[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-Q6_K.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-Q6_K.gguf: 678 tensors, data at offset 56864 [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-turbo-Q6_K.gguf: 678 tensors, data at offset 56864 [Synth-Load] Ready: turbo=yes, fa=no, batch_cfg=yes [Resolve-Params] max audio codes across batch: 434 (86.8s @ 5Hz) [Resolve-T] T=2170, S=1085 [Resolve-T] seed=42, steps=8, guidance=1.0, shift=3.0, duration=88.0s [BPE] Loaded from GGUF: 151643 vocab, 151387 merges ggml_cuda_init: found 1 CUDA devices (Total VRAM: 97247 MiB): Device 0: NVIDIA RTX PRO 6000 Blackwell Workstation Edition, compute capability 12.0, VMM: yes, VRAM: 97247 MiB load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = NVIDIA RTX PRO 6000 Blackwell Workstation Edition (NVIDIA) | uma: 0 | fp16: 1 | bf16: 0 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: NV_coopmat2 load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] TextEncoder backend: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/Qwen3-Embedding-0.6B-BF16.gguf: 310 tensors, data at offset 5337568 [Load] TextEncoder: 28L, H=1024, Nh=16/8 [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 310 tensors, 1136.5 MB into backend [Store] Load TextEnc: 5827 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-Q6_K.gguf: 678 tensors, data at offset 56864 [Load] LyricEncoder: 8L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [Load] TimbreEncoder: 4L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 140 tensors, 476.3 MB into backend [Load] CondEncoder: lyric(8L), timbre(4L), text_proj, null_cond [Store] Load CondEnc: 287 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, 27.3 ms [Debug] enc_hidden: [238, 2048] first4: 1.761451 -0.050624 -0.132928 0.058356 [Store] Unload CondEnc (476.3 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-Q6_K.gguf: 678 tensors, data at offset 56864 [WeightCtx] Loaded 30 tensors, 82.2 MB into backend [Load] Detokenizer: FSQ(6->2048) + 2L encoder(S=5, 2048->64) [Store] Load FSQ-Detok: 63 ms [Context] Decoded: 434 codes -> 2170 frames (86.8s @ 25Hz) [Build-Context Batch0] Detokenizer: 176.7 ms, 434 codes [Debug] detok_output: [2170, 64] first4: -0.141068 1.454457 0.315091 -0.623697 [Store] Unload FSQ-Detok (82.2 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: -0.141068 1.454457 0.315091 -0.623697 [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-Q6_K.gguf: 678 tensors, data at offset 56864 [DiT] Self-attn: Q+K+V fused [DiT] Cross-attn: Q+K+V fused [DiT] MLP: gate+up fused [Load] null_condition_emb found (CFG available) [WeightCtx] Loaded 478 tensors, 1237.2 MB into backend [Load] DiT: 24 layers, H=2048, Nh=16/8, D=128 [Store] Load DiT: 894 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: 0.261089 -0.161223 -0.098727 0.051901 [Debug] temb: [2048] first4: 0.000236 -0.132397 -0.035347 0.064653 [Debug] temb_t: [2048] first4: 0.001398 0.026957 -0.052741 0.063660 [Debug] temb_r: [2048] first4: -0.001162 -0.159353 0.017394 0.000993 [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: -0.049071 -0.051112 -0.017769 -0.037193 [Debug] temb_lin1_r: [2048] first4: -0.014408 -0.020609 -0.015729 0.003875 [Debug] hidden_after_proj_in: [2048, 1085] first4: -1.037598 -0.957581 0.541260 0.451874 [Debug] proj_in_input: [192, 2170] first4: -0.141068 1.454457 0.315091 -0.623697 [Debug] enc_after_cond_emb: [2048, 238] first4: -0.169739 0.815353 0.309631 -0.571152 [Debug] layer0_sa_input: [2048, 1085] first4: -0.716077 -0.756539 -0.048409 0.263530 [Debug] layer0_q_after_rope: [128, 16] first4: -1.541147 -1.051668 0.182985 0.461973 [Debug] layer0_k_after_rope: [128, 8] first4: -0.169739 0.815353 0.309631 -0.571152 [Debug] layer0_sa_output: [2048, 1085] first4: -1.505859 0.211914 -0.366699 0.520020 [Debug] layer0_attn_out: [2048, 1085] first4: -1.541147 -1.051668 0.182985 0.461973 [Debug] layer0_after_self_attn: [2048, 1085] first4: -1.541147 -1.051668 0.182985 0.461973 [Debug] layer0_after_cross_attn: [2048, 1085] first4: -1.584513 -0.808626 -0.317015 0.503476 [Debug] hidden_after_layer0: [2048, 1085] first4: -9.126452 0.527750 51.567387 -0.853296 [Debug] hidden_after_layer6: [2048, 1085] first4: -20.769812 -0.251942 34.724331 -4.333696 [Debug] hidden_after_layer12: [2048, 1085] first4: -14.690754 -17.032114 77.303764 30.865128 [Debug] hidden_after_layer18: [2048, 1085] first4: -27.899891 12.863169 65.200310 19.531570 [Debug] hidden_after_layer23: [2048, 1085] first4: -16.142891 45.194626 196.672028 135.397659 [Debug] dit_step0_vt: [2170, 64] first4: 0.095333 1.123520 0.322540 2.339783 [Debug] dit_step0_xt: [2170, 64] first4: 0.190003 2.105181 -0.186536 0.741302 [DiT] Step 1/8 t=1.000 [Debug] dit_step1_vt: [2170, 64] first4: -0.138275 1.330688 -0.186005 1.915009 [Debug] dit_step1_xt: [2170, 64] first4: 0.197545 2.032598 -0.176390 0.636847 [DiT] Step 2/8 t=0.955 [Debug] dit_step2_vt: [2170, 64] first4: 0.048317 1.242240 0.069305 2.385040 [Debug] dit_step2_xt: [2170, 64] first4: 0.194324 1.949782 -0.181011 0.477845 [DiT] Step 3/8 t=0.900 [Debug] dit_step3_vt: [2170, 64] first4: 0.279480 1.091293 0.230805 2.638580 [Debug] dit_step3_xt: [2170, 64] first4: 0.171034 1.858841 -0.200244 0.257963 [DiT] Step 4/8 t=0.833 [Debug] dit_step4_vt: [2170, 64] first4: 0.324036 1.029013 0.086426 2.721054 [Debug] dit_step4_xt: [2170, 64] first4: 0.136316 1.748590 -0.209504 -0.033578 [DiT] Step 5/8 t=0.750 [Debug] dit_step5_vt: [2170, 64] first4: 0.312531 0.919434 -0.286682 2.728294 [Debug] dit_step5_xt: [2170, 64] first4: 0.091668 1.617242 -0.168550 -0.423335 [DiT] Step 6/8 t=0.643 [Debug] dit_step6_vt: [2170, 64] first4: 0.209595 0.624629 -0.788853 2.786316 [Debug] dit_step6_xt: [2170, 64] first4: 0.049750 1.492316 -0.010779 -0.980598 [DiT] Step 7/8 t=0.500 [Debug] dit_step7_vt: [2170, 64] first4: 0.106659 0.098785 -1.340214 3.010498 [Debug] dit_x0: [2170, 64] first4: 0.017752 1.462680 0.391285 -1.883747 [DiT] Step 8/8 t=0.300 [DiT-Generate] Total: 305.4 ms (305.4 ms/sample) [Debug] dit_output: [2170, 64] first4: 0.017752 1.462680 0.391285 -1.883747 [Store] Unload DiT (1237.2 MB) [GGUF] /mnt/workspace/git/acestep.cpp/models/vae-BF16.gguf: 365 tensors, data at offset 30048 load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] VAE backend: Vulkan0 (CPU threads: 16) [VAE] Backend: Vulkan0, Weight buffer: 161.1 MB [VAE] Loaded: 5 blocks, upsample=1920x, F32 activations [Store] Load VAE-Dec: 711 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: 960.6 ms [Debug] vae_audio: [2, 4166400] first4: 0.000649 0.001364 0.001147 0.001676 [Store] Unload VAE-Dec (161.1 MB) [MP3] Encoding 86.8s @ 128 kbps, 48000 Hz stereo [MP3] 1388928 bytes (12.0:1), 444 ms (195.56x realtime), 32 threads [MP3] Wrote ggml-turbo/request00.mp3 [Ace-Synth] All done 2026-07-12 10:21:25.781 | 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 10:21:27.787 | WARNING | acestep.training.trainer::40 - bitsandbytes not installed. Using standard AdamW. 2026-07-12 10:21:27.792 | 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 10:21:27.887 | 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 10:21:30.517 | 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 10:21:30.517 | INFO | acestep.core.generation.handler.generate_music:generate_music:304 - [generate_music] Starting generation... 