Prompt crashes using the pre-trained models
#4
Closed
opened 
Loading…
Reference in New Issue
There is no content yet.
Delete Branch "%!s(<nil>)"
Deleting a branch is permanent. It CANNOT be undone. Continue?
python -m vall_e "The birch canoe slid on the smooth planks." '001.wav' --out-path="./output.wav" yaml="./data/config.yaml"
Error while opening HDF5 file: ./data/data.h5 [Errno 2] Unable to open file (unable to open file: name = './data/data.h5', errno = 2, error message = 'No such file or directory', flags = 0, o_flags = 0)
[2023-08-22 18:42:32,346] [INFO] [real_accelerator.py:158:get_accelerator] Setting ds_accelerator to cuda (auto detect)
[2023-08-22 18:42:34,005] [INFO] [comm.py:631:init_distributed] cdb=None
[2023-08-22 18:42:34,005] [INFO] [comm.py:662:init_distributed] Initializing TorchBackend in DeepSpeed with backend nccl
ar parameter count: 208996353
nar parameter count: 211091456
[2023-08-22 18:42:37,461] [INFO] [logging.py:96:log_dist] [Rank 0] Using quantized gradients
[2023-08-22 18:42:37,582] [INFO] [logging.py:96:log_dist] [Rank 0] DeepSpeed Flops Profiler Enabled: False
[WARNING] cpu_adam cuda is missing or is incompatible with installed torch, only cpu ops can be compiled!
Using /home/user/.cache/torch_extensions/py310_cu118 as PyTorch extensions root...
Emitting ninja build file /home/user/.cache/torch_extensions/py310_cu118/cpu_adam/build.ninja...
Building extension module cpu_adam...
Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N)
ninja: no work to do.
Loading extension module cpu_adam...
Time to load cpu_adam op: 2.196962356567383 seconds
Adam Optimizer #0 is created with AVX2 arithmetic capability.
Config: alpha=0.000001, betas=(0.900000, 0.999000), weight_decay=0.000000, adam_w=1
[2023-08-22 18:42:41,097] [INFO] [logging.py:96:log_dist] [Rank 0] Using DeepSpeed Optimizer param name adamw as basic optimizer
[2023-08-22 18:42:41,104] [INFO] [logging.py:96:log_dist] [Rank 0] DeepSpeed Basic Optimizer = DeepSpeedCPUAdam
[2023-08-22 18:42:41,104] [INFO] [utils.py:54:is_zero_supported_optimizer] Checking ZeRO support for optimizer=DeepSpeedCPUAdam type=<class 'deepspeed.ops.adam.cpu_adam.DeepSpeedCPUAdam'>
[2023-08-22 18:42:41,104] [INFO] [logging.py:96:log_dist] [Rank 0] Creating torch.bfloat16 ZeRO stage 2 optimizer
[2023-08-22 18:42:41,104] [INFO] [stage_1_and_2.py:146:init] Reduce bucket size 500000000
[2023-08-22 18:42:41,105] [INFO] [stage_1_and_2.py:147:init] Allgather bucket size 500000000
[2023-08-22 18:42:41,105] [INFO] [stage_1_and_2.py:148:init] CPU Offload: True
[2023-08-22 18:42:41,105] [INFO] [stage_1_and_2.py:149:init] Round robin gradient partitioning: True
Rank: 0 partition count [1] and sizes[(208996354, False)]
[2023-08-22 18:42:41,668] [INFO] [utils.py:803:see_memory_usage] Before initializing optimizer states
[2023-08-22 18:42:41,669] [INFO] [utils.py:804:see_memory_usage] MA 0.79 GB Max_MA 0.79 GB CA 0.8 GB Max_CA 1 GB
[2023-08-22 18:42:41,669] [INFO] [utils.py:811:see_memory_usage] CPU Virtual Memory: used = 5.27 GB, percent = 16.8%
[2023-08-22 18:42:42,416] [INFO] [utils.py:803:see_memory_usage] After initializing optimizer states
[2023-08-22 18:42:42,417] [INFO] [utils.py:804:see_memory_usage] MA 0.79 GB Max_MA 0.79 GB CA 0.8 GB Max_CA 1 GB
[2023-08-22 18:42:42,417] [INFO] [utils.py:811:see_memory_usage] CPU Virtual Memory: used = 7.84 GB, percent = 25.1%
[2023-08-22 18:42:42,417] [INFO] [stage_1_and_2.py:520:init] optimizer state initialized
[2023-08-22 18:42:42,508] [INFO] [utils.py:803:see_memory_usage] After initializing ZeRO optimizer
[2023-08-22 18:42:42,509] [INFO] [utils.py:804:see_memory_usage] MA 0.79 GB Max_MA 0.79 GB CA 0.8 GB Max_CA 1 GB
[2023-08-22 18:42:42,509] [INFO] [utils.py:811:see_memory_usage] CPU Virtual Memory: used = 7.84 GB, percent = 25.1%
[2023-08-22 18:42:42,514] [INFO] [logging.py:96:log_dist] [Rank 0] DeepSpeed Final Optimizer = adamw
[2023-08-22 18:42:42,515] [INFO] [logging.py:96:log_dist] [Rank 0] DeepSpeed using client LR scheduler
[2023-08-22 18:42:42,515] [INFO] [logging.py:96:log_dist] [Rank 0] DeepSpeed LR Scheduler = None
[2023-08-22 18:42:42,515] [INFO] [logging.py:96:log_dist] [Rank 0] step=0, skipped=0, lr=[1e-06], mom=[(0.9, 0.999)]
[2023-08-22 18:42:42,516] [INFO] [config.py:960:print] DeepSpeedEngine configuration:
