Hugging Face's logo

coml
/
childlex

Safetensors
Model card Files
xet
 Community
nhamilakis commited on
Commit
0e24e4b
·
verified ·
1 Parent(s): 42f9aec

Upload folder using huggingface_hub (part 2)

Browse files
Files changed (38) hide show
  1. stela/EN/30/01/orthographic_gpt2/generation_config.json +6 -0
  2. stela/EN/30/01/orthographic_gpt2/model.safetensors +3 -0
  3. stela/EN/30/01/orthographic_gpt2/training.logs +20 -0
  4. stela/EN/30/01/orthographic_gpt2/training_args.bin +3 -0
  5. stela/EN/30/01/orthographic_lstm/config.json +13 -0
  6. stela/EN/30/01/orthographic_lstm/generation_config.json +4 -0
  7. stela/EN/30/01/orthographic_lstm/model.safetensors +3 -0
  8. stela/EN/30/01/orthographic_lstm/training.logs +19 -0
  9. stela/EN/30/01/orthographic_lstm/training_args.bin +3 -0
  10. stela/EN/40/00/orthographic_gpt2/config.json +32 -0
  11. stela/EN/40/00/orthographic_gpt2/generation_config.json +6 -0
  12. stela/EN/40/00/orthographic_gpt2/model.safetensors +3 -0
  13. stela/EN/40/00/orthographic_gpt2/training.logs +20 -0
  14. stela/EN/40/00/orthographic_gpt2/training_args.bin +3 -0
  15. stela/EN/40/00/orthographic_lstm/config.json +13 -0
  16. stela/EN/40/00/orthographic_lstm/generation_config.json +4 -0
  17. stela/EN/40/00/orthographic_lstm/model.safetensors +3 -0
  18. stela/EN/40/00/orthographic_lstm/training.logs +19 -0
  19. stela/EN/40/00/orthographic_lstm/training_args.bin +3 -0
  20. stela/EN/50/00/orthographic_gpt2/config.json +32 -0
  21. stela/EN/50/00/orthographic_gpt2/generation_config.json +6 -0
  22. stela/EN/50/00/orthographic_gpt2/model.safetensors +3 -0
  23. stela/EN/50/00/orthographic_gpt2/training.logs +38 -0
  24. stela/EN/50/00/orthographic_gpt2/training_args.bin +3 -0
  25. stela/EN/50/00/orthographic_lstm/config.json +13 -0
  26. stela/EN/50/00/orthographic_lstm/generation_config.json +4 -0
  27. stela/EN/50/00/orthographic_lstm/model.safetensors +3 -0
  28. stela/EN/50/00/orthographic_lstm/training.logs +19 -0
  29. stela/EN/50/00/orthographic_lstm/training_args.bin +3 -0
  30. stela/EN/60/00/orthographic_gpt2/config.json +32 -0
  31. stela/EN/60/00/orthographic_gpt2/generation_config.json +6 -0
  32. stela/EN/60/00/orthographic_gpt2/model.safetensors +3 -0
  33. stela/EN/60/00/orthographic_gpt2/training.logs +38 -0
  34. stela/EN/60/00/orthographic_gpt2/training_args.bin +3 -0
  35. stela/EN/60/00/orthographic_lstm/config.json +13 -0
  36. stela/EN/60/00/orthographic_lstm/model.safetensors +3 -0
  37. stela/EN/60/00/orthographic_lstm/training.logs +18 -0
  38. stela/EN/60/00/orthographic_lstm/training_args.bin +3 -0
stela/EN/30/01/orthographic_gpt2/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": 50256,
5
+ "transformers_version": "4.49.0"
6
+ }
stela/EN/30/01/orthographic_gpt2/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:67dcd332e3f493cabd571eac6c6e82c8b42625537cb45a4bc36a99f2117af7b5
3
+ size 343737984
stela/EN/30/01/orthographic_gpt2/training.logs ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 05/28/2025 17:02:55 - lexical_benchmark.train.run_train - INFO - Loading gpt2 trainer class.
2
+ 05/28/2025 17:03:07 - datasets - INFO - PyTorch version 2.5.1 available.
3
+ 05/28/2025 17:03:07 - datasets - INFO - Polars version 1.26.0 available.
