[Request] Parsed ggml-turbo/request0.json [Request] seed=42 lm_batch=1 synth_batch=1 [Request] caption: An upbeat and anthemic pop-rock track driven by bright, slig... [Request] lyrics: 490 bytes [Request] bpm=83 dur=88 key=G major ts=4 lang=fr [Request] lm: temp=0.85 cfg=2.0 top_p=0.90 top_k=0 [Request] dit: steps=8 guidance=0.0 shift=3.0 [Request] task_type: text2music [Request] solver: euler (stork_substeps=10) [Request] lm_mode: generate [Request] output_format: mp3 [Request] synth_model: acestep-v15-turbo-BF16.gguf [Request] vae: vae-BF16.gguf [Request] audio_codes: (present) [Ace-Synth] Batch: 1 request(s) [Registry] Qwen3-Embedding-0.6B-BF16.gguf -> Text-Enc [Registry] Qwen3-Embedding-0.6B-Q8_0.gguf -> Text-Enc [Registry] acestep-5Hz-lm-0.6B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-0.6B-Q8_0.gguf -> LM [Registry] acestep-5Hz-lm-1.7B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-1.7B-Q8_0.gguf -> LM [Registry] acestep-5Hz-lm-4B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q5_K_M.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q6_K.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q8_0.gguf -> LM [Registry] acestep-v15-base-BF16.gguf -> DiT [Registry] acestep-v15-base-Q4_K_M.gguf -> DiT [Registry] acestep-v15-base-Q5_K_M.gguf -> DiT [Registry] acestep-v15-base-Q6_K.gguf -> DiT [Registry] acestep-v15-base-Q8_0.gguf -> DiT [Registry] acestep-v15-sft-BF16.gguf -> DiT [Registry] acestep-v15-sft-Q4_K_M.gguf -> DiT [Registry] acestep-v15-sft-Q5_K_M.gguf -> DiT [Registry] acestep-v15-sft-Q6_K.gguf -> DiT [Registry] acestep-v15-sft-Q8_0.gguf -> DiT [Registry] acestep-v15-sftturbo50-BF16.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q4_K_M.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q5_K_M.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q6_K.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-BF16.gguf -> DiT [Registry] acestep-v15-turbo-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-continuous-BF16.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-shift1-BF16.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-shift3-BF16.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-base-BF16.gguf -> DiT [Registry] acestep-v15-xl-base-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-base-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-base-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-base-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-sft-BF16.gguf -> DiT [Registry] acestep-v15-xl-sft-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-sft-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-sft-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-sft-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-BF16.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-turbo-BF16.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q8_0.gguf -> DiT [Registry] scragvae-BF16.gguf -> VAE [Registry] vae-BF16.gguf -> VAE [Store] Created (policy=STRICT) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-turbo-BF16.gguf: 678 tensors, data at offset 57024 [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-turbo-BF16.gguf: 678 tensors, data at offset 57024 [Synth-Load] Ready: turbo=yes, fa=no, batch_cfg=yes [Resolve-Params] max audio codes across batch: 434 (86.8s @ 5Hz) [Resolve-T] T=2170, S=1085 [Resolve-T] seed=42, steps=8, guidance=1.0, shift=3.0, duration=88.0s [BPE] Loaded from GGUF: 151643 vocab, 151387 merges ggml_cuda_init: found 1 CUDA devices (Total VRAM: 97247 MiB): Device 0: NVIDIA RTX PRO 6000 Blackwell Workstation Edition, compute capability 12.0, VMM: yes, VRAM: 97247 MiB load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = NVIDIA RTX PRO 6000 Blackwell Workstation Edition (NVIDIA) | uma: 0 | fp16: 1 | bf16: 0 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: NV_coopmat2 load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] TextEncoder backend: CUDA0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/Qwen3-Embedding-0.6B-BF16.gguf: 310 tensors, data at offset 5337568 [Load] TextEncoder: 28L, H=1024, Nh=16/8 [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 310 tensors, 1136.5 MB into backend [Store] Load TextEnc: 273 ms [Debug] text_hidden: [70, 1024] first4: 3.652379 1.048645 0.229843 -13.063987 [Debug] lyric_embed: [167, 1024] first4: 0.029175 0.032227 -0.022339 -0.028809 [Store] Unload TextEnc (1136.5 MB) load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] CondEncoder backend: CUDA0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-turbo-BF16.gguf: 678 tensors, data at offset 57024 [Load] LyricEncoder: 8L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [Load] TimbreEncoder: 4L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 140 tensors, 1160.5 MB into backend [Load] CondEncoder: lyric(8L), timbre(4L), text_proj, null_cond [Store] Load CondEnc: 535 ms [CondEnc] Lyric sliding mask: 167x167, window=128 [CondEnc] Timbre sliding mask: 1x1, window=128 [Encode] Packed: lyric=167 + timbre=1 + text=70 = 238 tokens [Encode-Text Batch0] 70+167 tokens -> enc_S=238, 19.0 ms [Debug] enc_hidden: [238, 2048] first4: 1.762581 -0.049432 -0.132992 0.058069 [Store] Unload CondEnc (1160.5 MB) load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] Detokenizer backend: CUDA0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-turbo-BF16.gguf: 678 tensors, data at offset 57024 [WeightCtx] Loaded 30 tensors, 200.3 MB into backend [Load] Detokenizer: FSQ(6->2048) + 2L encoder(S=5, 2048->64) [Store] Load FSQ-Detok: 91 ms CUDA Graph id 8 reused ggml_backend_cuda_graph_compute: CUDA graph warmup complete CUDA Graph id 8 reused [Context] Decoded: 434 codes -> 2170 frames (86.8s @ 25Hz) [Build-Context Batch0] Detokenizer: 107.5 ms, 434 codes [Debug] detok_output: [2170, 64] first4: -0.124076 1.434873 0.309935 -0.624868 [Store] Unload FSQ-Detok (200.3 MB) [Init-Noise Batch0] Philox noise seed=42, [2170, 64] solver=euler [Debug] noise: [2170, 64] first4: 0.194336 2.156250 -0.171875 0.847656 [Debug] context: [2170, 128] first4: -0.124076 1.434873 0.309935 -0.624868 [Init-Noise] Starting: T=2170, S=1085, enc_S=238, steps=8, batch=1 (cover) [Dedup] Previous line repeated 432 times total load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] DiT backend: CUDA0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-turbo-BF16.gguf: 678 tensors, data at offset 57024 [DiT] Self-attn: Q+K+V fused [DiT] Cross-attn: Q+K+V fused [DiT] MLP: gate+up fused [Load] null_condition_emb found (CFG available) [WeightCtx] Loaded 478 tensors, 3007.9 MB into backend [Load] DiT: 24 layers, H=2048, Nh=16/8, D=128 [Store] Load DiT: 1368 ms [DiT] Batch N=1, T=2170, S=1085, enc_S=238 [DiT] Graph: 2129 nodes [DiT] Solver: ODE Euler (1 NFE/step, order 1) [Debug] tproj: [12288] first4: 0.260062 -0.161562 -0.097030 0.052313 [Debug] temb: [2048] first4: 0.000069 -0.132499 -0.035430 0.064753 [Debug] temb_t: [2048] first4: 0.001065 0.026818 -0.052754 0.063717 [Debug] temb_r: [2048] first4: -0.000996 -0.159317 0.017323 0.001036 [Debug] sinusoidal_t: [256] first4: 0.562486 0.789701 0.439822 -0.023583 [Debug] sinusoidal_r: [256] first4: 1.000000 1.000000 1.000000 1.000000 [Debug] temb_lin1_t: [2048] first4: -0.049318 -0.051829 -0.014251 -0.038444 [Debug] temb_lin1_r: [2048] first4: -0.013266 -0.018319 -0.016375 0.008532 [Debug] hidden_after_proj_in: [2048, 1085] first4: -1.039807 -0.969741 0.533785 0.446624 [Debug] proj_in_input: [192, 2170] first4: -0.124076 1.434873 0.309935 -0.624868 [Debug] enc_after_cond_emb: [2048, 238] first4: -0.167358 0.814621 0.325745 -0.561218 [Debug] layer0_sa_input: [2048, 1085] first4: -0.719315 -0.764248 -0.047614 0.261743 [Debug] layer0_q_after_rope: [128, 16] first4: -1.540607 -1.044672 0.188601 0.456018 [Debug] layer0_k_after_rope: [128, 8] first4: -0.167358 0.814621 0.325745 -0.561218 [Debug] layer0_sa_output: [2048, 1085] first4: -1.500000 0.168945 -0.353516 0.515625 [Debug] layer0_attn_out: [2048, 1085] first4: -1.540607 -1.044672 0.188601 0.456018 [Debug] layer0_after_self_attn: [2048, 1085] first4: -1.540607 -1.044672 0.188601 0.456018 [Debug] layer0_after_cross_attn: [2048, 1085] first4: -1.596271 -0.821040 -0.295774 0.493371 [Debug] hidden_after_layer0: [2048, 1085] first4: -9.069268 0.562051 52.140285 -0.910502 [Debug] hidden_after_layer6: [2048, 1085] first4: -21.304485 0.122005 33.367920 -4.504872 [Debug] hidden_after_layer12: [2048, 1085] first4: -15.067718 -17.771275 71.619621 28.806711 [Debug] hidden_after_layer18: [2048, 1085] first4: -27.166594 15.731917 58.799187 21.235981 [Debug] hidden_after_layer23: [2048, 1085] first4: -9.282040 44.732018 196.654724 144.040543 [Debug] dit_step0_vt: [2170, 64] first4: 0.022634 1.115833 0.351516 2.377223 [Debug] dit_step0_xt: [2170, 64] first4: 0.193307 2.105530 -0.187853 0.739601 [DiT] Step 1/8 t=1.000 ggml_backend_cuda_graph_compute: CUDA graph warmup complete [Debug] dit_step1_vt: [2170, 64] first4: -0.228331 1.306460 -0.137135 1.866482 [Debug] dit_step1_xt: [2170, 64] first4: 0.205762 2.034269 -0.180373 0.637793 [DiT] Step 2/8 t=0.955 [Debug] dit_step2_vt: [2170, 64] first4: -0.014647 1.212786 0.103002 2.391060 [Debug] dit_step2_xt: [2170, 64] first4: 0.206738 1.953416 -0.187240 0.478389 [DiT] Step 3/8 t=0.900 [Debug] dit_step3_vt: [2170, 64] first4: 0.266804 1.086355 0.253257 2.641131 [Debug] dit_step3_xt: [2170, 64] first4: 0.184504 1.862887 -0.208344 0.258294 [DiT] Step 4/8 t=0.833 [Debug] dit_step4_vt: [2170, 64] first4: 0.297957 0.996606 0.111334 2.712721 [Debug] dit_step4_xt: [2170, 64] first4: 0.152580 1.756108 -0.220273 -0.032354 [DiT] Step 5/8 t=0.750 [Debug] dit_step5_vt: [2170, 64] first4: 0.283015 0.921376 -0.300600 2.771382 [Debug] dit_step5_xt: [2170, 64] first4: 0.112150 1.624483 -0.177330 -0.428266 [DiT] Step 6/8 t=0.643 [Debug] dit_step6_vt: [2170, 64] first4: 0.171021 0.636431 -0.827703 2.840888 [Debug] dit_step6_xt: [2170, 64] first4: 0.077946 1.497196 -0.011790 -0.996444 [DiT] Step 7/8 t=0.500 [Debug] dit_step7_vt: [2170, 64] first4: 0.000409 0.186355 -1.485624 3.123371 [Debug] dit_x0: [2170, 64] first4: 0.077823 1.441290 0.433898 -1.933455 [DiT] Step 8/8 t=0.300 [DiT-Generate] Total: 286.6 ms (286.6 ms/sample) [Debug] dit_output: [2170, 64] first4: 0.077823 1.441290 0.433898 -1.933455 [Store] Unload DiT (3007.9 MB) [GGUF] /mnt/workspace/git/acestep.cpp/models/vae-BF16.gguf: 365 tensors, data at offset 30048 load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] VAE backend: CUDA0 (CPU threads: 16) [VAE] Backend: CUDA0, Weight buffer: 161.1 MB [VAE] Loaded: 5 blocks, upsample=1920x, F32 activations [Store] Load VAE-Dec: 677 ms [VAE] Tiled decode: 3 tiles (chunk=1024, overlap=64, stride=896) [VAE] Graph: 479 nodes, T_latent=960 [VAE] Upsample factor: 1920.00 (expected ~1920) [VAE] Graph: 479 nodes, T_latent=1024 [VAE] Graph: 479 nodes, T_latent=442 [VAE] Tiled decode done: 3 tiles -> T_audio=4166400 (86.80s @ 48kHz) [VAE-Decode Batch0] Decode: 615.0 ms [Debug] vae_audio: [2, 4166400] first4: 0.000704 0.001266 0.001185 0.001547 [Store] Unload VAE-Dec (161.1 MB) [MP3] Encoding 86.8s @ 128 kbps, 48000 Hz stereo [MP3] 1388928 bytes (12.0:1), 449 ms (193.14x realtime), 32 threads [MP3] Wrote ggml-turbo/request00.mp3 [Ace-Synth] All done 2026-07-12 10:11:21.023 | DEBUG | acestep.core.generation.handler.init_service_memory_basic:_apply_malloc_mmap_threshold:50 - [memory] Set M_MMAP_THRESHOLD=131072 for immediate OS reclaim of large frees Skipping import of cpp extensions due to incompatible torch version. Please upgrade to torch >= 2.11.0 (found 2.10.0+cu128). 