Commit
•
9d46703
1
Parent(s):
39efcee
commit from root
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- README.md +20 -0
- adapter_config.json +21 -0
- adapter_model.bin +3 -0
- all_results.json +7 -0
- checkpoint-1000/README.md +20 -0
- checkpoint-1000/adapter_config.json +21 -0
- checkpoint-1000/adapter_model.bin +3 -0
- checkpoint-1000/finetuning_args.json +16 -0
- checkpoint-1000/optimizer.pt +3 -0
- checkpoint-1000/rng_state_0.pth +3 -0
- checkpoint-1000/rng_state_1.pth +3 -0
- checkpoint-1000/rng_state_2.pth +3 -0
- checkpoint-1000/scheduler.pt +3 -0
- checkpoint-1000/trainer_state.json +616 -0
- checkpoint-1000/training_args.bin +3 -0
- checkpoint-2000/README.md +20 -0
- checkpoint-2000/adapter_config.json +21 -0
- checkpoint-2000/adapter_model.bin +3 -0
- checkpoint-2000/finetuning_args.json +16 -0
- checkpoint-2000/optimizer.pt +3 -0
- checkpoint-2000/rng_state_0.pth +3 -0
- checkpoint-2000/rng_state_1.pth +3 -0
- checkpoint-2000/rng_state_2.pth +3 -0
- checkpoint-2000/scheduler.pt +3 -0
- checkpoint-2000/trainer_state.json +1216 -0
- checkpoint-2000/training_args.bin +3 -0
- checkpoint-3000/README.md +20 -0
- checkpoint-3000/adapter_config.json +21 -0
- checkpoint-3000/adapter_model.bin +3 -0
- checkpoint-3000/finetuning_args.json +16 -0
- checkpoint-3000/optimizer.pt +3 -0
- checkpoint-3000/rng_state_0.pth +3 -0
- checkpoint-3000/rng_state_1.pth +3 -0
- checkpoint-3000/rng_state_2.pth +3 -0
- checkpoint-3000/scheduler.pt +3 -0
- checkpoint-3000/trainer_state.json +1816 -0
- checkpoint-3000/training_args.bin +3 -0
- checkpoint-4000/README.md +20 -0
- checkpoint-4000/adapter_config.json +21 -0
- checkpoint-4000/adapter_model.bin +3 -0
- checkpoint-4000/finetuning_args.json +16 -0
- checkpoint-4000/optimizer.pt +3 -0
- checkpoint-4000/rng_state_0.pth +3 -0
- checkpoint-4000/rng_state_1.pth +3 -0
- checkpoint-4000/rng_state_2.pth +3 -0
- checkpoint-4000/scheduler.pt +3 -0
- checkpoint-4000/trainer_state.json +2416 -0
- checkpoint-4000/training_args.bin +3 -0
- checkpoint-5000/README.md +20 -0
- checkpoint-5000/adapter_config.json +21 -0
README.md
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
| 4 |
+
## Training procedure
|
| 5 |
+
|
| 6 |
+
|
| 7 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 8 |
+
- load_in_8bit: False
|
| 9 |
+
- load_in_4bit: True
|
| 10 |
+
- llm_int8_threshold: 6.0
|
| 11 |
+
- llm_int8_skip_modules: None
|
| 12 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 13 |
+
- llm_int8_has_fp16_weight: False
|
| 14 |
+
- bnb_4bit_quant_type: nf4
|
| 15 |
+
- bnb_4bit_use_double_quant: True
|
| 16 |
+
- bnb_4bit_compute_dtype: float16
|
| 17 |
+
### Framework versions
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
- PEFT 0.4.0
|
adapter_config.json
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "./Llama-2-7b-chat-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"fan_in_fan_out": false,
|
| 6 |
+
"inference_mode": true,
|
| 7 |
+
"init_lora_weights": true,
|
| 8 |
+
"layers_pattern": null,
|
| 9 |
+
"layers_to_transform": null,
|
| 10 |
+
"lora_alpha": 32.0,
|
| 11 |
+
"lora_dropout": 0.1,
|
| 12 |
+
"modules_to_save": null,
|
| 13 |
+
"peft_type": "LORA",
|
| 14 |
+
"r": 8,
|
| 15 |
+
"revision": null,
|
| 16 |
+
"target_modules": [
|
| 17 |
+
"q_proj",
|
| 18 |
+
"v_proj"
|
| 19 |
+
],
|
| 20 |
+
"task_type": "CAUSAL_LM"
|
| 21 |
+
}
|
adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d94d4e6266cd52f738d85bee34e2b0d0585cd6397ef83d164e2588e8ba35d35f
|
| 3 |
+
size 16821197
|
all_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"train_loss": 0.7592291734989016,
|
| 4 |
+
"train_runtime": 91923.4907,
|
| 5 |
+
"train_samples_per_second": 6.42,
|
| 6 |
+
"train_steps_per_second": 0.067
|
| 7 |
+
}
|
checkpoint-1000/README.md
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
| 4 |
+
## Training procedure
|
| 5 |
+
|
| 6 |
+
|
| 7 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 8 |
+
- load_in_8bit: False
|
| 9 |
+
- load_in_4bit: True
|
| 10 |
+
- llm_int8_threshold: 6.0
|
| 11 |
+
- llm_int8_skip_modules: None
|
| 12 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 13 |
+
- llm_int8_has_fp16_weight: False
|
| 14 |
+
- bnb_4bit_quant_type: nf4
|
| 15 |
+
- bnb_4bit_use_double_quant: True
|
| 16 |
+
- bnb_4bit_compute_dtype: float16
|
| 17 |
+
### Framework versions
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
- PEFT 0.4.0
|
checkpoint-1000/adapter_config.json
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "./Llama-2-7b-chat-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"fan_in_fan_out": false,
|
| 6 |
+
"inference_mode": true,
|
| 7 |
+
"init_lora_weights": true,
|
| 8 |
+
"layers_pattern": null,
|
| 9 |
+
"layers_to_transform": null,
|
| 10 |
+
"lora_alpha": 32.0,
|
| 11 |
+
"lora_dropout": 0.1,
|
| 12 |
+
"modules_to_save": null,
|
| 13 |
+
"peft_type": "LORA",
|
| 14 |
+
"r": 8,
|
| 15 |
+
"revision": null,
|
| 16 |
+
"target_modules": [
|
| 17 |
+
"q_proj",
|
| 18 |
+
"v_proj"
|
| 19 |
+
],
|
| 20 |
+
"task_type": "CAUSAL_LM"
|
| 21 |
+
}
|
checkpoint-1000/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f98b4cc2a828aff1dd5267d95be6fe53034890031c35689033e6eeafeb7d60e6
|
| 3 |
+
size 16821197
|
checkpoint-1000/finetuning_args.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"dpo_beta": 0.1,
|
| 3 |
+
"finetuning_type": "lora",
|
| 4 |
+
"lora_alpha": 32.0,
|
| 5 |
+
"lora_dropout": 0.1,
|
| 6 |
+
"lora_rank": 8,
|
| 7 |
+
"lora_target": [
|
| 8 |
+
"q_proj",
|
| 9 |
+
"v_proj"
|
| 10 |
+
],
|
| 11 |
+
"name_module_trainable": "mlp",
|
| 12 |
+
"num_hidden_layers": 32,
|
| 13 |
+
"num_layer_trainable": 3,
|
| 14 |
+
"ppo_score_norm": false,
|
| 15 |
+
"resume_lora_training": true
|
| 16 |
+
}
|
checkpoint-1000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:562775fcc4e95414c715fede9d871607f8a2118bdda55c5ba334aa92f0a3362c
|
| 3 |
+
size 33661637
|
checkpoint-1000/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6b31188a9e96043ff907f0fae1d7f1e48070209b8aba5a327972bfaf50e3a0ee
|
| 3 |
+
size 18679
|
checkpoint-1000/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0faa3bb92522a2d535dbc341e6f02d26880a9b1fa6a063079dbf6de31f5d9ad2
|
| 3 |
+
size 18679
|
checkpoint-1000/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f6d514f9f370ddc3811b7867487be2874d6d37b3abce9cb40d4bd360e354cbaa
|
| 3 |
+
size 18679
|
checkpoint-1000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:399e473545277d3bb53b675c2aee4ff085372e2cfcdd5a0b762fa92fbcfcfd6b
|
| 3 |
+
size 627
|
checkpoint-1000/trainer_state.json
ADDED
|
@@ -0,0 +1,616 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.32533550223668156,
|
| 5 |
+
"global_step": 1000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 4.999967339434433e-05,
|
| 13 |
+
"loss": 1.343,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 4.9998820960283476e-05,
|
| 19 |
+
"loss": 1.1858,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 4.99974394497948e-05,
|
| 25 |
+
"loss": 1.1057,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.01,
|
| 30 |
+
"learning_rate": 4.9995528892116685e-05,
|
| 31 |
+
"loss": 1.0698,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.02,
|
| 36 |
+
"learning_rate": 4.999308932768425e-05,
|
| 37 |
+
"loss": 1.0316,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.02,
|
| 42 |
+
"learning_rate": 4.99897583236926e-05,
|
| 43 |
+
"loss": 1.0046,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.02,
|
| 48 |
+
"learning_rate": 4.9985774375988764e-05,
|
| 49 |
+
"loss": 0.9848,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.03,
|
| 54 |
+
"learning_rate": 4.998113758866712e-05,
|
| 55 |
+
"loss": 0.9656,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.03,
|
| 60 |
+
"learning_rate": 4.997584808287975e-05,
|
| 61 |
+
"loss": 0.9648,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.03,
|
| 66 |
+
"learning_rate": 4.996990599683325e-05,
|
| 67 |
+
"loss": 0.9415,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.04,
|
| 72 |
+
"learning_rate": 4.996331148578514e-05,
|
| 73 |
+
"loss": 0.9388,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.04,
|
| 78 |
+
"learning_rate": 4.995606472203979e-05,
|
| 79 |
+
"loss": 0.9289,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.04,
|
| 84 |
+
"learning_rate": 4.994816589494391e-05,
|
| 85 |
+
"loss": 0.9252,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.05,
|
| 90 |
+
"learning_rate": 4.993961521088163e-05,
|
| 91 |
+
"loss": 0.91,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.05,
|
| 96 |
+
"learning_rate": 4.9930412893269106e-05,
|
| 97 |
+
"loss": 0.9033,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.05,
|
| 102 |
+
"learning_rate": 4.992055918254865e-05,
|
| 103 |
+
"loss": 0.9064,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.06,
|
| 108 |
+
"learning_rate": 4.9910054336182467e-05,
|
| 109 |
+
"loss": 0.8946,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.06,
|
| 114 |
+
"learning_rate": 4.989889862864594e-05,
|
| 115 |
+
"loss": 0.8914,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.06,
|
| 120 |
+
"learning_rate": 4.988709235142044e-05,
|
| 121 |
+
"loss": 0.9007,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.07,
|
| 126 |
+
"learning_rate": 4.9874635812985734e-05,
|
| 127 |
+
"loss": 0.9008,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.07,
|
| 132 |
+
"learning_rate": 4.9861529338811874e-05,
|
| 133 |
+
"loss": 0.8714,
|
| 134 |
+
"step": 210
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.07,
|
| 138 |
+
"learning_rate": 4.984777327135076e-05,
|
| 139 |
+
"loss": 0.8865,
|
| 140 |
+
"step": 220
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.07,
|
| 144 |
+
"learning_rate": 4.9833367970027145e-05,
|
| 145 |
+
"loss": 0.8716,
|
| 146 |
+
"step": 230
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.08,
|
| 150 |
+
"learning_rate": 4.981831381122926e-05,
|
| 151 |
+
"loss": 0.8817,
|
| 152 |
+
"step": 240
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.08,
|
| 156 |
+
"learning_rate": 4.980261118829896e-05,
|
| 157 |
+
"loss": 0.8658,
|
| 158 |
+
"step": 250
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.08,
|
| 162 |
+
"learning_rate": 4.978626051152152e-05,
|
| 163 |
+
"loss": 0.8722,
|
| 164 |
+
"step": 260
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.09,
|
| 168 |
+
"learning_rate": 4.976926220811479e-05,
|
| 169 |
+
"loss": 0.8541,
|
| 170 |
+
"step": 270
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.09,
|
| 174 |
+
"learning_rate": 4.975161672221813e-05,
|
| 175 |
+
"loss": 0.8504,
|
| 176 |
+
"step": 280
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.09,
|
| 180 |
+
"learning_rate": 4.97333245148808e-05,
|
| 181 |
+
"loss": 0.8467,
|
| 182 |
+
"step": 290
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.1,
|
| 186 |
+
"learning_rate": 4.971438606404986e-05,
|
| 187 |
+
"loss": 0.8533,
|
| 188 |
+
"step": 300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.1,
|
| 192 |
+
"learning_rate": 4.969480186455771e-05,
|
| 193 |
+
"loss": 0.853,
|
| 194 |
+
"step": 310
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.1,
|
| 198 |
+
"learning_rate": 4.9674572428109205e-05,
|
| 199 |
+
"loss": 0.8432,
|
| 200 |
+
"step": 320
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.11,
|
| 204 |
+
"learning_rate": 4.9653698283268183e-05,
|
| 205 |
+
"loss": 0.854,
|
| 206 |
+
"step": 330
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.11,
|
| 210 |
+
"learning_rate": 4.963217997544376e-05,
|
| 211 |
+
"loss": 0.8435,
|
| 212 |
+
"step": 340
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.11,
|
| 216 |
+
"learning_rate": 4.961001806687602e-05,
|
| 217 |
+
"loss": 0.8518,
|
| 218 |
+
"step": 350
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.12,
|
| 222 |
+
"learning_rate": 4.958721313662133e-05,
|
| 223 |
+
"loss": 0.848,
|
| 224 |
+
"step": 360
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.12,
|
| 228 |
+
"learning_rate": 4.956376578053723e-05,
|
| 229 |
+
"loss": 0.8436,
|
| 230 |
+
"step": 370
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.12,
|
| 234 |
+
"learning_rate": 4.9539676611266846e-05,
|
| 235 |
+
"loss": 0.8405,
|
| 236 |
+
"step": 380
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.13,
|
| 240 |
+
"learning_rate": 4.9514946258222895e-05,
|
| 241 |
+
"loss": 0.838,
|
| 242 |
+
"step": 390
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.13,
|
| 246 |
+
"learning_rate": 4.948957536757123e-05,
|
| 247 |
+
"loss": 0.8458,
|
| 248 |
+
"step": 400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.13,
|
| 252 |
+
"learning_rate": 4.946356460221395e-05,
|
| 253 |
+
"loss": 0.8379,
|
| 254 |
+
"step": 410
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.14,
|
| 258 |
+
"learning_rate": 4.943691464177212e-05,
|
| 259 |
+
"loss": 0.827,
|
| 260 |
+
"step": 420
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.14,
|
| 264 |
+
"learning_rate": 4.940962618256796e-05,
|
| 265 |
+
"loss": 0.8208,
|
| 266 |
+
"step": 430
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.14,
|
| 270 |
+
"learning_rate": 4.9381699937606674e-05,
|
| 271 |
+
"loss": 0.8242,
|
| 272 |
+
"step": 440
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.15,
|
| 276 |
+
"learning_rate": 4.935313663655782e-05,
|
| 277 |
+
"loss": 0.8362,
|
| 278 |
+
"step": 450
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.15,
|
| 282 |
+
"learning_rate": 4.932393702573627e-05,
|
| 283 |
+
"loss": 0.8363,
|
| 284 |
+
"step": 460
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.15,
|
| 288 |
+
"learning_rate": 4.929410186808264e-05,
|
| 289 |
+
"loss": 0.8275,
|
| 290 |
+
"step": 470
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.16,
|
| 294 |
+
"learning_rate": 4.926363194314344e-05,
|
| 295 |
+
"loss": 0.821,
|
| 296 |
+
"step": 480
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.16,
|
| 300 |
+
"learning_rate": 4.923252804705068e-05,
|
| 301 |
+
"loss": 0.8271,
|
| 302 |
+
"step": 490
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.16,
|
| 306 |
+
"learning_rate": 4.920079099250099e-05,
|
| 307 |
+
"loss": 0.8178,
|
| 308 |
+
"step": 500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.17,
|
| 312 |
+
"learning_rate": 4.916842160873451e-05,
|
| 313 |
+
"loss": 0.8293,
|
| 314 |
+
"step": 510
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.17,
|
| 318 |
+
"learning_rate": 4.913542074151315e-05,
|
| 319 |
+
"loss": 0.8294,
|
| 320 |
+
"step": 520
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.17,
|
| 324 |
+
"learning_rate": 4.910178925309849e-05,
|
| 325 |
+
"loss": 0.8202,
|
| 326 |
+
"step": 530
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.18,
|
| 330 |
+
"learning_rate": 4.906752802222928e-05,
|
| 331 |
+
"loss": 0.8301,
|
| 332 |
+
"step": 540
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.18,
|
| 336 |
+
"learning_rate": 4.903263794409846e-05,
|
| 337 |
+
"loss": 0.8031,
|
| 338 |
+
"step": 550
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.18,
|
| 342 |
+
"learning_rate": 4.8997119930329785e-05,
|
| 343 |
+
"loss": 0.821,
|
| 344 |
+
"step": 560
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.19,
|
| 348 |
+
"learning_rate": 4.896097490895398e-05,
|
| 349 |
+
"loss": 0.8132,
|
| 350 |
+
"step": 570
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.19,
|
| 354 |
+
"learning_rate": 4.8924203824384516e-05,
|
| 355 |
+
"loss": 0.8113,
|
| 356 |
+
"step": 580
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.19,
|
| 360 |
+
"learning_rate": 4.888680763739294e-05,
|
| 361 |
+
"loss": 0.8199,
|
| 362 |
+
"step": 590
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.2,
|
| 366 |
+
"learning_rate": 4.8848787325083734e-05,
|
| 367 |
+
"loss": 0.8196,
|
| 368 |
+
"step": 600
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.2,
|
| 372 |
+
"learning_rate": 4.8810143880868816e-05,
|
| 373 |
+
"loss": 0.8128,
|
| 374 |
+
"step": 610
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.2,
|
| 378 |
+
"learning_rate": 4.8770878314441606e-05,
|
| 379 |
+
"loss": 0.8123,
|
| 380 |
+
"step": 620
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.2,
|
| 384 |
+
"learning_rate": 4.8730991651750565e-05,
|
| 385 |
+
"loss": 0.8048,
|
| 386 |
+
"step": 630
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.21,
|
| 390 |
+
"learning_rate": 4.869048493497247e-05,
|
| 391 |
+
"loss": 0.8119,
|
| 392 |
+
"step": 640
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.21,
|
| 396 |
+
"learning_rate": 4.864935922248515e-05,
|
| 397 |
+
"loss": 0.8141,
|
| 398 |
+
"step": 650
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.21,
|
| 402 |
+
"learning_rate": 4.860761558883981e-05,
|
| 403 |
+
"loss": 0.8,
|
| 404 |
+
"step": 660
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.22,
|
| 408 |
+
"learning_rate": 4.856525512473302e-05,
|
| 409 |
+
"loss": 0.816,
|
| 410 |
+
"step": 670
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.22,
|
| 414 |
+
"learning_rate": 4.852227893697814e-05,
|
| 415 |
+
"loss": 0.7976,
|
| 416 |
+
"step": 680
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.22,
|
| 420 |
+
"learning_rate": 4.847868814847646e-05,
|
| 421 |
+
"loss": 0.8121,
|
| 422 |
+
"step": 690
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.23,
|
| 426 |
+
"learning_rate": 4.84344838981878e-05,
|
| 427 |
+
"loss": 0.8175,
|
| 428 |
+
"step": 700
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.23,
|
| 432 |
+
"learning_rate": 4.8389667341100844e-05,
|
| 433 |
+
"loss": 0.8109,
|
| 434 |
+
"step": 710
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.23,
|
| 438 |
+
"learning_rate": 4.834423964820284e-05,
|
| 439 |
+
"loss": 0.7948,
|
| 440 |
+
"step": 720
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.24,
|
| 444 |
+
"learning_rate": 4.8298202006449136e-05,
|
| 445 |
+
"loss": 0.7986,
|
| 446 |
+
"step": 730
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.24,
|
| 450 |
+
"learning_rate": 4.8251555618732036e-05,
|
| 451 |
+
"loss": 0.8049,
|
| 452 |
+
"step": 740
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.24,
|
| 456 |
+
"learning_rate": 4.820430170384947e-05,
|
| 457 |
+
"loss": 0.7981,
|
| 458 |
+
"step": 750
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.25,
|
| 462 |
+
"learning_rate": 4.8156441496473126e-05,
|
| 463 |
+
"loss": 0.799,
|
| 464 |
+
"step": 760
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.25,
|
| 468 |
+
"learning_rate": 4.8107976247116145e-05,
|
| 469 |
+
"loss": 0.804,
|
| 470 |
+
"step": 770
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.25,
|
| 474 |
+
"learning_rate": 4.805890722210048e-05,
|
| 475 |
+
"loss": 0.8053,
|
| 476 |
+
"step": 780
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.26,
|
| 480 |
+
"learning_rate": 4.8009235703523824e-05,
|
| 481 |
+
"loss": 0.7969,
|
| 482 |
+
"step": 790
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.26,
|
| 486 |
+
"learning_rate": 4.79589629892261e-05,
|
| 487 |
+
"loss": 0.8051,
|
| 488 |
+
"step": 800
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.26,
|
| 492 |
+
"learning_rate": 4.7908090392755526e-05,
|
| 493 |
+
"loss": 0.796,
|
| 494 |
+
"step": 810
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.27,
|
| 498 |
+
"learning_rate": 4.785661924333431e-05,
|
| 499 |
+
"loss": 0.7928,
|
| 500 |
+
"step": 820
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.27,
|
| 504 |
+
"learning_rate": 4.780455088582394e-05,
|
| 505 |
+
"loss": 0.8017,
|
| 506 |
+
"step": 830
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.27,
|
| 510 |
+
"learning_rate": 4.775188668069002e-05,
|
| 511 |
+
"loss": 0.7969,
|
| 512 |
+
"step": 840
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.28,
|
| 516 |
+
"learning_rate": 4.769862800396674e-05,
|
| 517 |
+
"loss": 0.7942,
|
| 518 |
+
"step": 850
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.28,
|
| 522 |
+
"learning_rate": 4.764477624722088e-05,
|
| 523 |
+
"loss": 0.7877,
|
| 524 |
+
"step": 860
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.28,
|
| 528 |
+
"learning_rate": 4.759033281751553e-05,
|
| 529 |
+
"loss": 0.7979,
|
| 530 |
+
"step": 870
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.29,
|
| 534 |
+
"learning_rate": 4.7535299137373236e-05,
|
| 535 |
+
"loss": 0.7986,
|
| 536 |
+
"step": 880
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.29,
|
| 540 |
+
"learning_rate": 4.7479676644738894e-05,
|
| 541 |
+
"loss": 0.8124,
|
| 542 |
+
"step": 890
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.29,
|
| 546 |
+
"learning_rate": 4.742346679294217e-05,
|
| 547 |
+
"loss": 0.7959,
|
| 548 |
+
"step": 900
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.3,
|
| 552 |
+
"learning_rate": 4.7366671050659505e-05,
|
| 553 |
+
"loss": 0.7884,
|
| 554 |
+
"step": 910
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.3,
|
| 558 |
+
"learning_rate": 4.730929090187573e-05,
|
| 559 |
+
"loss": 0.7911,
|
| 560 |
+
"step": 920
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.3,
|
| 564 |
+
"learning_rate": 4.725132784584536e-05,
|
| 565 |
+
"loss": 0.7937,
|
| 566 |
+
"step": 930
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.31,
|
| 570 |
+
"learning_rate": 4.719278339705333e-05,
|
| 571 |
+
"loss": 0.7994,
|
| 572 |
+
"step": 940
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.31,
|
| 576 |
+
"learning_rate": 4.71336590851755e-05,
|
| 577 |
+
"loss": 0.7905,
|
| 578 |
+
"step": 950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.31,
|
| 582 |
+
"learning_rate": 4.707395645503863e-05,
|
| 583 |
+
"loss": 0.7924,
|
| 584 |
+
"step": 960
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.32,
|
| 588 |
+
"learning_rate": 4.701367706658006e-05,
|
| 589 |
+
"loss": 0.7891,
|
| 590 |
+
"step": 970
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.32,
|
| 594 |
+
"learning_rate": 4.695282249480692e-05,
|
| 595 |
+
"loss": 0.7962,
|
| 596 |
+
"step": 980
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.32,
|
| 600 |
+
"learning_rate": 4.6891394329755e-05,
|
| 601 |
+
"loss": 0.7882,
|
| 602 |
+
"step": 990
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.33,
|
| 606 |
+
"learning_rate": 4.6829394176447195e-05,
|
| 607 |
+
"loss": 0.8011,
|
| 608 |
+
"step": 1000
|
| 609 |
+
}
|
| 610 |
+
],
|
| 611 |
+
"max_steps": 6146,
|
| 612 |
+
"num_train_epochs": 2,
|
| 613 |
+
"total_flos": 1.7954271898653164e+18,
|
| 614 |
+
"trial_name": null,
|
| 615 |
+
"trial_params": null
|
| 616 |
+
}
|
checkpoint-1000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dafdcf787495b020f1b3f6f6180b816b2753ea01048ce4086dbcd8a2c7a5f666
