Upload 12 files
Browse files- all_results.json +22 -0
- config.json +29 -0
- eval_results.json +10 -0
- predict_results.json +9 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +1 -0
- spiece.model +3 -0
- tokenizer.json +0 -0
- tokenizer_config.json +1 -0
- train_results.json +8 -0
- trainer_state.json +1303 -0
- training_args.bin +3 -0
all_results.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 50.0,
|
3 |
+
"eval_bleu": 10.8,
|
4 |
+
"eval_gen_len": 40.1719,
|
5 |
+
"eval_loss": 0.15596765279769897,
|
6 |
+
"eval_runtime": 2043.3124,
|
7 |
+
"eval_samples": 5357,
|
8 |
+
"eval_samples_per_second": 2.622,
|
9 |
+
"eval_steps_per_second": 0.656,
|
10 |
+
"predict_bleu": 10.7907,
|
11 |
+
"predict_gen_len": 44.1413,
|
12 |
+
"predict_loss": 0.16183890402317047,
|
13 |
+
"predict_runtime": 584.637,
|
14 |
+
"predict_samples": 1288,
|
15 |
+
"predict_samples_per_second": 2.203,
|
16 |
+
"predict_steps_per_second": 0.551,
|
17 |
+
"train_loss": 0.5778315416091034,
|
18 |
+
"train_runtime": 26224.0587,
|
19 |
+
"train_samples": 12704,
|
20 |
+
"train_samples_per_second": 24.222,
|
21 |
+
"train_steps_per_second": 0.757
|
22 |
+
}
|
config.json
ADDED
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "google/mt5-base",
|
3 |
+
"architectures": [
|
4 |
+
"MT5ForConditionalGeneration"
|
5 |
+
],
|
6 |
+
"d_ff": 2048,
|
7 |
+
"d_kv": 64,
|
8 |
+
"d_model": 768,
|
9 |
+
"decoder_start_token_id": 0,
|
10 |
+
"dropout_rate": 0.1,
|
11 |
+
"eos_token_id": 1,
|
12 |
+
"feed_forward_proj": "gated-gelu",
|
13 |
+
"initializer_factor": 1.0,
|
14 |
+
"is_encoder_decoder": true,
|
15 |
+
"layer_norm_epsilon": 1e-06,
|
16 |
+
"model_type": "mt5",
|
17 |
+
"num_decoder_layers": 12,
|
18 |
+
"num_heads": 12,
|
19 |
+
"num_layers": 12,
|
20 |
+
"output_past": true,
|
21 |
+
"pad_token_id": 0,
|
22 |
+
"relative_attention_num_buckets": 32,
|
23 |
+
"tie_word_embeddings": false,
|
24 |
+
"tokenizer_class": "T5Tokenizer",
|
25 |
+
"torch_dtype": "float32",
|
26 |
+
"transformers_version": "4.12.3",
|
27 |
+
"use_cache": true,
|
28 |
+
"vocab_size": 250100
|
29 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 50.0,
|
3 |
+
"eval_bleu": 10.8,
|
4 |
+
"eval_gen_len": 40.1719,
|
5 |
+
"eval_loss": 0.15596765279769897,
|
6 |
+
"eval_runtime": 2043.3124,
|
7 |
+
"eval_samples": 5357,
|
8 |
+
"eval_samples_per_second": 2.622,
|
9 |
+
"eval_steps_per_second": 0.656
|
10 |
+
}
|
predict_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"predict_bleu": 10.7907,
|
3 |
+
"predict_gen_len": 44.1413,
|
4 |
+
"predict_loss": 0.16183890402317047,
|
5 |
+
"predict_runtime": 584.637,
|
6 |
+
"predict_samples": 1288,
|
7 |
+
"predict_samples_per_second": 2.203,
|
8 |
+
"predict_steps_per_second": 0.551
|
9 |
+
}
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:07ba540d78bd078e05607bb7bba736eb1b30985c59083d80518c2da27572b3bb
|
3 |
+
size 2329626573
|
special_tokens_map.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>"}
|
spiece.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ef78f86560d809067d12bac6c09f19a462cb3af3f54d2b8acbba26e1433125d6
|
3 |
+
size 4309802
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>", "extra_ids": 0, "additional_special_tokens": null, "special_tokens_map_file": "/home/patrick/.cache/torch/transformers/685ac0ca8568ec593a48b61b0a3c272beee9bc194a3c7241d15dcadb5f875e53.f76030f3ec1b96a8199b2593390c610e76ca8028ef3d24680000619ffb646276", "name_or_path": "google/mt5-base", "sp_model_kwargs": {}, "tokenizer_class": "T5Tokenizer"}
|
train_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 50.0,
|
3 |
+
"train_loss": 0.5778315416091034,
|
4 |
+
"train_runtime": 26224.0587,
|
5 |
+
"train_samples": 12704,
|
6 |
+
"train_samples_per_second": 24.222,
|
7 |
+
"train_steps_per_second": 0.757
|
8 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,1303 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 50.0,
|
5 |
+
"global_step": 19850,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.25,
|
12 |
+
"learning_rate": 0.000994962216624685,
|
13 |
+
"loss": 4.6421,
|
14 |
+
"step": 100
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 0.5,
|
18 |
+
"learning_rate": 0.0009899244332493703,
|
19 |
+
"loss": 3.5433,
|
20 |
+
"step": 200
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"epoch": 0.76,
|
24 |
+
"learning_rate": 0.0009848866498740553,
|
25 |
+
"loss": 3.3518,
|
26 |
+
"step": 300
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"epoch": 1.01,
|
30 |
+
"learning_rate": 0.0009798488664987406,
|
31 |
+
"loss": 3.167,
|
32 |
+
"step": 400
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"epoch": 1.26,
|
36 |
+
"learning_rate": 0.0009748110831234257,
|
37 |
+
"loss": 2.8377,
|
38 |
+
"step": 500
