Commit
·
9cb590c
1
Parent(s):
e03102c
huggingartists
Browse files- README.md +3 -3
- config.json +3 -1
- evaluation.txt +1 -1
- flax_model.msgpack +1 -1
- optimizer.pt +1 -1
- pytorch_model.bin +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +1747 -7
- training_args.bin +2 -2
README.md
CHANGED
|
@@ -45,15 +45,15 @@ from datasets import load_dataset
|
|
| 45 |
dataset = load_dataset("huggingartists/boris-grebenshikov")
|
| 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 Борис Гребенщиков (Boris Grebenshikov)'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/boris-grebenshikov")
|
| 46 |
```
|
| 47 |
|
| 48 |
+
[Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/3nb43gls/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 Борис Гребенщиков (Boris Grebenshikov)'s lyrics.
|
| 53 |
|
| 54 |
+
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/34p8ye7k) for full transparency and reproducibility.
|
| 55 |
|
| 56 |
+
At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/34p8ye7k/artifacts) is logged and versioned.
|
| 57 |
|
| 58 |
## How to use
|
| 59 |
|
config.json
CHANGED
|
@@ -18,7 +18,9 @@
|
|
| 18 |
"n_inner": null,
|
| 19 |
"n_layer": 12,
|
| 20 |
"n_positions": 1024,
|
|
|
|
| 21 |
"resid_pdrop": 0.1,
|
|
|
|
| 22 |
"scale_attn_weights": true,
|
| 23 |
"summary_activation": null,
|
| 24 |
"summary_first_dropout": 0.1,
|
|
@@ -35,7 +37,7 @@
|
|
| 35 |
}
|
| 36 |
},
|
| 37 |
"torch_dtype": "float32",
|
| 38 |
-
"transformers_version": "4.
|
| 39 |
"use_cache": true,
|
| 40 |
"vocab_size": 50257
|
| 41 |
}
|
|
|
|
| 18 |
"n_inner": null,
|
| 19 |
"n_layer": 12,
|
| 20 |
"n_positions": 1024,
|
| 21 |
+
"reorder_and_upcast_attn": false,
|
| 22 |
"resid_pdrop": 0.1,
|
| 23 |
+
"scale_attn_by_inverse_layer_idx": false,
|
| 24 |
"scale_attn_weights": true,
|
| 25 |
"summary_activation": null,
|
| 26 |
"summary_first_dropout": 0.1,
|
|
|
|
| 37 |
}
|
| 38 |
},
|
| 39 |
"torch_dtype": "float32",
|
| 40 |
+
"transformers_version": "4.12.3",
|
| 41 |
"use_cache": true,
|
| 42 |
"vocab_size": 50257
|
| 43 |
}
|
evaluation.txt
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"eval_loss": 1.
|
|
|
|
| 1 |
+
{"eval_loss": 1.4596115350723267, "eval_runtime": 15.4471, "eval_samples_per_second": 14.76, "eval_steps_per_second": 1.877, "epoch": 13.0}
|
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 sha256:7b80a7b117fa8fcfdc92538bd1c749f3036846107cbc5d8e9f64eb849bb5ec92
|
| 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 sha256:5b2b56a72c94fe42a65e8817f7385e51070afc90c3af9858ca840d10ab763523
|
| 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 sha256:2ca27b4f0fac085f1d19ab9ee0ab56d92b931da2d0875538f42110a138b55842
|
| 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 14567
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:33e5560d3e336064cf05e59f7017fa9c91033de56df295788444612ddc529ffa
|
| 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 sha256:3ced6f9dc1a5421fbd523c43683a1b3e93bb93aa081e3667e9a65f022db0f820
|
| 3 |
size 623
|
trainer_state.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 1.
|
| 3 |
-
"best_model_checkpoint": "output/boris-grebenshikov/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -650,11 +650,1751 @@
|
|
| 650 |
"eval_samples_per_second": 22.761,
|
| 651 |
"eval_steps_per_second": 2.913,
|
| 652 |
"step": 510
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 653 |
}
|
| 654 |
],
|
| 655 |
-
"max_steps":
|
| 656 |
-
"num_train_epochs":
|
| 657 |
-
"total_flos":
|
| 658 |
"trial_name": null,
|
| 659 |
"trial_params": null
|
| 660 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.4596115350723267,
|
| 3 |
+
"best_model_checkpoint": "output/boris-grebenshikov/checkpoint-1903",
|
| 4 |
+
"epoch": 11.0,
|
| 5 |
+
"global_step": 1903,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 650 |
"eval_samples_per_second": 22.761,
|
| 651 |
"eval_steps_per_second": 2.913,
|
| 652 |
"step": 510
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 2.98,
|
| 656 |
+
"learning_rate": 1.8089671428491253e-07,
|
| 657 |
+
"loss": 1.7435,
|
| 658 |
+
"step": 515
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 3.0,
|
| 662 |
+
"eval_loss": 1.5615885257720947,
|
| 663 |
+
"eval_runtime": 14.4733,
|
| 664 |
+
"eval_samples_per_second": 15.753,
|
| 665 |
+
"eval_steps_per_second": 2.004,
|
| 666 |
+
"step": 519
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 3.01,
|
| 670 |
+
"learning_rate": 1.1310706275549354e-08,
|
| 671 |
+
"loss": 1.7668,
|
| 672 |
+
"step": 520
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 3.03,
|
| 676 |
+
"learning_rate": 4.0679393484475955e-07,
|
| 677 |
+
"loss": 1.6029,
|
| 678 |
+
"step": 525
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 3.06,
|
| 682 |
+
"learning_rate": 1.3640882038038144e-06,
|
| 683 |
+
"loss": 1.7397,
|
| 684 |
+
"step": 530
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 3.09,
|
| 688 |
+
"learning_rate": 2.875306825896957e-06,
|
| 689 |
+
"loss": 1.6337,
|
| 690 |
+
"step": 535
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 3.12,
|
| 694 |
+
"learning_rate": 4.927999597426121e-06,
|
| 695 |
+
"loss": 1.7185,
|
| 696 |
+
"step": 540
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 3.15,
|
| 700 |
+
"learning_rate": 7.505255369493709e-06,
|
| 701 |
+
"loss": 1.7042,
|
| 702 |
+
"step": 545
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.18,
|
| 706 |
+
"learning_rate": 1.0585841370828143e-05,
|
| 707 |
+
"loss": 1.6665,
|
| 708 |
+
"step": 550
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 3.21,
|
| 712 |
+
"learning_rate": 1.4144378134378619e-05,
|
| 713 |
+
"loss": 1.6517,
|
| 714 |
+
"step": 555
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 3.24,
|
| 718 |
+
"learning_rate": 1.8151548586542735e-05,
|
| 719 |
+
"loss": 1.7248,
|
| 720 |
+
"step": 560
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 3.27,
|
| 724 |
+
"learning_rate": 2.257433957644151e-05,
|
| 725 |
+
"loss": 1.7293,
|
| 726 |
+
"step": 565
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 3.29,
|
| 730 |
+
"learning_rate": 2.7376313855398193e-05,
|
| 731 |
+
"loss": 1.5925,
|
| 732 |
+
"step": 570
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 3.32,
|
| 736 |
+
"learning_rate": 3.25179102659126e-05,
|
| 737 |
+
"loss": 1.7049,
|
| 738 |
+
"step": 575
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 3.35,
|
| 742 |
+
"learning_rate": 3.795676966701845e-05,
|
| 743 |
+
"loss": 1.6687,
|
| 744 |
+
"step": 580
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.38,
|
| 748 |
+
"learning_rate": 4.3648083910881356e-05,
|
| 749 |
+
"loss": 1.6983,
|
| 750 |
+
