End of training
Browse files- README.md +19 -17
- adapter_1/adapter_config.json +32 -0
- adapter_1/adapter_model.safetensors +3 -0
- adapter_config.json +6 -3
- adapter_model.safetensors +1 -1
- trainer_state.json +1636 -0
- training_args.bin +2 -2
README.md
CHANGED
@@ -1,8 +1,9 @@
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
-
|
4 |
tags:
|
5 |
- generated_from_trainer
|
|
|
6 |
model-index:
|
7 |
- name: Whisper-large-v3-Arabic-phoneme
|
8 |
results: []
|
@@ -13,9 +14,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
13 |
|
14 |
# Whisper-large-v3-Arabic-phoneme
|
15 |
|
16 |
-
This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on
|
17 |
It achieves the following results on the evaluation set:
|
18 |
-
- Loss: 0.
|
19 |
|
20 |
## Model description
|
21 |
|
@@ -48,21 +49,22 @@ The following hyperparameters were used during training:
|
|
48 |
|
49 |
| Training Loss | Epoch | Step | Validation Loss |
|
50 |
|:-------------:|:-----:|:----:|:---------------:|
|
51 |
-
| 0.
|
52 |
-
| 0.
|
53 |
-
| 0.
|
54 |
-
| 0.
|
55 |
-
| 0.
|
56 |
-
| 0.
|
57 |
-
| 0.
|
58 |
-
| 0.
|
59 |
-
| 0.
|
60 |
-
| 0.
|
61 |
|
62 |
|
63 |
### Framework versions
|
64 |
|
65 |
-
-
|
66 |
-
-
|
67 |
-
-
|
68 |
-
-
|
|
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
+
library_name: peft
|
4 |
tags:
|
5 |
- generated_from_trainer
|
6 |
+
base_model: openai/whisper-large-v3
|
7 |
model-index:
|
8 |
- name: Whisper-large-v3-Arabic-phoneme
|
9 |
results: []
|
|
|
14 |
|
15 |
# Whisper-large-v3-Arabic-phoneme
|
16 |
|
17 |
+
This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on an unknown dataset.
|
18 |
It achieves the following results on the evaluation set:
|
19 |
+
- Loss: 0.2331
|
20 |
|
21 |
## Model description
|
22 |
|
|
|
49 |
|
50 |
| Training Loss | Epoch | Step | Validation Loss |
|
51 |
|:-------------:|:-----:|:----:|:---------------:|
|
52 |
+
| 0.0608 | 1.0 | 546 | 0.2074 |
|
53 |
+
| 0.0508 | 2.0 | 1092 | 0.2211 |
|
54 |
+
| 0.0287 | 3.0 | 1638 | 0.1681 |
|
55 |
+
| 0.0148 | 4.0 | 2184 | 0.1938 |
|
56 |
+
| 0.0263 | 5.0 | 2730 | 0.1846 |
|
57 |
+
| 0.0168 | 6.0 | 3276 | 0.1899 |
|
58 |
+
| 0.0086 | 7.0 | 3822 | 0.1975 |
|
59 |
+
| 0.0102 | 8.0 | 4368 | 0.2170 |
|
60 |
+
| 0.0023 | 9.0 | 4914 | 0.2294 |
|
61 |
+
| 0.0024 | 10.0 | 5460 | 0.2331 |
|
62 |
|
63 |
|
64 |
### Framework versions
|
65 |
|
66 |
+
- PEFT 0.10.0
|
67 |
+
- Transformers 4.38.2
|
68 |
+
- Pytorch 2.2.1+cu121
|
69 |
+
- Datasets 2.18.0
|
70 |
+
- Tokenizers 0.15.2
|
adapter_1/adapter_config.json
ADDED
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"alpha_pattern": {},
|
3 |
+
"auto_mapping": {
|
4 |
+
"base_model_class": "WhisperForConditionalGeneration",
|
5 |
+
"parent_library": "transformers.models.whisper.modeling_whisper"
|
6 |
+
},
|
7 |
+
"base_model_name_or_path": "openai/whisper-large-v3",
|
8 |
+
"bias": "none",
|
9 |
+
"fan_in_fan_out": false,
|
10 |
+
"inference_mode": true,
|
11 |
+
"init_lora_weights": true,
|
12 |
+
"layer_replication": null,
|
13 |
+
"layers_pattern": null,
|
14 |
+
"layers_to_transform": null,
|
15 |
+
"loftq_config": {},
|
16 |
+
"lora_alpha": 64,
|
17 |
+
"lora_dropout": 0.05,
|
18 |
+
"megatron_config": null,
|
19 |
+
