9e0a496 100644 --- a/README.md +++ b/README.md @@ -1,8 +1,8 @@ --- license: creativeml-openrail-m -base_model: stabilityai/stable-diffusion-2-1 -instance_prompt: a photo of ohwx person +base_model: SG161222/Realistic_Vision_V2.0 +instance_prompt: ohwx man tags: - stable-diffusion - stable-diffusion-diffusers @@ -14,7 +14,7 @@ inference: true # DreamBooth - yuanbit/model -This is a dreambooth model derived from stabilityai/stable-diffusion-2-1. The weights were trained on a photo of ohwx person using [DreamBooth](https://dreambooth.github.io/). +This is a dreambooth model derived from SG161222/Realistic_Vision_V2.0. The weights were trained on ohwx man using [DreamBooth](https://dreambooth.github.io/). You can find some example images in the following. diff --git a/checkpoint-1000/optimizer.bin b/checkpoint-1000/optimizer.bin index e3ff53c5e8450d2f18922e114a6d2335e5b6fc5c..697ec561b4e1628c5656adb3667f34bc4f7eab76 100644 --- a/checkpoint-1000/optimizer.bin +++ b/checkpoint-1000/optimizer.bin @@ -1,3 +1,3 @@ version https://git-lfs.github.com/spec/v1 -oid sha256:47142cab32d6769688e49c2fdbdd5c66e7775d834a25510fa9d05dedae93b6c0 -size 1737915525 +oid sha256:1a131ce48ebc96d440da4def00689966e8f77c46d27d6d6b3d17f8a1ec8b9cf6 +size 1725110981 diff --git a/checkpoint-1000/random_states_0.pkl b/checkpoint-1000/random_states_0.pkl index 7b34f567733ec0571ba3736de96326204234ec52..5107cbd4b5c257555f3635f3d1200b71d67749bc 100644 --- a/checkpoint-1000/random_states_0.pkl +++ b/checkpoint-1000/random_states_0.pkl @@ -1,3 +1,3 @@ version https://git-lfs.github.com/spec/v1 -oid sha256:b23433af9bb762ff0f8fc0121398922a04a5fc0cc4ecc066f11acdc8797678d0 -size 14727 +oid sha256:1f21db154c8cde738fcbb96197e1f156a4298d826f529948bc46bfc79ecdc923 +size 14599 diff --git a/checkpoint-1000/scaler.pt b/checkpoint-1000/scaler.pt new file mode 100644 index 0000000000000000000000000000000000000000..efdbd3c795f6b0d4144e68355e99c220ccdedd09 --- /dev/null +++ b/checkpoint-1000/scaler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:68cff80b680ddf6e7abbef98b5f336b97f9b5963e2209307f639383870e8cc71 +size 557 diff --git a/checkpoint-1000/scheduler.bin b/checkpoint-1000/scheduler.bin index 98ad17ca6fa67d9831589575e4d3fe4bf329fc9e..9e8e72d6d08c3d6a62a953beb6192a45ac6679fb 100644 --- a/checkpoint-1000/scheduler.bin +++ b/checkpoint-1000/scheduler.bin @@ -1,3 +1,3 @@ version https://git-lfs.github.com/spec/v1 -oid sha256:b8b7809f518a96a4fa4069f707e8433d0ae60949b56c65c39b0a64950d711aeb +oid sha256:b9d01d377e7ad0b99d75bd9b0223bbebdb7b53dab86375de089d5d92c547729d size 563 diff --git a/checkpoint-1000/unet/config.json b/checkpoint-1000/unet/config.json index 6389e554338110f210811f857aebeffe784d3f91..14e2e16ac7b0b32e261001f55127e59e4c91eee4 100644 --- a/checkpoint-1000/unet/config.json +++ b/checkpoint-1000/unet/config.json @@ -1,16 +1,11 @@ { "_class_name": "UNet2DConditionModel", "_diffusers_version": "0.18.0.dev0", - "_name_or_path": "stabilityai/stable-diffusion-2-1", + "_name_or_path": "SG161222/Realistic_Vision_V2.0", "act_fn": "silu", "addition_embed_type": null, "addition_embed_type_num_heads": 64, - "attention_head_dim": [ - 5, - 10, - 20, - 20 - ], + "attention_head_dim": 8, "block_out_channels": [ 320, 640, @@ -22,7 +17,7 @@ "class_embeddings_concat": false, "conv_in_kernel": 3, "conv_out_kernel": 3, - "cross_attention_dim": 1024, + "cross_attention_dim": 768, "cross_attention_norm": null, "down_block_types": [ "CrossAttnDownBlock2D", @@ -50,7 +45,7 @@ "resnet_out_scale_factor": 1.0, "resnet_skip_time_act": false, "resnet_time_scale_shift": "default", - "sample_size": 96, + "sample_size": 64, "time_cond_proj_dim": null, "time_embedding_act_fn": null, "time_embedding_dim": null, @@ -62,6 +57,6 @@ "CrossAttnUpBlock2D", "CrossAttnUpBlock2D" ], - "upcast_attention": true, - "use_linear_projection": true + "upcast_attention": false, + "use_linear_projection": false } diff --git a/checkpoint-1000/unet/diffusion_pytorch_model.bin b/checkpoint-1000/unet/diffusion_pytorch_model.bin index 8c5b76b89d7a67f5b190da04142dae4aa076dc01..e04f1293f20626987b4930ed00240a0021b04047 100644 --- a/checkpoint-1000/unet/diffusion_pytorch_model.bin +++ b/checkpoint-1000/unet/diffusion_pytorch_model.bin @@ -1,3 +1,3 @@ version https://git-lfs.github.com/spec/v1 -oid sha256:96de0ee4e02c38003f2fc28557a0a8114bc4fe824b8d676f33a3fc1842df0d6c -size 3463934693 +oid sha256:eec416e57b207456445793e84b260dc6b5bceea21a856d54b7d9faaec88797fe +size 3438375973 diff --git a/checkpoint-1500/optimizer.bin b/checkpoint-1500/optimizer.bin new file mode 100644 index 0000000000000000000000000000000000000000..8c16b2dad28ac385164a7de05d7c8cc70addd3a1 --- /dev/null +++ b/checkpoint-1500/optimizer.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a9cb6ecefa22e6bb6e16a6d13a3e2873ea956f8dd4d71e1bc8a7a6ec28817843 +size 1725110981 diff --git a/checkpoint-1500/random_states_0.pkl b/checkpoint-1500/random_states_0.pkl new file mode 100644 index 0000000000000000000000000000000000000000..ef3c31b45bc3bfce6df11271c87ef8b90f1c34b6 --- /dev/null +++ b/checkpoint-1500/random_states_0.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:af1dd3db952310b5ec5a0d7395612d676b8fca45283128aeacc33b40cfb1b69d +size 14599 diff --git a/checkpoint-1500/scaler.pt b/checkpoint-1500/scaler.pt new file mode 100644 index 0000000000000000000000000000000000000000..69c76b8f27f096d1d1a3d9d0e387af1a9cf5028d --- /dev/null +++ b/checkpoint-1500/scaler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:203a72d6c29f42a0e2964fdddc8d7a98df1eccee78fea9de0fa416613390f5c6 +size 557 diff --git a/checkpoint-1500/scheduler.bin b/checkpoint-1500/scheduler.bin new file mode 100644 index 0000000000000000000000000000000000000000..f45702ae693c6cb402fa3b43b07bf6223d4170e0 --- /dev/null +++ b/checkpoint-1500/scheduler.