Commit vesion 1 to the Hub
Browse files- checkpoint-164000/added_tokens.json +3 -0
- checkpoint-164000/config.json +38 -0
- checkpoint-164000/merges.txt +0 -0
- checkpoint-164000/optimizer.pt +3 -0
- checkpoint-164000/pytorch_model.bin +3 -0
- checkpoint-164000/rng_state.pth +3 -0
- checkpoint-164000/scheduler.pt +3 -0
- checkpoint-164000/special_tokens_map.json +23 -0
- checkpoint-164000/tokenizer_config.json +33 -0
- checkpoint-164000/trainer_state.json +1000 -0
- checkpoint-164000/training_args.bin +3 -0
- checkpoint-164000/vocab.json +0 -0
checkpoint-164000/added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"\n": 50257
|
| 3 |
+
}
|
checkpoint-164000/config.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_function": "gelu_new",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"GPT2LMHeadModel"
|
| 5 |
+
],
|
| 6 |
+
"attn_pdrop": 0.1,
|
| 7 |
+
"bos_token_id": 50256,
|
| 8 |
+
"embd_pdrop": 0.1,
|
| 9 |
+
"eos_token_id": 50256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"layer_norm_epsilon": 1e-05,
|
| 12 |
+
"model_type": "gpt2",
|
| 13 |
+
"n_ctx": 260,
|
| 14 |
+
"n_embd": 768,
|
| 15 |
+
"n_head": 12,
|
| 16 |
+
"n_inner": null,
|
| 17 |
+
"n_layer": 12,
|
| 18 |
+
"n_positions": 260,
|
| 19 |
+
"reorder_and_upcast_attn": false,
|
| 20 |
+
"resid_pdrop": 0.1,
|
| 21 |
+
"scale_attn_by_inverse_layer_idx": false,
|
| 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": 256
|
| 32 |
+
}
|
| 33 |
+
},
|
| 34 |
+
"torch_dtype": "float32",
|
| 35 |
+
"transformers_version": "4.21.2",
|
| 36 |
+
"use_cache": true,
|
| 37 |
+
"vocab_size": 50258
|
| 38 |
+
}
|
checkpoint-164000/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-164000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1baea63c869b3d8c125b29b2ad3a9ec06886c5212b0c671b09af088a7a9bc1de
|
| 3 |
+
size 990916401
|
checkpoint-164000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ecf0d23982e372836bfb4b41413040300da5feae74b264e783f583debc6c59f1
|
| 3 |
+
size 496280681
|
checkpoint-164000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d4db8c965d9190c1a2d7a6d62e9dbf7b719ede067cf4359d50cb48b94284da5
|
| 3 |
+
size 14503
|
checkpoint-164000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ba98fca09cb9f3bd5632ad199a76dfde4ac847bf76d064fdcdc21435a328a455
|
| 3 |
+
size 623
|
checkpoint-164000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<|endoftext|>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "<|endoftext|>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"unk_token": {
|
| 17 |
+
"content": "<|endoftext|>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": true,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
}
|
| 23 |
+
}
|
checkpoint-164000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_prefix_space": false,
|
| 4 |
+
"bos_token": {
|
| 5 |
+
"__type": "AddedToken",
|
| 6 |
+
"content": "<|endoftext|>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": true,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"eos_token": {
|
| 13 |
+
"__type": "AddedToken",
|
| 14 |
+
"content": "<|endoftext|>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": true,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false
|
| 19 |
+
},
|
| 20 |
+
"errors": "replace",
|
| 21 |
+
"name_or_path": "danghuy1999/gpt2-viwiki",
|
| 22 |
+
"pad_token": null,
|
| 23 |
+
"special_tokens_map_file": null,
|
| 24 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 25 |
+
"unk_token": {
|
| 26 |
+
"__type": "AddedToken",
|
| 27 |
+
"content": "<|endoftext|>",
|
| 28 |
+
"lstrip": false,
|
| 29 |
+
"normalized": true,
|
| 30 |
+
"rstrip": false,
|
| 31 |
+
"single_word": false
|
| 32 |
+
}
|
| 33 |
+
}
|
checkpoint-164000/trainer_state.json
ADDED
|
@@ -0,0 +1,1000 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 29.948867786705623,
|
| 5 |
+
"global_step": 164000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.18,
|
| 12 |
+
"learning_rate": 5e-05,
|
