Checkpoint 2000 - Epoch 0.35

.gitattributes

@@ -39,3 +39,5 @@ checkpoints/checkpoint-1000/source.spm filter=lfs diff=lfs merge=lfs -text
 checkpoints/checkpoint-1000/target.spm filter=lfs diff=lfs merge=lfs -text
 checkpoints/checkpoint-1500/source.spm filter=lfs diff=lfs merge=lfs -text
 checkpoints/checkpoint-1500/target.spm filter=lfs diff=lfs merge=lfs -text
+checkpoints/checkpoint-2000/source.spm filter=lfs diff=lfs merge=lfs -text
+checkpoints/checkpoint-2000/target.spm filter=lfs diff=lfs merge=lfs -text

checkpoints/checkpoint-2000/config.json

@@ -0,0 +1,54 @@
+{
+  "activation_dropout": 0.0,
+  "activation_function": "swish",
+  "add_bias_logits": false,
+  "add_final_layer_norm": false,
+  "architectures": [
+    "MarianMTModel"
+  ],
+  "attention_dropout": 0.0,
+  "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": 64171,
+  "decoder_vocab_size": 64172,
+  "dropout": 0.1,
+  "dtype": "float32",
+  "encoder_attention_heads": 8,
+  "encoder_ffn_dim": 2048,
+  "encoder_layerdrop": 0.0,
+  "encoder_layers": 6,
+  "eos_token_id": 0,
+  "extra_pos_embeddings": 64172,
+  "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": null,
+  "max_position_embeddings": 512,
+  "model_type": "marian",
+  "normalize_before": false,
+  "normalize_embedding": false,
+  "num_beams": null,
+  "num_hidden_layers": 6,
+  "pad_token_id": 64171,
+  "scale_embedding": true,
+  "share_encoder_decoder_embeddings": true,
+  "static_position_embeddings": true,
+  "transformers_version": "4.57.1",
+  "use_cache": true,
+  "vocab_size": 64172
+}

checkpoints/checkpoint-2000/generation_config.json

@@ -0,0 +1,17 @@
+{
+  "bad_words_ids": [
+    [
+      64171
+    ]
+  ],
+  "decoder_start_token_id": 64171,
+  "eos_token_id": [
+    0
+  ],
+  "forced_eos_token_id": 0,
+  "max_length": 512,
+  "num_beams": 6,
+  "pad_token_id": 64171,
+  "renormalize_logits": true,
+  "transformers_version": "4.57.1"
+}

checkpoints/checkpoint-2000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d60236664788abaeb746bca0997437a40e10fd5175c2e3641ea001bdcc0f6f02
+size 308263984

checkpoints/checkpoint-2000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1fe4553fc0bf95f2599ebb5277e29543cdb37ac36301298998c16b221b4c364
+size 616171979

checkpoints/checkpoint-2000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:72968c086f8b6a5526f3abb5ea42fda5ed853ab31471912dd84655006e20b36d
+size 14645

checkpoints/checkpoint-2000/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b225555a5e94470106810084f79e74ec08e088c95c1a7741adc8ff78ff299201
+size 1383

checkpoints/checkpoint-2000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:46e73072a61710db5bc09ac8430422aceab77b869d4ba1e52a8177105e118d8d
+size 1465

checkpoints/checkpoint-2000/source.spm

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c4a99ea3602b29fbf901ade8b93a45efa3d7c64eab8fc5fa812383efa327a87d
+size 706917

checkpoints/checkpoint-2000/special_tokens_map.json

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

checkpoints/checkpoint-2000/target.spm

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6dce5fa58fcd7dde9e81e279b8c075bf42ee558278f73d6fb48e342029d7f19
+size 791194

checkpoints/checkpoint-2000/tokenizer_config.json

@@ -0,0 +1,39 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "64171": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "separate_vocabs": false,
+  "source_lang": "mul",
+  "sp_model_kwargs": {},
+  "target_lang": "eng",
+  "tokenizer_class": "MarianTokenizer",
+  "unk_token": "<unk>"
+}

