Commit
•
dcec489
1
Parent(s):
1f979bf
Upload 60 files
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/README.md +21 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/adapter_config.json +28 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/adapter_model.bin +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/added_tokens.json +7 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/optimizer.pt +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/rng_state.pth +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/scheduler.pt +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/special_tokens_map.json +7 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/tokenizer_config.json +54 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/trainer_state.json +619 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/training_args.bin +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/vocab.txt +33 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/README.md +21 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/adapter_config.json +28 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/adapter_model.bin +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/added_tokens.json +7 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/optimizer.pt +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/rng_state.pth +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/scheduler.pt +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/special_tokens_map.json +7 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/tokenizer_config.json +54 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/trainer_state.json +1219 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/training_args.bin +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/vocab.txt +33 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/README.md +21 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/adapter_config.json +28 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/adapter_model.bin +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/added_tokens.json +7 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/optimizer.pt +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/rng_state.pth +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/scheduler.pt +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/special_tokens_map.json +7 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/tokenizer_config.json +54 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/trainer_state.json +1819 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/training_args.bin +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/vocab.txt +33 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/README.md +21 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/adapter_config.json +28 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/adapter_model.bin +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/added_tokens.json +7 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/optimizer.pt +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/rng_state.pth +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/scheduler.pt +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/special_tokens_map.json +7 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/tokenizer_config.json +54 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/trainer_state.json +2419 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/training_args.bin +3 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/vocab.txt +33 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-50000/README.md +21 -0
- esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-50000/adapter_config.json +28 -0
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/README.md
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
| 4 |
+
## Training procedure
|
| 5 |
+
|
| 6 |
+
|
| 7 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 8 |
+
- quant_method: bitsandbytes
|
| 9 |
+
- load_in_8bit: False
|
| 10 |
+
- load_in_4bit: True
|
| 11 |
+
- llm_int8_threshold: 6.0
|
| 12 |
+
- llm_int8_skip_modules: None
|
| 13 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 14 |
+
- llm_int8_has_fp16_weight: False
|
| 15 |
+
- bnb_4bit_quant_type: nf4
|
| 16 |
+
- bnb_4bit_use_double_quant: True
|
| 17 |
+
- bnb_4bit_compute_dtype: bfloat16
|
| 18 |
+
### Framework versions
|
| 19 |
+
|
| 20 |
+
|
| 21 |
+
- PEFT 0.6.0.dev0
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/adapter_config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "facebook/esm2_t33_650M_UR50D",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"fan_in_fan_out": false,
|
| 7 |
+
"inference_mode": true,
|
| 8 |
+
"init_lora_weights": true,
|
| 9 |
+
"layers_pattern": null,
|
| 10 |
+
"layers_to_transform": null,
|
| 11 |
+
"lora_alpha": 1,
|
| 12 |
+
"lora_dropout": 0.5,
|
| 13 |
+
"modules_to_save": null,
|
| 14 |
+
"peft_type": "LORA",
|
| 15 |
+
"r": 2,
|
| 16 |
+
"rank_pattern": {},
|
| 17 |
+
"revision": null,
|
| 18 |
+
"target_modules": [
|
| 19 |
+
"query",
|
| 20 |
+
"key",
|
| 21 |
+
"value",
|
| 22 |
+
"EsmSelfOutput.dense",
|
| 23 |
+
"EsmIntermediate.dense",
|
| 24 |
+
"EsmOutput.dense",
|
| 25 |
+
"classifier"
|
| 26 |
+
],
|
| 27 |
+
"task_type": "TOKEN_CLS"
|
| 28 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b4ed50a68c549fed28f1e2f2bc5b0c8a52fabaa8186ad7f5394582d716aee2ba
|
| 3 |
+
size 2133117
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/added_tokens.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<cls>": 0,
|
| 3 |
+
"<eos>": 2,
|
| 4 |
+
"<mask>": 32,
|
| 5 |
+
"<pad>": 1,
|
| 6 |
+
"<unk>": 3
|
| 7 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aa2974e9f434d3dc3a520d6f9eaa69474d0b34f5f18d4d7d13c9517f1715f00e
|
| 3 |
+
size 4215045
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b035daa3bac4ead994af5800c9d0816bbd69afcbb567b553cd55a2e6ace5c861
|
| 3 |
+
size 14575
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8ef2c8960e6680e2c188e8caf9b0a9a659698c51b86a9fe73ed60cafe81dd580
|
| 3 |
+
size 627
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "<cls>",
|
| 3 |
+
"eos_token": "<eos>",
|
| 4 |
+
"mask_token": "<mask>",
|
| 5 |
+
"pad_token": "<pad>",
|
| 6 |
+
"unk_token": "<unk>"
|
| 7 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<cls>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "<eos>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"32": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"additional_special_tokens": [],
|
| 45 |
+
"clean_up_tokenization_spaces": true,
|
| 46 |
+
"cls_token": "<cls>",
|
| 47 |
+
"eos_token": "<eos>",
|
| 48 |
+
"mask_token": "<mask>",
|
| 49 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 50 |
+
"pad_token": "<pad>",
|
| 51 |
+
"tokenizer_class": "EsmTokenizer",
|
| 52 |
+
"tokenizer_file": null,
|
| 53 |
+
"unk_token": "<unk>"
|
| 54 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/trainer_state.json
ADDED
|
@@ -0,0 +1,619 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.19692792437967704,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 10000,
|
| 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": 0.00017015520980041642,
|
| 14 |
+
"loss": 0.5575,
|
| 15 |
+
"step": 100
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.0,
|
| 19 |
+
"learning_rate": 0.00017015035788629153,
|
| 20 |
+
"loss": 0.3065,
|
| 21 |
+
"step": 200
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.01,
|
| 25 |
+
"learning_rate": 0.00017014224985982174,
|
| 26 |
+
"loss": 0.2229,
|
| 27 |
+
"step": 300
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.01,
|
| 31 |
+
"learning_rate": 0.0001701308860313402,
|
| 32 |
+
"loss": 0.1812,
|
| 33 |
+
"step": 400
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.01,
|
| 37 |
+
"learning_rate": 0.00017011626683579524,
|
| 38 |
+
"loss": 0.1572,
|
| 39 |
+
"step": 500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.01,
|
| 43 |
+
"learning_rate": 0.00017009839283273364,
|
| 44 |
+
"loss": 0.1335,
|
| 45 |
+
"step": 600
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.01,
|
| 49 |
+
"learning_rate": 0.00017007726470627936,
|
| 50 |
+
"loss": 0.1193,
|
| 51 |
+
"step": 700
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.02,
|
| 55 |
+
"learning_rate": 0.00017005288326510734,
|
| 56 |
+
"loss": 0.1075,
|
| 57 |
+
"step": 800
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.02,
|
| 61 |
+
"learning_rate": 0.0001700252494424124,
|
| 62 |
+
"loss": 0.0986,
|
| 63 |
+
"step": 900
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.02,
|
| 67 |
+
"learning_rate": 0.00016999436429587366,
|
| 68 |
+
"loss": 0.0952,
|
| 69 |
+
"step": 1000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.02,
|
| 73 |
+
"learning_rate": 0.00016996022900761407,
|
| 74 |
+
"loss": 0.0865,
|
| 75 |
+
"step": 1100
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.02,
|
| 79 |
+
"learning_rate": 0.000169922844884155,
|
| 80 |
+
"loss": 0.0816,
|
| 81 |
+
"step": 1200
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.03,
|
| 85 |
+
"learning_rate": 0.00016988221335636648,
|
| 86 |
+
"loss": 0.0775,
|
| 87 |
+
"step": 1300
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.03,
|
| 91 |
+
"learning_rate": 0.00016983833597941224,
|
| 92 |
+
"loss": 0.0744,
|
| 93 |
+
"step": 1400
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.03,
|
| 97 |
+
"learning_rate": 0.00016979121443269025,
|
| 98 |
+
"loss": 0.0694,
|
| 99 |
+
"step": 1500
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.03,
|
| 103 |
+
"learning_rate": 0.0001697408505197684,
|
| 104 |
+
"loss": 0.0641,
|
| 105 |
+
"step": 1600
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.03,
|
| 109 |
+
"learning_rate": 0.00016968724616831557,
|
| 110 |
+
"loss": 0.0629,
|
| 111 |
+
"step": 1700
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.04,
|
| 115 |
+
"learning_rate": 0.0001696304034300278,
|
| 116 |
+
"loss": 0.0696,
|
| 117 |
+
"step": 1800
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.04,
|
| 121 |
+
"learning_rate": 0.00016957032448054968,
|
| 122 |
+
"loss": 0.0591,
|
| 123 |
+
"step": 1900
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.04,
|
| 127 |
+
"learning_rate": 0.0001695070116193912,
|
| 128 |
+
"loss": 0.06,
|
| 129 |
+
"step": 2000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.04,
|
| 133 |
+
"learning_rate": 0.0001694404672698396,
|
| 134 |
+
"loss": 0.0558,
|
| 135 |
+
"step": 2100
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.04,
|
| 139 |
+
"learning_rate": 0.00016937069397886687,
|
| 140 |
+
"loss": 0.0529,
|
| 141 |
+
"step": 2200
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.05,
|
| 145 |
+
"learning_rate": 0.00016929769441703196,
|
| 146 |
+
"loss": 0.0536,
|
| 147 |
+
"step": 2300
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.05,
|
| 151 |
+
"learning_rate": 0.00016922147137837868,
|
| 152 |
+
"loss": 0.0537,
|
| 153 |
+
"step": 2400
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.05,
|
| 157 |
+
"learning_rate": 0.00016914202778032893,
|
| 158 |
+
"loss": 0.0508,
|
| 159 |
+
"step": 2500
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.05,
|
| 163 |
+
"learning_rate": 0.0001690593666635707,
|
| 164 |
+
"loss": 0.0524,
|
| 165 |
+
"step": 2600
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.05,
|
| 169 |
+
"learning_rate": 0.00016897349119194207,
|
| 170 |
+
"loss": 0.0494,
|
| 171 |
+
"step": 2700
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.06,
|
| 175 |
+
"learning_rate": 0.00016888440465230977,
|
| 176 |
+
"loss": 0.047,
|
| 177 |
+
"step": 2800
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.06,
|
| 181 |
+
"learning_rate": 0.00016879211045444354,
|
| 182 |
+
"loss": 0.0502,
|
| 183 |
+
"step": 2900
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.06,
|
| 187 |
+
"learning_rate": 0.00016869661213088575,
|
| 188 |
+
"loss": 0.046,
|
| 189 |
+
"step": 3000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.06,
|
| 193 |
+
"learning_rate": 0.00016859791333681583,
|
| 194 |
+
"loss": 0.0419,
|
| 195 |
+
"step": 3100
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.06,
|
| 199 |
+
"learning_rate": 0.0001684960178499108,
|
| 200 |
+
"loss": 0.0456,
|
| 201 |
+
"step": 3200
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.06,
|
| 205 |
+
"learning_rate": 0.00016839092957020028,
|
| 206 |
+
"loss": 0.0432,
|
| 207 |
+
"step": 3300
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.07,
|
| 211 |
+
"learning_rate": 0.00016828265251991761,
|
| 212 |
+
"loss": 0.0423,
|
| 213 |
+
"step": 3400
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.07,
|
| 217 |
+
"learning_rate": 0.00016817119084334555,
|
| 218 |
+
"loss": 0.0407,
|
| 219 |
+
"step": 3500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.07,
|
| 223 |
+
"learning_rate": 0.00016805654880665776,
|
| 224 |
+
"loss": 0.0409,
|
| 225 |
+
"step": 3600
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.07,
|
| 229 |
+
"learning_rate": 0.00016793873079775577,
|
| 230 |
+
"loss": 0.0397,
|
| 231 |
+
"step": 3700
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.07,
|
| 235 |
+
"learning_rate": 0.00016781774132610059,
|
| 236 |
+
"loss": 0.0419,
|
| 237 |
+
"step": 3800
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.08,
|
| 241 |
+
"learning_rate": 0.0001676935850225405,
|
| 242 |
+
"loss": 0.0399,
|
| 243 |
+
"step": 3900
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.08,
|
| 247 |
+
"learning_rate": 0.00016756626663913358,
|
| 248 |
+
"loss": 0.0389,
|
| 249 |
+
"step": 4000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.08,
|
| 253 |
+
"learning_rate": 0.00016743579104896593,
|
| 254 |
+
"loss": 0.0385,
|
| 255 |
+
"step": 4100
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.08,
|
| 259 |
+
"learning_rate": 0.00016730216324596504,
|
| 260 |
+
"loss": 0.0357,
|
| 261 |
+
"step": 4200
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.08,
|
| 265 |
+
"learning_rate": 0.0001671653883447088,
|
| 266 |
+
"loss": 0.0373,
|
| 267 |
+
"step": 4300
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.09,
|
| 271 |
+
"learning_rate": 0.00016702547158022968,
|
| 272 |
+
"loss": 0.0375,
|
| 273 |
+
"step": 4400
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.09,
|
| 277 |
+
"learning_rate": 0.0001668824183078143,
|
| 278 |
+
"loss": 0.0393,
|
| 279 |
+
"step": 4500
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.09,
|
| 283 |
+
"learning_rate": 0.00016673623400279849,
|
| 284 |
+
"loss": 0.0358,
|
| 285 |
+
"step": 4600
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.09,
|
| 289 |
+
"learning_rate": 0.00016658692426035782,
|
| 290 |
+
"loss": 0.0334,
|
| 291 |
+
"step": 4700
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.09,
|
| 295 |
+
"learning_rate": 0.00016643449479529325,
|
| 296 |
+
"loss": 0.035,
|
| 297 |
+
"step": 4800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.1,
|
| 301 |
+
"learning_rate": 0.00016627895144181258,
|
| 302 |
+
"loss": 0.0343,
|
| 303 |
+
"step": 4900
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.1,
|
| 307 |
+
"learning_rate": 0.0001661203001533071,
|
| 308 |
+
"loss": 0.0335,
|
| 309 |
+
"step": 5000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.1,
|
| 313 |
+
"learning_rate": 0.00016595854700212362,
|
| 314 |
+
"loss": 0.0353,
|
| 315 |
+
"step": 5100
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.1,
|
| 319 |
+
"learning_rate": 0.0001657936981793322,
|
| 320 |
+
"loss": 0.0331,
|
| 321 |
+
"step": 5200
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.1,
|
| 325 |
+
"learning_rate": 0.0001656257599944891,
|
| 326 |
+
"loss": 0.0387,
|
| 327 |
+
"step": 5300
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.11,
|
| 331 |
+
"learning_rate": 0.00016545473887539532,
|
| 332 |
+
"loss": 0.0327,
|
| 333 |
+
"step": 5400
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.11,
|
| 337 |
+
"learning_rate": 0.00016528064136785056,
|
| 338 |
+
"loss": 0.0316,
|
| 339 |
+
"step": 5500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.11,
|
| 343 |
+
"learning_rate": 0.00016510347413540262,
|
| 344 |
+
"loss": 0.0319,
|
| 345 |
+
"step": 5600
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.11,
|
| 349 |
+
"learning_rate": 0.0001649232439590925,
|
| 350 |
+
"loss": 0.0314,
|
| 351 |
+
"step": 5700
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.11,
|
| 355 |
+
"learning_rate": 0.0001647399577371947,
|
| 356 |
+
"loss": 0.0321,
|
| 357 |
+
"step": 5800
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.12,
|
| 361 |
+
"learning_rate": 0.00016455362248495338,
|
| 362 |
+
"loss": 0.0333,
|
| 363 |
+
"step": 5900
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.12,
|
| 367 |
+
"learning_rate": 0.00016436424533431362,
|
| 368 |
+
"loss": 0.0319,
|
| 369 |
+
"step": 6000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.12,
|
| 373 |
+
"learning_rate": 0.0001641718335336486,
|
| 374 |
+
"loss": 0.0315,
|
| 375 |
+
"step": 6100
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.12,
|
| 379 |
+
"learning_rate": 0.0001639763944474821,
|
| 380 |
+
"loss": 0.0311,
|
| 381 |
+
"step": 6200
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.12,
|
| 385 |
+
"learning_rate": 0.0001637779355562068,
|
| 386 |
+
"loss": 0.031,
|
| 387 |
+
"step": 6300
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.13,
|
| 391 |
+
"learning_rate": 0.00016357646445579763,
|
| 392 |
+
"loss": 0.0299,
|
| 393 |
+
"step": 6400
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.13,
|
| 397 |
+
"learning_rate": 0.00016337198885752133,
|
| 398 |
+
"loss": 0.0317,
|
| 399 |
+
"step": 6500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.13,
|
| 403 |
+
"learning_rate": 0.00016316451658764122,
|
| 404 |
+
"loss": 0.0302,
|
| 405 |
+
"step": 6600
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.13,
|
| 409 |
+
"learning_rate": 0.0001629540555871176,
|
| 410 |
+
"loss": 0.0295,
|
| 411 |
+
"step": 6700
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 0.13,
|
| 415 |
+
"learning_rate": 0.00016274061391130388,
|
| 416 |
+
"loss": 0.03,
|
| 417 |
+
"step": 6800
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.14,
|
| 421 |
+
"learning_rate": 0.0001625241997296382,
|
| 422 |
+
"loss": 0.0292,
|
| 423 |
+
"step": 6900
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.14,
|
| 427 |
+
"learning_rate": 0.00016230482132533077,
|
| 428 |
+
"loss": 0.0289,
|
| 429 |
+
"step": 7000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.14,
|
| 433 |
+
"learning_rate": 0.00016208472504084003,
|
| 434 |
+
"loss": 0.0318,
|
| 435 |
+
"step": 7100
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.14,
|
| 439 |
+
"learning_rate": 0.0001618594729250462,
|
| 440 |
+
"loss": 0.0301,
|
| 441 |
+
"step": 7200
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.14,
|
| 445 |
+
"learning_rate": 0.00016163128202889828,
|
| 446 |
+
"loss": 0.0295,
|
| 447 |
+
"step": 7300
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.15,
|
| 451 |
+
"learning_rate": 0.00016140016108635798,
|
| 452 |
+
"loss": 0.029,
|
| 453 |
+
"step": 7400
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 0.15,
|
| 457 |
+
"learning_rate": 0.00016116611894353386,
|
| 458 |
+
"loss": 0.0291,
|
| 459 |
+
"step": 7500
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.15,
|
| 463 |
+
"learning_rate": 0.00016092916455834295,
|
| 464 |
+
"loss": 0.0311,
|
| 465 |
+
"step": 7600
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.15,
|
| 469 |
+
"learning_rate": 0.00016068930700016766,
|
| 470 |
+
"loss": 0.0285,
|
| 471 |
+
"step": 7700
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.15,
|
| 475 |
+
"learning_rate": 0.00016044655544950889,
|
| 476 |
+
"loss": 0.0287,
|
| 477 |
+
"step": 7800
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.16,
|
| 481 |
+
"learning_rate": 0.00016020091919763445,
|
| 482 |
+
"loss": 0.0293,
|
| 483 |
+
"step": 7900
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 0.16,
|
| 487 |
+
"learning_rate": 0.00015995240764622357,
|
| 488 |
+
"loss": 0.0259,
|
| 489 |
+
"step": 8000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.16,
|
| 493 |
+
"learning_rate": 0.000159701030307007,
|
| 494 |
+
"loss": 0.0293,
|
| 495 |
+
"step": 8100
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.16,
|
| 499 |
+
"learning_rate": 0.00015944679680140295,
|
| 500 |
+
"loss": 0.0277,
|
| 501 |
+
"step": 8200
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.16,
|
| 505 |
+
"learning_rate": 0.0001591897168601488,
|
| 506 |
+
"loss": 0.0304,
|
| 507 |
+
"step": 8300
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.17,
|
| 511 |
+
"learning_rate": 0.00015892980032292876,
|
| 512 |
+
"loss": 0.026,
|
| 513 |
+
"step": 8400
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.17,
|
| 517 |
+
"learning_rate": 0.00015866705713799714,
|
| 518 |
+
"loss": 0.0294,
|
| 519 |
+
"step": 8500
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.17,
|
| 523 |
+
"learning_rate": 0.00015840149736179762,
|
| 524 |
+
"loss": 0.0321,
|
| 525 |
+
"step": 8600
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.17,
|
| 529 |
+
"learning_rate": 0.0001581331311585785,
|
| 530 |
+
"loss": 0.0279,
|
| 531 |
+
"step": 8700
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.17,
|
| 535 |
+
"learning_rate": 0.00015786196880000325,
|
| 536 |
+
"loss": 0.0277,
|
| 537 |
+
"step": 8800
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.18,
|
| 541 |
+
"learning_rate": 0.0001575880206647579,
|
| 542 |
+
"loss": 0.0268,
|
| 543 |
+
"step": 8900
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.18,
|
| 547 |
+
"learning_rate": 0.00015731129723815343,
|
| 548 |
+
"loss": 0.0281,
|
| 549 |
+
"step": 9000
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.18,
|
| 553 |
+
"learning_rate": 0.00015703180911172453,
|
| 554 |
+
"loss": 0.028,
|
| 555 |
+
"step": 9100
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.18,
|
| 559 |
+
"learning_rate": 0.0001567495669828243,
|
| 560 |
+
"loss": 0.026,
|
| 561 |
+
"step": 9200
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.18,
|
| 565 |
+
"learning_rate": 0.0001564645816542146,
|
| 566 |
+
"loss": 0.0256,
|
| 567 |
+
"step": 9300
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 0.19,
|
| 571 |
+
"learning_rate": 0.0001561768640336529,
|
| 572 |
+
"loss": 0.027,
|
| 573 |
+
"step": 9400
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 0.19,
|
| 577 |
+
"learning_rate": 0.0001558864251334745,
|
| 578 |
+
"loss": 0.0249,
|
| 579 |
+
"step": 9500
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.19,
|
| 583 |
+
"learning_rate": 0.00015559327607017119,
|
| 584 |
+
"loss": 0.0256,
|
| 585 |
+
"step": 9600
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.19,
|
| 589 |
+
"learning_rate": 0.00015529742806396564,
|
| 590 |
+
"loss": 0.0251,
|
| 591 |
+
"step": 9700
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.19,
|
| 595 |
+
"learning_rate": 0.00015499889243838211,
|
| 596 |
+
"loss": 0.0257,
|
| 597 |
+
"step": 9800
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.19,
|
| 601 |
+
"learning_rate": 0.00015469768061981295,
|
| 602 |
+
"loss": 0.0264,
|
| 603 |
+
"step": 9900
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.2,
|
| 607 |
+
"learning_rate": 0.00015439685605389073,
|
| 608 |
+
"loss": 0.0247,
|
| 609 |
+
"step": 10000
|
| 610 |
+
}
|
| 611 |
+
],
|
| 612 |
+
"logging_steps": 100,
|
| 613 |
+
"max_steps": 50780,
|
| 614 |
+
"num_train_epochs": 1,
|
| 615 |
+
"save_steps": 10000,
|
| 616 |
+
"total_flos": 3.992837363712e+18,
|
| 617 |
+
"trial_name": null,
|
| 618 |
+
"trial_params": null
|
| 619 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d17a2c3c8a65881aa45bfd2997dfe652a61cb20d8b49594dc95bff2d7669f6f
|
| 3 |
+
size 4155
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-10000/vocab.txt
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
<cls>
|
| 2 |
+
<pad>
|
| 3 |
+
<eos>
|
| 4 |
+
<unk>
|
| 5 |
+
L
|
| 6 |
+
A
|
| 7 |
+
G
|
| 8 |
+
V
|
| 9 |
+
S
|
| 10 |
+
E
|
| 11 |
+
R
|
| 12 |
+
T
|
| 13 |
+
I
|
| 14 |
+
D
|
| 15 |
+
P
|
| 16 |
+
K
|
| 17 |
+
Q
|
| 18 |
+
N
|
| 19 |
+
F
|
| 20 |
+
Y
|
| 21 |
+
M
|
| 22 |
+
H
|
| 23 |
+
W
|
| 24 |
+
C
|
| 25 |
+
X
|
| 26 |
+
B
|
| 27 |
+
U
|
| 28 |
+
Z
|
| 29 |
+
O
|
| 30 |
+
.