[2023-08-22 18:42:42,516] [INFO] [config.py:964:print] activation_checkpointing_config {
"partition_activations": false,
"contiguous_memory_optimization": false,
"cpu_checkpointing": false,
"number_checkpoints": null,
"synchronize_checkpoint_boundary": false,
"profile": false
}
[2023-08-22 18:42:42,516] [INFO] [config.py:964:print] aio_config ................... {'block_size': 1048576, 'queue_depth': 8, 'thread_count': 1, 'single_submit': False, 'overlap_events': True}
[2023-08-22 18:42:42,516] [INFO] [config.py:964:print] amp_enabled .................. False
[2023-08-22 18:42:42,516] [INFO] [config.py:964:print] amp_params ................... False
[2023-08-22 18:42:42,516] [INFO] [config.py:964:print] autotuning_config ............ {
"enabled": false,
"start_step": null,
"end_step": null,
"metric_path": null,
"arg_mappings": null,
"metric": "throughput",
"model_info": null,
"results_dir": "autotuning_results",
"exps_dir": "autotuning_exps",
"overwrite": true,
"fast": true,
"start_profile_step": 3,
"end_profile_step": 5,
"tuner_type": "gridsearch",
"tuner_early_stopping": 5,
"tuner_num_trials": 50,
"model_info_path": null,
"mp_size": 1,
"max_train_batch_size": null,
"min_train_batch_size": 1,
"max_train_micro_batch_size_per_gpu": 1.024000e+03,
"min_train_micro_batch_size_per_gpu": 1,
"num_tuning_micro_batch_sizes": 3
}
[2023-08-22 18:42:42,516] [INFO] [config.py:964:print] bfloat16_enabled ............. True
[2023-08-22 18:42:42,516] [INFO] [config.py:964:print] checkpoint_parallel_write_pipeline False
[2023-08-22 18:42:42,516] [INFO] [config.py:964:print] checkpoint_tag_validation_enabled True
[2023-08-22 18:42:42,516] [INFO] [config.py:964:print] checkpoint_tag_validation_fail False
[2023-08-22 18:42:42,516] [INFO] [config.py:964:print] comms_config ................. <deepspeed.comm.config.DeepSpeedCommsConfig object at 0x7f47e416f1c0>
[2023-08-22 18:42:42,516] [INFO] [config.py:964:print] communication_data_type ...... None
[2023-08-22 18:42:42,516] [INFO] [config.py:964:print] compression_config ........... {'weight_quantization': {'shared_parameters': {'enabled': True, 'quantizer_kernel': True, 'schedule_offset': 0, 'quantize_groups': 64, 'quantize_verbose': True, 'quantization_type': 'symmetric', 'quantize_weight_in_forward': True, 'rounding': 'nearest', 'fp16_mixed_quantize': False, 'quantize_change_ratio': 1}, 'different_groups': {'wq1': {'params': {'start_bits': 8, 'target_bits': 8, 'quantization_period': 0}, 'modules': ['blocks', 'retnet'], 'related_modules': None}}}, 'activation_quantization': {'shared_parameters': {'enabled': False, 'quantization_type': 'symmetric', 'range_calibration': 'dynamic', 'schedule_offset': 1000}, 'different_groups': {}}, 'sparse_pruning': {'shared_parameters': {'enabled': False, 'method': 'l1', 'schedule_offset': 1000}, 'different_groups': {}}, 'row_pruning': {'shared_parameters': {'enabled': False, 'method': 'l1', 'schedule_offset': 1000}, 'different_groups': {}}, 'head_pruning': {'shared_parameters': {'enabled': False, 'method': 'topk', 'schedule_offset': 1000}, 'different_groups': {}}, 'channel_pruning': {'shared_parameters': {'enabled': False, 'method': 'l1', 'schedule_offset': 1000}, 'different_groups': {}}, 'layer_reduction': {'enabled': False}}
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] curriculum_enabled_legacy .... False
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] curriculum_params_legacy ..... False
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] data_efficiency_config ....... {'enabled': False, 'seed': 1234, 'data_sampling': {'enabled': False, 'num_epochs': 1000, 'num_workers': 0, 'curriculum_learning': {'enabled': False}}, 'data_routing': {'enabled': False, 'random_ltd': {'enabled': False, 'layer_token_lr_schedule': {'enabled': False}}}}
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] data_efficiency_enabled ...... False
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] dataloader_drop_last ......... False
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] disable_allgather ............ False
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] dump_state ................... False
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] dynamic_loss_scale_args ...... None
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] eigenvalue_enabled ........... False