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 1.51,
|
42 |
+
"learning_rate": 0.0009697732997481109,
|
43 |
+
"loss": 2.8006,
|
44 |
+
"step": 600
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 1.76,
|
48 |
+
"learning_rate": 0.000964735516372796,
|
49 |
+
"loss": 2.7447,
|
50 |
+
"step": 700
|
51 |
+
},
|
52 |
+
{
|
53 |
+
"epoch": 2.02,
|
54 |
+
"learning_rate": 0.0009596977329974811,
|
55 |
+
"loss": 2.6706,
|
56 |
+
"step": 800
|
57 |
+
},
|
58 |
+
{
|
59 |
+
"epoch": 2.27,
|
60 |
+
"learning_rate": 0.0009546599496221663,
|
61 |
+
"loss": 2.3749,
|
62 |
+
"step": 900
|
63 |
+
},
|
64 |
+
{
|
65 |
+
"epoch": 2.52,
|
66 |
+
"learning_rate": 0.0009496221662468514,
|
67 |
+
"loss": 2.3879,
|
68 |
+
"step": 1000
|
69 |
+
},
|
70 |
+
{
|
71 |
+
"epoch": 2.77,
|
72 |
+
"learning_rate": 0.0009445843828715366,
|
73 |
+
"loss": 2.3724,
|
74 |
+
"step": 1100
|
75 |
+
},
|
76 |
+
{
|
77 |
+
"epoch": 3.02,
|
78 |
+
"learning_rate": 0.0009395465994962216,
|
79 |
+
"loss": 2.3291,
|
80 |
+
"step": 1200
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 3.27,
|
84 |
+
"learning_rate": 0.0009345088161209068,
|
85 |
+
"loss": 2.0652,
|
86 |
+
"step": 1300
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 3.53,
|
90 |
+
"learning_rate": 0.000929471032745592,
|
91 |
+
"loss": 2.053,
|
92 |
+
"step": 1400
|
93 |
+
},
|
94 |
+
{
|
95 |
+
"epoch": 3.78,
|
96 |
+
"learning_rate": 0.000924433249370277,
|
97 |
+
"loss": 2.0745,
|
98 |
+
"step": 1500
|
99 |
+
},
|
100 |
+
{
|
101 |
+
"epoch": 4.03,
|
102 |
+
"learning_rate": 0.0009193954659949623,
|
103 |
+
"loss": 2.0183,
|
104 |
+
"step": 1600
|
105 |
+
},
|
106 |
+
{
|
107 |
+
"epoch": 4.28,
|
108 |
+
"learning_rate": 0.0009143576826196473,
|
109 |
+
"loss": 1.7442,
|
110 |
+
"step": 1700
|
111 |
+
},
|
112 |
+
{
|
113 |
+
"epoch": 4.53,
|
114 |
+
"learning_rate": 0.0009093198992443325,
|
115 |
+
"loss": 1.8065,
|
116 |
+
"step": 1800
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"epoch": 4.79,
|
120 |
+
"learning_rate": 0.0009042821158690177,
|
121 |
+
"loss": 1.7861,
|
122 |
+
"step": 1900
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 5.04,
|
126 |
+
"learning_rate": 0.0008992443324937027,
|
127 |
+
"loss": 1.7369,
|
128 |
+
"step": 2000
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 5.04,
|
132 |
+
"eval_bleu": 2.1435,
|
133 |
+
"eval_gen_len": 15.286,
|
134 |
+
"eval_loss": 1.1994951963424683,
|
135 |
+
"eval_runtime": 384.6632,
|
136 |
+
"eval_samples_per_second": 13.926,
|
137 |
+
"eval_steps_per_second": 3.484,
|
138 |
+
"step": 2000
|
139 |
+
},
|
140 |
+
{
|
141 |
+
"epoch": 5.29,
|
142 |
+
"learning_rate": 0.000894206549118388,
|
143 |
+
"loss": 1.4835,
|
144 |
+
"step": 2100
|
145 |
+
},
|
146 |
+
{
|
147 |
+
"epoch": 5.54,
|
148 |
+
"learning_rate": 0.0008891687657430731,
|
149 |
+
"loss": 1.5012,
|
150 |
+
"step": 2200
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 5.79,
|
154 |
+
"learning_rate": 0.0008841309823677582,
|
155 |
+
"loss": 1.5058,
|
156 |
+
"step": 2300
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 6.05,
|
160 |
+
"learning_rate": 0.0008790931989924434,
|
161 |
+
"loss": 1.4738,
|
162 |
+
"step": 2400
|
163 |
+
},
|
164 |
+
{
|
165 |
+
"epoch": 6.3,
|
166 |
+
"learning_rate": 0.0008740554156171285,
|
167 |
+
"loss": 1.2644,
|
168 |
+
"step": 2500
|
169 |
+
},
|
170 |
+
{
|
171 |
+
"epoch": 6.55,
|
172 |
+
"learning_rate": 0.0008690176322418136,
|
173 |
+
"loss": 1.3161,
|
174 |
+
"step": 2600
|
175 |
+
},
|
176 |
+
{
|
177 |
+
"epoch": 6.8,
|
178 |
+
"learning_rate": 0.0008639798488664988,
|
179 |
+
"loss": 1.299,
|
180 |
+
"step": 2700
|
181 |
+
},
|
182 |
+
{
|
183 |
+
"epoch": 7.05,
|
184 |
+
"learning_rate": 0.0008589420654911839,
|
185 |
+
"loss": 1.2264,
|
186 |
+
"step": 2800
|
187 |
+
},
|
188 |
+
{
|
189 |
+
"epoch": 7.3,
|
190 |
+
"learning_rate": 0.000853904282115869,
|
191 |
+
"loss": 1.08,
|
192 |
+
"step": 2900
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 7.56,
|
196 |
+
"learning_rate": 0.0008488664987405543,
|
197 |
+
"loss": 1.0927,
|
198 |
+
"step": 3000
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 7.81,
|
202 |
+
"learning_rate": 0.0008438287153652393,
|
203 |
+
"loss": 1.1136,
|
204 |
+
"step": 3100
|
205 |
+
},
|
206 |
+
{
|
207 |
+
"epoch": 8.06,
|
208 |
+
"learning_rate": 0.0008387909319899244,
|
209 |
+
"loss": 1.0504,
|
210 |
+
"step": 3200
|
211 |
+
},
|
212 |
+
{
|
213 |
+
"epoch": 8.31,
|
214 |
+
"learning_rate": 0.0008337531486146096,
|
215 |
+
"loss": 0.9257,
|
216 |
+
"step": 3300
|
217 |
+
},
|
218 |
+
{
|
219 |
+
"epoch": 8.56,
|
220 |
+
"learning_rate": 0.0008287153652392947,
|
221 |
+
"loss": 0.9468,
|
222 |
+
"step": 3400
|
223 |
+
},
|
224 |
+
{
|
225 |
+
"epoch": 8.82,
|
