Add model weights and configuation
Browse files- config.json +25 -0
- merges.txt +0 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +976 -0
- training_args.bin +3 -0
- vocab.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"RobertaForMaskedLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"eos_token_id": 2,
|
| 8 |
+
"gradient_checkpointing": false,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 3072,
|
| 14 |
+
"layer_norm_eps": 1e-12,
|
| 15 |
+
"max_position_embeddings": 512,
|
| 16 |
+
"model_type": "roberta",
|
| 17 |
+
"num_attention_heads": 12,
|
| 18 |
+
"num_hidden_layers": 6,
|
| 19 |
+
"pad_token_id": 1,
|
| 20 |
+
"position_embedding_type": "absolute",
|
| 21 |
+
"transformers_version": "4.7.0",
|
| 22 |
+
"type_vocab_size": 1,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 52000
|
| 25 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d79df558f5635ef213ea35252b92b057e8e19909c87fe450060e8612552c6103
|
| 3 |
+
size 668089827
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:35a675d3a9e98d9076bed3c30605cdf0da745faf838cd717a4391b6b7e571159
|
| 3 |
+
size 334061362
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5c225307225715ecaf74c677143570d7348f3994ae5ad806f95a91433102735b
|
| 3 |
+
size 16723
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d5424c0c7c47bc61f62268037c3c6d3b8b40b13d5f1be76ee62452061c965ef5
|
| 3 |
+
size 623
|
trainer_state.json
ADDED
|
@@ -0,0 +1,976 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.15017728468489,
|
| 5 |
+
"global_step": 80000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.06,
|
| 12 |
+
"learning_rate": 4.97140569598536e-05,
|
| 13 |
+
"loss": 8.5722,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.11,
|
| 18 |
+
"learning_rate": 4.94281139197072e-05,
|
| 19 |
+
"loss": 7.9252,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.17,
|
| 24 |
+
"learning_rate": 4.91421708795608e-05,
|
| 25 |
+
"loss": 7.6163,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.23,
|
| 30 |
+
"learning_rate": 4.885622783941439e-05,
|
| 31 |
+
"loss": 7.3317,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.29,
|
| 36 |
+
"learning_rate": 4.857028479926799e-05,
|
| 37 |
+
"loss": 6.9857,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.34,
|
| 42 |
+
"learning_rate": 4.8284341759121586e-05,
|
| 43 |
+
"loss": 6.6297,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.4,
|
| 48 |
+
"learning_rate": 4.7998398718975185e-05,
|
| 49 |
+
"loss": 6.3251,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.46,
|
| 54 |
+
"learning_rate": 4.7712455678828777e-05,
|
| 55 |
+
"loss": 6.0786,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.51,
|
| 60 |
+
"learning_rate": 4.7426512638682375e-05,
|
| 61 |
+
"loss": 5.8383,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.57,
|
| 66 |
+
"learning_rate": 4.7140569598535973e-05,
|
| 67 |
+
"loss": 5.62,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.63,
|
| 72 |
+
"learning_rate": 4.685462655838957e-05,
|
| 73 |
+
"loss": 5.4495,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.69,
|
| 78 |
+
"learning_rate": 4.6568683518243164e-05,
|
| 79 |
+
"loss": 5.2876,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.74,
|
| 84 |
+
"learning_rate": 4.628274047809676e-05,
|
| 85 |
+
"loss": 5.1378,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.8,
|
| 90 |
+
"learning_rate": 4.599679743795036e-05,
|
| 91 |
+
"loss": 5.0281,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.86,
|
| 96 |
+
"learning_rate": 4.571085439780396e-05,
|
| 97 |
+
"loss": 4.9177,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.92,
|
| 102 |
+
"learning_rate": 4.542491135765756e-05,
|
| 103 |
+
"loss": 4.8254,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.97,
|
| 108 |
+
"learning_rate": 4.5138968317511156e-05,
|
| 109 |
+
"loss": 4.7354,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 1.03,