| 13 |
+
"loss": 7.0207,
|
| 14 |
+
"step": 1000
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.37,
|
| 18 |
+
"learning_rate": 4.96937775600196e-05,
|
| 19 |
+
"loss": 5.1703,
|
| 20 |
+
"step": 2000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.55,
|
| 24 |
+
"learning_rate": 4.93875551200392e-05,
|
| 25 |
+
"loss": 4.7048,
|
| 26 |
+
"step": 3000
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.73,
|
| 30 |
+
"learning_rate": 4.9081332680058794e-05,
|
| 31 |
+
"loss": 4.3894,
|
| 32 |
+
"step": 4000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.91,
|
| 36 |
+
"learning_rate": 4.87751102400784e-05,
|
| 37 |
+
"loss": 4.1348,
|
| 38 |
+
"step": 5000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 1.1,
|
| 42 |
+
"learning_rate": 4.8468887800097995e-05,
|
| 43 |
+
"loss": 3.9447,
|
| 44 |
+
"step": 6000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.28,
|
| 48 |
+
"learning_rate": 4.816266536011759e-05,
|
| 49 |
+
"loss": 3.8286,
|
| 50 |
+
"step": 7000
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.46,
|
| 54 |
+
"learning_rate": 4.785644292013719e-05,
|
| 55 |
+
"loss": 3.7462,
|
| 56 |
+
"step": 8000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.64,
|
| 60 |
+
"learning_rate": 4.7550220480156786e-05,
|
| 61 |
+
"loss": 3.693,
|
| 62 |
+
"step": 9000
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.83,
|
| 66 |
+
"learning_rate": 4.7243998040176384e-05,
|
| 67 |
+
"loss": 3.6311,
|
| 68 |
+
"step": 10000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 2.01,
|
| 72 |
+
"learning_rate": 4.693777560019598e-05,
|
| 73 |
+
"loss": 3.5843,
|
| 74 |
+
"step": 11000
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 2.19,
|
| 78 |
+
"learning_rate": 4.6631553160215585e-05,
|
| 79 |
+
"loss": 3.4624,
|
| 80 |
+
"step": 12000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 2.37,
|
| 84 |
+
"learning_rate": 4.632533072023518e-05,
|
| 85 |
+
"loss": 3.4432,
|
| 86 |
+
"step": 13000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.56,
|
| 90 |
+
"learning_rate": 4.601910828025478e-05,
|
| 91 |
+
"loss": 3.423,
|
| 92 |
+
"step": 14000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 2.74,
|
| 96 |
+
"learning_rate": 4.5712885840274376e-05,
|
| 97 |
+
"loss": 3.4042,
|
| 98 |
+
"step": 15000
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 2.92,
|
| 102 |
+
"learning_rate": 4.540666340029397e-05,
|
| 103 |
+
"loss": 3.3902,
|
| 104 |
+
"step": 16000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 3.1,
|
| 108 |
+
"learning_rate": 4.510044096031357e-05,
|
| 109 |
+
"loss": 3.2973,
|
| 110 |
+
"step": 17000
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 3.29,
|
| 114 |
+
"learning_rate": 4.479421852033317e-05,
|
| 115 |
+
"loss": 3.2482,
|
| 116 |
+
"step": 18000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 3.47,
|
| 120 |
+
"learning_rate": 4.448799608035277e-05,
|
| 121 |
+
"loss": 3.2427,
|
| 122 |
+
"step": 19000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 3.65,
|
| 126 |
+
"learning_rate": 4.418177364037237e-05,
|
| 127 |
+
"loss": 3.2309,
|
| 128 |
+
"step": 20000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 3.83,
|
| 132 |
+
"learning_rate": 4.3875551200391966e-05,
|
| 133 |
+
"loss": 3.2193,
|
| 134 |
+
"step": 21000
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 4.02,
|
| 138 |
+
"learning_rate": 4.356932876041156e-05,
|
| 139 |
+
"loss": 3.2065,
|
| 140 |
+
"step": 22000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 4.2,
|
| 144 |
+
"learning_rate": 4.326310632043116e-05,
|
| 145 |
+
"loss": 3.0733,
|
| 146 |
+
"step": 23000
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 4.38,
|
| 150 |
+
"learning_rate": 4.295688388045076e-05,
|
| 151 |
+
"loss": 3.0868,
|
| 152 |
+
"step": 24000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 4.57,
|
| 156 |
+
"learning_rate": 4.2650661440470355e-05,
|
| 157 |
+
"loss": 3.0884,
|
| 158 |
+
