[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-Q8_0.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-Q8_0.gguf: 678 tensors, data at offset 56864 [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-turbo-Q8_0.gguf: 678 tensors, data at offset 56864 [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: Vulkan0 (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: 498 ms [Debug] text_hidden: [70, 1024] first4: 3.705836 2.395382 0.221845 -13.145830 [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: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-turbo-Q8_0.gguf: 678 tensors, data at offset 56864 [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, 616.6 MB into backend [Load] CondEncoder: lyric(8L), timbre(4L), text_proj, null_cond [Store] Load CondEnc: 329 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, 3368.2 ms [Debug] enc_hidden: [238, 2048] first4: 1.758852 -0.049792 -0.133259 0.058312 [Store] Unload CondEnc (616.6 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: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-turbo-Q8_0.gguf: 678 tensors, data at offset 56864 [WeightCtx] Loaded 30 tensors, 106.5 MB into backend [Load] Detokenizer: FSQ(6->2048) + 2L encoder(S=5, 2048->64) [Store] Load FSQ-Detok: 188 ms [Context] Decoded: 434 codes -> 2170 frames (86.8s @ 25Hz) [Build-Context Batch0] Detokenizer: 337.4 ms, 434 codes [Debug] detok_output: [2170, 64] first4: -0.124812 1.441172 0.303914 -0.631591 [Store] Unload FSQ-Detok (106.5 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: -0.124812 1.441172 0.303914 -0.631591 [Init-Noise] Starting: T=2170, S=1085, enc_S=238, steps=8, batch=1 (cover) 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: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-turbo-Q8_0.gguf: 678 tensors, data at offset 56864 [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, 1600.7 MB into backend [Load] DiT: 24 layers, H=2048, Nh=16/8, D=128 [Store] Load DiT: 875 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: 0.260124 -0.161873 -0.097043 0.052039 [Debug] temb: [2048] first4: 0.000130 -0.132501 -0.035452 0.064788 [Debug] temb_t: [2048] first4: 0.001145 0.026826 -0.052770 0.063722 [Debug] temb_r: [2048] first4: -0.001015 -0.159327 0.017318 0.001066 [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: -0.048950 -0.051683 -0.015299 -0.038721 [Debug] temb_lin1_r: [2048] first4: -0.013066 -0.018836 -0.015732 0.008463 [Debug] hidden_after_proj_in: [2048, 1085] first4: -1.038574 -0.958557 0.537598 0.446259 [Debug] proj_in_input: [192, 2170] first4: -0.124812 1.441172 0.303914 -0.631591 [Debug] enc_after_cond_emb: [2048, 238] first4: -0.169296 0.817795 0.328674 -0.558029 [Debug] layer0_sa_input: [2048, 1085] first4: -0.718028 -0.757975 -0.047152 0.261293 [Debug] layer0_q_after_rope: [128, 16] first4: -1.541547 -1.031567 0.196720 0.455789 [Debug] layer0_k_after_rope: [128, 8] first4: -0.169296 0.817795 0.328674 -0.558029 [Debug] layer0_sa_output: [2048, 1085] first4: -1.507812 0.164551 -0.349121 0.512695 [Debug] layer0_attn_out: [2048, 1085] first4: -1.541547 -1.031567 0.196720 0.455789 [Debug] layer0_after_self_attn: [2048, 1085] first4: -1.541547 -1.031567 0.196720 0.455789 [Debug] layer0_after_cross_attn: [2048, 1085] first4: -1.582532 -0.791699 -0.283016 0.495675 [Debug] hidden_after_layer0: [2048, 1085] first4: -8.934134 0.561685 51.267479 -0.874029 [Debug] hidden_after_layer6: [2048, 1085] first4: -20.909487 -0.295922 34.080910 -4.367682 [Debug] hidden_after_layer12: [2048, 1085] first4: -14.458811 -18.314516 73.548622 30.143749 [Debug] hidden_after_layer18: [2048, 1085] first4: -26.918846 14.447463 61.928448 19.971241 [Debug] hidden_after_layer23: [2048, 1085] first4: -2.146482 41.510212 191.781036 143.801422 [Debug] dit_step0_vt: [2170, 64] first4: 0.035767 1.115242 0.349195 2.344635 [Debug] dit_step0_xt: [2170, 64] first4: 0.192710 2.105557 -0.187748 0.741082 [DiT] Step 1/8 t=1.000 [Debug] dit_step1_vt: [2170, 64] first4: -0.185852 1.308655 -0.125847 1.902275 [Debug] dit_step1_xt: [2170, 64] first4: 0.202848 2.034176 -0.180883 0.637321 [DiT] Step 2/8 t=0.955 [Debug] dit_step2_vt: [2170, 64] first4: -0.016571 1.210049 0.100586 2.365997 [Debug] dit_step2_xt: [2170, 64] first4: 0.203952 1.953506 -0.187589 0.479588 [DiT] Step 3/8 t=0.900 [Debug] dit_step3_vt: [2170, 64] first4: 0.249638 1.065643 0.254936 2.623840 [Debug] dit_step3_xt: [2170, 64] first4: 0.183149 1.864703 -0.208834 0.260935 [DiT] Step 4/8 t=0.833 [Debug] dit_step4_vt: [2170, 64] first4: 0.289047 1.007362 0.116272 2.690430 [Debug] dit_step4_xt: [2170, 64] first4: 0.152180 1.756771 -0.221291 -0.027325 [DiT] Step 5/8 t=0.750 [Debug] dit_step5_vt: [2170, 64] first4: 0.262039 0.930145 -0.241333 2.727318 [Debug] dit_step5_xt: [2170, 64] first4: 0.114746 1.623893 -0.186815 -0.416942 [DiT] Step 6/8 t=0.643 [Debug] dit_step6_vt: [2170, 64] first4: 0.158157 0.654114 -0.751740 2.777405 [Debug] dit_step6_xt: [2170, 64] first4: 0.083114 1.493070 -0.036467 -0.972423 [DiT] Step 7/8 t=0.500 [Debug] dit_step7_vt: [2170, 64] first4: -0.026886 0.187073 -1.401917 3.082886 [Debug] dit_x0: [2170, 64] first4: 0.091180 1.436948 0.384108 -1.897289 [DiT] Step 8/8 t=0.300 [DiT-Generate] Total: 281.9 ms (281.9 ms/sample) [Debug] dit_output: [2170, 64] first4: 0.091180 1.436948 0.384108 -1.897289 [Store] Unload DiT (1600.7 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: Vulkan0 (CPU threads: 16) [VAE] Backend: Vulkan0, Weight buffer: 161.1 MB [VAE] Loaded: 5 blocks, upsample=1920x, F32 activations [Store] Load VAE-Dec: 712 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: 787.3 ms [Debug] vae_audio: [2, 4166400] first4: 0.000762 0.001482 0.001247 0.001779 [Store] Unload VAE-Dec (161.1 MB) [MP3] Encoding 86.8s @ 128 kbps, 48000 Hz stereo [MP3] 1388928 bytes (12.0:1), 460 ms (188.66x realtime), 32 threads [MP3] Wrote ggml-turbo/request00.mp3 [Ace-Synth] All done 2026-07-12 10:16:19.328 | 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 10:16:21.454 | WARNING | acestep.training.trainer::40 - bitsandbytes not installed. Using standard AdamW. 2026-07-12 10:16:21.459 | 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 10:16:21.530 | 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 10:16:23.577 | 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 10:16:23.577 | INFO | acestep.core.generation.handler.generate_music:generate_music:304 - [generate_music] Starting generation... 2026-07-12 10:16:23.577 | INFO | acestep.core.generation.handler.generate_music:generate_music:307 - [generate_music] Preparing inputs... 2026-07-12 10:16:23.577 | 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 10:16:23.580 | 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 10:16:23.586 | INFO | acestep.core.generation.handler.conditioning_target:_prepare_target_latents_and_wavs:78 - [generate_music] Decoding audio codes for item 0... 