COV Current Status as of 11/4/2024
Database
Database Location and Current Database.
The database was located originally on the T7.
There was also a COV database on the home/gqc directory of the MSI that was created later than the T7 database but it exactly the same size so it is likely a copy.
I have moved the database that was in the home directory into a more suitable directory. It now mirrors the SD1 Directory path. '/home/gqc/CCTV/COV/Data/Video_DB/'
Looking at the database
The database is out of date it is not using the gold standard database. Starting with the Gold Standard Database.
Running COV
- Since I need a new database I need to make a full run for the database.
- Start with Gold Standard Database.
- Copy a new database.
- Reformat the Python runner.py so we are not longer running the files by commenting and uncommenting.
- I created 2 new args
- Utility - This sets up to go to the YAML.
- notebook_py_file - this is the step we are on for Creating a database.
- I created 2 new args
- Run Notebooks for COV.
- Make sure Video Groups is being handled. That should also be an argument?
- Start with Gold Standard Database.
- Create the Model
- Run Test and Train.
- Show what I have got in Streamlit
Notes on COV
Copy the Gold Standard database and place it /media/gqc/unionsine2/cctv/cov/Video_DB/
Rename the previous database to Video_DB_COV_Before_11_5_24.db
Rename the Gold Standard database to Video_DB_COV.db
Following the instructions from CCTV_COV_Workflow
We are skipping notebooks 1-6
- Used the Video List CSV files to fill inn the Videos table.
- Populated All Conditions Table from SD1 All Conditions Table.
- Populated the following from the Video_DB_COV_Before_11_5_24.db
- Conditions Standard table
- Video Type
- Video Group
- Frames table
- We could not copy the frames table due to the previous database only containing one video group.
- We did not want to Run Notebook 5. We had the resulting Blurred images already out in a ziped file. We used the file structure of these zip files to fill the Frames table. A Python file (
populate_frames_using_existing_files.py) was created to scrape the data from the file storage and fill the following columns of the frames table.- Video id
- frame id
- video Group
- dataset
- We also had the ORC_JSON files from a previous run. A second python file was generated (
populate_ocr_data.py) to scrape this data into a separate orc table. :::warning The ORC generated for COV was an earlier form where we were getting an all_text section and a textAnnotaions section. The current implementation uses text_and_bb_list instead and it is the textAnnotations renamed. So when getting the JSON Files data we are throwing out the all_text section and renaming the textAnnotations section. ::: {"all_text": "-1.60 m\nRPT: 4471 FR:396464 TO: 398817\nPSR:\nFJBZOF USE: CB SIZE: 200mm\n", "textAnnotations": [["-1.60 m", [81.0, 43.0, 162.0, 44.0, 162.0, 60.0, 81.0, 59.0]], ["RPT: 4471 FR:396464 TO: 398817", [33.0, 78.0, 380.0, 78.0, 380.0, 94.0, 33.0, 94.0]], ["PSR:", [33.0, 97.0, 79.0, 97.0, 79.0, 111.0, 33.0, 111.0]], ["FJBZOF USE: CB SIZE: 200mm", [113.0, 96.0, 417.0, 96.0, 417.0, 112.0, 113.0, 111.0]]]} {"text_and_bb_list": [["MH 0890045", [104.0, 38.0, 231.0, 37.0, 231.0, 59.0, 104.0, 59.0]], ["TO", [310.0, 41.0, 333.0, 40.0, 335.0, 56.0, 312.0, 57.0]], ["MH 0890043", [421.0, 37.0, 549.0, 34.0, 549.0, 58.0, 421.0, 60.0]], ["FT", [207.0, 399.0, 235.0, 399.0, 234.0, 416.0, 207.0, 416.0]], ["0 FPM", [411.0, 418.0, 477.0, 413.0, 478.0, 438.0, 414.0, 441.0]]]} The ocr table was then updated by using this sql Statement.So that brings us to notebook 5. Notebook 6 is about creating blurred images. We already have blurred images atUPDATE frames as f
Set
OCR_JSON = o.OCR_JSON
From ocr as ocr
where f.frame_id = o.frame_id/data0/COV_blurred_frames_unzipped_imagesThese images were originally from/media/gqc/unionsine2/cctv/cov/Data/COV_A/blurred_frames_oldA new BASH script was writtenextract_all_image_files.shand is saved in the compute msi project. This script is diffrent as it dumps all the images into 1 folder with no suborders at the end of the process.
- We did not want to Run Notebook 5. We had the resulting Blurred images already out in a ziped file. We used the file structure of these zip files to fill the Frames table. A Python file (
Notebook 7 This pulls the distance out of the OCR_JSON string. We use a regex expression based on what type of video it is.
distance_field_regex = r'(?<=^)-?\d?\d?\d?\d ?[\.,]? ?[\d@O][\d@O](?= ?[Mm])'
For more information on the Regex expression used please view in this page.
Video Types
Bounding Boxes
Bounding boxes are how we find the text in an image. After we have the OCR_JSON text we have the bounding boxes. This is an example of what we have in the OCR JSON
{"text_and_bb_list": [["0.89 m", [71.0, 23.0, 131.0, 24.0, 131.0, 40.0, 71.0, 38.0]], ["CL (Crack Longitudinal )", [41.0, 123.0, 268.0, 123.0, 268.0, 139.0, 41.0, 138.0]], ["FROM: 02", [42.0, 140.0, 122.0, 140.0, 122.0, 156.0, 42.0, 155.0]], ["TO :", [42.0, 158.0, 83.0, 158.0, 82.0, 171.0, 42.0, 171.0]], ["REMARKS :", [42.0, 173.0, 119.0, 173.0, 119.0, 187.0, 42.0, 187.0]], ["PHOTO: 0001", [42.0, 190.0, 152.0, 190.0, 152.0, 205.0, 42.0, 205.0]]]}
A bounding box starts at the left hand corner with width and height and goes around in a clockwise circle
Select Rectangle is defined by the top left corner, the width, and height. The units are in pixels. TODO: Rectangle_Distance and Rectangle_Condition each with width and height. The ids will be added into the videos table. distance_rectangle Same as SD1 distance bounding box columns: id, ... condition_rectangle Similar to distance_rectangle columns: id, ...
