commit from Ife
Browse files- config.json +59 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- source.spm +0 -0
- special_tokens_map.json +1 -0
- target.spm +0 -0
- tokenizer_config.json +1 -0
- trainer_state.json +480 -0
- training_args.bin +3 -0
- vocab.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "Helsinki-NLP/opus-mt-es-ca",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"activation_function": "swish",
|
| 5 |
+
"add_bias_logits": false,
|
| 6 |
+
"add_final_layer_norm": false,
|
| 7 |
+
"architectures": [
|
| 8 |
+
"MarianMTModel"
|
| 9 |
+
],
|
| 10 |
+
"attention_dropout": 0.0,
|
| 11 |
+
"bad_words_ids": [
|
| 12 |
+
[
|
| 13 |
+
51969
|
| 14 |
+
]
|
| 15 |
+
],
|
| 16 |
+
"bos_token_id": 0,
|
| 17 |
+
"classif_dropout": 0.0,
|
| 18 |
+
"classifier_dropout": 0.0,
|
| 19 |
+
"d_model": 512,
|
| 20 |
+
"decoder_attention_heads": 8,
|
| 21 |
+
"decoder_ffn_dim": 2048,
|
| 22 |
+
"decoder_layerdrop": 0.0,
|
| 23 |
+
"decoder_layers": 6,
|
| 24 |
+
"decoder_start_token_id": 51969,
|
| 25 |
+
"dropout": 0.1,
|
| 26 |
+
"encoder_attention_heads": 8,
|
| 27 |
+
"encoder_ffn_dim": 2048,
|
| 28 |
+
"encoder_layerdrop": 0.0,
|
| 29 |
+
"encoder_layers": 6,
|
| 30 |
+
"eos_token_id": 0,
|
| 31 |
+
"extra_pos_embeddings": 51970,
|
| 32 |
+
"forced_eos_token_id": 0,
|
| 33 |
+
"gradient_checkpointing": false,
|
| 34 |
+
"id2label": {
|
| 35 |
+
"0": "LABEL_0",
|
| 36 |
+
"1": "LABEL_1",
|
| 37 |
+
"2": "LABEL_2"
|
| 38 |
+
},
|
| 39 |
+
"init_std": 0.02,
|
| 40 |
+
"is_encoder_decoder": true,
|
| 41 |
+
"label2id": {
|
| 42 |
+
"LABEL_0": 0,
|
| 43 |
+
"LABEL_1": 1,
|
| 44 |
+
"LABEL_2": 2
|
| 45 |
+
},
|
| 46 |
+
"max_length": 512,
|
| 47 |
+
"max_position_embeddings": 512,
|
| 48 |
+
"model_type": "marian",
|
| 49 |
+
"normalize_before": false,
|
| 50 |
+
"normalize_embedding": false,
|
| 51 |
+
"num_beams": 4,
|
| 52 |
+
"num_hidden_layers": 6,
|
| 53 |
+
"pad_token_id": 51969,
|
| 54 |
+
"scale_embedding": true,
|
| 55 |
+
"static_position_embeddings": true,
|
| 56 |
+
"transformers_version": "4.7.0.dev0",
|
| 57 |
+
"use_cache": true,
|
| 58 |
+
"vocab_size": 51970
|
| 59 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3b6fd128fc7d92d55fa0fff8b84f2c02978b6fc093ab0c767cb93a8b102b9fbc
|
| 3 |
+
size 566136377
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7eb8b2110a6aeca21e5279e9c630e240e6492a0d5b44da87ef335983a24c9772
|
| 3 |
+
size 283287261
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:017fc28e3b75268ffab5f4c206d6d487e6bffa992beb7cb563eefd1f58cc0993
|
| 3 |
+
size 17756
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:02813dcc2085851c20da10a78140421ebd1a171ca24a5b6d5d6b8fe6d0f2dde9
|
| 3 |
+
size 623
|
source.spm
ADDED
|
Binary file (828 kB). View file
|
|
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>"}
|
target.spm
ADDED
|
Binary file (823 kB). View file
