Add files using upload-large-folder tool

pythia1b_v5_04_21/.hydra/config.yaml

@@ -0,0 +1,47 @@
+model:
+  name: EleutherAI/pythia-1b
+  checkpoint_path: null
+  from_scratch: false
+training:
+  epochs: 3
+  batch_size: 4
+  eval_batch_size: 12
+  gradient_accumulation_steps: 4
+  lr: 2.0e-05
+  weight_decay: 0.1
+  betas:
+  - 0.9
+  - 0.95
+  eps: 1.0e-08
+  lr_scheduler: wsd
+  warmup_ratio: 0.1
+  decay_ratio: 0.2
+  warmup_steps: 100
+  min_lr_ratio: 0.1
+  max_grad_norm: 1.0
+  use_amp: true
+  resume: false
+  resume_checkpoint: null
+data:
+  path: ${oc.env:PROJECT_ROOT}/code_completion_exp/datasets/data_V5_full
+  max_context_len: 4096
+  max_target_len: 256
+  num_workers: 4
+  pin_memory: true
+logging:
+  log_interval: 10
+  save_interval: 3000
+  eval_interval: 1000
+  save_every_epoch: true
+tracking:
+  enabled: true
+  backend: wandb
+  project: code-completion-full-docstring
+  run_name: pythia_train
+  entity: null
+  base_url: https://wandb.platun0v.ru
+  local_dir: ${paths.output_dir}
+paths:
+  output_dir: outputs/${now:%Y-%m-%d}/${now:%H-%M-%S}
+seed: 42
+device: cuda

pythia1b_v5_04_21/.hydra/hydra.yaml

@@ -0,0 +1,160 @@
+hydra:
+  run:
+    dir: ${paths.output_dir}
+  sweep:
+    dir: outputs/multirun/${now:%Y-%m-%d}/${now:%H-%M-%S}
+    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: null
+  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:
+      simple:
+        format: '[%(asctime)s][HYDRA] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    loggers:
+      logging_example:
+        level: DEBUG
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.runtime.output_dir}/${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: RUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=RUN
+    task: []
+  job:
+    name: train
+    chdir: false
+    override_dirname: ''
+    id: ???
+    num: ???
+    config_name: config
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.3.2
+    version_base: '1.3'
+    cwd: /home/test/byte-llms-code/code_completion_exp/train_pythia
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: /home/test/byte-llms-code/code_completion_exp/train_pythia/configs
+      schema: file
+      provider: main
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: /home/test/byte-llms-code/code_completion_exp/train_pythia/outputs/2026-04-21/20-28-37
+    choices:
+      paths: default
+      tracking: wandb
+      logging: default
+      data: default
+      training: default
+      model: pythia_1b
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: default
+      hydra/hydra_logging: default
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false

pythia1b_v5_04_21/.hydra/overrides.yaml

@@ -0,0 +1 @@
+[]

pythia1b_v5_04_21/eval_results/eval_config.yaml

@@ -0,0 +1,29 @@
+data:
+  path: /home/test/byte-llms-code/code_completion_exp/datasets/data_V5_full
+  max_context_len: 4096
+  max_target_len: 256
+  num_workers: 4
+  pin_memory: true
+model:
+  name: EleutherAI/pythia-1b
+  checkpoint_path: null
+  from_scratch: false
+paths:
+  checkpoints_dir: outputs/2026-04-21/20-28-37
+  initial_checkpoint: auto
+  output_dir: outputs/2026-04-21/20-28-37/eval_results
+evaluation:
+  batch_size: 16
+  max_samples: null
+  compute_perplexity: true
+  bleu_tokenize: none
+  save_predictions: true
+  use_amp: true
+generation:
+  max_new_tokens: 64
+  temperature: 0.1
+  top_k: 0
+  top_p: 1.0
+  do_sample: true
+seed: 42
+device: cuda

pythia1b_v5_04_21/eval_results/metrics_checkpoint_latest.txt

@@ -0,0 +1,17 @@
+Checkpoint: checkpoint_latest.pt
+================================================================================
+
+exact_match: 0.30150031921685466
+token_accuracy: 0.3239941940977367
+bleu: 16.280810769581016
+perplexity: 710.7840855707159
+num_samples: 37592
+gen_wall_time_s: 681.0312011489659
+gen_samples_per_s: 55.19864572515714
+gen_time_per_sample_ms: 18.11638649576947
+gen_chars_per_s: 1422.6602222708914
+gen_batch_mean_ms: 822.0316826109782
+gen_batch_p50_ms: 656.0139355024148
+gen_batch_p95_ms: 1683.2532198521196
+gen_batch_max_ms: 5435.805578999862
+gen_num_batches: 784

pythia1b_v5_04_21/eval_results/metrics_checkpoint_step_10591.txt

@@ -0,0 +1,17 @@
+Checkpoint: checkpoint_step_10591.pt
+================================================================================
+
+exact_match: 0.29402532453713554
+token_accuracy: 0.3131460051089516
+bleu: 15.46627135309508
+perplexity: 573.579940871295
+num_samples: 37592
+gen_wall_time_s: 697.1238539060578
+gen_samples_per_s: 53.924420731507176
+gen_time_per_sample_ms: 18.544473662110498
+gen_chars_per_s: 1355.0088620655529
+gen_batch_mean_ms: 852.3167421757888
+gen_batch_p50_ms: 675.3437760016823
+gen_batch_p95_ms: 1752.8990025475644
+gen_batch_max_ms: 5515.446357996552
+gen_num_batches: 784

pythia1b_v5_04_21/eval_results/metrics_checkpoint_step_12000.txt

@@ -0,0 +1,17 @@
+Checkpoint: checkpoint_step_12000.pt
+================================================================================
+
+exact_match: 0.29288146414130667
+token_accuracy: 0.3176458823850259
+bleu: 15.78001507522994
+perplexity: 743.1278507940478
+num_samples: 37592
+gen_wall_time_s: 678.8557107109154
+gen_samples_per_s: 55.375537698037
+gen_time_per_sample_ms: 18.058515394523184
+gen_chars_per_s: 1426.102756042461
+gen_batch_mean_ms: 840.9376154973704
+gen_batch_p50_ms: 666.0685060014657
+gen_batch_p95_ms: 1755.3121661472678
+gen_batch_max_ms: 5514.503185004287
+gen_num_batches: 784