CREATE TABLE "distance_rectangle" (
id INTEGER PRIMARY KEY AUTOINCREMENT,
distance_rectangle_top_left_x INTEGER,
distance_rectangle_top_left_y INTEGER,
distance_rectangle_width INTEGER,
distance_rectangle_height INTEGER,
CONSTRAINT unique_distance_box_combination
UNIQUE (distance_rectangle_top_left_x, distance_rectangle_top_left_y, distance_rectangle_width, distance_rectangle_height)
);
TODO: Rename SD1_distance_bounding_box to distance_rectangle.
All of this will get added into the videos table.
If the Top left hand corner of the bounding box exists in the selected Rectangle then it looks at the text in that section of the OCR_JSON.
Populate the Video Types table and then update videos table based on the video Type.
Distance Extraction
It compares the text against a Regex expression based on the video type of that video. TODO: Add Regex expressions to the video Types table. regex_distance Video Type gets changed to regex_distance table. ID and Regex are the 2 columns and this id will be populated into the videos table as regex_distance_id.
CREATE TABLE "regex_distance" (id INTEGER PRIMARY KEY AUTOINCREMENT, regex TEXT, PRIMARY KEY(id));
Condition Code Extraction
Look Through the Logic to see if we need a Table of condition Code Regex Table. Look at Metro Logic.
TODO: Possibly combine condition code extraction and distance extraction.
regex_condition Video Type gets changed to regex_condition table. ID and Regex are the 2 columns and this id will be populated into the videos table as regex_condition_id.
CREATE TABLE "condition_rectangle" (
id INTEGER PRIMARY KEY AUTOINCREMENT,
condition_rectangle_top_left_x INTEGER,
condition_rectangle_top_left_y INTEGER,
condition_rectangle_width INTEGER,
condition_rectangle_height INTEGER,
CONSTRAINT unique_distance_box_combination
UNIQUE (condition_rectangle_top_left_x, condition_rectangle_top_left_y, condition_rectangle_width, condition_rectangle_height)
);
Process: Modify the notebooks. Export the python files from the notebooks using nbdev_export Run Python Scripts. Clean the code with nbdev_clean. Commit the code.
Revised Videos Table regex_distance_id regex_condition_id "SELECT videos.dataset, videos.video_group, videos.video_id, distance_bounding_boxes.width, distance_bounding_boxes.height, distance_bounding_boxes.d_top_left_x, distance_bounding_boxes.d_top_left_y, distance_bounding_boxes.d_box_width, distance_bounding_boxes.d_box_height FROM videos JOIN distance_bounding_boxes ON videos.distance_bounding_box_id = distance_bounding_boxes.distance_bounding_box_id WHERE videos.dataset = 'COV_A' AND video_group = 'COV_1_to_11_videos';"
Notebook 8 Interpolation
We are using Pandas interpolate method to interpolate the data.
new_distance_df.interpolate(method='linear', limit_direction='forward', limit_area='inside')
We are getting some Null values at the end of videos. And only at the end of videos. It is happening at 2 videos.
video 001626-FJBXW0-D-2018-02-05 has 14 nulls at the end of the video
Video 004464-FJBZ0W-U-2021-12-19-P2
29700.png' SELECT FROM frames WHERE video_id = '001626-__FJBXW0-D-2018-02-05' SELECT FROM frames WHERE video_id ='004464-__FJBZ0W-U-2021-12-19-P2'
I saw that we were using Pandas Interpolate so asked Chat to give me an explanation of the code.
The section that is effecting this is the limit_area.
'inside': Only fills NaN values if surrounded by non-NaN values. 'outside': Fills NaN values at the edges as well.
So since we are using inside and there are not non-NaN values after the NaN values it is not filling in.
The .interpolate() method in pandas is used to fill missing values (NaN) in a DataFrame by estimating them using interpolation techniques. It's a powerful tool for cleaning and completing data where values are missing.
Here's a breakdown of how it works, specifically for DataFrames:
Key Parameters method:
Specifies the interpolation technique. Common options include: 'linear' (default): Interpolates linearly between missing and known values. 'time': Interpolates based on time indices (useful for time-series data). 'polynomial': Fits a polynomial of a specified order and uses it to estimate missing values. Other advanced methods include 'nearest', 'quadratic', and 'cubic'. limit_direction:
Specifies the direction of filling. 'forward': Fills NaN values using known values from before the NaN. 'backward': Fills NaN values using known values from after the NaN. 'both': Considers both forward and backward directions. limit_area:
Defines the region where interpolation is allowed. 'inside': Only fills NaN values if surrounded by non-NaN values. 'outside': Fills NaN values at the edges as well. axis:
Determines whether interpolation is applied column-wise (axis=0, default) or row-wise (axis=1). limit:
Maximum number of consecutive NaN values to fill.
Notebook 9 and why it is not working:
So the line where it is not working is
matching_des = condition_Abbr_df[condition_Abbr_df.shortened_description.eq(st)]
st is the string and we are truncating it to the first 17 char. st = st[:17] # Truncate to first 17 characters The reason for the truncation is that the values are truncated in the dataframe.
1 Access Point Clea 2 Access Point Clea 3 Access Point Clea 4 Access Point Disc ... 216 Weld Feature Long 217 Weld Feature Mult 218 Weld Feature Spir 219 Weld Feature Othe
But the ST is 'CL (Crack Longitudinal )' and truncating that to the first 17 gives us 'CL (Crack Longitu'
Why are we looking at the short description. We should be using the code which would be CL in this case we can easily split on a space or '(' char I need to look at the database and see what kind of values I need I might have to do a regex.