|
| 3 |
+
size 3265
|
checkpoint-2000/README.md
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
| 4 |
+
## Training procedure
|
| 5 |
+
|
| 6 |
+
|
| 7 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 8 |
+
- load_in_8bit: False
|
| 9 |
+
- load_in_4bit: True
|
| 10 |
+
- llm_int8_threshold: 6.0
|
| 11 |
+
- llm_int8_skip_modules: None
|
| 12 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 13 |
+
- llm_int8_has_fp16_weight: False
|
| 14 |
+
- bnb_4bit_quant_type: nf4
|
| 15 |
+
- bnb_4bit_use_double_quant: True
|
| 16 |
+
- bnb_4bit_compute_dtype: float16
|
| 17 |
+
### Framework versions
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
- PEFT 0.4.0
|
checkpoint-2000/adapter_config.json
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "./Llama-2-7b-chat-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"fan_in_fan_out": false,
|
| 6 |
+
"inference_mode": true,
|
| 7 |
+
"init_lora_weights": true,
|
| 8 |
+
"layers_pattern": null,
|
| 9 |
+
"layers_to_transform": null,
|
| 10 |
+
"lora_alpha": 32.0,
|
| 11 |
+
"lora_dropout": 0.1,
|
| 12 |
+
"modules_to_save": null,
|
| 13 |
+
"peft_type": "LORA",
|
| 14 |
+
"r": 8,
|
| 15 |
+
"revision": null,
|
| 16 |
+
"target_modules": [
|
| 17 |
+
"q_proj",
|
| 18 |
+
"v_proj"
|
| 19 |
+
],
|
| 20 |
+
"task_type": "CAUSAL_LM"
|
| 21 |
+
}
|
checkpoint-2000/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:38858458726a29aa27bd28fb767068b364aa849dec17d53de44459195794d47e
|
| 3 |
+
size 16821197
|
checkpoint-2000/finetuning_args.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"dpo_beta": 0.1,
|
| 3 |
+
"finetuning_type": "lora",
|
| 4 |
+
"lora_alpha": 32.0,
|
| 5 |
+
"lora_dropout": 0.1,
|
| 6 |
+
"lora_rank": 8,
|
| 7 |
+
"lora_target": [
|
| 8 |
+
"q_proj",
|
| 9 |
+
"v_proj"
|
| 10 |
+
],
|
| 11 |
+
"name_module_trainable": "mlp",
|
| 12 |
+
"num_hidden_layers": 32,
|
| 13 |
+
"num_layer_trainable": 3,
|
| 14 |
+
"ppo_score_norm": false,
|
| 15 |
+
"resume_lora_training": true
|
| 16 |
+
}
|
checkpoint-2000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:60448907929e2cce0be65ce1ce64291db486f251c3ce16181af73dbd56c4a888
|
| 3 |
+
size 33661637
|
checkpoint-2000/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7971cd06a53ab69433f290ec5afc5154601187e8be93b0ae873f493feabe6a7a
|
| 3 |
+
size 18679
|
checkpoint-2000/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:778e36af2f37a70f0a5cceeb1a49f58a443401f5197aa27d66864aea15c01840
|
| 3 |
+
size 18679
|
checkpoint-2000/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8919286e2f858a907cad2da6afa9792c4058e2bc8d3d4fd04899fd747e1b4265
|
| 3 |
+
size 18679
|
checkpoint-2000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9b8e8c4f25f78d8f230c5a497336e74accb960a2bad8b9f5f78a3895d126426f
|
| 3 |
+
size 627
|
checkpoint-2000/trainer_state.json
ADDED
|
@@ -0,0 +1,1216 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.6506710044733631,
|
| 5 |
+
"global_step": 2000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 4.999967339434433e-05,
|
| 13 |
+
"loss": 1.343,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 4.9998820960283476e-05,
|
| 19 |
+
"loss": 1.1858,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 4.99974394497948e-05,
|
| 25 |
+
"loss": 1.1057,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.01,
|
| 30 |
+
"learning_rate": 4.9995528892116685e-05,
|
| 31 |
+
"loss": 1.0698,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.02,
|
| 36 |
+
"learning_rate": 4.999308932768425e-05,
|
| 37 |
+
"loss": 1.0316,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.02,
|
| 42 |
+
"learning_rate": 4.99897583236926e-05,
|
| 43 |
+
"loss": 1.0046,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.02,
|
| 48 |
+
"learning_rate": 4.9985774375988764e-05,
|
| 49 |
+
"loss": 0.9848,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.03,
|
| 54 |
+
"learning_rate": 4.998113758866712e-05,
|
| 55 |
+
"loss": 0.9656,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.03,
|
| 60 |
+
"learning_rate": 4.997584808287975e-05,
|
| 61 |
+
"loss": 0.9648,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.03,
|
| 66 |
+
"learning_rate": 4.996990599683325e-05,
|
| 67 |
+
"loss": 0.9415,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.04,
|
| 72 |
+
"learning_rate": 4.996331148578514e-05,
|
| 73 |
+
"loss": 0.9388,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.04,
|
| 78 |
+
"learning_rate": 4.995606472203979e-05,
|
| 79 |
+
"loss": 0.9289,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.04,
|
| 84 |
+
"learning_rate": 4.994816589494391e-05,
|
| 85 |
+
"loss": 0.9252,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.05,
|
| 90 |
+
"learning_rate": 4.993961521088163e-05,
|
| 91 |
+
"loss": 0.91,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.05,
|
| 96 |
+
"learning_rate": 4.9930412893269106e-05,
|
| 97 |
+
"loss": 0.9033,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.05,
|
| 102 |
+
"learning_rate": 4.992055918254865e-05,
|
| 103 |
+
"loss": 0.9064,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.06,
|
| 108 |
+
"learning_rate": 4.9910054336182467e-05,
|
| 109 |
+
"loss": 0.8946,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.06,
|
| 114 |
+
"learning_rate": 4.989889862864594e-05,
|
| 115 |
+
"loss": 0.8914,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.06,
|
| 120 |
+
"learning_rate": 4.988709235142044e-05,
|
| 121 |
+
"loss": 0.9007,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.07,
|
| 126 |
+
"learning_rate": 4.9874635812985734e-05,
|
| 127 |
+
"loss": 0.9008,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.07,
|
| 132 |
+
"learning_rate": 4.9861529338811874e-05,
|
| 133 |
+
"loss": 0.8714,
|
| 134 |
+
"step": 210
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.07,
|
| 138 |
+
"learning_rate": 4.984777327135076e-05,
|
| 139 |
+
"loss": 0.8865,
|
| 140 |
+
"step": 220
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.07,
|
| 144 |
+
"learning_rate": 4.9833367970027145e-05,
|
| 145 |
+
"loss": 0.8716,
|
| 146 |
+
"step": 230
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.08,
|
| 150 |
+
"learning_rate": 4.981831381122926e-05,
|
| 151 |
+
"loss": 0.8817,
|
| 152 |
+
"step": 240
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.08,
|
| 156 |
+
"learning_rate": 4.980261118829896e-05,
|
| 157 |
+
"loss": 0.8658,
|
| 158 |
+
"step": 250
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.08,
|
| 162 |
+
"learning_rate": 4.978626051152152e-05,
|
| 163 |
+
"loss": 0.8722,
|
| 164 |
+
"step": 260
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.09,
|
| 168 |
+
"learning_rate": 4.976926220811479e-05,
|
| 169 |
+
"loss": 0.8541,
|
| 170 |
+
"step": 270
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.09,
|
| 174 |
+
"learning_rate": 4.975161672221813e-05,
|
| 175 |
+
"loss": 0.8504,
|
| 176 |
+
"step": 280
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.09,
|
| 180 |
+
"learning_rate": 4.97333245148808e-05,
|
| 181 |
+
"loss": 0.8467,
|
| 182 |
+
"step": 290
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.1,
|
| 186 |
+
"learning_rate": 4.971438606404986e-05,
|
| 187 |
+
"loss": 0.8533,
|
| 188 |
+
"step": 300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.1,
|
| 192 |
+
"learning_rate": 4.969480186455771e-05,
|
| 193 |
+
"loss": 0.853,
|
| 194 |
+
"step": 310
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.1,
|
| 198 |
+
"learning_rate": 4.9674572428109205e-05,
|
| 199 |
+
"loss": 0.8432,
|
| 200 |
+
"step": 320
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.11,
|
| 204 |
+
"learning_rate": 4.9653698283268183e-05,
|
| 205 |
+
"loss": 0.854,
|
| 206 |
+
"step": 330
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.11,
|
| 210 |
+
"learning_rate": 4.963217997544376e-05,
|
| 211 |
+
"loss": 0.8435,
|
| 212 |
+
"step": 340
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.11,
|
| 216 |
+
"learning_rate": 4.961001806687602e-05,
|
| 217 |
+
"loss": 0.8518,
|
| 218 |
+
"step": 350
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.12,
|
| 222 |
+
"learning_rate": 4.958721313662133e-05,
|
| 223 |
+
"loss": 0.848,
|
| 224 |
+
"step": 360
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.12,
|
| 228 |
+
"learning_rate": 4.956376578053723e-05,
|
| 229 |
+
"loss": 0.8436,
|
| 230 |
+
"step": 370
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.12,
|
| 234 |
+
"learning_rate": 4.9539676611266846e-05,
|
| 235 |
+
"loss": 0.8405,
|
| 236 |
+
"step": 380
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.13,
|
| 240 |
+
"learning_rate": 4.9514946258222895e-05,
|
| 241 |
+
"loss": 0.838,
|
| 242 |
+
"step": 390
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.13,
|
| 246 |
+
"learning_rate": 4.948957536757123e-05,
|
| 247 |
+
"loss": 0.8458,
|
| 248 |
+
"step": 400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.13,
|
| 252 |
+
"learning_rate": 4.946356460221395e-05,
|
| 253 |
+
"loss": 0.8379,
|
| 254 |
+
"step": 410
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.14,
|
| 258 |
+
"learning_rate": 4.943691464177212e-05,
|
| 259 |
+
"loss": 0.827,
|
| 260 |
+
"step": 420
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.14,
|
| 264 |
+
"learning_rate": 4.940962618256796e-05,
|
| 265 |
+
"loss": 0.8208,
|
| 266 |
+
"step": 430
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.14,
|
| 270 |
+
"learning_rate": 4.9381699937606674e-05,
|
| 271 |
+
"loss": 0.8242,
|
| 272 |
+
"step": 440
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.15,
|
| 276 |
+
"learning_rate": 4.935313663655782e-05,
|
| 277 |
+
"loss": 0.8362,
|
| 278 |
+
"step": 450
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.15,
|
| 282 |
+
"learning_rate": 4.932393702573627e-05,
|
| 283 |
+
"loss": 0.8363,
|
| 284 |
+
"step": 460
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.15,
|
| 288 |
+
"learning_rate": 4.929410186808264e-05,
|
| 289 |
+
"loss": 0.8275,
|
| 290 |
+
"step": 470
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.16,
|
| 294 |
+
"learning_rate": 4.926363194314344e-05,
|
| 295 |
+
"loss": 0.821,
|
| 296 |
+
"step": 480
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.16,
|
| 300 |
+
"learning_rate": 4.923252804705068e-05,
|
| 301 |
+
"loss": 0.8271,
|
| 302 |
+
"step": 490
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.16,
|
| 306 |
+
"learning_rate": 4.920079099250099e-05,
|
| 307 |
+
"loss": 0.8178,
|
| 308 |
+
"step": 500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.17,
|
| 312 |
+
"learning_rate": 4.916842160873451e-05,
|
| 313 |
+
"loss": 0.8293,
|
| 314 |
+
"step": 510
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.17,
|
| 318 |
+
"learning_rate": 4.913542074151315e-05,
|
| 319 |
+
"loss": 0.8294,
|
| 320 |
+
"step": 520
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.17,
|
| 324 |
+
"learning_rate": 4.910178925309849e-05,
|
| 325 |
+
"loss": 0.8202,
|
| 326 |
+
"step": 530
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.18,
|
| 330 |
+
"learning_rate": 4.906752802222928e-05,
|
| 331 |
+
"loss": 0.8301,
|
| 332 |
+
"step": 540
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.18,
|
| 336 |
+
"learning_rate": 4.903263794409846e-05,
|
| 337 |
+
"loss": 0.8031,
|
| 338 |
+
"step": 550
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.18,
|
| 342 |
+
"learning_rate": 4.8997119930329785e-05,
|
| 343 |
+
"loss": 0.821,
|
| 344 |
+
"step": 560
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.19,
|
| 348 |
+
"learning_rate": 4.896097490895398e-05,
|
| 349 |
+
"loss": 0.8132,
|
| 350 |
+
"step": 570
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.19,
|
| 354 |
+
"learning_rate": 4.8924203824384516e-05,
|
| 355 |
+
"loss": 0.8113,
|
| 356 |
+
"step": 580
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.19,
|
| 360 |
+
"learning_rate": 4.888680763739294e-05,
|
| 361 |
+
"loss": 0.8199,
|
| 362 |
+
"step": 590
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.2,
|
| 366 |
+
"learning_rate": 4.8848787325083734e-05,
|
| 367 |
+
"loss": 0.8196,
|
| 368 |
+
"step": 600
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.2,
|
| 372 |
+
"learning_rate": 4.8810143880868816e-05,
|
| 373 |
+
"loss": 0.8128,
|
| 374 |
+
"step": 610
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.2,
|
| 378 |
+
"learning_rate": 4.8770878314441606e-05,
|
| 379 |
+
"loss": 0.8123,
|
| 380 |
+
"step": 620
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.2,
|
| 384 |
+
"learning_rate": 4.8730991651750565e-05,
|
| 385 |
+
"loss": 0.8048,
|
| 386 |
+
"step": 630
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.21,
|
| 390 |
+
"learning_rate": 4.869048493497247e-05,
|
| 391 |
+
"loss": 0.8119,
|
| 392 |
+
"step": 640
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.21,
|
| 396 |
+
"learning_rate": 4.864935922248515e-05,
|
| 397 |
+
"loss": 0.8141,
|
| 398 |
+
"step": 650
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.21,
|
| 402 |
+
"learning_rate": 4.860761558883981e-05,
|
| 403 |
+
"loss": 0.8,
|
| 404 |
+
"step": 660
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.22,
|
| 408 |
+
"learning_rate": 4.856525512473302e-05,
|
| 409 |
+
"loss": 0.816,
|
| 410 |
+
"step": 670
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.22,
|
| 414 |
+
"learning_rate": 4.852227893697814e-05,
|
| 415 |
+
"loss": 0.7976,
|
| 416 |
+
"step": 680
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.22,
|
| 420 |
+
"learning_rate": 4.847868814847646e-05,
|
| 421 |
+
"loss": 0.8121,
|
| 422 |
+
"step": 690
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.23,
|
| 426 |
+
"learning_rate": 4.84344838981878e-05,
|
| 427 |
+
"loss": 0.8175,
|
| 428 |
+
"step": 700
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.23,
|
| 432 |
+
"learning_rate": 4.8389667341100844e-05,
|
| 433 |
+
"loss": 0.8109,
|
| 434 |
+
"step": 710
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.23,
|
| 438 |
+
"learning_rate": 4.834423964820284e-05,
|
| 439 |
+
"loss": 0.7948,
|
| 440 |
+
"step": 720
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.24,
|
| 444 |
+
"learning_rate": 4.8298202006449136e-05,
|
| 445 |
+
"loss": 0.7986,
|
| 446 |
+
"step": 730
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.24,
|
| 450 |
+
"learning_rate": 4.8251555618732036e-05,
|
| 451 |
+
"loss": 0.8049,
|
| 452 |
+
"step": 740
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.24,
|
| 456 |
+
"learning_rate": 4.820430170384947e-05,
|
| 457 |
+
"loss": 0.7981,
|
| 458 |
+
"step": 750
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.25,
|
| 462 |
+
"learning_rate": 4.8156441496473126e-05,
|
| 463 |
+
"loss": 0.799,
|
| 464 |
+
"step": 760
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.25,
|
| 468 |
+
"learning_rate": 4.8107976247116145e-05,
|
| 469 |
+
"loss": 0.804,
|
| 470 |
+
"step": 770
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.25,
|
| 474 |
+
"learning_rate": 4.805890722210048e-05,
|
| 475 |
+
"loss": 0.8053,
|
| 476 |
+
"step": 780
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.26,
|
| 480 |
+
"learning_rate": 4.8009235703523824e-05,
|
| 481 |
+
"loss": 0.7969,
|
| 482 |
+
"step": 790
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.26,
|
| 486 |
+
"learning_rate": 4.79589629892261e-05,
|
| 487 |
+
"loss": 0.8051,
|
| 488 |
+
"step": 800
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.26,
|
| 492 |
+
"learning_rate": 4.7908090392755526e-05,
|
| 493 |
+
"loss": 0.796,
|
| 494 |
+
"step": 810
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.27,
|
| 498 |
+
"learning_rate": 4.785661924333431e-05,
|
| 499 |
+
"loss": 0.7928,
|
| 500 |
+
"step": 820
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.27,
|
| 504 |
+
"learning_rate": 4.780455088582394e-05,
|
| 505 |
+
"loss": 0.8017,
|
| 506 |
+
"step": 830
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.27,
|
| 510 |
+
"learning_rate": 4.775188668069002e-05,
|
| 511 |
+
"loss": 0.7969,
|
| 512 |
+
"step": 840
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.28,
|
| 516 |
+
"learning_rate": 4.769862800396674e-05,
|
| 517 |
+
"loss": 0.7942,
|
| 518 |
+
"step": 850
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.28,
|
| 522 |
+
"learning_rate": 4.764477624722088e-05,
|
| 523 |
+
"loss": 0.7877,
|
| 524 |
+
"step": 860
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.28,
|
| 528 |
+
"learning_rate": 4.759033281751553e-05,
|
| 529 |
+
"loss": 0.7979,
|
| 530 |
+
"step": 870
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.29,
|
| 534 |
+
"learning_rate": 4.7535299137373236e-05,
|
| 535 |
+
"loss": 0.7986,
|
| 536 |
+
"step": 880
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.29,
|
| 540 |
+
"learning_rate": 4.7479676644738894e-05,
|
| 541 |
+
"loss": 0.8124,
|
| 542 |
+
"step": 890
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.29,
|
| 546 |
+
"learning_rate": 4.742346679294217e-05,
|
| 547 |
+
"loss": 0.7959,
|
| 548 |
+
"step": 900
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.3,
|
| 552 |
+
"learning_rate": 4.7366671050659505e-05,
|
| 553 |
+
"loss": 0.7884,
|
| 554 |
+
"step": 910
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.3,
|
| 558 |
+
"learning_rate": 4.730929090187573e-05,
|
| 559 |
+
"loss": 0.7911,
|
| 560 |
+
"step": 920
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.3,
|
| 564 |
+
"learning_rate": 4.725132784584536e-05,
|
| 565 |
+
"loss": 0.7937,
|
| 566 |
+
"step": 930
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.31,
|
| 570 |
+
"learning_rate": 4.719278339705333e-05,
|
| 571 |
+
"loss": 0.7994,
|
| 572 |
+
"step": 940
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.31,
|
| 576 |
+
"learning_rate": 4.71336590851755e-05,
|
| 577 |
+
"loss": 0.7905,
|
| 578 |
+
"step": 950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.31,
|
| 582 |
+
"learning_rate": 4.707395645503863e-05,
|
| 583 |
+
"loss": 0.7924,
|
| 584 |
+
"step": 960
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.32,
|
| 588 |
+
"learning_rate": 4.701367706658006e-05,
|
| 589 |
+
"loss": 0.7891,
|
| 590 |
+
"step": 970
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.32,
|
| 594 |
+
"learning_rate": 4.695282249480692e-05,
|
| 595 |
+
"loss": 0.7962,
|
| 596 |
+
"step": 980
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.32,
|
| 600 |
+
"learning_rate": 4.6891394329755e-05,
|
| 601 |
+
"loss": 0.7882,
|
| 602 |
+
"step": 990
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.33,
|
| 606 |
+
"learning_rate": 4.6829394176447195e-05,
|
| 607 |
+
"loss": 0.8011,
|
| 608 |
+
"step": 1000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.33,
|
| 612 |
+
"learning_rate": 4.676682365485154e-05,
|
| 613 |
+
"loss": 0.7955,
|
| 614 |
+
"step": 1010
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.33,
|
| 618 |
+
"learning_rate": 4.670368439983896e-05,
|
| 619 |
+
"loss": 0.7906,
|
| 620 |
+
"step": 1020
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.34,
|
| 624 |
+
"learning_rate": 4.663997806114048e-05,
|
| 625 |
+
"loss": 0.7853,
|
| 626 |
+
"step": 1030
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.34,
|
| 630 |
+
"learning_rate": 4.657570630330411e-05,
|
| 631 |
+
"loss": 0.7825,
|
| 632 |
+
"step": 1040
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.34,
|
| 636 |
+
"learning_rate": 4.6510870805651444e-05,
|
| 637 |
+
"loss": 0.7768,
|
| 638 |
+
"step": 1050
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.34,
|
| 642 |
+
"learning_rate": 4.6445473262233696e-05,
|
| 643 |
+
"loss": 0.7795,
|
| 644 |
+
"step": 1060
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.35,
|
| 648 |
+
"learning_rate": 4.637951538178746e-05,
|
| 649 |
+
"loss": 0.7832,
|
| 650 |
+
"step": 1070
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.35,
|
| 654 |
+
"learning_rate": 4.63129988876901e-05,
|
| 655 |
+
"loss": 0.7934,
|
| 656 |
+
"step": 1080
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.35,
|
| 660 |
+
"learning_rate": 4.624592551791465e-05,
|
| 661 |
+
"loss": 0.7892,
|
| 662 |
+
"step": 1090
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.36,
|
| 666 |
+
"learning_rate": 4.617829702498446e-05,
|
| 667 |
+
"loss": 0.7819,
|
| 668 |
+
"step": 1100
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.36,
|
| 672 |
+
"learning_rate": 4.611011517592741e-05,
|
| 673 |
+
"loss": 0.7868,
|
| 674 |
+
"step": 1110
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.36,
|
| 678 |
+
"learning_rate": 4.604138175222968e-05,
|
| 679 |
+
"loss": 0.7914,
|
| 680 |
+
"step": 1120
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.37,
|
| 684 |
+
"learning_rate": 4.597209854978929e-05,
|
| 685 |
+
"loss": 0.7817,
|
| 686 |
+
"step": 1130
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 0.37,
|
| 690 |
+
"learning_rate": 4.590226737886908e-05,
|
| 691 |
+
"loss": 0.7916,
|
| 692 |
+
"step": 1140
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.37,
|
| 696 |
+
"learning_rate": 4.583189006404949e-05,
|
| 697 |
+
"loss": 0.7799,
|
| 698 |
+
"step": 1150
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.38,
|
| 702 |
+
"learning_rate": 4.576096844418083e-05,
|
| 703 |
+
"loss": 0.7776,
|
| 704 |
+
"step": 1160
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.38,
|
| 708 |
+
"learning_rate": 4.568950437233529e-05,
|
| 709 |
+
"loss": 0.7803,
|
| 710 |
+
"step": 1170
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.38,
|
| 714 |
+
"learning_rate": 4.561749971575846e-05,
|
| 715 |
+
"loss": 0.7884,
|
| 716 |
+
"step": 1180
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.39,
|
| 720 |
+
"learning_rate": 4.5544956355820596e-05,
|
| 721 |
+
"loss": 0.7827,
|
| 722 |
+
"step": 1190
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.39,
|
| 726 |
+
"learning_rate": 4.547187618796742e-05,
|
| 727 |
+
"loss": 0.7857,
|
| 728 |
+
"step": 1200
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 0.39,
|
| 732 |
+
"learning_rate": 4.5398261121670634e-05,
|
| 733 |
+
"loss": 0.7797,
|
| 734 |
+
"step": 1210
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.4,
|
| 738 |
+
"learning_rate": 4.532411308037798e-05,
|
| 739 |
+
"loss": 0.7821,
|
| 740 |
+
"step": 1220
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.4,
|
| 744 |
+
"learning_rate": 4.524943400146306e-05,
|
| 745 |
+
"loss": 0.7747,
|
| 746 |
+
"step": 1230
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.4,
|
| 750 |
+
"learning_rate": 4.51742258361746e-05,
|
| 751 |
+
"loss": 0.7737,
|
| 752 |
+
"step": 1240
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.41,
