Hugging Face's logo

optimum-intel-internal-testing
/
tiny-random-ltx2.5

Diffusers
Safetensors
LTX2Pipeline
Model card Files
xet
 Community
1
h0witended commited on
Commit
de46413
·
verified ·
1 Parent(s): 3f43040

Add reference OpenVINO IRs

Browse files

Generated with:
- transformers==5.15.1
- optimum-intel==2.3.0.dev0+8c40189b
- openvino==2026.5.0-23697

Files changed (46) hide show
  1. audio_vae/config.json +0 -23
  2. audio_vae/diffusion_pytorch_model.safetensors → audio_vae_decoder/openvino_model.bin +2 -2
  3. audio_vae_decoder/openvino_model.xml +3205 -0
  4. connectors/config.json +0 -25
  5. transformer/diffusion_pytorch_model.safetensors → connectors/openvino_model.bin +2 -2
  6. connectors/openvino_model.xml +0 -0
  7. duration_head/config.json +0 -10
  8. duration_head/{diffusion_pytorch_model.safetensors → openvino_model.bin} +2 -2
  9. duration_head/openvino_model.xml +1010 -0
  10. metadata.json +22 -0
  11. model_index.json +0 -48
  12. processor/chat_template.jinja +0 -386
  13. processor/processor_config.json +0 -75
  14. processor/tokenizer_config.json +0 -141
  15. prompt_enhancer/config.json +0 -159
  16. prompt_enhancer/generation_config.json +0 -13
  17. prompt_enhancer/model.safetensors +0 -3
  18. text_encoder/model.safetensors → prompt_enhancer/openvino_audio_embeddings_model.bin +2 -2
  19. prompt_enhancer/openvino_audio_embeddings_model.xml +0 -0
  20. processor/tokenizer.json → prompt_enhancer/openvino_language_model.bin +2 -2
  21. prompt_enhancer/openvino_language_model.xml +0 -0
  22. prompt_enhancer/openvino_text_embeddings_model.bin +3 -0
  23. prompt_enhancer/openvino_text_embeddings_model.xml +129 -0
  24. prompt_enhancer/openvino_text_embeddings_per_layer_model.bin +3 -0
  25. prompt_enhancer/openvino_text_embeddings_per_layer_model.xml +513 -0
  26. prompt_enhancer/openvino_vision_embeddings_model.bin +3 -0
  27. prompt_enhancer/openvino_vision_embeddings_model.xml +0 -0
  28. scheduler/scheduler_config.json +0 -18
  29. text_encoder/config.json +0 -122
  30. text_encoder/generation_config.json +0 -13
  31. connectors/diffusion_pytorch_model.safetensors → text_encoder/openvino_model.bin +2 -2
  32. text_encoder/openvino_model.xml +0 -0
  33. tiny_model_marker.json +0 -13
  34. tokenizer/tokenizer.json +0 -1148
  35. tokenizer/tokenizer_config.json +0 -12
  36. transformer/config.json +0 -54
  37. transformer/openvino_model.bin +3 -0
  38. transformer/openvino_model.xml +0 -0
  39. vae/config.json +0 -366
  40. vae/diffusion_pytorch_model.safetensors +0 -3
  41. vae_decoder/openvino_model.bin +3 -0
  42. vae_decoder/openvino_model.xml +0 -0
  43. vocoder/config.json +0 -80
  44. vocoder/diffusion_pytorch_model.safetensors +0 -3
  45. vocoder/openvino_model.bin +3 -0
  46. vocoder/openvino_model.xml +0 -0
audio_vae/config.json DELETED
@@ -1,23 +0,0 @@
1
- {
2
- "_class_name": "AutoencoderKLLTX2Audio",
3
- "_diffusers_version": "0.40.0",
4
- "attn_resolutions": null,
5
- "base_channels": 128,
6
- "causality_axis": "height",
7
- "ch_mult": [
8
- 1
9
- ],
10
- "double_z": true,
11
- "dropout": 0.0,
12
- "in_channels": 2,
13
- "is_causal": true,
14
- "latent_channels": 8,
15
- "mel_bins": 64,
16
- "mel_hop_length": 160,
17
- "mid_block_add_attention": false,
18
- "norm_type": "pixel",
19
- "num_res_blocks": 1,
20
- "output_channels": 2,
21
- "resolution": 256,
22
- "sample_rate": 16000
23
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
audio_vae/diffusion_pytorch_model.safetensors → audio_vae_decoder/openvino_model.bin RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4879d6b1157d97f25ba9c8dc8b74258892f7fd3c43447e6ab329e4db571a237f
3
- size 8399760
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:42299eca5b6f08a5b6309747cb1108e20e3ce7445bf122d74a13c1c2579b1c7f
3
+ size 4769496
audio_vae_decoder/openvino_model.xml ADDED
@@ -0,0 +1,3205 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ <?xml version="1.0"?>
2
+ <net name="Model33" version="11">
3
+ <layers>
4
+ <layer id="0" name="latent_sample" type="Parameter" version="opset1">
5
+ <data shape="?,?,?,?" element_type="f32" />
6
+ <output>
7
+ <port id="0" precision="FP32" names="latent_sample">
8
+ <dim>-1</dim>
9
+ <dim>-1</dim>
10
+ <dim>-1</dim>
11
+ <dim>-1</dim>
12
+ </port>
13
+ </output>
14
+ </layer>
15
+ <layer id="1" name="self.decoder.conv_in.conv.weight" type="Const" version="opset1">
16
+ <data element_type="f32" shape="128, 8, 3, 3" offset="0" size="36864" />
17
+ <output>
18
+ <port id="0" precision="FP32" names="self.decoder.conv_in.conv.weight">
19
+ <dim>128</dim>
20
+ <dim>8</dim>
21
+ <dim>3</dim>
22
+ <dim>3</dim>
23
+ </port>
24
+ </output>
25
+ </layer>
26
+ <layer id="2" name="__module.decoder.conv_in.conv/aten::_convolution/Convolution" type="Convolution" version="opset1">
27
+ <data strides="1, 1" dilations="1, 1" pads_begin="2, 1" pads_end="0, 1" auto_pad="explicit" />
28
+ <input>
29
+ <port id="0" precision="FP32">
30
+ <dim>-1</dim>
31
+ <dim>-1</dim>
32
+ <dim>-1</dim>
33
+ <dim>-1</dim>
34
+ </port>
35
+ <port id="1" precision="FP32">
36
+ <dim>128</dim>
37
+ <dim>8</dim>
38
+ <dim>3</dim>
39
+ <dim>3</dim>
40
+ </port>
41
+ </input>
42
+ <output>
43
+ <port id="2" precision="FP32">
44
+ <dim>-1</dim>
45
+ <dim>128</dim>
46
+ <dim>-1</dim>
47
+ <dim>-1</dim>
48
+ </port>
49
+ </output>
50
+ </layer>
51
+ <layer id="3" name="__module.decoder.conv_in.conv/aten::_convolution/Reshape" type="Const" version="opset1">
52
+ <data element_type="f32" shape="1, 128, 1, 1" offset="36864" size="512" />
53
+ <output>
54
+ <port id="0" precision="FP32">
55
+ <dim>1</dim>
56
+ <dim>128</dim>
57
+ <dim>1</dim>
58
+ <dim>1</dim>
59
+ </port>
60
+ </output>
61
+ </layer>
62
+ <layer id="4" name="__module.decoder.conv_in.conv/aten::_convolution/Add" type="Add" version="opset1">
63
+ <data auto_broadcast="numpy" />
64
+ <input>
65
+ <port id="0" precision="FP32">
66
+ <dim>-1</dim>
67
+ <dim>128</dim>
68
+ <dim>-1</dim>
69
+ <dim>-1</dim>
70
+ </port>
71
+ <port id="1" precision="FP32">
72
+ <dim>1</dim>
73
+ <dim>128</dim>
74
+ <dim>1</dim>
75
+ <dim>1</dim>
76
+ </port>
77
+ </input>
78
+ <output>
79
+ <port id="2" precision="FP32" names="49,x.1">
80
+ <dim>-1</dim>
81
+ <dim>128</dim>
82
+ <dim>-1</dim>
83
+ <dim>-1</dim>
84
+ </port>
85
+ </output>
86
+ </layer>
87
+ <layer id="5" name="Constant_203646" type="Const" version="opset1">
88
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37376" size="4" />
89
+ <output>
90
+ <port id="0" precision="FP32">
91
+ <dim>1</dim>
92
+ <dim>1</dim>
93
+ <dim>1</dim>
94
+ <dim>1</dim>
95
+ </port>
96
+ </output>
97
+ </layer>
98
+ <layer id="6" name="__module.decoder.mid.block_1.norm1/aten::pow/Power" type="Power" version="opset1">
99
+ <data auto_broadcast="numpy" />
100
+ <input>
101
+ <port id="0" precision="FP32">
102
+ <dim>-1</dim>
103
+ <dim>128</dim>
104
+ <dim>-1</dim>
105
+ <dim>-1</dim>
106
+ </port>
107
+ <port id="1" precision="FP32">
108
+ <dim>1</dim>
109
+ <dim>1</dim>
110
+ <dim>1</dim>
111
+ <dim>1</dim>
112
+ </port>
113
+ </input>
114
+ <output>
115
+ <port id="2" precision="FP32" names="52">
116
+ <dim>-1</dim>
117
+ <dim>128</dim>
118
+ <dim>-1</dim>
119
+ <dim>-1</dim>
120
+ </port>
121
+ </output>
122
+ </layer>
123
+ <layer id="7" name="Constant_188123" type="Const" version="opset1">
124
+ <data element_type="i64" shape="1" offset="37384" size="8" />
125
+ <output>
126
+ <port id="0" precision="I64">
127
+ <dim>1</dim>
128
+ </port>
129
+ </output>
130
+ </layer>
131
+ <layer id="8" name="__module.decoder.mid.block_1.norm1/aten::mean/ReduceMean" type="ReduceMean" version="opset1">
132
+ <data keep_dims="true" />
133
+ <input>
134
+ <port id="0" precision="FP32">
135
+ <dim>-1</dim>
136
+ <dim>128</dim>
137
+ <dim>-1</dim>
138
+ <dim>-1</dim>
139
+ </port>
140
+ <port id="1" precision="I64">
141
+ <dim>1</dim>
142
+ </port>
143
+ </input>
144
+ <output>
145
+ <port id="2" precision="FP32" names="54,mean_sq.1">
146
+ <dim>-1</dim>
147
+ <dim>1</dim>
148
+ <dim>-1</dim>
149
+ <dim>-1</dim>
150
+ </port>
151
+ </output>
152
+ </layer>
153
+ <layer id="9" name="Constant_203647" type="Const" version="opset1">
154
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37392" size="4" />
155
+ <output>
156
+ <port id="0" precision="FP32">
157
+ <dim>1</dim>
158
+ <dim>1</dim>
159
+ <dim>1</dim>
160
+ <dim>1</dim>
161
+ </port>
162
+ </output>
163
+ </layer>
164
+ <layer id="10" name="__module.decoder.mid.block_1.norm1/aten::add/Add" type="Add" version="opset1">
165
+ <data auto_broadcast="numpy" />
166
+ <input>
167
+ <port id="0" precision="FP32">
168
+ <dim>-1</dim>
169
+ <dim>1</dim>
170
+ <dim>-1</dim>
171
+ <dim>-1</dim>
172
+ </port>
173
+ <port id="1" precision="FP32">
174
+ <dim>1</dim>
175
+ <dim>1</dim>
176
+ <dim>1</dim>
177
+ <dim>1</dim>
178
+ </port>
179
+ </input>
180
+ <output>
181
+ <port id="2" precision="FP32" names="55">
182
+ <dim>-1</dim>
183
+ <dim>1</dim>
184
+ <dim>-1</dim>
185
+ <dim>-1</dim>
186
+ </port>
187
+ </output>
188
+ </layer>
189
+ <layer id="11" name="__module.decoder.mid.block_1.norm1/aten::sqrt/Sqrt" type="Sqrt" version="opset1">
190
+ <input>
191
+ <port id="0" precision="FP32">
192
+ <dim>-1</dim>
193
+ <dim>1</dim>
194
+ <dim>-1</dim>
195
+ <dim>-1</dim>
196
+ </port>
197
+ </input>
198
+ <output>
199
+ <port id="1" precision="FP32" names="56,rms.1">
200
+ <dim>-1</dim>
201
+ <dim>1</dim>
202
+ <dim>-1</dim>
203
+ <dim>-1</dim>
204
+ </port>
205
+ </output>
206
+ </layer>
207
+ <layer id="12" name="__module.decoder.mid.block_1.norm1/aten::div/Divide" type="Divide" version="opset1">
208
+ <data auto_broadcast="numpy" m_pythondiv="true" />
209
+ <input>
210
+ <port id="0" precision="FP32">
211
+ <dim>-1</dim>
212
+ <dim>128</dim>
213
+ <dim>-1</dim>
214
+ <dim>-1</dim>
215
+ </port>
216
+ <port id="1" precision="FP32">
217
+ <dim>-1</dim>
218
+ <dim>1</dim>
219
+ <dim>-1</dim>
220
+ <dim>-1</dim>
221
+ </port>
222
+ </input>
223
+ <output>
224
+ <port id="2" precision="FP32" names="57,input.3">
225
+ <dim>-1</dim>
226
+ <dim>128</dim>
227
+ <dim>-1</dim>
228
+ <dim>-1</dim>
229
+ </port>
230
+ </output>
231
+ </layer>
232
+ <layer id="13" name="__module.decoder.mid.block_1.non_linearity/aten::silu/Swish" type="Swish" version="opset4">
233
+ <input>
234
+ <port id="0" precision="FP32">
235
+ <dim>-1</dim>
236
+ <dim>128</dim>
237
+ <dim>-1</dim>
238
+ <dim>-1</dim>
239
+ </port>
240
+ </input>
241
+ <output>
242
+ <port id="1" precision="FP32" names="58,input.5">
243
+ <dim>-1</dim>
244
+ <dim>128</dim>
245
+ <dim>-1</dim>
246
+ <dim>-1</dim>
247
+ </port>
248
+ </output>
249
+ </layer>
250
+ <layer id="14" name="self.decoder.mid.block_1.conv1.conv.weight" type="Const" version="opset1">
251
+ <data element_type="f32" shape="128, 128, 3, 3" offset="37396" size="589824" />
252
+ <output>
253
+ <port id="0" precision="FP32" names="self.decoder.mid.block_1.conv1.conv.weight">
254
+ <dim>128</dim>
255
+ <dim>128</dim>
256
+ <dim>3</dim>
257
+ <dim>3</dim>
258
+ </port>
259
+ </output>
260
+ </layer>
261
+ <layer id="15" name="__module.decoder.mid.block_1.conv1.conv/aten::_convolution/Convolution" type="Convolution" version="opset1">
262
+ <data strides="1, 1" dilations="1, 1" pads_begin="2, 1" pads_end="0, 1" auto_pad="explicit" />
263
+ <input>
264
+ <port id="0" precision="FP32">
265
+ <dim>-1</dim>
266
+ <dim>128</dim>
267
+ <dim>-1</dim>
268
+ <dim>-1</dim>
269
+ </port>
270
+ <port id="1" precision="FP32">
271
+ <dim>128</dim>
272
+ <dim>128</dim>
273
+ <dim>3</dim>
274
+ <dim>3</dim>
275
+ </port>
276
+ </input>
277
+ <output>
278
+ <port id="2" precision="FP32">
279
+ <dim>-1</dim>
280
+ <dim>128</dim>
281
+ <dim>-1</dim>
282
+ <dim>-1</dim>
283
+ </port>
284
+ </output>
285
+ </layer>
286
+ <layer id="16" name="__module.decoder.mid.block_1.conv1.conv/aten::_convolution/Reshape" type="Const" version="opset1">
287
+ <data element_type="f32" shape="1, 128, 1, 1" offset="627220" size="512" />
288
+ <output>
289
+ <port id="0" precision="FP32">
290
+ <dim>1</dim>
291
+ <dim>128</dim>
292
+ <dim>1</dim>
293
+ <dim>1</dim>
294
+ </port>
295
+ </output>
296
+ </layer>
297
+ <layer id="17" name="__module.decoder.mid.block_1.conv1.conv/aten::_convolution/Add" type="Add" version="opset1">
298
+ <data auto_broadcast="numpy" />
299
+ <input>
300
+ <port id="0" precision="FP32">
301
+ <dim>-1</dim>
302
+ <dim>128</dim>
303
+ <dim>-1</dim>
304
+ <dim>-1</dim>
305
+ </port>
306
+ <port id="1" precision="FP32">
307
+ <dim>1</dim>
308
+ <dim>128</dim>
309
+ <dim>1</dim>
310
+ <dim>1</dim>
311
+ </port>
312
+ </input>
313
+ <output>
314
+ <port id="2" precision="FP32" names="68,x.3">
315
+ <dim>-1</dim>
316
+ <dim>128</dim>
317
+ <dim>-1</dim>
318
+ <dim>-1</dim>
319
+ </port>
320
+ </output>
321
+ </layer>
322
+ <layer id="18" name="Constant_203648" type="Const" version="opset1">
323
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37376" size="4" />
324
+ <output>
325
+ <port id="0" precision="FP32">
326
+ <dim>1</dim>
327
+ <dim>1</dim>
328
+ <dim>1</dim>
329
+ <dim>1</dim>
330
+ </port>
331
+ </output>
332
+ </layer>
333
+ <layer id="19" name="__module.decoder.mid.block_1.norm2/aten::pow/Power" type="Power" version="opset1">
334
+ <data auto_broadcast="numpy" />
335
+ <input>
336
+ <port id="0" precision="FP32">
337
+ <dim>-1</dim>
338
+ <dim>128</dim>
339
+ <dim>-1</dim>
340
+ <dim>-1</dim>
341
+ </port>
342
+ <port id="1" precision="FP32">
343
+ <dim>1</dim>
344
+ <dim>1</dim>
345
+ <dim>1</dim>
346
+ <dim>1</dim>
347
+ </port>
348
+ </input>
349
+ <output>
350
+ <port id="2" precision="FP32" names="69">
351
+ <dim>-1</dim>
352
+ <dim>128</dim>
353
+ <dim>-1</dim>
354
+ <dim>-1</dim>
355
+ </port>
356
+ </output>
357
+ </layer>
358
+ <layer id="20" name="Constant_188273" type="Const" version="opset1">
359
+ <data element_type="i64" shape="1" offset="37384" size="8" />
360
+ <output>
361
+ <port id="0" precision="I64">
362
+ <dim>1</dim>
363
+ </port>
364
+ </output>
365
+ </layer>
366
+ <layer id="21" name="__module.decoder.mid.block_1.norm2/aten::mean/ReduceMean" type="ReduceMean" version="opset1">
367
+ <data keep_dims="true" />
368
+ <input>
369
+ <port id="0" precision="FP32">
370
+ <dim>-1</dim>
371
+ <dim>128</dim>
372
+ <dim>-1</dim>
373
+ <dim>-1</dim>
374
+ </port>
375
+ <port id="1" precision="I64">
376
+ <dim>1</dim>
377
+ </port>
378
+ </input>
379
+ <output>
380
+ <port id="2" precision="FP32" names="71,mean_sq.3">
381
+ <dim>-1</dim>
382
+ <dim>1</dim>
383
+ <dim>-1</dim>
384
+ <dim>-1</dim>
385
+ </port>
386
+ </output>
387
+ </layer>
388
+ <layer id="22" name="Constant_203649" type="Const" version="opset1">
389
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37392" size="4" />
390
+ <output>
391
+ <port id="0" precision="FP32">
392
+ <dim>1</dim>
393
+ <dim>1</dim>
394
+ <dim>1</dim>
395
+ <dim>1</dim>
396
+ </port>
397
+ </output>
398
+ </layer>
399
+ <layer id="23" name="__module.decoder.mid.block_1.norm2/aten::add/Add" type="Add" version="opset1">
400
+ <data auto_broadcast="numpy" />
401
+ <input>
402
+ <port id="0" precision="FP32">
403
+ <dim>-1</dim>
404
+ <dim>1</dim>
405
+ <dim>-1</dim>
406
+ <dim>-1</dim>
407
+ </port>
408
+ <port id="1" precision="FP32">
409
+ <dim>1</dim>
410
+ <dim>1</dim>
411
+ <dim>1</dim>
412
+ <dim>1</dim>
413
+ </port>
414
+ </input>
415
+ <output>
416
+ <port id="2" precision="FP32" names="72">
417
+ <dim>-1</dim>
418
+ <dim>1</dim>
419
+ <dim>-1</dim>
420
+ <dim>-1</dim>
421
+ </port>
422
+ </output>
423
+ </layer>
424
+ <layer id="24" name="__module.decoder.mid.block_1.norm2/aten::sqrt/Sqrt" type="Sqrt" version="opset1">
425
+ <input>
426
+ <port id="0" precision="FP32">
427
+ <dim>-1</dim>
428
+ <dim>1</dim>
429
+ <dim>-1</dim>
430
+ <dim>-1</dim>
431
+ </port>
432
+ </input>
433
+ <output>
434
+ <port id="1" precision="FP32" names="73,rms.3">
435
+ <dim>-1</dim>
436
+ <dim>1</dim>
437
+ <dim>-1</dim>
438
+ <dim>-1</dim>
439
+ </port>
440
+ </output>
441
+ </layer>
442
+ <layer id="25" name="__module.decoder.mid.block_1.norm2/aten::div/Divide" type="Divide" version="opset1">
443
+ <data auto_broadcast="numpy" m_pythondiv="true" />
444
+ <input>
445
+ <port id="0" precision="FP32">
446
+ <dim>-1</dim>
447
+ <dim>128</dim>
448
+ <dim>-1</dim>
449
+ <dim>-1</dim>
450
+ </port>
451
+ <port id="1" precision="FP32">
452
+ <dim>-1</dim>
453
+ <dim>1</dim>
454
+ <dim>-1</dim>
455
+ <dim>-1</dim>
456
+ </port>
457
+ </input>
458
+ <output>
459
+ <port id="2" precision="FP32" names="74,input.9">
460
+ <dim>-1</dim>
461
+ <dim>128</dim>
462
+ <dim>-1</dim>
463
+ <dim>-1</dim>
464
+ </port>
465
+ </output>
466
+ </layer>
467
+ <layer id="26" name="__module.decoder.mid.block_1.non_linearity/aten::silu/Swish_1" type="Swish" version="opset4">
468
+ <input>
469
+ <port id="0" precision="FP32">
470
+ <dim>-1</dim>
471
+ <dim>128</dim>
472
+ <dim>-1</dim>
473
+ <dim>-1</dim>
474
+ </port>
475
+ </input>
476
+ <output>
477
+ <port id="1" precision="FP32" names="75,input.11">
478
+ <dim>-1</dim>
479
+ <dim>128</dim>
480
+ <dim>-1</dim>
481
+ <dim>-1</dim>
482
+ </port>
483
+ </output>
484
+ </layer>
485
+ <layer id="27" name="self.decoder.mid.block_1.conv2.conv.weight" type="Const" version="opset1">
486
+ <data element_type="f32" shape="128, 128, 3, 3" offset="627732" size="589824" />
487
+ <output>
488
+ <port id="0" precision="FP32" names="self.decoder.mid.block_1.conv2.conv.weight">
489
+ <dim>128</dim>
490
+ <dim>128</dim>
491
+ <dim>3</dim>
492
+ <dim>3</dim>
493
+ </port>
494
+ </output>
495
+ </layer>
496
+ <layer id="28" name="__module.decoder.mid.block_1.conv2.conv/aten::_convolution/Convolution" type="Convolution" version="opset1">
497
+ <data strides="1, 1" dilations="1, 1" pads_begin="2, 1" pads_end="0, 1" auto_pad="explicit" />
498
+ <input>
499
+ <port id="0" precision="FP32">
500
+ <dim>-1</dim>
501
+ <dim>128</dim>
502
+ <dim>-1</dim>
503
+ <dim>-1</dim>
504
+ </port>
505
+ <port id="1" precision="FP32">
506
+ <dim>128</dim>
507
+ <dim>128</dim>
508
+ <dim>3</dim>
509
+ <dim>3</dim>
510
+ </port>
511
+ </input>
512
+ <output>
513
+ <port id="2" precision="FP32">
514
+ <dim>-1</dim>
515
+ <dim>128</dim>
516
+ <dim>-1</dim>
517
+ <dim>-1</dim>
518
+ </port>
519
+ </output>
520
+ </layer>
521
+ <layer id="29" name="__module.decoder.mid.block_1.conv2.conv/aten::_convolution/Reshape" type="Const" version="opset1">
522
+ <data element_type="f32" shape="1, 128, 1, 1" offset="1217556" size="512" />
523
+ <output>
524
+ <port id="0" precision="FP32">
525
+ <dim>1</dim>
526
+ <dim>128</dim>
527
+ <dim>1</dim>
528
+ <dim>1</dim>
529
+ </port>
530
+ </output>
531
+ </layer>
532
+ <layer id="30" name="__module.decoder.mid.block_1.conv2.conv/aten::_convolution/Add" type="Add" version="opset1">
533
+ <data auto_broadcast="numpy" />
534
+ <input>
535
+ <port id="0" precision="FP32">
536
+ <dim>-1</dim>
537
+ <dim>128</dim>
538
+ <dim>-1</dim>
539
+ <dim>-1</dim>
540
+ </port>
541
+ <port id="1" precision="FP32">
542
+ <dim>1</dim>
543
+ <dim>128</dim>
544
+ <dim>1</dim>
545
+ <dim>1</dim>
546
+ </port>
547
+ </input>
548
+ <output>
549
+ <port id="2" precision="FP32" names="86,h.1">
550
+ <dim>-1</dim>
551
+ <dim>128</dim>
552
+ <dim>-1</dim>
553
+ <dim>-1</dim>
554
+ </port>
555
+ </output>
556
+ </layer>
557
+ <layer id="31" name="__module.decoder.mid.block_1/aten::add/Add" type="Add" version="opset1">
558
+ <data auto_broadcast="numpy" />
559
+ <input>
560
+ <port id="0" precision="FP32">
561
+ <dim>-1</dim>
562
+ <dim>128</dim>
563
+ <dim>-1</dim>
564
+ <dim>-1</dim>
565
+ </port>
566
+ <port id="1" precision="FP32">
567
+ <dim>-1</dim>
568
+ <dim>128</dim>
569
+ <dim>-1</dim>
570
+ <dim>-1</dim>
571
+ </port>
572
+ </input>
573
+ <output>
574
+ <port id="2" precision="FP32" names="87,x.5">
575
+ <dim>-1</dim>
576
+ <dim>128</dim>
577
+ <dim>-1</dim>
578
+ <dim>-1</dim>
579
+ </port>
580
+ </output>
581
+ </layer>
582
+ <layer id="32" name="Constant_203650" type="Const" version="opset1">
583
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37376" size="4" />
584
+ <output>
585
+ <port id="0" precision="FP32">
586
+ <dim>1</dim>
587
+ <dim>1</dim>
588
+ <dim>1</dim>
589
+ <dim>1</dim>
590
+ </port>
591
+ </output>
592
+ </layer>
593
+ <layer id="33" name="__module.decoder.mid.block_2.norm1/aten::pow/Power" type="Power" version="opset1">
594
+ <data auto_broadcast="numpy" />
595
+ <input>
596
+ <port id="0" precision="FP32">
597
+ <dim>-1</dim>
598
+ <dim>128</dim>
599
+ <dim>-1</dim>
600
+ <dim>-1</dim>
601
+ </port>
602
+ <port id="1" precision="FP32">
603
+ <dim>1</dim>
604
+ <dim>1</dim>
605
+ <dim>1</dim>
606
+ <dim>1</dim>
607
+ </port>
608
+ </input>
609
+ <output>
610
+ <port id="2" precision="FP32" names="90">
611
+ <dim>-1</dim>
612
+ <dim>128</dim>
613
+ <dim>-1</dim>
614
+ <dim>-1</dim>
615
+ </port>
616
+ </output>
617
+ </layer>
618
+ <layer id="34" name="Constant_188431" type="Const" version="opset1">
619
+ <data element_type="i64" shape="1" offset="37384" size="8" />
620
+ <output>
621
+ <port id="0" precision="I64">
622
+ <dim>1</dim>
623
+ </port>
624
+ </output>
625
+ </layer>
626
+ <layer id="35" name="__module.decoder.mid.block_2.norm1/aten::mean/ReduceMean" type="ReduceMean" version="opset1">
627
+ <data keep_dims="true" />
628
+ <input>
629
+ <port id="0" precision="FP32">
630
+ <dim>-1</dim>
631
+ <dim>128</dim>
632
+ <dim>-1</dim>
633
+ <dim>-1</dim>
634
+ </port>
635
+ <port id="1" precision="I64">
636
+ <dim>1</dim>
637
+ </port>
638
+ </input>
639
+ <output>
640
+ <port id="2" precision="FP32" names="92,mean_sq.5">
641
+ <dim>-1</dim>
642
+ <dim>1</dim>
643
+ <dim>-1</dim>
644
+ <dim>-1</dim>
645
+ </port>
646
+ </output>
647
+ </layer>
648
+ <layer id="36" name="Constant_203651" type="Const" version="opset1">
649
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37392" size="4" />
650
+ <output>
651
+ <port id="0" precision="FP32">
652
+ <dim>1</dim>
653
+ <dim>1</dim>
654
+ <dim>1</dim>
655
+ <dim>1</dim>
656
+ </port>
657
+ </output>
658
+ </layer>
659
+ <layer id="37" name="__module.decoder.mid.block_2.norm1/aten::add/Add" type="Add" version="opset1">
660
+ <data auto_broadcast="numpy" />
661
+ <input>
662
+ <port id="0" precision="FP32">
663
+ <dim>-1</dim>
664
+ <dim>1</dim>
665
+ <dim>-1</dim>
666
+ <dim>-1</dim>
667
+ </port>
668
+ <port id="1" precision="FP32">
669
+ <dim>1</dim>
670
+ <dim>1</dim>
671
+ <dim>1</dim>
672
+ <dim>1</dim>
673
+ </port>
674
+ </input>
675
+ <output>
676
+ <port id="2" precision="FP32" names="93">
677
+ <dim>-1</dim>
678
+ <dim>1</dim>
679
+ <dim>-1</dim>
680
+ <dim>-1</dim>
681
+ </port>
682
+ </output>
683
+ </layer>
684
+ <layer id="38" name="__module.decoder.mid.block_2.norm1/aten::sqrt/Sqrt" type="Sqrt" version="opset1">
685
+ <input>
686
+ <port id="0" precision="FP32">
687
+ <dim>-1</dim>
688
+ <dim>1</dim>
689
+ <dim>-1</dim>
690
+ <dim>-1</dim>
691
+ </port>
692
+ </input>
693
+ <output>
694
+ <port id="1" precision="FP32" names="94,rms.5">
695
+ <dim>-1</dim>
696
+ <dim>1</dim>
697
+ <dim>-1</dim>
698
+ <dim>-1</dim>
699
+ </port>
700
+ </output>
701
+ </layer>
702
+ <layer id="39" name="__module.decoder.mid.block_2.norm1/aten::div/Divide" type="Divide" version="opset1">
703
+ <data auto_broadcast="numpy" m_pythondiv="true" />
704
+ <input>
705
+ <port id="0" precision="FP32">
706
+ <dim>-1</dim>
707
+ <dim>128</dim>
708
+ <dim>-1</dim>
709
+ <dim>-1</dim>
710
+ </port>
711
+ <port id="1" precision="FP32">
712
+ <dim>-1</dim>
713
+ <dim>1</dim>
714
+ <dim>-1</dim>
715
+ <dim>-1</dim>
716
+ </port>
717
+ </input>
718
+ <output>
719
+ <port id="2" precision="FP32" names="95,input.17">
720
+ <dim>-1</dim>
721
+ <dim>128</dim>
722
+ <dim>-1</dim>
723
+ <dim>-1</dim>
724
+ </port>
725
+ </output>
726
+ </layer>
727
+ <layer id="40" name="__module.decoder.mid.block_2.non_linearity/aten::silu/Swish" type="Swish" version="opset4">
728
+ <input>
729
+ <port id="0" precision="FP32">
730
+ <dim>-1</dim>
731
+ <dim>128</dim>
732
+ <dim>-1</dim>
733
+ <dim>-1</dim>
734
+ </port>
735
+ </input>
736
+ <output>
737
+ <port id="1" precision="FP32" names="96,input.19">
738
+ <dim>-1</dim>
739
+ <dim>128</dim>
740
+ <dim>-1</dim>
741
+ <dim>-1</dim>
742
+ </port>
743
+ </output>
744
+ </layer>
745
+ <layer id="41" name="self.decoder.mid.block_2.conv1.conv.weight" type="Const" version="opset1">
746
+ <data element_type="f32" shape="128, 128, 3, 3" offset="1218068" size="589824" />
747
+ <output>
748
+ <port id="0" precision="FP32" names="self.decoder.mid.block_2.conv1.conv.weight">
749
+ <dim>128</dim>
750
+ <dim>128</dim>
751
+ <dim>3</dim>
752
+ <dim>3</dim>
753
+ </port>
754
+ </output>
755
+ </layer>
756
+ <layer id="42" name="__module.decoder.mid.block_2.conv1.conv/aten::_convolution/Convolution" type="Convolution" version="opset1">
757
+ <data strides="1, 1" dilations="1, 1" pads_begin="2, 1" pads_end="0, 1" auto_pad="explicit" />
758
+ <input>
759
+ <port id="0" precision="FP32">
760
+ <dim>-1</dim>
761
+ <dim>128</dim>
762
+ <dim>-1</dim>
763
+ <dim>-1</dim>
764
+ </port>
765
+ <port id="1" precision="FP32">
766
+ <dim>128</dim>
767
+ <dim>128</dim>
768
+ <dim>3</dim>
769
+ <dim>3</dim>
770
+ </port>
771
+ </input>
772
+ <output>
773
+ <port id="2" precision="FP32">
774
+ <dim>-1</dim>
775
+ <dim>128</dim>
776
+ <dim>-1</dim>
777
+ <dim>-1</dim>
778
+ </port>
779
+ </output>
780
+ </layer>
781
+ <layer id="43" name="__module.decoder.mid.block_2.conv1.conv/aten::_convolution/Reshape" type="Const" version="opset1">
782
+ <data element_type="f32" shape="1, 128, 1, 1" offset="1807892" size="512" />
783
+ <output>
784
+ <port id="0" precision="FP32">
785
+ <dim>1</dim>
786
+ <dim>128</dim>
787
+ <dim>1</dim>
788
+ <dim>1</dim>
789
+ </port>
790
+ </output>
791
+ </layer>
792
+ <layer id="44" name="__module.decoder.mid.block_2.conv1.conv/aten::_convolution/Add" type="Add" version="opset1">
793
+ <data auto_broadcast="numpy" />
794
+ <input>
795
+ <port id="0" precision="FP32">
796
+ <dim>-1</dim>
797
+ <dim>128</dim>
798
+ <dim>-1</dim>
799
+ <dim>-1</dim>
800
+ </port>
801
+ <port id="1" precision="FP32">
802
+ <dim>1</dim>
803
+ <dim>128</dim>
804
+ <dim>1</dim>
805
+ <dim>1</dim>
806
+ </port>
807
+ </input>
808
+ <output>
809
+ <port id="2" precision="FP32" names="106,x.7">
810
+ <dim>-1</dim>
811
+ <dim>128</dim>
812
+ <dim>-1</dim>
813
+ <dim>-1</dim>
814
+ </port>
815
+ </output>
816
+ </layer>
817
+ <layer id="45" name="Constant_203652" type="Const" version="opset1">
818
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37376" size="4" />
819
+ <output>
820
+ <port id="0" precision="FP32">
