[Request] Parsed ggml-turbo/request0.json [Request] seed=42 lm_batch=1 synth_batch=1 [Request] caption: An upbeat and anthemic pop-rock track driven by bright, slig... [Request] lyrics: 490 bytes [Request] bpm=83 dur=88 key=G major ts=4 lang=fr [Request] lm: temp=0.85 cfg=2.0 top_p=0.90 top_k=0 [Request] dit: steps=8 guidance=0.0 shift=3.0 [Request] task_type: text2music [Request] solver: euler (stork_substeps=10) [Request] lm_mode: generate [Request] output_format: mp3 [Request] synth_model: acestep-v15-turbo-BF16.gguf [Request] vae: vae-BF16.gguf [Request] audio_codes: (present) [Ace-Synth] Batch: 1 request(s) [Registry] Qwen3-Embedding-0.6B-BF16.gguf -> Text-Enc [Registry] Qwen3-Embedding-0.6B-Q8_0.gguf -> Text-Enc [Registry] acestep-5Hz-lm-0.6B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-0.6B-Q8_0.gguf -> LM [Registry] acestep-5Hz-lm-1.7B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-1.7B-Q8_0.gguf -> LM [Registry] acestep-5Hz-lm-4B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q5_K_M.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q6_K.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q8_0.gguf -> LM [Registry] acestep-v15-base-BF16.gguf -> DiT [Registry] acestep-v15-base-Q4_K_M.gguf -> DiT [Registry] acestep-v15-base-Q5_K_M.gguf -> DiT [Registry] acestep-v15-base-Q6_K.gguf -> DiT [Registry] acestep-v15-base-Q8_0.gguf -> DiT [Registry] acestep-v15-sft-BF16.gguf -> DiT [Registry] acestep-v15-sft-Q4_K_M.gguf -> DiT [Registry] acestep-v15-sft-Q5_K_M.gguf -> DiT [Registry] acestep-v15-sft-Q6_K.gguf -> DiT [Registry] acestep-v15-sft-Q8_0.gguf -> DiT [Registry] acestep-v15-sftturbo50-BF16.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q4_K_M.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q5_K_M.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q6_K.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-BF16.gguf -> DiT [Registry] acestep-v15-turbo-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-continuous-BF16.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-shift1-BF16.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-shift3-BF16.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-base-BF16.gguf -> DiT [Registry] acestep-v15-xl-base-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-base-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-base-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-base-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-sft-BF16.gguf -> DiT [Registry] acestep-v15-xl-sft-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-sft-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-sft-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-sft-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-BF16.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-turbo-BF16.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q8_0.gguf -> DiT [Registry] scragvae-BF16.gguf -> VAE [Registry] vae-BF16.gguf -> VAE [Store] Created (policy=STRICT) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-turbo-BF16.gguf: 678 tensors, data at offset 57024 [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-turbo-BF16.gguf: 678 tensors, data at offset 57024 [Synth-Load] Ready: turbo=yes, fa=no, batch_cfg=yes [Resolve-Params] max audio codes across batch: 434 (86.8s @ 5Hz) [Resolve-T] T=2170, S=1085 [Resolve-T] seed=42, steps=8, guidance=1.0, shift=3.0, duration=88.0s [BPE] Loaded from GGUF: 151643 vocab, 151387 merges ggml_cuda_init: found 1 CUDA devices (Total VRAM: 97247 MiB): Device 0: NVIDIA RTX PRO 6000 Blackwell Workstation Edition, compute capability 12.0, VMM: yes, VRAM: 97247 MiB load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = NVIDIA RTX PRO 6000 Blackwell Workstation Edition (NVIDIA) | uma: 0 | fp16: 1 | bf16: 0 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: NV_coopmat2 load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] TextEncoder backend: 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: 356 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-BF16.gguf: 678 tensors, data at offset 57024 [Load] LyricEncoder: 8L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [Load] TimbreEncoder: 4L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 140 tensors, 1160.5 MB into backend [Load] CondEncoder: lyric(8L), timbre(4L), text_proj, null_cond [Store] Load CondEnc: 805 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, 23.5 ms [Debug] enc_hidden: [238, 2048] first4: 1.758170 -0.049675 -0.132771 0.058468 [Store] Unload CondEnc (1160.5 MB) load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] Detokenizer backend: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-turbo-BF16.gguf: 678 tensors, data at offset 57024 [WeightCtx] Loaded 30 tensors, 200.3 MB into backend [Load] Detokenizer: FSQ(6->2048) + 2L encoder(S=5, 2048->64) [Store] Load FSQ-Detok: 231 ms [Context] Decoded: 434 codes -> 2170 frames (86.8s @ 25Hz) [Build-Context Batch0] Detokenizer: 192.6 ms, 434 codes [Debug] detok_output: [2170, 64] first4: -0.125156 1.434936 0.308101 -0.624459 [Store] Unload FSQ-Detok (200.3 MB) [Init-Noise Batch0] Philox noise seed=42, [2170, 64] solver=euler [Debug] noise: [2170, 64] first4: 0.194336 2.156250 -0.171875 0.847656 [Debug] context: [2170, 128] first4: -0.125156 1.434936 0.308101 -0.624459 [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-BF16.gguf: 678 tensors, data at offset 57024 [DiT] Self-attn: Q+K+V fused [DiT] Cross-attn: Q+K+V fused [DiT] MLP: gate+up fused [Load] null_condition_emb found (CFG available) [WeightCtx] Loaded 478 tensors, 3007.9 MB into backend [Load] DiT: 24 layers, H=2048, Nh=16/8, D=128 [Store] Load DiT: 1576 ms [DiT] Batch N=1, T=2170, S=1085, enc_S=238 [DiT] Graph: 2129 nodes [DiT] Solver: ODE Euler (1 NFE/step, order 1) [Debug] tproj: [12288] first4: 0.260062 -0.161562 -0.097030 0.052313 [Debug] temb: [2048] first4: 0.000069 -0.132499 -0.035430 0.064753 [Debug] temb_t: [2048] first4: 0.001065 0.026818 -0.052754 0.063717 [Debug] temb_r: [2048] first4: -0.000996 -0.159317 0.017323 0.001036 [Debug] sinusoidal_t: [256] first4: 0.562486 0.789701 0.439822 -0.023583 [Debug] sinusoidal_r: [256] first4: 1.000000 1.000000 1.000000 1.000000 [Debug] temb_lin1_t: [2048] first4: -0.049318 -0.051829 -0.014251 -0.038444 [Debug] temb_lin1_r: [2048] first4: -0.013266 -0.018319 -0.016375 0.008532 [Debug] hidden_after_proj_in: [2048, 1085] first4: -1.039551 -0.968811 0.536133 0.446747 [Debug] proj_in_input: [192, 2170] first4: -0.125156 1.434936 0.308101 -0.624459 [Debug] enc_after_cond_emb: [2048, 238] first4: -0.168728 0.815327 0.326482 -0.561758 [Debug] layer0_sa_input: [2048, 1085] first4: -0.719109 -0.763732 -0.047328 0.261808 [Debug] layer0_q_after_rope: [128, 16] first4: -1.541024 -1.043470 0.190593 0.456097 [Debug] layer0_k_after_rope: [128, 8] first4: -0.168728 0.815327 0.326482 -0.561758 [Debug] layer0_sa_output: [2048, 1085] first4: -1.502017 0.168331 -0.353880 0.513236 [Debug] layer0_attn_out: [2048, 1085] first4: -1.541024 -1.043470 0.190593 0.456097 [Debug] layer0_after_self_attn: [2048, 1085] first4: -1.541024 -1.043470 0.190593 0.456097 [Debug] layer0_after_cross_attn: [2048, 1085] first4: -1.596669 -0.820366 -0.285782 0.493791 [Debug] hidden_after_layer0: [2048, 1085] first4: -9.084208 0.569350 52.448559 -0.903117 [Debug] hidden_after_layer6: [2048, 1085] first4: -21.330507 0.021903 33.695942 -4.504014 [Debug] hidden_after_layer12: [2048, 1085] first4: -15.037347 -17.892197 72.419395 28.877674 [Debug] hidden_after_layer18: [2048, 1085] first4: -27.284096 15.491617 60.516476 21.017166 [Debug] hidden_after_layer23: [2048, 1085] first4: -10.321899 45.379723 195.820038 145.172882 [Debug] dit_step0_vt: [2170, 64] first4: 0.020525 1.109125 0.339710 2.374054 [Debug] dit_step0_xt: [2170, 64] first4: 0.193403 2.105835 -0.187316 0.739745 [DiT] Step 1/8 t=1.000 [Debug] dit_step1_vt: [2170, 64] first4: -0.223236 1.308655 -0.131187 1.873806 [Debug] dit_step1_xt: [2170, 64] first4: 0.205580 2.034454 -0.180161 0.637537 [DiT] Step 2/8 t=0.955 [Debug] dit_step2_vt: [2170, 64] first4: -0.016788 1.210950 0.101990 2.389740 [Debug] dit_step2_xt: [2170, 64] first4: 0.206699 1.953724 -0.186960 0.478221 [DiT] Step 3/8 t=0.900 [Debug] dit_step3_vt: [2170, 64] first4: 0.256821 1.092899 0.261383 2.647095 [Debug] dit_step3_xt: [2170, 64] first4: 0.185297 1.862649 -0.208742 0.257630 [DiT] Step 4/8 t=0.833 [Debug] dit_step4_vt: [2170, 64] first4: 0.285187 1.001862 0.106995 2.708008 [Debug] dit_step4_xt: [2170, 64] first4: 0.154741 1.755307 -0.220206 -0.032514 [DiT] Step 5/8 t=0.750 [Debug] dit_step5_vt: [2170, 64] first4: 0.266907 0.920441 -0.298645 2.768204 [Debug] dit_step5_xt: [2170, 64] first4: 0.116612 1.623815 -0.177542 -0.427972 [DiT] Step 6/8 t=0.643 [Debug] dit_step6_vt: [2170, 64] first4: 0.172360 0.642120 -0.830612 2.822205 [Debug] dit_step6_xt: [2170, 64] first4: 0.082140 1.495391 -0.011420 -0.992412 [DiT] Step 7/8 t=0.500 [Debug] dit_step7_vt: [2170, 64] first4: 0.001678 0.189392 -1.477936 3.121216 [Debug] dit_x0: [2170, 64] first4: 0.081636 1.438574 0.431961 -1.928777 [DiT] Step 8/8 t=0.300 [DiT-Generate] Total: 766.1 ms (766.1 ms/sample) [Debug] dit_output: [2170, 64] first4: 0.081636 1.438574 0.431961 -1.928777 [Store] Unload DiT (3007.9 MB) [GGUF] /mnt/workspace/git/acestep.cpp/models/vae-BF16.gguf: 365 tensors, data at offset 30048 load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] VAE backend: Vulkan0 (CPU threads: 16) [VAE] Backend: Vulkan0, Weight buffer: 161.1 MB [VAE] Loaded: 5 blocks, upsample=1920x, F32 activations [Store] Load VAE-Dec: 688 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: 980.9 ms [Debug] vae_audio: [2, 4166400] first4: 0.000694 0.001404 0.001172 0.001701 [Store] Unload VAE-Dec (161.1 MB) [MP3] Encoding 86.8s @ 128 kbps, 48000 Hz stereo [MP3] 1388928 bytes (12.0:1), 450 ms (192.79x realtime), 32 threads [MP3] Wrote ggml-turbo/request00.mp3 [Ace-Synth] All done 2026-07-12 10:19:13.100 | 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 10:19:14.831 | WARNING | acestep.training.trainer::40 - bitsandbytes not installed. Using standard AdamW. 