Commit
•
5588683
1
Parent(s):
751deaa
First model version
Browse files- config.json +43 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +958 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "hivemind/gpt-j-6B-8bit",
|
| 3 |
+
"activation_function": "gelu_new",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"GPTJForCausalLM"
|
| 6 |
+
],
|
| 7 |
+
"attn_pdrop": 0.0,
|
| 8 |
+
"bos_token_id": 50256,
|
| 9 |
+
"eight_bit": true,
|
| 10 |
+
"embd_pdrop": 0.0,
|
| 11 |
+
"eos_token_id": 50256,
|
| 12 |
+
"gradient_checkpointing": false,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"layer_norm_epsilon": 1e-05,
|
| 15 |
+
"model_type": "gptj",
|
| 16 |
+
"n_embd": 4096,
|
| 17 |
+
"n_head": 16,
|
| 18 |
+
"n_inner": null,
|
| 19 |
+
"n_layer": 28,
|
| 20 |
+
"n_positions": 2048,
|
| 21 |
+
"resid_pdrop": 0.0,
|
| 22 |
+
"rotary": true,
|
| 23 |
+
"rotary_dim": 64,
|
| 24 |
+
"scale_attn_weights": true,
|
| 25 |
+
"summary_activation": null,
|
| 26 |
+
"summary_first_dropout": 0.1,
|
| 27 |
+
"summary_proj_to_labels": true,
|
| 28 |
+
"summary_type": "cls_index",
|
| 29 |
+
"summary_use_proj": true,
|
| 30 |
+
"task_specific_params": {
|
| 31 |
+
"text-generation": {
|
| 32 |
+
"do_sample": true,
|
| 33 |
+
"max_length": 50,
|
| 34 |
+
"temperature": 1.0
|
| 35 |
+
}
|
| 36 |
+
},
|
| 37 |
+
"tie_word_embeddings": false,
|
| 38 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 39 |
+
"torch_dtype": "float32",
|
| 40 |
+
"transformers_version": "4.19.2",
|
| 41 |
+
"use_cache": true,
|
| 42 |
+
"vocab_size": 50400
|
| 43 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:386800231a543bc12bb76866e8ba099599330560e42cb21eb2e62055ebbed64a
|
| 3 |
+
size 285351013
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bb88da72c050fe8130117594ac5340a13d930afb8c9a844fe688ae500668d814
|
| 3 |
+
size 6316424352
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ff0b7d688f72cecc3d6db8a5558dfafab2669c75076ea7f5e8d0318cba901963
|
| 3 |
+
size 14503
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:51fc000fe6c1805ee5d1a366836965daacd3c6792e46426bf653eb71aa8e61a5
|
| 3 |
+
size 623
|
trainer_state.json
ADDED
|
@@ -0,0 +1,958 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"global_step": 157480,
|
| 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": 4.97140678612276e-05,
|
| 13 |
+
"loss": 2.3365,
|
| 14 |
+
"step": 1000
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.13,
|
| 18 |
+
"learning_rate": 4.939636548481383e-05,
|
| 19 |
+
"loss": 1.9007,
|
| 20 |
+
"step": 2000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.19,
|
| 24 |
+
"learning_rate": 4.907866310840005e-05,
|
| 25 |
+
"loss": 1.8121,
|
| 26 |
+
"step": 3000
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.25,
|
| 30 |
+
"learning_rate": 4.8760960731986276e-05,
|
| 31 |
+
"loss": 1.7655,
|
| 32 |
+
"step": 4000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.32,
|
| 36 |
+
"learning_rate": 4.84432583555725e-05,
|
| 37 |
+
"loss": 1.7269,
|
| 38 |
+
"step": 5000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.38,
|
| 42 |
+
"learning_rate": 4.8125555979158726e-05,
|
| 43 |
+
"loss": 1.6977,
|
| 44 |
+
"step": 6000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.44,
|
| 48 |
+
"learning_rate": 4.780785360274495e-05,
|
| 49 |
+
"loss": 1.678,
|
| 50 |
+
"step": 7000
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.51,
|
| 54 |
+
"learning_rate": 4.7490151226331175e-05,
|
| 55 |
+
"loss": 1.6612,
|
| 56 |
+
"step": 8000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.57,
|
| 60 |
+
