Hugging Face's logo

Neoscopio-SA
/
Neo_EP

Automatic Speech Recognition
Transformers
Safetensors
Portuguese
whisper
pt_PT
Model card Files
xet
 Community
Neo_EP / training_stats.log
Nerpatime's picture
Nerpatime
Upload Neo_EP (Whisper fine-tuned for European Portuguese)
e0167dc verified
Raw
History Blame Contribute Delete
22.1 kB
+---------------------------------------------------------------------------+
| RELAT脫RIO DE EXECU脟脙O DE TREINO - DEUCALION |
+---------------------------------------------------------------------------+
| Data/Hora : 2026-05-29 16:32:26 |
| Script : train_neox_bracarense |
| Modelo Base : /projects/F202500024INOVIANEOSCOPIO/voicemodels/mo |
| Tempo Total : 0h 59m 54s |
+---------------------------------------------------------------------------+
| CONFIGURA脟脙O (Learning Rate: 5e-06 | Epochs: 1 | Batch: 16 ) |
+---------------------------------------------------------------------------+
| PROGRESS脙O E PERFORMANCE |
+---------------------------------------------------------------------------+
| Loss Inicial : 2.7326 | Loss Final : 1.4853 |
| Throughput : 31.58 spl/s | Audio M茅dio : 30s (Whisper Fixed) |
+---------------------------------------------------------------------------+
| AMBIENTE DE SOFTWARE |
+---------------------------------------------------------------------------+
| Py: 3.11.3 | Torch: 2.10.0+cu128 | Trans: 5.3.0 | CUDA: 12.8 |
+---------------------------------------------------------------------------+
| RECURSOS E I/O (LUSTRE/LOCAL) |
+---------------------------------------------------------------------------+
| Mem贸ria Total : RAM: 21.99GB | GPU Alocada: 11.52GB | Reservada: 18.10GB |
| Disco I/O : Leitura Proc: 194.83 MB | Escrita Proc: 2947.95 MB |
+---------------------------------------------------------------------------+
| SLURM: Nodo: 2 | Rank: 1 | Master: gnx518.deucalion.macc.fccn.pt:42171 |
+---------------------------------------------------------------------------+
| GPUs: UTILIZA脟脙O, ENERGIA E VRAM |
+---------------------------------------------------------------------------+
| GPU 0 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 16.89GB |
| GPU 1 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 2 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 3 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
+---------------------------------------------------------------------------+
| Output Final : /projects/F202500024INOVIANEOSCOPIO/voicemodels/models/hub/n |
+---------------------------------------------------------------------------+
+---------------------------------------------------------------------------+
| RELAT脫RIO DE EXECU脟脙O DE TREINO - DEUCALION |
+---------------------------------------------------------------------------+
| Data/Hora : 2026-05-29 16:32:34 |
| Script : train_neox_bracarense |
| Modelo Base : /projects/F202500024INOVIANEOSCOPIO/voicemodels/mo |
| Tempo Total : 1h 00m 02s |
+---------------------------------------------------------------------------+
| CONFIGURA脟脙O (Learning Rate: 5e-06 | Epochs: 1 | Batch: 16 ) |
+---------------------------------------------------------------------------+
| PROGRESS脙O E PERFORMANCE |
+---------------------------------------------------------------------------+
| Loss Inicial : 2.7326 | Loss Final : 1.4853 |
| Throughput : 31.58 spl/s | Audio M茅dio : 30s (Whisper Fixed) |
+---------------------------------------------------------------------------+
| AMBIENTE DE SOFTWARE |
+---------------------------------------------------------------------------+
| Py: 3.11.3 | Torch: 2.10.0+cu128 | Trans: 5.3.0 | CUDA: 12.8 |
+---------------------------------------------------------------------------+
| RECURSOS E I/O (LUSTRE/LOCAL) |
+---------------------------------------------------------------------------+
| Mem贸ria Total : RAM: 21.75GB | GPU Alocada: 11.52GB | Reservada: 17.94GB |
| Disco I/O : Leitura Proc: 194.83 MB | Escrita Proc: 2947.95 MB |
+---------------------------------------------------------------------------+
| SLURM: Nodo: 2 | Rank: 1 | Master: gnx518.deucalion.macc.fccn.pt:42171 |
+---------------------------------------------------------------------------+
| GPUs: UTILIZA脟脙O, ENERGIA E VRAM |
+---------------------------------------------------------------------------+
| GPU 0 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 1 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 16.87GB |
| GPU 2 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 3 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