"megatron_core": "megatron.core",
|
20 |
+
"modules_to_save": null,
|
21 |
+
"peft_type": "LORA",
|
22 |
+
"r": 32,
|
23 |
+
"rank_pattern": {},
|
24 |
+
"revision": null,
|
25 |
+
"target_modules": [
|
26 |
+
"v_proj",
|
27 |
+
"q_proj"
|
28 |
+
],
|
29 |
+
"task_type": null,
|
30 |
+
"use_dora": false,
|
31 |
+
"use_rslora": false
|
32 |
+
}
|
adapter_1/adapter_model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:08026702b7a1be7021b6a8e05de062904a5b406498ebe8c9213965d21f3676a2
|
3 |
+
size 62969640
|
adapter_config.json
CHANGED
@@ -9,6 +9,7 @@
|
|
9 |
"fan_in_fan_out": false,
|
10 |
"inference_mode": true,
|
11 |
"init_lora_weights": true,
|
|
|
12 |
"layers_pattern": null,
|
13 |
"layers_to_transform": null,
|
14 |
"loftq_config": {},
|
@@ -22,8 +23,10 @@
|
|
22 |
"rank_pattern": {},
|
23 |
"revision": null,
|
24 |
"target_modules": [
|
25 |
-
"
|
26 |
-
"
|
27 |
],
|
28 |
-
"task_type": null
|
|
|
|
|
29 |
}
|
|
|
9 |
"fan_in_fan_out": false,
|
10 |
"inference_mode": true,
|
11 |
"init_lora_weights": true,
|
12 |
+
"layer_replication": null,
|
13 |
"layers_pattern": null,
|
14 |
"layers_to_transform": null,
|
15 |
"loftq_config": {},
|
|
|
23 |
"rank_pattern": {},
|
24 |
"revision": null,
|
25 |
"target_modules": [
|
26 |
+
"v_proj",
|
27 |
+
"q_proj"
|
28 |
],
|
29 |
+
"task_type": null,
|
30 |
+
"use_dora": false,
|
31 |
+
"use_rslora": false
|
32 |
}
|
adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 62969640
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f5ec4edfc2c7ee47d1229478f8a8d2c11bff55bd08585acde03693a1a00498f2
|
3 |
size 62969640
|
trainer_state.json
ADDED
@@ -0,0 +1,1636 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 10.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 5460,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.05,
|
13 |
+
"grad_norm": 1.0899150371551514,
|
14 |
+
"learning_rate": 0.0005,
|
15 |
+
"loss": 2.9446,
|
16 |
+
"step": 25
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.09,
|
20 |
+
"grad_norm": 0.7281716465950012,
|
21 |
+
"learning_rate": 0.001,
|
22 |
+
"loss": 0.7356,
|
23 |
+
"step": 50
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.14,
|
27 |
+
"grad_norm": 0.8861620426177979,
|
28 |
+
"learning_rate": 0.0009963031423290203,
|
29 |
+
"loss": 0.5765,
|
30 |
+
"step": 75
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.18,
|
34 |
+
"grad_norm": 2.8832926750183105,
|
35 |
+
"learning_rate": 0.0009916820702402958,
|
36 |
+
"loss": 0.3944,
|
37 |
+
"step": 100
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.23,
|
41 |
+
"grad_norm": 1.077749252319336,
|
42 |
+
"learning_rate": 0.000987060998151571,
|
43 |
+
"loss": 0.1654,
|
44 |
+
"step": 125
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.27,
|
48 |
+
"grad_norm": 0.5631522536277771,
|
49 |
+
"learning_rate": 0.0009824399260628465,
|
50 |
+
"loss": 0.1044,
|
51 |
+
"step": 150
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.32,
|
55 |
+
"grad_norm": 0.25788772106170654,
|
56 |
+
"learning_rate": 0.000977818853974122,
|
57 |
+
"loss": 0.0853,
|
58 |
+
"step": 175
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.37,
|
62 |
+
"grad_norm": 0.7229248285293579,
|
63 |
+
"learning_rate": 0.0009731977818853974,
