Training in progress, epoch 0, checkpoint
Browse files- checkpoint-228960/config.json +28 -0
- checkpoint-228960/generation_config.json +7 -0
- checkpoint-228960/optimizer.pt +3 -0
- checkpoint-228960/pytorch_model.bin +3 -0
- checkpoint-228960/rng_state.pth +3 -0
- checkpoint-228960/scheduler.pt +3 -0
- checkpoint-228960/special_tokens_map.json +5 -0
- checkpoint-228960/tokenizer.json +0 -0
- checkpoint-228960/tokenizer_config.json +12 -0
- checkpoint-228960/trainer_state.json +2731 -0
- checkpoint-228960/training_args.bin +3 -0
checkpoint-228960/config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_remove_final_layer_norm": false,
|
| 3 |
+
"activation_function": "relu",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"OPTForCausalLM"
|
| 6 |
+
],
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"bos_token_id": 2,
|
| 9 |
+
"do_layer_norm_before": true,
|
| 10 |
+
"dropout": 0.1,
|
| 11 |
+
"enable_bias": true,
|
| 12 |
+
"eos_token_id": 2,
|
| 13 |
+
"ffn_dim": 768,
|
| 14 |
+
"hidden_size": 768,
|
| 15 |
+
"init_std": 0.02,
|
| 16 |
+
"layer_norm_elementwise_affine": true,
|
| 17 |
+
"layerdrop": 0.0,
|
| 18 |
+
"max_position_embeddings": 512,
|
| 19 |
+
"model_type": "opt",
|
| 20 |
+
"num_attention_heads": 8,
|
| 21 |
+
"num_hidden_layers": 8,
|
| 22 |
+
"pad_token_id": 1,
|
| 23 |
+
"torch_dtype": "float32",
|
| 24 |
+
"transformers_version": "4.33.3",
|
| 25 |
+
"use_cache": true,
|
| 26 |
+
"vocab_size": 50002,
|
| 27 |
+
"word_embed_proj_dim": 768
|
| 28 |
+
}
|
checkpoint-228960/generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 2,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"pad_token_id": 1,
|
| 6 |
+
"transformers_version": "4.33.3"
|
| 7 |
+
}
|
checkpoint-228960/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1389bfa93ccd2d83b2ed9bf3839e736d6af3c26ad06d17a782a9e3cab39ccc77
|
| 3 |
+
size 537476677
|
checkpoint-228960/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d8cdd0196cb97d273c42a55eceae15d6b2884c56b1a25a99fd4ab8de6c791be1
|
| 3 |
+
size 268727709
|
checkpoint-228960/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:78c700e70f9dc02c1da4d5d210fbbd709d73e7647ec58430fd8c9b48d6ef829d
|
| 3 |
+
size 14575
|
checkpoint-228960/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4e602f63a62c99957b6a86b5bd3f09322f89baaa6e0c64f4f003c399409a5ba
|
| 3 |
+
size 627
|
checkpoint-228960/special_tokens_map.json
ADDED
|
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"eos_token": "<eos>",
|
| 3 |
+
"pad_token": "<pad>",
|
| 4 |
+
"unk_token": "<unk>"
|
| 5 |
+
}
|
checkpoint-228960/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-228960/tokenizer_config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"clean_up_tokenization_spaces": true,
|
| 3 |
+
"eos_token": "<eos>",
|
| 4 |
+
"max_length": null,
|
| 5 |
+
"model_max_length": 512,
|
| 6 |
+
"pad_to_multiple_of": null,
|
| 7 |
+
"pad_token": "<pad>",
|
| 8 |
+
"pad_token_type_id": 0,
|
| 9 |
+
"padding_side": "right",
|
| 10 |
+
"tokenizer_class": "PreTrainedTokenizerFast",
|
| 11 |
+
"unk_token": "<unk>"
|
| 12 |
+
}
|
checkpoint-228960/trainer_state.json
ADDED
|
@@ -0,0 +1,2731 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 3.961202383041382,
|
| 3 |
+
"best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/rel-cl2/transformer/4/checkpoints/checkpoint-228960",
|
| 4 |
+
"epoch": 0.025000606015738065,
|
| 5 |
+
"eval_steps": 10,
|
| 6 |
+
"global_step": 228960,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0,
|
| 13 |
+
"learning_rate": 4.999998362119627e-05,
|
| 14 |
+
"loss": 11.0021,
|
| 15 |
+
"step": 1
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.0,
|
| 19 |
+
"learning_rate": 4.999161405248948e-05,
|
| 20 |
+
"loss": 6.8322,
|
| 21 |
+
"step": 512
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.0,
|
| 25 |
+
"learning_rate": 4.998322810497896e-05,
|
| 26 |
+
"loss": 6.2089,
|
| 27 |
+
"step": 1024
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.0,
|
| 31 |
+
"learning_rate": 4.997484215746844e-05,
|
| 32 |
+
"loss": 5.9977,
|
| 33 |
+
"step": 1536
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.0,
|
| 37 |
+
"learning_rate": 4.996645620995792e-05,
|
| 38 |
+
"loss": 5.8315,
|
| 39 |
+
"step": 2048
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.0,
|
| 43 |
+
"learning_rate": 4.99580702624474e-05,
|
| 44 |
+
"loss": 5.7266,
|
| 45 |
+
"step": 2560
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.0,
|
| 49 |
+
"learning_rate": 4.994968431493688e-05,
|
| 50 |
+
"loss": 5.6137,
|
| 51 |
+
"step": 3072
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.0,
|
| 55 |
+
"learning_rate": 4.994129836742636e-05,
|
| 56 |
+
"loss": 5.5416,
|
| 57 |
+
"step": 3584
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.0,
|
| 61 |
+
"learning_rate": 4.993291241991584e-05,
|
| 62 |
+
"loss": 5.4641,
|
| 63 |
+
"step": 4096
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.0,
|
| 67 |
+
"learning_rate": 4.992452647240532e-05,
|
| 68 |
+
"loss": 5.4047,
|
| 69 |
+
"step": 4608
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.0,
|
| 73 |
+
"learning_rate": 4.99161405248948e-05,
|
| 74 |
+
"loss": 5.3631,
|
| 75 |
+
"step": 5120
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.0,
|
| 79 |
+
"learning_rate": 4.990775457738428e-05,
|
| 80 |
+
"loss": 5.3173,
|
| 81 |
+
"step": 5632
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.0,
|
| 85 |
+
"learning_rate": 4.989936862987376e-05,
|
| 86 |
+
"loss": 5.2716,
|
| 87 |
+
"step": 6144
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.0,
|
| 91 |
+
"learning_rate": 4.989098268236324e-05,
|
| 92 |
+
"loss": 5.2186,
|
| 93 |
+
"step": 6656
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.0,
|
| 97 |
+
"learning_rate": 4.988259673485272e-05,
|
| 98 |
+
"loss": 5.1948,
|
| 99 |
+
"step": 7168
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.0,
|
| 103 |
+
"learning_rate": 4.98742107873422e-05,
|
| 104 |
+
"loss": 5.1472,
|
| 105 |
+
"step": 7680
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.0,
|
| 109 |
+
"learning_rate": 4.986584121863541e-05,
|
| 110 |
+
"loss": 5.1231,
|
| 111 |
+
"step": 8192
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.0,
|
| 115 |
+
"learning_rate": 4.985745527112489e-05,
|
| 116 |
+
"loss": 5.0978,
|
| 117 |
+
"step": 8704
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.0,
|
| 121 |
+
"learning_rate": 4.984906932361437e-05,
|
| 122 |
+
"loss": 5.0645,
|
| 123 |
+
"step": 9216
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.0,
|
| 127 |
+
"learning_rate": 4.984068337610385e-05,
|
| 128 |
+
"loss": 5.0348,
|
| 129 |
+
"step": 9728
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.0,
|
| 133 |
+
"learning_rate": 4.983229742859333e-05,
|
| 134 |
+
"loss": 5.0195,
|
| 135 |
+
"step": 10240
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.0,
|
| 139 |
+
"learning_rate": 4.9823927859886547e-05,
|
| 140 |
+
"loss": 4.9879,
|
| 141 |
+
"step": 10752
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.0,
|
| 145 |
+
"learning_rate": 4.9815541912376026e-05,
|
| 146 |
+
"loss": 4.9705,
|
| 147 |
+
"step": 11264
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.0,
|
| 151 |
+
"learning_rate": 4.9807155964865506e-05,
|
| 152 |
+
"loss": 4.9455,
|
| 153 |
+
"step": 11776
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.0,
|
| 157 |
+
"learning_rate": 4.9798770017354986e-05,
|
| 158 |
+
"loss": 4.9378,
|
| 159 |
+
"step": 12288
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.0,
|
| 163 |
+
"learning_rate": 4.9790384069844466e-05,
|
| 164 |
+
"loss": 4.9039,
|
| 165 |
+
"step": 12800
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.0,
|
| 169 |
+
"learning_rate": 4.9782014501137675e-05,
|
| 170 |
+
"loss": 4.8893,
|
| 171 |
+
"step": 13312
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.0,
|
| 175 |
+
"learning_rate": 4.9773628553627155e-05,
|
| 176 |
+
"loss": 4.8609,
|
| 177 |
+
"step": 13824
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.0,
|
| 181 |
+
"learning_rate": 4.9765242606116635e-05,
|
| 182 |
+
"loss": 4.8559,
|
| 183 |
+
"step": 14336
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.0,
|
| 187 |
+
"learning_rate": 4.9756856658606115e-05,
|
| 188 |
+
"loss": 4.8306,
|
| 189 |
+
"step": 14848
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.01,
|
| 193 |
+
"learning_rate": 4.9748470711095595e-05,
|
| 194 |
+
"loss": 4.8227,
|
| 195 |
+
"step": 15360
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.01,
|
| 199 |
+
"learning_rate": 4.9740101142388804e-05,
|
| 200 |
+
"loss": 4.813,
|
| 201 |
+
"step": 15872
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.01,
|
| 205 |
+
"learning_rate": 4.9731715194878284e-05,
|
| 206 |
+
"loss": 4.7958,
|
| 207 |
+
"step": 16384
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.01,
|
| 211 |
+
"learning_rate": 4.9723329247367764e-05,
|
| 212 |
+
"loss": 4.7863,
|
| 213 |
+
"step": 16896
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.01,
|
| 217 |
+
"learning_rate": 4.9714943299857244e-05,
|
| 218 |
+
"loss": 4.7715,
|
| 219 |
+
"step": 17408
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.01,
|
| 223 |
+
"learning_rate": 4.970657373115046e-05,
|
| 224 |
+
"loss": 4.7567,
|
| 225 |
+
"step": 17920
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.01,
|
| 229 |
+
"learning_rate": 4.969818778363994e-05,
|
| 230 |
+
"loss": 4.7541,
|
| 231 |
+
"step": 18432
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.01,
|
| 235 |
+
"learning_rate": 4.968981821493315e-05,
|
| 236 |
+
"loss": 4.7236,
|
| 237 |
+
"step": 18944
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.01,
|
| 241 |
+
"learning_rate": 4.968143226742263e-05,
|
| 242 |
+
"loss": 4.7143,
|
| 243 |
+
"step": 19456
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.01,
|
| 247 |
+
"learning_rate": 4.967304631991211e-05,
|
| 248 |
+
"loss": 4.6927,
|
| 249 |
+
"step": 19968
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.01,