2026-07-12 10:19:14.833 | 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 10:19:14.927 | 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 10:19:16.556 | 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 10:19:16.556 | INFO | acestep.core.generation.handler.generate_music:generate_music:304 - [generate_music] Starting generation... 2026-07-12 10:19:16.556 | INFO | acestep.core.generation.handler.generate_music:generate_music:307 - [generate_music] Preparing inputs... 2026-07-12 10:19:16.556 | 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 10:19:16.559 | 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 10:19:16.565 | INFO | acestep.core.generation.handler.conditioning_target:_prepare_target_latents_and_wavs:78 - [generate_music] Decoding audio codes for item 0... 2026-07-12 10:19:16.756 | 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 10:19:16.758 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:119 - ====================================================================== 2026-07-12 10:19:16.758 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:120 - πŸ” [DEBUG] DiT TEXT ENCODER INPUT (Inference) 2026-07-12 10:19:16.758 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:121 - ====================================================================== 2026-07-12 10:19:16.758 | 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 10:19:16.758 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:123 - ====================================================================== 2026-07-12 10:19:16.758 | 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 10:19:16.758 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:125 - ====================================================================== 2026-07-12 10:19:16.765 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:110 - [preprocess_batch] Inferring prompt embeddings... 2026-07-12 10:19:16.780 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:113 - [preprocess_batch] Inferring lyric embeddings... 2026-07-12 10:19:16.780 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:161 - [service_generate] Generating audio... (DiT backend: PyTorch (cuda)) 2026-07-12 10:19:16.815 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:263 - [service_generate] DiT diffusion via PyTorch (cuda)... 2026-07-12 10:19:16.823 | 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 10:19:16.907 | 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 10:19:17.133 | INFO | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:42 - [generate_music] Model generation completed. Decoding latents... 2026-07-12 10:19:17.134 | 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 10:19:17.134 | DEBUG | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:65 - [generate_music] time_costs: {'encoder_time_cost': 0.006901264190673828, 'diffusion_time_cost': 0.3100569248199463, 'diffusion_per_step_time_cost': 0.038757115602493286, 'total_time_cost': 0.3169581890106201, 'offload_time_cost': 0.0} 2026-07-12 10:19:17.148 | INFO | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:127 - [generate_music] Decoding latents with VAE... 2026-07-12 10:19:17.150 | 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 10:19:17.150 | 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 10:19:17.150 | 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 10:19:17.150 | 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 10:19:17.150 | 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 10:19:17.150 | 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 10:19:17.427 | 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 10:19:17.586 | 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 10:19:17.589 | INFO | acestep.core.generation.handler.generate_music_payload:_build_generate_music_success_payload:50 - [generate_music] Done! Generated 1 audio tensors. [Request] Parsed ggml-sft/request0.json [Request] seed=42 lm_batch=1 synth_batch=1 [Request] caption: An upbeat and anthemic pop-rock track driven by bright, slig... [Request] lyrics: 490 bytes [Request] bpm=83 dur=88 key=G major ts=4 lang=fr [Request] lm: temp=0.85 cfg=2.0 top_p=0.90 top_k=0 [Request] dit: steps=50 guidance=1.0 shift=1.0 [Request] task_type: text2music [Request] solver: euler (stork_substeps=10) [Request] lm_mode: generate [Request] output_format: mp3 [Request] synth_model: acestep-v15-sft-BF16.gguf [Request] vae: vae-BF16.gguf [Request] audio_codes: (present) [Ace-Synth] Batch: 1 request(s) [Registry] Qwen3-Embedding-0.6B-BF16.gguf -> Text-Enc [Registry] Qwen3-Embedding-0.6B-Q8_0.gguf -> Text-Enc [Registry] acestep-5Hz-lm-0.6B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-0.6B-Q8_0.gguf -> LM [Registry] acestep-5Hz-lm-1.7B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-1.7B-Q8_0.gguf -> LM [Registry] acestep-5Hz-lm-4B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q5_K_M.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q6_K.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q8_0.gguf -> LM [Registry] acestep-v15-base-BF16.gguf -> DiT [Registry] acestep-v15-base-Q4_K_M.gguf -> DiT [Registry] acestep-v15-base-Q5_K_M.gguf -> DiT [Registry] acestep-v15-base-Q6_K.gguf -> DiT [Registry] acestep-v15-base-Q8_0.gguf -> DiT [Registry] acestep-v15-sft-BF16.gguf -> DiT [Registry] acestep-v15-sft-Q4_K_M.gguf -> DiT [Registry] acestep-v15-sft-Q5_K_M.gguf -> DiT [Registry] acestep-v15-sft-Q6_K.gguf -> DiT [Registry] acestep-v15-sft-Q8_0.gguf -> DiT [Registry] acestep-v15-sftturbo50-BF16.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q4_K_M.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q5_K_M.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q6_K.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-BF16.gguf -> DiT [Registry] acestep-v15-turbo-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-continuous-BF16.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-shift1-BF16.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-shift3-BF16.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-base-BF16.gguf -> DiT [Registry] acestep-v15-xl-base-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-base-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-base-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-base-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-sft-BF16.gguf -> DiT [Registry] acestep-v15-xl-sft-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-sft-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-sft-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-sft-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-BF16.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-turbo-BF16.