Updated
Browse files- all_results.json +14 -0
- config.json +37 -0
- merges.txt +0 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +1 -0
- tokenizer.json +0 -0
- tokenizer_config.json +1 -0
- train_results.json +14 -0
- trainer_state.json +1031 -0
- training_args.bin +3 -0
- vocab.json +0 -0
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"init_mem_cpu_alloc_delta": 1005248512,
|
| 4 |
+
"init_mem_cpu_peaked_delta": 0,
|
| 5 |
+
"init_mem_gpu_alloc_delta": 511148032,
|
| 6 |
+
"init_mem_gpu_peaked_delta": 0,
|
| 7 |
+
"train_mem_cpu_alloc_delta": -17523781632,
|
| 8 |
+
"train_mem_cpu_peaked_delta": 17523781632,
|
| 9 |
+
"train_mem_gpu_alloc_delta": 1501863424,
|
| 10 |
+
"train_mem_gpu_peaked_delta": 12649742848,
|
| 11 |
+
"train_runtime": 27131.1773,
|
| 12 |
+
"train_samples": 33633,
|
| 13 |
+
"train_samples_per_second": 3.099
|
| 14 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "gpt2",
|
| 3 |
+
"activation_function": "gelu_new",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"GPT2LMHeadModel"
|
| 6 |
+
],
|
| 7 |
+
"attn_pdrop": 0.1,
|
| 8 |
+
"bos_token_id": 50256,
|
| 9 |
+
"embd_pdrop": 0.1,
|
| 10 |
+
"eos_token_id": 50256,
|
| 11 |
+
"gradient_checkpointing": false,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"layer_norm_epsilon": 1e-05,
|
| 14 |
+
"model_type": "gpt2",
|
| 15 |
+
"n_ctx": 1024,
|
| 16 |
+
"n_embd": 768,
|
| 17 |
+
"n_head": 12,
|
| 18 |
+
"n_inner": null,
|
| 19 |
+
"n_layer": 12,
|
| 20 |
+
"n_positions": 1024,
|
| 21 |
+
"resid_pdrop": 0.1,
|
| 22 |
+
"scale_attn_weights": true,
|
| 23 |
+
"summary_activation": null,
|
| 24 |
+
"summary_first_dropout": 0.1,
|
| 25 |
+
"summary_proj_to_labels": true,
|
| 26 |
+
"summary_type": "cls_index",
|
| 27 |
+
"summary_use_proj": true,
|
| 28 |
+
"task_specific_params": {
|
| 29 |
+
"text-generation": {
|
| 30 |
+
"do_sample": true,
|
| 31 |
+
"max_length": 50
|
| 32 |
+
}
|
| 33 |
+
},
|
| 34 |
+
"transformers_version": "4.6.0.dev0",
|
| 35 |
+
"use_cache": true,
|
| 36 |
+
"vocab_size": 50257
|
| 37 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4bd5ee42b00638fa3366a2fc8de6a953be9fb4a47c4649160949117af49153a7
|
| 3 |
+
size 510408315
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "unk_token": "<|endoftext|>"}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"unk_token": "<|endoftext|>", "bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "add_prefix_space": false, "model_max_length": 1024, "special_tokens_map_file": null, "name_or_path": "gpt2"}
|
train_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"init_mem_cpu_alloc_delta": 1005248512,
|
| 4 |
+
"init_mem_cpu_peaked_delta": 0,
|
| 5 |
+
"init_mem_gpu_alloc_delta": 511148032,
|
| 6 |
+
"init_mem_gpu_peaked_delta": 0,
|
| 7 |
+
"train_mem_cpu_alloc_delta": -17523781632,
|
| 8 |
+
"train_mem_cpu_peaked_delta": 17523781632,
|
| 9 |
+
"train_mem_gpu_alloc_delta": 1501863424,
|
| 10 |
+
"train_mem_gpu_peaked_delta": 12649742848,
|
| 11 |
+
"train_runtime": 27131.1773,
|
| 12 |
+
"train_samples": 33633,
|
| 13 |
+
"train_samples_per_second": 3.099
|
| 14 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1031 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"global_step": 84090,
|
| 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": 4.970448329171126e-05,
|
| 13 |
+
"loss": 3.6877,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.12,
|
| 18 |
+
"learning_rate": 4.9407777381377095e-05,
|
| 19 |
+
"loss": 3.4145,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.18,
|
| 24 |
+
"learning_rate": 4.911047687002022e-05,
|
| 25 |
+
"loss": 3.3234,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.24,
|
| 30 |
+
"learning_rate": 4.881317635866334e-05,
|
| 31 |
+
"loss": 3.2698,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.3,
|
| 36 |