821
+ <dim>1</dim>
822
+ <dim>1</dim>
823
+ <dim>1</dim>
824
+ <dim>1</dim>
825
+ </port>
826
+ </output>
827
+ </layer>
828
+ <layer id="46" name="__module.decoder.mid.block_2.norm2/aten::pow/Power" type="Power" version="opset1">
829
+ <data auto_broadcast="numpy" />
830
+ <input>
831
+ <port id="0" precision="FP32">
832
+ <dim>-1</dim>
833
+ <dim>128</dim>
834
+ <dim>-1</dim>
835
+ <dim>-1</dim>
836
+ </port>
837
+ <port id="1" precision="FP32">
838
+ <dim>1</dim>
839
+ <dim>1</dim>
840
+ <dim>1</dim>
841
+ <dim>1</dim>
842
+ </port>
843
+ </input>
844
+ <output>
845
+ <port id="2" precision="FP32" names="107">
846
+ <dim>-1</dim>
847
+ <dim>128</dim>
848
+ <dim>-1</dim>
849
+ <dim>-1</dim>
850
+ </port>
851
+ </output>
852
+ </layer>
853
+ <layer id="47" name="Constant_188581" type="Const" version="opset1">
854
+ <data element_type="i64" shape="1" offset="37384" size="8" />
855
+ <output>
856
+ <port id="0" precision="I64">
857
+ <dim>1</dim>
858
+ </port>
859
+ </output>
860
+ </layer>
861
+ <layer id="48" name="__module.decoder.mid.block_2.norm2/aten::mean/ReduceMean" type="ReduceMean" version="opset1">
862
+ <data keep_dims="true" />
863
+ <input>
864
+ <port id="0" precision="FP32">
865
+ <dim>-1</dim>
866
+ <dim>128</dim>
867
+ <dim>-1</dim>
868
+ <dim>-1</dim>
869
+ </port>
870
+ <port id="1" precision="I64">
871
+ <dim>1</dim>
872
+ </port>
873
+ </input>
874
+ <output>
875
+ <port id="2" precision="FP32" names="109,mean_sq.7">
876
+ <dim>-1</dim>
877
+ <dim>1</dim>
878
+ <dim>-1</dim>
879
+ <dim>-1</dim>
880
+ </port>
881
+ </output>
882
+ </layer>
883
+ <layer id="49" name="Constant_203653" type="Const" version="opset1">
884
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37392" size="4" />
885
+ <output>
886
+ <port id="0" precision="FP32">
887
+ <dim>1</dim>
888
+ <dim>1</dim>
889
+ <dim>1</dim>
890
+ <dim>1</dim>
891
+ </port>
892
+ </output>
893
+ </layer>
894
+ <layer id="50" name="__module.decoder.mid.block_2.norm2/aten::add/Add" type="Add" version="opset1">
895
+ <data auto_broadcast="numpy" />
896
+ <input>
897
+ <port id="0" precision="FP32">
898
+ <dim>-1</dim>
899
+ <dim>1</dim>
900
+ <dim>-1</dim>
901
+ <dim>-1</dim>
902
+ </port>
903
+ <port id="1" precision="FP32">
904
+ <dim>1</dim>
905
+ <dim>1</dim>
906
+ <dim>1</dim>
907
+ <dim>1</dim>
908
+ </port>
909
+ </input>
910
+ <output>
911
+ <port id="2" precision="FP32" names="110">
912
+ <dim>-1</dim>
913
+ <dim>1</dim>
914
+ <dim>-1</dim>
915
+ <dim>-1</dim>
916
+ </port>
917
+ </output>
918
+ </layer>
919
+ <layer id="51" name="__module.decoder.mid.block_2.norm2/aten::sqrt/Sqrt" type="Sqrt" version="opset1">
920
+ <input>
921
+ <port id="0" precision="FP32">
922
+ <dim>-1</dim>
923
+ <dim>1</dim>
924
+ <dim>-1</dim>
925
+ <dim>-1</dim>
926
+ </port>
927
+ </input>
928
+ <output>
929
+ <port id="1" precision="FP32" names="111,rms.7">
930
+ <dim>-1</dim>
931
+ <dim>1</dim>
932
+ <dim>-1</dim>
933
+ <dim>-1</dim>
934
+ </port>
935
+ </output>
936
+ </layer>
937
+ <layer id="52" name="__module.decoder.mid.block_2.norm2/aten::div/Divide" type="Divide" version="opset1">
938
+ <data auto_broadcast="numpy" m_pythondiv="true" />
939
+ <input>
940
+ <port id="0" precision="FP32">
941
+ <dim>-1</dim>
942
+ <dim>128</dim>
943
+ <dim>-1</dim>
944
+ <dim>-1</dim>
945
+ </port>
946
+ <port id="1" precision="FP32">
947
+ <dim>-1</dim>
948
+ <dim>1</dim>
949
+ <dim>-1</dim>
950
+ <dim>-1</dim>
951
+ </port>
952
+ </input>
953
+ <output>
954
+ <port id="2" precision="FP32" names="112,input.23">
955
+ <dim>-1</dim>
956
+ <dim>128</dim>
957
+ <dim>-1</dim>
958
+ <dim>-1</dim>
959
+ </port>
960
+ </output>
961
+ </layer>
962
+ <layer id="53" name="__module.decoder.mid.block_2.non_linearity/aten::silu/Swish_1" type="Swish" version="opset4">
963
+ <input>
964
+ <port id="0" precision="FP32">
965
+ <dim>-1</dim>
966
+ <dim>128</dim>
967
+ <dim>-1</dim>
968
+ <dim>-1</dim>
969
+ </port>
970
+ </input>
971
+ <output>
972
+ <port id="1" precision="FP32" names="113,input.25">
973
+ <dim>-1</dim>
974
+ <dim>128</dim>
975
+ <dim>-1</dim>
976
+ <dim>-1</dim>
977
+ </port>
978
+ </output>
979
+ </layer>
980
+ <layer id="54" name="self.decoder.mid.block_2.conv2.conv.weight" type="Const" version="opset1">
981
+ <data element_type="f32" shape="128, 128, 3, 3" offset="1808404" size="589824" />
982
+ <output>
983
+ <port id="0" precision="FP32" names="self.decoder.mid.block_2.conv2.conv.weight">
984
+ <dim>128</dim>
985
+ <dim>128</dim>
986
+ <dim>3</dim>
987
+ <dim>3</dim>
988
+ </port>
989
+ </output>
990
+ </layer>
991
+ <layer id="55" name="__module.decoder.mid.block_2.conv2.conv/aten::_convolution/Convolution" type="Convolution" version="opset1">
992
+ <data strides="1, 1" dilations="1, 1" pads_begin="2, 1" pads_end="0, 1" auto_pad="explicit" />
993
+ <input>
994
+ <port id="0" precision="FP32">
995
+ <dim>-1</dim>
996
+ <dim>128</dim>
997
+ <dim>-1</dim>
998
+ <dim>-1</dim>
999
+ </port>
1000
+ <port id="1" precision="FP32">
1001
+ <dim>128</dim>
1002
+ <dim>128</dim>
1003
+ <dim>3</dim>
1004
+ <dim>3</dim>
1005
+ </port>
1006
+ </input>
1007
+ <output>
1008
+ <port id="2" precision="FP32">
1009
+ <dim>-1</dim>
1010
+ <dim>128</dim>
1011
+ <dim>-1</dim>
1012
+ <dim>-1</dim>
1013
+ </port>
1014
+ </output>
1015
+ </layer>
1016
+ <layer id="56" name="__module.decoder.mid.block_2.conv2.conv/aten::_convolution/Reshape" type="Const" version="opset1">
1017
+ <data element_type="f32" shape="1, 128, 1, 1" offset="2398228" size="512" />
1018
+ <output>
1019
+ <port id="0" precision="FP32">
1020
+ <dim>1</dim>
1021
+ <dim>128</dim>
1022
+ <dim>1</dim>
1023
+ <dim>1</dim>
1024
+ </port>
1025
+ </output>
1026
+ </layer>
1027
+ <layer id="57" name="__module.decoder.mid.block_2.conv2.conv/aten::_convolution/Add" type="Add" version="opset1">
1028
+ <data auto_broadcast="numpy" />
1029
+ <input>
1030
+ <port id="0" precision="FP32">
1031
+ <dim>-1</dim>
1032
+ <dim>128</dim>
1033
+ <dim>-1</dim>
1034
+ <dim>-1</dim>
1035
+ </port>
1036
+ <port id="1" precision="FP32">
1037
+ <dim>1</dim>
1038
+ <dim>128</dim>
1039
+ <dim>1</dim>
1040
+ <dim>1</dim>
1041
+ </port>
1042
+ </input>
1043
+ <output>
1044
+ <port id="2" precision="FP32" names="124,h.3">
1045
+ <dim>-1</dim>
1046
+ <dim>128</dim>
1047
+ <dim>-1</dim>
1048
+ <dim>-1</dim>
1049
+ </port>
1050
+ </output>
1051
+ </layer>
1052
+ <layer id="58" name="__module.decoder.mid.block_2/aten::add/Add" type="Add" version="opset1">
1053
+ <data auto_broadcast="numpy" />
1054
+ <input>
1055
+ <port id="0" precision="FP32">
1056
+ <dim>-1</dim>
1057
+ <dim>128</dim>
1058
+ <dim>-1</dim>
1059
+ <dim>-1</dim>
1060
+ </port>
1061
+ <port id="1" precision="FP32">
1062
+ <dim>-1</dim>
1063
+ <dim>128</dim>
1064
+ <dim>-1</dim>
1065
+ <dim>-1</dim>
1066
+ </port>
1067
+ </input>
1068
+ <output>
1069
+ <port id="2" precision="FP32" names="125,x.9">
1070
+ <dim>-1</dim>
1071
+ <dim>128</dim>
1072
+ <dim>-1</dim>
1073
+ <dim>-1</dim>
1074
+ </port>
1075
+ </output>
1076
+ </layer>
1077
+ <layer id="59" name="Constant_203654" type="Const" version="opset1">
1078
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37376" size="4" />
1079
+ <output>
1080
+ <port id="0" precision="FP32">
1081
+ <dim>1</dim>
1082
+ <dim>1</dim>
1083
+ <dim>1</dim>
1084
+ <dim>1</dim>
1085
+ </port>
1086
+ </output>
1087
+ </layer>
1088
+ <layer id="60" name="__module.decoder.up.0.block.0.norm1/aten::pow/Power" type="Power" version="opset1">
1089
+ <data auto_broadcast="numpy" />
1090
+ <input>
1091
+ <port id="0" precision="FP32">
1092
+ <dim>-1</dim>
1093
+ <dim>128</dim>
1094
+ <dim>-1</dim>
1095
+ <dim>-1</dim>
1096
+ </port>
1097
+ <port id="1" precision="FP32">
1098
+ <dim>1</dim>
1099
+ <dim>1</dim>
1100
+ <dim>1</dim>
1101
+ <dim>1</dim>
1102
+ </port>
1103
+ </input>
1104
+ <output>
1105
+ <port id="2" precision="FP32" names="128">
1106
+ <dim>-1</dim>
1107
+ <dim>128</dim>
1108
+ <dim>-1</dim>
1109
+ <dim>-1</dim>
1110
+ </port>
1111
+ </output>
1112
+ </layer>
1113
+ <layer id="61" name="Constant_188739" type="Const" version="opset1">
1114
+ <data element_type="i64" shape="1" offset="37384" size="8" />
1115
+ <output>
1116
+ <port id="0" precision="I64">
1117
+ <dim>1</dim>
1118
+ </port>
1119
+ </output>
1120
+ </layer>
1121
+ <layer id="62" name="__module.decoder.up.0.block.0.norm1/aten::mean/ReduceMean" type="ReduceMean" version="opset1">
1122
+ <data keep_dims="true" />
1123
+ <input>
1124
+ <port id="0" precision="FP32">
1125
+ <dim>-1</dim>
1126
+ <dim>128</dim>
1127
+ <dim>-1</dim>
1128
+ <dim>-1</dim>
1129
+ </port>
1130
+ <port id="1" precision="I64">
1131
+ <dim>1</dim>
1132
+ </port>
1133
+ </input>
1134
+ <output>
1135
+ <port id="2" precision="FP32" names="130,mean_sq.9">
1136
+ <dim>-1</dim>
1137
+ <dim>1</dim>
1138
+ <dim>-1</dim>
1139
+ <dim>-1</dim>
1140
+ </port>
1141
+ </output>
1142
+ </layer>
1143
+ <layer id="63" name="Constant_203655" type="Const" version="opset1">
1144
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37392" size="4" />
1145
+ <output>
1146
+ <port id="0" precision="FP32">
1147
+ <dim>1</dim>
1148
+ <dim>1</dim>
1149
+ <dim>1</dim>
1150
+ <dim>1</dim>
1151
+ </port>
1152
+ </output>
1153
+ </layer>
1154
+ <layer id="64" name="__module.decoder.up.0.block.0.norm1/aten::add/Add" type="Add" version="opset1">
1155
+ <data auto_broadcast="numpy" />
1156
+ <input>
1157
+ <port id="0" precision="FP32">
1158
+ <dim>-1</dim>
1159
+ <dim>1</dim>
1160
+ <dim>-1</dim>
1161
+ <dim>-1</dim>
1162
+ </port>
1163
+ <port id="1" precision="FP32">
1164
+ <dim>1</dim>
1165
+ <dim>1</dim>
1166
+ <dim>1</dim>
1167
+ <dim>1</dim>
1168
+ </port>
1169
+ </input>
1170
+ <output>
1171
+ <port id="2" precision="FP32" names="131">
1172
+ <dim>-1</dim>
1173
+ <dim>1</dim>
1174
+ <dim>-1</dim>
1175
+ <dim>-1</dim>
1176
+ </port>
1177
+ </output>
1178
+ </layer>
1179
+ <layer id="65" name="__module.decoder.up.0.block.0.norm1/aten::sqrt/Sqrt" type="Sqrt" version="opset1">
1180
+ <input>
1181
+ <port id="0" precision="FP32">
1182
+ <dim>-1</dim>
1183
+ <dim>1</dim>
1184
+ <dim>-1</dim>
1185
+ <dim>-1</dim>
1186
+ </port>
1187
+ </input>
1188
+ <output>
1189
+ <port id="1" precision="FP32" names="132,rms.9">
1190
+ <dim>-1</dim>
1191
+ <dim>1</dim>
1192
+ <dim>-1</dim>
1193
+ <dim>-1</dim>
1194
+ </port>
1195
+ </output>
1196
+ </layer>
1197
+ <layer id="66" name="__module.decoder.up.0.block.0.norm1/aten::div/Divide" type="Divide" version="opset1">
1198
+ <data auto_broadcast="numpy" m_pythondiv="true" />
1199
+ <input>
1200
+ <port id="0" precision="FP32">
1201
+ <dim>-1</dim>
1202
+ <dim>128</dim>
1203
+ <dim>-1</dim>
1204
+ <dim>-1</dim>
1205
+ </port>
1206
+ <port id="1" precision="FP32">
1207
+ <dim>-1</dim>
1208
+ <dim>1</dim>
1209
+ <dim>-1</dim>
1210
+ <dim>-1</dim>
1211
+ </port>
1212
+ </input>
1213
+ <output>
1214
+ <port id="2" precision="FP32" names="133,input.31">
1215
+ <dim>-1</dim>
1216
+ <dim>128</dim>
1217
+ <dim>-1</dim>
1218
+ <dim>-1</dim>
1219
+ </port>
1220
+ </output>
1221
+ </layer>
1222
+ <layer id="67" name="__module.decoder.up.0.block.0.non_linearity/aten::silu/Swish" type="Swish" version="opset4">
1223
+ <input>
1224
+ <port id="0" precision="FP32">
1225
+ <dim>-1</dim>
1226
+ <dim>128</dim>
1227
+ <dim>-1</dim>
1228
+ <dim>-1</dim>
1229
+ </port>
1230
+ </input>
1231
+ <output>
1232
+ <port id="1" precision="FP32" names="134,input.33">
1233
+ <dim>-1</dim>
1234
+ <dim>128</dim>
1235
+ <dim>-1</dim>
1236
+ <dim>-1</dim>
1237
+ </port>
1238
+ </output>
1239
+ </layer>
1240
+ <layer id="68" name="self.decoder.up.0.block.0.conv1.conv.weight" type="Const" version="opset1">
1241
+ <data element_type="f32" shape="128, 128, 3, 3" offset="2398740" size="589824" />
1242
+ <output>
1243
+ <port id="0" precision="FP32" names="self.decoder.up.0.block.0.conv1.conv.weight">
1244
+ <dim>128</dim>
1245
+ <dim>128</dim>
1246
+ <dim>3</dim>
1247
+ <dim>3</dim>
1248
+ </port>
1249
+ </output>
1250
+ </layer>
1251
+ <layer id="69" name="__module.decoder.up.0.block.0.conv1.conv/aten::_convolution/Convolution" type="Convolution" version="opset1">
1252
+ <data strides="1, 1" dilations="1, 1" pads_begin="2, 1" pads_end="0, 1" auto_pad="explicit" />
1253
+ <input>
1254
+ <port id="0" precision="FP32">
1255
+ <dim>-1</dim>
1256
+ <dim>128</dim>
1257
+ <dim>-1</dim>
1258
+ <dim>-1</dim>
1259
+ </port>
1260
+ <port id="1" precision="FP32">
1261
+ <dim>128</dim>
1262
+ <dim>128</dim>
1263
+ <dim>3</dim>
1264
+ <dim>3</dim>
1265
+ </port>
1266
+ </input>
1267
+ <output>
1268
+ <port id="2" precision="FP32">
1269
+ <dim>-1</dim>
1270
+ <dim>128</dim>
1271
+ <dim>-1</dim>
1272
+ <dim>-1</dim>
1273
+ </port>
1274
+ </output>
1275
+ </layer>
1276
+ <layer id="70" name="__module.decoder.up.0.block.0.conv1.conv/aten::_convolution/Reshape" type="Const" version="opset1">
1277
+ <data element_type="f32" shape="1, 128, 1, 1" offset="2988564" size="512" />
1278
+ <output>
1279
+ <port id="0" precision="FP32">
1280
+ <dim>1</dim>
1281
+ <dim>128</dim>
1282
+ <dim>1</dim>
1283
+ <dim>1</dim>
1284
+ </port>
1285
+ </output>
1286
+ </layer>
1287
+ <layer id="71" name="__module.decoder.up.0.block.0.conv1.conv/aten::_convolution/Add" type="Add" version="opset1">
1288
+ <data auto_broadcast="numpy" />
1289
+ <input>
1290
+ <port id="0" precision="FP32">
1291
+ <dim>-1</dim>
1292
+ <dim>128</dim>
1293
+ <dim>-1</dim>
1294
+ <dim>-1</dim>
1295
+ </port>
1296
+ <port id="1" precision="FP32">
1297
+ <dim>1</dim>
1298
+ <dim>128</dim>
1299
+ <dim>1</dim>
1300
+ <dim>1</dim>
1301
+ </port>
1302
+ </input>
1303
+ <output>
1304
+ <port id="2" precision="FP32" names="144,x.11">
1305
+ <dim>-1</dim>
1306
+ <dim>128</dim>
1307
+ <dim>-1</dim>
1308
+ <dim>-1</dim>
1309
+ </port>
1310
+ </output>
1311
+ </layer>
1312
+ <layer id="72" name="Constant_203656" type="Const" version="opset1">
1313
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37376" size="4" />
1314
+ <output>
1315
+ <port id="0" precision="FP32">
1316
+ <dim>1</dim>
1317
+ <dim>1</dim>
1318
+ <dim>1</dim>
1319
+ <dim>1</dim>
1320
+ </port>
1321
+ </output>
1322
+ </layer>
1323
+ <layer id="73" name="__module.decoder.up.0.block.0.norm2/aten::pow/Power" type="Power" version="opset1">
1324
+ <data auto_broadcast="numpy" />
1325
+ <input>
1326
+ <port id="0" precision="FP32">
1327
+ <dim>-1</dim>
1328
+ <dim>128</dim>
1329
+ <dim>-1</dim>
1330
+ <dim>-1</dim>
1331
+ </port>
1332
+ <port id="1" precision="FP32">
1333
+ <dim>1</dim>
1334
+ <dim>1</dim>
1335
+ <dim>1</dim>
1336
+ <dim>1</dim>
1337
+ </port>
1338
+ </input>
1339
+ <output>
1340
+ <port id="2" precision="FP32" names="145">
1341
+ <dim>-1</dim>
1342
+ <dim>128</dim>
1343
+ <dim>-1</dim>
1344
+ <dim>-1</dim>
1345
+ </port>
1346
+ </output>
1347
+ </layer>
1348
+ <layer id="74" name="Constant_188889" type="Const" version="opset1">
1349
+ <data element_type="i64" shape="1" offset="37384" size="8" />
1350
+ <output>
1351
+ <port id="0" precision="I64">
1352
+ <dim>1</dim>
1353
+ </port>
1354
+ </output>
1355
+ </layer>
1356
+ <layer id="75" name="__module.decoder.up.0.block.0.norm2/aten::mean/ReduceMean" type="ReduceMean" version="opset1">
1357
+ <data keep_dims="true" />
1358
+ <input>
1359
+ <port id="0" precision="FP32">
1360
+ <dim>-1</dim>
1361
+ <dim>128</dim>
1362
+ <dim>-1</dim>
1363
+ <dim>-1</dim>
1364
+ </port>
1365
+ <port id="1" precision="I64">
1366
+ <dim>1</dim>
1367
+ </port>
1368
+ </input>
1369
+ <output>
1370
+ <port id="2" precision="FP32" names="147,mean_sq.11">
1371
+ <dim>-1</dim>
1372
+ <dim>1</dim>
1373
+ <dim>-1</dim>
1374
+ <dim>-1</dim>
1375
+ </port>
1376
+ </output>
1377
+ </layer>
1378
+ <layer id="76" name="Constant_203657" type="Const" version="opset1">
1379
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37392" size="4" />
1380
+ <output>
1381
+ <port id="0" precision="FP32">
1382
+ <dim>1</dim>
1383
+ <dim>1</dim>
1384
+ <dim>1</dim>
1385
+ <dim>1</dim>
1386
+ </port>
1387
+ </output>
1388
+ </layer>
1389
+ <layer id="77" name="__module.decoder.up.0.block.0.norm2/aten::add/Add" type="Add" version="opset1">
1390
+ <data auto_broadcast="numpy" />
1391
+ <input>
1392
+ <port id="0" precision="FP32">
1393
+ <dim>-1</dim>
1394
+ <dim>1</dim>
1395
+ <dim>-1</dim>
1396
+ <dim>-1</dim>
1397
+ </port>
1398
+ <port id="1" precision="FP32">
1399
+ <dim>1</dim>
1400
+ <dim>1</dim>
1401
+ <dim>1</dim>
1402
+ <dim>1</dim>
1403
+ </port>
1404
+ </input>
1405
+ <output>
1406
+ <port id="2" precision="FP32" names="148">
1407
+ <dim>-1</dim>
1408
+ <dim>1</dim>
1409
+ <dim>-1</dim>
1410
+ <dim>-1</dim>
1411
+ </port>
1412
+ </output>
1413
+ </layer>
1414
+ <layer id="78" name="__module.decoder.up.0.block.0.norm2/aten::sqrt/Sqrt" type="Sqrt" version="opset1">
1415
+ <input>
1416
+ <port id="0" precision="FP32">
1417
+ <dim>-1</dim>
1418
+ <dim>1</dim>
1419
+ <dim>-1</dim>
1420
+ <dim>-1</dim>
1421
+ </port>
1422
+ </input>
1423
+ <output>
1424
+ <port id="1" precision="FP32" names="149,rms.11">
1425
+ <dim>-1</dim>
1426
+ <dim>1</dim>
1427
+ <dim>-1</dim>
1428
+ <dim>-1</dim>
1429
+ </port>
1430
+ </output>
1431
+ </layer>
1432
+ <layer id="79" name="__module.decoder.up.0.block.0.norm2/aten::div/Divide" type="Divide" version="opset1">
1433
+ <data auto_broadcast="numpy" m_pythondiv="true" />
1434
+ <input>
1435
+ <port id="0" precision="FP32">
1436
+ <dim>-1</dim>
1437
+ <dim>128</dim>
1438
+ <dim>-1</dim>
1439
+ <dim>-1</dim>
1440
+ </port>
1441
+ <port id="1" precision="FP32">
1442
+ <dim>-1</dim>
1443
+ <dim>1</dim>
1444
+ <dim>-1</dim>
1445
+ <dim>-1</dim>
1446
+ </port>
1447
+ </input>
1448
+ <output>
1449
+ <port id="2" precision="FP32" names="150,input.37">
1450
+ <dim>-1</dim>
1451
+ <dim>128</dim>
1452
+ <dim>-1</dim>
1453
+ <dim>-1</dim>
1454
+ </port>
1455
+ </output>
1456
+ </layer>
1457
+ <layer id="80" name="__module.decoder.up.0.block.0.non_linearity/aten::silu/Swish_1" type="Swish" version="opset4">
1458
+ <input>
1459
+ <port id="0" precision="FP32">
1460
+ <dim>-1</dim>
1461
+ <dim>128</dim>
1462
+ <dim>-1</dim>
1463
+ <dim>-1</dim>
1464
+ </port>
1465
+ </input>
1466
+ <output>
1467
+ <port id="1" precision="FP32" names="151,input.39">
1468
+ <dim>-1</dim>
1469
+ <dim>128</dim>
1470
+ <dim>-1</dim>
1471
+ <dim>-1</dim>
1472
+ </port>
1473
+ </output>
1474
+ </layer>
1475
+ <layer id="81" name="self.decoder.up.0.block.0.conv2.conv.weight" type="Const" version="opset1">
1476
+ <data element_type="f32" shape="128, 128, 3, 3" offset="2989076" size="589824" />
1477
+ <output>
1478
+ <port id="0" precision="FP32" names="self.decoder.up.0.block.0.conv2.conv.weight">
1479
+ <dim>128</dim>
1480
+ <dim>128</dim>
1481
+ <dim>3</dim>
1482
+ <dim>3</dim>
1483
+ </port>
1484
+ </output>
1485
+ </layer>
1486
+ <layer id="82" name="__module.decoder.up.0.block.0.conv2.conv/aten::_convolution/Convolution" type="Convolution" version="opset1">
1487
+ <data strides="1, 1" dilations="1, 1" pads_begin="2, 1" pads_end="0, 1" auto_pad="explicit" />
1488
+ <input>
1489
+ <port id="0" precision="FP32">
1490
+ <dim>-1</dim>
1491
+ <dim>128</dim>
1492
+ <dim>-1</dim>
1493
+ <dim>-1</dim>
1494
+ </port>
1495
+ <port id="1" precision="FP32">
1496
+ <dim>128</dim>
1497
+ <dim>128</dim>
1498
+ <dim>3</dim>
1499
+ <dim>3</dim>
1500
+ </port>
1501
+ </input>
1502
+ <output>
1503
+ <port id="2" precision="FP32">
1504
+ <dim>-1</dim>
1505
+ <dim>128</dim>
1506
+ <dim>-1</dim>
1507
+ <dim>-1</dim>
1508
+ </port>
1509
+ </output>
1510
+ </layer>
1511
+ <layer id="83" name="__module.decoder.up.0.block.0.conv2.conv/aten::_convolution/Reshape" type="Const" version="opset1">
1512
+ <data element_type="f32" shape="1, 128, 1, 1" offset="3578900" size="512" />
1513
+ <output>
1514
+ <port id="0" precision="FP32">
1515
+ <dim>1</dim>
1516
+ <dim>128</dim>
1517
+ <dim>1</dim>
1518
+ <dim>1</dim>
1519
+ </port>
1520
+ </output>
1521
+ </layer>
1522
+ <layer id="84" name="__module.decoder.up.0.block.0.conv2.conv/aten::_convolution/Add" type="Add" version="opset1">
1523
+ <data auto_broadcast="numpy" />
1524
+ <input>
1525
+ <port id="0" precision="FP32">
1526
+ <dim>-1</dim>
1527
+ <dim>128</dim>
1528
+ <dim>-1</dim>
1529
+ <dim>-1</dim>
1530
+ </port>
1531
+ <port id="1" precision="FP32">
1532
+ <dim>1</dim>
1533
+ <dim>128</dim>
1534
+ <dim>1</dim>
1535
+ <dim>1</dim>
1536
+ </port>
1537
+ </input>
1538
+ <output>
1539
+ <port id="2" precision="FP32" names="162,h.5">
1540
+ <dim>-1</dim>
1541
+ <dim>128</dim>
1542
+ <dim>-1</dim>
1543
+ <dim>-1</dim>
1544
+ </port>
1545
+ </output>
1546
+ </layer>
1547
+ <layer id="85" name="__module.decoder.up.0.block.0/aten::add/Add" type="Add" version="opset1">
1548
+ <data auto_broadcast="numpy" />
1549
+ <input>
1550
+ <port id="0" precision="FP32">
1551
+ <dim>-1</dim>
1552
+ <dim>128</dim>
1553
+ <dim>-1</dim>
1554
+ <dim>-1</dim>
1555
+ </port>
1556
+ <port id="1" precision="FP32">
1557
+ <dim>-1</dim>
1558
+ <dim>128</dim>
1559
+ <dim>-1</dim>
1560
+ <dim>-1</dim>
1561
+ </port>
1562
+ </input>
1563
+ <output>
1564
+ <port id="2" precision="FP32" names="163,x.13">
1565
+ <dim>-1</dim>
1566
+ <dim>128</dim>
1567
+ <dim>-1</dim>
1568
+ <dim>-1</dim>
1569
+ </port>
1570
+ </output>
1571
+ </layer>
1572
+ <layer id="86" name="Constant_203658" type="Const" version="opset1">
1573
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37376" size="4" />
1574
+ <output>
1575
+ <port id="0" precision="FP32">
1576
+ <dim>1</dim>
1577
+ <dim>1</dim>
1578
+ <dim>1</dim>
1579
+ <dim>1</dim>
1580
+ </port>
1581
+ </output>
1582
+ </layer>
1583
+ <layer id="87" name="__module.decoder.up.0.block.1.norm1/aten::pow/Power" type="Power" version="opset1">
1584
+ <data auto_broadcast="numpy" />
1585
+ <input>
1586
+ <port id="0" precision="FP32">
1587
+ <dim>-1</dim>
1588
+ <dim>128</dim>
1589
+ <dim>-1</dim>
1590
+ <dim>-1</dim>
1591
+ </port>
1592
+ <port id="1" precision="FP32">
1593
+ <dim>1</dim>
1594
+ <dim>1</dim>
1595
+ <dim>1</dim>
1596
+ <dim>1</dim>
1597
+ </port>
1598
+ </input>
1599
+ <output>
1600
+ <port id="2" precision="FP32" names="166">
1601
+ <dim>-1</dim>
1602
+ <dim>128</dim>
1603
+ <dim>-1</dim>
1604
+ <dim>-1</dim>
1605
+ </port>
1606
+ </output>
1607
+ </layer>
1608
+ <layer id="88" name="Constant_189047" type="Const" version="opset1">
1609
+ <data element_type="i64" shape="1" offset="37384" size="8" />
1610
+ <output>
1611
+ <port id="0" precision="I64">
1612
+ <dim>1</dim>
1613
+ </port>
1614
+ </output>
1615
+ </layer>
1616
+ <layer id="89" name="__module.decoder.up.0.block.1.norm1/aten::mean/ReduceMean" type="ReduceMean" version="opset1">
1617
+ <data keep_dims="true" />
1618
+ <input>
1619
+ <port id="0" precision="FP32">
1620
+ <dim>-1</dim>
1621
+ <dim>128</dim>
1622
+ <dim>-1</dim>
1623
+ <dim>-1</dim>
1624
+ </port>
1625
+ <port id="1" precision="I64">
1626
+ <dim>1</dim>
1627
+ </port>
1628
+ </input>
1629
+ <output>
1630
+ <port id="2" precision="FP32" names="168,mean_sq.13">
1631
+ <dim>-1</dim>
1632
+ <dim>1</dim>
1633
+ <dim>-1</dim>
1634
+ <dim>-1</dim>
1635
+ </port>
1636
+ </output>
1637
+ </layer>
1638
+ <layer id="90" name="Constant_203659" type="Const" version="opset1">
1639
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37392" size="4" />
1640
+ <output>
1641
+ <port id="0" precision="FP32">
1642
+ <dim>1</dim>
1643
+ <dim>1</dim>
1644
+ <dim>1</dim>
1645
+ <dim>1</dim>
1646
+ </port>
1647
+ </output>
1648
+ </layer>
1649
+ <layer id="91" name="__module.decoder.up.0.block.1.norm1/aten::add/Add" type="Add" version="opset1">
1650
+ <data auto_broadcast="numpy" />
1651
+ <input>
1652
+ <port id="0" precision="FP32">
1653
+ <dim>-1</dim>
1654
+ <dim>1</dim>
1655
+ <dim>-1</dim>
1656
+ <dim>-1</dim>
1657
+ </port>
1658
+ <port id="1" precision="FP32">
1659
+ <dim>1</dim>
1660
+ <dim>1</dim>
1661
+ <dim>1</dim>
1662
+ <dim>1</dim>
1663
+ </port>
1664
+ </input>
1665
+ <output>
1666
+ <port id="2" precision="FP32" names="169">
1667
+ <dim>-1</dim>
1668
+ <dim>1</dim>
1669
+ <dim>-1</dim>
1670
+ <dim>-1</dim>
1671
+ </port>
1672
+ </output>
1673
+ </layer>
1674
+ <layer id="92" name="__module.decoder.up.0.block.1.norm1/aten::sqrt/Sqrt" type="Sqrt" version="opset1">
1675
+ <input>
1676
+ <port id="0" precision="FP32">
1677
+ <dim>-1</dim>
1678
+ <dim>1</dim>
1679
+ <dim>-1</dim>
1680
+ <dim>-1</dim>
1681
+ </port>
1682
+ </input>
1683
+ <output>
1684
+ <port id="1" precision="FP32" names="170,rms.13">
1685
+ <dim>-1</dim>
1686
+ <dim>1</dim>
1687
+ <dim>-1</dim>
1688
+ <dim>-1</dim>
1689
+ </port>
1690
+ </output>
1691
+ </layer>
1692
+ <layer id="93" name="__module.decoder.up.0.block.1.norm1/aten::div/Divide" type="Divide" version="opset1">
1693
+ <data auto_broadcast="numpy" m_pythondiv="true" />
1694
+ <input>
1695
+ <port id="0" precision="FP32">
1696
+ <dim>-1</dim>
1697
+ <dim>128</dim>
1698
+ <dim>-1</dim>
1699
+ <dim>-1</dim>
1700
+ </port>
1701
+ <port id="1" precision="FP32">
1702
+ <dim>-1</dim>
1703
+ <dim>1</dim>
1704
+ <dim>-1</dim>
1705
+ <dim>-1</dim>
1706
+ </port>
1707
+ </input>
1708
+ <output>
1709
+ <port id="2" precision="FP32" names="171,input.45">
1710
+ <dim>-1</dim>
1711
+ <dim>128</dim>
1712
+ <dim>-1</dim>
1713
+ <dim>-1</dim>
1714
+ </port>
1715
+ </output>
1716
+ </layer>
1717
+ <layer id="94" name="__module.decoder.up.0.block.1.non_linearity/aten::silu/Swish" type="Swish" version="opset4">
1718
+ <input>
1719
+ <port id="0" precision="FP32">
1720
+ <dim>-1</dim>
1721
+ <dim>128</dim>
1722
+ <dim>-1</dim>
1723
+ <dim>-1</dim>
1724
+ </port>
1725
+ </input>
1726
+ <output>
1727
+ <port id="1" precision="FP32" names="172,input.47">
1728
+ <dim>-1</dim>
1729
+ <dim>128</dim>
1730
+ <dim>-1</dim>
1731
+ <dim>-1</dim>
1732
+ </port>
1733
+ </output>
1734
+ </layer>
1735
+ <layer id="95" name="self.decoder.up.0.block.1.conv1.conv.weight" type="Const" version="opset1">
1736
+ <data element_type="f32" shape="128, 128, 3, 3" offset="3579412" size="589824" />
1737
+ <output>
1738
+ <port id="0" precision="FP32" names="self.decoder.up.0.block.1.conv1.conv.weight">
1739
+ <dim>128</dim>
1740
+ <dim>128</dim>
1741
+ <dim>3</dim>
1742
+ <dim>3</dim>
1743
+ </port>
1744
+ </output>
1745
+ </layer>
1746
+ <layer id="96" name="__module.decoder.up.0.block.1.conv1.conv/aten::_convolution/Convolution" type="Convolution" version="opset1">
1747
+ <data strides="1, 1" dilations="1, 1" pads_begin="2, 1" pads_end="0, 1" auto_pad="explicit" />
1748
+ <input>
1749
+ <port id="0" precision="FP32">
1750
+ <dim>-1</dim>
1751
+ <dim>128</dim>
1752
+ <dim>-1</dim>
1753
+ <dim>-1</dim>
1754
+ </port>
1755
+ <port id="1" precision="FP32">
1756
+ <dim>128</dim>
1757
+ <dim>128</dim>
1758
+ <dim>3</dim>
1759