|
| 31 |
+
-
|
| 32 |
+
<null_1>
|
| 33 |
+
<mask>
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/README.md
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
| 4 |
+
## Training procedure
|
| 5 |
+
|
| 6 |
+
|
| 7 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 8 |
+
- quant_method: bitsandbytes
|
| 9 |
+
- load_in_8bit: False
|
| 10 |
+
- load_in_4bit: True
|
| 11 |
+
- llm_int8_threshold: 6.0
|
| 12 |
+
- llm_int8_skip_modules: None
|
| 13 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 14 |
+
- llm_int8_has_fp16_weight: False
|
| 15 |
+
- bnb_4bit_quant_type: nf4
|
| 16 |
+
- bnb_4bit_use_double_quant: True
|
| 17 |
+
- bnb_4bit_compute_dtype: bfloat16
|
| 18 |
+
### Framework versions
|
| 19 |
+
|
| 20 |
+
|
| 21 |
+
- PEFT 0.6.0.dev0
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/adapter_config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "facebook/esm2_t33_650M_UR50D",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"fan_in_fan_out": false,
|
| 7 |
+
"inference_mode": true,
|
| 8 |
+
"init_lora_weights": true,
|
| 9 |
+
"layers_pattern": null,
|
| 10 |
+
"layers_to_transform": null,
|
| 11 |
+
"lora_alpha": 1,
|
| 12 |
+
"lora_dropout": 0.5,
|
| 13 |
+
"modules_to_save": null,
|
| 14 |
+
"peft_type": "LORA",
|
| 15 |
+
"r": 2,
|
| 16 |
+
"rank_pattern": {},
|
| 17 |
+
"revision": null,
|
| 18 |
+
"target_modules": [
|
| 19 |
+
"query",
|
| 20 |
+
"key",
|
| 21 |
+
"value",
|
| 22 |
+
"EsmSelfOutput.dense",
|
| 23 |
+
"EsmIntermediate.dense",
|
| 24 |
+
"EsmOutput.dense",
|
| 25 |
+
"classifier"
|
| 26 |
+
],
|
| 27 |
+
"task_type": "TOKEN_CLS"
|
| 28 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9d0569a2930474e7eb8b4baa3794f4dc925230d27ff471fa125b41c55ec90002
|
| 3 |
+
size 2133117
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/added_tokens.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<cls>": 0,
|
| 3 |
+
"<eos>": 2,
|
| 4 |
+
"<mask>": 32,
|
| 5 |
+
"<pad>": 1,
|
| 6 |
+
"<unk>": 3
|
| 7 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ba123d3d5300c04df61a19f973041d9aba2f60b21e2fa6150a534ca5a07fd298
|
| 3 |
+
size 4215045
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad4b33fd771a590bf68f30d331d8b8f0b3cae0df4abde1370ad68d21d8531255
|
| 3 |
+
size 14575
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3d4cc968ce155a88f3501215aa8a7e29778cb7885b3150bad2441fdd48871663
|
| 3 |
+
size 627
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "<cls>",
|
| 3 |
+
"eos_token": "<eos>",
|
| 4 |
+
"mask_token": "<mask>",
|
| 5 |
+
"pad_token": "<pad>",
|
| 6 |
+
"unk_token": "<unk>"
|
| 7 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<cls>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "<eos>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"32": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"additional_special_tokens": [],
|
| 45 |
+
"clean_up_tokenization_spaces": true,
|
| 46 |
+
"cls_token": "<cls>",
|
| 47 |
+
"eos_token": "<eos>",
|
| 48 |
+
"mask_token": "<mask>",
|
| 49 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 50 |
+
"pad_token": "<pad>",
|
| 51 |
+
"tokenizer_class": "EsmTokenizer",
|
| 52 |
+
"tokenizer_file": null,
|
| 53 |
+
"unk_token": "<unk>"
|
| 54 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/trainer_state.json
ADDED
|
@@ -0,0 +1,1219 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.3938558487593541,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 20000,
|
| 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": 0.00017015520980041642,
|
| 14 |
+
"loss": 0.5575,
|
| 15 |
+
"step": 100
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.0,
|
| 19 |
+
"learning_rate": 0.00017015035788629153,
|
| 20 |
+
"loss": 0.3065,
|
| 21 |
+
"step": 200
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.01,
|
| 25 |
+
"learning_rate": 0.00017014224985982174,
|
| 26 |
+
"loss": 0.2229,
|
| 27 |
+
"step": 300
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.01,
|
| 31 |
+
"learning_rate": 0.0001701308860313402,
|
| 32 |
+
"loss": 0.1812,
|
| 33 |
+
"step": 400
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.01,
|
| 37 |
+
"learning_rate": 0.00017011626683579524,
|
| 38 |
+
"loss": 0.1572,
|
| 39 |
+
"step": 500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.01,
|
| 43 |
+
"learning_rate": 0.00017009839283273364,
|
| 44 |
+
"loss": 0.1335,
|
| 45 |
+
"step": 600
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.01,
|
| 49 |
+
"learning_rate": 0.00017007726470627936,
|
| 50 |
+
"loss": 0.1193,
|
| 51 |
+
"step": 700
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.02,
|
| 55 |
+
"learning_rate": 0.00017005288326510734,
|
| 56 |
+
"loss": 0.1075,
|
| 57 |
+
"step": 800
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.02,
|
| 61 |
+
"learning_rate": 0.0001700252494424124,
|
| 62 |
+
"loss": 0.0986,
|
| 63 |
+
"step": 900
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.02,
|
| 67 |
+
"learning_rate": 0.00016999436429587366,
|
| 68 |
+
"loss": 0.0952,
|
| 69 |
+
"step": 1000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.02,
|
| 73 |
+
"learning_rate": 0.00016996022900761407,
|
| 74 |
+
"loss": 0.0865,
|
| 75 |
+
"step": 1100
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.02,
|
| 79 |
+
"learning_rate": 0.000169922844884155,
|
| 80 |
+
"loss": 0.0816,
|
| 81 |
+
"step": 1200
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.03,
|
| 85 |
+
"learning_rate": 0.00016988221335636648,
|
| 86 |
+
"loss": 0.0775,
|
| 87 |
+
"step": 1300
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.03,
|
| 91 |
+
"learning_rate": 0.00016983833597941224,
|
| 92 |
+
"loss": 0.0744,
|
| 93 |
+
"step": 1400
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.03,
|
| 97 |
+
"learning_rate": 0.00016979121443269025,
|
| 98 |
+
"loss": 0.0694,
|
| 99 |
+
"step": 1500
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.03,
|
| 103 |
+
"learning_rate": 0.0001697408505197684,
|
| 104 |
+
"loss": 0.0641,
|
| 105 |
+
"step": 1600
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.03,
|
| 109 |
+
"learning_rate": 0.00016968724616831557,
|
| 110 |
+
"loss": 0.0629,
|
| 111 |
+
"step": 1700
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.04,
|
| 115 |
+
"learning_rate": 0.0001696304034300278,
|
| 116 |
+
"loss": 0.0696,
|
| 117 |
+
"step": 1800
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.04,
|
| 121 |
+
"learning_rate": 0.00016957032448054968,
|
| 122 |
+
"loss": 0.0591,
|
| 123 |
+
"step": 1900
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.04,
|
| 127 |
+
"learning_rate": 0.0001695070116193912,
|
| 128 |
+
"loss": 0.06,
|
| 129 |
+
"step": 2000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.04,
|
| 133 |
+
"learning_rate": 0.0001694404672698396,
|
| 134 |
+
"loss": 0.0558,
|
| 135 |
+
"step": 2100
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.04,
|
| 139 |
+
"learning_rate": 0.00016937069397886687,
|
| 140 |
+
"loss": 0.0529,
|
| 141 |
+
"step": 2200
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.05,
|
| 145 |
+
"learning_rate": 0.00016929769441703196,
|
| 146 |
+
"loss": 0.0536,
|
| 147 |
+
"step": 2300
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.05,
|
| 151 |
+
"learning_rate": 0.00016922147137837868,
|
| 152 |
+
"loss": 0.0537,
|
| 153 |
+
"step": 2400
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.05,
|
| 157 |
+
"learning_rate": 0.00016914202778032893,
|
| 158 |
+
"loss": 0.0508,
|
| 159 |
+
"step": 2500
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.05,
|
| 163 |
+
"learning_rate": 0.0001690593666635707,
|
| 164 |
+
"loss": 0.0524,
|
| 165 |
+
"step": 2600
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.05,
|
| 169 |
+
"learning_rate": 0.00016897349119194207,
|
| 170 |
+
"loss": 0.0494,
|
| 171 |
+
"step": 2700
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.06,
|
| 175 |
+
"learning_rate": 0.00016888440465230977,
|
| 176 |
+
"loss": 0.047,
|
| 177 |
+
"step": 2800
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.06,
|
| 181 |
+
"learning_rate": 0.00016879211045444354,
|
| 182 |
+
"loss": 0.0502,
|
| 183 |
+
"step": 2900
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.06,
|
| 187 |
+
"learning_rate": 0.00016869661213088575,
|
| 188 |
+
"loss": 0.046,
|
| 189 |
+
"step": 3000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.06,
|
| 193 |
+
"learning_rate": 0.00016859791333681583,
|
| 194 |
+
"loss": 0.0419,
|
| 195 |
+
"step": 3100
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.06,
|
| 199 |
+
"learning_rate": 0.0001684960178499108,
|
| 200 |
+
"loss": 0.0456,
|
| 201 |
+
"step": 3200
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.06,
|
| 205 |
+
"learning_rate": 0.00016839092957020028,
|
| 206 |
+
"loss": 0.0432,
|
| 207 |
+
"step": 3300
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.07,
|
| 211 |
+
"learning_rate": 0.00016828265251991761,
|
| 212 |
+
"loss": 0.0423,
|
| 213 |
+
"step": 3400
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.07,
|
| 217 |
+
"learning_rate": 0.00016817119084334555,
|
| 218 |
+
"loss": 0.0407,
|
| 219 |
+
"step": 3500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.07,
|
| 223 |
+
"learning_rate": 0.00016805654880665776,
|
| 224 |
+
"loss": 0.0409,
|
| 225 |
+
"step": 3600
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.07,
|
| 229 |
+
"learning_rate": 0.00016793873079775577,
|
| 230 |
+
"loss": 0.0397,
|
| 231 |
+
"step": 3700
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.07,
|
| 235 |
+
"learning_rate": 0.00016781774132610059,
|
| 236 |
+
"loss": 0.0419,
|
| 237 |
+
"step": 3800
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.08,
|
| 241 |
+
"learning_rate": 0.0001676935850225405,
|
| 242 |
+
"loss": 0.0399,
|
| 243 |
+
"step": 3900
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.08,
|
| 247 |
+
"learning_rate": 0.00016756626663913358,
|
| 248 |
+
"loss": 0.0389,
|
| 249 |
+
"step": 4000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.08,
|
| 253 |
+
"learning_rate": 0.00016743579104896593,
|
| 254 |
+
"loss": 0.0385,
|
| 255 |
+
"step": 4100
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.08,
|
| 259 |
+
"learning_rate": 0.00016730216324596504,
|
| 260 |
+
"loss": 0.0357,
|
| 261 |
+
"step": 4200
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.08,
|
| 265 |
+
"learning_rate": 0.0001671653883447088,
|
| 266 |
+
"loss": 0.0373,
|
| 267 |
+
"step": 4300
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.09,
|
| 271 |
+
"learning_rate": 0.00016702547158022968,
|
| 272 |
+
"loss": 0.0375,
|
| 273 |
+
"step": 4400
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.09,
|
| 277 |
+
"learning_rate": 0.0001668824183078143,
|
| 278 |
+
"loss": 0.0393,
|
| 279 |
+
"step": 4500
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.09,
|
| 283 |
+
"learning_rate": 0.00016673623400279849,
|
| 284 |
+
"loss": 0.0358,
|
| 285 |
+
"step": 4600
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.09,
|
| 289 |
+
"learning_rate": 0.00016658692426035782,
|
| 290 |
+
"loss": 0.0334,
|
| 291 |
+
"step": 4700
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.09,
|
| 295 |
+
"learning_rate": 0.00016643449479529325,
|
| 296 |
+
"loss": 0.035,
|
| 297 |
+
"step": 4800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.1,
|
| 301 |
+
"learning_rate": 0.00016627895144181258,
|
| 302 |
+
"loss": 0.0343,
|
| 303 |
+
"step": 4900
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.1,
|
| 307 |
+
"learning_rate": 0.0001661203001533071,
|
| 308 |
+
"loss": 0.0335,
|
| 309 |
+
"step": 5000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.1,
|
| 313 |
+
"learning_rate": 0.00016595854700212362,
|
| 314 |
+
"loss": 0.0353,
|
| 315 |
+
"step": 5100
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.1,
|
| 319 |
+
"learning_rate": 0.0001657936981793322,
|
| 320 |
+
"loss": 0.0331,
|
| 321 |
+
"step": 5200
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.1,
|
| 325 |
+
"learning_rate": 0.0001656257599944891,
|
| 326 |
+
"loss": 0.0387,
|
| 327 |
+
"step": 5300
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.11,
|
| 331 |
+
"learning_rate": 0.00016545473887539532,
|
| 332 |
+
"loss": 0.0327,
|
| 333 |
+
"step": 5400
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.11,
|
| 337 |
+
"learning_rate": 0.00016528064136785056,
|
| 338 |
+
"loss": 0.0316,
|
| 339 |
+
"step": 5500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.11,
|
| 343 |
+
"learning_rate": 0.00016510347413540262,
|
| 344 |
+
"loss": 0.0319,
|
| 345 |
+
"step": 5600
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.11,
|
| 349 |
+
"learning_rate": 0.0001649232439590925,
|
| 350 |
+
"loss": 0.0314,
|
| 351 |
+
"step": 5700
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.11,
|
| 355 |
+
"learning_rate": 0.0001647399577371947,
|
| 356 |
+
"loss": 0.0321,
|
| 357 |
+
"step": 5800
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.12,
|
| 361 |
+
"learning_rate": 0.00016455362248495338,
|
| 362 |
+
"loss": 0.0333,
|
| 363 |
+
"step": 5900
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.12,
|
| 367 |
+
"learning_rate": 0.00016436424533431362,
|
| 368 |
+
"loss": 0.0319,
|
| 369 |
+
"step": 6000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.12,
|
| 373 |
+
"learning_rate": 0.0001641718335336486,
|
| 374 |
+
"loss": 0.0315,
|
| 375 |
+
"step": 6100
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.12,
|
| 379 |
+
"learning_rate": 0.0001639763944474821,
|
| 380 |
+
"loss": 0.0311,
|
| 381 |
+
"step": 6200
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.12,
|
| 385 |
+
"learning_rate": 0.0001637779355562068,
|
| 386 |
+
"loss": 0.031,
|
| 387 |
+
"step": 6300
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.13,
|
| 391 |
+
"learning_rate": 0.00016357646445579763,
|
| 392 |
+
"loss": 0.0299,
|
| 393 |
+
"step": 6400
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.13,
|
| 397 |
+
"learning_rate": 0.00016337198885752133,
|
| 398 |
+
"loss": 0.0317,
|
| 399 |
+
"step": 6500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.13,
|
| 403 |
+
"learning_rate": 0.00016316451658764122,
|
| 404 |
+
"loss": 0.0302,
|
| 405 |
+
"step": 6600
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.13,
|
| 409 |
+
"learning_rate": 0.0001629540555871176,
|
| 410 |
+
"loss": 0.0295,
|
| 411 |
+
"step": 6700
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 0.13,
|
| 415 |
+
"learning_rate": 0.00016274061391130388,
|
| 416 |
+
"loss": 0.03,
|
| 417 |
+
"step": 6800
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.14,
|
| 421 |
+
"learning_rate": 0.0001625241997296382,
|
| 422 |
+
"loss": 0.0292,
|
| 423 |
+
"step": 6900
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.14,
|
| 427 |
+
"learning_rate": 0.00016230482132533077,
|
| 428 |
+
"loss": 0.0289,
|
| 429 |
+
"step": 7000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.14,
|
| 433 |
+
"learning_rate": 0.00016208472504084003,
|
| 434 |
+
"loss": 0.0318,
|
| 435 |
+
"step": 7100
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.14,
|
| 439 |
+
"learning_rate": 0.0001618594729250462,
|
| 440 |
+
"loss": 0.0301,
|
| 441 |
+
"step": 7200
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.14,
|
| 445 |
+
"learning_rate": 0.00016163128202889828,
|
| 446 |
+
"loss": 0.0295,
|
| 447 |
+
"step": 7300
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.15,
|
| 451 |
+
"learning_rate": 0.00016140016108635798,
|
| 452 |
+
"loss": 0.029,
|
| 453 |
+
"step": 7400
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 0.15,
|
| 457 |
+
"learning_rate": 0.00016116611894353386,
|
| 458 |
+
"loss": 0.0291,
|
| 459 |
+
"step": 7500
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.15,
|
| 463 |
+
"learning_rate": 0.00016092916455834295,
|
| 464 |
+
"loss": 0.0311,
|
| 465 |
+
"step": 7600
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.15,
|
| 469 |
+
"learning_rate": 0.00016068930700016766,
|
| 470 |
+
"loss": 0.0285,
|
| 471 |
+
"step": 7700
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.15,
|
| 475 |
+
"learning_rate": 0.00016044655544950889,
|
| 476 |
+
"loss": 0.0287,
|
| 477 |
+
"step": 7800
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.16,
|
| 481 |
+
"learning_rate": 0.00016020091919763445,
|
| 482 |
+
"loss": 0.0293,
|
| 483 |
+
"step": 7900
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 0.16,
|
| 487 |
+
"learning_rate": 0.00015995240764622357,
|
| 488 |
+
"loss": 0.0259,
|
| 489 |
+
"step": 8000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.16,
|
| 493 |
+
"learning_rate": 0.000159701030307007,
|
| 494 |
+
"loss": 0.0293,
|
| 495 |
+
"step": 8100
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.16,
|
| 499 |
+
"learning_rate": 0.00015944679680140295,
|
| 500 |
+
"loss": 0.0277,
|
| 501 |
+
"step": 8200
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.16,
|
| 505 |
+
"learning_rate": 0.0001591897168601488,
|
| 506 |
+
"loss": 0.0304,
|
| 507 |
+
"step": 8300
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.17,
|
| 511 |
+
"learning_rate": 0.00015892980032292876,
|
| 512 |
+
"loss": 0.026,
|
| 513 |
+
"step": 8400
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.17,
|
| 517 |
+
"learning_rate": 0.00015866705713799714,
|
| 518 |
+
"loss": 0.0294,
|
| 519 |
+
"step": 8500
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.17,
|
| 523 |
+
"learning_rate": 0.00015840149736179762,
|
| 524 |
+
"loss": 0.0321,
|
| 525 |
+
"step": 8600
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.17,
|
| 529 |
+
"learning_rate": 0.0001581331311585785,
|
| 530 |
+
"loss": 0.0279,
|
| 531 |
+
"step": 8700
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.17,
|
| 535 |
+
"learning_rate": 0.00015786196880000325,
|
| 536 |
+
"loss": 0.0277,
|
| 537 |
+
"step": 8800
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.18,
|
| 541 |
+
"learning_rate": 0.0001575880206647579,
|
| 542 |
+
"loss": 0.0268,
|
| 543 |
+
"step": 8900
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.18,
|
| 547 |
+
"learning_rate": 0.00015731129723815343,
|
| 548 |
+
"loss": 0.0281,
|
| 549 |
+
"step": 9000
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.18,
|
| 553 |
+
"learning_rate": 0.00015703180911172453,
|
| 554 |
+
"loss": 0.028,
|
| 555 |
+
"step": 9100
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.18,
|
| 559 |
+
"learning_rate": 0.0001567495669828243,
|
| 560 |
+
"loss": 0.026,
|
| 561 |
+
"step": 9200
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.18,
|
| 565 |
+
"learning_rate": 0.0001564645816542146,
|
| 566 |
+
"loss": 0.0256,
|
| 567 |
+
"step": 9300
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 0.19,
|
| 571 |
+
"learning_rate": 0.0001561768640336529,
|
| 572 |
+
"loss": 0.027,
|
| 573 |
+
"step": 9400
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 0.19,
|
| 577 |
+
"learning_rate": 0.0001558864251334745,
|
| 578 |
+
"loss": 0.0249,
|
| 579 |
+
"step": 9500
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.19,
|
| 583 |
+
"learning_rate": 0.00015559327607017119,
|
| 584 |
+
"loss": 0.0256,
|
| 585 |
+
"step": 9600
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.19,
|
| 589 |
+
"learning_rate": 0.00015529742806396564,
|
| 590 |
+
"loss": 0.0251,
|
| 591 |
+
"step": 9700
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.19,
|
| 595 |
+
"learning_rate": 0.00015499889243838211,
|
| 596 |
+
"loss": 0.0257,
|
| 597 |
+
"step": 9800
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.19,
|
| 601 |
+
"learning_rate": 0.00015469768061981295,
|
| 602 |
+
"loss": 0.0264,
|
| 603 |
+
"step": 9900
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.2,
|
| 607 |
+
"learning_rate": 0.00015439685605389073,
|
| 608 |
+
"loss": 0.0247,
|
| 609 |
+
"step": 10000
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.2,
|
| 613 |
+
"learning_rate": 0.00015409035301023626,
|
| 614 |
+
"loss": 0.0256,
|
| 615 |
+
"step": 10100
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 0.2,
|
| 619 |
+
"learning_rate": 0.00015378120854776404,
|
| 620 |
+
"loss": 0.0247,
|
| 621 |
+
"step": 10200
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.2,
|
| 625 |
+
"learning_rate": 0.00015346943449891755,
|
| 626 |
+
"loss": 0.0254,
|
| 627 |
+
"step": 10300
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.2,
|
| 631 |
+
"learning_rate": 0.00015315504279678714,
|
| 632 |
+
"loss": 0.0246,
|
| 633 |
+
"step": 10400
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.21,
|
| 637 |
+
"learning_rate": 0.00015283804547465337,
|
| 638 |
+
"loss": 0.0249,
|
| 639 |
+
"step": 10500
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.21,
|
| 643 |
+
"learning_rate": 0.0001525184546655264,
|
| 644 |
+
"loss": 0.026,
|
| 645 |
+
"step": 10600
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 0.21,
|
| 649 |
+
"learning_rate": 0.0001521962826016816,
|
| 650 |
+
"loss": 0.0254,
|
| 651 |
+
"step": 10700
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 0.21,
|
| 655 |
+
"learning_rate": 0.00015187154161419122,
|
| 656 |
+
"loss": 0.0272,
|
| 657 |
+
"step": 10800
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.21,
|
| 661 |
+
"learning_rate": 0.0001515442441324528,
|
| 662 |
+
"loss": 0.0253,
|
| 663 |
+
"step": 10900
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 0.22,
|
| 667 |
+
"learning_rate": 0.00015121440268371297,
|
| 668 |
+
"loss": 0.0263,
|
| 669 |
+
"step": 11000
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.22,
|
| 673 |
+
"learning_rate": 0.00015088202989258835,
|
| 674 |
+
"loss": 0.0256,
|
| 675 |
+
"step": 11100
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.22,
|
| 679 |
+
"learning_rate": 0.00015054713848058212,
|
| 680 |
+
"loss": 0.023,
|
| 681 |
+
"step": 11200
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.22,
|
| 685 |
+
"learning_rate": 0.00015020974126559716,
|
| 686 |
+
"loss": 0.0236,
|
| 687 |
+
"step": 11300
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 0.22,
|
| 691 |
+
"learning_rate": 0.00014986985116144554,
|
| 692 |
+
"loss": 0.0241,
|
| 693 |
+
"step": 11400
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 0.23,
|
| 697 |
+
"learning_rate": 0.00014952748117735409,
|
| 698 |
+
"loss": 0.0235,
|
| 699 |
+
"step": 11500
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.23,
|
| 703 |
+
"learning_rate": 0.0001491826444174666,
|
| 704 |
+
"loss": 0.0242,
|
| 705 |
+
"step": 11600
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.23,
|
| 709 |
+
"learning_rate": 0.00014883535408034227,
|
| 710 |
+
"loss": 0.0233,
|
| 711 |
+
"step": 11700
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.23,
|
| 715 |
+
"learning_rate": 0.00014848562345845032,
|
| 716 |
+
"loss": 0.0228,
|
| 717 |
+
"step": 11800
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.23,
|
| 721 |
+
"learning_rate": 0.00014813346593766164,
|
| 722 |
+
"loss": 0.023,
|
| 723 |
+