+---------------------------------------------------------------------------+
| Output Final : /projects/F202500024INOVIANEOSCOPIO/voicemodels/models/hub/n |
+---------------------------------------------------------------------------+
+---------------------------------------------------------------------------+
| RELAT脫RIO DE EXECU脟脙O DE TREINO - DEUCALION |
+---------------------------------------------------------------------------+
| Data/Hora : 2026-05-29 16:32:35 |
| Script : train_neox_bracarense |
| Modelo Base : /projects/F202500024INOVIANEOSCOPIO/voicemodels/mo |
| Tempo Total : 1h 00m 03s |
+---------------------------------------------------------------------------+
| CONFIGURA脟脙O (Learning Rate: 5e-06 | Epochs: 1 | Batch: 16 ) |
+---------------------------------------------------------------------------+
| PROGRESS脙O E PERFORMANCE |
+---------------------------------------------------------------------------+
| Loss Inicial : 2.7326 | Loss Final : 1.4853 |
| Throughput : 31.58 spl/s | Audio M茅dio : 30s (Whisper Fixed) |
+---------------------------------------------------------------------------+
| AMBIENTE DE SOFTWARE |
+---------------------------------------------------------------------------+
| Py: 3.11.3 | Torch: 2.10.0+cu128 | Trans: 5.3.0 | CUDA: 12.8 |
+---------------------------------------------------------------------------+
| RECURSOS E I/O (LUSTRE/LOCAL) |
+---------------------------------------------------------------------------+
| Mem贸ria Total : RAM: 21.76GB | GPU Alocada: 11.52GB | Reservada: 17.95GB |
| Disco I/O : Leitura Proc: 194.82 MB | Escrita Proc: 2947.95 MB |
+---------------------------------------------------------------------------+
| SLURM: Nodo: 2 | Rank: 0 | Master: gnx518.deucalion.macc.fccn.pt:42171 |
+---------------------------------------------------------------------------+
| GPUs: UTILIZA脟脙O, ENERGIA E VRAM |
+---------------------------------------------------------------------------+
| GPU 0 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 1 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 2 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 3 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 16.89GB |
+---------------------------------------------------------------------------+
| Output Final : /projects/F202500024INOVIANEOSCOPIO/voicemodels/models/hub/n |
+---------------------------------------------------------------------------+
+---------------------------------------------------------------------------+
| RELAT脫RIO DE EXECU脟脙O DE TREINO - DEUCALION |
+---------------------------------------------------------------------------+
| Data/Hora : 2026-05-29 16:32:35 |
| Script : train_neox_bracarense |
| Modelo Base : /projects/F202500024INOVIANEOSCOPIO/voicemodels/mo |
| Tempo Total : 1h 00m 03s |
+---------------------------------------------------------------------------+
| CONFIGURA脟脙O (Learning Rate: 5e-06 | Epochs: 1 | Batch: 16 ) |
+---------------------------------------------------------------------------+
| PROGRESS脙O E PERFORMANCE |
+---------------------------------------------------------------------------+
| Loss Inicial : 2.7326 | Loss Final : 1.4853 |
| Throughput : 31.58 spl/s | Audio M茅dio : 30s (Whisper Fixed) |
+---------------------------------------------------------------------------+
| AMBIENTE DE SOFTWARE |
+---------------------------------------------------------------------------+
| Py: 3.11.3 | Torch: 2.10.0+cu128 | Trans: 5.3.0 | CUDA: 12.8 |
+---------------------------------------------------------------------------+
| RECURSOS E I/O (LUSTRE/LOCAL) |
+---------------------------------------------------------------------------+
| Mem贸ria Total : RAM: 21.76GB | GPU Alocada: 11.52GB | Reservada: 18.00GB |
| Disco I/O : Leitura Proc: 194.82 MB | Escrita Proc: 2947.95 MB |
+---------------------------------------------------------------------------+
| SLURM: Nodo: 2 | Rank: 0 | Master: gnx518.deucalion.macc.fccn.pt:42171 |
+---------------------------------------------------------------------------+
| GPUs: UTILIZA脟脙O, ENERGIA E VRAM |
+---------------------------------------------------------------------------+
| GPU 0 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 1 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 2 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 16.88GB |
| GPU 3 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
+---------------------------------------------------------------------------+