pythia1b_v5_04_21/eval_results/metrics_checkpoint_step_15000.txt

@@ -0,0 +1,17 @@
+Checkpoint: checkpoint_step_15000.pt
+================================================================================
+
+exact_match: 0.30064907427112153
+token_accuracy: 0.3238335924171765
+bleu: 16.369268121332667
+perplexity: 648.3400079564792
+num_samples: 37592
+gen_wall_time_s: 682.9705825539422
+gen_samples_per_s: 55.04190218475613
+gen_time_per_sample_ms: 18.167976765108058
+gen_chars_per_s: 1415.311324809008
+gen_batch_mean_ms: 816.9435996159968
+gen_batch_p50_ms: 650.5265364976367
+gen_batch_p95_ms: 1649.2040568959048
+gen_batch_max_ms: 5439.02037099906
+gen_num_batches: 784

pythia1b_v5_04_21/eval_results/metrics_checkpoint_step_18000.txt

@@ -0,0 +1,17 @@
+Checkpoint: checkpoint_step_18000.pt
+================================================================================
+
+exact_match: 0.3029367950627793
+token_accuracy: 0.32487901845101197
+bleu: 16.435138411892126
+perplexity: 540.7501954984347
+num_samples: 37592
+gen_wall_time_s: 665.5543457790118
+gen_samples_per_s: 56.48223956226996
+gen_time_per_sample_ms: 17.70468040484709
+gen_chars_per_s: 1447.5887748464947
+gen_batch_mean_ms: 566.4292304502227
+gen_batch_p50_ms: 467.9037870009779
+gen_batch_p95_ms: 1174.899779098632
+gen_batch_max_ms: 3922.1453899954213
+gen_num_batches: 1175

pythia1b_v5_04_21/eval_results/metrics_checkpoint_step_21000.txt

@@ -0,0 +1,17 @@
+Checkpoint: checkpoint_step_21000.pt
+================================================================================
+
+exact_match: 0.301393913598638
+token_accuracy: 0.32462144971803797
+bleu: 16.368019126858044
+perplexity: 541.3429525210204
+num_samples: 37592
+gen_wall_time_s: 671.3920430750659
+gen_samples_per_s: 55.991131244009956
+gen_time_per_sample_ms: 17.859971352284152
+gen_chars_per_s: 1442.0873913927937
+gen_batch_mean_ms: 571.3974834681411
+gen_batch_p50_ms: 470.4890410066582
+gen_batch_p95_ms: 1157.3573702989959
+gen_batch_max_ms: 3927.851617001579
+gen_num_batches: 1175

pythia1b_v5_04_21/eval_results/metrics_checkpoint_step_21182.txt

@@ -0,0 +1,17 @@
+Checkpoint: checkpoint_step_21182.pt
+================================================================================
+
+exact_match: 0.3023249627580336
+token_accuracy: 0.3244426667151502
+bleu: 16.474168997905824
+perplexity: 542.5514603476134
+num_samples: 37592
+gen_wall_time_s: 670.8535084038012
+gen_samples_per_s: 56.0360787103055
+gen_time_per_sample_ms: 17.8456455736274
+gen_chars_per_s: 1438.8566026832016
+gen_batch_mean_ms: 569.6230980945426
+gen_batch_p50_ms: 471.2959009993938
+gen_batch_p95_ms: 1174.0064269950378
+gen_batch_max_ms: 3921.823967997625
+gen_num_batches: 1175

pythia1b_v5_04_21/eval_results/metrics_checkpoint_step_24000.txt

@@ -0,0 +1,17 @@
+Checkpoint: checkpoint_step_24000.pt
+================================================================================
+
+exact_match: 0.3016599276441796
+token_accuracy: 0.3251093152004945
+bleu: 16.409958897349515
+perplexity: 566.5285793566808
+num_samples: 37592
+gen_wall_time_s: 676.4614610540084
+gen_samples_per_s: 55.57153239953558
+gen_time_per_sample_ms: 17.994824990796136
+gen_chars_per_s: 1434.8740554822307
+gen_batch_mean_ms: 575.7118817480923
+gen_batch_p50_ms: 472.5785279952106
+gen_batch_p95_ms: 1164.6154464004212
+gen_batch_max_ms: 3927.8858819961897
+gen_num_batches: 1175

pythia1b_v5_04_21/eval_results/metrics_checkpoint_step_27000.txt

@@ -0,0 +1,17 @@
+Checkpoint: checkpoint_step_27000.pt
+================================================================================
+
+exact_match: 0.3014737178123005
+token_accuracy: 0.3247578096354948
+bleu: 16.419003802997334
+perplexity: 572.8302545454646
+num_samples: 37592
+gen_wall_time_s: 670.9174906390836
+gen_samples_per_s: 56.03073481389146
+gen_time_per_sample_ms: 17.847347590952424
+gen_chars_per_s: 1447.2554577092376
+gen_batch_mean_ms: 569.2063897753015
+gen_batch_p50_ms: 469.24721299728844
+gen_batch_p95_ms: 1157.0710574953407
+gen_batch_max_ms: 3923.8531240043812
+gen_num_batches: 1175

pythia1b_v5_04_21/eval_results/metrics_checkpoint_step_3000.txt

@@ -0,0 +1,17 @@
+Checkpoint: checkpoint_step_3000.pt
+================================================================================
+
+exact_match: 0.28325175569270056
+token_accuracy: 0.3066916356826632
+bleu: 14.588953791669732
+perplexity: 527.8538214628118
+num_samples: 37592
+gen_wall_time_s: 689.2576722488011
+gen_samples_per_s: 54.539835410682855
+gen_time_per_sample_ms: 18.335222181549295
+gen_chars_per_s: 1389.1713339599544
+gen_batch_mean_ms: 586.6022742542988
+gen_batch_p50_ms: 484.3167610015371
+gen_batch_p95_ms: 1206.365606199688
+gen_batch_max_ms: 3868.3884359998046
+gen_num_batches: 1175