"step": 11900
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.24,
|
| 727 |
+
"learning_rate": 0.0001477788949967359,
|
| 728 |
+
"loss": 0.0236,
|
| 729 |
+
"step": 12000
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.24,
|
| 733 |
+
"learning_rate": 0.00014742192420680626,
|
| 734 |
+
"loss": 0.0237,
|
| 735 |
+
"step": 12100
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.24,
|
| 739 |
+
"learning_rate": 0.00014706256723085937,
|
| 740 |
+
"loss": 0.0237,
|
| 741 |
+
"step": 12200
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 0.24,
|
| 745 |
+
"learning_rate": 0.0001467008378232128,
|
| 746 |
+
"loss": 0.024,
|
| 747 |
+
"step": 12300
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.24,
|
| 751 |
+
"learning_rate": 0.00014633674982898854,
|
| 752 |
+
"loss": 0.0227,
|
| 753 |
+
"step": 12400
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.25,
|
| 757 |
+
"learning_rate": 0.00014597031718358285,
|
| 758 |
+
"loss": 0.0247,
|
| 759 |
+
"step": 12500
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.25,
|
| 763 |
+
"learning_rate": 0.00014560155391213318,
|
| 764 |
+
"loss": 0.0233,
|
| 765 |
+
"step": 12600
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.25,
|
| 769 |
+
"learning_rate": 0.0001452304741289812,
|
| 770 |
+
"loss": 0.0234,
|
| 771 |
+
"step": 12700
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 0.25,
|
| 775 |
+
"learning_rate": 0.00014485709203713263,
|
| 776 |
+
"loss": 0.0221,
|
| 777 |
+
"step": 12800
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.25,
|
| 781 |
+
"learning_rate": 0.00014448142192771354,
|
| 782 |
+
"loss": 0.0244,
|
| 783 |
+
"step": 12900
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.26,
|
| 787 |
+
"learning_rate": 0.00014410347817942347,
|
| 788 |
+
"loss": 0.0242,
|
| 789 |
+
"step": 13000
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.26,
|
| 793 |
+
"learning_rate": 0.00014372327525798503,
|
| 794 |
+
"loss": 0.0231,
|
| 795 |
+
"step": 13100
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.26,
|
| 799 |
+
"learning_rate": 0.00014334082771559026,
|
| 800 |
+
"loss": 0.0238,
|
| 801 |
+
"step": 13200
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.26,
|
| 805 |
+
"learning_rate": 0.00014295615019034358,
|
| 806 |
+
"loss": 0.0233,
|
| 807 |
+
"step": 13300
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.26,
|
| 811 |
+
"learning_rate": 0.0001425692574057016,
|
| 812 |
+
"loss": 0.024,
|
| 813 |
+
"step": 13400
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 0.27,
|
| 817 |
+
"learning_rate": 0.00014218016416990954,
|
| 818 |
+
"loss": 0.0249,
|
| 819 |
+
"step": 13500
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.27,
|
| 823 |
+
"learning_rate": 0.0001417928089327897,
|
| 824 |
+
"loss": 0.0265,
|
| 825 |
+
"step": 13600
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.27,
|
| 829 |
+
"learning_rate": 0.00014139938118717103,
|
| 830 |
+
"loss": 0.0206,
|
| 831 |
+
"step": 13700
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.27,
|
| 835 |
+
"learning_rate": 0.00014100379776718443,
|
| 836 |
+
"loss": 0.0216,
|
| 837 |
+
"step": 13800
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.27,
|
| 841 |
+
"learning_rate": 0.00014060607381370744,
|
| 842 |
+
"loss": 0.0239,
|
| 843 |
+
"step": 13900
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.28,
|
| 847 |
+
"learning_rate": 0.00014020622454954606,
|
| 848 |
+
"loss": 0.024,
|
| 849 |
+
"step": 14000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.28,
|
| 853 |
+
"learning_rate": 0.00013980426527885237,
|
| 854 |
+
"loss": 0.0234,
|
| 855 |
+
"step": 14100
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 0.28,
|
| 859 |
+
"learning_rate": 0.00013940021138653824,
|
| 860 |
+
"loss": 0.0228,
|
| 861 |
+
"step": 14200
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 0.28,
|
| 865 |
+
"learning_rate": 0.00013899407833768695,
|
| 866 |
+
"loss": 0.0228,
|
| 867 |
+
"step": 14300
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.28,
|
| 871 |
+
"learning_rate": 0.0001385858816769611,
|
| 872 |
+
"loss": 0.0234,
|
| 873 |
+
"step": 14400
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.29,
|
| 877 |
+
"learning_rate": 0.0001381756370280075,
|
| 878 |
+
"loss": 0.0233,
|
| 879 |
+
"step": 14500
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.29,
|
| 883 |
+
"learning_rate": 0.00013776336009285937,
|
| 884 |
+
"loss": 0.0216,
|
| 885 |
+
"step": 14600
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.29,
|
| 889 |
+
"learning_rate": 0.00013734906665133537,
|
| 890 |
+
"loss": 0.0268,
|
| 891 |
+
"step": 14700
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.29,
|
| 895 |
+
"learning_rate": 0.0001369327725604354,
|
| 896 |
+
"loss": 0.022,
|
| 897 |
+
"step": 14800
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.29,
|
| 901 |
+
"learning_rate": 0.00013651449375373396,
|
| 902 |
+
"loss": 0.0214,
|
| 903 |
+
"step": 14900
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.3,
|
| 907 |
+
"learning_rate": 0.00013609424624077,
|
| 908 |
+
"loss": 0.0235,
|
| 909 |
+
"step": 15000
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.3,
|
| 913 |
+
"learning_rate": 0.0001356720461064345,
|
| 914 |
+
"loss": 0.0225,
|
| 915 |
+
"step": 15100
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 0.3,
|
| 919 |
+
"learning_rate": 0.00013524790951035441,
|
| 920 |
+
"loss": 0.0218,
|
| 921 |
+
"step": 15200
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.3,
|
| 925 |
+
"learning_rate": 0.00013482185268627465,
|
| 926 |
+
"loss": 0.0225,
|
| 927 |
+
"step": 15300
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.3,
|
| 931 |
+
"learning_rate": 0.00013439389194143625,
|
| 932 |
+
"loss": 0.0216,
|
| 933 |
+
"step": 15400
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.31,
|
| 937 |
+
"learning_rate": 0.00013396404365595253,
|
| 938 |
+
"loss": 0.0223,
|
| 939 |
+
"step": 15500
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.31,
|
| 943 |
+
"learning_rate": 0.00013353232428218212,
|
| 944 |
+
"loss": 0.0212,
|
| 945 |
+
"step": 15600
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.31,
|
| 949 |
+
"learning_rate": 0.00013309875034409903,
|
| 950 |
+
"loss": 0.023,
|
| 951 |
+
"step": 15700
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 0.31,
|
| 955 |
+
"learning_rate": 0.0001326633384366605,
|
| 956 |
+
"loss": 0.0214,
|
| 957 |
+
"step": 15800
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 0.31,
|
| 961 |
+
"learning_rate": 0.0001322261052251716,
|
| 962 |
+
"loss": 0.021,
|
| 963 |
+
"step": 15900
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 0.32,
|
| 967 |
+
"learning_rate": 0.00013178706744464749,
|
| 968 |
+
"loss": 0.0211,
|
| 969 |
+
"step": 16000
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.32,
|
| 973 |
+
"learning_rate": 0.00013134624189917282,
|
| 974 |
+
"loss": 0.0212,
|
| 975 |
+
"step": 16100
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.32,
|
| 979 |
+
"learning_rate": 0.00013090808013598785,
|
| 980 |
+
"loss": 0.0252,
|
| 981 |
+
"step": 16200
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 0.32,
|
| 985 |
+
"learning_rate": 0.00013046374720137184,
|
| 986 |
+
"loss": 0.0214,
|
| 987 |
+
"step": 16300
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 0.32,
|
| 991 |
+
"learning_rate": 0.00013001767715162726,
|
| 992 |
+
"loss": 0.0202,
|
| 993 |
+
"step": 16400
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 0.32,
|
| 997 |
+
"learning_rate": 0.00012956988705999754,
|
| 998 |
+
"loss": 0.0205,
|
| 999 |
+
"step": 16500
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.33,
|
| 1003 |
+
"learning_rate": 0.00012912039406556028,
|
| 1004 |
+
"loss": 0.0235,
|
| 1005 |
+
"step": 16600
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 0.33,
|
| 1009 |
+
"learning_rate": 0.00012866921537257149,
|
| 1010 |
+
"loss": 0.0215,
|
| 1011 |
+
"step": 16700
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.33,
|
| 1015 |
+
"learning_rate": 0.00012821636824980682,
|
| 1016 |
+
"loss": 0.0198,
|
| 1017 |
+
"step": 16800
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.33,
|
| 1021 |
+
"learning_rate": 0.00012776187002990082,
|
| 1022 |
+
"loss": 0.0232,
|
| 1023 |
+
"step": 16900
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 0.33,
|
| 1027 |
+
"learning_rate": 0.00012730573810868347,
|
| 1028 |
+
"loss": 0.0235,
|
| 1029 |
+
"step": 17000
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 0.34,
|
| 1033 |
+
"learning_rate": 0.00012684798994451428,
|
| 1034 |
+
"loss": 0.0207,
|
| 1035 |
+
"step": 17100
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 0.34,
|
| 1039 |
+
"learning_rate": 0.00012638864305761417,
|
| 1040 |
+
"loss": 0.0218,
|
| 1041 |
+
"step": 17200
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.34,
|
| 1045 |
+
"learning_rate": 0.00012592771502939492,
|
| 1046 |
+
"loss": 0.0204,
|
| 1047 |
+
"step": 17300
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 0.34,
|
| 1051 |
+
"learning_rate": 0.0001254652235017861,
|
| 1052 |
+
"loss": 0.0212,
|
| 1053 |
+
"step": 17400
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.34,
|
| 1057 |
+
"learning_rate": 0.00012500118617655992,
|
| 1058 |
+
"loss": 0.0207,
|
| 1059 |
+
"step": 17500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.35,
|
| 1063 |
+
"learning_rate": 0.00012453562081465375,
|
| 1064 |
+
"loss": 0.0207,
|
| 1065 |
+
"step": 17600
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.35,
|
| 1069 |
+
"learning_rate": 0.0001240685452354902,
|
| 1070 |
+
"loss": 0.0204,
|
| 1071 |
+
"step": 17700
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 0.35,
|
| 1075 |
+
"learning_rate": 0.00012359997731629517,
|
| 1076 |
+
"loss": 0.0234,
|
| 1077 |
+
"step": 17800
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 0.35,
|
| 1081 |
+
"learning_rate": 0.00012312993499141364,
|
| 1082 |
+
"loss": 0.02,
|
| 1083 |
+
"step": 17900
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.35,
|
| 1087 |
+
"learning_rate": 0.0001226584362516231,
|
| 1088 |
+
"loss": 0.0191,
|
| 1089 |
+
"step": 18000
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.36,
|
| 1093 |
+
"learning_rate": 0.00012218549914344497,
|
| 1094 |
+
"loss": 0.0196,
|
| 1095 |
+
"step": 18100
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.36,
|
| 1099 |
+
"learning_rate": 0.00012171114176845411,
|
| 1100 |
+
"loss": 0.0203,
|
| 1101 |
+
"step": 18200
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.36,
|
| 1105 |
+
"learning_rate": 0.0001212353822825857,
|
| 1106 |
+
"loss": 0.0224,
|
| 1107 |
+
"step": 18300
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 0.36,
|
| 1111 |
+
"learning_rate": 0.00012075823889544048,
|
| 1112 |
+
"loss": 0.0211,
|
| 1113 |
+
"step": 18400
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 0.36,
|
| 1117 |
+
"learning_rate": 0.00012027972986958772,
|
| 1118 |
+
"loss": 0.0188,
|
| 1119 |
+
"step": 18500
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.37,
|
| 1123 |
+
"learning_rate": 0.0001197998735198662,
|
| 1124 |
+
"loss": 0.0214,
|
| 1125 |
+
"step": 18600
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 0.37,
|
| 1129 |
+
"learning_rate": 0.00011931868821268337,
|
| 1130 |
+
"loss": 0.0201,
|
| 1131 |
+
"step": 18700
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.37,
|
| 1135 |
+
"learning_rate": 0.00011883619236531218,
|
| 1136 |
+
"loss": 0.0206,
|
| 1137 |
+
"step": 18800
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.37,
|
| 1141 |
+
"learning_rate": 0.00011835240444518623,
|
| 1142 |
+
"loss": 0.0206,
|
| 1143 |
+
"step": 18900
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.37,
|
| 1147 |
+
"learning_rate": 0.000117867342969193,
|
| 1148 |
+
"loss": 0.0197,
|
| 1149 |
+
"step": 19000
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 0.38,
|
| 1153 |
+
"learning_rate": 0.00011738102650296509,
|
| 1154 |
+
"loss": 0.0198,
|
| 1155 |
+
"step": 19100
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 0.38,
|
| 1159 |
+
"learning_rate": 0.00011689347366016955,
|
| 1160 |
+
"loss": 0.0211,
|
| 1161 |
+
"step": 19200
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.38,
|
| 1165 |
+
"learning_rate": 0.0001164047031017955,
|
| 1166 |
+
"loss": 0.0204,
|
| 1167 |
+
"step": 19300
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.38,
|
| 1171 |
+
"learning_rate": 0.00011591473353543992,
|
| 1172 |
+
"loss": 0.0206,
|
| 1173 |
+
"step": 19400
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 0.38,
|
| 1177 |
+
"learning_rate": 0.00011542358371459156,
|
| 1178 |
+
"loss": 0.0192,
|
| 1179 |
+
"step": 19500
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.39,
|
| 1183 |
+
"learning_rate": 0.0001149312724379132,
|
| 1184 |
+
"loss": 0.0202,
|
| 1185 |
+
"step": 19600
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.39,
|
| 1189 |
+
"learning_rate": 0.00011443781854852212,
|
| 1190 |
+
"loss": 0.0197,
|
| 1191 |
+
"step": 19700
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 0.39,
|
| 1195 |
+
"learning_rate": 0.0001139432409332688,
|
| 1196 |
+
"loss": 0.021,
|
| 1197 |
+
"step": 19800
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 0.39,
|
| 1201 |
+
"learning_rate": 0.00011344755852201423,
|
| 1202 |
+
"loss": 0.022,
|
| 1203 |
+
"step": 19900
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.39,
|
| 1207 |
+
"learning_rate": 0.00011295079028690513,
|
| 1208 |
+
"loss": 0.0196,
|
| 1209 |
+
"step": 20000
|
| 1210 |
+
}
|
| 1211 |
+
],
|
| 1212 |
+
"logging_steps": 100,
|
| 1213 |
+
"max_steps": 50780,
|
| 1214 |
+
"num_train_epochs": 1,
|
| 1215 |
+
"save_steps": 10000,
|
| 1216 |
+
"total_flos": 7.985674727424e+18,
|
| 1217 |
+
"trial_name": null,
|
| 1218 |
+
"trial_params": null
|
| 1219 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d17a2c3c8a65881aa45bfd2997dfe652a61cb20d8b49594dc95bff2d7669f6f
|
| 3 |
+
size 4155
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-20000/vocab.txt
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
<cls>
|
| 2 |
+
<pad>
|
| 3 |
+
<eos>
|
| 4 |
+
<unk>
|
| 5 |
+
L
|
| 6 |
+
A
|
| 7 |
+
G
|
| 8 |
+
V
|
| 9 |
+
S
|
| 10 |
+
E
|
| 11 |
+
R
|
| 12 |
+
T
|
| 13 |
+
I
|
| 14 |
+
D
|
| 15 |
+
P
|
| 16 |
+
K
|
| 17 |
+
Q
|
| 18 |
+
N
|
| 19 |
+
F
|
| 20 |
+
Y
|
| 21 |
+
M
|
| 22 |
+
H
|
| 23 |
+
W
|
| 24 |
+
C
|
| 25 |
+
X
|
| 26 |
+
B
|
| 27 |
+
U
|
| 28 |
+
Z
|
| 29 |
+
O
|
| 30 |
+
.
|
| 31 |
+
-
|
| 32 |
+
<null_1>
|
| 33 |
+
<mask>
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/README.md
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
| 4 |
+
## Training procedure
|
| 5 |
+
|
| 6 |
+
|
| 7 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 8 |
+
- quant_method: bitsandbytes
|
| 9 |
+
- load_in_8bit: False
|
| 10 |
+
- load_in_4bit: True
|
| 11 |
+
- llm_int8_threshold: 6.0
|
| 12 |
+
- llm_int8_skip_modules: None
|
| 13 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 14 |
+
- llm_int8_has_fp16_weight: False
|
| 15 |
+
- bnb_4bit_quant_type: nf4
|
| 16 |
+
- bnb_4bit_use_double_quant: True
|
| 17 |
+
- bnb_4bit_compute_dtype: bfloat16
|
| 18 |
+
### Framework versions
|
| 19 |
+
|
| 20 |
+
|
| 21 |
+
- PEFT 0.6.0.dev0
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/adapter_config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "facebook/esm2_t33_650M_UR50D",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"fan_in_fan_out": false,
|
| 7 |
+
"inference_mode": true,
|
| 8 |
+
"init_lora_weights": true,
|
| 9 |
+
"layers_pattern": null,
|
| 10 |
+
"layers_to_transform": null,
|
| 11 |
+
"lora_alpha": 1,
|
| 12 |
+
"lora_dropout": 0.5,
|
| 13 |
+
"modules_to_save": null,
|
| 14 |
+
"peft_type": "LORA",
|
| 15 |
+
"r": 2,
|
| 16 |
+
"rank_pattern": {},
|
| 17 |
+
"revision": null,
|
| 18 |
+
"target_modules": [
|
| 19 |
+
"query",
|
| 20 |
+
"key",
|
| 21 |
+
"value",
|
| 22 |
+
"EsmSelfOutput.dense",
|
| 23 |
+
"EsmIntermediate.dense",
|
| 24 |
+
"EsmOutput.dense",
|
| 25 |
+
"classifier"
|
| 26 |
+
],
|
| 27 |
+
"task_type": "TOKEN_CLS"
|
| 28 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7e69ef69f9b32fffd90f8b84442da5df6abcbe5b65c8920aa28b43bdf626b909
|
| 3 |
+
size 2133117
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/added_tokens.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<cls>": 0,
|
| 3 |
+
"<eos>": 2,
|
| 4 |
+
"<mask>": 32,
|
| 5 |
+
"<pad>": 1,
|
| 6 |
+
"<unk>": 3
|
| 7 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a10e147fd00c89350f232720ea73d96536d1350be6ee6e230f682e6f9187feb4
|
| 3 |
+
size 4215045
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:601d282addfa26c7c4f37873b2314e1b151fe139d35103c2187ee704954f71d4
|
| 3 |
+
size 14575
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1697708f007c343ee0b9950cd53955c3e22105d9fb14648f4201068a99d7e02d
|
| 3 |
+
size 627
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "<cls>",
|
| 3 |
+
"eos_token": "<eos>",
|
| 4 |
+
"mask_token": "<mask>",
|
| 5 |
+
"pad_token": "<pad>",
|
| 6 |
+
"unk_token": "<unk>"
|
| 7 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<cls>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "<eos>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"32": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"additional_special_tokens": [],
|
| 45 |
+
"clean_up_tokenization_spaces": true,
|
| 46 |
+
"cls_token": "<cls>",
|
| 47 |
+
"eos_token": "<eos>",
|
| 48 |
+
"mask_token": "<mask>",
|
| 49 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 50 |
+
"pad_token": "<pad>",
|
| 51 |
+
"tokenizer_class": "EsmTokenizer",
|
| 52 |
+
"tokenizer_file": null,
|
| 53 |
+
"unk_token": "<unk>"
|
| 54 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/trainer_state.json
ADDED
|
@@ -0,0 +1,1819 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.5907837731390311,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 30000,
|
| 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": 0.00017015520980041642,
|
| 14 |
+
"loss": 0.5575,
|
| 15 |
+
"step": 100
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.0,
|
| 19 |
+
"learning_rate": 0.00017015035788629153,
|
| 20 |
+
"loss": 0.3065,
|
| 21 |
+
"step": 200
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.01,
|
| 25 |
+
"learning_rate": 0.00017014224985982174,
|
| 26 |
+
"loss": 0.2229,
|
| 27 |
+
"step": 300
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.01,
|
| 31 |
+
"learning_rate": 0.0001701308860313402,
|
| 32 |
+
"loss": 0.1812,
|
| 33 |
+
"step": 400
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.01,
|
| 37 |
+
"learning_rate": 0.00017011626683579524,
|
| 38 |
+
"loss": 0.1572,
|
| 39 |
+
"step": 500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.01,
|
| 43 |
+
"learning_rate": 0.00017009839283273364,
|
| 44 |
+
"loss": 0.1335,
|
| 45 |
+
"step": 600
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.01,
|
| 49 |
+
"learning_rate": 0.00017007726470627936,
|
| 50 |
+
"loss": 0.1193,
|
| 51 |
+
"step": 700
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.02,
|
| 55 |
+
"learning_rate": 0.00017005288326510734,
|
| 56 |
+
"loss": 0.1075,
|
| 57 |
+
"step": 800
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.02,
|
| 61 |
+
"learning_rate": 0.0001700252494424124,
|
| 62 |
+
"loss": 0.0986,
|
| 63 |
+
"step": 900
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.02,
|
| 67 |
+
"learning_rate": 0.00016999436429587366,
|
| 68 |
+
"loss": 0.0952,
|
| 69 |
+
"step": 1000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.02,
|
| 73 |
+
"learning_rate": 0.00016996022900761407,
|
| 74 |
+
"loss": 0.0865,
|
| 75 |
+
"step": 1100
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.02,
|
| 79 |
+
"learning_rate": 0.000169922844884155,
|
| 80 |
+
"loss": 0.0816,
|
| 81 |
+
"step": 1200
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.03,
|
| 85 |
+
"learning_rate": 0.00016988221335636648,
|
| 86 |
+
"loss": 0.0775,
|
| 87 |
+
"step": 1300
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.03,
|
| 91 |
+
"learning_rate": 0.00016983833597941224,
|
| 92 |
+
"loss": 0.0744,
|
| 93 |
+
"step": 1400
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.03,
|
| 97 |
+
"learning_rate": 0.00016979121443269025,
|
| 98 |
+
"loss": 0.0694,
|
| 99 |
+
"step": 1500
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.03,
|
| 103 |
+
"learning_rate": 0.0001697408505197684,
|
| 104 |
+
"loss": 0.0641,
|
| 105 |
+
"step": 1600
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.03,
|
| 109 |
+
"learning_rate": 0.00016968724616831557,
|
| 110 |
+
"loss": 0.0629,
|
| 111 |
+
"step": 1700
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.04,
|
| 115 |
+
"learning_rate": 0.0001696304034300278,
|
| 116 |
+
"loss": 0.0696,
|
| 117 |
+
"step": 1800
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.04,
|
| 121 |
+
"learning_rate": 0.00016957032448054968,
|
| 122 |
+
"loss": 0.0591,
|
| 123 |
+
"step": 1900
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.04,
|
| 127 |
+
"learning_rate": 0.0001695070116193912,
|
| 128 |
+
"loss": 0.06,
|
| 129 |
+
"step": 2000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.04,
|
| 133 |
+
"learning_rate": 0.0001694404672698396,
|
| 134 |
+
"loss": 0.0558,
|
| 135 |
+
"step": 2100
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.04,
|
| 139 |
+
"learning_rate": 0.00016937069397886687,
|
| 140 |
+
"loss": 0.0529,
|
| 141 |
+
"step": 2200
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.05,
|
| 145 |
+
"learning_rate": 0.00016929769441703196,
|
| 146 |
+
"loss": 0.0536,
|
| 147 |
+
"step": 2300
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.05,
|
| 151 |
+
"learning_rate": 0.00016922147137837868,
|
| 152 |
+
"loss": 0.0537,
|
| 153 |
+
"step": 2400
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.05,
|
| 157 |
+
"learning_rate": 0.00016914202778032893,
|
| 158 |
+
"loss": 0.0508,
|
| 159 |
+
"step": 2500
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.05,
|
| 163 |
+
"learning_rate": 0.0001690593666635707,
|
| 164 |
+
"loss": 0.0524,
|
| 165 |
+
"step": 2600
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.05,
|
| 169 |
+
"learning_rate": 0.00016897349119194207,
|
| 170 |
+
"loss": 0.0494,
|
| 171 |
+
"step": 2700
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.06,
|
| 175 |
+
"learning_rate": 0.00016888440465230977,
|
| 176 |
+
"loss": 0.047,
|
| 177 |
+
"step": 2800
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.06,
|
| 181 |
+
"learning_rate": 0.00016879211045444354,
|
| 182 |
+
"loss": 0.0502,
|
| 183 |
+
"step": 2900
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.06,
|
| 187 |
+
"learning_rate": 0.00016869661213088575,
|
| 188 |
+
"loss": 0.046,
|
| 189 |
+
"step": 3000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.06,
|
| 193 |
+
"learning_rate": 0.00016859791333681583,
|
| 194 |
+
"loss": 0.0419,
