Commit
•
474f148
1
Parent(s):
83b3858
Upload 10 files
Browse files- config.json +30 -0
- generation_config.json +6 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- tokenizer.json +0 -0
- tokenizer_config.json +1 -0
- trainer_state.json +2149 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "data/models/Mixtral-8x7B-v0.3",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"MixtralForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 1,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"hidden_act": "silu",
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 768,
|
| 13 |
+
"max_position_embeddings": 512,
|
| 14 |
+
"model_type": "mixtral",
|
| 15 |
+
"num_attention_heads": 8,
|
| 16 |
+
"num_experts_per_tok": 1,
|
| 17 |
+
"num_hidden_layers": 8,
|
| 18 |
+
"num_key_value_heads": 8,
|
| 19 |
+
"num_local_experts": 8,
|
| 20 |
+
"output_router_logits": false,
|
| 21 |
+
"rms_norm_eps": 1e-05,
|
| 22 |
+
"rope_theta": 1000000.0,
|
| 23 |
+
"router_aux_loss_coef": 0.02,
|
| 24 |
+
"sliding_window": null,
|
| 25 |
+
"tie_word_embeddings": false,
|
| 26 |
+
"torch_dtype": "bfloat16",
|
| 27 |
+
"transformers_version": "4.37.1",
|
| 28 |
+
"use_cache": true,
|
| 29 |
+
"vocab_size": 4096
|
| 30 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 1,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"transformers_version": "4.37.1"
|
| 6 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:10c5ed6a5dd737f64d43939371730dae712752a62d801d13e4b999ea000461d3
|
| 3 |
+
size 276979168
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1019e9e1537716e2ad64c5b6ed58f5b79624189f89b3227a77bf8b731a6be557
|
| 3 |
+
size 554108549
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1cfefee1204c1a5fb3e3901bd1ebf46a40b0e2aa0582720b87271309d529a880
|
| 3 |
+
size 14639
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4531945a00d959ff6a06d6f8b85a1b0372c7856d5f7bac506029f8fe7a1915d
|
| 3 |
+
size 627
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"tokenizer_class": "PreTrainedTokenizerFast", "unk_token": "[UNK]", "cls_token": "[CLS]", "sep_token": "[SEP]", "pad_token": "[PAD]", "mask_token": "[MASK]"}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2149 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 6.020912170410156,
|
| 3 |
+
"best_model_checkpoint": "./results/models/checkpoint-139568",
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 174460,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01,
|
| 13 |
+
"learning_rate": 0.0019994268027054914,
|
| 14 |
+
"loss": 6.8812,
|
| 15 |
+
"step": 500
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.03,
|
| 19 |
+
"learning_rate": 0.0019988536054109823,
|
| 20 |
+
"loss": 6.7318,
|
| 21 |
+
"step": 1000
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.04,
|
| 25 |
+
"learning_rate": 0.0019982804081164736,
|
| 26 |
+
"loss": 6.7006,
|
| 27 |
+
"step": 1500
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.06,
|
| 31 |
+
"learning_rate": 0.001997707210821965,
|
| 32 |
+
"loss": 6.6761,
|
| 33 |
+
"step": 2000
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.07,
|
| 37 |
+
"learning_rate": 0.0019971340135274563,
|
| 38 |
+
"loss": 6.6478,
|
| 39 |
+
"step": 2500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.09,
|
| 43 |
+
"learning_rate": 0.001996560816232947,
|
| 44 |
+
"loss": 6.6334,
|
| 45 |
+
"step": 3000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.1,
|
| 49 |
+
"learning_rate": 0.0019959876189384385,
|
| 50 |
+
"loss": 6.6285,
|
| 51 |
+
"step": 3500
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.11,
|
| 55 |
+
"learning_rate": 0.00199541442164393,
|
| 56 |
+
"loss": 6.6122,
|
| 57 |
+
"step": 4000
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.13,
|
| 61 |
+
"learning_rate": 0.0019948412243494212,
|
| 62 |
+
"loss": 6.6007,
|
| 63 |
+
"step": 4500
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.14,
|
| 67 |
+
"learning_rate": 0.001994268027054912,
|
| 68 |
+
"loss": 6.5815,
|
| 69 |
+
"step": 5000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.16,
|
| 73 |
+
"learning_rate": 0.0019936948297604035,
|
| 74 |
+
"loss": 6.5741,
|
| 75 |
+
"step": 5500
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.17,
|
| 79 |
+
"learning_rate": 0.001993121632465895,
|
| 80 |
+
"loss": 6.5711,
|
| 81 |
+
"step": 6000
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.19,
|
| 85 |
+
"learning_rate": 0.001992548435171386,
|
| 86 |
+
"loss": 6.5854,
|
| 87 |
+
"step": 6500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.2,
|
| 91 |
+
"learning_rate": 0.001991975237876877,
|
| 92 |
+
"loss": 6.5759,
|
| 93 |
+
"step": 7000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.21,
|
| 97 |
+
"learning_rate": 0.0019914020405823684,
|
| 98 |
+
"loss": 6.5474,
|
| 99 |
+
"step": 7500
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.23,
|
| 103 |
+
"learning_rate": 0.0019908288432878597,
|
| 104 |
+
"loss": 6.5431,
|
| 105 |
+
"step": 8000
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.24,
|
| 109 |
+
"learning_rate": 0.001990255645993351,
|
| 110 |
+
"loss": 6.5421,
|
| 111 |
+
"step": 8500
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.26,
|
| 115 |
+
"learning_rate": 0.001989682448698842,
|
| 116 |
+
"loss": 6.532,
|
| 117 |
+
"step": 9000
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.27,
|
| 121 |
+
"learning_rate": 0.0019891092514043333,
|
| 122 |
+
"loss": 6.5349,
|
| 123 |
+
"step": 9500
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.29,
|
| 127 |
+
"learning_rate": 0.0019885360541098246,
|
| 128 |
+
"loss": 6.5146,
|
| 129 |
+
"step": 10000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.3,
|
| 133 |
+
"learning_rate": 0.001987962856815316,
|
| 134 |
+
"loss": 6.5103,
|
| 135 |
+
"step": 10500
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.32,
|
| 139 |
+
"learning_rate": 0.0019873896595208073,
|
| 140 |
+
"loss": 6.5063,
|
| 141 |
+
"step": 11000
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.33,
|
| 145 |
+
"learning_rate": 0.0019868164622262982,
|
| 146 |
+
"loss": 6.494,
|
| 147 |
+
"step": 11500
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.34,
|
| 151 |
+
"learning_rate": 0.0019862432649317896,
|
| 152 |
+
"loss": 6.5092,
|
| 153 |
+
"step": 12000
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.36,
|
| 157 |
+
"learning_rate": 0.001985670067637281,
|
| 158 |
+
"loss": 6.4969,
|
| 159 |
+
"step": 12500
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.37,
|
| 163 |
+
"learning_rate": 0.0019850968703427722,
|
| 164 |
+
"loss": 6.4886,
|
| 165 |
+
"step": 13000
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.39,
|
| 169 |
+
"learning_rate": 0.001984523673048263,
|
| 170 |
+
"loss": 6.4776,
|
| 171 |
+
"step": 13500
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.4,
|
| 175 |
+
"learning_rate": 0.0019839504757537545,
|
| 176 |
+
"loss": 6.4729,
|
| 177 |
+
"step": 14000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.42,
|
| 181 |
+
"learning_rate": 0.001983377278459246,
|
| 182 |
+
"loss": 6.4706,
|
| 183 |
+
"step": 14500
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.43,
|
| 187 |
+
"learning_rate": 0.001982804081164737,
|
| 188 |
+
"loss": 6.4803,
|
| 189 |
+
"step": 15000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.44,
|
| 193 |
+
"learning_rate": 0.001982230883870228,
|
| 194 |
+
"loss": 6.4827,
|
| 195 |
+
"step": 15500
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.46,
|
| 199 |
+
"learning_rate": 0.0019816576865757194,
|
| 200 |
+
"loss": 6.4716,
|
| 201 |
+
"step": 16000
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.47,
|
| 205 |
+
"learning_rate": 0.0019810844892812107,
|
| 206 |
+
"loss": 6.4559,
|
| 207 |
+
"step": 16500
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.49,