2026-07-12 10:21:30.517 | INFO | acestep.core.generation.handler.generate_music:generate_music:307 - [generate_music] Preparing inputs... 2026-07-12 10:21:30.518 | 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 10:21:30.521 | 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 10:21:30.526 | INFO | acestep.core.generation.handler.conditioning_target:_prepare_target_latents_and_wavs:78 - [generate_music] Decoding audio codes for item 0... 2026-07-12 10:21:30.735 | 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 10:21:30.737 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:119 - ====================================================================== 2026-07-12 10:21:30.737 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:120 - πŸ” [DEBUG] DiT TEXT ENCODER INPUT (Inference) 2026-07-12 10:21:30.737 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:121 - ====================================================================== 2026-07-12 10:21:30.737 | 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 10:21:30.737 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:123 - ====================================================================== 2026-07-12 10:21:30.737 | 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 10:21:30.737 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:125 - ====================================================================== 2026-07-12 10:21:30.744 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:110 - [preprocess_batch] Inferring prompt embeddings... 2026-07-12 10:21:30.759 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:113 - [preprocess_batch] Inferring lyric embeddings... 2026-07-12 10:21:30.759 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:161 - [service_generate] Generating audio... (DiT backend: PyTorch (cuda)) 2026-07-12 10:21:30.784 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:263 - [service_generate] DiT diffusion via PyTorch (cuda)... 2026-07-12 10:21:30.792 | 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 10:21:30.878 | 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 10:21:31.103 | INFO | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:42 - [generate_music] Model generation completed. Decoding latents... 2026-07-12 10:21:31.104 | 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 10:21:31.104 | DEBUG | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:65 - [generate_music] time_costs: {'encoder_time_cost': 0.0069048404693603516, 'diffusion_time_cost': 0.3116879463195801, 'diffusion_per_step_time_cost': 0.03896099328994751, 'total_time_cost': 0.31859278678894043, 'offload_time_cost': 0.0} 2026-07-12 10:21:31.118 | INFO | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:127 - [generate_music] Decoding latents with VAE... 2026-07-12 10:21:31.128 | 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 10:21:31.129 | 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 10:21:31.129 | 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 10:21:31.129 | 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 10:21:31.129 | 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 10:21:31.129 | 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 10:21:31.405 | 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 10:21:31.563 | 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 10:21:31.567 | 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-Q6_K.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-Q6_K.gguf: 678 tensors, data at offset 56800 [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-sft-Q6_K.gguf: 678 tensors, data at offset 56800 [Synth-Load] Ready: turbo=no, fa=no, batch_cfg=yes [Resolve-Params] max audio codes across batch: 434 (86.8s @ 5Hz) [Resolve-T] T=2170, S=1085 [Resolve-T] seed=42, steps=50, guidance=1.0, shift=1.0, duration=88.0s [BPE] Loaded from GGUF: 151643 vocab, 151387 merges ggml_cuda_init: found 1 CUDA devices (Total VRAM: 97247 MiB): Device 0: NVIDIA RTX PRO 6000 Blackwell Workstation Edition, compute capability 12.0, VMM: yes, VRAM: 97247 MiB load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = NVIDIA RTX PRO 6000 Blackwell Workstation Edition (NVIDIA) | uma: 0 | fp16: 1 | bf16: 0 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: NV_coopmat2 load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] TextEncoder backend: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/Qwen3-Embedding-0.6B-BF16.gguf: 310 tensors, data at offset 5337568 [Load] TextEncoder: 28L, H=1024, Nh=16/8 [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 310 tensors, 1136.5 MB into backend [Store] Load TextEnc: 330 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-Q6_K.gguf: 678 tensors, data at offset 56800 [Load] LyricEncoder: 8L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [Load] TimbreEncoder: 4L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 140 tensors, 476.3 MB into backend [Load] CondEncoder: lyric(8L), timbre(4L), text_proj, null_cond [Store] Load CondEnc: 277 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, 7.6 ms [Debug] enc_hidden: [238, 2048] first4: 1.761451 -0.050624 -0.132928 0.058356 [Store] Unload CondEnc (476.3 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-Q6_K.gguf: 678 tensors, data at offset 56800 [WeightCtx] Loaded 30 tensors, 82.2 MB into backend [Load] Detokenizer: FSQ(6->2048) + 2L encoder(S=5, 2048->64) [Store] Load FSQ-Detok: 56 ms [Context] Decoded: 434 codes -> 2170 frames (86.8s @ 25Hz) [Build-Context Batch0] Detokenizer: 164.8 ms, 434 codes [Debug] detok_output: [2170, 64] first4: -0.141068 1.454457 0.315091 -0.623697 [Store] Unload FSQ-Detok (82.2 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: -0.141068 1.454457 0.315091 -0.623697 [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-Q6_K.gguf: 678 tensors, data at offset 56800 [DiT] Self-attn: Q+K+V fused [DiT] Cross-attn: Q+K+V fused [DiT] MLP: gate+up fused [Load] null_condition_emb found (CFG available) [WeightCtx] Loaded 478 tensors, 1237.2 MB into backend [Load] DiT: 24 layers, H=2048, Nh=16/8, D=128 [Store] Load DiT: 895 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: 0.153861 -0.117528 -0.090110 0.080834 [Debug] temb: [2048] first4: -0.002466 -0.176370 0.004370 -0.002069 [Debug] temb_t: [2048] first4: -0.000999 0.003474 -0.013219 -0.002130 [Debug] temb_r: [2048] first4: -0.001467 -0.179844 0.017589 0.000062 [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: -0.041140 0.030273 0.027836 -0.025459 [Debug] temb_lin1_r: [2048] first4: 0.004272 0.006720 0.000208 -0.005103 [Debug] hidden_after_proj_in: [2048, 1085] first4: -1.086914 -0.913330 0.502930 0.445343 [Debug] proj_in_input: [192, 2170] first4: -0.141068 1.454457 0.315091 -0.623697 [Debug] enc_after_cond_emb: [2048, 238] first4: -0.196411 0.882828 0.310303 -0.576378 [Debug] layer0_sa_input: [2048, 1085] first4: -0.912818 -0.710435 -0.040195 0.295128 [Debug] layer0_q_after_rope: [128, 16] first4: -1.803784 -1.364957 -0.180621 0.386244 [Debug] layer0_k_after_rope: [128, 8] first4: -0.196411 0.882828 0.310303 -0.576378 [Debug] layer0_sa_output: [2048, 1085] first4: -1.653320 0.799316 -0.604004 0.526855 [Debug] layer0_attn_out: [2048, 1085] first4: -1.803784 -1.364957 -0.180621 0.386244 [Debug] layer0_after_self_attn: [2048, 1085] first4: -1.803784 -1.364957 -0.180621 0.386244 [Debug] layer0_after_cross_attn: [2048, 1085] first4: -1.952343 -1.017789 -0.439410 0.404539 [Debug] hidden_after_layer0: [2048, 1085] first4: -9.334003 1.276022 56.610222 -1.684794 [Debug] hidden_after_layer6: [2048, 1085] first4: -16.521595 2.434341 58.004665 -1.119201 [Debug] hidden_after_layer12: [2048, 1085] first4: -9.769129 5.897276 -22.977211 -2.485344 [Debug] hidden_after_layer18: [2048, 1085] first4: -4.989435 20.700766 -39.799831 -14.226343 [Debug] hidden_after_layer23: [2048, 1085] first4: 33.624874 62.757225 54.192257 -11.231433 [Debug] dit_step0_vt: [2170, 64] first4: -0.598694 2.498535 -0.223083 1.457159 [Debug] dit_step0_xt: [2170, 64] first4: 0.206310 2.106279 -0.167413 0.818513 [DiT] Step 1/50 t=1.000 [Debug] dit_step1_vt: [2170, 64] first4: -0.656006 2.502686 -0.071869 1.283920 [Debug] dit_step1_xt: [2170, 64] first4: 0.219430 2.056226 -0.165976 0.792835 [DiT] Step 2/50 t=0.980 [Debug] dit_step2_vt: [2170, 64] first4: -0.648682 2.468994 -0.018890 1.305298 [Debug] dit_step2_xt: [2170, 64] first4: 0.232404 2.006846 -0.165598 0.766729 [DiT] Step 3/50 t=0.960 [Debug] dit_step3_vt: [2170, 64] first4: -0.640198 2.412476 0.001770 1.317688 [Debug] dit_step3_xt: [2170, 64] first4: 0.245208 1.958596 -0.165634 0.740375 [DiT] Step 4/50 t=0.940 [Debug] dit_step4_vt: [2170, 64] first4: -0.582153 2.326782 -0.013748 1.307663 [Debug] dit_step4_xt: [2170, 64] first4: 0.256851 1.912060 -0.165359 0.714222 [DiT] Step 5/50 t=0.920 [Debug] dit_step5_vt: [2170, 64] first4: -0.552185 2.269653 -0.039383 1.305725 [Debug] dit_step5_xt: [2170, 64] first4: 0.267894 1.866668 -0.164571 0.688107 [DiT] Step 6/50 t=0.900 [Debug] dit_step6_vt: [2170, 64] first4: -0.528503 2.216431 -0.068207 1.292084 [Debug] dit_step6_xt: [2170, 64] first4: 0.278464 1.822339 -0.163207 0.662266 [DiT] Step 7/50 t=0.880 [Debug] dit_step7_vt: [2170, 64] first4: -0.507782 2.161377 -0.090714 1.287567 [Debug] dit_step7_xt: [2170, 64] first4: 0.288620 1.779111 -0.161393 0.636514 [DiT] Step 8/50 t=0.860 [Debug] dit_step8_vt: [2170, 64] first4: -0.464355 2.081848 -0.098368 1.270142 [Debug] dit_step8_xt: [2170, 64] first4: 0.297907 1.737474 -0.159425 0.611111 [DiT] Step 9/50 t=0.840 [Debug] dit_step9_vt: [2170, 64] first4: -0.407471 2.000488 -0.111923 1.259506 [Debug] dit_step9_xt: [2170, 64] first4: 0.306056 1.697465 -0.157187 0.585921 [DiT] Step 10/50 t=0.820 [Debug] dit_step10_vt: [2170, 64] first4: -0.343811 1.912598 -0.101379 1.258804 [Debug] dit_step10_xt: [2170, 64] first4: 0.312933 1.659213 -0.155159 0.560745 [DiT] Step 11/50 t=0.800 [Debug] dit_step11_vt: [2170, 64] first4: -0.305786 1.846497 -0.084579 1.252747 [Debug] dit_step11_xt: [2170, 64] first4: 0.319048 1.622283 -0.153468 0.535690 [DiT] Step 12/50 t=0.780 [Debug] dit_step12_vt: [2170, 64] first4: -0.269653 1.768738 -0.092834 1.234390 [Debug] dit_step12_xt: [2170, 64] first4: 0.324441 1.586908 -0.151611 0.511002 [DiT] Step 13/50 t=0.760 [Debug] dit_step13_vt: [2170, 64] first4: -0.253357 1.685150 -0.105377 1.222870 [Debug] dit_step13_xt: [2170, 64] first4: 0.329509 1.553205 -0.149503 0.486545 [DiT] Step 14/50 t=0.740 [Debug] dit_step14_vt: [2170, 64] first4: -0.243454 1.614258 -0.129974 1.207916 [Debug] dit_step14_xt: [2170, 64] first4: 0.334378 1.520920 -0.146904 0.462387 [DiT] Step 15/50 t=0.720 [Debug] dit_step15_vt: [2170, 64] first4: -0.248703 1.531693 -0.174377 1.175507 [Debug] dit_step15_xt: [2170, 64] first4: 0.339352 1.490286 -0.143416 0.438877 [DiT] Step 16/50 t=0.700 [Debug] dit_step16_vt: [2170, 64] first4: -0.253792 1.454933 -0.196716 1.155853 [Debug] dit_step16_xt: [2170, 64] first4: 0.344428 1.461187 -0.139482 0.415759 [DiT] Step 17/50 t=0.680 [Debug] dit_step17_vt: [2170, 64] first4: -0.257452 1.370411 -0.226746 1.146973 [Debug] dit_step17_xt: [2170, 64] first4: 0.349577 1.433779 -0.134947 0.392820 [DiT] Step 18/50 t=0.660 [Debug] dit_step18_vt: [2170, 64] first4: -0.269676 1.294464 -0.256470 1.131332 [Debug] dit_step18_xt: [2170, 64] first4: 0.354970 1.407890 -0.129818 0.370193 [DiT] Step 19/50 t=0.640 [Debug] dit_step19_vt: [2170, 64] first4: -0.276497 1.215240 -0.280457 1.110878 [Debug] dit_step19_xt: [2170, 64] first4: 0.360500 1.383585 -0.124208 0.347976 [DiT] Step 20/50 t=0.620 [Debug] dit_step20_vt: [2170, 64] first4: -0.301437 1.132507 -0.296997 1.080811 [Debug] dit_step20_xt: [2170, 64] first4: 0.366529 1.360935 -0.118269 0.326360 [DiT] Step 21/50 t=0.600 [Debug] dit_step21_vt: [2170, 64] first4: -0.330734 1.029938 -0.296143 1.057058 [Debug] dit_step21_xt: [2170, 64] first4: 0.373144 1.340336 -0.112346 0.305218 [DiT] Step 22/50 t=0.580 [Debug] dit_step22_vt: [2170, 64] first4: -0.341087 0.931213 -0.318604 1.029957 [Debug] dit_step22_xt: [2170, 64] first4: 0.379965 1.321712 -0.105974 0.284619 [DiT] Step 23/50 t=0.560 [Debug] dit_step23_vt: [2170, 64] first4: -0.363670 0.836914 -0.313110 1.004951 [Debug] dit_step23_xt: [2170, 64] first4: 0.387239 1.304973 -0.099711 0.264520 [DiT] Step 24/50 t=0.540 [Debug] dit_step24_vt: [2170, 64] first4: -0.385925 0.736084 -0.319885 0.975868 [Debug] dit_step24_xt: [2170, 64] first4: 0.394957 1.290252 -0.093314 0.245003 [DiT] Step 25/50 t=0.520 [Debug] dit_step25_vt: [2170, 64] first4: -0.403404 0.619080 -0.324402 0.957748 [Debug] dit_step25_xt: [2170, 64] first4: 0.403025 1.277870 -0.086826 0.225848 [DiT] Step 26/50 t=0.500 [Debug] dit_step26_vt: [2170, 64] first4: -0.425167 0.489441 -0.318359 0.938599 [Debug] dit_step26_xt: [2170, 64] first4: 0.411529 1.268081 -0.080458 0.207076 [DiT] Step 27/50 t=0.480 [Debug] dit_step27_vt: [2170, 64] first4: -0.453316 0.390961 -0.313782 0.918137 [Debug] dit_step27_xt: [2170, 64] first4: 0.420595 1.260262 -0.074183 0.188713 [DiT] Step 28/50 t=0.460 [Debug] dit_step28_vt: [2170, 64] first4: -0.482693 0.277069 -0.312317 0.901062 [Debug] dit_step28_xt: [2170, 64] first4: 0.430249 1.254721 -0.067936 0.170692 [DiT] Step 29/50 t=0.440 [Debug] dit_step29_vt: [2170, 64] first4: -0.495075 0.169327 -0.303833 0.889343 [Debug] dit_step29_xt: [2170, 64] first4: 0.440150 1.251334 -0.061860 0.152905 [DiT] Step 30/50 t=0.420 [Debug] dit_step30_vt: [2170, 64] first4: -0.512227 0.048664 -0.287292 0.867859 [Debug] dit_step30_xt: [2170, 64] first4: 0.450395 1.250361 -0.056114 0.135548 [DiT] Step 31/50 t=0.400 [Debug] dit_step31_vt: [2170, 64] first4: -0.535218 -0.060868 -0.280182 0.854004 [Debug] dit_step31_xt: [2170, 64] first4: 0.461099 1.251579 -0.050510 0.118468 [DiT] Step 32/50 t=0.380 [Debug] dit_step32_vt: [2170, 64] first4: -0.547300 -0.171936 -0.269653 0.839539 [Debug] dit_step32_xt: [2170, 64] first4: 0.472045 1.255017 -0.045117 0.101677 [DiT] Step 33/50 t=0.360 [Debug] dit_step33_vt: [2170, 64] first4: -0.560158 -0.283630 -0.253296 0.814606 [Debug] dit_step33_xt: [2170, 64] first4: 0.483248 1.260690 -0.040051 0.085385 [DiT] Step 34/50 t=0.340 [Debug] dit_step34_vt: [2170, 64] first4: -0.572144 -0.371674 -0.229904 0.804047 [Debug] dit_step34_xt: [2170, 64] first4: 0.494691 1.268123 -0.035453 0.069304 [DiT] Step 35/50 t=0.320 [Debug] dit_step35_vt: [2170, 64] first4: -0.576477 -0.478790 -0.223656 0.791931 [Debug] dit_step35_xt: [2170, 64] first4: 0.506221 1.277699 -0.030980 0.053465 [DiT] Step 36/50 t=0.300 [Debug] dit_step36_vt: [2170, 64] first4: -0.587448 -0.568451 -0.188576 0.761688 [Debug] dit_step36_xt: [2170, 64] first4: 0.517970 