"step": 25000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 4.75,
|
| 162 |
+
"learning_rate": 4.234443900048996e-05,
|
| 163 |
+
"loss": 3.0843,
|
| 164 |
+
"step": 26000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 4.93,
|
| 168 |
+
"learning_rate": 4.2038216560509556e-05,
|
| 169 |
+
"loss": 3.0768,
|
| 170 |
+
"step": 27000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 5.11,
|
| 174 |
+
"learning_rate": 4.173199412052915e-05,
|
| 175 |
+
"loss": 2.9834,
|
| 176 |
+
"step": 28000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 5.3,
|
| 180 |
+
"learning_rate": 4.142577168054875e-05,
|
| 181 |
+
"loss": 2.9267,
|
| 182 |
+
"step": 29000
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 5.48,
|
| 186 |
+
"learning_rate": 4.111954924056835e-05,
|
| 187 |
+
"loss": 2.9463,
|
| 188 |
+
"step": 30000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 5.66,
|
| 192 |
+
"learning_rate": 4.081332680058795e-05,
|
| 193 |
+
"loss": 2.9523,
|
| 194 |
+
"step": 31000
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 5.84,
|
| 198 |
+
"learning_rate": 4.050710436060755e-05,
|
| 199 |
+
"loss": 2.9561,
|
| 200 |
+
"step": 32000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 6.03,
|
| 204 |
+
"learning_rate": 4.0200881920627145e-05,
|
| 205 |
+
"loss": 2.9237,
|
| 206 |
+
"step": 33000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 6.21,
|
| 210 |
+
"learning_rate": 3.989465948064674e-05,
|
| 211 |
+
"loss": 2.7803,
|
| 212 |
+
"step": 34000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 6.39,
|
| 216 |
+
"learning_rate": 3.958843704066634e-05,
|
| 217 |
+
"loss": 2.8068,
|
| 218 |
+
"step": 35000
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 6.57,
|
| 222 |
+
"learning_rate": 3.9282214600685944e-05,
|
| 223 |
+
"loss": 2.8223,
|
| 224 |
+
"step": 36000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 6.76,
|
| 228 |
+
"learning_rate": 3.897599216070554e-05,
|
| 229 |
+
"loss": 2.8238,
|
| 230 |
+
"step": 37000
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 6.94,
|
| 234 |
+
"learning_rate": 3.866976972072514e-05,
|
| 235 |
+
"loss": 2.8341,
|
| 236 |
+
"step": 38000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 7.12,
|
| 240 |
+
"learning_rate": 3.8363547280744735e-05,
|
| 241 |
+
"loss": 2.6976,
|
| 242 |
+
"step": 39000
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 7.3,
|
| 246 |
+
"learning_rate": 3.805732484076434e-05,
|
| 247 |
+
"loss": 2.6679,
|
| 248 |
+
"step": 40000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 7.49,
|
| 252 |
+
"learning_rate": 3.7751102400783936e-05,
|
| 253 |
+
"loss": 2.691,
|
| 254 |
+
"step": 41000
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 7.67,
|
| 258 |
+
"learning_rate": 3.744487996080353e-05,
|
| 259 |
+
"loss": 2.7001,
|
| 260 |
+
"step": 42000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 7.85,
|
| 264 |
+
"learning_rate": 3.713865752082313e-05,
|
| 265 |
+
"loss": 2.7189,
|
| 266 |
+
"step": 43000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 8.04,
|
| 270 |
+
"learning_rate": 3.683243508084273e-05,
|
| 271 |
+
"loss": 2.6829,
|
| 272 |
+
"step": 44000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 8.22,
|
| 276 |
+
"learning_rate": 3.6526212640862325e-05,
|
| 277 |
+
"loss": 2.5262,
|
| 278 |
+
"step": 45000
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 8.4,
|
| 282 |
+
"learning_rate": 3.621999020088192e-05,
|
| 283 |
+
"loss": 2.5547,
|
| 284 |
+
"step": 46000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 8.58,
|
| 288 |
+
"learning_rate": 3.5913767760901526e-05,
|
| 289 |
+
"loss": 2.5926,
|
| 290 |
+
"step": 47000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 8.77,
|
| 294 |
+
"learning_rate": 3.560754532092112e-05,
|
| 295 |
+
"loss": 2.5971,
|
| 296 |
+
"step": 48000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 8.95,
|
| 300 |
+
"learning_rate": 3.530132288094072e-05,
|
| 301 |
+
"loss": 2.6053,