+
"learning_rate": 4.851587584730646e-05,
|
| 37 |
+
"loss": 3.2274,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.36,
|
| 42 |
+
"learning_rate": 4.821857533594958e-05,
|
| 43 |
+
"loss": 3.1993,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.42,
|
| 48 |
+
"learning_rate": 4.79212748245927e-05,
|
| 49 |
+
"loss": 3.1753,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.48,
|
| 54 |
+
"learning_rate": 4.762397431323582e-05,
|
| 55 |
+
"loss": 3.1476,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.54,
|
| 60 |
+
"learning_rate": 4.732667380187894e-05,
|
| 61 |
+
"loss": 3.1343,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.59,
|
| 66 |
+
"learning_rate": 4.702937329052206e-05,
|
| 67 |
+
"loss": 3.1173,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.65,
|
| 72 |
+
"learning_rate": 4.673207277916518e-05,
|
| 73 |
+
"loss": 3.0942,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.71,
|
| 78 |
+
"learning_rate": 4.6435366868831015e-05,
|
| 79 |
+
"loss": 3.0873,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.77,
|
| 84 |
+
"learning_rate": 4.613806635747413e-05,
|
| 85 |
+
"loss": 3.0778,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.83,
|
| 90 |
+
"learning_rate": 4.584076584611726e-05,
|
| 91 |
+
"loss": 3.065,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.89,
|
| 96 |
+
"learning_rate": 4.5543465334760376e-05,
|
| 97 |
+
"loss": 3.0566,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.95,
|
| 102 |
+
"learning_rate": 4.5246164823403495e-05,
|
| 103 |
+
"loss": 3.0428,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 1.01,
|
| 108 |
+
"learning_rate": 4.494945891306933e-05,
|
| 109 |
+
"loss": 3.03,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 1.07,
|
| 114 |
+
"learning_rate": 4.4652158401712454e-05,
|
| 115 |
+
"loss": 2.9815,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.13,
|
| 120 |
+
"learning_rate": 4.435485789035557e-05,
|
| 121 |
+
"loss": 2.9757,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.19,
|
| 126 |
+
"learning_rate": 4.405755737899869e-05,
|
| 127 |
+
"loss": 2.9704,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.25,
|
| 132 |
+
"learning_rate": 4.3760851468664525e-05,
|
| 133 |
+
"loss": 2.9729,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 1.31,
|
| 138 |
+
"learning_rate": 4.346355095730765e-05,
|
| 139 |
+
"loss": 2.9675,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.37,
|
| 144 |
+
"learning_rate": 4.316625044595077e-05,
|
| 145 |
+
"loss": 2.9551,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 1.43,
|
| 150 |
+
"learning_rate": 4.286894993459389e-05,
|
| 151 |
+
"loss": 2.95,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.49,
|
| 156 |
+
"learning_rate": 4.257164942323701e-05,
|
| 157 |
+
"loss": 2.949,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.55,
|
| 162 |
+
"learning_rate": 4.2274943512902846e-05,
|
| 163 |
+
"loss": 2.9451,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.61,
|
| 168 |
+
"learning_rate": 4.1977643001545964e-05,
|
| 169 |
+
"loss": 2.9412,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.66,
|
| 174 |
+
"learning_rate": 4.168034249018908e-05,
|
| 175 |
+
"loss": 2.9367,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.72,
|
| 180 |
+
"learning_rate": 4.13830419788322e-05,
|
| 181 |
+
"loss": 2.9247,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.78,
|
| 186 |
+
"learning_rate": 4.1085741467475326e-05,
|
| 187 |
+
"loss": 2.9287,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 1.84,
|
| 192 |
+
"learning_rate": 4.078903555714116e-05,
|
| 193 |
+
"loss": 2.913,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 1.9,
|
| 198 |
+
"learning_rate": 4.049173504578428e-05,
|
| 199 |
+
"loss": 2.923,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.96,