226 |
+
"learning_rate": 0.0008236775818639799,
|
227 |
+
"loss": 0.9501,
|
228 |
+
"step": 3500
|
229 |
+
},
|
230 |
+
{
|
231 |
+
"epoch": 9.07,
|
232 |
+
"learning_rate": 0.000818639798488665,
|
233 |
+
"loss": 0.923,
|
234 |
+
"step": 3600
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 9.32,
|
238 |
+
"learning_rate": 0.0008136020151133502,
|
239 |
+
"loss": 0.8187,
|
240 |
+
"step": 3700
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 9.57,
|
244 |
+
"learning_rate": 0.0008085642317380352,
|
245 |
+
"loss": 0.8214,
|
246 |
+
"step": 3800
|
247 |
+
},
|
248 |
+
{
|
249 |
+
"epoch": 9.82,
|
250 |
+
"learning_rate": 0.0008035264483627204,
|
251 |
+
"loss": 0.8451,
|
252 |
+
"step": 3900
|
253 |
+
},
|
254 |
+
{
|
255 |
+
"epoch": 10.08,
|
256 |
+
"learning_rate": 0.0007984886649874056,
|
257 |
+
"loss": 0.8056,
|
258 |
+
"step": 4000
|
259 |
+
},
|
260 |
+
{
|
261 |
+
"epoch": 10.08,
|
262 |
+
"eval_bleu": 4.0133,
|
263 |
+
"eval_gen_len": 16.3603,
|
264 |
+
"eval_loss": 0.48705288767814636,
|
265 |
+
"eval_runtime": 378.0534,
|
266 |
+
"eval_samples_per_second": 14.17,
|
267 |
+
"eval_steps_per_second": 3.544,
|
268 |
+
"step": 4000
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 10.33,
|
272 |
+
"learning_rate": 0.0007934508816120906,
|
273 |
+
"loss": 0.7131,
|
274 |
+
"step": 4100
|
275 |
+
},
|
276 |
+
{
|
277 |
+
"epoch": 10.58,
|
278 |
+
"learning_rate": 0.0007884130982367759,
|
279 |
+
"loss": 0.7319,
|
280 |
+
"step": 4200
|
281 |
+
},
|
282 |
+
{
|
283 |
+
"epoch": 10.83,
|
284 |
+
"learning_rate": 0.000783375314861461,
|
285 |
+
"loss": 0.7542,
|
286 |
+
"step": 4300
|
287 |
+
},
|
288 |
+
{
|
289 |
+
"epoch": 11.08,
|
290 |
+
"learning_rate": 0.0007783375314861461,
|
291 |
+
"loss": 0.7128,
|
292 |
+
"step": 4400
|
293 |
+
},
|
294 |
+
{
|
295 |
+
"epoch": 11.34,
|
296 |
+
"learning_rate": 0.0007732997481108313,
|
297 |
+
"loss": 0.6346,
|
298 |
+
"step": 4500
|
299 |
+
},
|
300 |
+
{
|
301 |
+
"epoch": 11.59,
|
302 |
+
"learning_rate": 0.0007682619647355163,
|
303 |
+
"loss": 0.6691,
|
304 |
+
"step": 4600
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 11.84,
|
308 |
+
"learning_rate": 0.0007632241813602016,
|
309 |
+
"loss": 0.6781,
|
310 |
+
"step": 4700
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 12.09,
|
314 |
+
"learning_rate": 0.0007581863979848867,
|
315 |
+
"loss": 0.6235,
|
316 |
+
"step": 4800
|
317 |
+
},
|
318 |
+
{
|
319 |
+
"epoch": 12.34,
|
320 |
+
"learning_rate": 0.0007531486146095718,
|
321 |
+
"loss": 0.5506,
|
322 |
+
"step": 4900
|
323 |
+
},
|
324 |
+
{
|
325 |
+
"epoch": 12.59,
|
326 |
+
"learning_rate": 0.000748110831234257,
|
327 |
+
"loss": 0.581,
|
328 |
+
"step": 5000
|
329 |
+
},
|
330 |
+
{
|
331 |
+
"epoch": 12.85,
|
332 |
+
"learning_rate": 0.000743073047858942,
|
333 |
+
"loss": 0.5951,
|
334 |
+
"step": 5100
|
335 |
+
},
|
336 |
+
{
|
337 |
+
"epoch": 13.1,
|
338 |
+
"learning_rate": 0.0007380352644836272,
|
339 |
+
"loss": 0.5726,
|
340 |
+
"step": 5200
|
341 |
+
},
|
342 |
+
{
|
343 |
+
"epoch": 13.35,
|
344 |
+
"learning_rate": 0.0007329974811083124,
|
345 |
+
"loss": 0.499,
|
346 |
+
"step": 5300
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"epoch": 13.6,
|
350 |
+
"learning_rate": 0.0007279596977329975,
|
351 |
+
"loss": 0.538,
|
352 |
+
"step": 5400
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 13.85,
|
356 |
+
"learning_rate": 0.0007229219143576826,
|
357 |
+
"loss": 0.5296,
|
358 |
+
"step": 5500
|
359 |
+
},
|
360 |
+
{
|
361 |
+
"epoch": 14.11,
|
362 |
+
"learning_rate": 0.0007178841309823679,
|
363 |
+
"loss": 0.5083,
|
364 |
+
"step": 5600
|
365 |
+
},
|
366 |
+
{
|
367 |
+
"epoch": 14.36,
|
368 |
+
"learning_rate": 0.0007128463476070529,
|
369 |
+
"loss": 0.4574,
|
370 |
+
"step": 5700
|
371 |
+
},
|
372 |
+
{
|
373 |
+
"epoch": 14.61,
|
374 |
+
"learning_rate": 0.000707808564231738,
|
375 |
+
"loss": 0.4762,
|
376 |
+
"step": 5800
|
377 |
+
},
|
378 |
+
{
|
379 |
+
"epoch": 14.86,
|
380 |
+
"learning_rate": 0.0007027707808564232,
|
381 |
+
"loss": 0.4883,
|
382 |
+
"step": 5900
|
383 |
+
},
|
384 |
+
{
|
385 |
+
"epoch": 15.11,
|
386 |
+
"learning_rate": 0.0006977329974811083,
|
387 |
+
"loss": 0.4478,
|
388 |
+
"step": 6000
|
389 |
+
},
|
390 |
+
{
|
391 |
+
"epoch": 15.11,
|
392 |
+
"eval_bleu": 4.6327,
|
393 |
+
"eval_gen_len": 15.9266,
|
394 |
+
"eval_loss": 0.2581786513328552,
|
395 |
+
"eval_runtime": 361.4738,
|
396 |
+
"eval_samples_per_second": 14.82,
|
397 |
+
"eval_steps_per_second": 3.707,
|
398 |
+
"step": 6000
|
399 |
+
},
|
400 |
+
{
|
401 |
+
"epoch": 15.37,
|
402 |
+
"learning_rate": 0.0006926952141057935,
|
403 |
+
"loss": 0.4233,
|
404 |
+
"step": 6100
|
405 |
+
},
|
406 |
+
{
|
407 |
+
"epoch": 15.62,
|