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] eigenvalue_gas_boundary_resolution 1
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] eigenvalue_layer_name ........ bert.encoder.layer
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] eigenvalue_layer_num ......... 0
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] eigenvalue_max_iter .......... 100
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] eigenvalue_stability ......... 1e-06
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] eigenvalue_tol ............... 0.01
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] eigenvalue_verbose ........... False
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] elasticity_enabled ........... False
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] flops_profiler_config ........ {
"enabled": false,
"recompute_fwd_factor": 0.0,
"profile_step": 1,
"module_depth": -1,
"top_modules": 1,
"detailed": true,
"output_file": null
}
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] fp16_auto_cast ............... None
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] fp16_enabled ................. False
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] fp16_master_weights_and_gradients False
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] global_rank .................. 0
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] grad_accum_dtype ............. None
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] gradient_accumulation_steps .. 4
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] gradient_clipping ............ 100
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] gradient_predivide_factor .... 1.0
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] hybrid_engine ................ enabled=False max_out_tokens=512 inference_tp_size=1 release_inference_cache=False pin_parameters=True tp_gather_partition_size=8
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] initial_dynamic_scale ........ 1
[2023-08-22 18:42:42,517] [INFO] [config.py:964:print] load_universal_checkpoint .... False
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] loss_scale ................... 1.0
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] memory_breakdown ............. False
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] mics_hierarchial_params_gather False
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] mics_shard_size .............. -1
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] monitor_config ............... tensorboard=TensorBoardConfig(enabled=False, output_path='', job_name='DeepSpeedJobName') wandb=WandbConfig(enabled=False, group=None, team=None, project='deepspeed') csv_monitor=CSVConfig(enabled=False, output_path='', job_name='DeepSpeedJobName') enabled=False
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] nebula_config ................ {
"enabled": false,
"persistent_storage_path": null,
"persistent_time_interval": 100,
"num_of_version_in_retention": 2,
"enable_nebula_load": true,
"load_path": null
}
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] optimizer_legacy_fusion ...... False
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] optimizer_name ............... adamw
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] optimizer_params ............. {'lr': 1e-06}
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] pipeline ..................... {'stages': 'auto', 'partition': 'best', 'seed_layers': False, 'activation_checkpoint_interval': 0}
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] pld_enabled .................. False
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] pld_params ................... False
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] prescale_gradients ........... False
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] scheduler_name ............... None
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] scheduler_params ............. None
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] sparse_attention ............. None
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] sparse_gradients_enabled ..... False
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] steps_per_print .............. 10
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] train_batch_size ............. 128