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"source_lang": "spa", "target_lang": "cat", "unk_token": "<unk>", "eos_token": "</s>", "pad_token": "<pad>", "model_max_length": 512, "sp_model_kwargs": {}, "name_or_path": "Helsinki-NLP/opus-mt-es-ca"}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,480 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 47.3367,
|
| 3 |
+
"best_model_checkpoint": "/scratch/st-amuham01-1/fenimi/transformers/examples/pytorch/translation/similar-languages-task/model_out_es-pt-tok/checkpoint-1704014",
|
| 4 |
+
"epoch": 35.0,
|
| 5 |
+
"global_step": 3371314,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 1.0,
|
| 12 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 13 |
+
"loss": 0.8042,
|
| 14 |
+
"step": 131078
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.0,
|
| 18 |
+
"eval_bleu": 46.137,
|
| 19 |
+
"eval_gen_len": 73.7186,
|
| 20 |
+
"eval_loss": 0.7053773999214172,
|
| 21 |
+
"eval_runtime": 93.6018,
|
| 22 |
+
"eval_samples_per_second": 10.363,
|
| 23 |
+
"eval_steps_per_second": 0.331,
|
| 24 |
+
"step": 131078
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 2.0,
|
| 28 |
+
"learning_rate": 4.9e-05,
|
| 29 |
+
"loss": 0.6488,
|
| 30 |
+
"step": 262156
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 2.0,
|
| 34 |
+
"eval_bleu": 46.1742,
|
| 35 |
+
"eval_gen_len": 73.8876,
|
| 36 |
+
"eval_loss": 0.6584964990615845,
|
| 37 |
+
"eval_runtime": 91.805,
|
| 38 |
+
"eval_samples_per_second": 10.566,
|
| 39 |
+
"eval_steps_per_second": 0.338,
|
| 40 |
+
"step": 262156
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 3.0,
|
| 44 |
+
"learning_rate": 4.85e-05,
|
| 45 |
+
"loss": 0.6069,
|
| 46 |
+
"step": 393234
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 3.0,
|
| 50 |
+
"eval_bleu": 46.0956,
|
| 51 |
+
"eval_gen_len": 73.666,
|
| 52 |
+
"eval_loss": 0.6464588046073914,
|
| 53 |
+
"eval_runtime": 87.4692,
|
| 54 |
+
"eval_samples_per_second": 11.09,
|
| 55 |
+
"eval_steps_per_second": 0.354,
|
| 56 |
+
"step": 393234
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 4.0,
|
| 60 |
+
"learning_rate": 4.8e-05,
|
| 61 |
+
"loss": 0.5821,
|
| 62 |
+
"step": 524312
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 4.0,
|
| 66 |
+
"eval_bleu": 46.326,
|
| 67 |
+
"eval_gen_len": 73.932,
|
| 68 |
+
"eval_loss": 0.6305918097496033,
|
| 69 |
+
"eval_runtime": 90.9772,
|
| 70 |
+
"eval_samples_per_second": 10.662,
|
| 71 |
+
"eval_steps_per_second": 0.341,
|
| 72 |
+
"step": 524312
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 5.0,
|
| 76 |
+
"learning_rate": 4.75e-05,
|
| 77 |
+
"loss": 0.5648,
|
| 78 |
+
"step": 655390
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 5.0,
|
| 82 |
+
"eval_bleu": 46.3267,
|
| 83 |
+
"eval_gen_len": 73.6505,
|
| 84 |
+
"eval_loss": 0.617319643497467,
|
| 85 |
+
"eval_runtime": 86.5507,
|
| 86 |
+
"eval_samples_per_second": 11.207,
|
| 87 |
+
"eval_steps_per_second": 0.358,
|
| 88 |
+
"step": 655390
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 6.0,
|
| 92 |
+
"learning_rate": 4.7e-05,
|
| 93 |
+
"loss": 0.5515,
|
| 94 |
+
"step": 786468
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 6.0,
