Skip to content

Instantly share code, notes, and snippets.

View bveeramani's full-sized avatar

Balaji Veeramani bveeramani

  • Anyscale
  • Berkeley, California
  • X @B4110G
View GitHub Profile
@bveeramani
bveeramani / aws.txt
Created April 19, 2023 23:13
read_tfrecords_benchmark
[INFO 2023-04-13 16:18:26,072] anyscale_job_wrapper.py: 256 ### Starting ###
[INFO 2023-04-13 16:18:26,072] anyscale_job_wrapper.py: 280 ### Starting entrypoint ###
[INFO 2023-04-13 16:18:26,072] anyscale_job_wrapper.py: 162 Running command python read_tfrecords_benchmark.py
[INFO 2023-04-13 16:18:26,076] anyscale_job_wrapper.py: 176 Starting process 1561.
+ python read_tfrecords_benchmark.py
2023-04-13 16:18:27,133 INFO worker.py:1315 -- Using address 10.0.14.144:6379 set in the environment variable RAY_ADDRESS
2023-04-13 16:18:27,133 INFO worker.py:1432 -- Connecting to existing Ray cluster at address: 10.0.14.144:6379...
2023-04-13 16:18:27,139 INFO worker.py:1613 -- Connected to Ray cluster. View the dashboard at https://console.anyscale-staging.com/api/v2/sessions/ses_hapzispyymkvr3f7jsmrbi7txz/services?redirect_to=dashboard 
Running benchmark: read-tfrecords
2023-04-13 16:18:28,874 INFO streaming_executor.py:87 -- Executing DAG InputDataBuffer[Input] -> TaskPoolMapOperator[ReadIma
absl-py==1.3.0
accelerate==0.5.1
adal==1.2.7
aiohttp==3.8.3
aiohttp-cors==0.7.0
aiorwlock==1.3.0
aiosignal==1.3.1
ale-py==0.7.5
alembic==1.4.1
anyio==3.6.2
ray on  keras-callbacks [$?] via 🐍 v3.10.8 (.venv)
❯ python repro.py --smoke-test
/Users/balaji/Documents/GitHub/ray/.venv/lib/python3.10/site-packages/pl_bolts/models/self_supervised/amdim/amdim_module.py:35: UnderReviewWarning: The feature generate_power_seq is currently marked under review. The compatibility with other Lightning projects is not guaranteed and API may change at any time. The API and functionality may change without warning in future releases. More details: https://lightning-bolts.readthedocs.io/en/latest/stability.html
"lr_options": generate_power_seq(LEARNING_RATE_CIFAR, 11),
/Users/balaji/Documents/GitHub/ray/.venv/lib/python3.10/site-packages/pl_bolts/models/self_supervised/amdim/amdim_module.py:93: UnderReviewWarning: The feature FeatureMapContrastiveTask is currently marked under review. The compatibility with other Lightning projects is not guaranteed and API may change at any time. The API and functionality may change without warning in future releases. More details: https://ligh
@bveeramani
bveeramani / log.txt
Created December 27, 2022 20:40
linux://python/ray/train:test_base_trainer
=================================== FAILURES ===================================
__________________________ test_reserved_cpu_warnings __________________________
ray_start_4_cpus = RayContext(dashboard_url='127.0.0.1:8265', python_version='3.7.9', ray_version='3.0.0.dev0', ray_commit='{{RAY_COMMIT_...3:62813', 'dashboard_agent_listen_port': 52365, 'node_id': 'b30884dd0998047db0283b31753c2e3ceded2c0be637b6ce93a1d93d'})
mock_tuner_internal_logger = <python.ray.train.tests.test_base_trainer.mock_tuner_internal_logger.<locals>.MockLogger object at 0x7f9140bab590>
def test_reserved_cpu_warnings(ray_start_4_cpus, mock_tuner_internal_logger):
def train_loop(config):
pass
(RayTrainWorker pid=673, ip=172.31.210.32) 2022-11-30 09:33:55.742005: I tensorflow/compiler/jit/xla_gpu_device.cc:99] Not creating XLA devices, tf_xla_enable_xla_devices not set
(RayTrainWorker pid=673, ip=172.31.210.32) 2022-11-30 09:33:55.747631: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1261] Device interconnect StreamExecutor with strength 1 edge matrix:
(RayTrainWorker pid=673, ip=172.31.210.32) 2022-11-30 09:33:55.747656: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1267]
(RayTrainWorker pid=673, ip=172.31.210.32) 2022-11-30 09:33:55.747662: I tensorflow/compiler/jit/xla_gpu_device.cc:99] Not creating XLA devices, tf_xla_enable_xla_devices not set
(RayTrainWorker pid=743, ip=172.31.221.11) 2022-11-30 09:33:55.696012: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcufft.so.10
(RayTrainWorker pid=743, ip=172.31.221.11) 2022-11-30 09:33:55.701059: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened
absl-py==1.3.0
aeppl==0.0.33
aesara==2.7.9
aiohttp==3.8.3
aiohttp-cors==0.7.0
aiorwlock==1.3.0
aiosignal==1.3.1
alabaster==0.7.12
albumentations==1.2.1
altair==4.2.0
import numpy as np
import ray
from ray.data.preprocessors import BatchMapper
ds = ray.data.read_images("/Users/balaji/Datasets/COCO/val2017", mode="RGB", include_paths=True)
ds.fully_executed()
def fn(batch: dict[str, np.ndarray]) -> dict[str, np.ndarray]:
return batch
(_split_single_block pid=8869) Traceback (most recent call last):
(_split_single_block pid=8869) File "python/ray/_raylet.pyx", line 865, in ray._raylet.execute_task
(_split_single_block pid=8869) with ray._private.worker._changeproctitle(title, next_title):
(_split_single_block pid=8869) File "python/ray/_raylet.pyx", line 869, in ray._raylet.execute_task
(_split_single_block pid=8869) outputs = function_executor(*args, **kwargs)
(_split_single_block pid=8869) File "/Users/balaji/Documents/GitHub/ray/python/ray/data/_internal/split.py", line 121, in _split_single_block
(_split_single_block pid=8869) split_block = block_accessor.slice(prev_index, index, copy=True)
(_split_single_block pid=8869) File "/Users/balaji/Documents/GitHub/ray/python/ray/data/_internal/arrow_block.py", line 175, in slice
(_split_single_block pid=8869) view = _copy_table(view)
(_split_single_block pid=8869) File "/Users/balaji/Documents/GitHub/ray/python/ray/data/_internal/arrow_block.py", line 618, in _copy_t
============================= test session starts ==============================
platform darwin -- Python 3.9.12, pytest-7.1.2, pluggy-1.0.0
rootdir: /Users/bveeramani/GitHub/ray/python
plugins: anyio-3.6.1, lazy-fixture-0.6.3
collected 3 items
python/ray/data/tests/test_dataset_formats.py FFF [100%]
=================================== FAILURES ===================================
____________ TestReadImages.test_data_size_estimate[64-RGB-30000-4] ____________
@bveeramani
bveeramani / output.txt
Created September 11, 2022 06:26
read_tf_records bug
============================= test session starts ==============================
platform darwin -- Python 3.9.12, pytest-7.1.2, pluggy-1.0.0
rootdir: /Users/bveeramani/GitHub/ray/python
plugins: anyio-3.6.1, lazy-fixture-0.6.3
collected 1 item
python/ray/data/tests/test_dataset_formats.py F [100%]
=================================== FAILURES ===================================
_____________________________ test_read_tf_records _____________________________