nodes/hao-ai-lab-FastVideo
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

[bugfix]: commit the environment-variable golden snapshots

Browse Source 
The golden snapshots of the environment-variable cases lived under
config_snapshot_goldens/env/, which the repository .gitignore entry "env"
ignores, so they were never committed and the five cases would fail in CI
with a missing golden file. The case category is renamed environment, and its
golden files, generated from the previous commit, are added.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012M91wnVFmPEvJ9h39r5BH7
This commit is contained in:
Davids048
2026-10-02 19:04:40 +00:00
co-authored by Claude Opus 5.5
parent 1d7e7e2e35
commit 867f960bf0
6 changed files with 1878 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
fastvideo/tests/api/config_snapshot.py
+3 -3
View File
@@ -11,7 +11,7 @@ and a ``SamplingParam`` where the case has a request) through one of the public
- ``kwargs``: ``VideoGenerator.from_pretrained`` keywords, through ``legacy_from_pretrained_to_config`` and the same
resolution.
- ``cli``: argparse flags, through ``FastVideoArgs.add_cli_args`` and ``FastVideoArgs.from_cli_args``.
- ``env``: environment variables that ``FastVideoArgs.__post_init__`` folds into fields.
- ``environment``: environment variables that ``FastVideoArgs.__post_init__`` folds into fields.
- ``request``: typed ``GenerationRequest`` values, through ``request_to_sampling_param``.
Every case runs with all registered environment variables unset (except the ones an ``env`` case sets) and with model
@@ -422,11 +422,11 @@ def collect_cases() -> list[SnapshotCase]:
},
}
cases += [
SnapshotCase("env", name, _env_case(env_values, model_path=WAN_T2V))
SnapshotCase("environment", name, _env_case(env_values, model_path=WAN_T2V))
for name, env_values in env_cases.items()
]
cases.append(
SnapshotCase("env", "explicit_argument_wins_over_attention_backend",
SnapshotCase("environment", "explicit_argument_wins_over_attention_backend",
_env_case({"FASTVIDEO_ATTENTION_BACKEND": "TORCH_SDPA"},
model_path=WAN_T2V,
attention_backend="FLASH_ATTN")))
fastvideo/tests/api/config_snapshot_goldens/environment/attention_backend.json
+375
View File
@@ -0,0 +1,375 @@
{
"fastvideo_args": {
"VSA_sparsity": 0.0,
"VSA_tile_size": 256,
"attention_backend": "TORCH_SDPA",
"boundary_ratio": 0.875,
"disable_autocast": false,
"dist_timeout": null,
"distributed_executor_backend": "mp",
"dit_cpu_offload": true,
"dit_layerwise_offload": true,
"enable_stage_verification": true,
"enable_torch_compile": false,
"enable_torch_compile_audio_vae": false,
"enable_torch_compile_text_encoder": false,
"enable_torch_compile_vae": false,
"h3_sequential_load": null,
"hsdp_replicate_dim": 1,
"hsdp_shard_dim": 1,
"image_encoder_cpu_offload": true,
"inference_mode": true,
"inference_torch_compile": false,
"init_weights_from_safetensors": "",
"init_weights_from_safetensors_2": "",
"lazy_module_load": null,
"lora_nickname": "default",
"lora_path": null,
"lora_strength": 1.0,
"lora_target_modules": null,
"ltx2_audio_latent_path": null,
"ltx2_initial_latent_path": null,
"ltx2_legacy_native_noise_order": false,
"ltx2_refine_add_noise": true,
"ltx2_refine_audio_noise_path": null,
"ltx2_refine_enabled": false,
"ltx2_refine_guidance_scale": 1.0,
"ltx2_refine_lora_path": null,
"ltx2_refine_noise_path": null,
"ltx2_refine_num_inference_steps": 3,
"ltx2_refine_transformer_path": null,
"ltx2_refine_upsampler_path": null,
"ltx2_use_distilled_sigmas": true,
"ltx2_vae_spatial_tile_overlap_in_pixels": null,
"ltx2_vae_spatial_tile_size_in_pixels": null,
"ltx2_vae_temporal_tile_overlap_in_frames": null,
"ltx2_vae_temporal_tile_size_in_frames": null,
"ltx2_vae_tiling": null,
"master_port": null,
"moba_config": {},
"moba_config_path": null,
"mode": "inference",
"model_loaded": {
"transformer": true,
"upsampler": true,
"vae": true
},
"model_path": "Wan-AI/Wan2.1-T2V-1.3B-Diffusers",
"model_paths": {},
"num_gpus": 1,
"output_type": "pil",
"override_pipeline_cls_name": null,
"override_text_encoder_quant": null,
"override_text_encoder_safetensors": null,
"override_transformer_cls_name": null,
"pin_cpu_memory": true,
"preprocess_config": null,
"prompt_txt": null,
"ray_placement_group": null,
"ray_runtime_env": null,
"refine_add_noise": null,
"refine_audio_noise_path": null,
"refine_enabled": null,
"refine_guidance_scale": null,
"refine_lora_path": null,
"refine_noise_path": null,
"refine_num_inference_steps": null,
"refine_transformer_path": null,
"refine_upsampler_path": null,
"revision": null,
"sp_size": 1,
"taeh3_checkpoint": null,
"taeh3_chunk_size": 5,
"text_encoder_cpu_offload": true,
"torch_compile_kwargs": {},
"torch_compile_kwargs_audio_vae": {},
"torch_compile_kwargs_dit": {},
"torch_compile_kwargs_text_encoder": {},
"torch_compile_kwargs_vae": {},
"tp_size": 1,
"transformer_quant": null,
"trust_remote_code": false,
"use_fsdp_inference": false,
"vae_cpu_offload": true,
"vae_parallel_decode": false,
"vae_parallel_decode_strategy": "gather",
"vae_parallel_encode": false,
"video_decode_backend": "h3-vae",
"workload_type": "t2v"
},
"pipeline_config": {
"boundary_ratio": null,
"disable_autocast": false,
"dit_config": {
"_resolved_attention_backend": null,
"arch_config": {
"_supported_attention_backends": [
3,
1,
2,
7,
10,
4,
5,
6,
11,