pythia1b_v5_04_21/eval_results/metrics_checkpoint_step_30000.txt

@@ -0,0 +1,17 @@
+Checkpoint: checkpoint_step_30000.pt
+================================================================================
+
+exact_match: 0.3011278995530964
+token_accuracy: 0.3247790211782103
+bleu: 16.428889682390672
+perplexity: 576.8903125233649
+num_samples: 37592
+gen_wall_time_s: 668.1773826373537
+gen_samples_per_s: 56.26050952461326
+gen_time_per_sample_ms: 17.774456869476317
+gen_chars_per_s: 1450.5145268079564
+gen_batch_mean_ms: 568.6616022445563
+gen_batch_p50_ms: 472.4465320032323
+gen_batch_p95_ms: 1174.9766881941466
+gen_batch_max_ms: 3922.6198899996234
+gen_num_batches: 1175

pythia1b_v5_04_21/eval_results/metrics_checkpoint_step_31773.txt

@@ -0,0 +1,17 @@
+Checkpoint: checkpoint_step_31773.pt
+================================================================================
+
+exact_match: 0.3014737178123005
+token_accuracy: 0.3244669084782536
+bleu: 16.1888577028866
+perplexity: 576.6120524185981
+num_samples: 37592
+gen_wall_time_s: 675.8306533855648
+gen_samples_per_s: 55.62340182660163
+gen_time_per_sample_ms: 17.978044620812
+gen_chars_per_s: 1434.6759726609196
+gen_batch_mean_ms: 567.6975614385199
+gen_batch_p50_ms: 471.66150900011417
+gen_batch_p95_ms: 1157.574844303599
+gen_batch_max_ms: 3920.68699200172
+gen_num_batches: 1175

pythia1b_v5_04_21/eval_results/metrics_checkpoint_step_6000.txt

@@ -0,0 +1,17 @@
+Checkpoint: checkpoint_step_6000.pt
+================================================================================
+
+exact_match: 0.2844488188976378
+token_accuracy: 0.3113612053004615
+bleu: 15.080319950709026
+perplexity: 453.1818960521432
+num_samples: 37592
+gen_wall_time_s: 674.0668778871186
+gen_samples_per_s: 55.76894701877827
+gen_time_per_sample_ms: 17.93112571523512
+gen_chars_per_s: 1428.3389847279113
+gen_batch_mean_ms: 569.0444115582889
+gen_batch_p50_ms: 480.34932199516334
+gen_batch_p95_ms: 1182.4054761891603
+gen_batch_max_ms: 3867.8216240077745
+gen_num_batches: 1175

pythia1b_v5_04_21/eval_results/metrics_checkpoint_step_9000.txt

@@ -0,0 +1,17 @@
+Checkpoint: checkpoint_step_9000.pt
+================================================================================
+
+exact_match: 0.2909927644179613
+token_accuracy: 0.3142671866524852
+bleu: 15.966688789870538
+perplexity: 457.0689719416982
+num_samples: 37592
+gen_wall_time_s: 681.9195306949405
+gen_samples_per_s: 55.12673901815101
+gen_time_per_sample_ms: 18.140017309399354
+gen_chars_per_s: 1398.1063117937088
+gen_batch_mean_ms: 570.5794102875913
+gen_batch_p50_ms: 473.2587530015735
+gen_batch_p95_ms: 1157.3061183065875
+gen_batch_max_ms: 3924.258289000136
+gen_num_batches: 1175

pythia1b_v5_04_21/eval_results/metrics_initial_checkpoint.txt

@@ -0,0 +1,17 @@
+Checkpoint: initial_checkpoint
+================================================================================
+
+exact_match: 0.0
+token_accuracy: 0.25446275707632215
+bleu: 0.8505902631105438
+perplexity: 1221.5743555150027
+num_samples: 37592
+gen_wall_time_s: 1353.5502136530558
+gen_samples_per_s: 27.772889118419986
+gen_time_per_sample_ms: 36.00633681775526
+gen_chars_per_s: 6264.1983389124025
+gen_batch_mean_ms: 1691.747248192611
+gen_batch_p50_ms: 1425.062906499079
+gen_batch_p95_ms: 3282.010366447497
+gen_batch_max_ms: 5511.020823003491
+gen_num_batches: 784

pythia1b_v5_04_21/eval_results/metrics_model_best.txt

@@ -0,0 +1,17 @@
+Checkpoint: model_best.pt
+================================================================================
+
+exact_match: 0.30075547988933815
+token_accuracy: 0.32501840858885667
+bleu: 16.37759776617732
+perplexity: 541.3429525210204
+num_samples: 37592
+gen_wall_time_s: 676.7071885522455
+gen_samples_per_s: 55.55135313461753
+gen_time_per_sample_ms: 18.001361687386826
+gen_chars_per_s: 1433.1826473942099
+gen_batch_mean_ms: 572.0854987192062
+gen_batch_p50_ms: 468.9722350012744
+gen_batch_p95_ms: 1183.3017533019301
+gen_batch_max_ms: 3921.211963010137
+gen_num_batches: 1175

pythia1b_v5_04_21/eval_results/metrics_model_final.txt

@@ -0,0 +1,17 @@
+Checkpoint: model_final.pt
+================================================================================
+
+exact_match: 0.3001702489891466
+token_accuracy: 0.3246911447869604
+bleu: 16.324443990793565
+perplexity: 576.6120524185981
+num_samples: 37592
+gen_wall_time_s: 666.1244535958976
+gen_samples_per_s: 56.43389879634275
+gen_time_per_sample_ms: 17.71984607352356
+gen_chars_per_s: 1452.986742605242
+gen_batch_mean_ms: 565.5749287258368
+gen_batch_p50_ms: 468.90689300198574
+gen_batch_p95_ms: 1165.2756220020817
+gen_batch_max_ms: 3921.405152999796
+gen_num_batches: 1175

pythia1b_v5_04_21/eval_results/summary.txt

@@ -0,0 +1,17 @@
+EVALUATION SUMMARY
+==================================================================================================
+
+Checkpoint                      Exact Match    Token Acc     BLEU   PERPLEXITY  ms/sample   samp/s
+--------------------------------------------------------------------------------------------------
+checkpoint_step_18000                30.29%       32.49%    16.44       540.75       17.7    56.48
+checkpoint_step_21000                30.14%       32.46%    16.37       541.34       17.9    55.99
+checkpoint_step_21182                30.23%       32.44%    16.47       542.55       17.8    56.04
+checkpoint_step_24000                30.17%       32.51%    16.41       566.53       18.0    55.57
+checkpoint_step_27000                30.15%       32.48%    16.42       572.83       17.8    56.03
+checkpoint_step_3000                 28.33%       30.67%    14.59       527.85       18.3    54.54
+checkpoint_step_30000                30.11%       32.48%    16.43       576.89       17.8    56.26
+checkpoint_step_31773                30.15%       32.45%    16.19       576.61       18.0    55.62
+checkpoint_step_6000                 28.44%       31.14%    15.08       453.18       17.9    55.77
+checkpoint_step_9000                 29.10%       31.43%    15.97       457.07       18.1    55.13
+model_best                           30.08%       32.50%    16.38       541.34       18.0    55.55
+model_final                          30.02%       32.47%    16.32       576.61       17.7    56.43