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4ba37372eb32dd7950563bf32168bbfc20205f203aa6a37df6ade894c798557b +size 563 diff --git a/checkpoint-1500/unet/config.json b/checkpoint-1500/unet/config.json new file mode 100644 index 0000000000000000000000000000000000000000..14e2e16ac7b0b32e261001f55127e59e4c91eee4 --- /dev/null +++ b/checkpoint-1500/unet/config.json @@ -0,0 +1,62 @@ +{ + "_class_name": "UNet2DConditionModel", + "_diffusers_version": "0.18.0.dev0", + "_name_or_path": "SG161222/Realistic_Vision_V2.0", + "act_fn": "silu", + "addition_embed_type": null, + "addition_embed_type_num_heads": 64, + "attention_head_dim": 8, + "block_out_channels": [ + 320, + 640, + 1280, + 1280 + ], + "center_input_sample": false, + "class_embed_type": null, + "class_embeddings_concat": false, + "conv_in_kernel": 3, + "conv_out_kernel": 3, + "cross_attention_dim": 768, + "cross_attention_norm": null, + "down_block_types": [ + "CrossAttnDownBlock2D", + "CrossAttnDownBlock2D", + "CrossAttnDownBlock2D", + "DownBlock2D" + ], + "downsample_padding": 1, + "dual_cross_attention": false, + "encoder_hid_dim": null, + "encoder_hid_dim_type": null, + "flip_sin_to_cos": true, + "freq_shift": 0, + "in_channels": 4, + "layers_per_block": 2, + "mid_block_only_cross_attention": null, + "mid_block_scale_factor": 1, + "mid_block_type": "UNetMidBlock2DCrossAttn", + "norm_eps": 1e-05, + "norm_num_groups": 32, + "num_class_embeds": null, + "only_cross_attention": false, + "out_channels": 4, + "projection_class_embeddings_input_dim": null, + "resnet_out_scale_factor": 1.0, + "resnet_skip_time_act": false, + "resnet_time_scale_shift": "default", + "sample_size": 64, + "time_cond_proj_dim": null, + "time_embedding_act_fn": null, + "time_embedding_dim": null, + "time_embedding_type": "positional", + "timestep_post_act": null, + "up_block_types": [ + "UpBlock2D", + "CrossAttnUpBlock2D", + "CrossAttnUpBlock2D", + "CrossAttnUpBlock2D" + ], + "upcast_attention": false, + "use_linear_projection": false +} diff --git a/checkpoint-1500/unet/diffusion_pytorch_model.bin b/checkpoint-1500/unet/diffusion_pytorch_model.bin new file mode 100644 index 0000000000000000000000000000000000000000..a1446273d29670bae135e467eabfeaf9746709b7 --- /dev/null +++ b/checkpoint-1500/unet/diffusion_pytorch_model.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8d447bf7f1165724724e36872271e56b357559f38109249564a6ee3692ce6b56 +size 3438375973 diff --git a/checkpoint-2000/optimizer.bin b/checkpoint-2000/optimizer.bin new file mode 100644 index 0000000000000000000000000000000000000000..369fe78510d3fe7e5ddbdfe3dd114675e4fa691a --- /dev/null +++ b/checkpoint-2000/optimizer.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5494a569ad891003b00a30fd89a3b561c259e46353de683999e70afc3aba4b84 +size 1725110981 diff --git a/checkpoint-2000/random_states_0.pkl b/checkpoint-2000/random_states_0.pkl new file mode 100644 index 0000000000000000000000000000000000000000..2e784da43a0311be5d5873c9d49ed05e8dce3abf --- /dev/null +++ b/checkpoint-2000/random_states_0.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f787282e90035829dd56fc05ecb152579e581420d2b6dadddc723044c2fa2dd2 +size 14599 diff --git a/checkpoint-2000/scaler.pt b/checkpoint-2000/scaler.pt new file mode 100644 index 0000000000000000000000000000000000000000..21ece35416ce79724d347155f11cfa297b97cabc --- /dev/null +++ b/checkpoint-2000/scaler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dd2de9749828adacdf103bf6e9592702bb7067a2c1df27dd62ab38c1eb8c070f +size 557 diff --git a/checkpoint-2000/scheduler.bin b/checkpoint-2000/scheduler.bin new file mode 100644 index 0000000000000000000000000000000000000000..d8163c7c7fa07c224995c0f1d5641a616f361676 --- /dev/null +++ b/checkpoint-2000/scheduler.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6c0552b668d37e1982c999e28875a1750cad69238c57b443fee8417dd5194c3a +size 563 diff --git a/checkpoint-2000/unet/config.json b/checkpoint-2000/unet/config.json new file mode 100644 index 0000000000000000000000000000000000000000..14e2e16ac7b0b32e261001f55127e59e4c91eee4 --- /dev/null +++ b/checkpoint-2000/unet/config.json @@ -0,0 +1,62 @@ +{ + "_class_name": "UNet2DConditionModel", + "_diffusers_version": "0.18.0.dev0", + "_name_or_path": "SG161222/Realistic_Vision_V2.0", + "act_fn": "silu", + "addition_embed_type": null, + "addition_embed_type_num_heads": 64, + "attention_head_dim": 8, + "block_out_channels": [ + 320, + 640, + 1280, + 1280 + ], + "center_input_sample": false, + "class_embed_type": null, + "class_embeddings_concat": false, + "conv_in_kernel": 3, + "conv_out_kernel": 3, + "cross_attention_dim": 768, + "cross_attention_norm": null, + "down_block_types": [ + "CrossAttnDownBlock2D", + "CrossAttnDownBlock2D", + "CrossAttnDownBlock2D", + "DownBlock2D" + ], + "downsample_padding": 1, + "dual_cross_attention": false, + "encoder_hid_dim": null, + "encoder_hid_dim_type": null, + "flip_sin_to_cos": true, + "freq_shift": 0, + "in_channels": 4, + "layers_per_block": 2, + "mid_block_only_cross_attention": null, + "mid_block_scale_factor": 1, + "mid_block_type": "UNetMidBlock2DCrossAttn", + "norm_eps": 