|
| 114 |
+
"learning_rate": 4.4853025277364754e-05,
|
| 115 |
+
"loss": 4.6366,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.09,
|
| 120 |
+
"learning_rate": 4.4567082237218346e-05,
|
| 121 |
+
"loss": 4.5642,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.14,
|
| 126 |
+
"learning_rate": 4.4281139197071944e-05,
|
| 127 |
+
"loss": 4.5011,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.2,
|
| 132 |
+
"learning_rate": 4.399519615692554e-05,
|
| 133 |
+
"loss": 4.4285,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 1.26,
|
| 138 |
+
"learning_rate": 4.370925311677914e-05,
|
| 139 |
+
"loss": 4.3703,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.32,
|
| 144 |
+
"learning_rate": 4.342331007663274e-05,
|
| 145 |
+
"loss": 4.3203,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 1.37,
|
| 150 |
+
"learning_rate": 4.313736703648633e-05,
|
| 151 |
+
"loss": 4.2678,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.43,
|
| 156 |
+
"learning_rate": 4.285142399633993e-05,
|
| 157 |
+
"loss": 4.2177,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.49,
|
| 162 |
+
"learning_rate": 4.256548095619353e-05,
|
| 163 |
+
"loss": 4.1607,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.54,
|
| 168 |
+
"learning_rate": 4.227953791604713e-05,
|
| 169 |
+
"loss": 4.132,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.6,
|
| 174 |
+
"learning_rate": 4.199359487590072e-05,
|
| 175 |
+
"loss": 4.0738,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.66,
|
| 180 |
+
"learning_rate": 4.170765183575432e-05,
|
| 181 |
+
"loss": 4.0348,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.72,
|
| 186 |
+
"learning_rate": 4.1421708795607916e-05,
|
| 187 |
+
"loss": 3.9979,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 1.77,
|
| 192 |
+
"learning_rate": 4.1135765755461514e-05,
|
| 193 |
+
"loss": 3.9553,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 1.83,
|
| 198 |
+
"learning_rate": 4.084982271531511e-05,
|
| 199 |
+
"loss": 3.917,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.89,
|
| 204 |
+
"learning_rate": 4.056387967516871e-05,
|
| 205 |
+
"loss": 3.8866,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.94,
|
| 210 |
+
"learning_rate": 4.027793663502231e-05,
|
| 211 |
+
"loss": 3.8605,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 2.0,
|
| 216 |
+
"learning_rate": 3.99919935948759e-05,
|
| 217 |
+
"loss": 3.8163,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 2.06,
|
| 222 |
+
"learning_rate": 3.97060505547295e-05,
|
| 223 |
+
"loss": 3.7707,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 2.12,
|
| 228 |
+
"learning_rate": 3.94201075145831e-05,
|
| 229 |
+
"loss": 3.7489,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.17,
|
| 234 |
+
"learning_rate": 3.9134164474436696e-05,
|
| 235 |
+
"loss": 3.7237,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 2.23,
|
| 240 |
+
"learning_rate": 3.8848221434290295e-05,
|
| 241 |
+
"loss": 3.6926,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.29,
|
| 246 |
+
"learning_rate": 3.8562278394143887e-05,
|
| 247 |
+
"loss": 3.6643,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 2.34,
|
| 252 |
+
"learning_rate": 3.8276335353997485e-05,
|
| 253 |
+
"loss": 3.6353,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 2.4,
|
| 258 |
+
"learning_rate": 3.7990392313851084e-05,
|
| 259 |
+
"loss": 3.6286,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 2.46,
|
| 264 |
+
"learning_rate": 3.7704449273704675e-05,
|
| 265 |
+
"loss": 3.6052,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 2.52,
|
| 270 |
+
"learning_rate": 3.7418506233558274e-05,
|
| 271 |
+
"loss": 3.5797,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 2.57,
|
| 276 |
+
"learning_rate": 3.713256319341187e-05,
|
| 277 |
+
"loss": 3.5624,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 2.63,
