Commit
•
f82b15e
1
Parent(s):
98fedb0
huggingartists
Browse files- README.md +3 -3
- config.json +1 -1
- evaluation.txt +1 -1
- flax_model.msgpack +1 -1
- optimizer.pt +1 -1
- pytorch_model.bin +1 -1
- rng_state.pth +2 -2
- scheduler.pt +1 -1
- tokenizer_config.json +1 -1
- trainer_state.json +1211 -7
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -45,15 +45,15 @@ from datasets import load_dataset
|
|
| 45 |
dataset = load_dataset("huggingartists/slava-kpss")
|
| 46 |
```
|
| 47 |
|
| 48 |
-
[Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/
|
| 49 |
|
| 50 |
## Training procedure
|
| 51 |
|
| 52 |
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on Слава КПСС (Slava KPSS)'s lyrics.
|
| 53 |
|
| 54 |
-
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/
|
| 55 |
|
| 56 |
-
At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/
|
| 57 |
|
| 58 |
## How to use
|
| 59 |
|
|
|
|
| 45 |
dataset = load_dataset("huggingartists/slava-kpss")
|
| 46 |
```
|
| 47 |
|
| 48 |
+
[Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/2f2r3u3b/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
|
| 49 |
|
| 50 |
## Training procedure
|
| 51 |
|
| 52 |
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on Слава КПСС (Slava KPSS)'s lyrics.
|
| 53 |
|
| 54 |
+
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/pecxkpae) for full transparency and reproducibility.
|
| 55 |
|
| 56 |
+
At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/pecxkpae/artifacts) is logged and versioned.
|
| 57 |
|
| 58 |
## How to use
|
| 59 |
|
config.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "
|
| 3 |
"activation_function": "gelu_new",
|
| 4 |
"architectures": [
|
| 5 |
"GPT2LMHeadModel"
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "huggingartists/slava-kpss",
|
| 3 |
"activation_function": "gelu_new",
|
| 4 |
"architectures": [
|
| 5 |
"GPT2LMHeadModel"
|
evaluation.txt
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"eval_loss": 1.
|
|
|
|
| 1 |
+
{"eval_loss": 1.6512105464935303, "eval_runtime": 68.8621, "eval_samples_per_second": 20.476, "eval_steps_per_second": 2.57, "epoch": 2.0}
|
flax_model.msgpack
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 497764120
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99ac3f00b077f08b809d4c3048c2a2dd6f5e179f6d39456c91ba6fd59ca92749
|
| 3 |
size 497764120
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 995604017
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1e1f79ca37b89ce49476f1f6d2bdf308b33a81036d2a608862d052857fc05544
|
| 3 |
size 995604017
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 510403817
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:70d3a108d5b57c18b4dc086ddceadcd8fd6cb7447767943b3a401af1e7d26ec9
|
| 3 |
size 510403817
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cd584a84981d622a4a49aae5c6af7ee90b3c2b286a8932167387ee7992b325fe
|
| 3 |
+
size 14567
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 623
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ec253ddc8da1898d70eaa47720ce4ca3ff5eb5f957f1a07fb578eca051cb0b86
|
| 3 |
size 623
|
tokenizer_config.json
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"unk_token": "<|endoftext|>", "bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "add_prefix_space": false, "model_max_length": 1024, "special_tokens_map_file": null, "name_or_path": "
|
|
|
|
| 1 |
+
{"unk_token": "<|endoftext|>", "bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "add_prefix_space": false, "model_max_length": 1024, "special_tokens_map_file": null, "name_or_path": "huggingartists/slava-kpss", "tokenizer_class": "GPT2Tokenizer"}
|
trainer_state.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 1.
|
| 3 |
-
"best_model_checkpoint": "output/slava-kpss/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -1130,11 +1130,1215 @@
|
|
| 1130 |
"eval_samples_per_second": 20.494,
|
| 1131 |
"eval_steps_per_second": 2.562,
|
| 1132 |
"step": 933
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1133 |
}
|
| 1134 |
],
|
| 1135 |
-
"max_steps":
|
| 1136 |
-
"num_train_epochs":
|
| 1137 |
-
"total_flos":
|
| 1138 |
"trial_name": null,
|
| 1139 |
"trial_params": null
|
| 1140 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.6512105464935303,
|
| 3 |
+
"best_model_checkpoint": "output/slava-kpss/checkpoint-1922",
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"global_step": 1922,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 1130 |
"eval_samples_per_second": 20.494,
|
| 1131 |
"eval_steps_per_second": 2.562,
|
| 1132 |
"step": 933
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 0.97,
|
| 1136 |
+
"learning_rate": 2.476465669200449e-07,
|
| 1137 |
+
"loss": 1.7548,
|
| 1138 |
+
"step": 935
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 0.98,
|
| 1142 |
+
"learning_rate": 1.615902325712679e-07,