|
| 302 |
+
"step": 49000
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 9.13,
|
| 306 |
+
"learning_rate": 3.499510044096032e-05,
|
| 307 |
+
"loss": 2.4615,
|
| 308 |
+
"step": 50000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 9.31,
|
| 312 |
+
"learning_rate": 3.4688878000979914e-05,
|
| 313 |
+
"loss": 2.4412,
|
| 314 |
+
"step": 51000
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 9.5,
|
| 318 |
+
"learning_rate": 3.438265556099951e-05,
|
| 319 |
+
"loss": 2.4633,
|
| 320 |
+
"step": 52000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 9.68,
|
| 324 |
+
"learning_rate": 3.407643312101911e-05,
|
| 325 |
+
"loss": 2.469,
|
| 326 |
+
"step": 53000
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 9.86,
|
| 330 |
+
"learning_rate": 3.377021068103871e-05,
|
| 331 |
+
"loss": 2.5006,
|
| 332 |
+
"step": 54000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 10.04,
|
| 336 |
+
"learning_rate": 3.346398824105831e-05,
|
| 337 |
+
"loss": 2.4556,
|
| 338 |
+
"step": 55000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 10.23,
|
| 342 |
+
"learning_rate": 3.315776580107791e-05,
|
| 343 |
+
"loss": 2.3134,
|
| 344 |
+
"step": 56000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 10.41,
|
| 348 |
+
"learning_rate": 3.2851543361097504e-05,
|
| 349 |
+
"loss": 2.3445,
|
| 350 |
+
"step": 57000
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 10.59,
|
| 354 |
+
"learning_rate": 3.25453209211171e-05,
|
| 355 |
+
"loss": 2.3709,
|
| 356 |
+
"step": 58000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 10.77,
|
| 360 |
+
"learning_rate": 3.22390984811367e-05,
|
| 361 |
+
"loss": 2.3831,
|
| 362 |
+
"step": 59000
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 10.96,
|
| 366 |
+
"learning_rate": 3.1932876041156296e-05,
|
| 367 |
+
"loss": 2.4089,
|
| 368 |
+
"step": 60000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 11.14,
|
| 372 |
+
"learning_rate": 3.162665360117589e-05,
|
| 373 |
+
"loss": 2.248,
|
| 374 |
+
"step": 61000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 11.32,
|
| 378 |
+
"learning_rate": 3.13204311611955e-05,
|
| 379 |
+
"loss": 2.2344,
|
| 380 |
+
"step": 62000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 11.5,
|
| 384 |
+
"learning_rate": 3.1014208721215094e-05,
|
| 385 |
+
"loss": 2.2553,
|
| 386 |
+
"step": 63000
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 11.69,
|
| 390 |
+
"learning_rate": 3.070798628123469e-05,
|
| 391 |
+
"loss": 2.2816,
|
| 392 |
+
"step": 64000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 11.87,
|
| 396 |
+
"learning_rate": 3.0401763841254288e-05,
|
| 397 |
+
"loss": 2.3067,
|
| 398 |
+
"step": 65000
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 12.05,
|
| 402 |
+
"learning_rate": 3.0095541401273885e-05,
|
| 403 |
+
"loss": 2.248,
|
| 404 |
+
"step": 66000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 12.24,
|
| 408 |
+
"learning_rate": 2.9789318961293483e-05,
|
| 409 |
+
"loss": 2.1257,
|
| 410 |
+
"step": 67000
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 12.42,
|
| 414 |
+
"learning_rate": 2.948309652131308e-05,
|
| 415 |
+
"loss": 2.1559,
|
| 416 |
+
"step": 68000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 12.6,
|
| 420 |
+
"learning_rate": 2.9176874081332684e-05,
|
| 421 |
+
"loss": 2.1743,
|
| 422 |
+
"step": 69000
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 12.78,
|
| 426 |
+
"learning_rate": 2.887065164135228e-05,
|
| 427 |
+
"loss": 2.1997,
|
| 428 |
+
"step": 70000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 12.97,
|
| 432 |
+
"learning_rate": 2.8564429201371878e-05,
|
| 433 |
+
"loss": 2.2321,
|
| 434 |
+
"step": 71000
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 13.15,
|
| 438 |
+
"learning_rate": 2.8258206761391475e-05,
|
| 439 |
+
"loss": 2.0603,
|
| 440 |
+
"step": 72000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 13.33,
|
| 444 |
+
"learning_rate": 2.7951984321411072e-05,