"step": 585
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 3.41,
|
| 754 |
+
"learning_rate": 4.954496499558694e-05,
|
| 755 |
+
"loss": 1.785,
|
| 756 |
+
"step": 590
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 3.44,
|
| 760 |
+
"learning_rate": 5.559883135284689e-05,
|
| 761 |
+
"loss": 1.698,
|
| 762 |
+
"step": 595
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 3.47,
|
| 766 |
+
"learning_rate": 6.175980808818411e-05,
|
| 767 |
+
"loss": 1.6591,
|
| 768 |
+
"step": 600
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 3.5,
|
| 772 |
+
"learning_rate": 6.797713787621067e-05,
|
| 773 |
+
"loss": 1.7702,
|
| 774 |
+
"step": 605
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 3.53,
|
| 778 |
+
"learning_rate": 7.419959912582961e-05,
|
| 779 |
+
"loss": 1.7052,
|
| 780 |
+
"step": 610
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 3.55,
|
| 784 |
+
"learning_rate": 8.037592797029894e-05,
|
| 785 |
+
"loss": 1.6658,
|
| 786 |
+
"step": 615
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.58,
|
| 790 |
+
"learning_rate": 8.645524060558421e-05,
|
| 791 |
+
"loss": 1.6903,
|
| 792 |
+
"step": 620
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.61,
|
| 796 |
+
"learning_rate": 9.238745249755704e-05,
|
| 797 |
+
"loss": 1.7102,
|
| 798 |
+
"step": 625
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 3.64,
|
| 802 |
+
"learning_rate": 9.812369100439384e-05,
|
| 803 |
+
"loss": 1.7289,
|
| 804 |
+
"step": 630
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 3.67,
|
| 808 |
+
"learning_rate": 0.00010361669801477512,
|
| 809 |
+
"loss": 1.688,
|
| 810 |
+
"step": 635
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 3.7,
|
| 814 |
+
"learning_rate": 0.00010882121928474553,
|
| 815 |
+
"loss": 1.7392,
|
| 816 |
+
"step": 640
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 3.73,
|
| 820 |
+
"learning_rate": 0.00011369437726567332,
|
| 821 |
+
"loss": 1.7228,
|
| 822 |
+
"step": 645
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 3.76,
|
| 826 |
+
"learning_rate": 0.00011819602435175987,
|
| 827 |
+
"loss": 1.7298,
|
| 828 |
+
"step": 650
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.79,
|
| 832 |
+
"learning_rate": 0.0001222890736368625,
|
| 833 |
+
"loss": 1.6633,
|
| 834 |
+
"step": 655
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 3.82,
|
| 838 |
+
"learning_rate": 0.00012593980445568464,
|
| 839 |
+
"loss": 1.7014,
|
| 840 |
+
"step": 660
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 3.84,
|
| 844 |
+
"learning_rate": 0.0001291181401921254,
|
| 845 |
+
"loss": 1.6952,
|
| 846 |
+
"step": 665
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 3.87,
|
| 850 |
+
"learning_rate": 0.00013179789606606037,
|
| 851 |
+
"loss": 1.7324,
|
| 852 |
+
"step": 670
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 3.9,
|
| 856 |
+
"learning_rate": 0.0001339569948571563,
|
| 857 |
+
"loss": 1.6753,
|
| 858 |
+
"step": 675
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 3.93,
|
| 862 |
+
"learning_rate": 0.00013557764878847588,
|
| 863 |
+
"loss": 1.834,
|
| 864 |
+
"step": 680
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 3.96,
|
| 868 |
+
"learning_rate": 0.0001366465060714202,
|
| 869 |
+
"loss": 1.6832,
|
| 870 |
+
"step": 685
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.99,
|
| 874 |
+
"learning_rate": 0.00013715476090469595,
|
| 875 |
+
"loss": 1.703,
|
| 876 |
+
"step": 690
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 4.0,
|
| 880 |
+
"eval_loss": 1.588655710220337,
|
| 881 |
+
"eval_runtime": 14.8949,
|
| 882 |
+
"eval_samples_per_second": 15.307,
|
| 883 |
+
"eval_steps_per_second": 1.947,
|
| 884 |
+
"step": 692
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.02,
|
| 888 |
+
"learning_rate": 0.000137098226021079,
|
| 889 |
+
"loss": 1.6012,
|
| 890 |
+
"step": 695
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 4.05,
|
| 894 |
+
"learning_rate": 0.0001364773671842959,
|
| 895 |
+
"loss": 1.7298,
|
| 896 |
+
"step": 700
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 4.08,
|
| 900 |
+
"learning_rate": 0.0001352972993518206,
|
| 901 |
+
"loss": 1.6923,
|
| 902 |
+
"step": 705
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 4.1,
|
| 906 |
+
"learning_rate": 0.0001335677445351985,
|
| 907 |
+
"loss": 1.7106,
|
| 908 |
+
"step": 710
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 4.13,
|
| 912 |
+
"learning_rate": 0.00013130295170506775,
|
| 913 |
+
"loss": 1.6598,
|
| 914 |
+
"step": 715
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 4.16,
|
| 918 |
+
"learning_rate": 0.00012852157940074013,
|
| 919 |
+
"loss": 1.6285,
|
| 920 |
+
"step": 720
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 4.19,
|
| 924 |
+
"learning_rate": 0.00012524654201146593,
|
| 925 |
+
"loss": 1.6431,
|
| 926 |
+
"step": 725
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.22,
|
| 930 |
+
"learning_rate": 0.00012150482099579733,
|
| 931 |
+
"loss": 1.5908,
|
| 932 |
+
"step": 730
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 4.25,
|
| 936 |
+
"learning_rate": 0.00011732724259432273,
|
| 937 |
+
"loss": 1.6991,
|
| 938 |
+
"step": 735
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 4.28,
|
| 942 |
+
"learning_rate": 0.00011274822386709064,
|
| 943 |
+
"loss": 1.5654,
|
| 944 |
+
"step": 740
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 4.31,
|
| 948 |
+
"learning_rate": 0.00010780548914799808,
|
| 949 |
+
"loss": 1.6367,
|
| 950 |
+
"step": 745
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 4.34,
|
| 954 |
+
"learning_rate": 0.00010253975925213988,
|
| 955 |
+
"loss": 1.7132,
|
| 956 |
+
"step": 750
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 4.36,
|
| 960 |
+
"learning_rate": 9.699441599659e-05,
|
| 961 |
+
"loss": 1.6216,
|
| 962 |
+
"step": 755
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 4.39,
|
| 966 |
+
"learning_rate": 9.121514479846673e-05,
|
| 967 |
+
"loss": 1.6058,
|
| 968 |
+
"step": 760
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.42,
|
| 972 |
+
"learning_rate": 8.52495582947446e-05,
|
| 973 |
+
"loss": 1.5635,
|
| 974 |
+
"step": 765
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 4.45,
|
| 978 |
+
"learning_rate": 7.914680408462862e-05,
|
| 979 |
+
"loss": 1.6392,
|
| 980 |
+
"step": 770
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 4.48,
|
| 984 |
+
"learning_rate": 7.295715982611233e-05,
|
| 985 |
+
"loss": 1.5928,
|
| 986 |
+
"step": 775
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 4.51,
|
| 990 |
+
"learning_rate": 6.673161902252507e-05,
|
| 991 |
+
"loss": 1.5735,
|
| 992 |
+