1.289068 -0.027209 0.038232 [DiT] Step 37/50 t=0.280 [Debug] dit_step37_vt: [2170, 64] first4: -0.584106 -0.656303 -0.166092 0.745087 [Debug] dit_step37_xt: [2170, 64] first4: 0.529652 1.302194 -0.023887 0.023330 [DiT] Step 38/50 t=0.260 [Debug] dit_step38_vt: [2170, 64] first4: -0.587708 -0.738281 -0.133102 0.703278 [Debug] dit_step38_xt: [2170, 64] first4: 0.541406 1.316960 -0.021225 0.009264 [DiT] Step 39/50 t=0.240 [Debug] dit_step39_vt: [2170, 64] first4: -0.579498 -0.825516 -0.108063 0.675507 [Debug] dit_step39_xt: [2170, 64] first4: 0.552996 1.333470 -0.019063 -0.004246 [DiT] Step 40/50 t=0.220 [Debug] dit_step40_vt: [2170, 64] first4: -0.577513 -0.889908 -0.064224 0.632507 [Debug] dit_step40_xt: [2170, 64] first4: 0.564546 1.351268 -0.017779 -0.016896 [DiT] Step 41/50 t=0.200 [Debug] dit_step41_vt: [2170, 64] first4: -0.569004 -0.956856 -0.025662 0.580833 [Debug] dit_step41_xt: [2170, 64] first4: 0.575926 1.370405 -0.017266 -0.028513 [DiT] Step 42/50 t=0.180 [Debug] dit_step42_vt: [2170, 64] first4: -0.550995 -1.025384 0.015945 0.537682 [Debug] dit_step42_xt: [2170, 64] first4: 0.586946 1.390913 -0.017585 -0.039266 [DiT] Step 43/50 t=0.160 [Debug] dit_step43_vt: [2170, 64] first4: -0.529221 -1.077332 0.064934 0.485443 [Debug] dit_step43_xt: [2170, 64] first4: 0.597531 1.412460 -0.018883 -0.048975 [DiT] Step 44/50 t=0.140 [Debug] dit_step44_vt: [2170, 64] first4: -0.515263 -1.132950 0.104118 0.429138 [Debug] dit_step44_xt: [2170, 64] first4: 0.607836 1.435119 -0.020966 -0.057558 [DiT] Step 45/50 t=0.120 [Debug] dit_step45_vt: [2170, 64] first4: -0.500587 -1.171204 0.134445 0.372528 [Debug] dit_step45_xt: [2170, 64] first4: 0.617848 1.458543 -0.023655 -0.065008 [DiT] Step 46/50 t=0.100 [Debug] dit_step46_vt: [2170, 64] first4: -0.479355 -1.209061 0.175198 0.321899 [Debug] dit_step46_xt: [2170, 64] first4: 0.627435 1.482724 -0.027159 -0.071446 [DiT] Step 47/50 t=0.080 [Debug] dit_step47_vt: [2170, 64] first4: -0.451599 -1.253967 0.191666 0.278292 [Debug] dit_step47_xt: [2170, 64] first4: 0.636467 1.507803 -0.030992 -0.077012 [DiT] Step 48/50 t=0.060 [Debug] dit_step48_vt: [2170, 64] first4: -0.409851 -1.287491 0.200699 0.230682 [Debug] dit_step48_xt: [2170, 64] first4: 0.644664 1.533553 -0.035006 -0.081626 [DiT] Step 49/50 t=0.040 [Debug] dit_step49_vt: [2170, 64] first4: -0.448059 -1.309284 0.183105 0.233200 [Debug] dit_x0: [2170, 64] first4: 0.653625 1.559739 -0.038668 -0.086290 [DiT] Step 50/50 t=0.020 [DiT-Generate] Total: 1690.3 ms (1690.3 ms/sample) [Debug] dit_output: [2170, 64] first4: 0.653625 1.559739 -0.038668 -0.086290 [Store] Unload DiT (1237.2 MB) [GGUF] /mnt/workspace/git/acestep.cpp/models/vae-BF16.gguf: 365 tensors, data at offset 30048 load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] VAE backend: Vulkan0 (CPU threads: 16) [VAE] Backend: Vulkan0, Weight buffer: 161.1 MB [VAE] Loaded: 5 blocks, upsample=1920x, F32 activations [Store] Load VAE-Dec: 672 ms [VAE] Tiled decode: 3 tiles (chunk=1024, overlap=64, stride=896) [VAE] Graph: 479 nodes, T_latent=960 [VAE] Upsample factor: 1920.00 (expected ~1920) [VAE] Graph: 479 nodes, T_latent=1024 [VAE] Graph: 479 nodes, T_latent=442 [VAE] Tiled decode done: 3 tiles -> T_audio=4166400 (86.80s @ 48kHz) [VAE-Decode Batch0] Decode: 949.0 ms [Debug] vae_audio: [2, 4166400] first4: -0.001613 -0.001147 -0.001429 -0.001017 [Store] Unload VAE-Dec (161.1 MB) [MP3] Encoding 86.8s @ 128 kbps, 48000 Hz stereo [MP3] 1388928 bytes (12.0:1), 399 ms (217.50x realtime), 32 threads [MP3] Wrote ggml-sft/request00.mp3 [Ace-Synth] All done 2026-07-12 10:21:38.176 | 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 10:21:40.181 | INFO | acestep.core.generation.handler.generate_music:generate_music:304 - [generate_music] Starting generation... 2026-07-12 10:21:40.181 | INFO | acestep.core.generation.handler.generate_music:generate_music:307 - [generate_music] Preparing inputs... 2026-07-12 10:21:40.181 | 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 10:21:40.185 | 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 10:21:40.189 | INFO | acestep.core.generation.handler.conditioning_target:_prepare_target_latents_and_wavs:78 - [generate_music] Decoding audio codes for item 0... 2026-07-12 10:21:40.194 | 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 10:21:40.195 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:119 - ====================================================================== 2026-07-12 10:21:40.195 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:120 - πŸ” [DEBUG] DiT TEXT ENCODER INPUT (Inference) 2026-07-12 10:21:40.195 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:121 - ====================================================================== 2026-07-12 10:21:40.195 | 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 10:21:40.195 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:123 - ====================================================================== 2026-07-12 10:21:40.195 | 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 10:21:40.195 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:125 - ====================================================================== 2026-07-12 10:21:40.199 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:110 - [preprocess_batch] Inferring prompt embeddings... 2026-07-12 10:21:40.207 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:113 - [preprocess_batch] Inferring lyric embeddings... 2026-07-12 10:21:40.207 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:161 - [service_generate] Generating audio... (DiT backend: PyTorch (cuda)) 2026-07-12 10:21:40.213 | 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-Q6_K.gguf [GGML] Running acestep-v15-turbo-Q6_K.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 0.999669 detok_output 0.999971 context 0.999982 noise 1.000000 temb_t 0.999990 hidden_after_proj_in 0.999982 enc_after_cond_emb 0.999695 layer0_sa_output 0.999774 hidden_after_layer0 0.999725 hidden_after_layer6 0.999856 hidden_after_layer12 0.998840 hidden_after_layer18 0.995816 hidden_after_layer23 0.992138 dit_step0_vt 0.970320 dit_step0_xt 0.999935 dit_step1_vt 0.972968 dit_step1_xt 0.999793 dit_step2_vt 0.975971 dit_step2_xt 0.999443 dit_step3_vt 0.976319 dit_step3_xt 0.998652 dit_step4_vt 0.974951 dit_step4_xt 0.996845 dit_step5_vt 0.972674 dit_step5_xt 0.992851 dit_step6_vt 0.969512 dit_step6_xt 0.985330 dit_step7_vt 0.962142 dit_x0 0.974210 vae_audio 0.889683 vae_audio (STFT cosine) 0.969985 [Turbo] Error growth GGML vs Python stage cos max_err mean_err mean_A std_A mean_B std_B dit_step0_xt 0.999935 0.149187 0.007381 -0.002253 0.973074 -0.002342 0.972003 dit_step1_xt 0.999793 0.295445 0.012751 -0.005193 0.942934 -0.005313 0.941730 dit_step2_xt 0.999443 0.481156 0.019796 -0.009128 0.909520 -0.009311 0.908527 dit_step3_xt 0.998652 0.737996 0.029471 -0.014416 0.874223 -0.014577 0.873624 dit_step4_xt 0.996845 1.052053 0.043064 -0.021424 0.842418 -0.021660 0.841995 dit_step5_xt 0.992851 1.524985 0.063177 -0.031453 0.825720 -0.032109 0.824593 dit_step6_xt 0.985330 2.174533 0.093499 -0.045672 0.856514 -0.046482 0.855546 [SFT] steps=50, shift=1.0 | acestep-v15-sft-Q6_K.gguf [GGML] Running acestep-v15-sft-Q6_K.