+ <dim>3</dim>
1760
+ </port>
1761
+ </input>
1762
+ <output>
1763
+ <port id="2" precision="FP32">
1764
+ <dim>-1</dim>
1765
+ <dim>128</dim>
1766
+ <dim>-1</dim>
1767
+ <dim>-1</dim>
1768
+ </port>
1769
+ </output>
1770
+ </layer>
1771
+ <layer id="97" name="__module.decoder.up.0.block.1.conv1.conv/aten::_convolution/Reshape" type="Const" version="opset1">
1772
+ <data element_type="f32" shape="1, 128, 1, 1" offset="4169236" size="512" />
1773
+ <output>
1774
+ <port id="0" precision="FP32">
1775
+ <dim>1</dim>
1776
+ <dim>128</dim>
1777
+ <dim>1</dim>
1778
+ <dim>1</dim>
1779
+ </port>
1780
+ </output>
1781
+ </layer>
1782
+ <layer id="98" name="__module.decoder.up.0.block.1.conv1.conv/aten::_convolution/Add" type="Add" version="opset1">
1783
+ <data auto_broadcast="numpy" />
1784
+ <input>
1785
+ <port id="0" precision="FP32">
1786
+ <dim>-1</dim>
1787
+ <dim>128</dim>
1788
+ <dim>-1</dim>
1789
+ <dim>-1</dim>
1790
+ </port>
1791
+ <port id="1" precision="FP32">
1792
+ <dim>1</dim>
1793
+ <dim>128</dim>
1794
+ <dim>1</dim>
1795
+ <dim>1</dim>
1796
+ </port>
1797
+ </input>
1798
+ <output>
1799
+ <port id="2" precision="FP32" names="182,x.15">
1800
+ <dim>-1</dim>
1801
+ <dim>128</dim>
1802
+ <dim>-1</dim>
1803
+ <dim>-1</dim>
1804
+ </port>
1805
+ </output>
1806
+ </layer>
1807
+ <layer id="99" name="Constant_203660" type="Const" version="opset1">
1808
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37376" size="4" />
1809
+ <output>
1810
+ <port id="0" precision="FP32">
1811
+ <dim>1</dim>
1812
+ <dim>1</dim>
1813
+ <dim>1</dim>
1814
+ <dim>1</dim>
1815
+ </port>
1816
+ </output>
1817
+ </layer>
1818
+ <layer id="100" name="__module.decoder.up.0.block.1.norm2/aten::pow/Power" type="Power" version="opset1">
1819
+ <data auto_broadcast="numpy" />
1820
+ <input>
1821
+ <port id="0" precision="FP32">
1822
+ <dim>-1</dim>
1823
+ <dim>128</dim>
1824
+ <dim>-1</dim>
1825
+ <dim>-1</dim>
1826
+ </port>
1827
+ <port id="1" precision="FP32">
1828
+ <dim>1</dim>
1829
+ <dim>1</dim>
1830
+ <dim>1</dim>
1831
+ <dim>1</dim>
1832
+ </port>
1833
+ </input>
1834
+ <output>
1835
+ <port id="2" precision="FP32" names="183">
1836
+ <dim>-1</dim>
1837
+ <dim>128</dim>
1838
+ <dim>-1</dim>
1839
+ <dim>-1</dim>
1840
+ </port>
1841
+ </output>
1842
+ </layer>
1843
+ <layer id="101" name="Constant_189197" type="Const" version="opset1">
1844
+ <data element_type="i64" shape="1" offset="37384" size="8" />
1845
+ <output>
1846
+ <port id="0" precision="I64">
1847
+ <dim>1</dim>
1848
+ </port>
1849
+ </output>
1850
+ </layer>
1851
+ <layer id="102" name="__module.decoder.up.0.block.1.norm2/aten::mean/ReduceMean" type="ReduceMean" version="opset1">
1852
+ <data keep_dims="true" />
1853
+ <input>
1854
+ <port id="0" precision="FP32">
1855
+ <dim>-1</dim>
1856
+ <dim>128</dim>
1857
+ <dim>-1</dim>
1858
+ <dim>-1</dim>
1859
+ </port>
1860
+ <port id="1" precision="I64">
1861
+ <dim>1</dim>
1862
+ </port>
1863
+ </input>
1864
+ <output>
1865
+ <port id="2" precision="FP32" names="185,mean_sq.15">
1866
+ <dim>-1</dim>
1867
+ <dim>1</dim>
1868
+ <dim>-1</dim>
1869
+ <dim>-1</dim>
1870
+ </port>
1871
+ </output>
1872
+ </layer>
1873
+ <layer id="103" name="Constant_203661" type="Const" version="opset1">
1874
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37392" size="4" />
1875
+ <output>
1876
+ <port id="0" precision="FP32">
1877
+ <dim>1</dim>
1878
+ <dim>1</dim>
1879
+ <dim>1</dim>
1880
+ <dim>1</dim>
1881
+ </port>
1882
+ </output>
1883
+ </layer>
1884
+ <layer id="104" name="__module.decoder.up.0.block.1.norm2/aten::add/Add" type="Add" version="opset1">
1885
+ <data auto_broadcast="numpy" />
1886
+ <input>
1887
+ <port id="0" precision="FP32">
1888
+ <dim>-1</dim>
1889
+ <dim>1</dim>
1890
+ <dim>-1</dim>
1891
+ <dim>-1</dim>
1892
+ </port>
1893
+ <port id="1" precision="FP32">
1894
+ <dim>1</dim>
1895
+ <dim>1</dim>
1896
+ <dim>1</dim>
1897
+ <dim>1</dim>
1898
+ </port>
1899
+ </input>
1900
+ <output>
1901
+ <port id="2" precision="FP32" names="186">
1902
+ <dim>-1</dim>
1903
+ <dim>1</dim>
1904
+ <dim>-1</dim>
1905
+ <dim>-1</dim>
1906
+ </port>
1907
+ </output>
1908
+ </layer>
1909
+ <layer id="105" name="__module.decoder.up.0.block.1.norm2/aten::sqrt/Sqrt" type="Sqrt" version="opset1">
1910
+ <input>
1911
+ <port id="0" precision="FP32">
1912
+ <dim>-1</dim>
1913
+ <dim>1</dim>
1914
+ <dim>-1</dim>
1915
+ <dim>-1</dim>
1916
+ </port>
1917
+ </input>
1918
+ <output>
1919
+ <port id="1" precision="FP32" names="187,rms.15">
1920
+ <dim>-1</dim>
1921
+ <dim>1</dim>
1922
+ <dim>-1</dim>
1923
+ <dim>-1</dim>
1924
+ </port>
1925
+ </output>
1926
+ </layer>
1927
+ <layer id="106" name="__module.decoder.up.0.block.1.norm2/aten::div/Divide" type="Divide" version="opset1">
1928
+ <data auto_broadcast="numpy" m_pythondiv="true" />
1929
+ <input>
1930
+ <port id="0" precision="FP32">
1931
+ <dim>-1</dim>
1932
+ <dim>128</dim>
1933
+ <dim>-1</dim>
1934
+ <dim>-1</dim>
1935
+ </port>
1936
+ <port id="1" precision="FP32">
1937
+ <dim>-1</dim>
1938
+ <dim>1</dim>
1939
+ <dim>-1</dim>
1940
+ <dim>-1</dim>
1941
+ </port>
1942
+ </input>
1943
+ <output>
1944
+ <port id="2" precision="FP32" names="188,input.51">
1945
+ <dim>-1</dim>
1946
+ <dim>128</dim>
1947
+ <dim>-1</dim>
1948
+ <dim>-1</dim>
1949
+ </port>
1950
+ </output>
1951
+ </layer>
1952
+ <layer id="107" name="__module.decoder.up.0.block.1.non_linearity/aten::silu/Swish_1" type="Swish" version="opset4">
1953
+ <input>
1954
+ <port id="0" precision="FP32">
1955
+ <dim>-1</dim>
1956
+ <dim>128</dim>
1957
+ <dim>-1</dim>
1958
+ <dim>-1</dim>
1959
+ </port>
1960
+ </input>
1961
+ <output>
1962
+ <port id="1" precision="FP32" names="189,input.53">
1963
+ <dim>-1</dim>
1964
+ <dim>128</dim>
1965
+ <dim>-1</dim>
1966
+ <dim>-1</dim>
1967
+ </port>
1968
+ </output>
1969
+ </layer>
1970
+ <layer id="108" name="self.decoder.up.0.block.1.conv2.conv.weight" type="Const" version="opset1">
1971
+ <data element_type="f32" shape="128, 128, 3, 3" offset="4169748" size="589824" />
1972
+ <output>
1973
+ <port id="0" precision="FP32" names="self.decoder.up.0.block.1.conv2.conv.weight">
1974
+ <dim>128</dim>
1975
+ <dim>128</dim>
1976
+ <dim>3</dim>
1977
+ <dim>3</dim>
1978
+ </port>
1979
+ </output>
1980
+ </layer>
1981
+ <layer id="109" name="__module.decoder.up.0.block.1.conv2.conv/aten::_convolution/Convolution" type="Convolution" version="opset1">
1982
+ <data strides="1, 1" dilations="1, 1" pads_begin="2, 1" pads_end="0, 1" auto_pad="explicit" />
1983
+ <input>
1984
+ <port id="0" precision="FP32">
1985
+ <dim>-1</dim>
1986
+ <dim>128</dim>
1987
+ <dim>-1</dim>
1988
+ <dim>-1</dim>
1989
+ </port>
1990
+ <port id="1" precision="FP32">
1991
+ <dim>128</dim>
1992
+ <dim>128</dim>
1993
+ <dim>3</dim>
1994
+ <dim>3</dim>
1995
+ </port>
1996
+ </input>
1997
+ <output>
1998
+ <port id="2" precision="FP32">
1999
+ <dim>-1</dim>
2000
+ <dim>128</dim>
2001
+ <dim>-1</dim>
2002
+ <dim>-1</dim>
2003
+ </port>
2004
+ </output>
2005
+ </layer>
2006
+ <layer id="110" name="__module.decoder.up.0.block.1.conv2.conv/aten::_convolution/Reshape" type="Const" version="opset1">
2007
+ <data element_type="f32" shape="1, 128, 1, 1" offset="4759572" size="512" />
2008
+ <output>
2009
+ <port id="0" precision="FP32">
2010
+ <dim>1</dim>
2011
+ <dim>128</dim>
2012
+ <dim>1</dim>
2013
+ <dim>1</dim>
2014
+ </port>
2015
+ </output>
2016
+ </layer>
2017
+ <layer id="111" name="__module.decoder.up.0.block.1.conv2.conv/aten::_convolution/Add" type="Add" version="opset1">
2018
+ <data auto_broadcast="numpy" />
2019
+ <input>
2020
+ <port id="0" precision="FP32">
2021
+ <dim>-1</dim>
2022
+ <dim>128</dim>
2023
+ <dim>-1</dim>
2024
+ <dim>-1</dim>
2025
+ </port>
2026
+ <port id="1" precision="FP32">
2027
+ <dim>1</dim>
2028
+ <dim>128</dim>
2029
+ <dim>1</dim>
2030
+ <dim>1</dim>
2031
+ </port>
2032
+ </input>
2033
+ <output>
2034
+ <port id="2" precision="FP32" names="200,h">
2035
+ <dim>-1</dim>
2036
+ <dim>128</dim>
2037
+ <dim>-1</dim>
2038
+ <dim>-1</dim>
2039
+ </port>
2040
+ </output>
2041
+ </layer>
2042
+ <layer id="112" name="__module.decoder.up.0.block.1/aten::add/Add" type="Add" version="opset1">
2043
+ <data auto_broadcast="numpy" />
2044
+ <input>
2045
+ <port id="0" precision="FP32">
2046
+ <dim>-1</dim>
2047
+ <dim>128</dim>
2048
+ <dim>-1</dim>
2049
+ <dim>-1</dim>
2050
+ </port>
2051
+ <port id="1" precision="FP32">
2052
+ <dim>-1</dim>
2053
+ <dim>128</dim>
2054
+ <dim>-1</dim>
2055
+ <dim>-1</dim>
2056
+ </port>
2057
+ </input>
2058
+ <output>
2059
+ <port id="2" precision="FP32" names="201,x">
2060
+ <dim>-1</dim>
2061
+ <dim>128</dim>
2062
+ <dim>-1</dim>
2063
+ <dim>-1</dim>
2064
+ </port>
2065
+ </output>
2066
+ </layer>
2067
+ <layer id="113" name="Constant_203662" type="Const" version="opset1">
2068
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37376" size="4" />
2069
+ <output>
2070
+ <port id="0" precision="FP32">
2071
+ <dim>1</dim>
2072
+ <dim>1</dim>
2073
+ <dim>1</dim>
2074
+ <dim>1</dim>
2075
+ </port>
2076
+ </output>
2077
+ </layer>
2078
+ <layer id="114" name="__module.decoder.norm_out/aten::pow/Power" type="Power" version="opset1">
2079
+ <data auto_broadcast="numpy" />
2080
+ <input>
2081
+ <port id="0" precision="FP32">
2082
+ <dim>-1</dim>
2083
+ <dim>128</dim>
2084
+ <dim>-1</dim>
2085
+ <dim>-1</dim>
2086
+ </port>
2087
+ <port id="1" precision="FP32">
2088
+ <dim>1</dim>
2089
+ <dim>1</dim>
2090
+ <dim>1</dim>
2091
+ <dim>1</dim>
2092
+ </port>
2093
+ </input>
2094
+ <output>
2095
+ <port id="2" precision="FP32" names="202">
2096
+ <dim>-1</dim>
2097
+ <dim>128</dim>
2098
+ <dim>-1</dim>
2099
+ <dim>-1</dim>
2100
+ </port>
2101
+ </output>
2102
+ </layer>
2103
+ <layer id="115" name="Constant_189353" type="Const" version="opset1">
2104
+ <data element_type="i64" shape="1" offset="37384" size="8" />
2105
+ <output>
2106
+ <port id="0" precision="I64">
2107
+ <dim>1</dim>
2108
+ </port>
2109
+ </output>
2110
+ </layer>
2111
+ <layer id="116" name="__module.decoder.norm_out/aten::mean/ReduceMean" type="ReduceMean" version="opset1">
2112
+ <data keep_dims="true" />
2113
+ <input>
2114
+ <port id="0" precision="FP32">
2115
+ <dim>-1</dim>
2116
+ <dim>128</dim>
2117
+ <dim>-1</dim>
2118
+ <dim>-1</dim>
2119
+ </port>
2120
+ <port id="1" precision="I64">
2121
+ <dim>1</dim>
2122
+ </port>
2123
+ </input>
2124
+ <output>
2125
+ <port id="2" precision="FP32" names="204,mean_sq">
2126
+ <dim>-1</dim>
2127
+ <dim>1</dim>
2128
+ <dim>-1</dim>
2129
+ <dim>-1</dim>
2130
+ </port>
2131
+ </output>
2132
+ </layer>
2133
+ <layer id="117" name="Constant_203663" type="Const" version="opset1">
2134
+ <data element_type="f32" shape="1, 1, 1, 1" offset="37392" size="4" />
2135
+ <output>
2136
+ <port id="0" precision="FP32">
2137
+ <dim>1</dim>
2138
+ <dim>1</dim>
2139
+ <dim>1</dim>
2140
+ <dim>1</dim>
2141
+ </port>
2142
+ </output>
2143
+ </layer>
2144
+ <layer id="118" name="__module.decoder.norm_out/aten::add/Add" type="Add" version="opset1">
2145
+ <data auto_broadcast="numpy" />
2146
+ <input>
2147
+ <port id="0" precision="FP32">
2148
+ <dim>-1</dim>
2149
+ <dim>1</dim>
2150
+ <dim>-1</dim>
2151
+ <dim>-1</dim>
2152
+ </port>
2153
+ <port id="1" precision="FP32">
2154
+ <dim>1</dim>
2155
+ <dim>1</dim>
2156
+ <dim>1</dim>
2157
+ <dim>1</dim>
2158
+ </port>
2159
+ </input>
2160
+ <output>
2161
+ <port id="2" precision="FP32" names="205">
2162
+ <dim>-1</dim>
2163
+ <dim>1</dim>
2164
+ <dim>-1</dim>
2165
+ <dim>-1</dim>
2166
+ </port>
2167
+ </output>
2168
+ </layer>
2169
+ <layer id="119" name="__module.decoder.norm_out/aten::sqrt/Sqrt" type="Sqrt" version="opset1">
2170
+ <input>
2171
+ <port id="0" precision="FP32">
2172
+ <dim>-1</dim>
2173
+ <dim>1</dim>
2174
+ <dim>-1</dim>
2175
+ <dim>-1</dim>
2176
+ </port>
2177
+ </input>
2178
+ <output>
2179
+ <port id="1" precision="FP32" names="206,rms">
2180
+ <dim>-1</dim>
2181
+ <dim>1</dim>
2182
+ <dim>-1</dim>
2183
+ <dim>-1</dim>
2184
+ </port>
2185
+ </output>
2186
+ </layer>
2187
+ <layer id="120" name="__module.decoder.norm_out/aten::div/Divide" type="Divide" version="opset1">
2188
+ <data auto_broadcast="numpy" m_pythondiv="true" />
2189
+ <input>
2190
+ <port id="0" precision="FP32">
2191
+ <dim>-1</dim>
2192
+ <dim>128</dim>
2193
+ <dim>-1</dim>
2194
+ <dim>-1</dim>
2195
+ </port>
2196
+ <port id="1" precision="FP32">
2197
+ <dim>-1</dim>
2198
+ <dim>1</dim>
2199
+ <dim>-1</dim>
2200
+ <dim>-1</dim>
2201
+ </port>
2202
+ </input>
2203
+ <output>
2204
+ <port id="2" precision="FP32" names="207,input.59">
2205
+ <dim>-1</dim>
2206
+ <dim>128</dim>
2207
+ <dim>-1</dim>
2208
+ <dim>-1</dim>
2209
+ </port>
2210
+ </output>
2211
+ </layer>
2212
+ <layer id="121" name="__module.decoder.non_linearity/aten::silu/Swish" type="Swish" version="opset4">
2213
+ <input>
2214
+ <port id="0" precision="FP32">
2215
+ <dim>-1</dim>
2216
+ <dim>128</dim>
2217
+ <dim>-1</dim>
2218
+ <dim>-1</dim>
2219
+ </port>
2220
+ </input>
2221
+ <output>
2222
+ <port id="1" precision="FP32" names="208,input.61">
2223
+ <dim>-1</dim>
2224
+ <dim>128</dim>
2225
+ <dim>-1</dim>
2226
+ <dim>-1</dim>
2227
+ </port>
2228
+ </output>
2229
+ </layer>
2230
+ <layer id="122" name="self.decoder.conv_out.conv.weight" type="Const" version="opset1">
2231
+ <data element_type="f32" shape="2, 128, 3, 3" offset="4760084" size="9216" />
2232
+ <output>
2233
+ <port id="0" precision="FP32" names="self.decoder.conv_out.conv.weight">
2234
+ <dim>2</dim>
2235
+ <dim>128</dim>
2236
+ <dim>3</dim>
2237
+ <dim>3</dim>
2238
+ </port>
2239
+ </output>
2240
+ </layer>
2241
+ <layer id="123" name="__module.decoder.conv_out.conv/aten::_convolution/Convolution" type="Convolution" version="opset1">
2242
+ <data strides="1, 1" dilations="1, 1" pads_begin="2, 1" pads_end="0, 1" auto_pad="explicit" />
2243
+ <input>
2244
+ <port id="0" precision="FP32">
2245
+ <dim>-1</dim>
2246
+ <dim>128</dim>
2247
+ <dim>-1</dim>
2248
+ <dim>-1</dim>
2249
+ </port>
2250
+ <port id="1" precision="FP32">
2251
+ <dim>2</dim>
2252
+ <dim>128</dim>
2253
+ <dim>3</dim>
2254
+ <dim>3</dim>
2255
+ </port>
2256
+ </input>
2257
+ <output>
2258
+ <port id="2" precision="FP32">
2259
+ <dim>-1</dim>
2260
+ <dim>2</dim>
2261
+ <dim>-1</dim>
2262
+ <dim>-1</dim>
2263
+ </port>
2264
+ </output>
2265
+ </layer>
2266
+ <layer id="124" name="__module.decoder.conv_out.conv/aten::_convolution/Reshape" type="Const" version="opset1">
2267
+ <data element_type="f32" shape="1, 2, 1, 1" offset="4769300" size="8" />
2268
+ <output>
2269
+ <port id="0" precision="FP32">
2270
+ <dim>1</dim>
2271
+ <dim>2</dim>
2272
+ <dim>1</dim>
2273
+ <dim>1</dim>
2274
+ </port>
2275
+ </output>
2276
+ </layer>
2277
+ <layer id="125" name="__module.decoder.conv_out.conv/aten::_convolution/Add" type="Add" version="opset1">
2278
+ <data auto_broadcast="numpy" />
2279
+ <input>
2280
+ <port id="0" precision="FP32">
2281
+ <dim>-1</dim>
2282
+ <dim>2</dim>
2283
+ <dim>-1</dim>
2284
+ <dim>-1</dim>
2285
+ </port>
2286
+ <port id="1" precision="FP32">
2287
+ <dim>1</dim>
2288
+ <dim>2</dim>
2289
+ <dim>1</dim>
2290
+ <dim>1</dim>
2291
+ </port>
2292
+ </input>
2293
+ <output>
2294
+ <port id="2" precision="FP32" names="218,221,decoded_output.1">
2295
+ <dim>-1</dim>
2296
+ <dim>2</dim>
2297
+ <dim>-1</dim>
2298
+ <dim>-1</dim>
2299
+ </port>
2300
+ </output>
2301
+ </layer>
2302
+ <layer id="126" name="Constant_200852" type="Const" version="opset1">
2303
+ <data element_type="i64" shape="3" offset="4769312" size="24" />
2304
+ <output>
2305
+ <port id="0" precision="I64">
2306
+ <dim>3</dim>
2307
+ </port>
2308
+ </output>
2309
+ </layer>
2310
+ <layer id="127" name="Constant_189503" type="Const" version="opset1">
2311
+ <data element_type="i64" shape="1" offset="4769336" size="8" />
2312
+ <output>
2313
+ <port id="0" precision="I64">
2314
+ <dim>1</dim>
2315
+ </port>
2316
+ </output>
2317
+ </layer>
2318
+ <layer id="128" name="ShapeOf_204327" type="ShapeOf" version="opset3">
2319
+ <data output_type="i64" />
2320
+ <input>
2321
+ <port id="0" precision="FP32">
2322
+ <dim>-1</dim>
2323
+ <dim>2</dim>
2324
+ <dim>-1</dim>
2325
+ <dim>-1</dim>
2326
+ </port>
2327
+ </input>
2328
+ <output>
2329
+ <port id="1" precision="I64">
2330
+ <dim>4</dim>
2331
+ </port>
2332
+ </output>
2333
+ </layer>
2334
+ <layer id="129" name="Constant_204458" type="Const" version="opset1">
2335
+ <data element_type="i64" shape="2" offset="4769344" size="16" />
2336
+ <output>
2337
+ <port id="0" precision="I64">
2338
+ <dim>2</dim>
2339
+ </port>
2340
+ </output>
2341
+ </layer>
2342
+ <layer id="130" name="Constant_204459" type="Const" version="opset1">
2343
+ <data element_type="i64" shape="" offset="4769360" size="8" />
2344
+ <output>
2345
+ <port id="0" precision="I64" />
2346
+ </output>
2347
+ </layer>
2348
+ <layer id="131" name="Gather_204460" type="Gather" version="opset8">
2349
+ <data batch_dims="0" />
2350
+ <input>
2351
+ <port id="0" precision="I64">
2352
+ <dim>4</dim>
2353
+ </port>
2354
+ <port id="1" precision="I64">
2355
+ <dim>2</dim>
2356
+ </port>
2357
+ <port id="2" precision="I64" />
2358
+ </input>
2359
+ <output>
2360
+ <port id="3" precision="I64">
2361
+ <dim>2</dim>
2362
+ </port>
2363
+ </output>
2364
+ </layer>
2365
+ <layer id="132" name="Concat_200847" type="Concat" version="opset1">
2366
+ <data axis="0" />
2367
+ <input>
2368
+ <port id="0" precision="I64">
2369
+ <dim>1</dim>
2370
+ </port>
2371
+ <port id="1" precision="I64">
2372
+ <dim>2</dim>
2373
+ </port>
2374
+ </input>
2375
+ <output>
2376
+ <port id="2" precision="I64">
2377
+ <dim>3</dim>
2378
+ </port>
2379
+ </output>
2380
+ </layer>
2381
+ <layer id="133" name="Constant_200851" type="Const" version="opset1">
2382
+ <data element_type="i64" shape="3" offset="4769368" size="24" />
2383
+ <output>
2384
+ <port id="0" precision="I64">
2385
+ <dim>3</dim>
2386
+ </port>
2387
+ </output>
2388
+ </layer>
2389
+ <layer id="134" name="Constant_200850" type="Const" version="opset1">
2390
+ <data element_type="i64" shape="3" offset="4769392" size="24" />
2391
+ <output>
2392
+ <port id="0" precision="I64">
2393
+ <dim>3</dim>
2394
+ </port>
2395
+ </output>
2396
+ </layer>
2397
+ <layer id="135" name="__module.decoder/aten::slice/Slice_3" type="Slice" version="opset8">
2398
+ <input>
2399
+ <port id="0" precision="FP32">
2400
+ <dim>-1</dim>
2401
+ <dim>2</dim>
2402
+ <dim>-1</dim>
2403
+ <dim>-1</dim>
2404
+ </port>
2405
+ <port id="1" precision="I64">
2406
+ <dim>3</dim>
2407
+ </port>
2408
+ <port id="2" precision="I64">
2409
+ <dim>3</dim>
2410
+ </port>
2411
+ <port id="3" precision="I64">
2412
+ <dim>3</dim>
2413
+ </port>
2414
+ <port id="4" precision="I64">
2415
+ <dim>3</dim>
2416
+ </port>
2417
+ </input>
2418
+ <output>
2419
+ <port id="5" precision="FP32" names="224,decoded_output.3">
2420
+ <dim>-1</dim>
2421
+ <dim>2</dim>
2422
+ <dim>-1</dim>
2423
+ <dim>-1</dim>
2424
+ </port>
2425
+ </output>
2426
+ </layer>
2427
+ <layer id="136" name="__module.decoder/aten::pad/ConvertLike" type="Const" version="opset1">
2428
+ <data element_type="i64" shape="2" offset="4769416" size="16" />
2429
+ <output>
2430
+ <port id="0" precision="I64">
2431
+ <dim>2</dim>
2432
+ </port>
2433
+ </output>
2434
+ </layer>
2435
+ <layer id="137" name="Constant_189553" type="Const" version="opset1">
2436
+ <data element_type="i64" shape="1" offset="4769360" size="8" />
2437
+ <output>
2438
+ <port id="0" precision="I64">
2439
+ <dim>1</dim>
2440
+ </port>
2441
+ </output>
2442
+ </layer>
2443
+ <layer id="138" name="7" type="Const" version="opset1">
2444
+ <data element_type="i64" shape="" offset="4769432" size="8" />
2445
+ <output>
2446
+ <port id="0" precision="I64" names="7" />
2447
+ </output>
2448
+ </layer>
2449
+ <layer id="139" name="ShapeOf_204339" type="ShapeOf" version="opset3">
2450
+ <data output_type="i64" />
2451
+ <input>
2452
+ <port id="0" precision="FP32">
2453
+ <dim>-1</dim>
2454
+ <dim>2</dim>
2455
+ <dim>-1</dim>
2456
+ <dim>-1</dim>
2457
+ </port>
2458
+ </input>
2459
+ <output>
2460
+ <port id="1" precision="I64">
2461
+ <dim>4</dim>
2462
+ </port>
2463
+ </output>
2464
+ </layer>
2465
+ <layer id="140" name="Constant_204340" type="Const" version="opset1">
2466
+ <data element_type="i64" shape="" offset="4769440" size="8" />
2467
+ <output>
2468
+ <port id="0" precision="I64" />
2469
+ </output>
2470
+ </layer>
2471
+ <layer id="141" name="Constant_204341" type="Const" version="opset1">
2472
+ <data element_type="i64" shape="" offset="4769360" size="8" />
2473
+ <output>
2474
+ <port id="0" precision="I64" />
2475
+ </output>
2476
+ </layer>
2477
+ <layer id="142" name="Gather_204342" type="Gather" version="opset8">
2478
+ <data batch_dims="0" />
2479
+ <input>
2480
+ <port id="0" precision="I64">
2481
+ <dim>4</dim>
2482
+ </port>
2483
+ <port id="1" precision="I64" />
2484
+ <port id="2" precision="I64" />
2485
+ </input>
2486
+ <output>
2487
+ <port id="3" precision="I64" names="229" />
2488
+ </output>
2489
+ </layer>
2490
+ <layer id="143" name="__module.decoder/aten::rsub/Subtract" type="Subtract" version="opset1">
2491
+ <data auto_broadcast="numpy" />
2492
+ <input>
2493
+ <port id="0" precision="I64" />
2494
+ <port id="1" precision="I64" />
2495
+ </input>
2496
+ <output>
2497
+ <port id="2" precision="I64" names="231,232,freq_padding_needed" />
2498
+ </output>
2499
+ </layer>
2500
+ <layer id="144" name="Constant_189548" type="Const" version="opset1">
2501
+ <data element_type="i32" shape="" offset="4769448" size="4" />
2502
+ <output>
2503
+ <port id="0" precision="I32" />
2504
+ </output>
2505
+ </layer>
2506
+ <layer id="145" name="Unsqueeze_189550" type="Unsqueeze" version="opset1">
2507
+ <input>
2508
+ <port id="0" precision="I64" />
2509
+ <port id="1" precision="I32" />
2510
+ </input>
2511
+ <output>
2512
+ <port id="2" precision="I64">
2513
+ <dim>1</dim>
2514
+ </port>
2515
+ </output>
2516
+ </layer>
2517
+ <layer id="146" name="Constant_189554" type="Const" version="opset1">
2518
+ <data element_type="i64" shape="1" offset="4769360" size="8" />
2519
+ <output>
2520
+ <port id="0" precision="I64">
2521
+ <dim>1</dim>
2522
+ </port>
2523
+ </output>
2524
+ </layer>
2525
+ <layer id="147" name="ShapeOf_204345" type="ShapeOf" version="opset3">
2526
+ <data output_type="i64" />
2527
+ <input>
2528
+ <port id="0" precision="FP32">
2529
+ <dim>-1</dim>
2530
+ <dim>-1</dim>
2531
+ <dim>-1</dim>
2532
+ <dim>-1</dim>
2533
+ </port>
2534
+ </input>
2535
+ <output>
2536
+ <port id="1" precision="I64">
2537
+ <dim>4</dim>
2538
+ </port>
2539
+ </output>
2540
+ </layer>
2541
+ <layer id="148" name="Constant_204346" type="Const" version="opset1">
2542
+ <data element_type="i64" shape="" offset="4769336" size="8" />
2543
+ <output>
2544
+ <port id="0" precision="I64" />
2545
+ </output>
2546
+ </layer>
2547
+ <layer id="149" name="Constant_204347" type="Const" version="opset1">
2548
+ <data element_type="i64" shape="" offset="4769360" size="8" />
2549
+ <output>
2550
+ <port id="0" precision="I64" />
2551
+ </output>
2552
+ </layer>
2553
+ <layer id="150" name="Gather_204348" type="Gather" version="opset8">
2554
+ <data batch_dims="0" />
2555
+ <input>
2556
+ <port id="0" precision="I64">
2557
+ <dim>4</dim>
2558
+ </port>
2559
+ <port id="1" precision="I64" />
2560
+ <port id="2" precision="I64" />
2561
+ </input>
2562
+ <output>
2563
+ <port id="3" precision="I64" names="35" />
2564
+ </output>
2565
+ </layer>
2566
+ <layer id="151" name="Constant_187961" type="Const" version="opset1">
2567
+ <data element_type="i64" shape="" offset="4769456" size="8" />
2568
+ <output>
2569
+ <port id="0" precision="I64" names="19" />
2570
+ </output>
2571
+ </layer>
2572
+ <layer id="152" name="__module.decoder/aten::mul/Multiply" type="Multiply" version="opset1">
2573
+ <data auto_broadcast="numpy" />
2574
+ <input>
2575
+ <port id="0" precision="I64" />
2576
+ <port id="1" precision="I64" />
2577
+ </input>
2578
+ <output>
2579
+ <port id="2" precision="I64" names="37,target_frames.1" />
2580
+ </output>
2581
+ </layer>
2582
+ <layer id="153" name="Constant_203552" type="Const" version="opset1">
2583
+ <data element_type="i64" shape="" offset="4769464" size="8" />
2584
+ <output>
2585
+ <port id="0" precision="I64" />
2586
+ </output>
2587
+ </layer>
2588
+ <layer id="154" name="__module.decoder/aten::sub/Subtract" type="Add" version="opset1">
2589
+ <data auto_broadcast="numpy" />
2590
+ <input>
2591
+ <port id="0" precision="I64" />
2592
+ <port id="1" precision="I64" />
2593
+ </input>
2594
+ <output>
2595
+ <port id="2" precision="I64" names="38,39,target_frames" />
2596
+ </output>
2597
+ </layer>
2598
+ <layer id="155" name="Constant_204350" type="Const" version="opset1">
2599
+ <data element_type="i64" shape="" offset="4769336" size="8" />
2600
+ <output>
2601
+ <port id="0" precision="I64" />
2602
+ </output>
2603
+ </layer>
2604
+ <layer id="156" name="Constant_204351" type="Const" version="opset1">
2605
+ <data element_type="i64" shape="" offset="4769360" size="8" />
2606
+ <output>
2607
+ <port id="0" precision="I64" />
2608
+ </output>
2609
+ </layer>
2610
+ <layer id="157" name="Gather_204352" type="Gather" version="opset8">
2611
+ <data batch_dims="0" />
2612
+ <input>
2613
+ <port id="0" precision="I64">
2614
+ <dim>4</dim>
2615
+ </port>
2616
+ <port id="1" precision="I64" />
2617
+ <port id="2" precision="I64" />
2618
+ </input>
2619
+ <output>
2620
+ <port id="3" precision="I64" names="225" />
2621
+ </output>
2622
+ </layer>
2623
+ <layer id="158" name="__module.decoder/aten::sub/Subtract_1" type="Subtract" version="opset1">
2624
+ <data auto_broadcast="numpy" />
2625
+ <input>
2626
+ <port id="0" precision="I64" />
2627
+ <port id="1" precision="I64" />
2628
+ </input>
2629
+ <output>
2630
+ <port id="2" precision="I64" names="227,228,time_padding_needed" />
2631
+ </output>
2632
+ </layer>
2633
+ <layer id="159" name="Unsqueeze_189552" type="Unsqueeze" version="opset1">
2634
+ <input>
2635
+ <port id="0" precision="I64" />
2636
+ <port id="1" precision="I32" />
2637
+ </input>
2638
+ <output>
2639
+ <port id="2" precision="I64">
2640
+ <dim>1</dim>
2641
+ </port>
2642
+ </output>
2643
+ </layer>
2644
+ <layer id="160" name="__module.decoder/prim::ListConstruct/SequenceMark" type="Concat" version="opset1">
2645
+ <data axis="0" />
2646
+ <input>
2647
+ <port id="0" precision="I64">
2648
+ <dim>1</dim>
2649
+ </port>
2650
+ <port id="1" precision="I64">
2651
+ <dim>1</dim>
2652
+ </port>
2653
+ <port id="2" precision="I64">
2654
+ <dim>1</dim>
2655
+ </port>
2656