|
| 282 |
+
"learning_rate": 3.684662015326547e-05,
|
| 283 |
+
"loss": 3.5298,
|
| 284 |
+
"step": 23000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.69,
|
| 288 |
+
"learning_rate": 3.656067711311907e-05,
|
| 289 |
+
"loss": 3.5125,
|
| 290 |
+
"step": 23500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 2.75,
|
| 294 |
+
"learning_rate": 3.627473407297267e-05,
|
| 295 |
+
"loss": 3.4832,
|
| 296 |
+
"step": 24000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 2.8,
|
| 300 |
+
"learning_rate": 3.5988791032826266e-05,
|
| 301 |
+
"loss": 3.4755,
|
| 302 |
+
"step": 24500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 2.86,
|
| 306 |
+
"learning_rate": 3.570284799267986e-05,
|
| 307 |
+
"loss": 3.4514,
|
| 308 |
+
"step": 25000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 2.92,
|
| 312 |
+
"learning_rate": 3.5416904952533456e-05,
|
| 313 |
+
"loss": 3.441,
|
| 314 |
+
"step": 25500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 2.97,
|
| 318 |
+
"learning_rate": 3.5130961912387055e-05,
|
| 319 |
+
"loss": 3.4208,
|
| 320 |
+
"step": 26000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 3.03,
|
| 324 |
+
"learning_rate": 3.484501887224065e-05,
|
| 325 |
+
"loss": 3.3992,
|
| 326 |
+
"step": 26500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 3.09,
|
| 330 |
+
"learning_rate": 3.455907583209425e-05,
|
| 331 |
+
"loss": 3.3641,
|
| 332 |
+
"step": 27000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 3.15,
|
| 336 |
+
"learning_rate": 3.427313279194784e-05,
|
| 337 |
+
"loss": 3.3579,
|
| 338 |
+
"step": 27500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 3.2,
|
| 342 |
+
"learning_rate": 3.398718975180144e-05,
|
| 343 |
+
"loss": 3.3453,
|
| 344 |
+
"step": 28000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.26,
|
| 348 |
+
"learning_rate": 3.370124671165504e-05,
|
| 349 |
+
"loss": 3.3162,
|
| 350 |
+
"step": 28500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 3.32,
|
| 354 |
+
"learning_rate": 3.341530367150864e-05,
|
| 355 |
+
"loss": 3.2976,
|
| 356 |
+
"step": 29000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 3.37,
|
| 360 |
+
"learning_rate": 3.312936063136223e-05,
|
| 361 |
+
"loss": 3.2955,
|
| 362 |
+
"step": 29500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 3.43,
|
| 366 |
+
"learning_rate": 3.284341759121583e-05,
|
| 367 |
+
"loss": 3.2894,
|
| 368 |
+
"step": 30000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 3.49,
|
| 372 |
+
"learning_rate": 3.255747455106943e-05,
|
| 373 |
+
"loss": 3.2687,
|
| 374 |
+
"step": 30500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 3.55,
|
| 378 |
+
"learning_rate": 3.2271531510923026e-05,
|
| 379 |
+
"loss": 3.2548,
|
| 380 |
+
"step": 31000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 3.6,
|
| 384 |
+
"learning_rate": 3.1985588470776624e-05,
|
| 385 |
+
"loss": 3.2446,
|
| 386 |
+
"step": 31500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 3.66,
|
| 390 |
+
"learning_rate": 3.169964543063022e-05,
|
| 391 |
+
"loss": 3.2264,
|
| 392 |
+
"step": 32000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 3.72,
|
| 396 |
+
"learning_rate": 3.141370239048382e-05,
|
| 397 |
+
"loss": 3.2166,
|
| 398 |
+
"step": 32500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 3.77,
|
| 402 |
+
"learning_rate": 3.112775935033741e-05,
|
| 403 |
+
"loss": 3.2097,
|
| 404 |
+
"step": 33000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 3.83,
|
| 408 |
+
"learning_rate": 3.084181631019101e-05,
|
| 409 |
+
"loss": 3.1955,
|
| 410 |
+
"step": 33500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 3.89,
|
| 414 |
+
"learning_rate": 3.055587327004461e-05,
|
| 415 |
+
"loss": 3.181,
|
| 416 |
+
"step": 34000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 3.95,
|
| 420 |
+
"learning_rate": 3.0269930229898208e-05,
|
| 421 |
+
"loss": 3.1758,