2026-07-12 10:11:23.022 | WARNING | acestep.training.trainer::40 - bitsandbytes not installed. Using standard AdamW. 2026-07-12 10:11:23.027 | INFO | acestep.gpu_config:get_gpu_memory_gb:578 - CUDA GPU detected: NVIDIA RTX PRO 6000 Blackwell Workstation Edition (95.0 GB) 2026-07-12 10:11:23.127 | INFO | acestep.core.generation.handler.init_service_loader:_load_main_model_from_checkpoint:174 - [initialize_service] Attempting to load model with attention implementation: sdpa Unable to import `torchao` Tensor objects. This may affect loading checkpoints serialized with `torchao` 2026-07-12 10:11:24.692 | INFO | acestep.core.generation.handler.generate_music:generate_music:289 - [generate_music] Turbo model detected: overriding guidance_scale 0.0 -> 1.0 (turbo does not use CFG). 2026-07-12 10:11:24.692 | INFO | acestep.core.generation.handler.generate_music:generate_music:304 - [generate_music] Starting generation... 2026-07-12 10:11:24.692 | INFO | acestep.core.generation.handler.generate_music:generate_music:307 - [generate_music] Preparing inputs... 2026-07-12 10:11:24.692 | INFO | acestep.core.generation.handler.generate_music_request:_prepare_reference_and_source_audio:179 - [generate_music] cover task: no src_audio but audio codes provided, proceeding with codes 2026-07-12 10:11:24.696 | INFO | acestep.core.generation.handler.generate_music:_vram_preflight_check:156 - [generate_music] VRAM pre-flight: 87.83 GB free, ~0.94 GB needed (batch=1, duration=88s, mode=turbo). 2026-07-12 10:11:24.701 | INFO | acestep.core.generation.handler.conditioning_target:_prepare_target_latents_and_wavs:78 - [generate_music] Decoding audio codes for item 0... 2026-07-12 10:11:24.905 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_precomputed_lm_hints:31 - [generate_music] Decoding audio codes for LM hints for item 0... 2026-07-12 10:11:24.907 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:119 - ====================================================================== 2026-07-12 10:11:24.907 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:120 - πŸ” [DEBUG] DiT TEXT ENCODER INPUT (Inference) 2026-07-12 10:11:24.907 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:121 - ====================================================================== 2026-07-12 10:11:24.907 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:122 - text_prompt: # Instruction Generate audio semantic tokens based on the given conditions: # Caption An upbeat and anthemic pop-rock track driven by bright, slightly overdriven # Metas - bpm: 83 - timesignature: 4 - keyscale: G major - duration: 88 seconds <|endoftext|> 2026-07-12 10:11:24.907 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:123 - ====================================================================== 2026-07-12 10:11:24.907 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:124 - lyrics_text: # Languages fr # Lyric # Lyric [Intro - Guitar Riff] [Verse 1] Dans le monde des tutos virtuels G ta toise en nouvelle passion Avec Ggendoline et PumbΓ© Γ  midi La communautΓ©, c'est l'unitΓ© Quel joie, une clΓ© [Chorus] Dans le monde des tutos virtuels GΓ’ndoline et PumbΓ© Γ  midi Une famille Γ  connecter, c'est vrai D'un enfant qui voit toi fusionner [Guitar Solo] [Verse 2] Dans le monde des tutos virtuels GΓ’ndoline, PumbΓ© Γ  midi Une famille Γ  connecter, c'est vrai D'un enfant qui voit toi fusionner<|endoftext|> 2026-07-12 10:11:24.907 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:125 - ====================================================================== 2026-07-12 10:11:24.914 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:110 - [preprocess_batch] Inferring prompt embeddings... 2026-07-12 10:11:24.929 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:113 - [preprocess_batch] Inferring lyric embeddings... 2026-07-12 10:11:24.929 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:161 - [service_generate] Generating audio... (DiT backend: PyTorch (cuda)) 2026-07-12 10:11:24.964 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:263 - [service_generate] DiT diffusion via PyTorch (cuda)... 2026-07-12 10:11:24.972 | INFO | acestep.models.common.dcw_correction:__init__:98 - [DCW] Active β€” mode=double, scaler=0.0500, high_scaler=0.0200, wavelet='haar' 2026-07-12 10:11:25.057 | WARNING | acestep.models.common.dcw_loader:_try_import:39 - DCW is enabled but 'pytorch_wavelets' is not installed. Install with `pip install pytorch_wavelets PyWavelets` to use Differential Correction in Wavelet domain. Falling back to no-op for this generation. 2026-07-12 10:11:25.282 | INFO | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:42 - [generate_music] Model generation completed. Decoding latents... 2026-07-12 10:11:25.284 | DEBUG | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:64 - [generate_music] pred_latents: torch.Size([1, 2170, 64]), dtype=torch.bfloat16 2026-07-12 10:11:25.284 | DEBUG | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:65 - [generate_music] time_costs: {'encoder_time_cost': 0.006887912750244141, 'diffusion_time_cost': 0.3103818893432617, 'diffusion_per_step_time_cost': 0.038797736167907715, 'total_time_cost': 0.31726980209350586, 'offload_time_cost': 0.0} 2026-07-12 10:11:25.297 | INFO | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:127 - [generate_music] Decoding latents with VAE... 2026-07-12 10:11:25.300 | DEBUG | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:136 - [generate_music] Before VAE decode: allocated=5.98GB, max=7.29GB 2026-07-12 10:11:25.300 | INFO | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:154 - [generate_music] Effective free VRAM before VAE decode: 87.73 GB 2026-07-12 10:11:25.300 | INFO | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:172 - [generate_music] Using tiled VAE decode to reduce VRAM usage... 2026-07-12 10:11:25.300 | DEBUG | acestep.core.generation.handler.memory_utils:_get_auto_decode_chunk_size:108 - [_get_auto_decode_chunk_size] Effective free VRAM: 87.73 GB 2026-07-12 10:11:25.300 | DEBUG | acestep.core.generation.handler.memory_utils:_should_offload_wav_to_cpu:131 - [_should_offload_wav_to_cpu] Effective free VRAM: 87.73 GB 2026-07-12 10:11:25.300 | INFO | acestep.core.generation.handler.vae_decode:tiled_decode:56 - [tiled_decode] chunk_size=512, offload_wav_to_cpu=False, latents_shape=torch.Size([1, 64, 2170]) 2026-07-12 10:11:25.576 | DEBUG | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:193 - [generate_music] After VAE decode: allocated=6.15GB, max=7.44GB 2026-07-12 10:11:25.736 | INFO | acestep.core.generation.handler.generate_music_payload:_build_generate_music_success_payload:38 - [generate_music] VAE decode completed. Preparing audio tensors... 2026-07-12 10:11:25.739 | INFO | acestep.core.generation.handler.generate_music_payload:_build_generate_music_success_payload:50 - [generate_music] Done! Generated 1 audio tensors. [Request] Parsed ggml-sft/request0.json [Request] seed=42 lm_batch=1 synth_batch=1 [Request] caption: An upbeat and anthemic pop-rock track driven by bright, slig... [Request] lyrics: 490 bytes [Request] bpm=83 dur=88 key=G major ts=4 lang=fr [Request] lm: temp=0.85 cfg=2.0 top_p=0.90 top_k=0 [Request] dit: steps=50 guidance=1.0 shift=1.0 [Request] task_type: text2music [Request] solver: euler (stork_substeps=10) [Request] lm_mode: generate [Request] output_format: mp3 [Request] synth_model: acestep-v15-sft-BF16.gguf [Request] vae: vae-BF16.gguf [Request] audio_codes: (present) [Ace-Synth] Batch: 1 request(s) [Registry] Qwen3-Embedding-0.6B-BF16.gguf -> Text-Enc [Registry] Qwen3-Embedding-0.6B-Q8_0.gguf -> Text-Enc [Registry] acestep-5Hz-lm-0.6B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-0.6B-Q8_0.gguf -> LM [Registry] acestep-5Hz-lm-1.7B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-1.7B-Q8_0.gguf -> LM [Registry] acestep-5Hz-lm-4B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q5_K_M.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q6_K.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q8_0.gguf -> LM [Registry] acestep-v15-base-BF16.gguf -> DiT [Registry] acestep-v15-base-Q4_K_M.gguf -> DiT [Registry] acestep-v15-base-Q5_K_M.gguf -> DiT [Registry] acestep-v15-base-Q6_K.gguf -> DiT [Registry] acestep-v15-base-Q8_0.gguf -> DiT [Registry] acestep-v15-sft-BF16.gguf -> DiT [Registry] acestep-v15-sft-Q4_K_M.gguf -> DiT [Registry] acestep-v15-sft-Q5_K_M.gguf -> DiT [Registry] acestep-v15-sft-Q6_K.gguf -> DiT [Registry] acestep-v15-sft-Q8_0.gguf -> DiT [Registry] acestep-v15-sftturbo50-BF16.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q4_K_M.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q5_K_M.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q6_K.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-BF16.gguf -> DiT [Registry] acestep-v15-turbo-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-continuous-BF16.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-shift1-BF16.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-shift3-BF16.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-base-BF16.gguf -> DiT [Registry] acestep-v15-xl-base-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-base-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-base-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-base-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-sft-BF16.gguf -> DiT [Registry] acestep-v15-xl-sft-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-sft-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-sft-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-sft-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-BF16.