12
],
"added_kv_proj_dim": null,
"attention_head_dim": 128,
"boundary_ratio": null,
"cast_prompt_embeds_to_dit_dtype": false,
"cross_attn_norm": true,
"eps": 1e-06,
"exclude_lora_layers": [
"embedder"
],
"ffn_dim": 13824,
"freq_dim": 256,
"hidden_size": 5120,
"image_dim": null,
"in_channels": 16,
"local_attn_size": -1,
"num_attention_heads": 40,
"num_channels_latents": 16,
"num_frames_per_block": 3,
"num_layers": 40,
"out_channels": 16,
"patch_size": [
1,
2,
2
],
"pos_embed_seq_len": null,
"qk_norm": "rms_norm_across_heads",
"r_embedder": false,
"r_embedder_deltatime_type": "r",
"r_embedder_fusion": "additive",
"r_embedder_gate_value": 0.25,
"rope_cache_policy": "absolute",
"rope_max_seq_len": 1024,
"sink_size": 0,
"sliding_window_num_frames": 21,
"stacked_params_mapping": [],
"text_dim": 4096
},
"prefix": "Wan",
"quant_config": null
},
"dit_precision": "bf16",
"dmd_denoising_steps": null,
"embedded_cfg_scale": 6.0,
"flow_shift": 3.0,
"flow_shift_sr": null,
"image_encoder_config": {
"_resolved_attention_backend": null,
"arch_config": {
"_supported_attention_backends": [
1,
2
],
"architectures": [],
"output_hidden_states": false,
"stacked_params_mapping": [],
"use_return_dict": true
},
"lora_config": null,
"prefix": "",
"quant_config": null
},
"image_encoder_configs": null,
"image_encoder_precision": "fp32",
"image_encoder_precisions": null,
"is_causal": false,
"lucy_edit_task": false,
"model_path": "Wan-AI/Wan2.1-T2V-1.3B-Diffusers",
"pipeline_config_path": null,
"postprocess_text_funcs": [
"fastvideo.models.wan.pipeline_config.t5_postprocess_text"
],
"precision": "bf16",
"preprocess_text_funcs": [
"fastvideo.configs.pipelines.base.preprocess_text"
],
"scheduler_step_in_fp32": false,
"text_encoder_configs": [
{
"_resolved_attention_backend": null,
"arch_config": {
"_supported_attention_backends": [
1,
2
],
"architectures": [],
"classifier_dropout": 0.0,
"d_ff": 2048,
"d_kv": 64,
"d_model": 512,
"decoder_start_token_id": 0,
"dense_act_fn": "relu",
"dropout_rate": 0.1,
"dtype": null,
"eos_token_id": 1,
"feed_forward_proj": "relu",
"gradient_checkpointing": false,
"hidden_size": 512,
"hidden_state_skip_layer": 0,
"initializer_factor": 1.0,
"is_encoder_decoder": true,
"is_gated_act": false,
"layer_norm_epsilon": 1e-06,
"n_positions": 512,
"num_attention_heads": 0,
"num_decoder_layers": null,
"num_heads": 8,
"num_hidden_layers": 0,
"num_layers": 6,
"output_hidden_states": false,
"output_past": true,
"pad_token_id": 0,
"relative_attention_max_distance": 128,
"relative_attention_num_buckets": 32,
"require_processor": false,
"scalable_attention": true,
"stacked_params_mapping": [
[
".qkv_proj",
".q",
"q"
],
[
".qkv_proj",
".k",
"k"
],
[
".qkv_proj",
".v",
"v"
]
],
"task_specific_params": null,
"text_len": 512,
"tie_word_embeddings": false,
"tokenizer_kwargs": {
"add_special_tokens": true,
"max_length": 512,
"return_attention_mask": true,
"return_tensors": "pt",
"truncation": true
},
"use_cache": true,
"use_return_dict": true,
"vocab_size": 32128
},
"chat_template_enable_thinking": false,
"is_chat_model": false,
"lora_config": null,
"prefix": "t5",
"quant_config": null,
"treat_empty_as_dot": false
}
],
"text_encoder_precisions": [
"fp32"
],
"ti2v_task": false,
"upsampler_config": {
"_resolved_attention_backend": null,
"arch_config": {
"stacked_params_mapping": []
}
},
"upsampler_precision": "fp32",
"vae_config": {
"_resolved_attention_backend": null,
"arch_config": {
"attn_scales": [],
"base_dim": 96,
"clip_output": true,
"decoder_base_dim": null,
"dim_mult": [
1,
2,
4,
4
],
"dropout": 0.0,
"in_channels": 3,
"is_residual": false,
"latents_mean": [
-0.7571,
-0.7089,
-0.9113,
0.1075,
-0.1745,
0.9653,
-0.1517,
1.5508,
0.4134,
-0.0715,
0.5517,
-0.3632,
-0.1922,
-0.9497,
0.2503,
-0.2921
],
"latents_std": [
2.8184,
1.4541,
2.3275,
2.6558,
1.2196,
1.7708,
2.6052,
2.0743,
3.2687,
2.1526,
2.8652,
1.5579,
1.6382,
1.1253,
2.8251,
1.916
],
"num_res_blocks": 2,
"out_channels": 3,
"patch_size": null,
"scale_factor_spatial": 8,
"scale_factor_temporal": 4,
"scaling_factor": "<tensor shape=[1, 16, 1, 1, 1] dtype=torch.float32>",
"spatial_compression_ratio": 8,
"stacked_params_mapping": [],
"temperal_downsample": [
false,
true,
true
],
"temporal_compression_ratio": 4,
"z_dim": 16
},
"blend_num_frames": 8,
"load_decoder": true,
"load_encoder": false,
"tile_sample_min_height": 256,
"tile_sample_min_num_frames": 16,
"tile_sample_min_width": 256,
"tile_sample_stride_height": 192,
"tile_sample_stride_num_frames": 12,
"tile_sample_stride_width": 192,
"use_feature_cache": true,
"use_light_vae": false,
"use_parallel_tiling": false,
"use_temporal_scaling_frames": true,
"use_temporal_tiling": false,
"use_tiling": false
},
"vae_decode_precision": "bf16",
"vae_precision": "fp32",
"vae_sp": false,
"vae_tiling": false,
"warp_denoising_step": true
},
"pipeline_config_class": "fastvideo.models.wan.pipeline_config.WanT2V480PConfig"
}
fastvideo/tests/api/config_snapshot_goldens/environment/attention_backend_unknown_name.json
+375
View File
@@ -0,0 +1,375 @@
{
"fastvideo_args": {
"VSA_sparsity": 0.0,
"VSA_tile_size": 256,
"attention_backend": null,
"boundary_ratio": 0.875,
"disable_autocast": false,
"dist_timeout": null,
"distributed_executor_backend": "mp",
"dit_cpu_offload": true,
"dit_layerwise_offload": true,
"enable_stage_verification": true,