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q8_0.gguf -> DiT [Registry] scragvae-BF16.gguf -> VAE [Registry] vae-BF16.gguf -> VAE [Store] Created (policy=STRICT) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-sft-BF16.gguf: 678 tensors, data at offset 56960 [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-sft-BF16.gguf: 678 tensors, data at offset 56960 [Synth-Load] Ready: turbo=no, fa=no, batch_cfg=yes [Resolve-Params] max audio codes across batch: 434 (86.8s @ 5Hz) [Resolve-T] T=2170, S=1085 [Resolve-T] seed=42, steps=50, guidance=1.0, shift=1.0, duration=88.0s [BPE] Loaded from GGUF: 151643 vocab, 151387 merges ggml_cuda_init: found 1 CUDA devices (Total VRAM: 97247 MiB): Device 0: NVIDIA RTX PRO 6000 Blackwell Workstation Edition, compute capability 12.0, VMM: yes, VRAM: 97247 MiB load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = NVIDIA RTX PRO 6000 Blackwell Workstation Edition (NVIDIA) | uma: 0 | fp16: 1 | bf16: 0 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: NV_coopmat2 load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] TextEncoder backend: 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: 294 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-BF16.gguf: 678 tensors, data at offset 56960 [Load] LyricEncoder: 8L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [Load] TimbreEncoder: 4L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 140 tensors, 1160.5 MB into backend [Load] CondEncoder: lyric(8L), timbre(4L), text_proj, null_cond [Store] Load CondEnc: 784 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, 12.0 ms [Debug] enc_hidden: [238, 2048] first4: 1.758170 -0.049675 -0.132771 0.058468 [Store] Unload CondEnc (1160.5 MB) load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] Detokenizer backend: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-sft-BF16.gguf: 678 tensors, data at offset 56960 [WeightCtx] Loaded 30 tensors, 200.3 MB into backend [Load] Detokenizer: FSQ(6->2048) + 2L encoder(S=5, 2048->64) [Store] Load FSQ-Detok: 112 ms [Context] Decoded: 434 codes -> 2170 frames (86.8s @ 25Hz) [Build-Context Batch0] Detokenizer: 183.9 ms, 434 codes [Debug] detok_output: [2170, 64] first4: -0.125156 1.434936 0.308101 -0.624459 [Store] Unload FSQ-Detok (200.3 MB) [Init-Noise Batch0] Philox noise seed=42, [2170, 64] solver=euler [Debug] noise: [2170, 64] first4: 0.194336 2.156250 -0.171875 0.847656 [Debug] context: [2170, 128] first4: -0.125156 1.434936 0.308101 -0.624459 [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-BF16.gguf: 678 tensors, data at offset 56960 [DiT] Self-attn: Q+K+V fused [DiT] Cross-attn: Q+K+V fused [DiT] MLP: gate+up fused [Load] null_condition_emb found (CFG available) [WeightCtx] Loaded 478 tensors, 3007.9 MB into backend [Load] DiT: 24 layers, H=2048, Nh=16/8, D=128 [Store] Load DiT: 1515 ms [DiT] Batch N=1, T=2170, S=1085, enc_S=238 [DiT] Graph: 2129 nodes [DiT] Solver: ODE Euler (1 NFE/step, order 1) [Debug] tproj: [12288] first4: 0.154506 -0.116115 -0.086937 0.082315 [Debug] temb: [2048] first4: -0.002757 -0.176460 0.004187 -0.001983 [Debug] temb_t: [2048] first4: -0.001235 0.003330 -0.013135 -0.002086 [Debug] temb_r: [2048] first4: -0.001522 -0.179790 0.017322 0.000103 [Debug] sinusoidal_t: [256] first4: 0.562486 0.789701 0.439822 -0.023583 [Debug] sinusoidal_r: [256] first4: 1.000000 1.000000 1.000000 1.000000 [Debug] temb_lin1_t: [2048] first4: -0.039237 0.029644 0.028280 -0.024901 [Debug] temb_lin1_r: [2048] first4: 0.002274 0.002999 -0.001040 -0.000388 [Debug] hidden_after_proj_in: [2048, 1085] first4: -1.089844 -0.924561 0.498047 0.441437 [Debug] proj_in_input: [192, 2170] first4: -0.125156 1.434936 0.308101 -0.624459 [Debug] enc_after_cond_emb: [2048, 238] first4: -0.188777 0.876423 0.328321 -0.576760 [Debug] layer0_sa_input: [2048, 1085] first4: -0.915453 -0.715979 -0.037661 0.294912 [Debug] layer0_q_after_rope: [128, 16] first4: -1.802096 -1.382173 -0.130030 0.380215 [Debug] layer0_k_after_rope: [128, 8] first4: -0.188777 0.876423 0.328321 -0.576760 [Debug] layer0_sa_output: [2048, 1085] first4: -1.645017 0.807700 -0.555845 0.545196 [Debug] layer0_attn_out: [2048, 1085] first4: -1.802096 -1.382173 -0.130030 0.380215 [Debug] layer0_after_self_attn: [2048, 1085] first4: -1.802096 -1.382173 -0.130030 0.380215 [Debug] layer0_after_cross_attn: [2048, 1085] first4: -1.962266 -1.046275 -0.402873 0.386165 [Debug] hidden_after_layer0: [2048, 1085] first4: -9.440218 1.158031 57.820908 -1.608496 [Debug] hidden_after_layer6: [2048, 1085] first4: -17.081692 4.311724 57.786755 -2.362362 [Debug] hidden_after_layer12: [2048, 1085] first4: -9.821053 4.756193 -21.218874 -1.934380 [Debug] hidden_after_layer18: [2048, 1085] first4: -3.294344 19.068089 -36.475990 -11.643459 [Debug] hidden_after_layer23: [2048, 1085] first4: 34.190544 63.167038 58.399567 -9.609961 [Debug] dit_step0_vt: [2170, 64] first4: -0.577759 2.522583 -0.201225 1.467411 [Debug] dit_step0_xt: [2170, 64] first4: 0.205891 2.105798 -0.167850 0.818308 [DiT] Step 1/50 t=1.000 [Debug] dit_step1_vt: [2170, 64] first4: -0.644867 2.532471 -0.029938 1.302536 [Debug] dit_step1_xt: [2170, 64] first4: 0.218788 2.055149 -0.167252 0.792257 [DiT] Step 2/50 t=0.980 [Debug] dit_step2_vt: [2170, 64] first4: -0.641479 2.486816 0.004578 1.318710 [Debug] dit_step2_xt: [2170, 64] first4: 0.231618 2.005413 -0.167343 0.765883 [DiT] Step 3/50 t=0.960 [Debug] dit_step3_vt: [2170, 64] first4: -0.623413 2.432739 0.030487 1.326416 [Debug] dit_step3_xt: [2170, 64] first4: 0.244086 1.956758 -0.167953 0.739355 [DiT] Step 4/50 t=0.940 [Debug] dit_step4_vt: [2170, 64] first4: -0.567078 2.345581 0.010864 1.325562 [Debug] dit_step4_xt: [2170, 64] first4: 0.255428 1.909846 -0.168170 0.712844 [DiT] Step 5/50 t=0.920 [Debug] dit_step5_vt: [2170, 64] first4: -0.526855 2.277954 -0.016006 1.311691 [Debug] dit_step5_xt: [2170, 64] first4: 0.265965 1.864287 -0.167850 0.686610 [DiT] Step 6/50 t=0.900 [Debug] dit_step6_vt: [2170, 64] first4: -0.501343 2.225342 -0.050659 1.297791 [Debug] dit_step6_xt: [2170, 64] first4: 0.275992 1.819780 -0.166837 0.660654 [DiT] Step 7/50 t=0.880 [Debug] dit_step7_vt: [2170, 64] first4: -0.459778 2.151978 -0.071793 1.282288 [Debug] dit_step7_xt: [2170, 64] first4: 0.285187 1.776741 -0.165401 0.635008 [DiT] Step 8/50 t=0.860 [Debug] dit_step8_vt: [2170, 64] first4: -0.397400 2.068115 -0.082733 1.265411 [Debug] dit_step8_xt: [2170, 64] first4: 0.293135 1.735379 -0.163746 0.609700 [DiT] Step 9/50 t=0.840 [Debug] dit_step9_vt: [2170, 64] first4: -0.333618 1.978210 -0.083191 1.254608 [Debug] dit_step9_xt: [2170, 64] first4: 0.299808 1.695814 -0.162083 0.584608 [DiT] Step 10/50 t=0.820 [Debug] dit_step10_vt: [2170, 64] first4: -0.273193 1.893555 -0.068954 1.245728 [Debug] dit_step10_xt: [2170, 64] first4: 0.305272 1.657943 -0.160704 0.559693 [DiT] Step 11/50 t=0.800 [Debug] dit_step11_vt: [2170, 64] first4: -0.228699 1.815247 -0.067474 1.236221 [Debug] dit_step11_xt: [2170, 64] first4: 0.309846 1.621638 -0.159354 0.534969 [DiT] Step 12/50 t=0.780 [Debug] dit_step12_vt: [2170, 64] first4: -0.206116 1.743378 -0.078705 1.224182 [Debug] dit_step12_xt: [2170, 64] first4: 0.313968 1.586771 -0.157780 0.510485 [DiT] Step 13/50 t=0.760 [Debug] dit_step13_vt: [2170, 64] first4: -0.192352 1.670929 -0.091339 1.215729 [Debug] dit_step13_xt: [2170, 64] first4: 0.317815 1.553352 -0.155953 0.486171 [DiT] Step 14/50 t=0.740 [Debug] dit_step14_vt: [2170, 64] first4: -0.188141 1.592621 -0.121552 1.189362 [Debug] dit_step14_xt: [2170, 64] first4: 0.321578 1.521500 -0.153522 0.462383 [DiT] Step 15/50 t=0.720 [Debug] dit_step15_vt: [2170, 64] first4: -0.186563 1.511368 -0.158478 1.176605 [Debug] dit_step15_xt: [2170, 64] first4: 0.325309 1.491272 -0.150353 0.438851 [DiT] Step 16/50 t=0.700 [Debug] dit_step16_vt: [2170, 64] first4: -0.188347 1.429409 -0.185303 1.158127 [Debug] dit_step16_xt: [2170, 64] first4: 0.329076 1.462684 -0.146647 0.415689 [DiT] Step 17/50 t=0.680 [Debug] dit_step17_vt: [2170, 64] first4: -0.205124 1.354624 -0.211731 1.138115 [Debug] dit_step17_xt: [2170, 64] first4: 0.333178 1.435592 -0.142412 0.392926 [DiT] Step 18/50 t=0.660 [Debug] dit_step18_vt: [2170, 64] first4: -0.219360 1.273415 -0.237000 1.117798 [Debug] dit_step18_xt: [2170, 64] first4: 0.337566 1.410123 -0.137672 0.370570 [DiT] Step 19/50 t=0.640 [Debug] dit_step19_vt: [2170, 64] first4: -0.233444 1.194870 -0.259460 1.096102 [Debug] dit_step19_xt: [2170, 64] first4: 0.342234 1.386226 -0.132483 0.348648 [DiT] Step 20/50 t=0.620 [Debug] dit_step20_vt: [2170, 64] first4: -0.257217 1.107117 -0.274475 1.072788 [Debug] dit_step20_xt: [2170, 64] first4: 0.347379 1.364084 -0.126993 0.327193 [DiT] Step 21/50 t=0.600 [Debug] dit_step21_vt: [2170, 64] first4: -0.280533 1.013184 -0.282593 1.047379 [Debug] dit_step21_xt: [2170, 64] first4: 0.352990 1.343820 -0.121341 0.306245 [DiT] Step 22/50 t=0.580 [Debug] dit_step22_vt: [2170, 64] first4: -0.302643 0.913239 -0.291199 1.017487 [Debug] dit_step22_xt: [2170, 64] first4: 0.359042 1.325555 -0.115517 0.285895 [DiT] Step 23/50 t=0.560 [Debug] dit_step23_vt: [2170, 64] first4: -0.324524 0.808472 -0.291992 0.997696 [Debug] dit_step23_xt: [2170, 64] first4: 0.365533 1.309386 -0.109678 0.265941 [DiT] Step 24/50 t=0.540 [Debug] dit_step24_vt: [2170, 64] first4: -0.349060 0.704346 -0.299500 0.966568 [Debug] dit_step24_xt: [2170, 64] first4: 0.372514 1.295299 -0.103688 0.246610 [DiT] Step 25/50 t=0.520 [Debug] dit_step25_vt: [2170, 64] first4: -0.372643 0.591736 -0.297119 0.947220 [Debug] dit_step25_xt: [2170, 64] first4: 0.379967 1.283464 -0.097745 0.227666 [DiT] Step 26/50 t=0.500 [Debug] dit_step26_vt: [2170, 64] first4: -0.401146 0.476196 -0.290527 0.925568 [Debug] dit_step26_xt: [2170, 64] first4: 0.387990 1.273940 -0.091935 0.209154 [DiT] Step 27/50 t=0.480 [Debug] dit_step27_vt: [2170, 64] first4: -0.420921 0.364594 -0.290466 0.899475 [Debug] dit_step27_xt: [2170, 64] first4: 0.396408 1.266648 -0.086125 0.191165 [DiT] Step 28/50 t=0.460 [Debug] dit_step28_vt: [2170, 64] first4: -0.444221 0.245834 -0.281799 0.880463 [Debug] dit_step28_xt: [2170, 64] first4: 0.405293 1.261732 -0.080489 0.173556 [DiT] Step 29/50 t=0.440 [Debug] dit_step29_vt: [2170, 64] first4: -0.467411 0.127078 -0.268799 0.865479 [Debug] dit_step29_xt: [2170, 64] first4: 0.414641 1.259190 -0.075113 0.156246 [DiT] Step 30/50 t=0.420 [Debug] dit_step30_vt: [2170, 64] first4: -0.489529 0.012787 -0.263550 0.850281 [Debug] dit_step30_xt: [2170, 64] first4: 0.424432 1.258934 -0.069842 0.139240 [DiT] Step 31/50 t=0.400 [Debug] dit_step31_vt: [2170, 64] first4: -0.500151 -0.102364 -0.249603 0.824768 [Debug] dit_step31_xt: [2170, 64] first4: 0.434435 1.260982 -0.064850 0.122745 [DiT] Step 32/50 t=0.380 [Debug] dit_step32_vt: [2170, 64] first4: -0.516973 -0.210785 -0.236267 0.811188 [Debug] dit_step32_xt: [2170, 64] first4: 0.444774 1.265198 -0.060125 0.106521 [DiT] Step 33/50 t=0.360 [Debug] dit_step33_vt: [2170, 64] first4: -0.534254 -0.322571 -0.217377 0.779510 [Debug] dit_step33_xt: [2170, 64] first4: 0.455459 1.271649 -0.055777 0.090931 [DiT] Step 34/50 t=0.340 [Debug] dit_step34_vt: [2170, 64] first4: -0.544510 -0.427429 -0.194641 0.762451 [Debug] dit_step34_xt: [2170, 64] first4: 0.466349 1.280198 -0.051885 0.075682 [DiT] Step 35/50 t=0.320 [Debug] dit_step35_vt: [2170, 64] first4: -0.549725 -0.527931 -0.171284 0.735779 [Debug] dit_step35_xt: [2170, 64] first4: 0.477344 1.290756 -0.048459 0.060966 [DiT] Step 36/50 t=0.300 [Debug] dit_step36_vt: [2170, 64] first4: -0.555817 -0.624332 -0.148249 0.711670 [Debug] dit_step36_xt: [2170, 64] first4: 0.488460 1.303243 -0.045494 0.046733 [DiT] Step 37/50 t=0.280 [Debug] dit_step37_vt: [2170, 64] first4: -0.561295 -0.713199 -0.125015 0.677490 [Debug] dit_step37_xt: [2170, 64] first4: 0.499686 1.317507 -0.042994 0.033183 [DiT] Step 38/50 t=0.260 [Debug] dit_step38_vt: [2170, 64] first4: -0.553345 -0.796474 -0.095749 0.633972 [Debug] dit_step38_xt: [2170, 64] first4: 0.510753 1.333436 -0.041079 0.020504 [DiT] Step 39/50 t=0.240 [Debug] dit_step39_vt: [2170, 64] first4: -0.552345 -0.877983 -0.070312 0.598022 [Debug] dit_step39_xt: [2170, 64] first4: 0.521800 1.350996 -0.039672 0.008543 [DiT] Step 40/50 t=0.220 [Debug] dit_step40_vt: [2170, 64] first4: -0.540878 -0.945717 -0.033691 0.553314 [Debug] dit_step40_xt: [2170, 64] first4: 0.532617 1.369910 -0.038999 -0.002523 [DiT] Step 41/50 t=0.200 [Debug] dit_step41_vt: [2170, 64] first4: -0.536324 -1.011963 -0.001358 0.499817 [Debug] dit_step41_xt: [2170, 64] first4: 0.543344 1.390149 -0.038971 -0.012519 [DiT] Step 42/50 t=0.180 [Debug] dit_step42_vt: [2170, 64] first4: -0.522820 -1.070923 0.047810 0.460846 [Debug] dit_step42_xt: [2170, 64] first4: 0.553800 1.411568 -0.039928 -0.021736 [DiT] Step 43/50 t=0.160 [Debug] dit_step43_vt: [2170, 64] first4: -0.510788 -1.127708 0.086933 0.406555 [Debug] dit_step43_xt: [2170, 64] first4: 0.564016 1.434122 -0.041666 -0.029867 [DiT] Step 44/50 t=0.140 [Debug] dit_step44_vt: [2170, 64] first4: -0.485828 -1.173325 0.128616 0.353271 [Debug] dit_step44_xt: [2170, 64] first4: 0.573732 1.457588 -0.044239 -0.036933 [DiT] Step 45/50 t=0.120 [Debug] dit_step45_vt: [2170, 64] first4: -0.469360 -1.215393 0.166260 0.296356 [Debug] dit_step45_xt: [2170, 64] first4: 0.583120 1.481896 -0.047564 -0.042860 [DiT] Step 46/50 t=0.100 [Debug] dit_step46_vt: [2170, 64] first4: -0.445129 -1.255005 0.192190 0.236382 [Debug] dit_step46_xt: [2170, 64] first4: 0.592022 1.506997 -0.051408 -0.047587 [DiT] Step 47/50 t=0.080 [Debug] dit_step47_vt: [2170, 64] first4: -0.414398 -1.285965 0.211578 0.199039 [Debug] dit_step47_xt: [2170, 64] first4: 0.600310 1.532716 -0.055639 -0.051568 [DiT] Step 48/50 t=0.060 [Debug] dit_step48_vt: [2170, 64] first4: -0.368317 -1.309799 0.222214 0.152954 [Debug] dit_step48_xt: [2170, 64] first4: 0.607677 1.558912 -0.060083 -0.054627 [DiT] Step 49/50 t=0.040 [Debug] dit_step49_vt: [2170, 64] first4: -0.406799 -1.340286 0.210892 0.150299 [Debug] dit_x0: [2170, 64] first4: 0.615813 1.585717 -0.064301 -0.057633 [DiT] Step 50/50 t=0.020 [DiT-Generate] Total: 4619.6 ms (4619.6 ms/sample) [Debug] dit_output: [2170, 64] first4: 0.615813 1.585717 -0.064301 -0.057633 [Store] Unload DiT (3007.9 MB) [GGUF] /mnt/workspace/git/acestep.cpp/models/vae-BF16.gguf: 365 tensors, data at offset 30048 load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] VAE backend: Vulkan0 (CPU threads: 16) [VAE] Backend: Vulkan0, Weight buffer: 161.1 MB [VAE] Loaded: 5 blocks, upsample=1920x, F32 activations [Store] Load VAE-Dec: 682 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: 969.9 ms [Debug] vae_audio: [2, 4166400] first4: -0.001512 -0.000972 -0.001253 -0.000836 [Store] Unload VAE-Dec (161.1 MB) [MP3] Encoding 86.8s @ 128 kbps, 48000 Hz stereo [MP3] 1388928 bytes (12.0:1), 411 ms (211.29x realtime), 32 threads [MP3] Wrote ggml-sft/request00.mp3 [Ace-Synth] All done 2026-07-12 10:19:28.321 | 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 10:19:29.507 | INFO | acestep.core.generation.handler.generate_music:generate_music:304 - [generate_music] Starting generation... 2026-07-12 10:19:29.507 | INFO | acestep.core.generation.handler.generate_music:generate_music:307 - [generate_music] Preparing inputs... 2026-07-12 10:19:29.508 | 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 10:19:29.511 | 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 10:19:29.515 | INFO | acestep.core.generation.handler.conditioning_target:_prepare_target_latents_and_wavs:78 - [generate_music] Decoding audio codes for item 0... 2026-07-12 10:19:29.520 | 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 10:19:29.522 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:119 - ====================================================================== 2026-07-12 10:19:29.522 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:120 - πŸ” [DEBUG] DiT TEXT ENCODER INPUT (Inference) 2026-07-12 10:19:29.522 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:121 - ====================================================================== 2026-07-12 10:19:29.522 | 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 10:19:29.522 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:123 - ====================================================================== 2026-07-12 10:19:29.522 | 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 10:19:29.522 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:125 - ====================================================================== 2026-07-12 10:19:29.526 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:110 - [preprocess_batch] Inferring prompt embeddings... 2026-07-12 10:19:29.533 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:113 - [preprocess_batch] Inferring lyric embeddings... 2026-07-12 10:19:29.533 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:161 - [service_generate] Generating audio... (DiT backend: PyTorch (cuda)) 2026-07-12 10:19:29.540 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:263 - [service_generate] DiT diffusion via PyTorch (cuda)... [Request] Loaded request0.json [Turbo] steps=8, shift=3.0 | acestep-v15-turbo-BF16.gguf [GGML] Running acestep-v15-turbo-BF16.gguf... [GGML] Done, 47 dump files [Python] Initializing acestep-v15-turbo... [Python] Generating (acestep-v15-turbo, 8 steps)... Using precomputed LM hints Using precomputed LM hints [Python] Wrote python-turbo/output.wav: 4166400 samples (86.80s @ 48kHz stereo) [Python] Done, 40 dump files [Turbo] Cosine similarities GGML vs Python stage GGML vs Python text_hidden 0.999817 lyric_embed 1.000000 enc_hidden 0.999845 detok_output 0.999997 context 0.999998 noise 1.000000 temb_t 0.999999 hidden_after_proj_in 0.999988 enc_after_cond_emb 0.999833 layer0_sa_output 0.999960 hidden_after_layer0 0.999980 hidden_after_layer6 0.999916 hidden_after_layer12 0.999292 hidden_after_layer18 0.996684 hidden_after_layer23 0.993774 dit_step0_vt 0.975793 dit_step0_xt 0.999947 dit_step1_vt 0.980075 dit_step1_xt 0.999836 dit_step2_vt 0.982039 dit_step2_xt 0.999561 dit_step3_vt 0.982616 dit_step3_xt 0.998939 dit_step4_vt 0.981174 dit_step4_xt 0.997537 dit_step5_vt 0.978622 dit_step5_xt 0.994407 dit_step6_vt 0.975586 dit_step6_xt 0.988459 dit_step7_vt 0.970029 dit_x0 0.979715 vae_audio 0.904450 vae_audio (STFT cosine) 0.976230 [Turbo] Error growth GGML vs Python stage cos max_err mean_err mean_A std_A mean_B std_B dit_step0_xt 0.999947 0.138003 0.006611 -0.002312 0.972929 -0.002342 0.972003 dit_step1_xt 0.999836 0.265130 0.011195 -0.005309 0.942656 -0.005313 0.941730 dit_step2_xt 0.999561 0.451950 0.017352 -0.009353 0.909156 -0.009311 0.908527 dit_step3_xt 0.998939 0.639506 0.025796 -0.014724 0.873776 -0.014577 0.873624 dit_step4_xt 0.997537 0.767998 0.037600 -0.021774 0.841939 -0.021660 0.841995 dit_step5_xt 0.994407 1.219081 0.055160 -0.031844 0.825226 -0.032109 0.824593 dit_step6_xt 0.988459 2.081548 0.081774 -0.046131 0.856101 -0.046482 0.855546 [SFT] steps=50, shift=1.0 | acestep-v15-sft-BF16.gguf [GGML] Running acestep-v15-sft-BF16.gguf... [GGML] Done, 131 dump files [Python] Initializing acestep-v15-sft... [Python] Generating (acestep-v15-sft, 50 steps)... Using precomputed LM hints Using precomputed LM hints 0%| | 0/50 [00:00 Text-Enc [Registry] Qwen3-Embedding-0.6B-Q8_0.gguf -> Text-Enc [Registry] acestep-5Hz-lm-0.6B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-0.6B-Q8_0.gguf -> LM [Registry] acestep-5Hz-lm-1.7B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-1.7B-Q8_0.gguf -> LM [Registry] acestep-5Hz-lm-4B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q5_K_M.