Are the values code(description) or code (description) in the OCR_JSON?
No they are not sometimes it is code (description), code(Description), codeCdescription, codedescription,
Is the code always there? What are the standards of the Code? 3 - 5 letters, All caps
Notebook 9
This extracts the condition code from the OCR and put it into the condition code list and then set the Condition code list updated from 0 to 1.
Notebook 10 took about 8 min to run with COV. In this notebook we take the report data from the OCR and the inspection data from all conditions table and merge the two in the labels column. Notebook 3 fills the All Conditions table from a CSV. It is part of the supplemental table population
Database Changes these changes need to be backfilled into the Gold Standard:
- Created a Frame Height and Width for each video added into the videos table.
- Created the following tables:
- distance_rectangle - This contains the search rectangle for the distance bounding boxes.
- condition_code_rectangle - This contains the search rectangle for the condition code bounding boxes.
- regex_distance - This should be changed to distance_regex. It stores the regex used in distance notebook aka notebook 8.
- In the video table there are id's for the 3 tables in item number 2.
Running Training
Go into the Settings.py Change everything so that it points to the correct database and the correct
Then run the following script ./localrunner.sh -s fastai_multi_label_v3_generic -r cov_all_video_groups_MC -b 12 -e 10
Running Testing
The settings are the same for the Testing section. It is settings.py So that is good and yes I verified it.
Run this script. ./localrunner.sh -s fastai_multi_label_predict -r 123 -b 64 -m cctv-multilabel-cov_all_video_groups_MC
Set up for Streamlit.
./localrunner.sh -s fastai_multi_label_predict -r 123 -b 64 -m cctv-multilabel-all_video_groups_MC
SD1 Label Images with drop written in on label
drop is not in the Condtion Standards.
Drop is all over the file here are some of the examples. I would need to sit in the debugger to get a handle on this code.
Results of new Model.
Tue Dec 3 04:17:27 PM EST 2024 [326/351]
wandb: Currently logged in as: deven-gqc (lence_ubc). Use `wandb login --relogin` to force relogin
wandb: wandb version 0.18.7 is available! To upgrade, please run:
wandb: $ pip install wandb --upgrade
wandb: Tracking run with wandb version 0.17.4
wandb: Run data is saved locally in /home/gqc/git/gqc/compute-msi/sewer_ml/wandb/run-20241203_161729-mxec600b
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run cov_all_video_groups_MC_12_3_24_2024/12/03_16:17
wandb: ⭐️ View project at https://wandb.ai/lence_ubc/cctv-cov-multilabel
wandb: 🚀 View run at https://wandb.ai/lence_ubc/cctv-cov-multilabel/runs/mxec600b
wandb:
wandb:
wandb: Run history:
wandb: F1 Average ▁▃▃▄▄▅▅▆▇█
wandb: F1 validation ▁
wandb: F1 validation B ▁
wandb: F1 validation CC ▁
wandb: F1 validation CH2 ▁
wandb: F1 validation CL ▁
wandb: F1 validation CM ▁
wandb: F1 validation CS ▁
wandb: F1 validation DAE ▁
wandb: F1 validation DAGS ▁
wandb: F1 validation DD ▁
wandb: F1 validation DNF ▁
wandb: F1 validation DSF ▁
wandb: F1 validation DSGV ▁
wandb: F1 validation DSZ ▁
wandb: F1 validation FC ▁
wandb: F1 validation FL ▁ [297/351]
wandb: F1 validation FM ▁
wandb: F1 validation FS ▁
wandb: F1 validation H ▁
wandb: F1 validation HSV ▁
wandb: F1 validation HVV ▁
wandb: F1 validation ID ▁
wandb: F1 validation IG ▁
wandb: F1 validation IR ▁
wandb: F1 validation ISZ ▁
wandb: F1 validation JAL ▁
wandb: F1 validation JAM ▁
wandb: F1 validation JOL ▁
wandb: F1 validation JOM ▁
wandb: F1 validation LL ▁
wandb: F1 validation LR ▁
wandb: F1 validation MGO ▁
wandb: F1 validation MGP ▁
wandb: F1 validation MLC ▁
wandb: F1 validation MMC ▁
wandb: F1 validation MSC ▁
wandb: F1 validation MWL ▁
wandb: F1 validation MWLS ▁
wandb: F1 validation ND ▁
wandb: F1 validation OBM ▁
wandb: F1 validation OBR ▁
wandb: F1 validation OBZ ▁
wandb: F1 validation RBJ ▁
wandb: F1 validation RFB ▁
wandb: F1 validation RFC ▁
wandb: F1 validation RFJ ▁
wandb: F1 validation RFL ▁
wandb: F1 validation RMB ▁
wandb: F1 validation RMC ▁
wandb: F1 validation RMJ ▁
wandb: F1 validation RPL ▁
wandb: F1 validation RPP ▁
wandb: F1 validation RPR ▁
wandb: F1 validation RTB ▁
wandb: F1 validation SRP ▁
wandb: F1 validation SRV ▁
wandb: F1 validation SSS ▁
wandb: F1 validation TB ▁
wandb: F1 validation TBA ▁
