Adding regression benchmark for the transformers SHA 946bac798caefada3f5f1c9fecdcfd587ed24ac7

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/bert_1gpu_training/0/.config/config.yaml

@@ -0,0 +1,75 @@
+backend:
+  name: pytorch
+  version: ${pytorch_version:}
+  _target_: optimum_benchmark.backends.pytorch.backend.PyTorchBackend
+  seed: 42
+  inter_op_num_threads: null
+  intra_op_num_threads: null
+  initial_isolation_check: true
+  continous_isolation_check: true
+  delete_cache: false
+  no_weights: false
+  device_map: null
+  torch_dtype: float16
+  disable_grad: ${is_inference:${benchmark.name}}
+  eval_mode: ${is_inference:${benchmark.name}}
+  amp_autocast: false
+  amp_dtype: null
+  torch_compile: false
+  torch_compile_config: {}
+  bettertransformer: false
+  quantization_scheme: null
+  quantization_config: {}
+  use_ddp: false
+  ddp_config: {}
+  peft_strategy: null
+  peft_config: {}
+benchmark:
+  name: training
+  _target_: optimum_benchmark.benchmarks.training.benchmark.TrainingBenchmark
+  warmup_steps: 40
+  dataset_shapes:
+    dataset_size: 1500
+    sequence_length: 256
+    num_choices: 1
+    feature_size: 80
+    nb_max_frames: 3000
+    audio_sequence_length: 16000
+  training_arguments:
+    skip_memory_metrics: true
+    output_dir: ./trainer_output
+    use_cpu: ${is_cpu:${device}}
+    ddp_find_unused_parameters: false
+    do_train: true
+    do_eval: false
+    do_predict: false
+    report_to: none
+    per_device_train_batch_size: 32
+experiment_name: bert_1gpu_training
+model: bert-base-uncased
+device: cuda
+task: text-classification
+hub_kwargs:
+  revision: main
+  cache_dir: null
+  force_download: false
+  local_files_only: false
+environment:
+  optimum_version: 1.13.1
+  transformers_version: 4.34.0.dev0
+  accelerate_version: 0.23.0
+  diffusers_version: null
+  python_version: 3.10.12
+  system: Linux
+  cpu: ' AMD EPYC 7643 48-Core Processor'
+  cpu_count: 96
+  cpu_ram_mb: 1082028
+  gpus:
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/bert_1gpu_training/0/.config/hydra.yaml

@@ -0,0 +1,174 @@
+hydra:
+  run:
+    dir: runs/${oc.env:COMMIT_DATE_GMT}_${oc.env:COMMIT_SHA}/${experiment_name}
+  sweep:
+    dir: sweeps/${oc.env:COMMIT_DATE_GMT}_${oc.env:COMMIT_SHA}/${experiment_name}
+    subdir: ${hydra.job.num}
+  launcher:
+    _target_: hydra._internal.core_plugins.basic_launcher.BasicLauncher
+  sweeper:
+    _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
+    max_batch_size: null
+    params:
+      +benchmark.training_arguments.per_device_train_batch_size: '32'
+      backend.torch_dtype: float16,float32
+  help:
+    app_name: ${hydra.job.name}
+    header: '${hydra.help.app_name} is powered by Hydra.
+
+      '
+    footer: 'Powered by Hydra (https://hydra.cc)
+
+      Use --hydra-help to view Hydra specific help
+
+      '
+    template: '${hydra.help.header}
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (group=option)
+
+
+      $APP_CONFIG_GROUPS
+
+
+      == Config ==
+
+      Override anything in the config (foo.bar=value)
+
+
+      $CONFIG
+
+
+      ${hydra.help.footer}
+
+      '
+  hydra_help:
+    template: 'Hydra (${hydra.runtime.version})
+
+      See https://hydra.cc for more info.
+
+
+      == Flags ==
+
+      $FLAGS_HELP
+
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (For example, append hydra/job_logging=disabled
+      to command line)
+
+
+      $HYDRA_CONFIG_GROUPS
+
+
+      Use ''--cfg hydra'' to Show the Hydra config.
+
+      '
+    hydra_help: ???
+  hydra_logging:
+    version: 1
+    formatters:
+      colorlog:
+        (): colorlog.ColoredFormatter
+        format: '[%(cyan)s%(asctime)s%(reset)s][%(purple)sHYDRA%(reset)s] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: colorlog
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+      colorlog:
+        (): colorlog.ColoredFormatter
+        format: '[%(cyan)s%(asctime)s%(reset)s][%(blue)s%(name)s%(reset)s][%(log_color)s%(levelname)s%(reset)s]
+          - %(message)s'
+        log_colors:
+          DEBUG: purple
+          INFO: green
+          WARNING: yellow
+          ERROR: red
+          CRITICAL: red
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: colorlog
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: MULTIRUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=MULTIRUN
+    task:
+    - +benchmark.training_arguments.per_device_train_batch_size=32
+    - backend.torch_dtype=float16
+  job:
+    name: experiment
+    chdir: true
+    override_dirname: +benchmark.training_arguments.per_device_train_batch_size=32,backend.torch_dtype=float16
+    id: '0'
+    num: 0
+    config_name: bert_1gpu_training
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.3.2
+    version_base: '1.3'
+    cwd: /home/user/transformers-regression
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: optimum_benchmark
+      schema: pkg
+      provider: main
+    - path: hydra_plugins.hydra_colorlog.conf
+      schema: pkg
+      provider: hydra-colorlog
+    - path: /home/user/transformers-regression/configs
+      schema: file
+      provider: command-line
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: /home/user/transformers-regression/sweeps/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/bert_1gpu_training/0
+    choices:
+      benchmark: training
+      backend: pytorch
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: colorlog
+      hydra/hydra_logging: colorlog
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/bert_1gpu_training/0/.config/overrides.yaml

@@ -0,0 +1,2 @@
+- +benchmark.training_arguments.per_device_train_batch_size=32
+- backend.torch_dtype=float16

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/bert_1gpu_training/0/experiment.log

@@ -0,0 +1,17 @@
+[2023-09-27 12:00:57,716][experiment][WARNING] - Multiple GPUs detected but CUDA_DEVICE_ORDER is not set. This means that code might allocate resources from the wrong GPUs even if CUDA_VISIBLE_DEVICES is set. Pytorch uses the `FASTEST_FIRST` order by default, which is not guaranteed to be the same as nvidia-smi. `CUDA_DEVICE_ORDER` will be set to `PCI_BUS_ID` to ensure that the GPUs are allocated in the same order as nvidia-smi. 
+[2023-09-27 12:01:00,376][pytorch][INFO] - 	+ Inferred AutoModel class AutoModelForSequenceClassification for task text-classification and model_type bert
+[2023-09-27 12:01:00,376][backend][INFO] - Configuring pytorch backend
+[2023-09-27 12:01:00,379][backend][INFO] - 	+ Checking initial device(s) isolation of CUDA device(s): [0]
+[2023-09-27 12:01:00,505][backend][INFO] - 	+ Checking continuous device(s) isolation of CUDA device(s): [0]
+[2023-09-27 12:01:00,518][pytorch][INFO] - 	+ Loading model on device: cuda
+[2023-09-27 12:01:01,286][benchmark][INFO] - Configuring training benchmark
+[2023-09-27 12:01:01,287][training][INFO] - Running training benchmark
+[2023-09-27 12:01:01,287][dataset_generator][INFO] - Using text-classification task generator
+[2023-09-27 12:01:01,335][pytorch][INFO] - 	+ Setting dataset format to `torch`.
+[2023-09-27 12:01:01,336][pytorch][INFO] - 	+ Wrapping training arguments with transformers.TrainingArguments
+[2023-09-27 12:01:01,337][pytorch][INFO] - 	+ Wrapping model with transformers.Trainer
+[2023-09-27 12:01:01,341][pytorch][INFO] - 	+ Starting training
+[2023-09-27 12:01:21,019][pytorch][INFO] - 	+ Training finished successfully
+[2023-09-27 12:01:21,019][training][INFO] - Saving training results
+[2023-09-27 12:01:21,023][backend][INFO] - Cleaning pytorch backend
+[2023-09-27 12:01:21,023][backend][INFO] - 	+ Deleting pretrained model

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/bert_1gpu_training/0/hydra_config.yaml

@@ -0,0 +1,75 @@
+backend:
+  name: pytorch
+  version: 2.1.0+rocm5.6
+  _target_: optimum_benchmark.backends.pytorch.backend.PyTorchBackend
+  seed: 42
+  inter_op_num_threads: null
+  intra_op_num_threads: null
+  initial_isolation_check: true
+  continous_isolation_check: true
+  delete_cache: false
+  no_weights: false
+  device_map: null
+  torch_dtype: float16
+  disable_grad: false
+  eval_mode: false
+  amp_autocast: false
+  amp_dtype: null
+  torch_compile: false
+  torch_compile_config: {}
+  bettertransformer: false
+  quantization_scheme: null
+  quantization_config: {}
+  use_ddp: false
+  ddp_config: {}
+  peft_strategy: null
+  peft_config: {}
+benchmark:
+  name: training
+  _target_: optimum_benchmark.benchmarks.training.benchmark.TrainingBenchmark
+  warmup_steps: 40
+  dataset_shapes:
+    dataset_size: 1500
+    sequence_length: 256
+    num_choices: 1
+    feature_size: 80
+    nb_max_frames: 3000
+    audio_sequence_length: 16000
+  training_arguments:
+    skip_memory_metrics: true
+    output_dir: ./trainer_output
+    use_cpu: false
+    ddp_find_unused_parameters: false
+    do_train: true
+    do_eval: false
+    do_predict: false
+    report_to: none
+    per_device_train_batch_size: 32
+experiment_name: bert_1gpu_training
+model: bert-base-uncased
+device: cuda
+task: text-classification
+hub_kwargs:
+  revision: main
+  cache_dir: null
+  force_download: false
+  local_files_only: false
+environment:
+  optimum_version: 1.13.1
+  transformers_version: 4.34.0.dev0
+  accelerate_version: 0.23.0
+  diffusers_version: null
+  python_version: 3.10.12
+  system: Linux
+  cpu: ' AMD EPYC 7643 48-Core Processor'
+  cpu_count: 96
+  cpu_ram_mb: 1082028
+  gpus:
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/bert_1gpu_training/0/training_results.csv