1e-05, + "norm_num_groups": 32, + "num_class_embeds": null, + "only_cross_attention": false, + "out_channels": 4, + "projection_class_embeddings_input_dim": null, + "resnet_out_scale_factor": 1.0, + "resnet_skip_time_act": false, + "resnet_time_scale_shift": "default", + "sample_size": 64, + "time_cond_proj_dim": null, + "time_embedding_act_fn": null, + "time_embedding_dim": null, + "time_embedding_type": "positional", + "timestep_post_act": null, + "up_block_types": [ + "UpBlock2D", + "CrossAttnUpBlock2D", + "CrossAttnUpBlock2D", + "CrossAttnUpBlock2D" + ], + "upcast_attention": false, + "use_linear_projection": false +} diff --git a/checkpoint-2000/unet/diffusion_pytorch_model.bin b/checkpoint-2000/unet/diffusion_pytorch_model.bin new file mode 100644 index 0000000000000000000000000000000000000000..bab8f55b2c775fce45e093efb4ac1e36d32a8f63 --- /dev/null +++ b/checkpoint-2000/unet/diffusion_pytorch_model.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:751d5d12170c6a5a362a98a4790f5ba21996514b73846d1edfa4ad81b9e0500b +size 3438375973 diff --git a/checkpoint-2500/optimizer.bin b/checkpoint-2500/optimizer.bin new file mode 100644 index 0000000000000000000000000000000000000000..dafff385f4a7716462d35fedf05ad8b11d7de722 --- /dev/null +++ b/checkpoint-2500/optimizer.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4acb6dd08bd733703d7eb5bd4e82cee556c7643578e17402ee3e3489d28c32b6 +size 1725110981 diff --git a/checkpoint-2500/random_states_0.pkl b/checkpoint-2500/random_states_0.pkl new file mode 100644 index 0000000000000000000000000000000000000000..7c9596a568049861d822ec505ad283b04fa75918 --- /dev/null +++ b/checkpoint-2500/random_states_0.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cb9ce7a7462cdbd7ae17d58b6dae19bd37f122987f24c086b500084068204661 +size 14599 diff --git a/checkpoint-2500/scaler.pt b/checkpoint-2500/scaler.pt new file mode 100644 index 0000000000000000000000000000000000000000..9a3e6e139e26b18feddbe2f66ba98b73890dacfa --- /dev/null +++ b/checkpoint-2500/scaler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0fbcebc8f5487b0c117b5dd47f2ea304af3eebf408d297118d9307e1223927e1 +size 557 diff --git a/checkpoint-2500/scheduler.bin b/checkpoint-2500/scheduler.bin new file mode 100644 index 0000000000000000000000000000000000000000..033284ea96bb82e68b13955bb7490d461144fa03 --- /dev/null +++ b/checkpoint-2500/scheduler.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9f86c7ee8c22ef7b11ad1f3f144194c6b37d11ab6bfea37eb4b88a2cebb0f533 +size 563 diff --git a/checkpoint-2500/unet/config.json b/checkpoint-2500/unet/config.json new file mode 100644 index 0000000000000000000000000000000000000000..14e2e16ac7b0b32e261001f55127e59e4c91eee4 --- /dev/null +++ b/checkpoint-2500/unet/config.json @@ -0,0 +1,62 @@ +{ + "_class_name": "UNet2DConditionModel", + "_diffusers_version": "0.18.0.dev0", + "_name_or_path": "SG161222/Realistic_Vision_V2.0", + "act_fn": "silu", + "addition_embed_type": null, + "addition_embed_type_num_heads": 64, + "attention_head_dim": 8, + "block_out_channels": [ + 320, + 640, + 1280, + 1280 + ], + "center_input_sample": false, + "class_embed_type": null, + "class_embeddings_concat": false, + "conv_in_kernel": 3, + "conv_out_kernel": 3, + "cross_attention_dim": 768, + "cross_attention_norm": null, + "down_block_types": [ + "CrossAttnDownBlock2D", + "CrossAttnDownBlock2D", + "CrossAttnDownBlock2D", + "DownBlock2D" + ], + "downsample_padding": 1, + "dual_cross_attention": false, + "encoder_hid_dim": null, + "encoder_hid_dim_type": null, + "flip_sin_to_cos": true, + "freq_shift": 0, + "in_channels": 4, + "layers_per_block": 2, + "mid_block_only_cross_attention": null, + "mid_block_scale_factor": 1, + "mid_block_type": "UNetMidBlock2DCrossAttn", + "norm_eps": 1e-05, + "norm_num_groups": 32, + "num_class_embeds": null, + "only_cross_attention": false, + "out_channels": 4, + "projection_class_embeddings_input_dim": null, + "resnet_out_scale_factor": 1.0, + "resnet_skip_time_act": false, + "resnet_time_scale_shift": "default", + "sample_size": 64, + "time_cond_proj_dim": null, + "time_embedding_act_fn": null, + "time_embedding_dim": null, + "time_embedding_type": "positional", + "timestep_post_act": null, + "up_block_types": [ + "UpBlock2D", + "CrossAttnUpBlock2D", + "CrossAttnUpBlock2D", + "CrossAttnUpBlock2D" + ], + "upcast_attention": false, + "use_linear_projection": false +} diff --git a/checkpoint-2500/unet/diffusion_pytorch_model.bin b/checkpoint-2500/unet/diffusion_pytorch_model.bin new file mode 100644 index 0000000000000000000000000000000000000000..c26fa26d817a14a9b35bbcaf85955826f06ec829 --- /dev/null +++ b/checkpoint-2500/unet/diffusion_pytorch_model.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7f42153aa5dbc59b0428997fa7d67b531da29dd437ebbf45797d972a5c8f75bb +size 3438375973 diff --git a/checkpoint-3000/optimizer.bin b/checkpoint-3000/optimizer.bin new file mode 100644 index 0000000000000000000000000000000000000000..26780cdf71e92c0a13625445b8930bd8c9a19322 --- /dev/null +++ b/checkpoint-3000/optimizer.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c89a6fda3c35141a905e20ea494eb4e2dae3aac0e0d60476cef73dc9c8a04100 +size 1725110981 diff --git a/checkpoint-3000/random_states_0.pkl b/checkpoint-3000/random_states_0.pkl new file mode 100644 index 0000000000000000000000000000000000000000..219aa3d0e055837fbc85fa27f7f136382648c4de --- /dev/null +++ b/checkpoint-3000/random_states_0.