|
| 211 |
+
"learning_rate": 0.001980511291986702,
|
| 212 |
+
"loss": 6.4458,
|
| 213 |
+
"step": 17000
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.5,
|
| 217 |
+
"learning_rate": 0.001979938094692193,
|
| 218 |
+
"loss": 6.4338,
|
| 219 |
+
"step": 17500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.52,
|
| 223 |
+
"learning_rate": 0.0019793648973976843,
|
| 224 |
+
"loss": 6.4282,
|
| 225 |
+
"step": 18000
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.53,
|
| 229 |
+
"learning_rate": 0.0019787917001031757,
|
| 230 |
+
"loss": 6.438,
|
| 231 |
+
"step": 18500
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.54,
|
| 235 |
+
"learning_rate": 0.0019782185028086666,
|
| 236 |
+
"loss": 6.4408,
|
| 237 |
+
"step": 19000
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.56,
|
| 241 |
+
"learning_rate": 0.001977645305514158,
|
| 242 |
+
"loss": 6.4282,
|
| 243 |
+
"step": 19500
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.57,
|
| 247 |
+
"learning_rate": 0.0019770721082196492,
|
| 248 |
+
"loss": 6.43,
|
| 249 |
+
"step": 20000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.59,
|
| 253 |
+
"learning_rate": 0.0019764989109251406,
|
| 254 |
+
"loss": 6.4287,
|
| 255 |
+
"step": 20500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.6,
|
| 259 |
+
"learning_rate": 0.0019759257136306315,
|
| 260 |
+
"loss": 6.4267,
|
| 261 |
+
"step": 21000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.62,
|
| 265 |
+
"learning_rate": 0.001975352516336123,
|
| 266 |
+
"loss": 6.4148,
|
| 267 |
+
"step": 21500
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.63,
|
| 271 |
+
"learning_rate": 0.001974779319041614,
|
| 272 |
+
"loss": 6.4,
|
| 273 |
+
"step": 22000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.64,
|
| 277 |
+
"learning_rate": 0.0019742061217471055,
|
| 278 |
+
"loss": 6.3894,
|
| 279 |
+
"step": 22500
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.66,
|
| 283 |
+
"learning_rate": 0.0019736329244525964,
|
| 284 |
+
"loss": 6.395,
|
| 285 |
+
"step": 23000
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.67,
|
| 289 |
+
"learning_rate": 0.0019730597271580877,
|
| 290 |
+
"loss": 6.3977,
|
| 291 |
+
"step": 23500
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.69,
|
| 295 |
+
"learning_rate": 0.001972486529863579,
|
| 296 |
+
"loss": 6.4017,
|
| 297 |
+
"step": 24000
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.7,
|
| 301 |
+
"learning_rate": 0.0019719133325690704,
|
| 302 |
+
"loss": 6.3995,
|
| 303 |
+
"step": 24500
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.72,
|
| 307 |
+
"learning_rate": 0.0019713401352745613,
|
| 308 |
+
"loss": 6.3949,
|
| 309 |
+
"step": 25000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.73,
|
| 313 |
+
"learning_rate": 0.0019707669379800527,
|
| 314 |
+
"loss": 6.3947,
|
| 315 |
+
"step": 25500
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.75,
|
| 319 |
+
"learning_rate": 0.001970193740685544,
|
| 320 |
+
"loss": 6.4071,
|
| 321 |
+
"step": 26000
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.76,
|
| 325 |
+
"learning_rate": 0.0019696205433910353,
|
| 326 |
+
"loss": 6.3988,
|
| 327 |
+
"step": 26500
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.77,
|
| 331 |
+
"learning_rate": 0.0019690473460965262,
|
| 332 |
+
"loss": 6.3915,
|
| 333 |
+
"step": 27000
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.79,
|
| 337 |
+
"learning_rate": 0.0019684741488020176,
|
| 338 |
+
"loss": 6.3869,
|
| 339 |
+
"step": 27500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.8,
|
| 343 |
+
"learning_rate": 0.001967900951507509,
|
| 344 |
+
"loss": 6.3724,
|
| 345 |
+
"step": 28000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.82,
|
| 349 |
+
"learning_rate": 0.0019673277542130003,
|
| 350 |
+
"loss": 6.3722,
|
| 351 |
+
"step": 28500
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.83,
|
| 355 |
+
"learning_rate": 0.0019667545569184916,
|
| 356 |
+
"loss": 6.3742,
|
| 357 |
+
"step": 29000
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.85,
|
| 361 |
+
"learning_rate": 0.0019661813596239825,
|
| 362 |
+
"loss": 6.3672,
|
| 363 |
+
"step": 29500
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.86,
|
| 367 |
+
"learning_rate": 0.001965608162329474,
|
| 368 |
+
"loss": 6.3707,
|
| 369 |
+
"step": 30000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.87,
|
| 373 |
+
"learning_rate": 0.001965034965034965,
|
| 374 |
+
"loss": 6.3687,
|
| 375 |
+
"step": 30500
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.89,
|
| 379 |
+
"learning_rate": 0.0019644617677404565,
|
| 380 |
+
"loss": 6.3644,
|
| 381 |
+
"step": 31000
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.9,
|
| 385 |
+
"learning_rate": 0.0019638885704459474,
|
| 386 |
+
"loss": 6.3596,
|
| 387 |
+
"step": 31500
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.92,
|
| 391 |
+
"learning_rate": 0.0019633153731514388,
|
| 392 |
+
"loss": 6.3613,
|
| 393 |
+
"step": 32000
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.93,
|
| 397 |
+
"learning_rate": 0.00196274217585693,
|
| 398 |
+
"loss": 6.3542,
|
| 399 |
+
"step": 32500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.95,
|
| 403 |
+
"learning_rate": 0.0019621689785624214,
|
| 404 |
+
"loss": 6.3539,
|
| 405 |
+
"step": 33000
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.96,
|
| 409 |
+
"learning_rate": 0.0019615957812679123,
|
| 410 |
+
"loss": 6.3551,
|
| 411 |
+
"step": 33500
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 0.97,
|
| 415 |
+
"learning_rate": 0.0019610225839734037,
|
| 416 |
+
"loss": 6.3491,
|
| 417 |
+
"step": 34000
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.99,
|
| 421 |
+
"learning_rate": 0.001960449386678895,
|
| 422 |
+
"loss": 6.3431,
|
| 423 |
+
"step": 34500
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.0,
|
| 427 |
+
"eval_loss": 6.34874963760376,
|
| 428 |
+
"eval_runtime": 28.5,
|
| 429 |
+
"eval_samples_per_second": 39.228,
|
| 430 |
+
"eval_steps_per_second": 1.228,
|
| 431 |
+
"step": 34892
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.0,
|
| 435 |
+
"learning_rate": 0.0019598761893843864,
|
| 436 |
+
"loss": 6.3416,
|
| 437 |
+
"step": 35000
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.02,
|
| 441 |
+
"learning_rate": 0.0019593029920898773,
|
| 442 |
+
"loss": 6.343,
|
| 443 |
+
"step": 35500
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.03,
|
| 447 |
+
"learning_rate": 0.0019587297947953686,
|
| 448 |
+
"loss": 6.3372,
|
| 449 |
+
"step": 36000
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 1.05,
|
| 453 |
+
"learning_rate": 0.0019581565975008595,
|
| 454 |
+
"loss": 6.3392,
|
| 455 |
+
"step": 36500
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 1.06,
|
| 459 |
+
"learning_rate": 0.001957583400206351,
|
| 460 |
+
"loss": 6.3235,
|
| 461 |
+
"step": 37000
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 1.07,
|
| 465 |
+
"learning_rate": 0.001957010202911842,
|
| 466 |
+
"loss": 6.3273,
|
| 467 |
+
"step": 37500
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 1.09,
|
| 471 |
+
"learning_rate": 0.0019564370056173335,
|
| 472 |
+
"loss": 6.3301,
|
| 473 |
+
"step": 38000
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.1,
|
| 477 |
+
"learning_rate": 0.001955863808322825,
|
| 478 |
+
"loss": 6.3422,
|
| 479 |
+
"step": 38500
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.12,
|
| 483 |
+
"learning_rate": 0.0019552906110283158,
|
| 484 |
+
"loss": 6.3314,
|
| 485 |
+
"step": 39000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.13,
|
| 489 |
+
"learning_rate": 0.001954717413733807,
|
| 490 |
+
"loss": 6.3208,
|
| 491 |
+
"step": 39500
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 1.15,
|
| 495 |