"enable_torch_compile": false,
"enable_torch_compile_audio_vae": false,
"enable_torch_compile_text_encoder": false,
"enable_torch_compile_vae": false,
"h3_sequential_load": null,
"hsdp_replicate_dim": 1,
"hsdp_shard_dim": 1,
"image_encoder_cpu_offload": true,
"inference_mode": true,
"inference_torch_compile": false,
"init_weights_from_safetensors": "",
"init_weights_from_safetensors_2": "",
"lazy_module_load": null,
"lora_nickname": "default",
"lora_path": null,
"lora_strength": 1.0,
"lora_target_modules": null,
"ltx2_audio_latent_path": null,
"ltx2_initial_latent_path": null,
"ltx2_legacy_native_noise_order": false,
"ltx2_refine_add_noise": true,
"ltx2_refine_audio_noise_path": null,
"ltx2_refine_enabled": false,
"ltx2_refine_guidance_scale": 1.0,
"ltx2_refine_lora_path": null,
"ltx2_refine_noise_path": null,
"ltx2_refine_num_inference_steps": 3,
"ltx2_refine_transformer_path": null,
"ltx2_refine_upsampler_path": null,
"ltx2_use_distilled_sigmas": true,
"ltx2_vae_spatial_tile_overlap_in_pixels": null,
"ltx2_vae_spatial_tile_size_in_pixels": null,
"ltx2_vae_temporal_tile_overlap_in_frames": null,
"ltx2_vae_temporal_tile_size_in_frames": null,
"ltx2_vae_tiling": null,
"master_port": null,
"moba_config": {},
"moba_config_path": null,
"mode": "inference",
"model_loaded": {
"transformer": true,
"upsampler": true,
"vae": true
},
"model_path": "Wan-AI/Wan2.1-T2V-1.3B-Diffusers",
"model_paths": {},
"num_gpus": 1,
"output_type": "pil",
"override_pipeline_cls_name": null,
"override_text_encoder_quant": null,
"override_text_encoder_safetensors": null,
"override_transformer_cls_name": null,
"pin_cpu_memory": true,
"preprocess_config": null,
"prompt_txt": null,
"ray_placement_group": null,
"ray_runtime_env": null,
"refine_add_noise": null,
"refine_audio_noise_path": null,
"refine_enabled": null,
"refine_guidance_scale": null,
"refine_lora_path": null,
"refine_noise_path": null,
"refine_num_inference_steps": null,
"refine_transformer_path": null,
"refine_upsampler_path": null,
"revision": null,
"sp_size": 1,
"taeh3_checkpoint": null,
"taeh3_chunk_size": 5,
"text_encoder_cpu_offload": true,
"torch_compile_kwargs": {},
"torch_compile_kwargs_audio_vae": {},
"torch_compile_kwargs_dit": {},
"torch_compile_kwargs_text_encoder": {},
"torch_compile_kwargs_vae": {},
"tp_size": 1,
"transformer_quant": null,
"trust_remote_code": false,
"use_fsdp_inference": false,
"vae_cpu_offload": true,
"vae_parallel_decode": false,
"vae_parallel_decode_strategy": "gather",
"vae_parallel_encode": false,
"video_decode_backend": "h3-vae",
"workload_type": "t2v"
},
"pipeline_config": {
"boundary_ratio": null,
"disable_autocast": false,
"dit_config": {
"_resolved_attention_backend": null,
"arch_config": {
"_supported_attention_backends": [
3,
1,
2,
7,
10,
4,
5,
6,
11,
12
],
"added_kv_proj_dim": null,
"attention_head_dim": 128,
"boundary_ratio": null,
"cast_prompt_embeds_to_dit_dtype": false,
"cross_attn_norm": true,
"eps": 1e-06,
"exclude_lora_layers": [
"embedder"
],
"ffn_dim": 13824,
"freq_dim": 256,
"hidden_size": 5120,
"image_dim": null,
"in_channels": 16,
"local_attn_size": -1,
"num_attention_heads": 40,
"num_channels_latents": 16,
"num_frames_per_block": 3,
"num_layers": 40,
"out_channels": 16,
"patch_size": [
1,
2,
2
],
"pos_embed_seq_len": null,
"qk_norm": "rms_norm_across_heads",
"r_embedder": false,
"r_embedder_deltatime_type": "r",
"r_embedder_fusion": "additive",
"r_embedder_gate_value": 0.25,
"rope_cache_policy": "absolute",
"rope_max_seq_len": 1024,
"sink_size": 0,
"sliding_window_num_frames": 21,
"stacked_params_mapping": [],
"text_dim": 4096
},
"prefix": "Wan",
"quant_config": null
},
"dit_precision": "bf16",
"dmd_denoising_steps": null,
"embedded_cfg_scale": 6.0,
"flow_shift": 3.0,
"flow_shift_sr": null,
"image_encoder_config": {
"_resolved_attention_backend": null,
"arch_config": {
"_supported_attention_backends": [
1,
2
],
"architectures": [],
"output_hidden_states": false,
"stacked_params_mapping": [],
"use_return_dict": true
},
"lora_config": null,
"prefix": "",
"quant_config": null
},
"image_encoder_configs": null,
"image_encoder_precision": "fp32",
"image_encoder_precisions": null,
"is_causal": false,
"lucy_edit_task": false,
"model_path": "Wan-AI/Wan2.1-T2V-1.3B-Diffusers",
"pipeline_config_path": null,
"postprocess_text_funcs": [
"fastvideo.models.wan.pipeline_config.t5_postprocess_text"
],
"precision": "bf16",
"preprocess_text_funcs": [
"fastvideo.configs.pipelines.base.preprocess_text"
],
"scheduler_step_in_fp32": false,
"text_encoder_configs": [
{
"_resolved_attention_backend": null,
"arch_config": {
"_supported_attention_backends": [
1,
2
],
"architectures": [],
"classifier_dropout": 0.0,
"d_ff": 2048,
"d_kv": 64,
"d_model": 512,
"decoder_start_token_id": 0,
"dense_act_fn": "relu",
"dropout_rate": 0.1,
"dtype": null,
"eos_token_id": 1,
"feed_forward_proj": "relu",
"gradient_checkpointing": false,
"hidden_size": 512,
"hidden_state_skip_layer": 0,
"initializer_factor": 1.0,
"is_encoder_decoder": true,
"is_gated_act": false,
"layer_norm_epsilon": 1e-06,
"n_positions": 512,
"num_attention_heads": 0,
"num_decoder_layers": null,
"num_heads": 8,
"num_hidden_layers": 0,
"num_layers": 6,
"output_hidden_states": false,
"output_past": true,
"pad_token_id": 0,
"relative_attention_max_distance": 128,
"relative_attention_num_buckets": 32,