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4c74bf482a236a3c474472f0cc46f7d6592a1aff7f7320597aafbd3f7958b0da +size 14599 diff --git a/checkpoint-3000/scaler.pt b/checkpoint-3000/scaler.pt new file mode 100644 index 0000000000000000000000000000000000000000..2888c9ddc2c30e6b47ce0e6abcc8e2d1303aab00 --- /dev/null +++ b/checkpoint-3000/scaler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fb1f9398b77268202e8e1465734a63d123b1ef11c27f20f2473677e9883a6869 +size 557 diff --git a/checkpoint-3000/scheduler.bin b/checkpoint-3000/scheduler.bin new file mode 100644 index 0000000000000000000000000000000000000000..7d5134e8679645cb88e5a97569eb8c09b0bba450 --- /dev/null +++ b/checkpoint-3000/scheduler.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:19ceff562f482452bd53931b8722ca4cc070752efe38953d42cbcb586e0edb0c +size 563 diff --git a/checkpoint-3000/unet/config.json b/checkpoint-3000/unet/config.json new file mode 100644 index 0000000000000000000000000000000000000000..14e2e16ac7b0b32e261001f55127e59e4c91eee4 --- /dev/null +++ b/checkpoint-3000/unet/config.json @@ -0,0 +1,62 @@ +{ + "_class_name": "UNet2DConditionModel", + "_diffusers_version": "0.18.0.dev0", + "_name_or_path": "SG161222/Realistic_Vision_V2.0", + "act_fn": "silu", + "addition_embed_type": null, + "addition_embed_type_num_heads": 64, + "attention_head_dim": 8, + "block_out_channels": [ + 320, + 640, + 1280, + 1280 + ], + "center_input_sample": false, + "class_embed_type": null, + "class_embeddings_concat": false, + "conv_in_kernel": 3, + "conv_out_kernel": 3, + "cross_attention_dim": 768, + "cross_attention_norm": null, + "down_block_types": [ + "CrossAttnDownBlock2D", + "CrossAttnDownBlock2D", + "CrossAttnDownBlock2D", + "DownBlock2D" + ], + "downsample_padding": 1, + "dual_cross_attention": false, + "encoder_hid_dim": null, + "encoder_hid_dim_type": null, + "flip_sin_to_cos": true, + "freq_shift": 0, + "in_channels": 4, + "layers_per_block": 2, + "mid_block_only_cross_attention": null, + "mid_block_scale_factor": 1, + "mid_block_type": "UNetMidBlock2DCrossAttn", + "norm_eps": 1e-05, + "norm_num_groups": 32, + "num_class_embeds": null, + "only_cross_attention": false, + "out_channels": 4, + "projection_class_embeddings_input_dim": null, + "resnet_out_scale_factor": 1.0, + "resnet_skip_time_act": false, + "resnet_time_scale_shift": "default", + "sample_size": 64, + "time_cond_proj_dim": null, + "time_embedding_act_fn": null, + "time_embedding_dim": null, + "time_embedding_type": "positional", + "timestep_post_act": null, + "up_block_types": [ + "UpBlock2D", + "CrossAttnUpBlock2D", + "CrossAttnUpBlock2D", + "CrossAttnUpBlock2D" + ], + "upcast_attention": false, + "use_linear_projection": false +} diff --git a/checkpoint-3000/unet/diffusion_pytorch_model.bin b/checkpoint-3000/unet/diffusion_pytorch_model.bin new file mode 100644 index 0000000000000000000000000000000000000000..2762e748d0002e53ffe7288f09e710ed86b0f5d4 --- /dev/null +++ b/checkpoint-3000/unet/diffusion_pytorch_model.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:473dde96cb8db41d0f28d5c1f6ed759ef87256164b67903644834c0a2569be2d +size 3438375973 diff --git a/checkpoint-500/optimizer.bin b/checkpoint-500/optimizer.bin index 233038879065d92b108285285b1a851e44a08304..c60fa186fff4880ae1001f9c021e17629931b1ac 100644 --- a/checkpoint-500/optimizer.bin +++ b/checkpoint-500/optimizer.bin @@ -1,3 +1,3 @@ version https://git-lfs.github.com/spec/v1 -oid sha256:785b7d7f3fd6d392a84a0550a611a16c8b6ffb6ddff0432ee0139e6e7d9e3b5b -size 1737915525 +oid sha256:20fb4d135cd9355da1ab0402c913fa7fab921d9d50235089d2a73205e1068f2a +size 1725110981 diff --git a/checkpoint-500/random_states_0.pkl b/checkpoint-500/random_states_0.pkl index 348370c2e4b099346a70636a403f4bda6dd8816a..cbb88c485d0575302bae22286486f6a8f7b72e7f 100644 --- a/checkpoint-500/random_states_0.pkl +++ b/checkpoint-500/random_states_0.pkl @@ -1,3 +1,3 @@ version https://git-lfs.github.com/spec/v1 -oid sha256:0fec3edb12446b346038168ad3fd8555081aad03690d2a4ae16e5aac8b2f01f5 -size 14727 +oid sha256:7e544c0a171c53f6cf9757e9a22c9457dfaf2560f81c1f539ba056b3d4ea67df +size 14599 diff --git a/checkpoint-500/scaler.pt b/checkpoint-500/scaler.pt new file mode 100644 index 0000000000000000000000000000000000000000..b5f2e5ee77c2cef5eff816170558ce9209855685 --- /dev/null +++ b/checkpoint-500/scaler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a3f196a54202bb4ba1220e8c59f42f9cda0702d68ea83147d814c2fb2f36b8f2 +size 557 diff --git a/checkpoint-500/scheduler.bin b/checkpoint-500/scheduler.bin index 503978a95aa230bbe200d6b52fb4d6ab51ce76a8..b4a8d576b93909bcc1990fb3825c085357ea5dda 100644 --- a/checkpoint-500/scheduler.bin +++ b/checkpoint-500/scheduler.bin @@ -1,3 +1,3 @@ version https://git-lfs.github.com/spec/v1 -oid sha256:73fd5f417d1b160455bac391e57c588efe13ffb8e7c4ae93ce40d27dbbb261b5 +oid sha256:5de3bfcb8841128532d1d7410237097e891dd5cf86bd915dbe26b18aabc645ef size 563 diff --git a/checkpoint-500/unet/config.json b/checkpoint-500/unet/config.json index 6389e554338110f210811f857aebeffe784d3f91..14e2e16ac7b0b32e261001f55127e59e4c91eee4 100644 --- a/checkpoint-500/unet/config.json +++ b/checkpoint-500/unet/config.json @@ -1,16 +1,11 @@ { "_class_name": "UNet2DConditionModel", "_diffusers_version": "0.18.0.dev0", - "_name_or_path": "stabilityai/stable-diffusion-2-1", + "_name_or_path": "SG161222/Realistic_Vision_V2.0", "act_fn": "silu", "addition_embed_type": null, "addition_embed_type_num_heads": 