+ <port id="3" precision="I64">
2657
+ <dim>1</dim>
2658
+ </port>
2659
+ </input>
2660
+ <output>
2661
+ <port id="4" precision="I64">
2662
+ <dim>4</dim>
2663
+ </port>
2664
+ </output>
2665
+ </layer>
2666
+ <layer id="161" name="__module.decoder/aten::pad/Constant_1" type="Const" version="opset1">
2667
+ <data element_type="i32" shape="2" offset="4769472" size="8" />
2668
+ <output>
2669
+ <port id="0" precision="I32">
2670
+ <dim>2</dim>
2671
+ </port>
2672
+ </output>
2673
+ </layer>
2674
+ <layer id="162" name="__module.decoder/aten::pad/Reshape" type="Reshape" version="opset1">
2675
+ <data special_zero="false" />
2676
+ <input>
2677
+ <port id="0" precision="I64">
2678
+ <dim>4</dim>
2679
+ </port>
2680
+ <port id="1" precision="I32">
2681
+ <dim>2</dim>
2682
+ </port>
2683
+ </input>
2684
+ <output>
2685
+ <port id="2" precision="I64">
2686
+ <dim>2</dim>
2687
+ <dim>2</dim>
2688
+ </port>
2689
+ </output>
2690
+ </layer>
2691
+ <layer id="163" name="__module.decoder/aten::pad/Constant_2" type="Const" version="opset1">
2692
+ <data element_type="i32" shape="" offset="4769480" size="4" />
2693
+ <output>
2694
+ <port id="0" precision="I32" />
2695
+ </output>
2696
+ </layer>
2697
+ <layer id="164" name="__module.decoder/aten::pad/Split" type="Split" version="opset1">
2698
+ <data num_splits="2" />
2699
+ <input>
2700
+ <port id="0" precision="I64">
2701
+ <dim>2</dim>
2702
+ <dim>2</dim>
2703
+ </port>
2704
+ <port id="1" precision="I32" />
2705
+ </input>
2706
+ <output>
2707
+ <port id="2" precision="I64">
2708
+ <dim>2</dim>
2709
+ <dim>1</dim>
2710
+ </port>
2711
+ <port id="3" precision="I64">
2712
+ <dim>2</dim>
2713
+ <dim>1</dim>
2714
+ </port>
2715
+ </output>
2716
+ </layer>
2717
+ <layer id="165" name="__module.decoder/aten::pad/Squeeze" type="Squeeze" version="opset1">
2718
+ <input>
2719
+ <port id="0" precision="I64">
2720
+ <dim>2</dim>
2721
+ <dim>1</dim>
2722
+ </port>
2723
+ <port id="1" precision="I32" />
2724
+ </input>
2725
+ <output>
2726
+ <port id="2" precision="I64">
2727
+ <dim>2</dim>
2728
+ </port>
2729
+ </output>
2730
+ </layer>
2731
+ <layer id="166" name="__module.decoder/aten::pad/Range" type="Const" version="opset1">
2732
+ <data element_type="i32" shape="2" offset="37384" size="8" />
2733
+ <output>
2734
+ <port id="0" precision="I32">
2735
+ <dim>2</dim>
2736
+ </port>
2737
+ </output>
2738
+ </layer>
2739
+ <layer id="167" name="__module.decoder/aten::pad/Constant" type="Const" version="opset1">
2740
+ <data element_type="i32" shape="" offset="4769448" size="4" />
2741
+ <output>
2742
+ <port id="0" precision="I32" />
2743
+ </output>
2744
+ </layer>
2745
+ <layer id="168" name="__module.decoder/aten::pad/Gather" type="Gather" version="opset8">
2746
+ <data batch_dims="0" />
2747
+ <input>
2748
+ <port id="0" precision="I64">
2749
+ <dim>2</dim>
2750
+ </port>
2751
+ <port id="1" precision="I32">
2752
+ <dim>2</dim>
2753
+ </port>
2754
+ <port id="2" precision="I32" />
2755
+ </input>
2756
+ <output>
2757
+ <port id="3" precision="I64">
2758
+ <dim>2</dim>
2759
+ </port>
2760
+ </output>
2761
+ </layer>
2762
+ <layer id="169" name="__module.decoder/aten::pad/Concat" type="Concat" version="opset1">
2763
+ <data axis="0" />
2764
+ <input>
2765
+ <port id="0" precision="I64">
2766
+ <dim>2</dim>
2767
+ </port>
2768
+ <port id="1" precision="I64">
2769
+ <dim>2</dim>
2770
+ </port>
2771
+ </input>
2772
+ <output>
2773
+ <port id="2" precision="I64">
2774
+ <dim>4</dim>
2775
+ </port>
2776
+ </output>
2777
+ </layer>
2778
+ <layer id="170" name="__module.decoder/aten::pad/Squeeze_1" type="Squeeze" version="opset1">
2779
+ <input>
2780
+ <port id="0" precision="I64">
2781
+ <dim>2</dim>
2782
+ <dim>1</dim>
2783
+ </port>
2784
+ <port id="1" precision="I32" />
2785
+ </input>
2786
+ <output>
2787
+ <port id="2" precision="I64">
2788
+ <dim>2</dim>
2789
+ </port>
2790
+ </output>
2791
+ </layer>
2792
+ <layer id="171" name="__module.decoder/aten::pad/Gather_1" type="Gather" version="opset8">
2793
+ <data batch_dims="0" />
2794
+ <input>
2795
+ <port id="0" precision="I64">
2796
+ <dim>2</dim>
2797
+ </port>
2798
+ <port id="1" precision="I32">
2799
+ <dim>2</dim>
2800
+ </port>
2801
+ <port id="2" precision="I32" />
2802
+ </input>
2803
+ <output>
2804
+ <port id="3" precision="I64">
2805
+ <dim>2</dim>
2806
+ </port>
2807
+ </output>
2808
+ </layer>
2809
+ <layer id="172" name="__module.decoder/aten::pad/Concat_1" type="Concat" version="opset1">
2810
+ <data axis="0" />
2811
+ <input>
2812
+ <port id="0" precision="I64">
2813
+ <dim>2</dim>
2814
+ </port>
2815
+ <port id="1" precision="I64">
2816
+ <dim>2</dim>
2817
+ </port>
2818
+ </input>
2819
+ <output>
2820
+ <port id="2" precision="I64">
2821
+ <dim>4</dim>
2822
+ </port>
2823
+ </output>
2824
+ </layer>
2825
+ <layer id="173" name="__module.decoder/aten::pad/ConvertLike_1" type="Const" version="opset1">
2826
+ <data element_type="f32" shape="" offset="4769448" size="4" />
2827
+ <output>
2828
+ <port id="0" precision="FP32" />
2829
+ </output>
2830
+ </layer>
2831
+ <layer id="174" name="__module.decoder/aten::pad/Pad" type="Pad" version="opset12">
2832
+ <data pad_mode="constant" />
2833
+ <input>
2834
+ <port id="0" precision="FP32">
2835
+ <dim>-1</dim>
2836
+ <dim>2</dim>
2837
+ <dim>-1</dim>
2838
+ <dim>-1</dim>
2839
+ </port>
2840
+ <port id="1" precision="I64">
2841
+ <dim>4</dim>
2842
+ </port>
2843
+ <port id="2" precision="I64">
2844
+ <dim>4</dim>
2845
+ </port>
2846
+ <port id="3" precision="FP32" />
2847
+ </input>
2848
+ <output>
2849
+ <port id="4" precision="FP32" names="234,235,decoded_output">
2850
+ <dim>-1</dim>
2851
+ <dim>2</dim>
2852
+ <dim>-1</dim>
2853
+ <dim>-1</dim>
2854
+ </port>
2855
+ </output>
2856
+ </layer>
2857
+ <layer id="175" name="Constant_200992" type="Const" version="opset1">
2858
+ <data element_type="i64" shape="3" offset="4769312" size="24" />
2859
+ <output>
2860
+ <port id="0" precision="I64">
2861
+ <dim>3</dim>
2862
+ </port>
2863
+ </output>
2864
+ </layer>
2865
+ <layer id="176" name="Constant_189676" type="Const" version="opset1">
2866
+ <data element_type="i64" shape="1" offset="4769336" size="8" />
2867
+ <output>
2868
+ <port id="0" precision="I64">
2869
+ <dim>1</dim>
2870
+ </port>
2871
+ </output>
2872
+ </layer>
2873
+ <layer id="177" name="Constant_400243" type="Const" version="opset1">
2874
+ <data element_type="i64" shape="1" offset="4769488" size="8" />
2875
+ <output>
2876
+ <port id="0" precision="I64">
2877
+ <dim>1</dim>
2878
+ </port>
2879
+ </output>
2880
+ </layer>
2881
+ <layer id="178" name="__module.decoder/aten::slice/Reshape_2" type="Reshape" version="opset1">
2882
+ <data special_zero="false" />
2883
+ <input>
2884
+ <port id="0" precision="I64" />
2885
+ <port id="1" precision="I64">
2886
+ <dim>1</dim>
2887
+ </port>
2888
+ </input>
2889
+ <output>
2890
+ <port id="2" precision="I64">
2891
+ <dim>1</dim>
2892
+ </port>
2893
+ </output>
2894
+ </layer>
2895
+ <layer id="179" name="Constant_189819" type="Const" version="opset1">
2896
+ <data element_type="i64" shape="1" offset="4769432" size="8" />
2897
+ <output>
2898
+ <port id="0" precision="I64">
2899
+ <dim>1</dim>
2900
+ </port>
2901
+ </output>
2902
+ </layer>
2903
+ <layer id="180" name="Concat_200987" type="Concat" version="opset1">
2904
+ <data axis="0" />
2905
+ <input>
2906
+ <port id="0" precision="I64">
2907
+ <dim>1</dim>
2908
+ </port>
2909
+ <port id="1" precision="I64">
2910
+ <dim>1</dim>
2911
+ </port>
2912
+ <port id="2" precision="I64">
2913
+ <dim>1</dim>
2914
+ </port>
2915
+ </input>
2916
+ <output>
2917
+ <port id="3" precision="I64">
2918
+ <dim>3</dim>
2919
+ </port>
2920
+ </output>
2921
+ </layer>
2922
+ <layer id="181" name="Constant_200991" type="Const" version="opset1">
2923
+ <data element_type="i64" shape="3" offset="4769368" size="24" />
2924
+ <output>
2925
+ <port id="0" precision="I64">
2926
+ <dim>3</dim>
2927
+ </port>
2928
+ </output>
2929
+ </layer>
2930
+ <layer id="182" name="Constant_200990" type="Const" version="opset1">
2931
+ <data element_type="i64" shape="3" offset="4769392" size="24" />
2932
+ <output>
2933
+ <port id="0" precision="I64">
2934
+ <dim>3</dim>
2935
+ </port>
2936
+ </output>
2937
+ </layer>
2938
+ <layer id="183" name="__module.decoder/aten::slice/Slice_7" type="Slice" version="opset8">
2939
+ <input>
2940
+ <port id="0" precision="FP32">
2941
+ <dim>-1</dim>
2942
+ <dim>2</dim>
2943
+ <dim>-1</dim>
2944
+ <dim>-1</dim>
2945
+ </port>
2946
+ <port id="1" precision="I64">
2947
+ <dim>3</dim>
2948
+ </port>
2949
+ <port id="2" precision="I64">
2950
+ <dim>3</dim>
2951
+ </port>
2952
+ <port id="3" precision="I64">
2953
+ <dim>3</dim>
2954
+ </port>
2955
+ <port id="4" precision="I64">
2956
+ <dim>3</dim>
2957
+ </port>
2958
+ </input>
2959
+ <output>
2960
+ <port id="5" precision="FP32" names="sample">
2961
+ <dim>-1</dim>
2962
+ <dim>2</dim>
2963
+ <dim>-1</dim>
2964
+ <dim>-1</dim>
2965
+ </port>
2966
+ </output>
2967
+ </layer>
2968
+ <layer id="184" name="Result_189895" type="Result" version="opset1" output_names="sample">
2969
+ <input>
2970
+ <port id="0" precision="FP32">
2971
+ <dim>-1</dim>
2972
+ <dim>2</dim>
2973
+ <dim>-1</dim>
2974
+ <dim>-1</dim>
2975
+ </port>
2976
+ </input>
2977
+ </layer>
2978
+ </layers>
2979
+ <edges>
2980
+ <edge from-layer="0" from-port="0" to-layer="147" to-port="0" />
2981
+ <edge from-layer="0" from-port="0" to-layer="2" to-port="0" />
2982
+ <edge from-layer="1" from-port="0" to-layer="2" to-port="1" />
2983
+ <edge from-layer="2" from-port="2" to-layer="4" to-port="0" />
2984
+ <edge from-layer="3" from-port="0" to-layer="4" to-port="1" />
2985
+ <edge from-layer="4" from-port="2" to-layer="31" to-port="0" />
2986
+ <edge from-layer="4" from-port="2" to-layer="12" to-port="0" />
2987
+ <edge from-layer="4" from-port="2" to-layer="6" to-port="0" />
2988
+ <edge from-layer="5" from-port="0" to-layer="6" to-port="1" />
2989
+ <edge from-layer="6" from-port="2" to-layer="8" to-port="0" />
2990
+ <edge from-layer="7" from-port="0" to-layer="8" to-port="1" />
2991
+ <edge from-layer="8" from-port="2" to-layer="10" to-port="0" />
2992
+ <edge from-layer="9" from-port="0" to-layer="10" to-port="1" />
2993
+ <edge from-layer="10" from-port="2" to-layer="11" to-port="0" />
2994
+ <edge from-layer="11" from-port="1" to-layer="12" to-port="1" />
2995
+ <edge from-layer="12" from-port="2" to-layer="13" to-port="0" />
2996
+ <edge from-layer="13" from-port="1" to-layer="15" to-port="0" />
2997
+ <edge from-layer="14" from-port="0" to-layer="15" to-port="1" />
2998
+ <edge from-layer="15" from-port="2" to-layer="17" to-port="0" />
2999
+ <edge from-layer="16" from-port="0" to-layer="17" to-port="1" />
3000
+ <edge from-layer="17" from-port="2" to-layer="19" to-port="0" />
3001
+ <edge from-layer="17" from-port="2" to-layer="25" to-port="0" />
3002
+ <edge from-layer="18" from-port="0" to-layer="19" to-port="1" />
3003
+ <edge from-layer="19" from-port="2" to-layer="21" to-port="0" />
3004
+ <edge from-layer="20" from-port="0" to-layer="21" to-port="1" />
3005
+ <edge from-layer="21" from-port="2" to-layer="23" to-port="0" />
3006
+ <edge from-layer="22" from-port="0" to-layer="23" to-port="1" />
3007
+ <edge from-layer="23" from-port="2" to-layer="24" to-port="0" />
3008
+ <edge from-layer="24" from-port="1" to-layer="25" to-port="1" />
3009
+ <edge from-layer="25" from-port="2" to-layer="26" to-port="0" />
3010
+ <edge from-layer="26" from-port="1" to-layer="28" to-port="0" />
3011
+ <edge from-layer="27" from-port="0" to-layer="28" to-port="1" />
3012
+ <edge from-layer="28" from-port="2" to-layer="30" to-port="0" />
3013
+ <edge from-layer="29" from-port="0" to-layer="30" to-port="1" />
3014
+ <edge from-layer="30" from-port="2" to-layer="31" to-port="1" />
3015
+ <edge from-layer="31" from-port="2" to-layer="58" to-port="0" />
3016
+ <edge from-layer="31" from-port="2" to-layer="39" to-port="0" />
3017
+ <edge from-layer="31" from-port="2" to-layer="33" to-port="0" />
3018
+ <edge from-layer="32" from-port="0" to-layer="33" to-port="1" />
3019
+ <edge from-layer="33" from-port="2" to-layer="35" to-port="0" />
3020
+ <edge from-layer="34" from-port="0" to-layer="35" to-port="1" />
3021
+ <edge from-layer="35" from-port="2" to-layer="37" to-port="0" />
3022
+ <edge from-layer="36" from-port="0" to-layer="37" to-port="1" />
3023
+ <edge from-layer="37" from-port="2" to-layer="38" to-port="0" />
3024
+ <edge from-layer="38" from-port="1" to-layer="39" to-port="1" />
3025
+ <edge from-layer="39" from-port="2" to-layer="40" to-port="0" />
3026
+ <edge from-layer="40" from-port="1" to-layer="42" to-port="0" />
3027
+ <edge from-layer="41" from-port="0" to-layer="42" to-port="1" />
3028
+ <edge from-layer="42" from-port="2" to-layer="44" to-port="0" />
3029
+ <edge from-layer="43" from-port="0" to-layer="44" to-port="1" />
3030
+ <edge from-layer="44" from-port="2" to-layer="46" to-port="0" />
3031
+ <edge from-layer="44" from-port="2" to-layer="52" to-port="0" />
3032
+ <edge from-layer="45" from-port="0" to-layer="46" to-port="1" />
3033
+ <edge from-layer="46" from-port="2" to-layer="48" to-port="0" />
3034
+ <edge from-layer="47" from-port="0" to-layer="48" to-port="1" />
3035
+ <edge from-layer="48" from-port="2" to-layer="50" to-port="0" />
3036
+ <edge from-layer="49" from-port="0" to-layer="50" to-port="1" />
3037
+ <edge from-layer="50" from-port="2" to-layer="51" to-port="0" />
3038
+ <edge from-layer="51" from-port="1" to-layer="52" to-port="1" />
3039
+ <edge from-layer="52" from-port="2" to-layer="53" to-port="0" />
3040
+ <edge from-layer="53" from-port="1" to-layer="55" to-port="0" />
3041
+ <edge from-layer="54" from-port="0" to-layer="55" to-port="1" />
3042
+ <edge from-layer="55" from-port="2" to-layer="57" to-port="0" />
3043
+ <edge from-layer="56" from-port="0" to-layer="57" to-port="1" />
3044
+ <edge from-layer="57" from-port="2" to-layer="58" to-port="1" />
3045
+ <edge from-layer="58" from-port="2" to-layer="66" to-port="0" />
3046
+ <edge from-layer="58" from-port="2" to-layer="85" to-port="0" />
3047
+ <edge from-layer="58" from-port="2" to-layer="60" to-port="0" />
3048
+ <edge from-layer="59" from-port="0" to-layer="60" to-port="1" />
3049
+ <edge from-layer="60" from-port="2" to-layer="62" to-port="0" />
3050
+ <edge from-layer="61" from-port="0" to-layer="62" to-port="1" />
3051
+ <edge from-layer="62" from-port="2" to-layer="64" to-port="0" />
3052
+ <edge from-layer="63" from-port="0" to-layer="64" to-port="1" />
3053
+ <edge from-layer="64" from-port="2" to-layer="65" to-port="0" />
3054
+ <edge from-layer="65" from-port="1" to-layer="66" to-port="1" />
3055
+ <edge from-layer="66" from-port="2" to-layer="67" to-port="0" />
3056
+ <edge from-layer="67" from-port="1" to-layer="69" to-port="0" />
3057
+ <edge from-layer="68" from-port="0" to-layer="69" to-port="1" />
3058
+ <edge from-layer="69" from-port="2" to-layer="71" to-port="0" />
3059
+ <edge from-layer="70" from-port="0" to-layer="71" to-port="1" />
3060
+ <edge from-layer="71" from-port="2" to-layer="73" to-port="0" />
3061
+ <edge from-layer="71" from-port="2" to-layer="79" to-port="0" />
3062
+ <edge from-layer="72" from-port="0" to-layer="73" to-port="1" />
3063
+ <edge from-layer="73" from-port="2" to-layer="75" to-port="0" />
3064
+ <edge from-layer="74" from-port="0" to-layer="75" to-port="1" />
3065
+ <edge from-layer="75" from-port="2" to-layer="77" to-port="0" />
3066
+ <edge from-layer="76" from-port="0" to-layer="77" to-port="1" />
3067
+ <edge from-layer="77" from-port="2" to-layer="78" to-port="0" />
3068
+ <edge from-layer="78" from-port="1" to-layer="79" to-port="1" />
3069
+ <edge from-layer="79" from-port="2" to-layer="80" to-port="0" />
3070
+ <edge from-layer="80" from-port="1" to-layer="82" to-port="0" />
3071
+ <edge from-layer="81" from-port="0" to-layer="82" to-port="1" />
3072
+ <edge from-layer="82" from-port="2" to-layer="84" to-port="0" />
3073
+ <edge from-layer="83" from-port="0" to-layer="84" to-port="1" />
3074
+ <edge from-layer="84" from-port="2" to-layer="85" to-port="1" />
3075
+ <edge from-layer="85" from-port="2" to-layer="112" to-port="0" />
3076
+ <edge from-layer="85" from-port="2" to-layer="93" to-port="0" />
3077
+ <edge from-layer="85" from-port="2" to-layer="87" to-port="0" />
3078
+ <edge from-layer="86" from-port="0" to-layer="87" to-port="1" />
3079
+ <edge from-layer="87" from-port="2" to-layer="89" to-port="0" />
3080
+ <edge from-layer="88" from-port="0" to-layer="89" to-port="1" />
3081
+ <edge from-layer="89" from-port="2" to-layer="91" to-port="0" />
3082
+ <edge from-layer="90" from-port="0" to-layer="91" to-port="1" />
3083
+ <edge from-layer="91" from-port="2" to-layer="92" to-port="0" />
3084
+ <edge from-layer="92" from-port="1" to-layer="93" to-port="1" />
3085
+ <edge from-layer="93" from-port="2" to-layer="94" to-port="0" />
3086
+ <edge from-layer="94" from-port="1" to-layer="96" to-port="0" />
3087
+ <edge from-layer="95" from-port="0" to-layer="96" to-port="1" />
3088
+ <edge from-layer="96" from-port="2" to-layer="98" to-port="0" />
3089
+ <edge from-layer="97" from-port="0" to-layer="98" to-port="1" />
3090
+ <edge from-layer="98" from-port="2" to-layer="100" to-port="0" />
3091
+ <edge from-layer="98" from-port="2" to-layer="106" to-port="0" />
3092
+ <edge from-layer="99" from-port="0" to-layer="100" to-port="1" />
3093
+ <edge from-layer="100" from-port="2" to-layer="102" to-port="0" />
3094
+ <edge from-layer="101" from-port="0" to-layer="102" to-port="1" />
3095
+ <edge from-layer="102" from-port="2" to-layer="104" to-port="0" />
3096
+ <edge from-layer="103" from-port="0" to-layer="104" to-port="1" />
3097
+ <edge from-layer="104" from-port="2" to-layer="105" to-port="0" />
3098
+ <edge from-layer="105" from-port="1" to-layer="106" to-port="1" />
3099
+ <edge from-layer="106" from-port="2" to-layer="107" to-port="0" />
3100
+ <edge from-layer="107" from-port="1" to-layer="109" to-port="0" />
3101
+ <edge from-layer="108" from-port="0" to-layer="109" to-port="1" />
3102
+ <edge from-layer="109" from-port="2" to-layer="111" to-port="0" />
3103
+ <edge from-layer="110" from-port="0" to-layer="111" to-port="1" />
3104
+ <edge from-layer="111" from-port="2" to-layer="112" to-port="1" />
3105
+ <edge from-layer="112" from-port="2" to-layer="114" to-port="0" />
3106
+ <edge from-layer="112" from-port="2" to-layer="120" to-port="0" />
3107
+ <edge from-layer="113" from-port="0" to-layer="114" to-port="1" />
3108
+ <edge from-layer="114" from-port="2" to-layer="116" to-port="0" />
3109
+ <edge from-layer="115" from-port="0" to-layer="116" to-port="1" />
3110
+ <edge from-layer="116" from-port="2" to-layer="118" to-port="0" />
3111
+ <edge from-layer="117" from-port="0" to-layer="118" to-port="1" />
3112
+ <edge from-layer="118" from-port="2" to-layer="119" to-port="0" />
3113
+ <edge from-layer="119" from-port="1" to-layer="120" to-port="1" />
3114
+ <edge from-layer="120" from-port="2" to-layer="121" to-port="0" />
3115
+ <edge from-layer="121" from-port="1" to-layer="123" to-port="0" />
3116
+ <edge from-layer="122" from-port="0" to-layer="123" to-port="1" />
3117
+ <edge from-layer="123" from-port="2" to-layer="125" to-port="0" />
3118
+ <edge from-layer="124" from-port="0" to-layer="125" to-port="1" />
3119
+ <edge from-layer="125" from-port="2" to-layer="128" to-port="0" />
3120
+ <edge from-layer="125" from-port="2" to-layer="135" to-port="0" />
3121
+ <edge from-layer="126" from-port="0" to-layer="135" to-port="1" />
3122
+ <edge from-layer="127" from-port="0" to-layer="132" to-port="0" />
3123
+ <edge from-layer="128" from-port="1" to-layer="131" to-port="0" />
3124
+ <edge from-layer="129" from-port="0" to-layer="131" to-port="1" />
3125
+ <edge from-layer="130" from-port="0" to-layer="131" to-port="2" />
3126
+ <edge from-layer="131" from-port="3" to-layer="132" to-port="1" />
3127
+ <edge from-layer="132" from-port="2" to-layer="135" to-port="2" />
3128
+ <edge from-layer="133" from-port="0" to-layer="135" to-port="3" />
3129
+ <edge from-layer="134" from-port="0" to-layer="135" to-port="4" />
3130
+ <edge from-layer="135" from-port="5" to-layer="139" to-port="0" />
3131
+ <edge from-layer="135" from-port="5" to-layer="174" to-port="0" />
3132
+ <edge from-layer="136" from-port="0" to-layer="169" to-port="0" />
3133
+ <edge from-layer="136" from-port="0" to-layer="172" to-port="0" />
3134
+ <edge from-layer="137" from-port="0" to-layer="160" to-port="0" />
3135
+ <edge from-layer="138" from-port="0" to-layer="143" to-port="0" />
3136
+ <edge from-layer="139" from-port="1" to-layer="142" to-port="0" />
3137
+ <edge from-layer="139" from-port="1" to-layer="157" to-port="0" />
3138
+ <edge from-layer="140" from-port="0" to-layer="142" to-port="1" />
3139
+ <edge from-layer="141" from-port="0" to-layer="142" to-port="2" />
3140
+ <edge from-layer="142" from-port="3" to-layer="143" to-port="1" />
3141
+ <edge from-layer="143" from-port="2" to-layer="145" to-port="0" />
3142
+ <edge from-layer="144" from-port="0" to-layer="145" to-port="1" />
3143
+ <edge from-layer="144" from-port="0" to-layer="159" to-port="1" />
3144
+ <edge from-layer="145" from-port="2" to-layer="160" to-port="1" />
3145
+ <edge from-layer="146" from-port="0" to-layer="160" to-port="2" />
3146
+ <edge from-layer="147" from-port="1" to-layer="150" to-port="0" />
3147
+ <edge from-layer="148" from-port="0" to-layer="150" to-port="1" />
3148
+ <edge from-layer="149" from-port="0" to-layer="150" to-port="2" />
3149
+ <edge from-layer="150" from-port="3" to-layer="152" to-port="0" />
3150
+ <edge from-layer="151" from-port="0" to-layer="152" to-port="1" />
3151
+ <edge from-layer="152" from-port="2" to-layer="154" to-port="0" />
3152
+ <edge from-layer="153" from-port="0" to-layer="154" to-port="1" />
3153
+ <edge from-layer="154" from-port="2" to-layer="158" to-port="0" />
3154
+ <edge from-layer="154" from-port="2" to-layer="178" to-port="0" />
3155
+ <edge from-layer="155" from-port="0" to-layer="157" to-port="1" />
3156
+ <edge from-layer="156" from-port="0" to-layer="157" to-port="2" />
3157
+ <edge from-layer="157" from-port="3" to-layer="158" to-port="1" />
3158
+ <edge from-layer="158" from-port="2" to-layer="159" to-port="0" />
3159
+ <edge from-layer="159" from-port="2" to-layer="160" to-port="3" />
3160
+ <edge from-layer="160" from-port="4" to-layer="162" to-port="0" />
3161
+ <edge from-layer="161" from-port="0" to-layer="162" to-port="1" />
3162
+ <edge from-layer="162" from-port="2" to-layer="164" to-port="0" />
3163
+ <edge from-layer="163" from-port="0" to-layer="170" to-port="1" />
3164
+ <edge from-layer="163" from-port="0" to-layer="165" to-port="1" />
3165
+ <edge from-layer="163" from-port="0" to-layer="164" to-port="1" />
3166
+ <edge from-layer="164" from-port="3" to-layer="170" to-port="0" />
3167
+ <edge from-layer="164" from-port="2" to-layer="165" to-port="0" />
3168
+ <edge from-layer="165" from-port="2" to-layer="168" to-port="0" />
3169
+ <edge from-layer="166" from-port="0" to-layer="168" to-port="1" />
3170
+ <edge from-layer="166" from-port="0" to-layer="171" to-port="1" />
3171
+ <edge from-layer="167" from-port="0" to-layer="171" to-port="2" />
3172
+ <edge from-layer="167" from-port="0" to-layer="168" to-port="2" />
3173
+ <edge from-layer="168" from-port="3" to-layer="169" to-port="1" />
3174
+ <edge from-layer="169" from-port="2" to-layer="174" to-port="1" />
3175
+ <edge from-layer="170" from-port="2" to-layer="171" to-port="0" />
3176
+ <edge from-layer="171" from-port="3" to-layer="172" to-port="1" />
3177
+ <edge from-layer="172" from-port="2" to-layer="174" to-port="2" />
3178
+ <edge from-layer="173" from-port="0" to-layer="174" to-port="3" />
3179
+ <edge from-layer="174" from-port="4" to-layer="183" to-port="0" />
3180
+ <edge from-layer="175" from-port="0" to-layer="183" to-port="1" />
3181
+ <edge from-layer="176" from-port="0" to-layer="180" to-port="0" />
3182
+ <edge from-layer="177" from-port="0" to-layer="178" to-port="1" />
3183
+ <edge from-layer="178" from-port="2" to-layer="180" to-port="1" />
3184
+ <edge from-layer="179" from-port="0" to-layer="180" to-port="2" />
3185
+ <edge from-layer="180" from-port="3" to-layer="183" to-port="2" />
3186
+ <edge from-layer="181" from-port="0" to-layer="183" to-port="3" />
3187
+ <edge from-layer="182" from-port="0" to-layer="183" to-port="4" />
3188
+ <edge from-layer="183" from-port="5" to-layer="184" to-port="0" />
3189
+ </edges>
3190
+ <rt_info>
3191
+ <info name="OpenVINO Runtime" value="2026.5.0-23697-32776832f8d" />
3192
+ <Runtime_version value="2026.5.0-23697-32776832f8d" />
3193
+ <conversion_parameters>
3194
+ <framework value="pytorch" />
3195
+ <is_python_object value="True" />
3196
+ </conversion_parameters>
3197
+ <optimum>
3198
+ <diffusers_version value="0.41.0" />
3199
+ <optimum_intel_version value="2.3.0.dev0+8c40189b" />
3200
+ <optimum_version value="2.3.0" />
3201
+ <pytorch_version value="2.15.0" />
3202
+ <transformers_version value="5.15.1" />
3203
+ </optimum>
3204
+ </rt_info>
3205
+ </net>
connectors/config.json DELETED
@@ -1,25 +0,0 @@
1
- {
2
- "_class_name": "LTX2TextConnectors",
3
- "_diffusers_version": "0.40.0",
4
- "audio_connector_attention_head_dim": 8,
5
- "audio_connector_num_attention_heads": 4,
6
- "audio_connector_num_layers": 1,
7
- "audio_connector_num_learnable_registers": 8,
8
- "audio_gated_attn": true,
9
- "audio_hidden_dim": 32,
10
- "caption_channels": 32,
11
- "causal_temporal_positioning": false,
12
- "connector_rope_base_seq_len": 4096,
13
- "per_modality_projections": true,
14
- "proj_bias": true,
15
- "rope_double_precision": true,
16
- "rope_theta": 10000.0,
17
- "rope_type": "split",
18
- "text_proj_in_factor": 3,
19
- "video_connector_attention_head_dim": 8,
20
- "video_connector_num_attention_heads": 4,
21
- "video_connector_num_layers": 1,
22
- "video_connector_num_learnable_registers": 8,
23
- "video_gated_attn": true,
24
- "video_hidden_dim": 32
25
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
transformer/diffusion_pytorch_model.safetensors → connectors/openvino_model.bin RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4a84d5174365db1cb1697413aacb2428231bc0c3c1387972991a63c5fd42e195
3
- size 846144
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:35b2aeacc7a4a7c95a5cd104e3db3781d8dd8c5e77ae2402cd2d7ca6072b0c3e
3
+ size 128920
connectors/openvino_model.xml ADDED
The diff for this file is too large to render. See raw diff
 
