stage best big fine unsampled tfidf wce model

checkpoint-32000/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-32000/generation_config.json

@@ -0,0 +1,16 @@
+{
+  "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,
+  "renormalize_logits": true,
+  "transformers_version": "4.28.0"
+}

checkpoint-32000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b7d908550705ba53670f7691c48f9170f990bcf3907a0b1492d5a9d3515d8f0
+size 1845322757

checkpoint-32000/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6fe7f29734646ea64745e1fb9c697ec186d8002ecd41d780c0b05796245f8118
+size 922885701

checkpoint-32000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:130a4fc57c723d7800c5eac87bf28c38ce67d7d1bdf71c179873cbaabfcea0b1
+size 14639

checkpoint-32000/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97b776940b0c5587dab32e7e970f8a08b91b7b270a9b905b5c2df91aec615d84
+size 557

checkpoint-32000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:448b4a0ac8ff84d011d3c0924a8ddf1b053ab7eed882c9fbea385d7a1e97de77
+size 627

checkpoint-32000/special_tokens_map.json

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

checkpoint-32000/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-32000/trainer_state.json

@@ -0,0 +1,80 @@
+{
+  "best_metric": 42.9908,
+  "best_model_checkpoint": "opus_big_fine_tfidf_wce_unsampled/checkpoint-32000",
+  "epoch": 0.7886629698089956,
+  "global_step": 32000,
+  "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": 0.2036,
+      "step": 8000
+    },
+    {
+      "epoch": 0.2,
+      "eval_bleu": 41.9332,
+      "eval_gen_len": 33.7565,
+      "eval_loss": 0.09845346212387085,
+      "eval_runtime": 98.5797,
+      "eval_samples_per_second": 10.58,
+      "eval_steps_per_second": 0.67,
+      "step": 8000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.950723967960567e-05,
+      "loss": 0.1717,
+      "step": 16000
+    },
+    {
+      "epoch": 0.39,
+      "eval_bleu": 42.466,
+      "eval_gen_len": 34.3557,
+      "eval_loss": 0.09779991954565048,
+      "eval_runtime": 101.8795,
+      "eval_samples_per_second": 10.238,
+      "eval_steps_per_second": 0.648,
+      "step": 16000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.926090573012939e-05,
+      "loss": 0.1668,
+      "step": 24000
+    },
+    {
+      "epoch": 0.59,
+      "eval_bleu": 42.674,
+      "eval_gen_len": 33.7651,
+      "eval_loss": 0.09813275933265686,
+      "eval_runtime": 147.0398,
+      "eval_samples_per_second": 7.093,
+      "eval_steps_per_second": 0.449,
+      "step": 24000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.9014540973505854e-05,
+      "loss": 0.1623,
+      "step": 32000
+    },
+    {
+      "epoch": 0.79,
+      "eval_bleu": 42.9908,
+      "eval_gen_len": 33.8389,
+      "eval_loss": 0.09677895158529282,
+      "eval_runtime": 81.8819,
+      "eval_samples_per_second": 12.738,
+      "eval_steps_per_second": 0.806,
+      "step": 32000
+    }
+  ],
+  "max_steps": 649200,
+  "num_train_epochs": 16,
+  "total_flos": 1.0342418777269862e+17,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-32000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:318f091bf0bbc5a7ace3758915c85825376e2bde0e6c103735ea1e3a5831a158
+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 +1,16 @@
+{
+  "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,
+  "renormalize_logits": true,
+  "transformers_version": "4.28.0"
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6fe7f29734646ea64745e1fb9c697ec186d8002ecd41d780c0b05796245f8118
+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 sha256:318f091bf0bbc5a7ace3758915c85825376e2bde0e6c103735ea1e3a5831a158
+size 3771