64, - "attention_head_dim": [ - 5, - 10, - 20, - 20 - ], + "attention_head_dim": 8, "block_out_channels": [ 320, 640, @@ -22,7 +17,7 @@ "class_embeddings_concat": false, "conv_in_kernel": 3, "conv_out_kernel": 3, - "cross_attention_dim": 1024, + "cross_attention_dim": 768, "cross_attention_norm": null, "down_block_types": [ "CrossAttnDownBlock2D", @@ -50,7 +45,7 @@ "resnet_out_scale_factor": 1.0, "resnet_skip_time_act": false, "resnet_time_scale_shift": "default", - "sample_size": 96, + "sample_size": 64, "time_cond_proj_dim": null, "time_embedding_act_fn": null, "time_embedding_dim": null, @@ -62,6 +57,6 @@ "CrossAttnUpBlock2D", "CrossAttnUpBlock2D" ], - "upcast_attention": true, - "use_linear_projection": true + "upcast_attention": false, + "use_linear_projection": false } diff --git a/checkpoint-500/unet/diffusion_pytorch_model.bin b/checkpoint-500/unet/diffusion_pytorch_model.bin index ead0182c7d12da266a932aa0555ff6b062980130..4898f36d5e29f6c953e46e6d22a44abbdaa62bf4 100644 --- a/checkpoint-500/unet/diffusion_pytorch_model.bin +++ b/checkpoint-500/unet/diffusion_pytorch_model.bin @@ -1,3 +1,3 @@ version https://git-lfs.github.com/spec/v1 -oid sha256:bd434f097b9690261cb34032dda1e29a58c1a73c89d142cbd0a57361bf8aac9c -size 3463934693 +oid sha256:e7073e8fff22ef13c3f4f73e64d1047b31ef1b2d159b474fa1450f402632a466 +size 3438375973 diff --git a/logs/dreambooth/1687163565.1471028/events.out.tfevents.1687163565.ip-172-31-28-104.eu-west-1.compute.internal.16942.1 b/logs/dreambooth/1687163565.1471028/events.out.tfevents.1687163565.ip-172-31-28-104.eu-west-1.compute.internal.16942.1 new file mode 100644 index 0000000000000000000000000000000000000000..2255fa1b6966e53da4437fa368ea0c044b15a88c --- /dev/null +++ b/logs/dreambooth/1687163565.1471028/events.out.tfevents.1687163565.ip-172-31-28-104.eu-west-1.compute.internal.16942.1 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8a2720d00db1dfe1c70daf844b32155b09d1da3619ccbff80754e1ff64fca462 +size 2815 diff --git a/logs/dreambooth/1687163565.1719146/hparams.yml b/logs/dreambooth/1687163565.1719146/hparams.yml new file mode 100644 index 0000000000000000000000000000000000000000..f60ba26d44ef60c7bd0d672920a1fa5d2e7b2f37 --- /dev/null +++ b/logs/dreambooth/1687163565.1719146/hparams.yml @@ -0,0 +1,58 @@ +adam_beta1: 0.9 +adam_beta2: 0.999 +adam_epsilon: 1.0e-08 +adam_weight_decay: 0.01 +allow_tf32: false +center_crop: false +checkpointing_steps: 500 +checkpoints_total_limit: null +class_data_dir: StableDiffusion-Regularization-Imgs/man +class_labels_conditioning: null +class_prompt: a photo of person +dataloader_num_workers: 0 +enable_xformers_memory_efficient_attention: false +gradient_accumulation_steps: 1 +gradient_checkpointing: false +hub_model_id: null +hub_token: null +instance_data_dir: training_images/ +instance_prompt: a photo of ohwx person +learning_rate: 1.0e-06 +local_rank: -1 +logging_dir: logs +lr_num_cycles: 1 +lr_power: 1.0 +lr_scheduler: constant +lr_warmup_steps: 0 +max_grad_norm: 1.0 +max_train_steps: 1200 +mixed_precision: null +num_class_images: 378 +num_train_epochs: 4 +num_validation_images: 4 +offset_noise: false +output_dir: model/ +pre_compute_text_embeddings: false +pretrained_model_name_or_path: stabilityai/stable-diffusion-2-1 +prior_generation_precision: null +prior_loss_weight: 1.0 +push_to_hub: true +report_to: tensorboard +resolution: 768 +resume_from_checkpoint: null +revision: null +sample_batch_size: 4 +scale_lr: false +seed: null +set_grads_to_none: false +skip_save_text_encoder: false +text_encoder_use_attention_mask: false +tokenizer_max_length: null +tokenizer_name: null +train_batch_size: 1 +train_text_encoder: false +use_8bit_adam: false +validation_images: null +validation_prompt: null +validation_steps: 100 +with_prior_preservation: true diff --git a/logs/dreambooth/1687164760.464152/events.out.tfevents.1687164760.ip-172-31-28-104.eu-west-1.compute.internal.25014.1 b/logs/dreambooth/1687164760.464152/events.out.tfevents.1687164760.ip-172-31-28-104.eu-west-1.compute.internal.25014.1 new file mode 100644 index 0000000000000000000000000000000000000000..65aabc81b78ecd92e838f0c400156cb56ad31cfa --- /dev/null +++ b/logs/dreambooth/1687164760.464152/events.out.tfevents.1687164760.ip-172-31-28-104.eu-west-1.compute.internal.25014.1 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a1d9e3fa9fc5355bab961fee890f27b7e63905f7e1f11db00a36662c9758a54e +size 2815 diff --git a/logs/dreambooth/1687164760.4885616/hparams.yml b/logs/dreambooth/1687164760.4885616/hparams.yml new file mode 100644 index 0000000000000000000000000000000000000000..f60ba26d44ef60c7bd0d672920a1fa5d2e7b2f37 --- /dev/null +++ b/logs/dreambooth/1687164760.4885616/hparams.yml @@ -0,0 +1,58 @@ +adam_beta1: 0.9 +adam_beta2: 0.999 +adam_epsilon: 1.0e-08 +adam_weight_decay: 0.01 +allow_tf32: false +center_crop: false +checkpointing_steps: 500 +checkpoints_total_limit: null +class_data_dir: StableDiffusion-Regularization-Imgs/man +class_labels_conditioning: null +class_prompt: a photo of person +dataloader_num_workers: 0 +enable_xformers_memory_efficient_attention: false +gradient_accumulation_steps: 1 +gradient_checkpointing: false +hub_model_id: null +hub_token: null +instance_data_dir: training_images/ +instance_prompt: a photo of ohwx person +learning_rate: 1.0e-06 +local_rank: -1 +logging_dir: logs +lr_num_cycles: 1 +lr_power: 1.0 +lr_scheduler: constant +lr_warmup_steps: 0 +max_grad_norm: 