Stage best ailem 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 sha256:49ef285fe33721cadef5d9366ad4e7df65127db768ba214bbd970a220b23a886
+size 1845323269

checkpoint-80000/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ad3871e01e3f366a1adecf3c978774df4ce494521fbf87f670fe689d1c9f62b
+size 922885701

checkpoint-80000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2d1747e8e3e7bfc54feb68f2457c5f92d873054937c1572c6148a3187bd83604
+size 14511

checkpoint-80000/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5bcf3ecf6900daf4caf2abf0aaafae85803cec01c4337e21c56d5a754ec85631
+size 557

checkpoint-80000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3af67343aa0f8b4109961c1ed1ad806d5b136e6b0592737a57873f77d6c9558
+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": 43.1052,
+  "best_model_checkpoint": "opus_big_enfr_FT_ailem_2021/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": 0.1855,
+      "step": 8000
+    },
+    {
+      "epoch": 0.2,
+      "eval_bleu": 42.1782,
+      "eval_gen_len": 33.6491,
+      "eval_loss": 0.09836934506893158,
+      "eval_runtime": 122.585,
+      "eval_samples_per_second": 8.508,
+      "eval_steps_per_second": 0.538,
+      "step": 8000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.950723967960567e-05,
+      "loss": 0.1544,
+      "step": 16000
+    },
+    {
+      "epoch": 0.39,
+      "eval_bleu": 42.4911,
+      "eval_gen_len": 34.0786,
+      "eval_loss": 0.09750288724899292,
+      "eval_runtime": 142.1487,
+      "eval_samples_per_second": 7.337,
+      "eval_steps_per_second": 0.464,
+      "step": 16000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9260844115834877e-05,
+      "loss": 0.1501,
+      "step": 24000
+    },
+    {
+      "epoch": 0.59,
+      "eval_bleu": 42.8065,
+      "eval_gen_len": 33.7124,
+      "eval_loss": 0.09724840521812439,
+      "eval_runtime": 166.346,
+      "eval_samples_per_second": 6.27,
+      "eval_steps_per_second": 0.397,
+      "step": 24000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.901447935921134e-05,
+      "loss": 0.146,
+      "step": 32000
+    },
+    {
+      "epoch": 0.79,
+      "eval_bleu": 42.6121,
+      "eval_gen_len": 33.8476,
+      "eval_loss": 0.09660676121711731,
+      "eval_runtime": 115.7542,
+      "eval_samples_per_second": 9.01,
+      "eval_steps_per_second": 0.57,
+      "step": 32000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.8768114602587803e-05,
+      "loss": 0.1445,
+      "step": 40000
+    },
+    {
+      "epoch": 0.99,
+      "eval_bleu": 42.677,
+      "eval_gen_len": 34.2617,
+      "eval_loss": 0.09570582211017609,
+      "eval_runtime": 131.9831,
+      "eval_samples_per_second": 7.903,
+      "eval_steps_per_second": 0.5,
+      "step": 40000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.8521749845964266e-05,
+      "loss": 0.1288,
+      "step": 48000
+    },
+    {
+      "epoch": 1.18,
+      "eval_bleu": 43.0313,
+      "eval_gen_len": 34.024,
+      "eval_loss": 0.09699361026287079,
+      "eval_runtime": 123.262,
+      "eval_samples_per_second": 8.462,
+      "eval_steps_per_second": 0.535,
+      "step": 48000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.827535428219347e-05,
+      "loss": 0.1292,
+      "step": 56000
+    },
+    {
+      "epoch": 1.38,
+      "eval_bleu": 43.0386,
+      "eval_gen_len": 37.9616,
+      "eval_loss": 0.09719892591238022,
+      "eval_runtime": 189.4267,
+      "eval_samples_per_second": 5.506,
+      "eval_steps_per_second": 0.348,
+      "step": 56000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.8028989525569933e-05,
+      "loss": 0.1298,
+      "step": 64000
+    },
+    {
+      "epoch": 1.58,
+      "eval_bleu": 42.9287,
+      "eval_gen_len": 34.488,
+      "eval_loss": 0.09660369902849197,
+      "eval_runtime": 189.722,
+      "eval_samples_per_second": 5.498,
+      "eval_steps_per_second": 0.348,
+      "step": 64000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.7782624768946397e-05,
+      "loss": 0.1293,
+      "step": 72000
+    },
+    {
+      "epoch": 1.77,
+      "eval_bleu": 42.877,
+      "eval_gen_len": 34.4631,
+      "eval_loss": 0.0966239720582962,
+      "eval_runtime": 126.7169,
+      "eval_samples_per_second": 8.231,
+      "eval_steps_per_second": 0.521,
+      "step": 72000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.753626001232286e-05,
+      "loss": 0.129,
+      "step": 80000
+    },
+    {
+      "epoch": 1.97,
+      "eval_bleu": 43.1052,
+      "eval_gen_len": 34.2253,
+      "eval_loss": 0.09659728407859802,
+      "eval_runtime": 147.8544,
+      "eval_samples_per_second": 7.054,
+      "eval_steps_per_second": 0.446,
+      "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:d12adacb418dabf96d51aaf0c4442f64a5efd4ca525c4ebe20dde6e20ff047cd
+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 sha256:7ad3871e01e3f366a1adecf3c978774df4ce494521fbf87f670fe689d1c9f62b
+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:d12adacb418dabf96d51aaf0c4442f64a5efd4ca525c4ebe20dde6e20ff047cd
+size 3771