"learning_rate": 4.71724488499174e-05,
|
| 61 |
+
"loss": 1.6365,
|
| 62 |
+
"step": 9000
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.64,
|
| 66 |
+
"learning_rate": 4.6854746473503625e-05,
|
| 67 |
+
"loss": 1.6223,
|
| 68 |
+
"step": 10000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.7,
|
| 72 |
+
"learning_rate": 4.653704409708985e-05,
|
| 73 |
+
"loss": 1.6127,
|
| 74 |
+
"step": 11000
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.76,
|
| 78 |
+
"learning_rate": 4.6219341720676074e-05,
|
| 79 |
+
"loss": 1.5877,
|
| 80 |
+
"step": 12000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.83,
|
| 84 |
+
"learning_rate": 4.59016393442623e-05,
|
| 85 |
+
"loss": 1.5809,
|
| 86 |
+
"step": 13000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.89,
|
| 90 |
+
"learning_rate": 4.5583936967848524e-05,
|
| 91 |
+
"loss": 1.5737,
|
| 92 |
+
"step": 14000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.95,
|
| 96 |
+
"learning_rate": 4.526623459143475e-05,
|
| 97 |
+
"loss": 1.5611,
|
| 98 |
+
"step": 15000
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.02,
|
| 102 |
+
"learning_rate": 4.494853221502097e-05,
|
| 103 |
+
"loss": 1.5297,
|
| 104 |
+
"step": 16000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 1.08,
|
| 108 |
+
"learning_rate": 4.46308298386072e-05,
|
| 109 |
+
"loss": 1.4783,
|
| 110 |
+
"step": 17000
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 1.14,
|
| 114 |
+
"learning_rate": 4.431312746219342e-05,
|
| 115 |
+
"loss": 1.472,
|
| 116 |
+
"step": 18000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.21,
|
| 120 |
+
"learning_rate": 4.399542508577965e-05,
|
| 121 |
+
"loss": 1.4821,
|
| 122 |
+
"step": 19000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.27,
|
| 126 |
+
"learning_rate": 4.367772270936587e-05,
|
| 127 |
+
"loss": 1.4681,
|
| 128 |
+
"step": 20000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.33,
|
| 132 |
+
"learning_rate": 4.33600203329521e-05,
|
| 133 |
+
"loss": 1.4564,
|
| 134 |
+
"step": 21000
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 1.4,
|
| 138 |
+
"learning_rate": 4.304231795653832e-05,
|
| 139 |
+
"loss": 1.4509,
|
| 140 |
+
"step": 22000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.46,
|
| 144 |
+
"learning_rate": 4.2724615580124546e-05,
|
| 145 |
+
"loss": 1.4478,
|
| 146 |
+
"step": 23000
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 1.52,
|
| 150 |
+
"learning_rate": 4.2406913203710764e-05,
|
| 151 |
+
"loss": 1.4446,
|
| 152 |
+
"step": 24000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.59,
|
| 156 |
+
"learning_rate": 4.208921082729699e-05,
|
| 157 |
+
"loss": 1.439,
|
| 158 |
+
"step": 25000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.65,
|
| 162 |
+
"learning_rate": 4.1771508450883214e-05,
|
| 163 |
+
"loss": 1.4367,
|
| 164 |
+
"step": 26000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.71,
|
| 168 |
+
"learning_rate": 4.145380607446944e-05,
|
| 169 |
+
"loss": 1.4181,
|
| 170 |
+
"step": 27000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.78,
|
| 174 |
+
"learning_rate": 4.113610369805566e-05,
|
| 175 |
+
"loss": 1.417,
|
| 176 |
+
"step": 28000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.84,
|
| 180 |
+
"learning_rate": 4.081840132164189e-05,
|
| 181 |
+
"loss": 1.4056,
|
| 182 |
+
"step": 29000
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.91,
|
| 186 |
+
"learning_rate": 4.050069894522811e-05,
|
| 187 |
+
"loss": 1.4006,
|
| 188 |
+
"step": 30000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 1.97,