duration_head/config.json DELETED
@@ -1,10 +0,0 @@
1
- {
2
- "_class_name": "LTX2DurationHead",
3
- "_diffusers_version": "0.40.0",
4
- "audio_cross_attention_dim": 32,
5
- "mlp_hidden_dim": 16,
6
- "num_pooler_heads": 2,
7
- "num_queries": 1,
8
- "pooler_hidden_dim": 16,
9
- "video_cross_attention_dim": 32
10
- }
 
 
 
 
 
 
 
 
 
 
 
duration_head/{diffusion_pytorch_model.safetensors → openvino_model.bin} RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:87387b8f289b212b2f39da789f1961046a7e60006e63bba5371a81af6c83efcd
3
- size 11532
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca514c5c7fbc74f81f26c60bd1f0460624decf6bbaa91784f11b13788f1fcb51
3
+ size 9960
duration_head/openvino_model.xml ADDED
@@ -0,0 +1,1010 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ <?xml version="1.0"?>
2
+ <net name="Model24" version="11">
3
+ <layers>
4
+ <layer id="0" name="video_tokens" type="Parameter" version="opset1">
5
+ <data shape="?,?,32" element_type="f32" />
6
+ <output>
7
+ <port id="0" precision="FP32" names="video_tokens">
8
+ <dim>-1</dim>
9
+ <dim>-1</dim>
10
+ <dim>32</dim>
11
+ </port>
12
+ </output>
13
+ </layer>
14
+ <layer id="1" name="audio_tokens" type="Parameter" version="opset1">
15
+ <data shape="?,?,32" element_type="f32" />
16
+ <output>
17
+ <port id="0" precision="FP32" names="audio_tokens">
18
+ <dim>-1</dim>
19
+ <dim>-1</dim>
20
+ <dim>32</dim>
21
+ </port>
22
+ </output>
23
+ </layer>
24
+ <layer id="2" name="__module.attention_pooler/aten::unsqueeze/Unsqueeze" type="Const" version="opset1">
25
+ <data element_type="f32" shape="1, 1, 16" offset="0" size="64" />
26
+ <output>
27
+ <port id="0" precision="FP32" names="60">
28
+ <dim>1</dim>
29
+ <dim>1</dim>
30
+ <dim>16</dim>
31
+ </port>
32
+ </output>
33
+ </layer>
34
+ <layer id="3" name="self.video_input_proj.weight" type="Const" version="opset1">
35
+ <data element_type="f32" shape="16, 32" offset="64" size="2048" />
36
+ <output>
37
+ <port id="0" precision="FP32" names="self.video_input_proj.weight">
38
+ <dim>16</dim>
39
+ <dim>32</dim>
40
+ </port>
41
+ </output>
42
+ </layer>
43
+ <layer id="4" name="__module.video_input_proj/aten::linear/MatMul" type="MatMul" version="opset1">
44
+ <data transpose_a="false" transpose_b="true" />
45
+ <input>
46
+ <port id="0" precision="FP32">
47
+ <dim>-1</dim>
48
+ <dim>-1</dim>
49
+ <dim>32</dim>
50
+ </port>
51
+ <port id="1" precision="FP32">
52
+ <dim>16</dim>
53
+ <dim>32</dim>
54
+ </port>
55
+ </input>
56
+ <output>
57
+ <port id="2" precision="FP32">
58
+ <dim>-1</dim>
59
+ <dim>-1</dim>
60
+ <dim>16</dim>
61
+ </port>
62
+ </output>
63
+ </layer>
64
+ <layer id="5" name="Constant_163519" type="Const" version="opset1">
65
+ <data element_type="f32" shape="1, 1, 16" offset="2112" size="64" />
66
+ <output>
67
+ <port id="0" precision="FP32">
68
+ <dim>1</dim>
69
+ <dim>1</dim>
70
+ <dim>16</dim>
71
+ </port>
72
+ </output>
73
+ </layer>
74
+ <layer id="6" name="aten::add/Add" type="Add" version="opset1">
75
+ <data auto_broadcast="numpy" />
76
+ <input>
77
+ <port id="0" precision="FP32">
78
+ <dim>-1</dim>
79
+ <dim>-1</dim>
80
+ <dim>16</dim>
81
+ </port>
82
+ <port id="1" precision="FP32">
83
+ <dim>1</dim>
84
+ <dim>1</dim>
85
+ <dim>16</dim>
86
+ </port>
87
+ </input>
88
+ <output>
89
+ <port id="2" precision="FP32" names="17">
90
+ <dim>-1</dim>
91
+ <dim>-1</dim>
92
+ <dim>16</dim>
93
+ </port>
94
+ </output>
95
+ </layer>
96
+ <layer id="7" name="self.audio_input_proj.weight" type="Const" version="opset1">
97
+ <data element_type="f32" shape="16, 32" offset="2176" size="2048" />
98
+ <output>
99
+ <port id="0" precision="FP32" names="self.audio_input_proj.weight">
100
+ <dim>16</dim>
101
+ <dim>32</dim>
102
+ </port>
103
+ </output>
104
+ </layer>
105
+ <layer id="8" name="__module.audio_input_proj/aten::linear/MatMul" type="MatMul" version="opset1">
106
+ <data transpose_a="false" transpose_b="true" />
107
+ <input>
108
+ <port id="0" precision="FP32">
109
+ <dim>-1</dim>
110
+ <dim>-1</dim>
111
+ <dim>32</dim>
112
+ </port>
113
+ <port id="1" precision="FP32">
114
+ <dim>16</dim>
115
+ <dim>32</dim>
116
+ </port>
117
+ </input>
118
+ <output>
119
+ <port id="2" precision="FP32">
120
+ <dim>-1</dim>
121
+ <dim>-1</dim>
122
+ <dim>16</dim>
123
+ </port>
124
+ </output>
125
+ </layer>
126
+ <layer id="9" name="Constant_163520" type="Const" version="opset1">
127
+ <data element_type="f32" shape="1, 1, 16" offset="4224" size="64" />
128
+ <output>
129
+ <port id="0" precision="FP32">
130
+ <dim>1</dim>
131
+ <dim>1</dim>
132
+ <dim>16</dim>
133
+ </port>
134
+ </output>
135
+ </layer>
136
+ <layer id="10" name="aten::add/Add_1" type="Add" version="opset1">
137
+ <data auto_broadcast="numpy" />
138
+ <input>
139
+ <port id="0" precision="FP32">
140
+ <dim>-1</dim>
141
+ <dim>-1</dim>
142
+ <dim>16</dim>
143
+ </port>
144
+ <port id="1" precision="FP32">
145
+ <dim>1</dim>
146
+ <dim>1</dim>
147
+ <dim>16</dim>
148
+ </port>
149
+ </input>
150
+ <output>
151
+ <port id="2" precision="FP32" names="25">
152
+ <dim>-1</dim>
153
+ <dim>-1</dim>
154
+ <dim>16</dim>
155
+ </port>
156
+ </output>
157
+ </layer>
158
+ <layer id="11" name="aten::cat/Concat" type="Concat" version="opset1">
159
+ <data axis="1" />
160
+ <input>
161
+ <port id="0" precision="FP32">
162
+ <dim>-1</dim>
163
+ <dim>-1</dim>
164
+ <dim>16</dim>
165
+ </port>
166
+ <port id="1" precision="FP32">
167
+ <dim>-1</dim>
168
+ <dim>-1</dim>
169
+ <dim>16</dim>
170
+ </port>
171
+ </input>
172
+ <output>
173
+ <port id="2" precision="FP32" names="28,tokens">
174
+ <dim>-1</dim>
175
+ <dim>-1</dim>
176
+ <dim>16</dim>
177
+ </port>
178
+ </output>
179
+ </layer>
180
+ <layer id="12" name="self.attention_pooler.to_k.weight" type="Const" version="opset1">
181
+ <data element_type="f32" shape="16, 16" offset="4288" size="1024" />
182
+ <output>
183
+ <port id="0" precision="FP32" names="self.attention_pooler.to_k.weight">
184
+ <dim>16</dim>
185
+ <dim>16</dim>
186
+ </port>
187
+ </output>
188
+ </layer>
189
+ <layer id="13" name="__module.attention_pooler.to_k/aten::linear/MatMul" type="MatMul" version="opset1">
190
+ <data transpose_a="false" transpose_b="true" />
191
+ <input>
192
+ <port id="0" precision="FP32">
193
+ <dim>-1</dim>
194
+ <dim>-1</dim>
195
+ <dim>16</dim>
196
+ </port>
197
+ <port id="1" precision="FP32">
198
+ <dim>16</dim>
199
+ <dim>16</dim>
200
+ </port>
201
+ </input>
202
+ <output>
203
+ <port id="2" precision="FP32">
204
+ <dim>-1</dim>
205
+ <dim>-1</dim>
206
+ <dim>16</dim>
207
+ </port>
208
+ </output>
209
+ </layer>
210
+ <layer id="14" name="Constant_163522" type="Const" version="opset1">
211
+ <data element_type="f32" shape="1, 1, 16" offset="5312" size="64" />
212
+ <output>
213
+ <port id="0" precision="FP32">
214
+ <dim>1</dim>
215
+ <dim>1</dim>
216
+ <dim>16</dim>
217
+ </port>
218
+ </output>
219
+ </layer>
220
+ <layer id="15" name="__module.attention_pooler.to_k/aten::linear/Add" type="Add" version="opset1">
221
+ <data auto_broadcast="numpy" />
222
+ <input>
223
+ <port id="0" precision="FP32">
224
+ <dim>-1</dim>
225
+ <dim>-1</dim>
226
+ <dim>16</dim>
227
+ </port>
228
+ <port id="1" precision="FP32">
229
+ <dim>1</dim>
230
+ <dim>1</dim>
231
+ <dim>16</dim>
232
+ </port>
233
+ </input>
234
+ <output>
235
+ <port id="2" precision="FP32" names="71">
236
+ <dim>-1</dim>
237
+ <dim>-1</dim>
238
+ <dim>16</dim>
239
+ </port>
240
+ </output>
241
+ </layer>
242
+ <layer id="16" name="ShapeOf_163581" type="ShapeOf" version="opset3">
243
+ <data output_type="i64" />
244
+ <input>
245
+ <port id="0" precision="FP32">
246
+ <dim>-1</dim>
247
+ <dim>-1</dim>
248
+ <dim>16</dim>
249
+ </port>
250
+ </input>
251
+ <output>
252
+ <port id="1" precision="I64">
253
+ <dim>3</dim>
254
+ </port>
255
+ </output>
256
+ </layer>
257
+ <layer id="17" name="Constant_163582" type="Const" version="opset1">
258
+ <data element_type="i64" shape="1" offset="5376" size="8" />
259
+ <output>
260
+ <port id="0" precision="I64">
261
+ <dim>1</dim>
262
+ </port>
263
+ </output>
264
+ </layer>
265
+ <layer id="18" name="Constant_163583" type="Const" version="opset1">
266
+ <data element_type="i64" shape="" offset="5376" size="8" />
267
+ <output>
268
+ <port id="0" precision="I64" />
269
+ </output>
270
+ </layer>
271
+ <layer id="19" name="Gather_163584" type="Gather" version="opset8">
272
+ <data batch_dims="0" />
273
+ <input>
274
+ <port id="0" precision="I64">
275
+ <dim>3</dim>
276
+ </port>
277
+ <port id="1" precision="I64">
278
+ <dim>1</dim>
279
+ </port>
280
+ <port id="2" precision="I64" />
281
+ </input>
282
+ <output>
283
+ <port id="3" precision="I64" names="61">
284
+ <dim>1</dim>
285
+ </port>
286
+ </output>
287
+ </layer>
288
+ <layer id="20" name="Constant_162948" type="Const" version="opset1">
289
+ <data element_type="i64" shape="1" offset="5384" size="8" />
290
+ <output>
291
+ <port id="0" precision="I64">
292
+ <dim>1</dim>
293
+ </port>
294
+ </output>
295
+ </layer>
296
+ <layer id="21" name="Constant_162950" type="Const" version="opset1">
297
+ <data element_type="i64" shape="1" offset="5384" size="8" />
298
+ <output>
299
+ <port id="0" precision="I64">
300
+ <dim>1</dim>
301
+ </port>
302
+ </output>
303
+ </layer>
304
+ <layer id="22" name="__module.attention_pooler/prim::ListConstruct/SequenceMark" type="Concat" version="opset1">
305
+ <data axis="0" />
306
+ <input>
307
+ <port id="0" precision="I64">
308
+ <dim>1</dim>
309
+ </port>
310
+ <port id="1" precision="I64">
311
+ <dim>1</dim>
312
+ </port>
313
+ <port id="2" precision="I64">
314
+ <dim>1</dim>
315
+ </port>
316
+ </input>
317
+ <output>
318
+ <port id="3" precision="I64">
319
+ <dim>3</dim>
320
+ </port>
321
+ </output>
322
+ </layer>
323
+ <layer id="23" name="__module.attention_pooler/aten::expand/Broadcast" type="Broadcast" version="opset3">
324
+ <data mode="bidirectional" />
325
+ <input>
326
+ <port id="0" precision="FP32">
327
+ <dim>1</dim>
328
+ <dim>1</dim>
329
+ <dim>16</dim>
330
+ </port>
331
+ <port id="1" precision="I64">
332
+ <dim>3</dim>
333
+ </port>
334
+ </input>
335
+ <output>
336
+ <port id="2" precision="FP32" names="63,input.5">
337
+ <dim>-1</dim>
338
+ <dim>1</dim>
339
+ <dim>16</dim>
340
+ </port>
341
+ </output>
342
+ </layer>
343
+ <layer id="24" name="self.attention_pooler.to_q.weight" type="Const" version="opset1">
344
+ <data element_type="f32" shape="16, 16" offset="5392" size="1024" />
345
+ <output>
346
+ <port id="0" precision="FP32" names="self.attention_pooler.to_q.weight">
347
+ <dim>16</dim>
348
+ <dim>16</dim>
349
+ </port>
350
+ </output>
351
+ </layer>
352
+ <layer id="25" name="__module.attention_pooler.to_q/aten::linear/MatMul" type="MatMul" version="opset1">
353
+ <data transpose_a="false" transpose_b="true" />
354
+ <input>
355
+ <port id="0" precision="FP32">
356
+ <dim>-1</dim>
357
+ <dim>1</dim>
358
+ <dim>16</dim>
359
+ </port>
360
+ <port id="1" precision="FP32">
361
+ <dim>16</dim>
362
+ <dim>16</dim>
363
+ </port>
364
+ </input>
365
+ <output>
366
+ <port id="2" precision="FP32">
367
+ <dim>-1</dim>
368
+ <dim>1</dim>
369
+ <dim>16</dim>
370
+ </port>
371
+ </output>
372
+ </layer>
373
+ <layer id="26" name="Constant_163521" type="Const" version="opset1">
374
+ <data element_type="f32" shape="1, 1, 16" offset="6416" size="64" />
375
+ <output>
376
+ <port id="0" precision="FP32">
377
+ <dim>1</dim>
378
+ <dim>1</dim>
379
+ <dim>16</dim>
380
+ </port>
381
+ </output>
382
+ </layer>
383
+ <layer id="27" name="__module.attention_pooler.to_q/aten::linear/Add" type="Add" version="opset1">
384
+ <data auto_broadcast="numpy" />
385
+ <input>
386
+ <port id="0" precision="FP32">
387
+ <dim>-1</dim>
388
+ <dim>1</dim>
389
+ <dim>16</dim>
390
+ </port>
391
+ <port id="1" precision="FP32">
392
+ <dim>1</dim>
393
+ <dim>1</dim>
394
+ <dim>16</dim>
395
+ </port>
396
+ </input>
397
+ <output>
398
+ <port id="2" precision="FP32" names="66">
399
+ <dim>-1</dim>
400
+ <dim>1</dim>
401
+ <dim>16</dim>
402
+ </port>
403
+ </output>
404
+ </layer>
405
+ <layer id="28" name="Constant_163564" type="Const" version="opset1">
406
+ <data element_type="i64" shape="4" offset="6480" size="32" />
407
+ <output>
408
+ <port id="0" precision="I64">
409
+ <dim>4</dim>
410
+ </port>
411
+ </output>
412
+ </layer>
413
+ <layer id="29" name="__module.attention_pooler/aten::permute/Transpose" type="Reshape" version="opset1">
414
+ <data special_zero="false" />
415
+ <input>
416
+ <port id="0" precision="FP32">
417
+ <dim>-1</dim>
418
+ <dim>1</dim>
419
+ <dim>16</dim>
420
+ </port>
421
+ <port id="1" precision="I64">
422
+ <dim>4</dim>
423
+ </port>
424
+ </input>
425
+ <output>
426
+ <port id="2" precision="FP32" names="80,query">
427
+ <dim>-1</dim>
428
+ <dim>2</dim>
429
+ <dim>1</dim>
430
+ <dim>8</dim>
431
+ </port>
432
+ </output>
433
+ </layer>
434
+ <layer id="30" name="Constant_163593" type="Const" version="opset1">
435
+ <data element_type="i64" shape="4" offset="6512" size="32" />
436
+ <output>
437
+ <port id="0" precision="I64">
438
+ <dim>4</dim>
439
+ </port>
440
+ </output>
441
+ </layer>
442
+ <layer id="31" name="__module.attention_pooler/aten::unflatten/Reshape_1" type="Reshape" version="opset1">
443
+ <data special_zero="true" />
444
+ <input>
445
+ <port id="0" precision="FP32">
446
+ <dim>-1</dim>
447
+ <dim>-1</dim>
448
+ <dim>16</dim>
449
+ </port>
450
+ <port id="1" precision="I64">
451
+ <dim>4</dim>
452
+ </port>
453
+ </input>
454
+ <output>
455
+ <port id="2" precision="FP32" names="73,x.3">
456
+ <dim>-1</dim>
457
+ <dim>-1</dim>
458
+ <dim>2</dim>
459
+ <dim>8</dim>
460
+ </port>
461
+ </output>
462
+ </layer>
463
+ <layer id="32" name="Constant_160744" type="Const" version="opset1">
464
+ <data element_type="i64" shape="4" offset="6544" size="32" />
465
+ <output>
466
+ <port id="0" precision="I64">
467
+ <dim>4</dim>
468
+ </port>
469
+ </output>
470
+ </layer>
471
+ <layer id="33" name="__module.attention_pooler/aten::permute/Transpose_1" type="Transpose" version="opset1">
472
+ <input>
473
+ <port id="0" precision="FP32">
474
+ <dim>-1</dim>
475
+ <dim>-1</dim>
476
+ <dim>2</dim>
477
+ <dim>8</dim>
478
+ </port>
479
+ <port id="1" precision="I64">
480
+ <dim>4</dim>
481
+ </port>
482
+ </input>
483
+ <output>
484
+ <port id="2" precision="FP32" names="82,key">
485
+ <dim>-1</dim>
486
+ <dim>2</dim>
487
+ <dim>-1</dim>
488
+ <dim>8</dim>
489
+ </port>
490
+ </output>
491
+ </layer>
492
+ <layer id="34" name="self.attention_pooler.to_v.weight" type="Const" version="opset1">
493
+ <data element_type="f32" shape="16, 16" offset="6576" size="1024" />
494
+ <output>
495
+ <port id="0" precision="FP32" names="self.attention_pooler.to_v.weight">
496
+ <dim>16</dim>
497
+ <dim>16</dim>
498
+ </port>
499
+ </output>
500
+ </layer>
501
+ <layer id="35" name="__module.attention_pooler.to_v/aten::linear/MatMul" type="MatMul" version="opset1">
502
+ <data transpose_a="false" transpose_b="true" />
503
+ <input>
504
+ <port id="0" precision="FP32">
505
+ <dim>-1</dim>
506
+ <dim>-1</dim>
507
+ <dim>16</dim>
508
+ </port>
509
+ <port id="1" precision="FP32">
510
+ <dim>16</dim>
511
+ <dim>16</dim>
512
+ </port>
513
+ </input>
514
+ <output>
515
+ <port id="2" precision="FP32">
516
+ <dim>-1</dim>
517
+ <dim>-1</dim>
518
+ <dim>16</dim>
519
+ </port>
520
+ </output>
521
+ </layer>
522
+ <layer id="36" name="Constant_163523" type="Const" version="opset1">
523
+ <data element_type="f32" shape="1, 1, 16" offset="7600" size="64" />
524
+ <output>
525
+ <port id="0" precision="FP32">
526
+ <dim>1</dim>
527
+ <dim>1</dim>
528
+ <dim>16</dim>
529
+ </port>
530
+ </output>
531
+ </layer>
532
+ <layer id="37" name="__module.attention_pooler.to_v/aten::linear/Add" type="Add" version="opset1">
533
+ <data auto_broadcast="numpy" />
534
+ <input>
535
+ <port id="0" precision="FP32">
536
+ <dim>-1</dim>
537
+ <dim>-1</dim>
538
+ <dim>16</dim>
539
+ </port>
540
+ <port id="1" precision="FP32">
541
+ <dim>1</dim>
542
+ <dim>1</dim>
543
+ <dim>16</dim>
544
+ </port>
545
+ </input>
546
+ <output>
547
+ <port id="2" precision="FP32" names="76">
548
+ <dim>-1</dim>
549
+ <dim>-1</dim>
550
+ <dim>16</dim>
551
+ </port>
552
+ </output>
553
+ </layer>
554
+ <layer id="38" name="Constant_163594" type="Const" version="opset1">
555
+ <data element_type="i64" shape="4" offset="6512" size="32" />
556
+ <output>
557
+ <port id="0" precision="I64">
558
+ <dim>4</dim>
559
+ </port>
560
+ </output>
561
+ </layer>
562
+ <layer id="39" name="__module.attention_pooler/aten::unflatten/Reshape_2" type="Reshape" version="opset1">
563
+ <data special_zero="true" />
564
+ <input>
565
+ <port id="0" precision="FP32">
566
+ <dim>-1</dim>
567
+ <dim>-1</dim>
568
+ <dim>16</dim>
569
+ </port>
570
+ <port id="1" precision="I64">
571
+ <dim>4</dim>
572
+ </port>
573
+ </input>
574
+ <output>
575
+ <port id="2" precision="FP32" names="78,x">
576
+ <dim>-1</dim>
577
+ <dim>-1</dim>
578
+ <dim>2</dim>
579
+ <dim>8</dim>
580
+ </port>
581
+ </output>
582
+ </layer>
583
+ <layer id="40" name="Constant_160760" type="Const" version="opset1">
584
+ <data element_type="i64" shape="4" offset="6544" size="32" />
585
+ <output>
586
+ <port id="0" precision="I64">
587
+ <dim>4</dim>
588
+ </port>
589
+ </output>
590
+ </layer>
591
+ <layer id="41" name="__module.attention_pooler/aten::permute/Transpose_2" type="Transpose" version="opset1">
592
+ <input>
593
+ <port id="0" precision="FP32">
594
+ <dim>-1</dim>
595
+ <dim>-1</dim>
596
+ <dim>2</dim>
597
+ <dim>8</dim>
598
+ </port>
599
+ <port id="1" precision="I64">
600
+ <dim>4</dim>
601
+ </port>
602
+ </input>
603
+ <output>
604
+ <port id="2" precision="FP32" names="84,value">
605
+ <dim>-1</dim>
606
+ <dim>2</dim>
607
+ <dim>-1</dim>
608
+ <dim>8</dim>
609
+ </port>
610
+ </output>
611
+ </layer>
612
+ <layer id="42" name="__module.attention_pooler/aten::scaled_dot_product_attention/ScaledDotProductAttention" type="ScaledDotProductAttention" version="opset13">
613
+ <data causal="false" />
614
+ <input>
615
+ <port id="0" precision="FP32">
616
+ <dim>-1</dim>
617
+ <dim>2</dim>
618
+ <dim>1</dim>
619
+ <dim>8</dim>
620
+ </port>
621
+ <port id="1" precision="FP32">
622
+ <dim>-1</dim>
623
+ <dim>2</dim>
624
+ <dim>-1</dim>
625
+ <dim>8</dim>
626
+ </port>
627
+ <port id="2" precision="FP32">
628
+ <dim>-1</dim>
629
+ <dim>2</dim>
630
+ <dim>-1</dim>
631
+ <dim>8</dim>
632
+ </port>
633
+ </input>
634
+ <output>
635
+ <port id="3" precision="FP32" names="85,out">
636
+ <dim>-1</dim>
637
+ <dim>2</dim>
638
+ <dim>1</dim>
639
+ <dim>8</dim>
640
+ </port>
641
+ </output>
642
+ </layer>
643
+ <layer id="43" name="Constant_163699" type="Const" version="opset1">
644
+ <data element_type="i64" shape="3" offset="7664" size="24" />
645
+ <output>
646
+ <port id="0" precision="I64">
647
+ <dim>3</dim>
648
+ </port>
649
+ </output>
650
+ </layer>
651
+ <layer id="44" name="__module.attention_pooler/aten::flatten/Reshape" type="Reshape" version="opset1">
652
+ <data special_zero="false" />
653
+ <input>
654
+ <port id="0" precision="FP32">
655
+ <dim>-1</dim>
656
+ <dim>2</dim>
657
+ <dim>1</dim>
658
+ <dim>8</dim>
659
+ </port>
660
+ <port id="1" precision="I64">
661
+ <dim>3</dim>
662
+ </port>
663
+ </input>
664
+ <output>
665
+ <port id="2" precision="FP32" names="88,input.7">
666
+ <dim>-1</dim>
667
+ <dim>1</dim>
668
+ <dim>16</dim>
669
+ </port>
670
+ </output>
671
+ </layer>
672
+ <layer id="45" name="self.attention_pooler.to_out.weight" type="Const" version="opset1">
673
+ <data element_type="f32" shape="16, 16" offset="7688" size="1024" />
674
+ <output>
675
+ <port id="0" precision="FP32" names="self.attention_pooler.to_out.weight">
676
+ <dim>16</dim>
677
+ <dim>16</dim>
678
+ </port>
679
+ </output>
680
+ </layer>
681
+ <layer id="46" name="__module.attention_pooler.to_out/aten::linear/MatMul" type="MatMul" version="opset1">
682
+ <data transpose_a="false" transpose_b="true" />
683
+ <input>
684
+ <port id="0" precision="FP32">
685
+ <dim>-1</dim>
686
+ <dim>1</dim>
687
+ <dim>16</dim>
688
+ </port>
689
+ <port id="1" precision="FP32">
690
+ <dim>16</dim>
691
+ <dim>16</dim>
692
+ </port>
693
+ </input>
694
+ <output>
695
+ <port id="2" precision="FP32">
696
+ <dim>-1</dim>
697
+ <dim>1</dim>
698
+ <dim>16</dim>
699
+ </port>
700
+ </output>
701
+ </layer>
702
+ <layer id="47" name="Constant_163524" type="Const" version="opset1">
703
+ <data element_type="f32" shape="1, 1, 16" offset="8712" size="64" />
704
+ <output>
705
+ <port id="0" precision="FP32">
706
+ <dim>1</dim>
707
+ <dim>1</dim>
708
+ <dim>16</dim>
709
+ </port>
710
+ </output>
711
+ </layer>
712
+ <layer id="48" name="__module.attention_pooler.to_out/aten::linear/Add" type="Add" version="opset1">
713
+ <data auto_broadcast="numpy" />
714
+ <input>
715
+ <port id="0" precision="FP32">
716
+ <dim>-1</dim>
717
+ <dim>1</dim>
718
+ <dim>16</dim>
719
+ </port>
720
+ <port id="1" precision="FP32">
721
+ <dim>1</dim>
722
+ <dim>1</dim>
723
+ <dim>16</dim>
724
+ </port>
725
+ </input>
726
+ <output>
727
+ <port id="2" precision="FP32" names="91">
728
+ <dim>-1</dim>
729
+ <dim>1</dim>
730
+ <dim>16</dim>
731
+ </port>
732
+ </output>
733
+ </layer>
734
+ <layer id="49" name="Concat_161244" type="Const" version="opset1">
735
+ <data element_type="i64" shape="2" offset="8776" size="16" />
736
+ <output>
737
+ <port id="0" precision="I64">
738
+ <dim>2</dim>
739
+ </port>
740
+ </output>
741
+ </layer>
742
+ <layer id="50" name="aten::flatten/Reshape" type="Reshape" version="opset1">
743
+ <data special_zero="true" />
744
+ <input>
745
+ <port id="0" precision="FP32">
746
+ <dim>-1</dim>
747
+ <dim>1</dim>
748
+ <dim>16</dim>
749
+ </port>
750
+ <port id="1" precision="I64">
751
+ <dim>2</dim>
752
+ </port>
753
+ </input>
754
+ <output>
755
+ <port id="2" precision="FP32" names="32,input.9">
756
+ <dim>-1</dim>
757
+ <dim>16</dim>
758
+ </port>
759
+ </output>
760
+ </layer>
761
+ <layer id="51" name="self.mlp_hidden.weight" type="Const" version="opset1">
762
+ <data element_type="f32" shape="16, 16" offset="8792" size="1024" />
763
+ <output>
764
+ <port id="0" precision="FP32" names="self.mlp_hidden.weight">
765
+ <dim>16</dim>
766
+ <dim>16</dim>
767
+ </port>
768
+ </output>
769
+ </layer>
770
+ <layer id="52" name="__module.mlp_hidden/aten::linear/MatMul" type="MatMul" version="opset1">
771
+ <data transpose_a="false" transpose_b="true" />
772
+ <input>
773
+ <port id="0" precision="FP32">
774
+ <dim>-1</dim>
775
+ <dim>16</dim>
776
+ </port>
777
+ <port id="1" precision="FP32">
778
+ <dim>16</dim>
779
+ <dim>16</dim>
780
+ </port>
781
+ </input>
782
+ <output>
783
+ <port id="2" precision="FP32">
784
+ <dim>-1</dim>
785
+ <dim>16</dim>
786
+ </port>
787
+ </output>
788
+ </layer>
789
+ <layer id="53" name="Constant_163525" type="Const" version="opset1">
790
+ <data element_type="f32" shape="1, 16" offset="9816" size="64" />
791
+ <output>
792
+ <port id="0" precision="FP32">
793
+ <dim>1</dim>
794
+ <dim>16</dim>
795
+ </port>
796
+ </output>
797
+ </layer>
798
+ <layer id="54" name="__module.mlp_hidden/aten::linear/Add" type="Add" version="opset1">
799
+ <data auto_broadcast="numpy" />
800
+ <input>
801
+ <port id="0" precision="FP32">
802
+ <dim>-1</dim>
803
+ <dim>16</dim>
804
+ </port>
805
+ <port id="1" precision="FP32">
806
+ <dim>1</dim>
807
+ <dim>16</dim>
808
+ </port>
809
+ </input>
810
+ <output>
811
+ <port id="2" precision="FP32" names="94">
812
+ <dim>-1</dim>
813
+ <dim>16</dim>
814
+ </port>
815
+ </output>
816
+ </layer>
817
+ <layer id="55" name="aten::gelu/Gelu" type="Gelu" version="opset7">
818
+ <data approximation_mode="TANH" />
819
+ <input>
820
+ <port id="0" precision="FP32">
821
+ <dim>-1</dim>
822
+ <dim>16</dim>
823
+ </port>
824
+ </input>
825
+ <output>
826
+ <port id="1" precision="FP32" names="35,input">
827
+ <dim>-1</dim>
828
+ <dim>16</dim>
829
+ </port>
830
+ </output>
831
+ </layer>
832
+ <layer id="56" name="self.mlp_out.weight" type="Const" version="opset1">
833
+ <data element_type="f32" shape="1, 16" offset="9880" size="64" />
834
+ <output>
835
+ <port id="0" precision="FP32" names="self.mlp_out.weight">
836
+ <dim>1</dim>
837
+ <dim>16</dim>
838
+ </port>
839
+ </output>
840
+ </layer>
841
+ <layer id="57" name="__module.mlp_out/aten::linear/MatMul" type="MatMul" version="opset1">
842
+ <data transpose_a="false" transpose_b="true" />
843
+ <input>
844
+ <port id="0" precision="FP32">
845
+ <dim>-1</dim>
846
+ <dim>16</dim>
847
+ </port>
848
+ <port id="1" precision="FP32">
849
+ <dim>1</dim>
850
+ <dim>16</dim>
851
+ </port>
852
+ </input>
853
+ <output>
854
+ <port id="2" precision="FP32">
855
+ <dim>-1</dim>
856
+ <dim>1</dim>
857
+ </port>
858
+ </output>
859
+ </layer>
860
+ <layer id="58" name="Constant_163526" type="Const" version="opset1">
861
+ <data element_type="f32" shape="1, 1" offset="9944" size="4" />
862
+ <output>
863
+ <port id="0" precision="FP32">
864
+ <dim>1</dim>
865
+ <dim>1</dim>
866
+ </port>
867
+ </output>
868
+ </layer>
869
+ <layer id="59" name="__module.mlp_out/aten::linear/Add" type="Add" version="opset1">
870
+ <data auto_broadcast="numpy" />
871
+ <input>
872
+ <port id="0" precision="FP32">
873
+ <dim>-1</dim>
874
+ <dim>1</dim>
875
+ </port>
876
+ <port id="1" precision="FP32">
877
+ <dim>1</dim>
878
+ <dim>1</dim>
879
+ </port>
880
+ </input>
881
+ <output>
882
+ <port id="2" precision="FP32" names="97">
883
+ <dim>-1</dim>
884
+ <dim>1</dim>
885
+ </port>
886
+ </output>
887
+ </layer>
888
+ <layer id="60" name="37" type="Const" version="opset1">
889
+ <data element_type="i64" shape="" offset="9952" size="8" />
890
+ <output>
891
+ <port id="0" precision="I64" names="37" />
892
+ </output>
893
+ </layer>
894
+ <layer id="61" name="aten::squeeze/Squeeze" type="Squeeze" version="opset1">
895
+ <input>
896
+ <port id="0" precision="FP32">
897
+ <dim>-1</dim>
898
+ <dim>1</dim>
899
+ </port>
900
+ <port id="1" precision="I64" />
901
+ </input>
902
+ <output>
903
+ <port id="2" precision="FP32" names="38,log_duration">
904
+ <dim>-1</dim>
905
+ </port>
906
+ </output>
907
+ </layer>
908
+ <layer id="62" name="aten::exp/Exp" type="Exp" version="opset1">
909
+ <input>
910
+ <port id="0" precision="FP32">
911
+ <dim>-1</dim>
912
+ </port>
913
+ </input>
914
+ <output>
915
+ <port id="1" precision="FP32" names="duration">
916
+ <dim>-1</dim>
917
+ </port>
918
+ </output>
919
+ </layer>
920
+ <layer id="63" name="Result_160868" type="Result" version="opset1" output_names="duration">
921
+ <input>
922
+ <port id="0" precision="FP32">
923
+ <dim>-1</dim>
924
+ </port>
925
+ </input>
926
+ </layer>
927
+ </layers>
928
+ <edges>
929
+ <edge from-layer="0" from-port="0" to-layer="4" to-port="0" />
930
+ <edge from-layer="1" from-port="0" to-layer="8" to-port="0" />
931
+ <edge from-layer="2" from-port="0" to-layer="23" to-port="0" />
932
+ <edge from-layer="3" from-port="0" to-layer="4" to-port="1" />
933
+ <edge from-layer="4" from-port="2" to-layer="6" to-port="0" />
934
+ <edge from-layer="5" from-port="0" to-layer="6" to-port="1" />
935
+ <edge from-layer="6" from-port="2" to-layer="11" to-port="0" />
936
+ <edge from-layer="7" from-port="0" to-layer="8" to-port="1" />
937
+ <edge from-layer="8" from-port="2" to-layer="10" to-port="0" />
938
+ <edge from-layer="9" from-port="0" to-layer="10" to-port="1" />
939
+ <edge from-layer="10" from-port="2" to-layer="11" to-port="1" />
940
+ <edge from-layer="11" from-port="2" to-layer="13" to-port="0" />
941
+ <edge from-layer="11" from-port="2" to-layer="35" to-port="0" />
942
+ <edge from-layer="12" from-port="0" to-layer="13" to-port="1" />
943
+ <edge from-layer="13" from-port="2" to-layer="15" to-port="0" />
944
+ <edge from-layer="14" from-port="0" to-layer="15" to-port="1" />
945
+ <edge from-layer="15" from-port="2" to-layer="16" to-port="0" />
946
+ <edge from-layer="15" from-port="2" to-layer="31" to-port="0" />
947
+ <edge from-layer="16" from-port="1" to-layer="19" to-port="0" />
948
+ <edge from-layer="17" from-port="0" to-layer="19" to-port="1" />
949
+ <edge from-layer="18" from-port="0" to-layer="19" to-port="2" />
950
+ <edge from-layer="19" from-port="3" to-layer="22" to-port="0" />
951
+ <edge from-layer="20" from-port="0" to-layer="22" to-port="1" />
952
+ <edge from-layer="21" from-port="0" to-layer="22" to-port="2" />
953
+ <edge from-layer="22" from-port="3" to-layer="23" to-port="1" />
954
+ <edge from-layer="23" from-port="2" to-layer="25" to-port="0" />
955
+ <edge from-layer="24" from-port="0" to-layer="25" to-port="1" />
956
+ <edge from-layer="25" from-port="2" to-layer="27" to-port="0" />
957
+ <edge from-layer="26" from-port="0" to-layer="27" to-port="1" />
958
+ <edge from-layer="27" from-port="2" to-layer="29" to-port="0" />
959
+ <edge from-layer="28" from-port="0" to-layer="29" to-port="1" />
960
+ <edge from-layer="29" from-port="2" to-layer="42" to-port="0" />
961
+ <edge from-layer="30" from-port="0" to-layer="31" to-port="1" />
962
+ <edge from-layer="31" from-port="2" to-layer="33" to-port="0" />
963
+ <edge from-layer="32" from-port="0" to-layer="33" to-port="1" />
964
+ <edge from-layer="33" from-port="2" to-layer="42" to-port="1" />
965
+ <edge from-layer="34" from-port="0" to-layer="35" to-port="1" />
966
+ <edge from-layer="35" from-port="2" to-layer="37" to-port="0" />
967
+ <edge from-layer="36" from-port="0" to-layer="37" to-port="1" />
968
+ <edge from-layer="37" from-port="2" to-layer="39" to-port="0" />
969
+ <edge from-layer="38" from-port="0" to-layer="39" to-port="1" />
970
+ <edge from-layer="39" from-port="2" to-layer="41" to-port="0" />
971
+ <edge from-layer="40" from-port="0" to-layer="41" to-port="1" />
972
+ <edge from-layer="41" from-port="2" to-layer="42" to-port="2" />
973
+ <edge from-layer="42" from-port="3" to-layer="44" to-port="0" />
974
+ <edge from-layer="43" from-port="0" to-layer="44" to-port="1" />
975
+ <edge from-layer="44" from-port="2" to-layer="46" to-port="0" />
976