@@ -0,0 +1,2 @@
+warmup.runtime(s),warmup.throughput(samples/s),training.runtime(s),training.throughput(samples/s),overall_training.runtime(s),overall_training.throughput(samples/s)
+6.093268394470215,210.06788428384843,13.473761320114136,239.87362720869555,19.567030906677246,165.17580083634869

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/bert_1gpu_training/1/.config/config.yaml

@@ -0,0 +1,75 @@
+backend:
+  name: pytorch
+  version: ${pytorch_version:}
+  _target_: optimum_benchmark.backends.pytorch.backend.PyTorchBackend
+  seed: 42
+  inter_op_num_threads: null
+  intra_op_num_threads: null
+  initial_isolation_check: true
+  continous_isolation_check: true
+  delete_cache: false
+  no_weights: false
+  device_map: null
+  torch_dtype: float32
+  disable_grad: ${is_inference:${benchmark.name}}
+  eval_mode: ${is_inference:${benchmark.name}}
+  amp_autocast: false
+  amp_dtype: null
+  torch_compile: false
+  torch_compile_config: {}
+  bettertransformer: false
+  quantization_scheme: null
+  quantization_config: {}
+  use_ddp: false
+  ddp_config: {}
+  peft_strategy: null
+  peft_config: {}
+benchmark:
+  name: training
+  _target_: optimum_benchmark.benchmarks.training.benchmark.TrainingBenchmark
+  warmup_steps: 40
+  dataset_shapes:
+    dataset_size: 1500
+    sequence_length: 256
+    num_choices: 1
+    feature_size: 80
+    nb_max_frames: 3000
+    audio_sequence_length: 16000
+  training_arguments:
+    skip_memory_metrics: true
+    output_dir: ./trainer_output
+    use_cpu: ${is_cpu:${device}}
+    ddp_find_unused_parameters: false
+    do_train: true
+    do_eval: false
+    do_predict: false
+    report_to: none
+    per_device_train_batch_size: 32
+experiment_name: bert_1gpu_training
+model: bert-base-uncased
+device: cuda
+task: text-classification
+hub_kwargs:
+  revision: main
+  cache_dir: null
+  force_download: false
+  local_files_only: false
+environment:
+  optimum_version: 1.13.1
+  transformers_version: 4.34.0.dev0
+  accelerate_version: 0.23.0
+  diffusers_version: null
+  python_version: 3.10.12
+  system: Linux
+  cpu: ' AMD EPYC 7643 48-Core Processor'
+  cpu_count: 96
+  cpu_ram_mb: 1082028
+  gpus:
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/bert_1gpu_training/1/.config/hydra.yaml

@@ -0,0 +1,174 @@
+hydra:
+  run:
+    dir: runs/${oc.env:COMMIT_DATE_GMT}_${oc.env:COMMIT_SHA}/${experiment_name}
+  sweep:
+    dir: sweeps/${oc.env:COMMIT_DATE_GMT}_${oc.env:COMMIT_SHA}/${experiment_name}
+    subdir: ${hydra.job.num}
+  launcher:
+    _target_: hydra._internal.core_plugins.basic_launcher.BasicLauncher
+  sweeper:
+    _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
+    max_batch_size: null
+    params:
+      +benchmark.training_arguments.per_device_train_batch_size: '32'
+      backend.torch_dtype: float16,float32
+  help:
+    app_name: ${hydra.job.name}
+    header: '${hydra.help.app_name} is powered by Hydra.
+
+      '
+    footer: 'Powered by Hydra (https://hydra.cc)
+
+      Use --hydra-help to view Hydra specific help
+
+      '
+    template: '${hydra.help.header}
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (group=option)
+
+
+      $APP_CONFIG_GROUPS
+
+
+      == Config ==
+
+      Override anything in the config (foo.bar=value)
+
+
+      $CONFIG
+
+
+      ${hydra.help.footer}
+
+      '
+  hydra_help:
+    template: 'Hydra (${hydra.runtime.version})
+
+      See https://hydra.cc for more info.
+
+
+      == Flags ==
+
+      $FLAGS_HELP
+
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (For example, append hydra/job_logging=disabled
+      to command line)
+
+
+      $HYDRA_CONFIG_GROUPS
+
+
+      Use ''--cfg hydra'' to Show the Hydra config.
+
+      '
+    hydra_help: ???
+  hydra_logging:
+    version: 1
+    formatters:
+      colorlog:
+        (): colorlog.ColoredFormatter
+        format: '[%(cyan)s%(asctime)s%(reset)s][%(purple)sHYDRA%(reset)s] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: colorlog
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+      colorlog:
+        (): colorlog.ColoredFormatter
+        format: '[%(cyan)s%(asctime)s%(reset)s][%(blue)s%(name)s%(reset)s][%(log_color)s%(levelname)s%(reset)s]
+          - %(message)s'
+        log_colors:
+          DEBUG: purple
+          INFO: green
+          WARNING: yellow
+          ERROR: red
+          CRITICAL: red
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: colorlog
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: MULTIRUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=MULTIRUN
+    task:
+    - +benchmark.training_arguments.per_device_train_batch_size=32
+    - backend.torch_dtype=float32
+  job:
+    name: experiment
+    chdir: true
+    override_dirname: +benchmark.training_arguments.per_device_train_batch_size=32,backend.torch_dtype=float32
+    id: '1'
+    num: 1
+    config_name: bert_1gpu_training
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.3.2
+    version_base: '1.3'
+    cwd: /home/user/transformers-regression
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: optimum_benchmark
+      schema: pkg
+      provider: main
+    - path: hydra_plugins.hydra_colorlog.conf
+      schema: pkg
+      provider: hydra-colorlog
+    - path: /home/user/transformers-regression/configs
+      schema: file
+      provider: command-line
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: /home/user/transformers-regression/sweeps/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/bert_1gpu_training/1
+    choices:
+      benchmark: training
+      backend: pytorch
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: colorlog
+      hydra/hydra_logging: colorlog
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/bert_1gpu_training/1/.config/overrides.yaml

@@ -0,0 +1,2 @@
+- +benchmark.training_arguments.per_device_train_batch_size=32
+- backend.torch_dtype=float32

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/bert_1gpu_training/1/experiment.log

@@ -0,0 +1,16 @@
+[2023-09-27 12:01:22,703][pytorch][INFO] - 	+ Inferred AutoModel class AutoModelForSequenceClassification for task text-classification and model_type bert
+[2023-09-27 12:01:22,703][backend][INFO] - Configuring pytorch backend
+[2023-09-27 12:01:22,703][backend][INFO] - 	+ Checking initial device(s) isolation of CUDA device(s): [0]
+[2023-09-27 12:01:22,827][backend][INFO] - 	+ Checking continuous device(s) isolation of CUDA device(s): [0]
+[2023-09-27 12:01:22,842][pytorch][INFO] - 	+ Loading model on device: cuda
+[2023-09-27 12:01:23,406][benchmark][INFO] - Configuring training benchmark
+[2023-09-27 12:01:23,407][training][INFO] - Running training benchmark
+[2023-09-27 12:01:23,407][dataset_generator][INFO] - Using text-classification task generator
+[2023-09-27 12:01:23,437][pytorch][INFO] - 	+ Setting dataset format to `torch`.
+[2023-09-27 12:01:23,437][pytorch][INFO] - 	+ Wrapping training arguments with transformers.TrainingArguments
+[2023-09-27 12:01:23,438][pytorch][INFO] - 	+ Wrapping model with transformers.Trainer
+[2023-09-27 12:01:23,443][pytorch][INFO] - 	+ Starting training
+[2023-09-27 12:02:09,561][pytorch][INFO] - 	+ Training finished successfully
+[2023-09-27 12:02:09,561][training][INFO] - Saving training results
+[2023-09-27 12:02:09,563][backend][INFO] - Cleaning pytorch backend
+[2023-09-27 12:02:09,563][backend][INFO] - 	+ Deleting pretrained model

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/bert_1gpu_training/1/hydra_config.yaml