"step": 780
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 4.54,
|
| 996 |
+
"learning_rate": 6.052147091157344e-05,
|
| 997 |
+
"loss": 1.645,
|
| 998 |
+
"step": 785
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 4.57,
|
| 1002 |
+
"learning_rate": 5.4377877917969595e-05,
|
| 1003 |
+
"loss": 1.6497,
|
| 1004 |
+
"step": 790
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 4.6,
|
| 1008 |
+
"learning_rate": 4.8351454150810734e-05,
|
| 1009 |
+
"loss": 1.6139,
|
| 1010 |
+
"step": 795
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.62,
|
| 1014 |
+
"learning_rate": 4.249184841826729e-05,
|
| 1015 |
+
"loss": 1.5811,
|
| 1016 |
+
"step": 800
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 4.65,
|
| 1020 |
+
"learning_rate": 3.6847335194921544e-05,
|
| 1021 |
+
"loss": 1.5699,
|
| 1022 |
+
"step": 805
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 4.68,
|
| 1026 |
+
"learning_rate": 3.146441691157973e-05,
|
| 1027 |
+
"loss": 1.6003,
|
| 1028 |
+
"step": 810
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 4.71,
|
| 1032 |
+
"learning_rate": 2.638744084410121e-05,
|
| 1033 |
+
"loss": 1.6607,
|
| 1034 |
+
"step": 815
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 4.74,
|
| 1038 |
+
"learning_rate": 2.165823375751248e-05,
|
| 1039 |
+
"loss": 1.6131,
|
| 1040 |
+
"step": 820
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 4.77,
|
| 1044 |
+
"learning_rate": 1.7315757315399247e-05,
|
| 1045 |
+
"loss": 1.5716,
|
| 1046 |
+
"step": 825
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 4.8,
|
| 1050 |
+
"learning_rate": 1.3395787093490896e-05,
|
| 1051 |
+
"loss": 1.6196,
|
| 1052 |
+
"step": 830
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.83,
|
| 1056 |
+
"learning_rate": 9.930617841891443e-06,
|
| 1057 |
+
"loss": 1.6354,
|
| 1058 |
+
"step": 835
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 4.86,
|
| 1062 |
+
"learning_rate": 6.948797424159463e-06,
|
| 1063 |
+
"loss": 1.5669,
|
| 1064 |
+
"step": 840
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 4.88,
|
| 1068 |
+
"learning_rate": 4.474891625184719e-06,
|
| 1069 |
+
"loss": 1.604,
|
| 1070 |
+
"step": 845
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 4.91,
|
| 1074 |
+
"learning_rate": 2.5292817654967284e-06,
|
| 1075 |
+
"loss": 1.523,
|
| 1076 |
+
"step": 850
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 4.94,
|
| 1080 |
+
"learning_rate": 1.127996789363344e-06,
|
| 1081 |
+
"loss": 1.6719,
|
| 1082 |
+
"step": 855
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 4.97,
|
| 1086 |
+
"learning_rate": 2.8258121002494224e-07,
|
| 1087 |
+
"loss": 1.6116,
|
| 1088 |
+
"step": 860
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 5.0,
|
| 1092 |
+
"learning_rate": 0.0,
|
| 1093 |
+
"loss": 1.4979,
|
| 1094 |
+
"step": 865
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 5.0,
|
| 1098 |
+
"eval_loss": 1.516666054725647,
|
| 1099 |
+
"eval_runtime": 14.812,
|
| 1100 |
+
"eval_samples_per_second": 15.393,
|
| 1101 |
+
"eval_steps_per_second": 1.958,
|
| 1102 |
+
"step": 865
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 5.03,
|
| 1106 |
+
"learning_rate": 2.825812100249346e-07,
|
| 1107 |
+
"loss": 1.5673,
|
| 1108 |
+
"step": 870
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 5.06,
|
| 1112 |
+
"learning_rate": 1.127996789363329e-06,
|
| 1113 |
+
"loss": 1.5619,
|
| 1114 |
+
"step": 875
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 5.09,
|
| 1118 |
+
"learning_rate": 2.529281765496713e-06,
|
| 1119 |
+
"loss": 1.5238,
|
| 1120 |
+
"step": 880
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 5.12,
|
| 1124 |
+
"learning_rate": 4.4748916251846885e-06,
|
| 1125 |
+
"loss": 1.5144,
|
| 1126 |
+
"step": 885
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 5.14,
|
| 1130 |
+
"learning_rate": 6.948797424159478e-06,
|
| 1131 |
+
"loss": 1.5482,
|
| 1132 |
+
"step": 890
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 5.17,
|
| 1136 |
+
"learning_rate": 9.930617841891406e-06,
|
| 1137 |
+
"loss": 1.593,
|
| 1138 |
+
"step": 895
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 5.2,
|
| 1142 |
+
"learning_rate": 1.3395787093490912e-05,
|
| 1143 |
+
"loss": 1.6012,
|
| 1144 |
+
"step": 900
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 5.23,
|
| 1148 |
+
"learning_rate": 1.7315757315399193e-05,
|
| 1149 |
+
"loss": 1.5677,
|
| 1150 |
+
"step": 905
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.26,
|
| 1154 |
+
"learning_rate": 2.1658233757512504e-05,
|
| 1155 |
+
"loss": 1.5793,
|
| 1156 |
+
"step": 910
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 5.29,
|
| 1160 |
+
"learning_rate": 2.638744084410115e-05,
|
| 1161 |
+
"loss": 1.5031,
|
| 1162 |
+
"step": 915
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 5.32,
|
| 1166 |
+
"learning_rate": 3.1464416911579663e-05,
|
| 1167 |
+
"loss": 1.5173,
|
| 1168 |
+
"step": 920
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 5.35,
|
| 1172 |
+
"learning_rate": 3.684733519492136e-05,
|
| 1173 |
+
"loss": 1.489,
|
| 1174 |
+
"step": 925
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 5.38,
|
| 1178 |
+
"learning_rate": 4.249184841826722e-05,
|
| 1179 |
+
"loss": 1.6251,
|
| 1180 |
+
"step": 930
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 5.4,
|
| 1184 |
+
"learning_rate": 4.835145415081054e-05,
|
| 1185 |
+
"loss": 1.5882,
|
| 1186 |
+
"step": 935
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 5.43,
|
| 1190 |
+
"learning_rate": 5.4377877917969514e-05,
|
| 1191 |
+
"loss": 1.6055,
|
| 1192 |
+
"step": 940
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.46,
|
| 1196 |
+
"learning_rate": 6.052147091157324e-05,
|
| 1197 |
+
"loss": 1.5492,
|
| 1198 |
+
"step": 945
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 5.49,
|
| 1202 |
+
"learning_rate": 6.673161902252498e-05,
|
| 1203 |
+
"loss": 1.534,
|
| 1204 |
+
"step": 950
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 5.52,
|
| 1208 |
+
"learning_rate": 7.295715982611214e-05,
|
| 1209 |
+
"loss": 1.5354,
|
| 1210 |
+
"step": 955
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 5.55,
|
| 1214 |
+
"learning_rate": 7.914680408462853e-05,
|
| 1215 |
+
"loss": 1.6055,
|
| 1216 |
+
"step": 960
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 5.58,
|
| 1220 |
+
"learning_rate": 8.52495582947444e-05,
|
| 1221 |
+
"loss": 1.5166,
|
| 1222 |
+
"step": 965
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 5.61,
|
| 1226 |
+
"learning_rate": 9.121514479846663e-05,
|
| 1227 |
+
"loss": 1.5157,
|
| 1228 |
+
"step": 970
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 5.64,