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-Q6_K.gguf: 830 tensors, data at offset 69088 [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-turbo-Q6_K.gguf: 830 tensors, data at offset 69088 [Synth-Load] Ready: turbo=yes, fa=no, batch_cfg=yes [Resolve-Params] max audio codes across batch: 434 (86.8s @ 5Hz) [Resolve-T] T=2170, S=1085 [Resolve-T] seed=42, steps=8, guidance=1.0, shift=3.0, duration=88.0s [BPE] Loaded from GGUF: 151643 vocab, 151387 merges ggml_cuda_init: found 1 CUDA devices (Total VRAM: 97247 MiB): Device 0: NVIDIA RTX PRO 6000 Blackwell Workstation Edition, compute capability 12.0, VMM: yes, VRAM: 97247 MiB load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = NVIDIA RTX PRO 6000 Blackwell Workstation Edition (NVIDIA) | uma: 0 | fp16: 1 | bf16: 0 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: NV_coopmat2 load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] TextEncoder backend: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/Qwen3-Embedding-0.6B-BF16.gguf: 310 tensors, data at offset 5337568 [Load] TextEncoder: 28L, H=1024, Nh=16/8 [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 310 tensors, 1136.5 MB into backend [Store] Load TextEnc: 367 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-Q6_K.gguf: 830 tensors, data at offset 69088 [Load] LyricEncoder: 8L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [Load] TimbreEncoder: 4L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 141 tensors, 476.3 MB into backend [Load] CondEncoder: lyric(8L), timbre(4L, CLS), text_proj, null_cond [Store] Load CondEnc: 280 ms [CondEnc] Lyric sliding mask: 167x167, window=128 [CondEnc] Timbre sliding mask: 2x2, window=128 (CLS) [Encode] Packed: lyric=167 + timbre=1 + text=70 = 238 tokens [Encode-Text Batch0] 70+167 tokens -> enc_S=238, 7.7 ms [Debug] enc_hidden: [238, 2048] first4: 0.687665 -0.019437 -0.051735 0.022813 [Store] Unload CondEnc (476.3 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-Q6_K.gguf: 830 tensors, data at offset 69088 [WeightCtx] Loaded 30 tensors, 82.2 MB into backend [Load] Detokenizer: FSQ(6->2048) + 2L encoder(S=5, 2048->64) [Store] Load FSQ-Detok: 58 ms [Context] Decoded: 434 codes -> 2170 frames (86.8s @ 25Hz) [Build-Context Batch0] Detokenizer: 174.0 ms, 434 codes [Debug] detok_output: [2170, 64] first4: -0.141068 1.454457 0.315091 -0.623697 [Store] Unload FSQ-Detok (82.2 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: -0.141068 1.454457 0.315091 -0.623697 [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-Q6_K.gguf: 830 tensors, data at offset 69088 [DiT] Self-attn: Q+K+V fused [DiT] Cross-attn: Q+K+V fused [DiT] MLP: gate+up fused [Load] null_condition_emb found (CFG available) [WeightCtx] Loaded 630 tensors, 3267.8 MB into backend [Load] DiT: 32 layers, H=2560, Nh=32/8, D=128 [Store] Load DiT: 1974 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: 0.019342 0.105586 0.065691 -0.074838 [Debug] temb: [2560] first4: 0.026559 -0.022629 -0.016440 0.046999 [Debug] temb_t: [2560] first4: 0.014705 -0.022100 -0.020716 0.048496 [Debug] temb_r: [2560] first4: 0.011854 -0.000528 0.004277 -0.001497 [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: 0.062749 -0.017898 0.137387 -0.111467 [Debug] temb_lin1_r: [2560] first4: -0.011992 0.015308 -0.017417 0.011657 [Debug] hidden_after_proj_in: [2560, 1085] first4: 0.338501 -0.580811 0.254028 -0.190186 [Debug] proj_in_input: [192, 2170] first4: -0.141068 1.454457 0.315091 -0.623697 [Debug] enc_after_cond_emb: [2560, 238] first4: 0.030350 0.668533 0.102203 0.862045 [Debug] layer0_sa_input: [2560, 1085] first4: 0.042987 -0.049411 0.095042 -0.135277 [Debug] layer0_q_after_rope: [128, 32] first4: -0.256330 -2.178915 0.150258 0.053293 [Debug] layer0_k_after_rope: [128, 8] first4: 0.561523 2.197266 0.704590 0.185669 [Debug] layer0_sa_output: [2560, 1085] first4: 1.857422 4.464844 -0.962402 5.574219 [Debug] layer0_attn_out: [4096, 1085] first4: 0.487817 -2.312559 -0.447517 -2.875886 [Debug] layer0_after_self_attn: [2560, 1085] first4: 0.487817 -2.312559 -0.447517 -2.875886 [Debug] layer0_after_cross_attn: [2560, 1085] first4: 0.327051 -2.049131 -0.300178 -3.771882 [Debug] hidden_after_layer0: [2560, 1085] first4: 1.647189 -2.465074 238.475204 -3.746695 [Debug] hidden_after_layer6: [2560, 1085] first4: -6.249263 -2.887747 317.015594 -12.787234 [Debug] hidden_after_layer12: [2560, 1085] first4: -10.688093 29.653015 256.670197 31.328789 [Debug] hidden_after_layer18: [2560, 1085] first4: 42.621094 11.098314 273.272552 5.509533 [Debug] hidden_after_layer31: [2560, 1085] first4: -93.228546 -280.888794 -130.382141 214.595444 [Debug] dit_step0_vt: [2170, 64] first4: -1.055420 1.725586 0.818054 1.336182 [Debug] dit_step0_xt: [2170, 64] first4: 0.242310 2.077814 -0.209059 0.786921 [DiT] Step 1/8 t=1.000 [Debug] dit_step1_vt: [2170, 64] first4: -1.083984 1.956772 1.090912 1.571533 [Debug] dit_step1_xt: [2170, 64] first4: 0.301436 1.971081 -0.268564 0.701201 [DiT] Step 2/8 t=0.955 [Debug] dit_step2_vt: [2170, 64] first4: -0.479065 1.410583 0.879578 0.881226 [Debug] dit_step2_xt: [2170, 64] first4: 0.333374 1.877042 -0.327202 0.642452 [DiT] Step 3/8 t=0.900 [Debug] dit_step3_vt: [2170, 64] first4: -0.327209 1.375343 0.912430 0.938148 [Debug] dit_step3_xt: [2170, 64] first4: 0.360641 1.762431 -0.403238 0.564273 [DiT] Step 4/8 t=0.833 [Debug] dit_step4_vt: [2170, 64] first4: -0.773865 1.509817 1.334496 0.780823 [Debug] dit_step4_xt: [2170, 64] first4: 0.443555 1.600664 -0.546220 0.480614 [DiT] Step 5/8 t=0.750 [Debug] dit_step5_vt: [2170, 64] first4: -1.360840 2.323242 0.541870 0.573608 [Debug] dit_step5_xt: [2170, 64] first4: 0.637961 1.268773 -0.623630 0.398670 [DiT] Step 6/8 t=0.643 [Debug] dit_step6_vt: [2170, 64] first4: 0.811981 2.078766 0.582397 0.682983 [Debug] dit_step6_xt: [2170, 64] first4: 0.475565 0.853019 -0.740109 0.262073 [DiT] Step 7/8 t=0.500 [Debug] dit_step7_vt: [2170, 64] first4: -1.903809 1.918335 0.950150 0.200073 [Debug] dit_x0: [2170, 64] first4: 1.046707 0.277519 -1.025154 0.202051 [DiT] Step 8/8 t=0.300 [DiT-Generate] Total: 700.0 ms (700.0 ms/sample) [Debug] dit_output: [2170, 64] first4: 1.046707 0.277519 -1.025154 0.202051 [Store] Unload DiT (3267.8 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: 683 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: 954.2 ms [Debug] vae_audio: [2, 4166400] first4: 0.001064 0.002950 0.002825 0.002566 [Store] Unload VAE-Dec (161.1 MB) [MP3] Encoding 86.8s @ 128 kbps, 48000 Hz stereo [MP3] 1388928 bytes (12.0:1), 484 ms (179.26x realtime), 32 threads [MP3] Wrote ggml-xl-turbo/request00.mp3 [Ace-Synth] All done 2026-07-12 10:21:49.290 | 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 0.999669 detok_output 0.999971 context 0.999982 noise 1.000000 temb_t 0.999973 hidden_after_proj_in 0.999981 enc_after_cond_emb 0.999696 layer0_sa_output 0.999789 hidden_after_layer0 0.999786 hidden_after_layer6 0.999732 hidden_after_layer12 0.999287 hidden_after_layer18 0.998485 hidden_after_layer23 0.998801 dit_step0_vt 0.998680 dit_step0_xt 1.000000 dit_step5_vt 0.999085 dit_step5_xt 0.999973 dit_step10_vt 0.998441 dit_step10_xt 0.999906 dit_step15_vt 0.997256 dit_step15_xt 0.999749 dit_step20_vt 0.995494 dit_step20_xt 0.999398 dit_step25_vt 0.992245 dit_step25_xt 0.998771 dit_step30_vt 0.987685 dit_step30_xt 0.997718 dit_step35_vt 0.982445 dit_step35_xt 0.996310 dit_step40_vt 0.978842 dit_step40_xt 0.994808 dit_step45_vt 0.982232 dit_step45_xt 0.993696 dit_step49_vt 0.988152 dit_x0 0.993302 vae_audio 0.967914 vae_audio (STFT cosine) 0.987382 [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 0.020207 0.001499 -0.001652 0.978626 -0.001619 0.978872 dit_step5_xt 0.999973 0.085383 0.005190 -0.008320 0.885670 -0.008480 0.883811 dit_step10_xt 0.999906 0.107757 0.008434 -0.014877 0.802088 -0.015075 0.802324 dit_step15_xt 0.999749 0.151539 0.012549 -0.021338 0.731745 -0.021933 0.731304 dit_step20_xt 0.999398 0.214817 0.018007 -0.027801 0.679021 -0.028882 0.679066 dit_step25_xt 0.998771 0.355535 0.024503 -0.034302 0.648347 -0.035643 0.650780 dit_step30_xt 0.997718 0.594063 0.032916 -0.040816 0.642984 -0.042577 0.647107 dit_step35_xt 0.996310 0.858096 0.042919 -0.047313 0.663686 -0.049471 0.669256 dit_step40_xt 0.994808 1.121802 0.054162 -0.053810 0.708067 -0.056532 0.715783 dit_step45_xt 0.993696 1.356994 0.064881 -0.060381 0.771843 -0.063533 0.780575 [Xl-turbo] steps=8, shift=3.0 | acestep-v15-xl-turbo-Q6_K.gguf [GGML] Running acestep-v15-xl-turbo-Q6_K.