|
| 192 |
+
"learning_rate": 4.018299656881434e-05,
|
| 193 |
+
"loss": 1.4053,
|
| 194 |
+
"step": 31000
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 2.03,
|
| 198 |
+
"learning_rate": 3.986529419240056e-05,
|
| 199 |
+
"loss": 1.3491,
|
| 200 |
+
"step": 32000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 2.1,
|
| 204 |
+
"learning_rate": 3.954759181598679e-05,
|
| 205 |
+
"loss": 1.3028,
|
| 206 |
+
"step": 33000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 2.16,
|
| 210 |
+
"learning_rate": 3.922988943957301e-05,
|
| 211 |
+
"loss": 1.3012,
|
| 212 |
+
"step": 34000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 2.22,
|
| 216 |
+
"learning_rate": 3.8912187063159236e-05,
|
| 217 |
+
"loss": 1.2896,
|
| 218 |
+
"step": 35000
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 2.29,
|
| 222 |
+
"learning_rate": 3.859448468674546e-05,
|
| 223 |
+
"loss": 1.2955,
|
| 224 |
+
"step": 36000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 2.35,
|
| 228 |
+
"learning_rate": 3.8276782310331686e-05,
|
| 229 |
+
"loss": 1.2843,
|
| 230 |
+
"step": 37000
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.41,
|
| 234 |
+
"learning_rate": 3.7959079933917904e-05,
|
| 235 |
+
"loss": 1.2876,
|
| 236 |
+
"step": 38000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 2.48,
|
| 240 |
+
"learning_rate": 3.764137755750413e-05,
|
| 241 |
+
"loss": 1.2826,
|
| 242 |
+
"step": 39000
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.54,
|
| 246 |
+
"learning_rate": 3.732367518109035e-05,
|
| 247 |
+
"loss": 1.2706,
|
| 248 |
+
"step": 40000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 2.6,
|
| 252 |
+
"learning_rate": 3.700597280467658e-05,
|
| 253 |
+
"loss": 1.2743,
|
| 254 |
+
"step": 41000
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 2.67,
|
| 258 |
+
"learning_rate": 3.66882704282628e-05,
|
| 259 |
+
"loss": 1.2783,
|
| 260 |
+
"step": 42000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 2.73,
|
| 264 |
+
"learning_rate": 3.637056805184903e-05,
|
| 265 |
+
"loss": 1.264,
|
| 266 |
+
"step": 43000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 2.79,
|
| 270 |
+
"learning_rate": 3.605286567543525e-05,
|
| 271 |
+
"loss": 1.2614,
|
| 272 |
+
"step": 44000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 2.86,
|
| 276 |
+
"learning_rate": 3.573516329902148e-05,
|
| 277 |
+
"loss": 1.2549,
|
| 278 |
+
"step": 45000
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 2.92,
|
| 282 |
+
"learning_rate": 3.54174609226077e-05,
|
| 283 |
+
"loss": 1.2453,
|
| 284 |
+
"step": 46000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.98,
|
| 288 |
+
"learning_rate": 3.5099758546193926e-05,
|
| 289 |
+
"loss": 1.2425,
|
| 290 |
+
"step": 47000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 3.05,
|
| 294 |
+
"learning_rate": 3.478205616978015e-05,
|
| 295 |
+
"loss": 1.1538,
|
| 296 |
+
"step": 48000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 3.11,
|
| 300 |
+
"learning_rate": 3.4464353793366376e-05,
|
| 301 |
+
"loss": 1.1307,
|
| 302 |
+
"step": 49000
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 3.18,
|
| 306 |
+
"learning_rate": 3.41466514169526e-05,
|
| 307 |
+
"loss": 1.1375,
|
| 308 |
+
"step": 50000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 3.24,
|
| 312 |
+
"learning_rate": 3.3828949040538825e-05,
|
| 313 |
+
"loss": 1.1258,
|
| 314 |
+
"step": 51000
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 3.3,
|
| 318 |
+
"learning_rate": 3.351124666412505e-05,
|
| 319 |
+
"loss": 1.1237,
|
| 320 |
+