|
| 1143 |
+
"loss": 1.7767,
|
| 1144 |
+
"step": 940
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.98,
|
| 1148 |
+
"learning_rate": 9.38184037085626e-08,
|
| 1149 |
+
"loss": 1.7322,
|
| 1150 |
+
"step": 945
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.99,
|
| 1154 |
+
"learning_rate": 4.434918674879014e-08,
|
| 1155 |
+
"loss": 1.7019,
|
| 1156 |
+
"step": 950
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 0.99,
|
| 1160 |
+
"learning_rate": 1.319579824933328e-08,
|
| 1161 |
+
"loss": 1.6765,
|
| 1162 |
+
"step": 955
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 1.0,
|
| 1166 |
+
"learning_rate": 3.665613770225118e-10,
|
| 1167 |
+
"loss": 1.742,
|
| 1168 |
+
"step": 960
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 1.0,
|
| 1172 |
+
"eval_loss": 1.6864277124404907,
|
| 1173 |
+
"eval_runtime": 68.8946,
|
| 1174 |
+
"eval_samples_per_second": 20.466,
|
| 1175 |
+
"eval_steps_per_second": 2.569,
|
| 1176 |
+
"step": 961
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 1.0,
|
| 1180 |
+
"learning_rate": 5.8649036845078714e-09,
|
| 1181 |
+
"loss": 1.574,
|
| 1182 |
+
"step": 965
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 1.01,
|
| 1186 |
+
"learning_rate": 2.9689356194480253e-08,
|
| 1187 |
+
"loss": 1.7872,
|
| 1188 |
+
"step": 970
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 1.01,
|
| 1192 |
+
"learning_rate": 7.183355379217629e-08,
|
| 1193 |
+
"loss": 1.688,
|
| 1194 |
+
"step": 975
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 1.02,
|
| 1198 |
+
"learning_rate": 1.3228623692592306e-07,
|
| 1199 |
+
"loss": 1.7384,
|
| 1200 |
+
"step": 980
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.02,
|
| 1204 |
+
"learning_rate": 2.1103125461527332e-07,
|
| 1205 |
+
"loss": 1.672,
|
| 1206 |
+
"step": 985
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.03,
|
| 1210 |
+
"learning_rate": 3.0804756876609914e-07,
|
| 1211 |
+
"loss": 1.7357,
|
| 1212 |
+
"step": 990
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 1.04,
|
| 1216 |
+
"learning_rate": 4.233092597912044e-07,
|
| 1217 |
+
"loss": 1.7344,
|
| 1218 |
+
"step": 995
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 1.04,
|
| 1222 |
+
"learning_rate": 5.567855335352604e-07,
|
| 1223 |
+
"loss": 1.7574,
|
| 1224 |
+
"step": 1000
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 1.05,
|
| 1228 |
+
"learning_rate": 7.084407295019475e-07,
|
| 1229 |
+
"loss": 1.7031,
|
| 1230 |
+
"step": 1005
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 1.05,
|
| 1234 |
+
"learning_rate": 8.782343303812844e-07,
|
| 1235 |
+
"loss": 1.6877,
|
| 1236 |
+
"step": 1010
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 1.06,
|
| 1240 |
+
"learning_rate": 1.0661209728745555e-06,
|
| 1241 |
+
"loss": 1.6856,
|
| 1242 |
+
"step": 1015
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.06,
|
| 1246 |
+
"learning_rate": 1.272050459813843e-06,
|
| 1247 |
+
"loss": 1.7801,
|
| 1248 |
+
"step": 1020
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.07,
|
| 1252 |
+
"learning_rate": 1.495967773573164e-06,
|
| 1253 |
+
"loss": 1.7647,
|
| 1254 |
+
"step": 1025
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 1.07,
|
| 1258 |
+
"learning_rate": 1.7378130907672579e-06,
|
| 1259 |
+
"loss": 1.7437,
|
| 1260 |
+
"step": 1030
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 1.08,
|
| 1264 |
+
"learning_rate": 1.997521798234569e-06,
|
| 1265 |
+
"loss": 1.7193,
|
| 1266 |
+
"step": 1035
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 1.08,
|
| 1270 |
+
"learning_rate": 2.275024510299646e-06,
|
| 1271 |
+
"loss": 1.8438,
|
| 1272 |
+
"step": 1040
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 1.09,
|
| 1276 |
+
"learning_rate": 2.570247087310856e-06,
|
| 1277 |
+
"loss": 1.7861,
|
| 1278 |
+
"step": 1045
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 1.09,
|
| 1282 |
+
"learning_rate": 2.883110655447913e-06,
|
| 1283 |
+
"loss": 1.7835,
|
| 1284 |
+
"step": 1050
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.1,
|
| 1288 |
+
"learning_rate": 3.213531627794597e-06,
|
| 1289 |
+
"loss": 1.6692,
|
| 1290 |
+
"step": 1055
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.1,
|
| 1294 |
+
"learning_rate": 3.561421726670198e-06,
|
| 1295 |
+
"loss": 1.673,
|
| 1296 |
+
"step": 1060
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 1.11,
|
| 1300 |
+
"learning_rate": 3.926688007214648e-06,
|
| 1301 |
+
"loss": 1.7161,
|
| 1302 |
+
"step": 1065
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 1.11,
|
| 1306 |
+
"learning_rate": 4.3092328822202e-06,
|
| 1307 |
+
"loss": 1.725,
|
| 1308 |
+
"step": 1070
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 1.12,
|
| 1312 |
+
"learning_rate": 4.708954148203643e-06,
|
| 1313 |
+
"loss": 1.7126,
|
| 1314 |
+
"step": 1075
