stage best unsampled glossary fine adapt wce model

config.json

@@ -0,0 +1,62 @@
+{
+  "_name_or_path": "Helsinki-NLP/opus-mt-en-fr",
+  "_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": [
+    [
+      59513
+    ]
+  ],
+  "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": 59513,
+  "decoder_vocab_size": 59514,
+  "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,
+  "gradient_checkpointing": false,
+  "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": 59513,
+  "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": 59514
+}

generation_config.json

@@ -0,0 +1,15 @@
+{
+  "bad_words_ids": [
+    [
+      59513
+    ]
+  ],
+  "bos_token_id": 0,
+  "decoder_start_token_id": 59513,
+  "eos_token_id": 0,
+  "forced_eos_token_id": 0,
+  "max_length": 512,
+  "num_beams": 4,
+  "pad_token_id": 59513,
+  "transformers_version": "4.28.0"
+}

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:798bcd5a03038f2dbc7b283d0b4f6cdf5caf2bdd81ccf0dfb01a644f9a64f274
+size 597026309

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e255482a74a9085ba82922f26f8859ffc90e6721e3894cffbdd99e354123d1db
+size 298763205

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2eb8608536ff2f06f043c578921ef710ce4d295e077da17100e3e91b5050e6ed
+size 14575

scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f995fcf21d7c7551be1a3db97443d6f8a7fdfaa0e74ccd9e8616d6b4fc94f56c
+size 557

