stage big antagonistic wce model

checkpoint-80000/config.json

@@ -0,0 +1,46 @@
+{
+  "_name_or_path": "Helsinki-NLP/opus-mt-tc-big-en-fr",
+  "activation_dropout": 0.0,
+  "activation_function": "relu",
+  "architectures": [
+    "MarianMTModel"
+  ],
+  "attention_dropout": 0.0,
+  "bad_words_ids": [
+    [
+      53016
+    ]
+  ],
+  "bos_token_id": 0,
+  "classifier_dropout": 0.0,
+  "d_model": 1024,
+  "decoder_attention_heads": 16,
+  "decoder_ffn_dim": 4096,
+  "decoder_layerdrop": 0.0,
+  "decoder_layers": 6,
+  "decoder_start_token_id": 53016,
+  "decoder_vocab_size": 53017,
+  "dropout": 0.1,
+  "encoder_attention_heads": 16,
+  "encoder_ffn_dim": 4096,
+  "encoder_layerdrop": 0.0,
+  "encoder_layers": 6,
+  "eos_token_id": 43311,
+  "forced_eos_token_id": 43311,
+  "init_std": 0.02,
+  "is_encoder_decoder": true,
+  "max_length": 512,
+  "max_position_embeddings": 1024,
+  "model_type": "marian",
+  "normalize_embedding": false,
+  "num_beams": 4,
+  "num_hidden_layers": 6,
+  "pad_token_id": 53016,
+  "scale_embedding": true,
+  "share_encoder_decoder_embeddings": true,
+  "static_position_embeddings": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.28.0",
+  "use_cache": true,
+  "vocab_size": 53017
+}

checkpoint-80000/generation_config.json

@@ -0,0 +1,15 @@
+{
+  "bad_words_ids": [
+    [
+      53016
+    ]
+  ],
+  "bos_token_id": 0,
+  "decoder_start_token_id": 53016,
+  "eos_token_id": 43311,
+  "forced_eos_token_id": 43311,
+  "max_length": 512,
+  "num_beams": 4,
+  "pad_token_id": 53016,
+  "transformers_version": "4.28.0"
+}

checkpoint-80000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c9969af5213c1a33c7159412e86f552c71f6f0c6edaa6ff4bcaefd52821059e
+size 1845323269

checkpoint-80000/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c088c0183210a88f59fc36cbb0de8974d61b155f1442288c084ee6fdb3e6212
+size 922885701

checkpoint-80000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:570459e5297d9cfab841648d3a1490a0b22134d74778e3738065a1e2dfc9e879
+size 14511

checkpoint-80000/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fee1ca2c1f88e141257854ec538c62fac978cd9804204bb9d959c886a3694ad1
+size 557

checkpoint-80000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2d280d247d94ac98ee856efffc708e91babbdb3306480c0c30168fc32146e59
+size 627

checkpoint-80000/special_tokens_map.json

@@ -0,0 +1,5 @@
+{
+  "eos_token": "</s>",
+  "pad_token": "<pad>",
+  "unk_token": "<unk>"
+}

checkpoint-80000/tokenizer_config.json

@@ -0,0 +1,12 @@
+{
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "</s>",
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "separate_vocabs": false,
+  "source_lang": "en",
+  "sp_model_kwargs": {},
+  "target_lang": "fr",
+  "tokenizer_class": "MarianTokenizer",
+  "unk_token": "<unk>"
+}