|
64 |
+
"loss": 0.088,
|
65 |
+
"step": 200
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.41,
|
69 |
+
"grad_norm": 0.39239415526390076,
|
70 |
+
"learning_rate": 0.0009685767097966729,
|
71 |
+
"loss": 0.0781,
|
72 |
+
"step": 225
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.46,
|
76 |
+
"grad_norm": 0.23850314319133759,
|
77 |
+
"learning_rate": 0.0009639556377079483,
|
78 |
+
"loss": 0.0729,
|
79 |
+
"step": 250
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.5,
|
83 |
+
"grad_norm": 0.17475274205207825,
|
84 |
+
"learning_rate": 0.0009593345656192237,
|
85 |
+
"loss": 0.0671,
|
86 |
+
"step": 275
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.55,
|
90 |
+
"grad_norm": 0.7980681657791138,
|
91 |
+
"learning_rate": 0.0009547134935304991,
|
92 |
+
"loss": 0.0836,
|
93 |
+
"step": 300
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.6,
|
97 |
+
"grad_norm": 0.20785190165042877,
|
98 |
+
"learning_rate": 0.0009500924214417745,
|
99 |
+
"loss": 0.0717,
|
100 |
+
"step": 325
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.64,
|
104 |
+
"grad_norm": 0.22275149822235107,
|
105 |
+
"learning_rate": 0.0009454713493530499,
|
106 |
+
"loss": 0.0469,
|
107 |
+
"step": 350
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.69,
|
111 |
+
"grad_norm": 0.39296770095825195,
|
112 |
+
"learning_rate": 0.0009408502772643253,
|
113 |
+
"loss": 0.0815,
|
114 |
+
"step": 375
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.73,
|
118 |
+
"grad_norm": 0.3089096248149872,
|
119 |
+
"learning_rate": 0.0009362292051756007,
|
120 |
+
"loss": 0.0668,
|
121 |
+
"step": 400
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.78,
|
125 |
+
"grad_norm": 0.343288779258728,
|
126 |
+
"learning_rate": 0.0009316081330868762,
|
127 |
+
"loss": 0.0481,
|
128 |
+
"step": 425
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.82,
|
132 |
+
"grad_norm": 0.1959904432296753,
|
133 |
+
"learning_rate": 0.0009269870609981515,
|
134 |
+
"loss": 0.0554,
|
135 |
+
"step": 450
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.87,
|
139 |
+
"grad_norm": 0.5835228562355042,
|
140 |
+
"learning_rate": 0.000922365988909427,
|
141 |
+
"loss": 0.079,
|
142 |
+
"step": 475
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.92,
|
146 |
+
"grad_norm": 0.1368131786584854,
|
147 |
+
"learning_rate": 0.0009177449168207024,
|
148 |
+
"loss": 0.0575,
|
149 |
+
"step": 500
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.96,
|
153 |
+
"grad_norm": 0.24407006800174713,
|
154 |
+
"learning_rate": 0.0009131238447319779,
|
155 |
+
"loss": 0.0608,
|
156 |
+
"step": 525
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 1.0,
|
160 |
+
"eval_loss": 0.2074204683303833,
|
161 |
+
"eval_runtime": 457.9814,
|
162 |
+
"eval_samples_per_second": 1.775,
|
163 |
+
"eval_steps_per_second": 0.297,
|
164 |
+
"step": 546
|
165 |
+
},
|
166 |
+
{
|
167 |
+
"epoch": 1.01,
|
168 |
+
"grad_norm": 0.12714482843875885,
|
169 |
+
"learning_rate": 0.0009085027726432532,
|
170 |
+
"loss": 0.0468,
|
171 |
+
"step": 550
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 1.05,
|
175 |
+
"grad_norm": 0.3958694636821747,
|
176 |
+
"learning_rate": 0.0009038817005545287,