|
| 98 |
+
"eval_bleu": 46.6537,
|
| 99 |
+
"eval_gen_len": 73.8722,
|
| 100 |
+
"eval_loss": 0.6139377355575562,
|
| 101 |
+
"eval_runtime": 88.9058,
|
| 102 |
+
"eval_samples_per_second": 10.91,
|
| 103 |
+
"eval_steps_per_second": 0.349,
|
| 104 |
+
"step": 786468
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 7.0,
|
| 108 |
+
"learning_rate": 4.6500000000000005e-05,
|
| 109 |
+
"loss": 0.541,
|
| 110 |
+
"step": 917546
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 7.0,
|
| 114 |
+
"eval_bleu": 46.5628,
|
| 115 |
+
"eval_gen_len": 73.8412,
|
| 116 |
+
"eval_loss": 0.6046749353408813,
|
| 117 |
+
"eval_runtime": 87.2322,
|
| 118 |
+
"eval_samples_per_second": 11.12,
|
| 119 |
+
"eval_steps_per_second": 0.355,
|
| 120 |
+
"step": 917546
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 8.0,
|
| 124 |
+
"learning_rate": 4.600000000000001e-05,
|
| 125 |
+
"loss": 0.5323,
|
| 126 |
+
"step": 1048624
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 8.0,
|
| 130 |
+
"eval_bleu": 46.5538,
|
| 131 |
+
"eval_gen_len": 73.8216,
|
| 132 |
+
"eval_loss": 0.5961582064628601,
|
| 133 |
+
"eval_runtime": 88.6262,
|
| 134 |
+
"eval_samples_per_second": 10.945,
|
| 135 |
+
"eval_steps_per_second": 0.35,
|
| 136 |
+
"step": 1048624
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 9.0,
|
| 140 |
+
"learning_rate": 4.55e-05,
|
| 141 |
+
"loss": 0.525,
|
| 142 |
+
"step": 1179702
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 9.0,
|
| 146 |
+
"eval_bleu": 46.543,
|
| 147 |
+
"eval_gen_len": 74.0711,
|
| 148 |
+
"eval_loss": 0.5961570739746094,
|
| 149 |
+
"eval_runtime": 90.5754,
|
| 150 |
+
"eval_samples_per_second": 10.709,
|
| 151 |
+
"eval_steps_per_second": 0.342,
|
| 152 |
+
"step": 1179702
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 10.0,
|
| 156 |
+
"learning_rate": 4.5e-05,
|
| 157 |
+
"loss": 0.5184,
|
| 158 |
+
"step": 1310780
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 10.0,
|
| 162 |
+
"eval_bleu": 46.772,
|
| 163 |
+
"eval_gen_len": 73.8258,
|
| 164 |
+
"eval_loss": 0.5992215275764465,
|
| 165 |
+
"eval_runtime": 86.6687,
|
| 166 |
+
"eval_samples_per_second": 11.192,
|
| 167 |
+
"eval_steps_per_second": 0.358,
|
| 168 |
+
"step": 1310780
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 11.0,
|
| 172 |
+
"learning_rate": 4.4500000000000004e-05,
|
| 173 |
+
"loss": 0.5126,
|
| 174 |
+
"step": 1441858
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 11.0,
|
| 178 |
+
"eval_bleu": 46.875,
|
| 179 |
+
"eval_gen_len": 74.0124,
|
| 180 |
+
"eval_loss": 0.5917361378669739,
|
| 181 |
+
"eval_runtime": 91.2818,
|
| 182 |
+
"eval_samples_per_second": 10.626,
|
| 183 |
+
"eval_steps_per_second": 0.34,
|
| 184 |
+
"step": 1441858
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 12.0,
|
| 188 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 189 |
+
"loss": 0.5075,
|
| 190 |
+
"step": 1572936
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 12.0,
|
| 194 |
+
"eval_bleu": 46.5982,
|
| 195 |
+