"require_processor": false,
"scalable_attention": true,
"stacked_params_mapping": [
[
".qkv_proj",
".q",
"q"
],
[
".qkv_proj",
".k",
"k"
],
[
".qkv_proj",
".v",
"v"
]
],
"task_specific_params": null,
"text_len": 512,
"tie_word_embeddings": false,
"tokenizer_kwargs": {
"add_special_tokens": true,
"max_length": 512,
"return_attention_mask": true,
"return_tensors": "pt",
"truncation": true
},
"use_cache": true,
"use_return_dict": true,
"vocab_size": 32128
},
"chat_template_enable_thinking": false,
"is_chat_model": false,
"lora_config": null,
"prefix": "t5",
"quant_config": null,
"treat_empty_as_dot": false
}
],
"text_encoder_precisions": [
"fp32"
],
"ti2v_task": false,
"upsampler_config": {
"_resolved_attention_backend": null,
"arch_config": {
"stacked_params_mapping": []
}
},
"upsampler_precision": "fp32",
"vae_config": {
"_resolved_attention_backend": null,
"arch_config": {
"attn_scales": [],
"base_dim": 96,
"clip_output": true,
"decoder_base_dim": null,
"dim_mult": [
1,
2,
4,
4
],
"dropout": 0.0,
"in_channels": 3,
"is_residual": false,
"latents_mean": [
-0.7571,
-0.7089,
-0.9113,
0.1075,
-0.1745,
0.9653,
-0.1517,
1.5508,
0.4134,
-0.0715,
0.5517,
-0.3632,
-0.1922,
-0.9497,
0.2503,
-0.2921
],
"latents_std": [
2.8184,
1.4541,
2.3275,
2.6558,
1.2196,
1.7708,
2.6052,
2.0743,
3.2687,
2.1526,
2.8652,
1.5579,
1.6382,
1.1253,
2.8251,
1.916
],
"num_res_blocks": 2,
"out_channels": 3,
"patch_size": null,
"scale_factor_spatial": 8,
"scale_factor_temporal": 4,
"scaling_factor": "<tensor shape=[1, 16, 1, 1, 1] dtype=torch.float32>",
"spatial_compression_ratio": 8,
"stacked_params_mapping": [],
"temperal_downsample": [
false,
true,
true
],
"temporal_compression_ratio": 4,
"z_dim": 16
},
"blend_num_frames": 8,
"load_decoder": true,
"load_encoder": false,
"tile_sample_min_height": 256,
"tile_sample_min_num_frames": 16,
"tile_sample_min_width": 256,
"tile_sample_stride_height": 192,
"tile_sample_stride_num_frames": 12,
"tile_sample_stride_width": 192,
"use_feature_cache": true,
"use_light_vae": false,
"use_parallel_tiling": false,
"use_temporal_scaling_frames": true,
"use_temporal_tiling": false,
"use_tiling": false
},
"vae_decode_precision": "bf16",
"vae_precision": "fp32",
"vae_sp": false,
"vae_tiling": false,
"warp_denoising_step": true
},
"pipeline_config_class": "fastvideo.models.wan.pipeline_config.WanT2V480PConfig"
}
fastvideo/tests/api/config_snapshot_goldens/environment/explicit_argument_wins_over_attention_backend.json
+375
View File
@@ -0,0 +1,375 @@
{
"fastvideo_args": {
"VSA_sparsity": 0.0,
"VSA_tile_size": 256,
"attention_backend": "FLASH_ATTN",
"boundary_ratio": 0.875,
"disable_autocast": false,
"dist_timeout": null,
"distributed_executor_backend": "mp",
"dit_cpu_offload": true,
"dit_layerwise_offload": true,
"enable_stage_verification": true,
"enable_torch_compile": false,
"enable_torch_compile_audio_vae": false,
"enable_torch_compile_text_encoder": false,
"enable_torch_compile_vae": false,
"h3_sequential_load": null,
"hsdp_replicate_dim": 1,
"hsdp_shard_dim": 1,
"image_encoder_cpu_offload": true,
"inference_mode": true,
"inference_torch_compile": false,
"init_weights_from_safetensors": "",
"init_weights_from_safetensors_2": "",
"lazy_module_load": null,
"lora_nickname": "default",
"lora_path": null,
"lora_strength": 1.0,
"lora_target_modules": null,
"ltx2_audio_latent_path": null,
"ltx2_initial_latent_path": null,
"ltx2_legacy_native_noise_order": false,
"ltx2_refine_add_noise": true,
"ltx2_refine_audio_noise_path": null,
"ltx2_refine_enabled": false,
"ltx2_refine_guidance_scale": 1.0,
"ltx2_refine_lora_path": null,
"ltx2_refine_noise_path": null,
"ltx2_refine_num_inference_steps": 3,
"ltx2_refine_transformer_path": null,
"ltx2_refine_upsampler_path": null,
"ltx2_use_distilled_sigmas": true,
"ltx2_vae_spatial_tile_overlap_in_pixels": null,
"ltx2_vae_spatial_tile_size_in_pixels": null,
"ltx2_vae_temporal_tile_overlap_in_frames": null,
"ltx2_vae_temporal_tile_size_in_frames": null,
"ltx2_vae_tiling": null,
"master_port": null,
"moba_config": {},
"moba_config_path": null,
"mode": "inference",
"model_loaded": {
"transformer": true,
"upsampler": true,
"vae": true
},
"model_path": "Wan-AI/Wan2.1-T2V-1.3B-Diffusers",
"model_paths": {},
"num_gpus": 1,
"output_type": "pil",
"override_pipeline_cls_name": null,
"override_text_encoder_quant": null,
"override_text_encoder_safetensors": null,
"override_transformer_cls_name": null,
"pin_cpu_memory": true,
"preprocess_config": null,
"prompt_txt": null,
"ray_placement_group": null,
"ray_runtime_env": null,
"refine_add_noise": null,
"refine_audio_noise_path": null,
"refine_enabled": null,
"refine_guidance_scale": null,
"refine_lora_path": null,
"refine_noise_path": null,
"refine_num_inference_steps": null,
"refine_transformer_path": null,
"refine_upsampler_path": null,
"revision": null,
"sp_size": 1,
"taeh3_checkpoint": null,
"taeh3_chunk_size": 5,
"text_encoder_cpu_offload": true,
"torch_compile_kwargs": {},
"torch_compile_kwargs_audio_vae": {},
"torch_compile_kwargs_dit": {},
"torch_compile_kwargs_text_encoder": {},
"torch_compile_kwargs_vae": {},
"tp_size": 1,
"transformer_quant": null,
"trust_remote_code": false,
"use_fsdp_inference": false,
"vae_cpu_offload": true,
"vae_parallel_decode": false,
"vae_parallel_decode_strategy": "gather",
"vae_parallel_encode": false,
"video_decode_backend": "h3-vae",