wandb: F1 validation TBD ▁
wandb: F1 validation TBI ▁
wandb: F1 validation TF ▁
wandb: F1 validation TFA ▁
wandb: F1 validation TFC ▁
wandb: F1 validation TFD ▁
wandb: F1 validation TS ▁
wandb: F1 validation TSA ▁
wandb: F1 validation TSD ▁
wandb: F1 validation TSI ▁
wandb: F1 validation average no ND ▁
wandb: F1 validation normal ▁
wandb: F2 validation ▁
wandb: F2 validation B ▁
wandb: F2 validation CC ▁
wandb: F2 validation CH2 ▁ [238/351]
wandb: F2 validation CL ▁
wandb: F2 validation CM ▁
wandb: F2 validation CS ▁
wandb: F2 validation DAE ▁
wandb: F2 validation DAGS ▁
wandb: F2 validation DD ▁
wandb: F2 validation DNF ▁
wandb: F2 validation DSF ▁
wandb: F2 validation DSGV ▁
wandb: F2 validation DSZ ▁
wandb: F2 validation FC ▁
wandb: F2 validation FL ▁
wandb: F2 validation FM ▁
wandb: F2 validation FS ▁
wandb: F2 validation H ▁
wandb: F2 validation HSV ▁
wandb: F2 validation HVV ▁
wandb: F2 validation ID ▁
wandb: F2 validation IG ▁
wandb: F2 validation IR ▁
wandb: F2 validation ISZ ▁
wandb: F2 validation JAL ▁
wandb: F2 validation JAM ▁
wandb: F2 validation JOL ▁
wandb: F2 validation JOM ▁
wandb: F2 validation LL ▁
wandb: F2 validation LR ▁
wandb: F2 validation MGO ▁
wandb: F2 validation MGP ▁
wandb: F2 validation MLC ▁ [208/351]
wandb: F2 validation MMC ▁
wandb: F2 validation MSC ▁
wandb: F2 validation MWL ▁
wandb: F2 validation MWLS ▁
wandb: F2 validation ND ▁
wandb: F2 validation OBM ▁
wandb: F2 validation OBR ▁
wandb: F2 validation OBZ ▁
wandb: F2 validation RBJ ▁
wandb: F2 validation RFB ▁
wandb: F2 validation RFC ▁
wandb: F2 validation RFJ ▁
wandb: F2 validation RFL ▁
wandb: F2 validation RMB ▁
wandb: F2 validation RMC ▁
wandb: F2 validation RMJ ▁
wandb: F2 validation RPL ▁
wandb: F2 validation RPP ▁
wandb: F2 validation RPR ▁
wandb: F2 validation RTB ▁
wandb: F2 validation SRP ▁
wandb: F2 validation SRV ▁
wandb: F2 validation SSS ▁
wandb: F2 validation TB ▁
wandb: F2 validation TBA ▁
wandb: F2 validation TBD ▁
wandb: F2 validation TBI ▁
wandb: F2 validation TF ▁
wandb: F2 validation TFA ▁
wandb: F2 validation TFC ▁ [178/351]
wandb: F2 validation TFD ▁
wandb: F2 validation TS ▁
wandb: F2 validation TSA ▁
wandb: F2 validation TSD ▁
wandb: F2 validation TSI ▁
wandb: F2 validation average no ND ▁
wandb: accuracy_multi ▁▃▄▄▄▆▅▇▇█
wandb: epoch ▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███
wandb: eps_0 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: eps_1 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: eps_2 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: learning rate ▁
wandb: lr_0 ████████████████████████████████▇▆▅▄▃▂▁▁
wandb: lr_1 ████████████████████████████████▇▆▅▄▃▂▁▁
wandb: lr_2 ████████████████████████████████▇▆▅▄▃▂▁▁
wandb: mom_0 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: mom_1 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: mom_2 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: raw_loss ██▃▃▃▄▃▄▃▆▆▅▄▂▄▃▅▂▃▃▃▄▂▁▅▃▂▂▄▅▂▂▂▃▁▂▂▂▂▁
wandb: sqr_mom_0 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: sqr_mom_1 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: sqr_mom_2 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: train_loss █▇▆▅▅▅▄▄▄▄▄▄▃▃▃▃▃▃▃▄▃▂▃▃▃▃▃▂▃▃▃▂▂▂▂▂▁▁▁▁
wandb: train_samples_per_sec ▆▂▅▄▃▂▇▅▄█▂▃▄▃▃▃▃▅█▄▁▂▁▄▄▂▄▃▃▆▂▃▁▂▃▃▂▄▄▂
wandb: valid_loss █▆▅▄▅▃▃▄▁▃
wandb: wd_0 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: wd_1 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: wd_2 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb:
wandb: Run summary: [148/351]
wandb: F1 Average 0.26179
wandb: F1 validation 0.89492
wandb: F1 validation B 0.0
wandb: F1 validation CC 0.77876
wandb: F1 validation CH2 0.0
wandb: F1 validation CL 0.52632
wandb: F1 validation CM 0.28571
wandb: F1 validation CS 0.0
wandb: F1 validation DAE 0.84995
wandb: F1 validation DAGS 0.93806
wandb: F1 validation DD 0.23087
wandb: F1 validation DNF 0.0
wandb: F1 validation DSF 0.87395
wandb: F1 validation DSGV 0.82832
wandb: F1 validation DSZ 0.9502
wandb: F1 validation FC 0.1875
wandb: F1 validation FL 0.0
wandb: F1 validation FM 0.0
wandb: F1 validation FS 0.0
wandb: F1 validation H 0.0
wandb: F1 validation HSV 0.0
wandb: F1 validation HVV 0.0
wandb: F1 validation ID 0.0
wandb: F1 validation IG 0.26087
wandb: F1 validation IR 0.0
wandb: F1 validation ISZ 0.0
wandb: F1 validation JAL 0.47059
wandb: F1 validation JAM 0.0
wandb: F1 validation JOL 0.0
wandb: F1 validation JOM 0.0 [118/351]