408 |
+
"learning_rate": 0.0006876574307304786,
|
409 |
+
"loss": 0.4295,
|
410 |
+
"step": 6200
|
411 |
+
},
|
412 |
+
{
|
413 |
+
"epoch": 15.87,
|
414 |
+
"learning_rate": 0.0006826196473551638,
|
415 |
+
"loss": 0.4491,
|
416 |
+
"step": 6300
|
417 |
+
},
|
418 |
+
{
|
419 |
+
"epoch": 16.12,
|
420 |
+
"learning_rate": 0.0006775818639798488,
|
421 |
+
"loss": 0.4257,
|
422 |
+
"step": 6400
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 16.37,
|
426 |
+
"learning_rate": 0.000672544080604534,
|
427 |
+
"loss": 0.386,
|
428 |
+
"step": 6500
|
429 |
+
},
|
430 |
+
{
|
431 |
+
"epoch": 16.62,
|
432 |
+
"learning_rate": 0.0006675062972292192,
|
433 |
+
"loss": 0.399,
|
434 |
+
"step": 6600
|
435 |
+
},
|
436 |
+
{
|
437 |
+
"epoch": 16.88,
|
438 |
+
"learning_rate": 0.0006624685138539042,
|
439 |
+
"loss": 0.405,
|
440 |
+
"step": 6700
|
441 |
+
},
|
442 |
+
{
|
443 |
+
"epoch": 17.13,
|
444 |
+
"learning_rate": 0.0006574307304785895,
|
445 |
+
"loss": 0.3709,
|
446 |
+
"step": 6800
|
447 |
+
},
|
448 |
+
{
|
449 |
+
"epoch": 17.38,
|
450 |
+
"learning_rate": 0.0006523929471032746,
|
451 |
+
"loss": 0.3499,
|
452 |
+
"step": 6900
|
453 |
+
},
|
454 |
+
{
|
455 |
+
"epoch": 17.63,
|
456 |
+
"learning_rate": 0.0006473551637279596,
|
457 |
+
"loss": 0.3677,
|
458 |
+
"step": 7000
|
459 |
+
},
|
460 |
+
{
|
461 |
+
"epoch": 17.88,
|
462 |
+
"learning_rate": 0.0006423173803526449,
|
463 |
+
"loss": 0.3744,
|
464 |
+
"step": 7100
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 18.14,
|
468 |
+
"learning_rate": 0.0006372795969773299,
|
469 |
+
"loss": 0.3409,
|
470 |
+
"step": 7200
|
471 |
+
},
|
472 |
+
{
|
473 |
+
"epoch": 18.39,
|
474 |
+
"learning_rate": 0.0006322418136020152,
|
475 |
+
"loss": 0.3273,
|
476 |
+
"step": 7300
|
477 |
+
},
|
478 |
+
{
|
479 |
+
"epoch": 18.64,
|
480 |
+
"learning_rate": 0.0006272040302267003,
|
481 |
+
"loss": 0.3412,
|
482 |
+
"step": 7400
|
483 |
+
},
|
484 |
+
{
|
485 |
+
"epoch": 18.89,
|
486 |
+
"learning_rate": 0.0006221662468513854,
|
487 |
+
"loss": 0.3472,
|
488 |
+
"step": 7500
|
489 |
+
},
|
490 |
+
{
|
491 |
+
"epoch": 19.14,
|
492 |
+
"learning_rate": 0.0006171284634760706,
|
493 |
+
"loss": 0.3321,
|
494 |
+
"step": 7600
|
495 |
+
},
|
496 |
+
{
|
497 |
+
"epoch": 19.4,
|
498 |
+
"learning_rate": 0.0006120906801007557,
|
499 |
+
"loss": 0.3113,
|
500 |
+
"step": 7700
|
501 |
+
},
|
502 |
+
{
|
503 |
+
"epoch": 19.65,
|
504 |
+
"learning_rate": 0.0006070528967254408,
|
505 |
+
"loss": 0.3141,
|
506 |
+
"step": 7800
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 19.9,
|
510 |
+
"learning_rate": 0.000602015113350126,
|
511 |
+
"loss": 0.3326,
|
512 |
+
"step": 7900
|
513 |
+
},
|
514 |
+
{
|
515 |
+
"epoch": 20.15,
|
516 |
+
"learning_rate": 0.0005969773299748111,
|
517 |
+
"loss": 0.2949,
|
518 |
+
"step": 8000
|
519 |
+
},
|
520 |
+
{
|
521 |
+
"epoch": 20.15,
|
522 |
+
"eval_bleu": 5.0752,
|
523 |
+
"eval_gen_len": 16.3638,
|
524 |
+
"eval_loss": 0.19623734056949615,
|
525 |
+
"eval_runtime": 389.6723,
|
526 |
+
"eval_samples_per_second": 13.747,
|
527 |
+
"eval_steps_per_second": 3.439,
|
528 |
+
"step": 8000
|
529 |
+
},
|
530 |
+
{
|
531 |
+
"epoch": 20.4,
|
532 |
+
"learning_rate": 0.0005919395465994962,
|
533 |
+
"loss": 0.2852,
|
534 |
+
"step": 8100
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 20.65,
|
538 |
+
"learning_rate": 0.0005869017632241814,
|
539 |
+
"loss": 0.3026,
|
540 |
+
"step": 8200
|
541 |
+
},
|
542 |
+
{
|
543 |
+
"epoch": 20.91,
|
544 |
+
"learning_rate": 0.0005818639798488665,
|
545 |
+
"loss": 0.3194,
|
546 |
+
"step": 8300
|
547 |
+
},
|
548 |
+
{
|
549 |
+
"epoch": 21.16,
|
550 |
+
"learning_rate": 0.0005768261964735516,
|
551 |
+
"loss": 0.2843,
|
552 |
+
"step": 8400
|
553 |
+
},
|
554 |
+
{
|
555 |
+
"epoch": 21.41,
|
556 |
+
"learning_rate": 0.0005717884130982368,
|
557 |
+
"loss": 0.2761,
|
558 |
+
"step": 8500
|
559 |
+
},
|
560 |
+
{
|
561 |
+
"epoch": 21.66,
|
562 |
+
"learning_rate": 0.0005667506297229219,
|
563 |
+
"loss": 0.2826,
|
564 |
+
"step": 8600
|
565 |
+
},
|
566 |
+
{
|
567 |
+
"epoch": 21.91,
|
568 |
+
"learning_rate": 0.0005617128463476071,
|
569 |
+
"loss": 0.2929,
|
570 |
+
"step": 8700
|
571 |
+
},
|
572 |
+
{
|
573 |
+
"epoch": 22.17,
|
574 |
+
"learning_rate": 0.0005566750629722922,
|
575 |
+
"loss": 0.2638,
|
576 |
+
"step": 8800
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 22.42,
|
580 |
+
"learning_rate": 0.0005516372795969773,
|
581 |
+
"loss": 0.2595,
|
582 |
+
"step": 8900
|
583 |
+
},
|
584 |
+
{
|
585 |
+
"epoch": 22.67,
|
586 |
+
"learning_rate": 0.0005465994962216625,
|
587 |
+
"loss": 0.2747,
|
588 |
+
"step": 9000
|
589 |
+
},
|
590 |
+
{
|
591 |
+
"epoch": 22.92,
|
592 |
+