"eval_gen_len": 74.0639,
|
| 196 |
+
"eval_loss": 0.59002286195755,
|
| 197 |
+
"eval_runtime": 95.1271,
|
| 198 |
+
"eval_samples_per_second": 10.197,
|
| 199 |
+
"eval_steps_per_second": 0.326,
|
| 200 |
+
"step": 1572936
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 13.0,
|
| 204 |
+
"learning_rate": 4.35e-05,
|
| 205 |
+
"loss": 0.503,
|
| 206 |
+
"step": 1704014
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 13.0,
|
| 210 |
+
"eval_bleu": 47.3367,
|
| 211 |
+
"eval_gen_len": 73.9134,
|
| 212 |
+
"eval_loss": 0.591410756111145,
|
| 213 |
+
"eval_runtime": 87.7758,
|
| 214 |
+
"eval_samples_per_second": 11.051,
|
| 215 |
+
"eval_steps_per_second": 0.353,
|
| 216 |
+
"step": 1704014
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 14.0,
|
| 220 |
+
"learning_rate": 4.3e-05,
|
| 221 |
+
"loss": 0.4987,
|
| 222 |
+
"step": 1835092
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 14.0,
|
| 226 |
+
"eval_bleu": 46.5688,
|
| 227 |
+
"eval_gen_len": 73.9072,
|
| 228 |
+
"eval_loss": 0.5959733128547668,
|
| 229 |
+
"eval_runtime": 95.2654,
|
| 230 |
+
"eval_samples_per_second": 10.182,
|
| 231 |
+
"eval_steps_per_second": 0.325,
|
| 232 |
+
"step": 1835092
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 15.0,
|
| 236 |
+
"learning_rate": 4.25e-05,
|
| 237 |
+
"loss": 0.4947,
|
| 238 |
+
"step": 1966170
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 15.0,
|
| 242 |
+
"eval_bleu": 46.8246,
|
| 243 |
+
"eval_gen_len": 74.066,
|
| 244 |
+
"eval_loss": 0.5951239466667175,
|
| 245 |
+
"eval_runtime": 89.8542,
|
| 246 |
+
"eval_samples_per_second": 10.795,
|
| 247 |
+
"eval_steps_per_second": 0.345,
|
| 248 |
+
"step": 1966170
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 16.0,
|
| 252 |
+
"learning_rate": 4.2e-05,
|
| 253 |
+
"loss": 0.4912,
|
| 254 |
+
"step": 2097248
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 16.0,
|
| 258 |
+
"eval_bleu": 46.6078,
|
| 259 |
+
"eval_gen_len": 74.1052,
|
| 260 |
+
"eval_loss": 0.5862768888473511,
|
| 261 |
+
"eval_runtime": 90.8564,
|
| 262 |
+
"eval_samples_per_second": 10.676,
|
| 263 |
+
"eval_steps_per_second": 0.341,
|
| 264 |
+
"step": 2097248
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 17.0,
|
| 268 |
+
"learning_rate": 4.15e-05,
|
| 269 |
+
"loss": 0.4877,
|
| 270 |
+
"step": 2228326
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 17.0,
|
| 274 |
+
"eval_bleu": 46.4987,
|
| 275 |
+
"eval_gen_len": 73.8402,
|
| 276 |
+
"eval_loss": 0.5898135304450989,
|
| 277 |
+
"eval_runtime": 93.1614,
|
| 278 |
+
"eval_samples_per_second": 10.412,
|
| 279 |
+
"eval_steps_per_second": 0.333,
|
| 280 |
+
"step": 2228326
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 24.0,
|
| 284 |
+
"learning_rate": 3.780262784911128e-05,
|
| 285 |
+
"loss": 0.4915,
|
| 286 |
+
"step": 2323575
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 24.0,
|
| 290 |
+
"eval_bleu": 46.9695,
|
| 291 |
+
"eval_gen_len": 73.9588,
|
| 292 |
+
"eval_loss": 0.5856001973152161,
|
| 293 |
+
"eval_runtime": 77.7022,