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-turbo-BF16.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q8_0.gguf -> DiT [Registry] scragvae-BF16.gguf -> VAE [Registry] vae-BF16.gguf -> VAE [Store] Created (policy=STRICT) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-sft-BF16.gguf: 678 tensors, data at offset 56960 [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-sft-BF16.gguf: 678 tensors, data at offset 56960 [Synth-Load] Ready: turbo=no, fa=no, batch_cfg=yes [Resolve-Params] max audio codes across batch: 434 (86.8s @ 5Hz) [Resolve-T] T=2170, S=1085 [Resolve-T] seed=42, steps=50, guidance=1.0, shift=1.0, duration=88.0s [BPE] Loaded from GGUF: 151643 vocab, 151387 merges ggml_cuda_init: found 1 CUDA devices (Total VRAM: 97247 MiB): Device 0: NVIDIA RTX PRO 6000 Blackwell Workstation Edition, compute capability 12.0, VMM: yes, VRAM: 97247 MiB load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = NVIDIA RTX PRO 6000 Blackwell Workstation Edition (NVIDIA) | uma: 0 | fp16: 1 | bf16: 0 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: NV_coopmat2 load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] TextEncoder backend: CUDA0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/Qwen3-Embedding-0.6B-BF16.gguf: 310 tensors, data at offset 5337568 [Load] TextEncoder: 28L, H=1024, Nh=16/8 [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 310 tensors, 1136.5 MB into backend [Store] Load TextEnc: 268 ms [Debug] text_hidden: [70, 1024] first4: 3.652379 1.048645 0.229843 -13.063987 [Debug] lyric_embed: [167, 1024] first4: 0.029175 0.032227 -0.022339 -0.028809 [Store] Unload TextEnc (1136.5 MB) load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] CondEncoder backend: CUDA0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-sft-BF16.gguf: 678 tensors, data at offset 56960 [Load] LyricEncoder: 8L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [Load] TimbreEncoder: 4L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 140 tensors, 1160.5 MB into backend [Load] CondEncoder: lyric(8L), timbre(4L), text_proj, null_cond [Store] Load CondEnc: 518 ms [CondEnc] Lyric sliding mask: 167x167, window=128 [CondEnc] Timbre sliding mask: 1x1, window=128 [Encode] Packed: lyric=167 + timbre=1 + text=70 = 238 tokens [Encode-Text Batch0] 70+167 tokens -> enc_S=238, 8.7 ms [Debug] enc_hidden: [238, 2048] first4: 1.762581 -0.049432 -0.132992 0.058069 [Store] Unload CondEnc (1160.5 MB) load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] Detokenizer backend: CUDA0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-sft-BF16.gguf: 678 tensors, data at offset 56960 [WeightCtx] Loaded 30 tensors, 200.3 MB into backend [Load] Detokenizer: FSQ(6->2048) + 2L encoder(S=5, 2048->64) [Store] Load FSQ-Detok: 93 ms CUDA Graph id 8 reused ggml_backend_cuda_graph_compute: CUDA graph warmup complete CUDA Graph id 8 reused [Context] Decoded: 434 codes -> 2170 frames (86.8s @ 25Hz) [Build-Context Batch0] Detokenizer: 106.1 ms, 434 codes [Debug] detok_output: [2170, 64] first4: -0.124076 1.434873 0.309935 -0.624868 [Store] Unload FSQ-Detok (200.3 MB) [Init-Noise Batch0] Philox noise seed=42, [2170, 64] solver=euler [Debug] noise: [2170, 64] first4: 0.194336 2.156250 -0.171875 0.847656 [Debug] context: [2170, 128] first4: -0.124076 1.434873 0.309935 -0.624868 [Init-Noise] Starting: T=2170, S=1085, enc_S=238, steps=50, batch=1 (cover) [Dedup] Previous line repeated 432 times total load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] DiT backend: CUDA0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-sft-BF16.gguf: 678 tensors, data at offset 56960 [DiT] Self-attn: Q+K+V fused [DiT] Cross-attn: Q+K+V fused [DiT] MLP: gate+up fused [Load] null_condition_emb found (CFG available) [WeightCtx] Loaded 478 tensors, 3007.9 MB into backend [Load] DiT: 24 layers, H=2048, Nh=16/8, D=128 [Store] Load DiT: 1348 ms [DiT] Batch N=1, T=2170, S=1085, enc_S=238 [DiT] Graph: 2129 nodes [DiT] Solver: ODE Euler (1 NFE/step, order 1) [Debug] tproj: [12288] first4: 0.154506 -0.116115 -0.086937 0.082315 [Debug] temb: [2048] first4: -0.002757 -0.176460 0.004187 -0.001983 [Debug] temb_t: [2048] first4: -0.001235 0.003330 -0.013135 -0.002086 [Debug] temb_r: [2048] first4: -0.001522 -0.179790 0.017322 0.000103 [Debug] sinusoidal_t: [256] first4: 0.562486 0.789701 0.439822 -0.023583 [Debug] sinusoidal_r: [256] first4: 1.000000 1.000000 1.000000 1.000000 [Debug] temb_lin1_t: [2048] first4: -0.039237 0.029644 0.028280 -0.024901 [Debug] temb_lin1_r: [2048] first4: 0.002274 0.002999 -0.001040 -0.000388 [Debug] hidden_after_proj_in: [2048, 1085] first4: -1.090454 -0.926547 0.495885 0.440315 [Debug] proj_in_input: [192, 2170] first4: -0.124076 1.434873 0.309935 -0.624868 [Debug] enc_after_cond_emb: [2048, 238] first4: -0.187317 0.874771 0.328247 -0.574707 [Debug] layer0_sa_input: [2048, 1085] first4: -0.916055 -0.717175 -0.037946 0.294390 [Debug] layer0_q_after_rope: [128, 16] first4: -1.804187 -1.386879 -0.130883 0.379343 [Debug] layer0_k_after_rope: [128, 8] first4: -0.187317 0.874771 0.328247 -0.574707 [Debug] layer0_sa_output: [2048, 1085] first4: -1.648438 0.812500 -0.554688 0.542969 [Debug] layer0_attn_out: [2048, 1085] first4: -1.804187 -1.386879 -0.130883 0.379343 [Debug] layer0_after_self_attn: [2048, 1085] first4: -1.804187 -1.386879 -0.130883 0.379343 [Debug] layer0_after_cross_attn: [2048, 1085] first4: -1.963367 -1.050941 -0.412133 0.383799 [Debug] hidden_after_layer0: [2048, 1085] first4: -9.375484 1.139080 57.110970 -1.607987 [Debug] hidden_after_layer6: [2048, 1085] first4: -17.280643 4.597129 57.510284 -2.483391 [Debug] hidden_after_layer12: [2048, 1085] first4: -9.880440 4.828327 -20.658504 -1.613276 [Debug] hidden_after_layer18: [2048, 1085] first4: -3.342876 19.076633 -33.881344 -10.741245 [Debug] hidden_after_layer23: [2048, 1085] first4: 34.039001 64.193748 60.004242 -9.270440 [Debug] dit_step0_vt: [2170, 64] first4: -0.571996 2.508797 -0.199864 1.467649 [Debug] dit_step0_xt: [2170, 64] first4: 0.205776 2.106074 -0.167878 0.818303 [DiT] Step 1/50 t=1.000 ggml_backend_cuda_graph_compute: CUDA graph warmup complete [Debug] dit_step1_vt: [2170, 64] first4: -0.643908 2.536289 -0.036919 1.303706 [Debug] dit_step1_xt: [2170, 64] first4: 0.218654 2.055348 -0.167139 0.792229 [DiT] Step 2/50 t=0.980 [Debug] dit_step2_vt: [2170, 64] first4: -0.634547 2.482462 0.004237 1.316489 [Debug] dit_step2_xt: [2170, 64] first4: 0.231345 2.005699 -0.167224 0.765899 [DiT] Step 3/50 t=0.960 [Debug] dit_step3_vt: [2170, 64] first4: -0.608708 2.422676 0.030588 1.330857 [Debug] dit_step3_xt: [2170, 64] first4: 0.243519 1.957245 -0.167836 0.739282 [DiT] Step 4/50 t=0.940 [Debug] dit_step4_vt: [2170, 64] first4: -0.575322 2.344691 0.016715 1.312309 [Debug] dit_step4_xt: [2170, 64] first4: 0.255026 1.910352 -0.168170 0.713036 [DiT] Step 5/50 t=0.920 [Debug] dit_step5_vt: [2170, 64] first4: -0.527935 2.266378 -0.018465 1.294902 [Debug] dit_step5_xt: [2170, 64] first4: 0.265584 1.865024 -0.167801 0.687138 [DiT] Step 6/50 t=0.900 [Debug] dit_step6_vt: [2170, 64] first4: -0.494518 2.219774 -0.056412 1.304790 [Debug] dit_step6_xt: [2170, 64] first4: 0.275475 1.820629 -0.166673 0.661042 [DiT] Step 7/50 t=0.880 [Debug] dit_step7_vt: [2170, 64] first4: -0.464647 2.158529 -0.067376 1.284886 [Debug] dit_step7_xt: [2170, 64] first4: 0.284768 1.777458 -0.165325 0.635345 [DiT] Step 8/50 t=0.860 [Debug] dit_step8_vt: [2170, 64] first4: -0.398470 2.068550 -0.076446 1.265460 [Debug] dit_step8_xt: [2170, 64] first4: 0.292737 1.736087 -0.163796 0.610035 [DiT] Step 9/50 t=0.840 [Debug] dit_step9_vt: [2170, 64] first4: -0.329856 1.970060 -0.079246 1.257954 [Debug] dit_step9_xt: [2170, 64] first4: 0.299334 1.696686 -0.162211 0.584876 [DiT] Step 10/50 t=0.820 [Debug] dit_step10_vt: [2170, 64] first4: -0.270426 1.882727 -0.073412 1.245764 [Debug] dit_step10_xt: [2170, 64] first4: 0.304743 1.659031 -0.160743 0.559961 [DiT] Step 11/50 t=0.800 [Debug] dit_step11_vt: [2170, 64] first4: -0.230661 1.821196 -0.070251 1.230668 [Debug] dit_step11_xt: [2170, 64] first4: 0.309356 1.622607 -0.159338 0.535348 [DiT] Step 12/50 t=0.780 [Debug] dit_step12_vt: [2170, 64] first4: -0.202197 1.742254 -0.082008 1.221778 [Debug] dit_step12_xt: [2170, 64] first4: 0.313400 1.587762 -0.157698 0.510912 [DiT] Step 13/50 t=0.760 [Debug] dit_step13_vt: [2170, 64] first4: -0.187175 1.662718 -0.099400 1.211713 [Debug] dit_step13_xt: [2170, 64] first4: 0.317143 1.554508 -0.155710 0.486678 [DiT] Step 14/50 t=0.740 [Debug] dit_step14_vt: [2170, 64] first4: -0.195148 1.586830 -0.127878 1.186867 [Debug] dit_step14_xt: [2170, 64] first4: 0.321046 1.522771 -0.153152 0.462940 [DiT] Step 15/50 t=0.720 [Debug] dit_step15_vt: [2170, 64] first4: -0.185748 1.508590 -0.152303 1.169394 [Debug] dit_step15_xt: [2170, 64] first4: 0.324761 1.492599 -0.150106 0.439553 [DiT] Step 16/50 t=0.700 [Debug] dit_step16_vt: [2170, 64] first4: -0.189975 1.434674 -0.179299 1.146571 [Debug] dit_step16_xt: [2170, 64] first4: 0.328561 1.463906 -0.146520 0.416621 [DiT] Step 17/50 t=0.680 [Debug] dit_step17_vt: [2170, 64] first4: -0.212099 1.360305 -0.225475 1.137322 [Debug] dit_step17_xt: [2170, 64] first4: 0.332803 1.436700 -0.142011 0.393875 [DiT] Step 18/50 t=0.660 [Debug] dit_step18_vt: [2170, 64] first4: -0.217397 1.271699 -0.215731 1.131443 [Debug] dit_step18_xt: [2170, 64] first4: 0.337151 1.411266 -0.137696 0.371246 [DiT] Step 19/50 t=0.640 [Debug] dit_step19_vt: [2170, 64] first4: -0.242558 1.193631 -0.264060 1.086225 [Debug] dit_step19_xt: [2170, 64] first4: 0.342002 1.387393 -0.132415 0.349521 [DiT] Step 20/50 t=0.620 [Debug] dit_step20_vt: [2170, 64] first4: -0.252331 1.112728 -0.259616 1.080052 [Debug] dit_step20_xt: [2170, 64] first4: 0.347048 1.365139 -0.127223 0.327920 [DiT] Step 21/50 t=0.600 [Debug] dit_step21_vt: [2170, 64] first4: -0.283331 1.019205 -0.282729 1.053011 [Debug] dit_step21_xt: [2170, 64] first4: 0.352715 1.344754 -0.121568 0.306860 [DiT] Step 22/50 t=0.580 [Debug] dit_step22_vt: [2170, 64] first4: -0.297697 0.911692 -0.282147 1.031529 [Debug] dit_step22_xt: [2170, 64] first4: 0.358669 1.326521 -0.115925 0.286229 [DiT] Step 23/50 t=0.560 [Debug] dit_step23_vt: [2170, 64] first4: -0.324512 0.810913 -0.288747 0.992839 [Debug] dit_step23_xt: [2170, 64] first4: 0.365159 1.310302 -0.110150 0.266373 [DiT] Step 24/50 t=0.540 [Debug] dit_step24_vt: [2170, 64] first4: -0.358477 0.700207 -0.299200 0.964821 [Debug] dit_step24_xt: [2170, 64] first4: 0.372329 1.296298 -0.104166 0.247076 [DiT] Step 25/50 t=0.520 [Debug] dit_step25_vt: [2170, 