@@ -0,0 +1,75 @@
+backend:
+  name: pytorch
+  version: 2.1.0+rocm5.6
+  _target_: optimum_benchmark.backends.pytorch.backend.PyTorchBackend
+  seed: 42
+  inter_op_num_threads: null
+  intra_op_num_threads: null
+  initial_isolation_check: true
+  continous_isolation_check: true
+  delete_cache: false
+  no_weights: false
+  device_map: null
+  torch_dtype: float32
+  disable_grad: false
+  eval_mode: false
+  amp_autocast: false
+  amp_dtype: null
+  torch_compile: false
+  torch_compile_config: {}
+  bettertransformer: false
+  quantization_scheme: null
+  quantization_config: {}
+  use_ddp: false
+  ddp_config: {}
+  peft_strategy: null
+  peft_config: {}
+benchmark:
+  name: training
+  _target_: optimum_benchmark.benchmarks.training.benchmark.TrainingBenchmark
+  warmup_steps: 40
+  dataset_shapes:
+    dataset_size: 1500
+    sequence_length: 256
+    num_choices: 1
+    feature_size: 80
+    nb_max_frames: 3000
+    audio_sequence_length: 16000
+  training_arguments:
+    skip_memory_metrics: true
+    output_dir: ./trainer_output
+    use_cpu: false
+    ddp_find_unused_parameters: false
+    do_train: true
+    do_eval: false
+    do_predict: false
+    report_to: none
+    per_device_train_batch_size: 32
+experiment_name: bert_1gpu_training
+model: bert-base-uncased
+device: cuda
+task: text-classification
+hub_kwargs:
+  revision: main
+  cache_dir: null
+  force_download: false
+  local_files_only: false
+environment:
+  optimum_version: 1.13.1
+  transformers_version: 4.34.0.dev0
+  accelerate_version: 0.23.0
+  diffusers_version: null
+  python_version: 3.10.12
+  system: Linux
+  cpu: ' AMD EPYC 7643 48-Core Processor'
+  cpu_count: 96
+  cpu_ram_mb: 1082028
+  gpus:
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/bert_1gpu_training/1/training_results.csv

@@ -0,0 +1,2 @@
+warmup.runtime(s),warmup.throughput(samples/s),training.runtime(s),training.throughput(samples/s),overall_training.runtime(s),overall_training.throughput(samples/s)
+13.32169246673584,96.08388747872313,32.68557357788086,98.88154455356337,46.00726842880249,70.24976944679096

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/bert_1gpu_training/multirun.yaml

@@ -0,0 +1,246 @@
+hydra:
+  run:
+    dir: runs/${oc.env:COMMIT_DATE_GMT}_${oc.env:COMMIT_SHA}/${experiment_name}
+  sweep:
+    dir: sweeps/${oc.env:COMMIT_DATE_GMT}_${oc.env:COMMIT_SHA}/${experiment_name}
+    subdir: ${hydra.job.num}
+  launcher:
+    _target_: hydra._internal.core_plugins.basic_launcher.BasicLauncher
+  sweeper:
+    _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
+    max_batch_size: null
+    params:
+      +benchmark.training_arguments.per_device_train_batch_size: '32'
+      backend.torch_dtype: float16,float32
+  help:
+    app_name: ${hydra.job.name}
+    header: '${hydra.help.app_name} is powered by Hydra.
+
+      '
+    footer: 'Powered by Hydra (https://hydra.cc)
+
+      Use --hydra-help to view Hydra specific help
+
+      '
+    template: '${hydra.help.header}
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (group=option)
+
+
+      $APP_CONFIG_GROUPS
+
+
+      == Config ==
+
+      Override anything in the config (foo.bar=value)
+
+
+      $CONFIG
+
+
+      ${hydra.help.footer}
+
+      '
+  hydra_help:
+    template: 'Hydra (${hydra.runtime.version})
+
+      See https://hydra.cc for more info.
+
+
+      == Flags ==
+
+      $FLAGS_HELP
+
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (For example, append hydra/job_logging=disabled
+      to command line)
+
+
+      $HYDRA_CONFIG_GROUPS
+
+
+      Use ''--cfg hydra'' to Show the Hydra config.
+
+      '
+    hydra_help: ???
+  hydra_logging:
+    version: 1
+    formatters:
+      colorlog:
+        (): colorlog.ColoredFormatter
+        format: '[%(cyan)s%(asctime)s%(reset)s][%(purple)sHYDRA%(reset)s] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: colorlog
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+      colorlog:
+        (): colorlog.ColoredFormatter
+        format: '[%(cyan)s%(asctime)s%(reset)s][%(blue)s%(name)s%(reset)s][%(log_color)s%(levelname)s%(reset)s]
+          - %(message)s'
+        log_colors:
+          DEBUG: purple
+          INFO: green
+          WARNING: yellow
+          ERROR: red
+          CRITICAL: red
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: colorlog
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: MULTIRUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=MULTIRUN
+    task: []
+  job:
+    name: experiment
+    chdir: true
+    override_dirname: ''
+    id: ???
+    num: ???
+    config_name: bert_1gpu_training
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.3.2
+    version_base: '1.3'
+    cwd: /home/user/transformers-regression
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: optimum_benchmark
+      schema: pkg
+      provider: main
+    - path: hydra_plugins.hydra_colorlog.conf
+      schema: pkg
+      provider: hydra-colorlog
+    - path: /home/user/transformers-regression/configs
+      schema: file
+      provider: command-line
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: ???
+    choices:
+      benchmark: training
+      backend: pytorch
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: colorlog
+      hydra/hydra_logging: colorlog
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false
+backend:
+  name: pytorch
+  version: ${pytorch_version:}
+  _target_: optimum_benchmark.backends.pytorch.backend.PyTorchBackend
+  seed: 42
+  inter_op_num_threads: null
+  intra_op_num_threads: null
+  initial_isolation_check: true
+  continous_isolation_check: true
+  delete_cache: false
+  no_weights: false
+  device_map: null
+  torch_dtype: null
+  disable_grad: ${is_inference:${benchmark.name}}
+  eval_mode: ${is_inference:${benchmark.name}}
+  amp_autocast: false
+  amp_dtype: null
+  torch_compile: false
+  torch_compile_config: {}
+  bettertransformer: false
+  quantization_scheme: null
+  quantization_config: {}
+  use_ddp: false
+  ddp_config: {}
+  peft_strategy: null
+  peft_config: {}
+benchmark:
+  name: training
+  _target_: optimum_benchmark.benchmarks.training.benchmark.TrainingBenchmark
+  warmup_steps: 40
+  dataset_shapes:
+    dataset_size: 1500
+    sequence_length: 256
+    num_choices: 1
+    feature_size: 80
+    nb_max_frames: 3000
+    audio_sequence_length: 16000
+  training_arguments:
+    skip_memory_metrics: true
+    output_dir: ./trainer_output
+    use_cpu: ${is_cpu:${device}}
+    ddp_find_unused_parameters: false
+    do_train: true
+    do_eval: false
+    do_predict: false
+    report_to: none
+experiment_name: bert_1gpu_training
+model: bert-base-uncased
+device: cuda
+task: text-classification
+hub_kwargs:
+  revision: main
+  cache_dir: null
+  force_download: false
+  local_files_only: false
+environment:
+  optimum_version: 1.13.1
+  transformers_version: 4.34.0.dev0
+  accelerate_version: 0.23.0
+  diffusers_version: null
+  python_version: 3.10.12
+  system: Linux
+  cpu: ' AMD EPYC 7643 48-Core Processor'
+  cpu_count: 96
+  cpu_ram_mb: 1082028
+  gpus:
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/0/.config/config.yaml

@@ -0,0 +1,73 @@
+backend:
+  name: pytorch
+  version: ${pytorch_version:}
+  _target_: optimum_benchmark.backends.pytorch.backend.PyTorchBackend
+  seed: 42
+  inter_op_num_threads: null
+  intra_op_num_threads: null
+  initial_isolation_check: true
+  continous_isolation_check: true
+  delete_cache: false
+  no_weights: false
+  device_map: null
+  torch_dtype: float16
+  disable_grad: ${is_inference:${benchmark.name}}
+  eval_mode: ${is_inference:${benchmark.name}}
+  amp_autocast: false
+  amp_dtype: null
+  torch_compile: false
+  torch_compile_config: {}
+  bettertransformer: false
+  quantization_scheme: null
+  quantization_config: {}
+  use_ddp: false
+  ddp_config: {}
+  peft_strategy: null
+  peft_config: {}
+benchmark:
+  name: inference
+  _target_: optimum_benchmark.benchmarks.inference.benchmark.InferenceBenchmark
+  duration: 10
+  warmup_runs: 10
+  memory: false
+  energy: false
+  input_shapes:
+    batch_size: 1
+    sequence_length: 200
+    num_choices: 1
+    feature_size: 80
+    nb_max_frames: 3000
+    audio_sequence_length: 16000
+  new_tokens: 200
+  can_diffuse: ${can_diffuse:${task}}
+  can_generate: ${can_generate:${task}}
+  forward_kwargs: {}
+  generate_kwargs: {}
+experiment_name: llama_1gpu_inference
+model: fxmarty/tiny-llama-fast-tokenizer
+device: cuda
+task: ${infer_task:${model}}
+hub_kwargs:
+  revision: main
+  cache_dir: null
+  force_download: false
+  local_files_only: false
+environment:
+  optimum_version: 1.13.1
+  transformers_version: 4.34.0.dev0
+  accelerate_version: 0.23.0
+  diffusers_version: null
+  python_version: 3.10.12
+  system: Linux
+  cpu: ' AMD EPYC 7643 48-Core Processor'
+  cpu_count: 96
+  cpu_ram_mb: 1082028
+  gpus:
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/0/.config/hydra.yaml