|
| 195 |
+
"step": 3100
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.06,
|
| 199 |
+
"learning_rate": 0.0001684960178499108,
|
| 200 |
+
"loss": 0.0456,
|
| 201 |
+
"step": 3200
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.06,
|
| 205 |
+
"learning_rate": 0.00016839092957020028,
|
| 206 |
+
"loss": 0.0432,
|
| 207 |
+
"step": 3300
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.07,
|
| 211 |
+
"learning_rate": 0.00016828265251991761,
|
| 212 |
+
"loss": 0.0423,
|
| 213 |
+
"step": 3400
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.07,
|
| 217 |
+
"learning_rate": 0.00016817119084334555,
|
| 218 |
+
"loss": 0.0407,
|
| 219 |
+
"step": 3500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.07,
|
| 223 |
+
"learning_rate": 0.00016805654880665776,
|
| 224 |
+
"loss": 0.0409,
|
| 225 |
+
"step": 3600
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.07,
|
| 229 |
+
"learning_rate": 0.00016793873079775577,
|
| 230 |
+
"loss": 0.0397,
|
| 231 |
+
"step": 3700
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.07,
|
| 235 |
+
"learning_rate": 0.00016781774132610059,
|
| 236 |
+
"loss": 0.0419,
|
| 237 |
+
"step": 3800
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.08,
|
| 241 |
+
"learning_rate": 0.0001676935850225405,
|
| 242 |
+
"loss": 0.0399,
|
| 243 |
+
"step": 3900
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.08,
|
| 247 |
+
"learning_rate": 0.00016756626663913358,
|
| 248 |
+
"loss": 0.0389,
|
| 249 |
+
"step": 4000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.08,
|
| 253 |
+
"learning_rate": 0.00016743579104896593,
|
| 254 |
+
"loss": 0.0385,
|
| 255 |
+
"step": 4100
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.08,
|
| 259 |
+
"learning_rate": 0.00016730216324596504,
|
| 260 |
+
"loss": 0.0357,
|
| 261 |
+
"step": 4200
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.08,
|
| 265 |
+
"learning_rate": 0.0001671653883447088,
|
| 266 |
+
"loss": 0.0373,
|
| 267 |
+
"step": 4300
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.09,
|
| 271 |
+
"learning_rate": 0.00016702547158022968,
|
| 272 |
+
"loss": 0.0375,
|
| 273 |
+
"step": 4400
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.09,
|
| 277 |
+
"learning_rate": 0.0001668824183078143,
|
| 278 |
+
"loss": 0.0393,
|
| 279 |
+
"step": 4500
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.09,
|
| 283 |
+
"learning_rate": 0.00016673623400279849,
|
| 284 |
+
"loss": 0.0358,
|
| 285 |
+
"step": 4600
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.09,
|
| 289 |
+
"learning_rate": 0.00016658692426035782,
|
| 290 |
+
"loss": 0.0334,
|
| 291 |
+
"step": 4700
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.09,
|
| 295 |
+
"learning_rate": 0.00016643449479529325,
|
| 296 |
+
"loss": 0.035,
|
| 297 |
+
"step": 4800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.1,
|
| 301 |
+
"learning_rate": 0.00016627895144181258,
|
| 302 |
+
"loss": 0.0343,
|
| 303 |
+
"step": 4900
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.1,
|
| 307 |
+
"learning_rate": 0.0001661203001533071,
|
| 308 |
+
"loss": 0.0335,
|
| 309 |
+
"step": 5000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.1,
|
| 313 |
+
"learning_rate": 0.00016595854700212362,
|
| 314 |
+
"loss": 0.0353,
|
| 315 |
+
"step": 5100
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.1,
|
| 319 |
+
"learning_rate": 0.0001657936981793322,
|
| 320 |
+
"loss": 0.0331,
|
| 321 |
+
"step": 5200
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.1,
|
| 325 |
+
"learning_rate": 0.0001656257599944891,
|
| 326 |
+
"loss": 0.0387,
|
| 327 |
+
"step": 5300
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.11,
|
| 331 |
+
"learning_rate": 0.00016545473887539532,
|
| 332 |
+
"loss": 0.0327,
|
| 333 |
+
"step": 5400
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.11,
|
| 337 |
+
"learning_rate": 0.00016528064136785056,
|
| 338 |
+
"loss": 0.0316,
|
| 339 |
+
"step": 5500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.11,
|
| 343 |
+
"learning_rate": 0.00016510347413540262,
|
| 344 |
+
"loss": 0.0319,
|
| 345 |
+
"step": 5600
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.11,
|
| 349 |
+
"learning_rate": 0.0001649232439590925,
|
| 350 |
+
"loss": 0.0314,
|
| 351 |
+
"step": 5700
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.11,
|
| 355 |
+
"learning_rate": 0.0001647399577371947,
|
| 356 |
+
"loss": 0.0321,
|
| 357 |
+
"step": 5800
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.12,
|
| 361 |
+
"learning_rate": 0.00016455362248495338,
|
| 362 |
+
"loss": 0.0333,
|
| 363 |
+
"step": 5900
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.12,
|
| 367 |
+
"learning_rate": 0.00016436424533431362,
|
| 368 |
+
"loss": 0.0319,
|
| 369 |
+
"step": 6000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.12,
|
| 373 |
+
"learning_rate": 0.0001641718335336486,
|
| 374 |
+
"loss": 0.0315,
|
| 375 |
+
"step": 6100
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.12,
|
| 379 |
+
"learning_rate": 0.0001639763944474821,
|
| 380 |
+
"loss": 0.0311,
|
| 381 |
+
"step": 6200
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.12,
|
| 385 |
+
"learning_rate": 0.0001637779355562068,
|
| 386 |
+
"loss": 0.031,
|
| 387 |
+
"step": 6300
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.13,
|
| 391 |
+
"learning_rate": 0.00016357646445579763,
|
| 392 |
+
"loss": 0.0299,
|
| 393 |
+
"step": 6400
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.13,
|
| 397 |
+
"learning_rate": 0.00016337198885752133,
|
| 398 |
+
"loss": 0.0317,
|
| 399 |
+
"step": 6500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.13,
|
| 403 |
+
"learning_rate": 0.00016316451658764122,
|
| 404 |
+
"loss": 0.0302,
|
| 405 |
+
"step": 6600
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.13,
|
| 409 |
+
"learning_rate": 0.0001629540555871176,
|
| 410 |
+
"loss": 0.0295,
|
| 411 |
+
"step": 6700
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 0.13,
|
| 415 |
+
"learning_rate": 0.00016274061391130388,
|
| 416 |
+
"loss": 0.03,
|
| 417 |
+
"step": 6800
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.14,
|
| 421 |
+
"learning_rate": 0.0001625241997296382,
|
| 422 |
+
"loss": 0.0292,
|
| 423 |
+
"step": 6900
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.14,
|
| 427 |
+
"learning_rate": 0.00016230482132533077,
|
| 428 |
+
"loss": 0.0289,
|
| 429 |
+
"step": 7000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.14,
|
| 433 |
+
"learning_rate": 0.00016208472504084003,
|
| 434 |
+
"loss": 0.0318,
|
| 435 |
+
"step": 7100
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.14,
|
| 439 |
+
"learning_rate": 0.0001618594729250462,
|
| 440 |
+
"loss": 0.0301,
|
| 441 |
+
"step": 7200
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.14,
|
| 445 |
+
"learning_rate": 0.00016163128202889828,
|
| 446 |
+
"loss": 0.0295,
|
| 447 |
+
"step": 7300
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.15,
|
| 451 |
+
"learning_rate": 0.00016140016108635798,
|
| 452 |
+
"loss": 0.029,
|
| 453 |
+
"step": 7400
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 0.15,
|
| 457 |
+
"learning_rate": 0.00016116611894353386,
|
| 458 |
+
"loss": 0.0291,
|
| 459 |
+
"step": 7500
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.15,
|
| 463 |
+
"learning_rate": 0.00016092916455834295,
|
| 464 |
+
"loss": 0.0311,
|
| 465 |
+
"step": 7600
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.15,
|
| 469 |
+
"learning_rate": 0.00016068930700016766,
|
| 470 |
+
"loss": 0.0285,
|
| 471 |
+
"step": 7700
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.15,
|
| 475 |
+
"learning_rate": 0.00016044655544950889,
|
| 476 |
+
"loss": 0.0287,
|
| 477 |
+
"step": 7800
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.16,
|
| 481 |
+
"learning_rate": 0.00016020091919763445,
|
| 482 |
+
"loss": 0.0293,
|
| 483 |
+
"step": 7900
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 0.16,
|
| 487 |
+
"learning_rate": 0.00015995240764622357,
|
| 488 |
+
"loss": 0.0259,
|
| 489 |
+
"step": 8000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.16,
|
| 493 |
+
"learning_rate": 0.000159701030307007,
|
| 494 |
+
"loss": 0.0293,
|
| 495 |
+
"step": 8100
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.16,
|
| 499 |
+
"learning_rate": 0.00015944679680140295,
|
| 500 |
+
"loss": 0.0277,
|
| 501 |
+
"step": 8200
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.16,
|
| 505 |
+
"learning_rate": 0.0001591897168601488,
|
| 506 |
+
"loss": 0.0304,
|
| 507 |
+
"step": 8300
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.17,
|
| 511 |
+
"learning_rate": 0.00015892980032292876,
|
| 512 |
+
"loss": 0.026,
|
| 513 |
+
"step": 8400
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.17,
|
| 517 |
+
"learning_rate": 0.00015866705713799714,
|
| 518 |
+
"loss": 0.0294,
|
| 519 |
+
"step": 8500
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.17,
|
| 523 |
+
"learning_rate": 0.00015840149736179762,
|
| 524 |
+
"loss": 0.0321,
|
| 525 |
+
"step": 8600
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.17,
|
| 529 |
+
"learning_rate": 0.0001581331311585785,
|
| 530 |
+
"loss": 0.0279,
|
| 531 |
+
"step": 8700
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.17,
|
| 535 |
+
"learning_rate": 0.00015786196880000325,
|
| 536 |
+
"loss": 0.0277,
|
| 537 |
+
"step": 8800
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.18,
|
| 541 |
+
"learning_rate": 0.0001575880206647579,
|
| 542 |
+
"loss": 0.0268,
|
| 543 |
+
"step": 8900
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.18,
|
| 547 |
+
"learning_rate": 0.00015731129723815343,
|
| 548 |
+
"loss": 0.0281,
|
| 549 |
+
"step": 9000
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.18,
|
| 553 |
+
"learning_rate": 0.00015703180911172453,
|
| 554 |
+
"loss": 0.028,
|
| 555 |
+
"step": 9100
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.18,
|
| 559 |
+
"learning_rate": 0.0001567495669828243,
|
| 560 |
+
"loss": 0.026,
|
| 561 |
+
"step": 9200
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.18,
|
| 565 |
+
"learning_rate": 0.0001564645816542146,
|
| 566 |
+
"loss": 0.0256,
|
| 567 |
+
"step": 9300
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 0.19,
|
| 571 |
+
"learning_rate": 0.0001561768640336529,
|
| 572 |
+
"loss": 0.027,
|
| 573 |
+
"step": 9400
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 0.19,
|
| 577 |
+
"learning_rate": 0.0001558864251334745,
|
| 578 |
+
"loss": 0.0249,
|
| 579 |
+
"step": 9500
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.19,
|
| 583 |
+
"learning_rate": 0.00015559327607017119,
|
| 584 |
+
"loss": 0.0256,
|
| 585 |
+
"step": 9600
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.19,
|
| 589 |
+
"learning_rate": 0.00015529742806396564,
|
| 590 |
+
"loss": 0.0251,
|
| 591 |
+
"step": 9700
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.19,
|
| 595 |
+
"learning_rate": 0.00015499889243838211,
|
| 596 |
+
"loss": 0.0257,
|
| 597 |
+
"step": 9800
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.19,
|
| 601 |
+
"learning_rate": 0.00015469768061981295,
|
| 602 |
+
"loss": 0.0264,
|
| 603 |
+
"step": 9900
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.2,
|
| 607 |
+
"learning_rate": 0.00015439685605389073,
|
| 608 |
+
"loss": 0.0247,
|
| 609 |
+
"step": 10000
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.2,
|
| 613 |
+
"learning_rate": 0.00015409035301023626,
|
| 614 |
+
"loss": 0.0256,
|
| 615 |
+
"step": 10100
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 0.2,
|
| 619 |
+
"learning_rate": 0.00015378120854776404,
|
| 620 |
+
"loss": 0.0247,
|
| 621 |
+
"step": 10200
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.2,
|
| 625 |
+
"learning_rate": 0.00015346943449891755,
|
| 626 |
+
"loss": 0.0254,
|
| 627 |
+
"step": 10300
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.2,
|
| 631 |
+
"learning_rate": 0.00015315504279678714,
|
| 632 |
+
"loss": 0.0246,
|
| 633 |
+
"step": 10400
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.21,
|
| 637 |
+
"learning_rate": 0.00015283804547465337,
|
| 638 |
+
"loss": 0.0249,
|
| 639 |
+
"step": 10500
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.21,
|
| 643 |
+
"learning_rate": 0.0001525184546655264,
|
| 644 |
+
"loss": 0.026,
|
| 645 |
+
"step": 10600
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 0.21,
|
| 649 |
+
"learning_rate": 0.0001521962826016816,
|
| 650 |
+
"loss": 0.0254,
|
| 651 |
+
"step": 10700
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 0.21,
|
| 655 |
+
"learning_rate": 0.00015187154161419122,
|
| 656 |
+
"loss": 0.0272,
|
| 657 |
+
"step": 10800
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.21,
|
| 661 |
+
"learning_rate": 0.0001515442441324528,
|
| 662 |
+
"loss": 0.0253,
|
| 663 |
+
"step": 10900
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 0.22,
|
| 667 |
+
"learning_rate": 0.00015121440268371297,
|
| 668 |
+
"loss": 0.0263,
|
| 669 |
+
"step": 11000
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.22,
|
| 673 |
+
"learning_rate": 0.00015088202989258835,
|
| 674 |
+
"loss": 0.0256,
|
| 675 |
+
"step": 11100
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.22,
|
| 679 |
+
"learning_rate": 0.00015054713848058212,
|
| 680 |
+
"loss": 0.023,
|
| 681 |
+
"step": 11200
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.22,
|
| 685 |
+
"learning_rate": 0.00015020974126559716,
|
| 686 |
+
"loss": 0.0236,
|
| 687 |
+
"step": 11300
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 0.22,
|
| 691 |
+
"learning_rate": 0.00014986985116144554,
|
| 692 |
+
"loss": 0.0241,
|
| 693 |
+
"step": 11400
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 0.23,
|
| 697 |
+
"learning_rate": 0.00014952748117735409,
|
| 698 |
+
"loss": 0.0235,
|
| 699 |
+
"step": 11500
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.23,
|
| 703 |
+
"learning_rate": 0.0001491826444174666,
|
| 704 |
+
"loss": 0.0242,
|
| 705 |
+
"step": 11600
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.23,
|
| 709 |
+
"learning_rate": 0.00014883535408034227,
|
| 710 |
+
"loss": 0.0233,
|
| 711 |
+
"step": 11700
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.23,
|
| 715 |
+
"learning_rate": 0.00014848562345845032,
|
| 716 |
+
"loss": 0.0228,
|
| 717 |
+
"step": 11800
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.23,
|
| 721 |
+
"learning_rate": 0.00014813346593766164,
|
| 722 |
+
"loss": 0.023,
|
| 723 |
+
"step": 11900
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.24,
|
| 727 |
+
"learning_rate": 0.0001477788949967359,
|
| 728 |
+
"loss": 0.0236,
|
| 729 |
+
"step": 12000
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.24,
|
| 733 |
+
"learning_rate": 0.00014742192420680626,
|
| 734 |
+
"loss": 0.0237,
|
| 735 |
+
"step": 12100
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.24,
|
| 739 |
+
"learning_rate": 0.00014706256723085937,
|
| 740 |
+
"loss": 0.0237,
|
| 741 |
+
"step": 12200
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 0.24,
|
| 745 |
+
"learning_rate": 0.0001467008378232128,
|
| 746 |
+
"loss": 0.024,
|
| 747 |
+
"step": 12300
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.24,
|
| 751 |
+
"learning_rate": 0.00014633674982898854,
|
| 752 |
+
"loss": 0.0227,
|
| 753 |
+
"step": 12400
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.25,
|
| 757 |
+
"learning_rate": 0.00014597031718358285,
|
| 758 |
+
"loss": 0.0247,
|
| 759 |
+
"step": 12500
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.25,
|
| 763 |
+
"learning_rate": 0.00014560155391213318,
|
| 764 |
+
"loss": 0.0233,
|
| 765 |
+
"step": 12600
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.25,
|
| 769 |
+
"learning_rate": 0.0001452304741289812,
|
| 770 |
+
"loss": 0.0234,
|
| 771 |
+
"step": 12700
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 0.25,
|
| 775 |
+
"learning_rate": 0.00014485709203713263,
|
| 776 |
+
"loss": 0.0221,
|
| 777 |
+
"step": 12800
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.25,
|
| 781 |
+
"learning_rate": 0.00014448142192771354,
|
| 782 |
+
"loss": 0.0244,
|
| 783 |
+
"step": 12900
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.26,
|
| 787 |
+
"learning_rate": 0.00014410347817942347,
|
| 788 |
+
"loss": 0.0242,
|
| 789 |
+
"step": 13000
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.26,
|
| 793 |
+
"learning_rate": 0.00014372327525798503,
|
| 794 |
+
"loss": 0.0231,
|
| 795 |
+
"step": 13100
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.26,
|
| 799 |
+
"learning_rate": 0.00014334082771559026,
|
| 800 |
+
"loss": 0.0238,
|
| 801 |
+
"step": 13200
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.26,
|
| 805 |
+
"learning_rate": 0.00014295615019034358,
|
| 806 |
+
"loss": 0.0233,
|
| 807 |
+
"step": 13300
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.26,
|
| 811 |
+
"learning_rate": 0.0001425692574057016,
|
| 812 |
+
"loss": 0.024,
|
| 813 |
+
"step": 13400
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 0.27,
|
| 817 |
+
"learning_rate": 0.00014218016416990954,
|
| 818 |
+
"loss": 0.0249,
|
| 819 |
+
"step": 13500
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.27,
|
| 823 |
+
"learning_rate": 0.0001417928089327897,
|
| 824 |
+
"loss": 0.0265,
|
| 825 |
+
"step": 13600
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.27,
|
| 829 |
+
"learning_rate": 0.00014139938118717103,
|
| 830 |
+
"loss": 0.0206,
|
| 831 |
+
"step": 13700
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.27,
|
| 835 |
+
"learning_rate": 0.00014100379776718443,
|
| 836 |
+
"loss": 0.0216,
|
| 837 |
+
"step": 13800
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.27,
|
| 841 |
+
"learning_rate": 0.00014060607381370744,
|
| 842 |
+
"loss": 0.0239,
|
| 843 |
+
"step": 13900
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.28,
|
| 847 |
+
"learning_rate": 0.00014020622454954606,
|
| 848 |
+
"loss": 0.024,
|
| 849 |
+
"step": 14000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.28,
|
| 853 |
+
"learning_rate": 0.00013980426527885237,
|
| 854 |
+
"loss": 0.0234,
|
| 855 |
+
"step": 14100
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 0.28,
|
| 859 |
+
"learning_rate": 0.00013940021138653824,
|
| 860 |
+
"loss": 0.0228,
|
| 861 |
+
"step": 14200
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 0.28,
|
| 865 |
+
"learning_rate": 0.00013899407833768695,
|
| 866 |
+
"loss": 0.0228,
|
| 867 |
+
"step": 14300
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.28,
|
| 871 |
+
"learning_rate": 0.0001385858816769611,
|
| 872 |
+
"loss": 0.0234,
|
| 873 |
+
"step": 14400
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.29,
|
| 877 |
+
"learning_rate": 0.0001381756370280075,
|
| 878 |
+
"loss": 0.0233,
|
| 879 |
+
"step": 14500
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.29,
|
| 883 |
+
"learning_rate": 0.00013776336009285937,
|
| 884 |
+
"loss": 0.0216,
|
| 885 |
+
"step": 14600
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.29,
|
| 889 |
+
"learning_rate": 0.00013734906665133537,
|
| 890 |
+
"loss": 0.0268,
|
| 891 |
+
"step": 14700
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.29,
|
| 895 |
+
"learning_rate": 0.0001369327725604354,
|
| 896 |
+
"loss": 0.022,
|
| 897 |
+
"step": 14800
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.29,
|
| 901 |
+
"learning_rate": 0.00013651449375373396,
|
| 902 |
+
"loss": 0.0214,
|
| 903 |
+
"step": 14900
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.3,
|
| 907 |
+
"learning_rate": 0.00013609424624077,
|
| 908 |
+
"loss": 0.0235,
|
| 909 |
+
"step": 15000
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.3,
|
| 913 |
+
"learning_rate": 0.0001356720461064345,
|
| 914 |
+
"loss": 0.0225,
|
| 915 |
+
"step": 15100
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 0.3,
|
| 919 |
+
"learning_rate": 0.00013524790951035441,
|
| 920 |
+
"loss": 0.0218,
|
| 921 |
+
"step": 15200
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.3,
|
| 925 |
+
"learning_rate": 0.00013482185268627465,
|
| 926 |
+
"loss": 0.0225,
|
| 927 |
+
"step": 15300
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.3,
|
| 931 |
+
"learning_rate": 0.00013439389194143625,
|
| 932 |
+
"loss": 0.0216,
|
| 933 |
+
"step": 15400
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.31,
|
| 937 |
+
"learning_rate": 0.00013396404365595253,
|
| 938 |
+
"loss": 0.0223,
|
| 939 |
+
"step": 15500
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.31,
|
| 943 |
+
"learning_rate": 0.00013353232428218212,
|
| 944 |
+
"loss": 0.0212,
|
| 945 |
+
"step": 15600
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.31,
|
| 949 |
+
"learning_rate": 0.00013309875034409903,
|
| 950 |
+
"loss": 0.023,
|
| 951 |
+
"step": 15700
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 0.31,
|
| 955 |
+
"learning_rate": 0.0001326633384366605,
|
| 956 |
+
"loss": 0.0214,
|
| 957 |
+
"step": 15800
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 0.31,
|
| 961 |
+
"learning_rate": 0.0001322261052251716,
|
| 962 |
+
"loss": 0.021,
|
| 963 |
+
"step": 15900
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 0.32,
|
| 967 |
+
"learning_rate": 0.00013178706744464749,
|
| 968 |
+
"loss": 0.0211,
|
| 969 |
+
"step": 16000
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.32,
|
| 973 |
+
"learning_rate": 0.00013134624189917282,
|
| 974 |
+
"loss": 0.0212,
|
| 975 |
+
"step": 16100
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.32,
|
| 979 |
+
"learning_rate": 0.00013090808013598785,
|
| 980 |
+
"loss": 0.0252,
|
| 981 |
+
"step": 16200
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 0.32,
|
| 985 |
+
"learning_rate": 0.00013046374720137184,
|
| 986 |
+
"loss": 0.0214,
|
| 987 |
+
"step": 16300
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 0.32,
|
| 991 |
+
"learning_rate": 0.00013001767715162726,
|
| 992 |
+
"loss": 0.0202,
|
| 993 |
+
"step": 16400
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 0.32,
|
| 997 |
+
"learning_rate": 0.00012956988705999754,
|
| 998 |
+
"loss": 0.0205,
|
| 999 |
+
"step": 16500
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.33,
|
| 1003 |
+
"learning_rate": 0.00012912039406556028,
|
| 1004 |
+
"loss": 0.0235,
|
| 1005 |
+
"step": 16600
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 0.33,
|
| 1009 |
+
"learning_rate": 0.00012866921537257149,
|
| 1010 |
+
"loss": 0.0215,
|
| 1011 |
+
"step": 16700
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.33,
|
| 1015 |
+
"learning_rate": 0.00012821636824980682,
|
| 1016 |
+
"loss": 0.0198,
|
| 1017 |
+
"step": 16800
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.33,
|
| 1021 |
+
"learning_rate": 0.00012776187002990082,
|
| 1022 |
+
"loss": 0.0232,
|
| 1023 |
+
"step": 16900
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 0.33,
|
| 1027 |
+
"learning_rate": 0.00012730573810868347,
|
| 1028 |
+
"loss": 0.0235,