4
+ 05/28/2025 17:03:07 - lexical_benchmark.train.train_params - INFO - Loading params from /lustre/fsn1/projects/rech/hhb/uye44va/workspace/lm_benchmark/prod/code/src/lexical_benchmark/train/train-params.toml
5
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.gpt2 - INFO - Loading char tokenizer
6
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.gpt2 - INFO - Tokenizer loaded with vocabulary size: 115
7
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.gpt2 - INFO - Loading and processing datasets with joined utterances (max_length=1024)
8
+ 05/28/2025 17:03:10 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/30/01/train.tokenized
9
+ 05/28/2025 17:03:11 - lexical_benchmark.train.tokenizers - INFO - Read 1125142 lines from file
10
+ 05/28/2025 17:05:25 - lexical_benchmark.train.tokenizers - INFO - Created 110565 joined examples from 1125142 original lines
11
+ 05/28/2025 17:05:43 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/dev/dev.tokenized
12
+ 05/28/2025 17:05:43 - lexical_benchmark.train.tokenizers - INFO - Read 179806 lines from file
13
+ 05/28/2025 17:06:03 - lexical_benchmark.train.tokenizers - INFO - Created 18373 joined examples from 179806 original lines
14
+ 05/28/2025 17:06:06 - lexical_benchmark.train.trainers.gpt2 - INFO - Training dataset size: 110565
15
+ 05/28/2025 17:06:06 - lexical_benchmark.train.trainers.gpt2 - INFO - Validation dataset size: 18373
16
+ 05/28/2025 17:06:06 - lexical_benchmark.train.trainers.gpt2 - INFO - Creating GPT2 configuration
17
+ 05/28/2025 17:06:07 - lexical_benchmark.train.trainers.gpt2 - INFO - Creating standard Trainer
18
+ 05/28/2025 17:06:08 - lexical_benchmark.train.run_train - INFO - Running training of stela_gpt2_EN_30_01
19
+ 05/29/2025 02:07:31 - lexical_benchmark.train.run_train - INFO - Training of stela_gpt2_EN_30_01 has completed !
20
+ 05/29/2025 02:07:32 - lexical_benchmark.train.run_train - INFO - Model saved to /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/models/stela/EN/30/01/gpt2
stela/EN/30/01/orthographic_gpt2/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15e0b1aa73a92584466468e1600d39eb99a8ef1e18a9fade13b293d48562dfe7
3
+ size 5496
stela/EN/30/01/orthographic_lstm/config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LSTMForLanguageModeling"
4
+ ],
5
+ "dropout": 0.1,
6
+ "embedding_dim": 200,
7
+ "hidden_size": 1024,
8
+ "model_type": "lstm",
9
+ "num_layers": 3,
10
+ "torch_dtype": "float32",
11
+ "transformers_version": "4.49.0",
12
+ "vocab_size": 115
13
+ }
stela/EN/30/01/orthographic_lstm/generation_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "transformers_version": "4.49.0"
4
+ }
stela/EN/30/01/orthographic_lstm/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5d6e3b6ca5c9f5a411da6e43a8a9e33f5a2af0dea576a4d0ff8710bc9cbd2236
3
+ size 87825996
stela/EN/30/01/orthographic_lstm/training.logs ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 05/28/2025 17:02:55 - lexical_benchmark.train.run_train - INFO - Loading lstm trainer class.
2
+ 05/28/2025 17:03:07 - datasets - INFO - PyTorch version 2.5.1 available.
3
+ 05/28/2025 17:03:07 - datasets - INFO - Polars version 1.26.0 available.
4
+ 05/28/2025 17:03:07 - lexical_benchmark.train.train_params - INFO - Loading params from /lustre/fsn1/projects/rech/hhb/uye44va/workspace/lm_benchmark/prod/code/src/lexical_benchmark/train/train-params.toml
5
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.lstm - INFO - Loading char-tokenizer
6
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.lstm - INFO - Tokenizer loaded with vocabulary size: 115
7
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.lstm - INFO - Loading and processing datasets with joined utterances (max_length=1024)
8
+ 05/28/2025 17:03:11 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/30/01/train.tokenized
9
+ 05/28/2025 17:03:11 - lexical_benchmark.train.tokenizers - INFO - Read 1125142 lines from file
10
+ 05/28/2025 17:05:10 - lexical_benchmark.train.tokenizers - INFO - Created 110565 joined examples from 1125142 original lines
11
+ 05/28/2025 17:05:26 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/dev/dev.tokenized
12
+ 05/28/2025 17:05:26 - lexical_benchmark.train.tokenizers - INFO - Read 179806 lines from file
13
+ 05/28/2025 17:05:43 - lexical_benchmark.train.tokenizers - INFO - Created 18373 joined examples from 179806 original lines
14
+ 05/28/2025 17:05:46 - lexical_benchmark.train.trainers.lstm - INFO - Training dataset size: 110565
15
+ 05/28/2025 17:05:46 - lexical_benchmark.train.trainers.lstm - INFO - Validation dataset size: 18373
16
+ 05/28/2025 17:05:46 - lexical_benchmark.train.trainers.lstm - INFO - Loading configurations & initialising LSTM model trainer
17
+ 05/28/2025 17:05:46 - lexical_benchmark.train.run_train - INFO - Running training of stela_lstm_EN_30_01
18
+ 05/29/2025 02:03:07 - lexical_benchmark.train.run_train - INFO - Training of stela_lstm_EN_30_01 has completed !