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 1.12,
|
| 1318 |
+
"learning_rate": 5.1257450127116035e-06,
|
| 1319 |
+
"loss": 1.7486,
|
| 1320 |
+
"step": 1080
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 1.13,
|
| 1324 |
+
"learning_rate": 5.559494122852188e-06,
|
| 1325 |
+
"loss": 1.7327,
|
| 1326 |
+
"step": 1085
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.13,
|
| 1330 |
+
"learning_rate": 6.0100855950445935e-06,
|
| 1331 |
+
"loss": 1.7928,
|
| 1332 |
+
"step": 1090
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.14,
|
| 1336 |
+
"learning_rate": 6.477399045979628e-06,
|
| 1337 |
+
"loss": 1.7449,
|
| 1338 |
+
"step": 1095
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 1.14,
|
| 1342 |
+
"learning_rate": 6.961309624782102e-06,
|
| 1343 |
+
"loss": 1.7278,
|
| 1344 |
+
"step": 1100
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 1.15,
|
| 1348 |
+
"learning_rate": 7.461688046366857e-06,
|
| 1349 |
+
"loss": 1.7724,
|
| 1350 |
+
"step": 1105
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 1.16,
|
| 1354 |
+
"learning_rate": 7.97840062597962e-06,
|
| 1355 |
+
"loss": 1.7395,
|
| 1356 |
+
"step": 1110
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 1.16,
|
| 1360 |
+
"learning_rate": 8.511309314913076e-06,
|
| 1361 |
+
"loss": 1.7612,
|
| 1362 |
+
"step": 1115
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 1.17,
|
| 1366 |
+
"learning_rate": 9.060271737389124e-06,
|
| 1367 |
+
"loss": 1.7803,
|
| 1368 |
+
"step": 1120
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.17,
|
| 1372 |
+
"learning_rate": 9.625141228596819e-06,
|
| 1373 |
+
"loss": 1.698,
|
| 1374 |
+
"step": 1125
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.18,
|
| 1378 |
+
"learning_rate": 1.0205766873876643e-05,
|
| 1379 |
+
"loss": 1.7627,
|
| 1380 |
+
"step": 1130
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 1.18,
|
| 1384 |
+
"learning_rate": 1.08019935490398e-05,
|
| 1385 |
+
"loss": 1.615,
|
| 1386 |
+
"step": 1135
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 1.19,
|
| 1390 |
+
"learning_rate": 1.1413661961812419e-05,
|
| 1391 |
+
"loss": 1.7236,
|
| 1392 |
+
"step": 1140
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 1.19,
|
| 1396 |
+
"learning_rate": 1.2040608694393166e-05,
|
| 1397 |
+
"loss": 1.732,
|
| 1398 |
+
"step": 1145
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 1.2,
|
| 1402 |
+
"learning_rate": 1.2682666247113272e-05,
|
| 1403 |
+
"loss": 1.7578,
|
| 1404 |
+
"step": 1150
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 1.2,
|
| 1408 |
+
"learning_rate": 1.333966308318674e-05,
|
| 1409 |
+
"loss": 1.7636,
|
| 1410 |
+
"step": 1155
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.21,
|
| 1414 |
+
"learning_rate": 1.4011423674539631e-05,
|
| 1415 |
+
"loss": 1.6698,
|
| 1416 |
+
"step": 1160
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.21,
|
| 1420 |
+
"learning_rate": 1.4697768548705208e-05,
|
| 1421 |
+
"loss": 1.8001,
|
| 1422 |
+
"step": 1165
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 1.22,
|
| 1426 |
+
"learning_rate": 1.5398514336773276e-05,
|
| 1427 |
+
"loss": 1.7019,
|
| 1428 |
+
"step": 1170
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 1.22,
|
| 1432 |
+
"learning_rate": 1.611347382238021e-05,
|
| 1433 |
+
"loss": 1.7685,
|
| 1434 |
+
"step": 1175
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 1.23,
|
| 1438 |
+
"learning_rate": 1.684245599172719e-05,
|
| 1439 |
+
"loss": 1.8088,
|
| 1440 |
+
"step": 1180
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 1.23,
|
| 1444 |
+
"learning_rate": 1.7585266084612767e-05,
|
| 1445 |
+
"loss": 1.7428,
|
| 1446 |
+
"step": 1185
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 1.24,
|
| 1450 |
+
"learning_rate": 1.8341705646466328e-05,
|
| 1451 |
+
"loss": 1.712,
|
| 1452 |
+
"step": 1190
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.24,
|
| 1456 |
+
"learning_rate": 1.9111572581368946e-05,
|
| 1457 |
+
"loss": 1.7453,
|
| 1458 |
+
"step": 1195
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 1.25,
|
| 1462 |
+
"learning_rate": 1.9894661206046444e-05,
|
| 1463 |
+
"loss": 1.6769,
|
| 1464 |
+
"step": 1200
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 1.25,
|
| 1468 |
+
"learning_rate": 2.069076230482155e-05,
|
| 1469 |
+
"loss": 1.6822,
|
| 1470 |
+
"step": 1205
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 1.26,
|
| 1474 |
+
"learning_rate": 2.1499663185509284e-05,
|
| 1475 |
+
"loss": 1.7965,
|
| 1476 |
+
"step": 1210
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 1.26,
|
| 1480 |
+
"learning_rate": 2.232114773624152e-05,
|
| 1481 |
+
"loss": 1.6978,
|
| 1482 |
+
"step": 1215
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 1.27,
|
| 1486 |
+
"learning_rate": 2.3154996483204744e-05,
|
| 1487 |
+
"loss": 1.7852,
|
| 1488 |
+
"step": 1220