|
| 1232 |
+
"learning_rate": 9.699441599658982e-05,
|
| 1233 |
+
"loss": 1.6171,
|
| 1234 |
+
"step": 975
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.66,
|
| 1238 |
+
"learning_rate": 0.00010253975925213981,
|
| 1239 |
+
"loss": 1.5676,
|
| 1240 |
+
"step": 980
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 5.69,
|
| 1244 |
+
"learning_rate": 0.0001078054891479979,
|
| 1245 |
+
"loss": 1.5382,
|
| 1246 |
+
"step": 985
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 5.72,
|
| 1250 |
+
"learning_rate": 0.00011274822386709058,
|
| 1251 |
+
"loss": 1.6057,
|
| 1252 |
+
"step": 990
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 5.75,
|
| 1256 |
+
"learning_rate": 0.00011732724259432258,
|
| 1257 |
+
"loss": 1.5982,
|
| 1258 |
+
"step": 995
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 5.78,
|
| 1262 |
+
"learning_rate": 0.00012150482099579729,
|
| 1263 |
+
"loss": 1.5857,
|
| 1264 |
+
"step": 1000
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 5.81,
|
| 1268 |
+
"learning_rate": 0.00012524654201146596,
|
| 1269 |
+
"loss": 1.6016,
|
| 1270 |
+
"step": 1005
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 5.84,
|
| 1274 |
+
"learning_rate": 0.00012852157940074007,
|
| 1275 |
+
"loss": 1.552,
|
| 1276 |
+
"step": 1010
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 5.87,
|
| 1280 |
+
"learning_rate": 0.00013130295170506778,
|
| 1281 |
+
"loss": 1.6309,
|
| 1282 |
+
"step": 1015
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 5.9,
|
| 1286 |
+
"learning_rate": 0.00013356774453519848,
|
| 1287 |
+
"loss": 1.6005,
|
| 1288 |
+
"step": 1020
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 5.92,
|
| 1292 |
+
"learning_rate": 0.0001352972993518206,
|
| 1293 |
+
"loss": 1.5484,
|
| 1294 |
+
"step": 1025
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 5.95,
|
| 1298 |
+
"learning_rate": 0.00013647736718429588,
|
| 1299 |
+
"loss": 1.5604,
|
| 1300 |
+
"step": 1030
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 5.98,
|
| 1304 |
+
"learning_rate": 0.000137098226021079,
|
| 1305 |
+
"loss": 1.5962,
|
| 1306 |
+
"step": 1035
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 6.0,
|
| 1310 |
+
"eval_loss": 1.5382394790649414,
|
| 1311 |
+
"eval_runtime": 14.8295,
|
| 1312 |
+
"eval_samples_per_second": 15.375,
|
| 1313 |
+
"eval_steps_per_second": 1.956,
|
| 1314 |
+
"step": 1038
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 6.01,
|
| 1318 |
+
"learning_rate": 0.00013715476090469595,
|
| 1319 |
+
"loss": 1.5942,
|
| 1320 |
+
"step": 1040
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 6.04,
|
| 1324 |
+
"learning_rate": 0.0001366465060714202,
|
| 1325 |
+
"loss": 1.4645,
|
| 1326 |
+
"step": 1045
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 6.07,
|
| 1330 |
+
"learning_rate": 0.00013557764878847597,
|
| 1331 |
+
"loss": 1.5359,
|
| 1332 |
+
"step": 1050
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 6.1,
|
| 1336 |
+
"learning_rate": 0.00013395699485715633,
|
| 1337 |
+
"loss": 1.5554,
|
| 1338 |
+
"step": 1055
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 6.13,
|
| 1342 |
+
"learning_rate": 0.00013179789606606045,
|
| 1343 |
+
"loss": 1.5214,
|
| 1344 |
+
"step": 1060
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 6.16,
|
| 1348 |
+
"learning_rate": 0.00012911814019212547,
|
| 1349 |
+
"loss": 1.4883,
|
| 1350 |
+
"step": 1065
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 6.18,
|
| 1354 |
+
"learning_rate": 0.00012593980445568475,
|
| 1355 |
+
"loss": 1.536,
|
| 1356 |
+
"step": 1070
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 6.21,
|
| 1360 |
+
"learning_rate": 0.00012228907363686257,
|
| 1361 |
+
"loss": 1.4736,
|
| 1362 |
+
"step": 1075
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 6.24,
|
| 1366 |
+
"learning_rate": 0.00011819602435175999,
|
| 1367 |
+
"loss": 1.5588,
|
| 1368 |
+
"step": 1080
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 6.27,
|
| 1372 |
+
"learning_rate": 0.00011369437726567339,
|
| 1373 |
+
"loss": 1.6177,
|
| 1374 |
+
"step": 1085
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 6.3,
|
| 1378 |
+
"learning_rate": 0.00010882121928474568,
|
| 1379 |
+
"loss": 1.5433,
|
| 1380 |
+
"step": 1090
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 6.33,
|
| 1384 |
+
"learning_rate": 0.00010361669801477519,
|
| 1385 |
+
"loss": 1.4821,
|
| 1386 |
+
"step": 1095
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 6.36,
|
| 1390 |
+
"learning_rate": 9.812369100439403e-05,
|
| 1391 |
+
"loss": 1.5071,
|
| 1392 |
+
"step": 1100
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 6.39,
|
| 1396 |
+
"learning_rate": 9.238745249755713e-05,
|
| 1397 |
+
"loss": 1.5067,
|
| 1398 |
+
"step": 1105
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 6.42,
|
| 1402 |
+
"learning_rate": 8.645524060558428e-05,
|
| 1403 |
+
"loss": 1.5131,
|
| 1404 |
+
"step": 1110
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 6.45,
|
| 1408 |
+
"learning_rate": 8.037592797029891e-05,
|
| 1409 |
+
"loss": 1.5808,
|
| 1410 |
+
"step": 1115
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 6.47,
|
| 1414 |
+
"learning_rate": 7.419959912582969e-05,
|
| 1415 |
+
"loss": 1.5181,
|
| 1416 |
+
"step": 1120
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 6.5,
|
| 1420 |
+
"learning_rate": 6.797713787621063e-05,
|
| 1421 |
+
"loss": 1.5158,
|
| 1422 |
+
"step": 1125
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 6.53,
|
| 1426 |
+
"learning_rate": 6.17598080881842e-05,
|
| 1427 |
+
"loss": 1.5602,
|
| 1428 |
+
"step": 1130
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 6.56,
|
| 1432 |
+
"learning_rate": 5.559883135284685e-05,
|
| 1433 |
+
"loss": 1.5248,
|
| 1434 |
+
"step": 1135
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 6.59,
|
| 1438 |
+
"learning_rate": 4.954496499558702e-05,
|
| 1439 |
+
"loss": 1.6044,
|
| 1440 |
+
"step": 1140
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 6.62,
|
| 1444 |
+
"learning_rate": 4.364808391088155e-05,
|
| 1445 |
+
"loss": 1.5675,
|
| 1446 |
+
"step": 1145
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 6.65,
|
| 1450 |
+
"learning_rate": 3.795676966701853e-05,
|
| 1451 |
+
"loss": 1.5317,
|
| 1452 |
+
"step": 1150
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 6.68,
|
| 1456 |
+
"learning_rate": 3.2517910265912777e-05,
|
| 1457 |
+
"loss": 1.498,
|
| 1458 |
+
"step": 1155
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 6.71,
|
| 1462 |
+
"learning_rate": 2.737631385539826e-05,
|
| 1463 |
+
"loss": 1.5435,
|
| 1464 |
+
"step": 1160
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 6.73,