|
| 204 |
+
"learning_rate": 4.01944345344274e-05,
|
| 205 |
+
"loss": 2.9127,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 2.02,
|
| 210 |
+
"learning_rate": 3.989713402307052e-05,
|
| 211 |
+
"loss": 2.8935,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 2.08,
|
| 216 |
+
"learning_rate": 3.959983351171364e-05,
|
| 217 |
+
"loss": 2.8674,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 2.14,
|
| 222 |
+
"learning_rate": 3.930253300035676e-05,
|
| 223 |
+
"loss": 2.8602,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 2.2,
|
| 228 |
+
"learning_rate": 3.900582709002259e-05,
|
| 229 |
+
"loss": 2.8658,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.26,
|
| 234 |
+
"learning_rate": 3.870852657866572e-05,
|
| 235 |
+
"loss": 2.8655,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 2.32,
|
| 240 |
+
"learning_rate": 3.8411226067308836e-05,
|
| 241 |
+
"loss": 2.8594,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.38,
|
| 246 |
+
"learning_rate": 3.8113925555951955e-05,
|
| 247 |
+
"loss": 2.8506,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 2.44,
|
| 252 |
+
"learning_rate": 3.781662504459508e-05,
|
| 253 |
+
"loss": 2.8645,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 2.5,
|
| 258 |
+
"learning_rate": 3.75193245332382e-05,
|
| 259 |
+
"loss": 2.8637,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 2.56,
|
| 264 |
+
"learning_rate": 3.722202402188132e-05,
|
| 265 |
+
"loss": 2.8506,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 2.62,
|
| 270 |
+
"learning_rate": 3.692472351052444e-05,
|
| 271 |
+
"loss": 2.8443,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 2.68,
|
| 276 |
+
"learning_rate": 3.6628017600190276e-05,
|
| 277 |
+
"loss": 2.8417,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 2.74,
|
| 282 |
+
"learning_rate": 3.6330717088833394e-05,
|
| 283 |
+
"loss": 2.8418,
|
| 284 |
+
"step": 23000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.79,
|
| 288 |
+
"learning_rate": 3.603341657747651e-05,
|
| 289 |
+
"loss": 2.8462,
|
| 290 |
+
"step": 23500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 2.85,
|
| 294 |
+
"learning_rate": 3.573611606611964e-05,
|
| 295 |
+
"loss": 2.8516,
|
| 296 |
+
"step": 24000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 2.91,
|
| 300 |
+
"learning_rate": 3.5438815554762756e-05,
|
| 301 |
+
"loss": 2.8411,
|
| 302 |
+
"step": 24500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 2.97,
|
| 306 |
+
"learning_rate": 3.5141515043405875e-05,
|
| 307 |
+
"loss": 2.8285,
|
| 308 |
+
"step": 25000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 3.03,
|
| 312 |
+
"learning_rate": 3.4844214532049e-05,
|
| 313 |
+
"loss": 2.8064,
|
| 314 |
+
"step": 25500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 3.09,
|
| 318 |
+
"learning_rate": 3.454691402069212e-05,
|
| 319 |
+
"loss": 2.7888,
|
| 320 |
+
"step": 26000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 3.15,
|
| 324 |
+
"learning_rate": 3.425020811035795e-05,
|
| 325 |
+
"loss": 2.7971,
|
| 326 |
+
"step": 26500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 3.21,
|
| 330 |
+
"learning_rate": 3.395290759900107e-05,
|
| 331 |
+
"loss": 2.793,
|
| 332 |
+
"step": 27000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 3.27,
|
| 336 |
+
"learning_rate": 3.365560708764419e-05,
|
| 337 |
+
"loss": 2.7998,
|
| 338 |
+
"step": 27500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 3.33,
|
| 342 |
+
"learning_rate": 3.3358306576287314e-05,
|
| 343 |
+
"loss": 2.7935,
|
| 344 |
+
"step": 28000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.39,
|
| 348 |
+
"learning_rate": 3.306160066595315e-05,
|
| 349 |
+
"loss": 2.8,
|
| 350 |
+
"step": 28500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 3.45,
|
| 354 |