pythia1b_v5_04_21/wandb/debug-internal.log

@@ -0,0 +1,13 @@
+{"time":"2026-04-21T20:28:39.540287752Z","level":"INFO","msg":"stream: starting","core version":"0.24.0"}
+{"time":"2026-04-21T20:28:39.779234937Z","level":"INFO","msg":"stream: created new stream","id":"8ing6xdi"}
+{"time":"2026-04-21T20:28:39.779334686Z","level":"INFO","msg":"handler: started","stream_id":"8ing6xdi"}
+{"time":"2026-04-21T20:28:39.779468186Z","level":"INFO","msg":"stream: started","id":"8ing6xdi"}
+{"time":"2026-04-21T20:28:39.779542005Z","level":"INFO","msg":"sender: started","stream_id":"8ing6xdi"}
+{"time":"2026-04-21T20:28:39.779545215Z","level":"INFO","msg":"writer: started","stream_id":"8ing6xdi"}
+{"time":"2026-04-21T20:28:39.866766122Z","level":"ERROR","msg":"git repo not found","error":"repository does not exist"}
+{"time":"2026-04-22T03:07:37.46900739Z","level":"INFO","msg":"fileTransfer: Close: file transfer manager closed"}
+{"time":"2026-04-22T03:07:37.532953738Z","level":"INFO","msg":"handler: operation stats","stats":{}}
+{"time":"2026-04-22T03:07:37.536692894Z","level":"INFO","msg":"stream: closing","id":"8ing6xdi"}
+{"time":"2026-04-22T03:07:37.536702644Z","level":"INFO","msg":"handler: closed","stream_id":"8ing6xdi"}
+{"time":"2026-04-22T03:07:37.536758784Z","level":"INFO","msg":"sender: closed","stream_id":"8ing6xdi"}
+{"time":"2026-04-22T03:07:37.536765384Z","level":"INFO","msg":"stream: closed","id":"8ing6xdi"}

pythia1b_v5_04_21/wandb/debug.log

@@ -0,0 +1,24 @@
+2026-04-21 20:28:39,221 INFO    MainThread:13721 [wandb_setup.py:_flush():81] Current SDK version is 0.24.0
+2026-04-21 20:28:39,221 INFO    MainThread:13721 [wandb_setup.py:_flush():81] Configure stats pid to 13721
+2026-04-21 20:28:39,221 INFO    MainThread:13721 [wandb_setup.py:_flush():81] Loading settings from environment variables
+2026-04-21 20:28:39,222 INFO    MainThread:13721 [wandb_init.py:setup_run_log_directory():717] Logging user logs to outputs/2026-04-21/20-28-37/wandb/run-20260421_202839-8ing6xdi/logs/debug.log
+2026-04-21 20:28:39,222 INFO    MainThread:13721 [wandb_init.py:setup_run_log_directory():718] Logging internal logs to outputs/2026-04-21/20-28-37/wandb/run-20260421_202839-8ing6xdi/logs/debug-internal.log
+2026-04-21 20:28:39,222 INFO    MainThread:13721 [wandb_init.py:init():844] calling init triggers
+2026-04-21 20:28:39,222 INFO    MainThread:13721 [wandb_init.py:init():849] wandb.init called with sweep_config: {}
+config: {'model': {'name': 'EleutherAI/pythia-1b', 'checkpoint_path': None, 'from_scratch': False}, 'training': {'epochs': 3, 'batch_size': 4, 'eval_batch_size': 12, 'gradient_accumulation_steps': 4, 'lr': 2e-05, 'weight_decay': 0.1, 'betas': [0.9, 0.95], 'eps': 1e-08, 'lr_scheduler': 'wsd', 'warmup_ratio': 0.1, 'decay_ratio': 0.2, 'warmup_steps': 100, 'min_lr_ratio': 0.1, 'max_grad_norm': 1.0, 'use_amp': True, 'resume': False, 'resume_checkpoint': None}, 'data': {'path': '/home/test/byte-llms-code/code_completion_exp/datasets/data_V5_full', 'max_context_len': 4096, 'max_target_len': 256, 'num_workers': 4, 'pin_memory': True}, 'logging': {'log_interval': 10, 'save_interval': 3000, 'eval_interval': 1000, 'save_every_epoch': True}, 'tracking': {'enabled': True, 'backend': 'wandb', 'project': 'code-completion-full-docstring', 'run_name': 'pythia_train', 'entity': None, 'base_url': 'https://wandb.platun0v.ru', 'local_dir': 'outputs/2026-04-21/20-28-37'}, 'paths': {'output_dir': 'outputs/2026-04-21/20-28-37'}, 'seed': 42, 'device': 'cuda', '_wandb': {'code_path': 'code/code_completion_exp/train_pythia/train.py'}}
+2026-04-21 20:28:39,222 INFO    MainThread:13721 [wandb_init.py:init():892] starting backend
+2026-04-21 20:28:39,507 INFO    MainThread:13721 [wandb_init.py:init():895] sending inform_init request
+2026-04-21 20:28:39,538 INFO    MainThread:13721 [wandb_init.py:init():903] backend started and connected
+2026-04-21 20:28:39,541 INFO    MainThread:13721 [wandb_init.py:init():973] updated telemetry
+2026-04-21 20:28:39,573 INFO    MainThread:13721 [wandb_init.py:init():997] communicating run to backend with 90.0 second timeout
+2026-04-21 20:28:39,865 INFO    MainThread:13721 [wandb_init.py:init():1044] starting run threads in backend
+2026-04-21 20:28:40,023 INFO    MainThread:13721 [wandb_run.py:_console_start():2529] atexit reg
+2026-04-21 20:28:40,024 INFO    MainThread:13721 [wandb_run.py:_redirect():2377] redirect: wrap_raw
+2026-04-21 20:28:40,024 INFO    MainThread:13721 [wandb_run.py:_redirect():2446] Wrapping output streams.
+2026-04-21 20:28:40,024 INFO    MainThread:13721 [wandb_run.py:_redirect():2469] Redirects installed.
+2026-04-21 20:28:40,028 INFO    MainThread:13721 [wandb_init.py:init():1084] run started, returning control to user process
+2026-04-22 03:07:36,952 INFO    MainThread:13721 [wandb_run.py:_finish():2295] finishing run nikita/code-completion-full-docstring/8ing6xdi
+2026-04-22 03:07:36,952 INFO    MainThread:13721 [wandb_run.py:_atexit_cleanup():2494] got exitcode: 0
+2026-04-22 03:07:36,953 INFO    MainThread:13721 [wandb_run.py:_restore():2476] restore
+2026-04-22 03:07:36,953 INFO    MainThread:13721 [wandb_run.py:_restore():2482] restore done
+2026-04-22 03:07:37,536 INFO    MainThread:13721 [wandb_run.py:_footer_sync_info():3870] logging synced files