"workload_type": "t2v"
},
"pipeline_config": {
"boundary_ratio": null,
"disable_autocast": false,
"dit_config": {
"_resolved_attention_backend": null,
"arch_config": {
"_supported_attention_backends": [
3,
1,
2,
7,
10,
4,
5,
6,
11,
12
],
"added_kv_proj_dim": null,
"attention_head_dim": 128,
"boundary_ratio": null,
"cast_prompt_embeds_to_dit_dtype": false,
"cross_attn_norm": true,
"eps": 1e-06,
"exclude_lora_layers": [
"embedder"
],
"ffn_dim": 13824,
"freq_dim": 256,
"hidden_size": 5120,
"image_dim": null,
"in_channels": 16,
"local_attn_size": -1,
"num_attention_heads": 40,
"num_channels_latents": 16,
"num_frames_per_block": 3,
"num_layers": 40,
"out_channels": 16,
"patch_size": [
1,
2,
2
],
"pos_embed_seq_len": null,
"qk_norm": "rms_norm_across_heads",
"r_embedder": false,
"r_embedder_deltatime_type": "r",
"r_embedder_fusion": "additive",
"r_embedder_gate_value": 0.25,
"rope_cache_policy": "absolute",
"rope_max_seq_len": 1024,
"sink_size": 0,
"sliding_window_num_frames": 21,
"stacked_params_mapping": [],
"text_dim": 4096
},
"prefix": "Wan",
"quant_config": null
},
"dit_precision": "bf16",
"dmd_denoising_steps": null,
"embedded_cfg_scale": 6.0,
"flow_shift": 3.0,
"flow_shift_sr": null,
"image_encoder_config": {
"_resolved_attention_backend": null,
"arch_config": {
"_supported_attention_backends": [
1,
2
],
"architectures": [],
"output_hidden_states": false,
"stacked_params_mapping": [],
"use_return_dict": true
},
"lora_config": null,
"prefix": "",
"quant_config": null
},
"image_encoder_configs": null,
"image_encoder_precision": "fp32",
"image_encoder_precisions": null,
"is_causal": false,
"lucy_edit_task": false,
"model_path": "Wan-AI/Wan2.1-T2V-1.3B-Diffusers",
"pipeline_config_path": null,
"postprocess_text_funcs": [
"fastvideo.models.wan.pipeline_config.t5_postprocess_text"
],
"precision": "bf16",
"preprocess_text_funcs": [
"fastvideo.configs.pipelines.base.preprocess_text"
],
"scheduler_step_in_fp32": false,
"text_encoder_configs": [
{
"_resolved_attention_backend": null,
"arch_config": {
"_supported_attention_backends": [
1,
2
],
"architectures": [],
"classifier_dropout": 0.0,
"d_ff": 2048,
"d_kv": 64,
"d_model": 512,
"decoder_start_token_id": 0,
"dense_act_fn": "relu",
"dropout_rate": 0.1,
"dtype": null,
"eos_token_id": 1,
"feed_forward_proj": "relu",
"gradient_checkpointing": false,
"hidden_size": 512,
"hidden_state_skip_layer": 0,
"initializer_factor": 1.0,
"is_encoder_decoder": true,
"is_gated_act": false,
"layer_norm_epsilon": 1e-06,
"n_positions": 512,
"num_attention_heads": 0,
"num_decoder_layers": null,
"num_heads": 8,
"num_hidden_layers": 0,
"num_layers": 6,
"output_hidden_states": false,
"output_past": true,
"pad_token_id": 0,
"relative_attention_max_distance": 128,
"relative_attention_num_buckets": 32,
"require_processor": false,
"scalable_attention": true,
"stacked_params_mapping": [
[
".qkv_proj",
".q",
"q"
],
[
".qkv_proj",
".k",
"k"
],
[
".qkv_proj",
".v",
"v"
]
],
"task_specific_params": null,
"text_len": 512,
"tie_word_embeddings": false,
"tokenizer_kwargs": {
"add_special_tokens": true,
"max_length": 512,
"return_attention_mask": true,
"return_tensors": "pt",
"truncation": true
},
"use_cache": true,
"use_return_dict": true,
"vocab_size": 32128
},
"chat_template_enable_thinking": false,
"is_chat_model": false,
"lora_config": null,
"prefix": "t5",
"quant_config": null,
"treat_empty_as_dot": false
}
],
"text_encoder_precisions": [
"fp32"
],
"ti2v_task": false,
"upsampler_config": {
"_resolved_attention_backend": null,
"arch_config": {
"stacked_params_mapping": []
}
},
"upsampler_precision": "fp32",
"vae_config": {
"_resolved_attention_backend": null,
"arch_config": {
"attn_scales": [],
"base_dim": 96,
"clip_output": true,
"decoder_base_dim": null,
"dim_mult": [
1,
2,
4,
4
],
"dropout": 0.0,
"in_channels": 3,
"is_residual": false,
"latents_mean": [
-0.7571,
-0.7089,
-0.9113,
0.1075,
-0.1745,
0.9653,
-0.1517,
1.5508,
0.4134,
-0.0715,
0.5517,
-0.3632,
-0.1922,
-0.9497,
0.2503,
-0.2921
],
"latents_std": [
2.8184,
1.4541,
2.3275,
2.6558,
1.2196,
1.7708,
2.6052,
2.0743,
3.2687,
2.1526,
2.8652,
1.5579,
1.6382,
1.1253,
2.8251,
1.916
],
"num_res_blocks": 2,
"out_channels": 3,
"patch_size": null,
"scale_factor_spatial": 8,
"scale_factor_temporal": 4,
"scaling_factor": "<tensor shape=[1, 16, 1, 1, 1] dtype=torch.float32>",
"spatial_compression_ratio": 8,
"stacked_params_mapping": [],
"temperal_downsample": [
false,
true,
true
],
"temporal_compression_ratio": 4,
"z_dim": 16
},
"blend_num_frames": 8,
"load_decoder": true,
"load_encoder": false,
"tile_sample_min_height": 256,
"tile_sample_min_num_frames": 16,
"tile_sample_min_width": 256,
"tile_sample_stride_height": 192,
"tile_sample_stride_num_frames": 12,
"tile_sample_stride_width": 192,
"use_feature_cache": true,
"use_light_vae": false,
"use_parallel_tiling": false,
"use_temporal_scaling_frames": true,
"use_temporal_tiling": false,
"use_tiling": false
},
"vae_decode_precision": "bf16",
"vae_precision": "fp32",
"vae_sp": false,
"vae_tiling": false,
"warp_denoising_step": true
},
"pipeline_config_class": "fastvideo.models.wan.pipeline_config.WanT2V480PConfig"
}
fastvideo/tests/api/config_snapshot_goldens/environment/inference_torch_compile.json
+375
View File
@@ -0,0 +1,375 @@
{
"fastvideo_args": {
"VSA_sparsity": 0.0,
"VSA_tile_size": 256,