| Output Final : /projects/F202500024INOVIANEOSCOPIO/voicemodels/models/hub/n |
+---------------------------------------------------------------------------+
+---------------------------------------------------------------------------+
| RELAT脫RIO DE EXECU脟脙O DE TREINO - DEUCALION |
+---------------------------------------------------------------------------+
| Data/Hora : 2026-05-29 16:32:43 |
| Script : train_neox_bracarense |
| Modelo Base : /projects/F202500024INOVIANEOSCOPIO/voicemodels/mo |
| Tempo Total : 1h 00m 11s |
+---------------------------------------------------------------------------+
| CONFIGURA脟脙O (Learning Rate: 5e-06 | Epochs: 1 | Batch: 16 ) |
+---------------------------------------------------------------------------+
| PROGRESS脙O E PERFORMANCE |
+---------------------------------------------------------------------------+
| Loss Inicial : 2.7326 | Loss Final : 1.4853 |
| Throughput : 31.50 spl/s | Audio M茅dio : 30s (Whisper Fixed) |
+---------------------------------------------------------------------------+
| AMBIENTE DE SOFTWARE |
+---------------------------------------------------------------------------+
| Py: 3.11.3 | Torch: 2.10.0+cu128 | Trans: 5.3.0 | CUDA: 12.8 |
+---------------------------------------------------------------------------+
| RECURSOS E I/O (LUSTRE/LOCAL) |
+---------------------------------------------------------------------------+
| Mem贸ria Total : RAM: 21.74GB | GPU Alocada: 11.52GB | Reservada: 18.09GB |
| Disco I/O : Leitura Proc: 194.90 MB | Escrita Proc: 20621.82 MB |
+---------------------------------------------------------------------------+
| SLURM: Nodo: 2 | Rank: 0 | Master: gnx518.deucalion.macc.fccn.pt:42171 |
+---------------------------------------------------------------------------+
| GPUs: UTILIZA脟脙O, ENERGIA E VRAM |
+---------------------------------------------------------------------------+
| GPU 0 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 16.89GB |
| GPU 1 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 2 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 3 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
+---------------------------------------------------------------------------+
| Output Final : /projects/F202500024INOVIANEOSCOPIO/voicemodels/models/hub/n |
+---------------------------------------------------------------------------+
+---------------------------------------------------------------------------+
| RELAT脫RIO DE EXECU脟脙O DE TREINO - DEUCALION |
+---------------------------------------------------------------------------+
| Data/Hora : 2026-05-29 16:32:43 |
| Script : train_neox_bracarense |
| Modelo Base : /projects/F202500024INOVIANEOSCOPIO/voicemodels/mo |
| Tempo Total : 1h 00m 11s |
+---------------------------------------------------------------------------+
| CONFIGURA脟脙O (Learning Rate: 5e-06 | Epochs: 1 | Batch: 16 ) |
+---------------------------------------------------------------------------+
| PROGRESS脙O E PERFORMANCE |
+---------------------------------------------------------------------------+
| Loss Inicial : 2.7326 | Loss Final : 1.4853 |
| Throughput : 31.58 spl/s | Audio M茅dio : 30s (Whisper Fixed) |
+---------------------------------------------------------------------------+
| AMBIENTE DE SOFTWARE |
+---------------------------------------------------------------------------+
| Py: 3.11.3 | Torch: 2.10.0+cu128 | Trans: 5.3.0 | CUDA: 12.8 |
+---------------------------------------------------------------------------+
| RECURSOS E I/O (LUSTRE/LOCAL) |
+---------------------------------------------------------------------------+
| Mem贸ria Total : RAM: 21.96GB | GPU Alocada: 11.52GB | Reservada: 17.96GB |
| Disco I/O : Leitura Proc: 194.82 MB | Escrita Proc: 2947.95 MB |
+---------------------------------------------------------------------------+
| SLURM: Nodo: 2 | Rank: 0 | Master: gnx518.deucalion.macc.fccn.pt:42171 |
+---------------------------------------------------------------------------+
| GPUs: UTILIZA脟脙O, ENERGIA E VRAM |
+---------------------------------------------------------------------------+
| GPU 0 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 1 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 16.89GB |
| GPU 2 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 3 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
+---------------------------------------------------------------------------+
| Output Final : /projects/F202500024INOVIANEOSCOPIO/voicemodels/models/hub/n |
