stage big random ailem model

checkpoint-80000/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
+}

checkpoint-80000/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"
+}

checkpoint-80000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c4986300d92aa1f72e9f389662698e199a8933b692b08d244a7951f390a4b05c
+size 1845323269

checkpoint-80000/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8db6d6ed939b84dbdd0f6ef55dfee23517e455542417c4601c45786dfe156d80
+size 922885701

checkpoint-80000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf2d845f2c98845597f8fb73660d758bd6a2621b7f3fca2610d4eee48b4d9061
+size 14575

checkpoint-80000/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:142b7341abd2b4a012a47c78b94a847182093480e47e3e507e963cd68817944a
+size 557

checkpoint-80000/scheduler.pt

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

checkpoint-80000/special_tokens_map.json

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

checkpoint-80000/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>"
+}

checkpoint-80000/trainer_state.json

@@ -0,0 +1,176 @@
+{
+  "best_metric": 43.3155,
+  "best_model_checkpoint": "opus_big_ailem_random/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": 0.1794,
+      "step": 8000
+    },
+    {
+      "epoch": 0.2,
+      "eval_bleu": 42.1115,
+      "eval_gen_len": 33.8782,
+      "eval_loss": 0.09834092110395432,
+      "eval_runtime": 133.0059,
+      "eval_samples_per_second": 7.842,
+      "eval_steps_per_second": 0.496,
+      "step": 8000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.950723967960567e-05,
+      "loss": 0.1487,
+      "step": 16000
+    },
+    {
+      "epoch": 0.39,
+      "eval_bleu": 42.5044,
+      "eval_gen_len": 34.0422,
+      "eval_loss": 0.09755747765302658,
+      "eval_runtime": 123.2199,
+      "eval_samples_per_second": 8.465,
+      "eval_steps_per_second": 0.536,
+      "step": 16000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.926090573012939e-05,
+      "loss": 0.1445,
+      "step": 24000
+    },
+    {
+      "epoch": 0.59,
+      "eval_bleu": 42.1738,
+      "eval_gen_len": 33.3921,
+      "eval_loss": 0.09764789789915085,
+      "eval_runtime": 135.8363,
+      "eval_samples_per_second": 7.678,
+      "eval_steps_per_second": 0.486,
+      "step": 24000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.90145101663586e-05,
+      "loss": 0.1407,
+      "step": 32000
+    },
+    {
+      "epoch": 0.79,
+      "eval_bleu": 42.6761,
+      "eval_gen_len": 33.8092,
+      "eval_loss": 0.09669991582632065,
+      "eval_runtime": 112.788,
+      "eval_samples_per_second": 9.247,
+      "eval_steps_per_second": 0.585,
+      "step": 32000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.876814540973506e-05,
+      "loss": 0.1392,
+      "step": 40000
+    },
+    {
+      "epoch": 0.99,
+      "eval_bleu": 42.5089,
+      "eval_gen_len": 33.814,
+      "eval_loss": 0.09582150727510452,
+      "eval_runtime": 113.7375,
+      "eval_samples_per_second": 9.17,
+      "eval_steps_per_second": 0.58,
+      "step": 40000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.852178065311152e-05,
+      "loss": 0.1241,
+      "step": 48000
+    },
+    {
+      "epoch": 1.18,
+      "eval_bleu": 42.6428,
+      "eval_gen_len": 33.954,
+      "eval_loss": 0.09729909896850586,
+      "eval_runtime": 111.2244,
+      "eval_samples_per_second": 9.377,
+      "eval_steps_per_second": 0.593,
+      "step": 48000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.8275446703635243e-05,
+      "loss": 0.1244,
+      "step": 56000
+    },
+    {
+      "epoch": 1.38,
+      "eval_bleu": 42.5021,
+      "eval_gen_len": 33.7661,
+      "eval_loss": 0.09766314178705215,
+      "eval_runtime": 149.7488,
+      "eval_samples_per_second": 6.965,
+      "eval_steps_per_second": 0.441,
+      "step": 56000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.802908194701171e-05,
+      "loss": 0.125,
+      "step": 64000
+    },
+    {
+      "epoch": 1.58,
+      "eval_bleu": 42.8416,
+      "eval_gen_len": 33.999,
+      "eval_loss": 0.09677453339099884,
+      "eval_runtime": 156.1922,
+      "eval_samples_per_second": 6.678,
+      "eval_steps_per_second": 0.423,
+      "step": 64000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.7782717190388173e-05,
+      "loss": 0.1245,
+      "step": 72000
+    },
+    {
+      "epoch": 1.77,
+      "eval_bleu": 42.797,
+      "eval_gen_len": 34.5053,
+      "eval_loss": 0.09712178260087967,
+      "eval_runtime": 127.7296,
+      "eval_samples_per_second": 8.166,
+      "eval_steps_per_second": 0.517,
+      "step": 72000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7536321626617377e-05,
+      "loss": 0.1242,
+      "step": 80000
+    },
+    {
+      "epoch": 1.97,
+      "eval_bleu": 43.3155,
+      "eval_gen_len": 34.0355,
+      "eval_loss": 0.09631923586130142,
+      "eval_runtime": 125.7048,
+      "eval_samples_per_second": 8.297,
+      "eval_steps_per_second": 0.525,
+      "step": 80000
+    }
+  ],
+  "max_steps": 649200,
+  "num_train_epochs": 16,
+  "total_flos": 2.5844916958632346e+17,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-80000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7754c0c1d221803f1ec03c42c8ec824faf828e5ad475618050f1176fed7e195
+size 3771

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"
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8db6d6ed939b84dbdd0f6ef55dfee23517e455542417c4601c45786dfe156d80
+size 922885701

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>"
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7754c0c1d221803f1ec03c42c8ec824faf828e5ad475618050f1176fed7e195
+size 3771