"attention_backend": null,
"boundary_ratio": 0.875,
"disable_autocast": false,
"dist_timeout": null,
"distributed_executor_backend": "mp",
"dit_cpu_offload": true,
"dit_layerwise_offload": true,
"enable_stage_verification": true,
"enable_torch_compile": false,
"enable_torch_compile_audio_vae": false,
"enable_torch_compile_text_encoder": false,
"enable_torch_compile_vae": false,
"h3_sequential_load": null,
"hsdp_replicate_dim": 1,
"hsdp_shard_dim": 1,
"image_encoder_cpu_offload": true,
"inference_mode": true,
"inference_torch_compile": true,
"init_weights_from_safetensors": "",
"init_weights_from_safetensors_2": "",
"lazy_module_load": null,
"lora_nickname": "default",
"lora_path": null,
"lora_strength": 1.0,
"lora_target_modules": null,
"ltx2_audio_latent_path": null,
"ltx2_initial_latent_path": null,
"ltx2_legacy_native_noise_order": false,
"ltx2_refine_add_noise": true,
"ltx2_refine_audio_noise_path": null,
"ltx2_refine_enabled": false,
"ltx2_refine_guidance_scale": 1.0,
"ltx2_refine_lora_path": null,
"ltx2_refine_noise_path": null,
"ltx2_refine_num_inference_steps": 3,
"ltx2_refine_transformer_path": null,
"ltx2_refine_upsampler_path": null,
"ltx2_use_distilled_sigmas": true,
"ltx2_vae_spatial_tile_overlap_in_pixels": null,
"ltx2_vae_spatial_tile_size_in_pixels": null,
"ltx2_vae_temporal_tile_overlap_in_frames": null,
"ltx2_vae_temporal_tile_size_in_frames": null,
"ltx2_vae_tiling": null,
"master_port": null,
"moba_config": {},
"moba_config_path": null,
"mode": "inference",
"model_loaded": {
"transformer": true,
"upsampler": true,
"vae": true
},
"model_path": "Wan-AI/Wan2.1-T2V-1.3B-Diffusers",
"model_paths": {},
"num_gpus": 1,
"output_type": "pil",
"override_pipeline_cls_name": null,
"override_text_encoder_quant": null,
"override_text_encoder_safetensors": null,
"override_transformer_cls_name": null,
"pin_cpu_memory": true,
"preprocess_config": null,
"prompt_txt": null,
"ray_placement_group": null,
"ray_runtime_env": null,
"refine_add_noise": null,
"refine_audio_noise_path": null,
"refine_enabled": null,
"refine_guidance_scale": null,
"refine_lora_path": null,
"refine_noise_path": null,
"refine_num_inference_steps": null,
"refine_transformer_path": null,
"refine_upsampler_path": null,
"revision": null,
"sp_size": 1,
"taeh3_checkpoint": null,
"taeh3_chunk_size": 5,
"text_encoder_cpu_offload": true,
"torch_compile_kwargs": {},
"torch_compile_kwargs_audio_vae": {},
"torch_compile_kwargs_dit": {},
"torch_compile_kwargs_text_encoder": {},
"torch_compile_kwargs_vae": {},
"tp_size": 1,
"transformer_quant": null,
"trust_remote_code": false,
"use_fsdp_inference": false,
"vae_cpu_offload": true,
"vae_parallel_decode": false,
"vae_parallel_decode_strategy": "gather",
"vae_parallel_encode": false,
"video_decode_backend": "h3-vae",
"workload_type": "t2v"
},
"pipeline_config": {
"boundary_ratio": null,
"disable_autocast": false,
"dit_config": {
"_resolved_attention_backend": null,
"arch_config": {
"_supported_attention_backends": [
3,
1,
2,
7,
10,
4,
5,
6,
11,
12
],
"added_kv_proj_dim": null,
"attention_head_dim": 128,
"boundary_ratio": null,
"cast_prompt_embeds_to_dit_dtype": false,
"cross_attn_norm": true,
"eps": 1e-06,
"exclude_lora_layers": [
"embedder"
],
"ffn_dim": 13824,
"freq_dim": 256,
"hidden_size": 5120,
"image_dim": null,
"in_channels": 16,
"local_attn_size": -1,
"num_attention_heads": 40,
"num_channels_latents": 16,
"num_frames_per_block": 3,
"num_layers": 40,
"out_channels": 16,
"patch_size": [
1,
2,
2
],
"pos_embed_seq_len": null,
"qk_norm": "rms_norm_across_heads",
"r_embedder": false,
"r_embedder_deltatime_type": "r",
"r_embedder_fusion": "additive",
"r_embedder_gate_value": 0.25,
"rope_cache_policy": "absolute",
"rope_max_seq_len": 1024,
"sink_size": 0,
"sliding_window_num_frames": 21,
"stacked_params_mapping": [],
"text_dim": 4096
},
"prefix": "Wan",
"quant_config": null
},
"dit_precision": "bf16",
"dmd_denoising_steps": null,
"embedded_cfg_scale": 6.0,
"flow_shift": 3.0,
"flow_shift_sr": null,
"image_encoder_config": {
"_resolved_attention_backend": null,
"arch_config": {
"_supported_attention_backends": [
1,
2
],
"architectures": [],
"output_hidden_states": false,
"stacked_params_mapping": [],
"use_return_dict": true
},
"lora_config": null,
"prefix": "",
"quant_config": null
},
"image_encoder_configs": null,
"image_encoder_precision": "fp32",
"image_encoder_precisions": null,
"is_causal": false,
"lucy_edit_task": false,
"model_path": "Wan-AI/Wan2.1-T2V-1.3B-Diffusers",
"pipeline_config_path": null,
"postprocess_text_funcs": [
"fastvideo.models.wan.pipeline_config.t5_postprocess_text"
],
"precision": "bf16",
"preprocess_text_funcs": [
"fastvideo.configs.pipelines.base.preprocess_text"
],
"scheduler_step_in_fp32": false,
"text_encoder_configs": [
{
"_resolved_attention_backend": null,
"arch_config": {
"_supported_attention_backends": [
1,
2
],
"architectures": [],
"classifier_dropout": 0.0,
"d_ff": 2048,
"d_kv": 64,
"d_model": 512,
"decoder_start_token_id": 0,
"dense_act_fn": "relu",
"dropout_rate": 0.1,
"dtype": null,
"eos_token_id": 1,
"feed_forward_proj": "relu",
"gradient_checkpointing": false,
"hidden_size": 512,
"hidden_state_skip_layer": 0,
"initializer_factor": 1.0,
"is_encoder_decoder": true,
"is_gated_act": false,
"layer_norm_epsilon": 1e-06,
"n_positions": 512,
"num_attention_heads": 0,
"num_decoder_layers": null,
