de-en model

all_results.json

@@ -0,0 +1,15 @@
+{
+    "epoch": 12.0,
+    "eval_bleu": 20.0716,
+    "eval_gen_len": 231.5305,
+    "eval_loss": 2.73199200630188,
+    "eval_runtime": 916.5779,
+    "eval_samples": 1674,
+    "eval_samples_per_second": 1.826,
+    "eval_steps_per_second": 0.229,
+    "train_loss": 0.9599274822711565,
+    "train_runtime": 12720.632,
+    "train_samples": 3349,
+    "train_samples_per_second": 3.159,
+    "train_steps_per_second": 0.395
+}

config.json

@@ -0,0 +1,59 @@
+{
+  "_name_or_path": "opus-mt-de-en",
+  "_num_labels": 3,
+  "activation_dropout": 0.0,
+  "activation_function": "swish",
+  "add_bias_logits": false,
+  "add_final_layer_norm": false,
+  "architectures": [
+    "MarianMTModel"
+  ],
+  "attention_dropout": 0.0,
+  "bad_words_ids": [
+    [
+      58100
+    ]
+  ],
+  "bos_token_id": 0,
+  "classif_dropout": 0.0,
+  "classifier_dropout": 0.0,
+  "d_model": 512,
+  "decoder_attention_heads": 8,
+  "decoder_ffn_dim": 2048,
+  "decoder_layerdrop": 0.0,
+  "decoder_layers": 6,
+  "decoder_start_token_id": 58100,
+  "dropout": 0.1,
+  "encoder_attention_heads": 8,
+  "encoder_ffn_dim": 2048,
+  "encoder_layerdrop": 0.0,
+  "encoder_layers": 6,
+  "eos_token_id": 0,
+  "forced_eos_token_id": 0,
+  "id2label": {
+    "0": "LABEL_0",
+    "1": "LABEL_1",
+    "2": "LABEL_2"
+  },
+  "init_std": 0.02,
+  "is_encoder_decoder": true,
+  "label2id": {
+    "LABEL_0": 0,
+    "LABEL_1": 1,
+    "LABEL_2": 2
+  },
+  "max_length": 512,
+  "max_position_embeddings": 512,
+  "model_type": "marian",
+  "normalize_before": false,
+  "normalize_embedding": false,
+  "num_beams": 4,
+  "num_hidden_layers": 6,
+  "pad_token_id": 58100,
+  "scale_embedding": true,
+  "static_position_embeddings": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.13.0.dev0",
+  "use_cache": true,
+  "vocab_size": 58101
+}

eval_results.json

@@ -0,0 +1,10 @@
+{
+    "epoch": 12.0,
+    "eval_bleu": 20.0716,
+    "eval_gen_len": 231.5305,
+    "eval_loss": 2.73199200630188,
+    "eval_runtime": 916.5779,
+    "eval_samples": 1674,
+    "eval_samples_per_second": 1.826,
+    "eval_steps_per_second": 0.229
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e8c2fc1df88f34a116d1d1f378f812cff8b7f429f95f9151e9b77cc13facb56
+size 295861701

special_tokens_map.json

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

tokenizer_config.json

@@ -0,0 +1 @@
+{"source_lang": "de", "target_lang": "en", "unk_token": "<unk>", "eos_token": "</s>", "pad_token": "<pad>", "model_max_length": 512, "sp_model_kwargs": {}, "special_tokens_map_file": null, "tokenizer_file": null, "name_or_path": "opus-mt-de-en", "tokenizer_class": "MarianTokenizer"}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 12.0,
+    "train_loss": 0.9599274822711565,
+    "train_runtime": 12720.632,
+    "train_samples": 3349,
+    "train_samples_per_second": 3.159,
+    "train_steps_per_second": 0.395
+}