64] first4: -0.381883 0.601010 -0.298056 0.951873 [Debug] dit_step25_xt: [2170, 64] first4: 0.379966 1.284278 -0.098205 0.228039 [DiT] Step 26/50 t=0.500 [Debug] dit_step26_vt: [2170, 64] first4: -0.400225 0.467665 -0.287282 0.931475 [Debug] dit_step26_xt: [2170, 64] first4: 0.387971 1.274925 -0.092459 0.209409 [DiT] Step 27/50 t=0.480 [Debug] dit_step27_vt: [2170, 64] first4: -0.415849 0.359334 -0.292716 0.912017 [Debug] dit_step27_xt: [2170, 64] first4: 0.396288 1.267738 -0.086605 0.191169 [DiT] Step 28/50 t=0.460 [Debug] dit_step28_vt: [2170, 64] first4: -0.449091 0.246880 -0.279444 0.894237 [Debug] dit_step28_xt: [2170, 64] first4: 0.405270 1.262801 -0.081016 0.173284 [DiT] Step 29/50 t=0.440 [Debug] dit_step29_vt: [2170, 64] first4: -0.480181 0.129905 -0.265588 0.875099 [Debug] dit_step29_xt: [2170, 64] first4: 0.414873 1.260202 -0.075704 0.155782 [DiT] Step 30/50 t=0.420 [Debug] dit_step30_vt: [2170, 64] first4: -0.482887 0.005804 -0.261820 0.855353 [Debug] dit_step30_xt: [2170, 64] first4: 0.424531 1.260086 -0.070468 0.138675 [DiT] Step 31/50 t=0.400 [Debug] dit_step31_vt: [2170, 64] first4: -0.510049 -0.114843 -0.251913 0.828276 [Debug] dit_step31_xt: [2170, 64] first4: 0.434732 1.262383 -0.065430 0.122110 [DiT] Step 32/50 t=0.380 [Debug] dit_step32_vt: [2170, 64] first4: -0.520728 -0.210531 -0.236669 0.817334 [Debug] dit_step32_xt: [2170, 64] first4: 0.445147 1.266594 -0.060696 0.105763 [DiT] Step 33/50 t=0.360 [Debug] dit_step33_vt: [2170, 64] first4: -0.531485 -0.327356 -0.209233 0.794313 [Debug] dit_step33_xt: [2170, 64] first4: 0.455776 1.273141 -0.056512 0.089877 [DiT] Step 34/50 t=0.340 [Debug] dit_step34_vt: [2170, 64] first4: -0.554491 -0.423499 -0.195805 0.779239 [Debug] dit_step34_xt: [2170, 64] first4: 0.466866 1.281611 -0.052596 0.074292 [DiT] Step 35/50 t=0.320 [Debug] dit_step35_vt: [2170, 64] first4: -0.544407 -0.537084 -0.188109 0.742710 [Debug] dit_step35_xt: [2170, 64] first4: 0.477754 1.292352 -0.048833 0.059438 [DiT] Step 36/50 t=0.300 [Debug] dit_step36_vt: [2170, 64] first4: -0.551673 -0.639949 -0.152098 0.712391 [Debug] dit_step36_xt: [2170, 64] first4: 0.488788 1.305151 -0.045791 0.045190 [DiT] Step 37/50 t=0.280 [Debug] dit_step37_vt: [2170, 64] first4: -0.555337 -0.725967 -0.126220 0.685099 [Debug] dit_step37_xt: [2170, 64] first4: 0.499894 1.319671 -0.043267 0.031488 [DiT] Step 38/50 t=0.260 [Debug] dit_step38_vt: [2170, 64] first4: -0.560736 -0.797745 -0.093956 0.640889 [Debug] dit_step38_xt: [2170, 64] first4: 0.511109 1.335626 -0.041388 0.018670 [DiT] Step 39/50 t=0.240 [Debug] dit_step39_vt: [2170, 64] first4: -0.547260 -0.884292 -0.055918 0.604350 [Debug] dit_step39_xt: [2170, 64] first4: 0.522054 1.353312 -0.040270 0.006583 [DiT] Step 40/50 t=0.220 [Debug] dit_step40_vt: [2170, 64] first4: -0.543024 -0.945219 -0.034627 0.559281 [Debug] dit_step40_xt: [2170, 64] first4: 0.532915 1.372216 -0.039577 -0.004602 [DiT] Step 41/50 t=0.200 [Debug] dit_step41_vt: [2170, 64] first4: -0.527620 -1.012788 -0.003188 0.508355 [Debug] dit_step41_xt: [2170, 64] first4: 0.543467 1.392472 -0.039513 -0.014769 [DiT] Step 42/50 t=0.180 [Debug] dit_step42_vt: [2170, 64] first4: -0.517633 -1.084146 0.048687 0.463533 [Debug] dit_step42_xt: [2170, 64] first4: 0.553820 1.414155 -0.040487 -0.024040 [DiT] Step 43/50 t=0.160 [Debug] dit_step43_vt: [2170, 64] first4: -0.502495 -1.141670 0.085879 0.408264 [Debug] dit_step43_xt: [2170, 64] first4: 0.563870 1.436988 -0.042205 -0.032205 [DiT] Step 44/50 t=0.140 [Debug] dit_step44_vt: [2170, 64] first4: -0.486298 -1.167178 0.124588 0.350221 [Debug] dit_step44_xt: [2170, 64] first4: 0.573596 1.460331 -0.044696 -0.039210 [DiT] Step 45/50 t=0.120 [Debug] dit_step45_vt: [2170, 64] first4: -0.470233 -1.223217 0.154040 0.296348 [Debug] dit_step45_xt: [2170, 64] first4: 0.583000 1.484796 -0.047777 -0.045137 [DiT] Step 46/50 t=0.100 [Debug] dit_step46_vt: [2170, 64] first4: -0.438250 -1.263002 0.191674 0.239982 [Debug] dit_step46_xt: [2170, 64] first4: 0.591765 1.510056 -0.051611 -0.049936 [DiT] Step 47/50 t=0.080 [Debug] dit_step47_vt: [2170, 64] first4: -0.409247 -1.282963 0.216419 0.206215 [Debug] dit_step47_xt: [2170, 64] first4: 0.599950 1.535715 -0.055939 -0.054061 [DiT] Step 48/50 t=0.060 [Debug] dit_step48_vt: [2170, 64] first4: -0.369998 -1.318832 0.229462 0.157051 [Debug] dit_step48_xt: [2170, 64] first4: 0.607350 1.562092 -0.060528 -0.057202 [DiT] Step 49/50 t=0.040 [Debug] dit_step49_vt: [2170, 64] first4: -0.400588 -1.345626 0.197390 0.153967 [Debug] dit_x0: [2170, 64] first4: 0.615362 1.589004 -0.064476 -0.060281 [DiT] Step 50/50 t=0.020 [DiT-Generate] Total: 1616.0 ms (1616.0 ms/sample) [Debug] dit_output: [2170, 64] first4: 0.615362 1.589004 -0.064476 -0.060281 [Store] Unload DiT (3007.9 MB) [GGUF] /mnt/workspace/git/acestep.cpp/models/vae-BF16.gguf: 365 tensors, data at offset 30048 load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] VAE backend: CUDA0 (CPU threads: 16) [VAE] Backend: CUDA0, Weight buffer: 161.1 MB [VAE] Loaded: 5 blocks, upsample=1920x, F32 activations [Store] Load VAE-Dec: 665 ms [VAE] Tiled decode: 3 tiles (chunk=1024, overlap=64, stride=896) [VAE] Graph: 479 nodes, T_latent=960 [VAE] Upsample factor: 1920.00 (expected ~1920) [VAE] Graph: 479 nodes, T_latent=1024 [VAE] Graph: 479 nodes, T_latent=442 [VAE] Tiled decode done: 3 tiles -> T_audio=4166400 (86.80s @ 48kHz) [VAE-Decode Batch0] Decode: 615.1 ms [Debug] vae_audio: [2, 4166400] first4: -0.001537 -0.000900 -0.001131 -0.000862 [Store] Unload VAE-Dec (161.1 MB) [MP3] Encoding 86.8s @ 128 kbps, 48000 Hz stereo [MP3] 1388928 bytes (12.0:1), 396 ms (218.94x realtime), 32 threads [MP3] Wrote ggml-sft/request00.mp3 [Ace-Synth] All done 2026-07-12 10:11:32.494 | INFO | acestep.core.generation.handler.init_service_loader:_load_main_model_from_checkpoint:174 - [initialize_service] Attempting to load model with attention implementation: sdpa 2026-07-12 10:11:42.190 | INFO | acestep.core.generation.handler.generate_music:generate_music:304 - [generate_music] Starting generation... 2026-07-12 10:11:42.190 | INFO | acestep.core.generation.handler.generate_music:generate_music:307 - [generate_music] Preparing inputs... 2026-07-12 10:11:42.190 | INFO | acestep.core.generation.handler.generate_music_request:_prepare_reference_and_source_audio:179 - [generate_music] cover task: no src_audio but audio codes provided, proceeding with codes 2026-07-12 10:11:42.193 | INFO | acestep.core.generation.handler.generate_music:_vram_preflight_check:156 - [generate_music] VRAM pre-flight: 81.69 GB free, ~0.94 GB needed (batch=1, duration=88s, mode=turbo). 2026-07-12 10:11:42.199 | INFO | acestep.core.generation.handler.conditioning_target:_prepare_target_latents_and_wavs:78 - [generate_music] Decoding audio codes for item 0... 2026-07-12 10:11:42.204 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_precomputed_lm_hints:31 - [generate_music] Decoding audio codes for LM hints for item 0... 2026-07-12 10:11:42.206 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:119 - ====================================================================== 2026-07-12 10:11:42.206 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:120 - πŸ” [DEBUG] DiT TEXT ENCODER INPUT (Inference) 2026-07-12 10:11:42.206 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:121 - ====================================================================== 2026-07-12 10:11:42.206 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:122 - text_prompt: # Instruction Generate audio semantic tokens based on the given conditions: # Caption An upbeat and anthemic pop-rock track driven by bright, slightly overdriven # Metas - bpm: 83 - timesignature: 4 - keyscale: G major - duration: 88 seconds <|endoftext|> 2026-07-12 10:11:42.206 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:123 - ====================================================================== 2026-07-12 10:11:42.206 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:124 - lyrics_text: # Languages fr # Lyric # Lyric [Intro - Guitar Riff] [Verse 1] Dans le monde des tutos virtuels G ta toise en nouvelle passion Avec Ggendoline et PumbΓ© Γ  midi La communautΓ©, c'est l'unitΓ© Quel joie, une clΓ© [Chorus] Dans le monde des tutos virtuels GΓ’ndoline et PumbΓ© Γ  midi Une famille Γ  connecter, c'est vrai D'un enfant qui voit toi fusionner [Guitar Solo] [Verse 2] Dans le monde des tutos virtuels GΓ’ndoline, PumbΓ© Γ  midi Une famille Γ  connecter, c'est vrai D'un enfant qui voit toi fusionner<|endoftext|> 2026-07-12 10:11:42.206 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:125 - ====================================================================== 2026-07-12 10:11:42.210 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:110 - [preprocess_batch] Inferring prompt embeddings... 2026-07-12 10:11:42.218 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:113 - [preprocess_batch] Inferring lyric embeddings... 2026-07-12 10:11:42.218 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:161 - [service_generate] Generating audio... (DiT backend: PyTorch (cuda)) 2026-07-12 10:11:42.224 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:263 - [service_generate] DiT diffusion via PyTorch (cuda)... [Request] Loaded request0.json [Turbo] steps=8, shift=3.0 | acestep-v15-turbo-BF16.gguf [GGML] Running acestep-v15-turbo-BF16.gguf... [GGML] Done, 47 dump files [Python] Initializing acestep-v15-turbo... [Python] Generating (acestep-v15-turbo, 8 steps)... Using precomputed LM hints Using precomputed LM hints [Python] Wrote python-turbo/output.wav: 4166400 samples (86.80s @ 48kHz stereo) [Python] Done, 40 dump files [Turbo] Cosine similarities GGML vs Python stage GGML vs Python text_hidden 0.999797 lyric_embed 1.000000 enc_hidden 0.999823 detok_output 0.999996 context 0.999998 noise 1.000000 temb_t 0.999999 hidden_after_proj_in 0.999988 enc_after_cond_emb 0.999812 layer0_sa_output 0.999950 hidden_after_layer0 0.999978 hidden_after_layer6 0.999920 hidden_after_layer12 0.999370 hidden_after_layer18 0.996762 hidden_after_layer23 0.993859 dit_step0_vt 0.976086 dit_step0_xt 0.999947 dit_step1_vt 0.979915 dit_step1_xt 0.999836 dit_step2_vt 0.981870 dit_step2_xt 0.999558 dit_step3_vt 0.982533 dit_step3_xt 0.998935 dit_step4_vt 0.981136 dit_step4_xt 0.997528 dit_step5_vt 0.978563 dit_step5_xt 0.994391 dit_step6_vt 0.975507 dit_step6_xt 0.988435 dit_step7_vt 0.969822 dit_x0 0.979631 vae_audio 0.904109 vae_audio (STFT cosine) 0.976069 [Turbo] Error growth GGML vs Python stage cos max_err mean_err mean_A std_A mean_B std_B dit_step0_xt 0.999947 0.137818 0.006580 -0.002308 0.972940 -0.002342 0.972003 dit_step1_xt 0.999836 0.265643 0.011225 -0.005309 0.942700 -0.005313 0.941730 dit_step2_xt 0.999558 0.452632 0.017410 -0.009349 0.909222 -0.009311 0.908527 dit_step3_xt 0.998935 0.648021 0.025835 -0.014719 0.873854 -0.014577 0.873624 dit_step4_xt 0.997528 0.784556 0.037656 -0.021773 0.842038 -0.021660 0.841995 dit_step5_xt 0.994391 1.212388 0.055229 -0.031857 0.825323 -0.032109 0.824593 dit_step6_xt 0.988435 2.037127 0.081861 -0.046143 0.856229 -0.046482 0.855546 [SFT] steps=50, shift=1.0 | acestep-v15-sft-BF16.gguf [GGML] Running acestep-v15-sft-BF16.gguf... [GGML] Done, 131 dump files [Python] Initializing acestep-v15-sft... [Python] Generating (acestep-v15-sft, 50 steps)... Using precomputed LM hints Using precomputed LM hints 0%| | 0/50 [00:00 Text-Enc [Registry] Qwen3-Embedding-0.6B-Q8_0.gguf -> Text-Enc [Registry] acestep-5Hz-lm-0.6B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-0.6B-Q8_0.gguf -> LM [Registry] acestep-5Hz-lm-1.7B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-1.7B-Q8_0.gguf -> LM [Registry] acestep-5Hz-lm-4B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q5_K_M.