"num_heads": 8,
"num_hidden_layers": 0,
"num_layers": 6,
"output_hidden_states": false,
"output_past": true,
"pad_token_id": 0,
"relative_attention_max_distance": 128,
"relative_attention_num_buckets": 32,
"require_processor": false,
"scalable_attention": true,
"stacked_params_mapping": [
[
".qkv_proj",
".q",
"q"
],
[
".qkv_proj",
".k",
"k"
],
[
".qkv_proj",
".v",
"v"
]
],
"task_specific_params": null,
"text_len": 512,
"tie_word_embeddings": false,
"tokenizer_kwargs": {
"add_special_tokens": true,
"max_length": 512,
"return_attention_mask": true,
"return_tensors": "pt",
"truncation": true
},
"use_cache": true,
"use_return_dict": true,
"vocab_size": 32128
},
"chat_template_enable_thinking": false,
"is_chat_model": false,
"lora_config": null,
"prefix": "t5",
"quant_config": null,
"treat_empty_as_dot": false
}
],
"text_encoder_precisions": [
"fp32"
],
"ti2v_task": false,
"upsampler_config": {
"_resolved_attention_backend": null,
"arch_config": {
"stacked_params_mapping": []
}
},
"upsampler_precision": "fp32",
"vae_config": {
"_resolved_attention_backend": null,
"arch_config": {
"attn_scales": [],
"base_dim": 96,
"clip_output": true,
"decoder_base_dim": null,
"dim_mult": [
1,
2,
4,
4
],
"dropout": 0.0,
"in_channels": 3,
"is_residual": false,
"latents_mean": [
-0.7571,
-0.7089,
-0.9113,
0.1075,
-0.1745,
0.9653,
-0.1517,
1.5508,
0.4134,
-0.0715,
0.5517,
-0.3632,
-0.1922,
-0.9497,
0.2503,
-0.2921
],
"latents_std": [
2.8184,
1.4541,
2.3275,
2.6558,
1.2196,
1.7708,
2.6052,
2.0743,
3.2687,
2.1526,
2.8652,
1.5579,
1.6382,
1.1253,
2.8251,
1.916
],
"num_res_blocks": 2,
"out_channels": 3,
"patch_size": null,
"scale_factor_spatial": 8,
"scale_factor_temporal": 4,
"scaling_factor": "<tensor shape=[1, 16, 1, 1, 1] dtype=torch.float32>",
"spatial_compression_ratio": 8,
"stacked_params_mapping": [],
"temperal_downsample": [
false,
true,
true
],
"temporal_compression_ratio": 4,
"z_dim": 16
},
"blend_num_frames": 8,
"load_decoder": true,
"load_encoder": false,
"tile_sample_min_height": 256,
"tile_sample_min_num_frames": 16,
"tile_sample_min_width": 256,
"tile_sample_stride_height": 192,
"tile_sample_stride_num_frames": 12,
"tile_sample_stride_width": 192,
"use_feature_cache": true,
"use_light_vae": false,
"use_parallel_tiling": false,
"use_temporal_scaling_frames": true,
"use_temporal_tiling": false,
"use_tiling": false
},
"vae_decode_precision": "bf16",
"vae_precision": "fp32",
"vae_sp": false,
"vae_tiling": false,
"warp_denoising_step": true
},
"pipeline_config_class": "fastvideo.models.wan.pipeline_config.WanT2V480PConfig"
}
fastvideo/tests/api/config_snapshot_goldens/environment/vae_parallel.json
+375
View File
@@ -0,0 +1,375 @@
{
"fastvideo_args": {
"VSA_sparsity": 0.0,
"VSA_tile_size": 256,
"attention_backend": null,
"boundary_ratio": 0.875,
"disable_autocast": false,
"dist_timeout": null,
"distributed_executor_backend": "mp",
"dit_cpu_offload": true,
"dit_layerwise_offload": true,
"enable_stage_verification": true,
"enable_torch_compile": false,
"enable_torch_compile_audio_vae": false,
"enable_torch_compile_text_encoder": false,
"enable_torch_compile_vae": false,
"h3_sequential_load": null,
"hsdp_replicate_dim": 1,
"hsdp_shard_dim": 1,
"image_encoder_cpu_offload": true,
"inference_mode": true,
"inference_torch_compile": false,
"init_weights_from_safetensors": "",
"init_weights_from_safetensors_2": "",
"lazy_module_load": null,
"lora_nickname": "default",
"lora_path": null,
"lora_strength": 1.0,
"lora_target_modules": null,
"ltx2_audio_latent_path": null,
"ltx2_initial_latent_path": null,
"ltx2_legacy_native_noise_order": false,
"ltx2_refine_add_noise": true,
"ltx2_refine_audio_noise_path": null,
"ltx2_refine_enabled": false,
"ltx2_refine_guidance_scale": 1.0,
"ltx2_refine_lora_path": null,
"ltx2_refine_noise_path": null,
"ltx2_refine_num_inference_steps": 3,
"ltx2_refine_transformer_path": null,
"ltx2_refine_upsampler_path": null,
"ltx2_use_distilled_sigmas": true,
"ltx2_vae_spatial_tile_overlap_in_pixels": null,
"ltx2_vae_spatial_tile_size_in_pixels": null,
"ltx2_vae_temporal_tile_overlap_in_frames": null,
"ltx2_vae_temporal_tile_size_in_frames": null,
"ltx2_vae_tiling": null,
"master_port": null,
"moba_config": {},
"moba_config_path": null,
"mode": "inference",
"model_loaded": {
"transformer": true,
"upsampler": true,
"vae": true
},
"model_path": "Wan-AI/Wan2.1-T2V-1.3B-Diffusers",
"model_paths": {},
"num_gpus": 1,
"output_type": "pil",
"override_pipeline_cls_name": null,
"override_text_encoder_quant": null,
"override_text_encoder_safetensors": null,
"override_transformer_cls_name": null,
"pin_cpu_memory": true,
"preprocess_config": null,
"prompt_txt": null,
"ray_placement_group": null,
"ray_runtime_env": null,
"refine_add_noise": null,
"refine_audio_noise_path": null,
"refine_enabled": null,
"refine_guidance_scale": null,
"refine_lora_path": null,
"refine_noise_path": null,
"refine_num_inference_steps": null,
"refine_transformer_path": null,
"refine_upsampler_path": null,
"revision": null,
"sp_size": 1,
"taeh3_checkpoint": null,
"taeh3_chunk_size": 5,
"text_encoder_cpu_offload": true,
"torch_compile_kwargs": {},
"torch_compile_kwargs_audio_vae": {},
"torch_compile_kwargs_dit": {},
"torch_compile_kwargs_text_encoder": {},
"torch_compile_kwargs_vae": {},
"tp_size": 1,
"transformer_quant": null,
"trust_remote_code": false,