+ <edge from-layer="45" from-port="0" to-layer="46" to-port="1" />
977
+ <edge from-layer="46" from-port="2" to-layer="48" to-port="0" />
978
+ <edge from-layer="47" from-port="0" to-layer="48" to-port="1" />
979
+ <edge from-layer="48" from-port="2" to-layer="50" to-port="0" />
980
+ <edge from-layer="49" from-port="0" to-layer="50" to-port="1" />
981
+ <edge from-layer="50" from-port="2" to-layer="52" to-port="0" />
982
+ <edge from-layer="51" from-port="0" to-layer="52" to-port="1" />
983
+ <edge from-layer="52" from-port="2" to-layer="54" to-port="0" />
984
+ <edge from-layer="53" from-port="0" to-layer="54" to-port="1" />
985
+ <edge from-layer="54" from-port="2" to-layer="55" to-port="0" />
986
+ <edge from-layer="55" from-port="1" to-layer="57" to-port="0" />
987
+ <edge from-layer="56" from-port="0" to-layer="57" to-port="1" />
988
+ <edge from-layer="57" from-port="2" to-layer="59" to-port="0" />
989
+ <edge from-layer="58" from-port="0" to-layer="59" to-port="1" />
990
+ <edge from-layer="59" from-port="2" to-layer="61" to-port="0" />
991
+ <edge from-layer="60" from-port="0" to-layer="61" to-port="1" />
992
+ <edge from-layer="61" from-port="2" to-layer="62" to-port="0" />
993
+ <edge from-layer="62" from-port="1" to-layer="63" to-port="0" />
994
+ </edges>
995
+ <rt_info>
996
+ <info name="OpenVINO Runtime" value="2026.5.0-23697-32776832f8d" />
997
+ <Runtime_version value="2026.5.0-23697-32776832f8d" />
998
+ <conversion_parameters>
999
+ <framework value="pytorch" />
1000
+ <is_python_object value="True" />
1001
+ </conversion_parameters>
1002
+ <optimum>
1003
+ <diffusers_version value="0.41.0" />
1004
+ <optimum_intel_version value="2.3.0.dev0+8c40189b" />
1005
+ <optimum_version value="2.3.0" />
1006
+ <pytorch_version value="2.15.0" />
1007
+ <transformers_version value="5.15.1" />
1008
+ </optimum>
1009
+ </rt_info>
1010
+ </net>
metadata.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_id": "optimum-intel-internal-testing/tiny-random-ltx2.5",
3
+ "model_class": "OVLTX2Pipeline",
4
+ "transformers_version": "5.15.1",
5
+ "optimum_intel_version": "2.3.0.dev0+8c40189b",
6
+ "openvino_version": "2026.5.0-23697",
7
+ "generated_date": "2026-11-01",
8
+ "components": [
9
+ "audio_vae_decoder/openvino_model.xml",
10
+ "connectors/openvino_model.xml",
11
+ "duration_head/openvino_model.xml",
12
+ "prompt_enhancer/openvino_audio_embeddings_model.xml",
13
+ "prompt_enhancer/openvino_language_model.xml",
14
+ "prompt_enhancer/openvino_text_embeddings_model.xml",
15
+ "prompt_enhancer/openvino_text_embeddings_per_layer_model.xml",
16
+ "prompt_enhancer/openvino_vision_embeddings_model.xml",
17
+ "text_encoder/openvino_model.xml",
18
+ "transformer/openvino_model.xml",
19
+ "vae_decoder/openvino_model.xml",
20
+ "vocoder/openvino_model.xml"
21
+ ]
22
+ }
model_index.json DELETED
@@ -1,48 +0,0 @@
1
- {
2
- "_class_name": "LTX2Pipeline",
3
- "_diffusers_version": "0.40.0",
4
- "audio_vae": [
5
- "diffusers",
6
- "AutoencoderKLLTX2Audio"
7
- ],
8
- "connectors": [
9
- "ltx2",
10
- "LTX2TextConnectors"
11
- ],
12
- "duration_head": [
13
- "ltx2",
14
- "LTX2DurationHead"
15
- ],
16
- "processor": [
17
- "transformers",
18
- "Gemma4Processor"
19
- ],
20
- "prompt_enhancer": [
21
- "transformers",
22
- "Gemma4ForConditionalGeneration"
23
- ],
24
- "scheduler": [
25
- "diffusers",
26
- "FlowMatchEulerDiscreteScheduler"
27
- ],
28
- "text_encoder": [
29
- "transformers",
30
- "Gemma4UnifiedForConditionalGeneration"
31
- ],
32
- "tokenizer": [
33
- "transformers",
34
- "GemmaTokenizer"
35
- ],
36
- "transformer": [
37
- "diffusers",
38
- "LTX2VideoTransformer3DModel"
39
- ],
40
- "vae": [
41
- "diffusers",
42
- "AutoencoderKLLTX2Video"
43
- ],
44
- "vocoder": [
45
- "ltx2",
46
- "LTX2VocoderWithBWE"
47
- ]
48
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
processor/chat_template.jinja DELETED
@@ -1,386 +0,0 @@
1
- {#
2
- Template: Google Gemma 4 Canonical Chat Template
3
- Author: Google Gemma Engineering Team
4
- Published: 2026-07-09
5
- Context: Fixed tool-calling loops, turn closures, and thinking content-ordering.
6
- #}
7
- {%- macro format_parameters(properties, required, filter_keys=false) -%}
8
- {%- set standard_keys = ['description', 'type', 'properties', 'required', 'nullable'] -%}
9
- {%- set ns = namespace(found_first=false) -%}
10
- {%- for key, value in properties | dictsort -%}
11
- {%- set add_comma = false -%}
12
- {%- if not filter_keys or key not in standard_keys -%}
13
- {%- if ns.found_first %},{% endif -%}
14
- {%- set ns.found_first = true -%}
15
- {{ key }}:{
16
- {%- if value['description'] -%}
17
- description:<|"|>{{ value['description'] }}<|"|>
18
- {%- set add_comma = true -%}
19
- {%- endif -%}
20
- {%- if value['type'] | upper == 'STRING' -%}
21
- {%- if value['enum'] -%}
22
- {%- if add_comma %},{%- else -%} {%- set add_comma = true -%} {% endif -%}
23
- enum:{{ format_argument(value['enum']) }}
24
- {%- endif -%}
25
- {%- elif value['type'] | upper == 'ARRAY' -%}
26
- {%- if value['items'] is mapping and value['items'] -%}
27
- {%- if add_comma %},{%- else -%} {%- set add_comma = true -%} {% endif -%}
28
- items:{
29
- {%- set ns_items = namespace(found_first=false) -%}
30
- {%- for item_key, item_value in value['items'] | dictsort -%}
31
- {%- if item_value is not none -%}
32
- {%- if ns_items.found_first %},{% endif -%}
33
- {%- set ns_items.found_first = true -%}
34
- {%- if item_key == 'properties' -%}
35
- properties:{
36
- {%- if item_value is mapping -%}
37
- {{- format_parameters(item_value, value['items']['required'] | default([])) -}}
38
- {%- endif -%}
39
- }
40
- {%- elif item_key == 'required' -%}
41
- required:[
42
- {%- for req_item in item_value -%}
43
- <|"|>{{- req_item -}}<|"|>
44
- {%- if not loop.last %},{% endif -%}
45
- {%- endfor -%}
46
- ]
47
- {%- elif item_key == 'type' -%}
48
- {%- if item_value is string -%}
49
- type:{{ format_argument(item_value | upper) }}
50
- {%- else -%}
51
- type:{{ format_argument(item_value | map('upper') | list) }}
52
- {%- endif -%}
53
- {%- else -%}
54
- {{ item_key }}:{{ format_argument(item_value) }}
55
- {%- endif -%}
56
- {%- endif -%}
57
- {%- endfor -%}
58
- }
59
- {%- endif -%}
60
- {%- endif -%}
61
- {%- if value['nullable'] %}
62
- {%- if add_comma %},{%- else -%} {%- set add_comma = true -%} {% endif -%}
63
- nullable:true
64
- {%- endif -%}
65
- {%- if value['type'] | upper == 'OBJECT' -%}
66
- {%- if value['properties'] is defined and value['properties'] is mapping -%}
67
- {%- if add_comma %},{%- else -%} {%- set add_comma = true -%} {% endif -%}
68
- properties:{
69
- {{- format_parameters(value['properties'], value['required'] | default([])) -}}
70
- }
71
- {%- elif value is mapping -%}
72
- {%- if add_comma %},{%- else -%} {%- set add_comma = true -%} {% endif -%}
73
- properties:{
74
- {{- format_parameters(value, value['required'] | default([]), filter_keys=true) -}}
75
- }
76
- {%- endif -%}
77
- {%- if value['required'] -%}
78
- {%- if add_comma %},{%- else -%} {%- set add_comma = true -%} {% endif -%}
79
- required:[
80
- {%- for item in value['required'] | default([]) -%}
81
- <|"|>{{- item -}}<|"|>
82
- {%- if not loop.last %},{% endif -%}
83
- {%- endfor -%}
84
- ]
85
- {%- endif -%}
86
- {%- endif -%}
87
- {%- if add_comma %},{%- else -%} {%- set add_comma = true -%} {% endif -%}
88
- type:<|"|>{{ value['type'] | upper }}<|"|>}
89
- {%- endif -%}
90
- {%- endfor -%}
91
- {%- endmacro -%}
92
- {%- macro format_function_declaration(tool_data) -%}
93
- declaration:{{- tool_data['function']['name'] -}}{description:<|"|>{{- tool_data['function']['description'] -}}<|"|>
94
- {%- set params = tool_data['function']['parameters'] -%}
95
- {%- if params -%}
96
- ,parameters:{
97
- {%- if params['properties'] -%}
98
- properties:{ {{- format_parameters(params['properties'], params['required']) -}} },
99
- {%- endif -%}
100
- {%- if params['required'] -%}
101
- required:[
102
- {%- for item in params['required'] -%}
103
- <|"|>{{- item -}}<|"|>
104
- {{- ',' if not loop.last -}}
105
- {%- endfor -%}
106
- ],
107
- {%- endif -%}
108
- {%- if params['type'] -%}
109
- type:<|"|>{{- params['type'] | upper -}}<|"|>}
110
- {%- endif -%}
111
- {%- endif -%}
112
- {%- if 'response' in tool_data['function'] -%}
113
- {%- set response_declaration = tool_data['function']['response'] -%}
114
- ,response:{
115
- {%- if response_declaration['description'] -%}
116
- description:<|"|>{{- response_declaration['description'] -}}<|"|>,
117
- {%- endif -%}
118
- {%- if response_declaration['type'] | upper == 'OBJECT' -%}
119
- type:<|"|>{{- response_declaration['type'] | upper -}}<|"|>}
120
- {%- endif -%}
121
- {%- endif -%}
122
- }
123
- {%- endmacro -%}
124
- {%- macro format_argument(argument, escape_keys=True) -%}
125
- {%- if argument is none -%}
126
- {{- 'null' -}}
127
- {%- elif argument is string -%}
128
- {{- '<|"|>' + argument + '<|"|>' -}}
129
- {%- elif argument is boolean -%}
130
- {{- 'true' if argument else 'false' -}}
131
- {%- elif argument is mapping -%}
132
- {{- '{' -}}
133
- {%- set ns = namespace(found_first=false) -%}
134
- {%- for key, value in argument | dictsort -%}
135
- {%- if ns.found_first %},{% endif -%}
136
- {%- set ns.found_first = true -%}
137
- {%- if escape_keys -%}
138
- {{- '<|"|>' + key + '<|"|>' -}}
139
- {%- else -%}
140
- {{- key -}}
141
- {%- endif -%}
142
- :{{- format_argument(value, escape_keys=escape_keys) -}}
143
- {%- endfor -%}
144
- {{- '}' -}}
145
- {%- elif argument is sequence -%}
146
- {{- '[' -}}
147
- {%- for item in argument -%}
148
- {{- format_argument(item, escape_keys=escape_keys) -}}
149
- {%- if not loop.last %},{% endif -%}
150
- {%- endfor -%}
151
- {{- ']' -}}
152
- {%- else -%}
153
- {{- argument -}}
154
- {%- endif -%}
155
- {%- endmacro -%}
156
- {%- macro strip_thinking(text) -%}
157
- {%- set ns = namespace(result='') -%}
158
- {%- for part in text.split('<channel|>') -%}
159
- {%- if '<|channel>' in part -%}
160
- {%- set ns.result = ns.result + part.split('<|channel>')[0] -%}
161
- {%- else -%}
162
- {%- set ns.result = ns.result + part -%}
163
- {%- endif -%}
164
- {%- endfor -%}
165
- {{- ns.result | trim -}}
166
- {%- endmacro -%}
167
-
168
- {%- macro format_tool_response_block(tool_name, response) -%}
169
- {{- '<|tool_response>' -}}
170
- {%- if response is mapping -%}
171
- {{- 'response:' + tool_name + '{' -}}
172
- {%- for key, value in response | dictsort -%}
173
- {{- key -}}:{{- format_argument(value, escape_keys=False) -}}
174
- {%- if not loop.last %},{% endif -%}
175
- {%- endfor -%}
176
- {{- '}' -}}
177
- {%- else -%}
178
- {{- 'response:' + tool_name + '{value:' + format_argument(response, escape_keys=False) + '}' -}}
179
- {%- endif -%}
180
- {{- '<tool_response|>' -}}
181
- {%- endmacro -%}
182
-
183
- {#- ===== SETUP ===== -#}
184
- {%- set ns = namespace(prev_message_type=None, prev_non_tool_role=None) -%}
185
- {%- set loop_messages = messages -%}
186
- {%- set enable_thinking = enable_thinking | default(false) -%}
187
- {%- set preserve_thinking = preserve_thinking | default(false) -%}
188
- {{- bos_token -}}
189
- {#- Handle System/Tool Definitions Block -#}
190
- {%- if enable_thinking or tools or (messages and messages[0]['role'] in ['system', 'developer']) -%}
191
- {{- '<|turn>system\n' -}}
192
- {#- Inject Thinking token at the very top of the FIRST system turn -#}
193
- {%- if enable_thinking -%}
194
- {{- '<|think|>\n' -}}
195
- {%- set ns.prev_message_type = 'think' -%}
196
- {%- endif -%}
197
- {%- if messages and messages[0]['role'] in ['system', 'developer'] -%}
198
- {%- if messages[0]['content'] is string -%}
199
- {{- messages[0]['content'] | trim -}}
200
- {%- elif messages[0]['content'] is sequence -%}
201
- {%- for item in messages[0]['content'] -%}
202
- {{- item['text'] | trim + ' '-}}
203
- {%- endfor -%}
204
- {%- endif -%}
205
- {%- set loop_messages = messages[1:] -%}
206
- {%- endif -%}
207
- {%- if tools -%}
208
- {%- for tool in tools %}
209
- {{- '<|tool>' -}}
210
- {{- format_function_declaration(tool) | trim -}}
211
- {{- '<tool|>' -}}
212
- {%- endfor %}
213
- {%- set ns.prev_message_type = 'tool' -%}
214
- {%- endif -%}
215
- {{- '<turn|>\n' -}}
216
- {%- endif %}
217
-
218
- {#- Pre-scan: find last user message index for reasoning guard -#}
219
- {%- set ns_turn = namespace(last_user_idx=-1) -%}
220
- {%- for i in range(loop_messages | length) -%}
221
- {%- if loop_messages[i]['role'] == 'user' -%}
222
- {%- set ns_turn.last_user_idx = i -%}
223
- {%- endif -%}
224
- {%- endfor -%}
225
-
226
- {#- Loop through messages -#}
227
- {%- for message in loop_messages -%}
228
- {%- if message['role'] != 'tool' -%}
229
- {%- set ns.prev_message_type = None -%}
230
- {%- set role = 'model' if message['role'] == 'assistant' else message['role'] -%}
231
- {#- Detect continuation using tracked state — O(1) instead of O(n) backward scan -#}
232
- {%- set continue_same_model_turn = (role == 'model' and ns.prev_non_tool_role == 'assistant') -%}
233
- {%- if not continue_same_model_turn -%}
234
- {{- '<|turn>' + role + '\n' }}
235
- {%- endif -%}
236
-
237
- {#- Render reasoning/reasoning_content as thinking channel -#}
238
- {%- set thinking_text = message.get('reasoning') or message.get('reasoning_content') -%}
239
- {%- set thinking_gate = (loop.index0 > ns_turn.last_user_idx) or (preserve_thinking and message.get('tool_calls')) -%}
240
- {%- if thinking_text and thinking_gate -%}
241
- {{- '<|channel>thought\n' + thinking_text + '\n<channel|>' -}}
242
- {%- endif -%}
243
-
244
- {%- if message.get('tool_calls') -%}
245
- {%- for tool_call in message.get('tool_calls') -%}
246
- {%- set function = tool_call['function'] -%}
247
- {{- '<|tool_call>call:' + function['name'] + '{' -}}
248
- {%- if function['arguments'] is mapping -%}
249
- {%- set ns_args = namespace(found_first=false) -%}
250
- {%- for key, value in function['arguments'] | dictsort -%}
251
- {%- if ns_args.found_first %},{% endif -%}
252
- {%- set ns_args.found_first = true -%}
253
- {{- key -}}:{{- format_argument(value, escape_keys=False) -}}
254
- {%- endfor -%}
255
- {%- elif function['arguments'] is none -%}
256
- {%- else -%}
257
- {{- raise_exception(
258
- "chat_template: tool_calls[].function.arguments must be a "
259
- "JSON object (mapping), not a string. Deserialize arguments "
260
- "before passing to the template."
261
- ) -}}
262
- {%- endif -%}
263
- {{- '}<tool_call|>' -}}
264
- {%- endfor -%}
265
- {%- set ns.prev_message_type = 'tool_call' -%}
266
- {%- endif -%}
267
-
268
- {%- set ns_tr_out = namespace(flag=false) -%}
269
- {%- if message.get('tool_responses') -%}
270
- {#- Legacy: tool_responses embedded on the assistant message (Google/Gemma native) -#}
271
- {%- for tool_response in message.get('tool_responses') -%}
272
- {{- format_tool_response_block(tool_response['name'] | default('unknown', true), tool_response['response']) -}}
273
- {%- set ns_tr_out.flag = true -%}
274
- {%- set ns.prev_message_type = 'tool_response' -%}
275
- {%- endfor -%}
276
- {%- elif message.get('tool_calls') -%}
277
- {#- OpenAI Chat Completions: forward-scan consecutive role:tool messages -#}
278
- {%- set ns_tool_scan = namespace(stopped=false) -%}
279
- {%- for k in range(loop.index0 + 1, loop_messages | length) -%}
280
- {%- if ns_tool_scan.stopped -%}
281
- {%- elif loop_messages[k]['role'] != 'tool' -%}
282
- {%- set ns_tool_scan.stopped = true -%}
283
- {%- else -%}
284
- {%- set follow = loop_messages[k] -%}
285
- {#- Resolve tool_call_id to function name -#}
286
- {%- set ns_tname = namespace(name=follow.get('name') or 'unknown') -%}
287
- {%- for tc in message.get('tool_calls') -%}
288
- {%- if tc.get('id') == follow.get('tool_call_id') -%}
289
- {%- set ns_tname.name = tc['function']['name'] -%}
290
- {%- endif -%}
291
- {%- endfor -%}
292
- {#- Handle content as string or content-parts array -#}
293
- {%- set tool_body = follow.get('content') -%}
294
- {%- if tool_body is string -%}
295
- {{- format_tool_response_block(ns_tname.name, tool_body) -}}
296
- {%- elif tool_body is sequence and tool_body is not string -%}
297
- {%- set ns_txt = namespace(s='') -%}
298
- {%- for part in tool_body -%}
299
- {%- if part.get('type') == 'text' -%}
300
- {%- set ns_txt.s = ns_txt.s + (part.get('text') | default('')) -%}
301
- {%- endif -%}
302
- {%- endfor -%}
303
- {{- format_tool_response_block(ns_tname.name, ns_txt.s) -}}
304
- {%- for part in tool_body -%}
305
- {%- if part.get('type') in ['image', 'image_url'] -%}
306
- {{- '<|image|>' -}}
307
- {%- elif part.get('type') in ['audio', 'input_audio'] -%}
308
- {{- '<|audio|>' -}}
309
- {%- elif part.get('type') == 'video' -%}
310
- {{- '<|video|>' -}}
311
- {%- endif -%}
312
- {%- endfor -%}
313
- {%- else -%}
314
- {{- format_tool_response_block(ns_tname.name, tool_body) -}}
315
- {%- endif -%}
316
- {%- set ns_tr_out.flag = true -%}
317
- {%- set ns.prev_message_type = 'tool_response' -%}
318
- {%- endif -%}
319
- {%- endfor -%}
320
- {%- endif -%}
321
-
322
- {%- set captured_content -%}
323
- {%- if message.get('content') is string -%}
324
- {%- if role == 'model' -%}
325
- {{- strip_thinking(message['content']) -}}
326
- {%- else -%}
327
- {{- message['content'] | trim -}}
328
- {%- endif -%}
329
- {%- elif message.get('content') is sequence -%}
330
- {%- for item in message['content'] -%}
331
- {%- if item.get('type') == 'text' -%}
332
- {%- if role == 'model' -%}
333
- {{- strip_thinking(item['text']) -}}
334
- {%- else -%}
335
- {{- item['text'] | trim -}}
336
- {%- endif -%}
337
- {%- elif item.get('type') in ['image', 'image_url'] -%}
338
- {{- '<|image|>' -}}
339
- {%- elif item.get('type') in ['audio', 'input_audio'] -%}
340
- {{- '<|audio|>' -}}
341
- {%- elif item.get('type') == 'video' -%}
342
- {{- '<|video|>' -}}
343
- {%- endif -%}
344
- {%- endfor -%}
345
- {%- endif -%}
346
- {%- endset -%}
347
-
348
- {{- captured_content -}}
349
- {%- set has_content = captured_content | trim | length > 0 -%}
350
-
351
- {#- Forward-scan: find next non-tool message role for continuation detection -#}
352
- {%- set next_nt = namespace(role=None, found=false) -%}
353
- {%- for j in range(loop.index0 + 1, loop_messages | length) -%}
354
- {%- if not next_nt.found -%}
355
- {%- if loop_messages[j]['role'] != 'tool' -%}
356
- {%- set next_nt.role = loop_messages[j]['role'] -%}
357
- {%- set next_nt.found = true -%}
358
- {%- endif -%}
359
- {%- endif -%}
360
- {%- endfor -%}
361
-
362
- {%- set continues_into_next = (
363
- role == 'model'
364
- and next_nt.role == 'assistant'
365
- and (not message.get('tool_calls') or ns_tr_out.flag)
366
- ) -%}
367
-
368
- {%- if ns.prev_message_type == 'tool_call' and not ns_tr_out.flag -%}
369
- {{- '<|tool_response>' -}}
370
- {%- elif continues_into_next -%}
371
- {%- elif not (ns_tr_out.flag and not has_content and not next_nt.found) -%}
372
- {{- '<turn|>\n' -}}
373
- {%- endif -%}
374
-
375
- {#- Track previous non-tool role for next iteration (avoids O(n) backward scan) -#}
376
- {%- set ns.prev_non_tool_role = message['role'] -%}
377
- {%- endif -%}
378
- {%- endfor -%}
379
-
380
- {%- if add_generation_prompt -%}
381
- {%- if ns.prev_message_type != 'tool_response' and ns.prev_message_type != 'tool_call' -%}
382
- {{- '<|turn>model\n' -}}
383
- {%- elif ns.prev_message_type == 'tool_response' and enable_thinking -%}
384
- {{- '<|channel>thought\n' -}}
385
- {%- endif -%}
386
- {%- endif -%}
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
processor/processor_config.json DELETED
@@ -1,75 +0,0 @@
1
- {
2
- "audio_ms_per_token": 40,
3
- "audio_seq_length": 750,
4
- "feature_extractor": {
5
- "dither": 0.0,
6
- "feature_extractor_type": "Gemma4AudioFeatureExtractor",
7
- "feature_size": 128,
8
- "fft_length": 512,
9
- "fft_overdrive": false,
10
- "frame_length": 320,
11
- "hop_length": 160,
12
- "input_scale_factor": 1.0,
13
- "max_frequency": 8000.0,
14
- "mel_floor": 0.001,
15
- "min_frequency": 0.0,
16
- "padding_side": "right",
17
- "padding_value": 0.0,
18
- "per_bin_mean": null,
19
- "per_bin_stddev": null,
20
- "preemphasis": 0.0,
21
- "preemphasis_htk_flavor": true,
22
- "return_attention_mask": true,
23
- "sampling_rate": 16000
24
- },
25
- "image_processor": {
26
- "do_convert_rgb": true,
27
- "do_normalize": false,
28
- "do_rescale": true,
29
- "do_resize": true,
30
- "image_mean": [
31
- 0.0,
32
- 0.0,
33
- 0.0
34
- ],
35
- "image_processor_type": "Gemma4ImageProcessor",
36
- "image_seq_length": 280,
37
- "image_std": [
38
- 1.0,
39
- 1.0,
40
- 1.0
41
- ],
42
- "max_soft_tokens": 280,
43
- "patch_size": 16,
44
- "pooling_kernel_size": 3,
45
- "resample": 3,
46
- "rescale_factor": 0.00392156862745098
47
- },
48
- "image_seq_length": 280,
49
- "processor_class": "Gemma4Processor",
50
- "video_processor": {
51
- "do_convert_rgb": true,
52
- "do_normalize": true,
53
- "do_rescale": true,
54
- "do_resize": true,
55
- "do_sample_frames": true,
56
- "image_mean": [
57
- 0.0,
58
- 0.0,
59
- 0.0
60
- ],
61
- "image_std": [
62
- 1.0,
63
- 1.0,
64
- 1.0
65
- ],
66
- "max_soft_tokens": 70,
67
- "num_frames": 32,
68
- "patch_size": 16,
69
- "pooling_kernel_size": 3,
70
- "resample": 3,
71
- "rescale_factor": 0.00392156862745098,
72
- "return_metadata": false,
73
- "video_processor_type": "Gemma4VideoProcessor"
74
- }
75
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
processor/tokenizer_config.json DELETED
@@ -1,141 +0,0 @@
1
- {
2
- "audio_token": "<|audio|>",
3
- "backend": "tokenizers",
4
- "boa_token": "<|audio>",
5
- "boi_token": "<|image>",
6
- "bos_token": "<bos>",
7
- "eoa_token": "<audio|>",
8
- "eoc_token": "<channel|>",
9
- "eoi_token": "<image|>",
10
- "eos_token": "<eos>",
11
- "eot_token": "<turn|>",
12
- "escape_token": "<|\"|>",
13
- "etc_token": "<tool_call|>",
14
- "etd_token": "<tool|>",
15
- "etr_token": "<tool_response|>",
16
- "extra_special_tokens": [
17
- "<|video|>"
18
- ],
19
- "image_token": "<|image|>",
20
- "is_local": false,
21
- "mask_token": "<mask>",
22
- "model_max_length": 1000000000000000019884624838656,
23
- "model_specific_special_tokens": {
24
- "audio_token": "<|audio|>",
25
- "boa_token": "<|audio>",
26
- "boi_token": "<|image>",
27
- "eoa_token": "<audio|>",
28
- "eoc_token": "<channel|>",
29
- "eoi_token": "<image|>",
30
- "eot_token": "<turn|>",
31
- "escape_token": "<|\"|>",
32
- "etc_token": "<tool_call|>",
33
- "etd_token": "<tool|>",
34
- "etr_token": "<tool_response|>",
35
- "image_token": "<|image|>",
36
- "soc_token": "<|channel>",
37
- "sot_token": "<|turn>",
38
- "stc_token": "<|tool_call>",
39
- "std_token": "<|tool>",
40
- "str_token": "<|tool_response>",
41
- "think_token": "<|think|>"
42
- },
43
- "pad_token": "<pad>",
44
- "padding_side": "left",
45
- "processor_class": "Gemma4Processor",
46
- "response_schema": {
47
- "properties": {
48
- "content": {
49
- "type": "string"
50
- },
51
- "role": {
52
- "const": "assistant"
53
- },
54
- "thinking": {
55
- "type": "string"
56
- },
57
- "tool_calls": {
58
- "items": {
59
- "properties": {
60
- "function": {
61
- "properties": {
62
- "arguments": {
63
- "additionalProperties": {},
64
- "type": "object",
65
- "x-parser": "gemma4-tool-call"
66
- },
67
- "name": {
68
- "type": "string"
69
- }
70
- },
71
- "type": "object",
72
- "x-regex": "call\\:(?P<name>\\w+)(?P<arguments>\\{.*\\})"
73
- },
74
- "type": {
75
- "const": "function"
76
- }
77
- },
78
- "type": "object"
79
- },
80
- "type": "array",
81
- "x-regex-iterator": "<\\|tool_call>(.*?)<tool_call\\|>"
82
- }
83
- },
84
- "type": "object",
85
- "x-regex": "(\\<\\|channel\\>thought\\n(?P<thinking>.*?)\\<channel\\|\\>)?(?P<tool_calls>\\<\\|tool_call\\>.*\\<tool_call\\|\\>)?(?P<content>(?:(?!\\<turn\\|\\>)(?!\\<\\|tool_response\\>).)+)?(?:\\<turn\\|\\>|\\<\\|tool_response\\>)?"
86
- },
87
- "response_template": {
88
- "defaults": {
89
- "role": "assistant"
90
- },
91
- "fields": {
92
- "content": {
93
- "close": [
94
- "<turn|>",
95
- "<|tool_response>",
96
- "<eos>"
97
- ],
98
- "content": "text"
99
- },
100
- "thinking": {
101
- "close": "<channel|>",
102
- "content": "text",
103
- "open": "<|channel>thought\n"
104
- },
105
- "tool_calls": {
106
- "close": "<tool_call|>",
107
- "content": "json",
108
- "content_args": {
109
- "string_delims": [
110
- [
111
- "<|\"|>",
112
- "<|\"|>"
113
- ]
114
- ],
115
- "unquoted_keys": true
116
- },
117
- "open_pattern": "<\\|tool_call>call:(?P<name>\\w+)",
118
- "repeats": true,
119
- "transform": {
120
- "function": {
121
- "arguments": "{content}",
122
- "name": "{name}"
123
- },
124
- "type": "function"
125
- }
126
- }
127
- },
128
- "start_anchor": [
129
- "<|turn>model\n",
130
- "<tool_response|>"
131
- ]
132
- },
133
- "soc_token": "<|channel>",
134
- "sot_token": "<|turn>",
135
- "stc_token": "<|tool_call>",
136
- "std_token": "<|tool>",
137
- "str_token": "<|tool_response>",
138
- "think_token": "<|think|>",
139
- "tokenizer_class": "GemmaTokenizer",
140
- "unk_token": "<unk>"
141
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
prompt_enhancer/config.json DELETED
@@ -1,159 +0,0 @@
1
- {
2
- "architectures": [
3
- "Gemma4ForConditionalGeneration"
4
- ],
5
- "audio_config": {
6
- "_name_or_path": "",
7
- "architectures": null,
8
- "attention_chunk_size": 12,
9
- "attention_context_left": 13,
10
- "attention_context_right": 0,
11
- "attention_invalid_logits_value": -1000000000.0,
12
- "attention_logit_cap": 50.0,
13
- "chunk_size_feed_forward": 0,
14
- "conv_kernel_size": 5,
15
- "dtype": "float32",
16
- "gradient_clipping": 10000000000.0,
17
- "hidden_act": "silu",
18
- "hidden_size": 32,
19
- "id2label": {
20
- "0": "LABEL_0",
21
- "1": "LABEL_1"
22
- },
23
- "initializer_range": 0.02,
24
- "is_encoder_decoder": false,
25
- "label2id": {
26
- "LABEL_0": 0,
27
- "LABEL_1": 1
28
- },
29
- "model_type": "gemma4_audio",
30
- "num_attention_heads": 4,
31
- "num_hidden_layers": 1,
32
- "output_attentions": false,
33
- "output_hidden_states": false,
34
- "output_proj_dims": 32,
35
- "problem_type": null,
36
- "residual_weight": 0.5,
37
- "return_dict": true,
38
- "rms_norm_eps": 1e-06,
39
- "subsampling_conv_channels": [
40
- 128,
41
- 32
42
- ],
43
- "use_clipped_linears": true
44
- },
45
- "audio_token_id": 258881,
46
- "boa_token_id": 256000,
47
- "boi_token_id": 255999,
48
- "dtype": "float32",
49
- "eoa_token_id": 258883,
50
- "eoa_token_index": 258883,
51
- "eoi_token_id": 258882,
52
- "eos_token_id": [
53
- 1,
54
- 106
55
- ],
56
- "image_token_id": 258880,
57
- "initializer_range": 0.02,
58
- "model_type": "gemma4",
59
- "text_config": {
60
- "attention_bias": false,
61
- "attention_dropout": 0.0,
62
- "attention_k_eq_v": false,
63
- "bos_token_id": 2,
64
- "dtype": "float32",
65
- "enable_moe_block": false,
66
- "eos_token_id": 1,
67
- "expert_intermediate_size": null,
68
- "final_logit_softcapping": 30.0,
69
- "global_head_dim": 16,
70
- "head_dim": 8,
71
- "hidden_activation": "gelu_pytorch_tanh",
72
- "hidden_size": 32,
73
- "hidden_size_per_layer_input": 4,
74
- "initializer_range": 0.02,
75
- "intermediate_size": 64,
76
- "layer_types": [
77
- "sliding_attention",
78
- "full_attention",
79
- "full_attention"
80
- ],
81
- "max_position_embeddings": 131072,
82
- "model_type": "gemma4_text",
83
- "moe_intermediate_size": null,
84
- "num_attention_heads": 4,
85
- "num_experts": null,
86
- "num_global_key_value_heads": null,
87
- "num_hidden_layers": 3,
88
- "num_key_value_heads": 1,
89
- "num_kv_shared_layers": 1,
90
- "pad_token_id": 0,
91
- "rms_norm_eps": 1e-06,
92
- "rope_parameters": {
93
- "full_attention": {
94
- "partial_rotary_factor": 0.25,
95
- "rope_theta": 1000000.0,
96
- "rope_type": "proportional"
97
- },
98
- "sliding_attention": {
99
- "rope_theta": 10000.0,
100
- "rope_type": "default"
101
- }
102
- },
103
- "sliding_window": 512,
104
- "tie_word_embeddings": true,
105
- "top_k_experts": null,
106
- "use_bidirectional_attention": null,
107
- "use_cache": true,
108
- "use_double_wide_mlp": true,
109
- "vocab_size": 262144,
110
- "vocab_size_per_layer_input": 262144
111
- },
112
- "tie_word_embeddings": true,
113
- "transformers_version": "5.5.4",
114
- "video_token_id": 258884,
115
- "vision_config": {
116
- "_name_or_path": "",
117
- "architectures": null,
118
- "attention_bias": false,
119
- "attention_dropout": 0.0,
120
- "chunk_size_feed_forward": 0,
121
- "default_output_length": 280,
122
- "dtype": "float32",
123
- "global_head_dim": 8,
124
- "head_dim": 8,
125
- "hidden_activation": "gelu_pytorch_tanh",
126
- "hidden_size": 32,
127
- "id2label": {
128
- "0": "LABEL_0",
129
- "1": "LABEL_1"
130
- },
131
- "initializer_range": 0.02,
132
- "intermediate_size": 64,
133
- "is_encoder_decoder": false,
134
- "label2id": {
135
- "LABEL_0": 0,
136
- "LABEL_1": 1
137
- },
138
- "max_position_embeddings": 131072,
139
- "model_type": "gemma4_vision",
140
- "num_attention_heads": 4,
141
- "num_hidden_layers": 1,
142
- "num_key_value_heads": 4,
143
- "output_attentions": false,
144
- "output_hidden_states": false,
145
- "patch_size": 16,
146
- "pooling_kernel_size": 3,
147
- "position_embedding_size": 10240,
148
- "problem_type": null,
149
- "return_dict": true,
150
- "rms_norm_eps": 1e-06,
151
- "rope_parameters": {
152
- "rope_theta": 100.0,
153
- "rope_type": "default"
154
- },
155
- "standardize": false,
156
- "use_clipped_linears": true
157
- },
158
- "vision_soft_tokens_per_image": 280
159
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
prompt_enhancer/generation_config.json DELETED
@@ -1,13 +0,0 @@
1
- {
2
- "_from_model_config": true,
3
- "bos_token_id": 2,
4
- "eos_token_id": [
5
- 1,
6
- 106
7
- ],
8
- "output_attentions": false,
9
- "output_hidden_states": false,
10
- "pad_token_id": 0,
11
- "transformers_version": "5.5.4",
12
- "use_cache": true
13
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
prompt_enhancer/model.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:3f9988219ad792099f3cb4350a70e107f80d5804d75d4b56208fc8f7b2d9852e
3
- size 49467948
 
