Commit
•
c35fb66
1
Parent(s):
b54963c
Upload 13 files
Browse files- README.md +34 -1
- adapter_config.json +21 -0
- adapter_model.bin +3 -0
- checkpoint-3200/optimizer.pt +3 -0
- checkpoint-3200/pytorch_model.bin +3 -0
- checkpoint-3200/rng_state_0.pth +3 -0
- checkpoint-3200/rng_state_1.pth +3 -0
- checkpoint-3200/rng_state_2.pth +3 -0
- checkpoint-3200/rng_state_3.pth +3 -0
- checkpoint-3200/scaler.pt +3 -0
- checkpoint-3200/scheduler.pt +3 -0
- checkpoint-3200/trainer_state.json +2064 -0
- checkpoint-3200/training_args.bin +3 -0
README.md
CHANGED
|
@@ -1,3 +1,36 @@
|
|
| 1 |
---
|
| 2 |
-
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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: True
|
| 9 |
+
- llm_int8_threshold: 6.0
|
| 10 |
+
- llm_int8_skip_modules: None
|
| 11 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 12 |
+
|
| 13 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 14 |
+
- load_in_8bit: True
|
| 15 |
+
- llm_int8_threshold: 6.0
|
| 16 |
+
- llm_int8_skip_modules: None
|
| 17 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 18 |
+
|
| 19 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 20 |
+
- load_in_8bit: True
|
| 21 |
+
- llm_int8_threshold: 6.0
|
| 22 |
+
- llm_int8_skip_modules: None
|
| 23 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 24 |
+
|
| 25 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 26 |
+
- load_in_8bit: True
|
| 27 |
+
- llm_int8_threshold: 6.0
|
| 28 |
+
- llm_int8_skip_modules: None
|
| 29 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 30 |
+
### Framework versions
|
| 31 |
+
|
| 32 |
+
- PEFT 0.5.0.dev0
|
| 33 |
+
- PEFT 0.5.0.dev0
|
| 34 |
+
- PEFT 0.5.0.dev0
|
| 35 |
+
|
| 36 |
+
- PEFT 0.5.0.dev0
|
adapter_config.json
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "../llama/Llama-2-13b-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": 16,
|
| 11 |
+
"lora_dropout": 0.05,
|
| 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:e5e1621f48d9ad8feb1d6d31050275f0aafd080c5c07153301fe2f48411f4406
|
| 3 |
+
size 443
|
checkpoint-3200/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:18a7eec8ed156291603dfe367ff808a9d0b6f64fb3447ce6dce093f31537ab8c
|
| 3 |
+
size 52562757
|
checkpoint-3200/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a5c839c014f0c0b9f5ec7e0a109491d945403ff5b0f9ab59e686368b76a75487
|
| 3 |
+
size 26271757
|
checkpoint-3200/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ed1e172e64e488ec623938fd07511adaba291b041a121dae60e78bd286202b1c
|
| 3 |
+
size 17655
|
checkpoint-3200/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:47819534a1c7d1ffe91be76634083745ca98bfeaf704c2b83b091b256367942c
|
| 3 |
+
size 17655
|
checkpoint-3200/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e965dc1ab23dd00f1dbb2a2bb8d7d2e0e5d439070231715db5834e99d8ee347c
|
| 3 |
+
size 17655
|
checkpoint-3200/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9384c680bbb8d2b8ffdf05613d7317c80fa64cb67058ec9ef07c7d1b956fd6fb
|
| 3 |
+
size 17655
|
checkpoint-3200/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:91aa865b563025841a0478319d8fcedf45846c498a447b3a4f6f95b820f48fa1
|
| 3 |
+
size 557
|
checkpoint-3200/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f5644a99bc25bbc23aaebf311a6fe4201fca5758f83814351af9c6f176c477f5
|
| 3 |
+
size 627
|
checkpoint-3200/trainer_state.json
ADDED
|
@@ -0,0 +1,2064 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.5883459448814392,
|
| 3 |
+
"best_model_checkpoint": "../outputs/7b_lora_0.05/checkpoint-3200",
|
| 4 |
+
"epoch": 0.9644364074743822,
|
| 5 |
+
"global_step": 3200,
|
| 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": 2.9999999999999997e-05,
|
| 13 |
+
"loss": 2.247,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 5.9999999999999995e-05,
|
| 19 |
+
"loss": 2.1539,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 8.999999999999999e-05,
|
| 25 |
+
"loss": 1.7928,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.01,
|
| 30 |
+
"learning_rate": 0.00011999999999999999,
|
| 31 |
+
"loss": 1.2147,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.02,
|
| 36 |
+
"learning_rate": 0.00015,
|
| 37 |
+
"loss": 0.9432,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.02,
|
| 42 |
+
"learning_rate": 0.00017999999999999998,
|
| 43 |
+
"loss": 0.8265,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.02,
|
| 48 |
+
"learning_rate": 0.00020999999999999998,
|
| 49 |
+
"loss": 0.7876,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.02,
|
| 54 |
+
"learning_rate": 0.00023999999999999998,
|
| 55 |
+
"loss": 0.7653,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.03,
|
| 60 |
+
"learning_rate": 0.00027,
|
| 61 |
+
"loss": 0.7297,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.03,
|
| 66 |
+
"learning_rate": 0.0003,
|
| 67 |
+
"loss": 0.6621,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.03,
|
| 72 |
+
"learning_rate": 0.0002990677439403356,
|
| 73 |
+
"loss": 0.6617,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.04,
|
| 78 |
+
"learning_rate": 0.0002981354878806712,
|
| 79 |
+
"loss": 0.6567,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.04,
|
| 84 |
+
"learning_rate": 0.00029720323182100684,
|
| 85 |
+
"loss": 0.6495,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.04,
|
| 90 |
+
"learning_rate": 0.0002962709757613424,
|
| 91 |
+
"loss": 0.6454,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.05,
|
| 96 |
+
"learning_rate": 0.00029533871970167805,
|
| 97 |
+
"loss": 0.6387,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.05,
|
| 102 |
+
"learning_rate": 0.00029440646364201366,
|
| 103 |
+
"loss": 0.6421,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.05,
|
| 108 |
+
"learning_rate": 0.00029347420758234926,
|
| 109 |
+
"loss": 0.6382,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.05,
|
| 114 |
+
"learning_rate": 0.00029254195152268487,
|
| 115 |
+
"loss": 0.64,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.06,
|
| 120 |
+
"learning_rate": 0.0002916096954630205,
|
| 121 |
+
"loss": 0.651,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.06,
|
| 126 |
+
"learning_rate": 0.0002906774394033561,
|
| 127 |
+
"loss": 0.6477,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.06,
|
| 132 |
+
"eval_loss": 0.6271677017211914,
|
| 133 |
+
"eval_runtime": 38.9192,
|
| 134 |
+
"eval_samples_per_second": 51.388,
|
| 135 |
+
"eval_steps_per_second": 1.619,
|
| 136 |
+
"step": 200
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.06,
|
| 140 |
+
"learning_rate": 0.0002897451833436917,
|
| 141 |
+
"loss": 0.6264,
|
| 142 |
+
"step": 210
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.07,
|
| 146 |
+
"learning_rate": 0.00028881292728402734,
|
| 147 |
+
"loss": 0.6315,
|
| 148 |
+
"step": 220
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.07,
|
| 152 |
+
"learning_rate": 0.00028788067122436294,
|
| 153 |
+
"loss": 0.6303,
|
| 154 |
+
"step": 230
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.07,
|
| 158 |
+
"learning_rate": 0.00028694841516469855,
|
| 159 |
+
"loss": 0.6387,
|
| 160 |
+
"step": 240
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.08,
|
| 164 |
+
"learning_rate": 0.00028601615910503415,
|
| 165 |
+
"loss": 0.6267,
|
| 166 |
+
"step": 250
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.08,
|
| 170 |
+
"learning_rate": 0.00028508390304536976,
|
| 171 |
+
"loss": 0.6178,
|
| 172 |
+
"step": 260
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.08,
|
| 176 |
+
"learning_rate": 0.00028415164698570536,
|
| 177 |
+
"loss": 0.6268,
|