|
| 422 |
+
"step": 34500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 4.0,
|
| 426 |
+
"learning_rate": 2.9983987189751807e-05,
|
| 427 |
+
"loss": 3.1688,
|
| 428 |
+
"step": 35000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 4.06,
|
| 432 |
+
"learning_rate": 2.9698044149605398e-05,
|
| 433 |
+
"loss": 3.1284,
|
| 434 |
+
"step": 35500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 4.12,
|
| 438 |
+
"learning_rate": 2.9412101109458993e-05,
|
| 439 |
+
"loss": 3.1256,
|
| 440 |
+
"step": 36000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 4.17,
|
| 444 |
+
"learning_rate": 2.9126158069312592e-05,
|
| 445 |
+
"loss": 3.1062,
|
| 446 |
+
"step": 36500
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 4.23,
|
| 450 |
+
"learning_rate": 2.884021502916619e-05,
|
| 451 |
+
"loss": 3.1051,
|
| 452 |
+
"step": 37000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 4.29,
|
| 456 |
+
"learning_rate": 2.855427198901979e-05,
|
| 457 |
+
"loss": 3.108,
|
| 458 |
+
"step": 37500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 4.35,
|
| 462 |
+
"learning_rate": 2.8268328948873384e-05,
|
| 463 |
+
"loss": 3.0932,
|
| 464 |
+
"step": 38000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 4.4,
|
| 468 |
+
"learning_rate": 2.7982385908726982e-05,
|
| 469 |
+
"loss": 3.0766,
|
| 470 |
+
"step": 38500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 4.46,
|
| 474 |
+
"learning_rate": 2.769644286858058e-05,
|
| 475 |
+
"loss": 3.0501,
|
| 476 |
+
"step": 39000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 4.52,
|
| 480 |
+
"learning_rate": 2.7410499828434176e-05,
|
| 481 |
+
"loss": 3.0634,
|
| 482 |
+
"step": 39500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 4.58,
|
| 486 |
+
"learning_rate": 2.7124556788287774e-05,
|
| 487 |
+
"loss": 3.0524,
|
| 488 |
+
"step": 40000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 4.63,
|
| 492 |
+
"learning_rate": 2.6838613748141373e-05,
|
| 493 |
+
"loss": 3.0556,
|
| 494 |
+
"step": 40500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 4.69,
|
| 498 |
+
"learning_rate": 2.655267070799497e-05,
|
| 499 |
+
"loss": 3.0276,
|
| 500 |
+
"step": 41000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 4.75,
|
| 504 |
+
"learning_rate": 2.6266727667848566e-05,
|
| 505 |
+
"loss": 3.0364,
|
| 506 |
+
"step": 41500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 4.8,
|
| 510 |
+
"learning_rate": 2.5980784627702165e-05,
|
| 511 |
+
"loss": 3.0186,
|
| 512 |
+
"step": 42000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 4.86,
|
| 516 |
+
"learning_rate": 2.5694841587555763e-05,
|
| 517 |
+
"loss": 3.0071,
|
| 518 |
+
"step": 42500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 4.92,
|
| 522 |
+
"learning_rate": 2.540889854740936e-05,
|
| 523 |
+
"loss": 3.0144,
|
| 524 |
+
"step": 43000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 4.98,
|
| 528 |
+
"learning_rate": 2.5122955507262953e-05,
|
| 529 |
+
"loss": 3.0003,
|
| 530 |
+
"step": 43500
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 5.03,
|
| 534 |
+
"learning_rate": 2.4837012467116552e-05,
|
| 535 |
+
"loss": 2.9854,
|
| 536 |
+
"step": 44000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 5.09,
|
| 540 |
+
"learning_rate": 2.455106942697015e-05,
|
| 541 |
+
"loss": 2.9696,
|
| 542 |
+
"step": 44500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 5.15,
|
| 546 |
+
"learning_rate": 2.426512638682375e-05,
|
| 547 |
+
"loss": 2.9651,
|
| 548 |
+
"step": 45000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 5.2,
|
| 552 |
+
"learning_rate": 2.397918334667734e-05,
|
| 553 |
+
"loss": 2.9463,
|
| 554 |
+
"step": 45500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 5.26,
|
| 558 |
+
"learning_rate": 2.369324030653094e-05,
|
| 559 |
+
"loss": 2.9379,
|
| 560 |
+
"step": 46000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 5.32,