"learning_rate": 0.0005415617128463476,
|
593 |
+
"loss": 0.2806,
|
594 |
+
"step": 9100
|
595 |
+
},
|
596 |
+
{
|
597 |
+
"epoch": 23.17,
|
598 |
+
"learning_rate": 0.0005365239294710328,
|
599 |
+
"loss": 0.2588,
|
600 |
+
"step": 9200
|
601 |
+
},
|
602 |
+
{
|
603 |
+
"epoch": 23.43,
|
604 |
+
"learning_rate": 0.0005314861460957178,
|
605 |
+
"loss": 0.2559,
|
606 |
+
"step": 9300
|
607 |
+
},
|
608 |
+
{
|
609 |
+
"epoch": 23.68,
|
610 |
+
"learning_rate": 0.0005264483627204031,
|
611 |
+
"loss": 0.2597,
|
612 |
+
"step": 9400
|
613 |
+
},
|
614 |
+
{
|
615 |
+
"epoch": 23.93,
|
616 |
+
"learning_rate": 0.0005214105793450882,
|
617 |
+
"loss": 0.262,
|
618 |
+
"step": 9500
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 24.18,
|
622 |
+
"learning_rate": 0.0005163727959697732,
|
623 |
+
"loss": 0.2408,
|
624 |
+
"step": 9600
|
625 |
+
},
|
626 |
+
{
|
627 |
+
"epoch": 24.43,
|
628 |
+
"learning_rate": 0.0005113350125944585,
|
629 |
+
"loss": 0.244,
|
630 |
+
"step": 9700
|
631 |
+
},
|
632 |
+
{
|
633 |
+
"epoch": 24.69,
|
634 |
+
"learning_rate": 0.0005062972292191435,
|
635 |
+
"loss": 0.2435,
|
636 |
+
"step": 9800
|
637 |
+
},
|
638 |
+
{
|
639 |
+
"epoch": 24.94,
|
640 |
+
"learning_rate": 0.0005012594458438288,
|
641 |
+
"loss": 0.2535,
|
642 |
+
"step": 9900
|
643 |
+
},
|
644 |
+
{
|
645 |
+
"epoch": 25.19,
|
646 |
+
"learning_rate": 0.0004962216624685139,
|
647 |
+
"loss": 0.2261,
|
648 |
+
"step": 10000
|
649 |
+
},
|
650 |
+
{
|
651 |
+
"epoch": 25.19,
|
652 |
+
"eval_bleu": 4.9239,
|
653 |
+
"eval_gen_len": 16.3816,
|
654 |
+
"eval_loss": 0.17935612797737122,
|
655 |
+
"eval_runtime": 383.1714,
|
656 |
+
"eval_samples_per_second": 13.981,
|
657 |
+
"eval_steps_per_second": 3.497,
|
658 |
+
"step": 10000
|
659 |
+
},
|
660 |
+
{
|
661 |
+
"epoch": 25.44,
|
662 |
+
"learning_rate": 0.000491183879093199,
|
663 |
+
"loss": 0.2352,
|
664 |
+
"step": 10100
|
665 |
+
},
|
666 |
+
{
|
667 |
+
"epoch": 25.69,
|
668 |
+
"learning_rate": 0.00048614609571788416,
|
669 |
+
"loss": 0.2443,
|
670 |
+
"step": 10200
|
671 |
+
},
|
672 |
+
{
|
673 |
+
"epoch": 25.94,
|
674 |
+
"learning_rate": 0.0004811083123425693,
|
675 |
+
"loss": 0.2473,
|
676 |
+
"step": 10300
|
677 |
+
},
|
678 |
+
{
|
679 |
+
"epoch": 26.2,
|
680 |
+
"learning_rate": 0.00047607052896725437,
|
681 |
+
"loss": 0.2178,
|
682 |
+
"step": 10400
|
683 |
+
},
|
684 |
+
{
|
685 |
+
"epoch": 26.45,
|
686 |
+
"learning_rate": 0.00047103274559193956,
|
687 |
+
"loss": 0.2288,
|
688 |
+
"step": 10500
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 26.7,
|
692 |
+
"learning_rate": 0.0004659949622166247,
|
693 |
+
"loss": 0.2285,
|
694 |
+
"step": 10600
|
695 |
+
},
|
696 |
+
{
|
697 |
+
"epoch": 26.95,
|
698 |
+
"learning_rate": 0.00046095717884130983,
|
699 |
+
"loss": 0.2374,
|
700 |
+
"step": 10700
|
701 |
+
},
|
702 |
+
{
|
703 |
+
"epoch": 27.2,
|
704 |
+
"learning_rate": 0.00045591939546599496,
|
705 |
+
"loss": 0.2126,
|
706 |
+
"step": 10800
|
707 |
+
},
|
708 |
+
{
|
709 |
+
"epoch": 27.46,
|
710 |
+
"learning_rate": 0.00045088161209068015,
|
711 |
+
"loss": 0.2142,
|
712 |
+
"step": 10900
|
713 |
+
},
|
714 |
+
{
|
715 |
+
"epoch": 27.71,
|
716 |
+
"learning_rate": 0.0004458438287153653,
|
717 |
+
"loss": 0.2182,
|
718 |
+
"step": 11000
|
719 |
+
},
|
720 |
+
{
|
721 |
+
"epoch": 27.96,
|
722 |
+
"learning_rate": 0.00044080604534005037,
|
723 |
+
"loss": 0.2314,
|
724 |
+
"step": 11100
|
725 |
+
},
|
726 |
+
{
|
727 |
+
"epoch": 28.21,
|
728 |
+
"learning_rate": 0.0004357682619647355,
|
729 |
+
"loss": 0.2052,
|
730 |
+
"step": 11200
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 28.46,
|
734 |
+
"learning_rate": 0.00043073047858942063,
|
735 |
+
"loss": 0.2095,
|
736 |
+
"step": 11300
|
737 |
+
},
|
738 |
+
{
|
739 |
+
"epoch": 28.72,
|
740 |
+
"learning_rate": 0.0004256926952141058,
|
741 |
+
"loss": 0.2196,
|
742 |
+
"step": 11400
|
743 |
+
},
|
744 |
+
{
|
745 |
+
"epoch": 28.97,
|
746 |
+
"learning_rate": 0.00042065491183879096,
|
747 |
+
"loss": 0.2205,
|
748 |
+
"step": 11500
|
749 |
+
},
|
750 |
+
{
|
751 |
+
"epoch": 29.22,
|
752 |
+
"learning_rate": 0.0004156171284634761,
|
753 |
+
"loss": 0.2,
|
754 |
+
"step": 11600
|
755 |
+
},
|
756 |
+
{
|
757 |
+
"epoch": 29.47,
|
758 |
+
"learning_rate": 0.00041057934508816117,
|
759 |
+
"loss": 0.2051,
|
760 |
+
"step": 11700
|
761 |
+
},
|
762 |
+
{
|
763 |
+
"epoch": 29.72,
|
764 |
+
"learning_rate": 0.00040554156171284636,
|
765 |
+
"loss": 0.2124,
|
766 |
+
"step": 11800
|
767 |
+
},
|
768 |
+
{
|
769 |
+
"epoch": 29.97,
|
770 |
+
"learning_rate": 0.0004005037783375315,
|
771 |
+
"loss": 0.2159,
|
772 |
+
"step": 11900
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 30.23,
|
776 |
+