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 1.27,
|
| 1492 |
+
"learning_rate": 2.400098664927671e-05,
|
| 1493 |
+
"loss": 1.6758,
|
| 1494 |
+
"step": 1225
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 1.28,
|
| 1498 |
+
"learning_rate": 2.4858892213544702e-05,
|
| 1499 |
+
"loss": 1.742,
|
| 1500 |
+
"step": 1230
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.29,
|
| 1504 |
+
"learning_rate": 2.5728483971691396e-05,
|
| 1505 |
+
"loss": 1.7897,
|
| 1506 |
+
"step": 1235
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 1.29,
|
| 1510 |
+
"learning_rate": 2.660952959723034e-05,
|
| 1511 |
+
"loss": 1.6749,
|
| 1512 |
+
"step": 1240
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 1.3,
|
| 1516 |
+
"learning_rate": 2.750179370357635e-05,
|
| 1517 |
+
"loss": 1.7127,
|
| 1518 |
+
"step": 1245
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 1.3,
|
| 1522 |
+
"learning_rate": 2.8405037906932926e-05,
|
| 1523 |
+
"loss": 1.7259,
|
| 1524 |
+
"step": 1250
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 1.31,
|
| 1528 |
+
"learning_rate": 2.9319020889980675e-05,
|
| 1529 |
+
"loss": 1.6448,
|
| 1530 |
+
"step": 1255
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 1.31,
|
| 1534 |
+
"learning_rate": 3.0243498466349233e-05,
|
| 1535 |
+
"loss": 1.6744,
|
| 1536 |
+
"step": 1260
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.32,
|
| 1540 |
+
"learning_rate": 3.117822364585623e-05,
|
| 1541 |
+
"loss": 1.6893,
|
| 1542 |
+
"step": 1265
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.32,
|
| 1546 |
+
"learning_rate": 3.212294670049457e-05,
|
| 1547 |
+
"loss": 1.7822,
|
| 1548 |
+
"step": 1270
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 1.33,
|
| 1552 |
+
"learning_rate": 3.3077415231151914e-05,
|
| 1553 |
+
"loss": 1.7231,
|
| 1554 |
+
"step": 1275
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 1.33,
|
| 1558 |
+
"learning_rate": 3.4041374235043516e-05,
|
| 1559 |
+
"loss": 1.7438,
|
| 1560 |
+
"step": 1280
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 1.34,
|
| 1564 |
+
"learning_rate": 3.501456617384015e-05,
|
| 1565 |
+
"loss": 1.776,
|
| 1566 |
+
"step": 1285
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 1.34,
|
| 1570 |
+
"learning_rate": 3.599673104247454e-05,
|
| 1571 |
+
"loss": 1.7459,
|
| 1572 |
+
"step": 1290
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 1.35,
|
| 1576 |
+
"learning_rate": 3.6987606438605696e-05,
|
| 1577 |
+
"loss": 1.7127,
|
| 1578 |
+
"step": 1295
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.35,
|
| 1582 |
+
"learning_rate": 3.798692763272465e-05,
|
| 1583 |
+
"loss": 1.6762,
|
| 1584 |
+
"step": 1300
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.36,
|
| 1588 |
+
"learning_rate": 3.899442763888126e-05,
|
| 1589 |
+
"loss": 1.7619,
|
| 1590 |
+
"step": 1305
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 1.36,
|
| 1594 |
+
"learning_rate": 4.000983728601432e-05,
|
| 1595 |
+
"loss": 1.7135,
|
| 1596 |
+
"step": 1310
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 1.37,
|
| 1600 |
+
"learning_rate": 4.103288528986497e-05,
|
| 1601 |
+
"loss": 1.7328,
|
| 1602 |
+
"step": 1315
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 1.37,
|
| 1606 |
+
"learning_rate": 4.20632983254554e-05,
|
| 1607 |
+
"loss": 1.7118,
|
| 1608 |
+
"step": 1320
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 1.38,
|
| 1612 |
+
"learning_rate": 4.3100801100111876e-05,
|
| 1613 |
+
"loss": 1.801,
|
| 1614 |
+
"step": 1325
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 1.38,
|
| 1618 |
+
"learning_rate": 4.4145116427014376e-05,
|
| 1619 |
+
"loss": 1.7959,
|
| 1620 |
+
"step": 1330
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 1.39,
|
| 1624 |
+
"learning_rate": 4.519596529925159e-05,
|
| 1625 |
+
"loss": 1.7436,
|
| 1626 |
+
"step": 1335
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 1.39,
|
| 1630 |
+
"learning_rate": 4.625306696436258e-05,
|
| 1631 |
+
"loss": 1.7913,
|
| 1632 |
+
"step": 1340
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 1.4,
|
| 1636 |
+
"learning_rate": 4.7316138999344416e-05,
|
| 1637 |
+
"loss": 1.7002,
|
| 1638 |
+
"step": 1345
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 1.4,
|
| 1642 |
+
"learning_rate": 4.838489738610674e-05,
|
| 1643 |
+
"loss": 1.6758,
|
| 1644 |
+
"step": 1350
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 1.41,
|
| 1648 |
+
"learning_rate": 4.945905658735171e-05,
|
| 1649 |
+
"loss": 1.7027,
|
| 1650 |
+
"step": 1355
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 1.42,
|
| 1654 |
+
"learning_rate": 5.0538329622860605e-05,
|
| 1655 |
+
"loss": 1.7378,
|
| 1656 |
+
"step": 1360
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 1.42,
|
| 1660 |
+
"learning_rate": 5.1622428146165326e-05,
|
| 1661 |
+
"loss": 1.6332,