|
| 445 |
+
"loss": 2.0491,
|
| 446 |
+
"step": 73000
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 13.51,
|
| 450 |
+
"learning_rate": 2.764576188143067e-05,
|
| 451 |
+
"loss": 2.079,
|
| 452 |
+
"step": 74000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 13.7,
|
| 456 |
+
"learning_rate": 2.7339539441450267e-05,
|
| 457 |
+
"loss": 2.1119,
|
| 458 |
+
"step": 75000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 13.88,
|
| 462 |
+
"learning_rate": 2.703331700146987e-05,
|
| 463 |
+
"loss": 2.1355,
|
| 464 |
+
"step": 76000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 14.06,
|
| 468 |
+
"learning_rate": 2.6727094561489468e-05,
|
| 469 |
+
"loss": 2.068,
|
| 470 |
+
"step": 77000
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 14.24,
|
| 474 |
+
"learning_rate": 2.6420872121509065e-05,
|
| 475 |
+
"loss": 1.96,
|
| 476 |
+
"step": 78000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 14.43,
|
| 480 |
+
"learning_rate": 2.6114649681528662e-05,
|
| 481 |
+
"loss": 1.9827,
|
| 482 |
+
"step": 79000
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 14.61,
|
| 486 |
+
"learning_rate": 2.5808427241548263e-05,
|
| 487 |
+
"loss": 2.0163,
|
| 488 |
+
"step": 80000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 14.79,
|
| 492 |
+
"learning_rate": 2.550220480156786e-05,
|
| 493 |
+
"loss": 2.0496,
|
| 494 |
+
"step": 81000
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 14.97,
|
| 498 |
+
"learning_rate": 2.5195982361587457e-05,
|
| 499 |
+
"loss": 2.0579,
|
| 500 |
+
"step": 82000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 15.16,
|
| 504 |
+
"learning_rate": 2.4889759921607057e-05,
|
| 505 |
+
"loss": 1.8884,
|
| 506 |
+
"step": 83000
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 15.34,
|
| 510 |
+
"learning_rate": 2.4583537481626655e-05,
|
| 511 |
+
"loss": 1.8992,
|
| 512 |
+
"step": 84000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 15.52,
|
| 516 |
+
"learning_rate": 2.4277315041646255e-05,
|
| 517 |
+
"loss": 1.9383,
|
| 518 |
+
"step": 85000
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 15.7,
|
| 522 |
+
"learning_rate": 2.3971092601665852e-05,
|
| 523 |
+
"loss": 1.9512,
|
| 524 |
+
"step": 86000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 15.89,
|
| 528 |
+
"learning_rate": 2.366487016168545e-05,
|
| 529 |
+
"loss": 1.9814,
|
| 530 |
+
"step": 87000
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 16.07,
|
| 534 |
+
"learning_rate": 2.335864772170505e-05,
|
| 535 |
+
"loss": 1.9075,
|
| 536 |
+
"step": 88000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 16.25,
|
| 540 |
+
"learning_rate": 2.3052425281724647e-05,
|
| 541 |
+
"loss": 1.8141,
|
| 542 |
+
"step": 89000
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 16.44,
|
| 546 |
+
"learning_rate": 2.2746202841744244e-05,
|
| 547 |
+
"loss": 1.8431,
|
| 548 |
+
"step": 90000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 16.62,
|
| 552 |
+
"learning_rate": 2.243998040176384e-05,
|
| 553 |
+
"loss": 1.8723,
|
| 554 |
+
"step": 91000
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 16.8,
|
| 558 |
+
"learning_rate": 2.2133757961783442e-05,
|
| 559 |
+
"loss": 1.9034,
|
| 560 |
+
"step": 92000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 16.98,
|
| 564 |
+
"learning_rate": 2.182753552180304e-05,
|
| 565 |
+
"loss": 1.9112,
|
| 566 |
+
"step": 93000
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 17.17,
|
| 570 |
+
"learning_rate": 2.1521313081822636e-05,
|
| 571 |
+
"loss": 1.7565,
|
| 572 |
+
"step": 94000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 17.35,
|
| 576 |
+
"learning_rate": 2.1215090641842237e-05,
|
| 577 |
+
"loss": 1.7656,
|
| 578 |
+
"step": 95000
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 17.53,
|
| 582 |
+
"learning_rate": 2.0908868201861834e-05,
|
| 583 |
+
"loss": 1.7856,
|
| 584 |
+
"step": 96000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 17.71,