wandb: F1 validation LL 0.58537
wandb: F1 validation LR 0.74627
wandb: F1 validation MGO 0.0
wandb: F1 validation MGP 0.65306
wandb: F1 validation MLC 0.0
wandb: F1 validation MMC 0.0
wandb: F1 validation MSC 0.0
wandb: F1 validation MWL 0.61017
wandb: F1 validation MWLS 0.81333
wandb: F1 validation ND 0.95008
wandb: F1 validation OBM 0.0
wandb: F1 validation OBR 0.0
wandb: F1 validation OBZ 0.0
wandb: F1 validation RBJ 0.0
wandb: F1 validation RFB 0.0
wandb: F1 validation RFC 0.0
wandb: F1 validation RFJ 0.7951
wandb: F1 validation RFL 0.0
wandb: F1 validation RMB 0.0
wandb: F1 validation RMC 0.0
wandb: F1 validation RMJ 0.83577
wandb: F1 validation RPL 0.0
wandb: F1 validation RPP 0.0
wandb: F1 validation RPR 0.36364
wandb: F1 validation RTB 0.0
wandb: F1 validation SRP 0.0
wandb: F1 validation SRV 0.0
wandb: F1 validation SSS 0.0
wandb: F1 validation TB 0.04301
wandb: F1 validation TBA 0.0 [88/351]
wandb: F1 validation TBD 0.0
wandb: F1 validation TBI 0.0
wandb: F1 validation TF 0.87811
wandb: F1 validation TFA 0.64179
wandb: F1 validation TFC 0.41791
wandb: F1 validation TFD 0.47059
wandb: F1 validation TS 0.88983
wandb: F1 validation TSA 0.29358
wandb: F1 validation TSD 0.0
wandb: F1 validation TSI 0.0
wandb: F1 validation average no ND 0.25167
wandb: F1 validation normal 0.95008
wandb: F2 validation 0.88255
wandb: F2 validation B 0.0
wandb: F2 validation CC 0.79137
wandb: F2 validation CH2 0.0
wandb: F2 validation CL 0.43554
wandb: F2 validation CM 0.26316
wandb: F2 validation CS 0.0
wandb: F2 validation DAE 0.80114
wandb: F2 validation DAGS 0.93665
wandb: F2 validation DD 0.15862
wandb: F2 validation DNF 0.0
wandb: F2 validation DSF 0.84829
wandb: F2 validation DSGV 0.77725
wandb: F2 validation DSZ 0.93871
wandb: F2 validation FC 0.15789
wandb: F2 validation FL 0.0
wandb: F2 validation FM 0.0
wandb: F2 validation FS 0.0 [58/351]
wandb: F2 validation H 0.0
wandb: F2 validation HSV 0.0
wandb: F2 validation HVV 0.0
wandb: F2 validation ID 0.0
wandb: F2 validation IG 0.25424
wandb: F2 validation IR 0.0
wandb: F2 validation ISZ 0.0
wandb: F2 validation JAL 0.45455
wandb: F2 validation JAM 0.0
wandb: F2 validation JOL 0.0
wandb: F2 validation JOM 0.0
wandb: F2 validation LL 0.59406
wandb: F2 validation LR 0.76687
wandb: F2 validation MGO 0.0
wandb: F2 validation MGP 0.69565
wandb: F2 validation MLC 0.0
wandb: F2 validation MMC 0.0
wandb: F2 validation MSC 0.0
wandb: F2 validation MWL 0.56962
wandb: F2 validation MWLS 0.76441
wandb: F2 validation ND 0.96241
wandb: F2 validation OBM 0.0
wandb: F2 validation OBR 0.0
wandb: F2 validation OBZ 0.0
wandb: F2 validation RBJ 0.0
wandb: F2 validation RFB 0.0
wandb: F2 validation RFC 0.0
wandb: F2 validation RFJ 0.73739
wandb: F2 validation RFL 0.0
wandb: F2 validation RMB 0.0 [28/351]
wandb: F2 validation RMC 0.0
wandb: F2 validation RMJ 0.79305
wandb: F2 validation RPL 0.0
wandb: F2 validation RPP 0.0
wandb: F2 validation RPR 0.32
wandb: F2 validation RTB 0.0
wandb: F2 validation SRP 0.0
wandb: F2 validation SRV 0.0
wandb: F2 validation SSS 0.0
wandb: F2 validation TB 0.02825
wandb: F2 validation TBA 0.0
wandb: F2 validation TBD 0.0
wandb: F2 validation TBI 0.0
wandb: F2 validation TF 0.86743
wandb: F2 validation TFA 0.56136
wandb: F2 validation TFC 0.33898
wandb: F2 validation TFD 0.42553
wandb: F2 validation TS 0.90361
wandb: F2 validation TSA 0.22161
wandb: F2 validation TSD 0.0
wandb: F2 validation TSI 0.0
wandb: F2 validation average no ND 0.23953
wandb: accuracy_multi 0.99576
wandb: epoch 10
wandb: eps_0 1e-05
wandb: eps_1 1e-05
wandb: eps_2 1e-05
wandb: learning rate 0.001
wandb: lr_0 0.0
wandb: lr_1 0.0
wandb: lr_2 0.0
wandb: mom_0 0.9
wandb: mom_1 0.9
wandb: mom_2 0.9
wandb: raw_loss 0.01343
wandb: sqr_mom_0 0.99
wandb: sqr_mom_1 0.99
wandb: sqr_mom_2 0.99
wandb: train_loss 0.01391
wandb: train_samples_per_sec 45.52563
wandb: valid_loss 0.01673
wandb: wd_0 0.01
wandb: wd_1 0.01
wandb: wd_2 0.01
wandb:
wandb: 🚀 View run cov_all_video_groups_MC_12_3_24_2024/12/03_16:17 at: https://wandb.ai/lence_ubc/cctv-cov-multilabel/runs/mxec600b
wandb: ⭐️ View project at: https://wandb.ai/lence_ubc/cctv-cov-multilabel
wandb: Synced 5 W&B file(s), 0 media file(s), 2 artifact file(s) and 0 other file(s)
wandb: Find logs at: ./wandb/run-20241203_161729-mxec600b/logs
wandb: WARNING The new W&B backend becomes opt-out in version 0.18.0; try it out with `wandb.require("core")`! See https://wandb.me/wandb
-core for more information.