|
| 1662 |
+
"step": 1365
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 1.43,
|
| 1666 |
+
"learning_rate": 5.271106252158521e-05,
|
| 1667 |
+
"loss": 1.7335,
|
| 1668 |
+
"step": 1370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 1.43,
|
| 1672 |
+
"learning_rate": 5.3803941901608256e-05,
|
| 1673 |
+
"loss": 1.7126,
|
| 1674 |
+
"step": 1375
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 1.44,
|
| 1678 |
+
"learning_rate": 5.4900774304595864e-05,
|
| 1679 |
+
"loss": 1.7287,
|
| 1680 |
+
"step": 1380
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 1.44,
|
| 1684 |
+
"learning_rate": 5.600126669279115e-05,
|
| 1685 |
+
"loss": 1.6935,
|
| 1686 |
+
"step": 1385
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 1.45,
|
| 1690 |
+
"learning_rate": 5.710512505060865e-05,
|
| 1691 |
+
"loss": 1.7214,
|
| 1692 |
+
"step": 1390
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 1.45,
|
| 1696 |
+
"learning_rate": 5.8212054463186014e-05,
|
| 1697 |
+
"loss": 1.7355,
|
| 1698 |
+
"step": 1395
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 1.46,
|
| 1702 |
+
"learning_rate": 5.9321759195175425e-05,
|
| 1703 |
+
"loss": 1.7537,
|
| 1704 |
+
"step": 1400
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 1.46,
|
| 1708 |
+
"learning_rate": 6.043394276975451e-05,
|
| 1709 |
+
"loss": 1.776,
|
| 1710 |
+
"step": 1405
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 1.47,
|
| 1714 |
+
"learning_rate": 6.154830804783502e-05,
|
| 1715 |
+
"loss": 1.783,
|
| 1716 |
+
"step": 1410
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 1.47,
|
| 1720 |
+
"learning_rate": 6.266455730744911e-05,
|
| 1721 |
+
"loss": 1.7703,
|
| 1722 |
+
"step": 1415
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 1.48,
|
| 1726 |
+
"learning_rate": 6.378239232329044e-05,
|
| 1727 |
+
"loss": 1.7259,
|
| 1728 |
+
"step": 1420
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 1.48,
|
| 1732 |
+
"learning_rate": 6.490151444639059e-05,
|
| 1733 |
+
"loss": 1.6852,
|
| 1734 |
+
"step": 1425
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 1.49,
|
| 1738 |
+
"learning_rate": 6.602162468390815e-05,
|
| 1739 |
+
"loss": 1.675,
|
| 1740 |
+
"step": 1430
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 1.49,
|
| 1744 |
+
"learning_rate": 6.714242377900986e-05,
|
| 1745 |
+
"loss": 1.7367,
|
| 1746 |
+
"step": 1435
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 1.5,
|
| 1750 |
+
"learning_rate": 6.826361229082211e-05,
|
| 1751 |
+
"loss": 1.8171,
|
| 1752 |
+
"step": 1440
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 1.5,
|
| 1756 |
+
"learning_rate": 6.938489067443205e-05,
|
| 1757 |
+
"loss": 1.7388,
|
| 1758 |
+
"step": 1445
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 1.51,
|
| 1762 |
+
"learning_rate": 7.050595936091584e-05,
|
| 1763 |
+
"loss": 1.7459,
|
| 1764 |
+
"step": 1450
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 1.51,
|
| 1768 |
+
"learning_rate": 7.162651883737402e-05,
|
| 1769 |
+
"loss": 1.808,
|
| 1770 |
+
"step": 1455
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 1.52,
|
| 1774 |
+
"learning_rate": 7.274626972695132e-05,
|
| 1775 |
+
"loss": 1.7933,
|
| 1776 |
+
"step": 1460
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 1.52,
|
| 1780 |
+
"learning_rate": 7.386491286882046e-05,
|
| 1781 |
+
"loss": 1.7711,
|
| 1782 |
+
"step": 1465
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 1.53,
|
| 1786 |
+
"learning_rate": 7.498214939810823e-05,
|
| 1787 |
+
"loss": 1.7745,
|
| 1788 |
+
"step": 1470
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 1.53,
|
| 1792 |
+
"learning_rate": 7.609768082574206e-05,
|
| 1793 |
+
"loss": 1.6494,
|
| 1794 |
+
"step": 1475
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 1.54,
|
| 1798 |
+
"learning_rate": 7.721120911819707e-05,
|
| 1799 |
+
"loss": 1.7815,
|
| 1800 |
+
"step": 1480
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 1.55,
|
| 1804 |
+
"learning_rate": 7.83224367771204e-05,
|
| 1805 |
+
"loss": 1.7829,
|
| 1806 |
+
"step": 1485
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 1.55,
|
| 1810 |
+
"learning_rate": 7.943106691881334e-05,
|
| 1811 |
+
"loss": 1.7363,
|
| 1812 |
+
"step": 1490
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 1.56,
|
| 1816 |
+
"learning_rate": 8.053680335354889e-05,
|
| 1817 |
+
"loss": 1.7149,
|
| 1818 |
+
"step": 1495
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 1.56,
|
| 1822 |
+
"learning_rate": 8.163935066470398e-05,
|
| 1823 |
+
"loss": 1.7134,
|
| 1824 |
+
"step": 1500
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 1.57,
|
| 1828 |
+
"learning_rate": 8.273841428768484e-05,
|
| 1829 |
+
"loss": 1.7653,
|
| 1830 |
+
"step": 1505
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 1.57,
|
| 1834 |
+
"learning_rate": 8.383370058862563e-05,
|