|
| 253 |
+
"learning_rate": 4.966466037240159e-05,
|
| 254 |
+
"loss": 4.6956,
|
| 255 |
+
"step": 20480
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.01,
|
| 259 |
+
"learning_rate": 4.965627442489107e-05,
|
| 260 |
+
"loss": 4.6799,
|
| 261 |
+
"step": 20992
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.01,
|
| 265 |
+
"learning_rate": 4.964788847738054e-05,
|
| 266 |
+
"loss": 4.6817,
|
| 267 |
+
"step": 21504
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.01,
|
| 271 |
+
"learning_rate": 4.963950252987002e-05,
|
| 272 |
+
"loss": 4.661,
|
| 273 |
+
"step": 22016
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.01,
|
| 277 |
+
"learning_rate": 4.96311165823595e-05,
|
| 278 |
+
"loss": 4.6555,
|
| 279 |
+
"step": 22528
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.01,
|
| 283 |
+
"learning_rate": 4.962274701365272e-05,
|
| 284 |
+
"loss": 4.65,
|
| 285 |
+
"step": 23040
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.01,
|
| 289 |
+
"learning_rate": 4.96143610661422e-05,
|
| 290 |
+
"loss": 4.6449,
|
| 291 |
+
"step": 23552
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.01,
|
| 295 |
+
"learning_rate": 4.960597511863168e-05,
|
| 296 |
+
"loss": 4.6324,
|
| 297 |
+
"step": 24064
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.01,
|
| 301 |
+
"learning_rate": 4.959758917112116e-05,
|
| 302 |
+
"loss": 4.6142,
|
| 303 |
+
"step": 24576
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.01,
|
| 307 |
+
"learning_rate": 4.958920322361064e-05,
|
| 308 |
+
"loss": 4.6051,
|
| 309 |
+
"step": 25088
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.01,
|
| 313 |
+
"learning_rate": 4.958083365490385e-05,
|
| 314 |
+
"loss": 4.6232,
|
| 315 |
+
"step": 25600
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.01,
|
| 319 |
+
"learning_rate": 4.957244770739333e-05,
|
| 320 |
+
"loss": 4.6141,
|
| 321 |
+
"step": 26112
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.01,
|
| 325 |
+
"learning_rate": 4.956406175988281e-05,
|
| 326 |
+
"loss": 4.5911,
|
| 327 |
+
"step": 26624
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.01,
|
| 331 |
+
"learning_rate": 4.9555692191176016e-05,
|
| 332 |
+
"loss": 4.5789,
|
| 333 |
+
"step": 27136
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.01,
|
| 337 |
+
"learning_rate": 4.9547306243665496e-05,
|
| 338 |
+
"loss": 4.5871,
|
| 339 |
+
"step": 27648
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.01,
|
| 343 |
+
"learning_rate": 4.9538920296154976e-05,
|
| 344 |
+
"loss": 4.5626,
|
| 345 |
+
"step": 28160
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.01,
|
| 349 |
+
"learning_rate": 4.9530534348644456e-05,
|
| 350 |
+
"loss": 4.5747,
|
| 351 |
+
"step": 28672
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.01,
|
| 355 |
+
"learning_rate": 4.9522148401133936e-05,
|
| 356 |
+
"loss": 4.5429,
|
| 357 |
+
"step": 29184
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.01,
|
| 361 |
+
"learning_rate": 4.951377883242715e-05,
|
| 362 |
+
"loss": 4.5658,
|
| 363 |
+
"step": 29696
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.01,
|
| 367 |
+
"learning_rate": 4.950539288491663e-05,
|
| 368 |
+
"loss": 4.549,
|
| 369 |
+
"step": 30208
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.01,
|
| 373 |
+
"learning_rate": 4.949700693740611e-05,
|
| 374 |
+
"loss": 4.5287,
|
| 375 |
+
"step": 30720
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.01,
|
| 379 |
+
"learning_rate": 4.948862098989559e-05,
|
| 380 |
+
"loss": 4.5233,
|
| 381 |
+
"step": 31232
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.01,
|
| 385 |
+
"learning_rate": 4.94802514211888e-05,
|
| 386 |
+
"loss": 4.534,
|
| 387 |
+
"step": 31744
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.01,
|
| 391 |
+
"learning_rate": 4.947186547367828e-05,
|
| 392 |
+
"loss": 4.4946,
|
| 393 |
+
"step": 32256
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.01,
|
| 397 |
+
"learning_rate": 4.946347952616776e-05,
|
| 398 |
+
"loss": 4.5095,
|
| 399 |
+
"step": 32768
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.01,
|
| 403 |
+
"learning_rate": 4.945509357865724e-05,
|
| 404 |
+
"loss": 4.5155,
|
| 405 |
+
"step": 33280
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.01,
|
| 409 |
+
"learning_rate": 4.944672400995045e-05,
|
| 410 |
+
"loss": 4.5014,
|
| 411 |
+
"step": 33792
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 0.01,
|
| 415 |
+
"learning_rate": 4.943833806243993e-05,
|
| 416 |
+
"loss": 4.4933,
|
| 417 |
+
"step": 34304
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.01,
|
| 421 |
+
"learning_rate": 4.942995211492941e-05,
|
| 422 |
+
"loss": 4.4757,
|
| 423 |
+
"step": 34816
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.01,
|
| 427 |
+
"learning_rate": 4.9421582546222625e-05,
|
| 428 |
+
"loss": 4.4755,
|
| 429 |
+
"step": 35328
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.01,
|
| 433 |
+
"learning_rate": 4.9413196598712105e-05,
|
| 434 |
+
"loss": 4.472,
|
| 435 |
+
"step": 35840
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.01,
|
| 439 |
+
"learning_rate": 4.9404810651201585e-05,
|
| 440 |
+
"loss": 4.4865,
|
| 441 |
+
"step": 36352
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.01,
|
| 445 |
+
"learning_rate": 4.9396424703691065e-05,
|
| 446 |
+
"loss": 4.4634,
|
| 447 |
+
"step": 36864
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.01,
|
| 451 |
+
"learning_rate": 4.9388038756180545e-05,
|
| 452 |
+
"loss": 4.474,
|
| 453 |
+
"step": 37376
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 0.01,
|
| 457 |
+
"learning_rate": 4.9379652808670025e-05,
|
| 458 |
+
"loss": 4.4681,
|
| 459 |
+
"step": 37888
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.01,
|
| 463 |
+
"learning_rate": 4.9371266861159505e-05,
|
| 464 |
+
"loss": 4.4675,
|
| 465 |
+
"step": 38400
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.01,
|
| 469 |
+
"learning_rate": 4.9362897292452714e-05,
|
| 470 |
+
"loss": 4.4497,
|
| 471 |
+
"step": 38912
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.01,
|
| 475 |
+
"learning_rate": 4.9354511344942194e-05,
|
| 476 |
+
"loss": 4.4475,
|
| 477 |
+
"step": 39424
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.01,
|
| 481 |
+
"learning_rate": 4.9346125397431674e-05,
|
| 482 |
+
"loss": 4.444,
|
| 483 |
+
"step": 39936
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 0.01,
|
| 487 |
+
"learning_rate": 4.9337739449921154e-05,
|
| 488 |
+
"loss": 4.4412,
|
| 489 |
+
"step": 40448
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.01,
|
| 493 |
+
"learning_rate": 4.9329353502410634e-05,
|
| 494 |
+
"loss": 4.4318,
|
| 495 |
+
"step": 40960
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.01,
|
| 499 |
+
"learning_rate": 4.9320967554900114e-05,
|
| 500 |
+
"loss": 4.4261,
|
| 501 |
+
"step": 41472
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.01,
|
| 505 |
+
"learning_rate": 4.9312581607389594e-05,
|
| 506 |
+
"loss": 4.4272,
|
| 507 |
+
"step": 41984
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.01,
|
| 511 |
+
"learning_rate": 4.9304195659879074e-05,
|
| 512 |
+
"loss": 4.4262,
|
| 513 |
+
"step": 42496
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.01,
|
| 517 |
+
"learning_rate": 4.929582609117229e-05,
|
| 518 |
+
"loss": 4.4049,
|
| 519 |
+
"step": 43008
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.01,
|
| 523 |
+
"learning_rate": 4.928744014366177e-05,
|
| 524 |
+
"loss": 4.4137,
|
| 525 |
+
"step": 43520
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.01,
|
| 529 |
+
"learning_rate": 4.927905419615125e-05,
|
| 530 |
+
"loss": 4.4128,
|
| 531 |
+
"step": 44032
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.01,
|
| 535 |
+
"learning_rate": 4.927066824864073e-05,
|
| 536 |
+
"loss": 4.4128,
|
| 537 |
+
"step": 44544
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.01,
|
| 541 |
+
"learning_rate": 4.926229867993394e-05,
|
| 542 |
+
"loss": 4.3974,
|
| 543 |
+
"step": 45056
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.01,
|
| 547 |
+
"learning_rate": 4.925392911122715e-05,
|
| 548 |
+
"loss": 4.3968,
|
| 549 |
+
"step": 45568
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.02,
|
| 553 |
+
"learning_rate": 4.924554316371663e-05,
|
| 554 |
+
"loss": 4.3878,
|
| 555 |
+
"step": 46080
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.02,
|
| 559 |
+
"learning_rate": 4.923715721620611e-05,
|
| 560 |
+
"loss": 4.3828,
|
| 561 |
+
"step": 46592
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.02,
|
| 565 |
+
"learning_rate": 4.922877126869559e-05,
|
| 566 |
+
"loss": 4.3841,
|
| 567 |
+
"step": 47104
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 0.02,
|
| 571 |
+
"learning_rate": 4.922038532118507e-05,
|
| 572 |
+
"loss": 4.3874,
|
| 573 |
+
"step": 47616
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 0.02,
|
| 577 |
+
"learning_rate": 4.921199937367455e-05,
|
| 578 |
+
"loss": 4.3791,
|
| 579 |
+
"step": 48128
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.02,
|
| 583 |
+
"learning_rate": 4.920362980496776e-05,
|
| 584 |
+
"loss": 4.3883,
|
| 585 |
+
"step": 48640
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.02,
|
| 589 |
+
"learning_rate": 4.919524385745724e-05,
|
| 590 |
+
"loss": 4.3787,
|
| 591 |
+
"step": 49152
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.02,
|
| 595 |
+
"learning_rate": 4.918685790994672e-05,
|
| 596 |
+
"loss": 4.361,
|
| 597 |
+
"step": 49664
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.02,
|
| 601 |
+
"learning_rate": 4.91784719624362e-05,
|
| 602 |
+
"loss": 4.3605,
|
| 603 |
+
"step": 50176
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.02,
|
| 607 |
+
"learning_rate": 4.9170086014925676e-05,
|
| 608 |
+
"loss": 4.3702,
|
| 609 |
+
"step": 50688
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.02,
|
| 613 |
+
"learning_rate": 4.9161700067415156e-05,
|
| 614 |
+
"loss": 4.3666,
|
| 615 |
+