19
+ 05/29/2025 02:03:07 - lexical_benchmark.train.run_train - INFO - Model saved to /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/models/stela/EN/30/01/lstm
stela/EN/30/01/orthographic_lstm/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9804e0e2243170280dc6d5b9156bb8b1d513eaff93f03894eacfc33d4eb10cd3
3
+ size 5496
stela/EN/40/00/orthographic_gpt2/config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_function": "gelu_new",
3
+ "architectures": [
4
+ "GPT2LMHeadModel"
5
+ ],
6
+ "attn_pdrop": 0.1,
7
+ "bos_token_id": 50256,
8
+ "embd_pdrop": 0.1,
9
+ "eos_token_id": 50256,
10
+ "initializer_range": 0.02,
11
+ "layer_norm_epsilon": 1e-05,
12
+ "model_type": "gpt2",
13
+ "n_ctx": 1024,
14
+ "n_embd": 768,
15
+ "n_head": 12,
16
+ "n_inner": null,
17
+ "n_layer": 12,
18
+ "n_positions": 1024,
19
+ "reorder_and_upcast_attn": false,
20
+ "resid_pdrop": 0.1,
21
+ "scale_attn_by_inverse_layer_idx": false,
22
+ "scale_attn_weights": true,
23
+ "summary_activation": null,
24
+ "summary_first_dropout": 0.1,
25
+ "summary_proj_to_labels": true,
26
+ "summary_type": "cls_index",
27
+ "summary_use_proj": true,
28
+ "torch_dtype": "float32",
29
+ "transformers_version": "4.49.0",
30
+ "use_cache": true,
31
+ "vocab_size": 115
32
+ }
stela/EN/40/00/orthographic_gpt2/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": 50256,
5
+ "transformers_version": "4.49.0"
6
+ }
stela/EN/40/00/orthographic_gpt2/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:44a7588c948ad1be2e57c9c16e2bf6b74e63a0c7202270737d27b6ddb8b00865
3
+ size 343737984
stela/EN/40/00/orthographic_gpt2/training.logs ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 05/28/2025 17:02:55 - lexical_benchmark.train.run_train - INFO - Loading gpt2 trainer class.
2
+ 05/28/2025 17:03:06 - datasets - INFO - PyTorch version 2.5.1 available.
3
+ 05/28/2025 17:03:06 - datasets - INFO - Polars version 1.26.0 available.
4
+ 05/28/2025 17:03:07 - lexical_benchmark.train.train_params - INFO - Loading params from /lustre/fsn1/projects/rech/hhb/uye44va/workspace/lm_benchmark/prod/code/src/lexical_benchmark/train/train-params.toml
5
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.gpt2 - INFO - Loading char tokenizer
6
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.gpt2 - INFO - Tokenizer loaded with vocabulary size: 115
7
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.gpt2 - INFO - Loading and processing datasets with joined utterances (max_length=1024)
8
+ 05/28/2025 17:03:11 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/40/00/train.tokenized
9
+ 05/28/2025 17:03:11 - lexical_benchmark.train.tokenizers - INFO - Read 1489134 lines from file
10
+ 05/28/2025 17:05:53 - lexical_benchmark.train.tokenizers - INFO - Created 146119 joined examples from 1489134 original lines
11
+ 05/28/2025 17:06:14 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/dev/dev.tokenized
12
+ 05/28/2025 17:06:14 - lexical_benchmark.train.tokenizers - INFO - Read 179806 lines from file
13
+ 05/28/2025 17:06:31 - lexical_benchmark.train.tokenizers - INFO - Created 18373 joined examples from 179806 original lines
14
+ 05/28/2025 17:06:33 - lexical_benchmark.train.trainers.gpt2 - INFO - Training dataset size: 146119
15
+ 05/28/2025 17:06:33 - lexical_benchmark.train.trainers.gpt2 - INFO - Validation dataset size: 18373
16
+ 05/28/2025 17:06:33 - lexical_benchmark.train.trainers.gpt2 - INFO - Creating GPT2 configuration
17
+ 05/28/2025 17:06:34 - lexical_benchmark.train.trainers.gpt2 - INFO - Creating standard Trainer
18
+ 05/28/2025 17:06:34 - lexical_benchmark.train.run_train - INFO - Running training of stela_gpt2_EN_40_00
19
+ 05/29/2025 07:24:00 - lexical_benchmark.train.run_train - INFO - Training of stela_gpt2_EN_40_00 has completed !