|
| 1029 |
+
"step": 17000
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 0.34,
|
| 1033 |
+
"learning_rate": 0.00012684798994451428,
|
| 1034 |
+
"loss": 0.0207,
|
| 1035 |
+
"step": 17100
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 0.34,
|
| 1039 |
+
"learning_rate": 0.00012638864305761417,
|
| 1040 |
+
"loss": 0.0218,
|
| 1041 |
+
"step": 17200
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.34,
|
| 1045 |
+
"learning_rate": 0.00012592771502939492,
|
| 1046 |
+
"loss": 0.0204,
|
| 1047 |
+
"step": 17300
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 0.34,
|
| 1051 |
+
"learning_rate": 0.0001254652235017861,
|
| 1052 |
+
"loss": 0.0212,
|
| 1053 |
+
"step": 17400
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.34,
|
| 1057 |
+
"learning_rate": 0.00012500118617655992,
|
| 1058 |
+
"loss": 0.0207,
|
| 1059 |
+
"step": 17500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.35,
|
| 1063 |
+
"learning_rate": 0.00012453562081465375,
|
| 1064 |
+
"loss": 0.0207,
|
| 1065 |
+
"step": 17600
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.35,
|
| 1069 |
+
"learning_rate": 0.0001240685452354902,
|
| 1070 |
+
"loss": 0.0204,
|
| 1071 |
+
"step": 17700
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 0.35,
|
| 1075 |
+
"learning_rate": 0.00012359997731629517,
|
| 1076 |
+
"loss": 0.0234,
|
| 1077 |
+
"step": 17800
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 0.35,
|
| 1081 |
+
"learning_rate": 0.00012312993499141364,
|
| 1082 |
+
"loss": 0.02,
|
| 1083 |
+
"step": 17900
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.35,
|
| 1087 |
+
"learning_rate": 0.0001226584362516231,
|
| 1088 |
+
"loss": 0.0191,
|
| 1089 |
+
"step": 18000
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.36,
|
| 1093 |
+
"learning_rate": 0.00012218549914344497,
|
| 1094 |
+
"loss": 0.0196,
|
| 1095 |
+
"step": 18100
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.36,
|
| 1099 |
+
"learning_rate": 0.00012171114176845411,
|
| 1100 |
+
"loss": 0.0203,
|
| 1101 |
+
"step": 18200
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.36,
|
| 1105 |
+
"learning_rate": 0.0001212353822825857,
|
| 1106 |
+
"loss": 0.0224,
|
| 1107 |
+
"step": 18300
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 0.36,
|
| 1111 |
+
"learning_rate": 0.00012075823889544048,
|
| 1112 |
+
"loss": 0.0211,
|
| 1113 |
+
"step": 18400
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 0.36,
|
| 1117 |
+
"learning_rate": 0.00012027972986958772,
|
| 1118 |
+
"loss": 0.0188,
|
| 1119 |
+
"step": 18500
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.37,
|
| 1123 |
+
"learning_rate": 0.0001197998735198662,
|
| 1124 |
+
"loss": 0.0214,
|
| 1125 |
+
"step": 18600
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 0.37,
|
| 1129 |
+
"learning_rate": 0.00011931868821268337,
|
| 1130 |
+
"loss": 0.0201,
|
| 1131 |
+
"step": 18700
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.37,
|
| 1135 |
+
"learning_rate": 0.00011883619236531218,
|
| 1136 |
+
"loss": 0.0206,
|
| 1137 |
+
"step": 18800
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.37,
|
| 1141 |
+
"learning_rate": 0.00011835240444518623,
|
| 1142 |
+
"loss": 0.0206,
|
| 1143 |
+
"step": 18900
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.37,
|
| 1147 |
+
"learning_rate": 0.000117867342969193,
|
| 1148 |
+
"loss": 0.0197,
|
| 1149 |
+
"step": 19000
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 0.38,
|
| 1153 |
+
"learning_rate": 0.00011738102650296509,
|
| 1154 |
+
"loss": 0.0198,
|
| 1155 |
+
"step": 19100
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 0.38,
|
| 1159 |
+
"learning_rate": 0.00011689347366016955,
|
| 1160 |
+
"loss": 0.0211,
|
| 1161 |
+
"step": 19200
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.38,
|
| 1165 |
+
"learning_rate": 0.0001164047031017955,
|
| 1166 |
+
"loss": 0.0204,
|
| 1167 |
+
"step": 19300
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.38,
|
| 1171 |
+
"learning_rate": 0.00011591473353543992,
|
| 1172 |
+
"loss": 0.0206,
|
| 1173 |
+
"step": 19400
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 0.38,
|
| 1177 |
+
"learning_rate": 0.00011542358371459156,
|
| 1178 |
+
"loss": 0.0192,
|
| 1179 |
+
"step": 19500
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.39,
|
| 1183 |
+
"learning_rate": 0.0001149312724379132,
|
| 1184 |
+
"loss": 0.0202,
|
| 1185 |
+
"step": 19600
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.39,
|
| 1189 |
+
"learning_rate": 0.00011443781854852212,
|
| 1190 |
+
"loss": 0.0197,
|
| 1191 |
+
"step": 19700
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 0.39,
|
| 1195 |
+
"learning_rate": 0.0001139432409332688,
|
| 1196 |
+
"loss": 0.021,
|
| 1197 |
+
"step": 19800
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 0.39,
|
| 1201 |
+
"learning_rate": 0.00011344755852201423,
|
| 1202 |
+
"loss": 0.022,
|
| 1203 |
+
"step": 19900
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.39,
|
| 1207 |
+
"learning_rate": 0.00011295079028690513,
|
| 1208 |
+
"loss": 0.0196,
|
| 1209 |
+
"step": 20000
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 0.4,
|
| 1213 |
+
"learning_rate": 0.00011245295524164799,
|
| 1214 |
+
"loss": 0.0195,
|
| 1215 |
+
"step": 20100
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 0.4,
|
| 1219 |
+
"learning_rate": 0.00011195407244078124,
|
| 1220 |
+
"loss": 0.0197,
|
| 1221 |
+
"step": 20200
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.4,
|
| 1225 |
+
"learning_rate": 0.00011145916512264182,
|
| 1226 |
+
"loss": 0.0227,
|
| 1227 |
+
"step": 20300
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.4,
|
| 1231 |
+
"learning_rate": 0.0001109582541342795,
|
| 1232 |
+
"loss": 0.0197,
|
| 1233 |
+
"step": 20400
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 0.4,
|
| 1237 |
+
"learning_rate": 0.00011045635259969838,
|
| 1238 |
+
"loss": 0.0203,
|
| 1239 |
+
"step": 20500
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 0.41,
|
| 1243 |
+
"learning_rate": 0.0001099585132027323,
|
| 1244 |
+
"loss": 0.0216,
|
| 1245 |
+
"step": 20600
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.41,
|
| 1249 |
+
"learning_rate": 0.00010945469766893083,
|
| 1250 |
+
"loss": 0.0224,
|
| 1251 |
+
"step": 20700
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 0.41,
|
| 1255 |
+
"learning_rate": 0.00010894994913723849,
|
| 1256 |
+
"loss": 0.0193,
|
| 1257 |
+
"step": 20800
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 0.41,
|
| 1261 |
+
"learning_rate": 0.00010844428692680601,
|
| 1262 |
+
"loss": 0.0203,
|
| 1263 |
+
"step": 20900
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.41,
|
| 1267 |
+
"learning_rate": 0.000107937730391755,
|
| 1268 |
+
"loss": 0.02,
|
| 1269 |
+
"step": 21000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.42,
|
| 1273 |
+
"learning_rate": 0.00010743029892043725,
|
| 1274 |
+
"loss": 0.0197,
|
| 1275 |
+
"step": 21100
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 0.42,
|
| 1279 |
+
"learning_rate": 0.00010692201193469241,
|
| 1280 |
+
"loss": 0.0192,
|
| 1281 |
+
"step": 21200
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 0.42,
|
| 1285 |
+
"learning_rate": 0.00010641288888910492,
|
| 1286 |
+
"loss": 0.0189,
|
| 1287 |
+
"step": 21300
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.42,
|
| 1291 |
+
"learning_rate": 0.00010590294927025919,
|
| 1292 |
+
"loss": 0.02,
|
| 1293 |
+
"step": 21400
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 0.42,
|
| 1297 |
+
"learning_rate": 0.00010539221259599377,
|
| 1298 |
+
"loss": 0.0182,
|
| 1299 |
+
"step": 21500
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.43,
|
| 1303 |
+
"learning_rate": 0.00010488069841465444,
|
| 1304 |
+
"loss": 0.0191,
|
| 1305 |
+
"step": 21600
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.43,
|
| 1309 |
+
"learning_rate": 0.00010436842630434579,
|
| 1310 |
+
"loss": 0.0188,
|
| 1311 |
+
"step": 21700
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.43,
|
| 1315 |
+
"learning_rate": 0.0001038554158721821,
|
| 1316 |
+
"loss": 0.023,
|
| 1317 |
+
"step": 21800
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 0.43,
|
| 1321 |
+
"learning_rate": 0.00010334168675353674,
|
| 1322 |
+
"loss": 0.0206,
|
| 1323 |
+
"step": 21900
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 0.43,
|
| 1327 |
+
"learning_rate": 0.00010282725861129068,
|
| 1328 |
+
"loss": 0.0187,
|
| 1329 |
+
"step": 22000
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.44,
|
| 1333 |
+
"learning_rate": 0.00010231215113507977,
|
| 1334 |
+
"loss": 0.018,
|
| 1335 |
+
"step": 22100
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 0.44,
|
| 1339 |
+
"learning_rate": 0.00010179638404054145,
|
| 1340 |
+
"loss": 0.0202,
|
| 1341 |
+
"step": 22200
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 0.44,
|
| 1345 |
+
"learning_rate": 0.00010127997706855969,
|
| 1346 |
+
"loss": 0.0199,
|
| 1347 |
+
"step": 22300
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.44,
|
| 1351 |
+
"learning_rate": 0.00010076294998450981,
|
| 1352 |
+
"loss": 0.019,
|
| 1353 |
+
"step": 22400
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.44,
|
| 1357 |
+
"learning_rate": 0.0001002453225775017,
|
| 1358 |
+
"loss": 0.0193,
|
| 1359 |
+
"step": 22500
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 0.45,
|
| 1363 |
+
"learning_rate": 9.972711465962257e-05,
|
| 1364 |
+
"loss": 0.019,
|
| 1365 |
+
"step": 22600
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 0.45,
|
| 1369 |
+
"learning_rate": 9.920834606517847e-05,
|
| 1370 |
+
"loss": 0.0195,
|
| 1371 |
+
"step": 22700
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.45,
|
| 1375 |
+
"learning_rate": 9.868903664993534e-05,
|
| 1376 |
+
"loss": 0.018,
|
| 1377 |
+
"step": 22800
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 0.45,
|
| 1381 |
+
"learning_rate": 9.816920629035887e-05,
|
| 1382 |
+
"loss": 0.0202,
|
| 1383 |
+
"step": 22900
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 0.45,
|
| 1387 |
+
"learning_rate": 9.764887488285381e-05,
|
| 1388 |
+
"loss": 0.0185,
|
| 1389 |
+
"step": 23000
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.45,
|
| 1393 |
+
"learning_rate": 9.71280623430024e-05,
|
| 1394 |
+
"loss": 0.0189,
|
| 1395 |
+
"step": 23100
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.46,
|
| 1399 |
+
"learning_rate": 9.660678860480215e-05,
|
| 1400 |
+
"loss": 0.0186,
|
| 1401 |
+
"step": 23200
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 0.46,
|
| 1405 |
+
"learning_rate": 9.608507361990283e-05,
|
| 1406 |
+
"loss": 0.0195,
|
| 1407 |
+
"step": 23300
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 0.46,
|
| 1411 |
+
"learning_rate": 9.556293735684288e-05,
|
| 1412 |
+
"loss": 0.0206,
|
| 1413 |
+
"step": 23400
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.46,
|
| 1417 |
+
"learning_rate": 9.50456270965957e-05,
|
| 1418 |
+
"loss": 0.0204,
|
| 1419 |
+
"step": 23500
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 0.46,
|
| 1423 |
+
"learning_rate": 9.452271196044817e-05,
|
| 1424 |
+
"loss": 0.0189,
|
| 1425 |
+
"step": 23600
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 0.47,
|
| 1429 |
+
"learning_rate": 9.399943534522518e-05,
|
| 1430 |
+
"loss": 0.0198,
|
| 1431 |
+
"step": 23700
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 0.47,
|
| 1435 |
+
"learning_rate": 9.347581727923617e-05,
|
| 1436 |
+
"loss": 0.0197,
|
| 1437 |
+
"step": 23800
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.47,
|
| 1441 |
+
"learning_rate": 9.295187780385963e-05,
|
| 1442 |
+
"loss": 0.0187,
|
| 1443 |
+
"step": 23900
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 0.47,
|
| 1447 |
+
"learning_rate": 9.242763697277588e-05,
|
| 1448 |
+
"loss": 0.0201,
|
| 1449 |
+
"step": 24000
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 0.47,
|
| 1453 |
+
"learning_rate": 9.19031148511996e-05,
|
| 1454 |
+
"loss": 0.0194,
|
| 1455 |
+
"step": 24100
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 0.48,
|
| 1459 |
+
"learning_rate": 9.137833151511182e-05,
|
| 1460 |
+
"loss": 0.0186,
|
| 1461 |
+
"step": 24200
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 0.48,
|
| 1465 |
+
"learning_rate": 9.08533070504915e-05,
|
| 1466 |
+
"loss": 0.0215,
|
| 1467 |
+
"step": 24300
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 0.48,
|
| 1471 |
+
"learning_rate": 9.032806155254666e-05,
|
| 1472 |
+
"loss": 0.0205,
|
| 1473 |
+
"step": 24400
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.48,
|
| 1477 |
+
"learning_rate": 8.980261512494549e-05,
|
| 1478 |
+
"loss": 0.0188,
|
| 1479 |
+
"step": 24500
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.48,
|
| 1483 |
+
"learning_rate": 8.927698787904661e-05,
|
| 1484 |
+
"loss": 0.0188,
|
| 1485 |
+
"step": 24600
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 0.49,
|
| 1489 |
+
"learning_rate": 8.875119993312938e-05,
|
| 1490 |
+
"loss": 0.0204,
|
| 1491 |
+
"step": 24700
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 0.49,
|
| 1495 |
+
"learning_rate": 8.822527141162407e-05,
|
| 1496 |
+
"loss": 0.0191,
|
| 1497 |
+
"step": 24800
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 0.49,
|
| 1501 |
+
"learning_rate": 8.769922244434138e-05,
|
| 1502 |
+
"loss": 0.017,
|
| 1503 |
+
"step": 24900
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 0.49,
|
| 1507 |
+
"learning_rate": 8.717307316570196e-05,
|
| 1508 |
+
"loss": 0.0186,
|
| 1509 |
+
"step": 25000
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 0.49,
|
| 1513 |
+
"learning_rate": 8.664684371396603e-05,
|
| 1514 |
+
"loss": 0.0195,
|
| 1515 |
+
"step": 25100
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 0.5,
|
| 1519 |
+
"learning_rate": 8.612055423046226e-05,
|
| 1520 |
+
"loss": 0.0199,
|
| 1521 |
+
"step": 25200
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.5,
|
| 1525 |
+
"learning_rate": 8.559422485881711e-05,
|
| 1526 |
+
"loss": 0.0191,
|
| 1527 |
+
"step": 25300
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 0.5,
|
| 1531 |
+
"learning_rate": 8.506787574418376e-05,
|
| 1532 |
+
"loss": 0.0191,
|
| 1533 |
+
"step": 25400
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 0.5,
|
| 1537 |
+
"learning_rate": 8.4541527032471e-05,
|
| 1538 |
+
"loss": 0.0196,
|
| 1539 |
+
"step": 25500
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 0.5,
|
| 1543 |
+
"learning_rate": 8.401519886957223e-05,
|
| 1544 |
+
"loss": 0.0184,
|
| 1545 |
+
"step": 25600
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 0.51,
|
| 1549 |
+
"learning_rate": 8.348891140059429e-05,
|
| 1550 |
+
"loss": 0.0182,
|
| 1551 |
+
"step": 25700
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 0.51,
|
| 1555 |
+
"learning_rate": 8.29626847690866e-05,
|
| 1556 |
+
"loss": 0.0216,
|
| 1557 |
+
"step": 25800
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 0.51,
|
| 1561 |
+
"learning_rate": 8.243653911626992e-05,
|
| 1562 |
+
"loss": 0.0177,
|
| 1563 |
+
"step": 25900
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.51,
|
| 1567 |
+
"learning_rate": 8.191575445898302e-05,
|
| 1568 |
+
"loss": 0.019,
|
| 1569 |
+
"step": 26000
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 0.51,
|
| 1573 |
+
"learning_rate": 8.13898298618816e-05,
|
| 1574 |
+
"loss": 0.0187,
|
| 1575 |
+
"step": 26100
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 0.52,
|
| 1579 |
+
"learning_rate": 8.086404644418334e-05,
|
| 1580 |
+
"loss": 0.0189,
|
| 1581 |
+
"step": 26200
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 0.52,
|
| 1585 |
+
"learning_rate": 8.033842433014505e-05,
|
| 1586 |
+
"loss": 0.0178,
|
| 1587 |
+
"step": 26300
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 0.52,
|
| 1591 |
+
"learning_rate": 7.981298363784983e-05,
|
| 1592 |
+
"loss": 0.0182,
|
| 1593 |
+
"step": 26400
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 0.52,
|
| 1597 |
+
"learning_rate": 7.928774447843676e-05,
|
| 1598 |
+
"loss": 0.0197,
|
| 1599 |
+
"step": 26500
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 0.52,
|
| 1603 |
+
"learning_rate": 7.876272695533133e-05,
|
| 1604 |
+
"loss": 0.0182,
|
| 1605 |
+
"step": 26600
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.53,
|
| 1609 |
+
"learning_rate": 7.823795116347596e-05,
|
| 1610 |
+
"loss": 0.0203,
|
| 1611 |
+
"step": 26700
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 0.53,
|
| 1615 |
+
"learning_rate": 7.771343718856085e-05,
|
| 1616 |
+
"loss": 0.0193,
|
| 1617 |
+
"step": 26800
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 0.53,
|
| 1621 |
+
"learning_rate": 7.718920510625514e-05,
|
| 1622 |
+
"loss": 0.0182,
|
| 1623 |
+
"step": 26900
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 0.53,
|
| 1627 |
+
"learning_rate": 7.666527498143874e-05,
|
| 1628 |
+
"loss": 0.0202,
|
| 1629 |
+
"step": 27000
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 0.53,
|
| 1633 |
+
"learning_rate": 7.614166686743405e-05,
|
| 1634 |
+
"loss": 0.019,
|
| 1635 |
+
"step": 27100
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 0.54,
|
| 1639 |
+
"learning_rate": 7.561840080523864e-05,
|
| 1640 |
+
"loss": 0.0174,
|
| 1641 |
+
"step": 27200
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 0.54,
|
| 1645 |
+
"learning_rate": 7.509549682275809e-05,
|
| 1646 |
+
"loss": 0.018,
|
| 1647 |
+
"step": 27300
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.54,
|
| 1651 |
+
"learning_rate": 7.457297493403944e-05,
|
| 1652 |
+
"loss": 0.019,
|
| 1653 |
+
"step": 27400
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 0.54,
|
| 1657 |
+
"learning_rate": 7.405085513850521e-05,
|
| 1658 |
+
"loss": 0.0205,
|
| 1659 |
+
"step": 27500
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 0.54,
|
| 1663 |
+
"learning_rate": 7.352915742018785e-05,
|
| 1664 |
+
"loss": 0.0169,
|
| 1665 |
+
"step": 27600
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 0.55,
|
| 1669 |
+
"learning_rate": 7.30079017469649e-05,
|
| 1670 |
+
"loss": 0.0183,
|
| 1671 |
+
"step": 27700
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 0.55,
|
| 1675 |
+
"learning_rate": 7.248710806979468e-05,
|
| 1676 |
+
"loss": 0.0181,
|
| 1677 |
+
"step": 27800
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 0.55,
|
| 1681 |
+
"learning_rate": 7.196679632195274e-05,
|
| 1682 |
+
"loss": 0.0171,
|
| 1683 |
+
"step": 27900
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.55,
|
| 1687 |
+
"learning_rate": 7.144698641826882e-05,
|
| 1688 |
+
"loss": 0.0202,
|
| 1689 |
+
"step": 28000
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.55,
|
| 1693 |
+
"learning_rate": 7.092769825436483e-05,
|
| 1694 |
+
"loss": 0.0184,
|
| 1695 |
+
"step": 28100
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 0.56,
|
| 1699 |
+
"learning_rate": 7.040895170589299e-05,
|
| 1700 |
+
"loss": 0.0191,
|
| 1701 |
+
"step": 28200
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 0.56,
|
| 1705 |
+
"learning_rate": 6.989594563413708e-05,
|
| 1706 |
+
"loss": 0.0206,
|
| 1707 |
+
"step": 28300
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 0.56,
|
| 1711 |
+
"learning_rate": 6.937833594866479e-05,
|
| 1712 |
+
"loss": 0.0184,
|
| 1713 |
+
"step": 28400
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 0.56,
|
| 1717 |
+
"learning_rate": 6.886132718016212e-05,
|
| 1718 |
+
"loss": 0.0213,
|
| 1719 |
+
"step": 28500
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 0.56,
|
| 1723 |
+
"learning_rate": 6.834493911703791e-05,
|
| 1724 |
+
"loss": 0.0181,
|
| 1725 |
+
"step": 28600
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.57,
|
| 1729 |
+
"learning_rate": 6.78291915239438e-05,
|
| 1730 |
+
"loss": 0.0205,
|
| 1731 |
+
"step": 28700
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.57,
|
| 1735 |
+
"learning_rate": 6.731410414101738e-05,
|
| 1736 |
+
"loss": 0.0184,
|
| 1737 |
+
"step": 28800
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 0.57,
|
| 1741 |
+
"learning_rate": 6.679969668312693e-05,
|
| 1742 |
+
"loss": 0.0179,
|
| 1743 |
+
"step": 28900
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 0.57,
|
| 1747 |
+
"learning_rate": 6.628598883911667e-05,
|
| 1748 |
+
"loss": 0.0206,
|
| 1749 |
+
"step": 29000
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 0.57,
|
| 1753 |
+
"learning_rate": 6.577300027105329e-05,
|
| 1754 |
+
"loss": 0.0172,
|
| 1755 |
+
"step": 29100
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 0.58,
|
| 1759 |
+
"learning_rate": 6.526075061347325e-05,
|
| 1760 |
+
"loss": 0.0194,
|
| 1761 |
+
"step": 29200
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 0.58,
|
| 1765 |
+
"learning_rate": 6.474925947263152e-05,
|
| 1766 |
+
"loss": 0.0195,
|
| 1767 |
+
"step": 29300
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 0.58,
|
| 1771 |
+
"learning_rate": 6.423854642575082e-05,
|
| 1772 |
+
"loss": 0.0193,
|
| 1773 |
+
"step": 29400
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.58,
|
| 1777 |
+
"learning_rate": 6.372863102027257e-05,
|
| 1778 |
+
"loss": 0.0182,
|
| 1779 |
+
"step": 29500
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 0.58,
|
| 1783 |
+
"learning_rate": 6.321953277310858e-05,
|
| 1784 |
+
"loss": 0.0195,
|
| 1785 |
+
"step": 29600
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 0.58,
|
| 1789 |
+
"learning_rate": 6.27112711698941e-05,
|
| 1790 |
+
"loss": 0.0174,
|
| 1791 |
+
"step": 29700
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 0.59,
|
| 1795 |
+
"learning_rate": 6.220386566424204e-05,
|
| 1796 |
+
"loss": 0.0214,
|
| 1797 |
+
"step": 29800
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 0.59,
|
| 1801 |
+
"learning_rate": 6.169733567699826e-05,
|
| 1802 |
+
"loss": 0.0176,
|
| 1803 |
+
"step": 29900
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 0.59,
|
| 1807 |
+
"learning_rate": 6.11917005954984e-05,
|
| 1808 |
+
"loss": 0.017,
|
| 1809 |
+
"step": 30000
|
| 1810 |
+
}
|
| 1811 |
+
],
|
| 1812 |
+
"logging_steps": 100,
|
| 1813 |
+
"max_steps": 50780,
|
| 1814 |
+
"num_train_epochs": 1,
|
| 1815 |
+
"save_steps": 10000,
|
| 1816 |
+
"total_flos": 1.1978512091136e+19,
|
| 1817 |
+
"trial_name": null,
|
| 1818 |
+
"trial_params": null
|
| 1819 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d17a2c3c8a65881aa45bfd2997dfe652a61cb20d8b49594dc95bff2d7669f6f
|
| 3 |
+
size 4155
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-30000/vocab.txt
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
<cls>
|
| 2 |
+
<pad>
|
| 3 |
+
<eos>
|
| 4 |
+
<unk>
|
| 5 |
+
L
|
| 6 |
+
A
|
| 7 |
+
G
|
| 8 |
+
V
|
| 9 |
+
S
|
| 10 |
+
E
|
| 11 |
+
R
|
| 12 |
+
T
|
| 13 |
+
I
|
| 14 |
+
D
|
| 15 |
+
P
|
| 16 |
+
K
|
| 17 |
+
Q
|
| 18 |
+
N
|
| 19 |
+
F
|
| 20 |
+
Y
|
| 21 |
+
M
|
| 22 |
+
H
|
| 23 |
+
W
|
| 24 |
+
C
|
| 25 |
+
X
|
| 26 |
+
B
|
| 27 |
+
U
|
| 28 |
+
Z
|
| 29 |
+
O
|
| 30 |
+
.