2026-07-12 10:16:23.793 | 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 10:16:23.795 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:119 - ====================================================================== 2026-07-12 10:16:23.795 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:120 - πŸ” [DEBUG] DiT TEXT ENCODER INPUT (Inference) 2026-07-12 10:16:23.795 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:121 - ====================================================================== 2026-07-12 10:16:23.795 | 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 10:16:23.795 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:123 - ====================================================================== 2026-07-12 10:16:23.795 | 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 10:16:23.795 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:125 - ====================================================================== 2026-07-12 10:16:23.802 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:110 - [preprocess_batch] Inferring prompt embeddings... 2026-07-12 10:16:23.816 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:113 - [preprocess_batch] Inferring lyric embeddings... 2026-07-12 10:16:23.817 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:161 - [service_generate] Generating audio... (DiT backend: PyTorch (cuda)) 2026-07-12 10:16:23.852 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:263 - [service_generate] DiT diffusion via PyTorch (cuda)... 2026-07-12 10:16:23.859 | 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 10:16:23.944 | 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 10:16:24.187 | INFO | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:42 - [generate_music] Model generation completed. Decoding latents... 2026-07-12 10:16:24.188 | 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 10:16:24.188 | DEBUG | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:65 - [generate_music] time_costs: {'encoder_time_cost': 0.00689387321472168, 'diffusion_time_cost': 0.3277108669281006, 'diffusion_per_step_time_cost': 0.04096385836601257, 'total_time_cost': 0.33460474014282227, 'offload_time_cost': 0.0} 2026-07-12 10:16:24.202 | INFO | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:127 - [generate_music] Decoding latents with VAE... 2026-07-12 10:16:24.204 | 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 10:16:24.204 | 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 10:16:24.204 | 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 10:16:24.204 | 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 10:16:24.204 | 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 10:16:24.204 | 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 10:16:24.480 | 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 10:16:24.644 | 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 10:16:24.647 | 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-Q8_0.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-Q8_0.gguf: 678 tensors, data at offset 56800 [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-sft-Q8_0.gguf: 678 tensors, data at offset 56800 [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: Vulkan0 (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: 338 ms [Debug] text_hidden: [70, 1024] first4: 3.705836 2.395382 0.221845 -13.145830 [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: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-sft-Q8_0.gguf: 678 tensors, data at offset 56800 [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, 616.6 MB into backend [Load] CondEncoder: lyric(8L), timbre(4L), text_proj, null_cond [Store] Load CondEnc: 321 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, 26.8 ms [Debug] enc_hidden: [238, 2048] first4: 1.758852 -0.049792 -0.133259 0.058312 [Store] Unload CondEnc (616.6 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: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-sft-Q8_0.gguf: 678 tensors, data at offset 56800 [WeightCtx] Loaded 30 tensors, 106.5 MB into backend [Load] Detokenizer: FSQ(6->2048) + 2L encoder(S=5, 2048->64) [Store] Load FSQ-Detok: 174 ms [Context] Decoded: 434 codes -> 2170 frames (86.8s @ 25Hz) [Build-Context Batch0] Detokenizer: 102.4 ms, 434 codes [Debug] detok_output: [2170, 64] first4: -0.124812 1.441172 0.303914 -0.631591 [Store] Unload FSQ-Detok (106.5 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: -0.124812 1.441172 0.303914 -0.631591 [Init-Noise] Starting: T=2170, S=1085, enc_S=238, steps=50, batch=1 (cover) 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: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-sft-Q8_0.gguf: 678 tensors, data at offset 56800 [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, 1600.7 MB into backend [Load] DiT: 24 layers, H=2048, Nh=16/8, D=128 [Store] Load DiT: 947 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: 0.154891 -0.116111 -0.086457 0.081949 [Debug] temb: [2048] first4: -0.002756 -0.176432 0.004178 -0.001982 [Debug] temb_t: [2048] first4: -0.001185 0.003330 -0.013113 -0.002073 [Debug] temb_r: [2048] first4: -0.001571 -0.179762 0.017291 0.000091 [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: -0.038370 0.029874 0.028025 -0.024771 [Debug] temb_lin1_r: [2048] first4: 0.001863 0.003353 -0.000552 -0.000197 [Debug] hidden_after_proj_in: [2048, 1085] first4: -1.087891 -0.914795 0.499023 0.441193 [Debug] proj_in_input: [192, 2170] first4: -0.124812 1.441172 0.303914 -0.631591 [Debug] enc_after_cond_emb: [2048, 238] first4: -0.187347 0.873795 0.324829 -0.581207 [Debug] layer0_sa_input: [2048, 1085] first4: -0.912651 -0.709809 -0.037078 0.294389 [Debug] layer0_q_after_rope: [128, 16] first4: -1.803765 -1.369850 -0.116876 0.380624 [Debug] layer0_k_after_rope: [128, 8] first4: -0.187347 0.873795 0.324829 -0.581207 [Debug] layer0_sa_output: [2048, 1085] first4: -1.653320 0.803711 -0.544922 0.540527 [Debug] layer0_attn_out: [2048, 1085] first4: -1.803765 -1.369850 -0.116876 0.380624 [Debug] layer0_after_self_attn: [2048, 1085] first4: -1.803765 -1.369850 -0.116876 0.380624 [Debug] layer0_after_cross_attn: [2048, 1085] first4: -1.945000 -1.018043 -0.383234 0.395715 [Debug] hidden_after_layer0: [2048, 1085] first4: -9.240007 1.161420 55.883682 -1.572288 [Debug] hidden_after_layer6: [2048, 1085] first4: -16.968437 4.161700 56.720623 -2.322089 [Debug] hidden_after_layer12: [2048, 1085] first4: -9.775227 3.819345 -21.191896 -1.707024 [Debug] hidden_after_layer18: [2048, 1085] first4: -3.762543 19.361917 -37.082027 -11.004805 [Debug] hidden_after_layer23: [2048, 1085] first4: 34.009296 62.304276 58.564732 -9.942815 [Debug] dit_step0_vt: [2170, 64] first4: -0.573181 2.511108 -0.178902 1.475586 [Debug] dit_step0_xt: [2170, 64] first4: 0.205800 2.106028 -0.168297 0.818145 [DiT] Step 1/50 t=1.000 [Debug] dit_step1_vt: [2170, 64] first4: -0.628418 2.528076 -0.010651 1.298542 [Debug] dit_step1_xt: [2170, 64] first4: 0.218368 2.055466 -0.168084 0.792174 [DiT] Step 2/50 t=0.980 [Debug] dit_step2_vt: [2170, 64] first4: -0.614380 2.487061 0.026398 1.319626 [Debug] dit_step2_xt: [2170, 64] first4: 0.230656 2.005725 -0.168612 0.765781 [DiT] Step 3/50 t=0.960 [Debug] dit_step3_vt: [2170, 64] first4: -0.604492 2.427002 0.043579 1.331329 [Debug] dit_step3_xt: [2170, 64] first4: 0.242745 1.957185 -0.169483 0.739155 [DiT] Step 4/50 t=0.940 [Debug] dit_step4_vt: [2170, 64] first4: -0.545288 2.337524 0.026611 1.320969 [Debug] dit_step4_xt: [2170, 64] first4: 0.253651 1.910434 -0.170016 0.712735 [DiT] Step 5/50 t=0.920 [Debug] dit_step5_vt: [2170, 64] first4: -0.498840 2.264648 0.001862 1.309006 [Debug] dit_step5_xt: [2170, 64] first4: 0.263628 1.865141 -0.170053 0.686555 [DiT] Step 6/50 t=0.900 [Debug] dit_step6_vt: [2170, 64] first4: -0.476196 2.210449 -0.023140 1.298187 [Debug] dit_step6_xt: [2170, 64] first4: 0.273152 1.820932 -0.169590 0.660591 [DiT] Step 7/50 t=0.880 [Debug] dit_step7_vt: [2170, 64] first4: -0.439880 2.145264 -0.047558 1.279770 [Debug] dit_step7_xt: [2170, 64] first4: 0.281949 1.778027 -0.168639 0.634996 [DiT] Step 8/50 t=0.860 [Debug] dit_step8_vt: [2170, 64] first4: -0.381592 2.053040 -0.061089 1.271408 [Debug] dit_step8_xt: [2170, 64] first4: 0.289581 1.736966 -0.167417 0.609568 [DiT] Step 9/50 t=0.840 [Debug] dit_step9_vt: [2170, 64] first4: -0.315796 1.968384 -0.070114 1.255005 [Debug] dit_step9_xt: [2170, 64] first4: 0.295897 1.697599 -0.166015 0.584468 [DiT] Step 10/50 t=0.820 [Debug] dit_step10_vt: [2170, 64] first4: -0.260559 1.893799 -0.051498 1.248482 [Debug] dit_step10_xt: [2170, 64] first4: 0.301108 1.659723 -0.164985 0.559498 [DiT] Step 11/50 t=0.800 [Debug] dit_step11_vt: [2170, 64] first4: -0.225708 1.821716 -0.045013 1.239441 [Debug] dit_step11_xt: [2170, 64] first4: 0.305623 1.623288 -0.164085 0.534709 [DiT] Step 12/50 t=0.780 [Debug] dit_step12_vt: [2170, 64] first4: -0.203064 1.761200 -0.057892 1.236008 [Debug] dit_step12_xt: [2170, 64] first4: 0.309684 1.588064 -0.162927 0.509989 [DiT] Step 13/50 t=0.760 [Debug] dit_step13_vt: [2170, 64] first4: -0.188614 1.686340 -0.071213 1.218018 [Debug] dit_step13_xt: [2170, 64] first4: 0.313456 1.554338 -0.161503 0.485629 [DiT] Step 14/50 t=0.740 [Debug] dit_step14_vt: [2170, 64] first4: -0.179405 1.610931 -0.084076 1.206627 [Debug] dit_step14_xt: [2170, 64] first4: 0.317044 1.522119 -0.159821 0.461496 [DiT] Step 15/50 t=0.720 [Debug] dit_step15_vt: [2170, 64] first4: -0.177692 1.529755 -0.114746 1.191498 [Debug] dit_step15_xt: [2170, 64] first4: 0.320598 1.491524 -0.157526 0.437666 [DiT] Step 16/50 t=0.700 [Debug] dit_step16_vt: [2170, 64] first4: -0.185883 1.448570 -0.143921 1.171323 [Debug] dit_step16_xt: [2170, 64] first4: 0.324316 1.462552 -0.154648 0.414240 [DiT] Step 17/50 t=0.680 [Debug] dit_step17_vt: [2170, 64] first4: -0.187820 1.362694 -0.165771 1.155390 [Debug] dit_step17_xt: [2170, 64] first4: 0.328072 1.435298 -0.151332 0.391132 [DiT] Step 18/50 t=0.660 [Debug] dit_step18_vt: [2170, 64] first4: -0.208435 1.279594 -0.194092 1.143333 [Debug] dit_step18_xt: [2170, 64] first4: 0.332241 1.409707 -0.147450 0.368265 [DiT] Step 19/50 t=0.640 [Debug] dit_step19_vt: [2170, 64] first4: -0.227783 1.213791 -0.216187 1.134262 [Debug] dit_step19_xt: [2170, 64] first4: 0.336797 1.385431 -0.143127 0.345580 [DiT] Step 20/50 t=0.620 [Debug] dit_step20_vt: [2170, 64] first4: -0.242554 1.115173 -0.233337 1.102615 [Debug] dit_step20_xt: [2170, 64] first4: 0.341648 1.363127 -0.138460 0.323528 [DiT] Step 21/50 t=0.600 [Debug] dit_step21_vt: [2170, 64] first4: -0.268768 1.026794 -0.244019 1.083527 [Debug] dit_step21_xt: [2170, 64] first4: 0.347023 1.342591 -0.133580 0.301857 [DiT] Step 22/50 t=0.580 [Debug] dit_step22_vt: [2170, 64] first4: -0.302170 0.924164 -0.246887 1.053543 [Debug] dit_step22_xt: [2170, 64] first4: 0.353066 1.324108 -0.128642 0.280786 [DiT] Step 23/50 t=0.560 [Debug] dit_step23_vt: [2170, 64] first4: -0.314056 0.830017 -0.259277 1.032532 [Debug] dit_step23_xt: [2170, 64] first4: 0.359347 1.307508 -0.123456 0.260136 [DiT] Step 24/50 t=0.540 [Debug] dit_step24_vt: [2170, 64] first4: -0.334854 0.719666 -0.263367 1.000183 [Debug] dit_step24_xt: [2170, 64] first4: 0.366045 1.293114 -0.118189 0.240132 [DiT] Step 25/50 t=0.520 [Debug] dit_step25_vt: [2170, 64] first4: -0.363007 0.615540 -0.267517 0.985168 [Debug] dit_step25_xt: [2170, 64] first4: 0.373305 1.280804 -0.112839 0.220429 [DiT] Step 26/50 t=0.500 [Debug] dit_step26_vt: [2170, 64] first4: -0.388702 0.498962 -0.269531 0.966003 [Debug] dit_step26_xt: [2170, 64] first4: 0.381079 1.270824 -0.107448 0.201109 [DiT] Step 27/50 t=0.480 [Debug] dit_step27_vt: [2170, 64] first4: -0.408844 0.389923 -0.262878 0.931946 [Debug] dit_step27_xt: [2170, 64] first4: 0.389256 1.263026 -0.102190 0.182470 [DiT] Step 28/50 t=0.460 [Debug] dit_step28_vt: [2170, 64] first4: -0.433212 0.276031 -0.258545 0.926392 [Debug] dit_step28_xt: [2170, 64] first4: 0.397920 1.257505 -0.097020 0.163942 [DiT] Step 29/50 t=0.440 [Debug] dit_step29_vt: [2170, 64] first4: -0.463440 0.164532 -0.257111 0.895081 [Debug] dit_step29_xt: [2170, 64] first4: 0.407189 1.254215 -0.091877 0.146040 [DiT] Step 30/50 t=0.420 [Debug] dit_step30_vt: [2170, 64] first4: -0.481400 0.045422 -0.246948 0.879272 [Debug] dit_step30_xt: [2170, 64] first4: 0.416817 1.253306 -0.086938 0.128455 [DiT] Step 31/50 t=0.400 [Debug] dit_step31_vt: [2170, 64] first4: -0.502869 -0.073616 -0.240173 0.854126 [Debug] dit_step31_xt: [2170, 64] first4: 0.426874 1.254779 -0.082135 0.111372 [DiT] Step 32/50 t=0.380 [Debug] dit_step32_vt: [2170, 64] first4: -0.508575 -0.178040 -0.233368 0.832825 [Debug] dit_step32_xt: [2170, 64] first4: 0.437046 1.258340 -0.077468 0.094716 [DiT] Step 33/50 t=0.360 [Debug] dit_step33_vt: [2170, 64] first4: -0.514950 -0.301666 -0.220184 0.806152 [Debug] dit_step33_xt: [2170, 64] first4: 0.447345 1.264373 -0.073064 0.078593 [DiT] Step 34/50 t=0.340 [Debug] dit_step34_vt: [2170, 64] first4: -0.534730 -0.386261 -0.191406 0.786133 [Debug] dit_step34_xt: [2170, 64] first4: 0.458039 1.272098 -0.069236 0.062870 [DiT] Step 35/50 t=0.320 [Debug] dit_step35_vt: [2170, 64] first4: -0.538104 -0.492905 -0.186378 0.750183 [Debug] dit_step35_xt: [2170, 64] first4: 0.468801 1.281956 -0.065508 0.047866 [DiT] Step 36/50 t=0.300 [Debug] dit_step36_vt: [2170, 64] first4: -0.555412 -0.583405 -0.155294 0.733826 [Debug] dit_step36_xt: [2170, 64] first4: 0.479909 1.293624 -0.062402 0.033190 [DiT] Step 37/50 t=0.280 [Debug] dit_step37_vt: [2170, 64] first4: -0.558224 -0.687943 -0.133743 0.693604 [Debug] dit_step37_xt: [2170, 64] first4: 0.491074 1.307383 -0.059727 0.019318 [DiT] Step 38/50 t=0.260 [Debug] dit_step38_vt: [2170, 64] first4: -0.557346 -0.771755 -0.108200 0.645142 [Debug] dit_step38_xt: [2170, 64] first4: 0.502221 1.322818 -0.057563 0.006415 [DiT] Step 39/50 t=0.240 [Debug] dit_step39_vt: [2170, 64] first4: -0.563274 -0.852579 -0.072838 0.607056 [Debug] dit_step39_xt: [2170, 64] first4: 0.513486 1.339870 -0.056107 -0.005726 [DiT] Step 40/50 t=0.220 [Debug] dit_step40_vt: [2170, 64] first4: -0.548340 -0.917286 -0.043777 0.564209 [Debug] dit_step40_xt: [2170, 64] first4: 0.524453 1.358215 -0.055231 -0.017010 [DiT] Step 41/50 t=0.200 [Debug] dit_step41_vt: [2170, 64] first4: -0.539944 -0.988258 -0.011510 0.520813 [Debug] dit_step41_xt: [2170, 64] first4: 0.535252 1.377981 -0.055001 -0.027427 [DiT] Step 42/50 t=0.180 [Debug] dit_step42_vt: [2170, 64] first4: -0.534973 -1.051826 0.038669 0.468117 [Debug] dit_step42_xt: [2170, 64] first4: 0.545952 1.399017 -0.055774 -0.036789 [DiT] Step 43/50 t=0.160 [Debug] dit_step43_vt: [2170, 64] first4: -0.520775 -1.113958 0.088394 0.415894 [Debug] dit_step43_xt: [2170, 64] first4: 0.556367 1.421296 -0.057542 -0.045107 [DiT] Step 44/50 t=0.140 [Debug] dit_step44_vt: [2170, 64] first4: -0.493507 -1.163010 0.130157 0.355743 [Debug] dit_step44_xt: [2170, 64] first4: 0.566237 1.444556 -0.060145 -0.052222 [DiT] Step 45/50 t=0.120 [Debug] dit_step45_vt: [2170, 64] first4: -0.479034 -1.205872 0.173813 0.306442 [Debug] dit_step45_xt: [2170, 64] first4: 0.575818 1.468674 -0.063622 -0.058350 [DiT] Step 46/50 t=0.100 [Debug] dit_step46_vt: [2170, 64] first4: -0.447800 -1.245209 0.207451 0.245369 [Debug] dit_step46_xt: [2170, 64] first4: 0.584774 1.493578 -0.067771 -0.063258 [DiT] Step 47/50 t=0.080 [Debug] dit_step47_vt: [2170, 64] first4: -0.421173 -1.281021 0.225426 0.203156 [Debug] dit_step47_xt: [2170, 64] first4: 0.593197 1.519198 -0.072279 -0.067321 [DiT] Step 48/50 t=0.060 [Debug] dit_step48_vt: [2170, 64] first4: -0.373352 -1.307907 0.242920 0.173157 [Debug] dit_step48_xt: [2170, 64] first4: 0.600664 1.545357 -0.077138 -0.070784 [DiT] Step 49/50 t=0.040 [Debug] dit_step49_vt: [2170, 64] first4: -0.404266 -1.336517 0.225311 0.160156 [Debug] dit_x0: [2170, 64] first4: 0.608750 1.572087 -0.081644 -0.073987 [DiT] Step 50/50 t=0.020 [DiT-Generate] Total: 1486.1 ms (1486.1 ms/sample) [Debug] dit_output: [2170, 64] first4: 0.608750 1.572087 -0.081644 -0.073987 [Store] Unload DiT (1600.7 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: Vulkan0 (CPU threads: 16) [VAE] Backend: Vulkan0, Weight buffer: 161.1 MB [VAE] Loaded: 5 blocks, upsample=1920x, F32 activations [Store] Load VAE-Dec: 666 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: 786.0 ms [Debug] vae_audio: [2, 4166400] first4: -0.001594 -0.001085 -0.001352 -0.000940 [Store] Unload VAE-Dec (161.1 MB) [MP3] Encoding 86.8s @ 128 kbps, 48000 Hz stereo [MP3] 1388928 bytes (12.0:1), 407 ms (213.49x realtime), 32 threads [MP3] Wrote ggml-sft/request00.mp3 [Ace-Synth] All done 2026-07-12 10:16:31.034 | 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 10:16:33.687 | INFO | acestep.core.generation.handler.generate_music:generate_music:304 - [generate_music] Starting generation... 2026-07-12 10:16:33.687 | INFO | acestep.core.generation.handler.generate_music:generate_music:307 - [generate_music] Preparing inputs... 2026-07-12 10:16:33.687 | 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 10:16:33.690 | 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 10:16:33.695 | INFO | acestep.core.generation.handler.conditioning_target:_prepare_target_latents_and_wavs:78 - [generate_music] Decoding audio codes for item 0... 2026-07-12 10:16:33.700 | 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 10:16:33.701 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:119 - ====================================================================== 2026-07-12 10:16:33.701 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:120 - πŸ” [DEBUG] DiT TEXT ENCODER INPUT (Inference) 2026-07-12 10:16:33.701 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:121 - ====================================================================== 2026-07-12 10:16:33.701 | 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 10:16:33.701 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:123 - ====================================================================== 2026-07-12 10:16:33.701 | 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 10:16:33.701 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:125 - ====================================================================== 2026-07-12 10:16:33.705 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:110 - [preprocess_batch] Inferring prompt embeddings... 2026-07-12 10:16:33.713 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:113 - [preprocess_batch] Inferring lyric embeddings... 2026-07-12 10:16:33.713 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:161 - [service_generate] Generating audio... (DiT backend: PyTorch (cuda)) 2026-07-12 10:16:33.719 | 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-Q8_0.gguf [GGML] Running acestep-v15-turbo-Q8_0.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 0.999812 lyric_embed 1.000000 enc_hidden 0.999826 detok_output 0.999984 context 0.999990 noise 1.000000 temb_t 0.999998 hidden_after_proj_in 0.999985 enc_after_cond_emb 0.999818 layer0_sa_output 0.999938 hidden_after_layer0 0.999865 hidden_after_layer6 0.999893 hidden_after_layer12 0.999097 hidden_after_layer18 0.996362 hidden_after_layer23 0.993145 dit_step0_vt 0.973692 dit_step0_xt 0.999942 dit_step1_vt 0.975202 dit_step1_xt 0.999811 dit_step2_vt 0.976472 dit_step2_xt 0.999475 dit_step3_vt 0.978844 dit_step3_xt 0.998749 dit_step4_vt 0.977970 dit_step4_xt 0.997093 dit_step5_vt 0.975672 dit_step5_xt 0.993442 dit_step6_vt 0.972881 dit_step6_xt 0.986651 dit_step7_vt 0.966591 dit_x0 0.976811 vae_audio 0.897165 vae_audio (STFT cosine) 0.973873 [Turbo] Error growth GGML vs Python stage cos max_err mean_err mean_A std_A mean_B std_B dit_step0_xt 0.999942 0.140653 0.006960 -0.002313 0.973046 -0.002342 0.972003 dit_step1_xt 0.999811 0.270184 0.012096 -0.005305 0.942910 -0.005313 0.941730 dit_step2_xt 0.999475 0.461222 0.019001 -0.009329 0.909604 -0.009311 0.908527 dit_step3_xt 0.998749 0.641579 0.027927 -0.014679 0.874459 -0.014577 0.873624 dit_step4_xt 0.997093 0.963044 0.040519 -0.021686 0.842842 -0.021660 0.841995 dit_step5_xt 0.993442 1.433617 0.059086 -0.031715 0.826380 -0.032109 0.824593 dit_step6_xt 0.986651 2.153977 0.087109 -0.045875 0.857460 -0.046482 0.855546 [SFT] steps=50, shift=1.0 | acestep-v15-sft-Q8_0.gguf [GGML] Running acestep-v15-sft-Q8_0.