End of training

README.md

@@ -37,12 +37,12 @@ image.save("my_image.png")
 
 These are the key hyperparameters used during training:
 
-* Epochs: 30
+* Epochs: 20
 * Learning rate: 0.00025
 * Batch size: 128
-* Gradient accumulation steps: 32
+* Gradient accumulation steps: 4
 * Image resolution: 128
 * Mixed-precision: bf16
 
 
-More information on all the CLI arguments and the environment are available on your [`wandb` run page](https://wandb.ai/suvadityamuk/sakana-kanji/runs/of44igkm).
+More information on all the CLI arguments and the environment are available on your [`wandb` run page](https://wandb.ai/suvadityamuk/sakana-kanji/runs/ymtm4e77).

feature_extractor/preprocessor_config.json

@@ -8,7 +8,6 @@
   "do_normalize": true,
   "do_rescale": true,
   "do_resize": true,
-  "feature_extractor_type": "CLIPFeatureExtractor",
   "image_mean": [
     0.48145466,
     0.4578275,

text_encoder/config.json

@@ -20,6 +20,6 @@
   "pad_token_id": 1,
   "projection_dim": 512,
   "torch_dtype": "bfloat16",
-  "transformers_version": "4.36.1",
+  "transformers_version": "4.37.2",
   "vocab_size": 49408
 }

unet/diffusion_pytorch_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2d0e6bc33a9d379da597ba61e30d58e487c87b4a5be64262d495597e87d71ae7
+oid sha256:cc235bfc08ec00d752a1a9ae20340aea862036cbbeace7461b95663cb69ffa1f
 size 3463726504