"learning_rate": 0.00039546599496221663,
|
777 |
+
"loss": 0.1903,
|
778 |
+
"step": 12000
|
779 |
+
},
|
780 |
+
{
|
781 |
+
"epoch": 30.23,
|
782 |
+
"eval_bleu": 5.1033,
|
783 |
+
"eval_gen_len": 16.1934,
|
784 |
+
"eval_loss": 0.17088228464126587,
|
785 |
+
"eval_runtime": 381.8763,
|
786 |
+
"eval_samples_per_second": 14.028,
|
787 |
+
"eval_steps_per_second": 3.509,
|
788 |
+
"step": 12000
|
789 |
+
},
|
790 |
+
{
|
791 |
+
"epoch": 30.48,
|
792 |
+
"learning_rate": 0.00039042821158690176,
|
793 |
+
"loss": 0.1995,
|
794 |
+
"step": 12100
|
795 |
+
},
|
796 |
+
{
|
797 |
+
"epoch": 30.73,
|
798 |
+
"learning_rate": 0.00038539042821158695,
|
799 |
+
"loss": 0.2046,
|
800 |
+
"step": 12200
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 30.98,
|
804 |
+
"learning_rate": 0.00038035264483627203,
|
805 |
+
"loss": 0.2127,
|
806 |
+
"step": 12300
|
807 |
+
},
|
808 |
+
{
|
809 |
+
"epoch": 31.23,
|
810 |
+
"learning_rate": 0.00037531486146095717,
|
811 |
+
"loss": 0.1923,
|
812 |
+
"step": 12400
|
813 |
+
},
|
814 |
+
{
|
815 |
+
"epoch": 31.49,
|
816 |
+
"learning_rate": 0.0003702770780856423,
|
817 |
+
"loss": 0.1931,
|
818 |
+
"step": 12500
|
819 |
+
},
|
820 |
+
{
|
821 |
+
"epoch": 31.74,
|
822 |
+
"learning_rate": 0.0003652392947103275,
|
823 |
+
"loss": 0.1969,
|
824 |
+
"step": 12600
|
825 |
+
},
|
826 |
+
{
|
827 |
+
"epoch": 31.99,
|
828 |
+
"learning_rate": 0.0003602015113350126,
|
829 |
+
"loss": 0.2025,
|
830 |
+
"step": 12700
|
831 |
+
},
|
832 |
+
{
|
833 |
+
"epoch": 32.24,
|
834 |
+
"learning_rate": 0.00035516372795969776,
|
835 |
+
"loss": 0.1826,
|
836 |
+
"step": 12800
|
837 |
+
},
|
838 |
+
{
|
839 |
+
"epoch": 32.49,
|
840 |
+
"learning_rate": 0.0003501259445843829,
|
841 |
+
"loss": 0.1885,
|
842 |
+
"step": 12900
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 32.75,
|
846 |
+
"learning_rate": 0.00034508816120906797,
|
847 |
+
"loss": 0.1935,
|
848 |
+
"step": 13000
|
849 |
+
},
|
850 |
+
{
|
851 |
+
"epoch": 33.0,
|
852 |
+
"learning_rate": 0.00034005037783375316,
|
853 |
+
"loss": 0.2017,
|
854 |
+
"step": 13100
|
855 |
+
},
|
856 |
+
{
|
857 |
+
"epoch": 33.25,
|
858 |
+
"learning_rate": 0.0003350125944584383,
|
859 |
+
"loss": 0.1778,
|
860 |
+
"step": 13200
|
861 |
+
},
|
862 |
+
{
|
863 |
+
"epoch": 33.5,
|
864 |
+
"learning_rate": 0.00032997481108312343,
|
865 |
+
"loss": 0.1842,
|
866 |
+
"step": 13300
|
867 |
+
},
|
868 |
+
{
|
869 |
+
"epoch": 33.75,
|
870 |
+
"learning_rate": 0.00032493702770780856,
|
871 |
+
"loss": 0.1929,
|
872 |
+
"step": 13400
|
873 |
+
},
|
874 |
+
{
|
875 |
+
"epoch": 34.01,
|
876 |
+
"learning_rate": 0.00031989924433249375,
|
877 |
+
"loss": 0.1927,
|
878 |
+
"step": 13500
|
879 |
+
},
|
880 |
+
{
|
881 |
+
"epoch": 34.26,
|
882 |
+
"learning_rate": 0.00031486146095717883,
|
883 |
+
"loss": 0.1768,
|
884 |
+
"step": 13600
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 34.51,
|
888 |
+
"learning_rate": 0.00030982367758186397,
|
889 |
+
"loss": 0.1814,
|
890 |
+
"step": 13700
|
891 |
+
},
|
892 |
+
{
|
893 |
+
"epoch": 34.76,
|
894 |
+
"learning_rate": 0.0003047858942065491,
|
895 |
+
"loss": 0.1823,
|
896 |
+
"step": 13800
|
897 |
+
},
|
898 |
+
{
|
899 |
+
"epoch": 35.01,
|
900 |
+
"learning_rate": 0.0002997481108312343,
|
901 |
+
"loss": 0.1824,
|
902 |
+
"step": 13900
|
903 |
+
},
|
904 |
+
{
|
905 |
+
"epoch": 35.26,
|
906 |
+
"learning_rate": 0.0002947103274559194,
|
907 |
+
"loss": 0.1689,
|
908 |
+
"step": 14000
|
909 |
+
},
|
910 |
+
{
|
911 |
+
"epoch": 35.26,
|
912 |
+
"eval_bleu": 5.0783,
|
913 |
+
"eval_gen_len": 16.5555,
|
914 |
+
"eval_loss": 0.16545478999614716,
|
915 |
+
"eval_runtime": 405.9141,
|
916 |
+
"eval_samples_per_second": 13.197,
|
917 |
+
"eval_steps_per_second": 3.301,
|
918 |
+
"step": 14000
|
919 |
+
},
|
920 |
+
{
|
921 |
+
"epoch": 35.52,
|
922 |
+
"learning_rate": 0.00028967254408060456,
|
923 |
+
"loss": 0.1783,
|
924 |
+
"step": 14100
|
925 |
+
},
|
926 |
+
{
|
927 |
+
"epoch": 35.77,
|
928 |
+
"learning_rate": 0.00028463476070528964,
|
929 |
+
"loss": 0.1759,
|
930 |
+
"step": 14200
|
931 |
+
},
|
932 |
+
{
|
933 |
+
"epoch": 36.02,
|
934 |
+
"learning_rate": 0.0002795969773299748,
|
935 |
+
"loss": 0.1857,
|
936 |
+
"step": 14300
|
937 |
+
},
|
938 |
+
{
|
939 |
+
"epoch": 36.27,
|
940 |
+
"learning_rate": 0.00027455919395465996,
|
941 |
+
"loss": 0.1669,
|
942 |
+
"step": 14400
|
943 |
+
},
|
944 |
+
{
|
945 |
+
"epoch": 36.52,
|
946 |
+
"learning_rate": 0.0002695214105793451,
|
947 |
+
"loss": 0.173,
|
948 |
+
"step": 14500
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"epoch": 36.78,
|
952 |
+
"learning_rate": 0.00026448362720403023,
|
953 |
+
"loss": 0.1752,
|
954 |
+
"step": 14600
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 37.03,