+
"learning_rate": 3.2764300154596267e-05,
|
| 355 |
+
"loss": 2.7915,
|
| 356 |
+
"step": 29000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 3.51,
|
| 360 |
+
"learning_rate": 3.24675942442621e-05,
|
| 361 |
+
"loss": 2.7924,
|
| 362 |
+
"step": 29500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 3.57,
|
| 366 |
+
"learning_rate": 3.217029373290522e-05,
|
| 367 |
+
"loss": 2.7915,
|
| 368 |
+
"step": 30000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 3.63,
|
| 372 |
+
"learning_rate": 3.1872993221548344e-05,
|
| 373 |
+
"loss": 2.7827,
|
| 374 |
+
"step": 30500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 3.69,
|
| 378 |
+
"learning_rate": 3.157569271019146e-05,
|
| 379 |
+
"loss": 2.7958,
|
| 380 |
+
"step": 31000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 3.75,
|
| 384 |
+
"learning_rate": 3.127839219883458e-05,
|
| 385 |
+
"loss": 2.7905,
|
| 386 |
+
"step": 31500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 3.81,
|
| 390 |
+
"learning_rate": 3.0981686288500415e-05,
|
| 391 |
+
"loss": 2.778,
|
| 392 |
+
"step": 32000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 3.86,
|
| 396 |
+
"learning_rate": 3.068438577714354e-05,
|
| 397 |
+
"loss": 2.7901,
|
| 398 |
+
"step": 32500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 3.92,
|
| 402 |
+
"learning_rate": 3.038708526578666e-05,
|
| 403 |
+
"loss": 2.7851,
|
| 404 |
+
"step": 33000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 3.98,
|
| 408 |
+
"learning_rate": 3.0089784754429777e-05,
|
| 409 |
+
"loss": 2.7936,
|
| 410 |
+
"step": 33500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 4.04,
|
| 414 |
+
"learning_rate": 2.97924842430729e-05,
|
| 415 |
+
"loss": 2.765,
|
| 416 |
+
"step": 34000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 4.1,
|
| 420 |
+
"learning_rate": 2.9495778332738732e-05,
|
| 421 |
+
"loss": 2.7393,
|
| 422 |
+
"step": 34500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 4.16,
|
| 426 |
+
"learning_rate": 2.9198477821381854e-05,
|
| 427 |
+
"loss": 2.7426,
|
| 428 |
+
"step": 35000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 4.22,
|
| 432 |
+
"learning_rate": 2.8901177310024973e-05,
|
| 433 |
+
"loss": 2.7467,
|
| 434 |
+
"step": 35500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 4.28,
|
| 438 |
+
"learning_rate": 2.8603876798668094e-05,
|
| 439 |
+
"loss": 2.7458,
|
| 440 |
+
"step": 36000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 4.34,
|
| 444 |
+
"learning_rate": 2.8306576287311216e-05,
|
| 445 |
+
"loss": 2.7524,
|
| 446 |
+
"step": 36500
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 4.4,
|
| 450 |
+
"learning_rate": 2.8009275775954335e-05,
|
| 451 |
+
"loss": 2.7428,
|
| 452 |
+
"step": 37000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 4.46,
|
| 456 |
+
"learning_rate": 2.7711975264597456e-05,
|
| 457 |
+
"loss": 2.7437,
|
| 458 |
+
"step": 37500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 4.52,
|
| 462 |
+
"learning_rate": 2.7414674753240578e-05,
|
| 463 |
+
"loss": 2.7496,
|
| 464 |
+
"step": 38000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 4.58,
|
| 468 |
+
"learning_rate": 2.7117968842906412e-05,
|
| 469 |
+
"loss": 2.7503,
|
| 470 |
+
"step": 38500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 4.64,
|
| 474 |
+
"learning_rate": 2.682066833154953e-05,
|
| 475 |
+
"loss": 2.7481,
|
| 476 |
+
"step": 39000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 4.7,
|
| 480 |
+
"learning_rate": 2.6523962421215365e-05,
|
| 481 |
+
"loss": 2.7446,
|
| 482 |
+
"step": 39500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 4.76,
|
| 486 |
+
"learning_rate": 2.6226661909858486e-05,
|
| 487 |
+
"loss": 2.7428,
|
| 488 |
+
"step": 40000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 4.82,