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] train_micro_batch_size_per_gpu 32
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] use_node_local_storage ....... False
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] wall_clock_breakdown ......... False
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] world_size ................... 1
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] zero_allow_untested_optimizer False
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] zero_config .................. stage=2 contiguous_gradients=True reduce_scatter=True reduce_bucket_size=500000000 allgather_partitions=True allgather_bucket_size=500000000 overlap_comm=True load_from_fp32_weights=True elastic_checkpoint=False offload_param=DeepSpeedZeroOffloadParamConfig(device='cpu', nvme_path=None, buffer_count=5, buffer_size=100,000,000, max_in_cpu=1,000,000,000, pin_memory=True) offload_optimizer=DeepSpeedZeroOffloadOptimizerConfig(device='cpu', nvme_path=None, buffer_count=4, pin_memory=True, pipeline=False, pipeline_read=False, pipeline_write=False, fast_init=False) sub_group_size=500000000 cpu_offload_param=None cpu_offload_use_pin_memory=None cpu_offload=None prefetch_bucket_size=50,000,000 param_persistence_threshold=100,000 model_persistence_threshold=sys.maxsize max_live_parameters=1,000,000,000 max_reuse_distance=1,000,000,000 gather_16bit_weights_on_model_save=False stage3_gather_fp16_weights_on_model_save=False ignore_unused_parameters=True legacy_stage1=False round_robin_gradients=True zero_hpz_partition_size=1 zero_quantized_weights=True zero_quantized_gradients=True mics_shard_size=-1 mics_hierarchical_params_gather=False memory_efficient_linear=True pipeline_loading_checkpoint=False override_module_apply=True
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] zero_enabled ................. True
[2023-08-22 18:42:42,518] [INFO] [config.py:964:print] zero_force_ds_cpu_optimizer .. True
[2023-08-22 18:42:42,519] [INFO] [config.py:964:print] zero_optimization_stage ...... 2
[2023-08-22 18:42:42,519] [INFO] [config.py:950:print_user_config] json = {
"train_micro_batch_size_per_gpu": 32,
"gradient_accumulation_steps": 4,
"optimizer": {
"type": "AdamW",
"params": {
"lr": 1e-06
}
},
"gradient_clipping": 100,
"bf16": {
"enabled": true
},
"compression_training": {
"weight_quantization": {
"shared_parameters": {
"enabled": true,
"quantizer_kernel": true,
"schedule_offset": 0,
"quantize_groups": 64,
"quantize_verbose": true,
"quantization_type": "symmetric",
"rounding": "nearest",
"quantize_weight_in_forward": true,
"fp16_mixed_quantize": {
"enabled": false,
"quantize_change_ratio": 1
}
},
"different_groups": {
"wq1": {
"params": {
"start_bits": 8,
"target_bits": 8,
"quantization_period": 0
},
"modules": ["blocks", "retnet"]
}
}
},
"activation_quantization": {
"shared_parameters": {
},
"different_groups": {
}
},
"sparse_pruning": {
"shared_parameters": {
},
"different_groups": {
}
},
"row_pruning": {
"shared_parameters": {
},
"different_groups": {
}
},
"head_pruning": {
"shared_parameters": {
},
"different_groups": {
}
},
"channel_pruning": {
"shared_parameters": {
},
"different_groups": {
}
}
},
"zero_optimization": {
"stage": 2,
"contiguous_gradients": true,
"overlap_comm": true,
"reduce_scatter": true,
"reduce_bucket_size": 5.000000e+08,
"allgather_bucket_size": 5.000000e+08,
"sub_group_size": 5.000000e+08,
"round_robin_gradients": true,
"offload_optimizer": {
"device": "cpu",
"pin_memory": true
},
"offload_param": {
"device": "cpu",
"pin_memory": true
},
"zero_quantized_weights": true,
"zero_hpz_partition_size": 1,
"zero_quantized_gradients": true
},
"comms_logger": {
"enabled": false
}
}
[2023-08-22 18:42:42,964] [INFO] [logging.py:96:log_dist] [Rank 0] Using quantized gradients
[2023-08-22 18:42:43,005] [INFO] [logging.py:96:log_dist] [Rank 0] DeepSpeed Flops Profiler Enabled: False
[WARNING] cpu_adam cuda is missing or is incompatible with installed torch, only cpu ops can be compiled!
Using /home/user/.cache/torch_extensions/py310_cu118 as PyTorch extensions root...
No modifications detected for re-loaded extension module cpu_adam, skipping build step...
Loading extension module cpu_adam...
Time to load cpu_adam op: 2.1449809074401855 seconds
Adam Optimizer #1 is created with AVX2 arithmetic capability.
