stage big random wce model

checkpoint-40000/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-40000/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-40000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b45a2d0d3baa2160da1301e665844bb20244b7cd58c74fcc34db38eaa1704a2
+size 1845322757

checkpoint-40000/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec9d3a4fd497e14f5a741f6a4418a6d889e7ac3c0b72e2807cb16cf5e5e2ea62
+size 922885701

checkpoint-40000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:494859950e5c58b5a1ce02a8b8077c14722c165cd1bdb2b75ef384847914044c
+size 14511

checkpoint-40000/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76f24763d4153b9551d4f8f782d6bc084da794458c4e8df566c4649af17639e4
+size 557

checkpoint-40000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95098a1b669814751bc1c828ede8c95c6ad2e78c2107fea3566d08e47e43c93d
+size 627

checkpoint-40000/special_tokens_map.json

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

checkpoint-40000/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-40000/trainer_state.json

@@ -0,0 +1,96 @@
+{
+  "best_metric": 42.972,
+  "best_model_checkpoint": "opus_big_wce_random/checkpoint-40000",
+  "epoch": 0.9858287122612446,
+  "global_step": 40000,
+  "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": 0.181,
+      "step": 8000
+    },
+    {
+      "epoch": 0.2,
+      "eval_bleu": 42.0536,
+      "eval_gen_len": 33.651,
+      "eval_loss": 0.09812351316213608,
+      "eval_runtime": 122.051,
+      "eval_samples_per_second": 8.546,
+      "eval_steps_per_second": 0.541,
+      "step": 8000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9507208872458414e-05,
+      "loss": 0.1503,
+      "step": 16000
+    },
+    {
+      "epoch": 0.39,
+      "eval_bleu": 42.8859,
+      "eval_gen_len": 34.6251,
+      "eval_loss": 0.09743236750364304,
+      "eval_runtime": 150.32,
+      "eval_samples_per_second": 6.939,
+      "eval_steps_per_second": 0.439,
+      "step": 16000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9260844115834877e-05,
+      "loss": 0.1461,
+      "step": 24000
+    },
+    {
+      "epoch": 0.59,
+      "eval_bleu": 42.882,
+      "eval_gen_len": 33.5225,
+      "eval_loss": 0.09730042517185211,
+      "eval_runtime": 182.9066,
+      "eval_samples_per_second": 5.702,
+      "eval_steps_per_second": 0.361,
+      "step": 24000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.901447935921134e-05,
+      "loss": 0.1422,
+      "step": 32000
+    },
+    {
+      "epoch": 0.79,
+      "eval_bleu": 42.9206,
+      "eval_gen_len": 34.2148,
+      "eval_loss": 0.09637172520160675,
+      "eval_runtime": 115.4951,
+      "eval_samples_per_second": 9.031,
+      "eval_steps_per_second": 0.571,
+      "step": 32000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.8768114602587803e-05,
+      "loss": 0.1407,
+      "step": 40000
+    },
+    {
+      "epoch": 0.99,
+      "eval_bleu": 42.972,
+      "eval_gen_len": 34.4257,
+      "eval_loss": 0.09560249000787735,
+      "eval_runtime": 129.0503,
+      "eval_samples_per_second": 8.082,
+      "eval_steps_per_second": 0.511,
+      "step": 40000
+    }
+  ],
+  "max_steps": 649200,
+  "num_train_epochs": 16,
+  "total_flos": 1.2920984272738714e+17,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-40000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f36321e2af403d823f035a76d6b3eb270e75b7a3834592135a98e1009b049fb2
+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:ec9d3a4fd497e14f5a741f6a4418a6d889e7ac3c0b72e2807cb16cf5e5e2ea62
+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:f36321e2af403d823f035a76d6b3eb270e75b7a3834592135a98e1009b049fb2
+size 3771