|
| 492 |
+
"learning_rate": 2.5929361398501605e-05,
|
| 493 |
+
"loss": 2.7429,
|
| 494 |
+
"step": 40500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 4.88,
|
| 498 |
+
"learning_rate": 2.5632060887144726e-05,
|
| 499 |
+
"loss": 2.7472,
|
| 500 |
+
"step": 41000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 4.94,
|
| 504 |
+
"learning_rate": 2.5334760375787848e-05,
|
| 505 |
+
"loss": 2.7497,
|
| 506 |
+
"step": 41500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 4.99,
|
| 510 |
+
"learning_rate": 2.5037459864430967e-05,
|
| 511 |
+
"loss": 2.7491,
|
| 512 |
+
"step": 42000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 5.05,
|
| 516 |
+
"learning_rate": 2.474015935307409e-05,
|
| 517 |
+
"loss": 2.7089,
|
| 518 |
+
"step": 42500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 5.11,
|
| 522 |
+
"learning_rate": 2.444285884171721e-05,
|
| 523 |
+
"loss": 2.7098,
|
| 524 |
+
"step": 43000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 5.17,
|
| 528 |
+
"learning_rate": 2.4146152931383044e-05,
|
| 529 |
+
"loss": 2.71,
|
| 530 |
+
"step": 43500
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 5.23,
|
| 534 |
+
"learning_rate": 2.3848852420026163e-05,
|
| 535 |
+
"loss": 2.7139,
|
| 536 |
+
"step": 44000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 5.29,
|
| 540 |
+
"learning_rate": 2.3551551908669284e-05,
|
| 541 |
+
"loss": 2.7187,
|
| 542 |
+
"step": 44500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 5.35,
|
| 546 |
+
"learning_rate": 2.3254251397312403e-05,
|
| 547 |
+
"loss": 2.7163,
|
| 548 |
+
"step": 45000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 5.41,
|
| 552 |
+
"learning_rate": 2.2956950885955525e-05,
|
| 553 |
+
"loss": 2.7097,
|
| 554 |
+
"step": 45500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 5.47,
|
| 558 |
+
"learning_rate": 2.266024497562136e-05,
|
| 559 |
+
"loss": 2.7085,
|
| 560 |
+
"step": 46000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 5.53,
|
| 564 |
+
"learning_rate": 2.236294446426448e-05,
|
| 565 |
+
"loss": 2.7001,
|
| 566 |
+
"step": 46500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 5.59,
|
| 570 |
+
"learning_rate": 2.20656439529076e-05,
|
| 571 |
+
"loss": 2.7102,
|
| 572 |
+
"step": 47000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 5.65,
|
| 576 |
+
"learning_rate": 2.176834344155072e-05,
|
| 577 |
+
"loss": 2.7191,
|
| 578 |
+
"step": 47500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 5.71,
|
| 582 |
+
"learning_rate": 2.1471637531216554e-05,
|
| 583 |
+
"loss": 2.7124,
|
| 584 |
+
"step": 48000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 5.77,
|
| 588 |
+
"learning_rate": 2.1174337019859676e-05,
|
| 589 |
+
"loss": 2.7145,
|
| 590 |
+
"step": 48500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 5.83,
|
| 594 |
+
"learning_rate": 2.0877631109525507e-05,
|
| 595 |
+
"loss": 2.7162,
|
| 596 |
+
"step": 49000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 5.89,
|
| 600 |
+
"learning_rate": 2.058033059816863e-05,
|
| 601 |
+
"loss": 2.7112,
|
| 602 |
+
"step": 49500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 5.95,
|
| 606 |
+
"learning_rate": 2.028303008681175e-05,
|
| 607 |
+
"loss": 2.7122,
|
| 608 |
+
"step": 50000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 6.01,
|
| 612 |
+
"learning_rate": 1.998572957545487e-05,
|
| 613 |
+
"loss": 2.7143,
|
| 614 |
+
"step": 50500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 6.06,
|
| 618 |
+
"learning_rate": 1.968842906409799e-05,
|
| 619 |
+
"loss": 2.6759,
|
| 620 |
+
"step": 51000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 6.12,
|
| 624 |
+
"learning_rate": 1.9391128552741112e-05,
|
| 625 |
+
"loss": 2.6826,
|
| 626 |
+
"step": 51500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 6.18,