"step": 51200
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 0.02,
|
| 619 |
+
"learning_rate": 4.9153314119904636e-05,
|
| 620 |
+
"loss": 4.3651,
|
| 621 |
+
"step": 51712
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.02,
|
| 625 |
+
"learning_rate": 4.9144928172394116e-05,
|
| 626 |
+
"loss": 4.3532,
|
| 627 |
+
"step": 52224
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.02,
|
| 631 |
+
"learning_rate": 4.9136558603687325e-05,
|
| 632 |
+
"loss": 4.3443,
|
| 633 |
+
"step": 52736
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.02,
|
| 637 |
+
"learning_rate": 4.9128172656176805e-05,
|
| 638 |
+
"loss": 4.3571,
|
| 639 |
+
"step": 53248
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.02,
|
| 643 |
+
"learning_rate": 4.9119786708666285e-05,
|
| 644 |
+
"loss": 4.3348,
|
| 645 |
+
"step": 53760
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 0.02,
|
| 649 |
+
"learning_rate": 4.91114171399595e-05,
|
| 650 |
+
"loss": 4.3404,
|
| 651 |
+
"step": 54272
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 0.02,
|
| 655 |
+
"learning_rate": 4.910303119244898e-05,
|
| 656 |
+
"loss": 4.3354,
|
| 657 |
+
"step": 54784
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.02,
|
| 661 |
+
"learning_rate": 4.909464524493846e-05,
|
| 662 |
+
"loss": 4.3343,
|
| 663 |
+
"step": 55296
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 0.02,
|
| 667 |
+
"learning_rate": 4.908625929742794e-05,
|
| 668 |
+
"loss": 4.3249,
|
| 669 |
+
"step": 55808
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.02,
|
| 673 |
+
"learning_rate": 4.907787334991742e-05,
|
| 674 |
+
"loss": 4.3262,
|
| 675 |
+
"step": 56320
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.02,
|
| 679 |
+
"learning_rate": 4.90694874024069e-05,
|
| 680 |
+
"loss": 4.3201,
|
| 681 |
+
"step": 56832
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.02,
|
| 685 |
+
"learning_rate": 4.906110145489638e-05,
|
| 686 |
+
"loss": 4.3269,
|
| 687 |
+
"step": 57344
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 0.02,
|
| 691 |
+
"learning_rate": 4.905273188618959e-05,
|
| 692 |
+
"loss": 4.3217,
|
| 693 |
+
"step": 57856
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 0.02,
|
| 697 |
+
"learning_rate": 4.904434593867907e-05,
|
| 698 |
+
"loss": 4.3121,
|
| 699 |
+
"step": 58368
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.02,
|
| 703 |
+
"learning_rate": 4.903595999116855e-05,
|
| 704 |
+
"loss": 4.3238,
|
| 705 |
+
"step": 58880
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.02,
|
| 709 |
+
"learning_rate": 4.902757404365803e-05,
|
| 710 |
+
"loss": 4.3301,
|
| 711 |
+
"step": 59392
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.02,
|
| 715 |
+
"learning_rate": 4.901920447495124e-05,
|
| 716 |
+
"loss": 4.316,
|
| 717 |
+
"step": 59904
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.02,
|
| 721 |
+
"learning_rate": 4.901081852744072e-05,
|
| 722 |
+
"loss": 4.3032,
|
| 723 |
+
"step": 60416
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.02,
|
| 727 |
+
"learning_rate": 4.90024325799302e-05,
|
| 728 |
+
"loss": 4.3016,
|
| 729 |
+
"step": 60928
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.02,
|
| 733 |
+
"learning_rate": 4.8994046632419686e-05,
|
| 734 |
+
"loss": 4.2991,
|
| 735 |
+
"step": 61440
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.02,
|
| 739 |
+
"learning_rate": 4.8985677063712895e-05,
|
| 740 |
+
"loss": 4.3049,
|
| 741 |
+
"step": 61952
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 0.02,
|
| 745 |
+
"learning_rate": 4.8977291116202375e-05,
|
| 746 |
+
"loss": 4.3011,
|
| 747 |
+
"step": 62464
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.02,
|
| 751 |
+
"learning_rate": 4.8968905168691855e-05,
|
| 752 |
+
"loss": 4.2897,
|
| 753 |
+
"step": 62976
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.02,
|
| 757 |
+
"learning_rate": 4.8960519221181335e-05,
|
| 758 |
+
"loss": 4.2993,
|
| 759 |
+
"step": 63488
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.02,
|
| 763 |
+
"learning_rate": 4.8952133273670815e-05,
|
| 764 |
+
"loss": 4.2898,
|
| 765 |
+
"step": 64000
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.02,
|
| 769 |
+
"learning_rate": 4.8943763704964024e-05,
|
| 770 |
+
"loss": 4.2889,
|
| 771 |
+
"step": 64512
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 0.02,
|
| 775 |
+
"learning_rate": 4.8935377757453504e-05,
|
| 776 |
+
"loss": 4.2834,
|
| 777 |
+
"step": 65024
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.02,
|
| 781 |
+
"learning_rate": 4.8926991809942984e-05,
|
| 782 |
+
"loss": 4.2855,
|
| 783 |
+
"step": 65536
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.02,
|
| 787 |
+
"learning_rate": 4.8918605862432463e-05,
|
| 788 |
+
"loss": 4.2788,
|
| 789 |
+
"step": 66048
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.02,
|
| 793 |
+
"learning_rate": 4.8910219914921943e-05,
|
| 794 |
+
"loss": 4.288,
|
| 795 |
+
"step": 66560
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.02,
|
| 799 |
+
"learning_rate": 4.8901833967411423e-05,
|
| 800 |
+
"loss": 4.2854,
|
| 801 |
+
"step": 67072
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.02,
|
| 805 |
+
"learning_rate": 4.88934480199009e-05,
|
| 806 |
+
"loss": 4.2842,
|
| 807 |
+
"step": 67584
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.02,
|
| 811 |
+
"learning_rate": 4.888506207239038e-05,
|
| 812 |
+
"loss": 4.2617,
|
| 813 |
+
"step": 68096
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 0.02,
|
| 817 |
+
"learning_rate": 4.88766925036836e-05,
|
| 818 |
+
"loss": 4.2775,
|
| 819 |
+
"step": 68608
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.02,
|
| 823 |
+
"learning_rate": 4.886830655617308e-05,
|
| 824 |
+
"loss": 4.2715,
|
| 825 |
+
"step": 69120
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.02,
|
| 829 |
+
"learning_rate": 4.885992060866256e-05,
|
| 830 |
+
"loss": 4.2761,
|
| 831 |
+
"step": 69632
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.02,
|
| 835 |
+
"learning_rate": 4.885153466115204e-05,
|
| 836 |
+
"loss": 4.2636,
|
| 837 |
+
"step": 70144
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.02,
|
| 841 |
+
"learning_rate": 4.884316509244525e-05,
|
| 842 |
+
"loss": 4.2728,
|
| 843 |
+
"step": 70656
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.02,
|
| 847 |
+
"learning_rate": 4.883477914493473e-05,
|
| 848 |
+
"loss": 4.2619,
|
| 849 |
+
"step": 71168
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.02,
|
| 853 |
+
"learning_rate": 4.882639319742421e-05,
|
| 854 |
+
"loss": 4.2793,
|
| 855 |
+
"step": 71680
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 0.02,
|
| 859 |
+
"learning_rate": 4.881800724991369e-05,
|
| 860 |
+
"loss": 4.249,
|
| 861 |
+
"step": 72192
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 0.02,
|
| 865 |
+
"learning_rate": 4.880962130240316e-05,
|
| 866 |
+
"loss": 4.2662,
|
| 867 |
+
"step": 72704
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.02,
|
| 871 |
+
"learning_rate": 4.880123535489264e-05,
|
| 872 |
+
"loss": 4.2512,
|
| 873 |
+
"step": 73216
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.02,
|
| 877 |
+
"learning_rate": 4.879286578618586e-05,
|
| 878 |
+
"loss": 4.2479,
|
| 879 |
+
"step": 73728
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.02,
|
| 883 |
+
"learning_rate": 4.878447983867534e-05,
|
| 884 |
+
"loss": 4.2504,
|
| 885 |
+
"step": 74240
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.02,
|
| 889 |
+
"learning_rate": 4.877609389116482e-05,
|
| 890 |
+
"loss": 4.2386,
|
| 891 |
+
"step": 74752
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.02,
|
| 895 |
+
"learning_rate": 4.87677079436543e-05,
|
| 896 |
+
"loss": 4.2424,
|
| 897 |
+
"step": 75264
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.02,
|
| 901 |
+
"learning_rate": 4.875932199614378e-05,
|
| 902 |
+
"loss": 4.2488,
|
| 903 |
+
"step": 75776
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.02,
|
| 907 |
+
"learning_rate": 4.875093604863326e-05,
|
| 908 |
+
"loss": 4.2396,
|
| 909 |
+
"step": 76288
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.03,
|
| 913 |
+
"eval_loss": 4.205402851104736,
|
| 914 |
+
"eval_runtime": 302.2482,
|
| 915 |
+
"eval_samples_per_second": 1262.509,
|
| 916 |
+
"eval_steps_per_second": 39.454,
|
| 917 |
+
"step": 76320
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 1.0,
|
| 921 |
+
"learning_rate": 4.874255010112274e-05,
|
| 922 |
+
"loss": 4.2324,
|
| 923 |
+
"step": 76800
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 1.0,
|
| 927 |
+
"learning_rate": 4.8734180532415946e-05,
|
| 928 |
+
"loss": 4.2284,
|
| 929 |
+
"step": 77312
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 1.0,
|
| 933 |
+
"learning_rate": 4.8725794584905426e-05,
|
| 934 |
+
"loss": 4.2426,
|
| 935 |
+
"step": 77824
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.0,
|
| 939 |
+
"learning_rate": 4.8717408637394906e-05,
|
| 940 |
+
"loss": 4.2329,
|
| 941 |
+
"step": 78336
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.0,
|
| 945 |
+
"learning_rate": 4.8709022689884386e-05,
|
| 946 |
+
"loss": 4.2409,
|
| 947 |
+
"step": 78848
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.0,
|
| 951 |
+
"learning_rate": 4.8700636742373866e-05,
|
| 952 |
+
"loss": 4.22,
|
| 953 |
+
"step": 79360
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 1.0,
|
| 957 |
+
"learning_rate": 4.8692267173667075e-05,
|
| 958 |
+
"loss": 4.2222,
|
| 959 |
+
"step": 79872
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 1.0,
|
| 963 |
+
"learning_rate": 4.8683881226156555e-05,
|
| 964 |
+
"loss": 4.2103,
|
| 965 |
+
"step": 80384
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 1.0,
|
| 969 |
+
"learning_rate": 4.867549527864604e-05,
|
| 970 |
+
"loss": 4.2209,
|
| 971 |
+
"step": 80896
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 1.0,
|
| 975 |
+
"learning_rate": 4.866710933113552e-05,
|
| 976 |
+
"loss": 4.2185,
|