pythia1b_v5_04_21/wandb/run-20260421_202839-8ing6xdi/files/config.yaml

@@ -0,0 +1,126 @@
+_wandb:
+    value:
+        cli_version: 0.24.0
+        code_path: code/code_completion_exp/train_pythia/train.py
+        e:
+            5t2ii4n1vl6acag1rgxxlnfw2gymqoqp:
+                codePath: code_completion_exp/train_pythia/train.py
+                codePathLocal: train.py
+                cpu_count: 28
+                cpu_count_logical: 56
+                cudaVersion: "13.1"
+                disk:
+                    /:
+                        total: "265567551488"
+                        used: "47762849792"
+                email: nikita@local.ru
+                executable: /home/test/miniconda3/envs/bytellm/bin/python
+                git:
+                    commit: 01bfef6a1f209a8e7ab13b76c9be95616e3c74c4
+                    remote: https://github.com/naryst/byte-llms-code.git
+                gpu: NVIDIA A100-SXM4-80GB
+                gpu_count: 2
+                gpu_nvidia:
+                    - architecture: Ampere
+                      cudaCores: 6912
+                      memoryTotal: "85899345920"
+                      name: NVIDIA A100-SXM4-80GB
+                      uuid: GPU-b1db9754-65e5-affa-70f1-cf3cd600252a
+                    - architecture: Ampere
+                      cudaCores: 6912
+                      memoryTotal: "85899345920"
+                      name: NVIDIA A100-SXM4-80GB
+                      uuid: GPU-cf24f6d2-04cd-c64e-ebba-294de221c27e
+                host: compute-vm-56-238-256-ssd-1776796584508
+                memory:
+                    total: "251218849792"
+                os: Linux-6.8.0-110-generic-x86_64-with-glibc2.39
+                program: /home/test/byte-llms-code/code_completion_exp/train_pythia/train.py
+                python: CPython 3.12.0
+                root: outputs/2026-04-21/20-28-37
+                startedAt: "2026-04-21T20:28:39.219939Z"
+                writerId: 5t2ii4n1vl6acag1rgxxlnfw2gymqoqp
+        m: []
+        python_version: 3.12.0
+        t:
+            "1":
+                - 1
+                - 11
+                - 49
+                - 50
+                - 51
+                - 71
+                - 105
+            "2":
+                - 1
+                - 11
+                - 49
+                - 50
+                - 51
+                - 71
+                - 105
+            "3":
+                - 2
+                - 13
+                - 16
+                - 61
+            "4": 3.12.0
+            "5": 0.24.0
+            "6": 4.57.6
+            "12": 0.24.0
+            "13": linux-x86_64
+data:
+    value:
+        max_context_len: 4096
+        max_target_len: 256
+        num_workers: 4
+        path: /home/test/byte-llms-code/code_completion_exp/datasets/data_V5_full
+        pin_memory: true
+device:
+    value: cuda
+logging:
+    value:
+        eval_interval: 1000
+        log_interval: 10
+        save_every_epoch: true
+        save_interval: 3000
+model:
+    value:
+        checkpoint_path: null
+        from_scratch: false
+        name: EleutherAI/pythia-1b
+paths:
+    value:
+        output_dir: outputs/2026-04-21/20-28-37
+seed:
+    value: 42
+tracking:
+    value:
+        backend: wandb
+        base_url: https://wandb.platun0v.ru
+        enabled: true
+        entity: null
+        local_dir: outputs/2026-04-21/20-28-37
+        project: code-completion-full-docstring
+        run_name: pythia_train
+training:
+    value:
+        batch_size: 4
+        betas:
+            - 0.9
+            - 0.95
+        decay_ratio: 0.2
+        epochs: 3
+        eps: 1e-08
+        eval_batch_size: 12
+        gradient_accumulation_steps: 4
+        lr: 2e-05
+        lr_scheduler: wsd
+        max_grad_norm: 1
+        min_lr_ratio: 0.1
+        resume: false
+        resume_checkpoint: null
+        use_amp: true
+        warmup_ratio: 0.1
+        warmup_steps: 100
+        weight_decay: 0.1