|
| 756 |
+
"learning_rate": 4.509849054958559e-05,
|
| 757 |
+
"loss": 0.7868,
|
| 758 |
+
"step": 1250
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.41,
|
| 762 |
+
"learning_rate": 4.502223012054186e-05,
|
| 763 |
+
"loss": 0.7818,
|
| 764 |
+
"step": 1260
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.41,
|
| 768 |
+
"learning_rate": 4.4945446541610414e-05,
|
| 769 |
+
"loss": 0.7804,
|
| 770 |
+
"step": 1270
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 0.42,
|
| 774 |
+
"learning_rate": 4.486814181902734e-05,
|
| 775 |
+
"loss": 0.7759,
|
| 776 |
+
"step": 1280
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.42,
|
| 780 |
+
"learning_rate": 4.4790317972645396e-05,
|
| 781 |
+
"loss": 0.7744,
|
| 782 |
+
"step": 1290
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.42,
|
| 786 |
+
"learning_rate": 4.4711977035881256e-05,
|
| 787 |
+
"loss": 0.7814,
|
| 788 |
+
"step": 1300
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.43,
|
| 792 |
+
"learning_rate": 4.463312105566237e-05,
|
| 793 |
+
"loss": 0.7611,
|
| 794 |
+
"step": 1310
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.43,
|
| 798 |
+
"learning_rate": 4.4553752092373465e-05,
|
| 799 |
+
"loss": 0.7796,
|
| 800 |
+
"step": 1320
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.43,
|
| 804 |
+
"learning_rate": 4.447387221980272e-05,
|
| 805 |
+
"loss": 0.7839,
|
| 806 |
+
"step": 1330
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.44,
|
| 810 |
+
"learning_rate": 4.4393483525087584e-05,
|
| 811 |
+
"loss": 0.7726,
|
| 812 |
+
"step": 1340
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.44,
|
| 816 |
+
"learning_rate": 4.4312588108660246e-05,
|
| 817 |
+
"loss": 0.7763,
|
| 818 |
+
"step": 1350
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.44,
|
| 822 |
+
"learning_rate": 4.423118808419276e-05,
|
| 823 |
+
"loss": 0.766,
|
| 824 |
+
"step": 1360
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.45,
|
| 828 |
+
"learning_rate": 4.4149285578541785e-05,
|
| 829 |
+
"loss": 0.7887,
|
| 830 |
+
"step": 1370
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.45,
|
| 834 |
+
"learning_rate": 4.406688273169305e-05,
|
| 835 |
+
"loss": 0.7637,
|
| 836 |
+
"step": 1380
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.45,
|
| 840 |
+
"learning_rate": 4.3983981696705415e-05,
|
| 841 |
+
"loss": 0.7609,
|
| 842 |
+
"step": 1390
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.46,
|
| 846 |
+
"learning_rate": 4.390058463965464e-05,
|
| 847 |
+
"loss": 0.7729,
|
| 848 |
+
"step": 1400
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.46,
|
| 852 |
+
"learning_rate": 4.381669373957675e-05,
|
| 853 |
+
"loss": 0.7751,
|
| 854 |
+
"step": 1410
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.46,
|
| 858 |
+
"learning_rate": 4.3732311188411154e-05,
|
| 859 |
+
"loss": 0.7877,
|
| 860 |
+
"step": 1420
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.47,
|
| 864 |
+
"learning_rate": 4.3647439190943315e-05,
|
| 865 |
+
"loss": 0.7738,
|
| 866 |
+
"step": 1430
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.47,
|
| 870 |
+
"learning_rate": 4.3562079964747207e-05,
|
| 871 |
+
"loss": 0.7834,
|
| 872 |
+
"step": 1440
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.47,
|
| 876 |
+
"learning_rate": 4.347623574012728e-05,
|
| 877 |
+
"loss": 0.7835,
|
| 878 |
+
"step": 1450
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.47,
|
| 882 |
+
"learning_rate": 4.338990876006031e-05,
|
| 883 |
+
"loss": 0.7707,
|
| 884 |
+
"step": 1460
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.48,
|
| 888 |
+
"learning_rate": 4.330310128013666e-05,
|
| 889 |
+
"loss": 0.7674,
|
| 890 |
+
"step": 1470
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.48,
|
| 894 |
+
"learning_rate": 4.321581556850146e-05,
|
| 895 |
+
"loss": 0.7687,
|
| 896 |
+
"step": 1480
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.48,
|
| 900 |
+
"learning_rate": 4.312805390579527e-05,
|
| 901 |
+
"loss": 0.7663,
|
| 902 |
+
"step": 1490
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.49,
|
| 906 |
+
"learning_rate": 4.3039818585094524e-05,
|
| 907 |
+
"loss": 0.7662,
|
| 908 |
+
"step": 1500
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.49,
|
| 912 |
+
"learning_rate": 4.29511119118516e-05,
|
| 913 |
+
"loss": 0.7688,
|
| 914 |
+
"step": 1510
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.49,
|
| 918 |
+
"learning_rate": 4.28619362038346e-05,
|
| 919 |
+
"loss": 0.7652,
|
| 920 |
+
"step": 1520
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.5,
|
| 924 |
+
"learning_rate": 4.277229379106676e-05,
|
| 925 |
+
"loss": 0.7683,
|
| 926 |
+
"step": 1530
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.5,
|
| 930 |
+
"learning_rate": 4.2682187015765615e-05,
|
| 931 |
+
"loss": 0.7621,
|
| 932 |
+
"step": 1540
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.5,
|
| 936 |
+
"learning_rate": 4.259161823228174e-05,
|
| 937 |
+
"loss": 0.7778,
|
| 938 |
+
"step": 1550
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 0.51,
|
| 942 |
+
"learning_rate": 4.2500589807037306e-05,
|
| 943 |
+
"loss": 0.7809,
|
| 944 |
+
"step": 1560
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 0.51,
|
| 948 |
+
"learning_rate": 4.240910411846418e-05,
|
| 949 |
+
"loss": 0.7717,
|
| 950 |
+
"step": 1570
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.51,
|
| 954 |
+
"learning_rate": 4.231716355694184e-05,
|
| 955 |
+
"loss": 0.7584,
|
| 956 |
+
"step": 1580
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.52,
|
| 960 |
+
"learning_rate": 4.222477052473486e-05,
|
| 961 |
+
"loss": 0.7631,
|
| 962 |
+
"step": 1590
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.52,
|
| 966 |
+
"learning_rate": 4.21319274359302e-05,
|
| 967 |
+
"loss": 0.7681,
|
| 968 |
+
"step": 1600
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.52,
|
| 972 |
+
"learning_rate": 4.203863671637409e-05,
|
| 973 |
+
"loss": 0.7739,
|
| 974 |
+
"step": 1610
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 0.53,
|
| 978 |
+
"learning_rate": 4.194490080360866e-05,
|
| 979 |
+
"loss": 0.7646,
|
| 980 |
+
"step": 1620
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.53,
|
| 984 |
+
"learning_rate": 4.1850722146808254e-05,
|
| 985 |
+
"loss": 0.7647,
|
| 986 |
+
"step": 1630
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.53,
|
| 990 |
+
"learning_rate": 4.1756103206715416e-05,
|
| 991 |
+
"loss": 0.7675,
|
| 992 |
+
"step": 1640
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.54,
|
| 996 |
+
"learning_rate": 4.166104645557662e-05,
|
| 997 |
+
"loss": 0.7652,
|
| 998 |
+
"step": 1650
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 0.54,
|
| 1002 |
+
"learning_rate": 4.15655543770777e-05,
|
| 1003 |
+
"loss": 0.765,
|
| 1004 |
+
"step": 1660
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.54,
|
| 1008 |
+
"learning_rate": 4.146962946627886e-05,
|
| 1009 |
+
"loss": 0.7614,
|
| 1010 |
+
"step": 1670
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.55,
|
| 1014 |
+
"learning_rate": 4.137327422954958e-05,
|
| 1015 |
+
"loss": 0.7544,
|
| 1016 |
+
"step": 1680
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 0.55,
|
| 1020 |
+
"learning_rate": 4.127649118450308e-05,
|
| 1021 |
+
"loss": 0.7673,
|
| 1022 |
+
"step": 1690
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 0.55,
|
| 1026 |
+
"learning_rate": 4.117928285993055e-05,
|
| 1027 |
+
"loss": 0.7593,
|
| 1028 |
+
"step": 1700
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 0.56,
|
| 1032 |
+
"learning_rate": 4.108165179573508e-05,
|
| 1033 |
+
"loss": 0.7735,
|
| 1034 |
+
"step": 1710
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.56,
|
| 1038 |
+
"learning_rate": 4.0983600542865285e-05,
|
| 1039 |
+
"loss": 0.7666,
|
| 1040 |
+
"step": 1720
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.56,
|
| 1044 |
+
"learning_rate": 4.088513166324867e-05,
|
| 1045 |
+
"loss": 0.7627,
|
| 1046 |
+
"step": 1730
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.57,
|
| 1050 |
+
"learning_rate": 4.078624772972467e-05,
|
| 1051 |
+
"loss": 0.77,
|
| 1052 |
+
"step": 1740
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.57,
|
| 1056 |
+
"learning_rate": 4.068695132597744e-05,
|
| 1057 |
+
"loss": 0.7569,
|
| 1058 |
+
"step": 1750
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 0.57,
|
| 1062 |
+
"learning_rate": 4.058724504646834e-05,
|
| 1063 |
+
"loss": 0.7654,
|
| 1064 |
+
"step": 1760
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 0.58,
|
| 1068 |
+
"learning_rate": 4.048713149636816e-05,
|
| 1069 |
+
"loss": 0.753,
|
| 1070 |
+
"step": 1770
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.58,
|
| 1074 |
+
"learning_rate": 4.038661329148904e-05,
|
| 1075 |
+
"loss": 0.7705,
|
| 1076 |
+
"step": 1780
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.58,
|
| 1080 |
+
"learning_rate": 4.02856930582161e-05,
|
| 1081 |
+
"loss": 0.7542,
|
| 1082 |
+
"step": 1790
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.59,
|
| 1086 |
+
"learning_rate": 4.018437343343886e-05,
|
| 1087 |
+
"loss": 0.7532,
|
| 1088 |
+
"step": 1800
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.59,
|
| 1092 |
+
"learning_rate": 4.008265706448234e-05,
|
| 1093 |
+
"loss": 0.7662,
|
| 1094 |
+
"step": 1810
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.59,
|
| 1098 |
+
"learning_rate": 3.9980546609037825e-05,
|
| 1099 |
+
"loss": 0.7607,
|
| 1100 |
+
"step": 1820
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 0.6,
|
| 1104 |
+
"learning_rate": 3.9878044735093505e-05,
|
| 1105 |
+
"loss": 0.7552,
|
| 1106 |
+
"step": 1830
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 0.6,
|
| 1110 |
+
"learning_rate": 3.977515412086472e-05,
|
| 1111 |
+
"loss": 0.7645,
|
| 1112 |
+
"step": 1840
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.6,
|
| 1116 |
+
"learning_rate": 3.9671877454723985e-05,
|
| 1117 |
+
"loss": 0.7689,
|
| 1118 |
+
"step": 1850
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.61,
|
| 1122 |
+
"learning_rate": 3.956821743513078e-05,
|
| 1123 |
+
"loss": 0.7615,
|
| 1124 |
+
"step": 1860
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.61,
|
| 1128 |
+
"learning_rate": 3.946417677056097e-05,
|
| 1129 |
+
"loss": 0.7703,
|
| 1130 |
+
"step": 1870
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.61,
|
| 1134 |
+
"learning_rate": 3.935975817943614e-05,
|
| 1135 |
+
"loss": 0.766,
|
| 1136 |
+
"step": 1880
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.61,
|
| 1140 |
+
"learning_rate": 3.925496439005246e-05,
|
| 1141 |
+
"loss": 0.7632,
|
| 1142 |
+
"step": 1890
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 0.62,
|
| 1146 |
+
"learning_rate": 3.914979814050949e-05,
|
| 1147 |
+
"loss": 0.7646,
|
| 1148 |
+
"step": 1900
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 0.62,
|
| 1152 |
+
"learning_rate": 3.904426217863858e-05,
|
| 1153 |
+
"loss": 0.7624,
|
| 1154 |
+
"step": 1910
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.62,
|
| 1158 |
+
"learning_rate": 3.8938359261931076e-05,
|
| 1159 |
+
"loss": 0.7498,
|
| 1160 |
+
"step": 1920
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.63,
|
| 1164 |
+
"learning_rate": 3.883209215746633e-05,
|
| 1165 |
+
"loss": 0.7541,
|
| 1166 |
+
"step": 1930
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 0.63,
|
| 1170 |
+
"learning_rate": 3.8725463641839296e-05,
|
| 1171 |
+
"loss": 0.7622,
|
| 1172 |
+
"step": 1940
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.63,
|
| 1176 |
+
"learning_rate": 3.8618476501088094e-05,
|
| 1177 |
+
"loss": 0.7536,
|
| 1178 |
+
"step": 1950
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.64,
|
| 1182 |
+
"learning_rate": 3.851113353062115e-05,
|
| 1183 |
+
"loss": 0.7419,
|
| 1184 |
+
"step": 1960
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 0.64,
|
| 1188 |
+
"learning_rate": 3.840343753514414e-05,
|
| 1189 |
+
"loss": 0.756,
|
| 1190 |
+
"step": 1970
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 0.64,
|
| 1194 |
+
"learning_rate": 3.829539132858678e-05,
|
| 1195 |
+
"loss": 0.7511,
|
| 1196 |
+
"step": 1980
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.65,
|
| 1200 |
+
"learning_rate": 3.818699773402925e-05,
|
| 1201 |
+
"loss": 0.7505,
|
| 1202 |
+
"step": 1990
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.65,
|
| 1206 |
+
"learning_rate": 3.80782595836284e-05,
|
| 1207 |
+
"loss": 0.7396,
|
| 1208 |
+
"step": 2000
|
| 1209 |
+
}
|
| 1210 |
+
],
|
| 1211 |
+
"max_steps": 6146,
|
| 1212 |
+
"num_train_epochs": 2,
|
| 1213 |
+
"total_flos": 3.592019131911635e+18,
|
| 1214 |
+
"trial_name": null,
|
| 1215 |
+
"trial_params": null
|
| 1216 |
+
}
|
checkpoint-2000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dafdcf787495b020f1b3f6f6180b816b2753ea01048ce4086dbcd8a2c7a5f666
|
| 3 |
+
size 3265
|
checkpoint-3000/README.md
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
| 4 |
+
## Training procedure
|
| 5 |
+
|
| 6 |
+
|
| 7 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 8 |
+
- load_in_8bit: False
|
| 9 |
+
- load_in_4bit: True
|
| 10 |
+
- llm_int8_threshold: 6.0
|
| 11 |
+
- llm_int8_skip_modules: None
|
| 12 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 13 |
+
- llm_int8_has_fp16_weight: False
|
| 14 |
+
- bnb_4bit_quant_type: nf4
|
| 15 |
+
- bnb_4bit_use_double_quant: True
|
| 16 |
+
- bnb_4bit_compute_dtype: float16
|
| 17 |
+
### Framework versions
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
- PEFT 0.4.0
|
checkpoint-3000/adapter_config.json
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "./Llama-2-7b-chat-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"fan_in_fan_out": false,
|
| 6 |
+
"inference_mode": true,
|
| 7 |
+
"init_lora_weights": true,
|
| 8 |
+
"layers_pattern": null,
|
| 9 |
+
"layers_to_transform": null,
|
| 10 |
+
"lora_alpha": 32.0,
|
| 11 |
+
"lora_dropout": 0.1,
|
| 12 |
+
"modules_to_save": null,
|
| 13 |
+
"peft_type": "LORA",
|
| 14 |
+
"r": 8,
|
| 15 |
+
"revision": null,
|
| 16 |
+
"target_modules": [
|
| 17 |
+
"q_proj",
|
| 18 |
+
"v_proj"
|
| 19 |
+
],
|
| 20 |
+
"task_type": "CAUSAL_LM"
|
| 21 |
+
}
|
checkpoint-3000/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:499ec3e7b568f14ac8e621f300a8d1bc83dfebcde0368399d16760e784d09bc5
|
| 3 |
+
size 16821197
|
checkpoint-3000/finetuning_args.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"dpo_beta": 0.1,
|
| 3 |
+
"finetuning_type": "lora",
|
| 4 |
+
"lora_alpha": 32.0,
|
| 5 |
+
"lora_dropout": 0.1,
|
| 6 |
+
"lora_rank": 8,
|
| 7 |
+
"lora_target": [
|
| 8 |
+
"q_proj",
|
| 9 |
+
"v_proj"
|
| 10 |
+
],
|
| 11 |
+
"name_module_trainable": "mlp",
|
| 12 |
+
"num_hidden_layers": 32,
|
| 13 |
+
"num_layer_trainable": 3,
|
| 14 |
+
"ppo_score_norm": false,
|
| 15 |
+
"resume_lora_training": true
|
| 16 |
+
}
|
checkpoint-3000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d3d9effdc66e02ddb4d90e3853b9f83fe0d3eae77ce8fcc704aa9426d95aa124
|
| 3 |
+
size 33661637
|
checkpoint-3000/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5c4714e054267d9812f1cb05a99851d89d39d278f2ce8d02c5040e137d917671
|
| 3 |
+
size 18679
|
checkpoint-3000/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8f431207c4be949dbe34242eac350e08d4d64175be925d50bae5f64b579da6be
|
| 3 |
+
size 18679
|
checkpoint-3000/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1c015b7870fd9622533cc30be27a265b372068b75a1784db266144919911946c
|
| 3 |
+
size 18679
|
checkpoint-3000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:95f7b53177fec8a949bac40de334a438aba24a57af950f593701de5510c39cda
|
| 3 |
+
size 627
|
checkpoint-3000/trainer_state.json
ADDED
|
@@ -0,0 +1,1816 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9760065067100447,
|
| 5 |
+
"global_step": 3000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 4.999967339434433e-05,
|
| 13 |
+
"loss": 1.343,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 4.9998820960283476e-05,
|
| 19 |
+
"loss": 1.1858,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 4.99974394497948e-05,
|
| 25 |
+
"loss": 1.1057,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.01,
|
| 30 |
+
"learning_rate": 4.9995528892116685e-05,
|
| 31 |
+
"loss": 1.0698,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.02,
|
| 36 |
+
"learning_rate": 4.999308932768425e-05,
|
| 37 |
+
"loss": 1.0316,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.02,
|
| 42 |
+
"learning_rate": 4.99897583236926e-05,
|
| 43 |
+
"loss": 1.0046,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.02,
|
| 48 |
+
"learning_rate": 4.9985774375988764e-05,
|
| 49 |
+
"loss": 0.9848,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.03,
|
| 54 |
+
"learning_rate": 4.998113758866712e-05,
|
| 55 |
+
"loss": 0.9656,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.03,
|
| 60 |
+
"learning_rate": 4.997584808287975e-05,
|
| 61 |
+
"loss": 0.9648,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.03,
|
| 66 |
+
"learning_rate": 4.996990599683325e-05,
|
| 67 |
+
"loss": 0.9415,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.04,
|
| 72 |
+
"learning_rate": 4.996331148578514e-05,
|
| 73 |
+
"loss": 0.9388,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.04,
|
| 78 |
+
"learning_rate": 4.995606472203979e-05,
|
| 79 |
+
"loss": 0.9289,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.04,
|
| 84 |
+
"learning_rate": 4.994816589494391e-05,
|
| 85 |
+
"loss": 0.9252,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.05,
|
| 90 |
+
"learning_rate": 4.993961521088163e-05,
|
| 91 |
+
"loss": 0.91,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.05,
|
| 96 |
+
"learning_rate": 4.9930412893269106e-05,
|
| 97 |
+
"loss": 0.9033,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.05,
|
| 102 |
+
"learning_rate": 4.992055918254865e-05,
|
| 103 |
+
"loss": 0.9064,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.06,
|
| 108 |
+
"learning_rate": 4.9910054336182467e-05,
|
| 109 |
+
"loss": 0.8946,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.06,
|
| 114 |
+
"learning_rate": 4.989889862864594e-05,
|
| 115 |
+
"loss": 0.8914,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.06,
|
| 120 |
+
"learning_rate": 4.988709235142044e-05,
|
| 121 |
+
"loss": 0.9007,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.07,
|
| 126 |
+
"learning_rate": 4.9874635812985734e-05,
|
| 127 |
+
"loss": 0.9008,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.07,
|
| 132 |
+
"learning_rate": 4.9861529338811874e-05,
|
| 133 |
+
"loss": 0.8714,
|
| 134 |
+
"step": 210
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.07,
|
| 138 |
+
"learning_rate": 4.984777327135076e-05,
|
| 139 |
+
"loss": 0.8865,
|
| 140 |
+
"step": 220
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.07,
|
| 144 |
+
"learning_rate": 4.9833367970027145e-05,
|
| 145 |
+
"loss": 0.8716,
|
| 146 |
+
"step": 230
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.08,
|
| 150 |
+
"learning_rate": 4.981831381122926e-05,
|
| 151 |
+
"loss": 0.8817,
|
| 152 |
+
"step": 240
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.08,
|
| 156 |
+
"learning_rate": 4.980261118829896e-05,
|
| 157 |
+
"loss": 0.8658,
|
| 158 |
+
"step": 250
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.08,
|
| 162 |
+
"learning_rate": 4.978626051152152e-05,
|
| 163 |
+
"loss": 0.8722,
|
| 164 |
+
"step": 260
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.09,
|
| 168 |
+
"learning_rate": 4.976926220811479e-05,
|
| 169 |
+
"loss": 0.8541,
|
| 170 |
+
"step": 270
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.09,
|
| 174 |
+
"learning_rate": 4.975161672221813e-05,
|
| 175 |
+
"loss": 0.8504,
|
| 176 |
+
"step": 280
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.09,
|
| 180 |
+
"learning_rate": 4.97333245148808e-05,
|
| 181 |
+
"loss": 0.8467,
|
| 182 |
+
"step": 290
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.1,
|
| 186 |
+
"learning_rate": 4.971438606404986e-05,
|
| 187 |
+
"loss": 0.8533,
|
| 188 |
+
"step": 300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.1,
|
| 192 |
+
"learning_rate": 4.969480186455771e-05,
|
| 193 |
+
"loss": 0.853,
|
| 194 |
+
"step": 310
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.1,
|
| 198 |
+
"learning_rate": 4.9674572428109205e-05,
|
| 199 |
+
"loss": 0.8432,
|
| 200 |
+
"step": 320
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.11,
|
| 204 |
+
"learning_rate": 4.9653698283268183e-05,
|
| 205 |
+
"loss": 0.854,
|
| 206 |
+
"step": 330
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.11,
|
| 210 |
+
"learning_rate": 4.963217997544376e-05,
|
| 211 |
+
"loss": 0.8435,
|
| 212 |
+
"step": 340
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.11,
|
| 216 |
+
"learning_rate": 4.961001806687602e-05,
|
| 217 |
+
"loss": 0.8518,
|
| 218 |
+
"step": 350
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.12,
|
| 222 |
+
"learning_rate": 4.958721313662133e-05,
|
| 223 |
+
"loss": 0.848,
|
| 224 |
+
"step": 360
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.12,
|
| 228 |
+
"learning_rate": 4.956376578053723e-05,
|
| 229 |
+
"loss": 0.8436,
|
| 230 |
+
"step": 370
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.12,
|
| 234 |
+
"learning_rate": 4.9539676611266846e-05,
|
| 235 |
+
"loss": 0.8405,
|
| 236 |
+
"step": 380
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.13,
|
| 240 |
+
"learning_rate": 4.9514946258222895e-05,
|
| 241 |
+
"loss": 0.838,
|
| 242 |
+
"step": 390
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.13,
|
| 246 |
+
"learning_rate": 4.948957536757123e-05,
|
| 247 |
+
"loss": 0.8458,
|
| 248 |
+
"step": 400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.13,
|
| 252 |
+
"learning_rate": 4.946356460221395e-05,
|
| 253 |
+
"loss": 0.8379,
|
| 254 |
+
"step": 410
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.14,
|
| 258 |
+
"learning_rate": 4.943691464177212e-05,
|
| 259 |
+
"loss": 0.827,
|
| 260 |
+
"step": 420
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.14,
|
| 264 |
+
"learning_rate": 4.940962618256796e-05,
|
| 265 |
+
"loss": 0.8208,
|
| 266 |
+
"step": 430
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.14,
|
| 270 |
+
"learning_rate": 4.9381699937606674e-05,
|
| 271 |
+
"loss": 0.8242,
|
| 272 |
+
"step": 440
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.15,
|
| 276 |
+
"learning_rate": 4.935313663655782e-05,
|
| 277 |
+
"loss": 0.8362,
|
| 278 |
+
"step": 450
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.15,
|
| 282 |
+