|
| 1468 |
+
"learning_rate": 2.2574339576441664e-05,
|
| 1469 |
+
"loss": 1.4874,
|
| 1470 |
+
"step": 1165
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 6.76,
|
| 1474 |
+
"learning_rate": 1.815154858654279e-05,
|
| 1475 |
+
"loss": 1.4665,
|
| 1476 |
+
"step": 1170
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 6.79,
|
| 1480 |
+
"learning_rate": 1.4144378134378748e-05,
|
| 1481 |
+
"loss": 1.522,
|
| 1482 |
+
"step": 1175
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 6.82,
|
| 1486 |
+
"learning_rate": 1.0585841370828188e-05,
|
| 1487 |
+
"loss": 1.4874,
|
| 1488 |
+
"step": 1180
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 6.85,
|
| 1492 |
+
"learning_rate": 7.505255369493801e-06,
|
| 1493 |
+
"loss": 1.4275,
|
| 1494 |
+
"step": 1185
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 6.88,
|
| 1498 |
+
"learning_rate": 4.927999597426152e-06,
|
| 1499 |
+
"loss": 1.5011,
|
| 1500 |
+
"step": 1190
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 6.91,
|
| 1504 |
+
"learning_rate": 2.8753068258970103e-06,
|
| 1505 |
+
"loss": 1.5038,
|
| 1506 |
+
"step": 1195
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 6.94,
|
| 1510 |
+
"learning_rate": 1.3640882038038297e-06,
|
| 1511 |
+
"loss": 1.5051,
|
| 1512 |
+
"step": 1200
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 6.97,
|
| 1516 |
+
"learning_rate": 4.067939348447824e-07,
|
| 1517 |
+
"loss": 1.508,
|
| 1518 |
+
"step": 1205
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 6.99,
|
| 1522 |
+
"learning_rate": 1.1310706275549354e-08,
|
| 1523 |
+
"loss": 1.5261,
|
| 1524 |
+
"step": 1210
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 7.0,
|
| 1528 |
+
"eval_loss": 1.4838725328445435,
|
| 1529 |
+
"eval_runtime": 14.6674,
|
| 1530 |
+
"eval_samples_per_second": 15.545,
|
| 1531 |
+
"eval_steps_per_second": 1.977,
|
| 1532 |
+
"step": 1211
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 7.02,
|
| 1536 |
+
"learning_rate": 1.808967142848973e-07,
|
| 1537 |
+
"loss": 1.4801,
|
| 1538 |
+
"step": 1215
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 7.05,
|
| 1542 |
+
"learning_rate": 9.141548212825557e-07,
|
| 1543 |
+
"loss": 1.4927,
|
| 1544 |
+
"step": 1220
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 7.08,
|
| 1548 |
+
"learning_rate": 2.2050440662441005e-06,
|
| 1549 |
+
"loss": 1.4083,
|
| 1550 |
+
"step": 1225
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 7.11,
|
| 1554 |
+
"learning_rate": 4.0429294332848074e-06,
|
| 1555 |
+
"loss": 1.3969,
|
| 1556 |
+
"step": 1230
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 7.14,
|
| 1560 |
+
"learning_rate": 6.412669468442276e-06,
|
| 1561 |
+
"loss": 1.3985,
|
| 1562 |
+
"step": 1235
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 7.17,
|
| 1566 |
+
"learning_rate": 9.294741022836825e-06,
|
| 1567 |
+
"loss": 1.4522,
|
| 1568 |
+
"step": 1240
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 7.2,
|
| 1572 |
+
"learning_rate": 1.266540009450849e-05,
|
| 1573 |
+
"loss": 1.4855,
|
| 1574 |
+
"step": 1245
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 7.23,
|
| 1578 |
+
"learning_rate": 1.649687744383478e-05,
|
| 1579 |
+
"loss": 1.3859,
|
| 1580 |
+
"step": 1250
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 7.25,
|
| 1584 |
+
"learning_rate": 2.075760737094545e-05,
|
| 1585 |
+
"loss": 1.5523,
|
| 1586 |
+
"step": 1255
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 7.28,
|
| 1590 |
+
"learning_rate": 2.5412487770347686e-05,
|
| 1591 |
+
"loss": 1.4851,
|
| 1592 |
+
"step": 1260
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 7.31,
|
| 1596 |
+
"learning_rate": 3.0423169320298854e-05,
|
| 1597 |
+
"loss": 1.4955,
|
| 1598 |
+
"step": 1265
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 7.34,
|
| 1602 |
+
"learning_rate": 3.57483714244302e-05,
|
| 1603 |
+
"loss": 1.4801,
|
| 1604 |
+
"step": 1270
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 7.37,
|
| 1608 |
+
"learning_rate": 4.134422230273594e-05,
|
| 1609 |
+
"loss": 1.3948,
|
| 1610 |
+
"step": 1275
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 7.4,
|
| 1614 |
+
"learning_rate": 4.7164620430075375e-05,
|
| 1615 |
+
"loss": 1.5297,
|
| 1616 |
+
"step": 1280
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 7.43,
|
| 1620 |
+
"learning_rate": 5.316161434447666e-05,
|
| 1621 |
+
"loss": 1.5011,
|
| 1622 |
+
"step": 1285
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 7.46,
|
| 1626 |
+
"learning_rate": 5.92857976961771e-05,
|
| 1627 |
+
"loss": 1.4813,
|
| 1628 |
+
"step": 1290
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 7.49,
|
| 1632 |
+
"learning_rate": 6.548671628278832e-05,
|
| 1633 |
+
"loss": 1.476,
|
| 1634 |
+
"step": 1295
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 7.51,
|
| 1638 |
+
"learning_rate": 7.171328371721156e-05,
|
| 1639 |
+
"loss": 1.3971,
|
| 1640 |
+
"step": 1300
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 7.54,
|
| 1644 |
+
"learning_rate": 7.791420230382278e-05,
|
| 1645 |
+
"loss": 1.4009,
|
| 1646 |
+
"step": 1305
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 7.57,
|
| 1650 |
+
"learning_rate": 8.403838565552322e-05,
|
| 1651 |
+
"loss": 1.4851,
|
| 1652 |
+
"step": 1310
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 7.6,
|
| 1656 |
+
"learning_rate": 9.00353795699245e-05,
|
| 1657 |
+
"loss": 1.4137,
|
| 1658 |
+
"step": 1315
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 7.63,
|
| 1662 |
+
"learning_rate": 9.585577769726395e-05,
|
| 1663 |
+
"loss": 1.4647,
|
| 1664 |
+
"step": 1320
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 7.66,
|
| 1668 |
+
"learning_rate": 0.00010145162857556969,
|
| 1669 |
+
"loss": 1.3558,
|
| 1670 |
+
"step": 1325
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 7.69,
|
| 1674 |
+
"learning_rate": 0.00010677683067970103,
|
| 1675 |
+
"loss": 1.4209,
|
| 1676 |
+
"step": 1330
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 7.72,
|
| 1680 |
+
"learning_rate": 0.00011178751222965221,
|
| 1681 |
+
"loss": 1.5359,
|
| 1682 |
+
"step": 1335
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 7.75,
|
| 1686 |
+
"learning_rate": 0.00011644239262905447,
|
| 1687 |
+
"loss": 1.5243,
|
| 1688 |
+
"step": 1340
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 7.77,
|
| 1692 |
+
"learning_rate": 0.00012070312255616514,
|
| 1693 |
+
"loss": 1.5307,
|
| 1694 |
+
"step": 1345
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 7.8,
|
| 1698 |
+
"learning_rate": 0.00012453459990549145,
|
| 1699 |
+
"loss": 1.3905,
|
| 1700 |
+
"step": 1350