|
| 588 |
+
"learning_rate": 2.060264576188143e-05,
|
| 589 |
+
"loss": 1.8185,
|
| 590 |
+
"step": 97000
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 17.9,
|
| 594 |
+
"learning_rate": 2.029642332190103e-05,
|
| 595 |
+
"loss": 1.8397,
|
| 596 |
+
"step": 98000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 18.08,
|
| 600 |
+
"learning_rate": 1.999020088192063e-05,
|
| 601 |
+
"loss": 1.7726,
|
| 602 |
+
"step": 99000
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 18.26,
|
| 606 |
+
"learning_rate": 1.9683978441940226e-05,
|
| 607 |
+
"loss": 1.6883,
|
| 608 |
+
"step": 100000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 18.44,
|
| 612 |
+
"learning_rate": 1.9377756001959823e-05,
|
| 613 |
+
"loss": 1.724,
|
| 614 |
+
"step": 101000
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 18.63,
|
| 618 |
+
"learning_rate": 1.9071533561979424e-05,
|
| 619 |
+
"loss": 1.751,
|
| 620 |
+
"step": 102000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 18.81,
|
| 624 |
+
"learning_rate": 1.876531112199902e-05,
|
| 625 |
+
"loss": 1.7654,
|
| 626 |
+
"step": 103000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 18.99,
|
| 630 |
+
"learning_rate": 1.8459088682018618e-05,
|
| 631 |
+
"loss": 1.7807,
|
| 632 |
+
"step": 104000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 19.17,
|
| 636 |
+
"learning_rate": 1.8152866242038215e-05,
|
| 637 |
+
"loss": 1.6287,
|
| 638 |
+
"step": 105000
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 19.36,
|
| 642 |
+
"learning_rate": 1.7846643802057816e-05,
|
| 643 |
+
"loss": 1.6529,
|
| 644 |
+
"step": 106000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 19.54,
|
| 648 |
+
"learning_rate": 1.7540421362077413e-05,
|
| 649 |
+
"loss": 1.6782,
|
| 650 |
+
"step": 107000
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 19.72,
|
| 654 |
+
"learning_rate": 1.723419892209701e-05,
|
| 655 |
+
"loss": 1.6989,
|
| 656 |
+
"step": 108000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 19.91,
|
| 660 |
+
"learning_rate": 1.692797648211661e-05,
|
| 661 |
+
"loss": 1.7147,
|
| 662 |
+
"step": 109000
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 20.09,
|
| 666 |
+
"learning_rate": 1.6621754042136208e-05,
|
| 667 |
+
"loss": 1.6426,
|
| 668 |
+
"step": 110000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 20.27,
|
| 672 |
+
"learning_rate": 1.6315531602155805e-05,
|
| 673 |
+
"loss": 1.5852,
|
| 674 |
+
"step": 111000
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 20.45,
|
| 678 |
+
"learning_rate": 1.6009309162175405e-05,
|
| 679 |
+
"loss": 1.6156,
|
| 680 |
+
"step": 112000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 20.64,
|
| 684 |
+
"learning_rate": 1.5703086722195003e-05,
|
| 685 |
+
"loss": 1.6307,
|
| 686 |
+
"step": 113000
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 20.82,
|
| 690 |
+
"learning_rate": 1.53968642822146e-05,
|
| 691 |
+
"loss": 1.6493,
|
| 692 |
+
"step": 114000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 21.0,
|
| 696 |
+
"learning_rate": 1.5090641842234199e-05,
|
| 697 |
+
"loss": 1.6623,
|
| 698 |
+
"step": 115000
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 21.18,
|
| 702 |
+
"learning_rate": 1.47844194022538e-05,
|
| 703 |
+
"loss": 1.5248,
|
| 704 |
+
"step": 116000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 21.37,
|
| 708 |
+
"learning_rate": 1.4478196962273396e-05,
|
| 709 |
+
"loss": 1.551,
|
| 710 |
+
"step": 117000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 21.55,
|
| 714 |
+
"learning_rate": 1.4171974522292993e-05,
|
| 715 |
+
"loss": 1.5738,
|
| 716 |
+
"step": 118000
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 21.73,
|
| 720 |
+
"learning_rate": 1.386575208231259e-05,
|
| 721 |
+
"loss": 1.5854,
|
| 722 |
+
"step": 119000
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 21.91,
|
| 726 |
+
"learning_rate": 1.3559529642332191e-05,
|
| 727 |
+