checkpoints/checkpoint-2000/trainer_state.json

@@ -0,0 +1,367 @@
+{
+  "best_global_step": 2000,
+  "best_metric": 0.3400862027176887,
+  "best_model_checkpoint": "models/test-finetuned-sw-en/checkpoint-2000",
+  "epoch": 0.350385423966363,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.008759635599159075,
+      "grad_norm": 100.36164855957031,
+      "learning_rate": 2.5744308231173378e-06,
+      "loss": 14.1521,
+      "step": 50
+    },
+    {
+      "epoch": 0.01751927119831815,
+      "grad_norm": 11.487061500549316,
+      "learning_rate": 5.201401050788091e-06,
+      "loss": 5.8065,
+      "step": 100
+    },
+    {
+      "epoch": 0.026278906797477224,
+      "grad_norm": 10.166085243225098,
+      "learning_rate": 7.828371278458844e-06,
+      "loss": 4.411,
+      "step": 150
+    },
+    {
+      "epoch": 0.0350385423966363,
+      "grad_norm": 9.804460525512695,
+      "learning_rate": 1.0455341506129598e-05,
+      "loss": 4.1303,
+      "step": 200
+    },
+    {
+      "epoch": 0.04379817799579538,
+      "grad_norm": 8.944438934326172,
+      "learning_rate": 1.308231173380035e-05,
+      "loss": 4.0012,
+      "step": 250
+    },
+    {
+      "epoch": 0.05255781359495445,
+      "grad_norm": 9.135123252868652,
+      "learning_rate": 1.5709281961471103e-05,
+      "loss": 3.8622,
+      "step": 300
+    },
+    {
+      "epoch": 0.06131744919411353,
+      "grad_norm": 8.474159240722656,
+      "learning_rate": 1.8336252189141855e-05,
+      "loss": 3.749,
+      "step": 350
+    },
+    {
+      "epoch": 0.0700770847932726,
+      "grad_norm": 8.502720832824707,
+      "learning_rate": 2.096322241681261e-05,
+      "loss": 3.6698,
+      "step": 400
+    },
+    {
+      "epoch": 0.07883672039243167,
+      "grad_norm": 8.237638473510742,
+      "learning_rate": 2.3590192644483363e-05,
+      "loss": 3.5605,
+      "step": 450
+    },
+    {
+      "epoch": 0.08759635599159075,
+      "grad_norm": 8.184097290039062,
+      "learning_rate": 2.6217162872154118e-05,
+      "loss": 3.4528,
+      "step": 500
+    },
+    {
+      "epoch": 0.08759635599159075,
+      "eval_bleu": 0.18452195716495454,
+      "eval_chrf": 37.579073815894574,
+      "eval_loss": 3.3051514625549316,
+      "eval_model_preparation_time": 0.0024,
+      "eval_runtime": 312.3723,
+      "eval_samples_per_second": 68.79,
+      "eval_steps_per_second": 2.151,
+      "step": 500
+    },
+    {
+      "epoch": 0.09635599159074983,
+      "grad_norm": 8.083911895751953,
+      "learning_rate": 2.884413309982487e-05,
+      "loss": 3.4183,
+      "step": 550
+    },
+    {
+      "epoch": 0.1051156271899089,
+      "grad_norm": 7.500244617462158,
+      "learning_rate": 2.9997800888103602e-05,
+      "loss": 3.3458,
+      "step": 600
+    },
+    {
+      "epoch": 0.11387526278906797,
+      "grad_norm": 8.195356369018555,
+      "learning_rate": 2.9982937261717248e-05,
+      "loss": 3.3022,
+      "step": 650
+    },
+    {
+      "epoch": 0.12263489838822705,
+      "grad_norm": 7.267704010009766,
+      "learning_rate": 2.9954065389845778e-05,
+      "loss": 3.2882,
+      "step": 700
+    },
+    {
+      "epoch": 0.13139453398738613,
+      "grad_norm": 8.064687728881836,
+      "learning_rate": 2.9911212266146163e-05,
+      "loss": 3.1957,
+      "step": 750
+    },
+    {
+      "epoch": 0.1401541695865452,
+      "grad_norm": 7.6160688400268555,
+      "learning_rate": 2.985441795599852e-05,
+      "loss": 3.2081,
+      "step": 800
+    },
+    {
+      "epoch": 0.14891380518570427,
+      "grad_norm": 8.32299518585205,
+      "learning_rate": 2.978373555904712e-05,
+      "loss": 3.117,
+      "step": 850
+    },
+    {
+      "epoch": 0.15767344078486334,
+      "grad_norm": 7.199745178222656,
+      "learning_rate": 2.9699231159555054e-05,
+      "loss": 3.1089,
+      "step": 900
+    },
+    {
+      "epoch": 0.16643307638402244,
+      "grad_norm": 7.270870685577393,
+      "learning_rate": 2.9600983764618996e-05,
+      "loss": 3.0805,
+      "step": 950
+    },
+    {
+      "epoch": 0.1751927119831815,
+      "grad_norm": 7.554959297180176,
+      "learning_rate": 2.9489085230301778e-05,
+      "loss": 3.0778,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1751927119831815,
+      "eval_bleu": 0.2956498402636781,
+      "eval_chrf": 48.85756661559643,
+      "eval_loss": 2.9274492263793945,
+      "eval_model_preparation_time": 0.0024,
+      "eval_runtime": 292.5539,
+      "eval_samples_per_second": 73.45,
+      "eval_steps_per_second": 2.297,
+      "step": 1000
+    },
+    {
+      "epoch": 0.18395234758234058,