Config: alpha=0.000001, betas=(0.900000, 0.999000), weight_decay=0.000000, adam_w=1
[2023-08-22 18:42:46,223] [INFO] [logging.py:96:log_dist] [Rank 0] Using DeepSpeed Optimizer param name adamw as basic optimizer
[2023-08-22 18:42:46,229] [INFO] [logging.py:96:log_dist] [Rank 0] DeepSpeed Basic Optimizer = DeepSpeedCPUAdam
[2023-08-22 18:42:46,229] [INFO] [utils.py:54:is_zero_supported_optimizer] Checking ZeRO support for optimizer=DeepSpeedCPUAdam type=<class 'deepspeed.ops.adam.cpu_adam.DeepSpeedCPUAdam'>
[2023-08-22 18:42:46,229] [INFO] [logging.py:96:log_dist] [Rank 0] Creating torch.bfloat16 ZeRO stage 2 optimizer
[2023-08-22 18:42:46,229] [INFO] [stage_1_and_2.py:146:init] Reduce bucket size 500000000
[2023-08-22 18:42:46,229] [INFO] [stage_1_and_2.py:147:init] Allgather bucket size 500000000
[2023-08-22 18:42:46,229] [INFO] [stage_1_and_2.py:148:init] CPU Offload: True
[2023-08-22 18:42:46,229] [INFO] [stage_1_and_2.py:149:init] Round robin gradient partitioning: True
Rank: 0 partition count [1] and sizes[(211091456, False)]
[2023-08-22 18:42:46,771] [INFO] [utils.py:803:see_memory_usage] Before initializing optimizer states
[2023-08-22 18:42:46,771] [INFO] [utils.py:804:see_memory_usage] MA 1.19 GB Max_MA 1.19 GB CA 1.2 GB Max_CA 1 GB
[2023-08-22 18:42:46,771] [INFO] [utils.py:811:see_memory_usage] CPU Virtual Memory: used = 8.03 GB, percent = 25.7%
[2023-08-22 18:42:47,539] [INFO] [utils.py:803:see_memory_usage] After initializing optimizer states
[2023-08-22 18:42:47,540] [INFO] [utils.py:804:see_memory_usage] MA 1.19 GB Max_MA 1.19 GB CA 1.2 GB Max_CA 1 GB
[2023-08-22 18:42:47,540] [INFO] [utils.py:811:see_memory_usage] CPU Virtual Memory: used = 10.64 GB, percent = 34.0%
[2023-08-22 18:42:47,540] [INFO] [stage_1_and_2.py:520:init] optimizer state initialized
[2023-08-22 18:42:47,635] [INFO] [utils.py:803:see_memory_usage] After initializing ZeRO optimizer
[2023-08-22 18:42:47,635] [INFO] [utils.py:804:see_memory_usage] MA 1.19 GB Max_MA 1.19 GB CA 1.2 GB Max_CA 1 GB
[2023-08-22 18:42:47,636] [INFO] [utils.py:811:see_memory_usage] CPU Virtual Memory: used = 10.64 GB, percent = 34.0%
[2023-08-22 18:42:47,641] [INFO] [logging.py:96:log_dist] [Rank 0] DeepSpeed Final Optimizer = adamw
[2023-08-22 18:42:47,642] [INFO] [logging.py:96:log_dist] [Rank 0] DeepSpeed using client LR scheduler
[2023-08-22 18:42:47,642] [INFO] [logging.py:96:log_dist] [Rank 0] DeepSpeed LR Scheduler = None
[2023-08-22 18:42:47,642] [INFO] [logging.py:96:log_dist] [Rank 0] step=0, skipped=0, lr=[1e-06], mom=[(0.9, 0.999)]
[2023-08-22 18:42:47,642] [INFO] [config.py:960:print] DeepSpeedEngine configuration:
[2023-08-22 18:42:47,642] [INFO] [config.py:964:print] activation_checkpointing_config {
"partition_activations": false,
"contiguous_memory_optimization": false,
"cpu_checkpointing": false,
"number_checkpoints": null,
"synchronize_checkpoint_boundary": false,
"profile": false
}
[2023-08-22 18:42:47,642] [INFO] [config.py:964:print] aio_config ................... {'block_size': 1048576, 'queue_depth': 8, 'thread_count': 1, 'single_submit': False, 'overlap_events': True}
[2023-08-22 18:42:47,642] [INFO] [config.py:964:print] amp_enabled .................. False
[2023-08-22 18:42:47,642] [INFO] [config.py:964:print] amp_params ................... False
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] autotuning_config ............ {
"enabled": false,
"start_step": null,
"end_step": null,
"metric_path": null,
"arg_mappings": null,
"metric": "throughput",
"model_info": null,
"results_dir": "autotuning_results",
"exps_dir": "autotuning_exps",
"overwrite": true,
"fast": true,
"start_profile_step": 3,
"end_profile_step": 5,
"tuner_type": "gridsearch",
"tuner_early_stopping": 5,
"tuner_num_trials": 50,
"model_info_path": null,
"mp_size": 1,
"max_train_batch_size": null,
"min_train_batch_size": 1,
"max_train_micro_batch_size_per_gpu": 1.024000e+03,
"min_train_micro_batch_size_per_gpu": 1,
"num_tuning_micro_batch_sizes": 3
}
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] bfloat16_enabled ............. True
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] checkpoint_parallel_write_pipeline False
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] checkpoint_tag_validation_enabled True
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] checkpoint_tag_validation_fail False