"loss": 1.599,
|
| 728 |
+
"step": 120000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 22.1,
|
| 732 |
+
"learning_rate": 1.3253307202351788e-05,
|
| 733 |
+
"loss": 1.5402,
|
| 734 |
+
"step": 121000
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 22.28,
|
| 738 |
+
"learning_rate": 1.2947084762371387e-05,
|
| 739 |
+
"loss": 1.486,
|
| 740 |
+
"step": 122000
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 22.46,
|
| 744 |
+
"learning_rate": 1.2640862322390986e-05,
|
| 745 |
+
"loss": 1.5188,
|
| 746 |
+
"step": 123000
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 22.64,
|
| 750 |
+
"learning_rate": 1.2334639882410585e-05,
|
| 751 |
+
"loss": 1.5442,
|
| 752 |
+
"step": 124000
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 22.83,
|
| 756 |
+
"learning_rate": 1.2028417442430182e-05,
|
| 757 |
+
"loss": 1.5483,
|
| 758 |
+
"step": 125000
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 23.01,
|
| 762 |
+
"learning_rate": 1.1722195002449781e-05,
|
| 763 |
+
"loss": 1.5501,
|
| 764 |
+
"step": 126000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 23.19,
|
| 768 |
+
"learning_rate": 1.1415972562469378e-05,
|
| 769 |
+
"loss": 1.4471,
|
| 770 |
+
"step": 127000
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 23.37,
|
| 774 |
+
"learning_rate": 1.1109750122488977e-05,
|
| 775 |
+
"loss": 1.4578,
|
| 776 |
+
"step": 128000
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 23.56,
|
| 780 |
+
"learning_rate": 1.0803527682508574e-05,
|
| 781 |
+
"loss": 1.4712,
|
| 782 |
+
"step": 129000
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 23.74,
|
| 786 |
+
"learning_rate": 1.0497305242528173e-05,
|
| 787 |
+
"loss": 1.4918,
|
| 788 |
+
"step": 130000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 23.92,
|
| 792 |
+
"learning_rate": 1.0191082802547772e-05,
|
| 793 |
+
"loss": 1.5129,
|
| 794 |
+
"step": 131000
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 24.11,
|
| 798 |
+
"learning_rate": 9.884860362567369e-06,
|
| 799 |
+
"loss": 1.4455,
|
| 800 |
+
"step": 132000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 24.29,
|
| 804 |
+
"learning_rate": 9.578637922586968e-06,
|
| 805 |
+
"loss": 1.4141,
|
| 806 |
+
"step": 133000
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 24.47,
|
| 810 |
+
"learning_rate": 9.272415482606565e-06,
|
| 811 |
+
"loss": 1.4301,
|
| 812 |
+
"step": 134000
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 24.65,
|
| 816 |
+
"learning_rate": 8.966193042626164e-06,
|
| 817 |
+
"loss": 1.4454,
|
| 818 |
+
"step": 135000
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 24.84,
|
| 822 |
+
"learning_rate": 8.659970602645761e-06,
|
| 823 |
+
"loss": 1.4599,
|
| 824 |
+
"step": 136000
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 25.02,
|
| 828 |
+
"learning_rate": 8.35374816266536e-06,
|
| 829 |
+
"loss": 1.4573,
|
| 830 |
+
"step": 137000
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 25.2,
|
| 834 |
+
"learning_rate": 8.047525722684959e-06,
|
| 835 |
+
"loss": 1.3704,
|
| 836 |
+
"step": 138000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 25.38,
|
| 840 |
+
"learning_rate": 7.741303282704558e-06,
|
| 841 |
+
"loss": 1.3845,
|
| 842 |
+
"step": 139000
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 25.57,
|
| 846 |
+
"learning_rate": 7.4350808427241555e-06,
|
| 847 |
+
"loss": 1.4078,
|
| 848 |
+
"step": 140000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 25.75,
|
| 852 |
+
"learning_rate": 7.128858402743753e-06,
|
| 853 |
+
"loss": 1.4147,
|
| 854 |
+
"step": 141000
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 25.93,
|
| 858 |
+
"learning_rate": 6.8226359627633515e-06,
|
| 859 |
+
"loss": 1.4223,
|
| 860 |
+
"step": 142000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 26.11,
|
| 864 |
+
"learning_rate": 6.5164135227829495e-06,
|
| 865 |
+
"loss": 1.3655,
|
| 866 |
+
"step": 143000
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 26.3,