"learning_rate": 4.932393702573627e-05,
|
| 283 |
+
"loss": 0.8363,
|
| 284 |
+
"step": 460
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.15,
|
| 288 |
+
"learning_rate": 4.929410186808264e-05,
|
| 289 |
+
"loss": 0.8275,
|
| 290 |
+
"step": 470
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.16,
|
| 294 |
+
"learning_rate": 4.926363194314344e-05,
|
| 295 |
+
"loss": 0.821,
|
| 296 |
+
"step": 480
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.16,
|
| 300 |
+
"learning_rate": 4.923252804705068e-05,
|
| 301 |
+
"loss": 0.8271,
|
| 302 |
+
"step": 490
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.16,
|
| 306 |
+
"learning_rate": 4.920079099250099e-05,
|
| 307 |
+
"loss": 0.8178,
|
| 308 |
+
"step": 500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.17,
|
| 312 |
+
"learning_rate": 4.916842160873451e-05,
|
| 313 |
+
"loss": 0.8293,
|
| 314 |
+
"step": 510
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.17,
|
| 318 |
+
"learning_rate": 4.913542074151315e-05,
|
| 319 |
+
"loss": 0.8294,
|
| 320 |
+
"step": 520
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.17,
|
| 324 |
+
"learning_rate": 4.910178925309849e-05,
|
| 325 |
+
"loss": 0.8202,
|
| 326 |
+
"step": 530
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.18,
|
| 330 |
+
"learning_rate": 4.906752802222928e-05,
|
| 331 |
+
"loss": 0.8301,
|
| 332 |
+
"step": 540
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.18,
|
| 336 |
+
"learning_rate": 4.903263794409846e-05,
|
| 337 |
+
"loss": 0.8031,
|
| 338 |
+
"step": 550
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.18,
|
| 342 |
+
"learning_rate": 4.8997119930329785e-05,
|
| 343 |
+
"loss": 0.821,
|
| 344 |
+
"step": 560
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.19,
|
| 348 |
+
"learning_rate": 4.896097490895398e-05,
|
| 349 |
+
"loss": 0.8132,
|
| 350 |
+
"step": 570
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.19,
|
| 354 |
+
"learning_rate": 4.8924203824384516e-05,
|
| 355 |
+
"loss": 0.8113,
|
| 356 |
+
"step": 580
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.19,
|
| 360 |
+
"learning_rate": 4.888680763739294e-05,
|
| 361 |
+
"loss": 0.8199,
|
| 362 |
+
"step": 590
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.2,
|
| 366 |
+
"learning_rate": 4.8848787325083734e-05,
|
| 367 |
+
"loss": 0.8196,
|
| 368 |
+
"step": 600
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.2,
|
| 372 |
+
"learning_rate": 4.8810143880868816e-05,
|
| 373 |
+
"loss": 0.8128,
|
| 374 |
+
"step": 610
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.2,
|
| 378 |
+
"learning_rate": 4.8770878314441606e-05,
|
| 379 |
+
"loss": 0.8123,
|
| 380 |
+
"step": 620
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.2,
|
| 384 |
+
"learning_rate": 4.8730991651750565e-05,
|
| 385 |
+
"loss": 0.8048,
|
| 386 |
+
"step": 630
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.21,
|
| 390 |
+
"learning_rate": 4.869048493497247e-05,
|
| 391 |
+
"loss": 0.8119,
|
| 392 |
+
"step": 640
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.21,
|
| 396 |
+
"learning_rate": 4.864935922248515e-05,
|
| 397 |
+
"loss": 0.8141,
|
| 398 |
+
"step": 650
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.21,
|
| 402 |
+
"learning_rate": 4.860761558883981e-05,
|
| 403 |
+
"loss": 0.8,
|
| 404 |
+
"step": 660
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.22,
|
| 408 |
+
"learning_rate": 4.856525512473302e-05,
|
| 409 |
+
"loss": 0.816,
|
| 410 |
+
"step": 670
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.22,
|
| 414 |
+
"learning_rate": 4.852227893697814e-05,
|
| 415 |
+
"loss": 0.7976,
|
| 416 |
+
"step": 680
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.22,
|
| 420 |
+
"learning_rate": 4.847868814847646e-05,
|
| 421 |
+
"loss": 0.8121,
|
| 422 |
+
"step": 690
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.23,
|
| 426 |
+
"learning_rate": 4.84344838981878e-05,
|
| 427 |
+
"loss": 0.8175,
|
| 428 |
+
"step": 700
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.23,
|
| 432 |
+
"learning_rate": 4.8389667341100844e-05,
|
| 433 |
+
"loss": 0.8109,
|
| 434 |
+
"step": 710
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.23,
|
| 438 |
+
"learning_rate": 4.834423964820284e-05,
|
| 439 |
+
"loss": 0.7948,
|
| 440 |
+
"step": 720
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.24,
|
| 444 |
+
"learning_rate": 4.8298202006449136e-05,
|
| 445 |
+
"loss": 0.7986,
|
| 446 |
+
"step": 730
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.24,
|
| 450 |
+
"learning_rate": 4.8251555618732036e-05,
|
| 451 |
+
"loss": 0.8049,
|
| 452 |
+
"step": 740
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.24,
|
| 456 |
+
"learning_rate": 4.820430170384947e-05,
|
| 457 |
+
"loss": 0.7981,
|
| 458 |
+
"step": 750
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.25,
|
| 462 |
+
"learning_rate": 4.8156441496473126e-05,
|
| 463 |
+
"loss": 0.799,
|
| 464 |
+
"step": 760
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.25,
|
| 468 |
+
"learning_rate": 4.8107976247116145e-05,
|
| 469 |
+
"loss": 0.804,
|
| 470 |
+
"step": 770
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.25,
|
| 474 |
+
"learning_rate": 4.805890722210048e-05,
|
| 475 |
+
"loss": 0.8053,
|
| 476 |
+
"step": 780
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.26,
|
| 480 |
+
"learning_rate": 4.8009235703523824e-05,
|
| 481 |
+
"loss": 0.7969,
|
| 482 |
+
"step": 790
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.26,
|
| 486 |
+
"learning_rate": 4.79589629892261e-05,
|
| 487 |
+
"loss": 0.8051,
|
| 488 |
+
"step": 800
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.26,
|
| 492 |
+
"learning_rate": 4.7908090392755526e-05,
|
| 493 |
+
"loss": 0.796,
|
| 494 |
+
"step": 810
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.27,
|
| 498 |
+
"learning_rate": 4.785661924333431e-05,
|
| 499 |
+
"loss": 0.7928,
|
| 500 |
+
"step": 820
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.27,
|
| 504 |
+
"learning_rate": 4.780455088582394e-05,
|
| 505 |
+
"loss": 0.8017,
|
| 506 |
+
"step": 830
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.27,
|
| 510 |
+
"learning_rate": 4.775188668069002e-05,
|
| 511 |
+
"loss": 0.7969,
|
| 512 |
+
"step": 840
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.28,
|
| 516 |
+
"learning_rate": 4.769862800396674e-05,
|
| 517 |
+
"loss": 0.7942,
|
| 518 |
+
"step": 850
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.28,
|
| 522 |
+
"learning_rate": 4.764477624722088e-05,
|
| 523 |
+
"loss": 0.7877,
|
| 524 |
+
"step": 860
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.28,
|
| 528 |
+
"learning_rate": 4.759033281751553e-05,
|
| 529 |
+
"loss": 0.7979,
|
| 530 |
+
"step": 870
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.29,
|
| 534 |
+
"learning_rate": 4.7535299137373236e-05,
|
| 535 |
+
"loss": 0.7986,
|
| 536 |
+
"step": 880
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.29,
|
| 540 |
+
"learning_rate": 4.7479676644738894e-05,
|
| 541 |
+
"loss": 0.8124,
|
| 542 |
+
"step": 890
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.29,
|
| 546 |
+
"learning_rate": 4.742346679294217e-05,
|
| 547 |
+
"loss": 0.7959,
|
| 548 |
+
"step": 900
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.3,
|
| 552 |
+
"learning_rate": 4.7366671050659505e-05,
|
| 553 |
+
"loss": 0.7884,
|
| 554 |
+
"step": 910
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.3,
|
| 558 |
+
"learning_rate": 4.730929090187573e-05,
|
| 559 |
+
"loss": 0.7911,
|
| 560 |
+
"step": 920
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.3,
|
| 564 |
+
"learning_rate": 4.725132784584536e-05,
|
| 565 |
+
"loss": 0.7937,
|
| 566 |
+
"step": 930
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.31,
|
| 570 |
+
"learning_rate": 4.719278339705333e-05,
|
| 571 |
+
"loss": 0.7994,
|
| 572 |
+
"step": 940
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.31,
|
| 576 |
+
"learning_rate": 4.71336590851755e-05,
|
| 577 |
+
"loss": 0.7905,
|
| 578 |
+
"step": 950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.31,
|
| 582 |
+
"learning_rate": 4.707395645503863e-05,
|
| 583 |
+
"loss": 0.7924,
|
| 584 |
+
"step": 960
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.32,
|
| 588 |
+
"learning_rate": 4.701367706658006e-05,
|
| 589 |
+
"loss": 0.7891,
|
| 590 |
+
"step": 970
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.32,
|
| 594 |
+
"learning_rate": 4.695282249480692e-05,
|
| 595 |
+
"loss": 0.7962,
|
| 596 |
+
"step": 980
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.32,
|
| 600 |
+
"learning_rate": 4.6891394329755e-05,
|
| 601 |
+
"loss": 0.7882,
|
| 602 |
+
"step": 990
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.33,
|
| 606 |
+
"learning_rate": 4.6829394176447195e-05,
|
| 607 |
+
"loss": 0.8011,
|
| 608 |
+
"step": 1000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.33,
|
| 612 |
+
"learning_rate": 4.676682365485154e-05,
|
| 613 |
+
"loss": 0.7955,
|
| 614 |
+
"step": 1010
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.33,
|
| 618 |
+
"learning_rate": 4.670368439983896e-05,
|
| 619 |
+
"loss": 0.7906,
|
| 620 |
+
"step": 1020
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.34,
|
| 624 |
+
"learning_rate": 4.663997806114048e-05,
|
| 625 |
+
"loss": 0.7853,
|
| 626 |
+
"step": 1030
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.34,
|
| 630 |
+
"learning_rate": 4.657570630330411e-05,
|
| 631 |
+
"loss": 0.7825,
|
| 632 |
+
"step": 1040
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.34,
|
| 636 |
+
"learning_rate": 4.6510870805651444e-05,
|
| 637 |
+
"loss": 0.7768,
|
| 638 |
+
"step": 1050
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.34,
|
| 642 |
+
"learning_rate": 4.6445473262233696e-05,
|
| 643 |
+
"loss": 0.7795,
|
| 644 |
+
"step": 1060
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.35,
|
| 648 |
+
"learning_rate": 4.637951538178746e-05,
|
| 649 |
+
"loss": 0.7832,
|
| 650 |
+
"step": 1070
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.35,
|
| 654 |
+
"learning_rate": 4.63129988876901e-05,
|
| 655 |
+
"loss": 0.7934,
|
| 656 |
+
"step": 1080
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.35,
|
| 660 |
+
"learning_rate": 4.624592551791465e-05,
|
| 661 |
+
"loss": 0.7892,
|
| 662 |
+
"step": 1090
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.36,
|
| 666 |
+
"learning_rate": 4.617829702498446e-05,
|
| 667 |
+
"loss": 0.7819,
|
| 668 |
+
"step": 1100
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.36,
|
| 672 |
+
"learning_rate": 4.611011517592741e-05,
|
| 673 |
+
"loss": 0.7868,
|
| 674 |
+
"step": 1110
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.36,
|
| 678 |
+
"learning_rate": 4.604138175222968e-05,
|
| 679 |
+
"loss": 0.7914,
|
| 680 |
+
"step": 1120
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.37,
|
| 684 |
+
"learning_rate": 4.597209854978929e-05,
|
| 685 |
+
"loss": 0.7817,
|
| 686 |
+
"step": 1130
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 0.37,
|
| 690 |
+
"learning_rate": 4.590226737886908e-05,
|
| 691 |
+
"loss": 0.7916,
|
| 692 |
+
"step": 1140
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.37,
|
| 696 |
+
"learning_rate": 4.583189006404949e-05,
|
| 697 |
+
"loss": 0.7799,
|
| 698 |
+
"step": 1150
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.38,
|
| 702 |
+
"learning_rate": 4.576096844418083e-05,
|
| 703 |
+
"loss": 0.7776,
|
| 704 |
+
"step": 1160
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.38,
|
| 708 |
+
"learning_rate": 4.568950437233529e-05,
|
| 709 |
+
"loss": 0.7803,
|
| 710 |
+
"step": 1170
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.38,
|
| 714 |
+
"learning_rate": 4.561749971575846e-05,
|
| 715 |
+
"loss": 0.7884,
|
| 716 |
+
"step": 1180
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.39,
|
| 720 |
+
"learning_rate": 4.5544956355820596e-05,
|
| 721 |
+
"loss": 0.7827,
|
| 722 |
+
"step": 1190
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.39,
|
| 726 |
+
"learning_rate": 4.547187618796742e-05,
|
| 727 |
+
"loss": 0.7857,
|
| 728 |
+
"step": 1200
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 0.39,
|
| 732 |
+
"learning_rate": 4.5398261121670634e-05,
|
| 733 |
+
"loss": 0.7797,
|
| 734 |
+
"step": 1210
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.4,
|
| 738 |
+
"learning_rate": 4.532411308037798e-05,
|
| 739 |
+
"loss": 0.7821,
|
| 740 |
+
"step": 1220
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.4,
|
| 744 |
+
"learning_rate": 4.524943400146306e-05,
|
| 745 |
+
"loss": 0.7747,
|
| 746 |
+
"step": 1230
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.4,
|
| 750 |
+
"learning_rate": 4.51742258361746e-05,
|
| 751 |
+
"loss": 0.7737,
|
| 752 |
+
"step": 1240
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.41,
|
| 756 |
+
"learning_rate": 4.509849054958559e-05,
|
| 757 |
+
"loss": 0.7868,
|
| 758 |
+
"step": 1250
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.41,
|
| 762 |
+
"learning_rate": 4.502223012054186e-05,
|
| 763 |
+
"loss": 0.7818,
|
| 764 |
+
"step": 1260
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.41,
|
| 768 |
+
"learning_rate": 4.4945446541610414e-05,
|
| 769 |
+
"loss": 0.7804,
|
| 770 |
+
"step": 1270
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 0.42,
|
| 774 |
+
"learning_rate": 4.486814181902734e-05,
|
| 775 |
+
"loss": 0.7759,
|
| 776 |
+
"step": 1280
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.42,
|
| 780 |
+
"learning_rate": 4.4790317972645396e-05,
|
| 781 |
+
"loss": 0.7744,
|
| 782 |
+
"step": 1290
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.42,
|
| 786 |
+
"learning_rate": 4.4711977035881256e-05,
|
| 787 |
+
"loss": 0.7814,
|
| 788 |
+
"step": 1300
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.43,
|
| 792 |
+
"learning_rate": 4.463312105566237e-05,
|
| 793 |
+
"loss": 0.7611,
|
| 794 |
+
"step": 1310
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.43,
|
| 798 |
+
"learning_rate": 4.4553752092373465e-05,
|
| 799 |
+
"loss": 0.7796,
|
| 800 |
+
"step": 1320
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.43,
|
| 804 |
+
"learning_rate": 4.447387221980272e-05,
|
| 805 |
+
"loss": 0.7839,
|
| 806 |
+
"step": 1330
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.44,
|
| 810 |
+
"learning_rate": 4.4393483525087584e-05,
|
| 811 |
+
"loss": 0.7726,
|
| 812 |
+
"step": 1340
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.44,
|
| 816 |
+
"learning_rate": 4.4312588108660246e-05,
|
| 817 |
+
"loss": 0.7763,
|
| 818 |
+
"step": 1350
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.44,
|
| 822 |
+
"learning_rate": 4.423118808419276e-05,
|
| 823 |
+
"loss": 0.766,
|
| 824 |
+
"step": 1360
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.45,
|
| 828 |
+
"learning_rate": 4.4149285578541785e-05,
|
| 829 |
+
"loss": 0.7887,
|
| 830 |
+
"step": 1370
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.45,
|
| 834 |
+
"learning_rate": 4.406688273169305e-05,
|
| 835 |
+
"loss": 0.7637,
|
| 836 |
+
"step": 1380
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.45,
|
| 840 |
+
"learning_rate": 4.3983981696705415e-05,
|
| 841 |
+
"loss": 0.7609,
|
| 842 |
+
"step": 1390
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.46,
|
| 846 |
+
"learning_rate": 4.390058463965464e-05,
|
| 847 |
+
"loss": 0.7729,
|
| 848 |
+
"step": 1400
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.46,
|
| 852 |
+
"learning_rate": 4.381669373957675e-05,
|
| 853 |
+
"loss": 0.7751,
|
| 854 |
+
"step": 1410
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.46,
|
| 858 |
+
"learning_rate": 4.3732311188411154e-05,
|
| 859 |
+
"loss": 0.7877,
|
| 860 |
+
"step": 1420
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.47,
|
| 864 |
+
"learning_rate": 4.3647439190943315e-05,
|
| 865 |
+
"loss": 0.7738,
|
| 866 |
+
"step": 1430
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.47,
|
| 870 |
+
"learning_rate": 4.3562079964747207e-05,
|
| 871 |
+
"loss": 0.7834,
|
| 872 |
+
"step": 1440
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.47,
|
| 876 |
+
"learning_rate": 4.347623574012728e-05,
|
| 877 |
+
"loss": 0.7835,
|
| 878 |
+
"step": 1450
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.47,
|
| 882 |
+
"learning_rate": 4.338990876006031e-05,
|
| 883 |
+
"loss": 0.7707,
|
| 884 |
+
"step": 1460
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.48,
|
| 888 |
+
"learning_rate": 4.330310128013666e-05,
|
| 889 |
+
"loss": 0.7674,
|
| 890 |
+
"step": 1470
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.48,
|
| 894 |
+
"learning_rate": 4.321581556850146e-05,
|
| 895 |
+
"loss": 0.7687,
|
| 896 |
+
"step": 1480
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.48,
|
| 900 |
+
"learning_rate": 4.312805390579527e-05,
|
| 901 |
+
"loss": 0.7663,
|
| 902 |
+
"step": 1490
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.49,
|
| 906 |
+
"learning_rate": 4.3039818585094524e-05,
|
| 907 |
+
"loss": 0.7662,
|
| 908 |
+
"step": 1500
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.49,
|
| 912 |
+
"learning_rate": 4.29511119118516e-05,
|
| 913 |
+
"loss": 0.7688,
|
| 914 |
+
"step": 1510
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.49,
|
| 918 |
+
"learning_rate": 4.28619362038346e-05,
|
| 919 |
+
"loss": 0.7652,
|
| 920 |
+
"step": 1520
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.5,
|
| 924 |
+
"learning_rate": 4.277229379106676e-05,
|
| 925 |
+
"loss": 0.7683,
|
| 926 |
+
"step": 1530
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.5,
|
| 930 |
+
"learning_rate": 4.2682187015765615e-05,
|
| 931 |
+
"loss": 0.7621,
|
| 932 |
+
"step": 1540
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.5,
|
| 936 |
+
"learning_rate": 4.259161823228174e-05,
|
| 937 |
+
"loss": 0.7778,
|
| 938 |
+
"step": 1550
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 0.51,
|
| 942 |
+
"learning_rate": 4.2500589807037306e-05,
|
| 943 |
+
"loss": 0.7809,
|
| 944 |
+
"step": 1560
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 0.51,
|
| 948 |
+
"learning_rate": 4.240910411846418e-05,
|
| 949 |
+
"loss": 0.7717,
|
| 950 |
+
"step": 1570
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.51,
|
| 954 |
+
"learning_rate": 4.231716355694184e-05,
|
| 955 |
+
"loss": 0.7584,
|
| 956 |
+
"step": 1580
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.52,
|
| 960 |
+
"learning_rate": 4.222477052473486e-05,
|
| 961 |
+
"loss": 0.7631,
|
| 962 |
+
"step": 1590
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.52,
|
| 966 |
+
"learning_rate": 4.21319274359302e-05,
|
| 967 |
+
"loss": 0.7681,
|
| 968 |
+
"step": 1600
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.52,
|
| 972 |
+
"learning_rate": 4.203863671637409e-05,
|
| 973 |
+
"loss": 0.7739,
|
| 974 |
+
"step": 1610
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 0.53,
|
| 978 |
+
"learning_rate": 4.194490080360866e-05,
|
| 979 |
+
"loss": 0.7646,
|
| 980 |
+
"step": 1620
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.53,
|
| 984 |
+
"learning_rate": 4.1850722146808254e-05,
|
| 985 |
+
"loss": 0.7647,
|
| 986 |
+
"step": 1630
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.53,
|
| 990 |
+
"learning_rate": 4.1756103206715416e-05,
|
| 991 |
+
"loss": 0.7675,
|
| 992 |
+
"step": 1640
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.54,
|
| 996 |
+
"learning_rate": 4.166104645557662e-05,
|
| 997 |
+
"loss": 0.7652,
|
| 998 |
+
"step": 1650
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 0.54,
|
| 1002 |
+
"learning_rate": 4.15655543770777e-05,
|
| 1003 |
+
"loss": 0.765,
|
| 1004 |
+
"step": 1660
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.54,
|
| 1008 |
+
"learning_rate": 4.146962946627886e-05,
|
| 1009 |
+
"loss": 0.7614,
|
| 1010 |
+
"step": 1670
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.55,
|
| 1014 |
+
"learning_rate": 4.137327422954958e-05,
|
| 1015 |
+
"loss": 0.7544,
|
| 1016 |
+
"step": 1680
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 0.55,
|
| 1020 |
+
"learning_rate": 4.127649118450308e-05,
|
| 1021 |
+
"loss": 0.7673,
|
| 1022 |
+
"step": 1690
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 0.55,
|
| 1026 |
+
"learning_rate": 4.117928285993055e-05,
|
| 1027 |
+
"loss": 0.7593,
|
| 1028 |
+
"step": 1700
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 0.56,
|
| 1032 |
+
"learning_rate": 4.108165179573508e-05,
|
| 1033 |
+
"loss": 0.7735,
|
| 1034 |
+
"step": 1710
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.56,
|
| 1038 |
+
"learning_rate": 4.0983600542865285e-05,
|
| 1039 |
+
"loss": 0.7666,
|
| 1040 |
+
"step": 1720
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.56,
|
| 1044 |
+
"learning_rate": 4.088513166324867e-05,
|
| 1045 |
+
"loss": 0.7627,
|
| 1046 |
+
"step": 1730
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.57,
|
| 1050 |
+
"learning_rate": 4.078624772972467e-05,
|
| 1051 |
+
"loss": 0.77,
|
| 1052 |
+
"step": 1740
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.57,
|
| 1056 |
+
"learning_rate": 4.068695132597744e-05,
|
| 1057 |
+
"loss": 0.7569,
|
| 1058 |
+
"step": 1750
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 0.57,
|
| 1062 |
+
"learning_rate": 4.058724504646834e-05,
|
| 1063 |
+
"loss": 0.7654,
|
| 1064 |
+
"step": 1760
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 0.58,
|
| 1068 |
+
"learning_rate": 4.048713149636816e-05,
|
| 1069 |
+
"loss": 0.753,
|
| 1070 |
+
"step": 1770
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.58,
|
| 1074 |
+
"learning_rate": 4.038661329148904e-05,
|
| 1075 |
+
"loss": 0.7705,
|
| 1076 |
+
"step": 1780
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.58,
|
| 1080 |
+
"learning_rate": 4.02856930582161e-05,
|
| 1081 |
+
"loss": 0.7542,
|
| 1082 |
+
"step": 1790
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.59,
|
| 1086 |
+
"learning_rate": 4.018437343343886e-05,
|
| 1087 |
+
"loss": 0.7532,
|
| 1088 |
+
"step": 1800
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.59,
|
| 1092 |
+
"learning_rate": 4.008265706448234e-05,
|
| 1093 |
+
"loss": 0.7662,
|
| 1094 |
+
"step": 1810
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.59,
|
| 1098 |
+
"learning_rate": 3.9980546609037825e-05,
|
| 1099 |
+
"loss": 0.7607,
|
| 1100 |
+
"step": 1820
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 0.6,
|
| 1104 |
+
"learning_rate": 3.9878044735093505e-05,
|
| 1105 |
+
"loss": 0.7552,
|
| 1106 |
+
"step": 1830
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 0.6,
|
| 1110 |
+