20
+ 05/29/2025 07:24:00 - lexical_benchmark.train.run_train - INFO - Model saved to /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/models/stela/EN/40/00/gpt2
stela/EN/40/00/orthographic_gpt2/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb5eb1117af918ce8cf1d8e93faf46e29eb81c0e7f74ee3860c8536b252cd62e
3
+ size 5496
stela/EN/40/00/orthographic_lstm/config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LSTMForLanguageModeling"
4
+ ],
5
+ "dropout": 0.1,
6
+ "embedding_dim": 200,
7
+ "hidden_size": 1024,
8
+ "model_type": "lstm",
9
+ "num_layers": 3,
10
+ "torch_dtype": "float32",
11
+ "transformers_version": "4.49.0",
12
+ "vocab_size": 115
13
+ }
stela/EN/40/00/orthographic_lstm/generation_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "transformers_version": "4.49.0"
4
+ }
stela/EN/40/00/orthographic_lstm/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:973608333cf7bd4bfb5e36e8774eb452132e0d8da852aad2202e54f5814b8d04
3
+ size 87825996
stela/EN/40/00/orthographic_lstm/training.logs ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 05/28/2025 17:02:55 - lexical_benchmark.train.run_train - INFO - Loading lstm trainer class.
2
+ 05/28/2025 17:03:07 - datasets - INFO - PyTorch version 2.5.1 available.
3
+ 05/28/2025 17:03:07 - datasets - INFO - Polars version 1.26.0 available.
4
+ 05/28/2025 17:03:07 - lexical_benchmark.train.train_params - INFO - Loading params from /lustre/fsn1/projects/rech/hhb/uye44va/workspace/lm_benchmark/prod/code/src/lexical_benchmark/train/train-params.toml
5
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.lstm - INFO - Loading char-tokenizer
6
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.lstm - INFO - Tokenizer loaded with vocabulary size: 115
7
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.lstm - INFO - Loading and processing datasets with joined utterances (max_length=1024)
8
+ 05/28/2025 17:03:11 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/40/00/train.tokenized
9
+ 05/28/2025 17:03:11 - lexical_benchmark.train.tokenizers - INFO - Read 1489134 lines from file
10
+ 05/28/2025 17:06:17 - lexical_benchmark.train.tokenizers - INFO - Created 146119 joined examples from 1489134 original lines
11
+ 05/28/2025 17:06:42 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/dev/dev.tokenized
12
+ 05/28/2025 17:06:42 - lexical_benchmark.train.tokenizers - INFO - Read 179806 lines from file
13
+ 05/28/2025 17:07:01 - lexical_benchmark.train.tokenizers - INFO - Created 18373 joined examples from 179806 original lines
14
+ 05/28/2025 17:07:04 - lexical_benchmark.train.trainers.lstm - INFO - Training dataset size: 146119
15
+ 05/28/2025 17:07:04 - lexical_benchmark.train.trainers.lstm - INFO - Validation dataset size: 18373
16
+ 05/28/2025 17:07:04 - lexical_benchmark.train.trainers.lstm - INFO - Loading configurations & initialising LSTM model trainer
17
+ 05/28/2025 17:07:05 - lexical_benchmark.train.run_train - INFO - Running training of stela_lstm_EN_40_00
18
+ 05/29/2025 02:20:34 - lexical_benchmark.train.run_train - INFO - Training of stela_lstm_EN_40_00 has completed !
19
+ 05/29/2025 02:20:34 - lexical_benchmark.train.run_train - INFO - Model saved to /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/models/stela/EN/40/00/lstm
stela/EN/40/00/orthographic_lstm/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bbeffdb4b354e07c0cb35a30f3225990fdaf81ba7825c4e6b4c1850fcec4d9cc
3
+ size 5496
stela/EN/50/00/orthographic_gpt2/config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_function": "gelu_new",
3
+ "architectures": [
4
+ "GPT2LMHeadModel"
5
+ ],
6
+ "attn_pdrop": 0.1,
7
+ "bos_token_id": 50256,
8
+ "embd_pdrop": 0.1,
9
+ "eos_token_id": 50256,
10
+ "initializer_range": 0.02,
11
+ "layer_norm_epsilon": 1e-05,
12
+ "model_type": "gpt2",
13
+ "n_ctx": 1024,
14
+ "n_embd": 768,
15
+ "n_head": 12,
16
+ "n_inner": null,
17
+ "n_layer": 12,
18
+ "n_positions": 1024,
19
+ "reorder_and_upcast_attn": false,
20
+ "resid_pdrop": 0.1,
21
+ "scale_attn_by_inverse_layer_idx": false,
22
+ "scale_attn_weights": true,
23
+ "summary_activation": null,
24
+ "summary_first_dropout": 0.1,
25
+ "summary_proj_to_labels": true,
26
+ "summary_type": "cls_index",
27
+ "summary_use_proj": true,
28
+ "torch_dtype": "float32",
29
+ "transformers_version": "4.49.0",
30
+ "use_cache": true,
31
+ "vocab_size": 115
32
+ }
stela/EN/50/00/orthographic_gpt2/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": 50256,
5
+ "transformers_version": "4.49.0"
6
+ }
stela/EN/50/00/orthographic_gpt2/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cbc45627a3b13c86be116f2b0c5cf9a0a465861e810de090c09b97ded756578f
3
+ size 343737984
stela/EN/50/00/orthographic_gpt2/training.logs ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 05/28/2025 17:02:55 - lexical_benchmark.train.run_train - INFO - Loading gpt2 trainer class.