pythia1b_v5_04_21/wandb/run-20260421_202839-8ing6xdi/files/requirements.txt

@@ -0,0 +1,245 @@
+nest-asyncio==1.6.0
+safetensors==0.6.2
+sentry-sdk==2.50.0
+jupyter_server_terminals==0.5.3
+pandocfilters==1.5.1
+typer==0.21.1
+causal-conv1d==1.5.0.post8
+clearml==1.16.4
+jupyter-lsp==2.3.0
+pathlib2==2.3.7.post1
+mlstm_kernels==2.0.2
+fastapi==0.128.0
+antlr4-python3-runtime==4.9.3
+PyJWT==2.8.0
+ffmpy==1.0.0
+transformers==4.57.6
+shellingham==1.5.4
+filelock==3.19.1
+nvidia-nvtx-cu12==12.4.127
+nbclassic==1.3.3
+nbconvert==7.16.6
+traitlets==5.14.3
+colorama==0.4.6
+dacite==1.9.2
+argon2-cffi-bindings==25.1.0
+nvidia-nccl-cu12==2.21.5
+httpx==0.28.1
+xxhash==3.6.0
+wcwidth==0.2.14
+terminado==0.18.1
+librt==0.8.0
+notebook_shim==0.2.4
+nvidia-nvjitlink-cu12==12.4.127
+python-json-logger==4.0.0
+nbformat==5.10.4
+pip==25.2
+propcache==0.4.1
+attrs==25.4.0
+pytz==2025.2
+lxml==6.0.2
+executing==2.2.1
+orjson==3.11.6
+plotly==6.5.2
+MarkupSafe==3.0.3
+nvidia-ml-py==13.590.48
+types-python-dateutil==2.9.0.20251008
+nvidia-cufft-cu12==11.2.1.3
+jupyterlab_server==2.27.3
+anyio==4.11.0
+contourpy==1.3.3
+ipython-genutils==0.2.0
+jsonschema-specifications==2025.9.1
+aiohappyeyeballs==2.6.1
+starlette==0.50.0
+gradio_client==2.0.3
+nvidia-cublas-cu12==12.4.5.8
+seaborn==0.13.2
+fonttools==4.60.1
+einops==0.8.1
+omegaconf==2.3.0
+babel==2.17.0
+python-dateutil==2.9.0.post0
+jupyterlab_widgets==3.0.15
+nvidia-cusparse-cu12==12.3.1.170
+trackio==0.15.0
+fastjsonschema==2.21.2
+packaging==25.0
+matplotlib-inline==0.1.7
+pycparser==2.23
+ipykernel==6.30.1
+nvidia-cuda-cupti-cu12==12.4.127
+webencodings==0.5.1
+jupyterlab==4.4.9
+prompt_toolkit==3.0.52
+jupyter_client==8.6.3
+setuptools==78.1.1
+sacrebleu==2.6.0
+pillow==11.3.0
+gradio==6.5.1
+numpy==2.3.3
+certifi==2025.10.5
+fqdn==1.5.1
+hydra-core==1.3.2
+xlstm==2.0.4
+psutil==7.1.0
+six==1.17.0
+jedi==0.19.2
+reportlab==4.4.9
+click==8.3.1
+typing_extensions==4.15.0
+triton==3.2.0
+rpds-py==0.27.1
+pexpect==4.9.0
+jsonschema==4.25.1
+ipython==9.6.0
+jsonpointer==3.0.0
+gitdb==4.0.12
+websocket-client==1.9.0
+jupyter_server==2.17.0
+nvidia-cudnn-cu12==9.1.0.70
+mdurl==0.1.2
+async-lru==2.0.5
+torch==2.6.0
+tzdata==2025.2
+cffi==2.0.0
+tornado==6.5.2
+aiohttp==3.13.1
+nvidia-cuda-nvrtc-cu12==12.4.127
+annotated-types==0.7.0
+isoduration==20.11.0
+ptyprocess==0.7.0
+debugpy==1.8.17
+rfc3339-validator==0.1.4
+uvicorn==0.40.0
+safehttpx==0.1.7
+pydantic==2.12.5
+mypy==1.19.1
+mypy_extensions==1.1.0
+pydub==0.25.1
+markdown-it-py==4.0.0
+cycler==0.12.1
+tinycss2==1.4.0
+tokenizers==0.22.1
+jupyterlab_pygments==0.3.0
+joypy==0.2.6
+jupyter_contrib_core==0.4.2
+nbclient==0.10.2
+furl==2.1.4
+stack-data==0.6.3
+semantic-version==2.10.0
+requests==2.32.5
+nvidia-cufile-cu12==1.13.1.3
+ipython_pygments_lexers==1.1.1
+hf-xet==1.1.10
+arrow==1.3.0
+rfc3987-syntax==1.1.0
+tomlkit==0.13.3
+python-multipart==0.0.22
+jupyter==1.1.1
+idna==3.10
+optree==0.17.0
+h11==0.16.0
+nvidia-curand-cu12==10.3.5.147
+Pygments==2.19.2
+flash_attn==2.7.4.post1
+pandas==2.3.3
+mamba-ssm==2.2.4
+GitPython==3.1.46
+ftfy==6.3.1
+argon2-cffi==25.1.0
+cryptography==46.0.4
+json5==0.12.1
+tqdm==4.67.1
+annotated-doc==0.0.4
+aiosignal==1.4.0
+orderedmultidict==1.0.2
+jupyter-events==0.12.0
+pydantic_core==2.41.5
+jupyter_nbextensions_configurator==0.6.4
+jupyter_core==5.8.1
+multiprocess==0.70.16
+scipy==1.17.0
+frozenlist==1.8.0
+Jinja2==3.1.6
+protobuf==6.33.4
+wandb==0.24.0
+ipywidgets==8.1.7
+webcolors==24.11.1
+nvidia-cuda-runtime-cu12==12.4.127
+comm==0.2.3
+rfc3986-validator==0.1.1
+python-dotenv==1.2.1
+pyzmq==27.1.0
+sniffio==1.3.1
+widgetsnbextension==4.0.14
+kiwisolver==1.4.9
+PyYAML==6.0.3
+platformdirs==4.5.0
+wheel==0.45.1
+httpcore==1.0.9
+datasets==4.3.0
+prometheus_client==0.23.1
+ninja==1.13.0
+rich==14.2.0
+Send2Trash==1.8.3
+multidict==6.7.0
+accelerate==1.10.1
+bleach==6.2.0
+defusedxml==0.7.1
+huggingface-hub==0.35.3
+nvidia-cusparselt-cu12==0.6.2
+beautifulsoup4==4.14.2
+urllib3==2.5.0
+matplotlib==3.10.7
+nvidia-cusolver-cu12==11.6.1.9
+portalocker==3.2.0
+sympy==1.13.1
+brotli==1.2.0
+jupyter-console==6.6.3
+pathspec==1.0.4
+narwhals==2.15.0
+uri-template==1.3.0
+parso==0.8.5
+fsspec==2025.9.0
+typing-inspection==0.4.2
+asttokens==3.0.0
+mistune==3.1.4
+pyarrow==22.0.0
+Authlib==1.6.6
+yarl==1.22.0
+charset-normalizer==3.4.3
+lark==1.3.0
+tabulate==0.9.0
+regex==2025.9.18
+hnet==0.0.1
+dill==0.4.0
+referencing==0.36.2
+networkx==3.5
+notebook==7.4.7
+soupsieve==2.8
+itsdangerous==2.2.0
+opt_einsum==3.4.0
+aiofiles==24.1.0
+decorator==5.2.1
+pyparsing==3.2.5
+pure_eval==0.2.3
+groovy==0.1.2
+mpmath==1.3.0
+smmap==5.0.2
+inflect==7.3.1
+platformdirs==4.2.2
+jaraco.collections==5.1.0
+importlib_metadata==8.0.0
+tomli==2.0.1
+backports.tarfile==1.2.0
+jaraco.text==3.12.1
+typeguard==4.3.0
+autocommand==2.2.2
+wheel==0.45.1
+jaraco.context==5.3.0
+packaging==24.2
+more-itertools==10.3.0
+typing_extensions==4.12.2
+zipp==3.19.2
+jaraco.functools==4.0.1

