Upload folder using huggingface_hub

checkpoint-19272/config.json

@@ -0,0 +1,58 @@
+{
+  "activation_dropout": 0.1,
+  "activation_function": "gelu",
+  "add_bias_logits": false,
+  "add_final_layer_norm": true,
+  "architectures": [
+    "MBartForConditionalGeneration"
+  ],
+  "attention_dropout": 0.1,
+  "bos_token_id": 0,
+  "classif_dropout": 0.1,
+  "classifier_dropout": 0.0,
+  "d_model": 768,
+  "decoder_attention_heads": 12,
+  "decoder_ffn_dim": 3072,
+  "decoder_layerdrop": 0.0,
+  "decoder_layers": 6,
+  "decoder_start_token_id": 2,
+  "do_blenderbot_90_layernorm": false,
+  "dropout": 0.1,
+  "early_stopping": null,
+  "encoder_attention_heads": 12,
+  "encoder_ffn_dim": 3072,
+  "encoder_layerdrop": 0.0,
+  "encoder_layers": 6,
+  "eos_token_id": 2,
+  "extra_pos_embeddings": 2,
+  "force_bos_token_to_be_generated": false,
+  "forced_eos_token_id": 2,
+  "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_position_embeddings": 1024,
+  "model_type": "mbart",
+  "no_repeat_ngram_size": null,
+  "normalize_before": true,
+  "normalize_embedding": true,
+  "num_beams": null,
+  "num_hidden_layers": 6,
+  "pad_token_id": 1,
+  "scale_embedding": false,
+  "static_position_embeddings": false,
+  "tokenizer_class": "BarthezTokenizer",
+  "torch_dtype": "float32",
+  "transformers_version": "4.51.3",
+  "use_cache": true,
+  "vocab_size": 86001
+}

checkpoint-19272/generation_config.json

@@ -0,0 +1,8 @@
+{
+  "bos_token_id": 0,
+  "decoder_start_token_id": 2,
+  "eos_token_id": 2,
+  "forced_eos_token_id": 2,
+  "pad_token_id": 1,
+  "transformers_version": "4.51.3"
+}

checkpoint-19272/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:870d1aee170048624b89cd2a763f825ed09b064bc6b5453a613a768834bffda0
+size 667849236

checkpoint-19272/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f46c919b5e4ee90657a253df32f8c91cc95b518958a6c21d6e798fd0f0bbc5c
+size 1964811

checkpoint-19272/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b5c9a321d911642f4f39e094bc71f7f606a7c0729b70323a4e97b3ed73f476b
+size 14645

checkpoint-19272/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:788952817d7e20534058a75dacbdd9323070be90cfa509634b49d011404572c0
+size 1383

checkpoint-19272/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a317addef3c200d7a4f262c74fdc4a21ceba55a5883e574984d5309cb5ed2ad9
+size 1465

checkpoint-19272/special_tokens_map.json

@@ -0,0 +1,9 @@
+{
+  "pad_token": {
+    "content": "[PAD]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-19272/tokenizer_config.json

@@ -0,0 +1,25 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "86000": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": true,
+  "extra_special_tokens": {},
+  "model_max_length": 128,
+  "pad_token": "[PAD]",
+  "tokenizer_class": "PreTrainedTokenizer"
+}