| 1835 |
+
"loss": 1.7677,
|
| 1836 |
+
"step": 1510
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 1.58,
|
| 1840 |
+
"learning_rate": 8.492491694283713e-05,
|
| 1841 |
+
"loss": 1.6578,
|
| 1842 |
+
"step": 1515
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 1.58,
|
| 1846 |
+
"learning_rate": 8.601177181298704e-05,
|
| 1847 |
+
"loss": 1.7271,
|
| 1848 |
+
"step": 1520
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 1.59,
|
| 1852 |
+
"learning_rate": 8.709397482698893e-05,
|
| 1853 |
+
"loss": 1.6661,
|
| 1854 |
+
"step": 1525
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 1.59,
|
| 1858 |
+
"learning_rate": 8.817123685558034e-05,
|
| 1859 |
+
"loss": 1.7347,
|
| 1860 |
+
"step": 1530
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 1.6,
|
| 1864 |
+
"learning_rate": 8.92432700895683e-05,
|
| 1865 |
+
"loss": 1.7298,
|
| 1866 |
+
"step": 1535
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 1.6,
|
| 1870 |
+
"learning_rate": 9.03097881167231e-05,
|
| 1871 |
+
"loss": 1.7343,
|
| 1872 |
+
"step": 1540
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 1.61,
|
| 1876 |
+
"learning_rate": 9.137050599829776e-05,
|
| 1877 |
+
"loss": 1.7651,
|
| 1878 |
+
"step": 1545
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 1.61,
|
| 1882 |
+
"learning_rate": 9.242514034515462e-05,
|
| 1883 |
+
"loss": 1.7716,
|
| 1884 |
+
"step": 1550
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 1.62,
|
| 1888 |
+
"learning_rate": 9.34734093934778e-05,
|
| 1889 |
+
"loss": 1.8064,
|
| 1890 |
+
"step": 1555
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 1.62,
|
| 1894 |
+
"learning_rate": 9.451503308005074e-05,
|
| 1895 |
+
"loss": 1.8239,
|
| 1896 |
+
"step": 1560
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 1.63,
|
| 1900 |
+
"learning_rate": 9.554973311708053e-05,
|
| 1901 |
+
"loss": 1.7248,
|
| 1902 |
+
"step": 1565
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"epoch": 1.63,
|
| 1906 |
+
"learning_rate": 9.657723306654681e-05,
|
| 1907 |
+
"loss": 1.6457,
|
| 1908 |
+
"step": 1570
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 1.64,
|
| 1912 |
+
"learning_rate": 9.759725841405743e-05,
|
| 1913 |
+
"loss": 1.8182,
|
| 1914 |
+
"step": 1575
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 1.64,
|
| 1918 |
+
"learning_rate": 9.86095366421894e-05,
|
| 1919 |
+
"loss": 1.6211,
|
| 1920 |
+
"step": 1580
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 1.65,
|
| 1924 |
+
"learning_rate": 9.961379730329683e-05,
|
| 1925 |
+
"loss": 1.6631,
|
| 1926 |
+
"step": 1585
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 1.65,
|
| 1930 |
+
"learning_rate": 0.00010060977209176536,
|
| 1931 |
+
"loss": 1.7168,
|
| 1932 |
+
"step": 1590
|
| 1933 |
+
},
|
| 1934 |
+
{
|
| 1935 |
+
"epoch": 1.66,
|
| 1936 |
+
"learning_rate": 0.0001015971949156952,
|
| 1937 |
+
"loss": 1.752,
|
| 1938 |
+
"step": 1595
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 1.66,
|
| 1942 |
+
"learning_rate": 0.00010257580196799162,
|
| 1943 |
+
"loss": 1.7518,
|
| 1944 |
+
"step": 1600
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 1.67,
|
| 1948 |
+
"learning_rate": 0.00010354533179684596,
|
| 1949 |
+
"loss": 1.7591,
|
| 1950 |
+
"step": 1605
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 1.68,
|
| 1954 |
+
"learning_rate": 0.00010450552537558691,
|
| 1955 |
+
"loss": 1.7579,
|
| 1956 |
+
"step": 1610
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 1.68,
|
| 1960 |
+
"learning_rate": 0.00010545612617188394,
|
| 1961 |
+
"loss": 1.7364,
|
| 1962 |
+
"step": 1615
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 1.69,
|
| 1966 |
+
"learning_rate": 0.00010639688021628421,
|
| 1967 |
+
"loss": 1.7379,
|
| 1968 |
+
"step": 1620
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 1.69,
|
| 1972 |
+
"learning_rate": 0.00010732753617006524,
|
| 1973 |
+
"loss": 1.7443,
|
| 1974 |
+
"step": 1625
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 1.7,
|
| 1978 |
+
"learning_rate": 0.00010824784539238402,
|
| 1979 |
+
"loss": 1.6989,
|
| 1980 |
+
"step": 1630
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 1.7,
|
| 1984 |
+
"learning_rate": 0.00010915756200670606,
|
| 1985 |
+
"loss": 1.6362,
|
| 1986 |
+
"step": 1635
|
| 1987 |
+
},
|
| 1988 |
+
{
|
| 1989 |
+
"epoch": 1.71,
|
| 1990 |
+
"learning_rate": 0.00011005644296649529,
|
| 1991 |
+
"loss": 1.7417,
|
| 1992 |
+
"step": 1640
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 1.71,
|
| 1996 |
+
"learning_rate": 0.00011094424812014832,
|
| 1997 |
+
"loss": 1.635,
|
| 1998 |
+
"step": 1645
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 1.72,
|
| 2002 |
+
"learning_rate": 0.00011182074027515503,
|
| 2003 |
+
"loss": 1.6471,
|
| 2004 |
+
"step": 1650
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 1.72,
|
| 2008 |
+