|
| 564 |
+
"learning_rate": 2.3407297266384537e-05,
|
| 565 |
+
"loss": 2.9409,
|
| 566 |
+
"step": 46500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 5.38,
|
| 570 |
+
"learning_rate": 2.3121354226238136e-05,
|
| 571 |
+
"loss": 2.9315,
|
| 572 |
+
"step": 47000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 5.43,
|
| 576 |
+
"learning_rate": 2.283541118609173e-05,
|
| 577 |
+
"loss": 2.9407,
|
| 578 |
+
"step": 47500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 5.49,
|
| 582 |
+
"learning_rate": 2.254946814594533e-05,
|
| 583 |
+
"loss": 2.9327,
|
| 584 |
+
"step": 48000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 5.55,
|
| 588 |
+
"learning_rate": 2.2263525105798928e-05,
|
| 589 |
+
"loss": 2.9282,
|
| 590 |
+
"step": 48500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 5.6,
|
| 594 |
+
"learning_rate": 2.1977582065652523e-05,
|
| 595 |
+
"loss": 2.9177,
|
| 596 |
+
"step": 49000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 5.66,
|
| 600 |
+
"learning_rate": 2.1691639025506118e-05,
|
| 601 |
+
"loss": 2.9109,
|
| 602 |
+
"step": 49500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 5.72,
|
| 606 |
+
"learning_rate": 2.1405695985359716e-05,
|
| 607 |
+
"loss": 2.9009,
|
| 608 |
+
"step": 50000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 5.78,
|
| 612 |
+
"learning_rate": 2.1119752945213315e-05,
|
| 613 |
+
"loss": 2.8929,
|
| 614 |
+
"step": 50500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 5.83,
|
| 618 |
+
"learning_rate": 2.0833809905066913e-05,
|
| 619 |
+
"loss": 2.8863,
|
| 620 |
+
"step": 51000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 5.89,
|
| 624 |
+
"learning_rate": 2.0547866864920508e-05,
|
| 625 |
+
"loss": 2.8912,
|
| 626 |
+
"step": 51500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 5.95,
|
| 630 |
+
"learning_rate": 2.0261923824774107e-05,
|
| 631 |
+
"loss": 2.8757,
|
| 632 |
+
"step": 52000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 6.0,
|
| 636 |
+
"learning_rate": 1.9975980784627705e-05,
|
| 637 |
+
"loss": 2.874,
|
| 638 |
+
"step": 52500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 6.06,
|
| 642 |
+
"learning_rate": 1.96900377444813e-05,
|
| 643 |
+
"loss": 2.8468,
|
| 644 |
+
"step": 53000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 6.12,
|
| 648 |
+
"learning_rate": 1.9404094704334895e-05,
|
| 649 |
+
"loss": 2.866,
|
| 650 |
+
"step": 53500
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 6.18,
|
| 654 |
+
"learning_rate": 1.9118151664188494e-05,
|
| 655 |
+
"loss": 2.8428,
|
| 656 |
+
"step": 54000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 6.23,
|
| 660 |
+
"learning_rate": 1.8832208624042092e-05,
|
| 661 |
+
"loss": 2.8474,
|
| 662 |
+
"step": 54500
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 6.29,
|
| 666 |
+
"learning_rate": 1.8546265583895687e-05,
|
| 667 |
+
"loss": 2.8368,
|
| 668 |
+
"step": 55000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 6.35,
|
| 672 |
+
"learning_rate": 1.8260322543749286e-05,
|
| 673 |
+
"loss": 2.8377,
|
| 674 |
+
"step": 55500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 6.41,
|
| 678 |
+
"learning_rate": 1.7974379503602884e-05,
|
| 679 |
+
"loss": 2.838,
|
| 680 |
+
"step": 56000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 6.46,
|
| 684 |
+
"learning_rate": 1.7688436463456483e-05,
|
| 685 |
+
"loss": 2.83,
|
| 686 |
+
"step": 56500
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 6.52,
|
| 690 |
+
"learning_rate": 1.7402493423310078e-05,
|
| 691 |
+
"loss": 2.8275,
|
| 692 |
+
"step": 57000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 6.58,
|
| 696 |
+
"learning_rate": 1.7116550383163673e-05,
|
| 697 |
+
"loss": 2.8106,
|
| 698 |
+
"step": 57500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 6.63,
|
| 702 |
+
"learning_rate": 1.683060734301727e-05,
|
| 703 |