@@ -0,0 +1,174 @@
+hydra:
+  run:
+    dir: runs/${oc.env:COMMIT_DATE_GMT}_${oc.env:COMMIT_SHA}/${experiment_name}
+  sweep:
+    dir: sweeps/${oc.env:COMMIT_DATE_GMT}_${oc.env:COMMIT_SHA}/${experiment_name}
+    subdir: ${hydra.job.num}
+  launcher:
+    _target_: hydra._internal.core_plugins.basic_launcher.BasicLauncher
+  sweeper:
+    _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
+    max_batch_size: null
+    params:
+      benchmark.input_shapes.batch_size: 1,16
+      backend.torch_dtype: float16,float32
+  help:
+    app_name: ${hydra.job.name}
+    header: '${hydra.help.app_name} is powered by Hydra.
+
+      '
+    footer: 'Powered by Hydra (https://hydra.cc)
+
+      Use --hydra-help to view Hydra specific help
+
+      '
+    template: '${hydra.help.header}
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (group=option)
+
+
+      $APP_CONFIG_GROUPS
+
+
+      == Config ==
+
+      Override anything in the config (foo.bar=value)
+
+
+      $CONFIG
+
+
+      ${hydra.help.footer}
+
+      '
+  hydra_help:
+    template: 'Hydra (${hydra.runtime.version})
+
+      See https://hydra.cc for more info.
+
+
+      == Flags ==
+
+      $FLAGS_HELP
+
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (For example, append hydra/job_logging=disabled
+      to command line)
+
+
+      $HYDRA_CONFIG_GROUPS
+
+
+      Use ''--cfg hydra'' to Show the Hydra config.
+
+      '
+    hydra_help: ???
+  hydra_logging:
+    version: 1
+    formatters:
+      colorlog:
+        (): colorlog.ColoredFormatter
+        format: '[%(cyan)s%(asctime)s%(reset)s][%(purple)sHYDRA%(reset)s] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: colorlog
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+      colorlog:
+        (): colorlog.ColoredFormatter
+        format: '[%(cyan)s%(asctime)s%(reset)s][%(blue)s%(name)s%(reset)s][%(log_color)s%(levelname)s%(reset)s]
+          - %(message)s'
+        log_colors:
+          DEBUG: purple
+          INFO: green
+          WARNING: yellow
+          ERROR: red
+          CRITICAL: red
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: colorlog
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: MULTIRUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=MULTIRUN
+    task:
+    - benchmark.input_shapes.batch_size=1
+    - backend.torch_dtype=float16
+  job:
+    name: experiment
+    chdir: true
+    override_dirname: backend.torch_dtype=float16,benchmark.input_shapes.batch_size=1
+    id: '0'
+    num: 0
+    config_name: llama2_1gpu_inference
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.3.2
+    version_base: '1.3'
+    cwd: /home/user/transformers-regression
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: optimum_benchmark
+      schema: pkg
+      provider: main
+    - path: hydra_plugins.hydra_colorlog.conf
+      schema: pkg
+      provider: hydra-colorlog
+    - path: /home/user/transformers-regression/configs
+      schema: file
+      provider: command-line
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: /home/user/transformers-regression/sweeps/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/0
+    choices:
+      benchmark: inference
+      backend: pytorch
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: colorlog
+      hydra/hydra_logging: colorlog
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/0/.config/overrides.yaml

@@ -0,0 +1,2 @@
+- benchmark.input_shapes.batch_size=1
+- backend.torch_dtype=float16

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/0/experiment.log

@@ -0,0 +1,27 @@
+[2023-09-27 12:02:13,594][inference][INFO] - `new_tokens` was set to 200. `max_new_tokens` and `min_new_tokens` will be set to 200.
+[2023-09-27 12:02:13,737][experiment][WARNING] - Multiple GPUs detected but CUDA_DEVICE_ORDER is not set. This means that code might allocate resources from the wrong GPUs even if CUDA_VISIBLE_DEVICES is set. Pytorch uses the `FASTEST_FIRST` order by default, which is not guaranteed to be the same as nvidia-smi. `CUDA_DEVICE_ORDER` will be set to `PCI_BUS_ID` to ensure that the GPUs are allocated in the same order as nvidia-smi. 
+[2023-09-27 12:02:15,832][pytorch][INFO] - 	+ Inferred AutoModel class AutoModelForCausalLM for task text-generation and model_type llama
+[2023-09-27 12:02:15,832][backend][INFO] - Configuring pytorch backend
+[2023-09-27 12:02:15,833][backend][INFO] - 	+ Checking initial device(s) isolation of CUDA device(s): [0]
+[2023-09-27 12:02:15,962][backend][INFO] - 	+ Checking continuous device(s) isolation of CUDA device(s): [0]
+[2023-09-27 12:02:15,977][pytorch][INFO] - 	+ Disabling gradients
+[2023-09-27 12:02:15,978][pytorch][INFO] - 	+ Loading model on device: cuda
+[2023-09-27 12:02:16,614][pytorch][INFO] - 	+ Turning on model's eval mode
+[2023-09-27 12:02:16,621][benchmark][INFO] - Configuring inference benchmark
+[2023-09-27 12:02:16,621][inference][INFO] - Running inference benchmark
+[2023-09-27 12:02:16,622][input_generator][INFO] - Using llama model type generator
+[2023-09-27 12:02:16,642][inference][INFO] - 	+ Preparing input for the forward pass
+[2023-09-27 12:02:16,642][inference][INFO] - 	+ Warming up the forward pass
+[2023-09-27 12:02:16,970][inference][INFO] - 	+ Tracking forward pass latency and throughput
+[2023-09-27 12:02:16,971][latency_tracker][INFO] - Tracked Pytorch devices: [0]
+[2023-09-27 12:02:27,175][inference][INFO] - 	+ Forward pass latency: 3.32e-03 (s)
+[2023-09-27 12:02:27,177][inference][INFO] - 	+ Forward pass throughput: 301.00 (samples/s)
+[2023-09-27 12:02:27,177][inference][INFO] - 	+ Preparing input for the generation pass
+[2023-09-27 12:02:27,178][inference][INFO] - 	+ Warming up the generation pass
+[2023-09-27 12:02:28,309][inference][INFO] - 	+ Tracking generation latency and throughput
+[2023-09-27 12:02:28,309][latency_tracker][INFO] - Tracked Pytorch devices: [0]
+[2023-09-27 12:02:38,515][inference][INFO] - 	+ Generation pass latency: 5.37e-01 (s)
+[2023-09-27 12:02:38,515][inference][INFO] - 	+ Generation pass throughput: 372.00 (tokens/s)
+[2023-09-27 12:02:38,515][inference][INFO] - Saving inference results
+[2023-09-27 12:02:38,522][backend][INFO] - Cleaning pytorch backend
+[2023-09-27 12:02:38,522][backend][INFO] - 	+ Deleting pretrained model

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/0/hydra_config.yaml

@@ -0,0 +1,79 @@
+backend:
+  name: pytorch
+  version: 2.1.0+rocm5.6
+  _target_: optimum_benchmark.backends.pytorch.backend.PyTorchBackend
+  seed: 42
+  inter_op_num_threads: null
+  intra_op_num_threads: null
+  initial_isolation_check: true
+  continous_isolation_check: true
+  delete_cache: false
+  no_weights: false
+  device_map: null
+  torch_dtype: float16
+  disable_grad: true
+  eval_mode: true
+  amp_autocast: false
+  amp_dtype: null
+  torch_compile: false
+  torch_compile_config: {}
+  bettertransformer: false
+  quantization_scheme: null
+  quantization_config: {}
+  use_ddp: false
+  ddp_config: {}
+  peft_strategy: null
+  peft_config: {}
+benchmark:
+  name: inference
+  _target_: optimum_benchmark.benchmarks.inference.benchmark.InferenceBenchmark
+  duration: 10
+  warmup_runs: 10
+  memory: false
+  energy: false
+  input_shapes:
+    batch_size: 1
+    sequence_length: 200
+    num_choices: 1
+    feature_size: 80
+    nb_max_frames: 3000
+    audio_sequence_length: 16000
+  new_tokens: 200
+  can_diffuse: false
+  can_generate: true
+  forward_kwargs: {}
+  generate_kwargs:
+    max_new_tokens: 200
+    min_new_tokens: 200
+    do_sample: false
+    use_cache: true
+    pad_token_id: 0
+    num_beams: 1
+experiment_name: llama_1gpu_inference
+model: fxmarty/tiny-llama-fast-tokenizer
+device: cuda
+task: text-generation
+hub_kwargs:
+  revision: main
+  cache_dir: null
+  force_download: false
+  local_files_only: false
+environment:
+  optimum_version: 1.13.1
+  transformers_version: 4.34.0.dev0
+  accelerate_version: 0.23.0
+  diffusers_version: null
+  python_version: 3.10.12
+  system: Linux
+  cpu: ' AMD EPYC 7643 48-Core Processor'
+  cpu_count: 96
+  cpu_ram_mb: 1082028
+  gpus:
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/0/inference_results.csv

@@ -0,0 +1,2 @@
+forward.latency(s),forward.throughput(samples/s),generate.latency(s),generate.throughput(tokens/s)
+0.00332,301.0,0.537,372.0