| 178 |
+
"step": 270
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.08,
|
| 182 |
+
"learning_rate": 0.00028321939092604097,
|
| 183 |
+
"loss": 0.6212,
|
| 184 |
+
"step": 280
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.09,
|
| 188 |
+
"learning_rate": 0.00028228713486637663,
|
| 189 |
+
"loss": 0.6267,
|
| 190 |
+
"step": 290
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.09,
|
| 194 |
+
"learning_rate": 0.00028135487880671223,
|
| 195 |
+
"loss": 0.6205,
|
| 196 |
+
"step": 300
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.09,
|
| 200 |
+
"learning_rate": 0.00028042262274704784,
|
| 201 |
+
"loss": 0.6217,
|
| 202 |
+
"step": 310
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.1,
|
| 206 |
+
"learning_rate": 0.00027949036668738344,
|
| 207 |
+
"loss": 0.6191,
|
| 208 |
+
"step": 320
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.1,
|
| 212 |
+
"learning_rate": 0.00027855811062771905,
|
| 213 |
+
"loss": 0.6339,
|
| 214 |
+
"step": 330
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.1,
|
| 218 |
+
"learning_rate": 0.00027762585456805465,
|
| 219 |
+
"loss": 0.6272,
|
| 220 |
+
"step": 340
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.11,
|
| 224 |
+
"learning_rate": 0.00027669359850839026,
|
| 225 |
+
"loss": 0.6317,
|
| 226 |
+
"step": 350
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.11,
|
| 230 |
+
"learning_rate": 0.0002757613424487259,
|
| 231 |
+
"loss": 0.6142,
|
| 232 |
+
"step": 360
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.11,
|
| 236 |
+
"learning_rate": 0.0002748290863890615,
|
| 237 |
+
"loss": 0.6348,
|
| 238 |
+
"step": 370
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 0.11,
|
| 242 |
+
"learning_rate": 0.0002738968303293971,
|
| 243 |
+
"loss": 0.6161,
|
| 244 |
+
"step": 380
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.12,
|
| 248 |
+
"learning_rate": 0.00027296457426973273,
|
| 249 |
+
"loss": 0.6258,
|
| 250 |
+
"step": 390
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.12,
|
| 254 |
+
"learning_rate": 0.00027203231821006834,
|
| 255 |
+
"loss": 0.6176,
|
| 256 |
+
"step": 400
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.12,
|
| 260 |
+
"eval_loss": 0.6118931770324707,
|
| 261 |
+
"eval_runtime": 38.809,
|
| 262 |
+
"eval_samples_per_second": 51.534,
|
| 263 |
+
"eval_steps_per_second": 1.623,
|
| 264 |
+
"step": 400
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.12,
|
| 268 |
+
"learning_rate": 0.00027110006215040394,
|
| 269 |
+
"loss": 0.616,
|
| 270 |
+
"step": 410
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.13,
|
| 274 |
+
"learning_rate": 0.00027016780609073954,
|
| 275 |
+
"loss": 0.6126,
|
| 276 |
+
"step": 420
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.13,
|
| 280 |
+
"learning_rate": 0.0002692355500310752,
|
| 281 |
+
"loss": 0.6292,
|
| 282 |
+
"step": 430
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.13,
|
| 286 |
+
"learning_rate": 0.0002683032939714108,
|
| 287 |
+
"loss": 0.6165,
|
| 288 |
+
"step": 440
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.14,
|
| 292 |
+
"learning_rate": 0.0002673710379117464,
|
| 293 |
+
"loss": 0.6253,
|
| 294 |
+
"step": 450
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.14,
|
| 298 |
+
"learning_rate": 0.000266438781852082,
|
| 299 |
+
"loss": 0.6043,
|
| 300 |
+
"step": 460
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 0.14,
|
| 304 |
+
"learning_rate": 0.0002655065257924176,
|
| 305 |
+
"loss": 0.602,
|
| 306 |
+
"step": 470
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.14,
|
| 310 |
+
"learning_rate": 0.00026457426973275323,
|
| 311 |
+
"loss": 0.6207,
|
| 312 |
+
"step": 480
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.15,
|
| 316 |
+
"learning_rate": 0.00026364201367308883,
|
| 317 |
+
"loss": 0.6093,
|
| 318 |
+
"step": 490
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.15,
|
| 322 |
+
"learning_rate": 0.0002627097576134245,
|
| 323 |
+
"loss": 0.6087,
|
| 324 |
+
"step": 500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.15,
|
| 328 |
+
"learning_rate": 0.00026177750155376004,
|
| 329 |
+
"loss": 0.6149,
|
| 330 |
+
"step": 510
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 0.16,
|
| 334 |
+
"learning_rate": 0.0002608452454940957,
|
| 335 |
+
"loss": 0.6247,
|
| 336 |
+
"step": 520
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 0.16,
|
| 340 |
+
"learning_rate": 0.0002599129894344313,
|
| 341 |
+
"loss": 0.6071,
|
| 342 |
+
"step": 530
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.16,
|
| 346 |
+
"learning_rate": 0.0002589807333747669,
|
| 347 |
+
"loss": 0.6132,
|
| 348 |
+
"step": 540
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 0.17,
|
| 352 |
+
"learning_rate": 0.0002580484773151025,
|
| 353 |
+
"loss": 0.6248,
|
| 354 |
+
"step": 550
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.17,
|
| 358 |
+
"learning_rate": 0.0002571162212554382,
|
| 359 |
+
"loss": 0.6136,
|
| 360 |
+
"step": 560
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.17,
|
| 364 |
+
"learning_rate": 0.0002561839651957738,
|
| 365 |
+
"loss": 0.6133,
|
| 366 |
+
"step": 570
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.17,
|
| 370 |
+
"learning_rate": 0.00025525170913610933,
|
| 371 |
+
"loss": 0.6211,
|
| 372 |
+
"step": 580
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 0.18,
|
| 376 |
+
"learning_rate": 0.000254319453076445,
|
| 377 |
+
"loss": 0.6241,
|
| 378 |
+
"step": 590
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 0.18,
|
| 382 |
+
"learning_rate": 0.0002533871970167806,
|
| 383 |
+
"loss": 0.6193,
|
| 384 |
+
"step": 600
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.18,
|
| 388 |
+
"eval_loss": 0.6061509847640991,
|
| 389 |
+
"eval_runtime": 38.8987,
|
| 390 |
+
"eval_samples_per_second": 51.416,
|
| 391 |
+
"eval_steps_per_second": 1.62,
|
| 392 |
+
"step": 600
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.18,
|
| 396 |
+
"learning_rate": 0.0002524549409571162,
|
| 397 |
+
"loss": 0.6063,
|
| 398 |
+
"step": 610
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.19,
|
| 402 |
+
"learning_rate": 0.0002515226848974518,
|
| 403 |
+
"loss": 0.6095,
|
| 404 |
+
"step": 620
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.19,
|
| 408 |
+
"learning_rate": 0.00025059042883778746,
|
| 409 |
+
"loss": 0.6135,
|
| 410 |
+
"step": 630
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.19,
|
| 414 |
+
"learning_rate": 0.000249658172778123,
|
| 415 |
+
"loss": 0.6064,
|
| 416 |
+
"step": 640
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.2,
|
| 420 |
+
"learning_rate": 0.0002487259167184586,
|
| 421 |
+
"loss": 0.625,
|
| 422 |
+
"step": 650
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.2,
|
| 426 |
+
"learning_rate": 0.0002477936606587943,
|
| 427 |
+
"loss": 0.6062,
|
| 428 |
+
"step": 660
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.2,
|
| 432 |
+
"learning_rate": 0.0002468614045991299,
|
| 433 |
+
"loss": 0.6116,
|
| 434 |
+
"step": 670
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.2,
|
| 438 |
+
"learning_rate": 0.0002459291485394655,
|
| 439 |
+
"loss": 0.6064,
|
| 440 |
+
"step": 680
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.21,
|
| 444 |
+
"learning_rate": 0.0002449968924798011,
|
| 445 |
+
"loss": 0.6225,
|
| 446 |
+
"step": 690
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.21,
|
| 450 |
+
"learning_rate": 0.0002440646364201367,
|
| 451 |
+
"loss": 0.621,
|
| 452 |
+
"step": 700
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.21,