pythia1b_v5_04_21/wandb/run-20260421_202839-8ing6xdi/files/wandb-metadata.json

@@ -0,0 +1,47 @@
+{
+  "os": "Linux-6.8.0-110-generic-x86_64-with-glibc2.39",
+  "python": "CPython 3.12.0",
+  "startedAt": "2026-04-21T20:28:39.219939Z",
+  "program": "/home/test/byte-llms-code/code_completion_exp/train_pythia/train.py",
+  "codePath": "code_completion_exp/train_pythia/train.py",
+  "codePathLocal": "train.py",
+  "git": {
+    "remote": "https://github.com/naryst/byte-llms-code.git",
+    "commit": "01bfef6a1f209a8e7ab13b76c9be95616e3c74c4"
+  },
+  "email": "nikita@local.ru",
+  "root": "outputs/2026-04-21/20-28-37",
+  "host": "compute-vm-56-238-256-ssd-1776796584508",
+  "executable": "/home/test/miniconda3/envs/bytellm/bin/python",
+  "cpu_count": 28,
+  "cpu_count_logical": 56,
+  "gpu": "NVIDIA A100-SXM4-80GB",
+  "gpu_count": 2,
+  "disk": {
+    "/": {
+      "total": "265567551488",
+      "used": "47762849792"
+    }
+  },
+  "memory": {
+    "total": "251218849792"
+  },
+  "gpu_nvidia": [
+    {
+      "name": "NVIDIA A100-SXM4-80GB",
+      "memoryTotal": "85899345920",
+      "cudaCores": 6912,
+      "architecture": "Ampere",
+      "uuid": "GPU-b1db9754-65e5-affa-70f1-cf3cd600252a"
+    },
+    {
+      "name": "NVIDIA A100-SXM4-80GB",
+      "memoryTotal": "85899345920",
+      "cudaCores": 6912,
+      "architecture": "Ampere",
+      "uuid": "GPU-cf24f6d2-04cd-c64e-ebba-294de221c27e"
+    }
+  ],
+  "cudaVersion": "13.1",
+  "writerId": "5t2ii4n1vl6acag1rgxxlnfw2gymqoqp"
+}

pythia1b_v5_04_21/wandb/run-20260421_202839-8ing6xdi/files/wandb-summary.json

@@ -0,0 +1 @@
+{"epoch/loss":0.8193523640954421,"val/perplexity":2.954253480862506,"train/loss_avg":0.8192876343746018,"train/step_time":0.5378646373748779,"best/val_loss":1.0519802477912075,"train/lr":2.0000000000000003e-06,"_timestamp":1.776827170070497e+09,"_step":31770,"_wandb":{"runtime":23937},"train/loss":0.9364707842469215,"val/time":184.67540502548218,"train/epoch":3,"best/val_perplexity":2.912696143858297,"best/step":21000,"epoch/time":7722.56347155571,"_runtime":23937,"val/loss":1.0659428229359358}

pythia1b_v5_04_21/wandb/run-20260421_202839-8ing6xdi/logs/debug-core.log

@@ -0,0 +1,16 @@
+{"time":"2026-04-21T20:28:39.323073805Z","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmp0kqo2ro3/port-13721.txt","pid":13721,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2026-04-21T20:28:39.323771093Z","level":"INFO","msg":"server: will exit if parent process dies","ppid":13721}
+{"time":"2026-04-21T20:28:39.323710273Z","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-13721-13772-993102315/socket","Net":"unix"}}
+{"time":"2026-04-21T20:28:39.506674337Z","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2026-04-21T20:28:39.540165162Z","level":"INFO","msg":"handleInformInit: received","streamId":"8ing6xdi","id":"1(@)"}
+{"time":"2026-04-21T20:28:39.779483135Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"8ing6xdi","id":"1(@)"}
+{"time":"2026-04-22T03:07:37.536676435Z","level":"INFO","msg":"handleInformFinish: finish message received","streamId":"8ing6xdi","id":"1(@)"}
+{"time":"2026-04-22T03:07:40.251497019Z","level":"INFO","msg":"handleInformFinish: stream closed","streamId":"8ing6xdi","id":"1(@)"}
+{"time":"2026-04-22T03:07:40.251582799Z","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"1(@)"}
+{"time":"2026-04-22T03:07:40.251608679Z","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"1(@)"}
+{"time":"2026-04-22T03:07:40.251627789Z","level":"INFO","msg":"connection: closing","id":"1(@)"}
+{"time":"2026-04-22T03:07:40.251650818Z","level":"INFO","msg":"server is shutting down"}
+{"time":"2026-04-22T03:07:40.251686398Z","level":"INFO","msg":"connection: closed successfully","id":"1(@)"}
+{"time":"2026-04-22T03:07:40.251697698Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"1(@)"}
+{"time":"2026-04-22T03:07:40.251873857Z","level":"INFO","msg":"server: listener closed","addr":{"Name":"/tmp/wandb-13721-13772-993102315/socket","Net":"unix"}}
+{"time":"2026-04-22T03:07:40.251925627Z","level":"INFO","msg":"server is closed"}