trainer_state.json

@@ -0,0 +1,217 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 12.0,
+  "global_step": 5028,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.5833333333333334e-05,
+      "loss": 1.8646,
+      "step": 419
+    },
+    {
+      "epoch": 1.0,
+      "eval_bleu": 19.9719,
+      "eval_gen_len": 228.2151,
+      "eval_loss": 2.7007944583892822,
+      "eval_runtime": 943.6612,
+      "eval_samples_per_second": 1.774,
+      "eval_steps_per_second": 0.223,
+      "step": 419
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 1.4672,
+      "step": 838
+    },
+    {
+      "epoch": 2.0,
+      "eval_bleu": 19.5289,
+      "eval_gen_len": 226.9922,
+      "eval_loss": 2.629415273666382,
+      "eval_runtime": 944.6011,
+      "eval_samples_per_second": 1.772,
+      "eval_steps_per_second": 0.222,
+      "step": 838
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 1.254,
+      "step": 1257
+    },
+    {
+      "epoch": 3.0,
+      "eval_bleu": 19.8687,
+      "eval_gen_len": 228.9385,
+      "eval_loss": 2.6099565029144287,
+      "eval_runtime": 941.2139,
+      "eval_samples_per_second": 1.779,
+      "eval_steps_per_second": 0.223,
+      "step": 1257
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 1.1042,
+      "step": 1676
+    },
+    {
+      "epoch": 4.0,
+      "eval_bleu": 20.0491,
+      "eval_gen_len": 236.0472,
+      "eval_loss": 2.6182217597961426,
+      "eval_runtime": 963.8776,
+      "eval_samples_per_second": 1.737,
+      "eval_steps_per_second": 0.218,
+      "step": 1676
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 2.916666666666667e-05,
+      "loss": 0.9828,
+      "step": 2095
+    },
+    {
+      "epoch": 5.0,
+      "eval_bleu": 19.9041,
+      "eval_gen_len": 227.9576,
+      "eval_loss": 2.643587350845337,
+      "eval_runtime": 914.6993,
+      "eval_samples_per_second": 1.83,
+      "eval_steps_per_second": 0.23,
+      "step": 2095
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 2.5e-05,
+      "loss": 0.8701,
+      "step": 2514
+    },
+    {
+      "epoch": 6.0,
+      "eval_bleu": 19.7386,
+      "eval_gen_len": 237.0072,
+      "eval_loss": 2.644989013671875,
+      "eval_runtime": 967.0534,
+      "eval_samples_per_second": 1.731,
+      "eval_steps_per_second": 0.217,
+      "step": 2514
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.7886,
+      "step": 2933
+    },
+    {
+      "epoch": 7.0,
+      "eval_bleu": 20.0981,
+      "eval_gen_len": 234.6792,
+      "eval_loss": 2.6598494052886963,
+      "eval_runtime": 947.5271,
+      "eval_samples_per_second": 1.767,
+      "eval_steps_per_second": 0.222,
+      "step": 2933
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.7161,
+      "step": 3352
+    },
+    {
+      "epoch": 8.0,
+      "eval_bleu": 20.0204,
+      "eval_gen_len": 229.7993,
+      "eval_loss": 2.689500331878662,
+      "eval_runtime": 910.6374,
+      "eval_samples_per_second": 1.838,
+      "eval_steps_per_second": 0.231,
+      "step": 3352
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 1.25e-05,
+      "loss": 0.6671,
+      "step": 3771
+    },
+    {
+      "epoch": 9.0,
+      "eval_bleu": 19.6227,
+      "eval_gen_len": 229.2593,
+      "eval_loss": 2.7035083770751953,
+      "eval_runtime": 921.3104,
+      "eval_samples_per_second": 1.817,
+      "eval_steps_per_second": 0.228,
+      "step": 3771
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.62,
+      "step": 4190
+    },
+    {
+      "epoch": 10.0,
+      "eval_bleu": 20.0559,
+      "eval_gen_len": 229.6069,
+      "eval_loss": 2.727369785308838,
+      "eval_runtime": 913.7419,
+      "eval_samples_per_second": 1.832,
+      "eval_steps_per_second": 0.23,
+      "step": 4190
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.6028,
+      "step": 4609
+    },
+    {
+      "epoch": 11.0,
+      "eval_bleu": 19.9684,
+      "eval_gen_len": 230.6344,
+      "eval_loss": 2.731297254562378,
+      "eval_runtime": 915.0826,
+      "eval_samples_per_second": 1.829,
+      "eval_steps_per_second": 0.229,
+      "step": 4609
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 0.0,
+      "loss": 0.5815,
+      "step": 5028
+    },
+    {
+      "epoch": 12.0,
+      "eval_bleu": 20.0716,
+      "eval_gen_len": 231.5305,
+      "eval_loss": 2.73199200630188,
+      "eval_runtime": 916.305,
+      "eval_samples_per_second": 1.827,
+      "eval_steps_per_second": 0.229,
+      "step": 5028
+    },
+    {
+      "epoch": 12.0,
+      "step": 5028,
+      "total_flos": 5449229958905856.0,
+      "train_loss": 0.9599274822711565,
+      "train_runtime": 12720.632,
+      "train_samples_per_second": 3.159,
+      "train_steps_per_second": 0.395
+    }
+  ],
+  "max_steps": 5028,
+  "num_train_epochs": 12,
+  "total_flos": 5449229958905856.0,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba8227034f9e7977c8e9314aaaacbc2825e65a96b61e71117436ba00e66ccf2e
+size 2991