"step": 52000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 3.37,
|
| 324 |
+
"learning_rate": 3.3193544287711275e-05,
|
| 325 |
+
"loss": 1.1226,
|
| 326 |
+
"step": 53000
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 3.43,
|
| 330 |
+
"learning_rate": 3.28758419112975e-05,
|
| 331 |
+
"loss": 1.1246,
|
| 332 |
+
"step": 54000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 3.49,
|
| 336 |
+
"learning_rate": 3.2558139534883724e-05,
|
| 337 |
+
"loss": 1.1197,
|
| 338 |
+
"step": 55000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 3.56,
|
| 342 |
+
"learning_rate": 3.224043715846995e-05,
|
| 343 |
+
"loss": 1.117,
|
| 344 |
+
"step": 56000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.62,
|
| 348 |
+
"learning_rate": 3.1922734782056174e-05,
|
| 349 |
+
"loss": 1.1206,
|
| 350 |
+
"step": 57000
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 3.68,
|
| 354 |
+
"learning_rate": 3.160503240564239e-05,
|
| 355 |
+
"loss": 1.1023,
|
| 356 |
+
"step": 58000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 3.75,
|
| 360 |
+
"learning_rate": 3.1287330029228617e-05,
|
| 361 |
+
"loss": 1.1076,
|
| 362 |
+
"step": 59000
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 3.81,
|
| 366 |
+
"learning_rate": 3.096962765281484e-05,
|
| 367 |
+
"loss": 1.1061,
|
| 368 |
+
"step": 60000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 3.87,
|
| 372 |
+
"learning_rate": 3.0651925276401066e-05,
|
| 373 |
+
"loss": 1.1017,
|
| 374 |
+
"step": 61000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 3.94,
|
| 378 |
+
"learning_rate": 3.033422289998729e-05,
|
| 379 |
+
"loss": 1.1005,
|
| 380 |
+
"step": 62000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 4.0,
|
| 384 |
+
"learning_rate": 3.0016520523573516e-05,
|
| 385 |
+
"loss": 1.0891,
|
| 386 |
+
"step": 63000
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 4.06,
|
| 390 |
+
"learning_rate": 2.969881814715974e-05,
|
| 391 |
+
"loss": 0.9693,
|
| 392 |
+
"step": 64000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 4.13,
|
| 396 |
+
"learning_rate": 2.9381115770745965e-05,
|
| 397 |
+
"loss": 0.9703,
|
| 398 |
+
"step": 65000
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 4.19,
|
| 402 |
+
"learning_rate": 2.906341339433219e-05,
|
| 403 |
+
"loss": 0.9693,
|
| 404 |
+
"step": 66000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 4.25,
|
| 408 |
+
"learning_rate": 2.8745711017918414e-05,
|
| 409 |
+
"loss": 0.9698,
|
| 410 |
+
"step": 67000
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 4.32,
|
| 414 |
+
"learning_rate": 2.842800864150464e-05,
|
| 415 |
+
"loss": 0.9715,
|
| 416 |
+
"step": 68000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 4.38,
|
| 420 |
+
"learning_rate": 2.8110306265090864e-05,
|
| 421 |
+
"loss": 0.9734,
|
| 422 |
+
"step": 69000
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 4.45,
|
| 426 |
+
"learning_rate": 2.779260388867709e-05,
|
| 427 |
+
"loss": 0.9796,
|
| 428 |
+
"step": 70000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 4.51,
|
| 432 |
+
"learning_rate": 2.7474901512263313e-05,
|
| 433 |
+
"loss": 0.9729,
|
| 434 |
+
"step": 71000
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 4.57,
|
| 438 |
+
"learning_rate": 2.7157199135849538e-05,
|
| 439 |
+
"loss": 0.9727,
|
| 440 |
+
"step": 72000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 4.64,
|
| 444 |
+
"learning_rate": 2.6839496759435763e-05,
|
| 445 |
+
"loss": 0.9721,
|
| 446 |
+
"step": 73000
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 4.7,
|
| 450 |
+