|
| 456 |
+
"learning_rate": 0.0002431323803604723,
|
| 457 |
+
"loss": 0.622,
|
| 458 |
+
"step": 710
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.22,
|
| 462 |
+
"learning_rate": 0.00024220012430080793,
|
| 463 |
+
"loss": 0.6128,
|
| 464 |
+
"step": 720
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.22,
|
| 468 |
+
"learning_rate": 0.00024126786824114354,
|
| 469 |
+
"loss": 0.6186,
|
| 470 |
+
"step": 730
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.22,
|
| 474 |
+
"learning_rate": 0.00024033561218147917,
|
| 475 |
+
"loss": 0.6143,
|
| 476 |
+
"step": 740
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.23,
|
| 480 |
+
"learning_rate": 0.00023940335612181477,
|
| 481 |
+
"loss": 0.6224,
|
| 482 |
+
"step": 750
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.23,
|
| 486 |
+
"learning_rate": 0.00023847110006215038,
|
| 487 |
+
"loss": 0.5982,
|
| 488 |
+
"step": 760
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.23,
|
| 492 |
+
"learning_rate": 0.00023753884400248598,
|
| 493 |
+
"loss": 0.6107,
|
| 494 |
+
"step": 770
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.24,
|
| 498 |
+
"learning_rate": 0.0002366065879428216,
|
| 499 |
+
"loss": 0.6027,
|
| 500 |
+
"step": 780
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.24,
|
| 504 |
+
"learning_rate": 0.00023567433188315722,
|
| 505 |
+
"loss": 0.6086,
|
| 506 |
+
"step": 790
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.24,
|
| 510 |
+
"learning_rate": 0.00023474207582349283,
|
| 511 |
+
"loss": 0.6016,
|
| 512 |
+
"step": 800
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.24,
|
| 516 |
+
"eval_loss": 0.6028618216514587,
|
| 517 |
+
"eval_runtime": 38.8192,
|
| 518 |
+
"eval_samples_per_second": 51.521,
|
| 519 |
+
"eval_steps_per_second": 1.623,
|
| 520 |
+
"step": 800
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.24,
|
| 524 |
+
"learning_rate": 0.00023380981976382846,
|
| 525 |
+
"loss": 0.6162,
|
| 526 |
+
"step": 810
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.25,
|
| 530 |
+
"learning_rate": 0.00023287756370416406,
|
| 531 |
+
"loss": 0.618,
|
| 532 |
+
"step": 820
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 0.25,
|
| 536 |
+
"learning_rate": 0.00023194530764449967,
|
| 537 |
+
"loss": 0.6205,
|
| 538 |
+
"step": 830
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 0.25,
|
| 542 |
+
"learning_rate": 0.00023101305158483527,
|
| 543 |
+
"loss": 0.61,
|
| 544 |
+
"step": 840
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.26,
|
| 548 |
+
"learning_rate": 0.0002300807955251709,
|
| 549 |
+
"loss": 0.6099,
|
| 550 |
+
"step": 850
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.26,
|
| 554 |
+
"learning_rate": 0.0002291485394655065,
|
| 555 |
+
"loss": 0.6107,
|
| 556 |
+
"step": 860
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.26,
|
| 560 |
+
"learning_rate": 0.00022821628340584211,
|
| 561 |
+
"loss": 0.6038,
|
| 562 |
+
"step": 870
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.27,
|
| 566 |
+
"learning_rate": 0.00022728402734617775,
|
| 567 |
+
"loss": 0.6156,
|
| 568 |
+
"step": 880
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.27,
|
| 572 |
+
"learning_rate": 0.00022635177128651332,
|
| 573 |
+
"loss": 0.5985,
|
| 574 |
+
"step": 890
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 0.27,
|
| 578 |
+
"learning_rate": 0.00022541951522684896,
|
| 579 |
+
"loss": 0.6151,
|
| 580 |
+
"step": 900
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 0.27,
|
| 584 |
+
"learning_rate": 0.00022448725916718456,
|
| 585 |
+
"loss": 0.6065,
|
| 586 |
+
"step": 910
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 0.28,
|
| 590 |
+
"learning_rate": 0.0002235550031075202,
|
| 591 |
+
"loss": 0.6098,
|
| 592 |
+
"step": 920
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 0.28,
|
| 596 |
+
"learning_rate": 0.0002226227470478558,
|
| 597 |
+
"loss": 0.6095,
|
| 598 |
+
"step": 930
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.28,
|
| 602 |
+
"learning_rate": 0.00022169049098819143,
|
| 603 |
+
"loss": 0.6104,
|
| 604 |
+
"step": 940
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.29,
|
| 608 |
+
"learning_rate": 0.000220758234928527,
|
| 609 |
+
"loss": 0.6116,
|
| 610 |
+
"step": 950
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 0.29,
|
| 614 |
+
"learning_rate": 0.0002198259788688626,
|
| 615 |
+
"loss": 0.6018,
|
| 616 |
+
"step": 960
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 0.29,
|
| 620 |
+
"learning_rate": 0.00021889372280919824,
|
| 621 |
+
"loss": 0.613,
|
| 622 |
+
"step": 970
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 0.3,
|
| 626 |
+
"learning_rate": 0.00021796146674953385,
|
| 627 |
+
"loss": 0.6101,
|
| 628 |
+
"step": 980
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 0.3,
|
| 632 |
+
"learning_rate": 0.00021702921068986948,
|
| 633 |
+
"loss": 0.6087,
|
| 634 |
+
"step": 990
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 0.3,
|
| 638 |
+
"learning_rate": 0.00021609695463020508,
|
| 639 |
+
"loss": 0.606,
|
| 640 |
+
"step": 1000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.3,
|
| 644 |
+
"eval_loss": 0.5999864339828491,
|
| 645 |
+
"eval_runtime": 38.8621,
|
| 646 |
+
"eval_samples_per_second": 51.464,
|
| 647 |
+
"eval_steps_per_second": 1.621,
|
| 648 |
+
"step": 1000
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 0.3,
|
| 652 |
+
"learning_rate": 0.00021516469857054072,
|
| 653 |
+
"loss": 0.6162,
|
| 654 |
+
"step": 1010
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.31,
|
| 658 |
+
"learning_rate": 0.0002142324425108763,
|
| 659 |
+
"loss": 0.6055,
|
| 660 |
+
"step": 1020
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.31,
|
| 664 |
+
"learning_rate": 0.0002133001864512119,
|
| 665 |
+
"loss": 0.6071,
|
| 666 |
+
"step": 1030
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 0.31,
|
| 670 |
+
"learning_rate": 0.00021236793039154753,
|
| 671 |
+
"loss": 0.6127,
|
| 672 |
+
"step": 1040
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 0.32,
|
| 676 |
+
"learning_rate": 0.00021143567433188314,
|
| 677 |
+
"loss": 0.6111,
|
| 678 |
+
"step": 1050
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 0.32,
|
| 682 |
+
"learning_rate": 0.00021050341827221877,
|
| 683 |
+
"loss": 0.6108,
|
| 684 |
+
"step": 1060
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.32,
|
| 688 |
+
"learning_rate": 0.00020957116221255437,
|
| 689 |
+
"loss": 0.6036,
|
| 690 |
+
"step": 1070
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.33,
|
| 694 |
+
"learning_rate": 0.00020863890615288995,
|
| 695 |
+
"loss": 0.5965,
|
| 696 |
+
"step": 1080
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.33,
|
| 700 |
+
"learning_rate": 0.00020770665009322558,
|
| 701 |
+
"loss": 0.5953,
|
| 702 |
+
"step": 1090
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.33,
|
| 706 |
+
"learning_rate": 0.0002067743940335612,
|
| 707 |
+
"loss": 0.6056,
|
| 708 |
+
"step": 1100
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 0.33,
|
| 712 |
+
"learning_rate": 0.00020584213797389682,
|
| 713 |
+
"loss": 0.6036,
|
| 714 |
+
"step": 1110
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 0.34,
|
| 718 |
+
"learning_rate": 0.00020490988191423242,
|
| 719 |
+
"loss": 0.5897,
|
| 720 |
+
"step": 1120
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.34,
|
| 724 |
+
"learning_rate": 0.00020397762585456806,
|
| 725 |
+
"loss": 0.6104,
|
| 726 |
+
"step": 1130
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 0.34,