|
177 |
+
"loss": 0.0486,
|
178 |
+
"step": 575
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 1.1,
|
182 |
+
"grad_norm": 0.0621350072324276,
|
183 |
+
"learning_rate": 0.0008992606284658041,
|
184 |
+
"loss": 0.0531,
|
185 |
+
"step": 600
|
186 |
+
},
|
187 |
+
{
|
188 |
+
"epoch": 1.14,
|
189 |
+
"grad_norm": 0.019190674647688866,
|
190 |
+
"learning_rate": 0.0008946395563770795,
|
191 |
+
"loss": 0.0404,
|
192 |
+
"step": 625
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 1.19,
|
196 |
+
"grad_norm": 0.10872188955545425,
|
197 |
+
"learning_rate": 0.000890018484288355,
|
198 |
+
"loss": 0.0516,
|
199 |
+
"step": 650
|
200 |
+
},
|
201 |
+
{
|
202 |
+
"epoch": 1.24,
|
203 |
+
"grad_norm": 0.21620219945907593,
|
204 |
+
"learning_rate": 0.0008853974121996303,
|
205 |
+
"loss": 0.0352,
|
206 |
+
"step": 675
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 1.28,
|
210 |
+
"grad_norm": 0.09226030856370926,
|
211 |
+
"learning_rate": 0.0008807763401109058,
|
212 |
+
"loss": 0.046,
|
213 |
+
"step": 700
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 1.33,
|
217 |
+
"grad_norm": 0.28338831663131714,
|
218 |
+
"learning_rate": 0.0008761552680221812,
|
219 |
+
"loss": 0.0541,
|
220 |
+
"step": 725
|
221 |
+
},
|
222 |
+
{
|
223 |
+
"epoch": 1.37,
|
224 |
+
"grad_norm": 0.29886457324028015,
|
225 |
+
"learning_rate": 0.0008715341959334566,
|
226 |
+
"loss": 0.0469,
|
227 |
+
"step": 750
|
228 |
+
},
|
229 |
+
{
|
230 |
+
"epoch": 1.42,
|
231 |
+
"grad_norm": 0.18679702281951904,
|
232 |
+
"learning_rate": 0.000866913123844732,
|
233 |
+
"loss": 0.0479,
|
234 |
+
"step": 775
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 1.47,
|
238 |
+
"grad_norm": 0.37105533480644226,
|
239 |
+
"learning_rate": 0.0008622920517560074,
|
240 |
+
"loss": 0.0482,
|
241 |
+
"step": 800
|
242 |
+
},
|
243 |
+
{
|
244 |
+
"epoch": 1.51,
|
245 |
+
"grad_norm": 0.2705371081829071,
|
246 |
+
"learning_rate": 0.0008576709796672828,
|
247 |
+
"loss": 0.0339,
|
248 |
+
"step": 825
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 1.56,
|
252 |
+
"grad_norm": 0.23306304216384888,
|
253 |
+
"learning_rate": 0.0008530499075785582,
|
254 |
+
"loss": 0.0333,
|
255 |
+
"step": 850
|
256 |
+
},
|
257 |
+
{
|
258 |
+
"epoch": 1.6,
|
259 |
+
"grad_norm": 0.32292288541793823,
|
260 |
+
"learning_rate": 0.0008484288354898336,
|
261 |
+
"loss": 0.0631,
|
262 |
+
"step": 875
|
263 |
+
},
|
264 |
+
{
|
265 |
+
"epoch": 1.65,
|
266 |
+
"grad_norm": 1.583330512046814,
|
267 |
+
"learning_rate": 0.0008438077634011091,
|
268 |
+
"loss": 0.0554,
|
269 |
+
"step": 900
|
270 |
+
},
|
271 |
+
{
|
272 |
+
"epoch": 1.69,
|
273 |
+
"grad_norm": 0.2681499421596527,
|
274 |
+
"learning_rate": 0.0008391866913123844,
|
275 |
+
"loss": 0.0752,
|
276 |
+
"step": 925
|
277 |
+
},
|
278 |
+
{
|
279 |
+
"epoch": 1.74,
|
280 |
+
"grad_norm": 0.49241504073143005,
|
281 |
+
"learning_rate": 0.0008345656192236599,
|
282 |
+
"loss": 0.0393,
|
283 |
+
"step": 950
|
284 |
+
},
|
285 |
+
{
|
286 |
+
"epoch": 1.79,
|
287 |
+
"grad_norm": 0.3337648808956146,
|
288 |
+
"learning_rate": 0.0008299445471349352,
|
289 |
+
"loss": 0.0774,
|
290 |
+
"step": 975
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 1.83,