/usr/lib/python3.10/tempfile.py:1008: ResourceWarning: Implicitly cleaning up <TemporaryDirectory '/tmp/tmpr4ocp8su'>
_warnings.warn(warn_message, ResourceWarning)
Tue Dec 3 09:48:10 PM EST 2024
Finished
Started prediction.
Wed Dec 4 12:38:26 PM EST 2024
wandb: Currently logged in as: deven-gqc (lence_ubc). Use `wandb login --relogin` to force relogin
wandb: wandb version 0.18.7 is available! To upgrade, please run:
wandb: $ pip install wandb --upgrade
wandb: Tracking run with wandb version 0.17.4
wandb: Run data is saved locally in /home/gqc/git/gqc/compute-msi/sewer_ml/wandb/run-20241204_123828-j108u632
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run 123_2024/12/04_12:38
wandb: ⭐️ View project at https://wandb.ai/lence_ubc/cctv-cov-single-stage-model
wandb: 🚀 View run at https://wandb.ai/lence_ubc/cctv-cov-single-stage-model/runs/j108u632
Wed Dec 4 03:51:53 PM EST 2024
wandb: Currently logged in as: deven-gqc (lence_ubc). Use `wandb login --relogin` to force relogin
wandb: wandb version 0.18.7 is available! To upgrade, please run:
wandb: $ pip install wandb --upgrade
wandb: Tracking run with wandb version 0.17.4
wandb: Run data is saved locally in /home/gqc/git/gqc/compute-msi/sewer_ml/wandb/run-20241204_155156-zp2gcfhh
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run cov_all_video_groups_MC_12_4_24_2024/12/04_15:51
wandb: ⭐️ View project at https://wandb.ai/lence_ubc/cctv-cov-multilabel
wandb: 🚀 View run at https://wandb.ai/lence_ubc/cctv-cov-multilabel/runs/zp2gcfhh
wandb:
wandb:
wandb: Run history:
wandb: F1 Average ▁▂▂▃▄▅▄▅▇█
wandb: F1 validation ▁
wandb: F1 validation B ▁
wandb: F1 validation CC ▁
wandb: F1 validation CH2 ▁
wandb: F1 validation CL ▁
wandb: F1 validation CM ▁
wandb: F1 validation CS ▁
wandb: F1 validation DAE ▁
wandb: F1 validation DAGS ▁
wandb: F1 validation DNF ▁
wandb: F1 validation DSF ▁
wandb: F1 validation DSGV ▁
wandb: F1 validation DSZ ▁
wandb: F1 validation FC ▁
wandb: F1 validation FL ▁
wandb: F1 validation FM ▁
wandb: F1 validation FS ▁
wandb: F1 validation H ▁
wandb: F1 validation HSV ▁
wandb: F1 validation HVV ▁
wandb: F1 validation ID ▁ [888/937]
wandb: F1 validation IG ▁
wandb: F1 validation IR ▁
wandb: F1 validation ISZ ▁
wandb: F1 validation JAL ▁
wandb: F1 validation JAM ▁
wandb: F1 validation JOL ▁
wandb: F1 validation JOM ▁
wandb: F1 validation LL ▁
wandb: F1 validation LR ▁
wandb: F1 validation MGO ▁
wandb: F1 validation MGP ▁
wandb: F1 validation MLC ▁
wandb: F1 validation MMC ▁
wandb: F1 validation MSC ▁
wandb: F1 validation MWL ▁
wandb: F1 validation MWLS ▁
wandb: F1 validation ND ▁
wandb: F1 validation OBM ▁
wandb: F1 validation OBR ▁
wandb: F1 validation OBZ ▁
wandb: F1 validation RBJ ▁
wandb: F1 validation RFB ▁
wandb: F1 validation RFC ▁
wandb: F1 validation RFJ ▁
wandb: F1 validation RFL ▁
wandb: F1 validation RMB ▁
wandb: F1 validation RMC ▁
wandb: F1 validation RMJ ▁
wandb: F1 validation RPL ▁
wandb: F1 validation RPP ▁
wandb: F1 validation RPR ▁
wandb: F1 validation RTB ▁
wandb: F1 validation SRP ▁
wandb: F1 validation SRV ▁
wandb: F1 validation SSS ▁ [853/937]
wandb: F1 validation TB ▁
wandb: F1 validation TBA ▁
wandb: F1 validation TBD ▁
wandb: F1 validation TBI ▁
wandb: F1 validation TF ▁
wandb: F1 validation TFA ▁
wandb: F1 validation TFC ▁
wandb: F1 validation TFD ▁
wandb: F1 validation TS ▁
wandb: F1 validation TSA ▁
wandb: F1 validation TSD ▁
wandb: F1 validation TSI ▁
wandb: F1 validation average no ND ▁
wandb: F1 validation normal ▁
wandb: F2 validation ▁
wandb: F2 validation B ▁
wandb: F2 validation CC ▁
wandb: F2 validation CH2 ▁
wandb: F2 validation CL ▁
wandb: F2 validation CM ▁
wandb: F2 validation CS ▁
wandb: F2 validation DAE ▁
wandb: F2 validation DAGS ▁
wandb: F2 validation DNF ▁
wandb: F2 validation DSF ▁
wandb: F2 validation DSGV ▁
wandb: F2 validation DSZ ▁
wandb: F2 validation FC ▁
wandb: F2 validation FL ▁
wandb: F2 validation FM ▁
wandb: F2 validation FS ▁
wandb: F2 validation H ▁
wandb: F2 validation HSV ▁
wandb: F2 validation HVV ▁
wandb: F2 validation ID ▁ [818/937]
wandb: F2 validation IG ▁