|
| 630 |
+
"learning_rate": 1.909382804138423e-05,
|
| 631 |
+
"loss": 2.6815,
|
| 632 |
+
"step": 52000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 6.24,
|
| 636 |
+
"learning_rate": 1.8796527530027352e-05,
|
| 637 |
+
"loss": 2.6825,
|
| 638 |
+
"step": 52500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 6.3,
|
| 642 |
+
"learning_rate": 1.8499227018670474e-05,
|
| 643 |
+
"loss": 2.678,
|
| 644 |
+
"step": 53000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 6.36,
|
| 648 |
+
"learning_rate": 1.8202521108336308e-05,
|
| 649 |
+
"loss": 2.6829,
|
| 650 |
+
"step": 53500
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 6.42,
|
| 654 |
+
"learning_rate": 1.7905220596979427e-05,
|
| 655 |
+
"loss": 2.6912,
|
| 656 |
+
"step": 54000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 6.48,
|
| 660 |
+
"learning_rate": 1.760792008562255e-05,
|
| 661 |
+
"loss": 2.6928,
|
| 662 |
+
"step": 54500
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 6.54,
|
| 666 |
+
"learning_rate": 1.731061957426567e-05,
|
| 667 |
+
"loss": 2.6916,
|
| 668 |
+
"step": 55000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 6.6,
|
| 672 |
+
"learning_rate": 1.70139136639315e-05,
|
| 673 |
+
"loss": 2.6829,
|
| 674 |
+
"step": 55500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 6.66,
|
| 678 |
+
"learning_rate": 1.6716613152574623e-05,
|
| 679 |
+
"loss": 2.6827,
|
| 680 |
+
"step": 56000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 6.72,
|
| 684 |
+
"learning_rate": 1.6419312641217744e-05,
|
| 685 |
+
"loss": 2.684,
|
| 686 |
+
"step": 56500
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 6.78,
|
| 690 |
+
"learning_rate": 1.6122012129860863e-05,
|
| 691 |
+
"loss": 2.6825,
|
| 692 |
+
"step": 57000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 6.84,
|
| 696 |
+
"learning_rate": 1.5824711618503984e-05,
|
| 697 |
+
"loss": 2.6761,
|
| 698 |
+
"step": 57500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 6.9,
|
| 702 |
+
"learning_rate": 1.552800570816982e-05,
|
| 703 |
+
"loss": 2.6913,
|
| 704 |
+
"step": 58000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 6.96,
|
| 708 |
+
"learning_rate": 1.5230705196812939e-05,
|
| 709 |
+
"loss": 2.6902,
|
| 710 |
+
"step": 58500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 7.02,
|
| 714 |
+
"learning_rate": 1.4933404685456059e-05,
|
| 715 |
+
"loss": 2.6832,
|
| 716 |
+
"step": 59000
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 7.08,
|
| 720 |
+
"learning_rate": 1.463610417409918e-05,
|
| 721 |
+
"loss": 2.6619,
|
| 722 |
+
"step": 59500
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 7.14,
|
| 726 |
+
"learning_rate": 1.4339398263765014e-05,
|
| 727 |
+
"loss": 2.6594,
|
| 728 |
+
"step": 60000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 7.19,
|
| 732 |
+
"learning_rate": 1.4042097752408134e-05,
|
| 733 |
+
"loss": 2.6659,
|
| 734 |
+
"step": 60500
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 7.25,
|
| 738 |
+
"learning_rate": 1.3745391842073968e-05,
|
| 739 |
+
"loss": 2.6654,
|
| 740 |
+
"step": 61000
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 7.31,
|
| 744 |
+
"learning_rate": 1.3448091330717088e-05,
|
| 745 |
+
"loss": 2.6676,
|
| 746 |
+
"step": 61500
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 7.37,
|
| 750 |
+
"learning_rate": 1.3150790819360209e-05,
|
| 751 |
+
"loss": 2.6582,
|
| 752 |
+
"step": 62000
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 7.43,
|
| 756 |
+
"learning_rate": 1.285349030800333e-05,
|
| 757 |
+
"loss": 2.6585,
|
| 758 |
+
"step": 62500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 7.49,
|
| 762 |
+
"learning_rate": 1.255618979664645e-05,
|
| 763 |
+
"loss": 2.6636,
|
| 764 |
+
"step": 63000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 7.55,