"learning_rate": 0.0001126856852614686,
|
| 2009 |
+
"loss": 1.6803,
|
| 2010 |
+
"step": 1655
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 1.73,
|
| 2014 |
+
"learning_rate": 0.00011353885199406824,
|
| 2015 |
+
"loss": 1.6863,
|
| 2016 |
+
"step": 1660
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 1.73,
|
| 2020 |
+
"learning_rate": 0.00011438001253469733,
|
| 2021 |
+
"loss": 1.7498,
|
| 2022 |
+
"step": 1665
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 1.74,
|
| 2026 |
+
"learning_rate": 0.00011520894215276136,
|
| 2027 |
+
"loss": 1.7037,
|
| 2028 |
+
"step": 1670
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 1.74,
|
| 2032 |
+
"learning_rate": 0.00011602541938536831,
|
| 2033 |
+
"loss": 1.6918,
|
| 2034 |
+
"step": 1675
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 1.75,
|
| 2038 |
+
"learning_rate": 0.00011682922609649652,
|
| 2039 |
+
"loss": 1.7,
|
| 2040 |
+
"step": 1680
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 1.75,
|
| 2044 |
+
"learning_rate": 0.0001176201475352733,
|
| 2045 |
+
"loss": 1.7662,
|
| 2046 |
+
"step": 1685
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 1.76,
|
| 2050 |
+
"learning_rate": 0.00011839797239334955,
|
| 2051 |
+
"loss": 1.8041,
|
| 2052 |
+
"step": 1690
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 1.76,
|
| 2056 |
+
"learning_rate": 0.00011916249286135422,
|
| 2057 |
+
"loss": 1.7271,
|
| 2058 |
+
"step": 1695
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 1.77,
|
| 2062 |
+
"learning_rate": 0.00011991350468441439,
|
| 2063 |
+
"loss": 1.7033,
|
| 2064 |
+
"step": 1700
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 1.77,
|
| 2068 |
+
"learning_rate": 0.00012065080721672542,
|
| 2069 |
+
"loss": 1.6855,
|
| 2070 |
+
"step": 1705
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 1.78,
|
| 2074 |
+
"learning_rate": 0.00012137420347515691,
|
| 2075 |
+
"loss": 1.6554,
|
| 2076 |
+
"step": 1710
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 1.78,
|
| 2080 |
+
"learning_rate": 0.00012208350019188007,
|
| 2081 |
+
"loss": 1.7365,
|
| 2082 |
+
"step": 1715
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 1.79,
|
| 2086 |
+
"learning_rate": 0.00012277850786600282,
|
| 2087 |
+
"loss": 1.7106,
|
| 2088 |
+
"step": 1720
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 1.8,
|
| 2092 |
+
"learning_rate": 0.00012345904081419794,
|
| 2093 |
+
"loss": 1.735,
|
| 2094 |
+
"step": 1725
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 1.8,
|
| 2098 |
+
"learning_rate": 0.00012412491722031186,
|
| 2099 |
+
"loss": 1.6991,
|
| 2100 |
+
"step": 1730
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 1.81,
|
| 2104 |
+
"learning_rate": 0.00012477595918393978,
|
| 2105 |
+
"loss": 1.6981,
|
| 2106 |
+
"step": 1735
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 1.81,
|
| 2110 |
+
"learning_rate": 0.0001254119927679549,
|
| 2111 |
+
"loss": 1.7559,
|
| 2112 |
+
"step": 1740
|
| 2113 |
+
},
|
| 2114 |
+
{
|
| 2115 |
+
"epoch": 1.82,
|
| 2116 |
+
"learning_rate": 0.00012603284804497882,
|
| 2117 |
+
"loss": 1.7479,
|
| 2118 |
+
"step": 1745
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 1.82,
|
| 2122 |
+
"learning_rate": 0.00012663835914278047,
|
| 2123 |
+
"loss": 1.7712,
|
| 2124 |
+
"step": 1750
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 1.83,
|
| 2128 |
+
"learning_rate": 0.0001272283642885918,
|
| 2129 |
+
"loss": 1.7758,
|
| 2130 |
+
"step": 1755
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 1.83,
|
| 2134 |
+
"learning_rate": 0.00012780270585232808,
|
| 2135 |
+
"loss": 1.7353,
|
| 2136 |
+
"step": 1760
|
| 2137 |
+
},
|
| 2138 |
+
{
|
| 2139 |
+
"epoch": 1.84,
|
| 2140 |
+
"learning_rate": 0.00012836123038870173,
|
| 2141 |
+
"loss": 1.7297,
|
| 2142 |
+
"step": 1765
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 1.84,
|
| 2146 |
+
"learning_rate": 0.00012890378867821763,
|
| 2147 |
+
"loss": 1.8349,
|
| 2148 |
+
"step": 1770
|
| 2149 |
+
},
|
| 2150 |
+
{
|
| 2151 |
+
"epoch": 1.85,
|
| 2152 |
+
"learning_rate": 0.0001294302357670399,
|
| 2153 |
+
"loss": 1.8035,
|
| 2154 |
+
"step": 1775
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 1.85,
|
| 2158 |
+
"learning_rate": 0.00012994043100571866,
|
| 2159 |
+
"loss": 1.7481,
|
| 2160 |
+
"step": 1780
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 1.86,
|
| 2164 |
+
"learning_rate": 0.000130434238086767,
|
| 2165 |
+
"loss": 1.6094,
|
| 2166 |
+
"step": 1785
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 1.86,
|
| 2170 |
+
"learning_rate": 0.00013091152508107798,
|
| 2171 |
+
"loss": 1.7314,
|
| 2172 |
+
"step": 1790
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 1.87,
|
| 2176 |
+
"learning_rate": 0.00013137216447317167,
|
| 2177 |
+
"loss": 1.8012,
|
| 2178 |
+
"step": 1795
|
| 2179 |
+
},
|
| 2180 |
+
{
|
| 2181 |