|
| 294 |
+
"eval_samples_per_second": 12.484,
|
| 295 |
+
"eval_steps_per_second": 0.399,
|
| 296 |
+
"step": 2323575
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 25.0,
|
| 300 |
+
"learning_rate": 3.730262784911128e-05,
|
| 301 |
+
"loss": 0.4858,
|
| 302 |
+
"step": 2418824
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 25.0,
|
| 306 |
+
"eval_bleu": 46.9834,
|
| 307 |
+
"eval_gen_len": 73.7845,
|
| 308 |
+
"eval_loss": 0.5880559682846069,
|
| 309 |
+
"eval_runtime": 76.2261,
|
| 310 |
+
"eval_samples_per_second": 12.725,
|
| 311 |
+
"eval_steps_per_second": 0.407,
|
| 312 |
+
"step": 2418824
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 26.0,
|
| 316 |
+
"learning_rate": 3.6802627849111274e-05,
|
| 317 |
+
"loss": 0.4819,
|
| 318 |
+
"step": 2514073
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 26.0,
|
| 322 |
+
"eval_bleu": 47.2378,
|
| 323 |
+
"eval_gen_len": 73.7866,
|
| 324 |
+
"eval_loss": 0.5938757061958313,
|
| 325 |
+
"eval_runtime": 77.0503,
|
| 326 |
+
"eval_samples_per_second": 12.589,
|
| 327 |
+
"eval_steps_per_second": 0.402,
|
| 328 |
+
"step": 2514073
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 27.0,
|
| 332 |
+
"learning_rate": 3.6302627849111276e-05,
|
| 333 |
+
"loss": 0.4785,
|
| 334 |
+
"step": 2609322
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 27.0,
|
| 338 |
+
"eval_bleu": 47.1579,
|
| 339 |
+
"eval_gen_len": 74.1351,
|
| 340 |
+
"eval_loss": 0.593238115310669,
|
| 341 |
+
"eval_runtime": 79.7831,
|
| 342 |
+
"eval_samples_per_second": 12.158,
|
| 343 |
+
"eval_steps_per_second": 0.389,
|
| 344 |
+
"step": 2609322
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 28.0,
|
| 348 |
+
"learning_rate": 3.5802627849111284e-05,
|
| 349 |
+
"loss": 0.4757,
|
| 350 |
+
"step": 2704571
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 28.0,
|
| 354 |
+
"eval_bleu": 47.0974,
|
| 355 |
+
"eval_gen_len": 73.8289,
|
| 356 |
+
"eval_loss": 0.5946824550628662,
|
| 357 |
+
"eval_runtime": 77.1019,
|
| 358 |
+
"eval_samples_per_second": 12.581,
|
| 359 |
+
"eval_steps_per_second": 0.402,
|
| 360 |
+
"step": 2704571
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 29.0,
|
| 364 |
+
"learning_rate": 3.530262784911128e-05,
|
| 365 |
+
"loss": 0.4729,
|
| 366 |
+
"step": 2799820
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 29.0,
|
| 370 |
+
"eval_bleu": 47.2814,
|
| 371 |
+
"eval_gen_len": 74.1485,
|
| 372 |
+
"eval_loss": 0.5900229811668396,
|
| 373 |
+
"eval_runtime": 76.5406,
|
| 374 |
+
"eval_samples_per_second": 12.673,
|
| 375 |
+
"eval_steps_per_second": 0.405,
|
| 376 |
+
"step": 2799820
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 30.0,
|
| 380 |
+
"learning_rate": 3.480262784911128e-05,
|
| 381 |
+
"loss": 0.4702,
|
| 382 |
+
"step": 2895069
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 30.0,
|
| 386 |
+
"eval_bleu": 47.1382,
|
| 387 |
+
"eval_gen_len": 73.8495,
|
| 388 |
+
"eval_loss": 0.5961056351661682,
|
| 389 |
+
"eval_runtime": 77.883,
|
| 390 |
+