checkpoint-80000/trainer_state.json

@@ -0,0 +1,176 @@
+{
+  "best_metric": 43.4149,
+  "best_model_checkpoint": "opus_big_wce_antagonistic/checkpoint-80000",
+  "epoch": 1.9716574245224892,
+  "global_step": 80000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9753635243376465e-05,
+      "loss": 0.1851,
+      "step": 8000
+    },
+    {
+      "epoch": 0.2,
+      "eval_bleu": 42.2483,
+      "eval_gen_len": 33.7728,
+      "eval_loss": 0.09822206944227219,
+      "eval_runtime": 154.7836,
+      "eval_samples_per_second": 6.738,
+      "eval_steps_per_second": 0.426,
+      "step": 8000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9507208872458414e-05,
+      "loss": 0.1541,
+      "step": 16000
+    },
+    {
+      "epoch": 0.39,
+      "eval_bleu": 42.4928,
+      "eval_gen_len": 35.2013,
+      "eval_loss": 0.09750045835971832,
+      "eval_runtime": 126.2477,
+      "eval_samples_per_second": 8.262,
+      "eval_steps_per_second": 0.523,
+      "step": 16000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9260874922982132e-05,
+      "loss": 0.1499,
+      "step": 24000
+    },
+    {
+      "epoch": 0.59,
+      "eval_bleu": 42.3243,
+      "eval_gen_len": 33.5944,
+      "eval_loss": 0.09780784696340561,
+      "eval_runtime": 167.1696,
+      "eval_samples_per_second": 6.239,
+      "eval_steps_per_second": 0.395,
+      "step": 24000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.90145101663586e-05,
+      "loss": 0.1458,
+      "step": 32000
+    },
+    {
+      "epoch": 0.79,
+      "eval_bleu": 42.6185,
+      "eval_gen_len": 34.1783,
+      "eval_loss": 0.09651421010494232,
+      "eval_runtime": 122.2884,
+      "eval_samples_per_second": 8.529,
+      "eval_steps_per_second": 0.54,
+      "step": 32000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.876814540973506e-05,
+      "loss": 0.1443,
+      "step": 40000
+    },
+    {
+      "epoch": 0.99,
+      "eval_bleu": 42.436,
+      "eval_gen_len": 33.7565,
+      "eval_loss": 0.09578333050012589,
+      "eval_runtime": 123.1521,
+      "eval_samples_per_second": 8.469,
+      "eval_steps_per_second": 0.536,
+      "step": 40000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.852181146025878e-05,
+      "loss": 0.1286,
+      "step": 48000
+    },
+    {
+      "epoch": 1.18,
+      "eval_bleu": 42.5804,
+      "eval_gen_len": 34.0719,
+      "eval_loss": 0.09721639007329941,
+      "eval_runtime": 123.0622,
+      "eval_samples_per_second": 8.475,
+      "eval_steps_per_second": 0.536,
+      "step": 48000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.8275446703635243e-05,
+      "loss": 0.129,
+      "step": 56000
+    },
+    {
+      "epoch": 1.38,
+      "eval_bleu": 39.5838,
+      "eval_gen_len": 38.1592,
+      "eval_loss": 0.09705326706171036,
+      "eval_runtime": 187.8123,
+      "eval_samples_per_second": 5.553,
+      "eval_steps_per_second": 0.351,
+      "step": 56000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.802908194701171e-05,
+      "loss": 0.1296,
+      "step": 64000
+    },
+    {
+      "epoch": 1.58,
+      "eval_bleu": 42.7604,
+      "eval_gen_len": 34.4928,
+      "eval_loss": 0.09692135453224182,
+      "eval_runtime": 120.9945,
+      "eval_samples_per_second": 8.62,
+      "eval_steps_per_second": 0.545,
+      "step": 64000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.7782717190388173e-05,
+      "loss": 0.1291,
+      "step": 72000
+    },
+    {
+      "epoch": 1.77,
+      "eval_bleu": 42.5798,
+      "eval_gen_len": 34.0029,
+      "eval_loss": 0.0967508926987648,
+      "eval_runtime": 156.0025,
+      "eval_samples_per_second": 6.686,
+      "eval_steps_per_second": 0.423,
+      "step": 72000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7536352433764633e-05,
+      "loss": 0.1288,
+      "step": 80000
+    },
+    {
+      "epoch": 1.97,
+      "eval_bleu": 43.4149,
+      "eval_gen_len": 34.3289,
+      "eval_loss": 0.09629645943641663,
+      "eval_runtime": 113.2277,
+      "eval_samples_per_second": 9.212,
+      "eval_steps_per_second": 0.583,
+      "step": 80000
+    }
+  ],
+  "max_steps": 649200,
+  "num_train_epochs": 16,
+  "total_flos": 2.5844916958632346e+17,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-80000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c03829d7304a5ffb5384a19a365fff8e2d04781b01281f173571135e5838279
+size 3771

config.json

@@ -0,0 +1,46 @@
+{
+  "_name_or_path": "Helsinki-NLP/opus-mt-tc-big-en-fr",
+  "activation_dropout": 0.0,
+  "activation_function": "relu",
+  "architectures": [
+    "MarianMTModel"
+  ],
+  "attention_dropout": 0.0,
+  "bad_words_ids": [
+    [
+      53016
+    ]
+  ],
+  "bos_token_id": 0,
+  "classifier_dropout": 0.0,
+  "d_model": 1024,
+  "decoder_attention_heads": 16,
+  "decoder_ffn_dim": 4096,
+  "decoder_layerdrop": 0.0,
+  "decoder_layers": 6,
+  "decoder_start_token_id": 53016,
+  "decoder_vocab_size": 53017,
+  "dropout": 0.1,
+  "encoder_attention_heads": 16,
+  "encoder_ffn_dim": 4096,
+  "encoder_layerdrop": 0.0,
+  "encoder_layers": 6,
+  "eos_token_id": 43311,
+  "forced_eos_token_id": 43311,
+  "init_std": 0.02,
+  "is_encoder_decoder": true,
+  "max_length": 512,
+  "max_position_embeddings": 1024,
+  "model_type": "marian",
+  "normalize_embedding": false,
+  "num_beams": 4,
+  "num_hidden_layers": 6,
+  "pad_token_id": 53016,
+  "scale_embedding": true,
+  "share_encoder_decoder_embeddings": true,
+  "static_position_embeddings": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.28.0",
+  "use_cache": true,
+  "vocab_size": 53017
+}

generation_config.json

@@ -0,0 +1,15 @@
+{
+  "bad_words_ids": [
+    [
+      53016
+    ]
+  ],
+  "bos_token_id": 0,
+  "decoder_start_token_id": 53016,
+  "eos_token_id": 43311,
+  "forced_eos_token_id": 43311,
+  "max_length": 512,
+  "num_beams": 4,
+  "pad_token_id": 53016,
+  "transformers_version": "4.28.0"
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c088c0183210a88f59fc36cbb0de8974d61b155f1442288c084ee6fdb3e6212
+size 922885701

special_tokens_map.json

@@ -0,0 +1,5 @@
+{
+  "eos_token": "</s>",
+  "pad_token": "<pad>",
+  "unk_token": "<unk>"
+}

tokenizer_config.json

@@ -0,0 +1,12 @@
+{
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "</s>",
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "separate_vocabs": false,
+  "source_lang": "en",
+  "sp_model_kwargs": {},
+  "target_lang": "fr",
+  "tokenizer_class": "MarianTokenizer",
+  "unk_token": "<unk>"
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c03829d7304a5ffb5384a19a365fff8e2d04781b01281f173571135e5838279
+size 3771