"learning_rate": 3.977515412086472e-05,
|
| 1111 |
+
"loss": 0.7645,
|
| 1112 |
+
"step": 1840
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.6,
|
| 1116 |
+
"learning_rate": 3.9671877454723985e-05,
|
| 1117 |
+
"loss": 0.7689,
|
| 1118 |
+
"step": 1850
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.61,
|
| 1122 |
+
"learning_rate": 3.956821743513078e-05,
|
| 1123 |
+
"loss": 0.7615,
|
| 1124 |
+
"step": 1860
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.61,
|
| 1128 |
+
"learning_rate": 3.946417677056097e-05,
|
| 1129 |
+
"loss": 0.7703,
|
| 1130 |
+
"step": 1870
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.61,
|
| 1134 |
+
"learning_rate": 3.935975817943614e-05,
|
| 1135 |
+
"loss": 0.766,
|
| 1136 |
+
"step": 1880
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.61,
|
| 1140 |
+
"learning_rate": 3.925496439005246e-05,
|
| 1141 |
+
"loss": 0.7632,
|
| 1142 |
+
"step": 1890
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 0.62,
|
| 1146 |
+
"learning_rate": 3.914979814050949e-05,
|
| 1147 |
+
"loss": 0.7646,
|
| 1148 |
+
"step": 1900
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 0.62,
|
| 1152 |
+
"learning_rate": 3.904426217863858e-05,
|
| 1153 |
+
"loss": 0.7624,
|
| 1154 |
+
"step": 1910
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.62,
|
| 1158 |
+
"learning_rate": 3.8938359261931076e-05,
|
| 1159 |
+
"loss": 0.7498,
|
| 1160 |
+
"step": 1920
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.63,
|
| 1164 |
+
"learning_rate": 3.883209215746633e-05,
|
| 1165 |
+
"loss": 0.7541,
|
| 1166 |
+
"step": 1930
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 0.63,
|
| 1170 |
+
"learning_rate": 3.8725463641839296e-05,
|
| 1171 |
+
"loss": 0.7622,
|
| 1172 |
+
"step": 1940
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.63,
|
| 1176 |
+
"learning_rate": 3.8618476501088094e-05,
|
| 1177 |
+
"loss": 0.7536,
|
| 1178 |
+
"step": 1950
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.64,
|
| 1182 |
+
"learning_rate": 3.851113353062115e-05,
|
| 1183 |
+
"loss": 0.7419,
|
| 1184 |
+
"step": 1960
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 0.64,
|
| 1188 |
+
"learning_rate": 3.840343753514414e-05,
|
| 1189 |
+
"loss": 0.756,
|
| 1190 |
+
"step": 1970
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 0.64,
|
| 1194 |
+
"learning_rate": 3.829539132858678e-05,
|
| 1195 |
+
"loss": 0.7511,
|
| 1196 |
+
"step": 1980
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.65,
|
| 1200 |
+
"learning_rate": 3.818699773402925e-05,
|
| 1201 |
+
"loss": 0.7505,
|
| 1202 |
+
"step": 1990
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.65,
|
| 1206 |
+
"learning_rate": 3.80782595836284e-05,
|
| 1207 |
+
"loss": 0.7396,
|
| 1208 |
+
"step": 2000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 0.65,
|
| 1212 |
+
"learning_rate": 3.7969179718543855e-05,
|
| 1213 |
+
"loss": 0.7524,
|
| 1214 |
+
"step": 2010
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.66,
|
| 1218 |
+
"learning_rate": 3.7859760988863664e-05,
|
| 1219 |
+
"loss": 0.7486,
|
| 1220 |
+
"step": 2020
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.66,
|
| 1224 |
+
"learning_rate": 3.77500062535299e-05,
|
| 1225 |
+
"loss": 0.7599,
|
| 1226 |
+
"step": 2030
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 0.66,
|
| 1230 |
+
"learning_rate": 3.763991838026397e-05,
|
| 1231 |
+
"loss": 0.7526,
|
| 1232 |
+
"step": 2040
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 0.67,
|
| 1236 |
+
"learning_rate": 3.7529500245491616e-05,
|
| 1237 |
+
"loss": 0.7569,
|
| 1238 |
+
"step": 2050
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.67,
|
| 1242 |
+
"learning_rate": 3.741875473426783e-05,
|
| 1243 |
+
"loss": 0.7637,
|
| 1244 |
+
"step": 2060
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 0.67,
|
| 1248 |
+
"learning_rate": 3.730768474020144e-05,
|
| 1249 |
+
"loss": 0.757,
|
| 1250 |
+
"step": 2070
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 0.68,
|
| 1254 |
+
"learning_rate": 3.7196293165379486e-05,
|
| 1255 |
+
"loss": 0.765,
|
| 1256 |
+
"step": 2080
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.68,
|
| 1260 |
+
"learning_rate": 3.7084582920291456e-05,
|
| 1261 |
+
"loss": 0.7628,
|
| 1262 |
+
"step": 2090
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.68,
|
| 1266 |
+
"learning_rate": 3.697255692375317e-05,
|
| 1267 |
+
"loss": 0.7514,
|
| 1268 |
+
"step": 2100
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 0.69,
|
| 1272 |
+
"learning_rate": 3.686021810283055e-05,
|
| 1273 |
+
"loss": 0.7683,
|
| 1274 |
+
"step": 2110
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 0.69,
|
| 1278 |
+
"learning_rate": 3.6747569392763135e-05,
|
| 1279 |
+
"loss": 0.7664,
|
| 1280 |
+
"step": 2120
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.69,
|
| 1284 |
+
"learning_rate": 3.6634613736887404e-05,
|
| 1285 |
+
"loss": 0.7495,
|
| 1286 |
+
"step": 2130
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 0.7,
|
| 1290 |
+
"learning_rate": 3.652135408655982e-05,
|
| 1291 |
+
"loss": 0.7501,
|
| 1292 |
+
"step": 2140
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 0.7,
|
| 1296 |
+
"learning_rate": 3.6407793401079785e-05,
|
| 1297 |
+
"loss": 0.7581,
|
| 1298 |
+
"step": 2150
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.7,
|
| 1302 |
+
"learning_rate": 3.629393464761227e-05,
|
| 1303 |
+
"loss": 0.7564,
|
| 1304 |
+
"step": 2160
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.71,
|
| 1308 |
+
"learning_rate": 3.6179780801110305e-05,
|
| 1309 |
+
"loss": 0.7558,
|
| 1310 |
+
"step": 2170
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 0.71,
|
| 1314 |
+
"learning_rate": 3.606533484423722e-05,
|
| 1315 |
+
"loss": 0.7578,
|
| 1316 |
+
"step": 2180
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 0.71,
|
| 1320 |
+
"learning_rate": 3.5950599767288785e-05,
|
| 1321 |
+
"loss": 0.7585,
|
| 1322 |
+
"step": 2190
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.72,
|
| 1326 |
+
"learning_rate": 3.5835578568114985e-05,
|
| 1327 |
+
"loss": 0.7473,
|
| 1328 |
+
"step": 2200
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 0.72,
|
| 1332 |
+
"learning_rate": 3.572027425204176e-05,
|
| 1333 |
+
"loss": 0.746,
|
| 1334 |
+
"step": 2210
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 0.72,
|
| 1338 |
+
"learning_rate": 3.560468983179245e-05,
|
| 1339 |
+
"loss": 0.7517,
|
| 1340 |
+
"step": 2220
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.73,
|
| 1344 |
+
"learning_rate": 3.548882832740909e-05,
|
| 1345 |
+
"loss": 0.75,
|
| 1346 |
+
"step": 2230
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.73,
|
| 1350 |
+
"learning_rate": 3.537269276617347e-05,
|
| 1351 |
+
"loss": 0.7407,
|
| 1352 |
+
"step": 2240
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 0.73,
|
| 1356 |
+
"learning_rate": 3.5256286182528083e-05,
|
| 1357 |
+
"loss": 0.7465,
|
| 1358 |
+
"step": 2250
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 0.74,
|
| 1362 |
+
"learning_rate": 3.5139611617996846e-05,
|
| 1363 |
+
"loss": 0.7477,
|
| 1364 |
+
"step": 2260
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.74,
|
| 1368 |
+
"learning_rate": 3.5022672121105534e-05,
|
| 1369 |
+
"loss": 0.7494,
|
| 1370 |
+
"step": 2270
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.74,
|
| 1374 |
+
"learning_rate": 3.4905470747302256e-05,
|
| 1375 |
+
"loss": 0.7514,
|
| 1376 |
+
"step": 2280
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 0.75,
|
| 1380 |
+
"learning_rate": 3.478801055887752e-05,
|
| 1381 |
+
"loss": 0.7466,
|
| 1382 |
+
"step": 2290
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.75,
|
| 1386 |
+
"learning_rate": 3.4670294624884275e-05,
|
| 1387 |
+
"loss": 0.7552,
|
| 1388 |
+
"step": 2300
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.75,
|
| 1392 |
+
"learning_rate": 3.4552326021057716e-05,
|
| 1393 |
+
"loss": 0.7625,
|
| 1394 |
+
"step": 2310
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 0.75,
|
| 1398 |
+
"learning_rate": 3.443410782973489e-05,
|
| 1399 |
+
"loss": 0.7517,
|
| 1400 |
+
"step": 2320
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 0.76,
|
| 1404 |
+
"learning_rate": 3.431564313977419e-05,
|
| 1405 |
+
"loss": 0.7606,
|
| 1406 |
+
"step": 2330
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.76,
|
| 1410 |
+
"learning_rate": 3.419693504647464e-05,
|
| 1411 |
+
"loss": 0.7568,
|
| 1412 |
+
"step": 2340
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 0.76,
|
| 1416 |
+
"learning_rate": 3.4077986651495e-05,
|
| 1417 |
+
"loss": 0.7452,
|
| 1418 |
+
"step": 2350
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 0.77,
|
| 1422 |
+
"learning_rate": 3.395880106277277e-05,
|
| 1423 |
+
"loss": 0.746,
|
| 1424 |
+
"step": 2360
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 0.77,
|
| 1428 |
+
"learning_rate": 3.383938139444292e-05,
|
| 1429 |
+
"loss": 0.7602,
|
| 1430 |
+
"step": 2370
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.77,
|
| 1434 |
+
"learning_rate": 3.3719730766756596e-05,
|
| 1435 |
+
"loss": 0.7453,
|
| 1436 |
+
"step": 2380
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 0.78,
|
| 1440 |
+
"learning_rate": 3.3599852305999514e-05,
|
| 1441 |
+
"loss": 0.7606,
|
| 1442 |
+
"step": 2390
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 0.78,
|
| 1446 |
+
"learning_rate": 3.347974914441035e-05,
|
| 1447 |
+
"loss": 0.7545,
|
| 1448 |
+
"step": 2400
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 0.78,
|
| 1452 |
+
"learning_rate": 3.335942442009884e-05,
|
| 1453 |
+
"loss": 0.7379,
|
| 1454 |
+
"step": 2410
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 0.79,
|
| 1458 |
+
"learning_rate": 3.3238881276963835e-05,
|
| 1459 |
+
"loss": 0.7372,
|
| 1460 |
+
"step": 2420
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 0.79,
|
| 1464 |
+
"learning_rate": 3.311812286461111e-05,
|
| 1465 |
+
"loss": 0.7592,
|
| 1466 |
+
"step": 2430
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.79,
|
| 1470 |
+
"learning_rate": 3.299715233827111e-05,
|
| 1471 |
+
"loss": 0.7576,
|
| 1472 |
+
"step": 2440
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.8,
|
| 1476 |
+
"learning_rate": 3.2875972858716464e-05,
|
| 1477 |
+
"loss": 0.7449,
|
| 1478 |
+
"step": 2450
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 0.8,
|
| 1482 |
+
"learning_rate": 3.2754587592179454e-05,
|
| 1483 |
+
"loss": 0.7448,
|
| 1484 |
+
"step": 2460
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 0.8,
|
| 1488 |
+
"learning_rate": 3.263299971026925e-05,
|
| 1489 |
+
"loss": 0.7483,
|
| 1490 |
+
"step": 2470
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 0.81,
|
| 1494 |
+
"learning_rate": 3.251121238988902e-05,
|
| 1495 |
+
"loss": 0.7574,
|
| 1496 |
+
"step": 2480
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 0.81,
|
| 1500 |
+
"learning_rate": 3.2389228813153006e-05,
|
| 1501 |
+
"loss": 0.744,
|
| 1502 |
+
"step": 2490
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 0.81,
|
| 1506 |
+
"learning_rate": 3.2267052167303275e-05,
|
| 1507 |
+
"loss": 0.7445,
|
| 1508 |
+
"step": 2500
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 0.82,
|
| 1512 |
+
"learning_rate": 3.214468564462651e-05,
|
| 1513 |
+
"loss": 0.7486,
|
| 1514 |
+
"step": 2510
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.82,
|
| 1518 |
+
"learning_rate": 3.202213244237058e-05,
|
| 1519 |
+
"loss": 0.7631,
|
| 1520 |
+
"step": 2520
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 0.82,
|
| 1524 |
+
"learning_rate": 3.189939576266101e-05,
|
| 1525 |
+
"loss": 0.7538,
|
| 1526 |
+
"step": 2530
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 0.83,
|
| 1530 |
+
"learning_rate": 3.1776478812417295e-05,
|
| 1531 |
+
"loss": 0.7576,
|
| 1532 |
+
"step": 2540
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 0.83,
|
| 1536 |
+
"learning_rate": 3.165338480326912e-05,
|
| 1537 |
+
"loss": 0.7405,
|
| 1538 |
+
"step": 2550
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 0.83,
|
| 1542 |
+
"learning_rate": 3.1530116951472464e-05,
|
| 1543 |
+
"loss": 0.7362,
|
| 1544 |
+
"step": 2560
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 0.84,
|
| 1548 |
+
"learning_rate": 3.140667847782552e-05,
|
| 1549 |
+
"loss": 0.7503,
|
| 1550 |
+
"step": 2570
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 0.84,
|
| 1554 |
+
"learning_rate": 3.1283072607584575e-05,
|
| 1555 |
+
"loss": 0.7513,
|
| 1556 |
+
"step": 2580
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.84,
|
| 1560 |
+
"learning_rate": 3.115930257037975e-05,
|
| 1561 |
+
"loss": 0.7535,
|
| 1562 |
+
"step": 2590
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 0.85,
|
| 1566 |
+
"learning_rate": 3.1035371600130566e-05,
|
| 1567 |
+
"loss": 0.7386,
|
| 1568 |
+
"step": 2600
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 0.85,
|
| 1572 |
+
"learning_rate": 3.091128293496148e-05,
|
| 1573 |
+
"loss": 0.7617,
|
| 1574 |
+
"step": 2610
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 0.85,
|
| 1578 |
+
"learning_rate": 3.07870398171173e-05,
|
| 1579 |
+
"loss": 0.7519,
|
| 1580 |
+
"step": 2620
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 0.86,
|
| 1584 |
+
"learning_rate": 3.066264549287839e-05,
|
| 1585 |
+
"loss": 0.739,
|
| 1586 |
+
"step": 2630
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 0.86,
|
| 1590 |
+
"learning_rate": 3.053810321247597e-05,
|
| 1591 |
+
"loss": 0.7375,
|
| 1592 |
+
"step": 2640
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 0.86,
|
| 1596 |
+
"learning_rate": 3.041341623000708e-05,
|
| 1597 |
+
"loss": 0.7511,
|
| 1598 |
+
"step": 2650
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.87,
|
| 1602 |
+
"learning_rate": 3.028858780334961e-05,
|
| 1603 |
+
"loss": 0.7498,
|
| 1604 |
+
"step": 2660
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 0.87,
|
| 1608 |
+
"learning_rate": 3.0163621194077167e-05,
|
| 1609 |
+
"loss": 0.7504,
|
| 1610 |
+
"step": 2670
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 0.87,
|
| 1614 |
+
"learning_rate": 3.003851966737387e-05,
|
| 1615 |
+
"loss": 0.7508,
|
| 1616 |
+
"step": 2680
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 0.88,
|
| 1620 |
+
"learning_rate": 2.991328649194901e-05,
|
| 1621 |
+
"loss": 0.7465,
|
| 1622 |
+
"step": 2690
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 0.88,
|
| 1626 |
+
"learning_rate": 2.9787924939951644e-05,
|
| 1627 |
+
"loss": 0.741,
|
| 1628 |
+
"step": 2700
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 0.88,
|
| 1632 |
+
"learning_rate": 2.9662438286885146e-05,
|
| 1633 |
+
"loss": 0.7424,
|
| 1634 |
+
"step": 2710
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.88,
|
| 1638 |
+
"learning_rate": 2.953682981152154e-05,
|
| 1639 |
+
"loss": 0.7425,
|
| 1640 |
+
"step": 2720
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.89,
|
| 1644 |
+
"learning_rate": 2.941110279581591e-05,
|
| 1645 |
+
"loss": 0.7355,
|
| 1646 |
+
"step": 2730
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 0.89,
|
| 1650 |
+
"learning_rate": 2.9285260524820624e-05,
|
| 1651 |
+
"loss": 0.7374,
|
| 1652 |
+
"step": 2740
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 0.89,
|
| 1656 |
+
"learning_rate": 2.9159306286599446e-05,
|
| 1657 |
+
"loss": 0.7319,
|
| 1658 |
+
"step": 2750
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 0.9,
|
| 1662 |
+
"learning_rate": 2.9033243372141733e-05,
|
| 1663 |
+
"loss": 0.7515,
|
| 1664 |
+
"step": 2760
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 0.9,
|
| 1668 |
+
"learning_rate": 2.890707507527633e-05,
|
| 1669 |
+
"loss": 0.7391,
|
| 1670 |
+
"step": 2770
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 0.9,
|
| 1674 |
+
"learning_rate": 2.8780804692585585e-05,
|
| 1675 |
+
"loss": 0.7477,
|
| 1676 |
+
"step": 2780
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.91,
|
| 1680 |
+
"learning_rate": 2.86544355233192e-05,
|
| 1681 |
+
"loss": 0.7498,
|
| 1682 |
+
"step": 2790
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.91,
|
| 1686 |
+
"learning_rate": 2.8527970869307996e-05,
|
| 1687 |
+
"loss": 0.7417,
|
| 1688 |
+
"step": 2800
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 0.91,
|
| 1692 |
+
"learning_rate": 2.8401414034877683e-05,
|
| 1693 |
+
"loss": 0.7501,
|
| 1694 |
+
"step": 2810
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 0.92,
|
| 1698 |
+
"learning_rate": 2.827476832676247e-05,
|
| 1699 |
+
"loss": 0.7397,
|
| 1700 |
+
"step": 2820
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 0.92,
|
| 1704 |
+
"learning_rate": 2.814803705401874e-05,
|
| 1705 |
+
"loss": 0.7471,
|
| 1706 |
+
"step": 2830
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 0.92,
|
| 1710 |
+
"learning_rate": 2.8021223527938524e-05,
|
| 1711 |
+
"loss": 0.7494,
|
| 1712 |
+
"step": 2840
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 0.93,
|
| 1716 |
+
"learning_rate": 2.7894331061963004e-05,
|
| 1717 |
+
"loss": 0.745,
|
| 1718 |
+
"step": 2850
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.93,
|
| 1722 |
+
"learning_rate": 2.7767362971595946e-05,
|
| 1723 |
+
"loss": 0.7432,
|
| 1724 |
+
"step": 2860
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.93,
|
| 1728 |
+
"learning_rate": 2.7640322574317063e-05,
|
| 1729 |
+
"loss": 0.7431,
|
| 1730 |
+
"step": 2870
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 0.94,
|
| 1734 |
+
"learning_rate": 2.751321318949533e-05,
|
| 1735 |
+
"loss": 0.7471,
|
| 1736 |
+
"step": 2880
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 0.94,
|
| 1740 |
+
"learning_rate": 2.738603813830227e-05,
|
| 1741 |
+
"loss": 0.7417,
|
| 1742 |
+
"step": 2890
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 0.94,
|
| 1746 |
+
"learning_rate": 2.7258800743625167e-05,
|
| 1747 |
+
"loss": 0.7411,
|
| 1748 |
+
"step": 2900
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 0.95,
|
| 1752 |
+
"learning_rate": 2.7131504329980235e-05,
|
| 1753 |
+
"loss": 0.7546,
|
| 1754 |
+
"step": 2910
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 0.95,
|
| 1758 |
+
"learning_rate": 2.7004152223425753e-05,
|
| 1759 |
+
"loss": 0.7368,
|
| 1760 |
+
"step": 2920
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 0.95,
|
| 1764 |
+
"learning_rate": 2.687674775147519e-05,
|
| 1765 |
+
"loss": 0.7447,
|
| 1766 |
+
"step": 2930
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.96,
|
| 1770 |
+
"learning_rate": 2.6749294243010237e-05,
|
| 1771 |
+
"loss": 0.7364,
|
| 1772 |
+
"step": 2940
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 0.96,
|
| 1776 |
+
"learning_rate": 2.6621795028193837e-05,
|
| 1777 |
+
"loss": 0.7417,
|
| 1778 |
+
"step": 2950
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 0.96,
|
| 1782 |
+
"learning_rate": 2.649425343838314e-05,
|
| 1783 |
+
"loss": 0.7371,
|
| 1784 |
+
"step": 2960
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 0.97,
|
| 1788 |
+
"learning_rate": 2.6366672806042532e-05,
|
| 1789 |
+
"loss": 0.7452,
|
| 1790 |
+
"step": 2970
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 0.97,
|
| 1794 |
+
"learning_rate": 2.6239056464656485e-05,
|
| 1795 |
+
"loss": 0.7427,
|
| 1796 |
+
"step": 2980
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 0.97,
|
| 1800 |
+
"learning_rate": 2.611140774864252e-05,
|
| 1801 |
+
"loss": 0.7381,
|
| 1802 |
+
"step": 2990
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 0.98,
|
| 1806 |
+
"learning_rate": 2.5983729993264034e-05,
|
| 1807 |
+
"loss": 0.7476,
|
| 1808 |
+
"step": 3000
|
| 1809 |
+
}
|
| 1810 |
+
],
|
| 1811 |
+
"max_steps": 6146,
|
| 1812 |
+
"num_train_epochs": 2,
|
| 1813 |
+
"total_flos": 5.387941012888879e+18,
|
| 1814 |
+
"trial_name": null,
|
| 1815 |
+
"trial_params": null
|
| 1816 |
+
}
|
checkpoint-3000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dafdcf787495b020f1b3f6f6180b816b2753ea01048ce4086dbcd8a2c7a5f666
|
| 3 |
+
size 3265
|
checkpoint-4000/README.md
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
| 4 |
+
## Training procedure
|
| 5 |
+
|
| 6 |
+
|
| 7 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 8 |
+
- load_in_8bit: False
|
| 9 |
+
- load_in_4bit: True
|
| 10 |
+
- llm_int8_threshold: 6.0
|
| 11 |
+
- llm_int8_skip_modules: None
|
| 12 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 13 |
+
- llm_int8_has_fp16_weight: False
|
| 14 |
+
- bnb_4bit_quant_type: nf4
|
| 15 |
+
- bnb_4bit_use_double_quant: True
|
| 16 |
+
- bnb_4bit_compute_dtype: float16
|
| 17 |
+
### Framework versions
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
- PEFT 0.4.0
|
checkpoint-4000/adapter_config.json
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "./Llama-2-7b-chat-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"fan_in_fan_out": false,
|
| 6 |
+
"inference_mode": true,
|
| 7 |
+
"init_lora_weights": true,
|
| 8 |
+
"layers_pattern": null,
|
| 9 |
+
"layers_to_transform": null,
|
| 10 |
+
"lora_alpha": 32.0,
|
| 11 |
+
"lora_dropout": 0.1,
|
| 12 |
+
"modules_to_save": null,
|
| 13 |
+
"peft_type": "LORA",
|
| 14 |
+
"r": 8,
|
| 15 |
+
"revision": null,
|
| 16 |
+
"target_modules": [
|
| 17 |
+
"q_proj",
|
| 18 |
+
"v_proj"
|
| 19 |
+
],
|
| 20 |
+
"task_type": "CAUSAL_LM"
|
| 21 |
+
}
|
checkpoint-4000/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a5020df096936bc791c4ce017d5482982056ac801a7b83801f5d3cb19e4c888d
|
| 3 |
+
size 16821197
|
checkpoint-4000/finetuning_args.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"dpo_beta": 0.1,
|
| 3 |
+
"finetuning_type": "lora",
|
| 4 |
+
"lora_alpha": 32.0,
|
| 5 |
+
"lora_dropout": 0.1,
|
| 6 |
+
"lora_rank": 8,
|
| 7 |
+
"lora_target": [
|
| 8 |
+
"q_proj",
|
| 9 |
+
"v_proj"
|
| 10 |
+
],
|
| 11 |
+
"name_module_trainable": "mlp",
|
| 12 |
+
"num_hidden_layers": 32,
|
| 13 |
+
"num_layer_trainable": 3,
|
| 14 |
+
"ppo_score_norm": false,
|
| 15 |
+
"resume_lora_training": true
|
| 16 |
+
}
|
checkpoint-4000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:335fe6199a223e41d3b27400abb9edb1998acba8682256e989955cef4c60b81f
|
| 3 |
+
size 33661637
|
checkpoint-4000/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bc5c23ded368d76d795cbf4d4c132c70c3481b5f00f0109f38bc9f3105052c4c
|
| 3 |
+
size 18679
|
checkpoint-4000/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:880665380bf5cf710015454edca64bd385b756c1bf5504147701c5e158223240
|
| 3 |
+
size 18679
|
checkpoint-4000/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8a7dbae85d9036c69d7fcb3d58d9f928166451c8e4c732cf71c6398355c53e06
|
| 3 |
+
size 18679
|