+
"learning_rate": 0.0019541442164392984,
|
| 496 |
+
"loss": 6.3132,
|
| 497 |
+
"step": 40000
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 1.16,
|
| 501 |
+
"learning_rate": 0.00195357101914479,
|
| 502 |
+
"loss": 6.3034,
|
| 503 |
+
"step": 40500
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 1.18,
|
| 507 |
+
"learning_rate": 0.001952997821850281,
|
| 508 |
+
"loss": 6.2984,
|
| 509 |
+
"step": 41000
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 1.19,
|
| 513 |
+
"learning_rate": 0.0019524246245557722,
|
| 514 |
+
"loss": 6.3067,
|
| 515 |
+
"step": 41500
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.2,
|
| 519 |
+
"learning_rate": 0.0019518514272612634,
|
| 520 |
+
"loss": 6.2983,
|
| 521 |
+
"step": 42000
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.22,
|
| 525 |
+
"learning_rate": 0.0019512782299667547,
|
| 526 |
+
"loss": 6.2971,
|
| 527 |
+
"step": 42500
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.23,
|
| 531 |
+
"learning_rate": 0.0019507050326722458,
|
| 532 |
+
"loss": 6.2949,
|
| 533 |
+
"step": 43000
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 1.25,
|
| 537 |
+
"learning_rate": 0.0019501318353777372,
|
| 538 |
+
"loss": 6.2861,
|
| 539 |
+
"step": 43500
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 1.26,
|
| 543 |
+
"learning_rate": 0.0019495586380832283,
|
| 544 |
+
"loss": 6.2903,
|
| 545 |
+
"step": 44000
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 1.28,
|
| 549 |
+
"learning_rate": 0.0019489854407887196,
|
| 550 |
+
"loss": 6.2845,
|
| 551 |
+
"step": 44500
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 1.29,
|
| 555 |
+
"learning_rate": 0.0019484122434942107,
|
| 556 |
+
"loss": 6.2811,
|
| 557 |
+
"step": 45000
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.3,
|
| 561 |
+
"learning_rate": 0.001947839046199702,
|
| 562 |
+
"loss": 6.2832,
|
| 563 |
+
"step": 45500
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.32,
|
| 567 |
+
"learning_rate": 0.0019472658489051932,
|
| 568 |
+
"loss": 6.2748,
|
| 569 |
+
"step": 46000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.33,
|
| 573 |
+
"learning_rate": 0.0019466926516106843,
|
| 574 |
+
"loss": 6.2664,
|
| 575 |
+
"step": 46500
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 1.35,
|
| 579 |
+
"learning_rate": 0.0019461194543161757,
|
| 580 |
+
"loss": 6.2707,
|
| 581 |
+
"step": 47000
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 1.36,
|
| 585 |
+
"learning_rate": 0.0019455462570216668,
|
| 586 |
+
"loss": 6.2626,
|
| 587 |
+
"step": 47500
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 1.38,
|
| 591 |
+
"learning_rate": 0.0019449730597271581,
|
| 592 |
+
"loss": 6.2621,
|
| 593 |
+
"step": 48000
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 1.39,
|
| 597 |
+
"learning_rate": 0.0019443998624326493,
|
| 598 |
+
"loss": 6.2637,
|
| 599 |
+
"step": 48500
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.4,
|
| 603 |
+
"learning_rate": 0.0019438266651381406,
|
| 604 |
+
"loss": 6.2644,
|
| 605 |
+
"step": 49000
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.42,
|
| 609 |
+
"learning_rate": 0.0019432534678436317,
|
| 610 |
+
"loss": 6.2595,
|
| 611 |
+
"step": 49500
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.43,
|
| 615 |
+
"learning_rate": 0.001942680270549123,
|
| 616 |
+
"loss": 6.2502,
|
| 617 |
+
"step": 50000
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 1.45,
|
| 621 |
+
"learning_rate": 0.0019421070732546142,
|
| 622 |
+
"loss": 6.2373,
|
| 623 |
+
"step": 50500
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 1.46,
|
| 627 |
+
"learning_rate": 0.0019415338759601055,
|
| 628 |
+
"loss": 6.2397,
|
| 629 |
+
"step": 51000
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 1.48,
|
| 633 |
+
"learning_rate": 0.0019409606786655966,
|
| 634 |
+
"loss": 6.2358,
|
| 635 |
+
"step": 51500
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 1.49,
|
| 639 |
+
"learning_rate": 0.001940387481371088,
|
| 640 |
+
"loss": 6.2362,
|
| 641 |
+
"step": 52000
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.5,
|
| 645 |
+
"learning_rate": 0.001939814284076579,
|
| 646 |
+
"loss": 6.2348,
|
| 647 |
+
"step": 52500
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.52,
|
| 651 |
+
"learning_rate": 0.0019392410867820704,
|
| 652 |
+
"loss": 6.2401,
|
| 653 |
+
"step": 53000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.53,
|
| 657 |
+
"learning_rate": 0.0019386678894875616,
|
| 658 |
+
"loss": 6.2407,
|
| 659 |
+
"step": 53500
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 1.55,
|
| 663 |
+
"learning_rate": 0.0019380946921930529,
|
| 664 |
+
"loss": 6.2371,
|
| 665 |
+
"step": 54000
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 1.56,
|
| 669 |
+
"learning_rate": 0.001937521494898544,
|
| 670 |
+
"loss": 6.2278,
|
| 671 |
+
"step": 54500
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 1.58,
|
| 675 |
+
"learning_rate": 0.0019369482976040354,
|
| 676 |
+
"loss": 6.222,
|
| 677 |
+
"step": 55000
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 1.59,
|
| 681 |
+
"learning_rate": 0.0019363751003095267,
|
| 682 |
+
"loss": 6.219,
|
| 683 |
+
"step": 55500
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.6,
|
| 687 |
+
"learning_rate": 0.0019358019030150178,
|
| 688 |
+
"loss": 6.2236,
|
| 689 |
+
"step": 56000
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.62,
|
| 693 |
+
"learning_rate": 0.0019352287057205091,
|
| 694 |
+
"loss": 6.2238,
|
| 695 |
+
"step": 56500
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.63,
|
| 699 |
+
"learning_rate": 0.0019346555084260003,
|
| 700 |
+
"loss": 6.2226,
|
| 701 |
+
"step": 57000
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 1.65,
|
| 705 |
+
"learning_rate": 0.0019340823111314916,
|
| 706 |
+
"loss": 6.2154,
|
| 707 |
+
"step": 57500
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 1.66,
|
| 711 |
+
"learning_rate": 0.0019335091138369827,
|
| 712 |
+
"loss": 6.2112,
|
| 713 |
+
"step": 58000
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 1.68,
|
| 717 |
+
"learning_rate": 0.001932935916542474,
|
| 718 |
+
"loss": 6.2077,
|
| 719 |
+
"step": 58500
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 1.69,
|
| 723 |
+
"learning_rate": 0.0019323627192479652,
|
| 724 |
+
"loss": 6.211,
|
| 725 |
+
"step": 59000
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.71,
|
| 729 |
+
"learning_rate": 0.0019317895219534565,
|
| 730 |
+
"loss": 6.2161,
|
| 731 |
+
"step": 59500
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.72,
|
| 735 |
+
"learning_rate": 0.0019312163246589477,
|
| 736 |
+
"loss": 6.2132,
|
| 737 |
+
"step": 60000
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.73,
|
| 741 |
+
"learning_rate": 0.001930643127364439,
|
| 742 |
+
"loss": 6.2179,
|
| 743 |
+
"step": 60500
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 1.75,
|
| 747 |
+
"learning_rate": 0.0019300699300699301,
|
| 748 |
+
"loss": 6.222,
|
| 749 |
+
"step": 61000
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 1.76,
|
| 753 |
+
"learning_rate": 0.0019294967327754215,
|
| 754 |
+
"loss": 6.2136,
|
| 755 |
+
"step": 61500
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 1.78,
|
| 759 |
+
"learning_rate": 0.0019289235354809126,
|
| 760 |
+
"loss": 6.2057,
|
| 761 |
+
"step": 62000
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 1.79,
|
| 765 |
+
"learning_rate": 0.001928350338186404,
|
| 766 |
+
"loss": 6.2021,
|
| 767 |
+
"step": 62500
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.81,
|
| 771 |
+
"learning_rate": 0.001927777140891895,
|
| 772 |
+
"loss": 6.1955,
|
| 773 |
+
"step": 63000
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.82,
|
| 777 |
+
"learning_rate": 0.0019272039435973864,