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 7.83,
|
| 1704 |
+
"learning_rate": 0.0001279052589771631,
|
| 1705 |
+
"loss": 1.4955,
|
| 1706 |
+
"step": 1355
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 7.86,
|
| 1710 |
+
"learning_rate": 0.00013078733053155768,
|
| 1711 |
+
"loss": 1.501,
|
| 1712 |
+
"step": 1360
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 7.89,
|
| 1716 |
+
"learning_rate": 0.00013315707056671514,
|
| 1717 |
+
"loss": 1.4948,
|
| 1718 |
+
"step": 1365
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 7.92,
|
| 1722 |
+
"learning_rate": 0.00013499495593375586,
|
| 1723 |
+
"loss": 1.496,
|
| 1724 |
+
"step": 1370
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 7.95,
|
| 1728 |
+
"learning_rate": 0.00013628584517871742,
|
| 1729 |
+
"loss": 1.4929,
|
| 1730 |
+
"step": 1375
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 7.98,
|
| 1734 |
+
"learning_rate": 0.0001370191032857151,
|
| 1735 |
+
"loss": 1.5077,
|
| 1736 |
+
"step": 1380
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 8.0,
|
| 1740 |
+
"eval_loss": 1.5043175220489502,
|
| 1741 |
+
"eval_runtime": 14.7576,
|
| 1742 |
+
"eval_samples_per_second": 15.45,
|
| 1743 |
+
"eval_steps_per_second": 1.965,
|
| 1744 |
+
"step": 1384
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 8.01,
|
| 1748 |
+
"learning_rate": 0.00013718868929372445,
|
| 1749 |
+
"loss": 1.5034,
|
| 1750 |
+
"step": 1385
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 8.03,
|
| 1754 |
+
"learning_rate": 0.00013679320606515522,
|
| 1755 |
+
"loss": 1.4037,
|
| 1756 |
+
"step": 1390
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 8.06,
|
| 1760 |
+
"learning_rate": 0.00013583591179619614,
|
| 1761 |
+
"loss": 1.4957,
|
| 1762 |
+
"step": 1395
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 8.09,
|
| 1766 |
+
"learning_rate": 0.000134324693174103,
|
| 1767 |
+
"loss": 1.5076,
|
| 1768 |
+
"step": 1400
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 8.12,
|
| 1772 |
+
"learning_rate": 0.0001322720004025738,
|
| 1773 |
+
"loss": 1.503,
|
| 1774 |
+
"step": 1405
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 8.15,
|
| 1778 |
+
"learning_rate": 0.00012969474463050637,
|
| 1779 |
+
"loss": 1.4043,
|
| 1780 |
+
"step": 1410
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 8.18,
|
| 1784 |
+
"learning_rate": 0.00012661415862917202,
|
| 1785 |
+
"loss": 1.4017,
|
| 1786 |
+
"step": 1415
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 8.21,
|
| 1790 |
+
"learning_rate": 0.00012305562186562148,
|
| 1791 |
+
"loss": 1.4622,
|
| 1792 |
+
"step": 1420
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 8.24,
|
| 1796 |
+
"learning_rate": 0.00011904845141345745,
|
| 1797 |
+
"loss": 1.4555,
|
| 1798 |
+
"step": 1425
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 8.27,
|
| 1802 |
+
"learning_rate": 0.00011462566042355861,
|
| 1803 |
+
"loss": 1.474,
|
| 1804 |
+
"step": 1430
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 8.29,
|
| 1808 |
+
"learning_rate": 0.00010982368614460204,
|
| 1809 |
+
"loss": 1.4798,
|
| 1810 |
+
"step": 1435
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 8.32,
|
| 1814 |
+
"learning_rate": 0.00010468208973408754,
|
| 1815 |
+
"loss": 1.4409,
|
| 1816 |
+
"step": 1440
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 8.35,
|
| 1820 |
+
"learning_rate": 9.924323033298181e-05,
|
| 1821 |
+
"loss": 1.4283,
|
| 1822 |
+
"step": 1445
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 8.38,
|
| 1826 |
+
"learning_rate": 9.35519160891188e-05,
|
| 1827 |
+
"loss": 1.4248,
|
| 1828 |
+
"step": 1450
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 8.41,
|
| 1832 |
+
"learning_rate": 8.76550350044131e-05,
|
| 1833 |
+
"loss": 1.4627,
|
| 1834 |
+
"step": 1455
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 8.44,
|
| 1838 |
+
"learning_rate": 8.160116864715327e-05,
|
| 1839 |
+
"loss": 1.5012,
|
| 1840 |
+
"step": 1460
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 8.47,
|
| 1844 |
+
"learning_rate": 7.544019191181593e-05,
|
| 1845 |
+
"loss": 1.3866,
|
| 1846 |
+
"step": 1465
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 8.5,
|
| 1850 |
+
"learning_rate": 6.92228621237895e-05,
|
| 1851 |
+
"loss": 1.4909,
|
| 1852 |
+
"step": 1470
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 8.53,
|
| 1856 |
+
"learning_rate": 6.300040087417043e-05,
|
| 1857 |
+
"loss": 1.385,
|
| 1858 |
+
"step": 1475
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 8.55,
|
| 1862 |
+
"learning_rate": 5.682407202970122e-05,
|
| 1863 |
+
"loss": 1.4766,
|
| 1864 |
+
"step": 1480
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 8.58,
|
| 1868 |
+
"learning_rate": 5.0744759394415834e-05,
|
| 1869 |
+
"loss": 1.4607,
|
| 1870 |
+
"step": 1485
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 8.61,
|
| 1874 |
+
"learning_rate": 4.481254750244299e-05,
|
| 1875 |
+
"loss": 1.4235,
|
| 1876 |
+
"step": 1490
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 8.64,
|
| 1880 |
+
"learning_rate": 3.907630899560609e-05,
|
| 1881 |
+
"loss": 1.4085,
|
| 1882 |
+
"step": 1495
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 8.67,
|
| 1886 |
+
"learning_rate": 3.358330198522491e-05,
|
| 1887 |
+
"loss": 1.4548,
|
| 1888 |
+
"step": 1500
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 8.7,
|
| 1892 |
+
"learning_rate": 2.8378780715254406e-05,
|
| 1893 |
+
"loss": 1.4757,
|
| 1894 |
+
"step": 1505
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 8.73,
|
| 1898 |
+
"learning_rate": 2.3505622734326714e-05,
|
| 1899 |
+
"loss": 1.3891,
|
| 1900 |
+
"step": 1510
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 8.76,
|
| 1904 |
+
"learning_rate": 1.900397564824009e-05,
|
| 1905 |
+
"loss": 1.3819,
|
| 1906 |
+
"step": 1515
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 8.79,
|
| 1910 |
+
"learning_rate": 1.4910926363137506e-05,
|
| 1911 |
+
"loss": 1.4202,
|
| 1912 |
+
"step": 1520
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 8.82,
|
| 1916 |
+
"learning_rate": 1.1260195544315306e-05,
|
| 1917 |
+
"loss": 1.3403,
|
| 1918 |
+
"step": 1525
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 8.84,
|
| 1922 |
+
"learning_rate": 8.081859807874605e-06,
|
| 1923 |
+
"loss": 1.4776,
|
| 1924 |
+
"step": 1530
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 8.87,
|
| 1928 |
+
"learning_rate": 5.402103933939604e-06,
|
| 1929 |
+
"loss": 1.4531,
|
| 1930 |
+
"step": 1535
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 8.9,
|
| 1934 |
+
"learning_rate": 3.2430051428437045e-06,
|
| 1935 |
+
"loss": 1.373,
|
| 1936 |
+
"step": 1540