wandb: F2 validation IR ▁
wandb: F2 validation ISZ ▁
wandb: F2 validation JAL ▁
wandb: F2 validation JAM ▁
wandb: F2 validation JOL ▁
wandb: F2 validation JOM ▁
wandb: F2 validation LL ▁
wandb: F2 validation LR ▁
wandb: F2 validation MGO ▁
wandb: F2 validation MGP ▁
wandb: F2 validation MLC ▁
wandb: F2 validation MMC ▁
wandb: F2 validation MSC ▁
wandb: F2 validation MWL ▁
wandb: F2 validation MWLS ▁
wandb: F2 validation ND ▁
wandb: F2 validation OBM ▁
wandb: F2 validation OBR ▁
wandb: F2 validation OBZ ▁
wandb: F2 validation RBJ ▁
wandb: F2 validation RFB ▁
wandb: F2 validation RFC ▁
wandb: F2 validation RFJ ▁
wandb: F2 validation RFL ▁
wandb: F2 validation RMB ▁
wandb: F2 validation RMC ▁
wandb: F2 validation RMJ ▁
wandb: F2 validation RPL ▁
wandb: F2 validation RPP ▁
wandb: F2 validation RPR ▁
wandb: F2 validation RTB ▁
wandb: F2 validation SRP ▁
wandb: F2 validation SRV ▁
wandb: F2 validation SSS ▁ [783/937]
wandb: F2 validation TB ▁
wandb: F2 validation TBA ▁
wandb: F2 validation TBD ▁
wandb: F2 validation TBI ▁
wandb: F2 validation TF ▁
wandb: F2 validation TFA ▁
wandb: F2 validation TFC ▁
wandb: F2 validation TFD ▁
wandb: F2 validation TS ▁
wandb: F2 validation TSA ▁
wandb: F2 validation TSD ▁
wandb: F2 validation TSI ▁
wandb: F2 validation average no ND ▁
wandb: accuracy_multi ▁▂▃▃▄▅▅▆▇█
wandb: epoch ▁▁▁▂▂▂▂▂▂▃▃▃▃▃▄▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▇▇▇▇▇▇███
wandb: eps_0 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: eps_1 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: eps_2 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: learning rate ▁
wandb: lr_0 ████████████████████████████████▇▆▅▄▃▂▁▁
wandb: lr_1 ████████████████████████████████▇▆▅▄▃▂▁▁
wandb: lr_2 ████████████████████████████████▇▆▅▄▃▂▁▁
wandb: mom_0 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: mom_1 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: mom_2 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: raw_loss █▄▇▆▄▅▄▅▅▄▃▃▆▅▃▄▄▃▃▃▄▃▃▂▆▂▃▃▅▃▃▄▁▄▅▃▂▃▁▄
wandb: sqr_mom_0 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: sqr_mom_1 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: sqr_mom_2 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: train_loss █▆▅▅▄▄▄▄▄▄▃▃▃▄▃▃▄▃▃▃▃▃▂▃▂▂▂▃▃▂▂▂▂▂▂▂▁▁▁▁
wandb: train_samples_per_sec ▆▅▇▅▅▃▄█▄▆▇▇▇▇▄▄▅▆▅█▆▄▇█▄▄▆▃▅▆▆▄▅▇▄▅▅▇▆▁
wandb: valid_loss █▇▅▅▄▃▄▃▂▁
wandb: wd_0 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: wd_1 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: wd_2 ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁ [748/937]
wandb:
wandb: Run summary:
wandb: F1 Average 0.39622
wandb: F1 validation 0.89669
wandb: F1 validation B 0.0
wandb: F1 validation CC 0.91589
wandb: F1 validation CH2 0.0
wandb: F1 validation CL 0.72727
wandb: F1 validation CM 0.72222
wandb: F1 validation CS 0.0
wandb: F1 validation DAE 0.85548
wandb: F1 validation DAGS 0.93419
wandb: F1 validation DNF 0.0
wandb: F1 validation DSF 0.87751
wandb: F1 validation DSGV 0.8708
wandb: F1 validation DSZ 0.94574
wandb: F1 validation FC 0.38462
wandb: F1 validation FL 0.0
wandb: F1 validation FM 0.59459
wandb: F1 validation FS 0.0
wandb: F1 validation H 0.5
wandb: F1 validation HSV 0.0
wandb: F1 validation HVV 0.0
wandb: F1 validation ID 0.0
wandb: F1 validation IG 0.0
wandb: F1 validation IR 0.0
wandb: F1 validation ISZ 0.0
wandb: F1 validation JAL 0.90909
wandb: F1 validation JAM 0.41379
wandb: F1 validation JOL 0.0
wandb: F1 validation JOM 0.52941
wandb: F1 validation LL 0.91892
wandb: F1 validation LR 0.95238
wandb: F1 validation MGO 0.0
wandb: F1 validation MGP 0.9 [713/937]
wandb: F1 validation MLC 0.0
wandb: F1 validation MMC 0.0
wandb: F1 validation MSC 0.0
wandb: F1 validation MWL 0.54545
wandb: F1 validation MWLS 0.75
wandb: F1 validation ND 0.93689
wandb: F1 validation OBM 0.0
wandb: F1 validation OBR 0.0
wandb: F1 validation OBZ 0.16667
wandb: F1 validation RBJ 0.0
wandb: F1 validation RFB 0.0
wandb: F1 validation RFC 0.0
wandb: F1 validation RFJ 0.81739
wandb: F1 validation RFL 0.0
wandb: F1 validation RMB 0.33333