| 977 |
+
"step": 81408
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 1.0,
|
| 981 |
+
"learning_rate": 4.8658723383625e-05,
|
| 982 |
+
"loss": 4.2146,
|
| 983 |
+
"step": 81920
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.0,
|
| 987 |
+
"learning_rate": 4.865033743611448e-05,
|
| 988 |
+
"loss": 4.2183,
|
| 989 |
+
"step": 82432
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.0,
|
| 993 |
+
"learning_rate": 4.864196786740769e-05,
|
| 994 |
+
"loss": 4.1972,
|
| 995 |
+
"step": 82944
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 1.0,
|
| 999 |
+
"learning_rate": 4.863358191989717e-05,
|
| 1000 |
+
"loss": 4.2125,
|
| 1001 |
+
"step": 83456
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 1.0,
|
| 1005 |
+
"learning_rate": 4.862519597238665e-05,
|
| 1006 |
+
"loss": 4.197,
|
| 1007 |
+
"step": 83968
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 1.0,
|
| 1011 |
+
"learning_rate": 4.861681002487613e-05,
|
| 1012 |
+
"loss": 4.1975,
|
| 1013 |
+
"step": 84480
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 1.0,
|
| 1017 |
+
"learning_rate": 4.860842407736561e-05,
|
| 1018 |
+
"loss": 4.2034,
|
| 1019 |
+
"step": 84992
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 1.0,
|
| 1023 |
+
"learning_rate": 4.860003812985509e-05,
|
| 1024 |
+
"loss": 4.1958,
|
| 1025 |
+
"step": 85504
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.0,
|
| 1029 |
+
"learning_rate": 4.859165218234457e-05,
|
| 1030 |
+
"loss": 4.1979,
|
| 1031 |
+
"step": 86016
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.0,
|
| 1035 |
+
"learning_rate": 4.858328261363778e-05,
|
| 1036 |
+
"loss": 4.213,
|
| 1037 |
+
"step": 86528
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 1.0,
|
| 1041 |
+
"learning_rate": 4.857489666612726e-05,
|
| 1042 |
+
"loss": 4.188,
|
| 1043 |
+
"step": 87040
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 1.0,
|
| 1047 |
+
"learning_rate": 4.856651071861674e-05,
|
| 1048 |
+
"loss": 4.1929,
|
| 1049 |
+
"step": 87552
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 1.0,
|
| 1053 |
+
"learning_rate": 4.8558124771106226e-05,
|
| 1054 |
+
"loss": 4.1916,
|
| 1055 |
+
"step": 88064
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 1.0,
|
| 1059 |
+
"learning_rate": 4.8549755202399435e-05,
|
| 1060 |
+
"loss": 4.1969,
|
| 1061 |
+
"step": 88576
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.0,
|
| 1065 |
+
"learning_rate": 4.8541369254888915e-05,
|
| 1066 |
+
"loss": 4.1906,
|
| 1067 |
+
"step": 89088
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.0,
|
| 1071 |
+
"learning_rate": 4.8532983307378395e-05,
|
| 1072 |
+
"loss": 4.1887,
|
| 1073 |
+
"step": 89600
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.0,
|
| 1077 |
+
"learning_rate": 4.8524597359867875e-05,
|
| 1078 |
+
"loss": 4.1757,
|
| 1079 |
+
"step": 90112
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 1.0,
|
| 1083 |
+
"learning_rate": 4.8516227791161084e-05,
|
| 1084 |
+
"loss": 4.1824,
|
| 1085 |
+
"step": 90624
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 1.0,
|
| 1089 |
+
"learning_rate": 4.8507841843650564e-05,
|
| 1090 |
+
"loss": 4.1696,
|
| 1091 |
+
"step": 91136
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 1.01,
|
| 1095 |
+
"learning_rate": 4.8499455896140044e-05,
|
| 1096 |
+
"loss": 4.1799,
|
| 1097 |
+
"step": 91648
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 1.01,
|
| 1101 |
+
"learning_rate": 4.8491069948629524e-05,
|
| 1102 |
+
"loss": 4.1753,
|
| 1103 |
+
"step": 92160
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 1.01,
|
| 1107 |
+
"learning_rate": 4.848270037992273e-05,
|
| 1108 |
+
"loss": 4.1783,
|
| 1109 |
+
"step": 92672
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.01,
|
| 1113 |
+
"learning_rate": 4.847431443241221e-05,
|
| 1114 |
+
"loss": 4.1815,
|
| 1115 |
+
"step": 93184
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.01,
|
| 1119 |
+
"learning_rate": 4.846592848490169e-05,
|
| 1120 |
+
"loss": 4.1704,
|
| 1121 |
+
"step": 93696
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 1.01,
|
| 1125 |
+
"learning_rate": 4.845754253739117e-05,
|
| 1126 |
+
"loss": 4.1753,
|
| 1127 |
+
"step": 94208
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 1.01,
|
| 1131 |
+
"learning_rate": 4.844915658988065e-05,
|
| 1132 |
+
"loss": 4.1716,
|
| 1133 |
+
"step": 94720
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 1.01,
|
| 1137 |
+
"learning_rate": 4.844078702117387e-05,
|
| 1138 |
+
"loss": 4.1611,
|
| 1139 |
+
"step": 95232
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 1.01,
|
| 1143 |
+
"learning_rate": 4.843240107366335e-05,
|
| 1144 |
+
"loss": 4.1549,
|
| 1145 |
+
"step": 95744
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 1.01,
|
| 1149 |
+
"learning_rate": 4.842401512615282e-05,
|
| 1150 |
+
"loss": 4.147,
|
| 1151 |
+
"step": 96256
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.01,
|
| 1155 |
+
"learning_rate": 4.84156291786423e-05,
|
| 1156 |
+
"loss": 4.1548,
|
| 1157 |
+
"step": 96768
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.01,
|
| 1161 |
+
"learning_rate": 4.840724323113178e-05,
|
| 1162 |
+
"loss": 4.153,
|
| 1163 |
+
"step": 97280
|
| 1164 |
+
},
|
| 1165 |
+
{
|
| 1166 |
+
"epoch": 1.01,
|
| 1167 |
+
"learning_rate": 4.839885728362126e-05,
|
| 1168 |
+
"loss": 4.1635,
|
| 1169 |
+
"step": 97792
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 1.01,
|
| 1173 |
+
"learning_rate": 4.839047133611074e-05,
|
| 1174 |
+
"loss": 4.1494,
|
| 1175 |
+
"step": 98304
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 1.01,
|
| 1179 |
+
"learning_rate": 4.838208538860022e-05,
|
| 1180 |
+
"loss": 4.1493,
|
| 1181 |
+
"step": 98816
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 1.01,
|
| 1185 |
+
"learning_rate": 4.837371581989343e-05,
|
| 1186 |
+
"loss": 4.1528,
|
| 1187 |
+
"step": 99328
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 1.01,
|
| 1191 |
+
"learning_rate": 4.836532987238292e-05,
|
| 1192 |
+
"loss": 4.1484,
|
| 1193 |
+
"step": 99840
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.01,
|
| 1197 |
+
"learning_rate": 4.83569439248724e-05,
|
| 1198 |
+
"loss": 4.1487,
|
| 1199 |
+
"step": 100352
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.01,
|
| 1203 |
+
"learning_rate": 4.8348574356165606e-05,
|
| 1204 |
+
"loss": 4.1347,
|
| 1205 |
+
"step": 100864
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"epoch": 1.01,
|
| 1209 |
+
"learning_rate": 4.8340188408655086e-05,
|
| 1210 |
+
"loss": 4.1331,
|
| 1211 |
+
"step": 101376
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 1.01,
|
| 1215 |
+
"learning_rate": 4.8331802461144566e-05,
|
| 1216 |
+
"loss": 4.1565,
|
| 1217 |
+
"step": 101888
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 1.01,
|
| 1221 |
+
"learning_rate": 4.8323416513634046e-05,
|
| 1222 |
+
"loss": 4.1541,
|
| 1223 |
+
"step": 102400
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 1.01,
|
| 1227 |
+
"learning_rate": 4.8315030566123526e-05,
|
| 1228 |
+
"loss": 4.1367,
|
| 1229 |
+
"step": 102912
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 1.01,
|
| 1233 |
+
"learning_rate": 4.8306644618613006e-05,
|
| 1234 |
+
"loss": 4.1368,
|
| 1235 |
+
"step": 103424
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.01,
|
| 1239 |
+
"learning_rate": 4.8298275049906215e-05,
|
| 1240 |
+
"loss": 4.1413,
|
| 1241 |
+
"step": 103936
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.01,
|
| 1245 |
+
"learning_rate": 4.8289889102395695e-05,
|
| 1246 |
+
"loss": 4.1242,
|
| 1247 |
+
"step": 104448
|
| 1248 |
+
},
|
| 1249 |
+
{
|
| 1250 |
+
"epoch": 1.01,
|
| 1251 |
+
"learning_rate": 4.8281503154885175e-05,
|
| 1252 |
+
"loss": 4.1424,
|
| 1253 |
+
"step": 104960
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 1.01,
|
| 1257 |
+
"learning_rate": 4.8273117207374655e-05,
|
| 1258 |
+
"loss": 4.1169,
|
| 1259 |
+
"step": 105472
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 1.01,
|
| 1263 |
+
"learning_rate": 4.8264731259864135e-05,
|
| 1264 |
+
"loss": 4.1443,
|
| 1265 |
+
"step": 105984
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 1.01,
|
| 1269 |
+
"learning_rate": 4.8256345312353615e-05,
|
| 1270 |
+
"loss": 4.1396,
|
| 1271 |
+
"step": 106496
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 1.01,
|
| 1275 |
+
"learning_rate": 4.824797574364683e-05,
|
| 1276 |
+
"loss": 4.1207,
|
| 1277 |
+
"step": 107008
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.01,
|
| 1281 |
+
"learning_rate": 4.823958979613631e-05,
|
| 1282 |
+
"loss": 4.1167,
|
| 1283 |
+
"step": 107520
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.01,
|
| 1287 |
+
"learning_rate": 4.823120384862579e-05,
|
| 1288 |
+
"loss": 4.1344,
|
| 1289 |
+
"step": 108032
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 1.01,
|
| 1293 |
+
"learning_rate": 4.822281790111527e-05,
|
| 1294 |
+
"loss": 4.1,
|
| 1295 |
+
"step": 108544
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 1.01,
|
| 1299 |
+
"learning_rate": 4.821443195360475e-05,
|
| 1300 |
+
"loss": 4.1173,
|
| 1301 |
+
"step": 109056
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 1.01,
|
| 1305 |
+
"learning_rate": 4.820604600609423e-05,
|
| 1306 |
+
"loss": 4.1279,
|
| 1307 |
+
"step": 109568
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 1.01,
|
| 1311 |
+
"learning_rate": 4.819766005858371e-05,
|
| 1312 |
+
"loss": 4.1215,
|
| 1313 |
+
"step": 110080
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 1.01,
|
| 1317 |
+
"learning_rate": 4.8189274111073184e-05,
|
| 1318 |
+
"loss": 4.1126,
|
| 1319 |
+
"step": 110592
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.01,
|
| 1323 |
+
"learning_rate": 4.818092092117013e-05,
|
| 1324 |
+
"loss": 4.1014,
|
| 1325 |
+
"step": 111104
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.01,
|
| 1329 |
+
"learning_rate": 4.817253497365961e-05,
|
| 1330 |
+
"loss": 4.1068,
|
| 1331 |
+
"step": 111616