 
 
 
text_encoder/model.safetensors → prompt_enhancer/openvino_audio_embeddings_model.bin RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3bd9fce60b59497ee7ba57267e92c4486a511e519fb89cdac5cb5380bb2f44ea
3
- size 334216
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b5f99104a4f5505160862f2153afb9376121a7a049a8b5b17fa7bc84377a917d
3
+ size 388924
prompt_enhancer/openvino_audio_embeddings_model.xml ADDED
The diff for this file is too large to render. See raw diff
 
processor/tokenizer.json → prompt_enhancer/openvino_language_model.bin RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:cc8d3a0ce36466ccc1278bf987df5f71db1719b9ca6b4118264f45cb627bfe0f
3
- size 32169626
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:96e06dd4f40fb9e454c21dac820a7fb54d7667204a1e3c7c484c51dabae6c1c9
3
+ size 33705124
prompt_enhancer/openvino_language_model.xml ADDED
The diff for this file is too large to render. See raw diff
 
prompt_enhancer/openvino_text_embeddings_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c8f8ed51b0c7e49a986cf0f41b880d80ad9d682736f16b03205d4524ac55577c
3
+ size 33554440
prompt_enhancer/openvino_text_embeddings_model.xml ADDED
@@ -0,0 +1,129 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ <?xml version="1.0"?>
2
+ <net name="Model6" version="11">
3
+ <layers>
4
+ <layer id="0" name="input" type="Parameter" version="opset1">
5
+ <data shape="?,?" element_type="i64" />
6
+ <output>
7
+ <port id="0" precision="I64" names="input">
8
+ <dim>-1</dim>
9
+ <dim>-1</dim>
10
+ </port>
11
+ </output>
12
+ </layer>
13
+ <layer id="1" name="self.model.model.language_model.embed_tokens.weight" type="Const" version="opset1">
14
+ <data element_type="f32" shape="262144, 32" offset="0" size="33554432" />
15
+ <output>
16
+ <port id="0" precision="FP32" names="self.model.model.language_model.embed_tokens.weight">
17
+ <dim>262144</dim>
18
+ <dim>32</dim>
19
+ </port>
20
+ </output>
21
+ </layer>
22
+ <layer id="2" name="__module.model.model.language_model.embed_tokens/aten::embedding/Convert" type="Convert" version="opset1">
23
+ <data destination_type="i32" />
24
+ <input>
25
+ <port id="0" precision="I64">
26
+ <dim>-1</dim>
27
+ <dim>-1</dim>
28
+ </port>
29
+ </input>
30
+ <output>
31
+ <port id="1" precision="I32">
32
+ <dim>-1</dim>
33
+ <dim>-1</dim>
34
+ </port>
35
+ </output>
36
+ </layer>
37
+ <layer id="3" name="__module.model.model.language_model.embed_tokens/aten::embedding/Constant" type="Const" version="opset1">
38
+ <data element_type="i32" shape="" offset="33554432" size="4" />
39
+ <output>
40
+ <port id="0" precision="I32" />
41
+ </output>
42
+ </layer>
43
+ <layer id="4" name="__module.model.model.language_model.embed_tokens/aten::embedding/Gather" type="Gather" version="opset8">
44
+ <data batch_dims="0" />
45
+ <input>
46
+ <port id="0" precision="FP32">
47
+ <dim>262144</dim>
48
+ <dim>32</dim>
49
+ </port>
50
+ <port id="1" precision="I32">
51
+ <dim>-1</dim>
52
+ <dim>-1</dim>
53
+ </port>
54
+ <port id="2" precision="I32" />
55
+ </input>
56
+ <output>
57
+ <port id="3" precision="FP32" names="14">
58
+ <dim>-1</dim>
59
+ <dim>-1</dim>
60
+ <dim>32</dim>
61
+ </port>
62
+ </output>
63
+ </layer>
64
+ <layer id="5" name="Constant_65033" type="Const" version="opset1">
65
+ <data element_type="f32" shape="1, 1, 1" offset="33554436" size="4" />
66
+ <output>
67
+ <port id="0" precision="FP32">
68
+ <dim>1</dim>
69
+ <dim>1</dim>
70
+ <dim>1</dim>
71
+ </port>
72
+ </output>
73
+ </layer>
74
+ <layer id="6" name="__module.model.model.language_model.embed_tokens/aten::mul/Multiply" type="Multiply" version="opset1">
75
+ <data auto_broadcast="numpy" />
76
+ <input>
77
+ <port id="0" precision="FP32">
78
+ <dim>-1</dim>
79
+ <dim>-1</dim>
80
+ <dim>32</dim>
81
+ </port>
82
+ <port id="1" precision="FP32">
83
+ <dim>1</dim>
84
+ <dim>1</dim>
85
+ <dim>1</dim>
86
+ </port>
87
+ </input>
88
+ <output>
89
+ <port id="2" precision="FP32" names="inputs_embeds">
90
+ <dim>-1</dim>
91
+ <dim>-1</dim>
92
+ <dim>32</dim>
93
+ </port>
94
+ </output>
95
+ </layer>
96
+ <layer id="7" name="Result_62802" type="Result" version="opset1" output_names="inputs_embeds">
97
+ <input>
98
+ <port id="0" precision="FP32">
99
+ <dim>-1</dim>
100
+ <dim>-1</dim>
101
+ <dim>32</dim>
102
+ </port>
103
+ </input>
104
+ </layer>
105
+ </layers>
106
+ <edges>
107
+ <edge from-layer="0" from-port="0" to-layer="2" to-port="0" />
108
+ <edge from-layer="1" from-port="0" to-layer="4" to-port="0" />
109
+ <edge from-layer="2" from-port="1" to-layer="4" to-port="1" />
110
+ <edge from-layer="3" from-port="0" to-layer="4" to-port="2" />
111
+ <edge from-layer="4" from-port="3" to-layer="6" to-port="0" />
112
+ <edge from-layer="5" from-port="0" to-layer="6" to-port="1" />
113
+ <edge from-layer="6" from-port="2" to-layer="7" to-port="0" />
114
+ </edges>
115
+ <rt_info>
116
+ <info name="OpenVINO Runtime" value="2026.5.0-23697-32776832f8d" />
117
+ <Runtime_version value="2026.5.0-23697-32776832f8d" />
118
+ <conversion_parameters>
119
+ <framework value="pytorch" />
120
+ <is_python_object value="True" />
121
+ </conversion_parameters>
122
+ <optimum>
123
+ <optimum_intel_version value="2.3.0.dev0+8c40189b" />
124
+ <optimum_version value="2.3.0" />
125
+ <pytorch_version value="2.15.0" />
126
+ <transformers_version value="5.15.1" />
127
+ </optimum>
128
+ </rt_info>
129
+ </net>
prompt_enhancer/openvino_text_embeddings_per_layer_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:69120fa1c3bb387bd977b971ef1e7f866852065c6d72673d241354e0bade9c73
3
+ size 12582992
prompt_enhancer/openvino_text_embeddings_per_layer_model.xml ADDED
@@ -0,0 +1,513 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ <?xml version="1.0"?>
2
+ <net name="Model12" version="11">
3
+ <layers>
4
+ <layer id="0" name="input_ids" type="Parameter" version="opset1">
5
+ <data shape="?,?" element_type="i64" />
6
+ <output>
7
+ <port id="0" precision="I64" names="input_ids">
8
+ <dim>-1</dim>
9
+ <dim>-1</dim>
10
+ </port>
11
+ </output>
12
+ </layer>
13
+ <layer id="1" name="self.language_model.embed_tokens_per_layer.weight" type="Const" version="opset1">
14
+ <data element_type="f32" shape="262144, 12" offset="0" size="12582912" />
15
+ <output>
16
+ <port id="0" precision="FP32" names="self.language_model.embed_tokens_per_layer.weight">
17
+ <dim>262144</dim>
18
+ <dim>12</dim>
19
+ </port>
20
+ </output>
21
+ </layer>
22
+ <layer id="2" name="Constant_93977" type="Const" version="opset1">
23
+ <data element_type="i64" shape="1, 1" offset="12582912" size="8" />
24
+ <output>
25
+ <port id="0" precision="I64">
26
+ <dim>1</dim>
27
+ <dim>1</dim>
28
+ </port>
29
+ </output>
30
+ </layer>
31
+ <layer id="3" name="aten::eq/Equal" type="Equal" version="opset1">
32
+ <data auto_broadcast="numpy" />
33
+ <input>
34
+ <port id="0" precision="I64">
35
+ <dim>-1</dim>
36
+ <dim>-1</dim>
37
+ </port>
38
+ <port id="1" precision="I64">
39
+ <dim>1</dim>
40
+ <dim>1</dim>
41
+ </port>
42
+ </input>
43
+ <output>
44
+ <port id="2" precision="BOOL" names="7">
45
+ <dim>-1</dim>
46
+ <dim>-1</dim>
47
+ </port>
48
+ </output>
49
+ </layer>
50
+ <layer id="4" name="8" type="Const" version="opset1">
51
+ <data element_type="i64" shape="" offset="12582920" size="8" />
52
+ <output>
53
+ <port id="0" precision="I64" names="8" />
54
+ </output>
55
+ </layer>
56
+ <layer id="5" name="aten::full_like/ShapeOf" type="ShapeOf" version="opset3">
57
+ <data output_type="i32" />
58
+ <input>
59
+ <port id="0" precision="I64">
60
+ <dim>-1</dim>
61
+ <dim>-1</dim>
62
+ </port>
63
+ </input>
64
+ <output>
65
+ <port id="1" precision="I32">
66
+ <dim>2</dim>
67
+ </port>
68
+ </output>
69
+ </layer>
70
+ <layer id="6" name="aten::full_like/Broadcast" type="Broadcast" version="opset3">
71
+ <data mode="numpy" />
72
+ <input>
73
+ <port id="0" precision="I64" />
74
+ <port id="1" precision="I32">
75
+ <dim>2</dim>
76
+ </port>
77
+ </input>
78
+ <output>
79
+ <port id="2" precision="I64" names="14,24,34">
80
+ <dim>-1</dim>
81
+ <dim>-1</dim>
82
+ </port>
83
+ </output>
84
+ </layer>
85
+ <layer id="7" name="aten::where/Select" type="Select" version="opset1">
86
+ <data auto_broadcast="numpy" />
87
+ <input>
88
+ <port id="0" precision="BOOL">
89
+ <dim>-1</dim>
90
+ <dim>-1</dim>
91
+ </port>
92
+ <port id="1" precision="I64">
93
+ <dim>-1</dim>
94
+ <dim>-1</dim>
95
+ </port>
96
+ <port id="2" precision="I64">
97
+ <dim>-1</dim>
98
+ <dim>-1</dim>
99
+ </port>
100
+ </input>
101
+ <output>
102
+ <port id="3" precision="I64" names="15,per_layer_inputs_tokens.3">
103
+ <dim>-1</dim>
104
+ <dim>-1</dim>
105
+ </port>
106
+ </output>
107
+ </layer>
108
+ <layer id="8" name="Constant_93978" type="Const" version="opset1">
109
+ <data element_type="i64" shape="1, 1" offset="12582928" size="8" />
110
+ <output>
111
+ <port id="0" precision="I64">
112
+ <dim>1</dim>
113
+ <dim>1</dim>
114
+ </port>
115
+ </output>
116
+ </layer>
117
+ <layer id="9" name="aten::eq/Equal_1" type="Equal" version="opset1">
118
+ <data auto_broadcast="numpy" />
119
+ <input>
120
+ <port id="0" precision="I64">
121
+ <dim>-1</dim>
122
+ <dim>-1</dim>
123
+ </port>
124
+ <port id="1" precision="I64">
125
+ <dim>1</dim>
126
+ <dim>1</dim>
127
+ </port>
128
+ </input>
129
+ <output>
130
+ <port id="2" precision="BOOL" names="17">
131
+ <dim>-1</dim>
132
+ <dim>-1</dim>
133
+ </port>
134
+ </output>
135
+ </layer>
136
+ <layer id="10" name="aten::where/Select_1" type="Select" version="opset1">
137
+ <data auto_broadcast="numpy" />
138
+ <input>
139
+ <port id="0" precision="BOOL">
140
+ <dim>-1</dim>
141
+ <dim>-1</dim>
142
+ </port>
143
+ <port id="1" precision="I64">
144
+ <dim>-1</dim>
145
+ <dim>-1</dim>
146
+ </port>
147
+ <port id="2" precision="I64">
148
+ <dim>-1</dim>
149
+ <dim>-1</dim>
150
+ </port>
151
+ </input>
152
+ <output>
153
+ <port id="3" precision="I64" names="25,per_layer_inputs_tokens.5">
154
+ <dim>-1</dim>
155
+ <dim>-1</dim>
156
+ </port>
157
+ </output>
158
+ </layer>
159
+ <layer id="11" name="Constant_93979" type="Const" version="opset1">
160
+ <data element_type="i64" shape="1, 1" offset="12582936" size="8" />
161
+ <output>
162
+ <port id="0" precision="I64">
163
+ <dim>1</dim>
164
+ <dim>1</dim>
165
+ </port>
166
+ </output>
167
+ </layer>
168
+ <layer id="12" name="aten::eq/Equal_2" type="Equal" version="opset1">
169
+ <data auto_broadcast="numpy" />
170
+ <input>
171
+ <port id="0" precision="I64">
172
+ <dim>-1</dim>
173
+ <dim>-1</dim>
174
+ </port>
175
+ <port id="1" precision="I64">
176
+ <dim>1</dim>
177
+ <dim>1</dim>
178
+ </port>
179
+ </input>
180
+ <output>
181
+ <port id="2" precision="BOOL" names="27">
182
+ <dim>-1</dim>
183
+ <dim>-1</dim>
184
+ </port>
185
+ </output>
186
+ </layer>
187
+ <layer id="13" name="aten::where/Select_2" type="Select" version="opset1">
188
+ <data auto_broadcast="numpy" />
189
+ <input>
190
+ <port id="0" precision="BOOL">
191
+ <dim>-1</dim>
192
+ <dim>-1</dim>
193
+ </port>
194
+ <port id="1" precision="I64">
195
+ <dim>-1</dim>
196
+ <dim>-1</dim>
197
+ </port>
198
+ <port id="2" precision="I64">
199
+ <dim>-1</dim>
200
+ <dim>-1</dim>
201
+ </port>
202
+ </input>
203
+ <output>
204
+ <port id="3" precision="I64" names="35,per_layer_inputs_tokens">
205
+ <dim>-1</dim>
206
+ <dim>-1</dim>
207
+ </port>
208
+ </output>
209
+ </layer>
210
+ <layer id="14" name="Constant_93980" type="Const" version="opset1">
211
+ <data element_type="i64" shape="1, 1" offset="12582920" size="8" />
212
+ <output>
213
+ <port id="0" precision="I64">
214
+ <dim>1</dim>
215
+ <dim>1</dim>
216
+ </port>
217
+ </output>
218
+ </layer>
219
+ <layer id="15" name="aten::ge/GreaterEqual" type="GreaterEqual" version="opset1">
220
+ <data auto_broadcast="numpy" />
221
+ <input>
222
+ <port id="0" precision="I64">
223
+ <dim>-1</dim>
224
+ <dim>-1</dim>
225
+ </port>
226
+ <port id="1" precision="I64">
227
+ <dim>1</dim>
228
+ <dim>1</dim>
229
+ </port>
230
+ </input>
231
+ <output>
232
+ <port id="2" precision="BOOL" names="37">
233
+ <dim>-1</dim>
234
+ <dim>-1</dim>
235
+ </port>
236
+ </output>
237
+ </layer>
238
+ <layer id="16" name="Constant_93981" type="Const" version="opset1">
239
+ <data element_type="i64" shape="1, 1" offset="12582944" size="8" />
240
+ <output>
241
+ <port id="0" precision="I64">
242
+ <dim>1</dim>
243
+ <dim>1</dim>
244
+ </port>
245
+ </output>
246
+ </layer>
247
+ <layer id="17" name="aten::lt/Less" type="Less" version="opset1">
248
+ <data auto_broadcast="numpy" />
249
+ <input>
250
+ <port id="0" precision="I64">
251
+ <dim>-1</dim>
252
+ <dim>-1</dim>
253
+ </port>
254
+ <port id="1" precision="I64">
255
+ <dim>1</dim>
256
+ <dim>1</dim>
257
+ </port>
258
+ </input>
259
+ <output>
260
+ <port id="2" precision="BOOL" names="39">
261
+ <dim>-1</dim>
262
+ <dim>-1</dim>
263
+ </port>
264
+ </output>
265
+ </layer>
266
+ <layer id="18" name="aten::logical_and/LogicalAnd" type="LogicalAnd" version="opset1">
267
+ <data auto_broadcast="numpy" />
268
+ <input>
269
+ <port id="0" precision="BOOL">
270
+ <dim>-1</dim>
271
+ <dim>-1</dim>
272
+ </port>
273
+ <port id="1" precision="BOOL">
274
+ <dim>-1</dim>
275
+ <dim>-1</dim>
276
+ </port>
277
+ </input>
278
+ <output>
279
+ <port id="2" precision="BOOL" names="40,per_layer_inputs_mask">
280
+ <dim>-1</dim>
281
+ <dim>-1</dim>
282
+ </port>
283
+ </output>
284
+ </layer>
285
+ <layer id="19" name="aten::zeros_like/Constant" type="Const" version="opset1">
286
+ <data element_type="f32" shape="" offset="12582952" size="4" />
287
+ <output>
288
+ <port id="0" precision="FP32" />
289
+ </output>
290
+ </layer>
291
+ <layer id="20" name="aten::zeros_like/Broadcast" type="Broadcast" version="opset3">
292
+ <data mode="numpy" />
293
+ <input>
294
+ <port id="0" precision="FP32" />
295
+ <port id="1" precision="I32">
296
+ <dim>2</dim>
297
+ </port>
298
+ </input>
299
+ <output>
300
+ <port id="2" precision="FP32">
301
+ <dim>-1</dim>
302
+ <dim>-1</dim>
303
+ </port>
304
+ </output>
305
+ </layer>
306
+ <layer id="21" name="aten::zeros_like/ConvertLike" type="Convert" version="opset1">
307
+ <data destination_type="i64" />
308
+ <input>
309
+ <port id="0" precision="FP32">
310
+ <dim>-1</dim>
311
+ <dim>-1</dim>
312
+ </port>
313
+ </input>
314
+ <output>
315
+ <port id="1" precision="I64" names="46">
316
+ <dim>-1</dim>
317
+ <dim>-1</dim>
318
+ </port>
319
+ </output>
320
+ </layer>
321
+ <layer id="22" name="aten::where/Select_3" type="Select" version="opset1">
322
+ <data auto_broadcast="numpy" />
323
+ <input>
324
+ <port id="0" precision="BOOL">
325
+ <dim>-1</dim>
326
+ <dim>-1</dim>
327
+ </port>
328
+ <port id="1" precision="I64">
329
+ <dim>-1</dim>
330
+ <dim>-1</dim>
331
+ </port>
332
+ <port id="2" precision="I64">
333
+ <dim>-1</dim>
334
+ <dim>-1</dim>
335
+ </port>
336
+ </input>
337
+ <output>
338
+ <port id="3" precision="I64" names="47,input">
339
+ <dim>-1</dim>
340
+ <dim>-1</dim>
341
+ </port>
342
+ </output>
343
+ </layer>
344
+ <layer id="23" name="__module.language_model.embed_tokens_per_layer/aten::embedding/Convert" type="Convert" version="opset1">
345
+ <data destination_type="i32" />
346
+ <input>
347
+ <port id="0" precision="I64">
348
+ <dim>-1</dim>
349
+ <dim>-1</dim>
350
+ </port>
351
+ </input>
352
+ <output>
353
+ <port id="1" precision="I32">
354
+ <dim>-1</dim>
355
+ <dim>-1</dim>
356
+ </port>
357
+ </output>
358
+ </layer>
359
+ <layer id="24" name="__module.language_model.embed_tokens_per_layer/aten::embedding/Constant" type="Const" version="opset1">
360
+ <data element_type="i32" shape="" offset="12582952" size="4" />
361
+ <output>
362
+ <port id="0" precision="I32" />
363
+ </output>
364
+ </layer>
365
+ <layer id="25" name="__module.language_model.embed_tokens_per_layer/aten::embedding/Gather" type="Gather" version="opset8">
366
+ <data batch_dims="0" />
367
+ <input>
368
+ <port id="0" precision="FP32">
369
+ <dim>262144</dim>
370
+ <dim>12</dim>
371
+ </port>
372
+ <port id="1" precision="I32">
373
+ <dim>-1</dim>
374
+ <dim>-1</dim>
375
+ </port>
376
+ <port id="2" precision="I32" />
377
+ </input>
378
+ <output>
379
+ <port id="3" precision="FP32" names="68">
380
+ <dim>-1</dim>
381
+ <dim>-1</dim>
382
+ <dim>12</dim>
383
+ </port>
384
+ </output>
385
+ </layer>
386
+ <layer id="26" name="Constant_93982" type="Const" version="opset1">
387
+ <data element_type="f32" shape="1, 1, 1" offset="12582956" size="4" />
388
+ <output>
389
+ <port id="0" precision="FP32">
390
+ <dim>1</dim>
391
+ <dim>1</dim>
392
+ <dim>1</dim>
393
+ </port>
394
+ </output>
395
+ </layer>
396
+ <layer id="27" name="__module.language_model.embed_tokens_per_layer/aten::mul/Multiply" type="Multiply" version="opset1">
397
+ <data auto_broadcast="numpy" />
398
+ <input>
399
+ <port id="0" precision="FP32">
400
+ <dim>-1</dim>
401
+ <dim>-1</dim>
402
+ <dim>12</dim>
403
+ </port>
404
+ <port id="1" precision="FP32">
405
+ <dim>1</dim>
406
+ <dim>1</dim>
407
+ <dim>1</dim>
408
+ </port>
409
+ </input>
410
+ <output>
411
+ <port id="2" precision="FP32" names="70">
412
+ <dim>-1</dim>
413
+ <dim>-1</dim>
414
+ <dim>12</dim>
415
+ </port>
416
+ </output>
417
+ </layer>
418
+ <layer id="28" name="Constant_94014" type="Const" version="opset1">
419
+ <data element_type="i64" shape="4" offset="12582960" size="32" />
420
+ <output>
421
+ <port id="0" precision="I64">
422
+ <dim>4</dim>
423
+ </port>
424
+ </output>
425
+ </layer>
426
+ <layer id="29" name="aten::reshape/Reshape" type="Reshape" version="opset1">
427
+ <data special_zero="true" />
428
+ <input>
429
+ <port id="0" precision="FP32">
430
+ <dim>-1</dim>
431
+ <dim>-1</dim>
432
+ <dim>12</dim>
433
+ </port>
434
+ <port id="1" precision="I64">
435
+ <dim>4</dim>
436
+ </port>
437
+ </input>
438
+ <output>
439
+ <port id="2" precision="FP32" names="text_embeds_per_layer">
440
+ <dim>-1</dim>
441
+ <dim>-1</dim>
442
+ <dim>3</dim>
443
+ <dim>4</dim>
444
+ </port>
445
+ </output>
446
+ </layer>
447
+ <layer id="30" name="Result_91519" type="Result" version="opset1" output_names="text_embeds_per_layer">
448
+ <input>
449
+ <port id="0" precision="FP32">
450
+ <dim>-1</dim>
451
+ <dim>-1</dim>
452
+ <dim>3</dim>
453
+ <dim>4</dim>
454
+ </port>
455
+ </input>
456
+ </layer>
457
+ </layers>
458
+ <edges>
459
+ <edge from-layer="0" from-port="0" to-layer="5" to-port="0" />
460
+ <edge from-layer="0" from-port="0" to-layer="3" to-port="0" />
461
+ <edge from-layer="0" from-port="0" to-layer="7" to-port="2" />
462
+ <edge from-layer="1" from-port="0" to-layer="25" to-port="0" />
463
+ <edge from-layer="2" from-port="0" to-layer="3" to-port="1" />
464
+ <edge from-layer="3" from-port="2" to-layer="7" to-port="0" />
465
+ <edge from-layer="4" from-port="0" to-layer="6" to-port="0" />
466
+ <edge from-layer="5" from-port="1" to-layer="20" to-port="1" />
467
+ <edge from-layer="5" from-port="1" to-layer="6" to-port="1" />
468
+ <edge from-layer="6" from-port="2" to-layer="10" to-port="1" />
469
+ <edge from-layer="6" from-port="2" to-layer="13" to-port="1" />
470
+ <edge from-layer="6" from-port="2" to-layer="7" to-port="1" />
471
+ <edge from-layer="7" from-port="3" to-layer="9" to-port="0" />
472
+ <edge from-layer="7" from-port="3" to-layer="10" to-port="2" />
473
+ <edge from-layer="8" from-port="0" to-layer="9" to-port="1" />
474
+ <edge from-layer="9" from-port="2" to-layer="10" to-port="0" />
475
+ <edge from-layer="10" from-port="3" to-layer="12" to-port="0" />
476
+ <edge from-layer="10" from-port="3" to-layer="13" to-port="2" />
477
+ <edge from-layer="11" from-port="0" to-layer="12" to-port="1" />
478
+ <edge from-layer="12" from-port="2" to-layer="13" to-port="0" />
479
+ <edge from-layer="13" from-port="3" to-layer="15" to-port="0" />
480
+ <edge from-layer="13" from-port="3" to-layer="17" to-port="0" />
481
+ <edge from-layer="13" from-port="3" to-layer="22" to-port="1" />
482
+ <edge from-layer="14" from-port="0" to-layer="15" to-port="1" />
483
+ <edge from-layer="15" from-port="2" to-layer="18" to-port="0" />
484
+ <edge from-layer="16" from-port="0" to-layer="17" to-port="1" />
485
+ <edge from-layer="17" from-port="2" to-layer="18" to-port="1" />
486
+ <edge from-layer="18" from-port="2" to-layer="22" to-port="0" />
487
+ <edge from-layer="19" from-port="0" to-layer="20" to-port="0" />
488
+ <edge from-layer="20" from-port="2" to-layer="21" to-port="0" />
489
+ <edge from-layer="21" from-port="1" to-layer="22" to-port="2" />
490
+ <edge from-layer="22" from-port="3" to-layer="23" to-port="0" />
491
+ <edge from-layer="23" from-port="1" to-layer="25" to-port="1" />
492
+ <edge from-layer="24" from-port="0" to-layer="25" to-port="2" />
493
+ <edge from-layer="25" from-port="3" to-layer="27" to-port="0" />
494
+ <edge from-layer="26" from-port="0" to-layer="27" to-port="1" />
495
+ <edge from-layer="27" from-port="2" to-layer="29" to-port="0" />
496
+ <edge from-layer="28" from-port="0" to-layer="29" to-port="1" />
497
+ <edge from-layer="29" from-port="2" to-layer="30" to-port="0" />
498
+ </edges>
499
+ <rt_info>
500
+ <info name="OpenVINO Runtime" value="2026.5.0-23697-32776832f8d" />
501
+ <Runtime_version value="2026.5.0-23697-32776832f8d" />
502
+ <conversion_parameters>
503
+ <framework value="pytorch" />
504
+ <is_python_object value="True" />
505
+ </conversion_parameters>
506
+ <optimum>
507
+ <optimum_intel_version value="2.3.0.dev0+8c40189b" />
508
+ <optimum_version value="2.3.0" />
509
+ <pytorch_version value="2.15.0" />
510
+ <transformers_version value="5.15.1" />
511
+ </optimum>
512
+ </rt_info>
513
+ </net>
prompt_enhancer/openvino_vision_embeddings_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e8f3e5b3da06fa3771e583a7572e351e2fac47be7cbd5ed576ae3177b8fc0d68
3
+ size 1454300
prompt_enhancer/openvino_vision_embeddings_model.xml ADDED
The diff for this file is too large to render. See raw diff
 
scheduler/scheduler_config.json DELETED
@@ -1,18 +0,0 @@
1
- {
2
- "_class_name": "FlowMatchEulerDiscreteScheduler",
3
- "_diffusers_version": "0.40.0",
4
- "base_image_seq_len": 1024,
5
- "base_shift": 0.95,
6
- "invert_sigmas": false,
7
- "max_image_seq_len": 4096,
8
- "max_shift": 2.05,
9
- "num_train_timesteps": 1000,
10
- "shift": 1.0,
11
- "shift_terminal": null,
12
- "stochastic_sampling": false,
13
- "time_shift_type": "exponential",
14
- "use_beta_sigmas": false,
15
- "use_dynamic_shifting": false,
16
- "use_exponential_sigmas": false,
17
- "use_karras_sigmas": false
18
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
text_encoder/config.json DELETED
@@ -1,122 +0,0 @@
1
- {
2
- "architectures": [
3
- "Gemma4UnifiedForConditionalGeneration"
4
- ],
5
- "audio_config": {
6
- "_name_or_path": "",
7
- "architectures": null,
8
- "audio_embed_dim": 32,
9
- "chunk_size_feed_forward": 0,
10
- "dtype": null,
11
- "id2label": {
12
- "0": "LABEL_0",
13
- "1": "LABEL_1"
14
- },
15
- "initializer_range": 0.02,
16
- "is_encoder_decoder": false,
17
- "label2id": {
18
- "LABEL_0": 0,
19
- "LABEL_1": 1
20
- },
21
- "model_type": "gemma4_unified_audio",
22
- "output_attentions": false,
23
- "output_hidden_states": false,
24
- "problem_type": null,
25
- "return_dict": true,
26
- "rms_norm_eps": 1e-06
27
- },
28
- "audio_token_id": 258881,
29
- "boa_token_id": 256000,
30
- "boi_token_id": 255999,
31
- "dtype": "float32",
32
- "eoa_token_index": 258883,
33
- "eoi_token_id": 258882,
34
- "eos_token_id": [
35
- 1,
36
- 106
37
- ],
38
- "image_token_id": 258880,
39
- "initializer_range": 0.02,
40
- "model_type": "gemma4_unified",
41
- "text_config": {
42
- "attention_bias": false,
43
- "attention_dropout": 0.0,
44
- "attention_k_eq_v": true,
45
- "bos_token_id": 2,
46
- "enable_moe_block": false,
47
- "eos_token_id": 1,
48
- "final_logit_softcapping": 30.0,
49
- "global_head_dim": 16,
50
- "head_dim": 8,
51
- "hidden_activation": "gelu_pytorch_tanh",
52
- "hidden_size": 32,
53
- "hidden_size_per_layer_input": 0,
54
- "initializer_range": 0.02,
55
- "intermediate_size": 64,
56
- "layer_types": [
57
- "sliding_attention",
58
- "full_attention"
59
- ],
60
- "max_position_embeddings": 1024,
61
- "model_type": "gemma4_unified_text",
62
- "moe_intermediate_size": null,
63
- "num_attention_heads": 4,
64
- "num_experts": null,
65
- "num_global_key_value_heads": 1,
66
- "num_hidden_layers": 2,
67
- "num_key_value_heads": 2,
68
- "num_kv_shared_layers": 0,
69
- "pad_token_id": 0,
70
- "rms_norm_eps": 1e-06,
71
- "rope_parameters": {
72
- "full_attention": {
73
- "partial_rotary_factor": 0.25,
74
- "rope_theta": 1000000.0,
75
- "rope_type": "proportional"
76
- },
77
- "sliding_attention": {
78
- "rope_theta": 10000.0,
79
- "rope_type": "default"
80
- }
81
- },
82
- "sliding_window": 64,
83
- "tie_word_embeddings": true,
84
- "top_k_experts": null,
85
- "use_bidirectional_attention": "vision",
86
- "use_cache": true,
87
- "use_double_wide_mlp": false,
88
- "vocab_size": 1024,
89
- "vocab_size_per_layer_input": 1024
90
- },
91
- "tie_word_embeddings": true,
92
- "transformers_version": "5.10.4",
93
- "video_token_id": 258884,
94
- "vision_config": {
95
- "_name_or_path": "",
96
- "architectures": null,
97
- "chunk_size_feed_forward": 0,
98
- "dtype": null,
99
- "id2label": {
100
- "0": "LABEL_0",
101
- "1": "LABEL_1"
102
- },
103
- "initializer_range": 0.02,
104
- "is_encoder_decoder": false,
105
- "label2id": {
106
- "LABEL_0": 0,
107
- "LABEL_1": 1
108
- },
109
- "mm_embed_dim": 32,
110
- "mm_posemb_size": 16,
111
- "model_type": "gemma4_unified_vision",
112
- "num_soft_tokens": 4,
113
- "output_attentions": false,
114
- "output_hidden_states": false,
115
- "output_proj_dims": 32,
116
- "patch_size": 16,
117
- "pooling_kernel_size": 1,
118
- "problem_type": null,
119
- "return_dict": true,
120
- "rms_norm_eps": 1e-06
121
- }
122
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
text_encoder/generation_config.json DELETED
@@ -1,13 +0,0 @@
1
- {
2
- "_from_model_config": true,
3
- "bos_token_id": 2,
4
- "eos_token_id": [
5
- 1,
6
- 106
7
- ],
8
- "output_attentions": false,
9
- "output_hidden_states": false,
10
- "pad_token_id": 0,
11
- "transformers_version": "5.10.4",
12
- "use_cache": true
13
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
connectors/diffusion_pytorch_model.safetensors → text_encoder/openvino_model.bin RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:195116083dd54d3ef7f83fa39fab401005de53b76729c83393a2d00d2a7929fd
3
- size 133392
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d9ccbe773151093d8710a0e25f0835edc72da633bb7382183269361900038425
3
+ size 211440
text_encoder/openvino_model.xml ADDED
The diff for this file is too large to render. See raw diff
 