+
"epoch": 1.87,
|
| 2182 |
+
"learning_rate": 0.0001318160331952632,
|
| 2183 |
+
"loss": 1.7435,
|
| 2184 |
+
"step": 1800
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 1.88,
|
| 2188 |
+
"learning_rate": 0.0001322430126601424,
|
| 2189 |
+
"loss": 1.7132,
|
| 2190 |
+
"step": 1805
|
| 2191 |
+
},
|
| 2192 |
+
{
|
| 2193 |
+
"epoch": 1.88,
|
| 2194 |
+
"learning_rate": 0.00013265298879285635,
|
| 2195 |
+
"loss": 1.6928,
|
| 2196 |
+
"step": 1810
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 1.89,
|
| 2200 |
+
"learning_rate": 0.00013304585206118667,
|
| 2201 |
+
"loss": 1.6412,
|
| 2202 |
+
"step": 1815
|
| 2203 |
+
},
|
| 2204 |
+
{
|
| 2205 |
+
"epoch": 1.89,
|
| 2206 |
+
"learning_rate": 0.00013342149750491278,
|
| 2207 |
+
"loss": 1.7068,
|
| 2208 |
+
"step": 1820
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 1.9,
|
| 2212 |
+
"learning_rate": 0.0001337798247638538,
|
| 2213 |
+
"loss": 1.7385,
|
| 2214 |
+
"step": 1825
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 1.9,
|
| 2218 |
+
"learning_rate": 0.0001341207381046819,
|
| 2219 |
+
"loss": 1.612,
|
| 2220 |
+
"step": 1830
|
| 2221 |
+
},
|
| 2222 |
+
{
|
| 2223 |
+
"epoch": 1.91,
|
| 2224 |
+
"learning_rate": 0.00013444414644649843,
|
| 2225 |
+
"loss": 1.6917,
|
| 2226 |
+
"step": 1835
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 1.91,
|
| 2230 |
+
"learning_rate": 0.00013474996338516847,
|
| 2231 |
+
"loss": 1.7257,
|
| 2232 |
+
"step": 1840
|
| 2233 |
+
},
|
| 2234 |
+
{
|
| 2235 |
+
"epoch": 1.92,
|
| 2236 |
+
"learning_rate": 0.00013503810721640465,
|
| 2237 |
+
"loss": 1.7391,
|
| 2238 |
+
"step": 1845
|
| 2239 |
+
},
|
| 2240 |
+
{
|
| 2241 |
+
"epoch": 1.93,
|
| 2242 |
+
"learning_rate": 0.00013530850095759623,
|
| 2243 |
+
"loss": 1.7311,
|
| 2244 |
+
"step": 1850
|
| 2245 |
+
},
|
| 2246 |
+
{
|
| 2247 |
+
"epoch": 1.93,
|
| 2248 |
+
"learning_rate": 0.00013556107236837607,
|
| 2249 |
+
"loss": 1.671,
|
| 2250 |
+
"step": 1855
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 1.94,
|
| 2254 |
+
"learning_rate": 0.00013579575396992113,
|
| 2255 |
+
"loss": 1.716,
|
| 2256 |
+
"step": 1860
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 1.94,
|
| 2260 |
+
"learning_rate": 0.00013601248306298038,
|
| 2261 |
+
"loss": 1.7233,
|
| 2262 |
+
"step": 1865
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 1.95,
|
| 2266 |
+
"learning_rate": 0.00013621120174462615,
|
| 2267 |
+
"loss": 1.6546,
|
| 2268 |
+
"step": 1870
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 1.95,
|
| 2272 |
+
"learning_rate": 0.00013639185692372385,
|
| 2273 |
+
"loss": 1.7717,
|
| 2274 |
+
"step": 1875
|
| 2275 |
+
},
|
| 2276 |
+
{
|
| 2277 |
+
"epoch": 1.96,
|
| 2278 |
+
"learning_rate": 0.00013655440033511618,
|
| 2279 |
+
"loss": 1.7035,
|
| 2280 |
+
"step": 1880
|
| 2281 |
+
},
|
| 2282 |
+
{
|
| 2283 |
+
"epoch": 1.96,
|
| 2284 |
+
"learning_rate": 0.00013669878855251797,
|
| 2285 |
+
"loss": 1.7345,
|
| 2286 |
+
"step": 1885
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 1.97,
|
| 2290 |
+
"learning_rate": 0.00013682498300011836,
|
| 2291 |
+
"loss": 1.7678,
|
| 2292 |
+
"step": 1890
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 1.97,
|
| 2296 |
+
"learning_rate": 0.00013693294996288687,
|
| 2297 |
+
"loss": 1.6938,
|
| 2298 |
+
"step": 1895
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 1.98,
|
| 2302 |
+
"learning_rate": 0.00013702266059558108,
|
| 2303 |
+
"loss": 1.7334,
|
| 2304 |
+
"step": 1900
|
| 2305 |
+
},
|
| 2306 |
+
{
|
| 2307 |
+
"epoch": 1.98,
|
| 2308 |
+
"learning_rate": 0.00013709409093045299,
|
| 2309 |
+
"loss": 1.6722,
|
| 2310 |
+
"step": 1905
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"epoch": 1.99,
|
| 2314 |
+
"learning_rate": 0.00013714722188365257,
|
| 2315 |
+
"loss": 1.7148,
|
| 2316 |
+
"step": 1910
|
| 2317 |
+
},
|
| 2318 |
+
{
|
| 2319 |
+
"epoch": 1.99,
|
| 2320 |
+
"learning_rate": 0.00013718203926032623,
|
| 2321 |
+
"loss": 1.6982,
|
| 2322 |
+
"step": 1915
|
| 2323 |
+
},
|
| 2324 |
+
{
|
| 2325 |
+
"epoch": 2.0,
|
| 2326 |
+
"learning_rate": 0.00013719853375840932,
|
| 2327 |
+
"loss": 1.7054,
|
| 2328 |
+
"step": 1920
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 2.0,
|
| 2332 |
+
"eval_loss": 1.6512105464935303,
|
| 2333 |
+
"eval_runtime": 68.7702,
|
| 2334 |
+
"eval_samples_per_second": 20.503,
|
| 2335 |
+
"eval_steps_per_second": 2.574,
|
| 2336 |
+
"step": 1922
|
| 2337 |
}
|
| 2338 |
],
|
| 2339 |
+
"max_steps": 1922,
|
| 2340 |
+
"num_train_epochs": 2,
|
| 2341 |
+
"total_flos": 2008029265920000.0,
|
| 2342 |
"trial_name": null,
|
| 2343 |
"trial_params": null
|
| 2344 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2671
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:67ca83f5ce021a7b1aa7cd6b2ae011df0668facd380e0c315a7528650bbe5092
|
| 3 |
size 2671
|