add bf16 weights for flax, remove pt weights

model_index.json

@@ -7,15 +7,15 @@
   ],
   "safety_checker": [
     "stable_diffusion",
-    "StableDiffusionSafetyChecker"
+    "FlaxStableDiffusionSafetyChecker"
   ],
   "scheduler": [
     "diffusers",
-    "PNDMScheduler"
+    "FlaxPNDMScheduler"
   ],
   "text_encoder": [
     "transformers",
-    "CLIPTextModel"
+    "FlaxCLIPTextModel"
   ],
   "tokenizer": [
     "transformers",
@@ -23,10 +23,10 @@
   ],
   "unet": [
     "diffusers",
-    "UNet2DConditionModel"
+    "FlaxUNet2DConditionModel"
   ],
   "vae": [
     "diffusers",
-    "AutoencoderKL"
+    "FlaxAutoencoderKL"
   ]
 }

safety_checker/config.json

@@ -1,5 +1,6 @@
 {
-  "_name_or_path": "./safety_module",
+  "_commit_hash": null,
+  "_name_or_path": "fusing/sd-v1-4",
   "architectures": [
     "StableDiffusionSafetyChecker"
   ],
@@ -13,6 +14,7 @@
     "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,
@@ -66,8 +68,10 @@
     "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,
@@ -75,7 +79,7 @@
     "top_p": 1.0,
     "torch_dtype": null,
     "torchscript": false,
-    "transformers_version": "4.21.0.dev0",
+    "transformers_version": "4.23.1",
     "typical_p": 1.0,
     "use_bfloat16": false,
     "vocab_size": 49408
@@ -94,6 +98,7 @@
     "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,
@@ -133,6 +138,7 @@
     "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,
@@ -148,8 +154,10 @@
     "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,
@@ -157,7 +165,7 @@
     "top_p": 1.0,
     "torch_dtype": null,
     "torchscript": false,
-    "transformers_version": "4.21.0.dev0",
+    "transformers_version": "4.23.1",
     "typical_p": 1.0,
     "use_bfloat16": false
   },

text_encoder/pytorch_model.bin → safety_checker/flax_model.msgpack

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:770a47a9ffdcfda0b05506a7888ed714d06131d60267e6cf52765d61cf59fd67
-size 492305335
+oid sha256:1810a63be161ad95baecafae311c5d5cd8264ac0e2a357226d63240710759299
+size 607977215

text_encoder/config.json

@@ -19,6 +19,6 @@
   "num_hidden_layers": 12,
   "pad_token_id": 1,
   "torch_dtype": "float32",
-  "transformers_version": "4.21.0.dev0",
+  "transformers_version": "4.23.1",
   "vocab_size": 49408
 }

vae/diffusion_pytorch_model.bin → text_encoder/flax_model.msgpack

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1b134cded8eb78b184aefb8805b6b572f36fa77b255c483665dda931fa0130c5
-size 334707217
+oid sha256:38aca8345bad31819530d6291463fe6b38b07275ae38065f774b30e7a8e07d64
+size 246127918

unet/config.json

@@ -1,7 +1,6 @@
 {
-  "_class_name": "UNet2DConditionModel",
-  "_diffusers_version": "0.2.2",
-  "act_fn": "silu",
+  "_class_name": "FlaxUNet2DConditionModel",
+  "_diffusers_version": "0.5.0.dev0",
   "attention_head_dim": 8,
   "block_out_channels": [
     320,
@@ -9,7 +8,6 @@
     1280,
     1280
   ],
-  "center_input_sample": false,
   "cross_attention_dim": 768,
   "down_block_types": [
     "CrossAttnDownBlock2D",
@@ -17,14 +15,10 @@
     "CrossAttnDownBlock2D",
     "DownBlock2D"
   ],
-  "downsample_padding": 1,
-  "flip_sin_to_cos": true,
+  "dropout": 0.0,
   "freq_shift": 0,
   "in_channels": 4,
   "layers_per_block": 2,
-  "mid_block_scale_factor": 1,
-  "norm_eps": 1e-05,
-  "norm_num_groups": 32,
   "out_channels": 4,
   "sample_size": 64,
   "up_block_types": [

safety_checker/pytorch_model.bin → unet/diffusion_flax_model.msgpack

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:193490b58ef62739077262e833bf091c66c29488058681ac25cf7df3d8190974
-size 1216061799
+oid sha256:24fe3c11852cd27f6825387902e79ee044977b59654ae845ed99c795dc499a41
+size 1719067125

vae/config.json

@@ -1,6 +1,6 @@
 {
-  "_class_name": "AutoencoderKL",
-  "_diffusers_version": "0.2.2",
+  "_class_name": "FlaxAutoencoderKL",
+  "_diffusers_version": "0.5.0.dev0",
   "act_fn": "silu",
   "block_out_channels": [
     128,
@@ -17,6 +17,7 @@
   "in_channels": 3,
   "latent_channels": 4,
   "layers_per_block": 2,
+  "norm_num_groups": 32,
   "out_channels": 3,
   "sample_size": 512,
   "up_block_types": [

unet/diffusion_pytorch_model.bin → vae/diffusion_flax_model.msgpack

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:62d48b4d841a3178511fa453df0dae59b22089ace64609cc9d5353d0a7f37c26
-size 3438354725
+oid sha256:0e3e81680c6669960f086eca278c8a1f89dd2ed7eadef806e049b56e426c1809
+size 167316369