Upload folder using huggingface_hub

README.md

@@ -0,0 +1,48 @@
+## Run info
+
+complete_hash: 53fb0af286257eb95c3298a4ff2761e0
+short_hash: 53fb0761e0
+
+### Configuration
+
+```yaml
+data:
+  batch_size: 32
+  data_seed: 42
+  drop_last: false
+  eval_batch_size: 128
+  max_length: 512
+  multiprocessing_context: null
+  num_workers: 8
+  persistent_workers: false
+  pin_memory: true
+  replacement: false
+  shuffle: true
+dataset: mnli
+estimator:
+  accelerator: gpu
+  convert_to_bettertransformer: false
+  deterministic: true
+  precision: bf16-true
+  tf32_mode: high
+fit:
+  enable_progress_bar: true
+  limit_train_batches: null
+  limit_validation_batches: null
+  log_interval: 100
+  max_epochs: 20
+  optimizer_kwargs:
+    init_kwargs:
+      fused: true
+    lr: 3.0e-05
+    name: adamw
+  scheduler_kwargs:
+    name: constant_schedule_with_warmup
+    num_warmup_steps: 2000
+model:
+  name: bert-base-uncased
+  revision: null
+  seed: 42
+seed: 42
+
+```

hparams.yaml

@@ -0,0 +1,60 @@
+estimator:
+  accelerator: gpu
+  precision: bf16-true
+  deterministic: true
+  tf32_mode: high
+  convert_to_bettertransformer: false
+callbacks:
+  timer:
+    _target_: energizer.active_learning.callbacks.Timer
+  lr_monitor:
+    _target_: energizer.callbacks.lr_monitor.LearningRateMonitor
+  model_checkpoint:
+    _target_: energizer.callbacks.model_checkpoint.ModelCheckpoint
+    dirpath: .checkpoints
+    stage: train
+    frequency: 1:epoch
+loggers:
+  tensorboard:
+    _target_: energizer.loggers.TensorBoardLogger
+    root_dir: ./
+    name: tb_logs
+    version: null
+data:
+  batch_size: 32
+  eval_batch_size: 128
+  shuffle: true
+  replacement: false
+  data_seed: 42
+  drop_last: false
+  num_workers: 8
+  pin_memory: true
+  persistent_workers: false
+  multiprocessing_context: null
+  max_length: 512
+fit:
+  max_epochs: 20
+  optimizer_kwargs:
+    name: adamw
+    lr: 3.0e-05
+    init_kwargs:
+      fused: true
+  scheduler_kwargs:
+    name: constant_schedule_with_warmup
+    num_warmup_steps: 2000
+  log_interval: 100
+  enable_progress_bar: true
+  limit_train_batches: null
+  limit_validation_batches: null
+model:
+  name: bert-base-uncased
+  revision: null
+  seed: 42
+log_interval: 100
+enable_progress_bar: true
+limit_batches: null
+seed: 42
+experiment_group: training
+run_name: bert-base-uncased_2024-02-28T14-37-16
+data_path: /home/pl487/coreset-project/data/processed
+dataset: mnli

tb_logs/version_0/events.out.tfevents.1709131048.dev-gpu-pl487.689954.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86784ec596a8ef052c2556c40735fe550329af65024345d19579f0ac63e138e7
+size 404506

tensorboard_logs.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be203a8ce81ef65a419144cf03ac657d433d12be70c87bad3e51eb3438ddc2ca
+size 40873

train.log

@@ -0,0 +1,77 @@
+[2024-02-28 14:37:17,121][hydra][INFO] - 
+estimator:
+  accelerator: gpu
+  precision: bf16-true
+  deterministic: true
+  tf32_mode: high
+  convert_to_bettertransformer: false
+callbacks:
+  timer:
+    _target_: energizer.active_learning.callbacks.Timer
+  lr_monitor:
+    _target_: energizer.callbacks.lr_monitor.LearningRateMonitor
+  model_checkpoint:
+    _target_: energizer.callbacks.model_checkpoint.ModelCheckpoint
+    dirpath: .checkpoints
+    stage: train
+    frequency: 1:epoch
+loggers:
+  tensorboard:
+    _target_: energizer.loggers.TensorBoardLogger
+    root_dir: ./
+    name: tb_logs
+    version: null
+data:
+  batch_size: 32
+  eval_batch_size: 128
+  shuffle: true
+  replacement: false
+  data_seed: 42
+  drop_last: false
+  num_workers: 8
+  pin_memory: true
+  persistent_workers: false
+  multiprocessing_context: null
+  max_length: 512
+fit:
+  max_epochs: 20
+  optimizer_kwargs:
+    name: adamw
+    lr: 3.0e-05
+    init_kwargs:
+      fused: true
+  scheduler_kwargs:
+    name: constant_schedule_with_warmup
+    num_warmup_steps: 2000
+  log_interval: 100
+  enable_progress_bar: true
+  limit_train_batches: null
+  limit_validation_batches: null
+model:
+  name: bert-base-uncased
+  revision: null
+  seed: 42
+log_interval: 100
+enable_progress_bar: true
+limit_batches: null
+seed: 42
+experiment_group: training
+run_name: bert-base-uncased_2024-02-28T14-37-16
+data_path: /home/pl487/coreset-project/data/processed
+dataset: mnli
+
+======================================================================
+[2024-02-28 14:37:17,122][hydra][INFO] - Seed enabled: 42
+[2024-02-28 14:37:17,496][hydra][INFO] - Label distribution:
+{<RunningStage.TRAIN: 'train'>: {'0-(entailment)': 130899, '1-(neutral)': 130900, '2-(contradiction)': 130903}}
+[2024-02-28 14:37:20,088][hydra][INFO] - Loggers: [<energizer.loggers.tensorboard.TensorBoardLogger object at 0x7efde1ce9ea0>]
+[2024-02-28 14:37:20,088][hydra][INFO] - Callbacks: [<energizer.active_learning.callbacks.Timer object at 0x7efde11b6770>, <energizer.callbacks.lr_monitor.LearningRateMonitor object at 0x7efde11b67d0>, <energizer.callbacks.model_checkpoint.ModelCheckpoint object at 0x7efde11b5cf0>]
+[2024-02-28 14:37:20,091][hydra][INFO] - Model summary:
+Total num params: 109.5M
+Of which trainable: 109.5M
+With a memory footprint of 0.22GB
+Total memory allocated 0.68GB
+[2024-02-28 14:37:20,092][hydra][INFO] - Dataloading params:
+SequenceClassificationDataloaderArgs(batch_size=32, eval_batch_size=128, num_workers=8, pin_memory=True, drop_last=False, persistent_workers=False, shuffle=True, replacement=False, data_seed=42, multiprocessing_context=None, max_length=512)
+[2024-02-28 14:37:20,095][hydra][INFO] - Batch:
+{<InputKeys.INPUT_IDS: 'input_ids'>: tensor([[  101, 17842,   102,  5368,   102]]), <InputKeys.ATT_MASK: 'attention_mask'>: tensor([[1, 1, 1, 1, 1]]), <InputKeys.LABELS: 'labels'>: tensor([0]), <InputKeys.ON_CPU: 'on_cpu'>: {<SpecialKeys.ID: 'uid'>: [278780]}}