|
| 778 |
+
"loss": 6.2036,
|
| 779 |
+
"step": 63500
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.83,
|
| 783 |
+
"learning_rate": 0.0019266307463028773,
|
| 784 |
+
"loss": 6.196,
|
| 785 |
+
"step": 64000
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 1.85,
|
| 789 |
+
"learning_rate": 0.0019260575490083686,
|
| 790 |
+
"loss": 6.199,
|
| 791 |
+
"step": 64500
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 1.86,
|
| 795 |
+
"learning_rate": 0.00192548435171386,
|
| 796 |
+
"loss": 6.1965,
|
| 797 |
+
"step": 65000
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 1.88,
|
| 801 |
+
"learning_rate": 0.001924911154419351,
|
| 802 |
+
"loss": 6.199,
|
| 803 |
+
"step": 65500
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 1.89,
|
| 807 |
+
"learning_rate": 0.0019243379571248424,
|
| 808 |
+
"loss": 6.1923,
|
| 809 |
+
"step": 66000
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.91,
|
| 813 |
+
"learning_rate": 0.0019237647598303335,
|
| 814 |
+
"loss": 6.1953,
|
| 815 |
+
"step": 66500
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.92,
|
| 819 |
+
"learning_rate": 0.0019231915625358249,
|
| 820 |
+
"loss": 6.1907,
|
| 821 |
+
"step": 67000
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.93,
|
| 825 |
+
"learning_rate": 0.001922618365241316,
|
| 826 |
+
"loss": 6.195,
|
| 827 |
+
"step": 67500
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 1.95,
|
| 831 |
+
"learning_rate": 0.0019220451679468073,
|
| 832 |
+
"loss": 6.1954,
|
| 833 |
+
"step": 68000
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 1.96,
|
| 837 |
+
"learning_rate": 0.0019214719706522985,
|
| 838 |
+
"loss": 6.1866,
|
| 839 |
+
"step": 68500
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 1.98,
|
| 843 |
+
"learning_rate": 0.0019208987733577898,
|
| 844 |
+
"loss": 6.1891,
|
| 845 |
+
"step": 69000
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 1.99,
|
| 849 |
+
"learning_rate": 0.001920325576063281,
|
| 850 |
+
"loss": 6.1892,
|
| 851 |
+
"step": 69500
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 2.0,
|
| 855 |
+
"eval_loss": 6.193855285644531,
|
| 856 |
+
"eval_runtime": 29.902,
|
| 857 |
+
"eval_samples_per_second": 37.389,
|
| 858 |
+
"eval_steps_per_second": 1.17,
|
| 859 |
+
"step": 69784
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 2.01,
|
| 863 |
+
"learning_rate": 0.0019197523787687723,
|
| 864 |
+
"loss": 6.1899,
|
| 865 |
+
"step": 70000
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 2.02,
|
| 869 |
+
"learning_rate": 0.0019191791814742634,
|
| 870 |
+
"loss": 6.1933,
|
| 871 |
+
"step": 70500
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 2.03,
|
| 875 |
+
"learning_rate": 0.0019186059841797547,
|
| 876 |
+
"loss": 6.1924,
|
| 877 |
+
"step": 71000
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 2.05,
|
| 881 |
+
"learning_rate": 0.0019180327868852458,
|
| 882 |
+
"loss": 6.1888,
|
| 883 |
+
"step": 71500
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 2.06,
|
| 887 |
+
"learning_rate": 0.0019174595895907372,
|
| 888 |
+
"loss": 6.1943,
|
| 889 |
+
"step": 72000
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 2.08,
|
| 893 |
+
"learning_rate": 0.0019168863922962283,
|
| 894 |
+
"loss": 6.1907,
|
| 895 |
+
"step": 72500
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 2.09,
|
| 899 |
+
"learning_rate": 0.0019163131950017196,
|
| 900 |
+
"loss": 6.1904,
|
| 901 |
+
"step": 73000
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 2.11,
|
| 905 |
+
"learning_rate": 0.001915739997707211,
|
| 906 |
+
"loss": 6.1815,
|
| 907 |
+
"step": 73500
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 2.12,
|
| 911 |
+
"learning_rate": 0.001915166800412702,
|
| 912 |
+
"loss": 6.1816,
|
| 913 |
+
"step": 74000
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 2.14,
|
| 917 |
+
"learning_rate": 0.0019145936031181934,
|
| 918 |
+
"loss": 6.1742,
|
| 919 |
+
"step": 74500
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 2.15,
|
| 923 |
+
"learning_rate": 0.0019140204058236846,
|
| 924 |
+
"loss": 6.1747,
|
| 925 |
+
"step": 75000
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 2.16,
|
| 929 |
+
"learning_rate": 0.001913447208529176,
|
| 930 |
+
"loss": 6.1707,
|
| 931 |
+
"step": 75500
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 2.18,
|
| 935 |
+
"learning_rate": 0.001912874011234667,
|
| 936 |
+
"loss": 6.1718,
|
| 937 |
+
"step": 76000
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 2.19,
|
| 941 |
+
"learning_rate": 0.0019123008139401584,
|
| 942 |
+
"loss": 6.1677,
|
| 943 |
+
"step": 76500
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 2.21,
|
| 947 |
+
"learning_rate": 0.0019117276166456495,
|
| 948 |
+
"loss": 6.1584,
|
| 949 |
+
"step": 77000
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 2.22,
|
| 953 |
+
"learning_rate": 0.0019111544193511408,
|
| 954 |
+
"loss": 6.1722,
|
| 955 |
+
"step": 77500
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 2.24,
|
| 959 |
+
"learning_rate": 0.001910581222056632,
|
| 960 |
+
"loss": 6.1628,
|
| 961 |
+
"step": 78000
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 2.25,
|
| 965 |
+
"learning_rate": 0.0019100080247621233,
|
| 966 |
+
"loss": 6.156,
|
| 967 |
+
"step": 78500
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 2.26,
|
| 971 |
+
"learning_rate": 0.0019094348274676144,
|
| 972 |
+
"loss": 6.1512,
|
| 973 |
+
"step": 79000
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 2.28,
|
| 977 |
+
"learning_rate": 0.0019088616301731057,
|
| 978 |
+
"loss": 6.1462,
|
| 979 |
+
"step": 79500
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 2.29,
|
| 983 |
+
"learning_rate": 0.0019082884328785969,
|
| 984 |
+
"loss": 6.1388,
|
| 985 |
+
"step": 80000
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 2.31,
|
| 989 |
+
"learning_rate": 0.0019077152355840882,
|
| 990 |
+
"loss": 6.1404,
|
| 991 |
+
"step": 80500
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 2.32,
|
| 995 |
+
"learning_rate": 0.0019071420382895793,
|
| 996 |
+
"loss": 6.1397,
|
| 997 |
+
"step": 81000
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 2.34,
|
| 1001 |
+
"learning_rate": 0.0019065688409950707,
|
| 1002 |
+
"loss": 6.1493,
|
| 1003 |
+
"step": 81500
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 2.35,
|
| 1007 |
+
"learning_rate": 0.0019059956437005616,
|
| 1008 |
+
"loss": 6.1431,
|
| 1009 |
+
"step": 82000
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 2.36,
|
| 1013 |
+
"learning_rate": 0.001905422446406053,
|
| 1014 |
+
"loss": 6.1521,
|
| 1015 |
+
"step": 82500
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 2.38,
|
| 1019 |
+
"learning_rate": 0.0019048492491115442,
|
| 1020 |
+
"loss": 6.1409,
|
| 1021 |
+
"step": 83000
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 2.39,
|
| 1025 |
+
"learning_rate": 0.0019042760518170354,
|
| 1026 |
+
"loss": 6.1408,
|
| 1027 |
+
"step": 83500
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 2.41,
|
| 1031 |
+
"learning_rate": 0.0019037028545225267,
|
| 1032 |
+
"loss": 6.1416,
|
| 1033 |
+
"step": 84000
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 2.42,
|
| 1037 |
+
"learning_rate": 0.0019031296572280178,
|
| 1038 |
+
"loss": 6.1371,
|
| 1039 |
+
"step": 84500
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 2.44,
|
| 1043 |
+
"learning_rate": 0.0019025564599335092,
|
| 1044 |
+
"loss": 6.1367,
|
| 1045 |
+
"step": 85000
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 2.45,
|
| 1049 |
+
"learning_rate": 0.0019019832626390003,
|
| 1050 |
+
"loss": 6.1372,
|
| 1051 |
+
"step": 85500
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 2.46,
|
| 1055 |
+
"learning_rate": 0.0019014100653444916,
|
| 1056 |
+
"loss": 6.1358,
|
| 1057 |
+
"step": 86000
