initial commit

config.json

@@ -0,0 +1,30 @@
+{
+  "_name_or_path": "kravchenko/uk-mt5-large",
+  "architectures": [
+    "MT5ForConditionalGeneration"
+  ],
+  "d_ff": 2816,
+  "d_kv": 64,
+  "d_model": 1024,
+  "decoder_start_token_id": 0,
+  "dropout_rate": 0.1,
+  "eos_token_id": 1,
+  "feed_forward_proj": "gated-gelu",
+  "initializer_factor": 1.0,
+  "is_encoder_decoder": true,
+  "layer_norm_epsilon": 1e-06,
+  "model_type": "mt5",
+  "num_decoder_layers": 24,
+  "num_heads": 16,
+  "num_layers": 24,
+  "output_past": true,
+  "pad_token_id": 0,
+  "relative_attention_max_distance": 128,
+  "relative_attention_num_buckets": 32,
+  "tie_word_embeddings": false,
+  "tokenizer_class": "T5Tokenizer",
+  "torch_dtype": "float32",
+  "transformers_version": "4.19.4",
+  "use_cache": false,
+  "vocab_size": 30000
+}

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0da3b86c5415fa8e0a9a7a92b244de4b23442a053f2ebba9d3d8eb2d134214bc
+size 5755849

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1f26d899bd7f31507391ff6cbd010794e40bc0ae19b4bcb287dd5d61067cf0d9
+size 3115357157

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5244bb46684483c25851cf9b41df93d81cc47c42701ce95b85d18bd9b16df49e
+size 14503

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b648422d2cb14ab64e8e23dbd45cab432c48a50a2de3db4d20cff885c22026ea
+size 623

special_tokens_map.json

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

spiece.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa8ea53d190098f2632bd1c44b62731a90a9257a96081183c1a361a24f7ea8ff
+size 804017

tokenizer_config.json

@@ -0,0 +1 @@
+{"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>", "extra_ids": 0, "additional_special_tokens": null, "sp_model_kwargs": {}, "special_tokens_map_file": "/home/jupyter/.cache/huggingface/transformers/185abfb2369e8865a1629919f3937e6f094d5c5e4629c003bb05efae92ee060a.294ebaa4cd17bb284635004c92d2c4d522ec488c828dcce0c2471b6f28e3fe82", "name_or_path": "kravchenko/uk-mt5-large", "tokenizer_class": "T5Tokenizer"}