"use_fsdp_inference": false,
"vae_cpu_offload": true,
"vae_parallel_decode": true,
"vae_parallel_decode_strategy": "all_gather",
"vae_parallel_encode": true,
"video_decode_backend": "h3-vae",
"workload_type": "t2v"
},
"pipeline_config": {
"boundary_ratio": null,
"disable_autocast": false,
"dit_config": {
"_resolved_attention_backend": null,
"arch_config": {
"_supported_attention_backends": [
3,
1,
2,
7,
10,
4,
5,
6,
11,
12
],
"added_kv_proj_dim": null,
"attention_head_dim": 128,
"boundary_ratio": null,
"cast_prompt_embeds_to_dit_dtype": false,
"cross_attn_norm": true,
"eps": 1e-06,
"exclude_lora_layers": [
"embedder"
],
"ffn_dim": 13824,
"freq_dim": 256,
"hidden_size": 5120,
"image_dim": null,
"in_channels": 16,
"local_attn_size": -1,
"num_attention_heads": 40,
"num_channels_latents": 16,
"num_frames_per_block": 3,
"num_layers": 40,
"out_channels": 16,
"patch_size": [
1,
2,
2
],
"pos_embed_seq_len": null,
"qk_norm": "rms_norm_across_heads",
"r_embedder": false,
"r_embedder_deltatime_type": "r",
"r_embedder_fusion": "additive",
"r_embedder_gate_value": 0.25,
"rope_cache_policy": "absolute",
"rope_max_seq_len": 1024,
"sink_size": 0,
"sliding_window_num_frames": 21,
"stacked_params_mapping": [],
"text_dim": 4096
},
"prefix": "Wan",
"quant_config": null
},
"dit_precision": "bf16",
"dmd_denoising_steps": null,
"embedded_cfg_scale": 6.0,
"flow_shift": 3.0,
"flow_shift_sr": null,
"image_encoder_config": {
"_resolved_attention_backend": null,
"arch_config": {
"_supported_attention_backends": [
1,
2
],
"architectures": [],
"output_hidden_states": false,
"stacked_params_mapping": [],
"use_return_dict": true
},
"lora_config": null,
"prefix": "",
"quant_config": null
},
"image_encoder_configs": null,
"image_encoder_precision": "fp32",
"image_encoder_precisions": null,
"is_causal": false,
"lucy_edit_task": false,
"model_path": "Wan-AI/Wan2.1-T2V-1.3B-Diffusers",
"pipeline_config_path": null,
"postprocess_text_funcs": [
"fastvideo.models.wan.pipeline_config.t5_postprocess_text"
],
"precision": "bf16",
"preprocess_text_funcs": [
"fastvideo.configs.pipelines.base.preprocess_text"
],
"scheduler_step_in_fp32": false,
"text_encoder_configs": [
{
"_resolved_attention_backend": null,
"arch_config": {
"_supported_attention_backends": [
1,
2
],
"architectures": [],
"classifier_dropout": 0.0,
"d_ff": 2048,
"d_kv": 64,
"d_model": 512,
"decoder_start_token_id": 0,
"dense_act_fn": "relu",
"dropout_rate": 0.1,
"dtype": null,
"eos_token_id": 1,
"feed_forward_proj": "relu",
"gradient_checkpointing": false,
"hidden_size": 512,
"hidden_state_skip_layer": 0,
"initializer_factor": 1.0,
"is_encoder_decoder": true,
"is_gated_act": false,
"layer_norm_epsilon": 1e-06,
"n_positions": 512,
"num_attention_heads": 0,
"num_decoder_layers": null,
"num_heads": 8,
"num_hidden_layers": 0,
"num_layers": 6,
"output_hidden_states": false,
"output_past": true,
"pad_token_id": 0,
"relative_attention_max_distance": 128,
"relative_attention_num_buckets": 32,
"require_processor": false,
"scalable_attention": true,
"stacked_params_mapping": [
[
".qkv_proj",
".q",
"q"
],
[
".qkv_proj",
".k",
"k"
],
[
".qkv_proj",
".v",
"v"
]
],
"task_specific_params": null,
"text_len": 512,
"tie_word_embeddings": false,
"tokenizer_kwargs": {
"add_special_tokens": true,
"max_length": 512,
"return_attention_mask": true,
"return_tensors": "pt",
"truncation": true
},
"use_cache": true,
"use_return_dict": true,
"vocab_size": 32128
},
"chat_template_enable_thinking": false,
"is_chat_model": false,
"lora_config": null,
"prefix": "t5",
"quant_config": null,
"treat_empty_as_dot": false
}
],
"text_encoder_precisions": [
"fp32"
],
"ti2v_task": false,
"upsampler_config": {
"_resolved_attention_backend": null,
"arch_config": {
"stacked_params_mapping": []
}
},
"upsampler_precision": "fp32",
"vae_config": {
"_resolved_attention_backend": null,
"arch_config": {
"attn_scales": [],
"base_dim": 96,
"clip_output": true,
"decoder_base_dim": null,
"dim_mult": [
1,
2,
4,
4
],
"dropout": 0.0,
"in_channels": 3,
"is_residual": false,
"latents_mean": [
-0.7571,
-0.7089,
-0.9113,
0.1075,
-0.1745,
0.9653,
-0.1517,
1.5508,
0.4134,
-0.0715,
0.5517,
-0.3632,
-0.1922,
-0.9497,
0.2503,
-0.2921
],
"latents_std": [
2.8184,
1.4541,
2.3275,
2.6558,
1.2196,
1.7708,
2.6052,
2.0743,
3.2687,
2.1526,
2.8652,
1.5579,
1.6382,
1.1253,
2.8251,
1.916
],
"num_res_blocks": 2,
"out_channels": 3,
"patch_size": null,
"scale_factor_spatial": 8,
"scale_factor_temporal": 4,
"scaling_factor": "<tensor shape=[1, 16, 1, 1, 1] dtype=torch.float32>",
"spatial_compression_ratio": 8,
"stacked_params_mapping": [],
"temperal_downsample": [
false,
true,
true
],
"temporal_compression_ratio": 4,
"z_dim": 16
},
"blend_num_frames": 8,
"load_decoder": true,
"load_encoder": false,
"tile_sample_min_height": 256,
"tile_sample_min_num_frames": 16,
"tile_sample_min_width": 256,
"tile_sample_stride_height": 192,
"tile_sample_stride_num_frames": 12,
"tile_sample_stride_width": 192,
"use_feature_cache": true,
"use_light_vae": false,
"use_parallel_tiling": false,
"use_temporal_scaling_frames": true,
"use_temporal_tiling": false,
"use_tiling": false
},
"vae_decode_precision": "bf16",
"vae_precision": "fp32",
"vae_sp": false,
"vae_tiling": false,
"warp_denoising_step": true
},
"pipeline_config_class": "fastvideo.models.wan.pipeline_config.WanT2V480PConfig"
}