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q6_K.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q8_0.gguf -> LM [Registry] acestep-v15-base-BF16.gguf -> DiT [Registry] acestep-v15-base-Q4_K_M.gguf -> DiT [Registry] acestep-v15-base-Q5_K_M.gguf -> DiT [Registry] acestep-v15-base-Q6_K.gguf -> DiT [Registry] acestep-v15-base-Q8_0.gguf -> DiT [Registry] acestep-v15-sft-BF16.gguf -> DiT [Registry] acestep-v15-sft-Q4_K_M.gguf -> DiT [Registry] acestep-v15-sft-Q5_K_M.gguf -> DiT [Registry] acestep-v15-sft-Q6_K.gguf -> DiT [Registry] acestep-v15-sft-Q8_0.gguf -> DiT [Registry] acestep-v15-sftturbo50-BF16.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q4_K_M.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q5_K_M.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q6_K.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-BF16.gguf -> DiT [Registry] acestep-v15-turbo-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-continuous-BF16.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-shift1-BF16.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-shift3-BF16.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-base-BF16.gguf -> DiT [Registry] acestep-v15-xl-base-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-base-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-base-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-base-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-sft-BF16.gguf -> DiT [Registry] acestep-v15-xl-sft-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-sft-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-sft-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-sft-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-BF16.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-turbo-BF16.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q8_0.gguf -> DiT [Registry] scragvae-BF16.gguf -> VAE [Registry] vae-BF16.gguf -> VAE [Store] Created (policy=STRICT) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-turbo-BF16.gguf: 830 tensors, data at offset 69344 [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-turbo-BF16.gguf: 830 tensors, data at offset 69344 [Synth-Load] Ready: turbo=yes, fa=no, batch_cfg=yes [Resolve-Params] max audio codes across batch: 434 (86.8s @ 5Hz) [Resolve-T] T=2170, S=1085 [Resolve-T] seed=42, steps=8, guidance=1.0, shift=3.0, duration=88.0s [BPE] Loaded from GGUF: 151643 vocab, 151387 merges ggml_cuda_init: found 1 CUDA devices (Total VRAM: 97247 MiB): Device 0: NVIDIA RTX PRO 6000 Blackwell Workstation Edition, compute capability 12.0, VMM: yes, VRAM: 97247 MiB load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = NVIDIA RTX PRO 6000 Blackwell Workstation Edition (NVIDIA) | uma: 0 | fp16: 1 | bf16: 0 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: NV_coopmat2 load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] TextEncoder backend: CUDA0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/Qwen3-Embedding-0.6B-BF16.gguf: 310 tensors, data at offset 5337568 [Load] TextEncoder: 28L, H=1024, Nh=16/8 [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 310 tensors, 1136.5 MB into backend [Store] Load TextEnc: 270 ms [Debug] text_hidden: [70, 1024] first4: 3.652379 1.048645 0.229843 -13.063987 [Debug] lyric_embed: [167, 1024] first4: 0.029175 0.032227 -0.022339 -0.028809 [Store] Unload TextEnc (1136.5 MB) load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] CondEncoder backend: CUDA0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-turbo-BF16.gguf: 830 tensors, data at offset 69344 [Load] LyricEncoder: 8L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [Load] TimbreEncoder: 4L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 141 tensors, 1160.5 MB into backend [Load] CondEncoder: lyric(8L), timbre(4L, CLS), text_proj, null_cond [Store] Load CondEnc: 540 ms [CondEnc] Lyric sliding mask: 167x167, window=128 [CondEnc] Timbre sliding mask: 2x2, window=128 (CLS) [Encode] Packed: lyric=167 + timbre=1 + text=70 = 238 tokens [Encode-Text Batch0] 70+167 tokens -> enc_S=238, 11.7 ms [Debug] enc_hidden: [238, 2048] first4: 0.685622 -0.018933 -0.051580 0.022790 [Store] Unload CondEnc (1160.5 MB) load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] Detokenizer backend: CUDA0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-turbo-BF16.gguf: 830 tensors, data at offset 69344 [WeightCtx] Loaded 30 tensors, 200.3 MB into backend [Load] Detokenizer: FSQ(6->2048) + 2L encoder(S=5, 2048->64) [Store] Load FSQ-Detok: 89 ms CUDA Graph id 8 reused ggml_backend_cuda_graph_compute: CUDA graph warmup complete CUDA Graph id 8 reused [Context] Decoded: 434 codes -> 2170 frames (86.8s @ 25Hz) [Build-Context Batch0] Detokenizer: 104.2 ms, 434 codes [Debug] detok_output: [2170, 64] first4: -0.124076 1.434873 0.309935 -0.624868 [Store] Unload FSQ-Detok (200.3 MB) [Init-Noise Batch0] Philox noise seed=42, [2170, 64] solver=euler [Debug] noise: [2170, 64] first4: 0.194336 2.156250 -0.171875 0.847656 [Debug] context: [2170, 128] first4: -0.124076 1.434873 0.309935 -0.624868 [Init-Noise] Starting: T=2170, S=1085, enc_S=238, steps=8, batch=1 (cover) [Dedup] Previous line repeated 432 times total load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] DiT backend: CUDA0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-turbo-BF16.gguf: 830 tensors, data at offset 69344 [DiT] Self-attn: Q+K+V fused [DiT] Cross-attn: Q+K+V fused [DiT] MLP: gate+up fused [Load] null_condition_emb found (CFG available) [WeightCtx] Loaded 630 tensors, 7955.6 MB into backend [Load] DiT: 32 layers, H=2560, Nh=32/8, D=128 [Store] Load DiT: 4347 ms [DiT] Batch N=1, T=2170, S=1085, enc_S=238 [DiT] Graph: 2825 nodes [DiT] Solver: ODE Euler (1 NFE/step, order 1) [Debug] tproj: [15360] first4: 0.016458 0.108603 0.065523 -0.074042 [Debug] temb: [2560] first4: 0.026321 -0.023010 -0.016508 0.047175 [Debug] temb_t: [2560] first4: 0.014560 -0.022321 -0.020859 0.048784 [Debug] temb_r: [2560] first4: 0.011760 -0.000690 0.004351 -0.001609 [Debug] sinusoidal_t: [256] first4: 0.562486 0.789701 0.439822 -0.023583 [Debug] sinusoidal_r: [256] first4: 1.000000 1.000000 1.000000 1.000000 [Debug] temb_lin1_t: [2560] first4: 0.064010 -0.019467 0.138295 -0.113216 [Debug] temb_lin1_r: [2560] first4: -0.009879 0.015926 -0.017671 0.012016 [Debug] hidden_after_proj_in: [2560, 1085] first4: 0.336700 -0.577220 0.230202 -0.190902 [Debug] proj_in_input: [192, 2170] first4: -0.124076 1.434873 0.309935 -0.624868 [Debug] enc_after_cond_emb: [2560, 238] first4: 0.030151 0.654388 0.102570 0.872589 [Debug] layer0_sa_input: [2560, 1085] first4: 0.039990 -0.045584 0.094020 -0.134746 [Debug] layer0_q_after_rope: [128, 32] first4: -0.244144 -2.067986 0.283550 0.221414 [Debug] layer0_k_after_rope: [128, 8] first4: 0.024658 2.140625 0.730469 0.500000 [Debug] layer0_sa_output: [2560, 1085] first4: 1.671875 4.312500 -0.546875 5.437500 [Debug] layer0_attn_out: [4096, 1085] first4: 0.469895 -2.249235 -0.169631 -2.803040 [Debug] layer0_after_self_attn: [2560, 1085] first4: 0.469895 -2.249235 -0.169631 -2.803040 [Debug] layer0_after_cross_attn: [2560, 1085] first4: 0.338059 -2.004117 0.001267 -3.744447 [Debug] hidden_after_layer0: [2560, 1085] first4: 1.708741 -2.442633 242.353226 -3.716172 [Debug] hidden_after_layer6: [2560, 1085] first4: -6.204265 -2.193405 321.462402 -12.557486 [Debug] hidden_after_layer12: [2560, 1085] first4: -11.131212 29.302008 265.192352 28.704075 [Debug] hidden_after_layer18: [2560, 1085] first4: 42.067879 14.732556 279.249481 4.941826 [Debug] hidden_after_layer31: [2560, 1085] first4: -80.098793 -245.191711 -164.770691 186.469833 [Debug] dit_step0_vt: [2170, 64] first4: -1.246463 1.518559 0.394062 1.160819 [Debug] dit_step0_xt: [2170, 64] first4: 0.250993 2.087224 -0.189787 0.794892 [DiT] Step 1/8 t=1.000 ggml_backend_cuda_graph_compute: CUDA graph warmup complete [Debug] dit_step1_vt: [2170, 64] first4: -0.732804 2.505704 0.821889 1.114906 [Debug] dit_step1_xt: [2170, 64] first4: 0.290964 1.950550 -0.234617 0.734079 [DiT] Step 2/8 t=0.955 [Debug] dit_step2_vt: [2170, 64] first4: -1.133526 2.182483 0.953867 1.049844 [Debug] dit_step2_xt: [2170, 64] first4: 0.366533 1.805051 -0.298208 0.664089 [DiT] Step 3/8 t=0.900 [Debug] dit_step3_vt: [2170, 64] first4: -1.240928 2.303751 1.144955 0.553371 [Debug] dit_step3_xt: [2170, 64] first4: 0.469944 1.613072 -0.393621 0.617975 [DiT] Step 4/8 t=0.833 [Debug] dit_step4_vt: [2170, 64] first4: -1.149632 2.251340 1.166546 0.225735 [Debug] dit_step4_xt: [2170, 64] first4: 0.593118 1.371857 -0.518608 0.593789 [DiT] Step 5/8 t=0.750 [Debug] dit_step5_vt: [2170, 64] first4: -1.133823 2.250499 0.986205 1.019715 [Debug] dit_step5_xt: [2170, 64] first4: 0.755093 1.050357 -0.659495 0.448115 [DiT] Step 6/8 t=0.643 [Debug] dit_step6_vt: [2170, 64] first4: -0.972937 1.988650 0.639465 0.825770 [Debug] dit_step6_xt: [2170, 64] first4: 0.949681 0.652627 -0.787388 0.282961 [DiT] Step 7/8 t=0.500 [Debug] dit_step7_vt: [2170, 64] first4: -1.029946 1.507111 0.653788 0.531172 [Debug] dit_x0: [2170, 64] first4: 1.258664 0.200493 -0.983524 0.123610 [DiT] Step 8/8 t=0.300 [DiT-Generate] Total: 668.3 ms (668.3 ms/sample) [Debug] dit_output: [2170, 64] first4: 1.258664 0.200493 -0.983524 0.123610 [Store] Unload DiT (7955.6 MB) [GGUF] /mnt/workspace/git/acestep.cpp/models/vae-BF16.gguf: 365 tensors, data at offset 30048 load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] VAE