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 1.01,
|
| 1335 |
+
"learning_rate": 4.816414902614909e-05,
|
| 1336 |
+
"loss": 4.1081,
|
| 1337 |
+
"step": 112128
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 1.01,
|
| 1341 |
+
"learning_rate": 4.815576307863857e-05,
|
| 1342 |
+
"loss": 4.1168,
|
| 1343 |
+
"step": 112640
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 1.01,
|
| 1347 |
+
"learning_rate": 4.814737713112805e-05,
|
| 1348 |
+
"loss": 4.1066,
|
| 1349 |
+
"step": 113152
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 1.01,
|
| 1353 |
+
"learning_rate": 4.8139007562421265e-05,
|
| 1354 |
+
"loss": 4.115,
|
| 1355 |
+
"step": 113664
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 1.01,
|
| 1359 |
+
"learning_rate": 4.8130621614910745e-05,
|
| 1360 |
+
"loss": 4.1185,
|
| 1361 |
+
"step": 114176
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.01,
|
| 1365 |
+
"learning_rate": 4.8122235667400224e-05,
|
| 1366 |
+
"loss": 4.115,
|
| 1367 |
+
"step": 114688
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.01,
|
| 1371 |
+
"learning_rate": 4.8113849719889704e-05,
|
| 1372 |
+
"loss": 4.101,
|
| 1373 |
+
"step": 115200
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 1.01,
|
| 1377 |
+
"learning_rate": 4.8105480151182914e-05,
|
| 1378 |
+
"loss": 4.1077,
|
| 1379 |
+
"step": 115712
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 1.01,
|
| 1383 |
+
"learning_rate": 4.8097094203672393e-05,
|
| 1384 |
+
"loss": 4.1073,
|
| 1385 |
+
"step": 116224
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 1.01,
|
| 1389 |
+
"learning_rate": 4.8088708256161873e-05,
|
| 1390 |
+
"loss": 4.1023,
|
| 1391 |
+
"step": 116736
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 1.01,
|
| 1395 |
+
"learning_rate": 4.808032230865135e-05,
|
| 1396 |
+
"loss": 4.1028,
|
| 1397 |
+
"step": 117248
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 1.01,
|
| 1401 |
+
"learning_rate": 4.8071936361140827e-05,
|
| 1402 |
+
"loss": 4.097,
|
| 1403 |
+
"step": 117760
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.01,
|
| 1407 |
+
"learning_rate": 4.806356679243404e-05,
|
| 1408 |
+
"loss": 4.1033,
|
| 1409 |
+
"step": 118272
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.01,
|
| 1413 |
+
"learning_rate": 4.805518084492352e-05,
|
| 1414 |
+
"loss": 4.0998,
|
| 1415 |
+
"step": 118784
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 1.01,
|
| 1419 |
+
"learning_rate": 4.8046794897413e-05,
|
| 1420 |
+
"loss": 4.0842,
|
| 1421 |
+
"step": 119296
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 1.01,
|
| 1425 |
+
"learning_rate": 4.803840894990248e-05,
|
| 1426 |
+
"loss": 4.0923,
|
| 1427 |
+
"step": 119808
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 1.01,
|
| 1431 |
+
"learning_rate": 4.80300393811957e-05,
|
| 1432 |
+
"loss": 4.0941,
|
| 1433 |
+
"step": 120320
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 1.01,
|
| 1437 |
+
"learning_rate": 4.802165343368518e-05,
|
| 1438 |
+
"loss": 4.1003,
|
| 1439 |
+
"step": 120832
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 1.01,
|
| 1443 |
+
"learning_rate": 4.801326748617466e-05,
|
| 1444 |
+
"loss": 4.0889,
|
| 1445 |
+
"step": 121344
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.01,
|
| 1449 |
+
"learning_rate": 4.800488153866413e-05,
|
| 1450 |
+
"loss": 4.086,
|
| 1451 |
+
"step": 121856
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.02,
|
| 1455 |
+
"learning_rate": 4.799651196995735e-05,
|
| 1456 |
+
"loss": 4.0844,
|
| 1457 |
+
"step": 122368
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 1.02,
|
| 1461 |
+
"learning_rate": 4.798812602244683e-05,
|
| 1462 |
+
"loss": 4.0781,
|
| 1463 |
+
"step": 122880
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 1.02,
|
| 1467 |
+
"learning_rate": 4.79797400749363e-05,
|
| 1468 |
+
"loss": 4.0813,
|
| 1469 |
+
"step": 123392
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 1.02,
|
| 1473 |
+
"learning_rate": 4.797135412742578e-05,
|
| 1474 |
+
"loss": 4.0901,
|
| 1475 |
+
"step": 123904
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 1.02,
|
| 1479 |
+
"learning_rate": 4.7962984558718996e-05,
|
| 1480 |
+
"loss": 4.0861,
|
| 1481 |
+
"step": 124416
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 1.02,
|
| 1485 |
+
"learning_rate": 4.7954598611208476e-05,
|
| 1486 |
+
"loss": 4.0958,
|
| 1487 |
+
"step": 124928
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.02,
|
| 1491 |
+
"learning_rate": 4.7946212663697956e-05,
|
| 1492 |
+
"loss": 4.0858,
|
| 1493 |
+
"step": 125440
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.02,
|
| 1497 |
+
"learning_rate": 4.7937826716187436e-05,
|
| 1498 |
+
"loss": 4.0751,
|
| 1499 |
+
"step": 125952
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 1.02,
|
| 1503 |
+
"learning_rate": 4.792945714748065e-05,
|
| 1504 |
+
"loss": 4.0745,
|
| 1505 |
+
"step": 126464
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 1.02,
|
| 1509 |
+
"learning_rate": 4.792107119997013e-05,
|
| 1510 |
+
"loss": 4.0826,
|
| 1511 |
+
"step": 126976
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 1.02,
|
| 1515 |
+
"learning_rate": 4.7912685252459605e-05,
|
| 1516 |
+
"loss": 4.0864,
|
| 1517 |
+
"step": 127488
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 1.02,
|
| 1521 |
+
"learning_rate": 4.7904299304949085e-05,
|
| 1522 |
+
"loss": 4.0861,
|
| 1523 |
+
"step": 128000
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 1.02,
|
| 1527 |
+
"learning_rate": 4.78959297362423e-05,
|
| 1528 |
+
"loss": 4.0773,
|
| 1529 |
+
"step": 128512
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.02,
|
| 1533 |
+
"learning_rate": 4.7887543788731774e-05,
|
| 1534 |
+
"loss": 4.0677,
|
| 1535 |
+
"step": 129024
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.02,
|
| 1539 |
+
"learning_rate": 4.7879157841221254e-05,
|
| 1540 |
+
"loss": 4.086,
|
| 1541 |
+
"step": 129536
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 1.02,
|
| 1545 |
+
"learning_rate": 4.7870771893710734e-05,
|
| 1546 |
+
"loss": 4.0628,
|
| 1547 |
+
"step": 130048
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 1.02,
|
| 1551 |
+
"learning_rate": 4.786240232500395e-05,
|
| 1552 |
+
"loss": 4.0731,
|
| 1553 |
+
"step": 130560
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 1.02,
|
| 1557 |
+
"learning_rate": 4.785401637749343e-05,
|
| 1558 |
+
"loss": 4.0711,
|
| 1559 |
+
"step": 131072
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 1.02,
|
| 1563 |
+
"learning_rate": 4.784563042998291e-05,
|
| 1564 |
+
"loss": 4.0691,
|
| 1565 |
+
"step": 131584
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 1.02,
|
| 1569 |
+
"learning_rate": 4.783724448247239e-05,
|
| 1570 |
+
"loss": 4.0598,
|
| 1571 |
+
"step": 132096
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 1.02,
|
| 1575 |
+
"learning_rate": 4.7828874913765606e-05,
|
| 1576 |
+
"loss": 4.0688,
|
| 1577 |
+
"step": 132608
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.02,
|
| 1581 |
+
"learning_rate": 4.782048896625508e-05,
|
| 1582 |
+
"loss": 4.0594,
|
| 1583 |
+
"step": 133120
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 1.02,
|
| 1587 |
+
"learning_rate": 4.781210301874456e-05,
|
| 1588 |
+
"loss": 4.0688,
|
| 1589 |
+
"step": 133632
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 1.02,
|
| 1593 |
+
"learning_rate": 4.780371707123404e-05,
|
| 1594 |
+
"loss": 4.0635,
|
| 1595 |
+
"step": 134144
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 1.02,
|
| 1599 |
+
"learning_rate": 4.779534750252725e-05,
|
| 1600 |
+
"loss": 4.0558,
|
| 1601 |
+
"step": 134656
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 1.02,
|
| 1605 |
+
"learning_rate": 4.778696155501673e-05,
|
| 1606 |
+
"loss": 4.0766,
|
| 1607 |
+
"step": 135168
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 1.02,
|
| 1611 |
+
"learning_rate": 4.777857560750621e-05,
|
| 1612 |
+
"loss": 4.0772,
|
| 1613 |
+
"step": 135680
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 1.02,
|
| 1617 |
+
"learning_rate": 4.7770206038799423e-05,
|
| 1618 |
+
"loss": 4.0675,
|
| 1619 |
+
"step": 136192
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 1.02,
|
| 1623 |
+
"learning_rate": 4.7761820091288903e-05,
|
| 1624 |
+
"loss": 4.0579,
|
| 1625 |
+
"step": 136704
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 1.02,
|
| 1629 |
+
"learning_rate": 4.7753434143778383e-05,
|
| 1630 |
+
"loss": 4.0524,
|
| 1631 |
+
"step": 137216
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 1.02,
|
| 1635 |
+
"learning_rate": 4.774504819626786e-05,
|
| 1636 |
+
"loss": 4.0584,
|
| 1637 |
+
"step": 137728
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 1.02,
|
| 1641 |
+
"learning_rate": 4.773666224875734e-05,
|
| 1642 |
+
"loss": 4.0559,
|
| 1643 |
+
"step": 138240
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 1.02,
|
| 1647 |
+
"learning_rate": 4.772827630124682e-05,
|
| 1648 |
+
"loss": 4.0651,
|
| 1649 |
+
"step": 138752
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 1.02,
|
| 1653 |
+
"learning_rate": 4.77198903537363e-05,
|
| 1654 |
+
"loss": 4.0461,
|
| 1655 |
+
"step": 139264
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.02,
|
| 1659 |
+
"learning_rate": 4.771150440622578e-05,
|
| 1660 |
+
"loss": 4.0559,
|
| 1661 |
+
"step": 139776
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 1.02,
|
| 1665 |
+
"learning_rate": 4.770313483751899e-05,
|
| 1666 |
+
"loss": 4.0586,
|
| 1667 |
+
"step": 140288
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 1.02,
|
| 1671 |
+
"learning_rate": 4.769474889000847e-05,
|
| 1672 |
+
"loss": 4.0581,
|
| 1673 |
+
"step": 140800
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 1.02,
|
| 1677 |
+
"learning_rate": 4.768636294249795e-05,
|
| 1678 |
+
"loss": 4.0451,
|
| 1679 |
+
"step": 141312
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 1.02,
|
| 1683 |
+
"learning_rate": 4.767797699498743e-05,
|