2
+ 05/28/2025 17:03:06 - datasets - INFO - PyTorch version 2.5.1 available.
3
+ 05/28/2025 17:03:06 - datasets - INFO - Polars version 1.26.0 available.
4
+ 05/28/2025 17:03:07 - lexical_benchmark.train.train_params - INFO - Loading params from /lustre/fsn1/projects/rech/hhb/uye44va/workspace/lm_benchmark/prod/code/src/lexical_benchmark/train/train-params.toml
5
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.gpt2 - INFO - Loading char tokenizer
6
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.gpt2 - INFO - Tokenizer loaded with vocabulary size: 115
7
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.gpt2 - INFO - Loading and processing datasets with joined utterances (max_length=1024)
8
+ 05/28/2025 17:03:11 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/50/00/train.tokenized
9
+ 05/28/2025 17:03:11 - lexical_benchmark.train.tokenizers - INFO - Read 0 lines from file
10
+ 05/28/2025 17:03:11 - lexical_benchmark.train.tokenizers - INFO - Created 0 joined examples from 0 original lines
11
+ 05/28/2025 17:03:11 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/dev/dev.tokenized
12
+ 05/28/2025 17:03:12 - lexical_benchmark.train.tokenizers - INFO - Read 179806 lines from file
13
+ 05/28/2025 17:03:28 - lexical_benchmark.train.tokenizers - INFO - Created 18373 joined examples from 179806 original lines
14
+ 05/28/2025 17:03:30 - lexical_benchmark.train.trainers.gpt2 - INFO - Training dataset size: 0
15
+ 05/28/2025 17:03:30 - lexical_benchmark.train.trainers.gpt2 - INFO - Validation dataset size: 18373
16
+ 05/28/2025 17:03:30 - lexical_benchmark.train.trainers.gpt2 - INFO - Creating GPT2 configuration
17
+ 05/28/2025 17:03:31 - lexical_benchmark.train.trainers.gpt2 - INFO - Creating standard Trainer
18
+ 05/28/2025 17:03:32 - lexical_benchmark.train.run_train - INFO - Running training of stela_gpt2_EN_50_00
19
+ 06/03/2025 15:20:47 - lexical_benchmark.train.run_train - INFO - Loading gpt2 trainer class.
20
+ 06/03/2025 15:20:59 - datasets - INFO - PyTorch version 2.5.1 available.
21
+ 06/03/2025 15:20:59 - datasets - INFO - Polars version 1.26.0 available.
22
+ 06/03/2025 15:21:00 - lexical_benchmark.train.train_params - INFO - Loading params from /lustre/fsn1/projects/rech/hhb/uye44va/workspace/lm_benchmark/prod/code/src/lexical_benchmark/train/train-params.toml
23
+ 06/03/2025 15:21:00 - lexical_benchmark.train.trainers.gpt2 - INFO - Loading char tokenizer
24
+ 06/03/2025 15:21:00 - lexical_benchmark.train.trainers.gpt2 - INFO - Tokenizer loaded with vocabulary size: 115
25
+ 06/03/2025 15:21:00 - lexical_benchmark.train.trainers.gpt2 - INFO - Loading and processing datasets with joined utterances (max_length=1024)
26
+ 06/03/2025 15:21:00 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/50/00/train.tokenized
27
+ 06/03/2025 15:21:00 - lexical_benchmark.train.tokenizers - INFO - Read 1873535 lines from file
28
+ 06/03/2025 15:24:28 - lexical_benchmark.train.tokenizers - INFO - Created 182900 joined examples from 1873535 original lines
29
+ 06/03/2025 15:24:54 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/dev/dev.tokenized
30
+ 06/03/2025 15:24:55 - lexical_benchmark.train.tokenizers - INFO - Read 179806 lines from file
31
+ 06/03/2025 15:25:11 - lexical_benchmark.train.tokenizers - INFO - Created 18373 joined examples from 179806 original lines
32
+ 06/03/2025 15:25:13 - lexical_benchmark.train.trainers.gpt2 - INFO - Training dataset size: 182900
33
+ 06/03/2025 15:25:13 - lexical_benchmark.train.trainers.gpt2 - INFO - Validation dataset size: 18373
34
+ 06/03/2025 15:25:13 - lexical_benchmark.train.trainers.gpt2 - INFO - Creating GPT2 configuration
35
+ 06/03/2025 15:25:14 - lexical_benchmark.train.trainers.gpt2 - INFO - Creating standard Trainer
36
+ 06/03/2025 15:25:15 - lexical_benchmark.train.run_train - INFO - Running training of stela_gpt2_EN_50_00
37
+ 06/04/2025 07:20:00 - lexical_benchmark.train.run_train - INFO - Training of stela_gpt2_EN_50_00 has completed !