trainer_state.json

@@ -0,0 +1,271 @@
+{
+  "best_metric": 0.27745380997657776,
+  "best_model_checkpoint": "uk-mt5-base-gec/checkpoint-1350",
+  "epoch": 3.0177409816676524,
+  "global_step": 2550,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.000970414201183432,
+      "loss": 2.8559,
+      "step": 150
+    },
+    {
+      "epoch": 0.18,
+      "eval_google_bleu": 0.4161949252310698,
+      "eval_loss": 0.3385647237300873,
+      "eval_runtime": 574.9325,
+      "eval_samples_per_second": 5.227,
+      "eval_steps_per_second": 0.327,
+      "step": 150
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.000940828402366864,
+      "loss": 0.353,
+      "step": 300
+    },
+    {
+      "epoch": 0.35,
+      "eval_google_bleu": 0.4197968462318859,
+      "eval_loss": 0.30946752429008484,
+      "eval_runtime": 574.235,
+      "eval_samples_per_second": 5.233,
+      "eval_steps_per_second": 0.327,
+      "step": 300
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0009112426035502958,
+      "loss": 0.3433,
+      "step": 450
+    },
+    {
+      "epoch": 0.53,
+      "eval_google_bleu": 0.42082026039416087,
+      "eval_loss": 0.30230990052223206,
+      "eval_runtime": 573.7714,
+      "eval_samples_per_second": 5.237,
+      "eval_steps_per_second": 0.328,
+      "step": 450
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0008816568047337278,
+      "loss": 0.3248,
+      "step": 600
+    },
+    {
+      "epoch": 0.71,
+      "eval_google_bleu": 0.4194576852971206,
+      "eval_loss": 0.2984682619571686,
+      "eval_runtime": 574.3927,
+      "eval_samples_per_second": 5.232,
+      "eval_steps_per_second": 0.327,
+      "step": 600
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0008520710059171598,
+      "loss": 0.3046,
+      "step": 750
+    },
+    {
+      "epoch": 0.89,
+      "eval_google_bleu": 0.4217920913982863,
+      "eval_loss": 0.28489378094673157,
+      "eval_runtime": 574.7617,
+      "eval_samples_per_second": 5.228,
+      "eval_steps_per_second": 0.327,
+      "step": 750
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0008224852071005917,
+      "loss": 0.2625,
+      "step": 900
+    },
+    {
+      "epoch": 1.07,
+      "eval_google_bleu": 0.4213197969543147,
+      "eval_loss": 0.29553136229515076,
+      "eval_runtime": 573.965,
+      "eval_samples_per_second": 5.236,
+      "eval_steps_per_second": 0.328,
+      "step": 900
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0007928994082840238,
+      "loss": 0.2127,
+      "step": 1050
+    },
+    {
+      "epoch": 1.24,
+      "eval_google_bleu": 0.4211682670038433,
+      "eval_loss": 0.30292925238609314,
+      "eval_runtime": 574.6793,
+      "eval_samples_per_second": 5.229,
+      "eval_steps_per_second": 0.327,
+      "step": 1050
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0007633136094674556,
+      "loss": 0.224,
+      "step": 1200
+    },
+    {
+      "epoch": 1.42,
+      "eval_google_bleu": 0.4224696723929531,
+      "eval_loss": 0.3068563640117645,
+      "eval_runtime": 574.5278,
+      "eval_samples_per_second": 5.23,
+      "eval_steps_per_second": 0.327,
+      "step": 1200
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0007337278106508876,
+      "loss": 0.2332,
+      "step": 1350
+    },
+    {
+      "epoch": 1.6,
+      "eval_google_bleu": 0.41897146578336036,
+      "eval_loss": 0.27745380997657776,
+      "eval_runtime": 575.5901,
+      "eval_samples_per_second": 5.221,
+      "eval_steps_per_second": 0.327,
+      "step": 1350
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.0007041420118343196,
+      "loss": 0.238,
+      "step": 1500
+    },
+    {
+      "epoch": 1.77,
+      "eval_google_bleu": 0.4164072628882445,
+      "eval_loss": 0.2903579771518707,
+      "eval_runtime": 573.985,
+      "eval_samples_per_second": 5.235,
+      "eval_steps_per_second": 0.328,
+      "step": 1500
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.0006745562130177515,
+      "loss": 0.2297,
+      "step": 1650
+    },
+    {
+      "epoch": 1.95,
+      "eval_google_bleu": 0.41891172732452214,
+      "eval_loss": 0.2825988233089447,
+      "eval_runtime": 574.7646,
+      "eval_samples_per_second": 5.228,
+      "eval_steps_per_second": 0.327,
+      "step": 1650
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.0006449704142011834,
+      "loss": 0.1649,
+      "step": 1800
+    },
+    {
+      "epoch": 2.13,
+      "eval_google_bleu": 0.41988701131139533,
+      "eval_loss": 0.304867148399353,
+      "eval_runtime": 575.4878,
+      "eval_samples_per_second": 5.222,
+      "eval_steps_per_second": 0.327,
+      "step": 1800
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.0006153846153846154,
+      "loss": 0.1458,
+      "step": 1950
+    },
+    {
+      "epoch": 2.31,
+      "eval_google_bleu": 0.4201902185823143,
+      "eval_loss": 0.3138478994369507,
+      "eval_runtime": 575.2695,
+      "eval_samples_per_second": 5.224,
+      "eval_steps_per_second": 0.327,
+      "step": 1950
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.0005857988165680473,
+      "loss": 0.1564,
+      "step": 2100
+    },
+    {
+      "epoch": 2.48,
+      "eval_google_bleu": 0.415792735992587,
+      "eval_loss": 0.3027360141277313,
+      "eval_runtime": 566.4822,
+      "eval_samples_per_second": 5.305,
+      "eval_steps_per_second": 0.332,
+      "step": 2100
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.0005562130177514793,
+      "loss": 0.1572,
+      "step": 2250
+    },
+    {
+      "epoch": 2.66,
+      "eval_google_bleu": 0.4212085345156907,
+      "eval_loss": 0.3020596504211426,
+      "eval_runtime": 574.5482,
+      "eval_samples_per_second": 5.23,
+      "eval_steps_per_second": 0.327,
+      "step": 2250
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 0.0005266272189349113,
+      "loss": 0.159,
+      "step": 2400
+    },
+    {
+      "epoch": 2.84,
+      "eval_google_bleu": 0.41866493031642477,
+      "eval_loss": 0.29025933146476746,
+      "eval_runtime": 574.442,
+      "eval_samples_per_second": 5.231,
+      "eval_steps_per_second": 0.327,
+      "step": 2400
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 0.0004970414201183431,
+      "loss": 0.159,
+      "step": 2550
+    },
+    {
+      "epoch": 3.02,
+      "eval_google_bleu": 0.4191764521684672,
+      "eval_loss": 0.3248673677444458,
+      "eval_runtime": 573.9254,
+      "eval_samples_per_second": 5.236,
+      "eval_steps_per_second": 0.328,
+      "step": 2550
+    }
+  ],
+  "max_steps": 5070,
+  "num_train_epochs": 6,
+  "total_flos": 2.984077609323725e+16,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:391610fb80fe5f0515ce8c6c2d22591ef7350640f14bfa99e6eeb98e9dec8eb8
+size 3247