pythia1b_v5_04_21/wandb/run-20260421_202839-8ing6xdi/logs/debug-internal.log

@@ -0,0 +1,13 @@
+{"time":"2026-04-21T20:28:39.540287752Z","level":"INFO","msg":"stream: starting","core version":"0.24.0"}
+{"time":"2026-04-21T20:28:39.779234937Z","level":"INFO","msg":"stream: created new stream","id":"8ing6xdi"}
+{"time":"2026-04-21T20:28:39.779334686Z","level":"INFO","msg":"handler: started","stream_id":"8ing6xdi"}
+{"time":"2026-04-21T20:28:39.779468186Z","level":"INFO","msg":"stream: started","id":"8ing6xdi"}
+{"time":"2026-04-21T20:28:39.779542005Z","level":"INFO","msg":"sender: started","stream_id":"8ing6xdi"}
+{"time":"2026-04-21T20:28:39.779545215Z","level":"INFO","msg":"writer: started","stream_id":"8ing6xdi"}
+{"time":"2026-04-21T20:28:39.866766122Z","level":"ERROR","msg":"git repo not found","error":"repository does not exist"}
+{"time":"2026-04-22T03:07:37.46900739Z","level":"INFO","msg":"fileTransfer: Close: file transfer manager closed"}
+{"time":"2026-04-22T03:07:37.532953738Z","level":"INFO","msg":"handler: operation stats","stats":{}}
+{"time":"2026-04-22T03:07:37.536692894Z","level":"INFO","msg":"stream: closing","id":"8ing6xdi"}
+{"time":"2026-04-22T03:07:37.536702644Z","level":"INFO","msg":"handler: closed","stream_id":"8ing6xdi"}
+{"time":"2026-04-22T03:07:37.536758784Z","level":"INFO","msg":"sender: closed","stream_id":"8ing6xdi"}
+{"time":"2026-04-22T03:07:37.536765384Z","level":"INFO","msg":"stream: closed","id":"8ing6xdi"}

pythia1b_v5_04_21/wandb/run-20260421_202839-8ing6xdi/logs/debug.log

@@ -0,0 +1,24 @@
+2026-04-21 20:28:39,221 INFO    MainThread:13721 [wandb_setup.py:_flush():81] Current SDK version is 0.24.0
+2026-04-21 20:28:39,221 INFO    MainThread:13721 [wandb_setup.py:_flush():81] Configure stats pid to 13721
+2026-04-21 20:28:39,221 INFO    MainThread:13721 [wandb_setup.py:_flush():81] Loading settings from environment variables
+2026-04-21 20:28:39,222 INFO    MainThread:13721 [wandb_init.py:setup_run_log_directory():717] Logging user logs to outputs/2026-04-21/20-28-37/wandb/run-20260421_202839-8ing6xdi/logs/debug.log
+2026-04-21 20:28:39,222 INFO    MainThread:13721 [wandb_init.py:setup_run_log_directory():718] Logging internal logs to outputs/2026-04-21/20-28-37/wandb/run-20260421_202839-8ing6xdi/logs/debug-internal.log
+2026-04-21 20:28:39,222 INFO    MainThread:13721 [wandb_init.py:init():844] calling init triggers
+2026-04-21 20:28:39,222 INFO    MainThread:13721 [wandb_init.py:init():849] wandb.init called with sweep_config: {}
+config: {'model': {'name': 'EleutherAI/pythia-1b', 'checkpoint_path': None, 'from_scratch': False}, 'training': {'epochs': 3, 'batch_size': 4, 'eval_batch_size': 12, 'gradient_accumulation_steps': 4, 'lr': 2e-05, 'weight_decay': 0.1, 'betas': [0.9, 0.95], 'eps': 1e-08, 'lr_scheduler': 'wsd', 'warmup_ratio': 0.1, 'decay_ratio': 0.2, 'warmup_steps': 100, 'min_lr_ratio': 0.1, 'max_grad_norm': 1.0, 'use_amp': True, 'resume': False, 'resume_checkpoint': None}, 'data': {'path': '/home/test/byte-llms-code/code_completion_exp/datasets/data_V5_full', 'max_context_len': 4096, 'max_target_len': 256, 'num_workers': 4, 'pin_memory': True}, 'logging': {'log_interval': 10, 'save_interval': 3000, 'eval_interval': 1000, 'save_every_epoch': True}, 'tracking': {'enabled': True, 'backend': 'wandb', 'project': 'code-completion-full-docstring', 'run_name': 'pythia_train', 'entity': None, 'base_url': 'https://wandb.platun0v.ru', 'local_dir': 'outputs/2026-04-21/20-28-37'}, 'paths': {'output_dir': 'outputs/2026-04-21/20-28-37'}, 'seed': 42, 'device': 'cuda', '_wandb': {'code_path': 'code/code_completion_exp/train_pythia/train.py'}}
+2026-04-21 20:28:39,222 INFO    MainThread:13721 [wandb_init.py:init():892] starting backend
+2026-04-21 20:28:39,507 INFO    MainThread:13721 [wandb_init.py:init():895] sending inform_init request
+2026-04-21 20:28:39,538 INFO    MainThread:13721 [wandb_init.py:init():903] backend started and connected
+2026-04-21 20:28:39,541 INFO    MainThread:13721 [wandb_init.py:init():973] updated telemetry
+2026-04-21 20:28:39,573 INFO    MainThread:13721 [wandb_init.py:init():997] communicating run to backend with 90.0 second timeout
+2026-04-21 20:28:39,865 INFO    MainThread:13721 [wandb_init.py:init():1044] starting run threads in backend
+2026-04-21 20:28:40,023 INFO    MainThread:13721 [wandb_run.py:_console_start():2529] atexit reg
+2026-04-21 20:28:40,024 INFO    MainThread:13721 [wandb_run.py:_redirect():2377] redirect: wrap_raw
+2026-04-21 20:28:40,024 INFO    MainThread:13721 [wandb_run.py:_redirect():2446] Wrapping output streams.
+2026-04-21 20:28:40,024 INFO    MainThread:13721 [wandb_run.py:_redirect():2469] Redirects installed.
+2026-04-21 20:28:40,028 INFO    MainThread:13721 [wandb_init.py:init():1084] run started, returning control to user process
+2026-04-22 03:07:36,952 INFO    MainThread:13721 [wandb_run.py:_finish():2295] finishing run nikita/code-completion-full-docstring/8ing6xdi
+2026-04-22 03:07:36,952 INFO    MainThread:13721 [wandb_run.py:_atexit_cleanup():2494] got exitcode: 0
+2026-04-22 03:07:36,953 INFO    MainThread:13721 [wandb_run.py:_restore():2476] restore
+2026-04-22 03:07:36,953 INFO    MainThread:13721 [wandb_run.py:_restore():2482] restore done
+2026-04-22 03:07:37,536 INFO    MainThread:13721 [wandb_run.py:_footer_sync_info():3870] logging synced files