1.0 +max_train_steps: 1200 +mixed_precision: null +num_class_images: 378 +num_train_epochs: 4 +num_validation_images: 4 +offset_noise: false +output_dir: model/ +pre_compute_text_embeddings: false +pretrained_model_name_or_path: stabilityai/stable-diffusion-2-1 +prior_generation_precision: null +prior_loss_weight: 1.0 +push_to_hub: true +report_to: tensorboard +resolution: 768 +resume_from_checkpoint: null +revision: null +sample_batch_size: 4 +scale_lr: false +seed: null +set_grads_to_none: false +skip_save_text_encoder: false +text_encoder_use_attention_mask: false +tokenizer_max_length: null +tokenizer_name: null +train_batch_size: 1 +train_text_encoder: false +use_8bit_adam: false +validation_images: null +validation_prompt: null +validation_steps: 100 +with_prior_preservation: true diff --git a/logs/dreambooth/1690380698.735782/events.out.tfevents.1690380698.ip-172-31-28-104.eu-west-1.compute.internal.18483.1 b/logs/dreambooth/1690380698.735782/events.out.tfevents.1690380698.ip-172-31-28-104.eu-west-1.compute.internal.18483.1 new file mode 100644 index 0000000000000000000000000000000000000000..fe20088108f44baf40a6ce13dc2b4e57f10ae71b --- /dev/null +++ b/logs/dreambooth/1690380698.735782/events.out.tfevents.1690380698.ip-172-31-28-104.eu-west-1.compute.internal.18483.1 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9d23d27bc873296c454472c81a6531445418d4523a06c5be1d16d14d3b99d3cb +size 2800 diff --git a/logs/dreambooth/1690380698.7673254/hparams.yml b/logs/dreambooth/1690380698.7673254/hparams.yml new file mode 100644 index 0000000000000000000000000000000000000000..3ad16465040016ac2fbe4f24efea456ec6815ffe --- /dev/null +++ b/logs/dreambooth/1690380698.7673254/hparams.yml @@ -0,0 +1,58 @@ +adam_beta1: 0.9 +adam_beta2: 0.999 +adam_epsilon: 1.0e-08 +adam_weight_decay: 0.01 +allow_tf32: false +center_crop: false +checkpointing_steps: 500 +checkpoints_total_limit: null +class_data_dir: data-science/class-images/man-unsplash-2734 +class_labels_conditioning: null +class_prompt: a photo of man +dataloader_num_workers: 0 +enable_xformers_memory_efficient_attention: false +gradient_accumulation_steps: 1 +gradient_checkpointing: false +hub_model_id: null +hub_token: null +instance_data_dir: training_images/ +instance_prompt: ohwx man +learning_rate: 3.0e-06 +local_rank: -1 +logging_dir: logs +lr_num_cycles: 1 +lr_power: 1.0 +lr_scheduler: constant +lr_warmup_steps: 0 +max_grad_norm: 1.0 +max_train_steps: 3000 +mixed_precision: null +num_class_images: 100 +num_train_epochs: 30 +num_validation_images: 4 +offset_noise: false +output_dir: model/ +pre_compute_text_embeddings: false +pretrained_model_name_or_path: SG161222/Realistic_Vision_V2.0 +prior_generation_precision: null +prior_loss_weight: 1.0 +push_to_hub: true +report_to: tensorboard +resolution: 512 +resume_from_checkpoint: null +revision: null +sample_batch_size: 4 +scale_lr: false +seed: null +set_grads_to_none: true +skip_save_text_encoder: false +text_encoder_use_attention_mask: false +tokenizer_max_length: null +tokenizer_name: null +train_batch_size: 1 +train_text_encoder: false +use_8bit_adam: true +validation_images: null +validation_prompt: null +validation_steps: 100 +with_prior_preservation: true diff --git a/logs/dreambooth/events.out.tfevents.1687163565.ip-172-31-28-104.eu-west-1.compute.internal.16942.0 b/logs/dreambooth/events.out.tfevents.1687163565.ip-172-31-28-104.eu-west-1.compute.internal.16942.0 new file mode 100644 index 0000000000000000000000000000000000000000..10e66962a08e3fb77e0622262dd607673484849b --- /dev/null +++ b/logs/dreambooth/events.out.tfevents.1687163565.ip-172-31-28-104.eu-west-1.compute.internal.16942.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2823b6844e7c1bd9e6f2f0dc16a5b7e637fb67771cab6ff9277507c09de99aad +size 88 diff --git a/logs/dreambooth/events.out.tfevents.1687164760.ip-172-31-28-104.eu-west-1.compute.internal.25014.0 b/logs/dreambooth/events.out.tfevents.1687164760.ip-172-31-28-104.eu-west-1.compute.internal.25014.0 new file mode 100644 index 0000000000000000000000000000000000000000..6c0211063e9aea510beadf6b9865e84f692e6f57 --- /dev/null +++ b/logs/dreambooth/events.out.tfevents.1687164760.ip-172-31-28-104.eu-west-1.compute.internal.25014.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3728a16a0af0409d37a013bbe17978191439a315c4f895c25646a6ece1c6a336 +size 170 diff --git a/logs/dreambooth/events.out.tfevents.1690380698.ip-172-31-28-104.eu-west-1.compute.internal.18483.0 b/logs/dreambooth/events.out.tfevents.1690380698.ip-172-31-28-104.eu-west-1.compute.internal.18483.0 new file mode 100644 index 0000000000000000000000000000000000000000..f735fa4cff108e85918deea3932d2f23a673e6ad --- /dev/null +++ b/logs/dreambooth/events.out.tfevents.1690380698.ip-172-31-28-104.eu-west-1.compute.internal.18483.