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] comms_config ................. <deepspeed.comm.config.DeepSpeedCommsConfig object at 0x7f47dd429900>
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] communication_data_type ...... None
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] compression_config ........... {'weight_quantization': {'shared_parameters': {'enabled': True, 'quantizer_kernel': True, 'schedule_offset': 0, 'quantize_groups': 64, 'quantize_verbose': True, 'quantization_type': 'symmetric', 'quantize_weight_in_forward': True, 'rounding': 'nearest', 'fp16_mixed_quantize': False, 'quantize_change_ratio': 1}, 'different_groups': {'wq1': {'params': {'start_bits': 8, 'target_bits': 8, 'quantization_period': 0}, 'modules': ['blocks', 'retnet'], 'related_modules': None}}}, 'activation_quantization': {'shared_parameters': {'enabled': False, 'quantization_type': 'symmetric', 'range_calibration': 'dynamic', 'schedule_offset': 1000}, 'different_groups': {}}, 'sparse_pruning': {'shared_parameters': {'enabled': False, 'method': 'l1', 'schedule_offset': 1000}, 'different_groups': {}}, 'row_pruning': {'shared_parameters': {'enabled': False, 'method': 'l1', 'schedule_offset': 1000}, 'different_groups': {}}, 'head_pruning': {'shared_parameters': {'enabled': False, 'method': 'topk', 'schedule_offset': 1000}, 'different_groups': {}}, 'channel_pruning': {'shared_parameters': {'enabled': False, 'method': 'l1', 'schedule_offset': 1000}, 'different_groups': {}}, 'layer_reduction': {'enabled': False}}
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] curriculum_enabled_legacy .... False
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] curriculum_params_legacy ..... False
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] data_efficiency_config ....... {'enabled': False, 'seed': 1234, 'data_sampling': {'enabled': False, 'num_epochs': 1000, 'num_workers': 0, 'curriculum_learning': {'enabled': False}}, 'data_routing': {'enabled': False, 'random_ltd': {'enabled': False, 'layer_token_lr_schedule': {'enabled': False}}}}
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] data_efficiency_enabled ...... False
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] dataloader_drop_last ......... False
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] disable_allgather ............ False
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] dump_state ................... False
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] dynamic_loss_scale_args ...... None
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] eigenvalue_enabled ........... False
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] eigenvalue_gas_boundary_resolution 1
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] eigenvalue_layer_name ........ bert.encoder.layer
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] eigenvalue_layer_num ......... 0
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] eigenvalue_max_iter .......... 100
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] eigenvalue_stability ......... 1e-06
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] eigenvalue_tol ............... 0.01
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] eigenvalue_verbose ........... False
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] elasticity_enabled ........... False
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] flops_profiler_config ........ {
"enabled": false,
"recompute_fwd_factor": 0.0,
"profile_step": 1,
"module_depth": -1,
"top_modules": 1,
"detailed": true,
"output_file": null
}
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] fp16_auto_cast ............... None
[2023-08-22 18:42:47,643] [INFO] [config.py:964:print] fp16_enabled ................. False
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] fp16_master_weights_and_gradients False
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] global_rank .................. 0
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] grad_accum_dtype ............. None
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] gradient_accumulation_steps .. 4
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] gradient_clipping ............ 100
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] gradient_predivide_factor .... 1.0