+
"loss": 2.8158,
|
| 704 |
+
"step": 58000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 6.69,
|
| 708 |
+
"learning_rate": 1.654466430287087e-05,
|
| 709 |
+
"loss": 2.814,
|
| 710 |
+
"step": 58500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 6.75,
|
| 714 |
+
"learning_rate": 1.6258721262724465e-05,
|
| 715 |
+
"loss": 2.8101,
|
| 716 |
+
"step": 59000
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 6.81,
|
| 720 |
+
"learning_rate": 1.5972778222578063e-05,
|
| 721 |
+
"loss": 2.8015,
|
| 722 |
+
"step": 59500
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 6.86,
|
| 726 |
+
"learning_rate": 1.5686835182431662e-05,
|
| 727 |
+
"loss": 2.7999,
|
| 728 |
+
"step": 60000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 6.92,
|
| 732 |
+
"learning_rate": 1.540089214228526e-05,
|
| 733 |
+
"loss": 2.7919,
|
| 734 |
+
"step": 60500
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 6.98,
|
| 738 |
+
"learning_rate": 1.5114949102138854e-05,
|
| 739 |
+
"loss": 2.7953,
|
| 740 |
+
"step": 61000
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 7.03,
|
| 744 |
+
"learning_rate": 1.482900606199245e-05,
|
| 745 |
+
"loss": 2.7795,
|
| 746 |
+
"step": 61500
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 7.09,
|
| 750 |
+
"learning_rate": 1.4543063021846049e-05,
|
| 751 |
+
"loss": 2.7648,
|
| 752 |
+
"step": 62000
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 7.15,
|
| 756 |
+
"learning_rate": 1.4257119981699646e-05,
|
| 757 |
+
"loss": 2.773,
|
| 758 |
+
"step": 62500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 7.21,
|
| 762 |
+
"learning_rate": 1.3971176941553244e-05,
|
| 763 |
+
"loss": 2.7663,
|
| 764 |
+
"step": 63000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 7.26,
|
| 768 |
+
"learning_rate": 1.368523390140684e-05,
|
| 769 |
+
"loss": 2.7592,
|
| 770 |
+
"step": 63500
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 7.32,
|
| 774 |
+
"learning_rate": 1.3399290861260438e-05,
|
| 775 |
+
"loss": 2.7543,
|
| 776 |
+
"step": 64000
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 7.38,
|
| 780 |
+
"learning_rate": 1.3113347821114036e-05,
|
| 781 |
+
"loss": 2.7626,
|
| 782 |
+
"step": 64500
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 7.43,
|
| 786 |
+
"learning_rate": 1.2827404780967631e-05,
|
| 787 |
+
"loss": 2.755,
|
| 788 |
+
"step": 65000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 7.49,
|
| 792 |
+
"learning_rate": 1.2541461740821228e-05,
|
| 793 |
+
"loss": 2.7538,
|
| 794 |
+
"step": 65500
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 7.55,
|
| 798 |
+
"learning_rate": 1.2255518700674826e-05,
|
| 799 |
+
"loss": 2.7381,
|
| 800 |
+
"step": 66000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 7.61,
|
| 804 |
+
"learning_rate": 1.1969575660528423e-05,
|
| 805 |
+
"loss": 2.7491,
|
| 806 |
+
"step": 66500
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 7.66,
|
| 810 |
+
"learning_rate": 1.168363262038202e-05,
|
| 811 |
+
"loss": 2.7441,
|
| 812 |
+
"step": 67000
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 7.72,
|
| 816 |
+
"learning_rate": 1.1397689580235618e-05,
|
| 817 |
+
"loss": 2.7369,
|
| 818 |
+
"step": 67500
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 7.78,
|
| 822 |
+
"learning_rate": 1.1111746540089215e-05,
|
| 823 |
+
"loss": 2.7466,
|
| 824 |
+
"step": 68000
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 7.83,
|
| 828 |
+
"learning_rate": 1.0825803499942812e-05,
|
| 829 |
+
"loss": 2.7363,
|
| 830 |
+
"step": 68500
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 7.89,
|
| 834 |
+
"learning_rate": 1.0539860459796409e-05,
|
| 835 |
+
"loss": 2.7294,
|
| 836 |
+
"step": 69000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 7.95,
|
| 840 |
+
"learning_rate": 1.0253917419650007e-05,
|
| 841 |
+
"loss": 2.7304,
|
| 842 |
+