checkpoint-4000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8890d0af5c7e558d49b021915b5078d03193033b4d70a7f84f7bc7c08e3fe5c
|
| 3 |
+
size 627
|
checkpoint-4000/trainer_state.json
ADDED
|
@@ -0,0 +1,2416 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.3013420089467262,
|
| 5 |
+
"global_step": 4000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 4.999967339434433e-05,
|
| 13 |
+
"loss": 1.343,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 4.9998820960283476e-05,
|
| 19 |
+
"loss": 1.1858,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 4.99974394497948e-05,
|
| 25 |
+
"loss": 1.1057,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.01,
|
| 30 |
+
"learning_rate": 4.9995528892116685e-05,
|
| 31 |
+
"loss": 1.0698,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.02,
|
| 36 |
+
"learning_rate": 4.999308932768425e-05,
|
| 37 |
+
"loss": 1.0316,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.02,
|
| 42 |
+
"learning_rate": 4.99897583236926e-05,
|
| 43 |
+
"loss": 1.0046,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.02,
|
| 48 |
+
"learning_rate": 4.9985774375988764e-05,
|
| 49 |
+
"loss": 0.9848,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.03,
|
| 54 |
+
"learning_rate": 4.998113758866712e-05,
|
| 55 |
+
"loss": 0.9656,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.03,
|
| 60 |
+
"learning_rate": 4.997584808287975e-05,
|
| 61 |
+
"loss": 0.9648,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.03,
|
| 66 |
+
"learning_rate": 4.996990599683325e-05,
|
| 67 |
+
"loss": 0.9415,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.04,
|
| 72 |
+
"learning_rate": 4.996331148578514e-05,
|
| 73 |
+
"loss": 0.9388,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.04,
|
| 78 |
+
"learning_rate": 4.995606472203979e-05,
|
| 79 |
+
"loss": 0.9289,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.04,
|
| 84 |
+
"learning_rate": 4.994816589494391e-05,
|
| 85 |
+
"loss": 0.9252,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.05,
|
| 90 |
+
"learning_rate": 4.993961521088163e-05,
|
| 91 |
+
"loss": 0.91,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.05,
|
| 96 |
+
"learning_rate": 4.9930412893269106e-05,
|
| 97 |
+
"loss": 0.9033,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.05,
|
| 102 |
+
"learning_rate": 4.992055918254865e-05,
|
| 103 |
+
"loss": 0.9064,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.06,
|
| 108 |
+
"learning_rate": 4.9910054336182467e-05,
|
| 109 |
+
"loss": 0.8946,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.06,
|
| 114 |
+
"learning_rate": 4.989889862864594e-05,
|
| 115 |
+
"loss": 0.8914,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.06,
|
| 120 |
+
"learning_rate": 4.988709235142044e-05,
|
| 121 |
+
"loss": 0.9007,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.07,
|
| 126 |
+
"learning_rate": 4.9874635812985734e-05,
|
| 127 |
+
"loss": 0.9008,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.07,
|
| 132 |
+
"learning_rate": 4.9861529338811874e-05,
|
| 133 |
+
"loss": 0.8714,
|
| 134 |
+
"step": 210
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.07,
|
| 138 |
+
"learning_rate": 4.984777327135076e-05,
|
| 139 |
+
"loss": 0.8865,
|
| 140 |
+
"step": 220
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.07,
|
| 144 |
+
"learning_rate": 4.9833367970027145e-05,
|
| 145 |
+
"loss": 0.8716,
|
| 146 |
+
"step": 230
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.08,
|
| 150 |
+
"learning_rate": 4.981831381122926e-05,
|
| 151 |
+
"loss": 0.8817,
|
| 152 |
+
"step": 240
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.08,
|
| 156 |
+
"learning_rate": 4.980261118829896e-05,
|
| 157 |
+
"loss": 0.8658,
|
| 158 |
+
"step": 250
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.08,
|
| 162 |
+
"learning_rate": 4.978626051152152e-05,
|
| 163 |
+
"loss": 0.8722,
|
| 164 |
+
"step": 260
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.09,
|
| 168 |
+
"learning_rate": 4.976926220811479e-05,
|
| 169 |
+
"loss": 0.8541,
|
| 170 |
+
"step": 270
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.09,
|
| 174 |
+
"learning_rate": 4.975161672221813e-05,
|
| 175 |
+
"loss": 0.8504,
|
| 176 |
+
"step": 280
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.09,
|
| 180 |
+
"learning_rate": 4.97333245148808e-05,
|
| 181 |
+
"loss": 0.8467,
|
| 182 |
+
"step": 290
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.1,
|
| 186 |
+
"learning_rate": 4.971438606404986e-05,
|
| 187 |
+
"loss": 0.8533,
|
| 188 |
+
"step": 300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.1,
|
| 192 |
+
"learning_rate": 4.969480186455771e-05,
|
| 193 |
+
"loss": 0.853,
|
| 194 |
+
"step": 310
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.1,
|
| 198 |
+
"learning_rate": 4.9674572428109205e-05,
|
| 199 |
+
"loss": 0.8432,
|
| 200 |
+
"step": 320
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.11,
|
| 204 |
+
"learning_rate": 4.9653698283268183e-05,
|
| 205 |
+
"loss": 0.854,
|
| 206 |
+
"step": 330
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.11,
|
| 210 |
+
"learning_rate": 4.963217997544376e-05,
|
| 211 |
+
"loss": 0.8435,
|
| 212 |
+
"step": 340
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.11,
|
| 216 |
+
"learning_rate": 4.961001806687602e-05,
|
| 217 |
+
"loss": 0.8518,
|
| 218 |
+
"step": 350
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.12,
|
| 222 |
+
"learning_rate": 4.958721313662133e-05,
|
| 223 |
+
"loss": 0.848,
|
| 224 |
+
"step": 360
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.12,
|
| 228 |
+
"learning_rate": 4.956376578053723e-05,
|
| 229 |
+
"loss": 0.8436,
|
| 230 |
+
"step": 370
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.12,
|
| 234 |
+
"learning_rate": 4.9539676611266846e-05,
|
| 235 |
+
"loss": 0.8405,
|
| 236 |
+
"step": 380
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.13,
|
| 240 |
+
"learning_rate": 4.9514946258222895e-05,
|
| 241 |
+
"loss": 0.838,
|
| 242 |
+
"step": 390
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.13,
|
| 246 |
+
"learning_rate": 4.948957536757123e-05,
|
| 247 |
+
"loss": 0.8458,
|
| 248 |
+
"step": 400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.13,
|
| 252 |
+
"learning_rate": 4.946356460221395e-05,
|
| 253 |
+
"loss": 0.8379,
|
| 254 |
+
"step": 410
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.14,
|
| 258 |
+
"learning_rate": 4.943691464177212e-05,
|
| 259 |
+
"loss": 0.827,
|
| 260 |
+
"step": 420
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.14,
|
| 264 |
+
"learning_rate": 4.940962618256796e-05,
|
| 265 |
+
"loss": 0.8208,
|
| 266 |
+
"step": 430
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.14,
|
| 270 |
+
"learning_rate": 4.9381699937606674e-05,
|
| 271 |
+
"loss": 0.8242,
|
| 272 |
+
"step": 440
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.15,
|
| 276 |
+
"learning_rate": 4.935313663655782e-05,
|
| 277 |
+
"loss": 0.8362,
|
| 278 |
+
"step": 450
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.15,
|
| 282 |
+
"learning_rate": 4.932393702573627e-05,
|
| 283 |
+
"loss": 0.8363,
|
| 284 |
+
"step": 460
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.15,
|
| 288 |
+
"learning_rate": 4.929410186808264e-05,
|
| 289 |
+
"loss": 0.8275,
|
| 290 |
+
"step": 470
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.16,
|
| 294 |
+
"learning_rate": 4.926363194314344e-05,
|
| 295 |
+
"loss": 0.821,
|
| 296 |
+
"step": 480
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.16,
|
| 300 |
+
"learning_rate": 4.923252804705068e-05,
|
| 301 |
+
"loss": 0.8271,
|
| 302 |
+
"step": 490
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.16,
|
| 306 |
+
"learning_rate": 4.920079099250099e-05,
|
| 307 |
+
"loss": 0.8178,
|
| 308 |
+
"step": 500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.17,
|
| 312 |
+
"learning_rate": 4.916842160873451e-05,
|
| 313 |
+
"loss": 0.8293,
|
| 314 |
+
"step": 510
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.17,
|
| 318 |
+
"learning_rate": 4.913542074151315e-05,
|
| 319 |
+
"loss": 0.8294,
|
| 320 |
+
"step": 520
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.17,
|
| 324 |
+
"learning_rate": 4.910178925309849e-05,
|
| 325 |
+
"loss": 0.8202,
|
| 326 |
+
"step": 530
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.18,
|
| 330 |
+
"learning_rate": 4.906752802222928e-05,
|
| 331 |
+
"loss": 0.8301,
|
| 332 |
+
"step": 540
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.18,
|
| 336 |
+
"learning_rate": 4.903263794409846e-05,
|
| 337 |
+
"loss": 0.8031,
|
| 338 |
+
"step": 550
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.18,
|
| 342 |
+
"learning_rate": 4.8997119930329785e-05,
|
| 343 |
+
"loss": 0.821,
|
| 344 |
+
"step": 560
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.19,
|
| 348 |
+
"learning_rate": 4.896097490895398e-05,
|
| 349 |
+
"loss": 0.8132,
|
| 350 |
+
"step": 570
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.19,
|
| 354 |
+
"learning_rate": 4.8924203824384516e-05,
|
| 355 |
+
"loss": 0.8113,
|
| 356 |
+
"step": 580
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.19,
|
| 360 |
+
"learning_rate": 4.888680763739294e-05,
|
| 361 |
+
"loss": 0.8199,
|
| 362 |
+
"step": 590
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.2,
|
| 366 |
+
"learning_rate": 4.8848787325083734e-05,
|
| 367 |
+
"loss": 0.8196,
|
| 368 |
+
"step": 600
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.2,
|
| 372 |
+
"learning_rate": 4.8810143880868816e-05,
|
| 373 |
+
"loss": 0.8128,
|
| 374 |
+
"step": 610
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.2,
|
| 378 |
+
"learning_rate": 4.8770878314441606e-05,
|
| 379 |
+
"loss": 0.8123,
|
| 380 |
+
"step": 620
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.2,
|
| 384 |
+
"learning_rate": 4.8730991651750565e-05,
|
| 385 |
+
"loss": 0.8048,
|
| 386 |
+
"step": 630
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.21,
|
| 390 |
+
"learning_rate": 4.869048493497247e-05,
|
| 391 |
+
"loss": 0.8119,
|
| 392 |
+
"step": 640
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.21,
|
| 396 |
+
"learning_rate": 4.864935922248515e-05,
|
| 397 |
+
"loss": 0.8141,
|
| 398 |
+
"step": 650
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.21,
|
| 402 |
+
"learning_rate": 4.860761558883981e-05,
|
| 403 |
+
"loss": 0.8,
|
| 404 |
+
"step": 660
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.22,
|
| 408 |
+
"learning_rate": 4.856525512473302e-05,
|
| 409 |
+
"loss": 0.816,
|
| 410 |
+
"step": 670
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.22,
|
| 414 |
+
"learning_rate": 4.852227893697814e-05,
|
| 415 |
+
"loss": 0.7976,
|
| 416 |
+
"step": 680
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.22,
|
| 420 |
+
"learning_rate": 4.847868814847646e-05,
|
| 421 |
+
"loss": 0.8121,
|
| 422 |
+
"step": 690
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.23,
|
| 426 |
+
"learning_rate": 4.84344838981878e-05,
|
| 427 |
+
"loss": 0.8175,
|
| 428 |
+
"step": 700
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.23,
|
| 432 |
+
"learning_rate": 4.8389667341100844e-05,
|
| 433 |
+
"loss": 0.8109,
|
| 434 |
+
"step": 710
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.23,
|
| 438 |
+
"learning_rate": 4.834423964820284e-05,
|
| 439 |
+
"loss": 0.7948,
|
| 440 |
+
"step": 720
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.24,
|
| 444 |
+
"learning_rate": 4.8298202006449136e-05,
|
| 445 |
+
"loss": 0.7986,
|
| 446 |
+
"step": 730
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.24,
|
| 450 |
+
"learning_rate": 4.8251555618732036e-05,
|
| 451 |
+
"loss": 0.8049,
|
| 452 |
+
"step": 740
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.24,
|
| 456 |
+
"learning_rate": 4.820430170384947e-05,
|
| 457 |
+
"loss": 0.7981,
|
| 458 |
+
"step": 750
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.25,
|
| 462 |
+
"learning_rate": 4.8156441496473126e-05,
|
| 463 |
+
"loss": 0.799,
|
| 464 |
+
"step": 760
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.25,
|
| 468 |
+
"learning_rate": 4.8107976247116145e-05,
|
| 469 |
+
"loss": 0.804,
|
| 470 |
+
"step": 770
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.25,
|
| 474 |
+
"learning_rate": 4.805890722210048e-05,
|
| 475 |
+
"loss": 0.8053,
|
| 476 |
+
"step": 780
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.26,
|
| 480 |
+
"learning_rate": 4.8009235703523824e-05,
|
| 481 |
+
"loss": 0.7969,
|
| 482 |
+
"step": 790
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.26,
|
| 486 |
+
"learning_rate": 4.79589629892261e-05,
|
| 487 |
+
"loss": 0.8051,
|
| 488 |
+
"step": 800
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.26,
|
| 492 |
+
"learning_rate": 4.7908090392755526e-05,
|
| 493 |
+
"loss": 0.796,
|
| 494 |
+
"step": 810
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.27,
|
| 498 |
+
"learning_rate": 4.785661924333431e-05,
|
| 499 |
+
"loss": 0.7928,
|
| 500 |
+
"step": 820
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.27,
|
| 504 |
+
"learning_rate": 4.780455088582394e-05,
|
| 505 |
+
"loss": 0.8017,
|
| 506 |
+
"step": 830
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.27,
|
| 510 |
+
"learning_rate": 4.775188668069002e-05,
|
| 511 |
+
"loss": 0.7969,
|
| 512 |
+
"step": 840
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.28,
|
| 516 |
+
"learning_rate": 4.769862800396674e-05,
|
| 517 |
+
"loss": 0.7942,
|
| 518 |
+
"step": 850
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.28,
|
| 522 |
+
"learning_rate": 4.764477624722088e-05,
|
| 523 |
+
"loss": 0.7877,
|
| 524 |
+
"step": 860
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.28,
|
| 528 |
+
"learning_rate": 4.759033281751553e-05,
|
| 529 |
+
"loss": 0.7979,
|
| 530 |
+
"step": 870
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.29,
|
| 534 |
+
"learning_rate": 4.7535299137373236e-05,
|
| 535 |
+
"loss": 0.7986,
|
| 536 |
+
"step": 880
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.29,
|
| 540 |
+
"learning_rate": 4.7479676644738894e-05,
|
| 541 |
+
"loss": 0.8124,
|
| 542 |
+
"step": 890
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.29,
|
| 546 |
+
"learning_rate": 4.742346679294217e-05,
|
| 547 |
+
"loss": 0.7959,
|
| 548 |
+
"step": 900
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.3,
|
| 552 |
+
"learning_rate": 4.7366671050659505e-05,
|
| 553 |
+
"loss": 0.7884,
|
| 554 |
+
"step": 910
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.3,
|
| 558 |
+
"learning_rate": 4.730929090187573e-05,
|
| 559 |
+
"loss": 0.7911,
|
| 560 |
+
"step": 920
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.3,
|
| 564 |
+
"learning_rate": 4.725132784584536e-05,
|
| 565 |
+
"loss": 0.7937,
|
| 566 |
+
"step": 930
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.31,
|
| 570 |
+
"learning_rate": 4.719278339705333e-05,
|
| 571 |
+
"loss": 0.7994,
|
| 572 |
+
"step": 940
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.31,
|
| 576 |
+
"learning_rate": 4.71336590851755e-05,
|
| 577 |
+
"loss": 0.7905,
|
| 578 |
+
"step": 950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.31,
|
| 582 |
+
"learning_rate": 4.707395645503863e-05,
|
| 583 |
+
"loss": 0.7924,
|
| 584 |
+
"step": 960
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.32,
|
| 588 |
+
"learning_rate": 4.701367706658006e-05,
|
| 589 |
+
"loss": 0.7891,
|
| 590 |
+
"step": 970
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.32,
|
| 594 |
+
"learning_rate": 4.695282249480692e-05,
|
| 595 |
+
"loss": 0.7962,
|
| 596 |
+
"step": 980
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.32,
|
| 600 |
+
"learning_rate": 4.6891394329755e-05,
|
| 601 |
+
"loss": 0.7882,
|
| 602 |
+
"step": 990
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.33,
|
| 606 |
+
"learning_rate": 4.6829394176447195e-05,
|
| 607 |
+
"loss": 0.8011,
|
| 608 |
+
"step": 1000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.33,
|
| 612 |
+
"learning_rate": 4.676682365485154e-05,
|
| 613 |
+
"loss": 0.7955,
|
| 614 |
+
"step": 1010
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.33,
|
| 618 |
+
"learning_rate": 4.670368439983896e-05,
|
| 619 |
+
"loss": 0.7906,
|
| 620 |
+
"step": 1020
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.34,
|
| 624 |
+
"learning_rate": 4.663997806114048e-05,
|
| 625 |
+
"loss": 0.7853,
|
| 626 |
+
"step": 1030
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.34,
|
| 630 |
+
"learning_rate": 4.657570630330411e-05,
|
| 631 |
+
"loss": 0.7825,
|
| 632 |
+
"step": 1040
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.34,
|
| 636 |
+
"learning_rate": 4.6510870805651444e-05,
|
| 637 |
+
"loss": 0.7768,
|
| 638 |
+
"step": 1050
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.34,
|
| 642 |
+
"learning_rate": 4.6445473262233696e-05,
|
| 643 |
+
"loss": 0.7795,
|
| 644 |
+
"step": 1060
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.35,
|
| 648 |
+
"learning_rate": 4.637951538178746e-05,
|
| 649 |
+
"loss": 0.7832,
|
| 650 |
+
"step": 1070
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.35,
|
| 654 |
+
"learning_rate": 4.63129988876901e-05,
|
| 655 |
+
"loss": 0.7934,
|
| 656 |
+
"step": 1080
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.35,
|
| 660 |
+
"learning_rate": 4.624592551791465e-05,
|
| 661 |
+
"loss": 0.7892,
|
| 662 |
+
"step": 1090
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.36,
|
| 666 |
+
"learning_rate": 4.617829702498446e-05,
|
| 667 |
+
"loss": 0.7819,
|
| 668 |
+
"step": 1100
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.36,
|
| 672 |
+
"learning_rate": 4.611011517592741e-05,
|
| 673 |
+
"loss": 0.7868,
|
| 674 |
+
"step": 1110
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.36,
|
| 678 |
+
"learning_rate": 4.604138175222968e-05,
|
| 679 |
+
"loss": 0.7914,
|
| 680 |
+
"step": 1120
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.37,
|
| 684 |
+
"learning_rate": 4.597209854978929e-05,
|
| 685 |
+
"loss": 0.7817,
|
| 686 |
+
"step": 1130
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 0.37,
|
| 690 |
+
"learning_rate": 4.590226737886908e-05,
|
| 691 |
+
"loss": 0.7916,
|
| 692 |
+
"step": 1140
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.37,
|
| 696 |
+
"learning_rate": 4.583189006404949e-05,
|
| 697 |
+
"loss": 0.7799,
|
| 698 |
+
"step": 1150
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.38,
|
| 702 |
+
"learning_rate": 4.576096844418083e-05,
|
| 703 |
+
"loss": 0.7776,
|
| 704 |
+
"step": 1160
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.38,
|
| 708 |
+
"learning_rate": 4.568950437233529e-05,
|
| 709 |
+
"loss": 0.7803,
|
| 710 |
+
"step": 1170
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.38,
|
| 714 |
+
"learning_rate": 4.561749971575846e-05,
|
| 715 |
+
"loss": 0.7884,
|
| 716 |
+
"step": 1180
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.39,
|
| 720 |
+
"learning_rate": 4.5544956355820596e-05,
|
| 721 |
+
"loss": 0.7827,
|
| 722 |
+
"step": 1190
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.39,
|
| 726 |
+
"learning_rate": 4.547187618796742e-05,
|
| 727 |
+
"loss": 0.7857,
|
| 728 |
+
"step": 1200
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 0.39,
|
| 732 |
+
"learning_rate": 4.5398261121670634e-05,
|
| 733 |
+
"loss": 0.7797,
|
| 734 |
+
"step": 1210
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.4,
|
| 738 |
+
"learning_rate": 4.532411308037798e-05,
|
| 739 |
+
"loss": 0.7821,
|
| 740 |
+
"step": 1220
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.4,
|
| 744 |
+
"learning_rate": 4.524943400146306e-05,
|
| 745 |
+
"loss": 0.7747,
|
| 746 |
+
"step": 1230
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.4,
|
| 750 |
+
"learning_rate": 4.51742258361746e-05,
|
| 751 |
+
"loss": 0.7737,
|
| 752 |
+
"step": 1240
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.41,
|
| 756 |
+
"learning_rate": 4.509849054958559e-05,
|
| 757 |
+
"loss": 0.7868,
|
| 758 |
+
"step": 1250
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.41,
|
| 762 |
+
"learning_rate": 4.502223012054186e-05,
|
| 763 |
+
"loss": 0.7818,
|
| 764 |
+
"step": 1260
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.41,
|
| 768 |
+
"learning_rate": 4.4945446541610414e-05,
|
| 769 |
+
"loss": 0.7804,
|
| 770 |
+
"step": 1270
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 0.42,
|
| 774 |
+
"learning_rate": 4.486814181902734e-05,
|
| 775 |
+
"loss": 0.7759,
|
| 776 |
+
"step": 1280
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.42,
|
| 780 |
+
"learning_rate": 4.4790317972645396e-05,
|
| 781 |
+
"loss": 0.7744,
|
| 782 |
+
"step": 1290
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.42,
|
| 786 |
+
"learning_rate": 4.4711977035881256e-05,
|
| 787 |
+
"loss": 0.7814,
|
| 788 |
+
"step": 1300
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.43,
|
| 792 |
+
"learning_rate": 4.463312105566237e-05,
|
| 793 |
+
"loss": 0.7611,
|
| 794 |
+
"step": 1310
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.43,
|
| 798 |
+
"learning_rate": 4.4553752092373465e-05,
|
| 799 |
+
"loss": 0.7796,
|
| 800 |
+
"step": 1320
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.43,
|
| 804 |
+
"learning_rate": 4.447387221980272e-05,
|
| 805 |
+
"loss": 0.7839,
|
| 806 |
+
"step": 1330
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.44,
|
| 810 |
+
"learning_rate": 4.4393483525087584e-05,
|
| 811 |
+
"loss": 0.7726,
|
| 812 |
+
"step": 1340
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.44,
|
| 816 |
+
"learning_rate": 4.4312588108660246e-05,
|
| 817 |
+
"loss": 0.7763,
|
| 818 |
+
"step": 1350
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.44,
|
| 822 |
+
"learning_rate": 4.423118808419276e-05,
|
| 823 |
+
"loss": 0.766,
|
| 824 |
+
"step": 1360
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.45,
|
| 828 |
+
"learning_rate": 4.4149285578541785e-05,
|
| 829 |
+
"loss": 0.7887,
|
| 830 |
+
"step": 1370
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.45,
|
| 834 |
+
"learning_rate": 4.406688273169305e-05,
|
| 835 |
+
"loss": 0.7637,
|
| 836 |
+
"step": 1380
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.45,
|
| 840 |
+
"learning_rate": 4.3983981696705415e-05,
|
| 841 |
+
"loss": 0.7609,
|
| 842 |
+
"step": 1390
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.46,
|
| 846 |
+
"learning_rate": 4.390058463965464e-05,
|
| 847 |
+
"loss": 0.7729,
|
| 848 |
+
"step": 1400
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.46,
|
| 852 |
+
"learning_rate": 4.381669373957675e-05,
|
| 853 |
+
"loss": 0.7751,
|