|
| 768 |
+
"learning_rate": 1.225888928528957e-05,
|
| 769 |
+
"loss": 2.6627,
|
| 770 |
+
"step": 63500
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 7.61,
|
| 774 |
+
"learning_rate": 1.1961588773932692e-05,
|
| 775 |
+
"loss": 2.656,
|
| 776 |
+
"step": 64000
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 7.67,
|
| 780 |
+
"learning_rate": 1.1664288262575812e-05,
|
| 781 |
+
"loss": 2.6634,
|
| 782 |
+
"step": 64500
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 7.73,
|
| 786 |
+
"learning_rate": 1.1366987751218933e-05,
|
| 787 |
+
"loss": 2.6616,
|
| 788 |
+
"step": 65000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 7.79,
|
| 792 |
+
"learning_rate": 1.1070281840884766e-05,
|
| 793 |
+
"loss": 2.6614,
|
| 794 |
+
"step": 65500
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 7.85,
|
| 798 |
+
"learning_rate": 1.0772981329527887e-05,
|
| 799 |
+
"loss": 2.6709,
|
| 800 |
+
"step": 66000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 7.91,
|
| 804 |
+
"learning_rate": 1.0475680818171008e-05,
|
| 805 |
+
"loss": 2.6681,
|
| 806 |
+
"step": 66500
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 7.97,
|
| 810 |
+
"learning_rate": 1.0178380306814128e-05,
|
| 811 |
+
"loss": 2.6731,
|
| 812 |
+
"step": 67000
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 8.03,
|
| 816 |
+
"learning_rate": 9.881674396479962e-06,
|
| 817 |
+
"loss": 2.6451,
|
| 818 |
+
"step": 67500
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 8.09,
|
| 822 |
+
"learning_rate": 9.584968486145796e-06,
|
| 823 |
+
"loss": 2.6538,
|
| 824 |
+
"step": 68000
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 8.15,
|
| 828 |
+
"learning_rate": 9.287667974788918e-06,
|
| 829 |
+
"loss": 2.6463,
|
| 830 |
+
"step": 68500
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 8.21,
|
| 834 |
+
"learning_rate": 8.990367463432038e-06,
|
| 835 |
+
"loss": 2.6519,
|
| 836 |
+
"step": 69000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 8.26,
|
| 840 |
+
"learning_rate": 8.693066952075157e-06,
|
| 841 |
+
"loss": 2.6401,
|
| 842 |
+
"step": 69500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 8.32,
|
| 846 |
+
"learning_rate": 8.395766440718278e-06,
|
| 847 |
+
"loss": 2.6448,
|
| 848 |
+
"step": 70000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 8.38,
|
| 852 |
+
"learning_rate": 8.098465929361398e-06,
|
| 853 |
+
"loss": 2.6482,
|
| 854 |
+
"step": 70500
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 8.44,
|
| 858 |
+
"learning_rate": 7.801165418004519e-06,
|
| 859 |
+
"loss": 2.6444,
|
| 860 |
+
"step": 71000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 8.5,
|
| 864 |
+
"learning_rate": 7.5038649066476395e-06,
|
| 865 |
+
"loss": 2.6438,
|
| 866 |
+
"step": 71500
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 8.56,
|
| 870 |
+
"learning_rate": 7.2065643952907604e-06,
|
| 871 |
+
"loss": 2.6448,
|
| 872 |
+
"step": 72000
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 8.62,
|
| 876 |
+
"learning_rate": 6.9098584849565936e-06,
|
| 877 |
+
"loss": 2.6443,
|
| 878 |
+
"step": 72500
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 8.68,
|
| 882 |
+
"learning_rate": 6.6125579735997145e-06,
|
| 883 |
+
"loss": 2.6521,
|
| 884 |
+
"step": 73000
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 8.74,
|
| 888 |
+
"learning_rate": 6.3152574622428354e-06,
|
| 889 |
+
"loss": 2.6491,
|
| 890 |
+
"step": 73500
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 8.8,
|
| 894 |
+
"learning_rate": 6.018551551908669e-06,
|
| 895 |
+
"loss": 2.6341,
|
| 896 |
+
"step": 74000
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 8.86,
|
| 900 |
+
"learning_rate": 5.72125104055179e-06,
|
| 901 |
+
"loss": 2.6593,
|
| 902 |
+
"step": 74500
