mohammadali.sadraei
/
SuperposPrompt_Thesis
Watch 1
Star 0
Fork 0
Code Issues 0 Pull Requests 0 Releases 0 Wiki Activity
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
1 Commit
1 Branch
Tree: 6400586fcc
Branches Tags
${ item.name }
Create branch ${ searchTerm }
from '6400586fcc'
${ noResults }
SuperposPrompt_Thesis/14_thesis_run/config1.yaml

config1.yaml 2.3KB
Raw Blame History

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081 
  1. default: &default

  2.   use_tqdm: true 

  3.   random_seed: 42

  4.   base_save_path: /disks/ssd/trained_final/sing_thesis

  5.   model_name: google/t5-small-lm-adapt

  6.   project_name_prefix: sing_thesis

  7.   experiment_name_suffix: null

  8.   train_batch_size: 24

  9.   valid_batch_size: 24

  10.   remove_dropout: true

  11.   learning_rate: 0.01

  12.   weight_decay: 0.01

  13.   num_epochs: 20

  14.   peft_params: null  # no mutation

  15.   hot_modules:

  16.   - sadcl

  17.   best_finder:

  18.     save: True

  19.     metric: valid_mean

  20.     higher_better: true

  21.   tasks:

  22.   - glue:qqp

  23.   - glue:mnli

  24.   - glue:qnli

  25. 

  26. pp: &pp

  27. # - /disks/ssd/hzi_trained/hzi_cluster_t5_small_glue-mnli/10_combine_128

  28. # - /disks/ssd/hzi_trained/hzi_cluster_t5_small_glue-qqp/10_combine_128

  29. # - /disks/ssd/trained_final/cont_thesis/cont_thesis_t5_base_glue-mrpc/10_combine_128

  30. # - /disks/ssd/trained_final/cont_thesis/cont_thesis_t5_base_glue-cola/10_combine_128_simple

  31. # - /disks/ssd/trained_final/cont_thesis/cont_thesis_t5_base_glue-stsb/10_combine_128_simple

  32. # - /disks/ssd/trained_final/cont_thesis/cont_thesis_t5_base_glue-sst2/10_combine_128_simple

  33. # - /disks/ssd/trained_final/cont_thesis/cont_thesis_t5_base_superglue-rte/10_combine_128_simple

  34. # - /disks/ssd/trained_final/cont_thesis/cont_thesis_t5_base_superglue-cb/10_combine_128_simple

  35. # - /disks/ssd/trained_final/cont_thesis/cont_thesis_t5_base_superglue-copa/10_combine_128_simple

  36. 

  37. run_configs:

  38. - <<: *default

  39.   peft_params:

  40.     kind: combine

  41.     n_tokens: 50

  42.     n_comb_tokens: 128

  43.     # pretrained_paths: *pp

  44.     use_pretrained_mode: simple

  45. # - <<: *default

  46. #   peft_params:

  47. #     kind: combine

  48. #     n_tokens: 10

  49. #     n_comb_tokens: 128

  50. #     pretrained_paths: *pp

  51. #     use_pretrained_mode: gumbal

  52. # - <<: *default

  53. #   peft_params:

  54. #     kind: combine

  55. #     n_tokens: 10

  56. #     n_comb_tokens: 128

  57. #     pretrained_paths: *pp

  58. #     use_pretrained_mode: softmax

  59. #     tempreture: 0.2

  60. # - <<: *default

  61. #   peft_params:

  62. #     kind: combine

  63. #     n_tokens: 10

  64. #     n_comb_tokens: 128

  65. #     pretrained_paths: *pp

  66. #     use_pretrained_mode: softmax

  67. #     tempreture: 1.

  68. # - <<: *default

  69. #   peft_params:

  70. #     kind: combine

  71. #     n_tokens: 10

  72. #     n_comb_tokens: 128

  73. #     pretrained_paths: *pp

  74. #     use_pretrained_mode: softmax

  75. #     tempreture: 5.

  76. 

  77. # - <<: *default

  78. #   peft_params:

  79. #     kind: combine

  80. #     n_tokens: 10

  81. #     n_comb_tokens: 128