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 2.48,
|
| 1061 |
+
"learning_rate": 0.0019008368680499827,
|
| 1062 |
+
"loss": 6.1283,
|
| 1063 |
+
"step": 86500
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 2.49,
|
| 1067 |
+
"learning_rate": 0.001900263670755474,
|
| 1068 |
+
"loss": 6.1347,
|
| 1069 |
+
"step": 87000
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 2.51,
|
| 1073 |
+
"learning_rate": 0.0018996904734609652,
|
| 1074 |
+
"loss": 6.135,
|
| 1075 |
+
"step": 87500
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 2.52,
|
| 1079 |
+
"learning_rate": 0.0018991172761664565,
|
| 1080 |
+
"loss": 6.1352,
|
| 1081 |
+
"step": 88000
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 2.54,
|
| 1085 |
+
"learning_rate": 0.0018985440788719477,
|
| 1086 |
+
"loss": 6.1378,
|
| 1087 |
+
"step": 88500
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 2.55,
|
| 1091 |
+
"learning_rate": 0.001897970881577439,
|
| 1092 |
+
"loss": 6.1335,
|
| 1093 |
+
"step": 89000
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 2.57,
|
| 1097 |
+
"learning_rate": 0.0018973976842829301,
|
| 1098 |
+
"loss": 6.1267,
|
| 1099 |
+
"step": 89500
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 2.58,
|
| 1103 |
+
"learning_rate": 0.0018968244869884215,
|
| 1104 |
+
"loss": 6.1266,
|
| 1105 |
+
"step": 90000
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 2.59,
|
| 1109 |
+
"learning_rate": 0.0018962512896939126,
|
| 1110 |
+
"loss": 6.1195,
|
| 1111 |
+
"step": 90500
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 2.61,
|
| 1115 |
+
"learning_rate": 0.001895678092399404,
|
| 1116 |
+
"loss": 6.1167,
|
| 1117 |
+
"step": 91000
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 2.62,
|
| 1121 |
+
"learning_rate": 0.0018951048951048953,
|
| 1122 |
+
"loss": 6.1108,
|
| 1123 |
+
"step": 91500
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 2.64,
|
| 1127 |
+
"learning_rate": 0.0018945316978103864,
|
| 1128 |
+
"loss": 6.1093,
|
| 1129 |
+
"step": 92000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 2.65,
|
| 1133 |
+
"learning_rate": 0.0018939585005158777,
|
| 1134 |
+
"loss": 6.1124,
|
| 1135 |
+
"step": 92500
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"epoch": 2.67,
|
| 1139 |
+
"learning_rate": 0.0018933853032213688,
|
| 1140 |
+
"loss": 6.1069,
|
| 1141 |
+
"step": 93000
|
| 1142 |
+
},
|
| 1143 |
+
{
|
| 1144 |
+
"epoch": 2.68,
|
| 1145 |
+
"learning_rate": 0.0018928121059268602,
|
| 1146 |
+
"loss": 6.1051,
|
| 1147 |
+
"step": 93500
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 2.69,
|
| 1151 |
+
"learning_rate": 0.0018922389086323513,
|
| 1152 |
+
"loss": 6.1192,
|
| 1153 |
+
"step": 94000
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 2.71,
|
| 1157 |
+
"learning_rate": 0.0018916657113378426,
|
| 1158 |
+
"loss": 6.1231,
|
| 1159 |
+
"step": 94500
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 2.72,
|
| 1163 |
+
"learning_rate": 0.0018910925140433338,
|
| 1164 |
+
"loss": 6.1156,
|
| 1165 |
+
"step": 95000
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.74,
|
| 1169 |
+
"learning_rate": 0.001890519316748825,
|
| 1170 |
+
"loss": 6.1098,
|
| 1171 |
+
"step": 95500
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 2.75,
|
| 1175 |
+
"learning_rate": 0.0018899461194543162,
|
| 1176 |
+
"loss": 6.0978,
|
| 1177 |
+
"step": 96000
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 2.77,
|
| 1181 |
+
"learning_rate": 0.0018893729221598076,
|
| 1182 |
+
"loss": 6.0956,
|
| 1183 |
+
"step": 96500
|
| 1184 |
+
},
|
| 1185 |
+
{
|
| 1186 |
+
"epoch": 2.78,
|
| 1187 |
+
"learning_rate": 0.0018887997248652987,
|
| 1188 |
+
"loss": 6.0932,
|
| 1189 |
+
"step": 97000
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 2.79,
|
| 1193 |
+
"learning_rate": 0.00188822652757079,
|
| 1194 |
+
"loss": 6.0938,
|
| 1195 |
+
"step": 97500
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 2.81,
|
| 1199 |
+
"learning_rate": 0.0018876533302762811,
|
| 1200 |
+
"loss": 6.0921,
|
| 1201 |
+
"step": 98000
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 2.82,
|
| 1205 |
+
"learning_rate": 0.0018870801329817725,
|
| 1206 |
+
"loss": 6.0834,
|
| 1207 |
+
"step": 98500
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 2.84,
|
| 1211 |
+
"learning_rate": 0.0018865069356872636,
|
| 1212 |
+
"loss": 6.0833,
|
| 1213 |
+
"step": 99000
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 2.85,
|
| 1217 |
+
"learning_rate": 0.001885933738392755,
|
| 1218 |
+
"loss": 6.0827,
|
| 1219 |
+
"step": 99500
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 2.87,
|
| 1223 |
+
"learning_rate": 0.001885360541098246,
|
| 1224 |
+
"loss": 6.0858,
|
| 1225 |
+
"step": 100000
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 2.88,
|
| 1229 |
+
"learning_rate": 0.0018847873438037372,
|
| 1230 |
+
"loss": 6.0814,
|
| 1231 |
+
"step": 100500
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 2.89,
|
| 1235 |
+
"learning_rate": 0.0018842141465092285,
|
| 1236 |
+
"loss": 6.0793,
|
| 1237 |
+
"step": 101000
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 2.91,
|
| 1241 |
+
"learning_rate": 0.0018836409492147196,
|
| 1242 |
+
"loss": 6.0743,
|
| 1243 |
+
"step": 101500
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 2.92,
|
| 1247 |
+
"learning_rate": 0.001883067751920211,
|
| 1248 |
+
"loss": 6.0821,
|
| 1249 |
+
"step": 102000
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 2.94,
|
| 1253 |
+
"learning_rate": 0.001882494554625702,
|
| 1254 |
+
"loss": 6.0827,
|
| 1255 |
+
"step": 102500
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 2.95,
|
| 1259 |
+
"learning_rate": 0.0018819213573311934,
|
| 1260 |
+
"loss": 6.0723,
|
| 1261 |
+
"step": 103000
|
| 1262 |
+
},
|
| 1263 |
+
{
|
| 1264 |
+
"epoch": 2.97,
|
| 1265 |
+
"learning_rate": 0.0018813481600366846,
|
| 1266 |
+
"loss": 6.0692,
|
| 1267 |
+
"step": 103500
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 2.98,
|
| 1271 |
+
"learning_rate": 0.001880774962742176,
|
| 1272 |
+
"loss": 6.065,
|
| 1273 |
+
"step": 104000
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 2.99,
|
| 1277 |
+
"learning_rate": 0.001880201765447667,
|
| 1278 |
+
"loss": 6.0677,
|
| 1279 |
+
"step": 104500
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 3.0,
|
| 1283 |
+
"eval_loss": 6.062596321105957,
|
| 1284 |
+
"eval_runtime": 26.9494,
|
| 1285 |
+
"eval_samples_per_second": 41.485,
|
| 1286 |
+
"eval_steps_per_second": 1.299,
|
| 1287 |
+
"step": 104676
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 3.01,
|
| 1291 |
+
"learning_rate": 0.0018796285681531584,
|
| 1292 |
+
"loss": 6.0603,
|
| 1293 |
+
"step": 105000
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 3.02,
|
| 1297 |
+
"learning_rate": 0.0018790553708586495,
|
| 1298 |
+
"loss": 6.0606,
|
| 1299 |
+
"step": 105500
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 3.04,
|
| 1303 |
+
"learning_rate": 0.0018784821735641408,
|
| 1304 |
+
"loss": 6.0603,
|
| 1305 |
+
"step": 106000
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 3.05,
|
| 1309 |
+
"learning_rate": 0.001877908976269632,
|
| 1310 |
+
"loss": 6.066,
|
| 1311 |
+
"step": 106500
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 3.07,
|
| 1315 |
+
"learning_rate": 0.0018773357789751233,
|
| 1316 |
+
"loss": 6.0745,
|
| 1317 |
+
"step": 107000
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 3.08,
|
| 1321 |
+
"learning_rate": 0.0018767625816806144,
|
| 1322 |
+
"loss": 6.0723,
|
| 1323 |
+
"step": 107500
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 3.1,
|
| 1327 |
+
"learning_rate": 0.0018761893843861057,
|
| 1328 |
+
"loss": 6.0678,
|
| 1329 |
+
"step": 108000
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 3.11,
|
| 1333 |
+