| 1684 |
+
"loss": 4.053,
|
| 1685 |
+
"step": 141824
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 1.02,
|
| 1689 |
+
"learning_rate": 4.766960742628064e-05,
|
| 1690 |
+
"loss": 4.0514,
|
| 1691 |
+
"step": 142336
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 1.02,
|
| 1695 |
+
"learning_rate": 4.766122147877012e-05,
|
| 1696 |
+
"loss": 4.0542,
|
| 1697 |
+
"step": 142848
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 1.02,
|
| 1701 |
+
"learning_rate": 4.76528355312596e-05,
|
| 1702 |
+
"loss": 4.058,
|
| 1703 |
+
"step": 143360
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 1.02,
|
| 1707 |
+
"learning_rate": 4.764444958374909e-05,
|
| 1708 |
+
"loss": 4.0565,
|
| 1709 |
+
"step": 143872
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 1.02,
|
| 1713 |
+
"learning_rate": 4.76360800150423e-05,
|
| 1714 |
+
"loss": 4.0368,
|
| 1715 |
+
"step": 144384
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 1.02,
|
| 1719 |
+
"learning_rate": 4.762769406753178e-05,
|
| 1720 |
+
"loss": 4.051,
|
| 1721 |
+
"step": 144896
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 1.02,
|
| 1725 |
+
"learning_rate": 4.761930812002126e-05,
|
| 1726 |
+
"loss": 4.0526,
|
| 1727 |
+
"step": 145408
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 1.02,
|
| 1731 |
+
"learning_rate": 4.761092217251074e-05,
|
| 1732 |
+
"loss": 4.0494,
|
| 1733 |
+
"step": 145920
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 1.02,
|
| 1737 |
+
"learning_rate": 4.7602552603803946e-05,
|
| 1738 |
+
"loss": 4.0454,
|
| 1739 |
+
"step": 146432
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 1.02,
|
| 1743 |
+
"learning_rate": 4.7594166656293426e-05,
|
| 1744 |
+
"loss": 4.0511,
|
| 1745 |
+
"step": 146944
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 1.02,
|
| 1749 |
+
"learning_rate": 4.7585780708782906e-05,
|
| 1750 |
+
"loss": 4.0456,
|
| 1751 |
+
"step": 147456
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 1.02,
|
| 1755 |
+
"learning_rate": 4.7577394761272386e-05,
|
| 1756 |
+
"loss": 4.0609,
|
| 1757 |
+
"step": 147968
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 1.02,
|
| 1761 |
+
"learning_rate": 4.7569025192565595e-05,
|
| 1762 |
+
"loss": 4.0356,
|
| 1763 |
+
"step": 148480
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 1.02,
|
| 1767 |
+
"learning_rate": 4.7560639245055075e-05,
|
| 1768 |
+
"loss": 4.0501,
|
| 1769 |
+
"step": 148992
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 1.02,
|
| 1773 |
+
"learning_rate": 4.7552253297544555e-05,
|
| 1774 |
+
"loss": 4.0409,
|
| 1775 |
+
"step": 149504
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 1.02,
|
| 1779 |
+
"learning_rate": 4.7543867350034035e-05,
|
| 1780 |
+
"loss": 4.0349,
|
| 1781 |
+
"step": 150016
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 1.02,
|
| 1785 |
+
"learning_rate": 4.753549778132725e-05,
|
| 1786 |
+
"loss": 4.0418,
|
| 1787 |
+
"step": 150528
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 1.02,
|
| 1791 |
+
"learning_rate": 4.752711183381673e-05,
|
| 1792 |
+
"loss": 4.0326,
|
| 1793 |
+
"step": 151040
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 1.02,
|
| 1797 |
+
"learning_rate": 4.751872588630621e-05,
|
| 1798 |
+
"loss": 4.0335,
|
| 1799 |
+
"step": 151552
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 1.02,
|
| 1803 |
+
"learning_rate": 4.751033993879569e-05,
|
| 1804 |
+
"loss": 4.0357,
|
| 1805 |
+
"step": 152064
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 1.02,
|
| 1809 |
+
"learning_rate": 4.750195399128517e-05,
|
| 1810 |
+
"loss": 4.0324,
|
| 1811 |
+
"step": 152576
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 1.03,
|
| 1815 |
+
"eval_loss": 4.034352779388428,
|
| 1816 |
+
"eval_runtime": 302.1567,
|
| 1817 |
+
"eval_samples_per_second": 1262.891,
|
| 1818 |
+
"eval_steps_per_second": 39.466,
|
| 1819 |
+
"step": 152640
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"epoch": 0.0,
|
| 1823 |
+
"learning_rate": 4.749356804377465e-05,
|
| 1824 |
+
"loss": 4.0286,
|
| 1825 |
+
"step": 153088
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 0.0,
|
| 1829 |
+
"learning_rate": 4.748518209626413e-05,
|
| 1830 |
+
"loss": 4.0231,
|
| 1831 |
+
"step": 153600
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 0.0,
|
| 1835 |
+
"learning_rate": 4.747679614875361e-05,
|
| 1836 |
+
"loss": 4.0407,
|
| 1837 |
+
"step": 154112
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 0.0,
|
| 1841 |
+
"learning_rate": 4.7468410201243084e-05,
|
| 1842 |
+
"loss": 4.0303,
|
| 1843 |
+
"step": 154624
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.0,
|
| 1847 |
+
"learning_rate": 4.7460024253732563e-05,
|
| 1848 |
+
"loss": 4.0463,
|
| 1849 |
+
"step": 155136
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 0.0,
|
| 1853 |
+
"learning_rate": 4.7451638306222043e-05,
|
| 1854 |
+
"loss": 4.0213,
|
| 1855 |
+
"step": 155648
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 0.0,
|
| 1859 |
+
"learning_rate": 4.744325235871152e-05,
|
| 1860 |
+
"loss": 4.0274,
|
| 1861 |
+
"step": 156160
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 0.0,
|
| 1865 |
+
"learning_rate": 4.7434866411201e-05,
|
| 1866 |
+
"loss": 4.0121,
|
| 1867 |
+
"step": 156672
|
| 1868 |
+
},
|
| 1869 |
+
{
|
| 1870 |
+
"epoch": 0.0,
|
| 1871 |
+
"learning_rate": 4.742649684249422e-05,
|
| 1872 |
+
"loss": 4.0295,
|
| 1873 |
+
"step": 157184
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 0.0,
|
| 1877 |
+
"learning_rate": 4.74181108949837e-05,
|
| 1878 |
+
"loss": 4.0244,
|
| 1879 |
+
"step": 157696
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 0.0,
|
| 1883 |
+
"learning_rate": 4.740972494747318e-05,
|
| 1884 |
+
"loss": 4.0198,
|
| 1885 |
+
"step": 158208
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.0,
|
| 1889 |
+
"learning_rate": 4.740133899996266e-05,
|
| 1890 |
+
"loss": 4.0317,
|
| 1891 |
+
"step": 158720
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 0.0,
|
| 1895 |
+
"learning_rate": 4.739295305245214e-05,
|
| 1896 |
+
"loss": 4.0076,
|
| 1897 |
+
"step": 159232
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 0.0,
|
| 1901 |
+
"learning_rate": 4.738456710494162e-05,
|
| 1902 |
+
"loss": 4.0242,
|
| 1903 |
+
"step": 159744
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 0.0,
|
| 1907 |
+
"learning_rate": 4.73761811574311e-05,
|
| 1908 |
+
"loss": 4.0071,
|
| 1909 |
+
"step": 160256
|
| 1910 |
+
},
|
| 1911 |
+
{
|
| 1912 |
+
"epoch": 0.0,
|
| 1913 |
+
"learning_rate": 4.736779520992058e-05,
|
| 1914 |
+
"loss": 4.01,
|
| 1915 |
+
"step": 160768
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 0.0,
|
| 1919 |
+
"learning_rate": 4.735942564121379e-05,
|
| 1920 |
+
"loss": 4.0106,
|
| 1921 |
+
"step": 161280
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 0.0,
|
| 1925 |
+
"learning_rate": 4.735103969370327e-05,
|
| 1926 |
+
"loss": 4.0194,
|
| 1927 |
+
"step": 161792
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.0,
|
| 1931 |
+
"learning_rate": 4.734265374619275e-05,
|
| 1932 |
+
"loss": 4.0109,
|
| 1933 |
+
"step": 162304
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 0.0,
|
| 1937 |
+
"learning_rate": 4.733426779868223e-05,
|
| 1938 |
+
"loss": 4.0318,
|
| 1939 |
+
"step": 162816
|
| 1940 |
+
},
|
| 1941 |
+
{
|
| 1942 |
+
"epoch": 0.0,
|
| 1943 |
+
"learning_rate": 4.7325898229975444e-05,
|
| 1944 |
+
"loss": 4.0098,
|
| 1945 |
+
"step": 163328
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 0.0,
|
| 1949 |
+
"learning_rate": 4.7317512282464924e-05,
|
| 1950 |
+
"loss": 4.0137,
|
| 1951 |
+
"step": 163840
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 0.0,
|
| 1955 |
+
"learning_rate": 4.7309126334954404e-05,
|
| 1956 |
+
"loss": 4.0109,
|
| 1957 |
+
"step": 164352
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 0.0,
|
| 1961 |
+
"learning_rate": 4.7300740387443884e-05,
|
| 1962 |
+
"loss": 4.0195,
|
| 1963 |
+
"step": 164864
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 0.0,
|
| 1967 |
+
"learning_rate": 4.729237081873709e-05,
|
| 1968 |
+
"loss": 4.0086,
|
| 1969 |
+
"step": 165376
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.0,
|
| 1973 |
+
"learning_rate": 4.728398487122657e-05,
|
| 1974 |
+
"loss": 4.0136,
|
| 1975 |
+
"step": 165888
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 0.0,
|
| 1979 |
+
"learning_rate": 4.727559892371605e-05,
|
| 1980 |
+
"loss": 4.0028,
|
| 1981 |
+
"step": 166400
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 0.0,
|
| 1985 |
+
"learning_rate": 4.726721297620553e-05,
|
| 1986 |
+
"loss": 4.0067,
|
| 1987 |
+
"step": 166912
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 0.0,
|
| 1991 |
+
"learning_rate": 4.725884340749874e-05,
|
| 1992 |
+
"loss": 3.9938,
|
| 1993 |
+
"step": 167424
|
| 1994 |
+
},
|
| 1995 |
+
{
|
| 1996 |
+
"epoch": 0.01,
|
| 1997 |
+
"learning_rate": 4.725045745998822e-05,
|
| 1998 |
+
"loss": 4.0092,
|
| 1999 |
+
"step": 167936
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 0.01,
|
| 2003 |
+
"learning_rate": 4.72420715124777e-05,
|
| 2004 |
+
"loss": 4.0065,
|
| 2005 |
+
"step": 168448
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 0.01,
|
| 2009 |
+
"learning_rate": 4.723368556496718e-05,
|
| 2010 |
+
"loss": 4.0076,
|
| 2011 |
+
"step": 168960
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.01,
|
| 2015 |
+
"learning_rate": 4.72253159962604e-05,
|
| 2016 |
+
"loss": 4.0136,
|
| 2017 |
+
"step": 169472
|
| 2018 |
+
},
|
| 2019 |
+
{
|
| 2020 |
+
"epoch": 0.01,
|
| 2021 |
+
"learning_rate": 4.721693004874988e-05,
|
| 2022 |
+
"loss": 4.0018,
|
| 2023 |
+
"step": 169984
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 0.01,
|
| 2027 |
+
"learning_rate": 4.720854410123936e-05,
|
| 2028 |
+
"loss": 3.9994,
|
| 2029 |
+
"step": 170496
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 0.01,
|
| 2033 |
+
"learning_rate": 4.720015815372884e-05,
|
| 2034 |
+
"loss": 4.007,
|
| 2035 |
+