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/1/.config/config.yaml

@@ -0,0 +1,73 @@
+backend:
+  name: pytorch
+  version: ${pytorch_version:}
+  _target_: optimum_benchmark.backends.pytorch.backend.PyTorchBackend
+  seed: 42
+  inter_op_num_threads: null
+  intra_op_num_threads: null
+  initial_isolation_check: true
+  continous_isolation_check: true
+  delete_cache: false
+  no_weights: false
+  device_map: null
+  torch_dtype: float32
+  disable_grad: ${is_inference:${benchmark.name}}
+  eval_mode: ${is_inference:${benchmark.name}}
+  amp_autocast: false
+  amp_dtype: null
+  torch_compile: false
+  torch_compile_config: {}
+  bettertransformer: false
+  quantization_scheme: null
+  quantization_config: {}
+  use_ddp: false
+  ddp_config: {}
+  peft_strategy: null
+  peft_config: {}
+benchmark:
+  name: inference
+  _target_: optimum_benchmark.benchmarks.inference.benchmark.InferenceBenchmark
+  duration: 10
+  warmup_runs: 10
+  memory: false
+  energy: false
+  input_shapes:
+    batch_size: 1
+    sequence_length: 200
+    num_choices: 1
+    feature_size: 80
+    nb_max_frames: 3000
+    audio_sequence_length: 16000
+  new_tokens: 200
+  can_diffuse: ${can_diffuse:${task}}
+  can_generate: ${can_generate:${task}}
+  forward_kwargs: {}
+  generate_kwargs: {}
+experiment_name: llama_1gpu_inference
+model: fxmarty/tiny-llama-fast-tokenizer
+device: cuda
+task: ${infer_task:${model}}
+hub_kwargs:
+  revision: main
+  cache_dir: null
+  force_download: false
+  local_files_only: false
+environment:
+  optimum_version: 1.13.1
+  transformers_version: 4.34.0.dev0
+  accelerate_version: 0.23.0
+  diffusers_version: null
+  python_version: 3.10.12
+  system: Linux
+  cpu: ' AMD EPYC 7643 48-Core Processor'
+  cpu_count: 96
+  cpu_ram_mb: 1082028
+  gpus:
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/1/.config/hydra.yaml

@@ -0,0 +1,174 @@
+hydra:
+  run:
+    dir: runs/${oc.env:COMMIT_DATE_GMT}_${oc.env:COMMIT_SHA}/${experiment_name}
+  sweep:
+    dir: sweeps/${oc.env:COMMIT_DATE_GMT}_${oc.env:COMMIT_SHA}/${experiment_name}
+    subdir: ${hydra.job.num}
+  launcher:
+    _target_: hydra._internal.core_plugins.basic_launcher.BasicLauncher
+  sweeper:
+    _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
+    max_batch_size: null
+    params:
+      benchmark.input_shapes.batch_size: 1,16
+      backend.torch_dtype: float16,float32
+  help:
+    app_name: ${hydra.job.name}
+    header: '${hydra.help.app_name} is powered by Hydra.
+
+      '
+    footer: 'Powered by Hydra (https://hydra.cc)
+
+      Use --hydra-help to view Hydra specific help
+
+      '
+    template: '${hydra.help.header}
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (group=option)
+
+
+      $APP_CONFIG_GROUPS
+
+
+      == Config ==
+
+      Override anything in the config (foo.bar=value)
+
+
+      $CONFIG
+
+
+      ${hydra.help.footer}
+
+      '
+  hydra_help:
+    template: 'Hydra (${hydra.runtime.version})
+
+      See https://hydra.cc for more info.
+
+
+      == Flags ==
+
+      $FLAGS_HELP
+
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (For example, append hydra/job_logging=disabled
+      to command line)
+
+
+      $HYDRA_CONFIG_GROUPS
+
+
+      Use ''--cfg hydra'' to Show the Hydra config.
+
+      '
+    hydra_help: ???
+  hydra_logging:
+    version: 1
+    formatters:
+      colorlog:
+        (): colorlog.ColoredFormatter
+        format: '[%(cyan)s%(asctime)s%(reset)s][%(purple)sHYDRA%(reset)s] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: colorlog
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+      colorlog:
+        (): colorlog.ColoredFormatter
+        format: '[%(cyan)s%(asctime)s%(reset)s][%(blue)s%(name)s%(reset)s][%(log_color)s%(levelname)s%(reset)s]
+          - %(message)s'
+        log_colors:
+          DEBUG: purple
+          INFO: green
+          WARNING: yellow
+          ERROR: red
+          CRITICAL: red
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: colorlog
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: MULTIRUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=MULTIRUN
+    task:
+    - benchmark.input_shapes.batch_size=1
+    - backend.torch_dtype=float32
+  job:
+    name: experiment
+    chdir: true
+    override_dirname: backend.torch_dtype=float32,benchmark.input_shapes.batch_size=1
+    id: '1'
+    num: 1
+    config_name: llama2_1gpu_inference
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.3.2
+    version_base: '1.3'
+    cwd: /home/user/transformers-regression
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: optimum_benchmark
+      schema: pkg
+      provider: main
+    - path: hydra_plugins.hydra_colorlog.conf
+      schema: pkg
+      provider: hydra-colorlog
+    - path: /home/user/transformers-regression/configs
+      schema: file
+      provider: command-line
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: /home/user/transformers-regression/sweeps/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/1
+    choices:
+      benchmark: inference
+      backend: pytorch
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: colorlog
+      hydra/hydra_logging: colorlog
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/1/.config/overrides.yaml

@@ -0,0 +1,2 @@
+- benchmark.input_shapes.batch_size=1
+- backend.torch_dtype=float32

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/1/experiment.log

@@ -0,0 +1,26 @@
+[2023-09-27 12:02:39,149][inference][INFO] - `new_tokens` was set to 200. `max_new_tokens` and `min_new_tokens` will be set to 200.
+[2023-09-27 12:02:40,059][pytorch][INFO] - 	+ Inferred AutoModel class AutoModelForCausalLM for task text-generation and model_type llama
+[2023-09-27 12:02:40,059][backend][INFO] - Configuring pytorch backend
+[2023-09-27 12:02:40,059][backend][INFO] - 	+ Checking initial device(s) isolation of CUDA device(s): [0]
+[2023-09-27 12:02:40,182][backend][INFO] - 	+ Checking continuous device(s) isolation of CUDA device(s): [0]
+[2023-09-27 12:02:40,199][pytorch][INFO] - 	+ Disabling gradients
+[2023-09-27 12:02:40,200][pytorch][INFO] - 	+ Loading model on device: cuda
+[2023-09-27 12:02:40,560][pytorch][INFO] - 	+ Turning on model's eval mode
+[2023-09-27 12:02:40,561][benchmark][INFO] - Configuring inference benchmark
+[2023-09-27 12:02:40,561][inference][INFO] - Running inference benchmark
+[2023-09-27 12:02:40,562][input_generator][INFO] - Using llama model type generator
+[2023-09-27 12:02:40,562][inference][INFO] - 	+ Preparing input for the forward pass
+[2023-09-27 12:02:40,562][inference][INFO] - 	+ Warming up the forward pass
+[2023-09-27 12:02:40,783][inference][INFO] - 	+ Tracking forward pass latency and throughput
+[2023-09-27 12:02:40,783][latency_tracker][INFO] - Tracked Pytorch devices: [0]
+[2023-09-27 12:02:51,082][inference][INFO] - 	+ Forward pass latency: 3.12e-03 (s)
+[2023-09-27 12:02:51,084][inference][INFO] - 	+ Forward pass throughput: 321.00 (samples/s)
+[2023-09-27 12:02:51,084][inference][INFO] - 	+ Preparing input for the generation pass
+[2023-09-27 12:02:51,084][inference][INFO] - 	+ Warming up the generation pass
+[2023-09-27 12:02:51,711][inference][INFO] - 	+ Tracking generation latency and throughput
+[2023-09-27 12:02:51,711][latency_tracker][INFO] - Tracked Pytorch devices: [0]
+[2023-09-27 12:03:01,932][inference][INFO] - 	+ Generation pass latency: 6.39e-01 (s)
+[2023-09-27 12:03:01,933][inference][INFO] - 	+ Generation pass throughput: 313.00 (tokens/s)
+[2023-09-27 12:03:01,933][inference][INFO] - Saving inference results
+[2023-09-27 12:03:01,938][backend][INFO] - Cleaning pytorch backend
+[2023-09-27 12:03:01,939][backend][INFO] - 	+ Deleting pretrained model

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/1/hydra_config.yaml

@@ -0,0 +1,79 @@
+backend:
+  name: pytorch
+  version: 2.1.0+rocm5.6
+  _target_: optimum_benchmark.backends.pytorch.backend.PyTorchBackend
+  seed: 42
+  inter_op_num_threads: null
+  intra_op_num_threads: null
+  initial_isolation_check: true
+  continous_isolation_check: true
+  delete_cache: false
+  no_weights: false
+  device_map: null
+  torch_dtype: float32
+  disable_grad: true
+  eval_mode: true
+  amp_autocast: false
+  amp_dtype: null
+  torch_compile: false
+  torch_compile_config: {}
+  bettertransformer: false
+  quantization_scheme: null
+  quantization_config: {}
+  use_ddp: false
+  ddp_config: {}
+  peft_strategy: null
+  peft_config: {}
+benchmark:
+  name: inference
+  _target_: optimum_benchmark.benchmarks.inference.benchmark.InferenceBenchmark
+  duration: 10
+  warmup_runs: 10
+  memory: false
+  energy: false
+  input_shapes:
+    batch_size: 1
+    sequence_length: 200
+    num_choices: 1
+    feature_size: 80
+    nb_max_frames: 3000
+    audio_sequence_length: 16000
+  new_tokens: 200
+  can_diffuse: false
+  can_generate: true
+  forward_kwargs: {}
+  generate_kwargs:
+    max_new_tokens: 200
+    min_new_tokens: 200
+    do_sample: false
+    use_cache: true
+    pad_token_id: 0
+    num_beams: 1
+experiment_name: llama_1gpu_inference
+model: fxmarty/tiny-llama-fast-tokenizer
+device: cuda
+task: text-generation
+hub_kwargs:
+  revision: main
+  cache_dir: null
+  force_download: false
+  local_files_only: false
+environment:
+  optimum_version: 1.13.1
+  transformers_version: 4.34.0.dev0
+  accelerate_version: 0.23.0
+  diffusers_version: null
+  python_version: 3.10.12
+  system: Linux
+  cpu: ' AMD EPYC 7643 48-Core Processor'
+  cpu_count: 96
+  cpu_ram_mb: 1082028
+  gpus:
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/1/inference_results.csv