38
+ 06/04/2025 07:20:00 - lexical_benchmark.train.run_train - INFO - Model saved to /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/models/stela/EN/50/00/gpt2
stela/EN/50/00/orthographic_gpt2/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e86e2f522e8ab58ea050f8ac1159afc4c04229ad5337df52e7ee585cac277114
3
+ size 5496
stela/EN/50/00/orthographic_lstm/config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LSTMForLanguageModeling"
4
+ ],
5
+ "dropout": 0.1,
6
+ "embedding_dim": 200,
7
+ "hidden_size": 1024,
8
+ "model_type": "lstm",
9
+ "num_layers": 3,
10
+ "torch_dtype": "float32",
11
+ "transformers_version": "4.49.0",
12
+ "vocab_size": 115
13
+ }
stela/EN/50/00/orthographic_lstm/generation_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "transformers_version": "4.49.0"
4
+ }
stela/EN/50/00/orthographic_lstm/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1757aad48a06a962bae6eed959449f96feeade190f6032f2a82dfe371139a84
3
+ size 87825996
stela/EN/50/00/orthographic_lstm/training.logs ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 05/28/2025 17:02:55 - lexical_benchmark.train.run_train - INFO - Loading lstm trainer class.
2
+ 05/28/2025 17:03:07 - datasets - INFO - PyTorch version 2.5.1 available.
3
+ 05/28/2025 17:03:07 - datasets - INFO - Polars version 1.26.0 available.
4
+ 05/28/2025 17:03:07 - lexical_benchmark.train.train_params - INFO - Loading params from /lustre/fsn1/projects/rech/hhb/uye44va/workspace/lm_benchmark/prod/code/src/lexical_benchmark/train/train-params.toml
5
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.lstm - INFO - Loading char-tokenizer
6
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.lstm - INFO - Tokenizer loaded with vocabulary size: 115
7
+ 05/28/2025 17:03:07 - lexical_benchmark.train.trainers.lstm - INFO - Loading and processing datasets with joined utterances (max_length=1024)
8
+ 05/28/2025 17:03:12 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/50/00/train.tokenized
9
+ 05/28/2025 17:03:12 - lexical_benchmark.train.tokenizers - INFO - Read 1873535 lines from file
10
+ 05/28/2025 17:06:37 - lexical_benchmark.train.tokenizers - INFO - Created 182900 joined examples from 1873535 original lines
11
+ 05/28/2025 17:07:04 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/dev/dev.tokenized
12
+ 05/28/2025 17:07:04 - lexical_benchmark.train.tokenizers - INFO - Read 179806 lines from file
13
+ 05/28/2025 17:07:20 - lexical_benchmark.train.tokenizers - INFO - Created 18373 joined examples from 179806 original lines
14
+ 05/28/2025 17:07:23 - lexical_benchmark.train.trainers.lstm - INFO - Training dataset size: 182900
15
+ 05/28/2025 17:07:23 - lexical_benchmark.train.trainers.lstm - INFO - Validation dataset size: 18373
16
+ 05/28/2025 17:07:23 - lexical_benchmark.train.trainers.lstm - INFO - Loading configurations & initialising LSTM model trainer
17
+ 05/28/2025 17:07:23 - lexical_benchmark.train.run_train - INFO - Running training of stela_lstm_EN_50_00
18
+ 05/29/2025 01:56:26 - lexical_benchmark.train.run_train - INFO - Training of stela_lstm_EN_50_00 has completed !
19
+ 05/29/2025 01:56:26 - lexical_benchmark.train.run_train - INFO - Model saved to /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/models/stela/EN/50/00/lstm
stela/EN/50/00/orthographic_lstm/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4959ece801e057b0adbb2db7c83225ea80c0352b668eab2ada67d82bb39b4307
3
+ size 5496
stela/EN/60/00/orthographic_gpt2/config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_function": "gelu_new",
3
+ "architectures": [
4
+ "GPT2LMHeadModel"
5
+ ],
6
+ "attn_pdrop": 0.1,
7
+ "bos_token_id": 50256,
8
+ "embd_pdrop": 0.1,
9
+ "eos_token_id": 50256,
10
+ "initializer_range": 0.02,
11
+ "layer_norm_epsilon": 1e-05,
12
+ "model_type": "gpt2",
13
+ "n_ctx": 1024,
14
+ "n_embd": 768,
15
+ "n_head": 12,
16
+ "n_inner": null,
17
+ "n_layer": 12,
18
+ "n_positions": 1024,
19
+ "reorder_and_upcast_attn": false,
20
+ "resid_pdrop": 0.1,
21
+ "scale_attn_by_inverse_layer_idx": false,
22
+ "scale_attn_weights": true,
23
+ "summary_activation": null,
24
+ "summary_first_dropout": 0.1,
25
+ "summary_proj_to_labels": true,
26
+ "summary_type": "cls_index",
27
+ "summary_use_proj": true,
28
+ "torch_dtype": "float32",
29
+ "transformers_version": "4.49.0",
30
+ "use_cache": true,
31
+ "vocab_size": 115
32
+ }
stela/EN/60/00/orthographic_gpt2/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": 50256,
5
+ "transformers_version": "4.49.0"
6
+ }
stela/EN/60/00/orthographic_gpt2/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9af5252af8158675374e1f43a0b445cb6b29f23166666ab092c5062b3477c3e4
3
+ size 343737984
stela/EN/60/00/orthographic_gpt2/training.logs ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 05/29/2025 22:58:29 - lexical_benchmark.train.run_train - INFO - Loading gpt2 trainer class.