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc27b474d5e8eddc7e5d52760cfe96441eb43e8c06e69679f20e002f7a2deccb +size 251834 diff --git a/model_index.json b/model_index.json index c84a8aaebdb8d345b57b496c207db98aac10f6a7..3afc1ce68d1c38e2e75aa48ec97d6810f288afe2 100644 --- a/model_index.json +++ b/model_index.json @@ -5,14 +5,14 @@ "transformers", "CLIPImageProcessor" ], - "requires_safety_checker": false, + "requires_safety_checker": true, "safety_checker": [ - null, - null + "stable_diffusion", + "StableDiffusionSafetyChecker" ], "scheduler": [ "diffusers", - "DDIMScheduler" + "PNDMScheduler" ], "text_encoder": [ "transformers", diff --git a/safety_checker/config.json b/safety_checker/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f2ffbd0baf8a48774bb8cd6dd24c95e2b115a479 --- /dev/null +++ b/safety_checker/config.json @@ -0,0 +1,168 @@ +{ + "_commit_hash": "60c57a68e17cc9261e6031cb3babd7ffd6d6dde1", + "_name_or_path": "/home/ec2-user/.cache/huggingface/hub/models--SG161222--Realistic_Vision_V2.0/snapshots/60c57a68e17cc9261e6031cb3babd7ffd6d6dde1/safety_checker", + "architectures": [ + "StableDiffusionSafetyChecker" + ], + "initializer_factor": 1.0, + "logit_scale_init_value": 2.6592, + "model_type": "clip", + "projection_dim": 768, + "text_config": { + "_name_or_path": "", + "add_cross_attention": false, + "architectures": null, + "attention_dropout": 0.0, + "bad_words_ids": null, + "begin_suppress_tokens": null, + "bos_token_id": 0, + "chunk_size_feed_forward": 0, + "cross_attention_hidden_size": null, + "decoder_start_token_id": null, + "diversity_penalty": 0.0, + "do_sample": false, + "dropout": 0.0, + "early_stopping": false, + "encoder_no_repeat_ngram_size": 0, + "eos_token_id": 2, + "exponential_decay_length_penalty": null, + "finetuning_task": null, + "forced_bos_token_id": null, + "forced_eos_token_id": null, + "hidden_act": "quick_gelu", + "hidden_size": 768, + "id2label": { + "0": "LABEL_0", + "1": "LABEL_1" + }, + "initializer_factor": 1.0, + "initializer_range": 0.02, + "intermediate_size": 3072, + "is_decoder": false, + "is_encoder_decoder": false, + "label2id": { + "LABEL_0": 0, + "LABEL_1": 1 + }, + "layer_norm_eps": 1e-05, + "length_penalty": 1.0, + "max_length": 20, + "max_position_embeddings": 77, + "min_length": 0, + "model_type": "clip_text_model", + "no_repeat_ngram_size": 0, + "num_attention_heads": 12, + "num_beam_groups": 1, + "num_beams": 1, + "num_hidden_layers": 12, + "num_return_sequences": 1, + "output_attentions": false, + "output_hidden_states": false, + "output_scores": false, + "pad_token_id": 1, + "prefix": null, + "problem_type": null, + "projection_dim": 512, + "pruned_heads": {}, + "remove_invalid_values": false, + "repetition_penalty": 1.0, + "return_dict": true, + "return_dict_in_generate": false, + "sep_token_id": null, + "suppress_tokens": null, + "task_specific_params": null, + "temperature": 1.0, + "tf_legacy_loss": false, + "tie_encoder_decoder": false, + "tie_word_embeddings": true, + "tokenizer_class": null, + "top_k": 50, + "top_p": 1.0, + "torch_dtype": null, + "torchscript": false, + "transformers_version": "4.30.1", + "typical_p": 1.0, + "use_bfloat16": false, + "vocab_size": 49408 + }, + "torch_dtype": "float32", + "transformers_version": null, + "vision_config": { + "_name_or_path": "", + "add_cross_attention": false, + "architectures": null, + "attention_dropout": 0.0, + "bad_words_ids": null, + "begin_suppress_tokens": null, + "bos_token_id": null, + "chunk_size_feed_forward": 0, + "cross_attention_hidden_size": null, + "decoder_start_token_id": null, + "diversity_penalty": 0.0, + "do_sample": false, + "dropout": 0.0, + "early_stopping": false, + "encoder_no_repeat_ngram_size": 0, + "eos_token_id": null, + "exponential_decay_length_penalty": null, + "finetuning_task": null, + "forced_bos_token_id": null, + "forced_eos_token_id": null, + "hidden_act": "quick_gelu", + "hidden_size": 1024, + "id2label": { + "0": "LABEL_0", + "1": "LABEL_1" + }, + "image_size": 224, + "initializer_factor": 1.0, + "initializer_range": 0.02, + "intermediate_size": 4096, + "is_decoder": false, + "is_encoder_decoder": false, + "label2id": { + "LABEL_0": 0, + "LABEL_1": 1 + }, + "layer_norm_eps": 1e-05, + "length_penalty": 1.0, + "max_length": 20, + "min_length": 0, + "model_type": "clip_vision_model", + "no_repeat_ngram_size": 0, + "num_attention_heads": 16, + "num_beam_groups": 1, + "num_beams": 1, + "num_channels": 3, + "num_hidden_layers": 24, + "num_return_sequences": 1, + "output_attentions": false, + "output_hidden_states": false, + "output_scores": false, + "pad_token_id": null, + "patch_size": 14, + "prefix": null, + "problem_type": null, + "projection_dim": 512, + "pruned_heads": {}, + "remove_invalid_values": false, + "repetition_penalty": 1.0, + "return_dict": true, + "return_dict_in_generate": false, + "sep_token_id": null, + "suppress_tokens": null, + "task_specific_params": null, + "temperature": 1.0, + "tf_legacy_loss": false, + "tie_encoder_decoder": false, + "tie_word_embeddings": true, + "tokenizer_class": null, + "top_k": 50, + "top_p": 1.0, + "torch_dtype": null, + "torchscript": false, + "transformers_version": "4.30.1", + "typical_p": 1.0, + "use_bfloat16": false + } +} diff --git a/safety_checker/pytorch_model.bin b/safety_checker/pytorch_model.bin new file mode 100644 index 0000000000000000000000000000000000000000..c8835557a0d3af583cb06c7c154b7e54a069c41d --- /dev/null +++ b/safety_checker/pytorch_model.