@@ -0,0 +1,2 @@
+forward.latency(s),forward.throughput(samples/s),generate.latency(s),generate.throughput(tokens/s)
+0.00312,321.0,0.639,313.0

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/2/.config/config.yaml

@@ -0,0 +1,73 @@
+backend:
+  name: pytorch
+  version: ${pytorch_version:}
+  _target_: optimum_benchmark.backends.pytorch.backend.PyTorchBackend
+  seed: 42
+  inter_op_num_threads: null
+  intra_op_num_threads: null
+  initial_isolation_check: true
+  continous_isolation_check: true
+  delete_cache: false
+  no_weights: false
+  device_map: null
+  torch_dtype: float16
+  disable_grad: ${is_inference:${benchmark.name}}
+  eval_mode: ${is_inference:${benchmark.name}}
+  amp_autocast: false
+  amp_dtype: null
+  torch_compile: false
+  torch_compile_config: {}
+  bettertransformer: false
+  quantization_scheme: null
+  quantization_config: {}
+  use_ddp: false
+  ddp_config: {}
+  peft_strategy: null
+  peft_config: {}
+benchmark:
+  name: inference
+  _target_: optimum_benchmark.benchmarks.inference.benchmark.InferenceBenchmark
+  duration: 10
+  warmup_runs: 10
+  memory: false
+  energy: false
+  input_shapes:
+    batch_size: 16
+    sequence_length: 200
+    num_choices: 1
+    feature_size: 80
+    nb_max_frames: 3000
+    audio_sequence_length: 16000
+  new_tokens: 200
+  can_diffuse: ${can_diffuse:${task}}
+  can_generate: ${can_generate:${task}}
+  forward_kwargs: {}
+  generate_kwargs: {}
+experiment_name: llama_1gpu_inference
+model: fxmarty/tiny-llama-fast-tokenizer
+device: cuda
+task: ${infer_task:${model}}
+hub_kwargs:
+  revision: main
+  cache_dir: null
+  force_download: false
+  local_files_only: false
+environment:
+  optimum_version: 1.13.1
+  transformers_version: 4.34.0.dev0
+  accelerate_version: 0.23.0
+  diffusers_version: null
+  python_version: 3.10.12
+  system: Linux
+  cpu: ' AMD EPYC 7643 48-Core Processor'
+  cpu_count: 96
+  cpu_ram_mb: 1082028
+  gpus:
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/2/.config/hydra.yaml

@@ -0,0 +1,174 @@
+hydra:
+  run:
+    dir: runs/${oc.env:COMMIT_DATE_GMT}_${oc.env:COMMIT_SHA}/${experiment_name}
+  sweep:
+    dir: sweeps/${oc.env:COMMIT_DATE_GMT}_${oc.env:COMMIT_SHA}/${experiment_name}
+    subdir: ${hydra.job.num}
+  launcher:
+    _target_: hydra._internal.core_plugins.basic_launcher.BasicLauncher
+  sweeper:
+    _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
+    max_batch_size: null
+    params:
+      benchmark.input_shapes.batch_size: 1,16
+      backend.torch_dtype: float16,float32
+  help:
+    app_name: ${hydra.job.name}
+    header: '${hydra.help.app_name} is powered by Hydra.
+
+      '
+    footer: 'Powered by Hydra (https://hydra.cc)
+
+      Use --hydra-help to view Hydra specific help
+
+      '
+    template: '${hydra.help.header}
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (group=option)
+
+
+      $APP_CONFIG_GROUPS
+
+
+      == Config ==
+
+      Override anything in the config (foo.bar=value)
+
+
+      $CONFIG
+
+
+      ${hydra.help.footer}
+
+      '
+  hydra_help:
+    template: 'Hydra (${hydra.runtime.version})
+
+      See https://hydra.cc for more info.
+
+
+      == Flags ==
+
+      $FLAGS_HELP
+
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (For example, append hydra/job_logging=disabled
+      to command line)
+
+
+      $HYDRA_CONFIG_GROUPS
+
+
+      Use ''--cfg hydra'' to Show the Hydra config.
+
+      '
+    hydra_help: ???
+  hydra_logging:
+    version: 1
+    formatters:
+      colorlog:
+        (): colorlog.ColoredFormatter
+        format: '[%(cyan)s%(asctime)s%(reset)s][%(purple)sHYDRA%(reset)s] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: colorlog
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+      colorlog:
+        (): colorlog.ColoredFormatter
+        format: '[%(cyan)s%(asctime)s%(reset)s][%(blue)s%(name)s%(reset)s][%(log_color)s%(levelname)s%(reset)s]
+          - %(message)s'
+        log_colors:
+          DEBUG: purple
+          INFO: green
+          WARNING: yellow
+          ERROR: red
+          CRITICAL: red
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: colorlog
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: MULTIRUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=MULTIRUN
+    task:
+    - benchmark.input_shapes.batch_size=16
+    - backend.torch_dtype=float16
+  job:
+    name: experiment
+    chdir: true
+    override_dirname: backend.torch_dtype=float16,benchmark.input_shapes.batch_size=16
+    id: '2'
+    num: 2
+    config_name: llama2_1gpu_inference
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.3.2
+    version_base: '1.3'
+    cwd: /home/user/transformers-regression
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: optimum_benchmark
+      schema: pkg
+      provider: main
+    - path: hydra_plugins.hydra_colorlog.conf
+      schema: pkg
+      provider: hydra-colorlog
+    - path: /home/user/transformers-regression/configs
+      schema: file
+      provider: command-line
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: /home/user/transformers-regression/sweeps/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/2
+    choices:
+      benchmark: inference
+      backend: pytorch
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: colorlog
+      hydra/hydra_logging: colorlog
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/2/.config/overrides.yaml

@@ -0,0 +1,2 @@
+- benchmark.input_shapes.batch_size=16
+- backend.torch_dtype=float16

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/2/experiment.log

@@ -0,0 +1,26 @@
+[2023-09-27 12:03:02,564][inference][INFO] - `new_tokens` was set to 200. `max_new_tokens` and `min_new_tokens` will be set to 200.
+[2023-09-27 12:03:03,480][pytorch][INFO] - 	+ Inferred AutoModel class AutoModelForCausalLM for task text-generation and model_type llama
+[2023-09-27 12:03:03,480][backend][INFO] - Configuring pytorch backend
+[2023-09-27 12:03:03,481][backend][INFO] - 	+ Checking initial device(s) isolation of CUDA device(s): [0]
+[2023-09-27 12:03:03,603][backend][INFO] - 	+ Checking continuous device(s) isolation of CUDA device(s): [0]
+[2023-09-27 12:03:03,624][pytorch][INFO] - 	+ Disabling gradients
+[2023-09-27 12:03:03,625][pytorch][INFO] - 	+ Loading model on device: cuda
+[2023-09-27 12:03:03,982][pytorch][INFO] - 	+ Turning on model's eval mode
+[2023-09-27 12:03:03,982][benchmark][INFO] - Configuring inference benchmark
+[2023-09-27 12:03:03,982][inference][INFO] - Running inference benchmark
+[2023-09-27 12:03:03,983][input_generator][INFO] - Using llama model type generator
+[2023-09-27 12:03:03,983][inference][INFO] - 	+ Preparing input for the forward pass
+[2023-09-27 12:03:03,983][inference][INFO] - 	+ Warming up the forward pass
+[2023-09-27 12:03:04,016][inference][INFO] - 	+ Tracking forward pass latency and throughput
+[2023-09-27 12:03:04,016][latency_tracker][INFO] - Tracked Pytorch devices: [0]
+[2023-09-27 12:03:14,407][inference][INFO] - 	+ Forward pass latency: 6.09e-03 (s)
+[2023-09-27 12:03:14,409][inference][INFO] - 	+ Forward pass throughput: 2630.00 (samples/s)
+[2023-09-27 12:03:14,409][inference][INFO] - 	+ Preparing input for the generation pass
+[2023-09-27 12:03:14,409][inference][INFO] - 	+ Warming up the generation pass
+[2023-09-27 12:03:15,902][inference][INFO] - 	+ Tracking generation latency and throughput
+[2023-09-27 12:03:15,903][latency_tracker][INFO] - Tracked Pytorch devices: [0]
+[2023-09-27 12:03:26,584][inference][INFO] - 	+ Generation pass latency: 8.22e-01 (s)
+[2023-09-27 12:03:26,585][inference][INFO] - 	+ Generation pass throughput: 3890.00 (tokens/s)
+[2023-09-27 12:03:26,585][inference][INFO] - Saving inference results
+[2023-09-27 12:03:26,589][backend][INFO] - Cleaning pytorch backend
+[2023-09-27 12:03:26,589][backend][INFO] - 	+ Deleting pretrained model