2
+ 05/29/2025 22:58:34 - datasets - INFO - PyTorch version 2.5.1 available.
3
+ 05/29/2025 22:58:34 - datasets - INFO - Polars version 1.26.0 available.
4
+ 05/29/2025 22:58:34 - lexical_benchmark.train.train_params - INFO - Loading params from /lustre/fsn1/projects/rech/hhb/uye44va/workspace/lm_benchmark/prod/code/src/lexical_benchmark/train/train-params.toml
5
+ 05/29/2025 22:58:34 - lexical_benchmark.train.trainers.gpt2 - INFO - Loading char tokenizer
6
+ 05/29/2025 22:58:34 - lexical_benchmark.train.trainers.gpt2 - INFO - Tokenizer loaded with vocabulary size: 115
7
+ 05/29/2025 22:58:34 - lexical_benchmark.train.trainers.gpt2 - INFO - Loading and processing datasets with joined utterances (max_length=1024)
8
+ 05/29/2025 22:58:34 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/60/00/train.tokenized
9
+ 05/29/2025 22:58:34 - lexical_benchmark.train.tokenizers - INFO - Read 2244318 lines from file
10
+ 05/29/2025 23:02:51 - lexical_benchmark.train.tokenizers - INFO - Created 219952 joined examples from 2244318 original lines
11
+ 05/29/2025 23:03:25 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/dev/dev.tokenized
12
+ 05/29/2025 23:03:25 - lexical_benchmark.train.tokenizers - INFO - Read 179806 lines from file
13
+ 05/29/2025 23:03:41 - lexical_benchmark.train.tokenizers - INFO - Created 18373 joined examples from 179806 original lines
14
+ 05/29/2025 23:03:44 - lexical_benchmark.train.trainers.gpt2 - INFO - Training dataset size: 219952
15
+ 05/29/2025 23:03:44 - lexical_benchmark.train.trainers.gpt2 - INFO - Validation dataset size: 18373
16
+ 05/29/2025 23:03:44 - lexical_benchmark.train.trainers.gpt2 - INFO - Creating GPT2 configuration
17
+ 05/29/2025 23:03:45 - lexical_benchmark.train.trainers.gpt2 - INFO - Creating standard Trainer
18
+ 05/29/2025 23:03:45 - lexical_benchmark.train.run_train - INFO - Running training of stela_gpt2_EN_60_00
19
+ 05/29/2025 23:08:21 - lexical_benchmark.train.run_train - INFO - Loading gpt2 trainer class.
20
+ 05/29/2025 23:08:32 - datasets - INFO - PyTorch version 2.5.1 available.
21
+ 05/29/2025 23:08:32 - datasets - INFO - Polars version 1.26.0 available.
22
+ 05/29/2025 23:08:33 - lexical_benchmark.train.train_params - INFO - Loading params from /lustre/fsn1/projects/rech/hhb/uye44va/workspace/lm_benchmark/prod/code/src/lexical_benchmark/train/train-params.toml
23
+ 05/29/2025 23:08:33 - lexical_benchmark.train.trainers.gpt2 - INFO - Loading char tokenizer
24
+ 05/29/2025 23:08:33 - lexical_benchmark.train.trainers.gpt2 - INFO - Tokenizer loaded with vocabulary size: 115
25
+ 05/29/2025 23:08:33 - lexical_benchmark.train.trainers.gpt2 - INFO - Loading and processing datasets with joined utterances (max_length=1024)
26
+ 05/29/2025 23:08:33 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/60/00/train.tokenized
27
+ 05/29/2025 23:08:33 - lexical_benchmark.train.tokenizers - INFO - Read 2244318 lines from file
28
+ 05/29/2025 23:12:45 - lexical_benchmark.train.tokenizers - INFO - Created 219952 joined examples from 2244318 original lines
29
+ 05/29/2025 23:13:18 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/dev/dev.tokenized
30
+ 05/29/2025 23:13:18 - lexical_benchmark.train.tokenizers - INFO - Read 179806 lines from file
31
+ 05/29/2025 23:13:34 - lexical_benchmark.train.tokenizers - INFO - Created 18373 joined examples from 179806 original lines
32
+ 05/29/2025 23:13:37 - lexical_benchmark.train.trainers.gpt2 - INFO - Training dataset size: 219952
33
+ 05/29/2025 23:13:37 - lexical_benchmark.train.trainers.gpt2 - INFO - Validation dataset size: 18373
34
+ 05/29/2025 23:13:37 - lexical_benchmark.train.trainers.gpt2 - INFO - Creating GPT2 configuration
35
+ 05/29/2025 23:13:38 - lexical_benchmark.train.trainers.gpt2 - INFO - Creating standard Trainer
36
+ 05/29/2025 23:13:38 - lexical_benchmark.train.run_train - INFO - Running training of stela_gpt2_EN_60_00
37
+ 05/30/2025 17:47:02 - lexical_benchmark.train.run_train - INFO - Training of stela_gpt2_EN_60_00 has completed !