"learning_rate": 2.6521794383021988e-05,
|
| 451 |
+
"loss": 0.9643,
|
| 452 |
+
"step": 74000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 4.76,
|
| 456 |
+
"learning_rate": 2.6204092006608212e-05,
|
| 457 |
+
"loss": 0.9599,
|
| 458 |
+
"step": 75000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 4.83,
|
| 462 |
+
"learning_rate": 2.5886389630194434e-05,
|
| 463 |
+
"loss": 0.9651,
|
| 464 |
+
"step": 76000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 4.89,
|
| 468 |
+
"learning_rate": 2.556868725378066e-05,
|
| 469 |
+
"loss": 0.9587,
|
| 470 |
+
"step": 77000
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 4.95,
|
| 474 |
+
"learning_rate": 2.5250984877366883e-05,
|
| 475 |
+
"loss": 0.9568,
|
| 476 |
+
"step": 78000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 5.02,
|
| 480 |
+
"learning_rate": 2.4933282500953108e-05,
|
| 481 |
+
"loss": 0.9187,
|
| 482 |
+
"step": 79000
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 5.08,
|
| 486 |
+
"learning_rate": 2.4615580124539333e-05,
|
| 487 |
+
"loss": 0.8357,
|
| 488 |
+
"step": 80000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 5.14,
|
| 492 |
+
"learning_rate": 2.4297877748125557e-05,
|
| 493 |
+
"loss": 0.8324,
|
| 494 |
+
"step": 81000
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 5.21,
|
| 498 |
+
"learning_rate": 2.3980175371711782e-05,
|
| 499 |
+
"loss": 0.84,
|
| 500 |
+
"step": 82000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 5.27,
|
| 504 |
+
"learning_rate": 2.3662472995298007e-05,
|
| 505 |
+
"loss": 0.8392,
|
| 506 |
+
"step": 83000
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 5.33,
|
| 510 |
+
"learning_rate": 2.3344770618884228e-05,
|
| 511 |
+
"loss": 0.8439,
|
| 512 |
+
"step": 84000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 5.4,
|
| 516 |
+
"learning_rate": 2.3027068242470453e-05,
|
| 517 |
+
"loss": 0.8375,
|
| 518 |
+
"step": 85000
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 5.46,
|
| 522 |
+
"learning_rate": 2.2709365866056678e-05,
|
| 523 |
+
"loss": 0.8363,
|
| 524 |
+
"step": 86000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 5.52,
|
| 528 |
+
"learning_rate": 2.2391663489642903e-05,
|
| 529 |
+
"loss": 0.8387,
|
| 530 |
+
"step": 87000
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 5.59,
|
| 534 |
+
"learning_rate": 2.2073961113229127e-05,
|
| 535 |
+
"loss": 0.8458,
|
| 536 |
+
"step": 88000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 5.65,
|
| 540 |
+
"learning_rate": 2.1756258736815352e-05,
|
| 541 |
+
"loss": 0.84,
|
| 542 |
+
"step": 89000
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 5.72,
|
| 546 |
+
"learning_rate": 2.1438556360401577e-05,
|
| 547 |
+
"loss": 0.833,
|
| 548 |
+
"step": 90000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 5.78,
|
| 552 |
+
"learning_rate": 2.11208539839878e-05,
|
| 553 |
+
"loss": 0.8356,
|
| 554 |
+
"step": 91000
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 5.84,
|
| 558 |
+
"learning_rate": 2.0803151607574026e-05,
|
| 559 |
+
"loss": 0.8367,
|
| 560 |
+
"step": 92000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 5.91,
|
| 564 |
+
"learning_rate": 2.048544923116025e-05,
|
| 565 |
+
"loss": 0.8323,
|
| 566 |
+
"step": 93000
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 5.97,
|
| 570 |
+
"learning_rate": 2.0167746854746476e-05,
|
| 571 |
+
"loss": 0.8304,
|
| 572 |
+
"step": 94000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 6.03,
|
| 576 |
+
"learning_rate": 1.98500444783327e-05,
|
| 577 |
+
"loss": 0.7779,
|
| 578 |
+
"step": 95000