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 10:21:52.343 | INFO | acestep.core.generation.handler.generate_music:generate_music:304 - [generate_music] Starting generation... 2026-07-12 10:21:52.343 | INFO | acestep.core.generation.handler.generate_music:generate_music:307 - [generate_music] Preparing inputs... 2026-07-12 10:21:52.344 | 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 10:21:52.347 | 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 10:21:52.352 | INFO | acestep.core.generation.handler.conditioning_target:_prepare_target_latents_and_wavs:78 - [generate_music] Decoding audio codes for item 0... 2026-07-12 10:21:52.356 | 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 10:21:52.358 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:119 - ====================================================================== 2026-07-12 10:21:52.358 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:120 - πŸ” [DEBUG] DiT TEXT ENCODER INPUT (Inference) 2026-07-12 10:21:52.358 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:121 - ====================================================================== 2026-07-12 10:21:52.358 | 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 10:21:52.358 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:123 - ====================================================================== 2026-07-12 10:21:52.358 | 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 10:21:52.358 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:125 - ====================================================================== 2026-07-12 10:21:52.362 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:110 - [preprocess_batch] Inferring prompt embeddings... 2026-07-12 10:21:52.370 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:113 - [preprocess_batch] Inferring lyric embeddings... 2026-07-12 10:21:52.370 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:161 - [service_generate] Generating audio... (DiT backend: PyTorch (cuda)) 2026-07-12 10:21:52.377 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:263 - [service_generate] DiT diffusion via PyTorch (cuda)... 2026-07-12 10:21:52.383 | 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 10:21:52.958 | INFO | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:42 - [generate_music] Model generation completed. Decoding latents... 2026-07-12 10:21:52.959 | 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 10:21:52.959 | DEBUG | acestep.core.generation.handler.generate_music_decode:_prepare_generate_music_decode_state:65 - [generate_music] time_costs: {'encoder_time_cost': 0.006407976150512695, 'diffusion_time_cost': 0.5745298862457275, 'diffusion_per_step_time_cost': 0.07181623578071594, 'total_time_cost': 0.5809378623962402, 'offload_time_cost': 0.0} 2026-07-12 10:21:52.982 | INFO | acestep.core.generation.handler.generate_music_decode:_decode_generate_music_pred_latents:127 - [generate_music] Decoding latents with VAE... 2026-07-12 10:21:52.985 | 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 10:21:52.985 | 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 10:21:52.985 | 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 10:21:52.985 | 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 10:21:52.985 | 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 10:21:52.985 | 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 10:21:53.261 | 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 10:21:53.462 | 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 10:21:53.465 | 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-Q6_K.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-Q6_K.gguf: 830 tensors, data at offset 69056 [GGUF] /mnt/workspace/git/acestep.cpp/models/acestep-v15-xl-sft-Q6_K.gguf: 830 tensors, data at offset 69056 [Synth-Load] Ready: turbo=no, fa=no, batch_cfg=yes [Resolve-Params] max audio codes across batch: 434 (86.8s @ 5Hz) [Resolve-T] T=2170, S=1085 [Resolve-T] seed=42, steps=50, guidance=1.0, shift=1.0, duration=88.0s [BPE] Loaded from GGUF: 151643 vocab, 151387 merges ggml_cuda_init: found 1 CUDA devices (Total VRAM: 97247 MiB): Device 0: NVIDIA RTX PRO 6000 Blackwell Workstation Edition, compute capability 12.0, VMM: yes, VRAM: 97247 MiB load_backend: loaded CUDA backend from /mnt/workspace/git/acestep.cpp/build/libggml-cuda.so ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = NVIDIA RTX PRO 6000 Blackwell Workstation Edition (NVIDIA) | uma: 0 | fp16: 1 | bf16: 0 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: NV_coopmat2 load_backend: loaded Vulkan backend from /mnt/workspace/git/acestep.cpp/build/libggml-vulkan.so load_backend: loaded CPU backend from /mnt/workspace/git/acestep.cpp/build/libggml-cpu-zen4.so [Load] TextEncoder backend: Vulkan0 (CPU threads: 16) [GGUF] /mnt/workspace/git/acestep.cpp/models/Qwen3-Embedding-0.6B-BF16.gguf: 310 tensors, data at offset 5337568 [Load] TextEncoder: 28L, H=1024, Nh=16/8 [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 310 tensors, 1136.5 MB into backend [Store] Load TextEnc: 332 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-Q6_K.gguf: 830 tensors, data at offset 69056 [Load] LyricEncoder: 8L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [Load] TimbreEncoder: 4L [Qwen3] Attn: Q+K+V fused [Qwen3] MLP: gate+up fused [WeightCtx] Loaded 141 tensors, 476.3 MB into backend [Load] CondEncoder: lyric(8L), timbre(4L, CLS), text_proj, null_cond [Store] Load CondEnc: 298 ms [CondEnc] Lyric sliding mask: 167x167, window=128 [CondEnc] Timbre sliding mask: 2x2, window=128 (CLS) [Encode] Packed: lyric=167 + timbre=1 + text=70 = 238 tokens [Encode-Text Batch0] 70+167 tokens -> enc_S=238, 7.8 ms [Debug] enc_hidden: [238, 2048] first4: 0.687665 -0.019437 -0.051735 0.022813 [Store] Unload CondEnc (476.3 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-Q6_K.gguf: 830 tensors, data at offset 69056 [WeightCtx] Loaded 30 tensors, 82.2 MB into backend [Load] Detokenizer: FSQ(6->2048) + 2L encoder(S=5, 2048->64) [Store] Load FSQ-Detok: 60 ms [Context] Decoded: 434 codes -> 2170 frames (86.8s @ 25Hz) [Build-Context Batch0] Detokenizer: 164.6 ms, 434 codes [Debug] detok_output: [2170, 64] first4: -0.141068 1.454457 0.315091 -0.623697 [Store] Unload FSQ-Detok (82.2 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: -0.141068 1.454457 0.315091 -0.623697 [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-Q6_K.gguf: 830 tensors, data at offset 69056 [DiT] Self-attn: Q+K+V fused [DiT] Cross-attn: Q+K+V fused [DiT] MLP: gate+up fused [Load] null_condition_emb found (CFG available) [WeightCtx] Loaded 630 tensors, 3267.8 MB into backend [Load] DiT: 32 layers, H=2560, Nh=32/8, D=128 [Store] Load DiT: 2007 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: 0.010424 0.102531 0.074302 -0.034202 [Debug] temb: [2560] first4: 0.009747 -0.007826 -0.006004 0.004595 [Debug] temb_t: [2560] first4: 0.008220 -0.008037 -0.006556 0.003609 [Debug] temb_r: [2560] first4: 0.001527 0.000211 0.000553 0.000986 [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: 0.033989 -0.083181 0.088318 -0.064778 [Debug] temb_lin1_r: [2560] first4: 0.009306 -0.004837 -0.002965 -0.011076 [Debug] hidden_after_proj_in: [2560, 1085] first4: 0.328003 -0.496826 0.124268 -0.213867 [Debug] proj_in_input: [192, 2170] first4: -0.141068 1.454457 0.315091 -0.623697 [Debug] enc_after_cond_emb: [2560, 238] first4: 0.038025 0.641449 0.115051 0.859642 [Debug] layer0_sa_input: [2560, 1085] first4: 0.034297 -0.022830 0.100420 -0.099694 [Debug] layer0_q_after_rope: [128, 32] first4: -0.274132 -2.269409 0.038249 -0.161455 [Debug] layer0_k_after_rope: [128, 8] first4: 0.650391 3.011719 0.792969 1.776367 [Debug] layer0_sa_output: [2560, 1085] first4: 2.236328 4.519531 -1.233398 8.109375 [Debug] layer0_attn_out: [4096, 1085] first4: 0.475312 -2.229294 -0.791403 -4.241121 [Debug] layer0_after_self_attn: [2560, 1085] first4: 0.475312 -2.229294 -0.791403 -4.241121 [Debug] layer0_after_cross_attn: [2560, 1085] first4: 0.284638 -1.942917 -0.728964 -4.300966 [Debug] hidden_after_layer0: [2560, 1085] first4: -1.650337 -1.272049 248.750458 -4.150724 [Debug] hidden_after_layer6: [2560, 1085] first4: -7.702454 3.899193 308.505615 -21.138502 [Debug] hidden_after_layer12: [2560, 1085] first4: -21.748184 38.873428 226.090912 17.838795 [Debug] hidden_after_layer18: [2560, 1085] first4: 10.328800 -13.007311 234.017868 19.672960 [Debug] hidden_after_layer31: [2560, 1085] first4: 67.432190 -105.032913 9.238171 86.562019 [Debug] dit_step0_vt: [2170, 64] first4: -0.314575 2.208008 0.041260 1.676666 [Debug] dit_step0_xt: [2170, 64] first4: 0.200627 2.112090 -0.172700 0.814123 [DiT] Step 1/50 t=1.000 [Debug] dit_step1_vt: [2170, 64] first4: -0.345947 2.152954 0.106934 1.538635 [Debug] dit_step1_xt: [2170, 64] first4: 0.207546 2.069031 -0.174839 0.783350 [DiT] Step 2/50 t=0.980 [Debug] dit_step2_vt: [2170, 64] first4: -0.323853 2.163086 0.104248 1.548203 [Debug] dit_step2_xt: [2170, 64] first4: 0.214023 2.025769 -0.176924 0.752386 [DiT] Step 3/50 t=0.960 [Debug] dit_step3_vt: [2170, 64] first4: -0.269043 2.128418 0.111816 1.547470 [Debug] dit_step3_xt: [2170, 64] first4: 0.219404 1.983201 -0.179160 0.721437 [DiT] Step 4/50 t=0.940 [Debug] dit_step4_vt: [2170, 64] first4: -0.268677 2.036346 0.099731 1.565331 [Debug] dit_step4_xt: [2170, 64] first4: 0.224778 1.942474 -0.181155 0.690130 [DiT] Step 5/50 t=0.920 [Debug] dit_step5_vt: [2170, 64] first4: -0.250000 2.025787 0.102295 1.575562 [Debug] dit_step5_xt: [2170, 64] first4: 0.229778 1.901958 -0.183201 0.658619 [DiT] Step 6/50 t=0.900 [Debug] dit_step6_vt: [2170, 64] first4: -0.245361 1.995575 0.121826 1.592834 [Debug] dit_step6_xt: [2170, 64] first4: 0.234685 1.862046 -0.185637 0.626762 [DiT] Step 7/50 t=0.880 [Debug] dit_step7_vt: [2170, 64] first4: -0.258545 1.965607 0.144165 1.594177 [Debug] dit_step7_xt: [2170, 64] first4: 0.239856 1.822734 -0.188521 0.594879 [DiT] Step 8/50 t=0.860 [Debug] dit_step8_vt: [2170, 64] first4: -0.271729 1.928650 0.151611 1.587662 [Debug] dit_step8_xt: [2170, 64] first4: 0.245291 1.784161 -0.191553 0.563125 [DiT] Step 9/50 t=0.840 [Debug] dit_step9_vt: [2170, 64] first4: -0.304688 1.866760 0.122559 1.582617 [Debug] dit_step9_xt: [2170, 64] first4: 0.251384 1.746826 -0.194004 0.531473 [DiT] Step 10/50 t=0.820 [Debug] dit_step10_vt: [2170, 64] first4: -0.313599 1.826126 0.104492 1.574612 [Debug] dit_step10_xt: [2170, 64] first4: 0.257656 1.710304 -0.196094 0.499981 [DiT] Step 11/50 t=0.800 [Debug] dit_step11_vt: [2170, 64] first4: -0.309937 1.779251 0.089966 1.563148 [Debug] dit_step11_xt: [2170, 64] first4: 0.263855 1.674719 -0.197893 0.468718 [DiT] Step 12/50 t=0.780 [Debug] dit_step12_vt: [2170, 64] first4: -0.292480 1.730774 0.085205 1.553482 [Debug] dit_step12_xt: [2170, 64] first4: 0.269705 1.640103 -0.199597 0.437648 [DiT] Step 13/50 t=0.760 [Debug] dit_step13_vt: [2170, 64] first4: -0.292603 1.679630 0.063110 1.532211 [Debug] dit_step13_xt: [2170, 64] first4: 0.275557 1.606511 -0.200859 0.407004 [DiT] Step 14/50 t=0.740 [Debug] dit_step14_vt: [2170, 64] first4: -0.283813 1.621185 0.044067 1.520142 [Debug] dit_step14_xt: [2170, 64] first4: 0.281233 1.574087 -0.201741 0.376601 [DiT] Step 15/50 t=0.720 [Debug] dit_step15_vt: [2170, 64] first4: -0.272705 1.577637 0.039062 1.519104 [Debug] dit_step15_xt: [2170, 64] first4: 0.286687 1.542534 -0.202522 0.346219 [DiT] Step 16/50 t=0.700 [Debug] dit_step16_vt: [2170, 64] first4: -0.244995 1.527832 0.013916 1.510345 [Debug] dit_step16_xt: [2170, 64] first4: 0.291587 1.511978 -0.202800 0.316012 [DiT] Step 17/50 t=0.680 [Debug] dit_step17_vt: [2170, 64] first4: -0.214355 1.469971 -0.004028 1.506317 [Debug] dit_step17_xt: [2170, 64] first4: 0.295874 1.482578 -0.202720 0.285886 [DiT] Step 18/50 t=0.660 [Debug] dit_step18_vt: [2170, 64] first4: -0.178772 1.420837 -0.018311 1.499512 [Debug] dit_step18_xt: [2170, 64] first4: 0.299449 1.454161 -0.202354 0.255896 [DiT] Step 19/50 t=0.640 [Debug] dit_step19_vt: [2170, 64] first4: -0.142639 1.368042 -0.027832 1.486420 [Debug] dit_step19_xt: [2170, 64] first4: 0.302302 1.426801 -0.201797 0.226167 [DiT] Step 20/50 t=0.620 [Debug] dit_step20_vt: [2170, 64] first4: -0.118835 1.328308 -0.026611 1.480621 [Debug] dit_step20_xt: [2170, 64] first4: 0.304679 1.400234 -0.201265 0.196555 [DiT] Step 21/50 t=0.600 [Debug] dit_step21_vt: [2170, 64] first4: -0.081909 1.274170 -0.025269 1.467712 [Debug] dit_step21_xt: [2170, 64] first4: 0.306317 1.374751 -0.200759 0.167201 [DiT] Step 22/50 t=0.580 [Debug] dit_step22_vt: [2170, 64] first4: -0.062927 1.228699 -0.042969 1.455170 [Debug] dit_step22_xt: [2170, 64] first4: 0.307576 1.350177 -0.199900 0.138097 [DiT] Step 23/50 t=0.560 [Debug] dit_step23_vt: [2170, 64] first4: -0.056030 1.174011 -0.048828 1.435028 [Debug] dit_step23_xt: [2170, 64] first4: 0.308696 1.326697 -0.198923 0.109397 [DiT] Step 24/50 t=0.540 [Debug] dit_step24_vt: [2170, 64] first4: -0.034973 1.119141 -0.049805 1.423492 [Debug] dit_step24_xt: [2170, 64] first4: 0.309396 1.304314 -0.197927 0.080927 [DiT] Step 25/50 t=0.520 [Debug] dit_step25_vt: [2170, 64] first4: -0.031128 1.073425 -0.073486 1.394958 [Debug] dit_step25_xt: [2170, 64] first4: 0.310018 1.282845 -0.196458 0.053028 [DiT] Step 26/50 t=0.500 [Debug] dit_step26_vt: [2170, 64] first4: -0.033813 1.010132 -0.059814 1.364716 [Debug] dit_step26_xt: [2170, 64] first4: 0.310695 1.262643 -0.195261 0.025733 [DiT] Step 27/50 t=0.480 [Debug] dit_step27_vt: [2170, 64] first4: -0.040344 0.946167 -0.079224 1.354706 [Debug] dit_step27_xt: [2170, 64] first4: 0.311501 1.243720 -0.193677 -0.001361 [DiT] Step 28/50 t=0.460 [Debug] dit_step28_vt: [2170, 64] first4: -0.041992 0.883789 -0.086365 1.327759 [Debug] dit_step28_xt: [2170, 64] first4: 0.312341 1.226044 -0.191949 -0.027916 [DiT] Step 29/50 t=0.440 [Debug] dit_step29_vt: [2170, 64] first4: -0.057678 0.819550 -0.085571 1.306885 [Debug] dit_step29_xt: [2170, 64] first4: 0.313495 1.209653 -0.190238 -0.054054 [DiT] Step 30/50 t=0.420 [Debug] dit_step30_vt: [2170, 64] first4: -0.074585 0.744904 -0.103943 1.272522 [Debug] dit_step30_xt: [2170, 64] first4: 0.314987 1.194755 -0.188159 -0.079504 [DiT] Step 31/50 t=0.400 [Debug] dit_step31_vt: [2170, 64] first4: -0.082520 0.663101 -0.111267 1.266479 [Debug] dit_step31_xt: [2170, 64] first4: 0.316637 1.181493 -0.185934 -0.104834 [DiT] Step 32/50 t=0.380 [Debug] dit_step32_vt: [2170, 64] first4: -0.105469 0.578842 -0.115479 1.236633 [Debug] dit_step32_xt: [2170, 64] first4: 0.318746 1.169916 -0.183624 -0.129566 [DiT] Step 33/50 t=0.360 [Debug] dit_step33_vt: [2170, 64] first4: -0.128723 0.493504 -0.102448 1.200378 [Debug] dit_step33_xt: [2170, 64] first4: 0.321321 1.160046 -0.181575 -0.153574 [DiT] Step 34/50 t=0.340 [Debug] dit_step34_vt: [2170, 64] first4: -0.137238 0.403448 -0.109604 1.165649 [Debug] dit_step34_xt: [2170, 64] first4: 0.324066 1.151977 -0.179383 -0.176887 [DiT] Step 35/50 t=0.320 [Debug] dit_step35_vt: [2170, 64] first4: -0.130219 0.315063 -0.097340 1.150818 [Debug] dit_step35_xt: [2170, 64] first4: 0.326670 1.145676 -0.177436 -0.199903 [DiT] Step 36/50 t=0.300 [Debug] dit_step36_vt: [2170, 64] first4: -0.146214 0.206635 -0.097595 1.113098 [Debug] dit_step36_xt: [2170, 64] first4: 0.329594 1.141543 -0.175485 -0.222165 [DiT] Step 37/50 t=0.280 [Debug] dit_step37_vt: [2170, 64] first4: -0.169727 0.112862 -0.086578 1.076172 [Debug] dit_step37_xt: [2170, 64] first4: 0.332989 1.139286 -0.173753 -0.243689 [DiT] Step 38/50 t=0.260 [Debug] dit_step38_vt: [2170, 64] first4: -0.190903 -0.008713 -0.077499 1.026855 [Debug] dit_step38_xt: [2170, 64] first4: 0.336807 1.139460 -0.172203 -0.264226 [DiT] Step 39/50 t=0.240 [Debug] dit_step39_vt: [2170, 64] first4: -0.222321 -0.119934 -0.063576 0.996216 [Debug] dit_step39_xt: [2170, 64] first4: 0.341253 1.141859 -0.170931 -0.284150 [DiT] Step 40/50 t=0.220 [Debug] dit_step40_vt: [2170, 64] first4: -0.236450 -0.227615 -0.046600 0.956116 [Debug] dit_step40_xt: [2170, 64] first4: 0.345982 1.146411 -0.169999 -0.303272 [DiT] Step 41/50 t=0.200 [Debug] dit_step41_vt: [2170, 64] first4: -0.268433 -0.331869 -0.025604 0.919067 [Debug] dit_step41_xt: [2170, 64] first4: 0.351351 1.153048 -0.169487 -0.321654 [DiT] Step 42/50 t=0.180 [Debug] dit_step42_vt: [2170, 64] first4: -0.294067 -0.425491 0.008154 0.872803 [Debug] dit_step42_xt: [2170, 64] first4: 0.357232 1.161558 -0.169650 -0.339110 [DiT] Step 43/50 t=0.160 [Debug] dit_step43_vt: [2170, 64] first4: -0.325439 -0.533417 0.026215 0.817383 [Debug] dit_step43_xt: [2170, 64] first4: 0.363741 1.172226 -0.170175 -0.355457 [DiT] Step 44/50 t=0.140 [Debug] dit_step44_vt: [2170, 64] first4: -0.351013 -0.623505 0.077759 0.767273 [Debug] dit_step44_xt: [2170, 64] first4: 0.370761 1.184697 -0.171730 -0.370803 [DiT] Step 45/50 t=0.120 [Debug] dit_step45_vt: [2170, 64] first4: -0.390289 -0.699265 0.103607 0.713623 [Debug] dit_step45_xt: [2170, 64] first4: 0.378567 1.198682 -0.173802 -0.385075 [DiT] Step 46/50 t=0.100 [Debug] dit_step46_vt: [2170, 64] first4: -0.436661 -0.767632 0.147705 0.667175 [Debug] dit_step46_xt: [2170, 64] first4: 0.387300 1.214034 -0.176756 -0.398419 [DiT] Step 47/50 t=0.080 [Debug] dit_step47_vt: [2170, 64] first4: -0.488907 -0.844116 0.205383 0.636597 [Debug] dit_step47_xt: [2170, 64] first4: 0.397078 1.230917 -0.180864 -0.411151 [DiT] Step 48/50 t=0.060 [Debug] dit_step48_vt: [2170, 64] first4: -0.552174 -0.892790 0.256409 0.627930 [Debug] dit_step48_xt: [2170, 64] first4: 0.408122 1.248773 -0.185992 -0.423709 [DiT] Step 49/50 t=0.040 [Debug] dit_step49_vt: [2170, 64] first4: -0.538727 -0.881416 0.191162 0.646179 [Debug] dit_x0: [2170, 64] first4: 0.418896 1.266401 -0.189815 -0.436633 [DiT] Step 50/50 t=0.020 [DiT-Generate] Total: 4140.8 ms (4140.8 ms/sample) [Debug] dit_output: [2170, 64] first4: 0.418896 1.266401 -0.189815 -0.436633 [Store] Unload DiT (3267.8 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: 669 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: 946.6 ms [Debug] vae_audio: [2, 4166400] first4: -0.001417 -0.000703 -0.000822 -0.000342 [Store] Unload VAE-Dec (161.1 MB) [MP3] Encoding 86.8s @ 128 kbps, 48000 Hz stereo [MP3] 1388928 bytes (12.0:1), 409 ms (212.37x realtime), 32 threads [MP3] Wrote ggml-xl-sft/request00.mp3 [Ace-Synth] All done 2026-07-12 10:22:03.688 | 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 0.999400 detok_output 0.999971 context 0.999982 noise 1.000000 temb_t 0.999978 hidden_after_proj_in 0.999972 enc_after_cond_emb 0.999490 layer0_sa_output 0.989082 hidden_after_layer0 0.998997 hidden_after_layer6 0.999660 hidden_after_layer12 0.994961 hidden_after_layer18 0.995318 hidden_after_layer31 0.986970 dit_step0_vt 0.904608 dit_step0_xt 0.999828 dit_step1_vt 0.920345 dit_step1_xt 0.999458 dit_step2_vt 0.914869 dit_step2_xt 0.998525 dit_step3_vt 0.911524 dit_step3_xt 0.996267 dit_step4_vt 0.905879 dit_step4_xt 0.990748 dit_step5_vt 0.896796 dit_step5_xt 0.977685 dit_step6_vt 0.889228 dit_step6_xt 0.952309 dit_step7_vt 0.867227 dit_x0 0.911997 vae_audio 0.705133 vae_audio (STFT cosine) 0.904489 [Xl-turbo] Error growth GGML vs Python stage cos max_err mean_err mean_A std_A mean_B std_B dit_step0_xt 0.999828 0.151645 0.013047 -0.002896 0.978432 -0.002721 0.976534 dit_step1_xt 0.999458 0.305361 0.022692 -0.005845 0.952596 -0.005567 0.949004 dit_step2_xt 0.998525 0.522040 0.035784 -0.009679 0.924049 -0.009330 0.918931 dit_step3_xt 0.996267 0.830202 0.054031 -0.015262 0.893386 -0.014418 0.887241 dit_step4_xt 0.990748 1.335621 0.081541 -0.022903 0.864793 -0.021453 0.857796 dit_step5_xt 0.977685 2.013809 0.124032 -0.034237 0.848495 -0.032240 0.840033 dit_step6_xt 0.952309 2.851309 0.187583 -0.050979 0.872916 -0.049194 0.863964 [Xl-sft] steps=50, shift=1.0 | acestep-v15-xl-sft-Q6_K.gguf [GGML] Running acestep-v15-xl-sft-Q6_K.gguf... [GGML] Done, 131 dump files [Python] Initializing acestep-v15-xl-sft... Loading checkpoint shards: 0%| | 0/4 [00:00 2026-07-12 10:22:07.117 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:123 - ====================================================================== 2026-07-12 10:22:07.117 | 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 10:22:07.117 | INFO | acestep.core.generation.handler.conditioning_text:_prepare_text_conditioning_inputs:125 - ====================================================================== 2026-07-12 10:22:07.120 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:110 - [preprocess_batch] Inferring prompt embeddings... 2026-07-12 10:22:07.128 | INFO | acestep.core.generation.handler.conditioning_embed:preprocess_batch:113 - [preprocess_batch] Inferring lyric embeddings... 2026-07-12 10:22:07.128 | INFO | acestep.core.generation.handler.service_generate_execute:_execute_service_generate_diffusion:161 - [service_generate] Generating audio... (DiT backend: PyTorch (cuda)) 2026-07-12 10:22:07.134 | 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