"learning_rate": 0.0018756161870915969,
|
| 1334 |
+
"loss": 6.0639,
|
| 1335 |
+
"step": 108500
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 3.12,
|
| 1339 |
+
"learning_rate": 0.0018750429897970882,
|
| 1340 |
+
"loss": 6.0588,
|
| 1341 |
+
"step": 109000
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 3.14,
|
| 1345 |
+
"learning_rate": 0.0018744697925025793,
|
| 1346 |
+
"loss": 6.0602,
|
| 1347 |
+
"step": 109500
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 3.15,
|
| 1351 |
+
"learning_rate": 0.0018738965952080707,
|
| 1352 |
+
"loss": 6.054,
|
| 1353 |
+
"step": 110000
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 3.17,
|
| 1357 |
+
"learning_rate": 0.001873323397913562,
|
| 1358 |
+
"loss": 6.0581,
|
| 1359 |
+
"step": 110500
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 3.18,
|
| 1363 |
+
"learning_rate": 0.0018727502006190531,
|
| 1364 |
+
"loss": 6.0562,
|
| 1365 |
+
"step": 111000
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 3.2,
|
| 1369 |
+
"learning_rate": 0.0018721770033245445,
|
| 1370 |
+
"loss": 6.0482,
|
| 1371 |
+
"step": 111500
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 3.21,
|
| 1375 |
+
"learning_rate": 0.0018716038060300356,
|
| 1376 |
+
"loss": 6.0504,
|
| 1377 |
+
"step": 112000
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 3.22,
|
| 1381 |
+
"learning_rate": 0.001871030608735527,
|
| 1382 |
+
"loss": 6.0509,
|
| 1383 |
+
"step": 112500
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 3.24,
|
| 1387 |
+
"learning_rate": 0.001870457411441018,
|
| 1388 |
+
"loss": 6.0545,
|
| 1389 |
+
"step": 113000
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 3.25,
|
| 1393 |
+
"learning_rate": 0.0018698842141465094,
|
| 1394 |
+
"loss": 6.0529,
|
| 1395 |
+
"step": 113500
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 3.27,
|
| 1399 |
+
"learning_rate": 0.0018693110168520005,
|
| 1400 |
+
"loss": 6.0509,
|
| 1401 |
+
"step": 114000
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 3.28,
|
| 1405 |
+
"learning_rate": 0.0018687378195574918,
|
| 1406 |
+
"loss": 6.0604,
|
| 1407 |
+
"step": 114500
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 3.3,
|
| 1411 |
+
"learning_rate": 0.001868164622262983,
|
| 1412 |
+
"loss": 6.0547,
|
| 1413 |
+
"step": 115000
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 3.31,
|
| 1417 |
+
"learning_rate": 0.0018675914249684743,
|
| 1418 |
+
"loss": 6.0622,
|
| 1419 |
+
"step": 115500
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 3.32,
|
| 1423 |
+
"learning_rate": 0.0018670182276739654,
|
| 1424 |
+
"loss": 6.0639,
|
| 1425 |
+
"step": 116000
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 3.34,
|
| 1429 |
+
"learning_rate": 0.0018664450303794568,
|
| 1430 |
+
"loss": 6.0567,
|
| 1431 |
+
"step": 116500
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 3.35,
|
| 1435 |
+
"learning_rate": 0.0018658718330849479,
|
| 1436 |
+
"loss": 6.0493,
|
| 1437 |
+
"step": 117000
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 3.37,
|
| 1441 |
+
"learning_rate": 0.0018652986357904392,
|
| 1442 |
+
"loss": 6.0514,
|
| 1443 |
+
"step": 117500
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 3.38,
|
| 1447 |
+
"learning_rate": 0.0018647254384959303,
|
| 1448 |
+
"loss": 6.0456,
|
| 1449 |
+
"step": 118000
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 3.4,
|
| 1453 |
+
"learning_rate": 0.0018641522412014217,
|
| 1454 |
+
"loss": 6.0514,
|
| 1455 |
+
"step": 118500
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 3.41,
|
| 1459 |
+
"learning_rate": 0.0018635790439069128,
|
| 1460 |
+
"loss": 6.0442,
|
| 1461 |
+
"step": 119000
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 3.42,
|
| 1465 |
+
"learning_rate": 0.001863005846612404,
|
| 1466 |
+
"loss": 6.0402,
|
| 1467 |
+
"step": 119500
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 3.44,
|
| 1471 |
+
"learning_rate": 0.0018624326493178953,
|
| 1472 |
+
"loss": 6.0444,
|
| 1473 |
+
"step": 120000
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 3.45,
|
| 1477 |
+
"learning_rate": 0.0018618594520233864,
|
| 1478 |
+
"loss": 6.0423,
|
| 1479 |
+
"step": 120500
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 3.47,
|
| 1483 |
+
"learning_rate": 0.0018612862547288777,
|
| 1484 |
+
"loss": 6.0527,
|
| 1485 |
+
"step": 121000
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 3.48,
|
| 1489 |
+
"learning_rate": 0.0018607130574343688,
|
| 1490 |
+
"loss": 6.0419,
|
| 1491 |
+
"step": 121500
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 3.5,
|
| 1495 |
+
"learning_rate": 0.0018601398601398602,
|
| 1496 |
+
"loss": 6.0386,
|
| 1497 |
+
"step": 122000
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 3.51,
|
| 1501 |
+
"learning_rate": 0.0018595666628453513,
|
| 1502 |
+
"loss": 6.0389,
|
| 1503 |
+
"step": 122500
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 3.53,
|
| 1507 |
+
"learning_rate": 0.0018589934655508426,
|
| 1508 |
+
"loss": 6.0372,
|
| 1509 |
+
"step": 123000
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 3.54,
|
| 1513 |
+
"learning_rate": 0.0018584202682563338,
|
| 1514 |
+
"loss": 6.03,
|
| 1515 |
+
"step": 123500
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 3.55,
|
| 1519 |
+
"learning_rate": 0.001857847070961825,
|
| 1520 |
+
"loss": 6.031,
|
| 1521 |
+
"step": 124000
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 3.57,
|
| 1525 |
+
"learning_rate": 0.0018572738736673162,
|
| 1526 |
+
"loss": 6.038,
|
| 1527 |
+
"step": 124500
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 3.58,
|
| 1531 |
+
"learning_rate": 0.0018567006763728076,
|
| 1532 |
+
"loss": 6.0285,
|
| 1533 |
+
"step": 125000
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 3.6,
|
| 1537 |
+
"learning_rate": 0.0018561274790782987,
|
| 1538 |
+
"loss": 6.0309,
|
| 1539 |
+
"step": 125500
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 3.61,
|
| 1543 |
+
"learning_rate": 0.00185555428178379,
|
| 1544 |
+
"loss": 6.0351,
|
| 1545 |
+
"step": 126000
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 3.63,
|
| 1549 |
+
"learning_rate": 0.0018549810844892811,
|
| 1550 |
+
"loss": 6.0312,
|
| 1551 |
+
"step": 126500
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 3.64,
|
| 1555 |
+
"learning_rate": 0.0018544078871947725,
|
| 1556 |
+
"loss": 6.0369,
|
| 1557 |
+
"step": 127000
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 3.65,
|
| 1561 |
+
"learning_rate": 0.0018538346899002636,
|
| 1562 |
+
"loss": 6.0317,
|
| 1563 |
+
"step": 127500
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 3.67,
|
| 1567 |
+
"learning_rate": 0.001853261492605755,
|
| 1568 |
+
"loss": 6.036,
|
| 1569 |
+
"step": 128000
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 3.68,
|
| 1573 |
+
"learning_rate": 0.0018526882953112463,
|
| 1574 |
+
"loss": 6.038,
|
| 1575 |
+
"step": 128500
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 3.7,
|
| 1579 |
+
"learning_rate": 0.0018521150980167374,
|
| 1580 |
+
"loss": 6.034,
|
| 1581 |
+
"step": 129000
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 3.71,
|
| 1585 |
+
"learning_rate": 0.0018515419007222287,
|
| 1586 |
+
"loss": 6.0312,
|
| 1587 |
+
"step": 129500
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 3.73,
|
| 1591 |
+
"learning_rate": 0.0018509687034277199,
|
| 1592 |
+
"loss": 6.0305,
|
| 1593 |
+
"step": 130000
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 3.74,
|
| 1597 |
+
"learning_rate": 0.0018503955061332112,
|
| 1598 |
+
"loss": 6.0283,
|
| 1599 |
+
"step": 130500
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 3.75,
|
| 1603 |
+
"learning_rate": 0.0018498223088387023,
|
| 1604 |
+
"loss": 6.0298,
|
| 1605 |
+
"step": 131000
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 3.77,