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 8.93,
|
| 1940 |
+
"learning_rate": 1.6223512115240727e-06,
|
| 1941 |
+
"loss": 1.3284,
|
| 1942 |
+
"step": 1545
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 8.96,
|
| 1946 |
+
"learning_rate": 5.534939285798083e-07,
|
| 1947 |
+
"loss": 1.3615,
|
| 1948 |
+
"step": 1550
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 8.99,
|
| 1952 |
+
"learning_rate": 4.523909530404726e-08,
|
| 1953 |
+
"loss": 1.3743,
|
| 1954 |
+
"step": 1555
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 9.0,
|
| 1958 |
+
"eval_loss": 1.464684247970581,
|
| 1959 |
+
"eval_runtime": 15.3121,
|
| 1960 |
+
"eval_samples_per_second": 14.89,
|
| 1961 |
+
"eval_steps_per_second": 1.894,
|
| 1962 |
+
"step": 1557
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 9.02,
|
| 1966 |
+
"learning_rate": 1.0177397892099493e-07,
|
| 1967 |
+
"loss": 1.4264,
|
| 1968 |
+
"step": 1560
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 9.05,
|
| 1972 |
+
"learning_rate": 7.226328157040877e-07,
|
| 1973 |
+
"loss": 1.3427,
|
| 1974 |
+
"step": 1565
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 9.08,
|
| 1978 |
+
"learning_rate": 1.9027006481794198e-06,
|
| 1979 |
+
"loss": 1.4135,
|
| 1980 |
+
"step": 1570
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 9.1,
|
| 1984 |
+
"learning_rate": 3.6322554648015487e-06,
|
| 1985 |
+
"loss": 1.3408,
|
| 1986 |
+
"step": 1575
|
| 1987 |
+
},
|
| 1988 |
+
{
|
| 1989 |
+
"epoch": 9.13,
|
| 1990 |
+
"learning_rate": 5.897048294932268e-06,
|
| 1991 |
+
"loss": 1.355,
|
| 1992 |
+
"step": 1580
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 9.16,
|
| 1996 |
+
"learning_rate": 8.678420599259977e-06,
|
| 1997 |
+
"loss": 1.3903,
|
| 1998 |
+
"step": 1585
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 9.19,
|
| 2002 |
+
"learning_rate": 1.1953457988534113e-05,
|
| 2003 |
+
"loss": 1.3435,
|
| 2004 |
+
"step": 1590
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 9.22,
|
| 2008 |
+
"learning_rate": 1.5695179004202806e-05,
|
| 2009 |
+
"loss": 1.4126,
|
| 2010 |
+
"step": 1595
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 9.25,
|
| 2014 |
+
"learning_rate": 1.9872757405677322e-05,
|
| 2015 |
+
"loss": 1.3711,
|
| 2016 |
+
"step": 1600
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 9.28,
|
| 2020 |
+
"learning_rate": 2.445177613290951e-05,
|
| 2021 |
+
"loss": 1.4319,
|
| 2022 |
+
"step": 1605
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 9.31,
|
| 2026 |
+
"learning_rate": 2.9394510852001993e-05,
|
| 2027 |
+
"loss": 1.3731,
|
| 2028 |
+
"step": 1610
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 9.34,
|
| 2032 |
+
"learning_rate": 3.46602407478603e-05,
|
| 2033 |
+
"loss": 1.3164,
|
| 2034 |
+
"step": 1615
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 9.36,
|
| 2038 |
+
"learning_rate": 4.020558400341007e-05,
|
| 2039 |
+
"loss": 1.3927,
|
| 2040 |
+
"step": 1620
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 9.39,
|
| 2044 |
+
"learning_rate": 4.598485520153301e-05,
|
| 2045 |
+
"loss": 1.3304,
|
| 2046 |
+
"step": 1625
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 9.42,
|
| 2050 |
+
"learning_rate": 5.195044170525524e-05,
|
| 2051 |
+
"loss": 1.3834,
|
| 2052 |
+
"step": 1630
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 9.45,
|
| 2056 |
+
"learning_rate": 5.80531959153711e-05,
|
| 2057 |
+
"loss": 1.4126,
|
| 2058 |
+
"step": 1635
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 9.48,
|
| 2062 |
+
"learning_rate": 6.424284017388749e-05,
|
| 2063 |
+
"loss": 1.3473,
|
| 2064 |
+
"step": 1640
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 9.51,
|
| 2068 |
+
"learning_rate": 7.046838097747464e-05,
|
| 2069 |
+
"loss": 1.299,
|
| 2070 |
+
"step": 1645
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 9.54,
|
| 2074 |
+
"learning_rate": 7.66785290884264e-05,
|
| 2075 |
+
"loss": 1.3673,
|
| 2076 |
+
"step": 1650
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 9.57,
|
| 2080 |
+
"learning_rate": 8.282212208203013e-05,
|
| 2081 |
+
"loss": 1.3913,
|
| 2082 |
+
"step": 1655
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 9.6,
|
| 2086 |
+
"learning_rate": 8.884854584918912e-05,
|
| 2087 |
+
"loss": 1.3892,
|
| 2088 |
+
"step": 1660
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 9.62,
|
| 2092 |
+
"learning_rate": 9.470815158173245e-05,
|
| 2093 |
+
"loss": 1.3557,
|
| 2094 |
+
"step": 1665
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 9.65,
|
| 2098 |
+
"learning_rate": 0.00010035266480507833,
|
| 2099 |
+
"loss": 1.396,
|
| 2100 |
+
"step": 1670
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 9.68,
|
| 2104 |
+
"learning_rate": 0.00010573558308842004,
|
| 2105 |
+
"loss": 1.3608,
|
| 2106 |
+
"step": 1675
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 9.71,
|
| 2110 |
+
"learning_rate": 0.00011081255915589876,
|
| 2111 |
+
"loss": 1.3399,
|
| 2112 |
+
"step": 1680
|
| 2113 |
+
},
|
| 2114 |
+
{
|
| 2115 |
+
"epoch": 9.74,
|
| 2116 |
+
"learning_rate": 0.00011554176624248741,
|
| 2117 |
+
"loss": 1.3982,
|
| 2118 |
+
"step": 1685
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 9.77,
|
| 2122 |
+
"learning_rate": 0.00011988424268460073,
|
| 2123 |
+
"loss": 1.4392,
|
| 2124 |
+
"step": 1690
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 9.8,
|
| 2128 |
+
"learning_rate": 0.000123804212906509,
|
| 2129 |
+
"loss": 1.4083,
|
| 2130 |
+
"step": 1695
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 9.83,
|
| 2134 |
+
"learning_rate": 0.00012726938215810852,
|
| 2135 |
+
"loss": 1.434,
|
| 2136 |
+
"step": 1700
|
| 2137 |
+
},
|
| 2138 |
+
{
|
| 2139 |
+
"epoch": 9.86,
|
| 2140 |
+
"learning_rate": 0.00013025120257584047,
|
| 2141 |
+
"loss": 1.3252,
|
| 2142 |
+
"step": 1705
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 9.88,
|
| 2146 |
+
"learning_rate": 0.00013272510837481527,
|
| 2147 |
+
"loss": 1.3899,
|
| 2148 |
+
"step": 1710
|
| 2149 |
+
},
|
| 2150 |
+
{
|
| 2151 |
+
"epoch": 9.91,
|
| 2152 |
+
"learning_rate": 0.00013467071823450324,
|
| 2153 |
+
"loss": 1.3883,
|
| 2154 |
+
"step": 1715
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 9.94,
|
| 2158 |
+
"learning_rate": 0.00013607200321063664,
|
| 2159 |
+
"loss": 1.4395,
|
| 2160 |
+
"step": 1720
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 9.97,
|
| 2164 |
+
"learning_rate": 0.00013691741878997505,
|
| 2165 |
+
"loss": 1.4208,
|
| 2166 |
+
"step": 1725
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 10.0,
|
| 2170 |
+
"learning_rate": 0.0001372,
|
| 2171 |
+
"loss": 1.4499,
|
| 2172 |
+
"step": 1730