+      "grad_norm": 8.069275856018066,
+      "learning_rate": 2.9363640175751887e-05,
+      "loss": 3.0401,
+      "step": 1050
+    },
+    {
+      "epoch": 0.19271198318149965,
+      "grad_norm": 7.83376932144165,
+      "learning_rate": 2.922476588539015e-05,
+      "loss": 3.0312,
+      "step": 1100
+    },
+    {
+      "epoch": 0.20147161878065872,
+      "grad_norm": 7.48977518081665,
+      "learning_rate": 2.9072592199255066e-05,
+      "loss": 2.993,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2102312543798178,
+      "grad_norm": 7.3794779777526855,
+      "learning_rate": 2.8907261391609325e-05,
+      "loss": 3.0023,
+      "step": 1200
+    },
+    {
+      "epoch": 0.21899088997897687,
+      "grad_norm": 7.881261825561523,
+      "learning_rate": 2.8728928037920966e-05,
+      "loss": 2.9784,
+      "step": 1250
+    },
+    {
+      "epoch": 0.22775052557813594,
+      "grad_norm": 6.710996150970459,
+      "learning_rate": 2.853775887034356e-05,
+      "loss": 2.9662,
+      "step": 1300
+    },
+    {
+      "epoch": 0.23651016117729504,
+      "grad_norm": 7.729732990264893,
+      "learning_rate": 2.8333932621830594e-05,
+      "loss": 2.972,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2452697967764541,
+      "grad_norm": 6.742512226104736,
+      "learning_rate": 2.8117639859029685e-05,
+      "loss": 2.9373,
+      "step": 1400
+    },
+    {
+      "epoch": 0.25402943237561315,
+      "grad_norm": 6.744399547576904,
+      "learning_rate": 2.7889082804112972e-05,
+      "loss": 2.9438,
+      "step": 1450
+    },
+    {
+      "epoch": 0.26278906797477225,
+      "grad_norm": 6.606257438659668,
+      "learning_rate": 2.764847514571017e-05,
+      "loss": 2.9562,
+      "step": 1500
+    },
+    {
+      "epoch": 0.26278906797477225,
+      "eval_bleu": 0.32395940047060856,
+      "eval_chrf": 51.97366756294741,
+      "eval_loss": 2.8271327018737793,
+      "eval_model_preparation_time": 0.0024,
+      "eval_runtime": 299.5193,
+      "eval_samples_per_second": 71.742,
+      "eval_steps_per_second": 2.244,
+      "step": 1500
+    },
+    {
+      "epoch": 0.27154870357393135,
+      "grad_norm": 7.149145603179932,
+      "learning_rate": 2.7396041839121136e-05,
+      "loss": 2.9199,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2803083391730904,
+      "grad_norm": 7.462961673736572,
+      "learning_rate": 2.7132018895994697e-05,
+      "loss": 2.9294,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2890679747722495,
+      "grad_norm": 6.644341468811035,
+      "learning_rate": 2.685665316367035e-05,
+      "loss": 2.9246,
+      "step": 1650
+    },
+    {
+      "epoch": 0.29782761037140854,
+      "grad_norm": 6.9484429359436035,
+      "learning_rate": 2.6570202094389226e-05,
+      "loss": 2.9154,
+      "step": 1700
+    },
+    {
+      "epoch": 0.30658724597056763,
+      "grad_norm": 6.633726119995117,
+      "learning_rate": 2.6272933504589965e-05,
+      "loss": 2.8688,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3153468815697267,
+      "grad_norm": 7.676279544830322,
+      "learning_rate": 2.5965125324514702e-05,
+      "loss": 2.9137,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3241065171688858,
+      "grad_norm": 6.953280448913574,
+      "learning_rate": 2.564706533835911e-05,
+      "loss": 2.8974,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3328661527680449,
+      "grad_norm": 7.126858711242676,
+      "learning_rate": 2.5319050915209592e-05,
+      "loss": 2.9095,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3416257883672039,
+      "grad_norm": 6.376485824584961,
+      "learning_rate": 2.498138873101906e-05,
+      "loss": 2.8332,
+      "step": 1950
+    },
+    {
+      "epoch": 0.350385423966363,
+      "grad_norm": 7.114526748657227,
+      "learning_rate": 2.4634394481881312e-05,
+      "loss": 2.8781,
+      "step": 2000
+    },
+    {
+      "epoch": 0.350385423966363,
+      "eval_bleu": 0.3400862027176887,
+      "eval_chrf": 53.63135686349859,
+      "eval_loss": 2.7728657722473145,
+      "eval_model_preparation_time": 0.0024,
+      "eval_runtime": 459.505,
+      "eval_samples_per_second": 46.763,
+      "eval_steps_per_second": 1.462,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 50,
+  "max_steps": 5708,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.001
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1521367092559872.0,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoints/checkpoint-2000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3276994b4acd8c7aa25b182674cf1a04c7b0b3da28d839d1f1e286be4336383
+size 5969