|
| 1609 |
+
"learning_rate": 0.0018492491115441937,
|
| 1610 |
+
"loss": 6.0355,
|
| 1611 |
+
"step": 131500
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 3.78,
|
| 1615 |
+
"learning_rate": 0.0018486759142496848,
|
| 1616 |
+
"loss": 6.0299,
|
| 1617 |
+
"step": 132000
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 3.8,
|
| 1621 |
+
"learning_rate": 0.0018481027169551761,
|
| 1622 |
+
"loss": 6.0283,
|
| 1623 |
+
"step": 132500
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 3.81,
|
| 1627 |
+
"learning_rate": 0.0018475295196606672,
|
| 1628 |
+
"loss": 6.0302,
|
| 1629 |
+
"step": 133000
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 3.83,
|
| 1633 |
+
"learning_rate": 0.0018469563223661586,
|
| 1634 |
+
"loss": 6.022,
|
| 1635 |
+
"step": 133500
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 3.84,
|
| 1639 |
+
"learning_rate": 0.0018463831250716497,
|
| 1640 |
+
"loss": 6.023,
|
| 1641 |
+
"step": 134000
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 3.85,
|
| 1645 |
+
"learning_rate": 0.001845809927777141,
|
| 1646 |
+
"loss": 6.0229,
|
| 1647 |
+
"step": 134500
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 3.87,
|
| 1651 |
+
"learning_rate": 0.0018452367304826322,
|
| 1652 |
+
"loss": 6.0253,
|
| 1653 |
+
"step": 135000
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 3.88,
|
| 1657 |
+
"learning_rate": 0.0018446635331881235,
|
| 1658 |
+
"loss": 6.0343,
|
| 1659 |
+
"step": 135500
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 3.9,
|
| 1663 |
+
"learning_rate": 0.0018440903358936146,
|
| 1664 |
+
"loss": 6.0315,
|
| 1665 |
+
"step": 136000
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 3.91,
|
| 1669 |
+
"learning_rate": 0.001843517138599106,
|
| 1670 |
+
"loss": 6.031,
|
| 1671 |
+
"step": 136500
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 3.93,
|
| 1675 |
+
"learning_rate": 0.0018429439413045969,
|
| 1676 |
+
"loss": 6.0269,
|
| 1677 |
+
"step": 137000
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 3.94,
|
| 1681 |
+
"learning_rate": 0.0018423707440100882,
|
| 1682 |
+
"loss": 6.0263,
|
| 1683 |
+
"step": 137500
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 3.96,
|
| 1687 |
+
"learning_rate": 0.0018417975467155795,
|
| 1688 |
+
"loss": 6.0277,
|
| 1689 |
+
"step": 138000
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 3.97,
|
| 1693 |
+
"learning_rate": 0.0018412243494210707,
|
| 1694 |
+
"loss": 6.0346,
|
| 1695 |
+
"step": 138500
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 3.98,
|
| 1699 |
+
"learning_rate": 0.001840651152126562,
|
| 1700 |
+
"loss": 6.0296,
|
| 1701 |
+
"step": 139000
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 4.0,
|
| 1705 |
+
"learning_rate": 0.0018400779548320531,
|
| 1706 |
+
"loss": 6.0229,
|
| 1707 |
+
"step": 139500
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 4.0,
|
| 1711 |
+
"eval_loss": 6.020912170410156,
|
| 1712 |
+
"eval_runtime": 27.2309,
|
| 1713 |
+
"eval_samples_per_second": 41.056,
|
| 1714 |
+
"eval_steps_per_second": 1.285,
|
| 1715 |
+
"step": 139568
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 4.01,
|
| 1719 |
+
"learning_rate": 0.0018395047575375445,
|
| 1720 |
+
"loss": 6.0199,
|
| 1721 |
+
"step": 140000
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 4.03,
|
| 1725 |
+
"learning_rate": 0.0018389315602430356,
|
| 1726 |
+
"loss": 6.0325,
|
| 1727 |
+
"step": 140500
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 4.04,
|
| 1731 |
+
"learning_rate": 0.001838358362948527,
|
| 1732 |
+
"loss": 6.0448,
|
| 1733 |
+
"step": 141000
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 4.06,
|
| 1737 |
+
"learning_rate": 0.001837785165654018,
|
| 1738 |
+
"loss": 6.0367,
|
| 1739 |
+
"step": 141500
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 4.07,
|
| 1743 |
+
"learning_rate": 0.0018372119683595094,
|
| 1744 |
+
"loss": 6.0357,
|
| 1745 |
+
"step": 142000
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 4.08,
|
| 1749 |
+
"learning_rate": 0.0018366387710650005,
|
| 1750 |
+
"loss": 6.0411,
|
| 1751 |
+
"step": 142500
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 4.1,
|
| 1755 |
+
"learning_rate": 0.0018360655737704918,
|
| 1756 |
+
"loss": 6.0365,
|
| 1757 |
+
"step": 143000
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 4.11,
|
| 1761 |
+
"learning_rate": 0.001835492376475983,
|
| 1762 |
+
"loss": 6.0269,
|
| 1763 |
+
"step": 143500
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 4.13,
|
| 1767 |
+
"learning_rate": 0.0018349191791814743,
|
| 1768 |
+
"loss": 6.0342,
|
| 1769 |
+
"step": 144000
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 4.14,
|
| 1773 |
+
"learning_rate": 0.0018343459818869654,
|
| 1774 |
+
"loss": 6.0291,
|
| 1775 |
+
"step": 144500
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 4.16,
|
| 1779 |
+
"learning_rate": 0.0018337727845924568,
|
| 1780 |
+
"loss": 6.0281,
|
| 1781 |
+
"step": 145000
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 4.17,
|
| 1785 |
+
"learning_rate": 0.0018331995872979479,
|
| 1786 |
+
"loss": 6.027,
|
| 1787 |
+
"step": 145500
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 4.18,
|
| 1791 |
+
"learning_rate": 0.0018326263900034392,
|
| 1792 |
+
"loss": 6.0249,
|
| 1793 |
+
"step": 146000
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 4.2,
|
| 1797 |
+
"learning_rate": 0.0018320531927089306,
|
| 1798 |
+
"loss": 6.0276,
|
| 1799 |
+
"step": 146500
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 4.21,
|
| 1803 |
+
"learning_rate": 0.0018314799954144217,
|
| 1804 |
+
"loss": 6.0251,
|
| 1805 |
+
"step": 147000
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 4.23,
|
| 1809 |
+
"learning_rate": 0.001830906798119913,
|
| 1810 |
+
"loss": 6.021,
|
| 1811 |
+
"step": 147500
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 4.24,
|
| 1815 |
+
"learning_rate": 0.0018303336008254041,
|
| 1816 |
+
"loss": 6.0251,
|
| 1817 |
+
"step": 148000
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 4.26,
|
| 1821 |
+
"learning_rate": 0.0018297604035308955,
|
| 1822 |
+
"loss": 6.0278,
|
| 1823 |
+
"step": 148500
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 4.27,
|
| 1827 |
+
"learning_rate": 0.0018291872062363866,
|
| 1828 |
+
"loss": 6.0329,
|
| 1829 |
+
"step": 149000
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 4.28,
|
| 1833 |
+
"learning_rate": 0.001828614008941878,
|
| 1834 |
+
"loss": 6.0398,
|
| 1835 |
+
"step": 149500
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 4.3,
|
| 1839 |
+
"learning_rate": 0.001828040811647369,
|
| 1840 |
+
"loss": 6.0468,
|
| 1841 |
+
"step": 150000
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 4.31,
|
| 1845 |
+
"learning_rate": 0.0018274676143528604,
|
| 1846 |
+
"loss": 6.0485,
|
| 1847 |
+
"step": 150500
|
| 1848 |
+
},
|
| 1849 |
+
{
|
| 1850 |
+
"epoch": 4.33,
|
| 1851 |
+
"learning_rate": 0.0018268944170583515,
|
| 1852 |
+
"loss": 6.0523,
|
| 1853 |
+
"step": 151000
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 4.34,
|
| 1857 |
+
"learning_rate": 0.0018263212197638429,
|
| 1858 |
+
"loss": 6.0604,
|
| 1859 |
+
"step": 151500
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 4.36,
|
| 1863 |
+
"learning_rate": 0.001825748022469334,
|
| 1864 |
+
"loss": 6.0486,
|
| 1865 |
+
"step": 152000
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 4.37,
|
| 1869 |
+
"learning_rate": 0.0018251748251748253,
|
| 1870 |
+
"loss": 6.0446,
|
| 1871 |
+
"step": 152500
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 4.38,
|
| 1875 |
+
"learning_rate": 0.0018246016278803164,
|
| 1876 |
+
"loss": 6.0368,
|
| 1877 |
+
"step": 153000
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 4.4,
|
| 1881 |
+
"learning_rate": 0.0018240284305858078,
|