|
| 31 |
+
-
|
| 32 |
+
<null_1>
|
| 33 |
+
<mask>
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/README.md
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
| 4 |
+
## Training procedure
|
| 5 |
+
|
| 6 |
+
|
| 7 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 8 |
+
- quant_method: bitsandbytes
|
| 9 |
+
- load_in_8bit: False
|
| 10 |
+
- load_in_4bit: True
|
| 11 |
+
- llm_int8_threshold: 6.0
|
| 12 |
+
- llm_int8_skip_modules: None
|
| 13 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 14 |
+
- llm_int8_has_fp16_weight: False
|
| 15 |
+
- bnb_4bit_quant_type: nf4
|
| 16 |
+
- bnb_4bit_use_double_quant: True
|
| 17 |
+
- bnb_4bit_compute_dtype: bfloat16
|
| 18 |
+
### Framework versions
|
| 19 |
+
|
| 20 |
+
|
| 21 |
+
- PEFT 0.6.0.dev0
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/adapter_config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "facebook/esm2_t33_650M_UR50D",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"fan_in_fan_out": false,
|
| 7 |
+
"inference_mode": true,
|
| 8 |
+
"init_lora_weights": true,
|
| 9 |
+
"layers_pattern": null,
|
| 10 |
+
"layers_to_transform": null,
|
| 11 |
+
"lora_alpha": 1,
|
| 12 |
+
"lora_dropout": 0.5,
|
| 13 |
+
"modules_to_save": null,
|
| 14 |
+
"peft_type": "LORA",
|
| 15 |
+
"r": 2,
|
| 16 |
+
"rank_pattern": {},
|
| 17 |
+
"revision": null,
|
| 18 |
+
"target_modules": [
|
| 19 |
+
"query",
|
| 20 |
+
"key",
|
| 21 |
+
"value",
|
| 22 |
+
"EsmSelfOutput.dense",
|
| 23 |
+
"EsmIntermediate.dense",
|
| 24 |
+
"EsmOutput.dense",
|
| 25 |
+
"classifier"
|
| 26 |
+
],
|
| 27 |
+
"task_type": "TOKEN_CLS"
|
| 28 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fe0773ad58d9e904a3858633796c2742ac71afda0da63ebf6d4a178d951ce4fb
|
| 3 |
+
size 2133117
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/added_tokens.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<cls>": 0,
|
| 3 |
+
"<eos>": 2,
|
| 4 |
+
"<mask>": 32,
|
| 5 |
+
"<pad>": 1,
|
| 6 |
+
"<unk>": 3
|
| 7 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:369bf2563f08a76616fc3448106c6dfc038cb3a2fe6f80f47ede0bf272cd2095
|
| 3 |
+
size 4215045
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac84843979a304380eaef982db4fd0aa01fb1a7dafd3cdf6688e2c300c6b3f23
|
| 3 |
+
size 14575
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2963dcf1c9d9b00006d2790ed287a1d12d3120a01ec704c5500b55543bb05b48
|
| 3 |
+
size 627
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "<cls>",
|
| 3 |
+
"eos_token": "<eos>",
|
| 4 |
+
"mask_token": "<mask>",
|
| 5 |
+
"pad_token": "<pad>",
|
| 6 |
+
"unk_token": "<unk>"
|
| 7 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<cls>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "<eos>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"32": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"additional_special_tokens": [],
|
| 45 |
+
"clean_up_tokenization_spaces": true,
|
| 46 |
+
"cls_token": "<cls>",
|
| 47 |
+
"eos_token": "<eos>",
|
| 48 |
+
"mask_token": "<mask>",
|
| 49 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 50 |
+
"pad_token": "<pad>",
|
| 51 |
+
"tokenizer_class": "EsmTokenizer",
|
| 52 |
+
"tokenizer_file": null,
|
| 53 |
+
"unk_token": "<unk>"
|
| 54 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/trainer_state.json
ADDED
|
@@ -0,0 +1,2419 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.7877116975187082,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 40000,
|
| 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": 0.00017015520980041642,
|
| 14 |
+
"loss": 0.5575,
|
| 15 |
+
"step": 100
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.0,
|
| 19 |
+
"learning_rate": 0.00017015035788629153,
|
| 20 |
+
"loss": 0.3065,
|
| 21 |
+
"step": 200
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.01,
|
| 25 |
+
"learning_rate": 0.00017014224985982174,
|
| 26 |
+
"loss": 0.2229,
|
| 27 |
+
"step": 300
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.01,
|
| 31 |
+
"learning_rate": 0.0001701308860313402,
|
| 32 |
+
"loss": 0.1812,
|
| 33 |
+
"step": 400
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.01,
|
| 37 |
+
"learning_rate": 0.00017011626683579524,
|
| 38 |
+
"loss": 0.1572,
|
| 39 |
+
"step": 500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.01,
|
| 43 |
+
"learning_rate": 0.00017009839283273364,
|
| 44 |
+
"loss": 0.1335,
|
| 45 |
+
"step": 600
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.01,
|
| 49 |
+
"learning_rate": 0.00017007726470627936,
|
| 50 |
+
"loss": 0.1193,
|
| 51 |
+
"step": 700
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.02,
|
| 55 |
+
"learning_rate": 0.00017005288326510734,
|
| 56 |
+
"loss": 0.1075,
|
| 57 |
+
"step": 800
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.02,
|
| 61 |
+
"learning_rate": 0.0001700252494424124,
|
| 62 |
+
"loss": 0.0986,
|
| 63 |
+
"step": 900
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.02,
|
| 67 |
+
"learning_rate": 0.00016999436429587366,
|
| 68 |
+
"loss": 0.0952,
|
| 69 |
+
"step": 1000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.02,
|
| 73 |
+
"learning_rate": 0.00016996022900761407,
|
| 74 |
+
"loss": 0.0865,
|
| 75 |
+
"step": 1100
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.02,
|
| 79 |
+
"learning_rate": 0.000169922844884155,
|
| 80 |
+
"loss": 0.0816,
|
| 81 |
+
"step": 1200
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.03,
|
| 85 |
+
"learning_rate": 0.00016988221335636648,
|
| 86 |
+
"loss": 0.0775,
|
| 87 |
+
"step": 1300
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.03,
|
| 91 |
+
"learning_rate": 0.00016983833597941224,
|
| 92 |
+
"loss": 0.0744,
|
| 93 |
+
"step": 1400
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.03,
|
| 97 |
+
"learning_rate": 0.00016979121443269025,
|
| 98 |
+
"loss": 0.0694,
|
| 99 |
+
"step": 1500
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.03,
|
| 103 |
+
"learning_rate": 0.0001697408505197684,
|
| 104 |
+
"loss": 0.0641,
|
| 105 |
+
"step": 1600
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.03,
|
| 109 |
+
"learning_rate": 0.00016968724616831557,
|
| 110 |
+
"loss": 0.0629,
|
| 111 |
+
"step": 1700
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.04,
|
| 115 |
+
"learning_rate": 0.0001696304034300278,
|
| 116 |
+
"loss": 0.0696,
|
| 117 |
+
"step": 1800
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.04,
|
| 121 |
+
"learning_rate": 0.00016957032448054968,
|
| 122 |
+
"loss": 0.0591,
|
| 123 |
+
"step": 1900
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.04,
|
| 127 |
+
"learning_rate": 0.0001695070116193912,
|
| 128 |
+
"loss": 0.06,
|
| 129 |
+
"step": 2000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.04,
|
| 133 |
+
"learning_rate": 0.0001694404672698396,
|
| 134 |
+
"loss": 0.0558,
|
| 135 |
+
"step": 2100
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.04,
|
| 139 |
+
"learning_rate": 0.00016937069397886687,
|
| 140 |
+
"loss": 0.0529,
|
| 141 |
+
"step": 2200
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.05,
|
| 145 |
+
"learning_rate": 0.00016929769441703196,
|
| 146 |
+
"loss": 0.0536,
|
| 147 |
+
"step": 2300
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.05,
|
| 151 |
+
"learning_rate": 0.00016922147137837868,
|
| 152 |
+
"loss": 0.0537,
|
| 153 |
+
"step": 2400
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.05,
|
| 157 |
+
"learning_rate": 0.00016914202778032893,
|
| 158 |
+
"loss": 0.0508,
|
| 159 |
+
"step": 2500
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.05,
|
| 163 |
+
"learning_rate": 0.0001690593666635707,
|
| 164 |
+
"loss": 0.0524,
|
| 165 |
+
"step": 2600
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.05,
|
| 169 |
+
"learning_rate": 0.00016897349119194207,
|
| 170 |
+
"loss": 0.0494,
|
| 171 |
+
"step": 2700
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.06,
|
| 175 |
+
"learning_rate": 0.00016888440465230977,
|
| 176 |
+
"loss": 0.047,
|
| 177 |
+
"step": 2800
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.06,
|
| 181 |
+
"learning_rate": 0.00016879211045444354,
|
| 182 |
+
"loss": 0.0502,
|
| 183 |
+
"step": 2900
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.06,
|
| 187 |
+
"learning_rate": 0.00016869661213088575,
|
| 188 |
+
"loss": 0.046,
|
| 189 |
+
"step": 3000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.06,
|
| 193 |
+
"learning_rate": 0.00016859791333681583,
|
| 194 |
+
"loss": 0.0419,
|
| 195 |
+
"step": 3100
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.06,
|
| 199 |
+
"learning_rate": 0.0001684960178499108,
|
| 200 |
+
"loss": 0.0456,
|
| 201 |
+
"step": 3200
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.06,
|
| 205 |
+
"learning_rate": 0.00016839092957020028,
|
| 206 |
+
"loss": 0.0432,
|
| 207 |
+
"step": 3300
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.07,
|
| 211 |
+
"learning_rate": 0.00016828265251991761,
|
| 212 |
+
"loss": 0.0423,
|
| 213 |
+
"step": 3400
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.07,
|
| 217 |
+
"learning_rate": 0.00016817119084334555,
|
| 218 |
+
"loss": 0.0407,
|
| 219 |
+
"step": 3500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.07,
|
| 223 |
+
"learning_rate": 0.00016805654880665776,
|
| 224 |
+
"loss": 0.0409,
|
| 225 |
+
"step": 3600
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.07,
|
| 229 |
+
"learning_rate": 0.00016793873079775577,
|
| 230 |
+
"loss": 0.0397,
|
| 231 |
+
"step": 3700
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.07,
|
| 235 |
+
"learning_rate": 0.00016781774132610059,
|
| 236 |
+
"loss": 0.0419,
|
| 237 |
+
"step": 3800
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.08,
|
| 241 |
+
"learning_rate": 0.0001676935850225405,
|
| 242 |
+
"loss": 0.0399,
|
| 243 |
+
"step": 3900
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.08,
|
| 247 |
+
"learning_rate": 0.00016756626663913358,
|
| 248 |
+
"loss": 0.0389,
|
| 249 |
+
"step": 4000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.08,
|
| 253 |
+
"learning_rate": 0.00016743579104896593,
|
| 254 |
+
"loss": 0.0385,
|
| 255 |
+
"step": 4100
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.08,
|
| 259 |
+
"learning_rate": 0.00016730216324596504,
|
| 260 |
+
"loss": 0.0357,
|
| 261 |
+
"step": 4200
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.08,
|
| 265 |
+
"learning_rate": 0.0001671653883447088,
|
| 266 |
+
"loss": 0.0373,
|
| 267 |
+
"step": 4300
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.09,
|
| 271 |
+
"learning_rate": 0.00016702547158022968,
|
| 272 |
+
"loss": 0.0375,
|
| 273 |
+
"step": 4400
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.09,
|
| 277 |
+
"learning_rate": 0.0001668824183078143,
|
| 278 |
+
"loss": 0.0393,
|
| 279 |
+
"step": 4500
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.09,
|
| 283 |
+
"learning_rate": 0.00016673623400279849,
|
| 284 |
+
"loss": 0.0358,
|
| 285 |
+
"step": 4600
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.09,
|
| 289 |
+
"learning_rate": 0.00016658692426035782,
|
| 290 |
+
"loss": 0.0334,
|
| 291 |
+
"step": 4700
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.09,
|
| 295 |
+
"learning_rate": 0.00016643449479529325,
|
| 296 |
+
"loss": 0.035,
|
| 297 |
+
"step": 4800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.1,
|
| 301 |
+
"learning_rate": 0.00016627895144181258,
|
| 302 |
+
"loss": 0.0343,
|
| 303 |
+
"step": 4900
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.1,
|
| 307 |
+
"learning_rate": 0.0001661203001533071,
|
| 308 |
+
"loss": 0.0335,
|
| 309 |
+
"step": 5000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.1,
|
| 313 |
+
"learning_rate": 0.00016595854700212362,
|
| 314 |
+
"loss": 0.0353,
|
| 315 |
+
"step": 5100
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.1,
|
| 319 |
+
"learning_rate": 0.0001657936981793322,
|
| 320 |
+
"loss": 0.0331,
|
| 321 |
+
"step": 5200
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.1,
|
| 325 |
+
"learning_rate": 0.0001656257599944891,
|
| 326 |
+
"loss": 0.0387,
|
| 327 |
+
"step": 5300
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.11,
|
| 331 |
+
"learning_rate": 0.00016545473887539532,
|
| 332 |
+
"loss": 0.0327,
|
| 333 |
+
"step": 5400
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.11,
|
| 337 |
+
"learning_rate": 0.00016528064136785056,
|
| 338 |
+
"loss": 0.0316,
|
| 339 |
+
"step": 5500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.11,
|
| 343 |
+
"learning_rate": 0.00016510347413540262,
|
| 344 |
+
"loss": 0.0319,
|
| 345 |
+
"step": 5600
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.11,
|
| 349 |
+
"learning_rate": 0.0001649232439590925,
|
| 350 |
+
"loss": 0.0314,
|
| 351 |
+
"step": 5700
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.11,
|
| 355 |
+
"learning_rate": 0.0001647399577371947,
|
| 356 |
+
"loss": 0.0321,
|
| 357 |
+
"step": 5800
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.12,
|
| 361 |
+
"learning_rate": 0.00016455362248495338,
|
| 362 |
+
"loss": 0.0333,
|
| 363 |
+
"step": 5900
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.12,
|
| 367 |
+
"learning_rate": 0.00016436424533431362,
|
| 368 |
+
"loss": 0.0319,
|
| 369 |
+
"step": 6000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.12,
|
| 373 |
+
"learning_rate": 0.0001641718335336486,
|
| 374 |
+
"loss": 0.0315,
|
| 375 |
+
"step": 6100
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.12,
|
| 379 |
+
"learning_rate": 0.0001639763944474821,
|
| 380 |
+
"loss": 0.0311,
|
| 381 |
+
"step": 6200
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.12,
|
| 385 |
+
"learning_rate": 0.0001637779355562068,
|
| 386 |
+
"loss": 0.031,
|
| 387 |
+
"step": 6300
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.13,
|
| 391 |
+
"learning_rate": 0.00016357646445579763,
|
| 392 |
+
"loss": 0.0299,
|
| 393 |
+
"step": 6400
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.13,
|
| 397 |
+
"learning_rate": 0.00016337198885752133,
|
| 398 |
+
"loss": 0.0317,
|
| 399 |
+
"step": 6500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.13,
|
| 403 |
+
"learning_rate": 0.00016316451658764122,
|
| 404 |
+
"loss": 0.0302,
|
| 405 |
+
"step": 6600
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.13,
|
| 409 |
+
"learning_rate": 0.0001629540555871176,
|
| 410 |
+
"loss": 0.0295,
|
| 411 |
+
"step": 6700
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 0.13,
|
| 415 |
+
"learning_rate": 0.00016274061391130388,
|
| 416 |
+
"loss": 0.03,
|
| 417 |
+
"step": 6800
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.14,
|
| 421 |
+
"learning_rate": 0.0001625241997296382,
|
| 422 |
+
"loss": 0.0292,
|
| 423 |
+
"step": 6900
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.14,
|
| 427 |
+
"learning_rate": 0.00016230482132533077,
|
| 428 |
+
"loss": 0.0289,
|
| 429 |
+
"step": 7000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.14,
|
| 433 |
+
"learning_rate": 0.00016208472504084003,
|
| 434 |
+
"loss": 0.0318,
|
| 435 |
+
"step": 7100
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.14,
|
| 439 |
+
"learning_rate": 0.0001618594729250462,
|
| 440 |
+
"loss": 0.0301,
|
| 441 |
+
"step": 7200
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.14,
|
| 445 |
+
"learning_rate": 0.00016163128202889828,
|
| 446 |
+
"loss": 0.0295,
|
| 447 |
+
"step": 7300
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.15,
|
| 451 |
+
"learning_rate": 0.00016140016108635798,
|
| 452 |
+
"loss": 0.029,
|
| 453 |
+
"step": 7400
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 0.15,
|
| 457 |
+
"learning_rate": 0.00016116611894353386,
|
| 458 |
+
"loss": 0.0291,
|
| 459 |
+
"step": 7500
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.15,
|
| 463 |
+
"learning_rate": 0.00016092916455834295,
|
| 464 |
+
"loss": 0.0311,
|
| 465 |
+
"step": 7600
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.15,
|
| 469 |
+
"learning_rate": 0.00016068930700016766,
|
| 470 |
+
"loss": 0.0285,
|
| 471 |
+
"step": 7700
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.15,
|
| 475 |
+
"learning_rate": 0.00016044655544950889,
|
| 476 |
+
"loss": 0.0287,
|
| 477 |
+
"step": 7800
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.16,
|
| 481 |
+
"learning_rate": 0.00016020091919763445,
|
| 482 |
+
"loss": 0.0293,
|
| 483 |
+
"step": 7900
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 0.16,
|
| 487 |
+
"learning_rate": 0.00015995240764622357,
|
| 488 |
+
"loss": 0.0259,
|
| 489 |
+
"step": 8000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.16,
|
| 493 |
+
"learning_rate": 0.000159701030307007,
|
| 494 |
+
"loss": 0.0293,
|
| 495 |
+
"step": 8100
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.16,
|
| 499 |
+
"learning_rate": 0.00015944679680140295,
|
| 500 |
+
"loss": 0.0277,
|
| 501 |
+
"step": 8200
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.16,
|
| 505 |
+
"learning_rate": 0.0001591897168601488,
|
| 506 |
+
"loss": 0.0304,
|
| 507 |
+
"step": 8300
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.17,
|
| 511 |
+
"learning_rate": 0.00015892980032292876,
|
| 512 |
+
"loss": 0.026,
|
| 513 |
+
"step": 8400
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.17,
|
| 517 |
+
"learning_rate": 0.00015866705713799714,
|
| 518 |
+
"loss": 0.0294,
|
| 519 |
+
"step": 8500
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.17,
|
| 523 |
+
"learning_rate": 0.00015840149736179762,
|
| 524 |
+
"loss": 0.0321,
|
| 525 |
+
"step": 8600
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.17,
|
| 529 |
+
"learning_rate": 0.0001581331311585785,
|
| 530 |
+
"loss": 0.0279,
|
| 531 |
+
"step": 8700
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.17,
|
| 535 |
+
"learning_rate": 0.00015786196880000325,
|
| 536 |
+
"loss": 0.0277,
|
| 537 |
+
"step": 8800
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.18,
|
| 541 |
+
"learning_rate": 0.0001575880206647579,
|
| 542 |
+
"loss": 0.0268,
|
| 543 |
+
"step": 8900
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.18,
|
| 547 |
+
"learning_rate": 0.00015731129723815343,
|
| 548 |
+
"loss": 0.0281,
|
| 549 |
+
"step": 9000
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.18,
|
| 553 |
+
"learning_rate": 0.00015703180911172453,
|
| 554 |
+
"loss": 0.028,
|
| 555 |
+
"step": 9100
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.18,
|
| 559 |
+
"learning_rate": 0.0001567495669828243,
|
| 560 |
+
"loss": 0.026,
|
| 561 |
+
"step": 9200
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.18,
|
| 565 |
+
"learning_rate": 0.0001564645816542146,
|
| 566 |
+
"loss": 0.0256,
|
| 567 |
+
"step": 9300
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 0.19,
|
| 571 |
+
"learning_rate": 0.0001561768640336529,
|
| 572 |
+
"loss": 0.027,
|
| 573 |
+
"step": 9400
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 0.19,
|
| 577 |
+
"learning_rate": 0.0001558864251334745,
|
| 578 |
+
"loss": 0.0249,
|
| 579 |
+
"step": 9500
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.19,
|
| 583 |
+
"learning_rate": 0.00015559327607017119,
|
| 584 |
+
"loss": 0.0256,
|
| 585 |
+
"step": 9600
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.19,
|
| 589 |
+
"learning_rate": 0.00015529742806396564,
|
| 590 |
+
"loss": 0.0251,
|
| 591 |
+
"step": 9700
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.19,
|
| 595 |
+
"learning_rate": 0.00015499889243838211,
|
| 596 |
+
"loss": 0.0257,
|
| 597 |
+
"step": 9800
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.19,
|
| 601 |
+
"learning_rate": 0.00015469768061981295,
|
| 602 |
+
"loss": 0.0264,
|
| 603 |
+
"step": 9900
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.2,
|
| 607 |
+
"learning_rate": 0.00015439685605389073,
|
| 608 |
+
"loss": 0.0247,
|
| 609 |
+
"step": 10000
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.2,
|
| 613 |
+
"learning_rate": 0.00015409035301023626,
|
| 614 |
+
"loss": 0.0256,
|
| 615 |
+
"step": 10100
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 0.2,
|
| 619 |
+
"learning_rate": 0.00015378120854776404,
|
| 620 |
+
"loss": 0.0247,
|
| 621 |
+
"step": 10200
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.2,
|
| 625 |
+
"learning_rate": 0.00015346943449891755,
|
| 626 |
+
"loss": 0.0254,
|
| 627 |
+
"step": 10300
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.2,
|
| 631 |
+
"learning_rate": 0.00015315504279678714,
|
| 632 |
+
"loss": 0.0246,
|
| 633 |
+
"step": 10400
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.21,
|
| 637 |
+
"learning_rate": 0.00015283804547465337,
|
| 638 |
+
"loss": 0.0249,
|
| 639 |
+
"step": 10500
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.21,
|
| 643 |
+
"learning_rate": 0.0001525184546655264,
|
| 644 |
+
"loss": 0.026,
|
| 645 |
+
"step": 10600
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 0.21,
|
| 649 |
+
"learning_rate": 0.0001521962826016816,
|
| 650 |
+
"loss": 0.0254,
|
| 651 |
+
"step": 10700
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 0.21,
|
| 655 |
+
"learning_rate": 0.00015187154161419122,
|
| 656 |
+
"loss": 0.0272,
|
| 657 |
+
"step": 10800
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.21,
|
| 661 |
+
"learning_rate": 0.0001515442441324528,
|
| 662 |
+
"loss": 0.0253,
|
| 663 |
+
"step": 10900
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 0.22,
|
| 667 |
+
"learning_rate": 0.00015121440268371297,
|
| 668 |
+
"loss": 0.0263,
|
| 669 |
+
"step": 11000
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.22,
|
| 673 |
+
"learning_rate": 0.00015088202989258835,
|
| 674 |
+
"loss": 0.0256,
|
| 675 |
+
"step": 11100
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.22,
|
| 679 |
+
"learning_rate": 0.00015054713848058212,
|
| 680 |
+
"loss": 0.023,
|
| 681 |
+
"step": 11200
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.22,
|
| 685 |
+
"learning_rate": 0.00015020974126559716,
|
| 686 |
+
"loss": 0.0236,
|
| 687 |
+
"step": 11300
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 0.22,
|
| 691 |
+
"learning_rate": 0.00014986985116144554,
|
| 692 |
+
"loss": 0.0241,
|
| 693 |
+
"step": 11400
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 0.23,
|
| 697 |
+
"learning_rate": 0.00014952748117735409,
|
| 698 |
+
"loss": 0.0235,
|
| 699 |
+
"step": 11500
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.23,
|
| 703 |
+
"learning_rate": 0.0001491826444174666,
|
| 704 |
+
"loss": 0.0242,
|
| 705 |
+
"step": 11600
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.23,
|
| 709 |
+
"learning_rate": 0.00014883535408034227,
|
| 710 |
+
"loss": 0.0233,
|
| 711 |
+
"step": 11700
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.23,
|
| 715 |
+
"learning_rate": 0.00014848562345845032,
|
| 716 |
+
"loss": 0.0228,
|
| 717 |
+
"step": 11800
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.23,
|
| 721 |
+
"learning_rate": 0.00014813346593766164,
|
| 722 |
+
"loss": 0.023,
|
| 723 |
+
"step": 11900
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.24,
|
| 727 |
+
"learning_rate": 0.0001477788949967359,
|
| 728 |
+
"loss": 0.0236,
|
| 729 |
+
"step": 12000
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.24,
|
| 733 |
+
"learning_rate": 0.00014742192420680626,
|
| 734 |
+
"loss": 0.0237,
|
| 735 |
+
"step": 12100
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.24,
|
| 739 |
+
"learning_rate": 0.00014706256723085937,
|
| 740 |
+
"loss": 0.0237,
|
| 741 |
+
"step": 12200
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 0.24,
|
| 745 |
+
"learning_rate": 0.0001467008378232128,
|
| 746 |
+
"loss": 0.024,
|
| 747 |
+
"step": 12300
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.24,
|
| 751 |
+
"learning_rate": 0.00014633674982898854,
|
| 752 |
+
"loss": 0.0227,
|
| 753 |
+
"step": 12400
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.25,
|
| 757 |
+
"learning_rate": 0.00014597031718358285,
|
| 758 |
+
"loss": 0.0247,
|
| 759 |
+
"step": 12500
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.25,
|
| 763 |
+
"learning_rate": 0.00014560155391213318,
|
| 764 |
+
"loss": 0.0233,
|
| 765 |
+
"step": 12600
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.25,
|
| 769 |
+
"learning_rate": 0.0001452304741289812,
|
| 770 |
+
"loss": 0.0234,
|
| 771 |
+
"step": 12700
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 0.25,