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-Q8_0.gguf: 830 tensors, data at offset 69088 [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-turbo-Q8_0.gguf: 830 tensors, data at offset 69088 [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: Vulkan0 (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: 309 ms [Debug] text_hidden: [70, 1024] first4: 3.705836 2.395382 0.221845 -13.145830 [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: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-turbo-Q8_0.gguf: 830 tensors, data at offset 69088 [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, 616.6 MB into backend [Load] CondEncoder: lyric(8L), timbre(4L, CLS), text_proj, null_cond [Store] Load CondEnc: 329 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, 137.9 ms [Debug] enc_hidden: [238, 2048] first4: 0.686684 -0.019193 -0.051921 0.022813 [Store] Unload CondEnc (616.6 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: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-turbo-Q8_0.gguf: 830 tensors, data at offset 69088 [WeightCtx] Loaded 30 tensors, 106.5 MB into backend [Load] Detokenizer: FSQ(6->2048) + 2L encoder(S=5, 2048->64) [Store] Load FSQ-Detok: 184 ms [Context] Decoded: 434 codes -> 2170 frames (86.8s @ 25Hz) [Build-Context Batch0] Detokenizer: 101.2 ms, 434 codes [Debug] detok_output: [2170, 64] first4: -0.124812 1.441172 0.303914 -0.631591 [Store] Unload FSQ-Detok (106.5 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: -0.124812 1.441172 0.303914 -0.631591 [Init-Noise] Starting: T=2170, S=1085, enc_S=238, steps=8, batch=1 (cover) 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: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-turbo-Q8_0.gguf: 830 tensors, data at offset 69088 [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, 4230.2 MB into backend [Load] DiT: 32 layers, H=2560, Nh=32/8, D=128 [Store] Load DiT: 2440 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: 0.016348 0.109485 0.066159 -0.074161 [Debug] temb: [2560] first4: 0.026323 -0.023037 -0.016484 0.047127 [Debug] temb_t: [2560] first4: 0.014564 -0.022307 -0.020828 0.048741 [Debug] temb_r: [2560] first4: 0.011759 -0.000730 0.004344 -0.001614 [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: 0.064043 -0.018896 0.138192 -0.113571 [Debug] temb_lin1_r: [2560] first4: -0.009850 0.016688 -0.018317 0.012020 [Debug] hidden_after_proj_in: [2560, 1085] first4: 0.342407 -0.579834 0.230591 -0.188965 [Debug] proj_in_input: [192, 2170] first4: -0.124812 1.441172 0.303914 -0.631591 [Debug] enc_after_cond_emb: [2560, 238] first4: 0.031250 0.652512 0.100861 0.874588 [Debug] layer0_sa_input: [2560, 1085] first4: 0.040291 -0.045382 0.094641 -0.134234 [Debug] layer0_q_after_rope: [128, 32] first4: -0.256418 -2.171321 0.185056 0.020789 [Debug] layer0_k_after_rope: [128, 8] first4: 0.594727 2.179688 0.737305 0.156860 [Debug] layer0_sa_output: [2560, 1085] first4: 1.869141 4.457031 -0.947754 5.628906 [Debug] layer0_attn_out: [4096, 1085] first4: 0.490657 -2.306927 -0.464014 -2.894142 [Debug] layer0_after_self_attn: [2560, 1085] first4: 0.490657 -2.306927 -0.464014 -2.894142 [Debug] layer0_after_cross_attn: [2560, 1085] first4: 0.333553 -2.040814 -0.313502 -3.793556 [Debug] hidden_after_layer0: [2560, 1085] first4: 1.679857 -2.475510 241.296799 -3.764442 [Debug] hidden_after_layer6: [2560, 1085] first4: -6.018240 -2.451056 320.925598 -12.578479 [Debug] hidden_after_layer12: [2560, 1085] first4: -11.319883 29.579870 261.062927 29.462757 [Debug] hidden_after_layer18: [2560, 1085] first4: 41.428871 11.781424 276.454742 2.682156 [Debug] hidden_after_layer31: [2560, 1085] first4: -94.410431 -234.079666 -142.090698 249.054214 [Debug] dit_step0_vt: [2170, 64] first4: -1.106445 1.501389 0.492004 1.225586 [Debug] dit_step0_xt: [2170, 64] first4: 0.244629 2.088005 -0.194239 0.791948 [DiT] Step 1/8 t=1.000 [Debug] dit_step1_vt: [2170, 64] first4: -0.999390 1.313232 0.651428 1.117554 [Debug] dit_step1_xt: [2170, 64] first4: 0.299141 2.016374 -0.229771 0.730990 [DiT] Step 2/8 t=0.955 [Debug] dit_step2_vt: [2170, 64] first4: -0.414619 1.436302 0.854790 0.899231 [Debug] dit_step2_xt: [2170, 64] first4: 0.326782 1.920621 -0.286757 0.671042 [DiT] Step 3/8 t=0.900 [Debug] dit_step3_vt: [2170, 64] first4: -0.251343 1.429489 0.900108 0.944511 [Debug] dit_step3_xt: [2170, 64] first4: 0.347728 1.801497 -0.361766 0.592332 [DiT] Step 4/8 t=0.833 [Debug] dit_step4_vt: [2170, 64] first4: -1.024414 1.750908 1.480049 0.838379 [Debug] dit_step4_xt: [2170, 64] first4: 0.457486 1.613899 -0.520343 0.502506 [DiT] Step 5/8 t=0.750 [Debug] dit_step5_vt: [2170, 64] first4: 0.183815 2.631836 0.714966 0.507446 [Debug] dit_step5_xt: [2170, 64] first4: 0.431227 1.237923 -0.622481 0.430014 [DiT] Step 6/8 t=0.643 [Debug] dit_step6_vt: [2170, 64] first4: 0.822186 2.292429 0.498962 0.558716 [Debug] dit_step6_xt: [2170, 64] first4: 0.266790 0.779437 -0.722273 0.318271 [DiT] Step 7/8 t=0.500 [Debug] dit_step7_vt: [2170, 64] first4: -1.644043 2.355103 1.089111 0.584351 [Debug] dit_x0: [2170, 64] first4: 0.760003 0.072906 -1.049007 0.142965 [DiT] Step 8/8 t=0.300 [DiT-Generate] Total: 644.4 ms (644.4 ms/sample) [Debug] dit_output: [2170, 64] first4: 0.760003 0.072906 -1.049007 0.142965 [Store] Unload DiT (4230.2 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: Vulkan0 (CPU threads: 16) [VAE] Backend: Vulkan0, Weight buffer: 161.1 MB [VAE] Loaded: 5 blocks, upsample=1920x, F32 activations [Store] Load VAE-Dec: 672 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: 773.6 ms [Debug] vae_audio: [2, 4166400] first4: 0.003091 0.004970 0.004618 0.004363 [Store] Unload VAE-Dec (161.1 MB) [MP3] Encoding 86.8s @ 128 kbps, 48000 Hz stereo [MP3] 1388928 bytes (12.0:1), 460 ms (188.61x realtime), 32 threads [MP3] Wrote ggml-xl-turbo/request00.mp3 [Ace-Synth] All done 2026-07-12 10:16:43.337 | 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 0.999812 lyric_embed 1.000000 enc_hidden 0.999826 detok_output 0.999984 context 0.999990 noise 1.000000 temb_t 0.999994 hidden_after_proj_in 0.999985 enc_after_cond_emb 0.999819 layer0_sa_output 0.999931 hidden_after_layer0 0.999868 hidden_after_layer6 0.999812 hidden_after_layer12 0.999413 hidden_after_layer18 0.998637 hidden_after_layer23 0.998938 dit_step0_vt 0.998859 dit_step0_xt 1.000000 dit_step5_vt 0.999342 dit_step5_xt 0.999978 dit_step10_vt 0.998895 dit_step10_xt 0.999927 dit_step15_vt 0.997858 dit_step15_xt 0.999818 dit_step20_vt 0.995224 dit_step20_xt 0.999532 dit_step25_vt 0.991798 dit_step25_xt 0.998949 dit_step30_vt 0.987677 dit_step30_xt 0.997916 dit_step35_vt 0.983274 dit_step35_xt 0.996538 dit_step40_vt 0.980227 dit_step40_xt 0.995094 dit_step45_vt 0.983513 dit_step45_xt 0.994034 dit_step49_vt 0.989077 dit_x0 0.993663 vae_audio 0.970055 vae_audio (STFT cosine) 0.988686 [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 0.020817 0.001471 -0.001640 0.978490 -0.001619 0.978872 dit_step5_xt 0.999978 0.079015 0.004543 -0.008260 0.884941 -0.008480 0.883811 dit_step10_xt 0.999927 0.132437 0.007409 -0.014831 0.801089 -0.015075 0.802324 dit_step15_xt 0.999818 0.188064 0.010663 -0.021347 0.730531 -0.021933 0.731304 dit_step20_xt 0.999532 0.260642 0.015617 -0.027892 0.677526 -0.028882 0.679066 dit_step25_xt 0.998949 0.504863 0.021914 -0.034508 0.646504 -0.035643 0.650780 dit_step30_xt 0.997916 0.746289 0.030074 -0.041164 0.640807 -0.042577 0.647107 dit_step35_xt 0.996538 1.005137 0.039630 -0.047819 0.661249 -0.049471 0.669256 dit_step40_xt 0.995094 1.263057 0.050291 -0.054489 0.705492 -0.056532 0.715783 dit_step45_xt 0.994034 1.489052 0.060376 -0.061255 0.769194 -0.063533 0.780575 [Xl-turbo] steps=8, shift=3.0 | acestep-v15-xl-turbo-Q8_0.gguf [GGML] Running acestep-v15-xl-turbo-Q8_0.