|
958 |
+
"learning_rate": 0.00025944584382871536,
|
959 |
+
"loss": 0.1784,
|
960 |
+
"step": 14700
|
961 |
+
},
|
962 |
+
{
|
963 |
+
"epoch": 37.28,
|
964 |
+
"learning_rate": 0.00025440806045340055,
|
965 |
+
"loss": 0.1674,
|
966 |
+
"step": 14800
|
967 |
+
},
|
968 |
+
{
|
969 |
+
"epoch": 37.53,
|
970 |
+
"learning_rate": 0.00024937027707808563,
|
971 |
+
"loss": 0.1678,
|
972 |
+
"step": 14900
|
973 |
+
},
|
974 |
+
{
|
975 |
+
"epoch": 37.78,
|
976 |
+
"learning_rate": 0.00024433249370277077,
|
977 |
+
"loss": 0.1714,
|
978 |
+
"step": 15000
|
979 |
+
},
|
980 |
+
{
|
981 |
+
"epoch": 38.04,
|
982 |
+
"learning_rate": 0.00023929471032745593,
|
983 |
+
"loss": 0.1697,
|
984 |
+
"step": 15100
|
985 |
+
},
|
986 |
+
{
|
987 |
+
"epoch": 38.29,
|
988 |
+
"learning_rate": 0.00023425692695214106,
|
989 |
+
"loss": 0.1638,
|
990 |
+
"step": 15200
|
991 |
+
},
|
992 |
+
{
|
993 |
+
"epoch": 38.54,
|
994 |
+
"learning_rate": 0.0002292191435768262,
|
995 |
+
"loss": 0.1667,
|
996 |
+
"step": 15300
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 38.79,
|
1000 |
+
"learning_rate": 0.00022418136020151133,
|
1001 |
+
"loss": 0.1714,
|
1002 |
+
"step": 15400
|
1003 |
+
},
|
1004 |
+
{
|
1005 |
+
"epoch": 39.04,
|
1006 |
+
"learning_rate": 0.0002191435768261965,
|
1007 |
+
"loss": 0.1642,
|
1008 |
+
"step": 15500
|
1009 |
+
},
|
1010 |
+
{
|
1011 |
+
"epoch": 39.29,
|
1012 |
+
"learning_rate": 0.0002141057934508816,
|
1013 |
+
"loss": 0.1585,
|
1014 |
+
"step": 15600
|
1015 |
+
},
|
1016 |
+
{
|
1017 |
+
"epoch": 39.55,
|
1018 |
+
"learning_rate": 0.00020906801007556676,
|
1019 |
+
"loss": 0.1607,
|
1020 |
+
"step": 15700
|
1021 |
+
},
|
1022 |
+
{
|
1023 |
+
"epoch": 39.8,
|
1024 |
+
"learning_rate": 0.0002040302267002519,
|
1025 |
+
"loss": 0.1643,
|
1026 |
+
"step": 15800
|
1027 |
+
},
|
1028 |
+
{
|
1029 |
+
"epoch": 40.05,
|
1030 |
+
"learning_rate": 0.00019899244332493706,
|
1031 |
+
"loss": 0.1662,
|
1032 |
+
"step": 15900
|
1033 |
+
},
|
1034 |
+
{
|
1035 |
+
"epoch": 40.3,
|
1036 |
+
"learning_rate": 0.00019395465994962217,
|
1037 |
+
"loss": 0.1571,
|
1038 |
+
"step": 16000
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 40.3,
|
1042 |
+
"eval_bleu": 5.197,
|
1043 |
+
"eval_gen_len": 16.6603,
|
1044 |
+
"eval_loss": 0.16155509650707245,
|
1045 |
+
"eval_runtime": 388.2766,
|
1046 |
+
"eval_samples_per_second": 13.797,
|
1047 |
+
"eval_steps_per_second": 3.451,
|
1048 |
+
"step": 16000
|
1049 |
+
},
|
1050 |
+
{
|
1051 |
+
"epoch": 40.55,
|
1052 |
+
"learning_rate": 0.0001889168765743073,
|
1053 |
+
"loss": 0.1597,
|
1054 |
+
"step": 16100
|
1055 |
+
},
|
1056 |
+
{
|
1057 |
+
"epoch": 40.81,
|
1058 |
+
"learning_rate": 0.00018387909319899246,
|
1059 |
+
"loss": 0.1634,
|
1060 |
+
"step": 16200
|
1061 |
+
},
|
1062 |
+
{
|
1063 |
+
"epoch": 41.06,
|
1064 |
+
"learning_rate": 0.00017884130982367757,
|
1065 |
+
"loss": 0.1609,
|
1066 |
+
"step": 16300
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 41.31,
|
1070 |
+
"learning_rate": 0.00017380352644836273,
|
1071 |
+
"loss": 0.1526,
|
1072 |
+
"step": 16400
|
1073 |
+
},
|
1074 |
+
{
|
1075 |
+
"epoch": 41.56,
|
1076 |
+
"learning_rate": 0.00016876574307304786,
|
1077 |
+
"loss": 0.1536,
|
1078 |
+
"step": 16500
|
1079 |
+
},
|
1080 |
+
{
|
1081 |
+
"epoch": 41.81,
|
1082 |
+
"learning_rate": 0.000163727959697733,
|
1083 |
+
"loss": 0.1606,
|
1084 |
+
"step": 16600
|
1085 |
+
},
|
1086 |
+
{
|
1087 |
+
"epoch": 42.07,
|
1088 |
+
"learning_rate": 0.00015869017632241813,
|
1089 |
+
"loss": 0.1603,
|
1090 |
+
"step": 16700
|
1091 |
+
},
|
1092 |
+
{
|
1093 |
+
"epoch": 42.32,
|
1094 |
+
"learning_rate": 0.0001536523929471033,
|
1095 |
+
"loss": 0.1522,
|
1096 |
+
"step": 16800
|
1097 |
+
},
|
1098 |
+
{
|
1099 |
+
"epoch": 42.57,
|
1100 |
+
"learning_rate": 0.0001486146095717884,
|
1101 |
+
"loss": 0.1554,
|
1102 |
+
"step": 16900
|
1103 |
+
},
|
1104 |
+
{
|
1105 |
+
"epoch": 42.82,
|
1106 |
+
"learning_rate": 0.00014357682619647356,
|
1107 |
+
"loss": 0.1567,
|
1108 |
+
"step": 17000
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 43.07,
|
1112 |
+
"learning_rate": 0.0001385390428211587,
|
1113 |
+
"loss": 0.1537,
|
1114 |
+
"step": 17100
|
1115 |
+
},
|
1116 |
+
{
|
1117 |
+
"epoch": 43.32,
|
1118 |
+
"learning_rate": 0.00013350125944584383,
|
1119 |
+
"loss": 0.1506,
|
1120 |
+
"step": 17200
|
1121 |
+
},
|
1122 |
+
{
|
1123 |
+
"epoch": 43.58,
|
1124 |
+
"learning_rate": 0.00012846347607052897,
|
1125 |
+
"loss": 0.1513,
|
1126 |
+
"step": 17300
|
1127 |
+
},
|
1128 |
+
{
|
1129 |
+
"epoch": 43.83,
|
1130 |
+
"learning_rate": 0.00012342569269521413,
|
1131 |
+
"loss": 0.1563,
|
1132 |
+
"step": 17400
|
1133 |
+
},
|
1134 |
+
{
|
1135 |
+
"epoch": 44.08,