|
| 870 |
+
"learning_rate": 6.210191082802548e-06,
|
| 871 |
+
"loss": 1.3528,
|
| 872 |
+
"step": 144000
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 26.48,
|
| 876 |
+
"learning_rate": 5.903968642822146e-06,
|
| 877 |
+
"loss": 1.3626,
|
| 878 |
+
"step": 145000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 26.66,
|
| 882 |
+
"learning_rate": 5.597746202841744e-06,
|
| 883 |
+
"loss": 1.3719,
|
| 884 |
+
"step": 146000
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 26.84,
|
| 888 |
+
"learning_rate": 5.291523762861342e-06,
|
| 889 |
+
"loss": 1.3808,
|
| 890 |
+
"step": 147000
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 27.03,
|
| 894 |
+
"learning_rate": 4.985301322880941e-06,
|
| 895 |
+
"loss": 1.3752,
|
| 896 |
+
"step": 148000
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 27.21,
|
| 900 |
+
"learning_rate": 4.679078882900539e-06,
|
| 901 |
+
"loss": 1.3171,
|
| 902 |
+
"step": 149000
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 27.39,
|
| 906 |
+
"learning_rate": 4.372856442920137e-06,
|
| 907 |
+
"loss": 1.3321,
|
| 908 |
+
"step": 150000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 27.57,
|
| 912 |
+
"learning_rate": 4.066634002939736e-06,
|
| 913 |
+
"loss": 1.3365,
|
| 914 |
+
"step": 151000
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 27.76,
|
| 918 |
+
"learning_rate": 3.7604115629593337e-06,
|
| 919 |
+
"loss": 1.3424,
|
| 920 |
+
"step": 152000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 27.94,
|
| 924 |
+
"learning_rate": 3.454189122978932e-06,
|
| 925 |
+
"loss": 1.3471,
|
| 926 |
+
"step": 153000
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 28.12,
|
| 930 |
+
"learning_rate": 3.14796668299853e-06,
|
| 931 |
+
"loss": 1.3095,
|
| 932 |
+
"step": 154000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 28.31,
|
| 936 |
+
"learning_rate": 2.8417442430181285e-06,
|
| 937 |
+
"loss": 1.3113,
|
| 938 |
+
"step": 155000
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 28.49,
|
| 942 |
+
"learning_rate": 2.5355218030377265e-06,
|
| 943 |
+
"loss": 1.3017,
|
| 944 |
+
"step": 156000
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 28.67,
|
| 948 |
+
"learning_rate": 2.229299363057325e-06,
|
| 949 |
+
"loss": 1.3147,
|
| 950 |
+
"step": 157000
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 28.85,
|
| 954 |
+
"learning_rate": 1.9230769230769234e-06,
|
| 955 |
+
"loss": 1.3135,
|
| 956 |
+
"step": 158000
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 29.04,
|
| 960 |
+
"learning_rate": 1.6168544830965214e-06,
|
| 961 |
+
"loss": 1.309,
|
| 962 |
+
"step": 159000
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 29.22,
|
| 966 |
+
"learning_rate": 1.3106320431161196e-06,
|
| 967 |
+
"loss": 1.2784,
|
| 968 |
+
"step": 160000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 29.4,
|
| 972 |
+
"learning_rate": 1.0044096031357178e-06,
|
| 973 |
+
"loss": 1.2903,
|
| 974 |
+
"step": 161000
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 29.58,
|
| 978 |
+
"learning_rate": 6.98187163155316e-07,
|
| 979 |
+
"loss": 1.2919,
|
| 980 |
+
"step": 162000
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 29.77,
|
| 984 |
+
"learning_rate": 3.9196472317491427e-07,
|
| 985 |
+
"loss": 1.2802,
|
| 986 |
+
"step": 163000
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 29.95,
|
| 990 |
+
"learning_rate": 8.57422831945125e-08,
|
| 991 |
+
"loss": 1.2904,
|
| 992 |
+
"step": 164000
|
| 993 |
+
}
|
| 994 |
+
],
|
| 995 |
+
"max_steps": 164280,
|
| 996 |
+
"num_train_epochs": 30,
|
| 997 |
+
"total_flos": 3.45466660974336e+17,
|
| 998 |
+
"trial_name": null,
|
| 999 |
+
"trial_params": null
|
| 1000 |
+
}
|
checkpoint-164000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0283ddbcd339cecb5d476c9f4b9397cd19c1055249bb40cd00611f944b7d5771
|
| 3 |
+
size 3311
|
checkpoint-164000/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|