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e3349122e2399c65deb4183b16094c6d030d1768282a8e017f61e944aead1de
+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 +1,336 @@
+{
+  "best_metric": 42.7722,
+  "best_model_checkpoint": "opus_base_adapt_wce_gloss_unsampled_precision_3_ubweight_1.25/checkpoint-80000",
+  "epoch": 3.943217665615142,
+  "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": 1.9753918572555207e-05,
+      "loss": 0.2076,
+      "step": 4000
+    },
+    {
+      "epoch": 0.2,
+      "eval_bleu": 35.6766,
+      "eval_gen_len": 39.8754,
+      "eval_loss": 0.10901036113500595,
+      "eval_runtime": 190.0619,
+      "eval_samples_per_second": 5.488,
+      "eval_steps_per_second": 0.174,
+      "step": 4000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.950746746845426e-05,
+      "loss": 0.1743,
+      "step": 8000
+    },
+    {
+      "epoch": 0.39,
+      "eval_bleu": 37.712,
+      "eval_gen_len": 37.931,
+      "eval_loss": 0.10654148459434509,
+      "eval_runtime": 187.1025,
+      "eval_samples_per_second": 5.574,
+      "eval_steps_per_second": 0.176,
+      "step": 8000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9261016364353314e-05,
+      "loss": 0.1672,
+      "step": 12000
+    },
+    {
+      "epoch": 0.59,
+      "eval_bleu": 41.1339,
+      "eval_gen_len": 34.9118,
+      "eval_loss": 0.10531440377235413,
+      "eval_runtime": 148.6317,
+      "eval_samples_per_second": 7.017,
+      "eval_steps_per_second": 0.222,
+      "step": 12000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.9014565260252367e-05,
+      "loss": 0.1618,
+      "step": 16000
+    },
+    {
+      "epoch": 0.79,
+      "eval_bleu": 41.355,
+      "eval_gen_len": 35.2416,
+      "eval_loss": 0.10421621054410934,
+      "eval_runtime": 124.1237,
+      "eval_samples_per_second": 8.403,
+      "eval_steps_per_second": 0.266,
+      "step": 16000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.876811415615142e-05,
+      "loss": 0.1594,
+      "step": 20000
+    },
+    {
+      "epoch": 0.99,
+      "eval_bleu": 41.6651,
+      "eval_gen_len": 35.4228,
+      "eval_loss": 0.10366757214069366,
+      "eval_runtime": 159.7806,
+      "eval_samples_per_second": 6.528,
+      "eval_steps_per_second": 0.207,
+      "step": 20000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.8521786277602524e-05,
+      "loss": 0.1477,
+      "step": 24000
+    },
+    {
+      "epoch": 1.18,
+      "eval_bleu": 41.9522,
+      "eval_gen_len": 35.0451,
+      "eval_loss": 0.10360125452280045,
+      "eval_runtime": 138.0407,
+      "eval_samples_per_second": 7.556,
+      "eval_steps_per_second": 0.239,
+      "step": 24000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.827545839905363e-05,
+      "loss": 0.1475,
+      "step": 28000
+    },
+    {
+      "epoch": 1.38,
+      "eval_bleu": 40.2121,
+      "eval_gen_len": 36.4669,
+      "eval_loss": 0.10278471559286118,
+      "eval_runtime": 128.7397,
+      "eval_samples_per_second": 8.102,
+      "eval_steps_per_second": 0.256,
+      "step": 28000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.8029130520504733e-05,
+      "loss": 0.147,
+      "step": 32000
+    },
+    {
+      "epoch": 1.58,
+      "eval_bleu": 41.9375,
+      "eval_gen_len": 35.2301,
+      "eval_loss": 0.10222012549638748,
+      "eval_runtime": 132.4166,
+      "eval_samples_per_second": 7.877,
+      "eval_steps_per_second": 0.249,
+      "step": 32000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.7782802641955836e-05,
+      "loss": 0.1456,
+      "step": 36000
+    },
+    {
+      "epoch": 1.77,
+      "eval_bleu": 40.9742,
+      "eval_gen_len": 37.1965,
+      "eval_loss": 0.10183104127645493,
+      "eval_runtime": 119.5559,
+      "eval_samples_per_second": 8.724,
+      "eval_steps_per_second": 0.276,
+      "step": 36000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7536474763406942e-05,
+      "loss": 0.1448,
+      "step": 40000
+    },
+    {
+      "epoch": 1.97,
+      "eval_bleu": 41.7528,
+      "eval_gen_len": 35.7248,
+      "eval_loss": 0.10195966809988022,
+      "eval_runtime": 123.0307,
+      "eval_samples_per_second": 8.478,
+      "eval_steps_per_second": 0.268,
+      "step": 40000
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.7290146884858045e-05,
+      "loss": 0.1372,
+      "step": 44000
+    },
+    {
+      "epoch": 2.17,
+      "eval_bleu": 42.1909,
+      "eval_gen_len": 35.2924,
+      "eval_loss": 0.1022593304514885,
+      "eval_runtime": 159.0132,
+      "eval_samples_per_second": 6.559,
+      "eval_steps_per_second": 0.208,
+      "step": 44000
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.7043757393533123e-05,
+      "loss": 0.1363,
+      "step": 48000
+    },
+    {
+      "epoch": 2.37,
+      "eval_bleu": 42.6162,
+      "eval_gen_len": 34.9866,
+      "eval_loss": 0.10216603428125381,
+      "eval_runtime": 116.2982,
+      "eval_samples_per_second": 8.968,
+      "eval_steps_per_second": 0.284,
+      "step": 48000
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.679742951498423e-05,
+      "loss": 0.1352,
+      "step": 52000
+    },
+    {
+      "epoch": 2.56,
+      "eval_bleu": 42.4525,
+      "eval_gen_len": 34.8888,
+      "eval_loss": 0.10181364417076111,
+      "eval_runtime": 134.9256,
+      "eval_samples_per_second": 7.73,
+      "eval_steps_per_second": 0.245,
+      "step": 52000
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.6551101636435333e-05,
+      "loss": 0.1355,
+      "step": 56000
+    },
+    {
+      "epoch": 2.76,
+      "eval_bleu": 41.9729,
+      "eval_gen_len": 35.9051,
+      "eval_loss": 0.10166899114847183,
+      "eval_runtime": 135.9018,
+      "eval_samples_per_second": 7.675,
+      "eval_steps_per_second": 0.243,
+      "step": 56000
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.6304773757886436e-05,
+      "loss": 0.1358,
+      "step": 60000
+    },
+    {
+      "epoch": 2.96,
+      "eval_bleu": 42.3275,
+      "eval_gen_len": 34.8514,
+      "eval_loss": 0.10106752812862396,
+      "eval_runtime": 160.3679,
+      "eval_samples_per_second": 6.504,
+      "eval_steps_per_second": 0.206,
+      "step": 60000
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.6058445879337542e-05,
+      "loss": 0.1294,
+      "step": 64000
+    },
+    {
+      "epoch": 3.15,
+      "eval_bleu": 42.2988,
+      "eval_gen_len": 34.8188,
+      "eval_loss": 0.10183101147413254,
+      "eval_runtime": 137.4561,
+      "eval_samples_per_second": 7.588,
+      "eval_steps_per_second": 0.24,
+      "step": 64000
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.5812118000788645e-05,
+      "loss": 0.1289,
+      "step": 68000
+    },
+    {
+      "epoch": 3.35,
+      "eval_bleu": 42.7527,
+      "eval_gen_len": 34.7766,
+      "eval_loss": 0.10201847553253174,
+      "eval_runtime": 136.3856,
+      "eval_samples_per_second": 7.647,
+      "eval_steps_per_second": 0.242,
+      "step": 68000
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 1.5565728509463723e-05,
+      "loss": 0.1277,
+      "step": 72000
+    },
+    {
+      "epoch": 3.55,
+      "eval_bleu": 42.3528,
+      "eval_gen_len": 35.2416,
+      "eval_loss": 0.10199479013681412,
+      "eval_runtime": 110.1197,
+      "eval_samples_per_second": 9.472,
+      "eval_steps_per_second": 0.3,
+      "step": 72000
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.5319400630914826e-05,
+      "loss": 0.1282,
+      "step": 76000
+    },
+    {
+      "epoch": 3.75,
+      "eval_bleu": 42.4538,
+      "eval_gen_len": 35.4746,
+      "eval_loss": 0.10173474997282028,
+      "eval_runtime": 126.659,
+      "eval_samples_per_second": 8.235,
+      "eval_steps_per_second": 0.261,
+      "step": 76000
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 1.5073072752365931e-05,
+      "loss": 0.129,
+      "step": 80000
+    },
+    {
+      "epoch": 3.94,
+      "eval_bleu": 42.7722,
+      "eval_gen_len": 34.8581,
+      "eval_loss": 0.10149160027503967,
+      "eval_runtime": 151.6781,
+      "eval_samples_per_second": 6.876,
+      "eval_steps_per_second": 0.218,
+      "step": 80000
+    }
+  ],
+  "max_steps": 324608,
+  "num_train_epochs": 16,
+  "total_flos": 1.3796012702564352e+17,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95403e3b8a17ad59b55b7a5c77ea0c35560a3f8a0621ae589da0ae3348a38c98
+size 3835