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/2/hydra_config.yaml

@@ -0,0 +1,79 @@
+backend:
+  name: pytorch
+  version: 2.1.0+rocm5.6
+  _target_: optimum_benchmark.backends.pytorch.backend.PyTorchBackend
+  seed: 42
+  inter_op_num_threads: null
+  intra_op_num_threads: null
+  initial_isolation_check: true
+  continous_isolation_check: true
+  delete_cache: false
+  no_weights: false
+  device_map: null
+  torch_dtype: float16
+  disable_grad: true
+  eval_mode: true
+  amp_autocast: false
+  amp_dtype: null
+  torch_compile: false
+  torch_compile_config: {}
+  bettertransformer: false
+  quantization_scheme: null
+  quantization_config: {}
+  use_ddp: false
+  ddp_config: {}
+  peft_strategy: null
+  peft_config: {}
+benchmark:
+  name: inference
+  _target_: optimum_benchmark.benchmarks.inference.benchmark.InferenceBenchmark
+  duration: 10
+  warmup_runs: 10
+  memory: false
+  energy: false
+  input_shapes:
+    batch_size: 16
+    sequence_length: 200
+    num_choices: 1
+    feature_size: 80
+    nb_max_frames: 3000
+    audio_sequence_length: 16000
+  new_tokens: 200
+  can_diffuse: false
+  can_generate: true
+  forward_kwargs: {}
+  generate_kwargs:
+    max_new_tokens: 200
+    min_new_tokens: 200
+    do_sample: false
+    use_cache: true
+    pad_token_id: 0
+    num_beams: 1
+experiment_name: llama_1gpu_inference
+model: fxmarty/tiny-llama-fast-tokenizer
+device: cuda
+task: text-generation
+hub_kwargs:
+  revision: main
+  cache_dir: null
+  force_download: false
+  local_files_only: false
+environment:
+  optimum_version: 1.13.1
+  transformers_version: 4.34.0.dev0
+  accelerate_version: 0.23.0
+  diffusers_version: null
+  python_version: 3.10.12
+  system: Linux
+  cpu: ' AMD EPYC 7643 48-Core Processor'
+  cpu_count: 96
+  cpu_ram_mb: 1082028
+  gpus:
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/2/inference_results.csv

@@ -0,0 +1,2 @@
+forward.latency(s),forward.throughput(samples/s),generate.latency(s),generate.throughput(tokens/s)
+0.00609,2630.0,0.822,3890.0

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/3/.config/config.yaml

@@ -0,0 +1,73 @@
+backend:
+  name: pytorch
+  version: ${pytorch_version:}
+  _target_: optimum_benchmark.backends.pytorch.backend.PyTorchBackend
+  seed: 42
+  inter_op_num_threads: null
+  intra_op_num_threads: null
+  initial_isolation_check: true
+  continous_isolation_check: true
+  delete_cache: false
+  no_weights: false
+  device_map: null
+  torch_dtype: float32
+  disable_grad: ${is_inference:${benchmark.name}}
+  eval_mode: ${is_inference:${benchmark.name}}
+  amp_autocast: false
+  amp_dtype: null
+  torch_compile: false
+  torch_compile_config: {}
+  bettertransformer: false
+  quantization_scheme: null
+  quantization_config: {}
+  use_ddp: false
+  ddp_config: {}
+  peft_strategy: null
+  peft_config: {}
+benchmark:
+  name: inference
+  _target_: optimum_benchmark.benchmarks.inference.benchmark.InferenceBenchmark
+  duration: 10
+  warmup_runs: 10
+  memory: false
+  energy: false
+  input_shapes:
+    batch_size: 16
+    sequence_length: 200
+    num_choices: 1
+    feature_size: 80
+    nb_max_frames: 3000
+    audio_sequence_length: 16000
+  new_tokens: 200
+  can_diffuse: ${can_diffuse:${task}}
+  can_generate: ${can_generate:${task}}
+  forward_kwargs: {}
+  generate_kwargs: {}
+experiment_name: llama_1gpu_inference
+model: fxmarty/tiny-llama-fast-tokenizer
+device: cuda
+task: ${infer_task:${model}}
+hub_kwargs:
+  revision: main
+  cache_dir: null
+  force_download: false
+  local_files_only: false
+environment:
+  optimum_version: 1.13.1
+  transformers_version: 4.34.0.dev0
+  accelerate_version: 0.23.0
+  diffusers_version: null
+  python_version: 3.10.12
+  system: Linux
+  cpu: ' AMD EPYC 7643 48-Core Processor'
+  cpu_count: 96
+  cpu_ram_mb: 1082028
+  gpus:
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/3/.config/hydra.yaml

@@ -0,0 +1,174 @@
+hydra:
+  run:
+    dir: runs/${oc.env:COMMIT_DATE_GMT}_${oc.env:COMMIT_SHA}/${experiment_name}
+  sweep:
+    dir: sweeps/${oc.env:COMMIT_DATE_GMT}_${oc.env:COMMIT_SHA}/${experiment_name}
+    subdir: ${hydra.job.num}
+  launcher:
+    _target_: hydra._internal.core_plugins.basic_launcher.BasicLauncher
+  sweeper:
+    _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
+    max_batch_size: null
+    params:
+      benchmark.input_shapes.batch_size: 1,16
+      backend.torch_dtype: float16,float32
+  help:
+    app_name: ${hydra.job.name}
+    header: '${hydra.help.app_name} is powered by Hydra.
+
+      '
+    footer: 'Powered by Hydra (https://hydra.cc)
+
+      Use --hydra-help to view Hydra specific help
+
+      '
+    template: '${hydra.help.header}
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (group=option)
+
+
+      $APP_CONFIG_GROUPS
+
+
+      == Config ==
+
+      Override anything in the config (foo.bar=value)
+
+
+      $CONFIG
+
+
+      ${hydra.help.footer}
+
+      '
+  hydra_help:
+    template: 'Hydra (${hydra.runtime.version})
+
+      See https://hydra.cc for more info.
+
+
+      == Flags ==
+
+      $FLAGS_HELP
+
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (For example, append hydra/job_logging=disabled
+      to command line)
+
+
+      $HYDRA_CONFIG_GROUPS
+
+
+      Use ''--cfg hydra'' to Show the Hydra config.
+
+      '
+    hydra_help: ???
+  hydra_logging:
+    version: 1
+    formatters:
+      colorlog:
+        (): colorlog.ColoredFormatter
+        format: '[%(cyan)s%(asctime)s%(reset)s][%(purple)sHYDRA%(reset)s] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: colorlog
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+      colorlog:
+        (): colorlog.ColoredFormatter
+        format: '[%(cyan)s%(asctime)s%(reset)s][%(blue)s%(name)s%(reset)s][%(log_color)s%(levelname)s%(reset)s]
+          - %(message)s'
+        log_colors:
+          DEBUG: purple
+          INFO: green
+          WARNING: yellow
+          ERROR: red
+          CRITICAL: red
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: colorlog
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: MULTIRUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=MULTIRUN
+    task:
+    - benchmark.input_shapes.batch_size=16
+    - backend.torch_dtype=float32
+  job:
+    name: experiment
+    chdir: true
+    override_dirname: backend.torch_dtype=float32,benchmark.input_shapes.batch_size=16
+    id: '3'
+    num: 3
+    config_name: llama2_1gpu_inference
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.3.2
+    version_base: '1.3'
+    cwd: /home/user/transformers-regression
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: optimum_benchmark
+      schema: pkg
+      provider: main
+    - path: hydra_plugins.hydra_colorlog.conf
+      schema: pkg
+      provider: hydra-colorlog
+    - path: /home/user/transformers-regression/configs
+      schema: file
+      provider: command-line
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: /home/user/transformers-regression/sweeps/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/3
+    choices:
+      benchmark: inference
+      backend: pytorch
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: colorlog
+      hydra/hydra_logging: colorlog
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/3/.config/overrides.yaml

@@ -0,0 +1,2 @@
+- benchmark.input_shapes.batch_size=16
+- backend.torch_dtype=float32