|
| 730 |
+
"learning_rate": 0.00020304536979490363,
|
| 731 |
+
"loss": 0.6159,
|
| 732 |
+
"step": 1140
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.35,
|
| 736 |
+
"learning_rate": 0.00020211311373523924,
|
| 737 |
+
"loss": 0.6044,
|
| 738 |
+
"step": 1150
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.35,
|
| 742 |
+
"learning_rate": 0.00020118085767557487,
|
| 743 |
+
"loss": 0.5959,
|
| 744 |
+
"step": 1160
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.35,
|
| 748 |
+
"learning_rate": 0.00020024860161591048,
|
| 749 |
+
"loss": 0.6102,
|
| 750 |
+
"step": 1170
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 0.36,
|
| 754 |
+
"learning_rate": 0.0001993163455562461,
|
| 755 |
+
"loss": 0.5941,
|
| 756 |
+
"step": 1180
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 0.36,
|
| 760 |
+
"learning_rate": 0.0001983840894965817,
|
| 761 |
+
"loss": 0.6033,
|
| 762 |
+
"step": 1190
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 0.36,
|
| 766 |
+
"learning_rate": 0.00019745183343691732,
|
| 767 |
+
"loss": 0.605,
|
| 768 |
+
"step": 1200
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 0.36,
|
| 772 |
+
"eval_loss": 0.5981193780899048,
|
| 773 |
+
"eval_runtime": 38.9735,
|
| 774 |
+
"eval_samples_per_second": 51.317,
|
| 775 |
+
"eval_steps_per_second": 1.616,
|
| 776 |
+
"step": 1200
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.36,
|
| 780 |
+
"learning_rate": 0.00019651957737725292,
|
| 781 |
+
"loss": 0.5959,
|
| 782 |
+
"step": 1210
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.37,
|
| 786 |
+
"learning_rate": 0.00019558732131758855,
|
| 787 |
+
"loss": 0.5907,
|
| 788 |
+
"step": 1220
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.37,
|
| 792 |
+
"learning_rate": 0.00019465506525792416,
|
| 793 |
+
"loss": 0.5995,
|
| 794 |
+
"step": 1230
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.37,
|
| 798 |
+
"learning_rate": 0.00019372280919825976,
|
| 799 |
+
"loss": 0.609,
|
| 800 |
+
"step": 1240
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.38,
|
| 804 |
+
"learning_rate": 0.0001927905531385954,
|
| 805 |
+
"loss": 0.6117,
|
| 806 |
+
"step": 1250
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.38,
|
| 810 |
+
"learning_rate": 0.000191858297078931,
|
| 811 |
+
"loss": 0.6081,
|
| 812 |
+
"step": 1260
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.38,
|
| 816 |
+
"learning_rate": 0.0001909260410192666,
|
| 817 |
+
"loss": 0.6005,
|
| 818 |
+
"step": 1270
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.39,
|
| 822 |
+
"learning_rate": 0.0001899937849596022,
|
| 823 |
+
"loss": 0.5991,
|
| 824 |
+
"step": 1280
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.39,
|
| 828 |
+
"learning_rate": 0.00018906152889993784,
|
| 829 |
+
"loss": 0.5992,
|
| 830 |
+
"step": 1290
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.39,
|
| 834 |
+
"learning_rate": 0.00018812927284027345,
|
| 835 |
+
"loss": 0.6054,
|
| 836 |
+
"step": 1300
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.39,
|
| 840 |
+
"learning_rate": 0.00018719701678060908,
|
| 841 |
+
"loss": 0.6078,
|
| 842 |
+
"step": 1310
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.4,
|
| 846 |
+
"learning_rate": 0.00018626476072094468,
|
| 847 |
+
"loss": 0.5962,
|
| 848 |
+
"step": 1320
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.4,
|
| 852 |
+
"learning_rate": 0.00018533250466128026,
|
| 853 |
+
"loss": 0.5958,
|
| 854 |
+
"step": 1330
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.4,
|
| 858 |
+
"learning_rate": 0.0001844002486016159,
|
| 859 |
+
"loss": 0.6063,
|
| 860 |
+
"step": 1340
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.41,
|
| 864 |
+
"learning_rate": 0.0001834679925419515,
|
| 865 |
+
"loss": 0.603,
|
| 866 |
+
"step": 1350
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.41,
|
| 870 |
+
"learning_rate": 0.00018253573648228713,
|
| 871 |
+
"loss": 0.6126,
|
| 872 |
+
"step": 1360
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.41,
|
| 876 |
+
"learning_rate": 0.00018160348042262273,
|
| 877 |
+
"loss": 0.5996,
|
| 878 |
+
"step": 1370
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.42,
|
| 882 |
+
"learning_rate": 0.00018067122436295837,
|
| 883 |
+
"loss": 0.6068,
|
| 884 |
+
"step": 1380
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.42,
|
| 888 |
+
"learning_rate": 0.00017973896830329394,
|
| 889 |
+
"loss": 0.6055,
|
| 890 |
+
"step": 1390
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.42,
|
| 894 |
+
"learning_rate": 0.00017880671224362955,
|
| 895 |
+
"loss": 0.5939,
|
| 896 |
+
"step": 1400
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.42,
|
| 900 |
+
"eval_loss": 0.596312940120697,
|
| 901 |
+
"eval_runtime": 39.1962,
|
| 902 |
+
"eval_samples_per_second": 51.025,
|
| 903 |
+
"eval_steps_per_second": 1.607,
|
| 904 |
+
"step": 1400
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 0.42,
|
| 908 |
+
"learning_rate": 0.00017787445618396518,
|
| 909 |
+
"loss": 0.6049,
|
| 910 |
+
"step": 1410
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 0.43,
|
| 914 |
+
"learning_rate": 0.00017694220012430079,
|
| 915 |
+
"loss": 0.5989,
|
| 916 |
+
"step": 1420
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 0.43,
|
| 920 |
+
"learning_rate": 0.00017600994406463642,
|
| 921 |
+
"loss": 0.6064,
|
| 922 |
+
"step": 1430
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 0.43,
|
| 926 |
+
"learning_rate": 0.00017507768800497202,
|
| 927 |
+
"loss": 0.6158,
|
| 928 |
+
"step": 1440
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 0.44,
|
| 932 |
+
"learning_rate": 0.00017414543194530765,
|
| 933 |
+
"loss": 0.5992,
|
| 934 |
+
"step": 1450
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.44,
|
| 938 |
+
"learning_rate": 0.00017321317588564323,
|
| 939 |
+
"loss": 0.6047,
|
| 940 |
+
"step": 1460
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.44,
|
| 944 |
+
"learning_rate": 0.00017228091982597884,
|
| 945 |
+
"loss": 0.6053,
|
| 946 |
+
"step": 1470
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 0.45,
|
| 950 |
+
"learning_rate": 0.00017134866376631447,
|
| 951 |
+
"loss": 0.6072,
|
| 952 |
+
"step": 1480
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 0.45,
|
| 956 |
+
"learning_rate": 0.00017041640770665007,
|
| 957 |
+
"loss": 0.5979,
|
| 958 |
+
"step": 1490
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 0.45,
|
| 962 |
+
"learning_rate": 0.0001694841516469857,
|
| 963 |
+
"loss": 0.6035,
|
| 964 |
+
"step": 1500
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.46,
|
| 968 |
+
"learning_rate": 0.0001685518955873213,
|
| 969 |
+
"loss": 0.6015,
|
| 970 |
+
"step": 1510
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 0.46,
|
| 974 |
+
"learning_rate": 0.00016761963952765691,
|
| 975 |
+
"loss": 0.6049,
|
| 976 |
+
"step": 1520
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.46,
|
| 980 |
+
"learning_rate": 0.00016668738346799252,
|
| 981 |
+
"loss": 0.5886,
|
| 982 |
+
"step": 1530
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.46,
|
| 986 |
+
"learning_rate": 0.00016575512740832812,
|
| 987 |
+
"loss": 0.6108,
|
| 988 |
+
"step": 1540
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 0.47,
|
| 992 |
+
"learning_rate": 0.00016482287134866376,
|
| 993 |
+
"loss": 0.6053,
|
| 994 |
+
"step": 1550
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 0.47,
|
| 998 |
+
"learning_rate": 0.00016389061528899936,
|
| 999 |
+
"loss": 0.6172,
|
| 1000 |
+
"step": 1560