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 6.1,
|
| 582 |
+
"learning_rate": 1.9532342101918925e-05,
|
| 583 |
+
"loss": 0.7192,
|
| 584 |
+
"step": 96000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 6.16,
|
| 588 |
+
"learning_rate": 1.921463972550515e-05,
|
| 589 |
+
"loss": 0.7189,
|
| 590 |
+
"step": 97000
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 6.22,
|
| 594 |
+
"learning_rate": 1.8896937349091375e-05,
|
| 595 |
+
"loss": 0.7193,
|
| 596 |
+
"step": 98000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 6.29,
|
| 600 |
+
"learning_rate": 1.85792349726776e-05,
|
| 601 |
+
"loss": 0.7281,
|
| 602 |
+
"step": 99000
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 6.35,
|
| 606 |
+
"learning_rate": 1.826153259626382e-05,
|
| 607 |
+
"loss": 0.7303,
|
| 608 |
+
"step": 100000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 6.41,
|
| 612 |
+
"learning_rate": 1.7943830219850045e-05,
|
| 613 |
+
"loss": 0.7308,
|
| 614 |
+
"step": 101000
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 6.48,
|
| 618 |
+
"learning_rate": 1.762612784343627e-05,
|
| 619 |
+
"loss": 0.7285,
|
| 620 |
+
"step": 102000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 6.54,
|
| 624 |
+
"learning_rate": 1.730842546702249e-05,
|
| 625 |
+
"loss": 0.7279,
|
| 626 |
+
"step": 103000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 6.6,
|
| 630 |
+
"learning_rate": 1.6990723090608716e-05,
|
| 631 |
+
"loss": 0.7323,
|
| 632 |
+
"step": 104000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 6.67,
|
| 636 |
+
"learning_rate": 1.667302071419494e-05,
|
| 637 |
+
"loss": 0.7309,
|
| 638 |
+
"step": 105000
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 6.73,
|
| 642 |
+
"learning_rate": 1.6355318337781166e-05,
|
| 643 |
+
"loss": 0.729,
|
| 644 |
+
"step": 106000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 6.79,
|
| 648 |
+
"learning_rate": 1.603761596136739e-05,
|
| 649 |
+
"loss": 0.7285,
|
| 650 |
+
"step": 107000
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 6.86,
|
| 654 |
+
"learning_rate": 1.5719913584953615e-05,
|
| 655 |
+
"loss": 0.7269,
|
| 656 |
+
"step": 108000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 6.92,
|
| 660 |
+
"learning_rate": 1.540221120853984e-05,
|
| 661 |
+
"loss": 0.7236,
|
| 662 |
+
"step": 109000
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 6.99,
|
| 666 |
+
"learning_rate": 1.5084508832126065e-05,
|
| 667 |
+
"loss": 0.7238,
|
| 668 |
+
"step": 110000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 7.05,
|
| 672 |
+
"learning_rate": 1.476680645571229e-05,
|
| 673 |
+
"loss": 0.6498,
|
| 674 |
+
"step": 111000
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 7.11,
|
| 678 |
+
"learning_rate": 1.4449104079298514e-05,
|
| 679 |
+
"loss": 0.6289,
|
| 680 |
+
"step": 112000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 7.18,
|
| 684 |
+
"learning_rate": 1.4131401702884739e-05,
|
| 685 |
+
"loss": 0.629,
|
| 686 |
+
"step": 113000
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 7.24,
|
| 690 |
+
"learning_rate": 1.3813699326470964e-05,
|
| 691 |
+
"loss": 0.6331,
|
| 692 |
+
"step": 114000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 7.3,
|
| 696 |
+
"learning_rate": 1.3495996950057188e-05,
|
| 697 |
+
"loss": 0.6354,
|
| 698 |
+
"step": 115000
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 7.37,
|
| 702 |
+
"learning_rate": 1.3178294573643413e-05,
|
| 703 |
+
"loss": 0.634,
|
| 704 |
+
"step": 116000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 7.43,
|