|
1136 |
+
"learning_rate": 0.00011838790931989925,
|
1137 |
+
"loss": 0.1503,
|
1138 |
+
"step": 17500
|
1139 |
+
},
|
1140 |
+
{
|
1141 |
+
"epoch": 44.33,
|
1142 |
+
"learning_rate": 0.00011335012594458438,
|
1143 |
+
"loss": 0.1456,
|
1144 |
+
"step": 17600
|
1145 |
+
},
|
1146 |
+
{
|
1147 |
+
"epoch": 44.58,
|
1148 |
+
"learning_rate": 0.00010831234256926953,
|
1149 |
+
"loss": 0.1486,
|
1150 |
+
"step": 17700
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 44.84,
|
1154 |
+
"learning_rate": 0.00010327455919395466,
|
1155 |
+
"loss": 0.1501,
|
1156 |
+
"step": 17800
|
1157 |
+
},
|
1158 |
+
{
|
1159 |
+
"epoch": 45.09,
|
1160 |
+
"learning_rate": 9.82367758186398e-05,
|
1161 |
+
"loss": 0.1497,
|
1162 |
+
"step": 17900
|
1163 |
+
},
|
1164 |
+
{
|
1165 |
+
"epoch": 45.34,
|
1166 |
+
"learning_rate": 9.319899244332495e-05,
|
1167 |
+
"loss": 0.1463,
|
1168 |
+
"step": 18000
|
1169 |
+
},
|
1170 |
+
{
|
1171 |
+
"epoch": 45.34,
|
1172 |
+
"eval_bleu": 5.2322,
|
1173 |
+
"eval_gen_len": 16.6093,
|
1174 |
+
"eval_loss": 0.15787681937217712,
|
1175 |
+
"eval_runtime": 391.3483,
|
1176 |
+
"eval_samples_per_second": 13.689,
|
1177 |
+
"eval_steps_per_second": 3.424,
|
1178 |
+
"step": 18000
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 45.59,
|
1182 |
+
"learning_rate": 8.816120906801008e-05,
|
1183 |
+
"loss": 0.1457,
|
1184 |
+
"step": 18100
|
1185 |
+
},
|
1186 |
+
{
|
1187 |
+
"epoch": 45.84,
|
1188 |
+
"learning_rate": 8.312342569269522e-05,
|
1189 |
+
"loss": 0.1471,
|
1190 |
+
"step": 18200
|
1191 |
+
},
|
1192 |
+
{
|
1193 |
+
"epoch": 46.1,
|
1194 |
+
"learning_rate": 7.808564231738035e-05,
|
1195 |
+
"loss": 0.1492,
|
1196 |
+
"step": 18300
|
1197 |
+
},
|
1198 |
+
{
|
1199 |
+
"epoch": 46.35,
|
1200 |
+
"learning_rate": 7.304785894206548e-05,
|
1201 |
+
"loss": 0.1398,
|
1202 |
+
"step": 18400
|
1203 |
+
},
|
1204 |
+
{
|
1205 |
+
"epoch": 46.6,
|
1206 |
+
"learning_rate": 6.801007556675063e-05,
|
1207 |
+
"loss": 0.1436,
|
1208 |
+
"step": 18500
|
1209 |
+
},
|
1210 |
+
{
|
1211 |
+
"epoch": 46.85,
|
1212 |
+
"learning_rate": 6.297229219143577e-05,
|
1213 |
+
"loss": 0.1452,
|
1214 |
+
"step": 18600
|
1215 |
+
},
|
1216 |
+
{
|
1217 |
+
"epoch": 47.1,
|
1218 |
+
"learning_rate": 5.793450881612091e-05,
|
1219 |
+
"loss": 0.1452,
|
1220 |
+
"step": 18700
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 47.36,
|
1224 |
+
"learning_rate": 5.289672544080604e-05,
|
1225 |
+
"loss": 0.1411,
|
1226 |
+
"step": 18800
|
1227 |
+
},
|
1228 |
+
{
|
1229 |
+
"epoch": 47.61,
|
1230 |
+
"learning_rate": 4.785894206549118e-05,
|
1231 |
+
"loss": 0.1436,
|
1232 |
+
"step": 18900
|
1233 |
+
},
|
1234 |
+
{
|
1235 |
+
"epoch": 47.86,
|
1236 |
+
"learning_rate": 4.2821158690176324e-05,
|
1237 |
+
"loss": 0.1442,
|
1238 |
+
"step": 19000
|
1239 |
+
},
|
1240 |
+
{
|
1241 |
+
"epoch": 48.11,
|
1242 |
+
"learning_rate": 3.7783375314861465e-05,
|
1243 |
+
"loss": 0.1434,
|
1244 |
+
"step": 19100
|
1245 |
+
},
|
1246 |
+
{
|
1247 |
+
"epoch": 48.36,
|
1248 |
+
"learning_rate": 3.27455919395466e-05,
|
1249 |
+
"loss": 0.1379,
|
1250 |
+
"step": 19200
|
1251 |
+
},
|
1252 |
+
{
|
1253 |
+
"epoch": 48.61,
|
1254 |
+
"learning_rate": 2.7707808564231737e-05,
|
1255 |
+
"loss": 0.1403,
|
1256 |
+
"step": 19300
|
1257 |
+
},
|
1258 |
+
{
|
1259 |
+
"epoch": 48.87,
|
1260 |
+
"learning_rate": 2.2670025188916875e-05,
|
1261 |
+
"loss": 0.1437,
|
1262 |
+
"step": 19400
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 49.12,
|
1266 |
+
"learning_rate": 1.7632241813602016e-05,
|
1267 |
+
"loss": 0.14,
|
1268 |
+
"step": 19500
|
1269 |
+
},
|
1270 |
+
{
|
1271 |
+
"epoch": 49.37,
|
1272 |
+
"learning_rate": 1.2594458438287154e-05,
|
1273 |
+
"loss": 0.1398,
|
1274 |
+
"step": 19600
|
1275 |
+
},
|
1276 |
+
{
|
1277 |
+
"epoch": 49.62,
|
1278 |
+
"learning_rate": 7.5566750629722926e-06,
|
1279 |
+
"loss": 0.1365,
|
1280 |
+
"step": 19700
|
1281 |
+
},
|
1282 |
+
{
|
1283 |
+
"epoch": 49.87,
|
1284 |
+
"learning_rate": 2.5188916876574307e-06,
|
1285 |
+
"loss": 0.1407,
|
1286 |
+
"step": 19800
|
1287 |
+
},
|
1288 |
+
{
|
1289 |
+
"epoch": 50.0,
|
1290 |
+
"step": 19850,
|
1291 |
+
"total_flos": 1.4648767908102144e+17,
|
1292 |
+
"train_loss": 0.5778315416091034,
|
1293 |
+
"train_runtime": 26224.0587,
|
1294 |
+
"train_samples_per_second": 24.222,
|
1295 |
+
"train_steps_per_second": 0.757
|
1296 |
+
}
|
1297 |
+
],
|
1298 |
+
"max_steps": 19850,
|
1299 |
+
"num_train_epochs": 50,
|
1300 |
+
"total_flos": 1.4648767908102144e+17,
|
1301 |
+
"trial_name": null,
|
1302 |
+
"trial_params": null
|
1303 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7d613f2a4ed6c75eb8951f51092dec11ed78b3bcd6f44f9dd74e7e58394b0fa9
|
3 |
+
size 2927
|