backend: CUDA0 (CPU threads: 16) [VAE] Backend: CUDA0, Weight buffer: 161.1 MB [VAE] Loaded: 5 blocks, upsample=1920x, F32 activations [Store] Load VAE-Dec: 651 ms [VAE] Tiled decode: 3 tiles (chunk=1024, overlap=64, stride=896) [VAE] Graph: 479 nodes, T_latent=960 [VAE] Upsample factor: 1920.00 (expected ~1920) [VAE] Graph: 479 nodes, T_latent=1024 [VAE] Graph: 479 nodes, T_latent=442 [VAE] Tiled decode done: 3 tiles -> T_audio=4166400 (86.80s @ 48kHz) [VAE-Decode Batch0] Decode: 630.9 ms [Debug] vae_audio: [2, 4166400] first4: 0.001822 0.004719 0.004745 0.003986 [Store] Unload VAE-Dec (161.1 MB) [MP3] Encoding 86.8s @ 128 kbps, 48000 Hz stereo [MP3] 1388928 bytes (12.0:1), 468 ms (185.37x realtime), 32 threads [MP3] Wrote ggml-xl-turbo/request00.mp3 [Ace-Synth] All done 2026-07-12 10:11:53.415 | INFO | acestep.core.generation.handler.init_service_loader:_load_main_model_from_checkpoint:174 - [initialize_service] Attempting to load model with attention implementation: sdpa [Python] Wrote python-sft/output.wav: 4166400 samples (86.80s @ 48kHz stereo) [Python] Done, 166 dump files [SFT] Cosine similarities GGML vs Python stage GGML vs Python text_hidden 0.999797 lyric_embed 1.000000 enc_hidden 0.999823 detok_output 0.999996 context 0.999998 noise 1.000000 temb_t 0.999997 hidden_after_proj_in 0.999988 enc_after_cond_emb 0.999813 layer0_sa_output 0.999941 hidden_after_layer0 0.999978 hidden_after_layer6 0.999852 hidden_after_layer12 0.999478 hidden_after_layer18 0.998700 hidden_after_layer23 0.998984 dit_step0_vt 0.998939 dit_step0_xt 1.000000 dit_step5_vt 0.999519 dit_step5_xt 0.999981 dit_step10_vt 0.999178 dit_step10_xt 0.999938 dit_step15_vt 0.998320 dit_step15_xt 0.999857 dit_step20_vt 0.996618 dit_step20_xt 0.999643 dit_step25_vt 0.993376 dit_step25_xt 0.999200 dit_step30_vt 0.989240 dit_step30_xt 0.998369 dit_step35_vt 0.984833 dit_step35_xt 0.997194 dit_step40_vt 0.982210 dit_step40_xt 0.995919 dit_step45_vt 0.985444 dit_step45_xt 0.994969 dit_step49_vt 0.990479 dit_x0 0.994634 vae_audio 0.970580 vae_audio (STFT cosine) 0.988233 [SFT] Error growth GGML vs Python stage cos max_err mean_err mean_A std_A mean_B std_B dit_step0_xt 1.000000 0.020084 0.001453 -0.001655 0.978459 -0.001619 0.978872 dit_step5_xt 0.999981 0.077753 0.004220 -0.008386 0.884766 -0.008480 0.883811 dit_step10_xt 0.999938 0.117825 0.006671 -0.015115 0.800909 -0.015075 0.802324 dit_step15_xt 0.999857 0.163803 0.009333 -0.021815 0.730457 -0.021933 0.731304 dit_step20_xt 0.999643 0.224397 0.013631 -0.028555 0.677686 -0.028882 0.679066 dit_step25_xt 0.999200 0.330851 0.019336 -0.035370 0.647041 -0.035643 0.650780 dit_step30_xt 0.998369 0.563946 0.026939 -0.042225 0.641798 -0.042577 0.647107 dit_step35_xt 0.997194 0.819768 0.036056 -0.049089 0.662661 -0.049471 0.669256 dit_step40_xt 0.995919 1.074785 0.046244 -0.055981 0.707262 -0.056532 0.715783 dit_step45_xt 0.994969 1.302280 0.055871 -0.062975 0.771253 -0.063533 0.780575 [Xl-turbo] steps=8, shift=3.0 | acestep-v15-xl-turbo-BF16.gguf [GGML] Running acestep-v15-xl-turbo-BF16.gguf... [GGML] Done, 47 dump files [Python] Initializing acestep-v15-xl-turbo... Loading checkpoint shards: 0%| | 0/4 [00:00 1.0 (turbo does not use CFG). 2026-07-12 10:11:57.000 | INFO | acestep.core.generation.handler.generate_music:generate_music:304 - [generate_music] Starting generation... 2026-07-12 10:11:57.000 | INFO | acestep.core.generation.handler.generate_music:generate_music:307 - [generate_music] Preparing inputs... 2026-07-12 10:11:57.001 | INFO | acestep.core.generation.handler.generate_music_request:_prepare_reference_and_source_audio:179 - [generate_music] cover task: no src_audio but audio codes provided, proceeding with codes 2026-07-12 10:11:57.007 | INFO | acestep.core.generation.handler.generate_music:_vram_preflight_check:156 - [generate_music] VRAM pre-flight: 76.76 GB free, ~1.23 GB needed (batch=1, duration=88s, mode=xl_turbo). 2026-07-12 10:11:57.011 | INFO | acestep.core.generation.handler.conditioning_target:_prepare_target_latents_and_wavs:78 - [generate_music] Decoding audio codes for item 0... 2026-07-12 10:11:57.016 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_precomputed_lm_hints:31 - [generate_music] Decoding audio codes for LM hints for item 0... 2026-07-12 10:11:57.018 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:119 - ====================================================================== 2026-07-12 10:11:57.018 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:120 - πŸ” [DEBUG] DiT TEXT ENCODER INPUT (Inference) 2026-07-12 10:11:57.018 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:121 - ====================================================================== 2026-07-12 10:11:57.018 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:122 - text_prompt: # Instruction Generate audio semantic tokens based on the given conditions: # Caption An upbeat and anthemic pop-rock track driven by bright, slightly overdriven # Metas - bpm: 83 - timesignature: 4 - keyscale: G major - duration: 88 seconds <|endoftext|> 2026-07-12 10:11:57.018 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:123 - ====================================================================== 2026-07-12 10:11:57.018 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:124 - lyrics_text: # Languages fr # Lyric # Lyric [Intro - Guitar Riff] [Verse 1] Dans le monde des tutos virtuels G ta toise en nouvelle passion Avec Ggendoline et PumbΓ© Γ  midi La communautΓ©, c'est l'unitΓ© Quel joie, une clΓ© [Chorus] Dans le monde des tutos virtuels GΓ’ndoline et PumbΓ© Γ  midi Une famille Γ  connecter, c'est vrai D'un enfant qui voit toi fusionner [Guitar Solo] [Verse 2] Dans le monde des tutos virtuels GΓ’ndoline, PumbΓ© Γ  midi Une famille Γ  connecter, c'est vrai D'un enfant qui voit toi fusionner<|endoftext|> 2026-07-12 10:11:57.018 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:125 - ====================================================================== 2026-07-12 10:11:57.022 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:110 - [preprocess_batch] Inferring prompt embeddings... 2026-07-12 10:11:57.029 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:113 - [preprocess_batch] Inferring lyric embeddings... 2026-07-12 10:11:57.029 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:161 - [service_generate] Generating audio... (DiT backend: PyTorch (cuda)) 2026-07-12 10:11:57.036 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:263 - [service_generate] DiT diffusion via PyTorch (cuda)... 2026-07-12 10:11:57.042 | INFO | acestep.models.common.dcw_correction:__init__:98 - [DCW] Active β€” mode=double, scaler=0.0500, high_scaler=0.0200, wavelet='haar' 2026-07-12 10:11:57.598 | INFO | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:42 - [generate_music] Model generation completed. Decoding latents... 2026-07-12 10:11:57.599 | DEBUG | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:64 - [generate_music] pred_latents: torch.Size([1, 2170, 64]), dtype=torch.bfloat16 2026-07-12 10:11:57.599 | DEBUG | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:65 - [generate_music] time_costs: {'encoder_time_cost': 0.006384134292602539, 'diffusion_time_cost': 0.5552432537078857, 'diffusion_per_step_time_cost': 0.06940540671348572, 'total_time_cost': 0.5616273880004883, 'offload_time_cost': 0.0} 2026-07-12 10:11:57.622 | INFO | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:127 - [generate_music] Decoding latents with VAE... 2026-07-12 10:11:57.623 | DEBUG | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:136 - [generate_music] Before VAE decode: allocated=17.03GB, max=18.33GB 2026-07-12 10:11:57.623 | INFO | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:154 - [generate_music] Effective free VRAM before VAE decode: 76.67 GB 2026-07-12 10:11:57.623 | INFO | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:172 - [generate_music] Using tiled VAE decode to reduce VRAM usage... 2026-07-12 10:11:57.623 | DEBUG | acestep.core.generation.handler.memory_utils:_get_auto_decode_chunk_size:108 - [_get_auto_decode_chunk_size] Effective free VRAM: 76.67 GB 2026-07-12 10:11:57.624 | DEBUG | acestep.core.generation.handler.memory_utils:_should_offload_wav_to_cpu:131 - [_should_offload_wav_to_cpu] Effective free VRAM: 76.67 GB 2026-07-12 10:11:57.624 | INFO | acestep.core.generation.handler.vae_decode:tiled_decode:56 - [tiled_decode] chunk_size=512, offload_wav_to_cpu=False, latents_shape=torch.Size([1, 64, 2170]) 2026-07-12 10:11:57.895 | DEBUG | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:193 - [generate_music] After VAE decode: allocated=17.20GB, max=18.49GB 2026-07-12 10:11:58.097 | INFO | acestep.core.generation.handler.generate_music_payload:_build_generate_music_success_payload:38 - [generate_music] VAE decode completed. Preparing audio tensors... 2026-07-12 10:11:58.101 | INFO | acestep.core.generation.handler.generate_music_payload:_build_generate_music_success_payload:50 - [generate_music] Done! Generated 1 audio tensors. [Request] Parsed ggml-xl-sft/request0.json [Request] seed=42 lm_batch=1 synth_batch=1 [Request] caption: An upbeat and anthemic pop-rock track driven by bright, slig... [Request] lyrics: 490 bytes [Request] bpm=83 dur=88 key=G major ts=4 lang=fr [Request] lm: temp=0.85 cfg=2.0 top_p=0.90 top_k=0 [Request] dit: steps=50 guidance=1.0 shift=1.0 [Request] task_type: text2music [Request] solver: euler (stork_substeps=10) [Request] lm_mode: generate [Request] output_format: mp3 [Request] synth_model: acestep-v15-xl-sft-BF16.gguf [Request] vae: vae-BF16.gguf [Request] audio_codes: (present) [Ace-Synth] Batch: 1 request(s) [Registry] Qwen3-Embedding-0.6B-BF16.gguf -> Text-Enc [Registry] Qwen3-Embedding-0.6B-Q8_0.gguf -> Text-Enc [Registry] acestep-5Hz-lm-0.6B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-0.6B-Q8_0.gguf -> LM [Registry] acestep-5Hz-lm-1.7B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-1.7B-Q8_0.gguf -> LM [Registry] acestep-5Hz-lm-4B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q5_K_M.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q6_K.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q8_0.gguf -> LM [Registry] acestep-v15-base-BF16.gguf -> DiT [Registry] acestep-v15-base-Q4_K_M.gguf -> DiT [Registry] acestep-v15-base-Q5_K_M.gguf -> DiT [Registry] acestep-v15-base-Q6_K.gguf -> DiT [Registry] acestep-v15-base-Q8_0.gguf -> DiT [Registry] acestep-v15-sft-BF16.gguf -> DiT [Registry] acestep-v15-sft-Q4_K_M.gguf -> DiT [Registry] acestep-v15-sft-Q5_K_M.gguf -> DiT [Registry] acestep-v15-sft-Q6_K.gguf -> DiT [Registry] acestep-v15-sft-Q8_0.gguf -> DiT [Registry] acestep-v15-sftturbo50-BF16.