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/3/experiment.log

@@ -0,0 +1,26 @@
+[2023-09-27 12:03:27,204][inference][INFO] - `new_tokens` was set to 200. `max_new_tokens` and `min_new_tokens` will be set to 200.
+[2023-09-27 12:03:28,110][pytorch][INFO] - 	+ Inferred AutoModel class AutoModelForCausalLM for task text-generation and model_type llama
+[2023-09-27 12:03:28,110][backend][INFO] - Configuring pytorch backend
+[2023-09-27 12:03:28,110][backend][INFO] - 	+ Checking initial device(s) isolation of CUDA device(s): [0]
+[2023-09-27 12:03:28,228][backend][INFO] - 	+ Checking continuous device(s) isolation of CUDA device(s): [0]
+[2023-09-27 12:03:28,248][pytorch][INFO] - 	+ Disabling gradients
+[2023-09-27 12:03:28,249][pytorch][INFO] - 	+ Loading model on device: cuda
+[2023-09-27 12:03:28,622][pytorch][INFO] - 	+ Turning on model's eval mode
+[2023-09-27 12:03:28,622][benchmark][INFO] - Configuring inference benchmark
+[2023-09-27 12:03:28,623][inference][INFO] - Running inference benchmark
+[2023-09-27 12:03:28,623][input_generator][INFO] - Using llama model type generator
+[2023-09-27 12:03:28,624][inference][INFO] - 	+ Preparing input for the forward pass
+[2023-09-27 12:03:28,660][inference][INFO] - 	+ Warming up the forward pass
+[2023-09-27 12:03:28,686][inference][INFO] - 	+ Tracking forward pass latency and throughput
+[2023-09-27 12:03:28,686][latency_tracker][INFO] - Tracked Pytorch devices: [0]
+[2023-09-27 12:03:38,803][inference][INFO] - 	+ Forward pass latency: 4.87e-03 (s)
+[2023-09-27 12:03:38,804][inference][INFO] - 	+ Forward pass throughput: 3290.00 (samples/s)
+[2023-09-27 12:03:38,805][inference][INFO] - 	+ Preparing input for the generation pass
+[2023-09-27 12:03:38,805][inference][INFO] - 	+ Warming up the generation pass
+[2023-09-27 12:03:39,433][inference][INFO] - 	+ Tracking generation latency and throughput
+[2023-09-27 12:03:39,433][latency_tracker][INFO] - Tracked Pytorch devices: [0]
+[2023-09-27 12:03:50,151][inference][INFO] - 	+ Generation pass latency: 7.65e-01 (s)
+[2023-09-27 12:03:50,151][inference][INFO] - 	+ Generation pass throughput: 4180.00 (tokens/s)
+[2023-09-27 12:03:50,151][inference][INFO] - Saving inference results
+[2023-09-27 12:03:50,156][backend][INFO] - Cleaning pytorch backend
+[2023-09-27 12:03:50,156][backend][INFO] - 	+ Deleting pretrained model

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/3/hydra_config.yaml

@@ -0,0 +1,79 @@
+backend:
+  name: pytorch
+  version: 2.1.0+rocm5.6
+  _target_: optimum_benchmark.backends.pytorch.backend.PyTorchBackend
+  seed: 42
+  inter_op_num_threads: null
+  intra_op_num_threads: null
+  initial_isolation_check: true
+  continous_isolation_check: true
+  delete_cache: false
+  no_weights: false
+  device_map: null
+  torch_dtype: float32
+  disable_grad: true
+  eval_mode: true
+  amp_autocast: false
+  amp_dtype: null
+  torch_compile: false
+  torch_compile_config: {}
+  bettertransformer: false
+  quantization_scheme: null
+  quantization_config: {}
+  use_ddp: false
+  ddp_config: {}
+  peft_strategy: null
+  peft_config: {}
+benchmark:
+  name: inference
+  _target_: optimum_benchmark.benchmarks.inference.benchmark.InferenceBenchmark
+  duration: 10
+  warmup_runs: 10
+  memory: false
+  energy: false
+  input_shapes:
+    batch_size: 16
+    sequence_length: 200
+    num_choices: 1
+    feature_size: 80
+    nb_max_frames: 3000
+    audio_sequence_length: 16000
+  new_tokens: 200
+  can_diffuse: false
+  can_generate: true
+  forward_kwargs: {}
+  generate_kwargs:
+    max_new_tokens: 200
+    min_new_tokens: 200
+    do_sample: false
+    use_cache: true
+    pad_token_id: 0
+    num_beams: 1
+experiment_name: llama_1gpu_inference
+model: fxmarty/tiny-llama-fast-tokenizer
+device: cuda
+task: text-generation
+hub_kwargs:
+  revision: main
+  cache_dir: null
+  force_download: false
+  local_files_only: false
+environment:
+  optimum_version: 1.13.1
+  transformers_version: 4.34.0.dev0
+  accelerate_version: 0.23.0
+  diffusers_version: null
+  python_version: 3.10.12
+  system: Linux
+  cpu: ' AMD EPYC 7643 48-Core Processor'
+  cpu_count: 96
+  cpu_ram_mb: 1082028
+  gpus:
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/3/inference_results.csv

@@ -0,0 +1,2 @@
+forward.latency(s),forward.throughput(samples/s),generate.latency(s),generate.throughput(tokens/s)
+0.00487,3290.0,0.765,4180.0

raw_results/2023-09-27_10:28:40_946bac798caefada3f5f1c9fecdcfd587ed24ac7/llama_1gpu_inference/multirun.yaml

@@ -0,0 +1,245 @@
+hydra:
+  run:
+    dir: runs/${oc.env:COMMIT_DATE_GMT}_${oc.env:COMMIT_SHA}/${experiment_name}
+  sweep:
+    dir: sweeps/${oc.env:COMMIT_DATE_GMT}_${oc.env:COMMIT_SHA}/${experiment_name}
+    subdir: ${hydra.job.num}
+  launcher:
+    _target_: hydra._internal.core_plugins.basic_launcher.BasicLauncher
+  sweeper:
+    _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
+    max_batch_size: null
+    params:
+      benchmark.input_shapes.batch_size: 1,16
+      backend.torch_dtype: float16,float32
+  help:
+    app_name: ${hydra.job.name}
+    header: '${hydra.help.app_name} is powered by Hydra.
+
+      '
+    footer: 'Powered by Hydra (https://hydra.cc)
+
+      Use --hydra-help to view Hydra specific help
+
+      '
+    template: '${hydra.help.header}
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (group=option)
+
+
+      $APP_CONFIG_GROUPS
+
+
+      == Config ==
+
+      Override anything in the config (foo.bar=value)
+
+
+      $CONFIG
+
+
+      ${hydra.help.footer}
+
+      '
+  hydra_help:
+    template: 'Hydra (${hydra.runtime.version})
+
+      See https://hydra.cc for more info.
+
+
+      == Flags ==
+
+      $FLAGS_HELP
+
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (For example, append hydra/job_logging=disabled
+      to command line)
+
+
+      $HYDRA_CONFIG_GROUPS
+
+
+      Use ''--cfg hydra'' to Show the Hydra config.
+
+      '
+    hydra_help: ???
+  hydra_logging:
+    version: 1
+    formatters:
+      colorlog:
+        (): colorlog.ColoredFormatter
+        format: '[%(cyan)s%(asctime)s%(reset)s][%(purple)sHYDRA%(reset)s] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: colorlog
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+      colorlog:
+        (): colorlog.ColoredFormatter
+        format: '[%(cyan)s%(asctime)s%(reset)s][%(blue)s%(name)s%(reset)s][%(log_color)s%(levelname)s%(reset)s]
+          - %(message)s'
+        log_colors:
+          DEBUG: purple
+          INFO: green
+          WARNING: yellow
+          ERROR: red
+          CRITICAL: red
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: colorlog
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: MULTIRUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=MULTIRUN
+    task: []
+  job:
+    name: experiment
+    chdir: true
+    override_dirname: ''
+    id: ???
+    num: ???
+    config_name: llama2_1gpu_inference
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.3.2
+    version_base: '1.3'
+    cwd: /home/user/transformers-regression
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: optimum_benchmark
+      schema: pkg
+      provider: main
+    - path: hydra_plugins.hydra_colorlog.conf
+      schema: pkg
+      provider: hydra-colorlog
+    - path: /home/user/transformers-regression/configs
+      schema: file
+      provider: command-line
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: ???
+    choices:
+      benchmark: inference
+      backend: pytorch
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: colorlog
+      hydra/hydra_logging: colorlog
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false
+backend:
+  name: pytorch
+  version: ${pytorch_version:}
+  _target_: optimum_benchmark.backends.pytorch.backend.PyTorchBackend
+  seed: 42
+  inter_op_num_threads: null
+  intra_op_num_threads: null
+  initial_isolation_check: true
+  continous_isolation_check: true
+  delete_cache: false
+  no_weights: false
+  device_map: null
+  torch_dtype: null
+  disable_grad: ${is_inference:${benchmark.name}}
+  eval_mode: ${is_inference:${benchmark.name}}
+  amp_autocast: false
+  amp_dtype: null
+  torch_compile: false
+  torch_compile_config: {}
+  bettertransformer: false
+  quantization_scheme: null
+  quantization_config: {}
+  use_ddp: false
+  ddp_config: {}
+  peft_strategy: null
+  peft_config: {}
+benchmark:
+  name: inference
+  _target_: optimum_benchmark.benchmarks.inference.benchmark.InferenceBenchmark
+  duration: 10
+  warmup_runs: 10
+  memory: false
+  energy: false
+  input_shapes:
+    batch_size: 2
+    sequence_length: 200
+    num_choices: 1
+    feature_size: 80
+    nb_max_frames: 3000
+    audio_sequence_length: 16000
+  new_tokens: 200
+  can_diffuse: ${can_diffuse:${task}}
+  can_generate: ${can_generate:${task}}
+  forward_kwargs: {}
+  generate_kwargs: {}
+experiment_name: llama_1gpu_inference
+model: fxmarty/tiny-llama-fast-tokenizer
+device: cuda
+task: ${infer_task:${model}}
+hub_kwargs:
+  revision: main
+  cache_dir: null
+  force_download: false
+  local_files_only: false
+environment:
+  optimum_version: 1.13.1
+  transformers_version: 4.34.0.dev0
+  accelerate_version: 0.23.0
+  diffusers_version: null
+  python_version: 3.10.12
+  system: Linux
+  cpu: ' AMD EPYC 7643 48-Core Processor'
+  cpu_count: 96
+  cpu_ram_mb: 1082028
+  gpus:
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210
+  - Instinct MI210