"eval_samples_per_second": 12.455,
|
| 391 |
+
"eval_steps_per_second": 0.398,
|
| 392 |
+
"step": 2895069
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 31.0,
|
| 396 |
+
"learning_rate": 3.430262784911128e-05,
|
| 397 |
+
"loss": 0.4678,
|
| 398 |
+
"step": 2990318
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 31.0,
|
| 402 |
+
"eval_bleu": 46.9917,
|
| 403 |
+
"eval_gen_len": 74.0526,
|
| 404 |
+
"eval_loss": 0.5983180999755859,
|
| 405 |
+
"eval_runtime": 77.7736,
|
| 406 |
+
"eval_samples_per_second": 12.472,
|
| 407 |
+
"eval_steps_per_second": 0.399,
|
| 408 |
+
"step": 2990318
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 32.0,
|
| 412 |
+
"learning_rate": 3.3802627849111276e-05,
|
| 413 |
+
"loss": 0.4654,
|
| 414 |
+
"step": 3085567
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 32.0,
|
| 418 |
+
"eval_bleu": 47.1391,
|
| 419 |
+
"eval_gen_len": 73.8381,
|
| 420 |
+
"eval_loss": 0.5979074835777283,
|
| 421 |
+
"eval_runtime": 79.7831,
|
| 422 |
+
"eval_samples_per_second": 12.158,
|
| 423 |
+
"eval_steps_per_second": 0.389,
|
| 424 |
+
"step": 3085567
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 33.0,
|
| 428 |
+
"learning_rate": 3.330262784911128e-05,
|
| 429 |
+
"loss": 0.4632,
|
| 430 |
+
"step": 3180816
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 33.0,
|
| 434 |
+
"eval_bleu": 46.5598,
|
| 435 |
+
"eval_gen_len": 73.8753,
|
| 436 |
+
"eval_loss": 0.5982452630996704,
|
| 437 |
+
"eval_runtime": 78.0544,
|
| 438 |
+
"eval_samples_per_second": 12.427,
|
| 439 |
+
"eval_steps_per_second": 0.397,
|
| 440 |
+
"step": 3180816
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 34.0,
|
| 444 |
+
"learning_rate": 3.280262784911128e-05,
|
| 445 |
+
"loss": 0.461,
|
| 446 |
+
"step": 3276065
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 34.0,
|
| 450 |
+
"eval_bleu": 46.503,
|
| 451 |
+
"eval_gen_len": 73.9577,
|
| 452 |
+
"eval_loss": 0.5941172242164612,
|
| 453 |
+
"eval_runtime": 77.6074,
|
| 454 |
+
"eval_samples_per_second": 12.499,
|
| 455 |
+
"eval_steps_per_second": 0.399,
|
| 456 |
+
"step": 3276065
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 35.0,
|
| 460 |
+
"learning_rate": 3.230262784911128e-05,
|
| 461 |
+
"loss": 0.459,
|
| 462 |
+
"step": 3371314
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 35.0,
|
| 466 |
+
"eval_bleu": 47.3154,
|
| 467 |
+
"eval_gen_len": 74.0649,
|
| 468 |
+
"eval_loss": 0.598816990852356,
|
| 469 |
+
"eval_runtime": 79.9962,
|
| 470 |
+
"eval_samples_per_second": 12.126,
|
| 471 |
+
"eval_steps_per_second": 0.388,
|
| 472 |
+
"step": 3371314
|
| 473 |
+
}
|
| 474 |
+
],
|
| 475 |
+
"max_steps": 9524900,
|
| 476 |
+
"num_train_epochs": 100,
|
| 477 |
+
"total_flos": 1.0411885225994306e+19,
|
| 478 |
+
"trial_name": null,
|
| 479 |
+
"trial_params": null
|
| 480 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8767a06a5b9079e8e6a5e89983f6a6d694bbcdf76e0cb8f35b40cd5f830bb448
|
| 3 |
+
size 2607
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|