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q4_K_M.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q5_K_M.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q6_K.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-BF16.gguf -> DiT [Registry] acestep-v15-turbo-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-continuous-BF16.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-shift1-BF16.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-shift3-BF16.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-base-BF16.gguf -> DiT [Registry] acestep-v15-xl-base-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-base-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-base-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-base-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-sft-BF16.gguf -> DiT [Registry] acestep-v15-xl-sft-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-sft-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-sft-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-sft-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-BF16.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-turbo-BF16.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q8_0.gguf -> DiT [Registry] scragvae-BF16.gguf -> VAE [Registry] vae-BF16.gguf -> VAE [Store] Created (policy=STRICT) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-sft-BF16.gguf: 830 tensors, data at offset 69312 [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-sft-BF16.gguf: 830 tensors, data at offset 69312 [Synth-Load] Ready: turbo=no, fa=no, batch_cfg=yes [Resolve-Params] max audio codes across batch: 434 (86.8s @ 5Hz) [Resolve-T] T=2170, S=1085 [Resolve-T] seed=42, steps=50, guidance=1.0, shift=1.0, duration=88.0s [BPE] Loaded from GGUF: 151643 vocab, 151387 merges ggml_cuda_init: found 1 CUDA devices (Total VRAM: 97247 MiB): Device 0: NVIDIA RTX PRO 6000 Blackwell Workstation Edition, compute capability 12.0, VMM: yes, VRAM: 97247 MiB load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = NVIDIA RTX PRO 6000 Blackwell Workstation Edition (NVIDIA) | uma: 0 | fp16: 1 | bf16: 0 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: NV_coopmat2 load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] TextEncoder backend: CUDA0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/Qwen3-Embedding-0.6B-BF16.gguf: 310 tensors, data at offset 5337568 [Load] TextEncoder: 28L, H=1024, Nh=16/8 [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 310 tensors, 1136.5 MB into backend [Store] Load TextEnc: 285 ms [Debug] text_hidden: [70, 1024] first4: 3.652379 1.048645 0.229843 -13.063987 [Debug] lyric_embed: [167, 1024] first4: 0.029175 0.032227 -0.022339 -0.028809 [Store] Unload TextEnc (1136.5 MB) load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] CondEncoder backend: CUDA0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-sft-BF16.gguf: 830 tensors, data at offset 69312 [Load] LyricEncoder: 8L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [Load] TimbreEncoder: 4L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 141 tensors, 1160.5 MB into backend [Load] CondEncoder: lyric(8L), timbre(4L, CLS), text_proj, null_cond [Store] Load CondEnc: 539 ms [CondEnc] Lyric sliding mask: 167x167, window=128 [CondEnc] Timbre sliding mask: 2x2, window=128 (CLS) [Encode] Packed: lyric=167 + timbre=1 + text=70 = 238 tokens [Encode-Text Batch0] 70+167 tokens -> enc_S=238, 11.3 ms [Debug] enc_hidden: [238, 2048] first4: 0.685622 -0.018933 -0.051580 0.022790 [Store] Unload CondEnc (1160.5 MB) load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] Detokenizer backend: CUDA0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-sft-BF16.gguf: 830 tensors, data at offset 69312 [WeightCtx] Loaded 30 tensors, 200.3 MB into backend [Load] Detokenizer: FSQ(6->2048) + 2L encoder(S=5, 2048->64) [Store] Load FSQ-Detok: 93 ms CUDA Graph id 8 reused ggml_backend_cuda_graph_compute: CUDA graph warmup complete CUDA Graph id 8 reused [Context] Decoded: 434 codes -> 2170 frames (86.8s @ 25Hz) [Build-Context Batch0] Detokenizer: 124.4 ms, 434 codes [Debug] detok_output: [2170, 64] first4: -0.124076 1.434873 0.309935 -0.624868 [Store] Unload FSQ-Detok (200.3 MB) [Init-Noise Batch0] Philox noise seed=42, [2170, 64] solver=euler [Debug] noise: [2170, 64] first4: 0.194336 2.156250 -0.171875 0.847656 [Debug] context: [2170, 128] first4: -0.124076 1.434873 0.309935 -0.624868 [Init-Noise] Starting: T=2170, S=1085, enc_S=238, steps=50, batch=1 (cover) [Dedup] Previous line repeated 432 times total load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] DiT backend: CUDA0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-sft-BF16.gguf: 830 tensors, data at offset 69312 [DiT] Self-attn: Q+K+V fused [DiT] Cross-attn: Q+K+V fused [DiT] MLP: gate+up fused [Load] null_condition_emb found (CFG available) [WeightCtx] Loaded 630 tensors, 7955.6 MB into backend [Load] DiT: 32 layers, H=2560, Nh=32/8, D=128 [Store] Load DiT: 3968 ms [DiT] Batch N=1, T=2170, S=1085, enc_S=238 [DiT] Graph: 2825 nodes [DiT] Solver: ODE Euler (1 NFE/step, order 1) [Debug] tproj: [15360] first4: 0.012546 0.105322 0.073633 -0.031913 [Debug] temb: [2560] first4: 0.009488 -0.007695 -0.006104 0.004570 [Debug] temb_t: [2560] first4: 0.007949 -0.007936 -0.006669 0.003595 [Debug] temb_r: [2560] first4: 0.001539 0.000241 0.000566 0.000975 [Debug] sinusoidal_t: [256] first4: 0.562486 0.789701 0.439822 -0.023583 [Debug] sinusoidal_r: [256] first4: 1.000000 1.000000 1.000000 1.000000 [Debug] temb_lin1_t: [2560] first4: 0.032232 -0.081851 0.088236 -0.064962 [Debug] temb_lin1_r: [2560] first4: 0.006956 -0.005390 -0.002806 -0.011722 [Debug] hidden_after_proj_in: [2560, 1085] first4: 0.326133 -0.492742 0.101310 -0.214511 [Debug] proj_in_input: [192, 2170] first4: -0.124076 1.434873 0.309935 -0.624868 [Debug] enc_after_cond_emb: [2560, 238] first4: 0.034821 0.643005 0.107605 0.873055 [Debug] layer0_sa_input: [2560, 1085] first4: 0.036269 -0.018961 0.098650 -0.097624 [Debug] layer0_q_after_rope: [128, 32] first4: -0.278736 -2.132909 0.214724 -0.008532 [Debug] layer0_k_after_rope: [128, 8] first4: 0.738281 2.906250 1.171875 2.046875 [Debug] layer0_sa_output: [2560, 1085] first4: 1.945312 4.406250 -0.808594 7.843750 [Debug] layer0_attn_out: [4096, 1085] first4: 0.453824 -2.182210 -0.499636 -4.120116 [Debug] layer0_after_self_attn: [2560, 1085] first4: 0.453824 -2.182210 -0.499636 -4.120116 [Debug] layer0_after_cross_attn: [2560, 1085] first4: 0.289761 -1.912679 -0.415652 -4.224120 [Debug] hidden_after_layer0: [2560, 1085] first4: -1.721187 -1.255613 247.916580 -3.996658 [Debug] hidden_after_layer6: [2560, 1085] first4: -7.834574 3.589973 314.930420 -21.910683 [Debug] hidden_after_layer12: [2560, 1085] first4: -23.644529 39.430508 231.298721 16.490387 [Debug] hidden_after_layer18: [2560, 1085] first4: 9.473307 -12.984356 232.645172 19.724997 [Debug] hidden_after_layer31: [2560, 1085] first4: 62.285324 -107.379547 4.018387 83.120201 [Debug] dit_step0_vt: [2170, 64] first4: -0.316202 2.209554 0.016607 1.703432 [Debug] dit_step0_xt: [2170, 64] first4: 0.200660 2.112059 -0.172207 0.813588 [DiT] Step 1/50 t=1.000 ggml_backend_cuda_graph_compute: CUDA graph warmup complete [Debug] dit_step1_vt: [2170, 64] first4: -0.343204 2.130481 0.072172 1.557741 [Debug] dit_step1_xt: [2170, 64] first4: 0.207524 2.069449 -0.173651 0.782433 [DiT] Step 2/50 t=0.980 [Debug] dit_step2_vt: [2170, 64] first4: -0.323388 2.172076 0.050186 1.584516 [Debug] dit_step2_xt: [2170, 64] first4: 0.213992 2.026008 -0.174654 0.750742 [DiT] Step 3/50 t=0.960 [Debug] dit_step3_vt: [2170, 64] first4: -0.278304 2.133125 0.037878 1.578382 [Debug] dit_step3_xt: [2170, 64] first4: 0.219558 1.983345 -0.175412 0.719175 [DiT] Step 4/50 t=0.940 [Debug] dit_step4_vt: [2170, 64] first4: -0.267278 2.069202 0.015467 1.591200 [Debug] dit_step4_xt: [2170, 64] first4: 0.224903 1.941961 -0.175721 0.687351 [DiT] Step 5/50 t=0.920 [Debug] dit_step5_vt: [2170, 64] first4: -0.263141 2.050870 0.022551 1.596046 [Debug] dit_step5_xt: [2170, 64] first4: 0.230166 1.900944 -0.176172 0.655430 [DiT] Step 6/50 t=0.900 [Debug] dit_step6_vt: [2170, 64] first4: -0.251216 2.023096 0.060205 1.618396 [Debug] dit_step6_xt: [2170, 64] first4: 0.235191 1.860482 -0.177376 0.623062 [DiT] Step 7/50 t=0.880 [Debug] dit_step7_vt: [2170, 64] first4: -0.250367 1.985753 0.107183 1.622868 [Debug] dit_step7_xt: [2170, 64] first4: 0.240198 1.820767 -0.179520 0.590605 [DiT] Step 8/50 t=0.860 [Debug] dit_step8_vt: [2170, 64] first4: -0.262136 1.932814 0.132220 1.619447 [Debug] dit_step8_xt: [2170, 64] first4: 0.245441 1.782110 -0.182164 0.558216 [DiT] Step 9/50 t=0.840 [Debug] dit_step9_vt: [2170, 64] first4: -0.293373 1.881742 0.120795 1.615607 [Debug] dit_step9_xt: [2170, 64] first4: 0.251308 1.744475 -0.184580 0.525904 [DiT] Step 10/50 t=0.820 [Debug] dit_step10_vt: [2170, 64] first4: -0.296031 1.837028 0.113188 1.602298 [Debug] dit_step10_xt: [2170, 64] first4: 0.257229 1.707735 -0.186844 0.493858 [DiT] Step 11/50 t=0.800 [Debug] dit_step11_vt: [2170, 64] first4: -0.291505 1.785168 0.100330 1.587226 [Debug] dit_step11_xt: [2170, 64] first4: 0.263059 1.672032 -0.188851 0.462113 [DiT] Step 12/50 t=0.780 [Debug] dit_step12_vt: [2170, 64] first4: -0.287889 1.743116 0.091006 1.586907 [Debug] dit_step12_xt: [2170, 64] first4: 0.268817 1.637169 -0.190671 0.430375 [DiT] Step 13/50 t=0.760 [Debug] dit_step13_vt: [2170, 64] first4: -0.284759 1.682417 0.072042 1.565709 [Debug] dit_step13_xt: [2170, 64] first4: 0.274512 1.603521 -0.192112 0.399061 [DiT] Step 14/50 t=0.740 [Debug] dit_step14_vt: [2170, 64] first4: -0.278153 1.631876 0.059444 1.565421 [Debug] dit_step14_xt: [2170, 64] first4: 0.280075 1.570883 -0.193300 0.367752 [DiT] Step 15/50 t=0.720 [Debug] dit_step15_vt: [2170, 64] first4: -0.261671 1.582477 0.032380 1.561401 [Debug] dit_step15_xt: [2170, 64] first4: 0.285308 1.539234 -0.193948 0.336524 [DiT] Step 16/50 t=0.700 [Debug] dit_step16_vt: [2170, 64] first4: -0.244664 1.540197 0.015157 1.554451 [Debug] dit_step16_xt: [2170, 64] first4: 0.290202 1.508430 -0.194251 0.305435 [DiT] Step 17/50 t=0.680 [Debug] dit_step17_vt: [2170, 64] first4: -0.222339 1.488901 -0.002147 1.539411 [Debug] dit_step17_xt: [2170, 64] first4: 0.294648 1.478652 -0.194208 0.274647 [DiT] Step 18/50 t=0.660 [Debug] dit_step18_vt: [2170, 64] first4: -0.184409 1.442553 -0.011353 1.540462 [Debug] dit_step18_xt: [2170, 64] first4: 0.298337 1.449801 -0.193981 