wandb: F1 validation RMC 0.26087
wandb: F1 validation RMJ 0.83403
wandb: F1 validation RPL 0.0
wandb: F1 validation RPP 1.0
wandb: F1 validation RPR 0.81633
wandb: F1 validation RTB 0.0
wandb: F1 validation SRP 0.0
wandb: F1 validation SRV 0.0
wandb: F1 validation SSS 0.0
wandb: F1 validation TB 0.24
wandb: F1 validation TBA 0.6015
wandb: F1 validation TBD 0.0
wandb: F1 validation TBI 0.52174
wandb: F1 validation TF 0.9071
wandb: F1 validation TFA 0.84514
wandb: F1 validation TFC 0.67143
wandb: F1 validation TFD 0.8
wandb: F1 validation TS 0.925
wandb: F1 validation TSA 0.74648
wandb: F1 validation TSD 0.0 [678/937]
wandb: F1 validation TSI 0.375
wandb: F1 validation average no ND 0.38815
wandb: F1 validation normal 0.93689
wandb: F2 validation 0.87022
wandb: F2 validation B 0.0
wandb: F2 validation CC 0.89091
wandb: F2 validation CH2 0.0
wandb: F2 validation CL 0.625
wandb: F2 validation CM 0.65657
wandb: F2 validation CS 0.0
wandb: F2 validation DAE 0.83318
wandb: F2 validation DAGS 0.92315
wandb: F2 validation DNF 0.0
wandb: F2 validation DSF 0.87168
wandb: F2 validation DSGV 0.83704
wandb: F2 validation DSZ 0.94136
wandb: F2 validation FC 0.2809
wandb: F2 validation FL 0.0
wandb: F2 validation FM 0.47826
wandb: F2 validation FS 0.0
wandb: F2 validation H 0.38462
wandb: F2 validation HSV 0.0
wandb: F2 validation HVV 0.0
wandb: F2 validation ID 0.0
wandb: F2 validation IG 0.0
wandb: F2 validation IR 0.0
wandb: F2 validation ISZ 0.0
wandb: F2 validation JAL 0.86207
wandb: F2 validation JAM 0.30612
wandb: F2 validation JOL 0.0
wandb: F2 validation JOM 0.41284
wandb: F2 validation LL 0.87629
wandb: F2 validation LR 0.92593
wandb: F2 validation MGO 0.0
wandb: F2 validation MGP 0.84906 [643/937]
wandb: F2 validation MLC 0.0
wandb: F2 validation MMC 0.0
wandb: F2 validation MSC 0.0
wandb: F2 validation MWL 0.42857
wandb: F2 validation MWLS 0.67742
wandb: F2 validation ND 0.93379
wandb: F2 validation OBM 0.0
wandb: F2 validation OBR 0.0
wandb: F2 validation OBZ 0.11111
wandb: F2 validation RBJ 0.0
wandb: F2 validation RFB 0.0
wandb: F2 validation RFC 0.0
wandb: F2 validation RFJ 0.78952
wandb: F2 validation RFL 0.0
wandb: F2 validation RMB 0.2381
wandb: F2 validation RMC 0.18072
wandb: F2 validation RMJ 0.79777
wandb: F2 validation RPL 0.0
wandb: F2 validation RPP 1.0
wandb: F2 validation RPR 0.76923
wandb: F2 validation RTB 0.0
wandb: F2 validation SRP 0.0
wandb: F2 validation SRV 0.0
wandb: F2 validation SSS 0.0
wandb: F2 validation TB 0.1662
wandb: F2 validation TBA 0.50761
wandb: F2 validation TBD 0.0
wandb: F2 validation TBI 0.40541
wandb: F2 validation TF 0.89394
wandb: F2 validation TFA 0.80219
wandb: F2 validation TFC 0.5649
wandb: F2 validation TFD 0.71429
wandb: F2 validation TS 0.94872
wandb: F2 validation TSA 0.66751
wandb: F2 validation TSD 0.0 [608/937]
wandb: F2 validation TSI 0.27273
wandb: F2 validation average no ND 0.35464
wandb: accuracy_multi 0.99642
wandb: epoch 10
wandb: eps_0 1e-05
wandb: eps_1 1e-05
wandb: eps_2 1e-05
wandb: learning rate 0.001
wandb: lr_0 0.0
wandb: lr_1 0.0
wandb: lr_2 0.0
wandb: mom_0 0.9
wandb: mom_1 0.9
wandb: mom_2 0.9
wandb: raw_loss 0.00773
wandb: sqr_mom_0 0.99
wandb: sqr_mom_1 0.99
wandb: sqr_mom_2 0.99
wandb: train_loss 0.0106
wandb: train_samples_per_sec 45.47407
wandb: valid_loss 0.00951
wandb: wd_0 0.01
wandb: wd_1 0.01
wandb: wd_2 0.01
wandb:
wandb: 🚀 View run cov_all_video_groups_MC_12_4_24_2024/12/04_15:51 at: https://wandb.ai/lence_ubc/cctv-cov-multilabel/runs/zp2gcfhh
wandb: ⭐️ View project at: https://wandb.ai/lence_ubc/cctv-cov-multilabel
wandb: Synced 5 W&B file(s), 0 media file(s), 2 artifact file(s) and 0 other file(s)
wandb: Find logs at: ./wandb/run-20241204_155156-zp2gcfhh/logs
wandb: WARNING The new W&B backend becomes opt-out in version 0.18.0; try it out with `wandb.require("core")`! See https://wandb.me/wandb-core for more information
.
/usr/lib/python3.10/tempfile.py:1008: ResourceWarning: Implicitly cleaning up <TemporaryDirectory '/tmp/tmp000os2a4'>
_warnings.warn(warn_message, ResourceWarning)
Wed Dec 4 09:15:03 PM EST 2024
Finished