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 0.47,
|
| 1004 |
+
"learning_rate": 0.000162958359229335,
|
| 1005 |
+
"loss": 0.6019,
|
| 1006 |
+
"step": 1570
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.48,
|
| 1010 |
+
"learning_rate": 0.00016202610316967057,
|
| 1011 |
+
"loss": 0.5979,
|
| 1012 |
+
"step": 1580
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 0.48,
|
| 1016 |
+
"learning_rate": 0.0001610938471100062,
|
| 1017 |
+
"loss": 0.5978,
|
| 1018 |
+
"step": 1590
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.48,
|
| 1022 |
+
"learning_rate": 0.0001601615910503418,
|
| 1023 |
+
"loss": 0.6038,
|
| 1024 |
+
"step": 1600
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.48,
|
| 1028 |
+
"eval_loss": 0.5952789187431335,
|
| 1029 |
+
"eval_runtime": 38.92,
|
| 1030 |
+
"eval_samples_per_second": 51.387,
|
| 1031 |
+
"eval_steps_per_second": 1.619,
|
| 1032 |
+
"step": 1600
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.49,
|
| 1036 |
+
"learning_rate": 0.00015922933499067744,
|
| 1037 |
+
"loss": 0.6076,
|
| 1038 |
+
"step": 1610
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.49,
|
| 1042 |
+
"learning_rate": 0.00015829707893101304,
|
| 1043 |
+
"loss": 0.5901,
|
| 1044 |
+
"step": 1620
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 0.49,
|
| 1048 |
+
"learning_rate": 0.00015736482287134865,
|
| 1049 |
+
"loss": 0.6019,
|
| 1050 |
+
"step": 1630
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 0.49,
|
| 1054 |
+
"learning_rate": 0.00015643256681168425,
|
| 1055 |
+
"loss": 0.6007,
|
| 1056 |
+
"step": 1640
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 0.5,
|
| 1060 |
+
"learning_rate": 0.00015550031075201986,
|
| 1061 |
+
"loss": 0.5994,
|
| 1062 |
+
"step": 1650
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 0.5,
|
| 1066 |
+
"learning_rate": 0.0001545680546923555,
|
| 1067 |
+
"loss": 0.5997,
|
| 1068 |
+
"step": 1660
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 0.5,
|
| 1072 |
+
"learning_rate": 0.0001536357986326911,
|
| 1073 |
+
"loss": 0.606,
|
| 1074 |
+
"step": 1670
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.51,
|
| 1078 |
+
"learning_rate": 0.00015270354257302673,
|
| 1079 |
+
"loss": 0.5983,
|
| 1080 |
+
"step": 1680
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.51,
|
| 1084 |
+
"learning_rate": 0.00015177128651336233,
|
| 1085 |
+
"loss": 0.5898,
|
| 1086 |
+
"step": 1690
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 0.51,
|
| 1090 |
+
"learning_rate": 0.00015083903045369796,
|
| 1091 |
+
"loss": 0.6042,
|
| 1092 |
+
"step": 1700
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 0.52,
|
| 1096 |
+
"learning_rate": 0.00014990677439403354,
|
| 1097 |
+
"loss": 0.5931,
|
| 1098 |
+
"step": 1710
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.52,
|
| 1102 |
+
"learning_rate": 0.00014897451833436915,
|
| 1103 |
+
"loss": 0.5964,
|
| 1104 |
+
"step": 1720
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 0.52,
|
| 1108 |
+
"learning_rate": 0.00014804226227470478,
|
| 1109 |
+
"loss": 0.6073,
|
| 1110 |
+
"step": 1730
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 0.52,
|
| 1114 |
+
"learning_rate": 0.00014711000621504038,
|
| 1115 |
+
"loss": 0.5911,
|
| 1116 |
+
"step": 1740
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.53,
|
| 1120 |
+
"learning_rate": 0.000146177750155376,
|
| 1121 |
+
"loss": 0.5921,
|
| 1122 |
+
"step": 1750
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.53,
|
| 1126 |
+
"learning_rate": 0.0001452454940957116,
|
| 1127 |
+
"loss": 0.6007,
|
| 1128 |
+
"step": 1760
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 0.53,
|
| 1132 |
+
"learning_rate": 0.00014431323803604723,
|
| 1133 |
+
"loss": 0.5974,
|
| 1134 |
+
"step": 1770
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 0.54,
|
| 1138 |
+
"learning_rate": 0.00014338098197638283,
|
| 1139 |
+
"loss": 0.6019,
|
| 1140 |
+
"step": 1780
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 0.54,
|
| 1144 |
+
"learning_rate": 0.00014244872591671843,
|
| 1145 |
+
"loss": 0.6049,
|
| 1146 |
+
"step": 1790
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 0.54,
|
| 1150 |
+
"learning_rate": 0.00014151646985705407,
|
| 1151 |
+
"loss": 0.5953,
|
| 1152 |
+
"step": 1800
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 0.54,
|
| 1156 |
+
"eval_loss": 0.5934386253356934,
|
| 1157 |
+
"eval_runtime": 38.9572,
|
| 1158 |
+
"eval_samples_per_second": 51.338,
|
| 1159 |
+
"eval_steps_per_second": 1.617,
|
| 1160 |
+
"step": 1800
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.55,
|
| 1164 |
+
"learning_rate": 0.00014058421379738967,
|
| 1165 |
+
"loss": 0.597,
|
| 1166 |
+
"step": 1810
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 0.55,
|
| 1170 |
+
"learning_rate": 0.00013965195773772528,
|
| 1171 |
+
"loss": 0.6122,
|
| 1172 |
+
"step": 1820
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.55,
|
| 1176 |
+
"learning_rate": 0.0001387197016780609,
|
| 1177 |
+
"loss": 0.6017,
|
| 1178 |
+
"step": 1830
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.55,
|
| 1182 |
+
"learning_rate": 0.0001377874456183965,
|
| 1183 |
+
"loss": 0.6007,
|
| 1184 |
+
"step": 1840
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 0.56,
|
| 1188 |
+
"learning_rate": 0.00013685518955873212,
|
| 1189 |
+
"loss": 0.5982,
|
| 1190 |
+
"step": 1850
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 0.56,
|
| 1194 |
+
"learning_rate": 0.00013592293349906772,
|
| 1195 |
+
"loss": 0.6025,
|
| 1196 |
+
"step": 1860
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.56,
|
| 1200 |
+
"learning_rate": 0.00013499067743940335,
|
| 1201 |
+
"loss": 0.5958,
|
| 1202 |
+
"step": 1870
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.57,
|
| 1206 |
+
"learning_rate": 0.00013405842137973896,
|
| 1207 |
+
"loss": 0.5963,
|
| 1208 |
+
"step": 1880
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 0.57,
|
| 1212 |
+
"learning_rate": 0.00013312616532007456,
|
| 1213 |
+
"loss": 0.6054,
|
| 1214 |
+
"step": 1890
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.57,
|
| 1218 |
+
"learning_rate": 0.0001321939092604102,
|
| 1219 |
+
"loss": 0.5936,
|
| 1220 |
+
"step": 1900
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.58,
|
| 1224 |
+
"learning_rate": 0.00013126165320074577,
|
| 1225 |
+
"loss": 0.5929,
|
| 1226 |
+
"step": 1910
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 0.58,
|
| 1230 |
+
"learning_rate": 0.0001303293971410814,
|
| 1231 |
+
"loss": 0.602,
|
| 1232 |
+
"step": 1920
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 0.58,
|
| 1236 |
+
"learning_rate": 0.000129397141081417,
|
| 1237 |
+
"loss": 0.6012,
|
| 1238 |
+
"step": 1930
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.58,
|
| 1242 |
+
"learning_rate": 0.00012846488502175262,
|
| 1243 |
+
"loss": 0.5999,
|
| 1244 |
+
"step": 1940
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 0.59,
|
| 1248 |
+
"learning_rate": 0.00012753262896208825,
|
| 1249 |
+
"loss": 0.5978,
|
| 1250 |
+
"step": 1950
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 0.59,
|
| 1254 |
+
"learning_rate": 0.00012660037290242385,
|
| 1255 |
+
"loss": 0.6024,
|
| 1256 |
+
"step": 1960
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.59,
|
| 1260 |
+
"learning_rate": 0.00012566811684275946,
|
| 1261 |
+
"loss": 0.5854,
|
| 1262 |
+
"step": 1970
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.6,
|
| 1266 |
+
"learning_rate": 0.0001247358607830951,
|
| 1267 |
+
"loss": 0.5979,
|
| 1268 |
+
"step": 1980