checkpoint-19272/trainer_state.json

@@ -0,0 +1,444 @@
+{
+  "best_global_step": 19272,
+  "best_metric": 0.029430464545251396,
+  "best_model_checkpoint": "checkpoints/checkpoint-19272",
+  "epoch": 2.9999610849515506,
+  "eval_steps": 1606,
+  "global_step": 19272,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.07783009689847063,
+      "grad_norm": 0.2925865948200226,
+      "learning_rate": 4.991766654390158e-05,
+      "loss": 2.2108,
+      "step": 500
+    },
+    {
+      "epoch": 0.15566019379694127,
+      "grad_norm": 0.11884481459856033,
+      "learning_rate": 4.966988962239142e-05,
+      "loss": 0.3886,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2334902906954119,
+      "grad_norm": 0.09392867982387543,
+      "learning_rate": 4.925831306092169e-05,
+      "loss": 0.2271,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2499902712378877,
+      "eval_avg": 26.749637281629344,
+      "eval_cer": 0.6640905490705347,
+      "eval_der": 78.87487905513909,
+      "eval_loss": 0.13775022327899933,
+      "eval_runtime": 1439.866,
+      "eval_samples_per_second": 94.261,
+      "eval_steps_per_second": 0.369,
+      "eval_wer": 0.7099422406784007,
+      "step": 1606
+    },
+    {
+      "epoch": 0.31132038759388253,
+      "grad_norm": 0.0717502012848854,
+      "learning_rate": 4.8685669587267704e-05,
+      "loss": 0.1657,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3891504844923532,
+      "grad_norm": 0.0638226792216301,
+      "learning_rate": 4.795576135848184e-05,
+      "loss": 0.1334,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4669805813908238,
+      "grad_norm": 0.05441045016050339,
+      "learning_rate": 4.707343471586959e-05,
+      "loss": 0.1135,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4999805424757754,
+      "eval_avg": 26.867260983703588,
+      "eval_cer": 0.6597599235984197,
+      "eval_der": 79.24906581788363,
+      "eval_loss": 0.07722582668066025,
+      "eval_runtime": 1442.4649,
+      "eval_samples_per_second": 94.091,
+      "eval_steps_per_second": 0.368,
+      "eval_wer": 0.6929572096287142,
+      "step": 3212
+    },
+    {
+      "epoch": 0.5448106782892945,
+      "grad_norm": 0.050636373460292816,
+      "learning_rate": 4.604454800693874e-05,
+      "loss": 0.1002,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6226407751877651,
+      "grad_norm": 0.0581156350672245,
+      "learning_rate": 4.48759326879731e-05,
+      "loss": 0.0906,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7004708720862357,
+      "grad_norm": 0.050500743091106415,
+      "learning_rate": 4.3575347965496405e-05,
+      "loss": 0.0828,
+      "step": 4500
+    },
+    {
+      "epoch": 0.749970813713663,
+      "eval_avg": 26.888053782081823,
+      "eval_cer": 0.6586451367564623,
+      "eval_der": 79.3192047037025,
+      "eval_loss": 0.05902250111103058,
+      "eval_runtime": 1431.2383,
+      "eval_samples_per_second": 94.829,
+      "eval_steps_per_second": 0.371,
+      "eval_wer": 0.6863115057865005,
+      "step": 4818
+    },
+    {
+      "epoch": 0.7783009689847064,
+      "grad_norm": 0.04594194516539574,
+      "learning_rate": 4.9989118673869795e-05,
+      "loss": 0.0765,
+      "step": 5000
+    },
+    {
+      "epoch": 0.856131065883177,
+      "grad_norm": 0.04987065866589546,
+      "learning_rate": 4.984611214237347e-05,
+      "loss": 0.0724,
+      "step": 5500
+    },
+    {
+      "epoch": 0.9339611627816476,
+      "grad_norm": 0.04362853616476059,
+      "learning_rate": 4.953813591358179e-05,
+      "loss": 0.0683,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9999610849515508,
+      "eval_avg": 0.708712044715695,
+      "eval_cer": 0.6546741642910598,
+      "eval_der": 0.7934032551594553,
+      "eval_loss": 0.049141544848680496,
+      "eval_runtime": 305.8639,
+      "eval_samples_per_second": 443.737,
+      "eval_steps_per_second": 1.736,
+      "eval_wer": 0.6780587146965699,
+      "step": 6424
+    },
+    {
+      "epoch": 1.0118301747285676,
+      "grad_norm": 0.04417261481285095,