|
| 775 |
+
"learning_rate": 0.00014485709203713263,
|
| 776 |
+
"loss": 0.0221,
|
| 777 |
+
"step": 12800
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.25,
|
| 781 |
+
"learning_rate": 0.00014448142192771354,
|
| 782 |
+
"loss": 0.0244,
|
| 783 |
+
"step": 12900
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.26,
|
| 787 |
+
"learning_rate": 0.00014410347817942347,
|
| 788 |
+
"loss": 0.0242,
|
| 789 |
+
"step": 13000
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.26,
|
| 793 |
+
"learning_rate": 0.00014372327525798503,
|
| 794 |
+
"loss": 0.0231,
|
| 795 |
+
"step": 13100
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.26,
|
| 799 |
+
"learning_rate": 0.00014334082771559026,
|
| 800 |
+
"loss": 0.0238,
|
| 801 |
+
"step": 13200
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.26,
|
| 805 |
+
"learning_rate": 0.00014295615019034358,
|
| 806 |
+
"loss": 0.0233,
|
| 807 |
+
"step": 13300
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.26,
|
| 811 |
+
"learning_rate": 0.0001425692574057016,
|
| 812 |
+
"loss": 0.024,
|
| 813 |
+
"step": 13400
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 0.27,
|
| 817 |
+
"learning_rate": 0.00014218016416990954,
|
| 818 |
+
"loss": 0.0249,
|
| 819 |
+
"step": 13500
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.27,
|
| 823 |
+
"learning_rate": 0.0001417928089327897,
|
| 824 |
+
"loss": 0.0265,
|
| 825 |
+
"step": 13600
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.27,
|
| 829 |
+
"learning_rate": 0.00014139938118717103,
|
| 830 |
+
"loss": 0.0206,
|
| 831 |
+
"step": 13700
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.27,
|
| 835 |
+
"learning_rate": 0.00014100379776718443,
|
| 836 |
+
"loss": 0.0216,
|
| 837 |
+
"step": 13800
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.27,
|
| 841 |
+
"learning_rate": 0.00014060607381370744,
|
| 842 |
+
"loss": 0.0239,
|
| 843 |
+
"step": 13900
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.28,
|
| 847 |
+
"learning_rate": 0.00014020622454954606,
|
| 848 |
+
"loss": 0.024,
|
| 849 |
+
"step": 14000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.28,
|
| 853 |
+
"learning_rate": 0.00013980426527885237,
|
| 854 |
+
"loss": 0.0234,
|
| 855 |
+
"step": 14100
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 0.28,
|
| 859 |
+
"learning_rate": 0.00013940021138653824,
|
| 860 |
+
"loss": 0.0228,
|
| 861 |
+
"step": 14200
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 0.28,
|
| 865 |
+
"learning_rate": 0.00013899407833768695,
|
| 866 |
+
"loss": 0.0228,
|
| 867 |
+
"step": 14300
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.28,
|
| 871 |
+
"learning_rate": 0.0001385858816769611,
|
| 872 |
+
"loss": 0.0234,
|
| 873 |
+
"step": 14400
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.29,
|
| 877 |
+
"learning_rate": 0.0001381756370280075,
|
| 878 |
+
"loss": 0.0233,
|
| 879 |
+
"step": 14500
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.29,
|
| 883 |
+
"learning_rate": 0.00013776336009285937,
|
| 884 |
+
"loss": 0.0216,
|
| 885 |
+
"step": 14600
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.29,
|
| 889 |
+
"learning_rate": 0.00013734906665133537,
|
| 890 |
+
"loss": 0.0268,
|
| 891 |
+
"step": 14700
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.29,
|
| 895 |
+
"learning_rate": 0.0001369327725604354,
|
| 896 |
+
"loss": 0.022,
|
| 897 |
+
"step": 14800
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.29,
|
| 901 |
+
"learning_rate": 0.00013651449375373396,
|
| 902 |
+
"loss": 0.0214,
|
| 903 |
+
"step": 14900
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.3,
|
| 907 |
+
"learning_rate": 0.00013609424624077,
|
| 908 |
+
"loss": 0.0235,
|
| 909 |
+
"step": 15000
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.3,
|
| 913 |
+
"learning_rate": 0.0001356720461064345,
|
| 914 |
+
"loss": 0.0225,
|
| 915 |
+
"step": 15100
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 0.3,
|
| 919 |
+
"learning_rate": 0.00013524790951035441,
|
| 920 |
+
"loss": 0.0218,
|
| 921 |
+
"step": 15200
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.3,
|
| 925 |
+
"learning_rate": 0.00013482185268627465,
|
| 926 |
+
"loss": 0.0225,
|
| 927 |
+
"step": 15300
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.3,
|
| 931 |
+
"learning_rate": 0.00013439389194143625,
|
| 932 |
+
"loss": 0.0216,
|
| 933 |
+
"step": 15400
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.31,
|
| 937 |
+
"learning_rate": 0.00013396404365595253,
|
| 938 |
+
"loss": 0.0223,
|
| 939 |
+
"step": 15500
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.31,
|
| 943 |
+
"learning_rate": 0.00013353232428218212,
|
| 944 |
+
"loss": 0.0212,
|
| 945 |
+
"step": 15600
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.31,
|
| 949 |
+
"learning_rate": 0.00013309875034409903,
|
| 950 |
+
"loss": 0.023,
|
| 951 |
+
"step": 15700
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 0.31,
|
| 955 |
+
"learning_rate": 0.0001326633384366605,
|
| 956 |
+
"loss": 0.0214,
|
| 957 |
+
"step": 15800
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 0.31,
|
| 961 |
+
"learning_rate": 0.0001322261052251716,
|
| 962 |
+
"loss": 0.021,
|
| 963 |
+
"step": 15900
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 0.32,
|
| 967 |
+
"learning_rate": 0.00013178706744464749,
|
| 968 |
+
"loss": 0.0211,
|
| 969 |
+
"step": 16000
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.32,
|
| 973 |
+
"learning_rate": 0.00013134624189917282,
|
| 974 |
+
"loss": 0.0212,
|
| 975 |
+
"step": 16100
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.32,
|
| 979 |
+
"learning_rate": 0.00013090808013598785,
|
| 980 |
+
"loss": 0.0252,
|
| 981 |
+
"step": 16200
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 0.32,
|
| 985 |
+
"learning_rate": 0.00013046374720137184,
|
| 986 |
+
"loss": 0.0214,
|
| 987 |
+
"step": 16300
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 0.32,
|
| 991 |
+
"learning_rate": 0.00013001767715162726,
|
| 992 |
+
"loss": 0.0202,
|
| 993 |
+
"step": 16400
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 0.32,
|
| 997 |
+
"learning_rate": 0.00012956988705999754,
|
| 998 |
+
"loss": 0.0205,
|
| 999 |
+
"step": 16500
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.33,
|
| 1003 |
+
"learning_rate": 0.00012912039406556028,
|
| 1004 |
+
"loss": 0.0235,
|
| 1005 |
+
"step": 16600
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 0.33,
|
| 1009 |
+
"learning_rate": 0.00012866921537257149,
|
| 1010 |
+
"loss": 0.0215,
|
| 1011 |
+
"step": 16700
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.33,
|
| 1015 |
+
"learning_rate": 0.00012821636824980682,
|
| 1016 |
+
"loss": 0.0198,
|
| 1017 |
+
"step": 16800
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.33,
|
| 1021 |
+
"learning_rate": 0.00012776187002990082,
|
| 1022 |
+
"loss": 0.0232,
|
| 1023 |
+
"step": 16900
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 0.33,
|
| 1027 |
+
"learning_rate": 0.00012730573810868347,
|
| 1028 |
+
"loss": 0.0235,
|
| 1029 |
+
"step": 17000
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 0.34,
|
| 1033 |
+
"learning_rate": 0.00012684798994451428,
|
| 1034 |
+
"loss": 0.0207,
|
| 1035 |
+
"step": 17100
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 0.34,
|
| 1039 |
+
"learning_rate": 0.00012638864305761417,
|
| 1040 |
+
"loss": 0.0218,
|
| 1041 |
+
"step": 17200
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.34,
|
| 1045 |
+
"learning_rate": 0.00012592771502939492,
|
| 1046 |
+
"loss": 0.0204,
|
| 1047 |
+
"step": 17300
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 0.34,
|
| 1051 |
+
"learning_rate": 0.0001254652235017861,
|
| 1052 |
+
"loss": 0.0212,
|
| 1053 |
+
"step": 17400
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.34,
|
| 1057 |
+
"learning_rate": 0.00012500118617655992,
|
| 1058 |
+
"loss": 0.0207,
|
| 1059 |
+
"step": 17500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.35,
|
| 1063 |
+
"learning_rate": 0.00012453562081465375,
|
| 1064 |
+
"loss": 0.0207,
|
| 1065 |
+
"step": 17600
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.35,
|
| 1069 |
+
"learning_rate": 0.0001240685452354902,
|
| 1070 |
+
"loss": 0.0204,
|
| 1071 |
+
"step": 17700
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 0.35,
|
| 1075 |
+
"learning_rate": 0.00012359997731629517,
|
| 1076 |
+
"loss": 0.0234,
|
| 1077 |
+
"step": 17800
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 0.35,
|
| 1081 |
+
"learning_rate": 0.00012312993499141364,
|
| 1082 |
+
"loss": 0.02,
|
| 1083 |
+
"step": 17900
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.35,
|
| 1087 |
+
"learning_rate": 0.0001226584362516231,
|
| 1088 |
+
"loss": 0.0191,
|
| 1089 |
+
"step": 18000
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.36,
|
| 1093 |
+
"learning_rate": 0.00012218549914344497,
|
| 1094 |
+
"loss": 0.0196,
|
| 1095 |
+
"step": 18100
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.36,
|
| 1099 |
+
"learning_rate": 0.00012171114176845411,
|
| 1100 |
+
"loss": 0.0203,
|
| 1101 |
+
"step": 18200
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.36,
|
| 1105 |
+
"learning_rate": 0.0001212353822825857,
|
| 1106 |
+
"loss": 0.0224,
|
| 1107 |
+
"step": 18300
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 0.36,
|
| 1111 |
+
"learning_rate": 0.00012075823889544048,
|
| 1112 |
+
"loss": 0.0211,
|
| 1113 |
+
"step": 18400
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 0.36,
|
| 1117 |
+
"learning_rate": 0.00012027972986958772,
|
| 1118 |
+
"loss": 0.0188,
|
| 1119 |
+
"step": 18500
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.37,
|
| 1123 |
+
"learning_rate": 0.0001197998735198662,
|
| 1124 |
+
"loss": 0.0214,
|
| 1125 |
+
"step": 18600
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 0.37,
|
| 1129 |
+
"learning_rate": 0.00011931868821268337,
|
| 1130 |
+
"loss": 0.0201,
|
| 1131 |
+
"step": 18700
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.37,
|
| 1135 |
+
"learning_rate": 0.00011883619236531218,
|
| 1136 |
+
"loss": 0.0206,
|
| 1137 |
+
"step": 18800
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.37,
|
| 1141 |
+
"learning_rate": 0.00011835240444518623,
|
| 1142 |
+
"loss": 0.0206,
|
| 1143 |
+
"step": 18900
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.37,
|
| 1147 |
+
"learning_rate": 0.000117867342969193,
|
| 1148 |
+
"loss": 0.0197,
|
| 1149 |
+
"step": 19000
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 0.38,
|
| 1153 |
+
"learning_rate": 0.00011738102650296509,
|
| 1154 |
+
"loss": 0.0198,
|
| 1155 |
+
"step": 19100
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 0.38,
|
| 1159 |
+
"learning_rate": 0.00011689347366016955,
|
| 1160 |
+
"loss": 0.0211,
|
| 1161 |
+
"step": 19200
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.38,
|
| 1165 |
+
"learning_rate": 0.0001164047031017955,
|
| 1166 |
+
"loss": 0.0204,
|
| 1167 |
+
"step": 19300
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.38,
|
| 1171 |
+
"learning_rate": 0.00011591473353543992,
|
| 1172 |
+
"loss": 0.0206,
|
| 1173 |
+
"step": 19400
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 0.38,
|
| 1177 |
+
"learning_rate": 0.00011542358371459156,
|
| 1178 |
+
"loss": 0.0192,
|
| 1179 |
+
"step": 19500
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.39,
|
| 1183 |
+
"learning_rate": 0.0001149312724379132,
|
| 1184 |
+
"loss": 0.0202,
|
| 1185 |
+
"step": 19600
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.39,
|
| 1189 |
+
"learning_rate": 0.00011443781854852212,
|
| 1190 |
+
"loss": 0.0197,
|
| 1191 |
+
"step": 19700
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 0.39,
|
| 1195 |
+
"learning_rate": 0.0001139432409332688,
|
| 1196 |
+
"loss": 0.021,
|
| 1197 |
+
"step": 19800
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 0.39,
|
| 1201 |
+
"learning_rate": 0.00011344755852201423,
|
| 1202 |
+
"loss": 0.022,
|
| 1203 |
+
"step": 19900
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.39,
|
| 1207 |
+
"learning_rate": 0.00011295079028690513,
|
| 1208 |
+
"loss": 0.0196,
|
| 1209 |
+
"step": 20000
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 0.4,
|
| 1213 |
+
"learning_rate": 0.00011245295524164799,
|
| 1214 |
+
"loss": 0.0195,
|
| 1215 |
+
"step": 20100
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 0.4,
|
| 1219 |
+
"learning_rate": 0.00011195407244078124,
|
| 1220 |
+
"loss": 0.0197,
|
| 1221 |
+
"step": 20200
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.4,
|
| 1225 |
+
"learning_rate": 0.00011145916512264182,
|
| 1226 |
+
"loss": 0.0227,
|
| 1227 |
+
"step": 20300
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.4,
|
| 1231 |
+
"learning_rate": 0.0001109582541342795,
|
| 1232 |
+
"loss": 0.0197,
|
| 1233 |
+
"step": 20400
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 0.4,
|
| 1237 |
+
"learning_rate": 0.00011045635259969838,
|
| 1238 |
+
"loss": 0.0203,
|
| 1239 |
+
"step": 20500
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 0.41,
|
| 1243 |
+
"learning_rate": 0.0001099585132027323,
|
| 1244 |
+
"loss": 0.0216,
|
| 1245 |
+
"step": 20600
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.41,
|
| 1249 |
+
"learning_rate": 0.00010945469766893083,
|
| 1250 |
+
"loss": 0.0224,
|
| 1251 |
+
"step": 20700
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 0.41,
|
| 1255 |
+
"learning_rate": 0.00010894994913723849,
|
| 1256 |
+
"loss": 0.0193,
|
| 1257 |
+
"step": 20800
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 0.41,
|
| 1261 |
+
"learning_rate": 0.00010844428692680601,
|
| 1262 |
+
"loss": 0.0203,
|
| 1263 |
+
"step": 20900
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.41,
|
| 1267 |
+
"learning_rate": 0.000107937730391755,
|
| 1268 |
+
"loss": 0.02,
|
| 1269 |
+
"step": 21000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.42,
|
| 1273 |
+
"learning_rate": 0.00010743029892043725,
|
| 1274 |
+
"loss": 0.0197,
|
| 1275 |
+
"step": 21100
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 0.42,
|
| 1279 |
+
"learning_rate": 0.00010692201193469241,
|
| 1280 |
+
"loss": 0.0192,
|
| 1281 |
+
"step": 21200
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 0.42,
|
| 1285 |
+
"learning_rate": 0.00010641288888910492,
|
| 1286 |
+
"loss": 0.0189,
|
| 1287 |
+
"step": 21300
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.42,
|
| 1291 |
+
"learning_rate": 0.00010590294927025919,
|
| 1292 |
+
"loss": 0.02,
|
| 1293 |
+
"step": 21400
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 0.42,
|
| 1297 |
+
"learning_rate": 0.00010539221259599377,
|
| 1298 |
+
"loss": 0.0182,
|
| 1299 |
+
"step": 21500
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.43,
|
| 1303 |
+
"learning_rate": 0.00010488069841465444,
|
| 1304 |
+
"loss": 0.0191,
|
| 1305 |
+
"step": 21600
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.43,
|
| 1309 |
+
"learning_rate": 0.00010436842630434579,
|
| 1310 |
+
"loss": 0.0188,
|
| 1311 |
+
"step": 21700
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.43,
|
| 1315 |
+
"learning_rate": 0.0001038554158721821,
|
| 1316 |
+
"loss": 0.023,
|
| 1317 |
+
"step": 21800
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 0.43,
|
| 1321 |
+
"learning_rate": 0.00010334168675353674,
|
| 1322 |
+
"loss": 0.0206,
|
| 1323 |
+
"step": 21900
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 0.43,
|
| 1327 |
+
"learning_rate": 0.00010282725861129068,
|
| 1328 |
+
"loss": 0.0187,
|
| 1329 |
+
"step": 22000
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.44,
|
| 1333 |
+
"learning_rate": 0.00010231215113507977,
|
| 1334 |
+
"loss": 0.018,
|
| 1335 |
+
"step": 22100
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 0.44,
|
| 1339 |
+
"learning_rate": 0.00010179638404054145,
|
| 1340 |
+
"loss": 0.0202,
|
| 1341 |
+
"step": 22200
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 0.44,
|
| 1345 |
+
"learning_rate": 0.00010127997706855969,
|
| 1346 |
+
"loss": 0.0199,
|
| 1347 |
+
"step": 22300
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.44,
|
| 1351 |
+
"learning_rate": 0.00010076294998450981,
|
| 1352 |
+
"loss": 0.019,
|
| 1353 |
+
"step": 22400
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.44,
|
| 1357 |
+
"learning_rate": 0.0001002453225775017,
|
| 1358 |
+
"loss": 0.0193,
|
| 1359 |
+
"step": 22500
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 0.45,
|
| 1363 |
+
"learning_rate": 9.972711465962257e-05,
|
| 1364 |
+
"loss": 0.019,
|
| 1365 |
+
"step": 22600
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 0.45,
|
| 1369 |
+
"learning_rate": 9.920834606517847e-05,
|
| 1370 |
+
"loss": 0.0195,
|
| 1371 |
+
"step": 22700
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.45,
|
| 1375 |
+
"learning_rate": 9.868903664993534e-05,
|
| 1376 |
+
"loss": 0.018,
|
| 1377 |
+
"step": 22800
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 0.45,
|
| 1381 |
+
"learning_rate": 9.816920629035887e-05,
|
| 1382 |
+
"loss": 0.0202,
|
| 1383 |
+
"step": 22900
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 0.45,
|
| 1387 |
+
"learning_rate": 9.764887488285381e-05,
|
| 1388 |
+
"loss": 0.0185,
|
| 1389 |
+
"step": 23000
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.45,
|
| 1393 |
+
"learning_rate": 9.71280623430024e-05,
|
| 1394 |
+
"loss": 0.0189,
|
| 1395 |
+
"step": 23100
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.46,
|
| 1399 |
+
"learning_rate": 9.660678860480215e-05,
|
| 1400 |
+
"loss": 0.0186,
|
| 1401 |
+
"step": 23200
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 0.46,
|
| 1405 |
+
"learning_rate": 9.608507361990283e-05,
|
| 1406 |
+
"loss": 0.0195,
|
| 1407 |
+
"step": 23300
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 0.46,
|
| 1411 |
+
"learning_rate": 9.556293735684288e-05,
|
| 1412 |
+
"loss": 0.0206,
|
| 1413 |
+
"step": 23400
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.46,
|
| 1417 |
+
"learning_rate": 9.50456270965957e-05,
|
| 1418 |
+
"loss": 0.0204,
|
| 1419 |
+
"step": 23500
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 0.46,
|
| 1423 |
+
"learning_rate": 9.452271196044817e-05,
|
| 1424 |
+
"loss": 0.0189,
|
| 1425 |
+
"step": 23600
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 0.47,
|
| 1429 |
+
"learning_rate": 9.399943534522518e-05,
|
| 1430 |
+
"loss": 0.0198,
|
| 1431 |
+
"step": 23700
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 0.47,
|
| 1435 |
+
"learning_rate": 9.347581727923617e-05,
|
| 1436 |
+
"loss": 0.0197,
|
| 1437 |
+
"step": 23800
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.47,
|
| 1441 |
+
"learning_rate": 9.295187780385963e-05,
|
| 1442 |
+
"loss": 0.0187,
|
| 1443 |
+
"step": 23900
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 0.47,
|
| 1447 |
+
"learning_rate": 9.242763697277588e-05,
|
| 1448 |
+
"loss": 0.0201,
|
| 1449 |
+
"step": 24000
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 0.47,
|
| 1453 |
+
"learning_rate": 9.19031148511996e-05,
|
| 1454 |
+
"loss": 0.0194,
|
| 1455 |
+
"step": 24100
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 0.48,
|
| 1459 |
+
"learning_rate": 9.137833151511182e-05,
|
| 1460 |
+
"loss": 0.0186,
|
| 1461 |
+
"step": 24200
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 0.48,
|
| 1465 |
+
"learning_rate": 9.08533070504915e-05,
|
| 1466 |
+
"loss": 0.0215,
|
| 1467 |
+
"step": 24300
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 0.48,
|
| 1471 |
+
"learning_rate": 9.032806155254666e-05,
|
| 1472 |
+
"loss": 0.0205,
|
| 1473 |
+
"step": 24400
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.48,
|
| 1477 |
+
"learning_rate": 8.980261512494549e-05,
|
| 1478 |
+
"loss": 0.0188,
|
| 1479 |
+
"step": 24500
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.48,
|
| 1483 |
+
"learning_rate": 8.927698787904661e-05,
|
| 1484 |
+
"loss": 0.0188,
|
| 1485 |
+
"step": 24600
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 0.49,
|
| 1489 |
+
"learning_rate": 8.875119993312938e-05,
|
| 1490 |
+
"loss": 0.0204,
|
| 1491 |
+
"step": 24700
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 0.49,
|
| 1495 |
+
"learning_rate": 8.822527141162407e-05,
|
| 1496 |
+
"loss": 0.0191,
|
| 1497 |
+
"step": 24800
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 0.49,
|
| 1501 |
+
"learning_rate": 8.769922244434138e-05,
|
| 1502 |
+
"loss": 0.017,
|
| 1503 |
+
"step": 24900
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 0.49,
|
| 1507 |
+
"learning_rate": 8.717307316570196e-05,
|
| 1508 |
+
"loss": 0.0186,
|
| 1509 |
+
"step": 25000
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 0.49,
|
| 1513 |
+
"learning_rate": 8.664684371396603e-05,
|
| 1514 |
+
"loss": 0.0195,
|
| 1515 |
+
"step": 25100
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 0.5,
|
| 1519 |
+
"learning_rate": 8.612055423046226e-05,
|
| 1520 |
+
"loss": 0.0199,
|
| 1521 |
+
"step": 25200
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.5,
|
| 1525 |
+
"learning_rate": 8.559422485881711e-05,
|
| 1526 |
+
"loss": 0.0191,
|
| 1527 |
+
"step": 25300
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 0.5,
|
| 1531 |
+
"learning_rate": 8.506787574418376e-05,
|
| 1532 |
+
"loss": 0.0191,
|
| 1533 |
+
"step": 25400
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 0.5,
|
| 1537 |
+
"learning_rate": 8.4541527032471e-05,
|
| 1538 |
+
"loss": 0.0196,
|
| 1539 |
+
"step": 25500
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 0.5,
|
| 1543 |
+
"learning_rate": 8.401519886957223e-05,
|
| 1544 |
+
"loss": 0.0184,
|
| 1545 |
+
"step": 25600
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 0.51,
|
| 1549 |
+
"learning_rate": 8.348891140059429e-05,
|
| 1550 |
+
"loss": 0.0182,
|
| 1551 |
+
"step": 25700
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 0.51,
|
| 1555 |
+
"learning_rate": 8.29626847690866e-05,
|
| 1556 |
+
"loss": 0.0216,
|
| 1557 |
+
"step": 25800
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 0.51,
|
| 1561 |
+
"learning_rate": 8.243653911626992e-05,
|
| 1562 |
+
"loss": 0.0177,
|
| 1563 |
+
"step": 25900
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.51,
|
| 1567 |
+
"learning_rate": 8.191575445898302e-05,
|
| 1568 |
+
"loss": 0.019,
|
| 1569 |
+
"step": 26000
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 0.51,
|
| 1573 |
+
"learning_rate": 8.13898298618816e-05,
|
| 1574 |
+
"loss": 0.0187,
|
| 1575 |
+
"step": 26100
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 0.52,
|
| 1579 |
+
"learning_rate": 8.086404644418334e-05,
|
| 1580 |
+
"loss": 0.0189,
|
| 1581 |
+
"step": 26200
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 0.52,
|
| 1585 |
+
"learning_rate": 8.033842433014505e-05,
|
| 1586 |
+
"loss": 0.0178,
|
| 1587 |
+
"step": 26300
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 0.52,
|
| 1591 |
+
"learning_rate": 7.981298363784983e-05,
|
| 1592 |
+
"loss": 0.0182,
|
| 1593 |
+
"step": 26400
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 0.52,
|
| 1597 |
+
"learning_rate": 7.928774447843676e-05,
|
| 1598 |
+
"loss": 0.0197,
|
| 1599 |
+
"step": 26500
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 0.52,
|
| 1603 |
+
"learning_rate": 7.876272695533133e-05,
|
| 1604 |
+
"loss": 0.0182,
|
| 1605 |
+
"step": 26600
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.53,
|
| 1609 |
+
"learning_rate": 7.823795116347596e-05,
|
| 1610 |
+
"loss": 0.0203,
|
| 1611 |
+
"step": 26700
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 0.53,
|
| 1615 |
+
"learning_rate": 7.771343718856085e-05,
|
| 1616 |
+
"loss": 0.0193,
|
| 1617 |
+
"step": 26800
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 0.53,
|
| 1621 |
+
"learning_rate": 7.718920510625514e-05,
|
| 1622 |
+
"loss": 0.0182,
|
| 1623 |
+
"step": 26900
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 0.53,
|
| 1627 |
+
"learning_rate": 7.666527498143874e-05,
|
| 1628 |
+
"loss": 0.0202,
|
| 1629 |
+
"step": 27000
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 0.53,
|
| 1633 |
+
"learning_rate": 7.614166686743405e-05,
|
| 1634 |
+
"loss": 0.019,
|
| 1635 |
+
"step": 27100
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 0.54,
|
| 1639 |
+
"learning_rate": 7.561840080523864e-05,
|
| 1640 |
+
"loss": 0.0174,
|