| 854 |
+
"step": 1410
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.46,
|
| 858 |
+
"learning_rate": 4.3732311188411154e-05,
|
| 859 |
+
"loss": 0.7877,
|
| 860 |
+
"step": 1420
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.47,
|
| 864 |
+
"learning_rate": 4.3647439190943315e-05,
|
| 865 |
+
"loss": 0.7738,
|
| 866 |
+
"step": 1430
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.47,
|
| 870 |
+
"learning_rate": 4.3562079964747207e-05,
|
| 871 |
+
"loss": 0.7834,
|
| 872 |
+
"step": 1440
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.47,
|
| 876 |
+
"learning_rate": 4.347623574012728e-05,
|
| 877 |
+
"loss": 0.7835,
|
| 878 |
+
"step": 1450
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.47,
|
| 882 |
+
"learning_rate": 4.338990876006031e-05,
|
| 883 |
+
"loss": 0.7707,
|
| 884 |
+
"step": 1460
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.48,
|
| 888 |
+
"learning_rate": 4.330310128013666e-05,
|
| 889 |
+
"loss": 0.7674,
|
| 890 |
+
"step": 1470
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.48,
|
| 894 |
+
"learning_rate": 4.321581556850146e-05,
|
| 895 |
+
"loss": 0.7687,
|
| 896 |
+
"step": 1480
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.48,
|
| 900 |
+
"learning_rate": 4.312805390579527e-05,
|
| 901 |
+
"loss": 0.7663,
|
| 902 |
+
"step": 1490
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.49,
|
| 906 |
+
"learning_rate": 4.3039818585094524e-05,
|
| 907 |
+
"loss": 0.7662,
|
| 908 |
+
"step": 1500
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.49,
|
| 912 |
+
"learning_rate": 4.29511119118516e-05,
|
| 913 |
+
"loss": 0.7688,
|
| 914 |
+
"step": 1510
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.49,
|
| 918 |
+
"learning_rate": 4.28619362038346e-05,
|
| 919 |
+
"loss": 0.7652,
|
| 920 |
+
"step": 1520
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.5,
|
| 924 |
+
"learning_rate": 4.277229379106676e-05,
|
| 925 |
+
"loss": 0.7683,
|
| 926 |
+
"step": 1530
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.5,
|
| 930 |
+
"learning_rate": 4.2682187015765615e-05,
|
| 931 |
+
"loss": 0.7621,
|
| 932 |
+
"step": 1540
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.5,
|
| 936 |
+
"learning_rate": 4.259161823228174e-05,
|
| 937 |
+
"loss": 0.7778,
|
| 938 |
+
"step": 1550
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 0.51,
|
| 942 |
+
"learning_rate": 4.2500589807037306e-05,
|
| 943 |
+
"loss": 0.7809,
|
| 944 |
+
"step": 1560
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 0.51,
|
| 948 |
+
"learning_rate": 4.240910411846418e-05,
|
| 949 |
+
"loss": 0.7717,
|
| 950 |
+
"step": 1570
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.51,
|
| 954 |
+
"learning_rate": 4.231716355694184e-05,
|
| 955 |
+
"loss": 0.7584,
|
| 956 |
+
"step": 1580
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.52,
|
| 960 |
+
"learning_rate": 4.222477052473486e-05,
|
| 961 |
+
"loss": 0.7631,
|
| 962 |
+
"step": 1590
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.52,
|
| 966 |
+
"learning_rate": 4.21319274359302e-05,
|
| 967 |
+
"loss": 0.7681,
|
| 968 |
+
"step": 1600
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.52,
|
| 972 |
+
"learning_rate": 4.203863671637409e-05,
|
| 973 |
+
"loss": 0.7739,
|
| 974 |
+
"step": 1610
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 0.53,
|
| 978 |
+
"learning_rate": 4.194490080360866e-05,
|
| 979 |
+
"loss": 0.7646,
|
| 980 |
+
"step": 1620
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.53,
|
| 984 |
+
"learning_rate": 4.1850722146808254e-05,
|
| 985 |
+
"loss": 0.7647,
|
| 986 |
+
"step": 1630
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.53,
|
| 990 |
+
"learning_rate": 4.1756103206715416e-05,
|
| 991 |
+
"loss": 0.7675,
|
| 992 |
+
"step": 1640
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.54,
|
| 996 |
+
"learning_rate": 4.166104645557662e-05,
|
| 997 |
+
"loss": 0.7652,
|
| 998 |
+
"step": 1650
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 0.54,
|
| 1002 |
+
"learning_rate": 4.15655543770777e-05,
|
| 1003 |
+
"loss": 0.765,
|
| 1004 |
+
"step": 1660
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.54,
|
| 1008 |
+
"learning_rate": 4.146962946627886e-05,
|
| 1009 |
+
"loss": 0.7614,
|
| 1010 |
+
"step": 1670
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.55,
|
| 1014 |
+
"learning_rate": 4.137327422954958e-05,
|
| 1015 |
+
"loss": 0.7544,
|
| 1016 |
+
"step": 1680
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 0.55,
|
| 1020 |
+
"learning_rate": 4.127649118450308e-05,
|
| 1021 |
+
"loss": 0.7673,
|
| 1022 |
+
"step": 1690
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 0.55,
|
| 1026 |
+
"learning_rate": 4.117928285993055e-05,
|
| 1027 |
+
"loss": 0.7593,
|
| 1028 |
+
"step": 1700
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 0.56,
|
| 1032 |
+
"learning_rate": 4.108165179573508e-05,
|
| 1033 |
+
"loss": 0.7735,
|
| 1034 |
+
"step": 1710
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.56,
|
| 1038 |
+
"learning_rate": 4.0983600542865285e-05,
|
| 1039 |
+
"loss": 0.7666,
|
| 1040 |
+
"step": 1720
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.56,
|
| 1044 |
+
"learning_rate": 4.088513166324867e-05,
|
| 1045 |
+
"loss": 0.7627,
|
| 1046 |
+
"step": 1730
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.57,
|
| 1050 |
+
"learning_rate": 4.078624772972467e-05,
|
| 1051 |
+
"loss": 0.77,
|
| 1052 |
+
"step": 1740
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.57,
|
| 1056 |
+
"learning_rate": 4.068695132597744e-05,
|
| 1057 |
+
"loss": 0.7569,
|
| 1058 |
+
"step": 1750
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 0.57,
|
| 1062 |
+
"learning_rate": 4.058724504646834e-05,
|
| 1063 |
+
"loss": 0.7654,
|
| 1064 |
+
"step": 1760
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 0.58,
|
| 1068 |
+
"learning_rate": 4.048713149636816e-05,
|
| 1069 |
+
"loss": 0.753,
|
| 1070 |
+
"step": 1770
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.58,
|
| 1074 |
+
"learning_rate": 4.038661329148904e-05,
|
| 1075 |
+
"loss": 0.7705,
|
| 1076 |
+
"step": 1780
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.58,
|
| 1080 |
+
"learning_rate": 4.02856930582161e-05,
|
| 1081 |
+
"loss": 0.7542,
|
| 1082 |
+
"step": 1790
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.59,
|
| 1086 |
+
"learning_rate": 4.018437343343886e-05,
|
| 1087 |
+
"loss": 0.7532,
|
| 1088 |
+
"step": 1800
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.59,
|
| 1092 |
+
"learning_rate": 4.008265706448234e-05,
|
| 1093 |
+
"loss": 0.7662,
|
| 1094 |
+
"step": 1810
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.59,
|
| 1098 |
+
"learning_rate": 3.9980546609037825e-05,
|
| 1099 |
+
"loss": 0.7607,
|
| 1100 |
+
"step": 1820
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 0.6,
|
| 1104 |
+
"learning_rate": 3.9878044735093505e-05,
|
| 1105 |
+
"loss": 0.7552,
|
| 1106 |
+
"step": 1830
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 0.6,
|
| 1110 |
+
"learning_rate": 3.977515412086472e-05,
|
| 1111 |
+
"loss": 0.7645,
|
| 1112 |
+
"step": 1840
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.6,
|
| 1116 |
+
"learning_rate": 3.9671877454723985e-05,
|
| 1117 |
+
"loss": 0.7689,
|
| 1118 |
+
"step": 1850
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.61,
|
| 1122 |
+
"learning_rate": 3.956821743513078e-05,
|
| 1123 |
+
"loss": 0.7615,
|
| 1124 |
+
"step": 1860
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.61,
|
| 1128 |
+
"learning_rate": 3.946417677056097e-05,
|
| 1129 |
+
"loss": 0.7703,
|
| 1130 |
+
"step": 1870
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.61,
|
| 1134 |
+
"learning_rate": 3.935975817943614e-05,
|
| 1135 |
+
"loss": 0.766,
|
| 1136 |
+
"step": 1880
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.61,
|
| 1140 |
+
"learning_rate": 3.925496439005246e-05,
|
| 1141 |
+
"loss": 0.7632,
|
| 1142 |
+
"step": 1890
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 0.62,
|
| 1146 |
+
"learning_rate": 3.914979814050949e-05,
|
| 1147 |
+
"loss": 0.7646,
|
| 1148 |
+
"step": 1900
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 0.62,
|
| 1152 |
+
"learning_rate": 3.904426217863858e-05,
|
| 1153 |
+
"loss": 0.7624,
|
| 1154 |
+
"step": 1910
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.62,
|
| 1158 |
+
"learning_rate": 3.8938359261931076e-05,
|
| 1159 |
+
"loss": 0.7498,
|
| 1160 |
+
"step": 1920
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.63,
|
| 1164 |
+
"learning_rate": 3.883209215746633e-05,
|
| 1165 |
+
"loss": 0.7541,
|
| 1166 |
+
"step": 1930
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 0.63,
|
| 1170 |
+
"learning_rate": 3.8725463641839296e-05,
|
| 1171 |
+
"loss": 0.7622,
|
| 1172 |
+
"step": 1940
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.63,
|
| 1176 |
+
"learning_rate": 3.8618476501088094e-05,
|
| 1177 |
+
"loss": 0.7536,
|
| 1178 |
+
"step": 1950
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.64,
|
| 1182 |
+
"learning_rate": 3.851113353062115e-05,
|
| 1183 |
+
"loss": 0.7419,
|
| 1184 |
+
"step": 1960
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 0.64,
|
| 1188 |
+
"learning_rate": 3.840343753514414e-05,
|
| 1189 |
+
"loss": 0.756,
|
| 1190 |
+
"step": 1970
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 0.64,
|
| 1194 |
+
"learning_rate": 3.829539132858678e-05,
|
| 1195 |
+
"loss": 0.7511,
|
| 1196 |
+
"step": 1980
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.65,
|
| 1200 |
+
"learning_rate": 3.818699773402925e-05,
|
| 1201 |
+
"loss": 0.7505,
|
| 1202 |
+
"step": 1990
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.65,
|
| 1206 |
+
"learning_rate": 3.80782595836284e-05,
|
| 1207 |
+
"loss": 0.7396,
|
| 1208 |
+
"step": 2000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 0.65,
|
| 1212 |
+
"learning_rate": 3.7969179718543855e-05,
|
| 1213 |
+
"loss": 0.7524,
|
| 1214 |
+
"step": 2010
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.66,
|
| 1218 |
+
"learning_rate": 3.7859760988863664e-05,
|
| 1219 |
+
"loss": 0.7486,
|
| 1220 |
+
"step": 2020
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.66,
|
| 1224 |
+
"learning_rate": 3.77500062535299e-05,
|
| 1225 |
+
"loss": 0.7599,
|
| 1226 |
+
"step": 2030
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 0.66,
|
| 1230 |
+
"learning_rate": 3.763991838026397e-05,
|
| 1231 |
+
"loss": 0.7526,
|
| 1232 |
+
"step": 2040
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 0.67,
|
| 1236 |
+
"learning_rate": 3.7529500245491616e-05,
|
| 1237 |
+
"loss": 0.7569,
|
| 1238 |
+
"step": 2050
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.67,
|
| 1242 |
+
"learning_rate": 3.741875473426783e-05,
|
| 1243 |
+
"loss": 0.7637,
|
| 1244 |
+
"step": 2060
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 0.67,
|
| 1248 |
+
"learning_rate": 3.730768474020144e-05,
|
| 1249 |
+
"loss": 0.757,
|
| 1250 |
+
"step": 2070
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 0.68,
|
| 1254 |
+
"learning_rate": 3.7196293165379486e-05,
|
| 1255 |
+
"loss": 0.765,
|
| 1256 |
+
"step": 2080
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.68,
|
| 1260 |
+
"learning_rate": 3.7084582920291456e-05,
|
| 1261 |
+
"loss": 0.7628,
|
| 1262 |
+
"step": 2090
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.68,
|
| 1266 |
+
"learning_rate": 3.697255692375317e-05,
|
| 1267 |
+
"loss": 0.7514,
|
| 1268 |
+
"step": 2100
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 0.69,
|
| 1272 |
+
"learning_rate": 3.686021810283055e-05,
|
| 1273 |
+
"loss": 0.7683,
|
| 1274 |
+
"step": 2110
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 0.69,
|
| 1278 |
+
"learning_rate": 3.6747569392763135e-05,
|
| 1279 |
+
"loss": 0.7664,
|
| 1280 |
+
"step": 2120
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.69,
|
| 1284 |
+
"learning_rate": 3.6634613736887404e-05,
|
| 1285 |
+
"loss": 0.7495,
|
| 1286 |
+
"step": 2130
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 0.7,
|
| 1290 |
+
"learning_rate": 3.652135408655982e-05,
|
| 1291 |
+
"loss": 0.7501,
|
| 1292 |
+
"step": 2140
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 0.7,
|
| 1296 |
+
"learning_rate": 3.6407793401079785e-05,
|
| 1297 |
+
"loss": 0.7581,
|
| 1298 |
+
"step": 2150
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.7,
|
| 1302 |
+
"learning_rate": 3.629393464761227e-05,
|
| 1303 |
+
"loss": 0.7564,
|
| 1304 |
+
"step": 2160
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.71,
|
| 1308 |
+
"learning_rate": 3.6179780801110305e-05,
|
| 1309 |
+
"loss": 0.7558,
|
| 1310 |
+
"step": 2170
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 0.71,
|
| 1314 |
+
"learning_rate": 3.606533484423722e-05,
|
| 1315 |
+
"loss": 0.7578,
|
| 1316 |
+
"step": 2180
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 0.71,
|
| 1320 |
+
"learning_rate": 3.5950599767288785e-05,
|
| 1321 |
+
"loss": 0.7585,
|
| 1322 |
+
"step": 2190
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.72,
|
| 1326 |
+
"learning_rate": 3.5835578568114985e-05,
|
| 1327 |
+
"loss": 0.7473,
|
| 1328 |
+
"step": 2200
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 0.72,
|
| 1332 |
+
"learning_rate": 3.572027425204176e-05,
|
| 1333 |
+
"loss": 0.746,
|
| 1334 |
+
"step": 2210
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 0.72,
|
| 1338 |
+
"learning_rate": 3.560468983179245e-05,
|
| 1339 |
+
"loss": 0.7517,
|
| 1340 |
+
"step": 2220
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.73,
|
| 1344 |
+
"learning_rate": 3.548882832740909e-05,
|
| 1345 |
+
"loss": 0.75,
|
| 1346 |
+
"step": 2230
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.73,
|
| 1350 |
+
"learning_rate": 3.537269276617347e-05,
|
| 1351 |
+
"loss": 0.7407,
|
| 1352 |
+
"step": 2240
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 0.73,
|
| 1356 |
+
"learning_rate": 3.5256286182528083e-05,
|
| 1357 |
+
"loss": 0.7465,
|
| 1358 |
+
"step": 2250
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 0.74,
|
| 1362 |
+
"learning_rate": 3.5139611617996846e-05,
|
| 1363 |
+
"loss": 0.7477,
|
| 1364 |
+
"step": 2260
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.74,
|
| 1368 |
+
"learning_rate": 3.5022672121105534e-05,
|
| 1369 |
+
"loss": 0.7494,
|
| 1370 |
+
"step": 2270
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.74,
|
| 1374 |
+
"learning_rate": 3.4905470747302256e-05,
|
| 1375 |
+
"loss": 0.7514,
|
| 1376 |
+
"step": 2280
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 0.75,
|
| 1380 |
+
"learning_rate": 3.478801055887752e-05,
|
| 1381 |
+
"loss": 0.7466,
|
| 1382 |
+
"step": 2290
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.75,
|
| 1386 |
+
"learning_rate": 3.4670294624884275e-05,
|
| 1387 |
+
"loss": 0.7552,
|
| 1388 |
+
"step": 2300
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.75,
|
| 1392 |
+
"learning_rate": 3.4552326021057716e-05,
|
| 1393 |
+
"loss": 0.7625,
|
| 1394 |
+
"step": 2310
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 0.75,
|
| 1398 |
+
"learning_rate": 3.443410782973489e-05,
|
| 1399 |
+
"loss": 0.7517,
|
| 1400 |
+
"step": 2320
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 0.76,
|
| 1404 |
+
"learning_rate": 3.431564313977419e-05,
|
| 1405 |
+
"loss": 0.7606,
|
| 1406 |
+
"step": 2330
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.76,
|
| 1410 |
+
"learning_rate": 3.419693504647464e-05,
|
| 1411 |
+
"loss": 0.7568,
|
| 1412 |
+
"step": 2340
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 0.76,
|
| 1416 |
+
"learning_rate": 3.4077986651495e-05,
|
| 1417 |
+
"loss": 0.7452,
|
| 1418 |
+
"step": 2350
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 0.77,
|
| 1422 |
+
"learning_rate": 3.395880106277277e-05,
|
| 1423 |
+
"loss": 0.746,
|
| 1424 |
+
"step": 2360
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 0.77,
|
| 1428 |
+
"learning_rate": 3.383938139444292e-05,
|
| 1429 |
+
"loss": 0.7602,
|
| 1430 |
+
"step": 2370
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.77,
|
| 1434 |
+
"learning_rate": 3.3719730766756596e-05,
|
| 1435 |
+
"loss": 0.7453,
|
| 1436 |
+
"step": 2380
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 0.78,
|
| 1440 |
+
"learning_rate": 3.3599852305999514e-05,
|
| 1441 |
+
"loss": 0.7606,
|
| 1442 |
+
"step": 2390
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 0.78,
|
| 1446 |
+
"learning_rate": 3.347974914441035e-05,
|
| 1447 |
+
"loss": 0.7545,
|
| 1448 |
+
"step": 2400
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 0.78,
|
| 1452 |
+
"learning_rate": 3.335942442009884e-05,
|
| 1453 |
+
"loss": 0.7379,
|
| 1454 |
+
"step": 2410
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 0.79,
|
| 1458 |
+
"learning_rate": 3.3238881276963835e-05,
|
| 1459 |
+
"loss": 0.7372,
|
| 1460 |
+
"step": 2420
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 0.79,
|
| 1464 |
+
"learning_rate": 3.311812286461111e-05,
|
| 1465 |
+
"loss": 0.7592,
|
| 1466 |
+
"step": 2430
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.79,
|
| 1470 |
+
"learning_rate": 3.299715233827111e-05,
|
| 1471 |
+
"loss": 0.7576,
|
| 1472 |
+
"step": 2440
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.8,
|
| 1476 |
+
"learning_rate": 3.2875972858716464e-05,
|
| 1477 |
+
"loss": 0.7449,
|
| 1478 |
+
"step": 2450
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 0.8,
|
| 1482 |
+
"learning_rate": 3.2754587592179454e-05,
|
| 1483 |
+
"loss": 0.7448,
|
| 1484 |
+
"step": 2460
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 0.8,
|
| 1488 |
+
"learning_rate": 3.263299971026925e-05,
|
| 1489 |
+
"loss": 0.7483,
|
| 1490 |
+
"step": 2470
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 0.81,
|
| 1494 |
+
"learning_rate": 3.251121238988902e-05,
|
| 1495 |
+
"loss": 0.7574,
|
| 1496 |
+
"step": 2480
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 0.81,
|
| 1500 |
+
"learning_rate": 3.2389228813153006e-05,
|
| 1501 |
+
"loss": 0.744,
|
| 1502 |
+
"step": 2490
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 0.81,
|
| 1506 |
+
"learning_rate": 3.2267052167303275e-05,
|
| 1507 |
+
"loss": 0.7445,
|
| 1508 |
+
"step": 2500
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 0.82,
|
| 1512 |
+
"learning_rate": 3.214468564462651e-05,
|
| 1513 |
+
"loss": 0.7486,
|
| 1514 |
+
"step": 2510
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.82,
|
| 1518 |
+
"learning_rate": 3.202213244237058e-05,
|
| 1519 |
+
"loss": 0.7631,
|
| 1520 |
+
"step": 2520
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 0.82,
|
| 1524 |
+
"learning_rate": 3.189939576266101e-05,
|
| 1525 |
+
"loss": 0.7538,
|
| 1526 |
+
"step": 2530
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 0.83,
|
| 1530 |
+
"learning_rate": 3.1776478812417295e-05,
|
| 1531 |
+
"loss": 0.7576,
|
| 1532 |
+
"step": 2540
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 0.83,
|
| 1536 |
+
"learning_rate": 3.165338480326912e-05,
|
| 1537 |
+
"loss": 0.7405,
|
| 1538 |
+
"step": 2550
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 0.83,
|
| 1542 |
+
"learning_rate": 3.1530116951472464e-05,
|
| 1543 |
+
"loss": 0.7362,
|
| 1544 |
+
"step": 2560
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 0.84,
|
| 1548 |
+
"learning_rate": 3.140667847782552e-05,
|
| 1549 |
+
"loss": 0.7503,
|
| 1550 |
+
"step": 2570
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 0.84,
|
| 1554 |
+
"learning_rate": 3.1283072607584575e-05,
|
| 1555 |
+
"loss": 0.7513,
|
| 1556 |
+
"step": 2580
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.84,
|
| 1560 |
+
"learning_rate": 3.115930257037975e-05,
|
| 1561 |
+
"loss": 0.7535,
|
| 1562 |
+
"step": 2590
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 0.85,
|
| 1566 |
+
"learning_rate": 3.1035371600130566e-05,
|
| 1567 |
+
"loss": 0.7386,
|
| 1568 |
+
"step": 2600
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 0.85,
|
| 1572 |
+
"learning_rate": 3.091128293496148e-05,
|
| 1573 |
+
"loss": 0.7617,
|
| 1574 |
+
"step": 2610
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 0.85,
|
| 1578 |
+
"learning_rate": 3.07870398171173e-05,
|
| 1579 |
+
"loss": 0.7519,
|
| 1580 |
+
"step": 2620
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 0.86,
|
| 1584 |
+
"learning_rate": 3.066264549287839e-05,
|
| 1585 |
+
"loss": 0.739,
|
| 1586 |
+
"step": 2630
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 0.86,
|
| 1590 |
+
"learning_rate": 3.053810321247597e-05,
|
| 1591 |
+
"loss": 0.7375,
|
| 1592 |
+
"step": 2640
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 0.86,
|
| 1596 |
+
"learning_rate": 3.041341623000708e-05,
|
| 1597 |
+
"loss": 0.7511,
|
| 1598 |
+
"step": 2650
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.87,
|
| 1602 |
+
"learning_rate": 3.028858780334961e-05,
|
| 1603 |
+
"loss": 0.7498,
|
| 1604 |
+
"step": 2660
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 0.87,
|
| 1608 |
+
"learning_rate": 3.0163621194077167e-05,
|
| 1609 |
+
"loss": 0.7504,
|
| 1610 |
+
"step": 2670
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 0.87,
|
| 1614 |
+
"learning_rate": 3.003851966737387e-05,
|
| 1615 |
+
"loss": 0.7508,
|
| 1616 |
+
"step": 2680
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 0.88,
|
| 1620 |
+
"learning_rate": 2.991328649194901e-05,
|
| 1621 |
+
"loss": 0.7465,
|
| 1622 |
+
"step": 2690
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 0.88,
|
| 1626 |
+
"learning_rate": 2.9787924939951644e-05,
|
| 1627 |
+
"loss": 0.741,
|
| 1628 |
+
"step": 2700
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 0.88,
|
| 1632 |
+
"learning_rate": 2.9662438286885146e-05,
|
| 1633 |
+
"loss": 0.7424,
|
| 1634 |
+
"step": 2710
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.88,
|
| 1638 |
+
"learning_rate": 2.953682981152154e-05,
|
| 1639 |
+
"loss": 0.7425,
|
| 1640 |
+
"step": 2720
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.89,
|
| 1644 |
+
"learning_rate": 2.941110279581591e-05,
|
| 1645 |
+
"loss": 0.7355,
|
| 1646 |
+
"step": 2730
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 0.89,
|
| 1650 |
+
"learning_rate": 2.9285260524820624e-05,
|
| 1651 |
+
"loss": 0.7374,
|
| 1652 |
+
"step": 2740
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 0.89,
|
| 1656 |
+
"learning_rate": 2.9159306286599446e-05,
|
| 1657 |
+
"loss": 0.7319,
|
| 1658 |
+
"step": 2750
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 0.9,
|
| 1662 |
+
"learning_rate": 2.9033243372141733e-05,
|
| 1663 |
+
"loss": 0.7515,
|
| 1664 |
+
"step": 2760