| 708 |
+
"learning_rate": 1.2860592197229638e-05,
|
| 709 |
+
"loss": 0.636,
|
| 710 |
+
"step": 117000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 7.49,
|
| 714 |
+
"learning_rate": 1.254288982081586e-05,
|
| 715 |
+
"loss": 0.6422,
|
| 716 |
+
"step": 118000
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 7.56,
|
| 720 |
+
"learning_rate": 1.2225187444402086e-05,
|
| 721 |
+
"loss": 0.6371,
|
| 722 |
+
"step": 119000
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 7.62,
|
| 726 |
+
"learning_rate": 1.1907485067988309e-05,
|
| 727 |
+
"loss": 0.6383,
|
| 728 |
+
"step": 120000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 7.68,
|
| 732 |
+
"learning_rate": 1.1589782691574534e-05,
|
| 733 |
+
"loss": 0.6401,
|
| 734 |
+
"step": 121000
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 7.75,
|
| 738 |
+
"learning_rate": 1.1272080315160757e-05,
|
| 739 |
+
"loss": 0.6332,
|
| 740 |
+
"step": 122000
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 7.81,
|
| 744 |
+
"learning_rate": 1.0954377938746981e-05,
|
| 745 |
+
"loss": 0.6408,
|
| 746 |
+
"step": 123000
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 7.87,
|
| 750 |
+
"learning_rate": 1.0636675562333206e-05,
|
| 751 |
+
"loss": 0.6337,
|
| 752 |
+
"step": 124000
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 7.94,
|
| 756 |
+
"learning_rate": 1.031897318591943e-05,
|
| 757 |
+
"loss": 0.6406,
|
| 758 |
+
"step": 125000
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 8.0,
|
| 762 |
+
"learning_rate": 1.0001270809505656e-05,
|
| 763 |
+
"loss": 0.6372,
|
| 764 |
+
"step": 126000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 8.06,
|
| 768 |
+
"learning_rate": 9.68356843309188e-06,
|
| 769 |
+
"loss": 0.5542,
|
| 770 |
+
"step": 127000
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 8.13,
|
| 774 |
+
"learning_rate": 9.365866056678105e-06,
|
| 775 |
+
"loss": 0.5541,
|
| 776 |
+
"step": 128000
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 8.19,
|
| 780 |
+
"learning_rate": 9.04816368026433e-06,
|
| 781 |
+
"loss": 0.565,
|
| 782 |
+
"step": 129000
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 8.26,
|
| 786 |
+
"learning_rate": 8.730461303850553e-06,
|
| 787 |
+
"loss": 0.5644,
|
| 788 |
+
"step": 130000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 8.32,
|
| 792 |
+
"learning_rate": 8.412758927436778e-06,
|
| 793 |
+
"loss": 0.5651,
|
| 794 |
+
"step": 131000
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 8.38,
|
| 798 |
+
"learning_rate": 8.095056551023002e-06,
|
| 799 |
+
"loss": 0.5661,
|
| 800 |
+
"step": 132000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 8.45,
|
| 804 |
+
"learning_rate": 7.777354174609227e-06,
|
| 805 |
+
"loss": 0.5704,
|
| 806 |
+
"step": 133000
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 8.51,
|
| 810 |
+
"learning_rate": 7.459651798195451e-06,
|
| 811 |
+
"loss": 0.5669,
|
| 812 |
+
"step": 134000
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 8.57,
|
| 816 |
+
"learning_rate": 7.141949421781676e-06,
|
| 817 |
+
"loss": 0.5694,
|
| 818 |
+
"step": 135000
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 8.64,
|
| 822 |
+
"learning_rate": 6.8242470453679e-06,
|
| 823 |
+
"loss": 0.5705,
|
| 824 |
+
"step": 136000
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 8.7,
|
| 828 |
+
"learning_rate": 6.506544668954124e-06,
|
| 829 |
+
"loss": 0.5682,
|
| 830 |
+
"step": 137000
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 8.76,
|
| 834 |
+
"learning_rate": 6.188842292540348e-06,
|
| 835 |
+