+      "learning_rate": 4.99981315851881e-05,
+      "loss": 0.0626,
+      "step": 6500
+    },
+    {
+      "epoch": 1.0896602716270383,
+      "grad_norm": 0.039167486131191254,
+      "learning_rate": 4.9890257744619245e-05,
+      "loss": 0.061,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1674903685255087,
+      "grad_norm": 0.035239290446043015,
+      "learning_rate": 4.961712109503609e-05,
+      "loss": 0.0583,
+      "step": 7500
+    },
+    {
+      "epoch": 1.2453204654239793,
+      "grad_norm": 0.03707383945584297,
+      "learning_rate": 4.918053517048949e-05,
+      "loss": 0.056,
+      "step": 8000
+    },
+    {
+      "epoch": 1.2499902712378876,
+      "eval_avg": 0.7079419812877328,
+      "eval_cer": 0.6543333833811898,
+      "eval_der": 0.793713077370868,
+      "eval_loss": 0.043703265488147736,
+      "eval_runtime": 307.3798,
+      "eval_samples_per_second": 441.548,
+      "eval_steps_per_second": 1.728,
+      "eval_wer": 0.6757794831111406,
+      "step": 8030
+    },
+    {
+      "epoch": 1.32315056232245,
+      "grad_norm": 0.035194575786590576,
+      "learning_rate": 4.8583398752382485e-05,
+      "loss": 0.0539,
+      "step": 8500
+    },
+    {
+      "epoch": 1.4009806592209206,
+      "grad_norm": 0.03347332403063774,
+      "learning_rate": 4.782967662255196e-05,
+      "loss": 0.0523,
+      "step": 9000
+    },
+    {
+      "epoch": 1.4788107561193913,
+      "grad_norm": 0.06645191460847855,
+      "learning_rate": 4.692437323847159e-05,
+      "loss": 0.0507,
+      "step": 9500
+    },
+    {
+      "epoch": 1.4999805424757753,
+      "eval_avg": 0.7073388677263619,
+      "eval_cer": 0.6541002942380697,
+      "eval_der": 0.7938397323961762,
+      "eval_loss": 0.03928952291607857,
+      "eval_runtime": 307.3123,
+      "eval_samples_per_second": 441.645,
+      "eval_steps_per_second": 1.728,
+      "eval_wer": 0.6740765765448399,
+      "step": 9636
+    },
+    {
+      "epoch": 1.556640853017862,
+      "grad_norm": 0.031485334038734436,
+      "learning_rate": 4.587349950536374e-05,
+      "loss": 0.0492,
+      "step": 10000
+    },
+    {
+      "epoch": 1.6344709499163326,
+      "grad_norm": 0.031592607498168945,
+      "learning_rate": 4.46840328658421e-05,
+      "loss": 0.0481,
+      "step": 10500
+    },
+    {
+      "epoch": 1.7123010468148032,
+      "grad_norm": 0.031952131539583206,
+      "learning_rate": 4.336387097207617e-05,
+      "loss": 0.0468,
+      "step": 11000
+    },
+    {
+      "epoch": 1.749970813713663,
+      "eval_avg": 0.7068964281422255,
+      "eval_cer": 0.6539495257646198,
+      "eval_der": 0.7937636951733035,
+      "eval_loss": 0.036437951028347015,
+      "eval_runtime": 308.1694,
+      "eval_samples_per_second": 440.417,
+      "eval_steps_per_second": 1.723,
+      "eval_wer": 0.6729760634887532,
+      "step": 11242
+    },
+    {
+      "epoch": 1.7901311437132739,
+      "grad_norm": 0.030866818502545357,
+      "learning_rate": 4.9978063925651226e-05,
+      "loss": 0.0457,
+      "step": 11500
+    },
+    {
+      "epoch": 1.8679612406117445,
+      "grad_norm": 0.029339410364627838,
+      "learning_rate": 4.980989359026424e-05,
+      "loss": 0.0451,
+      "step": 12000
+    },
+    {
+      "epoch": 1.9457913375102152,
+      "grad_norm": 0.02953779138624668,
+      "learning_rate": 4.94769940363958e-05,
+      "loss": 0.0439,
+      "step": 12500
+    },
+    {
+      "epoch": 1.9999610849515508,
+      "eval_avg": 0.706664474285978,
+      "eval_cer": 0.6538952007882788,
+      "eval_der": 0.7938883608526655,
+      "eval_loss": 0.0344935841858387,
+      "eval_runtime": 304.0677,
+      "eval_samples_per_second": 446.358,
+      "eval_steps_per_second": 1.746,
+      "eval_wer": 0.6722098612169899,
+      "step": 12848
+    },
+    {
+      "epoch": 2.0236603494571352,
+      "grad_norm": 0.028502434492111206,
+      "learning_rate": 4.898157560336646e-05,
+      "loss": 0.043,
+      "step": 13000
+    },
+    {
+      "epoch": 2.101490446355606,
+      "grad_norm": 0.030778545886278152,
