thanks to text_encoder ❤
Browse files
transformer/config.json
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "CogVideoXTransformer3DModel",
|
| 3 |
+
"_diffusers_version": "0.30.1",
|
| 4 |
+
"_name_or_path": "/kaggle/working/CogVideoX-5b",
|
| 5 |
+
"activation_fn": "gelu-approximate",
|
| 6 |
+
"attention_bias": true,
|
| 7 |
+
"attention_head_dim": 64,
|
| 8 |
+
"dropout": 0.0,
|
| 9 |
+
"flip_sin_to_cos": true,
|
| 10 |
+
"freq_shift": 0,
|
| 11 |
+
"in_channels": 16,
|
| 12 |
+
"max_text_seq_length": 226,
|
| 13 |
+
"norm_elementwise_affine": true,
|
| 14 |
+
"norm_eps": 1e-05,
|
| 15 |
+
"num_attention_heads": 48,
|
| 16 |
+
"num_layers": 42,
|
| 17 |
+
"out_channels": 16,
|
| 18 |
+
"patch_size": 2,
|
| 19 |
+
"sample_frames": 49,
|
| 20 |
+
"sample_height": 60,
|
| 21 |
+
"sample_width": 90,
|
| 22 |
+
"spatial_interpolation_scale": 1.875,
|
| 23 |
+
"temporal_compression_ratio": 4,
|
| 24 |
+
"temporal_interpolation_scale": 1.0,
|
| 25 |
+
"text_embed_dim": 4096,
|
| 26 |
+
"time_embed_dim": 512,
|
| 27 |
+
"timestep_activation_fn": "silu",
|
| 28 |
+
"use_rotary_positional_embeddings": true
|
| 29 |
+
}
|
transformer/diffusion_pytorch_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:33dccb517bdef035b7fe4bee4043032266078ed5adca396341d026b61e249140
|
| 3 |
+
size 5570402552
|