38
+ 05/30/2025 17:47:02 - lexical_benchmark.train.run_train - INFO - Model saved to /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/models/stela/EN/60/00/gpt2
stela/EN/60/00/orthographic_gpt2/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1d377d205ece785b31cb186b01f4af676daf68f5128f8b2d13e60ab50ea43350
3
+ size 5496
stela/EN/60/00/orthographic_lstm/config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LSTMForLanguageModeling"
4
+ ],
5
+ "dropout": 0.1,
6
+ "embedding_dim": 200,
7
+ "hidden_size": 1024,
8
+ "model_type": "lstm",
9
+ "num_layers": 3,
10
+ "torch_dtype": "float32",
11
+ "transformers_version": "4.49.0",
12
+ "vocab_size": 115
13
+ }
stela/EN/60/00/orthographic_lstm/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:43a907aba22645b88c73e112ef38344d7a7ed684e58d8cf0a911b6fd494c04be
3
+ size 87825996
stela/EN/60/00/orthographic_lstm/training.logs ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 06/03/2025 15:20:47 - lexical_benchmark.train.run_train - INFO - Loading lstm trainer class.
2
+ 06/03/2025 15:20:59 - datasets - INFO - PyTorch version 2.5.1 available.
3
+ 06/03/2025 15:20:59 - datasets - INFO - Polars version 1.26.0 available.
4
+ 06/03/2025 15:21:00 - lexical_benchmark.train.train_params - INFO - Loading params from /lustre/fsn1/projects/rech/hhb/uye44va/workspace/lm_benchmark/prod/code/src/lexical_benchmark/train/train-params.toml
5
+ 06/03/2025 15:21:00 - lexical_benchmark.train.trainers.lstm - INFO - Loading char-tokenizer
6
+ 06/03/2025 15:21:00 - lexical_benchmark.train.trainers.lstm - INFO - Tokenizer loaded with vocabulary size: 115
7
+ 06/03/2025 15:21:00 - lexical_benchmark.train.trainers.lstm - INFO - Loading and processing datasets with joined utterances (max_length=1024)
8
+ 06/03/2025 15:21:00 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/60/00/train.tokenized
9
+ 06/03/2025 15:21:00 - lexical_benchmark.train.tokenizers - INFO - Read 2244318 lines from file
10
+ 06/03/2025 15:25:24 - lexical_benchmark.train.tokenizers - INFO - Created 219952 joined examples from 2244318 original lines
11
+ 06/03/2025 15:25:58 - lexical_benchmark.train.tokenizers - INFO - Reading text from /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/datasets/stela/by_size/EN/dev/dev.tokenized
12
+ 06/03/2025 15:25:58 - lexical_benchmark.train.tokenizers - INFO - Read 179806 lines from file
13
+ 06/03/2025 15:26:15 - lexical_benchmark.train.tokenizers - INFO - Created 18373 joined examples from 179806 original lines
14
+ 06/03/2025 15:26:17 - lexical_benchmark.train.trainers.lstm - INFO - Training dataset size: 219952
15
+ 06/03/2025 15:26:17 - lexical_benchmark.train.trainers.lstm - INFO - Validation dataset size: 18373
16
+ 06/03/2025 15:26:17 - lexical_benchmark.train.trainers.lstm - INFO - Loading configurations & initialising LSTM model trainer
17
+ 06/03/2025 15:26:18 - lexical_benchmark.train.run_train - INFO - Resuming previous training using /lustre/fswork/projects/rech/hhb/commun/lexical-benchmark/models/stela/EN/60/00/lstm/checkpoint-100000
18
+ 06/03/2025 15:26:18 - lexical_benchmark.train.run_train - INFO - Running training of stela_lstm_EN_60_00
stela/EN/60/00/orthographic_lstm/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4959ece801e057b0adbb2db7c83225ea80c0352b668eab2ada67d82bb39b4307
3
+ size 5496