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q6_K.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q8_0.gguf -> LM [Registry] acestep-v15-base-BF16.gguf -> DiT [Registry] acestep-v15-base-Q4_K_M.gguf -> DiT [Registry] acestep-v15-base-Q5_K_M.gguf -> DiT [Registry] acestep-v15-base-Q6_K.gguf -> DiT [Registry] acestep-v15-base-Q8_0.gguf -> DiT [Registry] acestep-v15-sft-BF16.gguf -> DiT [Registry] acestep-v15-sft-Q4_K_M.gguf -> DiT [Registry] acestep-v15-sft-Q5_K_M.gguf -> DiT [Registry] acestep-v15-sft-Q6_K.gguf -> DiT [Registry] acestep-v15-sft-Q8_0.gguf -> DiT [Registry] acestep-v15-sftturbo50-BF16.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q4_K_M.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q5_K_M.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q6_K.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-BF16.gguf -> DiT [Registry] acestep-v15-turbo-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-continuous-BF16.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-shift1-BF16.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-shift3-BF16.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-base-BF16.gguf -> DiT [Registry] acestep-v15-xl-base-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-base-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-base-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-base-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-sft-BF16.gguf -> DiT [Registry] acestep-v15-xl-sft-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-sft-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-sft-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-sft-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-BF16.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-turbo-BF16.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q8_0.gguf -> DiT [Registry] scragvae-BF16.gguf -> VAE [Registry] vae-BF16.gguf -> VAE [Store] Created (policy=STRICT) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-turbo-BF16.gguf: 830 tensors, data at offset 69344 [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-turbo-BF16.gguf: 830 tensors, data at offset 69344 [Synth-Load] Ready: turbo=yes, fa=no, batch_cfg=yes [Resolve-Params] max audio codes across batch: 434 (86.8s @ 5Hz) [Resolve-T] T=2170, S=1085 [Resolve-T] seed=42, steps=8, guidance=1.0, shift=3.0, duration=88.0s [BPE] Loaded from GGUF: 151643 vocab, 151387 merges ggml_cuda_init: found 1 CUDA devices (Total VRAM: 97247 MiB): Device 0: NVIDIA RTX PRO 6000 Blackwell Workstation Edition, compute capability 12.0, VMM: yes, VRAM: 97247 MiB load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = NVIDIA RTX PRO 6000 Blackwell Workstation Edition (NVIDIA) | uma: 0 | fp16: 1 | bf16: 0 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: NV_coopmat2 load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] TextEncoder backend: 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: 299 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-BF16.gguf: 830 tensors, data at offset 69344 [Load] LyricEncoder: 8L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [Load] TimbreEncoder: 4L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 141 tensors, 1160.5 MB into backend [Load] CondEncoder: lyric(8L), timbre(4L, CLS), text_proj, null_cond [Store] Load CondEnc: 698 ms [CondEnc] Lyric sliding mask: 167x167, window=128 [CondEnc] Timbre sliding mask: 2x2, window=128 (CLS) [Encode] Packed: lyric=167 + timbre=1 + text=70 = 238 tokens [Encode-Text Batch0] 70+167 tokens -> enc_S=238, 11.4 ms [Debug] enc_hidden: [238, 2048] first4: 0.686579 -0.019135 -0.051539 0.022756 [Store] Unload CondEnc (1160.5 MB) load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] Detokenizer backend: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-turbo-BF16.gguf: 830 tensors, data at offset 69344 [WeightCtx] Loaded 30 tensors, 200.3 MB into backend [Load] Detokenizer: FSQ(6->2048) + 2L encoder(S=5, 2048->64) [Store] Load FSQ-Detok: 93 ms [Context] Decoded: 434 codes -> 2170 frames (86.8s @ 25Hz) [Build-Context Batch0] Detokenizer: 180.7 ms, 434 codes [Debug] detok_output: [2170, 64] first4: -0.125156 1.434936 0.308101 -0.624459 [Store] Unload FSQ-Detok (200.3 MB) [Init-Noise Batch0] Philox noise seed=42, [2170, 64] solver=euler [Debug] noise: [2170, 64] first4: 0.194336 2.156250 -0.171875 0.847656 [Debug] context: [2170, 128] first4: -0.125156 1.434936 0.308101 -0.624459 [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-BF16.gguf: 830 tensors, data at offset 69344 [DiT] Self-attn: Q+K+V fused [DiT] Cross-attn: Q+K+V fused [DiT] MLP: gate+up fused [Load] null_condition_emb found (CFG available) [WeightCtx] Loaded 630 tensors, 7955.6 MB into backend [Load] DiT: 32 layers, H=2560, Nh=32/8, D=128 [Store] Load DiT: 5237 ms [DiT] Batch N=1, T=2170, S=1085, enc_S=238 [DiT] Graph: 2825 nodes [DiT] Solver: ODE Euler (1 NFE/step, order 1) [Debug] tproj: [15360] first4: 0.016458 0.108603 0.065523 -0.074042 [Debug] temb: [2560] first4: 0.026321 -0.023010 -0.016508 0.047175 [Debug] temb_t: [2560] first4: 0.014560 -0.022321 -0.020859 0.048784 [Debug] temb_r: [2560] first4: 0.011760 -0.000690 0.004351 -0.001609 [Debug] sinusoidal_t: [256] first4: 0.562486 0.789701 0.439822 -0.023583 [Debug] sinusoidal_r: [256] first4: 1.000000 1.000000 1.000000 1.000000 [Debug] temb_lin1_t: [2560] first4: 0.064010 -0.019467 0.138295 -0.113216 [Debug] temb_lin1_r: [2560] first4: -0.009879 0.015926 -0.017671 0.012016 [Debug] hidden_after_proj_in: [2560, 1085] first4: 0.338257 -0.576416 0.228638 -0.191162 [Debug] proj_in_input: [192, 2170] first4: -0.125156 1.434936 0.308101 -0.624459 [Debug] enc_after_cond_emb: [2560, 238] first4: 0.029157 0.652797 0.103845 0.870175 [Debug] layer0_sa_input: [2560, 1085] first4: 0.040103 -0.045358 0.093961 -0.134828 [Debug] layer0_q_after_rope: [128, 32] first4: -0.227182 -2.177052 0.197650 -0.065409 [Debug] layer0_k_after_rope: [128, 8] first4: 0.576342 2.180944 0.746347 0.177076 [Debug] layer0_sa_output: [2560, 1085] first4: 1.863274 4.426660 -0.946620 5.656957 [Debug] layer0_attn_out: [4096, 1085] first4: 0.486701 -2.292692 -0.463459 -2.908726 [Debug] layer0_after_self_attn: [2560, 1085] first4: 0.486701 -2.292692 -0.463459 -2.908726 [Debug] layer0_after_cross_attn: [2560, 1085] first4: 0.352633 -2.046512 -0.288305 -3.871271 [Debug] hidden_after_layer0: [2560, 1085] first4: 1.702097 -2.483572 241.016083 -3.841073 [Debug] hidden_after_layer6: [2560, 1085] first4: -6.205515 -2.173777 320.014587 -12.230325 [Debug] hidden_after_layer12: [2560, 1085] first4: -10.401177 29.480074 260.742523 29.244408 [Debug] hidden_after_layer18: [2560, 1085] first4: 41.523281 13.234173 278.211273 4.641582 [Debug] hidden_after_layer31: [2560, 1085] first4: -89.614502 -232.100800 -167.380676 238.108551 [Debug] dit_step0_vt: [2170, 64] first4: -1.172607 1.473038 0.329620 1.149292 [Debug] dit_step0_xt: [2170, 64] first4: 0.247636 2.089294 -0.186858 0.795416 [DiT] Step 1/8 t=1.000 [Debug] dit_step1_vt: [2170, 64] first4: -0.803955 1.302063 0.499481 0.959717 [Debug] dit_step1_xt: [2170, 64] first4: 0.291488 2.018272 -0.214102 0.743068 [DiT] Step 2/8 t=0.955 [Debug] dit_step2_vt: [2170, 64] first4: -1.380737 1.851887 1.411713 1.357910 [Debug] dit_step2_xt: [2170, 64] first4: 0.383538 1.894813 -0.308216 0.652540 [DiT] Step 3/8 t=0.900 [Debug] dit_step3_vt: [2170, 64] first4: -0.661774 1.566681 1.178055 0.879150 [Debug] dit_step3_xt: [2170, 64] first4: 0.438685 1.764256 -0.406388 0.579278 [DiT] Step 4/8 t=0.833 [Debug] dit_step4_vt: [2170, 64] first4: -0.041496 2.802368 1.023010 0.592896 [Debug] dit_step4_xt: [2170, 64] first4: 0.443131 1.464003 -0.515996 0.515753 [DiT] Step 5/8 t=0.750 [Debug] dit_step5_vt: [2170, 64] first4: 0.570618 2.336487 1.062622 0.401855 [Debug] dit_step5_xt: [2170, 64] first4: 0.361615 1.130219 -0.667799 0.458345 [DiT] Step 6/8 t=0.643 [Debug] dit_step6_vt: [2170, 64] first4: 0.953094 1.994577 0.434647 0.650635 [Debug] dit_step6_xt: [2170, 64] first4: 0.170996 0.731303 -0.754728 0.328218 [DiT] Step 7/8 t=0.500 [Debug] dit_step7_vt: [2170, 64] first4: -1.075439 2.842133 1.517700 0.775513 [Debug] dit_x0: [2170, 64] first4: 0.493627 -0.121337 -1.210038 0.095564 [DiT] Step 8/8 t=0.300 [DiT-Generate] Total: 1906.7 ms (1906.7 ms/sample) [Debug] dit_output: [2170, 64] first4: 0.493627 -0.121337 -1.210038 0.095564 [Store] Unload DiT (7955.6 MB) [GGUF] /mnt/workspace/git/acestep.cpp/models/vae-BF16.gguf: 365 tensors, data at offset 30048 load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] VAE backend: Vulkan0 (CPU threads: 16) [VAE] Backend: Vulkan0, Weight buffer: 161.1 MB [VAE] Loaded: 5 blocks, upsample=1920x, F32 activations [Store] Load VAE-Dec: 674 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: 949.5 ms [Debug] vae_audio: [2, 4166400] first4: 0.005018 0.007553 0.007136 0.006468 [Store] Unload VAE-Dec (161.1 MB) [MP3] Encoding 86.8s @ 128 kbps, 48000 Hz stereo [MP3] 1388928 bytes (12.0:1), 449 ms (193.27x realtime), 32 threads [MP3] Wrote ggml-xl-turbo/request00.mp3 [Ace-Synth] All done 2026-07-12 10:19:43.550 | 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 0.999817 lyric_embed 1.000000 enc_hidden 0.999845 detok_output 0.999997 context 0.999998 noise 1.000000 temb_t 0.999998 hidden_after_proj_in 0.999987 enc_after_cond_emb 0.999834 layer0_sa_output 0.999950 hidden_after_layer0 0.999982 hidden_after_layer6 0.999854 hidden_after_layer12 0.999479 hidden_after_layer18 0.998728 hidden_after_layer23 0.998987 dit_step0_vt 0.998944 dit_step0_xt 1.000000 dit_step5_vt 0.999527 dit_step5_xt 0.999981 dit_step10_vt 0.999180 dit_step10_xt 0.999938 dit_step15_vt 0.998274 dit_step15_xt 0.999856 dit_step20_vt 0.996623 dit_step20_xt 0.999642 dit_step25_vt 0.993607 dit_step25_xt 0.999204 dit_step30_vt 0.989609 dit_step30_xt 0.998392 dit_step35_vt 0.985259 dit_step35_xt 0.997247 dit_step40_vt 0.982577 dit_step40_xt 0.996004 dit_step45_vt 0.985765 dit_step45_xt 0.995075 dit_step49_vt 0.990676 dit_x0 0.994748 vae_audio 0.970357 vae_audio (STFT cosine) 0.988134 [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 0.020037 0.001452 -0.001656 0.978450 -0.001619 0.978872 dit_step5_xt 0.999981 0.079046 0.004206 -0.008388 0.884726 -0.008480 0.883811 dit_step10_xt 0.999938 0.118885 0.006669 -0.015119 0.800846 -0.015075 0.802324 dit_step15_xt 0.999856 0.164940 0.009344 -0.021822 0.730399 -0.021933 0.731304 dit_step20_xt 0.999642 0.227404 0.013633 -0.028563 0.677648 -0.028882 0.679066 dit_step25_xt 0.999204 0.331498 0.019314 -0.035376 0.647040 -0.035643 0.650780 dit_step30_xt 0.998392 0.566910 0.026867 -0.042226 0.641845 -0.042577 0.647107 dit_step35_xt 0.997247 0.825760 0.035935 -0.049084 0.662760 -0.049471 0.669256 dit_step40_xt 0.996004 1.083266 0.046088 -0.055971 0.707409 -0.056532 0.715783 dit_step45_xt 0.995075 1.312466 0.055676 -0.062961 0.771444 -0.063533 0.780575 [Xl-turbo] steps=8, shift=3.0 | acestep-v15-xl-turbo-BF16.gguf [GGML] Running acestep-v15-xl-turbo-BF16.gguf... [GGML] Done, 47 dump files [Python] Initializing acestep-v15-xl-turbo... Loading checkpoint shards: 0%| | 0/4 [00:00 1.0 (turbo does not use CFG). 2026-07-12 10:19:47.272 | INFO | acestep.core.generation.handler.generate_music:generate_music:304 - [generate_music] Starting generation... 2026-07-12 10:19:47.272 | INFO | acestep.core.generation.handler.generate_music:generate_music:307 - [generate_music] Preparing inputs... 2026-07-12 10:19:47.272 | 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 10:19:47.275 | 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 10:19:47.280 | INFO | acestep.core.generation.handler.conditioning_target:_prepare_target_latents_and_wavs:78 - [generate_music] Decoding audio codes for item 0... 2026-07-12 10:19:47.286 | 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 10:19:47.288 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:119 - ====================================================================== 2026-07-12 10:19:47.288 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:120 - πŸ” [DEBUG] DiT TEXT ENCODER INPUT (Inference) 2026-07-12 10:19:47.288 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:121 - ====================================================================== 2026-07-12 10:19:47.289 | 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 10:19:47.289 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:123 - ====================================================================== 2026-07-12 10:19:47.289 | 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 10:19:47.289 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:125 - ====================================================================== 2026-07-12 10:19:47.292 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:110 - [preprocess_batch] Inferring prompt embeddings... 2026-07-12 10:19:47.300 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:113 - [preprocess_batch] Inferring lyric embeddings... 2026-07-12 10:19:47.300 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:161 - [service_generate] Generating audio... (DiT backend: PyTorch (cuda)) 2026-07-12 10:19:47.306 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:263 - [service_generate] DiT diffusion via PyTorch (cuda)... 2026-07-12 10:19:47.313 | 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 10:19:47.868 | INFO | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:42 - [generate_music] Model generation completed. Decoding latents... 2026-07-12 10:19:47.869 | 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 10:19:47.869 | DEBUG | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:65 - [generate_music] time_costs: {'encoder_time_cost': 0.006394624710083008, 'diffusion_time_cost': 0.5549852848052979, 'diffusion_per_step_time_cost': 0.06937316060066223, 'total_time_cost': 0.5613799095153809, 'offload_time_cost': 0.0} 2026-07-12 10:19:47.892 | INFO | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:127 - [generate_music] Decoding latents with VAE... 2026-07-12 10:19:47.895 | 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 10:19:47.895 | 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 10:19:47.895 | 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 10:19:47.895 | 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 10:19:47.895 | 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 10:19:47.895 | 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 10:19:48.172 | 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 10:19:48.373 | 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 10:19:48.376 | 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-BF16.gguf [Request] vae: vae-BF16.gguf [Request] audio_codes: (present) [Ace-Synth] Batch: 1 request(s) [Registry] Qwen3-Embedding-0.6B-BF16.gguf -> Text-Enc [Registry] Qwen3-Embedding-0.6B-Q8_0.gguf -> Text-Enc [Registry] acestep-5Hz-lm-0.6B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-0.6B-Q8_0.gguf -> LM [Registry] acestep-5Hz-lm-1.7B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-1.7B-Q8_0.gguf -> LM [Registry] acestep-5Hz-lm-4B-BF16.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q5_K_M.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q6_K.gguf -> LM [Registry] acestep-5Hz-lm-4B-Q8_0.gguf -> LM [Registry] acestep-v15-base-BF16.gguf -> DiT [Registry] acestep-v15-base-Q4_K_M.gguf -> DiT [Registry] acestep-v15-base-Q5_K_M.gguf -> DiT [Registry] acestep-v15-base-Q6_K.gguf -> DiT [Registry] acestep-v15-base-Q8_0.gguf -> DiT [Registry] acestep-v15-sft-BF16.gguf -> DiT [Registry] acestep-v15-sft-Q4_K_M.gguf -> DiT [Registry] acestep-v15-sft-Q5_K_M.gguf -> DiT [Registry] acestep-v15-sft-Q6_K.gguf -> DiT [Registry] acestep-v15-sft-Q8_0.gguf -> DiT [Registry] acestep-v15-sftturbo50-BF16.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q4_K_M.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q5_K_M.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q6_K.gguf -> DiT [Registry] acestep-v15-sftturbo50-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-BF16.gguf -> DiT [Registry] acestep-v15-turbo-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-continuous-BF16.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-continuous-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-shift1-BF16.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-shift1-Q8_0.gguf -> DiT [Registry] acestep-v15-turbo-shift3-BF16.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q4_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q5_K_M.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q6_K.gguf -> DiT [Registry] acestep-v15-turbo-shift3-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-base-BF16.gguf -> DiT [Registry] acestep-v15-xl-base-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-base-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-base-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-base-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-sft-BF16.gguf -> DiT [Registry] acestep-v15-xl-sft-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-sft-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-sft-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-sft-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-BF16.