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] hybrid_engine ................ enabled=False max_out_tokens=512 inference_tp_size=1 release_inference_cache=False pin_parameters=True tp_gather_partition_size=8
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] initial_dynamic_scale ........ 1
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] load_universal_checkpoint .... False
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] loss_scale ................... 1.0
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] memory_breakdown ............. False
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] mics_hierarchial_params_gather False
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] mics_shard_size .............. -1
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] monitor_config ............... tensorboard=TensorBoardConfig(enabled=False, output_path='', job_name='DeepSpeedJobName') wandb=WandbConfig(enabled=False, group=None, team=None, project='deepspeed') csv_monitor=CSVConfig(enabled=False, output_path='', job_name='DeepSpeedJobName') enabled=False
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] nebula_config ................ {
"enabled": false,
"persistent_storage_path": null,
"persistent_time_interval": 100,
"num_of_version_in_retention": 2,
"enable_nebula_load": true,
"load_path": null
}
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] optimizer_legacy_fusion ...... False
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] optimizer_name ............... adamw
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] optimizer_params ............. {'lr': 1e-06}
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] pipeline ..................... {'stages': 'auto', 'partition': 'best', 'seed_layers': False, 'activation_checkpoint_interval': 0}
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] pld_enabled .................. False
[2023-08-22 18:42:47,644] [INFO] [config.py:964:print] pld_params ................... False
[2023-08-22 18:42:47,645] [INFO] [config.py:964:print] prescale_gradients ........... False
[2023-08-22 18:42:47,645] [INFO] [config.py:964:print] scheduler_name ............... None
[2023-08-22 18:42:47,645] [INFO] [config.py:964:print] scheduler_params ............. None
[2023-08-22 18:42:47,645] [INFO] [config.py:964:print] sparse_attention ............. None
[2023-08-22 18:42:47,645] [INFO] [config.py:964:print] sparse_gradients_enabled ..... False
[2023-08-22 18:42:47,645] [INFO] [config.py:964:print] steps_per_print .............. 10
[2023-08-22 18:42:47,645] [INFO] [config.py:964:print] train_batch_size ............. 128
[2023-08-22 18:42:47,645] [INFO] [config.py:964:print] train_micro_batch_size_per_gpu 32
[2023-08-22 18:42:47,645] [INFO] [config.py:964:print] use_node_local_storage ....... False
[2023-08-22 18:42:47,645] [INFO] [config.py:964:print] wall_clock_breakdown ......... False
[2023-08-22 18:42:47,645] [INFO] [config.py:964:print] world_size ................... 1
[2023-08-22 18:42:47,645] [INFO] [config.py:964:print] zero_allow_untested_optimizer False
[2023-08-22 18:42:47,645] [INFO] [config.py:964:print] zero_config .................. stage=2 contiguous_gradients=True reduce_scatter=True reduce_bucket_size=500000000 allgather_partitions=True allgather_bucket_size=500000000 overlap_comm=True load_from_fp32_weights=True elastic_checkpoint=False offload_param=DeepSpeedZeroOffloadParamConfig(device='cpu', nvme_path=None, buffer_count=5, buffer_size=100,000,000, max_in_cpu=1,000,000,000, pin_memory=True) offload_optimizer=DeepSpeedZeroOffloadOptimizerConfig(device='cpu', nvme_path=None, buffer_count=4, pin_memory=True, pipeline=False, pipeline_read=False, pipeline_write=False, fast_init=False) sub_group_size=500000000 cpu_offload_param=None cpu_offload_use_pin_memory=None cpu_offload=None prefetch_bucket_size=50,000,000 param_persistence_threshold=100,000 model_persistence_threshold=sys.maxsize max_live_parameters=1,000,000,000 max_reuse_distance=1,000,000,000 gather_16bit_weights_on_model_save=False stage3_gather_fp16_weights_on_model_save=False ignore_unused_parameters=True legacy_stage1=False round_robin_gradients=True zero_hpz_partition_size=1 zero_quantized_weights=True zero_quantized_gradients=True mics_shard_size=-1 mics_hierarchical_params_gather=False memory_efficient_linear=True pipeline_loading_checkpoint=False override_module_apply=True