0.243838 [DiT] Step 19/50 t=0.640 [Debug] dit_step19_vt: [2170, 64] first4: -0.162454 1.392205 -0.021457 1.525675 [Debug] dit_step19_xt: [2170, 64] first4: 0.301586 1.421957 -0.193552 0.213324 [DiT] Step 20/50 t=0.620 [Debug] dit_step20_vt: [2170, 64] first4: -0.119116 1.348247 -0.029730 1.518041 [Debug] dit_step20_xt: [2170, 64] first4: 0.303968 1.394992 -0.192957 0.182964 [DiT] Step 21/50 t=0.600 [Debug] dit_step21_vt: [2170, 64] first4: -0.091729 1.289829 -0.037363 1.503773 [Debug] dit_step21_xt: [2170, 64] first4: 0.305803 1.369195 -0.192210 0.152888 [DiT] Step 22/50 t=0.580 [Debug] dit_step22_vt: [2170, 64] first4: -0.070102 1.241824 -0.049521 1.496528 [Debug] dit_step22_xt: [2170, 64] first4: 0.307205 1.344359 -0.191220 0.122957 [DiT] Step 23/50 t=0.560 [Debug] dit_step23_vt: [2170, 64] first4: -0.050569 1.191121 -0.063131 1.485620 [Debug] dit_step23_xt: [2170, 64] first4: 0.308216 1.320536 -0.189957 0.093245 [DiT] Step 24/50 t=0.540 [Debug] dit_step24_vt: [2170, 64] first4: -0.041034 1.137858 -0.071927 1.475013 [Debug] dit_step24_xt: [2170, 64] first4: 0.309037 1.297779 -0.188519 0.063745 [DiT] Step 25/50 t=0.520 [Debug] dit_step25_vt: [2170, 64] first4: -0.029203 1.082175 -0.078827 1.440275 [Debug] dit_step25_xt: [2170, 64] first4: 0.309621 1.276136 -0.186942 0.034939 [DiT] Step 26/50 t=0.500 [Debug] dit_step26_vt: [2170, 64] first4: -0.026016 1.028872 -0.089483 1.424736 [Debug] dit_step26_xt: [2170, 64] first4: 0.310141 1.255558 -0.185152 0.006445 [DiT] Step 27/50 t=0.480 [Debug] dit_step27_vt: [2170, 64] first4: -0.032095 0.958167 -0.099371 1.404521 [Debug] dit_step27_xt: [2170, 64] first4: 0.310783 1.236395 -0.183165 -0.021646 [DiT] Step 28/50 t=0.460 [Debug] dit_step28_vt: [2170, 64] first4: -0.041671 0.897095 -0.111367 1.387364 [Debug] dit_step28_xt: [2170, 64] first4: 0.311616 1.218453 -0.180938 -0.049393 [DiT] Step 29/50 t=0.440 [Debug] dit_step29_vt: [2170, 64] first4: -0.048523 0.825854 -0.122376 1.355994 [Debug] dit_step29_xt: [2170, 64] first4: 0.312587 1.201936 -0.178490 -0.076513 [DiT] Step 30/50 t=0.420 [Debug] dit_step30_vt: [2170, 64] first4: -0.067228 0.743517 -0.146208 1.339129 [Debug] dit_step30_xt: [2170, 64] first4: 0.313931 1.187065 -0.175566 -0.103296 [DiT] Step 31/50 t=0.400 [Debug] dit_step31_vt: [2170, 64] first4: -0.076558 0.672226 -0.148805 1.307325 [Debug] dit_step31_xt: [2170, 64] first4: 0.315462 1.173621 -0.172590 -0.129442 [DiT] Step 32/50 t=0.380 [Debug] dit_step32_vt: [2170, 64] first4: -0.100921 0.575673 -0.176446 1.269109 [Debug] dit_step32_xt: [2170, 64] first4: 0.317481 1.162107 -0.169061 -0.154824 [DiT] Step 33/50 t=0.360 [Debug] dit_step33_vt: [2170, 64] first4: -0.112492 0.489739 -0.170804 1.252692 [Debug] dit_step33_xt: [2170, 64] first4: 0.319731 1.152313 -0.165645 -0.179878 [DiT] Step 34/50 t=0.340 [Debug] dit_step34_vt: [2170, 64] first4: -0.133039 0.409635 -0.177201 1.214676 [Debug] dit_step34_xt: [2170, 64] first4: 0.322392 1.144120 -0.162101 -0.204172 [DiT] Step 35/50 t=0.320 [Debug] dit_step35_vt: [2170, 64] first4: -0.157808 0.312022 -0.175914 1.201879 [Debug] dit_step35_xt: [2170, 64] first4: 0.325548 1.137880 -0.158583 -0.228209 [DiT] Step 36/50 t=0.300 [Debug] dit_step36_vt: [2170, 64] first4: -0.176505 0.211753 -0.163521 1.156443 [Debug] dit_step36_xt: [2170, 64] first4: 0.329078 1.133645 -0.155312 -0.251338 [DiT] Step 37/50 t=0.280 [Debug] dit_step37_vt: [2170, 64] first4: -0.170152 0.106741 -0.151048 1.120881 [Debug] dit_step37_xt: [2170, 64] first4: 0.332481 1.131510 -0.152291 -0.273756 [DiT] Step 38/50 t=0.260 [Debug] dit_step38_vt: [2170, 64] first4: -0.171724 0.014132 -0.146436 1.076355 [Debug] dit_step38_xt: [2170, 64] first4: 0.335915 1.131227 -0.149362 -0.295283 [DiT] Step 39/50 t=0.240 [Debug] dit_step39_vt: [2170, 64] first4: -0.194143 -0.099352 -0.128456 1.053998 [Debug] dit_step39_xt: [2170, 64] first4: 0.339798 1.133214 -0.146793 -0.316363 [DiT] Step 40/50 t=0.220 [Debug] dit_step40_vt: [2170, 64] first4: -0.227044 -0.210392 -0.126014 1.001273 [Debug] dit_step40_xt: [2170, 64] first4: 0.344339 1.137422 -0.144273 -0.336388 [DiT] Step 41/50 t=0.200 [Debug] dit_step41_vt: [2170, 64] first4: -0.253378 -0.300455 -0.097250 0.965406 [Debug] dit_step41_xt: [2170, 64] first4: 0.349407 1.143431 -0.142328 -0.355696 [DiT] Step 42/50 t=0.180 [Debug] dit_step42_vt: [2170, 64] first4: -0.265390 -0.396163 -0.060336 0.917468 [Debug] dit_step42_xt: [2170, 64] first4: 0.354714 1.151354 -0.141121 -0.374046 [DiT] Step 43/50 t=0.160 [Debug] dit_step43_vt: [2170, 64] first4: -0.295512 -0.494991 -0.042881 0.871955 [Debug] dit_step43_xt: [2170, 64] first4: 0.360625 1.161254 -0.140264 -0.391485 [DiT] Step 44/50 t=0.140 [Debug] dit_step44_vt: [2170, 64] first4: -0.313489 -0.586699 0.022644 0.813441 [Debug] dit_step44_xt: [2170, 64] first4: 0.366894 1.172988 -0.140717 -0.407753 [DiT] Step 45/50 t=0.120 [Debug] dit_step45_vt: [2170, 64] first4: -0.351881 -0.658069 0.069285 0.750789 [Debug] dit_step45_xt: [2170, 64] first4: 0.373932 1.186150 -0.142102 -0.422769 [DiT] Step 46/50 t=0.100 [Debug] dit_step46_vt: [2170, 64] first4: -0.402397 -0.726261 0.104158 0.707855 [Debug] dit_step46_xt: [2170, 64] first4: 0.381980 1.200675 -0.144185 -0.436926 [DiT] Step 47/50 t=0.080 [Debug] dit_step47_vt: [2170, 64] first4: -0.441808 -0.807912 0.180271 0.671096 [Debug] dit_step47_xt: [2170, 64] first4: 0.390816 1.216833 -0.147791 -0.450348 [DiT] Step 48/50 t=0.060 [Debug] dit_step48_vt: [2170, 64] first4: -0.497726 -0.866669 0.246025 0.663294 [Debug] dit_step48_xt: [2170, 64] first4: 0.400771 1.234167 -0.152711 -0.463614 [DiT] Step 49/50 t=0.040 [Debug] dit_step49_vt: [2170, 64] first4: -0.502782 -0.851709 0.169530 0.674803 [Debug] dit_x0: [2170, 64] first4: 0.410826 1.251201 -0.156102 -0.477110 [DiT] Step 50/50 t=0.020 [DiT-Generate] Total: 3926.1 ms (3926.1 ms/sample) [Debug] dit_output: [2170, 64] first4: 0.410826 1.251201 -0.156102 -0.477110 [Store] Unload DiT (7955.6 MB) [GGUF] /mnt/workspace/git/acestep.cpp/models/vae-BF16.gguf: 365 tensors, data at offset 30048 load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] VAE backend: CUDA0 (CPU threads: 16) [VAE] Backend: CUDA0, Weight buffer: 161.1 MB [VAE] Loaded: 5 blocks, upsample=1920x, F32 activations [Store] Load VAE-Dec: 671 ms [VAE] Tiled decode: 3 tiles (chunk=1024, overlap=64, stride=896) [VAE] Graph: 479 nodes, T_latent=960 [VAE] Upsample factor: 1920.00 (expected ~1920) [VAE] Graph: 479 nodes, T_latent=1024 [VAE] Graph: 479 nodes, T_latent=442 [VAE] Tiled decode done: 3 tiles -> T_audio=4166400 (86.80s @ 48kHz) [VAE-Decode Batch0] Decode: 615.2 ms [Debug] vae_audio: [2, 4166400] first4: -0.001324 -0.000638 -0.000648 -0.000196 [Store] Unload VAE-Dec (161.1 MB) [MP3] Encoding 86.8s @ 128 kbps, 48000 Hz stereo [MP3] 1388928 bytes (12.0:1), 416 ms (208.41x realtime), 32 threads [MP3] Wrote ggml-xl-sft/request00.mp3 [Ace-Synth] All done 2026-07-12 10:12:09.884 | INFO | acestep.core.generation.handler.init_service_loader:_load_main_model_from_checkpoint:174 - [initialize_service] Attempting to load model with attention implementation: sdpa [Python] Generating (acestep-v15-xl-turbo, 8 steps)... Using precomputed LM hints Using precomputed LM hints [Python] Wrote python-xl-turbo/output.wav: 4166400 samples (86.80s @ 48kHz stereo) [Python] Done, 40 dump files [Xl-turbo] Cosine similarities GGML vs Python stage GGML vs Python text_hidden 0.999797 lyric_embed 1.000000 enc_hidden 0.999540 detok_output 0.999996 context 0.999998 noise 1.000000 temb_t 0.999996 hidden_after_proj_in 0.999980 enc_after_cond_emb 0.999666 layer0_sa_output 0.999555 hidden_after_layer0 0.999429 hidden_after_layer6 0.999828 hidden_after_layer12 0.997396 hidden_after_layer18 0.996287 hidden_after_layer31 0.988172 dit_step0_vt 0.911727 dit_step0_xt 0.999841 dit_step1_vt 0.930988 dit_step1_xt 0.999534 dit_step2_vt 0.932361 dit_step2_xt 0.998799 dit_step3_vt 0.928919 dit_step3_xt 0.997041 dit_step4_vt 0.922962 dit_step4_xt 0.992628 dit_step5_vt 0.915350 dit_step5_xt 0.982147 dit_step6_vt 0.907975 dit_step6_xt 0.961540 dit_step7_vt 0.887508 dit_x0 0.928000 vae_audio 0.732799 vae_audio (STFT cosine) 0.923088 [Xl-turbo] Error growth GGML vs Python stage cos max_err mean_err mean_A std_A mean_B std_B dit_step0_xt 0.999841 0.160453 0.012482 -0.002975 0.978347 -0.002721 0.976534 dit_step1_xt 0.999534 0.370658 0.021102 -0.005913 0.952143 -0.005567 0.949004 dit_step2_xt 0.998799 0.547901 0.032335 -0.009705 0.923198 -0.009330 0.918931 dit_step3_xt 0.997041 0.812332 0.048155 -0.015330 0.891828 -0.014418 0.887241 dit_step4_xt 0.992628 1.181622 0.072818 -0.023067 0.862653 -0.021453 0.857796 dit_step5_xt 0.982147 1.769622 0.110946 -0.034550 0.846362 -0.032240 0.840033 dit_step6_xt 0.961540 2.534719 0.168391 -0.051596 0.870536 -0.049194 0.863964 [Xl-sft] steps=50, shift=1.0 | acestep-v15-xl-sft-BF16.gguf [GGML] Running acestep-v15-xl-sft-BF16.gguf... [GGML] Done, 131 dump files [Python] Initializing acestep-v15-xl-sft... Loading checkpoint shards: 0%| | 0/4 [00:00 2026-07-12 10:12:14.596 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:123 - ====================================================================== 2026-07-12 10:12:14.596 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:124 - lyrics_text: # Languages fr # Lyric # Lyric [Intro - Guitar Riff] [Verse 1] Dans le monde des tutos virtuels G ta toise en nouvelle passion Avec Ggendoline et PumbΓ© Γ  midi La communautΓ©, c'est l'unitΓ© Quel joie, une clΓ© [Chorus] Dans le monde des tutos virtuels GΓ’ndoline et PumbΓ© Γ  midi Une famille Γ  connecter, c'est vrai D'un enfant qui voit toi fusionner [Guitar Solo] [Verse 2] Dans le monde des tutos virtuels GΓ’ndoline, PumbΓ© Γ  midi Une famille Γ  connecter, c'est vrai D'un enfant qui voit toi fusionner<|endoftext|> 2026-07-12 10:12:14.596 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:125 - ====================================================================== 2026-07-12 10:12:14.599 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:110 - [preprocess_batch] Inferring prompt embeddings... 2026-07-12 10:12:14.607 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:113 - [preprocess_batch] Inferring lyric embeddings... 2026-07-12 10:12:14.607 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:161 - [service_generate] Generating audio... (DiT backend: PyTorch (cuda)) 2026-07-12 10:12:14.614 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:263 - [service_generate] DiT diffusion via PyTorch (cuda)... [Python] Generating (acestep-v15-xl-sft, 50 steps)... Using precomputed LM hints Using precomputed LM hints 0%| | 0/50 [00:00