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 10.0,
|
| 2176 |
+
"eval_loss": 1.4852185249328613,
|
| 2177 |
+
"eval_runtime": 15.5041,
|
| 2178 |
+
"eval_samples_per_second": 14.706,
|
| 2179 |
+
"eval_steps_per_second": 1.87,
|
| 2180 |
+
"step": 1730
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 10.03,
|
| 2184 |
+
"learning_rate": 0.00013691741878997505,
|
| 2185 |
+
"loss": 1.3342,
|
| 2186 |
+
"step": 1735
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 10.06,
|
| 2190 |
+
"learning_rate": 0.0001360720032106367,
|
| 2191 |
+
"loss": 1.3537,
|
| 2192 |
+
"step": 1740
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 10.09,
|
| 2196 |
+
"learning_rate": 0.00013467071823450327,
|
| 2197 |
+
"loss": 1.376,
|
| 2198 |
+
"step": 1745
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 10.12,
|
| 2202 |
+
"learning_rate": 0.00013272510837481532,
|
| 2203 |
+
"loss": 1.3746,
|
| 2204 |
+
"step": 1750
|
| 2205 |
+
},
|
| 2206 |
+
{
|
| 2207 |
+
"epoch": 10.14,
|
| 2208 |
+
"learning_rate": 0.00013025120257584055,
|
| 2209 |
+
"loss": 1.3632,
|
| 2210 |
+
"step": 1755
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 10.17,
|
| 2214 |
+
"learning_rate": 0.00012726938215810863,
|
| 2215 |
+
"loss": 1.4038,
|
| 2216 |
+
"step": 1760
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 10.2,
|
| 2220 |
+
"learning_rate": 0.00012380421290650925,
|
| 2221 |
+
"loss": 1.3788,
|
| 2222 |
+
"step": 1765
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 10.23,
|
| 2226 |
+
"learning_rate": 0.00011988424268460084,
|
| 2227 |
+
"loss": 1.3642,
|
| 2228 |
+
"step": 1770
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 10.26,
|
| 2232 |
+
"learning_rate": 0.00011554176624248753,
|
| 2233 |
+
"loss": 1.3539,
|
| 2234 |
+
"step": 1775
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 10.29,
|
| 2238 |
+
"learning_rate": 0.00011081255915589869,
|
| 2239 |
+
"loss": 1.371,
|
| 2240 |
+
"step": 1780
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 10.32,
|
| 2244 |
+
"learning_rate": 0.00010573558308842038,
|
| 2245 |
+
"loss": 1.3995,
|
| 2246 |
+
"step": 1785
|
| 2247 |
+
},
|
| 2248 |
+
{
|
| 2249 |
+
"epoch": 10.35,
|
| 2250 |
+
"learning_rate": 0.00010035266480507848,
|
| 2251 |
+
"loss": 1.4698,
|
| 2252 |
+
"step": 1790
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 10.38,
|
| 2256 |
+
"learning_rate": 9.47081515817326e-05,
|
| 2257 |
+
"loss": 1.3803,
|
| 2258 |
+
"step": 1795
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 10.4,
|
| 2262 |
+
"learning_rate": 8.884854584918904e-05,
|
| 2263 |
+
"loss": 1.4079,
|
| 2264 |
+
"step": 1800
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 10.43,
|
| 2268 |
+
"learning_rate": 8.282212208203053e-05,
|
| 2269 |
+
"loss": 1.4,
|
| 2270 |
+
"step": 1805
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 10.46,
|
| 2274 |
+
"learning_rate": 7.667852908842657e-05,
|
| 2275 |
+
"loss": 1.3182,
|
| 2276 |
+
"step": 1810
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 10.49,
|
| 2280 |
+
"learning_rate": 7.04683809774748e-05,
|
| 2281 |
+
"loss": 1.3298,
|
| 2282 |
+
"step": 1815
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 10.52,
|
| 2286 |
+
"learning_rate": 6.424284017388742e-05,
|
| 2287 |
+
"loss": 1.348,
|
| 2288 |
+
"step": 1820
|
| 2289 |
+
},
|
| 2290 |
+
{
|
| 2291 |
+
"epoch": 10.55,
|
| 2292 |
+
"learning_rate": 5.8053195915371506e-05,
|
| 2293 |
+
"loss": 1.3953,
|
| 2294 |
+
"step": 1825
|
| 2295 |
+
},
|
| 2296 |
+
{
|
| 2297 |
+
"epoch": 10.58,
|
| 2298 |
+
"learning_rate": 5.1950441705255396e-05,
|
| 2299 |
+
"loss": 1.3451,
|
| 2300 |
+
"step": 1830
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 10.61,
|
| 2304 |
+
"learning_rate": 4.5984855201533164e-05,
|
| 2305 |
+
"loss": 1.3076,
|
| 2306 |
+
"step": 1835
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 10.64,
|
| 2310 |
+
"learning_rate": 4.0205584003410225e-05,
|
| 2311 |
+
"loss": 1.4287,
|
| 2312 |
+
"step": 1840
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 10.66,
|
| 2316 |
+
"learning_rate": 3.4660240747860236e-05,
|
| 2317 |
+
"loss": 1.3216,
|
| 2318 |
+
"step": 1845
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 10.69,
|
| 2322 |
+
"learning_rate": 2.939451085200233e-05,
|
| 2323 |
+
"loss": 1.341,
|
| 2324 |
+
"step": 1850
|
| 2325 |
+
},
|
| 2326 |
+
{
|
| 2327 |
+
"epoch": 10.72,
|
| 2328 |
+
"learning_rate": 2.4451776132909644e-05,
|
| 2329 |
+
"loss": 1.3848,
|
| 2330 |
+
"step": 1855
|
| 2331 |
+
},
|
| 2332 |
+
{
|
| 2333 |
+
"epoch": 10.75,
|
| 2334 |
+
"learning_rate": 1.9872757405677438e-05,
|
| 2335 |
+
"loss": 1.328,
|
| 2336 |
+
"step": 1860
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 10.78,
|
| 2340 |
+
"learning_rate": 1.5695179004202758e-05,
|
| 2341 |
+
"loss": 1.3147,
|
| 2342 |
+
"step": 1865
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 10.81,
|
| 2346 |
+
"learning_rate": 1.1953457988534349e-05,
|
| 2347 |
+
"loss": 1.2886,
|
| 2348 |
+
"step": 1870
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 10.84,
|
| 2352 |
+
"learning_rate": 8.678420599260062e-06,
|
| 2353 |
+
"loss": 1.2943,
|
| 2354 |
+
"step": 1875
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 10.87,
|
| 2358 |
+
"learning_rate": 5.8970482949323365e-06,
|
| 2359 |
+
"loss": 1.3852,
|
| 2360 |
+
"step": 1880
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 10.9,
|
| 2364 |
+
"learning_rate": 3.6322554648015263e-06,
|
| 2365 |
+
"loss": 1.3121,
|
| 2366 |
+
"step": 1885
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 10.92,
|
| 2370 |
+
"learning_rate": 1.902700648179519e-06,
|
| 2371 |
+
"loss": 1.37,
|
| 2372 |
+
"step": 1890
|
| 2373 |
+
},
|
| 2374 |
+
{
|
| 2375 |
+
"epoch": 10.95,
|
| 2376 |
+
"learning_rate": 7.226328157041105e-07,
|
| 2377 |
+
"loss": 1.3568,
|
| 2378 |
+
"step": 1895
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 10.98,
|
| 2382 |
+
"learning_rate": 1.0177397892100256e-07,
|
| 2383 |
+
"loss": 1.2291,
|
| 2384 |
+
"step": 1900
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 11.0,
|
| 2388 |
+
"eval_loss": 1.4596115350723267,
|
| 2389 |
+
"eval_runtime": 15.3008,
|
| 2390 |
+
"eval_samples_per_second": 14.901,
|
| 2391 |
+
"eval_steps_per_second": 1.895,
|
| 2392 |
+
"step": 1903
|
| 2393 |
}
|
| 2394 |
],
|
| 2395 |
+
"max_steps": 2249,
|
| 2396 |
+
"num_train_epochs": 13,
|
| 2397 |
+
"total_flos": 1979287142400000.0,
|
| 2398 |
"trial_name": null,
|
| 2399 |
"trial_params": null
|
| 2400 |
}
|
training_args.bin
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 sha256:8dadfe9fb117ded7156087ae0bf3af1eed8739c4ee1dba2436f9ab382c9e3d92
|
| 3 |
+
size 2863
|