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 0.6,
|
| 1272 |
+
"learning_rate": 0.0001238036047234307,
|
| 1273 |
+
"loss": 0.5964,
|
| 1274 |
+
"step": 1990
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 0.6,
|
| 1278 |
+
"learning_rate": 0.0001228713486637663,
|
| 1279 |
+
"loss": 0.6042,
|
| 1280 |
+
"step": 2000
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.6,
|
| 1284 |
+
"eval_loss": 0.5928096771240234,
|
| 1285 |
+
"eval_runtime": 38.9501,
|
| 1286 |
+
"eval_samples_per_second": 51.348,
|
| 1287 |
+
"eval_steps_per_second": 1.617,
|
| 1288 |
+
"step": 2000
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 0.61,
|
| 1292 |
+
"learning_rate": 0.00012193909260410192,
|
| 1293 |
+
"loss": 0.6045,
|
| 1294 |
+
"step": 2010
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 0.61,
|
| 1298 |
+
"learning_rate": 0.00012100683654443754,
|
| 1299 |
+
"loss": 0.5995,
|
| 1300 |
+
"step": 2020
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 0.61,
|
| 1304 |
+
"learning_rate": 0.00012007458048477315,
|
| 1305 |
+
"loss": 0.6149,
|
| 1306 |
+
"step": 2030
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 0.61,
|
| 1310 |
+
"learning_rate": 0.00011914232442510874,
|
| 1311 |
+
"loss": 0.5979,
|
| 1312 |
+
"step": 2040
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 0.62,
|
| 1316 |
+
"learning_rate": 0.00011821006836544436,
|
| 1317 |
+
"loss": 0.5905,
|
| 1318 |
+
"step": 2050
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.62,
|
| 1322 |
+
"learning_rate": 0.00011727781230577998,
|
| 1323 |
+
"loss": 0.5985,
|
| 1324 |
+
"step": 2060
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 0.62,
|
| 1328 |
+
"learning_rate": 0.00011634555624611559,
|
| 1329 |
+
"loss": 0.6088,
|
| 1330 |
+
"step": 2070
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 0.63,
|
| 1334 |
+
"learning_rate": 0.0001154133001864512,
|
| 1335 |
+
"loss": 0.6097,
|
| 1336 |
+
"step": 2080
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 0.63,
|
| 1340 |
+
"learning_rate": 0.00011448104412678682,
|
| 1341 |
+
"loss": 0.6012,
|
| 1342 |
+
"step": 2090
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 0.63,
|
| 1346 |
+
"learning_rate": 0.00011354878806712241,
|
| 1347 |
+
"loss": 0.603,
|
| 1348 |
+
"step": 2100
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 0.64,
|
| 1352 |
+
"learning_rate": 0.00011261653200745803,
|
| 1353 |
+
"loss": 0.5894,
|
| 1354 |
+
"step": 2110
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 0.64,
|
| 1358 |
+
"learning_rate": 0.00011168427594779365,
|
| 1359 |
+
"loss": 0.5879,
|
| 1360 |
+
"step": 2120
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.64,
|
| 1364 |
+
"learning_rate": 0.00011075201988812926,
|
| 1365 |
+
"loss": 0.602,
|
| 1366 |
+
"step": 2130
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 0.64,
|
| 1370 |
+
"learning_rate": 0.00010981976382846487,
|
| 1371 |
+
"loss": 0.5899,
|
| 1372 |
+
"step": 2140
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 0.65,
|
| 1376 |
+
"learning_rate": 0.00010888750776880049,
|
| 1377 |
+
"loss": 0.5918,
|
| 1378 |
+
"step": 2150
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.65,
|
| 1382 |
+
"learning_rate": 0.0001079552517091361,
|
| 1383 |
+
"loss": 0.5957,
|
| 1384 |
+
"step": 2160
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 0.65,
|
| 1388 |
+
"learning_rate": 0.00010702299564947172,
|
| 1389 |
+
"loss": 0.5982,
|
| 1390 |
+
"step": 2170
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 0.66,
|
| 1394 |
+
"learning_rate": 0.00010609073958980733,
|
| 1395 |
+
"loss": 0.5832,
|
| 1396 |
+
"step": 2180
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 0.66,
|
| 1400 |
+
"learning_rate": 0.00010515848353014293,
|
| 1401 |
+
"loss": 0.5965,
|
| 1402 |
+
"step": 2190
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.66,
|
| 1406 |
+
"learning_rate": 0.00010422622747047854,
|
| 1407 |
+
"loss": 0.5968,
|
| 1408 |
+
"step": 2200
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 0.66,
|
| 1412 |
+
"eval_loss": 0.5914697647094727,
|
| 1413 |
+
"eval_runtime": 38.9209,
|
| 1414 |
+
"eval_samples_per_second": 51.386,
|
| 1415 |
+
"eval_steps_per_second": 1.619,
|
| 1416 |
+
"step": 2200
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.67,
|
| 1420 |
+
"learning_rate": 0.00010329397141081416,
|
| 1421 |
+
"loss": 0.5961,
|
| 1422 |
+
"step": 2210
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 0.67,
|
| 1426 |
+
"learning_rate": 0.00010236171535114978,
|
| 1427 |
+
"loss": 0.591,
|
| 1428 |
+
"step": 2220
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 0.67,
|
| 1432 |
+
"learning_rate": 0.00010142945929148539,
|
| 1433 |
+
"loss": 0.6009,
|
| 1434 |
+
"step": 2230
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.68,
|
| 1438 |
+
"learning_rate": 0.000100497203231821,
|
| 1439 |
+
"loss": 0.581,
|
| 1440 |
+
"step": 2240
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 0.68,
|
| 1444 |
+
"learning_rate": 9.956494717215662e-05,
|
| 1445 |
+
"loss": 0.58,
|
| 1446 |
+
"step": 2250
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 0.68,
|
| 1450 |
+
"learning_rate": 9.863269111249221e-05,
|
| 1451 |
+
"loss": 0.5942,
|
| 1452 |
+
"step": 2260
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.68,
|
| 1456 |
+
"learning_rate": 9.770043505282783e-05,
|
| 1457 |
+
"loss": 0.6004,
|
| 1458 |
+
"step": 2270
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.69,
|
| 1462 |
+
"learning_rate": 9.676817899316345e-05,
|
| 1463 |
+
"loss": 0.6008,
|
| 1464 |
+
"step": 2280
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 0.69,
|
| 1468 |
+
"learning_rate": 9.583592293349906e-05,
|
| 1469 |
+
"loss": 0.5915,
|
| 1470 |
+
"step": 2290
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 0.69,
|
| 1474 |
+
"learning_rate": 9.490366687383467e-05,
|
| 1475 |
+
"loss": 0.5905,
|
| 1476 |
+
"step": 2300
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 0.7,
|
| 1480 |
+
"learning_rate": 9.397141081417029e-05,
|
| 1481 |
+
"loss": 0.6123,
|
| 1482 |
+
"step": 2310
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 0.7,
|
| 1486 |
+
"learning_rate": 9.30391547545059e-05,
|
| 1487 |
+
"loss": 0.5946,
|
| 1488 |
+
"step": 2320
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 0.7,
|
| 1492 |
+
"learning_rate": 9.21068986948415e-05,
|
| 1493 |
+
"loss": 0.5816,
|
| 1494 |
+
"step": 2330
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 0.71,
|
| 1498 |
+
"learning_rate": 9.117464263517712e-05,
|
| 1499 |
+
"loss": 0.5996,
|
| 1500 |
+
"step": 2340
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.71,
|
| 1504 |
+
"learning_rate": 9.024238657551272e-05,
|
| 1505 |
+
"loss": 0.5959,
|
| 1506 |
+
"step": 2350
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 0.71,
|
| 1510 |
+
"learning_rate": 8.931013051584834e-05,
|
| 1511 |
+
"loss": 0.6013,
|
| 1512 |
+
"step": 2360
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 0.71,
|
| 1516 |
+
"learning_rate": 8.837787445618396e-05,
|
| 1517 |
+
"loss": 0.5863,
|
| 1518 |
+
"step": 2370
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 0.72,
|
| 1522 |
+
"learning_rate": 8.744561839651957e-05,
|
| 1523 |
+
"loss": 0.6047,
|
| 1524 |
+
"step": 2380
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 0.72,
|
| 1528 |
+
"learning_rate": 8.651336233685518e-05,
|
| 1529 |
+
"loss": 0.601,
|
| 1530 |
+
"step": 2390
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 0.72,
|
| 1534 |
+
"learning_rate": 8.55811062771908e-05,
|
| 1535 |
+
"loss": 0.599,
|
| 1536 |
+
"step": 2400