tiny_model_marker.json DELETED
@@ -1,13 +0,0 @@
1
- {
2
- "marker_version": 1,
3
- "base_repo_id": "optimum-intel-internal-testing/tiny-random-ltx2.5",
4
- "base_revision": "5e56afb1e7f81f48e6d2aeaf0f3acb55fc4cf68e",
5
- "source_model_id": "Lightricks/LTX-2.5-Diffusers",
6
- "enhancer_repo_id": "optimum-intel-internal-testing/tiny-random-gemma4",
7
- "seed": 42,
8
- "duration_head": {
9
- "pooler_hidden_dim": 16,
10
- "num_pooler_heads": 2,
11
- "mlp_hidden_dim": 16
12
- }
13
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
tokenizer/tokenizer.json DELETED
@@ -1,1148 +0,0 @@
1
- {
2
- "version": "1.0",
3
- "truncation": null,
4
- "padding": null,
5
- "added_tokens": [
6
- {
7
- "id": 0,
8
- "content": "<pad>",
9
- "single_word": false,
10
- "lstrip": false,
11
- "rstrip": false,
12
- "normalized": false,
13
- "special": true
14
- },
15
- {
16
- "id": 1,
17
- "content": "<eos>",
18
- "single_word": false,
19
- "lstrip": false,
20
- "rstrip": false,
21
- "normalized": false,
22
- "special": true
23
- },
24
- {
25
- "id": 2,
26
- "content": "<bos>",
27
- "single_word": false,
28
- "lstrip": false,
29
- "rstrip": false,
30
- "normalized": false,
31
- "special": true
32
- },
33
- {
34
- "id": 3,
35
- "content": "<unk>",
36
- "single_word": false,
37
- "lstrip": false,
38
- "rstrip": false,
39
- "normalized": false,
40
- "special": true
41
- },
42
- {
43
- "id": 1024,
44
- "content": "<mask>",
45
- "single_word": false,
46
- "lstrip": false,
47
- "rstrip": false,
48
- "normalized": false,
49
- "special": true
50
- }
51
- ],
52
- "normalizer": {
53
- "type": "Replace",
54
- "pattern": {
55
- "String": " "
56
- },
57
- "content": "▁"
58
- },
59
- "pre_tokenizer": {
60
- "type": "Split",
61
- "pattern": {
62
- "String": " "
63
- },
64
- "behavior": "MergedWithPrevious",
65
- "invert": false
66
- },
67
- "post_processor": {
68
- "type": "TemplateProcessing",
69
- "single": [
70
- {
71
- "Sequence": {
72
- "id": "A",
73
- "type_id": 0
74
- }
75
- }
76
- ],
77
- "pair": [
78
- {
79
- "Sequence": {
80
- "id": "A",
81
- "type_id": 0
82
- }
83
- },
84
- {
85
- "Sequence": {
86
- "id": "B",
87
- "type_id": 1
88
- }
89
- }
90
- ],
91
- "special_tokens": {}
92
- },
93
- "decoder": {
94
- "type": "Sequence",
95
- "decoders": [
96
- {
97
- "type": "Replace",
98
- "pattern": {
99
- "String": "▁"
100
- },
101
- "content": " "
102
- },
103
- {
104
- "type": "ByteFallback"
105
- },
106
- {
107
- "type": "Fuse"
108
- }
109
- ]
110
- },
111
- "model": {
112
- "type": "BPE",
113
- "dropout": null,
114
- "unk_token": "<unk>",
115
- "continuing_subword_prefix": null,
116
- "end_of_word_suffix": null,
117
- "fuse_unk": true,
118
- "byte_fallback": true,
119
- "ignore_merges": false,
120
- "vocab": {
121
- "<pad>": 0,
122
- "<eos>": 1,
123
- "<bos>": 2,
124
- "<unk>": 3,
125
- "v4": 4,
126
- "v5": 5,
127
- "v6": 6,
128
- "v7": 7,
129
- "v8": 8,
130
- "v9": 9,
131
- "v10": 10,
132
- "v11": 11,
133
- "v12": 12,
134
- "v13": 13,
135
- "v14": 14,
136
- "v15": 15,
137
- "v16": 16,
138
- "v17": 17,
139
- "v18": 18,
140
- "v19": 19,
141
- "v20": 20,
142
- "v21": 21,
143
- "v22": 22,
144
- "v23": 23,
145
- "v24": 24,
146
- "v25": 25,
147
- "v26": 26,
148
- "v27": 27,
149
- "v28": 28,
150
- "v29": 29,
151
- "v30": 30,
152
- "v31": 31,
153
- "v32": 32,
154
- "v33": 33,
155
- "v34": 34,
156
- "v35": 35,
157
- "v36": 36,
158
- "v37": 37,
159
- "v38": 38,
160
- "v39": 39,
161
- "v40": 40,
162
- "v41": 41,
163
- "v42": 42,
164
- "v43": 43,
165
- "v44": 44,
166
- "v45": 45,
167
- "v46": 46,
168
- "v47": 47,
169
- "v48": 48,
170
- "v49": 49,
171
- "v50": 50,
172
- "v51": 51,
173
- "v52": 52,
174
- "v53": 53,
175
- "v54": 54,
176
- "v55": 55,
177
- "v56": 56,
178
- "v57": 57,
179
- "v58": 58,
180
- "v59": 59,
181
- "v60": 60,
182
- "v61": 61,
183
- "v62": 62,
184
- "v63": 63,
185
- "v64": 64,
186
- "v65": 65,
187
- "v66": 66,
188
- "v67": 67,
189
- "v68": 68,
190
- "v69": 69,
191
- "v70": 70,
192
- "v71": 71,
193
- "v72": 72,
194
- "v73": 73,
195
- "v74": 74,
196
- "v75": 75,
197
- "v76": 76,
198
- "v77": 77,
199
- "v78": 78,
200
- "v79": 79,
201
- "v80": 80,
202
- "v81": 81,
203
- "v82": 82,
204
- "v83": 83,
205
- "v84": 84,
206
- "v85": 85,
207
- "v86": 86,
208
- "v87": 87,
209
- "v88": 88,
210
- "v89": 89,
211
- "v90": 90,
212
- "v91": 91,
213
- "v92": 92,
214
- "v93": 93,
215
- "v94": 94,
216
- "v95": 95,
217
- "v96": 96,
218
- "v97": 97,
219
- "v98": 98,
220
- "v99": 99,
221
- "v100": 100,
222
- "v101": 101,
223
- "v102": 102,
224
- "v103": 103,
225
- "v104": 104,
226
- "v105": 105,
227
- "v106": 106,
228
- "v107": 107,
229
- "v108": 108,
230
- "v109": 109,
231
- "v110": 110,
232
- "v111": 111,
233
- "v112": 112,
234
- "v113": 113,
235
- "v114": 114,
236
- "v115": 115,
237
- "v116": 116,
238
- "v117": 117,
239
- "v118": 118,
240
- "v119": 119,
241
- "v120": 120,
242
- "v121": 121,
243
- "v122": 122,
244
- "v123": 123,
245
- "v124": 124,
246
- "v125": 125,
247
- "v126": 126,
248
- "v127": 127,
249
- "v128": 128,
250
- "v129": 129,
251
- "v130": 130,
252
- "v131": 131,
253
- "v132": 132,
254
- "v133": 133,
255
- "v134": 134,
256
- "v135": 135,
257
- "v136": 136,
258
- "v137": 137,
259
- "v138": 138,
260
- "v139": 139,
261
- "v140": 140,
262
- "v141": 141,
263
- "v142": 142,
264
- "v143": 143,
265
- "v144": 144,
266
- "v145": 145,
267
- "v146": 146,
268
- "v147": 147,
269
- "v148": 148,
270
- "v149": 149,
271
- "v150": 150,
272
- "v151": 151,
273
- "v152": 152,
274
- "v153": 153,
275
- "v154": 154,
276
- "v155": 155,
277
- "v156": 156,
278
- "v157": 157,
279
- "v158": 158,
280
- "v159": 159,
281
- "v160": 160,
282
- "v161": 161,
283
- "v162": 162,
284
- "v163": 163,
285
- "v164": 164,
286
- "v165": 165,
287
- "v166": 166,
288
- "v167": 167,
289
- "v168": 168,
290
- "v169": 169,
291
- "v170": 170,
292
- "v171": 171,
293
- "v172": 172,
294
- "v173": 173,
295
- "v174": 174,
296
- "v175": 175,
297
- "v176": 176,
298
- "v177": 177,
299
- "v178": 178,
300
- "v179": 179,
301
- "v180": 180,
302
- "v181": 181,
303
- "v182": 182,
304
- "v183": 183,
305
- "v184": 184,
306
- "v185": 185,
307
- "v186": 186,
308
- "v187": 187,
309
- "v188": 188,
310
- "v189": 189,
311
- "v190": 190,
312
- "v191": 191,
313
- "v192": 192,
314
- "v193": 193,
315
- "v194": 194,
316
- "v195": 195,
317
- "v196": 196,
318
- "v197": 197,
319
- "v198": 198,
320
- "v199": 199,
321
- "v200": 200,
322
- "v201": 201,
323
- "v202": 202,
324
- "v203": 203,
325
- "v204": 204,
326
- "v205": 205,
327
- "v206": 206,
328
- "v207": 207,
329
- "v208": 208,
330
- "v209": 209,
331
- "v210": 210,
332
- "v211": 211,
333
- "v212": 212,
334
- "v213": 213,
335
- "v214": 214,
336
- "v215": 215,
337
- "v216": 216,
338
- "v217": 217,
339
- "v218": 218,
340
- "v219": 219,
341
- "v220": 220,
342
- "v221": 221,
343
- "v222": 222,
344
- "v223": 223,
345
- "v224": 224,
346
- "v225": 225,
347
- "v226": 226,
348
- "v227": 227,
349
- "v228": 228,
350
- "v229": 229,
351
- "v230": 230,
352
- "v231": 231,
353
- "v232": 232,
354
- "v233": 233,
355
- "v234": 234,
356
- "v235": 235,
357
- "v236": 236,
358
- "v237": 237,
359
- "v238": 238,
360
- "v239": 239,
361
- "v240": 240,
362
- "v241": 241,
363
- "v242": 242,
364
- "v243": 243,
365
- "v244": 244,
366
- "v245": 245,
367
- "v246": 246,
368
- "v247": 247,
369
- "v248": 248,
370
- "v249": 249,
371
- "v250": 250,
372
- "v251": 251,
373
- "v252": 252,
374
- "v253": 253,
375
- "v254": 254,
376
- "v255": 255,
377
- "v256": 256,
378
- "v257": 257,
379
- "v258": 258,
380
- "v259": 259,
381
- "v260": 260,
382
- "v261": 261,
383
- "v262": 262,
384
- "v263": 263,
385
- "v264": 264,
386
- "v265": 265,
387
- "v266": 266,
388
- "v267": 267,
389
- "v268": 268,
390
- "v269": 269,
391
- "v270": 270,
392
- "v271": 271,
393
- "v272": 272,
394
- "v273": 273,
395
- "v274": 274,
396
- "v275": 275,
397
- "v276": 276,
398
- "v277": 277,
399
- "v278": 278,
400
- "v279": 279,
401
- "v280": 280,
402
- "v281": 281,
403
- "v282": 282,
404
- "v283": 283,
405
- "v284": 284,
406
- "v285": 285,
407
- "v286": 286,
408
- "v287": 287,
409
- "v288": 288,
410
- "v289": 289,
411
- "v290": 290,
412
- "v291": 291,
413
- "v292": 292,
414
- "v293": 293,
415
- "v294": 294,
416
- "v295": 295,
417
- "v296": 296,
418
- "v297": 297,
419
- "v298": 298,
420
- "v299": 299,
421
- "v300": 300,
422
- "v301": 301,
423
- "v302": 302,
424
- "v303": 303,
425
- "v304": 304,
426
- "v305": 305,
427
- "v306": 306,
428
- "v307": 307,
429
- "v308": 308,
430
- "v309": 309,
431
- "v310": 310,
432
- "v311": 311,
433
- "v312": 312,
434
- "v313": 313,
435
- "v314": 314,
436
- "v315": 315,
437
- "v316": 316,
438
- "v317": 317,
439
- "v318": 318,
440
- "v319": 319,
441
- "v320": 320,
442
- "v321": 321,
443
- "v322": 322,
444
- "v323": 323,
445
- "v324": 324,
446
- "v325": 325,
447
- "v326": 326,
448
- "v327": 327,
449
- "v328": 328,
450
- "v329": 329,
451
- "v330": 330,
452
- "v331": 331,
453
- "v332": 332,
454
- "v333": 333,
455
- "v334": 334,
456
- "v335": 335,
457
- "v336": 336,
458
- "v337": 337,
459
- "v338": 338,
460
- "v339": 339,
461
- "v340": 340,
462
- "v341": 341,
463
- "v342": 342,
464
- "v343": 343,
465
- "v344": 344,
466
- "v345": 345,
467
- "v346": 346,
468
- "v347": 347,
469
- "v348": 348,
470
- "v349": 349,
471
- "v350": 350,
472
- "v351": 351,
473
- "v352": 352,
474
- "v353": 353,
475
- "v354": 354,
476
- "v355": 355,
477
- "v356": 356,
478
- "v357": 357,
479
- "v358": 358,
480
- "v359": 359,
481
- "v360": 360,
482
- "v361": 361,
483
- "v362": 362,
484
- "v363": 363,
485
- "v364": 364,
486
- "v365": 365,
487
- "v366": 366,
488
- "v367": 367,
489
- "v368": 368,
490
- "v369": 369,
491
- "v370": 370,
492
- "v371": 371,
493
- "v372": 372,
494
- "v373": 373,
495
- "v374": 374,
496
- "v375": 375,
497
- "v376": 376,
498
- "v377": 377,
499
- "v378": 378,
500
- "v379": 379,
501
- "v380": 380,
502
- "v381": 381,
503
- "v382": 382,
504
- "v383": 383,
505
- "v384": 384,
506
- "v385": 385,
507
- "v386": 386,
508
- "v387": 387,
509
- "v388": 388,
510
- "v389": 389,
511
- "v390": 390,
512
- "v391": 391,
513
- "v392": 392,
514
- "v393": 393,
515
- "v394": 394,
516
- "v395": 395,
517
- "v396": 396,
518
- "v397": 397,
519
- "v398": 398,
520
- "v399": 399,
521
- "v400": 400,
522
- "v401": 401,
523
- "v402": 402,
524
- "v403": 403,
525
- "v404": 404,
526
- "v405": 405,
527
- "v406": 406,
528
- "v407": 407,
529
- "v408": 408,
530
- "v409": 409,
531
- "v410": 410,
532
- "v411": 411,
533
- "v412": 412,
534
- "v413": 413,
535
- "v414": 414,
536
- "v415": 415,
537
- "v416": 416,
538
- "v417": 417,
539
- "v418": 418,
540
- "v419": 419,
541
- "v420": 420,
542
- "v421": 421,
543
- "v422": 422,
544
- "v423": 423,
545
- "v424": 424,
546
- "v425": 425,
547
- "v426": 426,
548
- "v427": 427,
549
- "v428": 428,
550
- "v429": 429,
551
- "v430": 430,
552
- "v431": 431,
553
- "v432": 432,
554
- "v433": 433,
555
- "v434": 434,
556
- "v435": 435,
557
- "v436": 436,
558
- "v437": 437,
559
- "v438": 438,
560
- "v439": 439,
561
- "v440": 440,
562
- "v441": 441,
563
- "v442": 442,
564
- "v443": 443,
565
- "v444": 444,
566
- "v445": 445,
567
- "v446": 446,
568
- "v447": 447,
569
- "v448": 448,
570
- "v449": 449,
571
- "v450": 450,
572
- "v451": 451,
573
- "v452": 452,
574
- "v453": 453,
575
- "v454": 454,
576
- "v455": 455,
577
- "v456": 456,
578
- "v457": 457,
579
- "v458": 458,
580
- "v459": 459,
581
- "v460": 460,
582
- "v461": 461,
583
- "v462": 462,
584
- "v463": 463,
585
- "v464": 464,
586
- "v465": 465,
587
- "v466": 466,
588
- "v467": 467,
589
- "v468": 468,
590
- "v469": 469,
591
- "v470": 470,
592
- "v471": 471,
593
- "v472": 472,
594
- "v473": 473,
595
- "v474": 474,
596
- "v475": 475,
597
- "v476": 476,
598
- "v477": 477,
599
- "v478": 478,
600
- "v479": 479,
601
- "v480": 480,
602
- "v481": 481,
603
- "v482": 482,
604
- "v483": 483,
605
- "v484": 484,
606
- "v485": 485,
607
- "v486": 486,
608
- "v487": 487,
609
- "v488": 488,
610
- "v489": 489,
611
- "v490": 490,
612
- "v491": 491,
613
- "v492": 492,
614
- "v493": 493,
615
- "v494": 494,
616
- "v495": 495,
617
- "v496": 496,
618
- "v497": 497,
619
- "v498": 498,
620
- "v499": 499,
621
- "v500": 500,
622
- "v501": 501,
623
- "v502": 502,
624
- "v503": 503,
625
- "v504": 504,
626
- "v505": 505,
627
- "v506": 506,
628
- "v507": 507,
629
- "v508": 508,
630
- "v509": 509,
631
- "v510": 510,
632
- "v511": 511,
633
- "v512": 512,
634
- "v513": 513,
635
- "v514": 514,
636
- "v515": 515,
637
- "v516": 516,
638
- "v517": 517,
639
- "v518": 518,
640
- "v519": 519,
641
- "v520": 520,
642
- "v521": 521,
643
- "v522": 522,
644
- "v523": 523,
645
- "v524": 524,
646
- "v525": 525,
647
- "v526": 526,
648
- "v527": 527,
649
- "v528": 528,
650
- "v529": 529,
651
- "v530": 530,
652
- "v531": 531,
653
- "v532": 532,
654
- "v533": 533,
655
- "v534": 534,
656
- "v535": 535,
657
- "v536": 536,
658
- "v537": 537,
659
- "v538": 538,
660
- "v539": 539,
661
- "v540": 540,
662
- "v541": 541,
663
- "v542": 542,
664
- "v543": 543,
665
- "v544": 544,
666
- "v545": 545,
667
- "v546": 546,
668
- "v547": 547,
669
- "v548": 548,
670
- "v549": 549,
671
- "v550": 550,
672
- "v551": 551,
673
- "v552": 552,
674
- "v553": 553,
675
- "v554": 554,
676
- "v555": 555,
677
- "v556": 556,
678
- "v557": 557,
679
- "v558": 558,
680
- "v559": 559,
681
- "v560": 560,
682
- "v561": 561,
683
- "v562": 562,
684
- "v563": 563,
685
- "v564": 564,
686
- "v565": 565,
687
- "v566": 566,
688
- "v567": 567,
689
- "v568": 568,
690
- "v569": 569,
691
- "v570": 570,
692
- "v571": 571,
693
- "v572": 572,
694
- "v573": 573,
695
- "v574": 574,
696
- "v575": 575,
697
- "v576": 576,
698
- "v577": 577,
699
- "v578": 578,
700
- "v579": 579,
701
- "v580": 580,
702
- "v581": 581,
703
- "v582": 582,
704
- "v583": 583,
705
- "v584": 584,
706
- "v585": 585,
707
- "v586": 586,
708
- "v587": 587,
709
- "v588": 588,
710
- "v589": 589,
711
- "v590": 590,
712
- "v591": 591,
713
- "v592": 592,
714
- "v593": 593,
715
- "v594": 594,
716
- "v595": 595,
717
- "v596": 596,
718
- "v597": 597,
719
- "v598": 598,
720
- "v599": 599,
721
- "v600": 600,
722
- "v601": 601,
723
- "v602": 602,
724
- "v603": 603,
725
- "v604": 604,
726
- "v605": 605,
727
- "v606": 606,
728
- "v607": 607,
729
- "v608": 608,
730
- "v609": 609,
731
- "v610": 610,
732
- "v611": 611,
733
- "v612": 612,
734
- "v613": 613,
735
- "v614": 614,
736
- "v615": 615,
737
- "v616": 616,
738
- "v617": 617,
739
- "v618": 618,
740
- "v619": 619,
741
- "v620": 620,
742
- "v621": 621,
743
- "v622": 622,
744
- "v623": 623,
745
- "v624": 624,
746
- "v625": 625,
747
- "v626": 626,
748
- "v627": 627,
749
- "v628": 628,
750
- "v629": 629,
751
- "v630": 630,
752
- "v631": 631,
753
- "v632": 632,
754
- "v633": 633,
755
- "v634": 634,
756
- "v635": 635,
757
- "v636": 636,
758
- "v637": 637,
759
- "v638": 638,
760
- "v639": 639,
761
- "v640": 640,
762
- "v641": 641,
763
- "v642": 642,
764
- "v643": 643,
765
- "v644": 644,
766
- "v645": 645,
767
- "v646": 646,
768
- "v647": 647,
769
- "v648": 648,
770
- "v649": 649,
771
- "v650": 650,
772
- "v651": 651,
773
- "v652": 652,
774
- "v653": 653,
775
- "v654": 654,
776
- "v655": 655,
777
- "v656": 656,
778
- "v657": 657,
779
- "v658": 658,
780
- "v659": 659,
781
- "v660": 660,
782
- "v661": 661,
783
- "v662": 662,
784
- "v663": 663,
785
- "v664": 664,
786
- "v665": 665,
787
- "v666": 666,
788
- "v667": 667,
789
- "v668": 668,
790
- "v669": 669,
791
- "v670": 670,
792
- "v671": 671,
793
- "v672": 672,
794
- "v673": 673,
795
- "v674": 674,
796
- "v675": 675,
797
- "v676": 676,
798
- "v677": 677,
799
- "v678": 678,
800
- "v679": 679,
801
- "v680": 680,
802
- "v681": 681,
803
- "v682": 682,
804
- "v683": 683,
805
- "v684": 684,
806
- "v685": 685,
807
- "v686": 686,
808
- "v687": 687,
809
- "v688": 688,
810
- "v689": 689,
811
- "v690": 690,
812
- "v691": 691,
813
- "v692": 692,
814
- "v693": 693,
815
- "v694": 694,
816
- "v695": 695,
817
- "v696": 696,
818
- "v697": 697,
819
- "v698": 698,
820
- "v699": 699,
821
- "v700": 700,
822
- "v701": 701,
823
- "v702": 702,
824
- "v703": 703,
825
- "v704": 704,
826
- "v705": 705,
827
- "v706": 706,
828
- "v707": 707,
829
- "v708": 708,
830
- "v709": 709,
831
- "v710": 710,
832
- "v711": 711,
833
- "v712": 712,
834
- "v713": 713,
835
- "v714": 714,
836
- "v715": 715,
837
- "v716": 716,
838
- "v717": 717,
839
- "v718": 718,
840
- "v719": 719,
841
- "v720": 720,
842
- "v721": 721,
843
- "v722": 722,
844
- "v723": 723,
845
- "v724": 724,
846
- "v725": 725,
847
- "v726": 726,
848
- "v727": 727,
849
- "v728": 728,
850
- "v729": 729,
851
- "v730": 730,
852
- "v731": 731,
853
- "v732": 732,
854
- "v733": 733,
855
- "v734": 734,
856
- "v735": 735,
857
- "v736": 736,
858
- "v737": 737,
859
- "v738": 738,
860
- "v739": 739,
861
- "v740": 740,
862
- "v741": 741,
863
- "v742": 742,
864
- "v743": 743,
865
- "v744": 744,
866
- "v745": 745,
867
- "v746": 746,
868
- "v747": 747,
869
- "v748": 748,
870
- "v749": 749,
871
- "v750": 750,
872
- "v751": 751,
873
- "v752": 752,
874
- "v753": 753,
875
- "v754": 754,
876
- "v755": 755,
877
- "v756": 756,
878
- "v757": 757,
879
- "v758": 758,
880
- "v759": 759,
881
- "v760": 760,
882
- "v761": 761,
883
- "v762": 762,
884
- "v763": 763,
885
- "v764": 764,
886
- "v765": 765,
887
- "v766": 766,
888
- "v767": 767,
889
- "v768": 768,
890
- "v769": 769,
891
- "v770": 770,
892
- "v771": 771,
893
- "v772": 772,
894
- "v773": 773,
895
- "v774": 774,
896
- "v775": 775,
897
- "v776": 776,
898
- "v777": 777,
899
- "v778": 778,
900
- "v779": 779,
901
- "v780": 780,
902
- "v781": 781,
903
- "v782": 782,
904
- "v783": 783,
905
- "v784": 784,
906
- "v785": 785,
907
- "v786": 786,
908
- "v787": 787,
909
- "v788": 788,
910
- "v789": 789,
911
- "v790": 790,
912
- "v791": 791,
913
- "v792": 792,
914
- "v793": 793,
915
- "v794": 794,
916
- "v795": 795,
917
- "v796": 796,
918
- "v797": 797,
919
- "v798": 798,
920
- "v799": 799,
921
- "v800": 800,
922
- "v801": 801,
923
- "v802": 802,
924
- "v803": 803,
925
- "v804": 804,
926
- "v805": 805,
927
- "v806": 806,
928
- "v807": 807,
929
- "v808": 808,
930
- "v809": 809,
931
- "v810": 810,
932
- "v811": 811,
933
- "v812": 812,
934
- "v813": 813,
935
- "v814": 814,
936
- "v815": 815,
937
- "v816": 816,
938
- "v817": 817,
939
- "v818": 818,
940
- "v819": 819,
941
- "v820": 820,
942
- "v821": 821,
943
- "v822": 822,
944
- "v823": 823,
945
- "v824": 824,
946
- "v825": 825,
947
- "v826": 826,
948
- "v827": 827,
949
- "v828": 828,
950
- "v829": 829,
951
- "v830": 830,
952
- "v831": 831,
953
- "v832": 832,
954
- "v833": 833,
955
- "v834": 834,
956
- "v835": 835,
957
- "v836": 836,
958
- "v837": 837,
959
- "v838": 838,
960
- "v839": 839,
961
- "v840": 840,
962
- "v841": 841,
963
- "v842": 842,
964
- "v843": 843,
965
- "v844": 844,
966
- "v845": 845,
967
- "v846": 846,
968
- "v847": 847,
969
- "v848": 848,
970
- "v849": 849,
971
- "v850": 850,
972
- "v851": 851,
973
- "v852": 852,
974
- "v853": 853,
975
- "v854": 854,
976
- "v855": 855,
977
- "v856": 856,
978
- "v857": 857,
979
- "v858": 858,
980
- "v859": 859,
981
- "v860": 860,
982
- "v861": 861,
983
- "v862": 862,
984
- "v863": 863,
985
- "v864": 864,
986
- "v865": 865,
987
- "v866": 866,
988
- "v867": 867,
989
- "v868": 868,
990
- "v869": 869,
991
- "v870": 870,
992
- "v871": 871,
993
- "v872": 872,
994
- "v873": 873,
995
- "v874": 874,
996
- "v875": 875,
997
- "v876": 876,
998
- "v877": 877,
999
- "v878": 878,
1000
- "v879": 879,
1001
- "v880": 880,
1002
- "v881": 881,
1003
- "v882": 882,
1004
- "v883": 883,
1005
- "v884": 884,
1006
- "v885": 885,
1007
- "v886": 886,
1008
- "v887": 887,
1009
- "v888": 888,
1010
- "v889": 889,
1011
- "v890": 890,
1012
- "v891": 891,
1013
- "v892": 892,
1014
- "v893": 893,
1015
- "v894": 894,
1016
- "v895": 895,
1017
- "v896": 896,
1018
- "v897": 897,
1019
- "v898": 898,
1020
- "v899": 899,
1021
- "v900": 900,
1022
- "v901": 901,
1023
- "v902": 902,
1024
- "v903": 903,
1025
- "v904": 904,
1026
- "v905": 905,
1027
- "v906": 906,
1028
- "v907": 907,
1029
- "v908": 908,
1030
- "v909": 909,
1031
- "v910": 910,
1032
- "v911": 911,
1033
- "v912": 912,
1034
- "v913": 913,
1035
- "v914": 914,
1036
- "v915": 915,
1037
- "v916": 916,
1038
- "v917": 917,
1039
- "v918": 918,
1040
- "v919": 919,
1041
- "v920": 920,
1042
- "v921": 921,
1043
- "v922": 922,
1044
- "v923": 923,
1045
- "v924": 924,
1046
- "v925": 925,
1047
- "v926": 926,
1048
- "v927": 927,
1049
- "v928": 928,
1050
- "v929": 929,
1051
- "v930": 930,
1052
- "v931": 931,
1053
- "v932": 932,
1054
- "v933": 933,
1055
- "v934": 934,
1056
- "v935": 935,
1057
- "v936": 936,
1058
- "v937": 937,
1059
- "v938": 938,
1060
- "v939": 939,
1061
- "v940": 940,
1062
- "v941": 941,
1063
- "v942": 942,
1064
- "v943": 943,
1065
- "v944": 944,
1066
- "v945": 945,
1067
- "v946": 946,
1068
- "v947": 947,
1069
- "v948": 948,
1070
- "v949": 949,
1071
- "v950": 950,
1072
- "v951": 951,
1073
- "v952": 952,
1074
- "v953": 953,
1075
- "v954": 954,
1076
- "v955": 955,
1077
- "v956": 956,
1078
- "v957": 957,
1079
- "v958": 958,
1080
- "v959": 959,
1081
- "v960": 960,
1082
- "v961": 961,
1083
- "v962": 962,
1084
- "v963": 963,
1085
- "v964": 964,
1086
- "v965": 965,
1087
- "v966": 966,
1088
- "v967": 967,
1089
- "v968": 968,
1090
- "v969": 969,
1091
- "v970": 970,
1092
- "v971": 971,
1093
- "v972": 972,
1094
- "v973": 973,
1095
- "v974": 974,
1096
- "v975": 975,
1097
- "v976": 976,
1098
- "v977": 977,
1099
- "v978": 978,
1100
- "v979": 979,
1101
- "v980": 980,
1102
- "v981": 981,
1103
- "v982": 982,
1104
- "v983": 983,
1105
- "v984": 984,
1106
- "v985": 985,
1107
- "v986": 986,
1108
- "v987": 987,
1109
- "v988": 988,
1110
- "v989": 989,
1111
- "v990": 990,
1112
- "v991": 991,
1113
- "v992": 992,
1114
- "v993": 993,
1115
- "v994": 994,
1116
- "v995": 995,
1117
- "v996": 996,
1118
- "v997": 997,
1119
- "v998": 998,
1120
- "v999": 999,
1121
- "v1000": 1000,
1122
- "v1001": 1001,
1123
- "v1002": 1002,
1124
- "v1003": 1003,
1125
- "v1004": 1004,
1126
- "v1005": 1005,
1127
- "v1006": 1006,
1128
- "v1007": 1007,
1129
- "v1008": 1008,
1130
- "v1009": 1009,
1131
- "v1010": 1010,
1132
- "v1011": 1011,
1133
- "v1012": 1012,
1134
- "v1013": 1013,
1135
- "v1014": 1014,
1136
- "v1015": 1015,
1137
- "v1016": 1016,
1138
- "v1017": 1017,
1139
- "v1018": 1018,
1140
- "v1019": 1019,
1141
- "v1020": 1020,
1142
- "v1021": 1021,
1143
- "v1022": 1022,
1144
- "v1023": 1023
1145
- },
1146
- "merges": []
1147
- }
1148
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
tokenizer/tokenizer_config.json DELETED
@@ -1,12 +0,0 @@
1
- {
2
- "backend": "tokenizers",
3
- "bos_token": "<bos>",
4
- "eos_token": "<eos>",
5
- "is_local": true,
6
- "local_files_only": false,
7
- "mask_token": "<mask>",
8
- "model_max_length": 1024,
9
- "pad_token": "<pad>",
10
- "tokenizer_class": "GemmaTokenizer",
11
- "unk_token": "<unk>"
12
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
transformer/config.json DELETED
@@ -1,54 +0,0 @@
1
- {
2
- "_class_name": "LTX2VideoTransformer3DModel",
3
- "_diffusers_version": "0.40.0",
4
- "activation_fn": "gelu-approximate",
5
- "attention_bias": true,
6
- "attention_head_dim": 8,
7
- "attention_out_bias": true,
8
- "audio_attention_head_dim": 8,
9
- "audio_cross_attention_dim": 32,
10
- "audio_cross_attn_mod": true,
11
- "audio_ff_bias": true,
12
- "audio_gated_attn": true,
13
- "audio_hop_length": 160,
14
- "audio_in_channels": 128,
15
- "audio_num_attention_heads": 4,
16
- "audio_out_channels": 128,
17
- "audio_patch_size": 1,
18
- "audio_patch_size_t": 1,
19
- "audio_pos_embed_max_pos": 20,
20
- "audio_sampling_rate": 16000,
21
- "audio_scale_factor": 4,
22
- "base_height": 2048,
23
- "base_width": 2048,
24
- "caption_channels": 32,
25
- "causal_offset": 1,
26
- "cross_attention_dim": 32,
27
- "cross_attn_mod": true,
28
- "cross_attn_timestep_scale_multiplier": 1000,
29
- "ff_bias": false,
30
- "gated_attn": true,
31
- "in_channels": 8,
32
- "norm_elementwise_affine": false,
33
- "norm_eps": 1e-06,
34
- "num_attention_heads": 4,
35
- "num_layers": 2,
36
- "out_channels": 8,
37
- "patch_size": 1,
38
- "patch_size_t": 1,
39
- "perturbed_attn": true,
40
- "pos_embed_max_pos": 20,
41
- "qk_norm": "rms_norm_across_heads",
42
- "rope_double_precision": true,
43
- "rope_theta": 10000.0,
44
- "rope_type": "split",
45
- "timestep_scale_multiplier": 1000,
46
- "use_keyframes_abs_pos_embedding": true,
47
- "use_prompt_adaln_single": true,
48
- "use_prompt_embeddings": false,
49
- "vae_scale_factors": [
50
- 8,
51
- 32,
52
- 32
53
- ]
54
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
transformer/openvino_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3604ebaa7aecda36e94634278d55c671662f5c753ac11f335ea7ed6213572e6b
3
+ size 818832
transformer/openvino_model.xml ADDED
The diff for this file is too large to render. See raw diff
 
vae/config.json DELETED
@@ -1,366 +0,0 @@
1
- {
2
- "_class_name": "AutoencoderKLLTX2Video",
3
- "_diffusers_version": "0.40.0",
4
- "audio_latents_mean_data": [
5
- 0.0,
6
- 0.0,
7
- 0.0,
8
- 0.0,
9
- 0.0,
10
- 0.0,
11
- 0.0,
12
- 0.0,
13
- 0.0,
14
- 0.0,
15
- 0.0,
16
- 0.0,
17
- 0.0,
18
- 0.0,
19
- 0.0,
20
- 0.0,
21
- 0.0,
22
- 0.0,
23
- 0.0,
24
- 0.0,
25
- 0.0,
26
- 0.0,
27
- 0.0,
28
- 0.0,
29
- 0.0,
30
- 0.0,
31
- 0.0,
32
- 0.0,
33
- 0.0,
34
- 0.0,
35
- 0.0,
36
- 0.0,
37
- 0.0,
38
- 0.0,
39
- 0.0,
40
- 0.0,
41
- 0.0,
42
- 0.0,
43
- 0.0,
44
- 0.0,
45
- 0.0,
46
- 0.0,
47
- 0.0,
48
- 0.0,
49
- 0.0,
50
- 0.0,
51
- 0.0,
52
- 0.0,
53
- 0.0,
54
- 0.0,
55
- 0.0,
56
- 0.0,
57
- 0.0,
58
- 0.0,
59
- 0.0,
60
- 0.0,
61
- 0.0,
62
- 0.0,
63
- 0.0,
64
- 0.0,
65
- 0.0,
66
- 0.0,
67
- 0.0,
68
- 0.0,
69
- 0.0,
70
- 0.0,
71
- 0.0,
72
- 0.0,
73
- 0.0,
74
- 0.0,
75
- 0.0,
76
- 0.0,
77
- 0.0,
78
- 0.0,
79
- 0.0,
80
- 0.0,
81
- 0.0,
82
- 0.0,
83
- 0.0,
84
- 0.0,
85
- 0.0,
86
- 0.0,
87
- 0.0,
88
- 0.0,
89
- 0.0,
90
- 0.0,
91
- 0.0,
92
- 0.0,
93
- 0.0,
94
- 0.0,
95
- 0.0,
96
- 0.0,
97
- 0.0,
98
- 0.0,
99
- 0.0,
100
- 0.0,
101
- 0.0,
102
- 0.0,
103
- 0.0,
104
- 0.0,
105
- 0.0,
106
- 0.0,
107
- 0.0,
108
- 0.0,
109
- 0.0,
110
- 0.0,
111
- 0.0,
112
- 0.0,
113
- 0.0,
114
- 0.0,
115
- 0.0,
116
- 0.0,
117
- 0.0,
118
- 0.0,
119
- 0.0,
120
- 0.0,
121
- 0.0,
122
- 0.0,
123
- 0.0,
124
- 0.0,
125
- 0.0,
126
- 0.0,
127
- 0.0,
128
- 0.0,
129
- 0.0,
130
- 0.0,
131
- 0.0,
132
- 0.0
133
- ],
134
- "audio_latents_std_data": [
135
- 1.0,
136
- 1.0,
137
- 1.0,
138
- 1.0,
139
- 1.0,
140
- 1.0,
141
- 1.0,
142
- 1.0,
143
- 1.0,
144
- 1.0,
145
- 1.0,
146
- 1.0,
147
- 1.0,
148
- 1.0,
149
- 1.0,
150
- 1.0,
151
- 1.0,
152
- 1.0,
153
- 1.0,
154
- 1.0,
155
- 1.0,
156
- 1.0,
157
- 1.0,
158
- 1.0,
159
- 1.0,
160
- 1.0,
161
- 1.0,
162
- 1.0,
163
- 1.0,
164
- 1.0,
165
- 1.0,
166
- 1.0,
167
- 1.0,
168
- 1.0,
169
- 1.0,
170
- 1.0,
171
- 1.0,
172
- 1.0,
173
- 1.0,
174
- 1.0,
175
- 1.0,
176
- 1.0,
177
- 1.0,
178
- 1.0,
179
- 1.0,
180
- 1.0,
181
- 1.0,
182
- 1.0,
183
- 1.0,
184
- 1.0,
185
- 1.0,
186
- 1.0,
187
- 1.0,
188
- 1.0,
189
- 1.0,
190
- 1.0,
191
- 1.0,
192
- 1.0,
193
- 1.0,
194
- 1.0,
195
- 1.0,
196
- 1.0,
197
- 1.0,
198
- 1.0,
199
- 1.0,
200
- 1.0,
201
- 1.0,
202
- 1.0,
203
- 1.0,
204
- 1.0,
205
- 1.0,
206
- 1.0,
207
- 1.0,
208
- 1.0,
209
- 1.0,
210
- 1.0,
211
- 1.0,
212
- 1.0,
213
- 1.0,
214
- 1.0,
215
- 1.0,
216
- 1.0,
217
- 1.0,
218
- 1.0,
219
- 1.0,
220
- 1.0,
221
- 1.0,
222
- 1.0,
223
- 1.0,
224
- 1.0,
225
- 1.0,
226
- 1.0,
227
- 1.0,
228
- 1.0,
229
- 1.0,
230
- 1.0,
231
- 1.0,
232
- 1.0,
233
- 1.0,
234
- 1.0,
235
- 1.0,
236
- 1.0,
237
- 1.0,
238
- 1.0,
239
- 1.0,
240
- 1.0,
241
- 1.0,
242
- 1.0,
243
- 1.0,
244
- 1.0,
245
- 1.0,
246
- 1.0,
247
- 1.0,
248
- 1.0,
249
- 1.0,
250
- 1.0,
251
- 1.0,
252
- 1.0,
253
- 1.0,
254
- 1.0,
255
- 1.0,
256
- 1.0,
257
- 1.0,
258
- 1.0,
259
- 1.0,
260
- 1.0,
261
- 1.0,
262
- 1.0
263
- ],
264
- "block_out_channels": [
265
- 8,
266
- 8,
267
- 8,
268
- 8
269
- ],
270
- "decoder_block_out_channels": [
271
- 8,
272
- 16,
273
- 32
274
- ],
275
- "decoder_causal": false,
276
- "decoder_inject_noise": [
277
- false,
278
- false,
279
- false,
280
- false
281
- ],
282
- "decoder_layers_per_block": [
283
- 1,
284
- 1,
285
- 1,
286
- 1
287
- ],
288
- "decoder_spatial_padding_mode": "zeros",
289
- "decoder_spatio_temporal_scaling": [
290
- true,
291
- true,
292
- true
293
- ],
294
- "down_block_types": [
295
- "LTX2VideoDownBlock3D",
296
- "LTX2VideoDownBlock3D",
297
- "LTX2VideoDownBlock3D",
298
- "LTX2VideoDownBlock3D"
299
- ],
300
- "downsample_type": [
301
- "spatial",
302
- "temporal",
303
- "spatiotemporal",
304
- "spatiotemporal"
305
- ],
306
- "encoder_causal": true,
307
- "encoder_spatial_padding_mode": "zeros",
308
- "in_channels": 3,
309
- "latent_channels": 8,
310
- "latents_mean_data": [
311
- 0.0,
312
- 0.0,
313
- 0.0,
314
- 0.0,
315
- 0.0,
316
- 0.0,
317
- 0.0,
318
- 0.0
319
- ],
320
- "latents_std_data": [
321
- 1.0,
322
- 1.0,
323
- 1.0,
324
- 1.0,
325
- 1.0,
326
- 1.0,
327
- 1.0,
328
- 1.0
329
- ],
330
- "layers_per_block": [
331
- 1,
332
- 1,
333
- 1,
334
- 1,
335
- 1
336
- ],
337
- "out_channels": 3,
338
- "patch_size": 4,
339
- "patch_size_t": 1,
340
- "resnet_norm_eps": 1e-06,
341
- "scaling_factor": 1.0,
342
- "spatial_compression_ratio": 32,
343
- "spatio_temporal_scaling": [
344
- true,
345
- true,
346
- true,
347
- true
348
- ],
349
- "temporal_compression_ratio": 8,
350
- "timestep_conditioning": false,
351
- "upsample_factor": [
352
- 2,
353
- 2,
354
- 2
355
- ],
356
- "upsample_residual": [
357
- true,
358
- true,
359
- true
360
- ],
361
- "upsample_type": [
362
- "spatiotemporal",
363
- "spatiotemporal",
364
- "spatiotemporal"
365
- ]
366
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
vae/diffusion_pytorch_model.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:23aa7e896e48f467bd088b00d4a95cee1efba97483b5b8ace3210e4a23f7fe93
3
- size 1056860
 
 
 
 
vae_decoder/openvino_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7efa86ff89e0b5bcc28757545c9af3a4c44cab29df15a445844a020903bbd3e6
3
+ size 925008
vae_decoder/openvino_model.xml ADDED
The diff for this file is too large to render. See raw diff
 
vocoder/config.json DELETED
@@ -1,80 +0,0 @@
1
- {
2
- "_class_name": "LTX2VocoderWithBWE",
3
- "_diffusers_version": "0.40.0",
4
- "act_fn": "snakebeta",
5
- "antialias": true,
6
- "antialias_kernel_size": 12,
7
- "antialias_ratio": 2,
8
- "bwe_act_fn": "snakebeta",
9
- "bwe_antialias": true,
10
- "bwe_antialias_kernel_size": 12,
11
- "bwe_antialias_ratio": 2,
12
- "bwe_final_act_fn": null,
13
- "bwe_final_bias": false,
14
- "bwe_hidden_channels": 32,
15
- "bwe_in_channels": 128,
16
- "bwe_leaky_relu_negative_slope": 0.1,
17
- "bwe_out_channels": 2,
18
- "bwe_resnet_dilations": [
19
- [
20
- 1,
21
- 3,
22
- 5
23
- ]
24
- ],
25
- "bwe_resnet_kernel_sizes": [
26
- 3
27
- ],
28
- "bwe_upsample_factors": [
29
- 6,
30
- 5,
31
- 2,
32
- 2,
33
- 2
34
- ],
35
- "bwe_upsample_kernel_sizes": [
36
- 12,
37
- 11,
38
- 4,
39
- 4,
40
- 4
41
- ],
42
- "filter_length": 512,
43
- "final_act_fn": null,
44
- "final_bias": false,
45
- "hidden_channels": 64,
46
- "hop_length": 80,
47
- "in_channels": 128,
48
- "input_sampling_rate": 16000,
49
- "leaky_relu_negative_slope": 0.1,
50
- "num_mel_channels": 64,
51
- "out_channels": 2,
52
- "output_sampling_rate": 48000,
53
- "resnet_dilations": [
54
- [
55
- 1,
56
- 3,
57
- 5
58
- ]
59
- ],
60
- "resnet_kernel_sizes": [
61
- 3
62
- ],
63
- "upsample_factors": [
64
- 5,
65
- 2,
66
- 2,
67
- 2,
68
- 2,
69
- 2
70
- ],
71
- "upsample_kernel_sizes": [
72
- 11,
73
- 4,
74
- 4,
75
- 4,
76
- 4,
77
- 4
78
- ],
79
- "window_length": 512
80
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
vocoder/diffusion_pytorch_model.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:89b083230b9eaa8ce3dc583ce5806de3f86fa253ca5c1f33d17bc9942eb04bc7
3
- size 2829208
 
 
 
 
vocoder/openvino_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:31b8e7f10c415b664f5d3b5ccfd76c8a4ce38dbe5fe3363f873b2130c66b04fc
3
+ size 1727024
vocoder/openvino_model.xml ADDED
The diff for this file is too large to render. See raw diff