"loss": 0.5676,
|
| 836 |
+
"step": 138000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 8.83,
|
| 840 |
+
"learning_rate": 5.871139916126573e-06,
|
| 841 |
+
"loss": 0.5674,
|
| 842 |
+
"step": 139000
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 8.89,
|
| 846 |
+
"learning_rate": 5.553437539712798e-06,
|
| 847 |
+
"loss": 0.5681,
|
| 848 |
+
"step": 140000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 8.95,
|
| 852 |
+
"learning_rate": 5.2357351632990216e-06,
|
| 853 |
+
"loss": 0.5636,
|
| 854 |
+
"step": 141000
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 9.02,
|
| 858 |
+
"learning_rate": 4.918032786885246e-06,
|
| 859 |
+
"loss": 0.5495,
|
| 860 |
+
"step": 142000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 9.08,
|
| 864 |
+
"learning_rate": 4.60033041047147e-06,
|
| 865 |
+
"loss": 0.5124,
|
| 866 |
+
"step": 143000
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 9.14,
|
| 870 |
+
"learning_rate": 4.282628034057695e-06,
|
| 871 |
+
"loss": 0.514,
|
| 872 |
+
"step": 144000
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 9.21,
|
| 876 |
+
"learning_rate": 3.964925657643919e-06,
|
| 877 |
+
"loss": 0.517,
|
| 878 |
+
"step": 145000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 9.27,
|
| 882 |
+
"learning_rate": 3.6472232812301436e-06,
|
| 883 |
+
"loss": 0.5173,
|
| 884 |
+
"step": 146000
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 9.33,
|
| 888 |
+
"learning_rate": 3.3295209048163683e-06,
|
| 889 |
+
"loss": 0.5164,
|
| 890 |
+
"step": 147000
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 9.4,
|
| 894 |
+
"learning_rate": 3.0118185284025926e-06,
|
| 895 |
+
"loss": 0.5202,
|
| 896 |
+
"step": 148000
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 9.46,
|
| 900 |
+
"learning_rate": 2.694116151988817e-06,
|
| 901 |
+
"loss": 0.5183,
|
| 902 |
+
"step": 149000
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 9.53,
|
| 906 |
+
"learning_rate": 2.3764137755750413e-06,
|
| 907 |
+
"loss": 0.517,
|
| 908 |
+
"step": 150000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 9.59,
|
| 912 |
+
"learning_rate": 2.058711399161266e-06,
|
| 913 |
+
"loss": 0.5131,
|
| 914 |
+
"step": 151000
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 9.65,
|
| 918 |
+
"learning_rate": 1.7410090227474903e-06,
|
| 919 |
+
"loss": 0.516,
|
| 920 |
+
"step": 152000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 9.72,
|
| 924 |
+
"learning_rate": 1.4233066463337146e-06,
|
| 925 |
+
"loss": 0.5135,
|
| 926 |
+
"step": 153000
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 9.78,
|
| 930 |
+
"learning_rate": 1.105604269919939e-06,
|
| 931 |
+
"loss": 0.5121,
|
| 932 |
+
"step": 154000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 9.84,
|
| 936 |
+
"learning_rate": 7.879018935061634e-07,
|
| 937 |
+
"loss": 0.514,
|
| 938 |
+
"step": 155000
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 9.91,
|
| 942 |
+
"learning_rate": 4.701995170923879e-07,
|
| 943 |
+
"loss": 0.5134,
|
| 944 |
+
"step": 156000
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 9.97,
|
| 948 |
+
"learning_rate": 1.5249714067861227e-07,
|
| 949 |
+
"loss": 0.5109,
|
| 950 |
+
"step": 157000
|
| 951 |
+
}
|
| 952 |
+
],
|
| 953 |
+
"max_steps": 157480,
|
| 954 |
+
"num_train_epochs": 10,
|
| 955 |
+
"total_flos": 1.68490001031168e+16,
|
| 956 |
+
"trial_name": null,
|
| 957 |
+
"trial_params": null
|
| 958 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b79b66fabbbf14ac0fa4635c63ca9914fd435f821bef98b62c09f01970470f94
|
| 3 |
+
size 3119
|