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 0.72,
|
| 1540 |
+
"eval_loss": 0.5905566215515137,
|
| 1541 |
+
"eval_runtime": 38.9566,
|
| 1542 |
+
"eval_samples_per_second": 51.339,
|
| 1543 |
+
"eval_steps_per_second": 1.617,
|
| 1544 |
+
"step": 2400
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 0.73,
|
| 1548 |
+
"learning_rate": 8.46488502175264e-05,
|
| 1549 |
+
"loss": 0.5969,
|
| 1550 |
+
"step": 2410
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 0.73,
|
| 1554 |
+
"learning_rate": 8.371659415786201e-05,
|
| 1555 |
+
"loss": 0.5864,
|
| 1556 |
+
"step": 2420
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.73,
|
| 1560 |
+
"learning_rate": 8.278433809819763e-05,
|
| 1561 |
+
"loss": 0.5924,
|
| 1562 |
+
"step": 2430
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 0.74,
|
| 1566 |
+
"learning_rate": 8.185208203853325e-05,
|
| 1567 |
+
"loss": 0.6041,
|
| 1568 |
+
"step": 2440
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 0.74,
|
| 1572 |
+
"learning_rate": 8.091982597886885e-05,
|
| 1573 |
+
"loss": 0.5869,
|
| 1574 |
+
"step": 2450
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 0.74,
|
| 1578 |
+
"learning_rate": 7.998756991920447e-05,
|
| 1579 |
+
"loss": 0.588,
|
| 1580 |
+
"step": 2460
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 0.74,
|
| 1584 |
+
"learning_rate": 7.905531385954009e-05,
|
| 1585 |
+
"loss": 0.5971,
|
| 1586 |
+
"step": 2470
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 0.75,
|
| 1590 |
+
"learning_rate": 7.812305779987568e-05,
|
| 1591 |
+
"loss": 0.5946,
|
| 1592 |
+
"step": 2480
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 0.75,
|
| 1596 |
+
"learning_rate": 7.71908017402113e-05,
|
| 1597 |
+
"loss": 0.5881,
|
| 1598 |
+
"step": 2490
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.75,
|
| 1602 |
+
"learning_rate": 7.625854568054692e-05,
|
| 1603 |
+
"loss": 0.6019,
|
| 1604 |
+
"step": 2500
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 0.76,
|
| 1608 |
+
"learning_rate": 7.532628962088252e-05,
|
| 1609 |
+
"loss": 0.5923,
|
| 1610 |
+
"step": 2510
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 0.76,
|
| 1614 |
+
"learning_rate": 7.439403356121814e-05,
|
| 1615 |
+
"loss": 0.5933,
|
| 1616 |
+
"step": 2520
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 0.76,
|
| 1620 |
+
"learning_rate": 7.346177750155375e-05,
|
| 1621 |
+
"loss": 0.602,
|
| 1622 |
+
"step": 2530
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 0.77,
|
| 1626 |
+
"learning_rate": 7.252952144188937e-05,
|
| 1627 |
+
"loss": 0.6092,
|
| 1628 |
+
"step": 2540
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 0.77,
|
| 1632 |
+
"learning_rate": 7.159726538222498e-05,
|
| 1633 |
+
"loss": 0.5907,
|
| 1634 |
+
"step": 2550
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.77,
|
| 1638 |
+
"learning_rate": 7.066500932256059e-05,
|
| 1639 |
+
"loss": 0.5993,
|
| 1640 |
+
"step": 2560
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.77,
|
| 1644 |
+
"learning_rate": 6.97327532628962e-05,
|
| 1645 |
+
"loss": 0.5954,
|
| 1646 |
+
"step": 2570
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 0.78,
|
| 1650 |
+
"learning_rate": 6.880049720323181e-05,
|
| 1651 |
+
"loss": 0.6,
|
| 1652 |
+
"step": 2580
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 0.78,
|
| 1656 |
+
"learning_rate": 6.786824114356743e-05,
|
| 1657 |
+
"loss": 0.603,
|
| 1658 |
+
"step": 2590
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 0.78,
|
| 1662 |
+
"learning_rate": 6.693598508390303e-05,
|
| 1663 |
+
"loss": 0.5915,
|
| 1664 |
+
"step": 2600
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 0.78,
|
| 1668 |
+
"eval_loss": 0.5895615220069885,
|
| 1669 |
+
"eval_runtime": 38.9648,
|
| 1670 |
+
"eval_samples_per_second": 51.328,
|
| 1671 |
+
"eval_steps_per_second": 1.617,
|
| 1672 |
+
"step": 2600
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 0.79,
|
| 1676 |
+
"learning_rate": 6.600372902423865e-05,
|
| 1677 |
+
"loss": 0.5956,
|
| 1678 |
+
"step": 2610
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 0.79,
|
| 1682 |
+
"learning_rate": 6.507147296457427e-05,
|
| 1683 |
+
"loss": 0.5891,
|
| 1684 |
+
"step": 2620
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 0.79,
|
| 1688 |
+
"learning_rate": 6.413921690490988e-05,
|
| 1689 |
+
"loss": 0.5896,
|
| 1690 |
+
"step": 2630
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.8,
|
| 1694 |
+
"learning_rate": 6.320696084524548e-05,
|
| 1695 |
+
"loss": 0.5994,
|
| 1696 |
+
"step": 2640
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.8,
|
| 1700 |
+
"learning_rate": 6.22747047855811e-05,
|
| 1701 |
+
"loss": 0.5935,
|
| 1702 |
+
"step": 2650
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"epoch": 0.8,
|
| 1706 |
+
"learning_rate": 6.134244872591672e-05,
|
| 1707 |
+
"loss": 0.5937,
|
| 1708 |
+
"step": 2660
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 0.8,
|
| 1712 |
+
"learning_rate": 6.041019266625232e-05,
|
| 1713 |
+
"loss": 0.5925,
|
| 1714 |
+
"step": 2670
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 0.81,
|
| 1718 |
+
"learning_rate": 5.947793660658794e-05,
|
| 1719 |
+
"loss": 0.584,
|
| 1720 |
+
"step": 2680
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 0.81,
|
| 1724 |
+
"learning_rate": 5.8545680546923546e-05,
|
| 1725 |
+
"loss": 0.5897,
|
| 1726 |
+
"step": 2690
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 0.81,
|
| 1730 |
+
"learning_rate": 5.7613424487259164e-05,
|
| 1731 |
+
"loss": 0.5952,
|
| 1732 |
+
"step": 2700
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 0.82,
|
| 1736 |
+
"learning_rate": 5.6681168427594776e-05,
|
| 1737 |
+
"loss": 0.5935,
|
| 1738 |
+
"step": 2710
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.82,
|
| 1742 |
+
"learning_rate": 5.574891236793039e-05,
|
| 1743 |
+
"loss": 0.5965,
|
| 1744 |
+
"step": 2720
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 0.82,
|
| 1748 |
+
"learning_rate": 5.490988191423244e-05,
|
| 1749 |
+
"loss": 0.6008,
|
| 1750 |
+
"step": 2730
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 0.83,
|
| 1754 |
+
"learning_rate": 5.397762585456805e-05,
|
| 1755 |
+
"loss": 0.6001,
|
| 1756 |
+
"step": 2740
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 0.83,
|
| 1760 |
+
"learning_rate": 5.304536979490367e-05,
|
| 1761 |
+
"loss": 0.5923,
|
| 1762 |
+
"step": 2750
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 0.83,
|
| 1766 |
+
"learning_rate": 5.211311373523927e-05,
|
| 1767 |
+
"loss": 0.5929,
|
| 1768 |
+
"step": 2760
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 0.83,
|
| 1772 |
+
"learning_rate": 5.118085767557489e-05,
|
| 1773 |
+
"loss": 0.5931,
|
| 1774 |
+
"step": 2770
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 0.84,
|
| 1778 |
+
"learning_rate": 5.02486016159105e-05,
|
| 1779 |
+
"loss": 0.5961,
|
| 1780 |
+
"step": 2780
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.84,
|
| 1784 |
+
"learning_rate": 4.931634555624611e-05,
|
| 1785 |
+
"loss": 0.5871,
|
| 1786 |
+
"step": 2790
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 0.84,
|
| 1790 |
+
"learning_rate": 4.8384089496581725e-05,
|
| 1791 |
+
"loss": 0.5975,
|
| 1792 |
+
"step": 2800
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 0.84,
|
| 1796 |
+
"eval_loss": 0.5893495082855225,
|
| 1797 |
+
"eval_runtime": 38.9212,
|
| 1798 |
+
"eval_samples_per_second": 51.386,
|
| 1799 |
+
"eval_steps_per_second": 1.619,
|
| 1800 |
+
"step": 2800
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 0.85,
|
| 1804 |
+