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 0.9,
|
| 1668 |
+
"learning_rate": 2.890707507527633e-05,
|
| 1669 |
+
"loss": 0.7391,
|
| 1670 |
+
"step": 2770
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 0.9,
|
| 1674 |
+
"learning_rate": 2.8780804692585585e-05,
|
| 1675 |
+
"loss": 0.7477,
|
| 1676 |
+
"step": 2780
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.91,
|
| 1680 |
+
"learning_rate": 2.86544355233192e-05,
|
| 1681 |
+
"loss": 0.7498,
|
| 1682 |
+
"step": 2790
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.91,
|
| 1686 |
+
"learning_rate": 2.8527970869307996e-05,
|
| 1687 |
+
"loss": 0.7417,
|
| 1688 |
+
"step": 2800
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 0.91,
|
| 1692 |
+
"learning_rate": 2.8401414034877683e-05,
|
| 1693 |
+
"loss": 0.7501,
|
| 1694 |
+
"step": 2810
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 0.92,
|
| 1698 |
+
"learning_rate": 2.827476832676247e-05,
|
| 1699 |
+
"loss": 0.7397,
|
| 1700 |
+
"step": 2820
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 0.92,
|
| 1704 |
+
"learning_rate": 2.814803705401874e-05,
|
| 1705 |
+
"loss": 0.7471,
|
| 1706 |
+
"step": 2830
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 0.92,
|
| 1710 |
+
"learning_rate": 2.8021223527938524e-05,
|
| 1711 |
+
"loss": 0.7494,
|
| 1712 |
+
"step": 2840
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 0.93,
|
| 1716 |
+
"learning_rate": 2.7894331061963004e-05,
|
| 1717 |
+
"loss": 0.745,
|
| 1718 |
+
"step": 2850
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.93,
|
| 1722 |
+
"learning_rate": 2.7767362971595946e-05,
|
| 1723 |
+
"loss": 0.7432,
|
| 1724 |
+
"step": 2860
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.93,
|
| 1728 |
+
"learning_rate": 2.7640322574317063e-05,
|
| 1729 |
+
"loss": 0.7431,
|
| 1730 |
+
"step": 2870
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 0.94,
|
| 1734 |
+
"learning_rate": 2.751321318949533e-05,
|
| 1735 |
+
"loss": 0.7471,
|
| 1736 |
+
"step": 2880
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 0.94,
|
| 1740 |
+
"learning_rate": 2.738603813830227e-05,
|
| 1741 |
+
"loss": 0.7417,
|
| 1742 |
+
"step": 2890
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 0.94,
|
| 1746 |
+
"learning_rate": 2.7258800743625167e-05,
|
| 1747 |
+
"loss": 0.7411,
|
| 1748 |
+
"step": 2900
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 0.95,
|
| 1752 |
+
"learning_rate": 2.7131504329980235e-05,
|
| 1753 |
+
"loss": 0.7546,
|
| 1754 |
+
"step": 2910
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 0.95,
|
| 1758 |
+
"learning_rate": 2.7004152223425753e-05,
|
| 1759 |
+
"loss": 0.7368,
|
| 1760 |
+
"step": 2920
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 0.95,
|
| 1764 |
+
"learning_rate": 2.687674775147519e-05,
|
| 1765 |
+
"loss": 0.7447,
|
| 1766 |
+
"step": 2930
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.96,
|
| 1770 |
+
"learning_rate": 2.6749294243010237e-05,
|
| 1771 |
+
"loss": 0.7364,
|
| 1772 |
+
"step": 2940
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 0.96,
|
| 1776 |
+
"learning_rate": 2.6621795028193837e-05,
|
| 1777 |
+
"loss": 0.7417,
|
| 1778 |
+
"step": 2950
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 0.96,
|
| 1782 |
+
"learning_rate": 2.649425343838314e-05,
|
| 1783 |
+
"loss": 0.7371,
|
| 1784 |
+
"step": 2960
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 0.97,
|
| 1788 |
+
"learning_rate": 2.6366672806042532e-05,
|
| 1789 |
+
"loss": 0.7452,
|
| 1790 |
+
"step": 2970
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 0.97,
|
| 1794 |
+
"learning_rate": 2.6239056464656485e-05,
|
| 1795 |
+
"loss": 0.7427,
|
| 1796 |
+
"step": 2980
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 0.97,
|
| 1800 |
+
"learning_rate": 2.611140774864252e-05,
|
| 1801 |
+
"loss": 0.7381,
|
| 1802 |
+
"step": 2990
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 0.98,
|
| 1806 |
+
"learning_rate": 2.5983729993264034e-05,
|
| 1807 |
+
"loss": 0.7476,
|
| 1808 |
+
"step": 3000
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.98,
|
| 1812 |
+
"learning_rate": 2.5856026534543193e-05,
|
| 1813 |
+
"loss": 0.7429,
|
| 1814 |
+
"step": 3010
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 0.98,
|
| 1818 |
+
"learning_rate": 2.5728300709173747e-05,
|
| 1819 |
+
"loss": 0.742,
|
| 1820 |
+
"step": 3020
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 0.99,
|
| 1824 |
+
"learning_rate": 2.5600555854433844e-05,
|
| 1825 |
+
"loss": 0.7426,
|
| 1826 |
+
"step": 3030
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 0.99,
|
| 1830 |
+
"learning_rate": 2.5472795308098857e-05,
|
| 1831 |
+
"loss": 0.7503,
|
| 1832 |
+
"step": 3040
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 0.99,
|
| 1836 |
+
"learning_rate": 2.5345022408354142e-05,
|
| 1837 |
+
"loss": 0.7454,
|
| 1838 |
+
"step": 3050
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 1.0,
|
| 1842 |
+
"learning_rate": 2.521724049370784e-05,
|
| 1843 |
+
"loss": 0.742,
|
| 1844 |
+
"step": 3060
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 1.0,
|
| 1848 |
+
"learning_rate": 2.5089452902903616e-05,
|
| 1849 |
+
"loss": 0.7444,
|
| 1850 |
+
"step": 3070
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 1.0,
|
| 1854 |
+
"learning_rate": 2.4961662974833476e-05,
|
| 1855 |
+
"loss": 0.736,
|
| 1856 |
+
"step": 3080
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 1.01,
|
| 1860 |
+
"learning_rate": 2.4833874048450476e-05,
|
| 1861 |
+
"loss": 0.7477,
|
| 1862 |
+
"step": 3090
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 1.01,
|
| 1866 |
+
"learning_rate": 2.4706089462681502e-05,
|
| 1867 |
+
"loss": 0.7311,
|
| 1868 |
+
"step": 3100
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 1.01,
|
| 1872 |
+
"learning_rate": 2.4578312556340026e-05,
|
| 1873 |
+
"loss": 0.7276,
|
| 1874 |
+
"step": 3110
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 1.02,
|
| 1878 |
+
"learning_rate": 2.4450546668038875e-05,
|
| 1879 |
+
"loss": 0.7389,
|
| 1880 |
+
"step": 3120
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 1.02,
|
| 1884 |
+
"learning_rate": 2.4322795136102985e-05,
|
| 1885 |
+
"loss": 0.7442,
|
| 1886 |
+
"step": 3130
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 1.02,
|
| 1890 |
+
"learning_rate": 2.4195061298482178e-05,
|
| 1891 |
+
"loss": 0.7416,
|
| 1892 |
+
"step": 3140
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 1.02,
|
| 1896 |
+
"learning_rate": 2.4067348492663965e-05,
|
| 1897 |
+
"loss": 0.7446,
|
| 1898 |
+
"step": 3150
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 1.03,
|
| 1902 |
+
"learning_rate": 2.3939660055586318e-05,
|
| 1903 |
+
"loss": 0.7352,
|
| 1904 |
+
"step": 3160
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 1.03,
|
| 1908 |
+
"learning_rate": 2.38119993235505e-05,
|
| 1909 |
+
"loss": 0.7446,
|
| 1910 |
+
"step": 3170
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 1.03,
|
| 1914 |
+
"learning_rate": 2.3684369632133856e-05,
|
| 1915 |
+
"loss": 0.7265,
|
| 1916 |
+
"step": 3180
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 1.04,
|
| 1920 |
+
"learning_rate": 2.3556774316102735e-05,
|
| 1921 |
+
"loss": 0.7365,
|
| 1922 |
+
"step": 3190
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 1.04,
|
| 1926 |
+
"learning_rate": 2.342921670932528e-05,
|
| 1927 |
+
"loss": 0.7357,
|
| 1928 |
+
"step": 3200
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 1.04,
|
| 1932 |
+
"learning_rate": 2.3301700144684338e-05,
|
| 1933 |
+
"loss": 0.7325,
|
| 1934 |
+
"step": 3210
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 1.05,
|
| 1938 |
+
"learning_rate": 2.3174227953990423e-05,
|
| 1939 |
+
"loss": 0.7349,
|
| 1940 |
+
"step": 3220
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 1.05,
|
| 1944 |
+
"learning_rate": 2.3046803467894616e-05,
|
| 1945 |
+
"loss": 0.7401,
|
| 1946 |
+
"step": 3230
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 1.05,
|
| 1950 |
+
"learning_rate": 2.291943001580151e-05,
|
| 1951 |
+
"loss": 0.7388,
|
| 1952 |
+
"step": 3240
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 1.06,
|
| 1956 |
+
"learning_rate": 2.2792110925782324e-05,
|
| 1957 |
+
"loss": 0.7374,
|
| 1958 |
+
"step": 3250
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 1.06,
|
| 1962 |
+
"learning_rate": 2.2664849524487843e-05,
|
| 1963 |
+
"loss": 0.7327,
|
| 1964 |
+
"step": 3260
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 1.06,
|
| 1968 |
+
"learning_rate": 2.253764913706153e-05,
|
| 1969 |
+
"loss": 0.7454,
|
| 1970 |
+
"step": 3270
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 1.07,
|
| 1974 |
+
"learning_rate": 2.241051308705265e-05,
|
| 1975 |
+
"loss": 0.7346,
|
| 1976 |
+
"step": 3280
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 1.07,
|
| 1980 |
+
"learning_rate": 2.2283444696329473e-05,
|
| 1981 |
+
"loss": 0.731,
|
| 1982 |
+
"step": 3290
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 1.07,
|
| 1986 |
+
"learning_rate": 2.215644728499238e-05,
|
| 1987 |
+
"loss": 0.7428,
|
| 1988 |
+
"step": 3300
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 1.08,
|
| 1992 |
+
"learning_rate": 2.2029524171287194e-05,
|
| 1993 |
+
"loss": 0.7332,
|
| 1994 |
+
"step": 3310
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 1.08,
|
| 1998 |
+
"learning_rate": 2.190267867151847e-05,
|
| 1999 |
+
"loss": 0.732,
|
| 2000 |
+
"step": 3320
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 1.08,
|
| 2004 |
+
"learning_rate": 2.1775914099962814e-05,
|
| 2005 |
+
"loss": 0.7334,
|
| 2006 |
+
"step": 3330
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 1.09,
|
| 2010 |
+
"learning_rate": 2.164923376878229e-05,
|
| 2011 |
+
"loss": 0.742,
|
| 2012 |
+
"step": 3340
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 1.09,
|
| 2016 |
+
"learning_rate": 2.1522640987937937e-05,
|
| 2017 |
+
"loss": 0.7381,
|
| 2018 |
+
"step": 3350
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 1.09,
|
| 2022 |
+
"learning_rate": 2.1396139065103193e-05,
|
| 2023 |
+
"loss": 0.7255,
|
| 2024 |
+
"step": 3360
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 1.1,
|
| 2028 |
+
"learning_rate": 2.126973130557753e-05,
|
| 2029 |
+
"loss": 0.7232,
|
| 2030 |
+
"step": 3370
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 1.1,
|
| 2034 |
+
"learning_rate": 2.1143421012200092e-05,
|
| 2035 |
+
"loss": 0.7288,
|
| 2036 |
+
"step": 3380
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 1.1,
|
| 2040 |
+
"learning_rate": 2.1017211485263376e-05,
|
| 2041 |
+
"loss": 0.7373,
|
| 2042 |
+
"step": 3390
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 1.11,
|
| 2046 |
+
"learning_rate": 2.0891106022426997e-05,
|
| 2047 |
+
"loss": 0.7399,
|
| 2048 |
+
"step": 3400
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 1.11,
|
| 2052 |
+
"learning_rate": 2.0765107918631542e-05,
|
| 2053 |
+
"loss": 0.7413,
|
| 2054 |
+
"step": 3410
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 1.11,
|
| 2058 |
+
"learning_rate": 2.063922046601249e-05,
|
| 2059 |
+
"loss": 0.733,
|
| 2060 |
+
"step": 3420
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 1.12,
|
| 2064 |
+
"learning_rate": 2.0513446953814154e-05,
|
| 2065 |
+
"loss": 0.7261,
|
| 2066 |
+
"step": 3430
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 1.12,
|
| 2070 |
+
"learning_rate": 2.0387790668303758e-05,
|
| 2071 |
+
"loss": 0.7331,
|
| 2072 |
+
"step": 3440
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 1.12,
|
| 2076 |
+
"learning_rate": 2.02622548926856e-05,
|
| 2077 |
+
"loss": 0.7363,
|
| 2078 |
+
"step": 3450
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 1.13,
|
| 2082 |
+
"learning_rate": 2.0136842907015214e-05,
|
| 2083 |
+
"loss": 0.7245,
|
| 2084 |
+
"step": 3460
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 1.13,
|
| 2088 |
+
"learning_rate": 2.0011557988113704e-05,
|
| 2089 |
+
"loss": 0.739,
|
| 2090 |
+
"step": 3470
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 1.13,
|
| 2094 |
+
"learning_rate": 1.9886403409482125e-05,
|
| 2095 |
+
"loss": 0.7398,
|
| 2096 |
+
"step": 3480
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 1.14,
|
| 2100 |
+
"learning_rate": 1.976138244121593e-05,
|
| 2101 |
+
"loss": 0.7304,
|
| 2102 |
+
"step": 3490
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 1.14,
|
| 2106 |
+
"learning_rate": 1.963649834991954e-05,
|
| 2107 |
+
"loss": 0.7308,
|
| 2108 |
+
"step": 3500
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 1.14,
|
| 2112 |
+
"learning_rate": 1.9511754398620996e-05,
|
| 2113 |
+
"loss": 0.7362,
|
| 2114 |
+
"step": 3510
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 1.15,
|
| 2118 |
+
"learning_rate": 1.938715384668671e-05,
|
| 2119 |
+
"loss": 0.7335,
|
| 2120 |
+
"step": 3520
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 1.15,
|
| 2124 |
+
"learning_rate": 1.9262699949736273e-05,
|
| 2125 |
+
"loss": 0.7334,
|
| 2126 |
+
"step": 3530
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 1.15,
|
| 2130 |
+
"learning_rate": 1.9138395959557398e-05,
|
| 2131 |
+
"loss": 0.7313,
|
| 2132 |
+
"step": 3540
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 1.15,
|
| 2136 |
+
"learning_rate": 1.9014245124021007e-05,
|
| 2137 |
+
"loss": 0.7268,
|
| 2138 |
+
"step": 3550
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 1.16,
|
| 2142 |
+
"learning_rate": 1.8890250686996298e-05,
|
| 2143 |
+
"loss": 0.7211,
|
| 2144 |
+
"step": 3560
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 1.16,
|
| 2148 |
+
"learning_rate": 1.8766415888266004e-05,
|
| 2149 |
+
"loss": 0.7338,
|
| 2150 |
+
"step": 3570
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 1.16,
|
| 2154 |
+
"learning_rate": 1.8642743963441798e-05,
|
| 2155 |
+
"loss": 0.7289,
|
| 2156 |
+
"step": 3580
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 1.17,
|
| 2160 |
+
"learning_rate": 1.8519238143879687e-05,
|
| 2161 |
+
"loss": 0.7242,
|
| 2162 |
+
"step": 3590
|
| 2163 |
+
},
|
| 2164 |
+
{
|
| 2165 |
+
"epoch": 1.17,
|
| 2166 |
+
"learning_rate": 1.839590165659559e-05,
|
| 2167 |
+
"loss": 0.7247,
|
| 2168 |
+
"step": 3600
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 1.17,
|
| 2172 |
+
"learning_rate": 1.827273772418105e-05,
|
| 2173 |
+
"loss": 0.728,
|
| 2174 |
+
"step": 3610
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 1.18,
|
| 2178 |
+
"learning_rate": 1.8149749564719042e-05,
|
| 2179 |
+
"loss": 0.7247,
|
| 2180 |
+
"step": 3620
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 1.18,
|
| 2184 |
+
"learning_rate": 1.8026940391699827e-05,
|
| 2185 |
+
"loss": 0.7211,
|
| 2186 |
+
"step": 3630
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 1.18,
|
| 2190 |
+
"learning_rate": 1.790431341393703e-05,
|
| 2191 |
+
"loss": 0.7289,
|
| 2192 |
+
"step": 3640
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 1.19,
|
| 2196 |
+
"learning_rate": 1.7781871835483843e-05,
|
| 2197 |
+
"loss": 0.7412,
|
| 2198 |
+
"step": 3650
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 1.19,
|
| 2202 |
+
"learning_rate": 1.7659618855549198e-05,
|
| 2203 |
+
"loss": 0.7245,
|
| 2204 |
+
"step": 3660
|
| 2205 |
+
},
|
| 2206 |
+
{
|
| 2207 |
+
"epoch": 1.19,
|
| 2208 |
+
"learning_rate": 1.7537557668414273e-05,
|
| 2209 |
+
"loss": 0.7317,
|
| 2210 |
+
"step": 3670
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 1.2,
|
| 2214 |
+
"learning_rate": 1.741569146334901e-05,
|
| 2215 |
+
"loss": 0.7372,
|
| 2216 |
+
"step": 3680
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 1.2,
|
| 2220 |
+
"learning_rate": 1.729402342452874e-05,
|
| 2221 |
+
"loss": 0.742,
|
| 2222 |
+
"step": 3690
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 1.2,
|
| 2226 |
+
"learning_rate": 1.7172556730951027e-05,
|
| 2227 |
+
"loss": 0.7454,
|
| 2228 |
+
"step": 3700
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 1.21,
|
| 2232 |
+
"learning_rate": 1.7051294556352603e-05,
|
| 2233 |
+
"loss": 0.741,
|
| 2234 |
+
"step": 3710
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 1.21,
|
| 2238 |
+
"learning_rate": 1.6930240069126435e-05,
|
| 2239 |
+
"loss": 0.7297,
|
| 2240 |
+
"step": 3720
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 1.21,
|
| 2244 |
+
"learning_rate": 1.6809396432238932e-05,
|
| 2245 |
+
"loss": 0.74,
|
| 2246 |
+
"step": 3730
|
| 2247 |
+
},
|
| 2248 |
+
{
|
| 2249 |
+
"epoch": 1.22,
|
| 2250 |
+
"learning_rate": 1.6688766803147315e-05,
|
| 2251 |
+
"loss": 0.7344,
|
| 2252 |
+
"step": 3740
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 1.22,
|
| 2256 |
+
"learning_rate": 1.6568354333717117e-05,
|
| 2257 |
+
"loss": 0.7337,
|
| 2258 |
+
"step": 3750
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 1.22,
|
| 2262 |
+
"learning_rate": 1.644816217013982e-05,
|
| 2263 |
+
"loss": 0.731,
|
| 2264 |
+
"step": 3760
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 1.23,
|
| 2268 |
+
"learning_rate": 1.632819345285065e-05,
|
| 2269 |
+
"loss": 0.7272,
|
| 2270 |
+
"step": 3770
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 1.23,
|
| 2274 |
+
"learning_rate": 1.6208451316446537e-05,
|
| 2275 |
+
"loss": 0.7309,
|
| 2276 |
+
"step": 3780
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 1.23,
|
| 2280 |
+
"learning_rate": 1.6088938889604198e-05,
|
| 2281 |
+
"loss": 0.726,
|
| 2282 |
+
"step": 3790
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 1.24,
|
| 2286 |
+
"learning_rate": 1.5969659294998392e-05,
|
| 2287 |
+
"loss": 0.7305,
|
| 2288 |
+
"step": 3800
|
| 2289 |
+
},
|
| 2290 |
+
{
|
| 2291 |
+
"epoch": 1.24,
|
| 2292 |
+
"learning_rate": 1.5850615649220333e-05,
|
| 2293 |
+
"loss": 0.731,
|
| 2294 |
+
"step": 3810
|
| 2295 |
+
},
|
| 2296 |
+
{
|
| 2297 |
+
"epoch": 1.24,
|
| 2298 |
+
"learning_rate": 1.5731811062696275e-05,
|
| 2299 |
+
"loss": 0.7275,
|
| 2300 |
+
"step": 3820
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 1.25,
|
| 2304 |
+
"learning_rate": 1.5613248639606192e-05,
|
| 2305 |
+
"loss": 0.7363,
|
| 2306 |
+
"step": 3830
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 1.25,
|
| 2310 |
+
"learning_rate": 1.5494931477802725e-05,
|
| 2311 |
+
"loss": 0.7305,
|
| 2312 |
+
"step": 3840
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 1.25,
|
| 2316 |
+
"learning_rate": 1.5376862668730212e-05,
|
| 2317 |
+
"loss": 0.7283,
|
| 2318 |
+
"step": 3850
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 1.26,
|
| 2322 |
+
"learning_rate": 1.5259045297343917e-05,
|
| 2323 |
+
"loss": 0.7362,
|
| 2324 |
+
"step": 3860
|
| 2325 |
+
},
|
| 2326 |
+
{
|
| 2327 |
+
"epoch": 1.26,
|
| 2328 |
+
"learning_rate": 1.5141482442029415e-05,
|
| 2329 |
+
"loss": 0.7211,
|
| 2330 |
+
"step": 3870
|
| 2331 |
+
},
|
| 2332 |
+
{
|
| 2333 |
+
"epoch": 1.26,
|
| 2334 |
+
"learning_rate": 1.5024177174522197e-05,
|
| 2335 |
+
"loss": 0.726,
|
| 2336 |
+
"step": 3880
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 1.27,
|
| 2340 |
+
"learning_rate": 1.4907132559827364e-05,
|
| 2341 |
+
"loss": 0.7326,
|
| 2342 |
+
"step": 3890
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 1.27,
|
| 2346 |
+
"learning_rate": 1.479035165613957e-05,
|
| 2347 |
+
"loss": 0.7265,
|
| 2348 |
+
"step": 3900
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 1.27,
|
| 2352 |
+
"learning_rate": 1.467383751476308e-05,
|
| 2353 |
+
"loss": 0.7278,
|
| 2354 |
+
"step": 3910
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 1.28,
|
| 2358 |
+
"learning_rate": 1.455759318003213e-05,
|
| 2359 |
+
"loss": 0.7196,
|
| 2360 |
+
"step": 3920
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 1.28,
|
| 2364 |
+
"learning_rate": 1.4441621689231271e-05,
|
| 2365 |
+
"loss": 0.7279,
|
| 2366 |
+
"step": 3930
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 1.28,
|
| 2370 |
+
"learning_rate": 1.4325926072516094e-05,
|
| 2371 |
+
"loss": 0.7329,
|
| 2372 |
+
"step": 3940
|
| 2373 |
+
},
|
| 2374 |
+
{
|
| 2375 |
+
"epoch": 1.29,
|
| 2376 |
+
"learning_rate": 1.4210509352834029e-05,
|
| 2377 |
+
"loss": 0.729,
|
| 2378 |
+
"step": 3950
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 1.29,
|
| 2382 |
+
"learning_rate": 1.4095374545845327e-05,
|
| 2383 |
+
"loss": 0.7458,
|
| 2384 |
+
"step": 3960
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 1.29,
|
| 2388 |
+
"learning_rate": 1.3980524659844333e-05,
|
| 2389 |
+
"loss": 0.7398,
|
| 2390 |
+
"step": 3970
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 1.29,
|
| 2394 |
+
"learning_rate": 1.3865962695680836e-05,
|
| 2395 |
+
"loss": 0.7355,
|
| 2396 |
+
"step": 3980
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 1.3,
|
| 2400 |
+
"learning_rate": 1.3751691646681655e-05,
|
| 2401 |
+
"loss": 0.7286,
|
| 2402 |
+
"step": 3990
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 1.3,
|
| 2406 |
+
"learning_rate": 1.363771449857248e-05,
|
| 2407 |
+
"loss": 0.7265,
|
| 2408 |
+
"step": 4000
|
| 2409 |
+
}
|
| 2410 |
+
],
|
| 2411 |
+
"max_steps": 6146,
|
| 2412 |
+
"num_train_epochs": 2,
|
| 2413 |
+
"total_flos": 7.183103765912748e+18,
|
| 2414 |
+
"trial_name": null,
|
| 2415 |
+
"trial_params": null
|
| 2416 |
+
}
|
checkpoint-4000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dafdcf787495b020f1b3f6f6180b816b2753ea01048ce4086dbcd8a2c7a5f666
|
| 3 |
+
size 3265
|
checkpoint-5000/README.md
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
| 4 |
+
## Training procedure
|
| 5 |
+
|
| 6 |
+
|
| 7 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 8 |
+
- load_in_8bit: False
|
| 9 |
+
- load_in_4bit: True
|
| 10 |
+
- llm_int8_threshold: 6.0
|
| 11 |
+
- llm_int8_skip_modules: None
|
| 12 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 13 |
+
- llm_int8_has_fp16_weight: False
|
| 14 |
+
- bnb_4bit_quant_type: nf4
|
| 15 |
+
- bnb_4bit_use_double_quant: True
|
| 16 |
+
- bnb_4bit_compute_dtype: float16
|
| 17 |
+
### Framework versions
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
- PEFT 0.4.0
|
checkpoint-5000/adapter_config.json
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "./Llama-2-7b-chat-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"fan_in_fan_out": false,
|
| 6 |
+
"inference_mode": true,
|
| 7 |
+
"init_lora_weights": true,
|
| 8 |
+
"layers_pattern": null,
|
| 9 |
+
"layers_to_transform": null,
|
| 10 |
+
"lora_alpha": 32.0,
|
| 11 |
+
"lora_dropout": 0.1,
|
| 12 |
+
"modules_to_save": null,
|
| 13 |
+
"peft_type": "LORA",
|
| 14 |
+
"r": 8,
|
| 15 |
+
"revision": null,
|
| 16 |
+
"target_modules": [
|
| 17 |
+
"q_proj",
|
| 18 |
+
"v_proj"
|
| 19 |
+
],
|
| 20 |
+
"task_type": "CAUSAL_LM"
|
| 21 |
+
}
|