| 1641 |
+
"step": 27200
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 0.54,
|
| 1645 |
+
"learning_rate": 7.509549682275809e-05,
|
| 1646 |
+
"loss": 0.018,
|
| 1647 |
+
"step": 27300
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.54,
|
| 1651 |
+
"learning_rate": 7.457297493403944e-05,
|
| 1652 |
+
"loss": 0.019,
|
| 1653 |
+
"step": 27400
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 0.54,
|
| 1657 |
+
"learning_rate": 7.405085513850521e-05,
|
| 1658 |
+
"loss": 0.0205,
|
| 1659 |
+
"step": 27500
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 0.54,
|
| 1663 |
+
"learning_rate": 7.352915742018785e-05,
|
| 1664 |
+
"loss": 0.0169,
|
| 1665 |
+
"step": 27600
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 0.55,
|
| 1669 |
+
"learning_rate": 7.30079017469649e-05,
|
| 1670 |
+
"loss": 0.0183,
|
| 1671 |
+
"step": 27700
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 0.55,
|
| 1675 |
+
"learning_rate": 7.248710806979468e-05,
|
| 1676 |
+
"loss": 0.0181,
|
| 1677 |
+
"step": 27800
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 0.55,
|
| 1681 |
+
"learning_rate": 7.196679632195274e-05,
|
| 1682 |
+
"loss": 0.0171,
|
| 1683 |
+
"step": 27900
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.55,
|
| 1687 |
+
"learning_rate": 7.144698641826882e-05,
|
| 1688 |
+
"loss": 0.0202,
|
| 1689 |
+
"step": 28000
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.55,
|
| 1693 |
+
"learning_rate": 7.092769825436483e-05,
|
| 1694 |
+
"loss": 0.0184,
|
| 1695 |
+
"step": 28100
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 0.56,
|
| 1699 |
+
"learning_rate": 7.040895170589299e-05,
|
| 1700 |
+
"loss": 0.0191,
|
| 1701 |
+
"step": 28200
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 0.56,
|
| 1705 |
+
"learning_rate": 6.989594563413708e-05,
|
| 1706 |
+
"loss": 0.0206,
|
| 1707 |
+
"step": 28300
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 0.56,
|
| 1711 |
+
"learning_rate": 6.937833594866479e-05,
|
| 1712 |
+
"loss": 0.0184,
|
| 1713 |
+
"step": 28400
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 0.56,
|
| 1717 |
+
"learning_rate": 6.886132718016212e-05,
|
| 1718 |
+
"loss": 0.0213,
|
| 1719 |
+
"step": 28500
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 0.56,
|
| 1723 |
+
"learning_rate": 6.834493911703791e-05,
|
| 1724 |
+
"loss": 0.0181,
|
| 1725 |
+
"step": 28600
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.57,
|
| 1729 |
+
"learning_rate": 6.78291915239438e-05,
|
| 1730 |
+
"loss": 0.0205,
|
| 1731 |
+
"step": 28700
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.57,
|
| 1735 |
+
"learning_rate": 6.731410414101738e-05,
|
| 1736 |
+
"loss": 0.0184,
|
| 1737 |
+
"step": 28800
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 0.57,
|
| 1741 |
+
"learning_rate": 6.679969668312693e-05,
|
| 1742 |
+
"loss": 0.0179,
|
| 1743 |
+
"step": 28900
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 0.57,
|
| 1747 |
+
"learning_rate": 6.628598883911667e-05,
|
| 1748 |
+
"loss": 0.0206,
|
| 1749 |
+
"step": 29000
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 0.57,
|
| 1753 |
+
"learning_rate": 6.577300027105329e-05,
|
| 1754 |
+
"loss": 0.0172,
|
| 1755 |
+
"step": 29100
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 0.58,
|
| 1759 |
+
"learning_rate": 6.526075061347325e-05,
|
| 1760 |
+
"loss": 0.0194,
|
| 1761 |
+
"step": 29200
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 0.58,
|
| 1765 |
+
"learning_rate": 6.474925947263152e-05,
|
| 1766 |
+
"loss": 0.0195,
|
| 1767 |
+
"step": 29300
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 0.58,
|
| 1771 |
+
"learning_rate": 6.423854642575082e-05,
|
| 1772 |
+
"loss": 0.0193,
|
| 1773 |
+
"step": 29400
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.58,
|
| 1777 |
+
"learning_rate": 6.372863102027257e-05,
|
| 1778 |
+
"loss": 0.0182,
|
| 1779 |
+
"step": 29500
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 0.58,
|
| 1783 |
+
"learning_rate": 6.321953277310858e-05,
|
| 1784 |
+
"loss": 0.0195,
|
| 1785 |
+
"step": 29600
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 0.58,
|
| 1789 |
+
"learning_rate": 6.27112711698941e-05,
|
| 1790 |
+
"loss": 0.0174,
|
| 1791 |
+
"step": 29700
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 0.59,
|
| 1795 |
+
"learning_rate": 6.220386566424204e-05,
|
| 1796 |
+
"loss": 0.0214,
|
| 1797 |
+
"step": 29800
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 0.59,
|
| 1801 |
+
"learning_rate": 6.169733567699826e-05,
|
| 1802 |
+
"loss": 0.0176,
|
| 1803 |
+
"step": 29900
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 0.59,
|
| 1807 |
+
"learning_rate": 6.11917005954984e-05,
|
| 1808 |
+
"loss": 0.017,
|
| 1809 |
+
"step": 30000
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 0.59,
|
| 1813 |
+
"learning_rate": 6.068697977282572e-05,
|
| 1814 |
+
"loss": 0.0185,
|
| 1815 |
+
"step": 30100
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.59,
|
| 1819 |
+
"learning_rate": 6.018319252707041e-05,
|
| 1820 |
+
"loss": 0.0175,
|
| 1821 |
+
"step": 30200
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 0.6,
|
| 1825 |
+
"learning_rate": 5.968035814059016e-05,
|
| 1826 |
+
"loss": 0.0166,
|
| 1827 |
+
"step": 30300
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 0.6,
|
| 1831 |
+
"learning_rate": 5.9178495859272244e-05,
|
| 1832 |
+
"loss": 0.0167,
|
| 1833 |
+
"step": 30400
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 0.6,
|
| 1837 |
+
"learning_rate": 5.8677624891796735e-05,
|
| 1838 |
+
"loss": 0.018,
|
| 1839 |
+
"step": 30500
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 0.6,
|
| 1843 |
+
"learning_rate": 5.8177764408901366e-05,
|
| 1844 |
+
"loss": 0.0182,
|
| 1845 |
+
"step": 30600
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 0.6,
|
| 1849 |
+
"learning_rate": 5.767893354264776e-05,
|
| 1850 |
+
"loss": 0.019,
|
| 1851 |
+
"step": 30700
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 0.61,
|
| 1855 |
+
"learning_rate": 5.7186123953555236e-05,
|
| 1856 |
+
"loss": 0.0198,
|
| 1857 |
+
"step": 30800
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.61,
|
| 1861 |
+
"learning_rate": 5.668939878661137e-05,
|
| 1862 |
+
"loss": 0.0177,
|
| 1863 |
+
"step": 30900
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 0.61,
|
| 1867 |
+
"learning_rate": 5.619376020321073e-05,
|
| 1868 |
+
"loss": 0.0165,
|
| 1869 |
+
"step": 31000
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 0.61,
|
| 1873 |
+
"learning_rate": 5.5699227173822495e-05,
|
| 1874 |
+
"loss": 0.0205,
|
| 1875 |
+
"step": 31100
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 0.61,
|
| 1879 |
+
"learning_rate": 5.5205818626601115e-05,
|
| 1880 |
+
"loss": 0.0181,
|
| 1881 |
+
"step": 31200
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 0.62,
|
| 1885 |
+
"learning_rate": 5.4713553446661644e-05,
|
| 1886 |
+
"loss": 0.0187,
|
| 1887 |
+
"step": 31300
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 0.62,
|
| 1891 |
+
"learning_rate": 5.4222450475357026e-05,
|
| 1892 |
+
"loss": 0.0169,
|
| 1893 |
+
"step": 31400
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 0.62,
|
| 1897 |
+
"learning_rate": 5.373252850955681e-05,
|
| 1898 |
+
"loss": 0.0174,
|
| 1899 |
+
"step": 31500
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.62,
|
| 1903 |
+
"learning_rate": 5.3243806300927926e-05,
|
| 1904 |
+
"loss": 0.0179,
|
| 1905 |
+
"step": 31600
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 0.62,
|
| 1909 |
+
"learning_rate": 5.2756302555216735e-05,
|
| 1910 |
+
"loss": 0.0175,
|
| 1911 |
+
"step": 31700
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 0.63,
|
| 1915 |
+
"learning_rate": 5.2270035931533207e-05,
|
| 1916 |
+
"loss": 0.0199,
|
| 1917 |
+
"step": 31800
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 0.63,
|
| 1921 |
+
"learning_rate": 5.1785025041636665e-05,
|
| 1922 |
+
"loss": 0.0184,
|
| 1923 |
+
"step": 31900
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 0.63,
|
| 1927 |
+
"learning_rate": 5.130128844922355e-05,
|
| 1928 |
+
"loss": 0.0185,
|
| 1929 |
+
"step": 32000
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 0.63,
|
| 1933 |
+
"learning_rate": 5.081884466921669e-05,
|
| 1934 |
+
"loss": 0.0174,
|
| 1935 |
+
"step": 32100
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 0.63,
|
| 1939 |
+
"learning_rate": 5.0337712167056944e-05,
|
| 1940 |
+
"loss": 0.0201,
|
| 1941 |
+
"step": 32200
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.64,
|
| 1945 |
+
"learning_rate": 4.985790935799614e-05,
|
| 1946 |
+
"loss": 0.0178,
|
| 1947 |
+
"step": 32300
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 0.64,
|
| 1951 |
+
"learning_rate": 4.937945460639236e-05,
|
| 1952 |
+
"loss": 0.0176,
|
| 1953 |
+
"step": 32400
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 0.64,
|
| 1957 |
+
"learning_rate": 4.890236622500717e-05,
|
| 1958 |
+
"loss": 0.0239,
|
| 1959 |
+
"step": 32500
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 0.64,
|
| 1963 |
+
"learning_rate": 4.8426662474304465e-05,
|
| 1964 |
+
"loss": 0.0177,
|
| 1965 |
+
"step": 32600
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 0.64,
|
| 1969 |
+
"learning_rate": 4.795236156175173e-05,
|
| 1970 |
+
"loss": 0.0182,
|
| 1971 |
+
"step": 32700
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 0.65,
|
| 1975 |
+
"learning_rate": 4.7479481641123066e-05,
|
| 1976 |
+
"loss": 0.016,
|
| 1977 |
+
"step": 32800
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 0.65,
|
| 1981 |
+
"learning_rate": 4.7008040811804424e-05,
|
| 1982 |
+
"loss": 0.0175,
|
| 1983 |
+
"step": 32900
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.65,
|
| 1987 |
+
"learning_rate": 4.653805711810078e-05,
|
| 1988 |
+
"loss": 0.0173,
|
| 1989 |
+
"step": 33000
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 0.65,
|
| 1993 |
+
"learning_rate": 4.6069548548545565e-05,
|
| 1994 |
+
"loss": 0.0179,
|
| 1995 |
+
"step": 33100
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 0.65,
|
| 1999 |
+
"learning_rate": 4.560253303521206e-05,
|
| 2000 |
+
"loss": 0.0186,
|
| 2001 |
+
"step": 33200
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 0.66,
|
| 2005 |
+
"learning_rate": 4.513702845302723e-05,
|
| 2006 |
+
"loss": 0.019,
|
| 2007 |
+
"step": 33300
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 0.66,
|
| 2011 |
+
"learning_rate": 4.467305261908736e-05,
|
| 2012 |
+
"loss": 0.0164,
|
| 2013 |
+
"step": 33400
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 0.66,
|
| 2017 |
+
"learning_rate": 4.42106232919762e-05,
|
| 2018 |
+
"loss": 0.019,
|
| 2019 |
+
"step": 33500
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 0.66,
|
| 2023 |
+
"learning_rate": 4.374975817108527e-05,
|
| 2024 |
+
"loss": 0.0174,
|
| 2025 |
+
"step": 33600
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.66,
|
| 2029 |
+
"learning_rate": 4.3295059840780387e-05,
|
| 2030 |
+
"loss": 0.0187,
|
| 2031 |
+
"step": 33700
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 0.67,
|
| 2035 |
+
"learning_rate": 4.283735990928788e-05,
|
| 2036 |
+
"loss": 0.019,
|
| 2037 |
+
"step": 33800
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 0.67,
|
| 2041 |
+
"learning_rate": 4.2381276745401456e-05,
|
| 2042 |
+
"loss": 0.0175,
|
| 2043 |
+
"step": 33900
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 0.67,
|
| 2047 |
+
"learning_rate": 4.192682780561444e-05,
|
| 2048 |
+
"loss": 0.0186,
|
| 2049 |
+
"step": 34000
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 0.67,
|
| 2053 |
+
"learning_rate": 4.147403048387059e-05,
|
| 2054 |
+
"loss": 0.02,
|
| 2055 |
+
"step": 34100
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 0.67,
|
| 2059 |
+
"learning_rate": 4.1022902110898326e-05,
|
| 2060 |
+
"loss": 0.0169,
|
| 2061 |
+
"step": 34200
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 0.68,
|
| 2065 |
+
"learning_rate": 4.057345995354738e-05,
|
| 2066 |
+
"loss": 0.0176,
|
| 2067 |
+
"step": 34300
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.68,
|
| 2071 |
+
"learning_rate": 4.0125721214127854e-05,
|
| 2072 |
+
"loss": 0.0167,
|
| 2073 |
+
"step": 34400
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 0.68,
|
| 2077 |
+
"learning_rate": 3.9679703029752e-05,
|
| 2078 |
+
"loss": 0.0179,
|
| 2079 |
+
"step": 34500
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 0.68,
|
| 2083 |
+
"learning_rate": 3.923542247167802e-05,
|
| 2084 |
+
"loss": 0.0162,
|
| 2085 |
+
"step": 34600
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 0.68,
|
| 2089 |
+
"learning_rate": 3.879289654465689e-05,
|
| 2090 |
+
"loss": 0.0194,
|
| 2091 |
+
"step": 34700
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 0.69,
|
| 2095 |
+
"learning_rate": 3.835214218628141e-05,
|
| 2096 |
+
"loss": 0.0187,
|
| 2097 |
+
"step": 34800
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 0.69,
|
| 2101 |
+
"learning_rate": 3.7913176266337885e-05,
|
| 2102 |
+
"loss": 0.0165,
|
| 2103 |
+
"step": 34900
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 0.69,
|
| 2107 |
+
"learning_rate": 3.747601558616062e-05,
|
| 2108 |
+
"loss": 0.0162,
|
| 2109 |
+
"step": 35000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.69,
|
| 2113 |
+
"learning_rate": 3.70406768779886e-05,
|
| 2114 |
+
"loss": 0.0169,
|
| 2115 |
+
"step": 35100
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 0.69,
|
| 2119 |
+
"learning_rate": 3.660717680432526e-05,
|
| 2120 |
+
"loss": 0.0173,
|
| 2121 |
+
"step": 35200
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 0.7,
|
| 2125 |
+
"learning_rate": 3.61755319573006e-05,
|
| 2126 |
+
"loss": 0.016,
|
| 2127 |
+
"step": 35300
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 0.7,
|
| 2131 |
+
"learning_rate": 3.5745758858036255e-05,
|
| 2132 |
+
"loss": 0.0188,
|
| 2133 |
+
"step": 35400
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 0.7,
|
| 2137 |
+
"learning_rate": 3.5317873956012995e-05,
|
| 2138 |
+
"loss": 0.0198,
|
| 2139 |
+
"step": 35500
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 0.7,
|
| 2143 |
+
"learning_rate": 3.4891893628441334e-05,
|
| 2144 |
+
"loss": 0.0171,
|
| 2145 |
+
"step": 35600
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 0.7,
|
| 2149 |
+
"learning_rate": 3.446783417963448e-05,
|
| 2150 |
+
"loss": 0.0179,
|
| 2151 |
+
"step": 35700
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.71,
|
| 2155 |
+
"learning_rate": 3.4045711840384404e-05,
|
| 2156 |
+
"loss": 0.0158,
|
| 2157 |
+
"step": 35800
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 0.71,
|
| 2161 |
+
"learning_rate": 3.36255427673406e-05,
|
| 2162 |
+
"loss": 0.0171,
|
| 2163 |
+
"step": 35900
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 0.71,
|
| 2167 |
+
"learning_rate": 3.320734304239168e-05,
|
| 2168 |
+
"loss": 0.0162,
|
| 2169 |
+
"step": 36000
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 0.71,
|
| 2173 |
+
"learning_rate": 3.279112867204983e-05,
|
| 2174 |
+
"loss": 0.0172,
|
| 2175 |
+
"step": 36100
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 0.71,
|
| 2179 |
+
"learning_rate": 3.2376915586838195e-05,
|
| 2180 |
+
"loss": 0.0166,
|
| 2181 |
+
"step": 36200
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 0.71,
|
| 2185 |
+
"learning_rate": 3.196471964068113e-05,
|
| 2186 |
+
"loss": 0.0171,
|
| 2187 |
+
"step": 36300
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 0.72,
|
| 2191 |
+
"learning_rate": 3.1554556610297393e-05,
|
| 2192 |
+
"loss": 0.019,
|
| 2193 |
+
"step": 36400
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.72,
|
| 2197 |
+
"learning_rate": 3.1146442194596243e-05,
|
| 2198 |
+
"loss": 0.0167,
|
| 2199 |
+
"step": 36500
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 0.72,
|
| 2203 |
+
"learning_rate": 3.074039201407669e-05,
|
| 2204 |
+
"loss": 0.0172,
|
| 2205 |
+
"step": 36600
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 0.72,
|
| 2209 |
+
"learning_rate": 3.0336421610229454e-05,
|
| 2210 |
+
"loss": 0.0182,
|
| 2211 |
+
"step": 36700
|
| 2212 |
+
},
|
| 2213 |
+
{
|
| 2214 |
+
"epoch": 0.72,
|
| 2215 |
+
"learning_rate": 2.9934546444942337e-05,
|
| 2216 |
+
"loss": 0.0171,
|
| 2217 |
+
"step": 36800
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 0.73,
|
| 2221 |
+
"learning_rate": 2.9534781899908195e-05,
|
| 2222 |
+
"loss": 0.0165,
|
| 2223 |
+
"step": 36900
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 0.73,
|
| 2227 |
+
"learning_rate": 2.9141109088924195e-05,
|
| 2228 |
+
"loss": 0.02,
|
| 2229 |
+
"step": 37000
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 0.73,
|
| 2233 |
+
"learning_rate": 2.87455901192796e-05,
|
| 2234 |
+
"loss": 0.0174,
|
| 2235 |
+
"step": 37100
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.73,
|
| 2239 |
+
"learning_rate": 2.835222727695746e-05,
|
| 2240 |
+
"loss": 0.0171,
|
| 2241 |
+
"step": 37200
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 0.73,
|
| 2245 |
+
"learning_rate": 2.7961035617843112e-05,
|
| 2246 |
+
"loss": 0.0172,
|
| 2247 |
+
"step": 37300
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 0.74,
|
| 2251 |
+
"learning_rate": 2.757203011472043e-05,
|
| 2252 |
+
"loss": 0.0164,
|
| 2253 |
+
"step": 37400
|
| 2254 |
+
},
|
| 2255 |
+
{
|
| 2256 |
+
"epoch": 0.74,
|
| 2257 |
+
"learning_rate": 2.718522565669847e-05,
|
| 2258 |
+
"loss": 0.017,
|
| 2259 |
+
"step": 37500
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 0.74,
|
| 2263 |
+
"learning_rate": 2.6800637048641775e-05,
|
| 2264 |
+
"loss": 0.0167,
|
| 2265 |
+
"step": 37600
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 0.74,
|
| 2269 |
+
"learning_rate": 2.6418279010603663e-05,
|
| 2270 |
+
"loss": 0.0184,
|
| 2271 |
+
"step": 37700
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 0.74,
|
| 2275 |
+
"learning_rate": 2.6038166177262784e-05,
|
| 2276 |
+
"loss": 0.0158,
|
| 2277 |
+
"step": 37800
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.75,
|
| 2281 |
+
"learning_rate": 2.566031309736311e-05,
|
| 2282 |
+
"loss": 0.0165,
|
| 2283 |
+
"step": 37900
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 0.75,
|
| 2287 |
+
"learning_rate": 2.5284734233156903e-05,
|
| 2288 |
+
"loss": 0.0165,
|
| 2289 |
+
"step": 38000
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 0.75,
|
| 2293 |
+
"learning_rate": 2.4911443959851313e-05,
|
| 2294 |
+
"loss": 0.017,
|
| 2295 |
+
"step": 38100
|
| 2296 |
+
},
|
| 2297 |
+
{
|
| 2298 |
+
"epoch": 0.75,
|
| 2299 |
+
"learning_rate": 2.4540456565058115e-05,
|
| 2300 |
+
"loss": 0.0177,
|
| 2301 |
+
"step": 38200
|
| 2302 |
+
},
|
| 2303 |
+
{
|
| 2304 |
+
"epoch": 0.75,
|
| 2305 |
+
"learning_rate": 2.417178624824684e-05,
|
| 2306 |
+
"loss": 0.0165,
|
| 2307 |
+
"step": 38300
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 0.76,
|
| 2311 |
+
"learning_rate": 2.3805447120201323e-05,
|
| 2312 |
+
"loss": 0.0173,
|
| 2313 |
+
"step": 38400
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 0.76,
|
| 2317 |
+
"learning_rate": 2.344145320247968e-05,
|
| 2318 |
+
"loss": 0.0189,
|
| 2319 |
+
"step": 38500
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.76,
|
| 2323 |
+
"learning_rate": 2.3079818426877448e-05,
|
| 2324 |
+
"loss": 0.0183,
|
| 2325 |
+
"step": 38600
|
| 2326 |
+
},
|
| 2327 |
+
{
|
| 2328 |
+
"epoch": 0.76,
|
| 2329 |
+
"learning_rate": 2.2720556634894542e-05,
|
| 2330 |
+
"loss": 0.0161,
|
| 2331 |
+
"step": 38700
|
| 2332 |
+
},
|
| 2333 |
+
{
|
| 2334 |
+
"epoch": 0.76,
|
| 2335 |
+
"learning_rate": 2.2363681577205363e-05,
|
| 2336 |
+
"loss": 0.0161,
|
| 2337 |
+
"step": 38800
|
| 2338 |
+
},
|
| 2339 |
+
{
|
| 2340 |
+
"epoch": 0.77,
|
| 2341 |
+
"learning_rate": 2.2009206913132545e-05,
|
| 2342 |
+
"loss": 0.017,
|
| 2343 |
+
"step": 38900
|
| 2344 |
+
},
|
| 2345 |
+
{
|
| 2346 |
+
"epoch": 0.77,
|
| 2347 |
+
"learning_rate": 2.1657146210124072e-05,
|
| 2348 |
+
"loss": 0.0175,
|
| 2349 |
+
"step": 39000
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 0.77,
|
| 2353 |
+
"learning_rate": 2.1307512943234157e-05,
|
| 2354 |
+
"loss": 0.0162,
|
| 2355 |
+
"step": 39100
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 0.77,
|
| 2359 |
+
"learning_rate": 2.0960320494607276e-05,
|
| 2360 |
+
"loss": 0.0161,
|
| 2361 |
+
"step": 39200
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.77,
|
| 2365 |
+
"learning_rate": 2.0615582152966087e-05,
|
| 2366 |
+
"loss": 0.0174,
|
| 2367 |
+
"step": 39300
|
| 2368 |
+
},
|
| 2369 |
+
{
|
| 2370 |
+
"epoch": 0.78,
|
| 2371 |
+
"learning_rate": 2.027672156726409e-05,
|
| 2372 |
+
"loss": 0.0194,
|
| 2373 |
+
"step": 39400
|
| 2374 |
+
},
|
| 2375 |
+
{
|
| 2376 |
+
"epoch": 0.78,
|
| 2377 |
+
"learning_rate": 1.9936906060978636e-05,
|
| 2378 |
+
"loss": 0.0185,
|
| 2379 |
+
"step": 39500
|
| 2380 |
+
},
|
| 2381 |
+
{
|
| 2382 |
+
"epoch": 0.78,
|
| 2383 |
+
"learning_rate": 1.9599583832664966e-05,
|
| 2384 |
+
"loss": 0.0169,
|
| 2385 |
+
"step": 39600
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 0.78,
|
| 2389 |
+
"learning_rate": 1.9264767793265046e-05,
|
| 2390 |
+
"loss": 0.0174,
|
| 2391 |
+
"step": 39700
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 0.78,
|
| 2395 |
+
"learning_rate": 1.8932470757796908e-05,
|
| 2396 |
+
"loss": 0.0172,
|
| 2397 |
+
"step": 39800
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 0.79,
|
| 2401 |
+
"learning_rate": 1.86027054448642e-05,
|
| 2402 |
+
"loss": 0.0157,
|
| 2403 |
+
"step": 39900
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.79,
|
| 2407 |
+
"learning_rate": 1.8275484476169405e-05,
|
| 2408 |
+
"loss": 0.0166,
|
| 2409 |
+
"step": 40000
|
| 2410 |
+
}
|
| 2411 |
+
],
|
| 2412 |
+
"logging_steps": 100,
|
| 2413 |
+
"max_steps": 50780,
|
| 2414 |
+
"num_train_epochs": 1,
|
| 2415 |
+
"save_steps": 10000,
|
| 2416 |
+
"total_flos": 1.5971349454848e+19,
|
| 2417 |
+
"trial_name": null,
|
| 2418 |
+
"trial_params": null
|
| 2419 |
+
}
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d17a2c3c8a65881aa45bfd2997dfe652a61cb20d8b49594dc95bff2d7669f6f
|
| 3 |
+
size 4155
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-40000/vocab.txt
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
<cls>
|
| 2 |
+
<pad>
|
| 3 |
+
<eos>
|
| 4 |
+
<unk>
|
| 5 |
+
L
|
| 6 |
+
A
|
| 7 |
+
G
|
| 8 |
+
V
|
| 9 |
+
S
|
| 10 |
+
E
|
| 11 |
+
R
|
| 12 |
+
T
|
| 13 |
+
I
|
| 14 |
+
D
|
| 15 |
+
P
|
| 16 |
+
K
|
| 17 |
+
Q
|
| 18 |
+
N
|
| 19 |
+
F
|
| 20 |
+
Y
|
| 21 |
+
M
|
| 22 |
+
H
|
| 23 |
+
W
|
| 24 |
+
C
|
| 25 |
+
X
|
| 26 |
+
B
|
| 27 |
+
U
|
| 28 |
+
Z
|
| 29 |
+
O
|
| 30 |
+
.
|
| 31 |
+
-
|
| 32 |
+
<null_1>
|
| 33 |
+
<mask>
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-50000/README.md
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
| 4 |
+
## Training procedure
|
| 5 |
+
|
| 6 |
+
|
| 7 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 8 |
+
- quant_method: bitsandbytes
|
| 9 |
+
- load_in_8bit: False
|
| 10 |
+
- load_in_4bit: True
|
| 11 |
+
- llm_int8_threshold: 6.0
|
| 12 |
+
- llm_int8_skip_modules: None
|
| 13 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 14 |
+
- llm_int8_has_fp16_weight: False
|
| 15 |
+
- bnb_4bit_quant_type: nf4
|
| 16 |
+
- bnb_4bit_use_double_quant: True
|
| 17 |
+
- bnb_4bit_compute_dtype: bfloat16
|
| 18 |
+
### Framework versions
|
| 19 |
+
|
| 20 |
+
|
| 21 |
+
- PEFT 0.6.0.dev0
|
esm2_t33_650M_qlora_binding_sites_2023-10-18_02-14-48/checkpoint-50000/adapter_config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "facebook/esm2_t33_650M_UR50D",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"fan_in_fan_out": false,
|
| 7 |
+
"inference_mode": true,
|
| 8 |
+
"init_lora_weights": true,
|
| 9 |
+
"layers_pattern": null,
|
| 10 |
+
"layers_to_transform": null,
|
| 11 |
+
"lora_alpha": 1,
|
| 12 |
+
"lora_dropout": 0.5,
|
| 13 |
+
"modules_to_save": null,
|
| 14 |
+
"peft_type": "LORA",
|
| 15 |
+
"r": 2,
|
| 16 |
+
"rank_pattern": {},
|
| 17 |
+
"revision": null,
|
| 18 |
+
"target_modules": [
|
| 19 |
+
"query",
|
| 20 |
+
"key",
|
| 21 |
+
"value",
|
| 22 |
+
"EsmSelfOutput.dense",
|
| 23 |
+
"EsmIntermediate.dense",
|
| 24 |
+
"EsmOutput.dense",
|
| 25 |
+
"classifier"
|
| 26 |
+
],
|
| 27 |
+
"task_type": "TOKEN_CLS"
|
| 28 |
+
}
|