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 8.92,
|
| 906 |
+
"learning_rate": 5.42395052919491e-06,
|
| 907 |
+
"loss": 2.6501,
|
| 908 |
+
"step": 75000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 8.98,
|
| 912 |
+
"learning_rate": 5.126650017838031e-06,
|
| 913 |
+
"loss": 2.6496,
|
| 914 |
+
"step": 75500
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 9.04,
|
| 918 |
+
"learning_rate": 4.8293495064811514e-06,
|
| 919 |
+
"loss": 2.6418,
|
| 920 |
+
"step": 76000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 9.1,
|
| 924 |
+
"learning_rate": 4.5320489951242715e-06,
|
| 925 |
+
"loss": 2.6389,
|
| 926 |
+
"step": 76500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 9.16,
|
| 930 |
+
"learning_rate": 4.2347484837673925e-06,
|
| 931 |
+
"loss": 2.6309,
|
| 932 |
+
"step": 77000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 9.22,
|
| 936 |
+
"learning_rate": 3.9374479724105126e-06,
|
| 937 |
+
"loss": 2.6315,
|
| 938 |
+
"step": 77500
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 9.28,
|
| 942 |
+
"learning_rate": 3.6407420620763465e-06,
|
| 943 |
+
"loss": 2.6371,
|
| 944 |
+
"step": 78000
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 9.34,
|
| 948 |
+
"learning_rate": 3.343441550719467e-06,
|
| 949 |
+
"loss": 2.6414,
|
| 950 |
+
"step": 78500
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 9.39,
|
| 954 |
+
"learning_rate": 3.046141039362588e-06,
|
| 955 |
+
"loss": 2.6361,
|
| 956 |
+
"step": 79000
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 9.45,
|
| 960 |
+
"learning_rate": 2.7488405280057085e-06,
|
| 961 |
+
"loss": 2.632,
|
| 962 |
+
"step": 79500
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 9.51,
|
| 966 |
+
"learning_rate": 2.4521346176715424e-06,
|
| 967 |
+
"loss": 2.6387,
|
| 968 |
+
"step": 80000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 9.57,
|
| 972 |
+
"learning_rate": 2.154834106314663e-06,
|
| 973 |
+
"loss": 2.6371,
|
| 974 |
+
"step": 80500
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 9.63,
|
| 978 |
+
"learning_rate": 1.8575335949577837e-06,
|
| 979 |
+
"loss": 2.6383,
|
| 980 |
+
"step": 81000
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 9.69,
|
| 984 |
+
"learning_rate": 1.5602330836009038e-06,
|
| 985 |
+
"loss": 2.6379,
|
| 986 |
+
"step": 81500
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 9.75,
|
| 990 |
+
"learning_rate": 1.2629325722440245e-06,
|
| 991 |
+
"loss": 2.6272,
|
| 992 |
+
"step": 82000
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 9.81,
|
| 996 |
+
"learning_rate": 9.668212629325723e-07,
|
| 997 |
+
"loss": 2.6353,
|
| 998 |
+
"step": 82500
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 9.87,
|
| 1002 |
+
"learning_rate": 6.695207515756928e-07,
|
| 1003 |
+
"loss": 2.6404,
|
| 1004 |
+
"step": 83000
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 9.93,
|
| 1008 |
+
"learning_rate": 3.722202402188132e-07,
|
| 1009 |
+
"loss": 2.6373,
|
| 1010 |
+
"step": 83500
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 9.99,
|
| 1014 |
+
"learning_rate": 7.491972886193364e-08,
|
| 1015 |
+
"loss": 2.6304,
|
| 1016 |
+
"step": 84000
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 10.0,
|
| 1020 |
+
"step": 84090,
|
| 1021 |
+
"total_flos": 2.5714385279778816e+17,
|
| 1022 |
+
"train_runtime": 27131.1773,
|
| 1023 |
+
"train_samples_per_second": 3.099
|
| 1024 |
+
}
|
| 1025 |
+
],
|
| 1026 |
+
"max_steps": 84090,
|
| 1027 |
+
"num_train_epochs": 10,
|
| 1028 |
+
"total_flos": 2.5714385279778816e+17,
|
| 1029 |
+
"trial_name": null,
|
| 1030 |
+
"trial_params": null
|
| 1031 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:82ae42394fa03cb8bfb525dc392d949fbb410a186c86f8b4c84fe5d326dc7b13
|
| 3 |
+
size 2415
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|