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:16d28f2b37109f222cdc33620fdd262102ac32112be0352a7f77e9614b35a394 +size 1216064769 diff --git a/scheduler/scheduler_config.json b/scheduler/scheduler_config.json index ec16fc0603a89e4814f05c3fc14478f076280056..ddff71b94f6990b371337c0864bf40a972de5c8b 100644 --- a/scheduler/scheduler_config.json +++ b/scheduler/scheduler_config.json @@ -1,20 +1,14 @@ { - "_class_name": "DDIMScheduler", + "_class_name": "PNDMScheduler", "_diffusers_version": "0.18.0.dev0", "beta_end": 0.012, "beta_schedule": "scaled_linear", "beta_start": 0.00085, "clip_sample": false, - "clip_sample_range": 1.0, - "dynamic_thresholding_ratio": 0.995, "num_train_timesteps": 1000, - "prediction_type": "v_prediction", - "rescale_betas_zero_snr": false, - "sample_max_value": 1.0, + "prediction_type": "epsilon", "set_alpha_to_one": false, "skip_prk_steps": true, "steps_offset": 1, - "thresholding": false, - "timestep_spacing": "leading", "trained_betas": null } diff --git a/text_encoder/config.json b/text_encoder/config.json index 326741d84a35be1642c320536ea1d91b86e01142..d2c236d3935bdbd493e6d3cfc482e975cce31df7 100644 --- a/text_encoder/config.json +++ b/text_encoder/config.json @@ -1,5 +1,5 @@ { - "_name_or_path": "stabilityai/stable-diffusion-2-1", + "_name_or_path": "SG161222/Realistic_Vision_V2.0", "architectures": [ "CLIPTextModel" ], @@ -7,19 +7,19 @@ "bos_token_id": 0, "dropout": 0.0, "eos_token_id": 2, - "hidden_act": "gelu", - "hidden_size": 1024, + "hidden_act": "quick_gelu", + "hidden_size": 768, "initializer_factor": 1.0, "initializer_range": 0.02, - "intermediate_size": 4096, + "intermediate_size": 3072, "layer_norm_eps": 1e-05, "max_position_embeddings": 77, "model_type": "clip_text_model", - "num_attention_heads": 16, - "num_hidden_layers": 23, + "num_attention_heads": 12, + "num_hidden_layers": 12, "pad_token_id": 1, - "projection_dim": 512, - "torch_dtype": "float32", - "transformers_version": "4.30.2", + "projection_dim": 768, + "torch_dtype": "float16", + "transformers_version": "4.30.1", "vocab_size": 49408 } diff --git a/text_encoder/pytorch_model.bin b/text_encoder/pytorch_model.bin index a4ccd982de47235ab9475a5b6e19c4ddd34ed94f..c878fa83a48c685a3b161fd1914c2a8ffb39054e 100644 --- a/text_encoder/pytorch_model.bin +++ b/text_encoder/pytorch_model.bin @@ -1,3 +1,3 @@ version https://git-lfs.github.com/spec/v1 -oid sha256:e9c787e9388134c1a25dc69934a51a32a2683b38b8a9b017e1f3a692b8ed6b98 -size 1361679905 +oid sha256:4f3b80fe3f244f6136effd16df75196dba6060369bc4c88acbe63a9c84f6d395 +size 246188833 diff --git a/tokenizer/special_tokens_map.json b/tokenizer/special_tokens_map.json index ae0c5be6f35217e51c4c000fd325d8de0294e99c..2c2130b544c0c5a72d5d00da071ba130a9800fb2 100644 --- a/tokenizer/special_tokens_map.json +++ b/tokenizer/special_tokens_map.json @@ -13,7 +13,7 @@ "rstrip": false, "single_word": false }, - "pad_token": "!", + "pad_token": "<|endoftext|>", "unk_token": { "content": "<|endoftext|>", "lstrip": false, diff --git a/unet/config.json b/unet/config.json index 6389e554338110f210811f857aebeffe784d3f91..14e2e16ac7b0b32e261001f55127e59e4c91eee4 100644 --- a/unet/config.json +++ b/unet/config.json @@ -1,16 +1,11 @@ { "_class_name": "UNet2DConditionModel", "_diffusers_version": "0.18.0.dev0", - "_name_or_path": "stabilityai/stable-diffusion-2-1", + "_name_or_path": "SG161222/Realistic_Vision_V2.0", "act_fn": "silu", "addition_embed_type": null, "addition_embed_type_num_heads": 64, - "attention_head_dim": [ - 5, - 10, - 20, - 20 - ], + "attention_head_dim": 8, "block_out_channels": [ 320, 640, @@ -22,7 +17,7 @@ "class_embeddings_concat": false, "conv_in_kernel": 3, "conv_out_kernel": 3, - "cross_attention_dim": 1024, + "cross_attention_dim": 768, "cross_attention_norm": null, "down_block_types": [ "CrossAttnDownBlock2D", @@ -50,7 +45,7 @@ "resnet_out_scale_factor": 1.0, "resnet_skip_time_act": false, "resnet_time_scale_shift": "default", - "sample_size": 96, + "sample_size": 64, "time_cond_proj_dim": null, "time_embedding_act_fn": null, "time_embedding_dim": null, @@ -62,6 +57,6 @@ "CrossAttnUpBlock2D", "CrossAttnUpBlock2D" ], - "upcast_attention": true, - "use_linear_projection": true + "upcast_attention": false, + "use_linear_projection": false } diff --git a/unet/diffusion_pytorch_model.bin b/unet/diffusion_pytorch_model.bin index ab021d961393c39c7fa4c1bb9d6acb6fdc7b8329..2762e748d0002e53ffe7288f09e710ed86b0f5d4 100644 --- a/unet/diffusion_pytorch_model.bin +++ b/unet/diffusion_pytorch_model.bin @@ -1,3 +1,3 @@ version https://git-lfs.github.com/spec/v1 -oid sha256:7f583381f72cf608a92c8b77a082a8200c6936219743d1b57a69ebffcca14078 -size 3463934693 +oid sha256:473dde96cb8db41d0f28d5c1f6ed759ef87256164b67903644834c0a2569be2d +size 3438375973 diff --git a/vae/config.json b/vae/config.json index 837b4d1318ae6ab5063cb4396bd87c8e13b5e03c..1421ac7a7a31f37dfdf366c4f1ea2bef54914965 100644 --- a/vae/config.json +++ b/vae/config.json @@ -1,7 +1,7 @@ { "_class_name": "AutoencoderKL", "_diffusers_version": "0.18.0.dev0", - "_name_or_path": "/home/ubuntu/.cache/huggingface/hub/models--stabilityai--stable-diffusion-2-1/snapshots/845609e6cf0a060d8cd837297e5c169df5bff72c/vae", + "_name_or_path": "/home/ec2-user/.cache/huggingface/hub/models--SG161222--Realistic_Vision_V2.0/snapshots/60c57a68e17cc9261e6031cb3babd7ffd6d6dde1/vae", "act_fn": "silu", "block_out_channels": [ 128, @@ -20,7 +20,7 @@ "layers_per_block": 2, "norm_num_groups": 32, "out_channels": 3, - "sample_size": 768, + "sample_size": 512, "scaling_factor": 0.18215, "up_block_types": [ "UpDecoderBlock2D", diff --git a/vae/diffusion_pytorch_model.bin b/vae/diffusion_pytorch_model.bin index 66e5e70dec004f9bddcfa088b1cec180c1ff97e7..455ed67dbdcf79e2c64fe667f4d247fc74521ea8 100644 --- a/vae/diffusion_pytorch_model.bin +++ b/vae/diffusion_pytorch_model.bin @@ -1,3 +1,3 @@ version https://git-lfs.github.com/spec/v1 -oid sha256:1b01618945554d9840701d3453d4a9fe3db0db090164a5ed6305641306285b6f +oid sha256:251086e7c7793410070d11a421db3886a7b2a7ff27cdea006a67a4cd76a7a899 size 334712113