"step": 171008
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"epoch": 0.01,
|
| 2039 |
+
"learning_rate": 4.7191788585022046e-05,
|
| 2040 |
+
"loss": 3.9991,
|
| 2041 |
+
"step": 171520
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 0.01,
|
| 2045 |
+
"learning_rate": 4.7183402637511526e-05,
|
| 2046 |
+
"loss": 3.9889,
|
| 2047 |
+
"step": 172032
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 0.01,
|
| 2051 |
+
"learning_rate": 4.7175016690001006e-05,
|
| 2052 |
+
"loss": 3.9814,
|
| 2053 |
+
"step": 172544
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.01,
|
| 2057 |
+
"learning_rate": 4.7166630742490486e-05,
|
| 2058 |
+
"loss": 3.9897,
|
| 2059 |
+
"step": 173056
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 0.01,
|
| 2063 |
+
"learning_rate": 4.7158261173783695e-05,
|
| 2064 |
+
"loss": 3.9918,
|
| 2065 |
+
"step": 173568
|
| 2066 |
+
},
|
| 2067 |
+
{
|
| 2068 |
+
"epoch": 0.01,
|
| 2069 |
+
"learning_rate": 4.7149875226273175e-05,
|
| 2070 |
+
"loss": 3.9991,
|
| 2071 |
+
"step": 174080
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"epoch": 0.01,
|
| 2075 |
+
"learning_rate": 4.7141489278762655e-05,
|
| 2076 |
+
"loss": 3.9934,
|
| 2077 |
+
"step": 174592
|
| 2078 |
+
},
|
| 2079 |
+
{
|
| 2080 |
+
"epoch": 0.01,
|
| 2081 |
+
"learning_rate": 4.7133103331252135e-05,
|
| 2082 |
+
"loss": 3.9826,
|
| 2083 |
+
"step": 175104
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 0.01,
|
| 2087 |
+
"learning_rate": 4.712473376254535e-05,
|
| 2088 |
+
"loss": 3.997,
|
| 2089 |
+
"step": 175616
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 0.01,
|
| 2093 |
+
"learning_rate": 4.711634781503483e-05,
|
| 2094 |
+
"loss": 3.9898,
|
| 2095 |
+
"step": 176128
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.01,
|
| 2099 |
+
"learning_rate": 4.710796186752431e-05,
|
| 2100 |
+
"loss": 3.9865,
|
| 2101 |
+
"step": 176640
|
| 2102 |
+
},
|
| 2103 |
+
{
|
| 2104 |
+
"epoch": 0.01,
|
| 2105 |
+
"learning_rate": 4.709957592001379e-05,
|
| 2106 |
+
"loss": 3.9797,
|
| 2107 |
+
"step": 177152
|
| 2108 |
+
},
|
| 2109 |
+
{
|
| 2110 |
+
"epoch": 0.01,
|
| 2111 |
+
"learning_rate": 4.7091206351307e-05,
|
| 2112 |
+
"loss": 3.9741,
|
| 2113 |
+
"step": 177664
|
| 2114 |
+
},
|
| 2115 |
+
{
|
| 2116 |
+
"epoch": 0.01,
|
| 2117 |
+
"learning_rate": 4.708282040379648e-05,
|
| 2118 |
+
"loss": 4.0005,
|
| 2119 |
+
"step": 178176
|
| 2120 |
+
},
|
| 2121 |
+
{
|
| 2122 |
+
"epoch": 0.01,
|
| 2123 |
+
"learning_rate": 4.707443445628596e-05,
|
| 2124 |
+
"loss": 3.9971,
|
| 2125 |
+
"step": 178688
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 0.01,
|
| 2129 |
+
"learning_rate": 4.706604850877544e-05,
|
| 2130 |
+
"loss": 3.9826,
|
| 2131 |
+
"step": 179200
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 0.01,
|
| 2135 |
+
"learning_rate": 4.705767894006865e-05,
|
| 2136 |
+
"loss": 3.9823,
|
| 2137 |
+
"step": 179712
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.01,
|
| 2141 |
+
"learning_rate": 4.704930937136186e-05,
|
| 2142 |
+
"loss": 3.9911,
|
| 2143 |
+
"step": 180224
|
| 2144 |
+
},
|
| 2145 |
+
{
|
| 2146 |
+
"epoch": 0.01,
|
| 2147 |
+
"learning_rate": 4.704092342385134e-05,
|
| 2148 |
+
"loss": 3.9684,
|
| 2149 |
+
"step": 180736
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 0.01,
|
| 2153 |
+
"learning_rate": 4.703253747634082e-05,
|
| 2154 |
+
"loss": 3.9894,
|
| 2155 |
+
"step": 181248
|
| 2156 |
+
},
|
| 2157 |
+
{
|
| 2158 |
+
"epoch": 0.01,
|
| 2159 |
+
"learning_rate": 4.7024151528830305e-05,
|
| 2160 |
+
"loss": 3.9684,
|
| 2161 |
+
"step": 181760
|
| 2162 |
+
},
|
| 2163 |
+
{
|
| 2164 |
+
"epoch": 0.01,
|
| 2165 |
+
"learning_rate": 4.7015765581319785e-05,
|
| 2166 |
+
"loss": 3.9887,
|
| 2167 |
+
"step": 182272
|
| 2168 |
+
},
|
| 2169 |
+
{
|
| 2170 |
+
"epoch": 0.01,
|
| 2171 |
+
"learning_rate": 4.7007379633809265e-05,
|
| 2172 |
+
"loss": 3.9947,
|
| 2173 |
+
"step": 182784
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 0.01,
|
| 2177 |
+
"learning_rate": 4.6998993686298745e-05,
|
| 2178 |
+
"loss": 3.9747,
|
| 2179 |
+
"step": 183296
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.01,
|
| 2183 |
+
"learning_rate": 4.6990624117591954e-05,
|
| 2184 |
+
"loss": 3.9653,
|
| 2185 |
+
"step": 183808
|
| 2186 |
+
},
|
| 2187 |
+
{
|
| 2188 |
+
"epoch": 0.01,
|
| 2189 |
+
"learning_rate": 4.6982238170081434e-05,
|
| 2190 |
+
"loss": 3.9869,
|
| 2191 |
+
"step": 184320
|
| 2192 |
+
},
|
| 2193 |
+
{
|
| 2194 |
+
"epoch": 0.01,
|
| 2195 |
+
"learning_rate": 4.6973852222570914e-05,
|
| 2196 |
+
"loss": 3.9578,
|
| 2197 |
+
"step": 184832
|
| 2198 |
+
},
|
| 2199 |
+
{
|
| 2200 |
+
"epoch": 0.01,
|
| 2201 |
+
"learning_rate": 4.6965466275060394e-05,
|
| 2202 |
+
"loss": 3.9714,
|
| 2203 |
+
"step": 185344
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 0.01,
|
| 2207 |
+
"learning_rate": 4.69570967063536e-05,
|
| 2208 |
+
"loss": 3.9798,
|
| 2209 |
+
"step": 185856
|
| 2210 |
+
},
|
| 2211 |
+
{
|
| 2212 |
+
"epoch": 0.01,
|
| 2213 |
+
"learning_rate": 4.694871075884308e-05,
|
| 2214 |
+
"loss": 3.9792,
|
| 2215 |
+
"step": 186368
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 0.01,
|
| 2219 |
+
"learning_rate": 4.694032481133256e-05,
|
| 2220 |
+
"loss": 3.9716,
|
| 2221 |
+
"step": 186880
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.01,
|
| 2225 |
+
"learning_rate": 4.693193886382204e-05,
|
| 2226 |
+
"loss": 3.9576,
|
| 2227 |
+
"step": 187392
|
| 2228 |
+
},
|
| 2229 |
+
{
|
| 2230 |
+
"epoch": 0.01,
|
| 2231 |
+
"learning_rate": 4.692356929511526e-05,
|
| 2232 |
+
"loss": 3.9612,
|
| 2233 |
+
"step": 187904
|
| 2234 |
+
},
|
| 2235 |
+
{
|
| 2236 |
+
"epoch": 0.01,
|
| 2237 |
+
"learning_rate": 4.691518334760474e-05,
|
| 2238 |
+
"loss": 3.9644,
|
| 2239 |
+
"step": 188416
|
| 2240 |
+
},
|
| 2241 |
+
{
|
| 2242 |
+
"epoch": 0.01,
|
| 2243 |
+
"learning_rate": 4.690679740009422e-05,
|
| 2244 |
+
"loss": 3.9755,
|
| 2245 |
+
"step": 188928
|
| 2246 |
+
},
|
| 2247 |
+
{
|
| 2248 |
+
"epoch": 0.01,
|
| 2249 |
+
"learning_rate": 4.689841145258369e-05,
|
| 2250 |
+
"loss": 3.9638,
|
| 2251 |
+
"step": 189440
|
| 2252 |
+
},
|
| 2253 |
+
{
|
| 2254 |
+
"epoch": 0.01,
|
| 2255 |
+
"learning_rate": 4.689004188387691e-05,
|
| 2256 |
+
"loss": 3.9758,
|
| 2257 |
+
"step": 189952
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 0.01,
|
| 2261 |
+
"learning_rate": 4.688165593636639e-05,
|
| 2262 |
+
"loss": 3.9772,
|
| 2263 |
+
"step": 190464
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.01,
|
| 2267 |
+
"learning_rate": 4.687326998885587e-05,
|
| 2268 |
+
"loss": 3.9766,
|
| 2269 |
+
"step": 190976
|
| 2270 |
+
},
|
| 2271 |
+
{
|
| 2272 |
+
"epoch": 0.01,
|
| 2273 |
+
"learning_rate": 4.686488404134534e-05,
|
| 2274 |
+
"loss": 3.9605,
|
| 2275 |
+
"step": 191488
|
| 2276 |
+
},
|
| 2277 |
+
{
|
| 2278 |
+
"epoch": 0.01,
|
| 2279 |
+
"learning_rate": 4.6856514472638556e-05,
|
| 2280 |
+
"loss": 3.9698,
|
| 2281 |
+
"step": 192000
|
| 2282 |
+
},
|
| 2283 |
+
{
|
| 2284 |
+
"epoch": 0.01,
|
| 2285 |
+
"learning_rate": 4.6848128525128036e-05,
|
| 2286 |
+
"loss": 3.973,
|
| 2287 |
+
"step": 192512
|
| 2288 |
+
},
|
| 2289 |
+
{
|
| 2290 |
+
"epoch": 0.01,
|
| 2291 |
+
"learning_rate": 4.683974257761751e-05,
|
| 2292 |
+
"loss": 3.9625,
|
| 2293 |
+
"step": 193024
|
| 2294 |
+
},
|
| 2295 |
+
{
|
| 2296 |
+
"epoch": 0.01,
|
| 2297 |
+
"learning_rate": 4.6831356630106996e-05,
|
| 2298 |
+
"loss": 3.9665,
|
| 2299 |
+
"step": 193536
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 0.01,
|
| 2303 |
+
"learning_rate": 4.6822987061400205e-05,
|
| 2304 |
+
"loss": 3.9644,
|
| 2305 |
+
"step": 194048
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.01,
|
| 2309 |
+
"learning_rate": 4.681460111388969e-05,
|
| 2310 |
+
"loss": 3.9622,
|
| 2311 |
+
"step": 194560
|
| 2312 |
+
},
|
| 2313 |
+
{
|
| 2314 |
+
"epoch": 0.01,
|
| 2315 |
+
"learning_rate": 4.6806215166379165e-05,
|
| 2316 |
+
"loss": 3.9669,
|
| 2317 |
+
"step": 195072
|
| 2318 |
+
},
|
| 2319 |
+
{
|
| 2320 |
+
"epoch": 0.01,
|
| 2321 |
+
"learning_rate": 4.6797829218868645e-05,
|
| 2322 |
+
"loss": 3.9485,
|
| 2323 |
+
"step": 195584
|
| 2324 |
+
},
|
| 2325 |
+
{
|
| 2326 |
+
"epoch": 0.01,
|
| 2327 |
+
"learning_rate": 4.678945965016186e-05,
|
| 2328 |
+
"loss": 3.9605,
|
| 2329 |
+
"step": 196096
|
| 2330 |
+
},
|
| 2331 |
+
{
|
| 2332 |
+
"epoch": 0.01,
|
| 2333 |
+
"learning_rate": 4.678107370265134e-05,
|
| 2334 |
+
"loss": 3.9588,
|
| 2335 |
+
"step": 196608
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 0.01,
|
| 2339 |
+
"learning_rate": 4.6772687755140814e-05,
|
| 2340 |
+
"loss": 3.9662,
|
| 2341 |
+
"step": 197120
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 0.01,
|
| 2345 |
+
"learning_rate": 4.6764301807630294e-05,
|
| 2346 |
+
"loss": 3.9569,
|
| 2347 |
+
"step": 197632
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.01,
|
| 2351 |
+
"learning_rate": 4.675593223892351e-05,
|
| 2352 |
+
"loss": 3.9549,
|
| 2353 |
+
"step": 198144
|
| 2354 |
+
},
|
| 2355 |
+
{
|
| 2356 |
+
"epoch": 0.02,
|
| 2357 |
+
"learning_rate": 4.674754629141298e-05,
|
| 2358 |
+
"loss": 3.9548,
|
| 2359 |
+
"step": 198656
|
| 2360 |
+
},
|
| 2361 |
+
{
|
| 2362 |
+
"epoch": 0.02,
|
| 2363 |
+
"learning_rate": 4.673916034390246e-05,
|
| 2364 |
+
"loss": 3.9493,
|
| 2365 |
+
"step": 199168
|
| 2366 |
+
},
|
| 2367 |
+
{
|
| 2368 |
+
"epoch": 0.02,
|
| 2369 |
+
"learning_rate": 4.673079077519568e-05,
|
| 2370 |
+
"loss": 3.9467,
|
| 2371 |
+
"step": 199680
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 0.02,
|
| 2375 |
+
"learning_rate": 4.672240482768516e-05,
|
| 2376 |
+
"loss": 3.9569,
|
| 2377 |
+
"step": 200192
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 0.02,
|
| 2381 |
+
"learning_rate": 4.671401888017464e-05,
|
| 2382 |
+
"loss": 3.9674,
|
| 2383 |
+
"step": 200704
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 0.02,
|
| 2387 |
+
"learning_rate": 4.670563293266412e-05,