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 10:16:48.270 | INFO | acestep.core.generation.handler.generate_music:generate_music:304 - [generate_music] Starting generation... 2026-07-12 10:16:48.270 | INFO | acestep.core.generation.handler.generate_music:generate_music:307 - [generate_music] Preparing inputs... 2026-07-12 10:16:48.271 | 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 10:16:48.275 | INFO | acestep.core.generation.handler.generate_music:_vram_preflight_check:156 - [generate_music] VRAM pre-flight: 76.75 GB free, ~1.23 GB needed (batch=1, duration=88s, mode=xl_turbo). 2026-07-12 10:16:48.279 | INFO | acestep.core.generation.handler.conditioning_target:_prepare_target_latents_and_wavs:78 - [generate_music] Decoding audio codes for item 0... 2026-07-12 10:16:48.284 | 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 10:16:48.286 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:119 - ====================================================================== 2026-07-12 10:16:48.286 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:120 - πŸ” [DEBUG] DiT TEXT ENCODER INPUT (Inference) 2026-07-12 10:16:48.286 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:121 - ====================================================================== 2026-07-12 10:16:48.286 | 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 10:16:48.286 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:123 - ====================================================================== 2026-07-12 10:16:48.286 | 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 10:16:48.286 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:125 - ====================================================================== 2026-07-12 10:16:48.290 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:110 - [preprocess_batch] Inferring prompt embeddings... 2026-07-12 10:16:48.298 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:113 - [preprocess_batch] Inferring lyric embeddings... 2026-07-12 10:16:48.298 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:161 - [service_generate] Generating audio... (DiT backend: PyTorch (cuda)) 2026-07-12 10:16:48.304 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:263 - [service_generate] DiT diffusion via PyTorch (cuda)... 2026-07-12 10:16:48.311 | 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 10:16:48.865 | INFO | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:42 - [generate_music] Model generation completed. Decoding latents... 2026-07-12 10:16:48.866 | 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 10:16:48.866 | DEBUG | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:65 - [generate_music] time_costs: {'encoder_time_cost': 0.006372213363647461, 'diffusion_time_cost': 0.5539402961730957, 'diffusion_per_step_time_cost': 0.06924253702163696, 'total_time_cost': 0.5603125095367432, 'offload_time_cost': 0.0} 2026-07-12 10:16:48.889 | INFO | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:127 - [generate_music] Decoding latents with VAE... 2026-07-12 10:16:48.891 | DEBUG | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:136 - [generate_music] Before VAE decode: allocated=17.04GB, max=18.34GB 2026-07-12 10:16:48.892 | INFO | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:154 - [generate_music] Effective free VRAM before VAE decode: 76.66 GB 2026-07-12 10:16:48.892 | 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 10:16:48.892 | DEBUG | acestep.core.generation.handler.memory_utils:_get_auto_decode_chunk_size:108 - [_get_auto_decode_chunk_size] Effective free VRAM: 76.66 GB 2026-07-12 10:16:48.892 | DEBUG | acestep.core.generation.handler.memory_utils:_should_offload_wav_to_cpu:131 - [_should_offload_wav_to_cpu] Effective free VRAM: 76.66 GB 2026-07-12 10:16:48.892 | 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 10:16:49.168 | DEBUG | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:193 - [generate_music] After VAE decode: allocated=17.21GB, max=18.50GB 2026-07-12 10:16:49.372 | 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 10:16:49.376 | 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-Q8_0.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-Q8_0.gguf: 830 tensors, data at offset 69056 [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-sft-Q8_0.gguf: 830 tensors, data at offset 69056 [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: Vulkan0 (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: 333 ms [Debug] text_hidden: [70, 1024] first4: 3.705836 2.395382 0.221845 -13.145830 [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: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-sft-Q8_0.gguf: 830 tensors, data at offset 69056 [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, 616.6 MB into backend [Load] CondEncoder: lyric(8L), timbre(4L, CLS), text_proj, null_cond [Store] Load CondEnc: 375 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, 7.5 ms [Debug] enc_hidden: [238, 2048] first4: 0.686684 -0.019193 -0.051921 0.022813 [Store] Unload CondEnc (616.6 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: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-sft-Q8_0.gguf: 830 tensors, data at offset 69056 [WeightCtx] Loaded 30 tensors, 106.5 MB into backend [Load] Detokenizer: FSQ(6->2048) + 2L encoder(S=5, 2048->64) [Store] Load FSQ-Detok: 180 ms [Context] Decoded: 434 codes -> 2170 frames (86.8s @ 25Hz) [Build-Context Batch0] Detokenizer: 101.1 ms, 434 codes [Debug] detok_output: [2170, 64] first4: -0.124812 1.441172 0.303914 -0.631591 [Store] Unload FSQ-Detok (106.5 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: -0.124812 1.441172 0.303914 -0.631591 [Init-Noise] Starting: T=2170, S=1085, enc_S=238, steps=50, batch=1 (cover) 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: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-sft-Q8_0.gguf: 830 tensors, data at offset 69056 [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, 4230.2 MB into backend [Load] DiT: 32 layers, H=2560, Nh=32/8, D=128 [Store] Load DiT: 2381 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: 0.013019 0.105941 0.074809 -0.031887 [Debug] temb: [2560] first4: 0.009503 -0.007658 -0.006166 0.004559 [Debug] temb_t: [2560] first4: 0.007870 -0.007938 -0.006686 0.003566 [Debug] temb_r: [2560] first4: 0.001633 0.000279 0.000520 0.000994 [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: 0.033182 -0.081874 0.089398 -0.065044 [Debug] temb_lin1_r: [2560] first4: 0.006798 -0.005077 -0.002197 -0.011158 [Debug] hidden_after_proj_in: [2560, 1085] first4: 0.331421 -0.495850 0.102783 -0.213135 [Debug] proj_in_input: [192, 2170] first4: -0.124812 1.441172 0.303914 -0.631591 [Debug] enc_after_cond_emb: [2560, 238] first4: 0.033188 0.641914 0.113708 0.871178 [Debug] layer0_sa_input: [2560, 1085] first4: 0.037142 -0.019126 0.099905 -0.097168 [Debug] layer0_q_after_rope: [128, 32] first4: -0.283023 -2.258710 0.026662 -0.190462 [Debug] layer0_k_after_rope: [128, 8] first4: 0.639160 2.990234 0.689941 1.805664 [Debug] layer0_sa_output: [2560, 1085] first4: 2.169922 4.503906 -1.257812 8.242188 [Debug] layer0_attn_out: [4096, 1085] first4: 0.473449 -2.221286 -0.833027 -4.311768 [Debug] layer0_after_self_attn: [2560, 1085] first4: 0.473449 -2.221286 -0.833027 -4.311768 [Debug] layer0_after_cross_attn: [2560, 1085] first4: 0.287902 -1.939792 -0.766377 -4.366394 [Debug] hidden_after_layer0: [2560, 1085] first4: -1.724620 -1.262485 251.529099 -4.071287 [Debug] hidden_after_layer6: [2560, 1085] first4: -8.027910 3.572001 317.164886 -21.245775 [Debug] hidden_after_layer12: [2560, 1085] first4: -23.195108 39.854141 231.530487 17.963764 [Debug] hidden_after_layer18: [2560, 1085] first4: 9.784201 -12.514282 239.251495 19.611496 [Debug] hidden_after_layer31: [2560, 1085] first4: 63.149628 -104.865341 12.949883 82.768150 [Debug] dit_step0_vt: [2170, 64] first4: -0.312866 2.219116 -0.005371 1.693268 [Debug] dit_step0_xt: [2170, 64] first4: 0.200593 2.111868 -0.171768 0.813791 [DiT] Step 1/50 t=1.000 [Debug] dit_step1_vt: [2170, 64] first4: -0.335449 2.158203 0.077637 1.557648 [Debug] dit_step1_xt: [2170, 64] first4: 0.207302 2.068703 -0.173320 0.782638 [DiT] Step 2/50 t=0.980 [Debug] dit_step2_vt: [2170, 64] first4: -0.322998 2.174011 0.062500 1.565033 [Debug] dit_step2_xt: [2170, 64] first4: 0.213762 2.025223 -0.174570 0.751337 [DiT] Step 3/50 t=0.960 [Debug] dit_step3_vt: [2170, 64] first4: -0.274902 2.138550 0.058594 1.567963 [Debug] dit_step3_xt: [2170, 64] first4: 0.219260 1.982452 -0.175742 0.719978 [DiT] Step 4/50 t=0.940 [Debug] dit_step4_vt: [2170, 64] first4: -0.263794 2.058533 0.035522 1.585815 [Debug] dit_step4_xt: [2170, 64] first4: 0.224536 1.941282 -0.176453 0.688262 [DiT] Step 5/50 t=0.920 [Debug] dit_step5_vt: [2170, 64] first4: -0.252930 2.037415 0.052490 1.588287 [Debug] dit_step5_xt: [2170, 64] first4: 0.229595 1.900533 -0.177502 0.656496 [DiT] Step 6/50 t=0.900 [Debug] dit_step6_vt: [2170, 64] first4: -0.247803 2.001617 0.082886 1.615204 [Debug] dit_step6_xt: [2170, 64] first4: 0.234551 1.860501 -0.179160 0.624192 [DiT] Step 7/50 t=0.880 [Debug] dit_step7_vt: [2170, 64] first4: -0.250488 1.962677 0.121582 1.612152 [Debug] dit_step7_xt: [2170, 64] first4: 0.239561 1.821248 -0.181592 0.591949 [DiT] Step 8/50 t=0.860 [Debug] dit_step8_vt: [2170, 64] first4: -0.254272 1.926514 0.151611 1.616211 [Debug] dit_step8_xt: [2170, 64] first4: 0.244646 1.782717 -0.184624 0.559625 [DiT] Step 9/50 t=0.840 [Debug] dit_step9_vt: [2170, 64] first4: -0.280151 1.871216 0.120117 1.614929 [Debug] dit_step9_xt: [2170, 64] first4: 0.250249 1.745293 -0.187026 0.527326 [DiT] Step 10/50 t=0.820 [Debug] dit_step10_vt: [2170, 64] first4: -0.275757 1.835327 0.104492 1.609467 [Debug] dit_step10_xt: [2170, 64] first4: 0.255764 1.708586 -0.189116 0.495137 [DiT] Step 11/50 t=0.800 [Debug] dit_step11_vt: [2170, 64] first4: -0.280151 1.792511 0.095581 1.595942 [Debug] dit_step11_xt: [2170, 64] first4: 0.261367 1.672736 -0.191028 0.463218 [DiT] Step 12/50 t=0.780 [Debug] dit_step12_vt: [2170, 64] first4: -0.273560 1.743523 0.089111 1.587654 [Debug] dit_step12_xt: [2170, 64] first4: 0.266838 1.637866 -0.192810 0.431465 [DiT] Step 13/50 t=0.760 [Debug] dit_step13_vt: [2170, 64] first4: -0.262817 1.697700 0.082031 1.573341 [Debug] dit_step13_xt: [2170, 64] first4: 0.272095 1.603912 -0.194451 0.399998 [DiT] Step 14/50 t=0.740 [Debug] dit_step14_vt: [2170, 64] first4: -0.260010 1.652367 0.064819 1.567505 [Debug] dit_step14_xt: [2170, 64] first4: 0.277295 1.570864 -0.195747 0.368648 [DiT] Step 15/50 t=0.720 [Debug] dit_step15_vt: [2170, 64] first4: -0.252441 1.605118 0.049194 1.562317 [Debug] dit_step15_xt: [2170, 64] first4: 0.282344 1.538762 -0.196731 0.337402 [DiT] Step 16/50 t=0.700 [Debug] dit_step16_vt: [2170, 64] first4: -0.238037 1.558075 0.032715 1.561096 [Debug] dit_step16_xt: [2170, 64] first4: 0.287104 1.507600 -0.197385 0.306180 [DiT] Step 17/50 t=0.680 [Debug] dit_step17_vt: [2170, 64] first4: -0.210327 1.509552 0.015747 1.554138 [Debug] dit_step17_xt: [2170, 64] first4: 0.291311 1.477409 -0.197700 0.275097 [DiT] Step 18/50 t=0.660 [Debug] dit_step18_vt: [2170, 64] first4: -0.175110 1.453186 0.000854 1.540283 [Debug] dit_step18_xt: [2170, 64] first4: 0.294813 1.448346 -0.197717 0.244291 [DiT] Step 19/50 t=0.640 [Debug] dit_step19_vt: [2170, 64] first4: -0.154297 1.404602 -0.011353 1.533661 [Debug] dit_step19_xt: [2170, 64] first4: 0.297899 1.420254 -0.197490 0.213618 [DiT] Step 20/50 t=0.620 [Debug] dit_step20_vt: [2170, 64] first4: -0.126526 1.351440 -0.023438 1.535828 [Debug] dit_step20_xt: [2170, 64] first4: 0.300430 1.393225 -0.197021 0.182902 [DiT] Step 21/50 t=0.600 [Debug] dit_step21_vt: [2170, 64] first4: -0.092773 1.302368 -0.041626 1.521454 [Debug] dit_step21_xt: [2170, 64] first4: 0.302285 1.367177 -0.196189 0.152472 [DiT] Step 22/50 t=0.580 [Debug] dit_step22_vt: [2170, 64] first4: -0.070251 1.253723 -0.049805 1.507751 [Debug] dit_step22_xt: [2170, 64] first4: 0.303690 1.342103 -0.195193 0.122317 [DiT] Step 23/50 t=0.560 [Debug] dit_step23_vt: [2170, 64] first4: -0.051575 1.204895 -0.058472 1.496765 [Debug] dit_step23_xt: [2170, 64] first4: 0.304722 1.318005 -0.194023 0.092382 [DiT] Step 24/50 t=0.540 [Debug] dit_step24_vt: [2170, 64] first4: -0.036499 1.145264 -0.058838 1.472687 [Debug] dit_step24_xt: [2170, 64] first4: 0.305452 1.295100 -0.192847 0.062928 [DiT] Step 25/50 t=0.520 [Debug] dit_step25_vt: [2170, 64] first4: -0.032959 1.090088 -0.056763 1.463440 [Debug] dit_step25_xt: [2170, 64] first4: 0.306111 1.273298 -0.191711 0.033660 [DiT] Step 26/50 t=0.500 [Debug] dit_step26_vt: [2170, 64] first4: -0.032349 1.037537 -0.067871 1.444763 [Debug] dit_step26_xt: [2170, 64] first4: 0.306758 1.252547 -0.190354 0.004764 [DiT] Step 27/50 t=0.480 [Debug] dit_step27_vt: [2170, 64] first4: -0.044678 0.971283 -0.073059 1.415710 [Debug] dit_step27_xt: [2170, 64] first4: 0.307651 1.233122 -0.188893 -0.023550 [DiT] Step 28/50 t=0.460 [Debug] dit_step28_vt: [2170, 64] first4: -0.045898 0.904144 -0.090881 1.401794 [Debug] dit_step28_xt: [2170, 64] first4: 0.308569 1.215039 -0.187075 -0.051586 [DiT] Step 29/50 t=0.440 [Debug] dit_step29_vt: [2170, 64] first4: -0.063171 0.842102 -0.096558 1.379669 [Debug] dit_step29_xt: [2170, 64] first4: 0.309833 1.198197 -0.185144 -0.079179 [DiT] Step 30/50 t=0.420 [Debug] dit_step30_vt: [2170, 64] first4: -0.066345 0.771667 -0.101501 1.367737 [Debug] dit_step30_xt: [2170, 64] first4: 0.311160 1.182763 -0.183114 -0.106534 [DiT] Step 31/50 t=0.400 [Debug] dit_step31_vt: [2170, 64] first4: -0.081909 0.690781 -0.101379 1.337952 [Debug] dit_step31_xt: [2170, 64] first4: 0.312798 1.168948 -0.181086 -0.133293 [DiT] Step 32/50 t=0.380 [Debug] dit_step32_vt: [2170, 64] first4: -0.107605 0.595661 -0.115417 1.307465 [Debug] dit_step32_xt: [2170, 64] first4: 0.314950 1.157034 -0.178778 -0.159442 [DiT] Step 33/50 t=0.360 [Debug] dit_step33_vt: [2170, 64] first4: -0.114624 0.517342 -0.125671 1.291382 [Debug] dit_step33_xt: [2170, 64] first4: 0.317242 1.146688 -0.176265 -0.185270 [DiT] Step 34/50 t=0.340 [Debug] dit_step34_vt: [2170, 64] first4: -0.150330 0.430939 -0.130646 1.274536 [Debug] dit_step34_xt: [2170, 64] first4: 0.320249 1.138069 -0.173652 -0.210761 [DiT] Step 35/50 t=0.320 [Debug] dit_step35_vt: [2170, 64] first4: -0.162689 0.327362 -0.122498 1.243225 [Debug] dit_step35_xt: [2170, 64] first4: 0.323503 1.131522 -0.171202 -0.235625 [DiT] Step 36/50 t=0.300 [Debug] dit_step36_vt: [2170, 64] first4: -0.168289 0.229401 -0.117371 1.193787 [Debug] dit_step36_xt: [2170, 64] first4: 0.326869 1.126934 -0.168854 -0.259501 [DiT] Step 37/50 t=0.280 [Debug] dit_step37_vt: [2170, 64] first4: -0.189658 0.128876 -0.105804 1.173523 [Debug] dit_step37_xt: [2170, 64] first4: 0.330662 1.124356 -0.166738 -0.282971 [DiT] Step 38/50 t=0.260 [Debug] dit_step38_vt: [2170, 64] first4: -0.182701 0.032379 -0.087402 1.131287 [Debug] dit_step38_xt: [2170, 64] first4: 0.334316 1.123709 -0.164990 -0.305597 [DiT] Step 39/50 t=0.240 [Debug] dit_step39_vt: [2170, 64] first4: -0.218983 -0.073792 -0.062225 1.102051 [Debug] dit_step39_xt: [2170, 64] first4: 0.338695 1.125184 -0.163746 -0.327638 [DiT] Step 40/50 t=0.220 [Debug] dit_step40_vt: [2170, 64] first4: -0.243668 -0.180450 -0.043983 1.059204 [Debug] dit_step40_xt: [2170, 64] first4: 0.343569 1.128793 -0.162866 -0.348822 [DiT] Step 41/50 t=0.200 [Debug] dit_step41_vt: [2170, 64] first4: -0.269714 -0.284355 -0.024353 1.012207 [Debug] dit_step41_xt: [2170, 64] first4: 0.348963 1.134481 -0.162379 -0.369066 [DiT] Step 42/50 t=0.180 [Debug] dit_step42_vt: [2170, 64] first4: -0.295532 -0.379089 0.005066 0.973877 [Debug] dit_step42_xt: [2170, 64] first4: 0.354874 1.142062 -0.162480 -0.388544 [DiT] Step 43/50 t=0.160 [Debug] dit_step43_vt: [2170, 64] first4: -0.314880 -0.481735 0.032804 0.917908 [Debug] dit_step43_xt: [2170, 64] first4: 0.361171 1.151697 -0.163136 -0.406902 [DiT] Step 44/50 t=0.140 [Debug] dit_step44_vt: [2170, 64] first4: -0.342957 -0.569763 0.084442 0.856750 [Debug] dit_step44_xt: [2170, 64] first4: 0.368030 1.163092 -0.164825 -0.424037 [DiT] Step 45/50 t=0.120 [Debug] dit_step45_vt: [2170, 64] first4: -0.381866 -0.653900 0.110291 0.804565 [Debug] dit_step45_xt: [2170, 64] first4: 0.375668 1.176170 -0.167031 -0.440128 [DiT] Step 46/50 t=0.100 [Debug] dit_step46_vt: [2170, 64] first4: -0.420502 -0.725250 0.157288 0.751709 [Debug] dit_step46_xt: [2170, 64] first4: 0.384078 1.190675 -0.170177 -0.455162 [DiT] Step 47/50 t=0.080 [Debug] dit_step47_vt: [2170, 64] first4: -0.467569 -0.803406 0.217163 0.713074 [Debug] dit_step47_xt: [2170, 64] first4: 0.393429 1.206743 -0.174520 -0.469424 [DiT] Step 48/50 t=0.060 [Debug] dit_step48_vt: [2170, 64] first4: -0.527222 -0.859070 0.266846 0.694092 [Debug] dit_step48_xt: [2170, 64] first4: 0.403974 1.223925 -0.179857 -0.483306 [DiT] Step 49/50 t=0.040 [Debug] dit_step49_vt: [2170, 64] first4: -0.521616 -0.849487 0.192810 0.710449 [Debug] dit_x0: [2170, 64] first4: 0.414406 1.240915 -0.183713 -0.497515 [DiT] Step 50/50 t=0.020 [DiT-Generate] Total: 3680.1 ms (3680.1 ms/sample) [Debug] dit_output: [2170, 64] first4: 0.414406 1.240915 -0.183713 -0.497515 [Store] Unload DiT (4230.2 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: Vulkan0 (CPU threads: 16) [VAE] Backend: Vulkan0, Weight buffer: 161.1 MB [VAE] Loaded: 5 blocks, upsample=1920x, F32 activations [Store] Load VAE-Dec: 667 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: 786.4 ms [Debug] vae_audio: [2, 4166400] first4: -0.001415 -0.000694 -0.000796 -0.000290 [Store] Unload VAE-Dec (161.1 MB) [MP3] Encoding 86.8s @ 128 kbps, 48000 Hz stereo [MP3] 1388928 bytes (12.0:1), 423 ms (205.29x realtime), 32 threads [MP3] Wrote ggml-xl-sft/request00.mp3 [Ace-Synth] All done 2026-07-12 10:16:59.572 | 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 0.999812 lyric_embed 1.000000 enc_hidden 0.999545 detok_output 0.999984 context 0.999990 noise 1.000000 temb_t 0.999996 hidden_after_proj_in 0.999977 enc_after_cond_emb 0.999667 layer0_sa_output 0.989298 hidden_after_layer0 0.999109 hidden_after_layer6 0.999671 hidden_after_layer12 0.995082 hidden_after_layer18 0.995633 hidden_after_layer31 0.986932 dit_step0_vt 0.902634 dit_step0_xt 0.999824 dit_step1_vt 0.914318 dit_step1_xt 0.999420 dit_step2_vt 0.904140 dit_step2_xt 0.998337 dit_step3_vt 0.897550 dit_step3_xt 0.995675 dit_step4_vt 0.893874 dit_step4_xt 0.989287 dit_step5_vt 0.883779 dit_step5_xt 0.974320 dit_step6_vt 0.872880 dit_step6_xt 0.944957 dit_step7_vt 0.850781 dit_x0 0.898408 vae_audio 0.687474 vae_audio (STFT cosine) 0.897207 [Xl-turbo] Error growth GGML vs Python stage cos max_err mean_err mean_A std_A mean_B std_B dit_step0_xt 0.999824 0.152588 0.013150 -0.002917 0.978410 -0.002721 0.976534 dit_step1_xt 0.999420 0.323140 0.023467 -0.005813 0.952816 -0.005567 0.949004 dit_step2_xt 0.998337 0.535469 0.037881 -0.009571 0.924432 -0.009330 0.918931 dit_step3_xt 0.995675 0.830990 0.058102 -0.014962 0.893948 -0.014418 0.887241 dit_step4_xt 0.989287 1.288611 0.087588 -0.022717 0.864961 -0.021453 0.857796 dit_step5_xt 0.974320 1.962143 0.133203 -0.033761 0.848435 -0.032240 0.840033 dit_step6_xt 0.944957 2.975263 0.202123 -0.049875 0.870998 -0.049194 0.863964 [Xl-sft] steps=50, shift=1.0 | acestep-v15-xl-sft-Q8_0.gguf [GGML] Running acestep-v15-xl-sft-Q8_0.gguf... [GGML] Done, 131 dump files [Python] Initializing acestep-v15-xl-sft... Loading checkpoint shards: 0%| | 0/4 [00:00 2026-07-12 10:17:04.242 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:123 - ====================================================================== 2026-07-12 10:17:04.242 | 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 10:17:04.242 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:125 - ====================================================================== 2026-07-12 10:17:04.246 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:110 - [preprocess_batch] Inferring prompt embeddings... 2026-07-12 10:17:04.254 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:113 - [preprocess_batch] Inferring lyric embeddings... 2026-07-12 10:17:04.254 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:161 - [service_generate] Generating audio... (DiT backend: PyTorch (cuda)) 2026-07-12 10:17:04.260 | 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