stage best unsampled glossary fine adapt wce model

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"
+}

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6a23ad3e8845beaf9fa4e1153e13c93fb24495bfe9ffbc2f4cd420131f4b2ce
+size 1845323269

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c91091e1d99a274f665eb51013f1f1fc8446c396c57ed4d4a36b5d104a1e768
+size 922885701

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3135b4b2b5b6421ed765b6de6788be2793682866627de507316c6f0956b67c3b
+size 14575

scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01f30a0038add223303f40959e0d8f8a5056bff6ba6c6bff2f7bcef69e68348c
+size 557

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d841877dae97f3484921bd550ae3afec1339fe3ea6e45d722c6eee90ccd95ea9
+size 627

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>"
+}

trainer_state.json

@@ -0,0 +1,176 @@
+{
+  "best_metric": 43.218,
+  "best_model_checkpoint": "opus_big_adapt_wce_gloss_unsampled_precision_2_ubweight_1.5/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": 0.1832,
+      "step": 8000
+    },
+    {
+      "epoch": 0.2,
+      "eval_bleu": 42.311,
+      "eval_gen_len": 33.3634,
+      "eval_loss": 0.0981452614068985,
+      "eval_runtime": 129.4343,
+      "eval_samples_per_second": 8.058,
+      "eval_steps_per_second": 0.51,
+      "step": 8000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.950723967960567e-05,
+      "loss": 0.152,
+      "step": 16000
+    },
+    {
+      "epoch": 0.39,
+      "eval_bleu": 42.5149,
+      "eval_gen_len": 33.8552,
+      "eval_loss": 0.09721900522708893,
+      "eval_runtime": 121.8178,
+      "eval_samples_per_second": 8.562,
+      "eval_steps_per_second": 0.542,
+      "step": 16000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9260844115834877e-05,
+      "loss": 0.1477,
+      "step": 24000
+    },
+    {
+      "epoch": 0.59,
+      "eval_bleu": 42.3053,
+      "eval_gen_len": 33.1083,
+      "eval_loss": 0.09773427993059158,
+      "eval_runtime": 113.6446,
+      "eval_samples_per_second": 9.178,
+      "eval_steps_per_second": 0.581,
+      "step": 24000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.901447935921134e-05,
+      "loss": 0.1437,
+      "step": 32000
+    },
+    {
+      "epoch": 0.79,
+      "eval_bleu": 42.7689,
+      "eval_gen_len": 34.2637,
+      "eval_loss": 0.09667054563760757,
+      "eval_runtime": 127.749,
+      "eval_samples_per_second": 8.164,
+      "eval_steps_per_second": 0.517,
+      "step": 32000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.8768114602587803e-05,
+      "loss": 0.1422,
+      "step": 40000
+    },
+    {
+      "epoch": 0.99,
+      "eval_bleu": 42.1023,
+      "eval_gen_len": 34.2656,
+      "eval_loss": 0.09601164609193802,
+      "eval_runtime": 127.2919,
+      "eval_samples_per_second": 8.194,
+      "eval_steps_per_second": 0.518,
+      "step": 40000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.8521749845964266e-05,
+      "loss": 0.1265,
+      "step": 48000
+    },
+    {
+      "epoch": 1.18,
+      "eval_bleu": 42.4637,
+      "eval_gen_len": 34.1064,
+      "eval_loss": 0.09702879190444946,
+      "eval_runtime": 125.1346,
+      "eval_samples_per_second": 8.335,
+      "eval_steps_per_second": 0.527,
+      "step": 48000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.8275415896487988e-05,
+      "loss": 0.1269,
+      "step": 56000
+    },
+    {
+      "epoch": 1.38,
+      "eval_bleu": 42.9093,
+      "eval_gen_len": 33.7507,
+      "eval_loss": 0.09716643393039703,
+      "eval_runtime": 127.4089,
+      "eval_samples_per_second": 8.186,
+      "eval_steps_per_second": 0.518,
+      "step": 56000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.8029020332717192e-05,
+      "loss": 0.1275,
+      "step": 64000
+    },
+    {
+      "epoch": 1.58,
+      "eval_bleu": 42.6221,
+      "eval_gen_len": 34.4372,
+      "eval_loss": 0.09658621996641159,
+      "eval_runtime": 249.3691,
+      "eval_samples_per_second": 4.183,
+      "eval_steps_per_second": 0.265,
+      "step": 64000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.7782655576093655e-05,
+      "loss": 0.127,
+      "step": 72000
+    },
+    {
+      "epoch": 1.77,
+      "eval_bleu": 42.3284,
+      "eval_gen_len": 34.0441,
+      "eval_loss": 0.09654054045677185,
+      "eval_runtime": 197.536,
+      "eval_samples_per_second": 5.28,
+      "eval_steps_per_second": 0.334,
+      "step": 72000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7536321626617377e-05,
+      "loss": 0.1266,
+      "step": 80000
+    },
+    {
+      "epoch": 1.97,
+      "eval_bleu": 43.218,
+      "eval_gen_len": 34.1035,
+      "eval_loss": 0.09614226967096329,
+      "eval_runtime": 124.4592,
+      "eval_samples_per_second": 8.38,
+      "eval_steps_per_second": 0.53,
+      "step": 80000
+    }
+  ],
+  "max_steps": 649200,
+  "num_train_epochs": 16,
+  "total_flos": 2.5844916958632346e+17,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c1e3e973ab9c641205085b0da2fa097a725aaa3c27d0d8b28b392b3501e503e
+size 3835