"learning_rate": 4.745183343691734e-05,
|
| 1805 |
+
"loss": 0.6009,
|
| 1806 |
+
"step": 2810
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 0.85,
|
| 1810 |
+
"learning_rate": 4.651957737725295e-05,
|
| 1811 |
+
"loss": 0.5906,
|
| 1812 |
+
"step": 2820
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 0.85,
|
| 1816 |
+
"learning_rate": 4.558732131758856e-05,
|
| 1817 |
+
"loss": 0.5948,
|
| 1818 |
+
"step": 2830
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 0.86,
|
| 1822 |
+
"learning_rate": 4.465506525792417e-05,
|
| 1823 |
+
"loss": 0.6,
|
| 1824 |
+
"step": 2840
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 0.86,
|
| 1828 |
+
"learning_rate": 4.372280919825978e-05,
|
| 1829 |
+
"loss": 0.5898,
|
| 1830 |
+
"step": 2850
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 0.86,
|
| 1834 |
+
"learning_rate": 4.27905531385954e-05,
|
| 1835 |
+
"loss": 0.6005,
|
| 1836 |
+
"step": 2860
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.86,
|
| 1840 |
+
"learning_rate": 4.1858297078931006e-05,
|
| 1841 |
+
"loss": 0.5949,
|
| 1842 |
+
"step": 2870
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 0.87,
|
| 1846 |
+
"learning_rate": 4.0926041019266625e-05,
|
| 1847 |
+
"loss": 0.6213,
|
| 1848 |
+
"step": 2880
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 0.87,
|
| 1852 |
+
"learning_rate": 3.9993784959602236e-05,
|
| 1853 |
+
"loss": 0.5966,
|
| 1854 |
+
"step": 2890
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 0.87,
|
| 1858 |
+
"learning_rate": 3.906152889993784e-05,
|
| 1859 |
+
"loss": 0.5955,
|
| 1860 |
+
"step": 2900
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 0.88,
|
| 1864 |
+
"learning_rate": 3.812927284027346e-05,
|
| 1865 |
+
"loss": 0.6001,
|
| 1866 |
+
"step": 2910
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 0.88,
|
| 1870 |
+
"learning_rate": 3.719701678060907e-05,
|
| 1871 |
+
"loss": 0.5951,
|
| 1872 |
+
"step": 2920
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 0.88,
|
| 1876 |
+
"learning_rate": 3.626476072094468e-05,
|
| 1877 |
+
"loss": 0.5946,
|
| 1878 |
+
"step": 2930
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.89,
|
| 1882 |
+
"learning_rate": 3.5332504661280294e-05,
|
| 1883 |
+
"loss": 0.6006,
|
| 1884 |
+
"step": 2940
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 0.89,
|
| 1888 |
+
"learning_rate": 3.4400248601615906e-05,
|
| 1889 |
+
"loss": 0.593,
|
| 1890 |
+
"step": 2950
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 0.89,
|
| 1894 |
+
"learning_rate": 3.346799254195152e-05,
|
| 1895 |
+
"loss": 0.5961,
|
| 1896 |
+
"step": 2960
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 0.9,
|
| 1900 |
+
"learning_rate": 3.2535736482287136e-05,
|
| 1901 |
+
"loss": 0.579,
|
| 1902 |
+
"step": 2970
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"epoch": 0.9,
|
| 1906 |
+
"learning_rate": 3.160348042262274e-05,
|
| 1907 |
+
"loss": 0.6025,
|
| 1908 |
+
"step": 2980
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 0.9,
|
| 1912 |
+
"learning_rate": 3.067122436295836e-05,
|
| 1913 |
+
"loss": 0.601,
|
| 1914 |
+
"step": 2990
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 0.9,
|
| 1918 |
+
"learning_rate": 2.973896830329397e-05,
|
| 1919 |
+
"loss": 0.5875,
|
| 1920 |
+
"step": 3000
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.9,
|
| 1924 |
+
"eval_loss": 0.5886192917823792,
|
| 1925 |
+
"eval_runtime": 38.9814,
|
| 1926 |
+
"eval_samples_per_second": 51.306,
|
| 1927 |
+
"eval_steps_per_second": 1.616,
|
| 1928 |
+
"step": 3000
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 0.91,
|
| 1932 |
+
"learning_rate": 2.8806712243629582e-05,
|
| 1933 |
+
"loss": 0.5942,
|
| 1934 |
+
"step": 3010
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.91,
|
| 1938 |
+
"learning_rate": 2.7874456183965194e-05,
|
| 1939 |
+
"loss": 0.5854,
|
| 1940 |
+
"step": 3020
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 0.91,
|
| 1944 |
+
"learning_rate": 2.6942200124300805e-05,
|
| 1945 |
+
"loss": 0.6094,
|
| 1946 |
+
"step": 3030
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 0.92,
|
| 1950 |
+
"learning_rate": 2.600994406463642e-05,
|
| 1951 |
+
"loss": 0.5852,
|
| 1952 |
+
"step": 3040
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 0.92,
|
| 1956 |
+
"learning_rate": 2.507768800497203e-05,
|
| 1957 |
+
"loss": 0.5987,
|
| 1958 |
+
"step": 3050
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 0.92,
|
| 1962 |
+
"learning_rate": 2.414543194530764e-05,
|
| 1963 |
+
"loss": 0.5986,
|
| 1964 |
+
"step": 3060
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 0.93,
|
| 1968 |
+
"learning_rate": 2.3213175885643255e-05,
|
| 1969 |
+
"loss": 0.5948,
|
| 1970 |
+
"step": 3070
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 0.93,
|
| 1974 |
+
"learning_rate": 2.228091982597887e-05,
|
| 1975 |
+
"loss": 0.5955,
|
| 1976 |
+
"step": 3080
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.93,
|
| 1980 |
+
"learning_rate": 2.1348663766314478e-05,
|
| 1981 |
+
"loss": 0.5843,
|
| 1982 |
+
"step": 3090
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 0.93,
|
| 1986 |
+
"learning_rate": 2.041640770665009e-05,
|
| 1987 |
+
"loss": 0.5928,
|
| 1988 |
+
"step": 3100
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 0.94,
|
| 1992 |
+
"learning_rate": 1.9484151646985705e-05,
|
| 1993 |
+
"loss": 0.6071,
|
| 1994 |
+
"step": 3110
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 0.94,
|
| 1998 |
+
"learning_rate": 1.8551895587321316e-05,
|
| 1999 |
+
"loss": 0.5952,
|
| 2000 |
+
"step": 3120
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 0.94,
|
| 2004 |
+
"learning_rate": 1.7619639527656928e-05,
|
| 2005 |
+
"loss": 0.5893,
|
| 2006 |
+
"step": 3130
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 0.95,
|
| 2010 |
+
"learning_rate": 1.668738346799254e-05,
|
| 2011 |
+
"loss": 0.5913,
|
| 2012 |
+
"step": 3140
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 0.95,
|
| 2016 |
+
"learning_rate": 1.575512740832815e-05,
|
| 2017 |
+
"loss": 0.5957,
|
| 2018 |
+
"step": 3150
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.95,
|
| 2022 |
+
"learning_rate": 1.4822871348663764e-05,
|
| 2023 |
+
"loss": 0.5921,
|
| 2024 |
+
"step": 3160
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 0.96,
|
| 2028 |
+
"learning_rate": 1.3890615288999376e-05,
|
| 2029 |
+
"loss": 0.5943,
|
| 2030 |
+
"step": 3170
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 0.96,
|
| 2034 |
+
"learning_rate": 1.295835922933499e-05,
|
| 2035 |
+
"loss": 0.598,
|
| 2036 |
+
"step": 3180
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 0.96,
|
| 2040 |
+
"learning_rate": 1.2026103169670601e-05,
|
| 2041 |
+
"loss": 0.5911,
|
| 2042 |
+
"step": 3190
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 0.96,
|
| 2046 |
+
"learning_rate": 1.1093847110006214e-05,
|
| 2047 |
+
"loss": 0.5963,
|
| 2048 |
+
"step": 3200
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 0.96,
|
| 2052 |
+
"eval_loss": 0.5883459448814392,
|
| 2053 |
+
"eval_runtime": 38.935,
|
| 2054 |
+
"eval_samples_per_second": 51.368,
|
| 2055 |
+
"eval_steps_per_second": 1.618,
|
| 2056 |
+
"step": 3200
|
| 2057 |
+
}
|
| 2058 |
+
],
|
| 2059 |
+
"max_steps": 3318,
|
| 2060 |
+
"num_train_epochs": 1,
|
| 2061 |
+
"total_flos": 1.0078203805888938e+19,
|
| 2062 |
+
"trial_name": null,
|
| 2063 |
+
"trial_params": null
|
| 2064 |
+
}
|
checkpoint-3200/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4830016fae72af86a9ebb5e0bef10ca94e44ed3f248ccef5abacaf783b5c8b08
|
| 3 |
+
size 3899
|