[2023-08-22 18:42:47,645] [INFO] [config.py:964:print] zero_enabled ................. True
[2023-08-22 18:42:47,645] [INFO] [config.py:964:print] zero_force_ds_cpu_optimizer .. True
[2023-08-22 18:42:47,645] [INFO] [config.py:964:print] zero_optimization_stage ...... 2
[2023-08-22 18:42:47,645] [INFO] [config.py:950:print_user_config] json = {
"train_micro_batch_size_per_gpu": 32,
"gradient_accumulation_steps": 4,
"optimizer": {
"type": "AdamW",
"params": {
"lr": 1e-06
}
},
"gradient_clipping": 100,
"bf16": {
"enabled": true
},
"compression_training": {
"weight_quantization": {
"shared_parameters": {
"enabled": true,
"quantizer_kernel": true,
"schedule_offset": 0,
"quantize_groups": 64,
"quantize_verbose": true,
"quantization_type": "symmetric",
"rounding": "nearest",
"quantize_weight_in_forward": true,
"fp16_mixed_quantize": {
"enabled": false,
"quantize_change_ratio": 1
}
},
"different_groups": {
"wq1": {
"params": {
"start_bits": 8,
"target_bits": 8,
"quantization_period": 0
},
"modules": ["blocks", "retnet"]
}
}
},
"activation_quantization": {
"shared_parameters": {
},
"different_groups": {
}
},
"sparse_pruning": {
"shared_parameters": {
},
"different_groups": {
}
},
"row_pruning": {
"shared_parameters": {
},
"different_groups": {
}
},
"head_pruning": {
"shared_parameters": {
},
"different_groups": {
}
},
"channel_pruning": {
"shared_parameters": {
},
"different_groups": {
}
}
},
"zero_optimization": {
"stage": 2,
"contiguous_gradients": true,
"overlap_comm": true,
"reduce_scatter": true,
"reduce_bucket_size": 5.000000e+08,
"allgather_bucket_size": 5.000000e+08,
"sub_group_size": 5.000000e+08,
"round_robin_gradients": true,
"offload_optimizer": {
"device": "cpu",
"pin_memory": true
},
"offload_param": {
"device": "cpu",
"pin_memory": true
},
"zero_quantized_weights": true,
"zero_hpz_partition_size": 1,
"zero_quantized_gradients": true
},
"comms_logger": {
"enabled": false
}
}
0%| | 0/450 [00:00<?, ?it/s]
Traceback (most recent call last):
File "/usr/lib/python3.10/runpy.py", line 196, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/usr/lib/python3.10/runpy.py", line 86, in _run_code
exec(code, run_globals)
File "/home/user/ai-voice-cloning python3 latest version/modules/vall-e/vall_e/main.py", line 27, in
main()
File "/home/user/ai-voice-cloning python3 latest version/modules/vall-e/vall_e/main.py", line 24, in main
tts.inference( text=args.text, references=args.references, out_path=args.out_path, max_ar_steps=args.max_ar_steps, ar_temp=args.ar_temp, nar_temp=args.nar_temp )
File "/home/user/ai-voice-cloning/venv/lib/python3.10/site-packages/torch/utils/_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
File "/home/user/ai-voice-cloning python3 latest version/modules/vall-e/vall_e/inference.py", line 107, in inference
resps_list = self.ar(text_list=[phns], proms_list=[prom], max_steps=max_ar_steps, sampling_temperature=ar_temp)
File "/home/user/ai-voice-cloning/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "/home/user/ai-voice-cloning python3 latest version/modules/vall-e/vall_e/models/ar.py", line 94, in forward
r, state = super().forward(
File "/home/user/ai-voice-cloning python3 latest version/modules/vall-e/vall_e/models/base.py", line 292, in forward
x, _ = self.retnet(x, incremental_state=state, token_embeddings=x, features_only=True)
File "/home/user/ai-voice-cloning/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "/home/user/ai-voice-cloning python3 latest version/modules/vall-e/vall_e/models/retnet.py", line 59, in forward
return super().forward(src_tokens, **kwargs)
File "/home/user/ai-voice-cloning/venv/lib/python3.10/site-packages/torchscale/architecture/retnet.py", line 362, in forward
retention_rel_pos = self.retnet_rel_pos(slen, incremental_state is not None and not is_first_step, chunkwise_recurrent=self.chunkwise_recurrent)
File "/home/user/ai-voice-cloning/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "/home/user/ai-voice-cloning/venv/lib/python3.10/site-packages/torchscale/architecture/retnet.py", line 45, in forward
mask = torch.tril(torch.ones(self.recurrent_chunk_size, self.recurrent_chunk_size).to(self.decay))
RuntimeError: "triu_tril_cuda_template" not implemented for 'BFloat16'
Try updating the repo to the latest version. I neglected to re-include the code to convert the weights to whatever is specified in the
config.yamlunder cfg.inference.weight_dtype.