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-sftturbo50-Q8_0.gguf -> DiT [Registry] acestep-v15-xl-turbo-BF16.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q4_K_M.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q5_K_M.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q6_K.gguf -> DiT [Registry] acestep-v15-xl-turbo-Q8_0.gguf -> DiT [Registry] scragvae-BF16.gguf -> VAE [Registry] vae-BF16.gguf -> VAE [Store] Created (policy=STRICT) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-sft-BF16.gguf: 830 tensors, data at offset 69312 [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-sft-BF16.gguf: 830 tensors, data at offset 69312 [Synth-Load] Ready: turbo=no, fa=no, batch_cfg=yes [Resolve-Params] max audio codes across batch: 434 (86.8s @ 5Hz) [Resolve-T] T=2170, S=1085 [Resolve-T] seed=42, steps=50, guidance=1.0, shift=1.0, duration=88.0s [BPE] Loaded from GGUF: 151643 vocab, 151387 merges ggml_cuda_init: found 1 CUDA devices (Total VRAM: 97247 MiB): Device 0: NVIDIA RTX PRO 6000 Blackwell Workstation Edition, compute capability 12.0, VMM: yes, VRAM: 97247 MiB load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = NVIDIA RTX PRO 6000 Blackwell Workstation Edition (NVIDIA) | uma: 0 | fp16: 1 | bf16: 0 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: NV_coopmat2 load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] TextEncoder backend: 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: 325 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-BF16.gguf: 830 tensors, data at offset 69312 [Load] LyricEncoder: 8L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [Load] TimbreEncoder: 4L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 141 tensors, 1160.5 MB into backend [Load] CondEncoder: lyric(8L), timbre(4L, CLS), text_proj, null_cond [Store] Load CondEnc: 707 ms [CondEnc] Lyric sliding mask: 167x167, window=128 [CondEnc] Timbre sliding mask: 2x2, window=128 (CLS) [Encode] Packed: lyric=167 + timbre=1 + text=70 = 238 tokens [Encode-Text Batch0] 70+167 tokens -> enc_S=238, 11.4 ms [Debug] enc_hidden: [238, 2048] first4: 0.686579 -0.019135 -0.051539 0.022756 [Store] Unload CondEnc (1160.5 MB) load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] Detokenizer backend: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-sft-BF16.gguf: 830 tensors, data at offset 69312 [WeightCtx] Loaded 30 tensors, 200.3 MB into backend [Load] Detokenizer: FSQ(6->2048) + 2L encoder(S=5, 2048->64) [Store] Load FSQ-Detok: 97 ms [Context] Decoded: 434 codes -> 2170 frames (86.8s @ 25Hz) [Build-Context Batch0] Detokenizer: 191.5 ms, 434 codes [Debug] detok_output: [2170, 64] first4: -0.125156 1.434936 0.308101 -0.624459 [Store] Unload FSQ-Detok (200.3 MB) [Init-Noise Batch0] Philox noise seed=42, [2170, 64] solver=euler [Debug] noise: [2170, 64] first4: 0.194336 2.156250 -0.171875 0.847656 [Debug] context: [2170, 128] first4: -0.125156 1.434936 0.308101 -0.624459 [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-BF16.gguf: 830 tensors, data at offset 69312 [DiT] Self-attn: Q+K+V fused [DiT] Cross-attn: Q+K+V fused [DiT] MLP: gate+up fused [Load] null_condition_emb found (CFG available) [WeightCtx] Loaded 630 tensors, 7955.6 MB into backend [Load] DiT: 32 layers, H=2560, Nh=32/8, D=128 [Store] Load DiT: 3885 ms [DiT] Batch N=1, T=2170, S=1085, enc_S=238 [DiT] Graph: 2825 nodes [DiT] Solver: ODE Euler (1 NFE/step, order 1) [Debug] tproj: [15360] first4: 0.012546 0.105322 0.073633 -0.031913 [Debug] temb: [2560] first4: 0.009488 -0.007695 -0.006104 0.004570 [Debug] temb_t: [2560] first4: 0.007949 -0.007936 -0.006669 0.003595 [Debug] temb_r: [2560] first4: 0.001539 0.000241 0.000566 0.000975 [Debug] sinusoidal_t: [256] first4: 0.562486 0.789701 0.439822 -0.023583 [Debug] sinusoidal_r: [256] first4: 1.000000 1.000000 1.000000 1.000000 [Debug] temb_lin1_t: [2560] first4: 0.032232 -0.081851 0.088236 -0.064962 [Debug] temb_lin1_r: [2560] first4: 0.006956 -0.005390 -0.002806 -0.011722 [Debug] hidden_after_proj_in: [2560, 1085] first4: 0.327515 -0.491455 0.102783 -0.215088 [Debug] proj_in_input: [192, 2170] first4: -0.125156 1.434936 0.308101 -0.624459 [Debug] enc_after_cond_emb: [2560, 238] first4: 0.033762 0.641460 0.109139 0.869468 [Debug] layer0_sa_input: [2560, 1085] first4: 0.036374 -0.018622 0.098722 -0.097802 [Debug] layer0_q_after_rope: [128, 32] first4: -0.253460 -2.265648 0.030818 -0.232458 [Debug] layer0_k_after_rope: [128, 8] first4: 0.647806 2.980026 0.673072 1.817613 [Debug] layer0_sa_output: [2560, 1085] first4: 2.196771 4.505235 -1.249555 8.230964 [Debug] layer0_attn_out: [4096, 1085] first4: 0.471712 -2.218876 -0.825885 -4.313496 [Debug] layer0_after_self_attn: [2560, 1085] first4: 0.471712 -2.218876 -0.825885 -4.313496 [Debug] layer0_after_cross_attn: [2560, 1085] first4: 0.310287 -1.951492 -0.739202 -4.388470 [Debug] hidden_after_layer0: [2560, 1085] first4: -1.703167 -1.277268 249.988998 -4.098482 [Debug] hidden_after_layer6: [2560, 1085] first4: -7.904379 3.749470 316.075043 -21.383392 [Debug] hidden_after_layer12: [2560, 1085] first4: -23.674171 40.319260 229.033401 17.408325 [Debug] hidden_after_layer18: [2560, 1085] first4: 9.057741 -12.742176 233.615005 19.633303 [Debug] hidden_after_layer31: [2560, 1085] first4: 62.296417 -104.924454 7.629906 90.108032 [Debug] dit_step0_vt: [2170, 64] first4: -0.313110 2.213501 0.002686 1.701447 [Debug] dit_step0_xt: [2170, 64] first4: 0.200598 2.111980 -0.171929 0.813627 [DiT] Step 1/50 t=1.000 [Debug] dit_step1_vt: [2170, 64] first4: -0.350342 2.149414 0.083496 1.558319 [Debug] dit_step1_xt: [2170, 64] first4: 0.207605 2.068992 -0.173599 0.782461 [DiT] Step 2/50 t=0.980 [Debug] dit_step2_vt: [2170, 64] first4: -0.324707 2.165955 0.060059 1.567017 [Debug] dit_step2_xt: [2170, 64] first4: 0.214099 2.025673 -0.174800 0.751121 [DiT] Step 3/50 t=0.960 [Debug] dit_step3_vt: [2170, 64] first4: -0.277588 2.128662 0.048706 1.567612 [Debug] dit_step3_xt: [2170, 64] first4: 0.219651 1.983099 -0.175774 0.719768 [DiT] Step 4/50 t=0.940 [Debug] dit_step4_vt: [2170, 64] first4: -0.265381 2.065125 0.023560 1.586212 [Debug] dit_step4_xt: [2170, 64] first4: 0.224959 1.941797 -0.176245 0.688044 [DiT] Step 5/50 t=0.920 [Debug] dit_step5_vt: [2170, 64] first4: -0.258545 2.042694 0.030762 1.587585 [Debug] dit_step5_xt: [2170, 64] first4: 0.230129 1.900943 -0.176860 0.656292 [DiT] Step 6/50 t=0.900 [Debug] dit_step6_vt: [2170, 64] first4: -0.252686 2.011414 0.068359 1.607635 [Debug] dit_step6_xt: [2170, 64] first4: 0.235183 1.860715 -0.178228 0.624140 [DiT] Step 7/50 t=0.880 [Debug] dit_step7_vt: [2170, 64] first4: -0.249023 1.972534 0.110474 1.615204 [Debug] dit_step7_xt: [2170, 64] first4: 0.240164 1.821264 -0.180437 0.591836 [DiT] Step 8/50 t=0.860 [Debug] dit_step8_vt: [2170, 64] first4: -0.258179 1.927551 0.140869 1.609680 [Debug] dit_step8_xt: [2170, 64] first4: 0.245327 1.782713 -0.183254 0.559642 [DiT] Step 9/50 t=0.840 [Debug] dit_step9_vt: [2170, 64] first4: -0.277100 1.877106 0.127441 1.612549 [Debug] dit_step9_xt: [2170, 64] first4: 0.250869 1.745171 -0.185803 0.527391 [DiT] Step 10/50 t=0.820 [Debug] dit_step10_vt: [2170, 64] first4: -0.280151 1.837173 0.118164 1.608665 [Debug] dit_step10_xt: [2170, 64] first4: 0.256472 1.708427 -0.188166 0.495218 [DiT] Step 11/50 t=0.800 [Debug] dit_step11_vt: [2170, 64] first4: -0.276245 1.791702 0.109619 1.601681 [Debug] dit_step11_xt: [2170, 64] first4: 0.261997 1.672593 -0.190359 0.463184 [DiT] Step 12/50 t=0.780 [Debug] dit_step12_vt: [2170, 64] first4: -0.267944 1.748299 0.102905 1.588501 [Debug] dit_step12_xt: [2170, 64] first4: 0.267356 1.637627 -0.192417 0.431414 [DiT] Step 13/50 t=0.760 [Debug] dit_step13_vt: [2170, 64] first4: -0.264771 1.699287 0.089111 1.576660 [Debug] dit_step13_xt: [2170, 64] first4: 0.272651 1.603642 -0.194199 0.399881 [DiT] Step 14/50 t=0.740 [Debug] dit_step14_vt: [2170, 64] first4: -0.256836 1.643967 0.073608 1.571655 [Debug] dit_step14_xt: [2170, 64] first4: 0.277788 1.570762 -0.195671 0.368448 [DiT] Step 15/50 t=0.720 [Debug] dit_step15_vt: [2170, 64] first4: -0.251831 1.598694 0.055664 1.566559 [Debug] dit_step15_xt: [2170, 64] first4: 0.282825 1.538789 -0.196785 0.337117 [DiT] Step 16/50 t=0.700 [Debug] dit_step16_vt: [2170, 64] first4: -0.235840 1.550934 0.035767 1.562042 [Debug] dit_step16_xt: [2170, 64] first4: 0.287542 1.507770 -0.197500 0.305876 [DiT] Step 17/50 t=0.680 [Debug] dit_step17_vt: [2170, 64] first4: -0.212280 1.505524 0.019531 1.555420 [Debug] dit_step17_xt: [2170, 64] first4: 0.291787 1.477659 -0.197891 0.274767 [DiT] Step 18/50 t=0.660 [Debug] dit_step18_vt: [2170, 64] first4: -0.182312 1.450500 0.000610 1.544220 [Debug] dit_step18_xt: [2170, 64] first4: 0.295433 1.448649 -0.197903 0.243883 [DiT] Step 19/50 t=0.640 [Debug] dit_step19_vt: [2170, 64] first4: -0.149658 1.401062 -0.008789 1.535645 [Debug] dit_step19_xt: [2170, 64] first4: 0.298427 1.420628 -0.197727 0.213170 [DiT] Step 20/50 t=0.620 [Debug] dit_step20_vt: [2170, 64] first4: -0.120544 1.353394 -0.020996 1.532990 [Debug] dit_step20_xt: [2170, 64] first4: 0.300837 1.393560 -0.197307 0.182510 [DiT] Step 21/50 t=0.600 [Debug] dit_step21_vt: [2170, 64] first4: -0.091309 1.296631 -0.028320 1.525208 [Debug] dit_step21_xt: [2170, 64] first4: 0.302664 1.367628 -0.196741 0.152006 [DiT] Step 22/50 t=0.580 [Debug] dit_step22_vt: [2170, 64] first4: -0.065735 1.255371 -0.041138 1.514435 [Debug] dit_step22_xt: [2170, 64] first4: 0.303978 1.342520 -0.195918 0.121717 [DiT] Step 23/50 t=0.560 [Debug] dit_step23_vt: [2170, 64] first4: -0.055420 1.201904 -0.056641 1.501465 [Debug] dit_step23_xt: [2170, 64] first4: 0.305087 1.318482 -0.194785 0.091688 [DiT] Step 24/50 t=0.540 [Debug] dit_step24_vt: [2170, 64] first4: -0.037781 1.152771 -0.051880 1.479919 [Debug] dit_step24_xt: [2170, 64] first4: 0.305842 1.295427 -0.193748 0.062090 [DiT] Step 25/50 t=0.520 [Debug] dit_step25_vt: [2170, 64] first4: -0.033203 1.098389 -0.063721 1.462921 [Debug] dit_step25_xt: [2170, 64] first4: 0.306506 1.273459 -0.192473 0.032831 [DiT] Step 26/50 t=0.500 [Debug] dit_step26_vt: [2170, 64] first4: -0.033386 1.036316 -0.069336 1.444336 [Debug] dit_step26_xt: [2170, 64] first4: 0.307174 1.252733 -0.191086 0.003945 [DiT] Step 27/50 t=0.480 [Debug] dit_step27_vt: [2170, 64] first4: -0.038574 0.977386 -0.077026 1.422424 [Debug] dit_step27_xt: [2170, 64] first4: 0.307946 1.233185 -0.189546 -0.024504 [DiT] Step 28/50 t=0.460 [Debug] dit_step28_vt: [2170, 64] first4: -0.044434 0.916473 -0.087891 1.403503 [Debug] dit_step28_xt: [2170, 64] first4: 0.308834 1.214855 -0.187788 -0.052574 [DiT] Step 29/50 t=0.440 [Debug] dit_step29_vt: [2170, 64] first4: -0.053589 0.845337 -0.099609 1.378265 [Debug] dit_step29_xt: [2170, 64] first4: 0.309906 1.197949 -0.185796 -0.080139 [DiT] Step 30/50 t=0.420 [Debug] dit_step30_vt: [2170, 64] first4: -0.066589 0.772202 -0.109863 1.363068 [Debug] dit_step30_xt: [2170, 64] first4: 0.311238 1.182505 -0.183599 -0.107401 [DiT] Step 31/50 t=0.400 [Debug] dit_step31_vt: [2170, 64] first4: -0.084961 0.698105 -0.119080 1.339355 [Debug] dit_step31_xt: [2170, 64] first4: 0.312937 1.168543 -0.181217 -0.134188 [DiT] Step 32/50 t=0.380 [Debug] dit_step32_vt: [2170, 64] first4: -0.103943 0.612225 -0.126709 1.305908 [Debug] dit_step32_xt: [2170, 64] first4: 0.315016 1.156298 -0.178683 -0.160306 [DiT] Step 33/50 t=0.360 [Debug] dit_step33_vt: [2170, 64] first4: -0.114197 0.530327 -0.137573 1.292358 [Debug] dit_step33_xt: [2170, 64] first4: 0.317300 1.145692 -0.175931 -0.186153 [DiT] Step 34/50 t=0.340 [Debug] dit_step34_vt: [2170, 64] first4: -0.127075 0.445534 -0.147491 1.260437 [Debug] dit_step34_xt: [2170, 64] first4: 0.319841 1.136781 -0.172982 -0.211362 [DiT] Step 35/50 t=0.320 [Debug] dit_step35_vt: [2170, 64] first4: -0.146942 0.343864 -0.144806 1.227966 [Debug] dit_step35_xt: [2170, 64] first4: 0.322780 1.129904 -0.170085 -0.235921 [DiT] Step 36/50 t=0.300 [Debug] dit_step36_vt: [2170, 64] first4: -0.158447 0.247955 -0.134750 1.203735 [Debug] dit_step36_xt: [2170, 64] first4: 0.325949 1.124944 -0.167390 -0.259996 [DiT] Step 37/50 t=0.280 [Debug] dit_step37_vt: [2170, 64] first4: -0.157616 0.150543 -0.133377 1.160828 [Debug] dit_step37_xt: [2170, 64] first4: 0.329101 1.121934 -0.164723 -0.283212 [DiT] Step 38/50 t=0.260 [Debug] dit_step38_vt: [2170, 64] first4: -0.163857 0.054321 -0.116840 1.123413 [Debug] dit_step38_xt: [2170, 64] first4: 0.332379 1.120847 -0.162386 -0.305681 [DiT] Step 39/50 t=0.240 [Debug] dit_step39_vt: [2170, 64] first4: -0.187918 -0.058502 -0.095459 1.090332 [Debug] dit_step39_xt: [2170, 64] first4: 0.336137 1.122017 -0.160477 -0.327487 [DiT] Step 40/50 t=0.220 [Debug] dit_step40_vt: [2170, 64] first4: -0.213722 -0.160797 -0.086387 1.046936 [Debug] dit_step40_xt: [2170, 64] first4: 0.340411 1.125233 -0.158749 -0.348426 [DiT] Step 41/50 t=0.200 [Debug] dit_step41_vt: [2170, 64] first4: -0.240204 -0.272194 -0.059448 1.003662 [Debug] dit_step41_xt: [2170, 64] first4: 0.345215 1.130677 -0.157560 -0.368499 [DiT] Step 42/50 t=0.180 [Debug] dit_step42_vt: [2170, 64] first4: -0.261749 -0.363136 -0.026398 0.955444 [Debug] dit_step42_xt: [2170, 64] first4: 0.350450 1.137940 -0.157032 -0.387608 [DiT] Step 43/50 t=0.160 [Debug] dit_step43_vt: [2170, 64] first4: -0.290985 -0.459885 -0.004700 0.910767 [Debug] dit_step43_xt: [2170, 64] first4: 0.356270 1.147137 -0.156938 -0.405823 [DiT] Step 44/50 t=0.140 [Debug] dit_step44_vt: [2170, 64] first4: -0.316406 -0.555389 0.055618 0.845398 [Debug] dit_step44_xt: [2170, 64] first4: 0.362598 1.158245 -0.158051 -0.422731 [DiT] Step 45/50 t=0.120 [Debug] dit_step45_vt: [2170, 64] first4: -0.353668 -0.632721 0.093445 0.794556 [Debug] dit_step45_xt: [2170, 64] first4: 0.369672 1.170900 -0.159920 -0.438623 [DiT] Step 46/50 t=0.100 [Debug] dit_step46_vt: [2170, 64] first4: -0.394104 -0.705872 0.139832 0.744507 [Debug] dit_step46_xt: [2170, 64] first4: 0.377554 1.185017 -0.162716 -0.453513 [DiT] Step 47/50 t=0.080 [Debug] dit_step47_vt: [2170, 64] first4: -0.441751 -0.785477 0.199951 0.702515 [Debug] dit_step47_xt: [2170, 64] first4: 0.386389 1.200727 -0.166715 -0.467563 [DiT] Step 48/50 t=0.060 [Debug] dit_step48_vt: [2170, 64] first4: -0.495434 -0.846634 0.264160 0.685120 [Debug] dit_step48_xt: [2170, 64] first4: 0.396297 1.217659 -0.171998 -0.481265 [DiT] Step 49/50 t=0.040 [Debug] dit_step49_vt: [2170, 64] first4: -0.499638 -0.830109 0.186890 0.696899 [Debug] dit_x0: [2170, 64] first4: 0.406290 1.234261 -0.175736 -0.495203 [DiT] Step 50/50 t=0.020 [DiT-Generate] Total: 11709.0 ms (11709.0 ms/sample) [Debug] dit_output: [2170, 64] first4: 0.406290 1.234261 -0.175736 -0.495203 [Store] Unload DiT (7955.6 MB) [GGUF] /mnt/workspace/git/acestep.cpp/models/vae-BF16.gguf: 365 tensors, data at offset 30048 load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] VAE backend: Vulkan0 (CPU threads: 16) [VAE] Backend: Vulkan0, Weight buffer: 161.1 MB [VAE] Loaded: 5 blocks, upsample=1920x, F32 activations [Store] Load VAE-Dec: 675 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: 958.9 ms [Debug] vae_audio: [2, 4166400] first4: -0.001391 -0.000663 -0.000788 -0.000318 [Store] Unload VAE-Dec (161.1 MB) [MP3] Encoding 86.8s @ 128 kbps, 48000 Hz stereo [MP3] 1388928 bytes (12.0:1), 419 ms (206.95x realtime), 32 threads [MP3] Wrote ggml-xl-sft/request00.mp3 [Ace-Synth] All done 2026-07-12 10:20:08.607 | 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 0.999817 lyric_embed 1.000000 enc_hidden 0.999575 detok_output 0.999997 context 0.999998 noise 1.000000 temb_t 0.999996 hidden_after_proj_in 0.999980 enc_after_cond_emb 0.999692 layer0_sa_output 0.989369 hidden_after_layer0 0.999258 hidden_after_layer6 0.999714 hidden_after_layer12 0.997139 hidden_after_layer18 0.996173 hidden_after_layer31 0.987843 dit_step0_vt 0.908839 dit_step0_xt 0.999836 dit_step1_vt 0.922933 dit_step1_xt 0.999499 dit_step2_vt 0.927696 dit_step2_xt 0.998710 dit_step3_vt 0.924563 dit_step3_xt 0.996866 dit_step4_vt 0.919998 dit_step4_xt 0.992327 dit_step5_vt 0.909287 dit_step5_xt 0.981312 dit_step6_vt 0.899922 dit_step6_xt 0.959377 dit_step7_vt 0.879879 dit_x0 0.923441 vae_audio 0.728236 vae_audio (STFT cosine) 0.920632 [Xl-turbo] Error growth GGML vs Python stage cos max_err mean_err mean_A std_A mean_B std_B dit_step0_xt 0.999836 0.152072 0.012725 -0.002987 0.978406 -0.002721 0.976534 dit_step1_xt 0.999499 0.319018 0.022084 -0.005855 0.952734 -0.005567 0.949004 dit_step2_xt 0.998710 0.549332 0.033951 -0.009539 0.924220 -0.009330 0.918931 dit_step3_xt 0.996866 0.811508 0.050287 -0.015060 0.893551 -0.014418 0.887241 dit_step4_xt 0.992327 1.183722 0.075117 -0.022735 0.864343 -0.021453 0.857796 dit_step5_xt 0.981312 2.421603 0.114743 -0.033921 0.847894 -0.032240 0.840033 dit_step6_xt 0.959377 4.100842 0.175064 -0.050079 0.871132 -0.049194 0.863964 [Xl-sft] steps=50, shift=1.0 | acestep-v15-xl-sft-BF16.gguf [GGML] Running acestep-v15-xl-sft-BF16.gguf... [GGML] Done, 131 dump files [Python] Initializing acestep-v15-xl-sft... Loading checkpoint shards: 0%| | 0/4 [00:00 2026-07-12 10:20:14.540 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:123 - ====================================================================== 2026-07-12 10:20:14.540 | 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 10:20:14.540 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:125 - ====================================================================== 2026-07-12 10:20:14.543 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:110 - [preprocess_batch] Inferring prompt embeddings... 2026-07-12 10:20:14.551 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:113 - [preprocess_batch] Inferring lyric embeddings... 2026-07-12 10:20:14.551 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:161 - [service_generate] Generating audio... (DiT backend: PyTorch (cuda)) 2026-07-12 10:20:14.557 | 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