| 1882 |
+
"loss": 6.0323,
|
| 1883 |
+
"step": 153500
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 4.41,
|
| 1887 |
+
"learning_rate": 0.001823455233291299,
|
| 1888 |
+
"loss": 6.0308,
|
| 1889 |
+
"step": 154000
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 4.43,
|
| 1893 |
+
"learning_rate": 0.0018228820359967902,
|
| 1894 |
+
"loss": 6.0291,
|
| 1895 |
+
"step": 154500
|
| 1896 |
+
},
|
| 1897 |
+
{
|
| 1898 |
+
"epoch": 4.44,
|
| 1899 |
+
"learning_rate": 0.0018223088387022812,
|
| 1900 |
+
"loss": 6.0372,
|
| 1901 |
+
"step": 155000
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 4.46,
|
| 1905 |
+
"learning_rate": 0.0018217356414077725,
|
| 1906 |
+
"loss": 6.0355,
|
| 1907 |
+
"step": 155500
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 4.47,
|
| 1911 |
+
"learning_rate": 0.0018211624441132638,
|
| 1912 |
+
"loss": 6.0415,
|
| 1913 |
+
"step": 156000
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 4.49,
|
| 1917 |
+
"learning_rate": 0.001820589246818755,
|
| 1918 |
+
"loss": 6.0396,
|
| 1919 |
+
"step": 156500
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 4.5,
|
| 1923 |
+
"learning_rate": 0.0018200160495242463,
|
| 1924 |
+
"loss": 6.034,
|
| 1925 |
+
"step": 157000
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 4.51,
|
| 1929 |
+
"learning_rate": 0.0018194428522297374,
|
| 1930 |
+
"loss": 6.0322,
|
| 1931 |
+
"step": 157500
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 4.53,
|
| 1935 |
+
"learning_rate": 0.0018188696549352287,
|
| 1936 |
+
"loss": 6.0336,
|
| 1937 |
+
"step": 158000
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 4.54,
|
| 1941 |
+
"learning_rate": 0.0018182964576407199,
|
| 1942 |
+
"loss": 6.0386,
|
| 1943 |
+
"step": 158500
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 4.56,
|
| 1947 |
+
"learning_rate": 0.0018177232603462112,
|
| 1948 |
+
"loss": 6.0294,
|
| 1949 |
+
"step": 159000
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 4.57,
|
| 1953 |
+
"learning_rate": 0.0018171500630517023,
|
| 1954 |
+
"loss": 6.0288,
|
| 1955 |
+
"step": 159500
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 4.59,
|
| 1959 |
+
"learning_rate": 0.0018165768657571937,
|
| 1960 |
+
"loss": 6.0205,
|
| 1961 |
+
"step": 160000
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 4.6,
|
| 1965 |
+
"learning_rate": 0.0018160036684626848,
|
| 1966 |
+
"loss": 6.0317,
|
| 1967 |
+
"step": 160500
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 4.61,
|
| 1971 |
+
"learning_rate": 0.0018154304711681761,
|
| 1972 |
+
"loss": 6.0411,
|
| 1973 |
+
"step": 161000
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 4.63,
|
| 1977 |
+
"learning_rate": 0.0018148572738736672,
|
| 1978 |
+
"loss": 6.0329,
|
| 1979 |
+
"step": 161500
|
| 1980 |
+
},
|
| 1981 |
+
{
|
| 1982 |
+
"epoch": 4.64,
|
| 1983 |
+
"learning_rate": 0.0018142840765791586,
|
| 1984 |
+
"loss": 6.0292,
|
| 1985 |
+
"step": 162000
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 4.66,
|
| 1989 |
+
"learning_rate": 0.0018137108792846497,
|
| 1990 |
+
"loss": 6.0329,
|
| 1991 |
+
"step": 162500
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 4.67,
|
| 1995 |
+
"learning_rate": 0.001813137681990141,
|
| 1996 |
+
"loss": 6.0332,
|
| 1997 |
+
"step": 163000
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 4.69,
|
| 2001 |
+
"learning_rate": 0.0018125644846956322,
|
| 2002 |
+
"loss": 6.0351,
|
| 2003 |
+
"step": 163500
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 4.7,
|
| 2007 |
+
"learning_rate": 0.0018119912874011235,
|
| 2008 |
+
"loss": 6.0342,
|
| 2009 |
+
"step": 164000
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 4.71,
|
| 2013 |
+
"learning_rate": 0.0018114180901066146,
|
| 2014 |
+
"loss": 6.0428,
|
| 2015 |
+
"step": 164500
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 4.73,
|
| 2019 |
+
"learning_rate": 0.001810844892812106,
|
| 2020 |
+
"loss": 6.039,
|
| 2021 |
+
"step": 165000
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 4.74,
|
| 2025 |
+
"learning_rate": 0.0018102716955175973,
|
| 2026 |
+
"loss": 6.0341,
|
| 2027 |
+
"step": 165500
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 4.76,
|
| 2031 |
+
"learning_rate": 0.0018096984982230884,
|
| 2032 |
+
"loss": 6.035,
|
| 2033 |
+
"step": 166000
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 4.77,
|
| 2037 |
+
"learning_rate": 0.0018091253009285798,
|
| 2038 |
+
"loss": 6.0465,
|
| 2039 |
+
"step": 166500
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 4.79,
|
| 2043 |
+
"learning_rate": 0.0018085521036340709,
|
| 2044 |
+
"loss": 6.0363,
|
| 2045 |
+
"step": 167000
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 4.8,
|
| 2049 |
+
"learning_rate": 0.0018079789063395622,
|
| 2050 |
+
"loss": 6.045,
|
| 2051 |
+
"step": 167500
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 4.81,
|
| 2055 |
+
"learning_rate": 0.0018074057090450533,
|
| 2056 |
+
"loss": 6.0442,
|
| 2057 |
+
"step": 168000
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 4.83,
|
| 2061 |
+
"learning_rate": 0.0018068325117505447,
|
| 2062 |
+
"loss": 6.0453,
|
| 2063 |
+
"step": 168500
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 4.84,
|
| 2067 |
+
"learning_rate": 0.0018062593144560358,
|
| 2068 |
+
"loss": 6.0351,
|
| 2069 |
+
"step": 169000
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 4.86,
|
| 2073 |
+
"learning_rate": 0.0018056861171615271,
|
| 2074 |
+
"loss": 6.0323,
|
| 2075 |
+
"step": 169500
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 4.87,
|
| 2079 |
+
"learning_rate": 0.0018051129198670183,
|
| 2080 |
+
"loss": 6.0308,
|
| 2081 |
+
"step": 170000
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 4.89,
|
| 2085 |
+
"learning_rate": 0.0018045397225725096,
|
| 2086 |
+
"loss": 6.0335,
|
| 2087 |
+
"step": 170500
|
| 2088 |
+
},
|
| 2089 |
+
{
|
| 2090 |
+
"epoch": 4.9,
|
| 2091 |
+
"learning_rate": 0.0018039665252780007,
|
| 2092 |
+
"loss": 6.037,
|
| 2093 |
+
"step": 171000
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 4.92,
|
| 2097 |
+
"learning_rate": 0.001803393327983492,
|
| 2098 |
+
"loss": 6.0301,
|
| 2099 |
+
"step": 171500
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 4.93,
|
| 2103 |
+
"learning_rate": 0.0018028201306889832,
|
| 2104 |
+
"loss": 6.0363,
|
| 2105 |
+
"step": 172000
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 4.94,
|
| 2109 |
+
"learning_rate": 0.0018022469333944745,
|
| 2110 |
+
"loss": 6.0376,
|
| 2111 |
+
"step": 172500
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 4.96,
|
| 2115 |
+
"learning_rate": 0.0018016737360999656,
|
| 2116 |
+
"loss": 6.0319,
|
| 2117 |
+
"step": 173000
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 4.97,
|
| 2121 |
+
"learning_rate": 0.0018011005388054568,
|
| 2122 |
+
"loss": 6.0311,
|
| 2123 |
+
"step": 173500
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 4.99,
|
| 2127 |
+
"learning_rate": 0.0018005273415109481,
|
| 2128 |
+
"loss": 6.0321,
|
| 2129 |
+
"step": 174000
|
| 2130 |
+
},
|
| 2131 |
+
{
|
| 2132 |
+
"epoch": 5.0,
|
| 2133 |
+
"eval_loss": 6.036427974700928,
|
| 2134 |
+
"eval_runtime": 27.0381,
|
| 2135 |
+
"eval_samples_per_second": 41.349,
|
| 2136 |
+
"eval_steps_per_second": 1.294,
|
| 2137 |
+
"step": 174460
|
| 2138 |
+
}
|
| 2139 |
+
],
|
| 2140 |
+
"logging_steps": 500,
|
| 2141 |
+
"max_steps": 1744600,
|
| 2142 |
+
"num_input_tokens_seen": 0,
|
| 2143 |
+
"num_train_epochs": 50,
|
| 2144 |
+
"save_steps": 500,
|
| 2145 |
+
"total_flos": 2.390379476809689e+19,
|
| 2146 |
+
"train_batch_size": 32,
|
| 2147 |
+
"trial_name": null,
|
| 2148 |
+
"trial_params": null
|
| 2149 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d51e4081e7b4ed59fc54a424ce2eeae083ef712ec31040c6b906df43761a7f48
|
| 3 |
+
size 4155
|