+      "learning_rate": 4.832692770033863e-05,
+      "loss": 0.0417,
+      "step": 13500
+    },
+    {
+      "epoch": 2.1793205432540765,
+      "grad_norm": 0.0269626472145319,
+      "learning_rate": 4.7517396965763394e-05,
+      "loss": 0.041,
+      "step": 14000
+    },
+    {
+      "epoch": 2.2499902712378876,
+      "eval_avg": 0.7064455694431908,
+      "eval_cer": 0.6538135050224774,
+      "eval_der": 0.7939745658437141,
+      "eval_loss": 0.032875534147024155,
+      "eval_runtime": 304.6489,
+      "eval_samples_per_second": 445.506,
+      "eval_steps_per_second": 1.743,
+      "eval_wer": 0.6715486374633808,
+      "step": 14454
+    },
+    {
+      "epoch": 2.257150640152547,
+      "grad_norm": 0.03818966820836067,
+      "learning_rate": 4.6558358407185946e-05,
+      "loss": 0.0402,
+      "step": 14500
+    },
+    {
+      "epoch": 2.3349807370510174,
+      "grad_norm": 0.029628828167915344,
+      "learning_rate": 4.5456179713031586e-05,
+      "loss": 0.0395,
+      "step": 15000
+    },
+    {
+      "epoch": 2.4128108339494885,
+      "grad_norm": 0.02792350761592388,
+      "learning_rate": 4.4218178973329804e-05,
+      "loss": 0.0392,
+      "step": 15500
+    },
+    {
+      "epoch": 2.4906409308479587,
+      "grad_norm": 0.03364017978310585,
+      "learning_rate": 4.2852576090096216e-05,
+      "loss": 0.0384,
+      "step": 16000
+    },
+    {
+      "epoch": 2.4999805424757753,
+      "eval_avg": 0.7061941356307186,
+      "eval_cer": 0.6537081045699137,
+      "eval_der": 0.7939727975362082,
+      "eval_loss": 0.031167298555374146,
+      "eval_runtime": 303.0246,
+      "eval_samples_per_second": 447.894,
+      "eval_steps_per_second": 1.752,
+      "eval_wer": 0.670901504786034,
+      "step": 16060
+    },
+    {
+      "epoch": 2.5684710277464298,
+      "grad_norm": 0.032605357468128204,
+      "learning_rate": 4.136843819999117e-05,
+      "loss": 0.038,
+      "step": 16500
+    },
+    {
+      "epoch": 2.6463011246449,
+      "grad_norm": 0.027949590235948563,
+      "learning_rate": 3.977561947162954e-05,
+      "loss": 0.0376,
+      "step": 17000
+    },
+    {
+      "epoch": 2.724131221543371,
+      "grad_norm": 0.026584528386592865,
+      "learning_rate": 3.80846956772673e-05,
+      "loss": 0.0372,
+      "step": 17500
+    },
+    {
+      "epoch": 2.749970813713663,
+      "eval_avg": 0.706001258948505,
+      "eval_cer": 0.6536609451579365,
+      "eval_der": 0.7939834810607371,
+      "eval_loss": 0.030004331842064857,
+      "eval_runtime": 304.4012,
+      "eval_samples_per_second": 445.869,
+      "eval_steps_per_second": 1.744,
+      "eval_wer": 0.6703593506268415,
+      "step": 17666
+    },
+    {
+      "epoch": 2.8019613184418413,
+      "grad_norm": 0.027681950479745865,
+      "learning_rate": 4.9963175402849494e-05,
+      "loss": 0.0369,
+      "step": 18000
+    },
+    {
+      "epoch": 2.8797914153403124,
+      "grad_norm": 0.037826769053936005,
+      "learning_rate": 4.976986707530646e-05,
+      "loss": 0.0367,
+      "step": 18500
+    },
+    {
+      "epoch": 2.9576215122387826,
+      "grad_norm": 0.024277370423078537,
+      "learning_rate": 4.94120952916684e-05,
+      "loss": 0.0361,
+      "step": 19000
+    },
+    {
+      "epoch": 2.9999610849515506,
+      "eval_avg": 0.029430464545251396,
+      "eval_cer": 0.014860838907614712,
+      "eval_der": 0.7972269254290619,
+      "eval_loss": 0.029503749683499336,
+      "eval_runtime": 898.4638,
+      "eval_samples_per_second": 151.061,
+      "eval_steps_per_second": 0.591,
+      "eval_wer": 0.04400009018288808,
+      "step": 19272
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 19272,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1606,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.5041733416362967e+18,
+  "train_batch_size": 256,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-19272/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da7b800b927eb24b13d6f1efc7e805de8dee2e88614da9aa3f5e72a4c20aff1f
+size 5905

trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 17666,
-  "best_metric": 0.706001258948505,
-  "best_model_checkpoint": "checkpoints/checkpoint-17666",
-  "epoch": 2.749970813713663,
+  "best_global_step": 19272,
+  "best_metric": 0.029430464545251396,
+  "best_model_checkpoint": "checkpoints/checkpoint-19272",
+  "epoch": 2.9999610849515506,
   "eval_steps": 1606,
-  "global_step": 17666,
+  "global_step": 19272,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -385,6 +385,39 @@
       "eval_steps_per_second": 1.744,
       "eval_wer": 0.6703593506268415,
       "step": 17666
+    },
+    {
+      "epoch": 2.8019613184418413,
+      "grad_norm": 0.027681950479745865,
+      "learning_rate": 4.9963175402849494e-05,
+      "loss": 0.0369,
+      "step": 18000
+    },
+    {
+      "epoch": 2.8797914153403124,
+      "grad_norm": 0.037826769053936005,
+      "learning_rate": 4.976986707530646e-05,
+      "loss": 0.0367,
+      "step": 18500
+    },
+    {
+      "epoch": 2.9576215122387826,
+      "grad_norm": 0.024277370423078537,
+      "learning_rate": 4.94120952916684e-05,
+      "loss": 0.0361,
+      "step": 19000
+    },
+    {
+      "epoch": 2.9999610849515506,
+      "eval_avg": 0.029430464545251396,
+      "eval_cer": 0.014860838907614712,
+      "eval_der": 0.7972269254290619,
+      "eval_loss": 0.029503749683499336,
+      "eval_runtime": 898.4638,
+      "eval_samples_per_second": 151.061,
+      "eval_steps_per_second": 0.591,
+      "eval_wer": 0.04400009018288808,
+      "step": 19272
     }
   ],
   "logging_steps": 500,
@@ -399,12 +432,12 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }
   },
-  "total_flos": 1.3788271574227354e+18,
+  "total_flos": 1.5041733416362967e+18,
   "train_batch_size": 256,
   "trial_name": null,
   "trial_params": null