|
| 2388 |
+
"loss": 3.9633,
|
| 2389 |
+
"step": 201216
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.02,
|
| 2393 |
+
"learning_rate": 4.66972469851536e-05,
|
| 2394 |
+
"loss": 3.9546,
|
| 2395 |
+
"step": 201728
|
| 2396 |
+
},
|
| 2397 |
+
{
|
| 2398 |
+
"epoch": 0.02,
|
| 2399 |
+
"learning_rate": 4.668886103764308e-05,
|
| 2400 |
+
"loss": 3.9485,
|
| 2401 |
+
"step": 202240
|
| 2402 |
+
},
|
| 2403 |
+
{
|
| 2404 |
+
"epoch": 0.02,
|
| 2405 |
+
"learning_rate": 4.668047509013256e-05,
|
| 2406 |
+
"loss": 3.9502,
|
| 2407 |
+
"step": 202752
|
| 2408 |
+
},
|
| 2409 |
+
{
|
| 2410 |
+
"epoch": 0.02,
|
| 2411 |
+
"learning_rate": 4.667208914262204e-05,
|
| 2412 |
+
"loss": 3.9523,
|
| 2413 |
+
"step": 203264
|
| 2414 |
+
},
|
| 2415 |
+
{
|
| 2416 |
+
"epoch": 0.02,
|
| 2417 |
+
"learning_rate": 4.666371957391525e-05,
|
| 2418 |
+
"loss": 3.9599,
|
| 2419 |
+
"step": 203776
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"epoch": 0.02,
|
| 2423 |
+
"learning_rate": 4.665533362640473e-05,
|
| 2424 |
+
"loss": 3.9587,
|
| 2425 |
+
"step": 204288
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 0.02,
|
| 2429 |
+
"learning_rate": 4.664694767889421e-05,
|
| 2430 |
+
"loss": 3.9512,
|
| 2431 |
+
"step": 204800
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 0.02,
|
| 2435 |
+
"learning_rate": 4.663856173138369e-05,
|
| 2436 |
+
"loss": 3.9427,
|
| 2437 |
+
"step": 205312
|
| 2438 |
+
},
|
| 2439 |
+
{
|
| 2440 |
+
"epoch": 0.02,
|
| 2441 |
+
"learning_rate": 4.663017578387317e-05,
|
| 2442 |
+
"loss": 3.9631,
|
| 2443 |
+
"step": 205824
|
| 2444 |
+
},
|
| 2445 |
+
{
|
| 2446 |
+
"epoch": 0.02,
|
| 2447 |
+
"learning_rate": 4.662178983636265e-05,
|
| 2448 |
+
"loss": 3.9399,
|
| 2449 |
+
"step": 206336
|
| 2450 |
+
},
|
| 2451 |
+
{
|
| 2452 |
+
"epoch": 0.02,
|
| 2453 |
+
"learning_rate": 4.661340388885213e-05,
|
| 2454 |
+
"loss": 3.951,
|
| 2455 |
+
"step": 206848
|
| 2456 |
+
},
|
| 2457 |
+
{
|
| 2458 |
+
"epoch": 0.02,
|
| 2459 |
+
"learning_rate": 4.660503432014534e-05,
|
| 2460 |
+
"loss": 3.9429,
|
| 2461 |
+
"step": 207360
|
| 2462 |
+
},
|
| 2463 |
+
{
|
| 2464 |
+
"epoch": 0.02,
|
| 2465 |
+
"learning_rate": 4.659664837263482e-05,
|
| 2466 |
+
"loss": 3.9472,
|
| 2467 |
+
"step": 207872
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 0.02,
|
| 2471 |
+
"learning_rate": 4.65882624251243e-05,
|
| 2472 |
+
"loss": 3.9371,
|
| 2473 |
+
"step": 208384
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 0.02,
|
| 2477 |
+
"learning_rate": 4.657987647761378e-05,
|
| 2478 |
+
"loss": 3.9481,
|
| 2479 |
+
"step": 208896
|
| 2480 |
+
},
|
| 2481 |
+
{
|
| 2482 |
+
"epoch": 0.02,
|
| 2483 |
+
"learning_rate": 4.657149053010326e-05,
|
| 2484 |
+
"loss": 3.9401,
|
| 2485 |
+
"step": 209408
|
| 2486 |
+
},
|
| 2487 |
+
{
|
| 2488 |
+
"epoch": 0.02,
|
| 2489 |
+
"learning_rate": 4.656310458259274e-05,
|
| 2490 |
+
"loss": 3.9411,
|
| 2491 |
+
"step": 209920
|
| 2492 |
+
},
|
| 2493 |
+
{
|
| 2494 |
+
"epoch": 0.02,
|
| 2495 |
+
"learning_rate": 4.655473501388595e-05,
|
| 2496 |
+
"loss": 3.9507,
|
| 2497 |
+
"step": 210432
|
| 2498 |
+
},
|
| 2499 |
+
{
|
| 2500 |
+
"epoch": 0.02,
|
| 2501 |
+
"learning_rate": 4.654634906637543e-05,
|
| 2502 |
+
"loss": 3.9335,
|
| 2503 |
+
"step": 210944
|
| 2504 |
+
},
|
| 2505 |
+
{
|
| 2506 |
+
"epoch": 0.02,
|
| 2507 |
+
"learning_rate": 4.653796311886491e-05,
|
| 2508 |
+
"loss": 3.9572,
|
| 2509 |
+
"step": 211456
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 0.02,
|
| 2513 |
+
"learning_rate": 4.652957717135439e-05,
|
| 2514 |
+
"loss": 3.9556,
|
| 2515 |
+
"step": 211968
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 0.02,
|
| 2519 |
+
"learning_rate": 4.65212076026476e-05,
|
| 2520 |
+
"loss": 3.9481,
|
| 2521 |
+
"step": 212480
|
| 2522 |
+
},
|
| 2523 |
+
{
|
| 2524 |
+
"epoch": 0.02,
|
| 2525 |
+
"learning_rate": 4.651282165513708e-05,
|
| 2526 |
+
"loss": 3.9392,
|
| 2527 |
+
"step": 212992
|
| 2528 |
+
},
|
| 2529 |
+
{
|
| 2530 |
+
"epoch": 0.02,
|
| 2531 |
+
"learning_rate": 4.650443570762657e-05,
|
| 2532 |
+
"loss": 3.9359,
|
| 2533 |
+
"step": 213504
|
| 2534 |
+
},
|
| 2535 |
+
{
|
| 2536 |
+
"epoch": 0.02,
|
| 2537 |
+
"learning_rate": 4.649604976011605e-05,
|
| 2538 |
+
"loss": 3.9391,
|
| 2539 |
+
"step": 214016
|
| 2540 |
+
},
|
| 2541 |
+
{
|
| 2542 |
+
"epoch": 0.02,
|
| 2543 |
+
"learning_rate": 4.648768019140926e-05,
|
| 2544 |
+
"loss": 3.9381,
|
| 2545 |
+
"step": 214528
|
| 2546 |
+
},
|
| 2547 |
+
{
|
| 2548 |
+
"epoch": 0.02,
|
| 2549 |
+
"learning_rate": 4.647929424389874e-05,
|
| 2550 |
+
"loss": 3.9453,
|
| 2551 |
+
"step": 215040
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 0.02,
|
| 2555 |
+
"learning_rate": 4.647090829638822e-05,
|
| 2556 |
+
"loss": 3.9282,
|
| 2557 |
+
"step": 215552
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 0.02,
|
| 2561 |
+
"learning_rate": 4.64625223488777e-05,
|
| 2562 |
+
"loss": 3.9405,
|
| 2563 |
+
"step": 216064
|
| 2564 |
+
},
|
| 2565 |
+
{
|
| 2566 |
+
"epoch": 0.02,
|
| 2567 |
+
"learning_rate": 4.645413640136718e-05,
|
| 2568 |
+
"loss": 3.9443,
|
| 2569 |
+
"step": 216576
|
| 2570 |
+
},
|
| 2571 |
+
{
|
| 2572 |
+
"epoch": 0.02,
|
| 2573 |
+
"learning_rate": 4.644575045385665e-05,
|
| 2574 |
+
"loss": 3.9426,
|
| 2575 |
+
"step": 217088
|
| 2576 |
+
},
|
| 2577 |
+
{
|
| 2578 |
+
"epoch": 0.02,
|
| 2579 |
+
"learning_rate": 4.643736450634613e-05,
|
| 2580 |
+
"loss": 3.9308,
|
| 2581 |
+
"step": 217600
|
| 2582 |
+
},
|
| 2583 |
+
{
|
| 2584 |
+
"epoch": 0.02,
|
| 2585 |
+
"learning_rate": 4.6428994937639346e-05,
|
| 2586 |
+
"loss": 3.9364,
|
| 2587 |
+
"step": 218112
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 0.02,
|
| 2591 |
+
"learning_rate": 4.642060899012882e-05,
|
| 2592 |
+
"loss": 3.9383,
|
| 2593 |
+
"step": 218624
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 0.02,
|
| 2597 |
+
"learning_rate": 4.6412223042618306e-05,
|
| 2598 |
+
"loss": 3.9402,
|
| 2599 |
+
"step": 219136
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 0.02,
|
| 2603 |
+
"learning_rate": 4.6403837095107786e-05,
|
| 2604 |
+
"loss": 3.9401,
|
| 2605 |
+
"step": 219648
|
| 2606 |
+
},
|
| 2607 |
+
{
|
| 2608 |
+
"epoch": 0.02,
|
| 2609 |
+
"learning_rate": 4.6395451147597266e-05,
|
| 2610 |
+
"loss": 3.9456,
|
| 2611 |
+
"step": 220160
|
| 2612 |
+
},
|
| 2613 |
+
{
|
| 2614 |
+
"epoch": 0.02,
|
| 2615 |
+
"learning_rate": 4.6387065200086746e-05,
|
| 2616 |
+
"loss": 3.9227,
|
| 2617 |
+
"step": 220672
|
| 2618 |
+
},
|
| 2619 |
+
{
|
| 2620 |
+
"epoch": 0.02,
|
| 2621 |
+
"learning_rate": 4.6378679252576226e-05,
|
| 2622 |
+
"loss": 3.938,
|
| 2623 |
+
"step": 221184
|
| 2624 |
+
},
|
| 2625 |
+
{
|
| 2626 |
+
"epoch": 0.02,
|
| 2627 |
+
"learning_rate": 4.6370309683869435e-05,
|
| 2628 |
+
"loss": 3.9415,
|
| 2629 |
+
"step": 221696
|
| 2630 |
+
},
|
| 2631 |
+
{
|
| 2632 |
+
"epoch": 0.02,
|
| 2633 |
+
"learning_rate": 4.6361923736358915e-05,
|
| 2634 |
+
"loss": 3.935,
|
| 2635 |
+
"step": 222208
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 0.02,
|
| 2639 |
+
"learning_rate": 4.6353537788848395e-05,
|
| 2640 |
+
"loss": 3.9334,
|
| 2641 |
+
"step": 222720
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 0.02,
|
| 2645 |
+
"learning_rate": 4.6345151841337874e-05,
|
| 2646 |
+
"loss": 3.943,
|
| 2647 |
+
"step": 223232
|
| 2648 |
+
},
|
| 2649 |
+
{
|
| 2650 |
+
"epoch": 0.02,
|
| 2651 |
+
"learning_rate": 4.6336782272631084e-05,
|
| 2652 |
+
"loss": 3.935,
|
| 2653 |
+
"step": 223744
|
| 2654 |
+
},
|
| 2655 |
+
{
|
| 2656 |
+
"epoch": 0.02,
|
| 2657 |
+
"learning_rate": 4.6328396325120564e-05,
|
| 2658 |
+
"loss": 3.9494,
|
| 2659 |
+
"step": 224256
|
| 2660 |
+
},
|
| 2661 |
+
{
|
| 2662 |
+
"epoch": 0.02,
|
| 2663 |
+
"learning_rate": 4.6320010377610043e-05,
|
| 2664 |
+
"loss": 3.9258,
|
| 2665 |
+
"step": 224768
|
| 2666 |
+
},
|
| 2667 |
+
{
|
| 2668 |
+
"epoch": 0.02,
|
| 2669 |
+
"learning_rate": 4.6311624430099523e-05,
|
| 2670 |
+
"loss": 3.9347,
|
| 2671 |
+
"step": 225280
|
| 2672 |
+
},
|
| 2673 |
+
{
|
| 2674 |
+
"epoch": 0.02,
|
| 2675 |
+
"learning_rate": 4.630325486139274e-05,
|
| 2676 |
+
"loss": 3.9342,
|
| 2677 |
+
"step": 225792
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 0.02,
|
| 2681 |
+
"learning_rate": 4.629486891388222e-05,
|
| 2682 |
+
"loss": 3.9267,
|
| 2683 |
+
"step": 226304
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.02,
|
| 2687 |
+
"learning_rate": 4.628649934517543e-05,
|
| 2688 |
+
"loss": 3.9313,
|
| 2689 |
+
"step": 226816
|
| 2690 |
+
},
|
| 2691 |
+
{
|
| 2692 |
+
"epoch": 0.02,
|
| 2693 |
+
"learning_rate": 4.627811339766491e-05,
|
| 2694 |
+
"loss": 3.926,
|
| 2695 |
+
"step": 227328
|
| 2696 |
+
},
|
| 2697 |
+
{
|
| 2698 |
+
"epoch": 0.02,
|
| 2699 |
+
"learning_rate": 4.626972745015439e-05,
|
| 2700 |
+
"loss": 3.9223,
|
| 2701 |
+
"step": 227840
|
| 2702 |
+
},
|
| 2703 |
+
{
|
| 2704 |
+
"epoch": 0.02,
|
| 2705 |
+
"learning_rate": 4.626134150264387e-05,
|
| 2706 |
+
"loss": 3.9272,
|
| 2707 |
+
"step": 228352
|
| 2708 |
+
},
|
| 2709 |
+
{
|
| 2710 |
+
"epoch": 0.02,
|
| 2711 |
+
"learning_rate": 4.625295555513335e-05,
|
| 2712 |
+
"loss": 3.9268,
|
| 2713 |
+
"step": 228864
|
| 2714 |
+
},
|
| 2715 |
+
{
|
| 2716 |
+
"epoch": 0.03,
|
| 2717 |
+
"eval_loss": 3.961202383041382,
|
| 2718 |
+
"eval_runtime": 302.9755,
|
| 2719 |
+
"eval_samples_per_second": 1259.478,
|
| 2720 |
+
"eval_steps_per_second": 39.36,
|
| 2721 |
+
"step": 228960
|
| 2722 |
+
}
|
| 2723 |
+
],
|
| 2724 |
+
"logging_steps": 512,
|
| 2725 |
+
"max_steps": 3052726,
|
| 2726 |
+
"num_train_epochs": 9223372036854775807,
|
| 2727 |
+
"save_steps": 10,
|
| 2728 |
+
"total_flos": 1.563515463083858e+17,
|
| 2729 |
+
"trial_name": null,
|
| 2730 |
+
"trial_params": null
|
| 2731 |
+
}
|
checkpoint-228960/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:119a39cddde44478d69c3f16e0bca4bf55e2742c4863b38c5e93c30f15aa5332
|
| 3 |
+
size 4219
|