+---------------------------------------------------------------------------+
+---------------------------------------------------------------------------+
| RELAT脫RIO DE EXECU脟脙O DE TREINO - DEUCALION |
+---------------------------------------------------------------------------+
| Data/Hora : 2026-05-29 16:32:43 |
| Script : train_neox_bracarense |
| Modelo Base : /projects/F202500024INOVIANEOSCOPIO/voicemodels/mo |
| Tempo Total : 1h 00m 11s |
+---------------------------------------------------------------------------+
| CONFIGURA脟脙O (Learning Rate: 5e-06 | Epochs: 1 | Batch: 16 ) |
+---------------------------------------------------------------------------+
| PROGRESS脙O E PERFORMANCE |
+---------------------------------------------------------------------------+
| Loss Inicial : 2.7326 | Loss Final : 1.4853 |
| Throughput : 31.58 spl/s | Audio M茅dio : 30s (Whisper Fixed) |
+---------------------------------------------------------------------------+
| AMBIENTE DE SOFTWARE |
+---------------------------------------------------------------------------+
| Py: 3.11.3 | Torch: 2.10.0+cu128 | Trans: 5.3.0 | CUDA: 12.8 |
+---------------------------------------------------------------------------+
| RECURSOS E I/O (LUSTRE/LOCAL) |
+---------------------------------------------------------------------------+
| Mem贸ria Total : RAM: 21.78GB | GPU Alocada: 11.52GB | Reservada: 18.17GB |
| Disco I/O : Leitura Proc: 194.83 MB | Escrita Proc: 2947.95 MB |
+---------------------------------------------------------------------------+
| SLURM: Nodo: 2 | Rank: 1 | Master: gnx518.deucalion.macc.fccn.pt:42171 |
+---------------------------------------------------------------------------+
| GPUs: UTILIZA脟脙O, ENERGIA E VRAM |
+---------------------------------------------------------------------------+
| GPU 0 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 1 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 2 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 16.88GB |
| GPU 3 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
+---------------------------------------------------------------------------+
| Output Final : /projects/F202500024INOVIANEOSCOPIO/voicemodels/models/hub/n |
+---------------------------------------------------------------------------+
+---------------------------------------------------------------------------+
| RELAT脫RIO DE EXECU脟脙O DE TREINO - DEUCALION |
+---------------------------------------------------------------------------+
| Data/Hora : 2026-05-29 16:32:43 |
| Script : train_neox_bracarense |
| Modelo Base : /projects/F202500024INOVIANEOSCOPIO/voicemodels/mo |
| Tempo Total : 1h 00m 11s |
+---------------------------------------------------------------------------+
| CONFIGURA脟脙O (Learning Rate: 5e-06 | Epochs: 1 | Batch: 16 ) |
+---------------------------------------------------------------------------+
| PROGRESS脙O E PERFORMANCE |
+---------------------------------------------------------------------------+
| Loss Inicial : 2.7326 | Loss Final : 1.4853 |
| Throughput : 31.58 spl/s | Audio M茅dio : 30s (Whisper Fixed) |
+---------------------------------------------------------------------------+
| AMBIENTE DE SOFTWARE |
+---------------------------------------------------------------------------+
| Py: 3.11.3 | Torch: 2.10.0+cu128 | Trans: 5.3.0 | CUDA: 12.8 |
+---------------------------------------------------------------------------+
| RECURSOS E I/O (LUSTRE/LOCAL) |
+---------------------------------------------------------------------------+
| Mem贸ria Total : RAM: 21.82GB | GPU Alocada: 11.52GB | Reservada: 17.92GB |
| Disco I/O : Leitura Proc: 194.83 MB | Escrita Proc: 2947.95 MB |
+---------------------------------------------------------------------------+
| SLURM: Nodo: 2 | Rank: 1 | Master: gnx518.deucalion.macc.fccn.pt:42171 |
+---------------------------------------------------------------------------+
| GPUs: UTILIZA脟脙O, ENERGIA E VRAM |
+---------------------------------------------------------------------------+
| GPU 0 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 1 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 2 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 0.00GB |
| GPU 3 (NVIDIA A100-SXM): Erro leitura Util | MaxVRAM: 16.89GB |
+---------------------------------------------------------------------------+
| Output Final : /projects/F202500024INOVIANEOSCOPIO/voicemodels/models/hub/n |
+---------------------------------------------------------------------------+