"step": 69500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 8.01,
|
| 846 |
+
"learning_rate": 9.967974379503602e-06,
|
| 847 |
+
"loss": 2.7278,
|
| 848 |
+
"step": 70000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 8.06,
|
| 852 |
+
"learning_rate": 9.6820313393572e-06,
|
| 853 |
+
"loss": 2.7231,
|
| 854 |
+
"step": 70500
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 8.12,
|
| 858 |
+
"learning_rate": 9.396088299210797e-06,
|
| 859 |
+
"loss": 2.7267,
|
| 860 |
+
"step": 71000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 8.18,
|
| 864 |
+
"learning_rate": 9.110145259064396e-06,
|
| 865 |
+
"loss": 2.7133,
|
| 866 |
+
"step": 71500
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 8.24,
|
| 870 |
+
"learning_rate": 8.824202218917991e-06,
|
| 871 |
+
"loss": 2.7088,
|
| 872 |
+
"step": 72000
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 8.29,
|
| 876 |
+
"learning_rate": 8.53825917877159e-06,
|
| 877 |
+
"loss": 2.7115,
|
| 878 |
+
"step": 72500
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 8.35,
|
| 882 |
+
"learning_rate": 8.252316138625186e-06,
|
| 883 |
+
"loss": 2.7049,
|
| 884 |
+
"step": 73000
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 8.41,
|
| 888 |
+
"learning_rate": 7.966373098478785e-06,
|
| 889 |
+
"loss": 2.7083,
|
| 890 |
+
"step": 73500
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 8.46,
|
| 894 |
+
"learning_rate": 7.68043005833238e-06,
|
| 895 |
+
"loss": 2.6946,
|
| 896 |
+
"step": 74000
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 8.52,
|
| 900 |
+
"learning_rate": 7.394487018185977e-06,
|
| 901 |
+
"loss": 2.691,
|
| 902 |
+
"step": 74500
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 8.58,
|
| 906 |
+
"learning_rate": 7.108543978039575e-06,
|
| 907 |
+
"loss": 2.7066,
|
| 908 |
+
"step": 75000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 8.64,
|
| 912 |
+
"learning_rate": 6.8226009378931726e-06,
|
| 913 |
+
"loss": 2.6896,
|
| 914 |
+
"step": 75500
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 8.69,
|
| 918 |
+
"learning_rate": 6.5366578977467685e-06,
|
| 919 |
+
"loss": 2.6976,
|
| 920 |
+
"step": 76000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 8.75,
|
| 924 |
+
"learning_rate": 6.250714857600366e-06,
|
| 925 |
+
"loss": 2.6989,
|
| 926 |
+
"step": 76500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 8.81,
|
| 930 |
+
"learning_rate": 5.964771817453964e-06,
|
| 931 |
+
"loss": 2.6869,
|
| 932 |
+
"step": 77000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 8.86,
|
| 936 |
+
"learning_rate": 5.6788287773075605e-06,
|
| 937 |
+
"loss": 2.6898,
|
| 938 |
+
"step": 77500
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 8.92,
|
| 942 |
+
"learning_rate": 5.392885737161157e-06,
|
| 943 |
+
"loss": 2.6851,
|
| 944 |
+
"step": 78000
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 8.98,
|
| 948 |
+
"learning_rate": 5.106942697014755e-06,
|
| 949 |
+
"loss": 2.6871,
|
| 950 |
+
"step": 78500
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 9.04,
|
| 954 |
+
"learning_rate": 4.820999656868352e-06,
|
| 955 |
+
"loss": 2.6719,
|
| 956 |
+
"step": 79000
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 9.09,
|
| 960 |
+
"learning_rate": 4.535056616721949e-06,
|
| 961 |
+
"loss": 2.6704,
|
| 962 |
+
"step": 79500
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 9.15,
|
| 966 |
+
"learning_rate": 4.249113576575546e-06,
|
| 967 |
+
"loss": 2.6776,
|
| 968 |
+
"step": 80000
|
| 969 |
+
}
|
| 970 |
+
],
|
| 971 |
+
"max_steps": 87430,
|
| 972 |
+
"num_train_epochs": 10,
|
| 973 |
+
"total_flos": 9.648556042711978e+17,
|
| 974 |
+
"trial_name": null,
|
| 975 |
+
"trial_params": null
|
| 976 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:db390b123ce3ded507bae971b2e3591ba48f74410acab774be268a32083b121f
|
| 3 |
+
size 2479
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|