|
294 |
+
"grad_norm": 0.3321288228034973,
|
295 |
+
"learning_rate": 0.0008253234750462108,
|
296 |
+
"loss": 0.0513,
|
297 |
+
"step": 1000
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 1.88,
|
301 |
+
"grad_norm": 0.03459502011537552,
|
302 |
+
"learning_rate": 0.0008207024029574861,
|
303 |
+
"loss": 0.0389,
|
304 |
+
"step": 1025
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 1.92,
|
308 |
+
"grad_norm": 0.3316815197467804,
|
309 |
+
"learning_rate": 0.0008160813308687616,
|
310 |
+
"loss": 0.0556,
|
311 |
+
"step": 1050
|
312 |
+
},
|
313 |
+
{
|
314 |
+
"epoch": 1.97,
|
315 |
+
"grad_norm": 0.3402951657772064,
|
316 |
+
"learning_rate": 0.000811460258780037,
|
317 |
+
"loss": 0.0508,
|
318 |
+
"step": 1075
|
319 |
+
},
|
320 |
+
{
|
321 |
+
"epoch": 2.0,
|
322 |
+
"eval_loss": 0.22112932801246643,
|
323 |
+
"eval_runtime": 457.2021,
|
324 |
+
"eval_samples_per_second": 1.778,
|
325 |
+
"eval_steps_per_second": 0.297,
|
326 |
+
"step": 1092
|
327 |
+
},
|
328 |
+
{
|
329 |
+
"epoch": 2.01,
|
330 |
+
"grad_norm": 0.23782561719417572,
|
331 |
+
"learning_rate": 0.0008068391866913124,
|
332 |
+
"loss": 0.0493,
|
333 |
+
"step": 1100
|
334 |
+
},
|
335 |
+
{
|
336 |
+
"epoch": 2.06,
|
337 |
+
"grad_norm": 1.8971781730651855,
|
338 |
+
"learning_rate": 0.0008022181146025879,
|
339 |
+
"loss": 0.0332,
|
340 |
+
"step": 1125
|
341 |
+
},
|
342 |
+
{
|
343 |
+
"epoch": 2.11,
|
344 |
+
"grad_norm": 0.3028140664100647,
|
345 |
+
"learning_rate": 0.0007975970425138632,
|
346 |
+
"loss": 0.0718,
|
347 |
+
"step": 1150
|
348 |
+
},
|
349 |
+
{
|
350 |
+
"epoch": 2.15,
|
351 |
+
"grad_norm": 0.9446136355400085,
|
352 |
+
"learning_rate": 0.0007929759704251387,
|
353 |
+
"loss": 0.0541,
|
354 |
+
"step": 1175
|
355 |
+
},
|
356 |
+
{
|
357 |
+
"epoch": 2.2,
|
358 |
+
"grad_norm": 0.1498022973537445,
|
359 |
+
"learning_rate": 0.000788354898336414,
|
360 |
+
"loss": 0.0383,
|
361 |
+
"step": 1200
|
362 |
+
},
|
363 |
+
{
|
364 |
+
"epoch": 2.24,
|
365 |
+
"grad_norm": 0.2323082983493805,
|
366 |
+
"learning_rate": 0.0007837338262476895,
|
367 |
+
"loss": 0.0326,
|
368 |
+
"step": 1225
|
369 |
+
},
|
370 |
+
{
|
371 |
+
"epoch": 2.29,
|
372 |
+
"grad_norm": 0.15694022178649902,
|
373 |
+
"learning_rate": 0.0007791127541589649,
|
374 |
+
"loss": 0.0743,
|
375 |
+
"step": 1250
|
376 |
+
},
|
377 |
+
{
|
378 |
+
"epoch": 2.34,
|
379 |
+
"grad_norm": 0.18881258368492126,
|
380 |
+
"learning_rate": 0.0007744916820702403,
|
381 |
+
"loss": 0.0398,
|
382 |
+
"step": 1275
|
383 |
+
},
|
384 |
+
{
|
385 |
+
"epoch": 2.38,
|
386 |
+
"grad_norm": 0.5623793005943298,
|
387 |
+
"learning_rate": 0.0007698706099815157,
|
388 |
+
"loss": 0.0375,
|
389 |
+
"step": 1300
|
390 |
+
},
|
391 |
+
{
|
392 |
+
"epoch": 2.43,
|
393 |
+
"grad_norm": 0.5120436549186707,
|
394 |
+
"learning_rate": 0.0007652495378927911,
|
395 |
+
"loss": 0.0303,
|
396 |
+
"step": 1325
|
397 |
+
},
|
398 |
+
{
|
399 |
+
"epoch": 2.47,
|
400 |
+
"grad_norm": 0.22715122997760773,
|
401 |
+
"learning_rate": 0.0007606284658040665,
|
402 |
+
"loss": 0.0324,
|
403 |
+
"step": 1350
|
404 |
+
},
|
405 |
+
{
|
406 |
+
"epoch": 2.52,
|
407 |
+
"grad_norm": 0.235545352101326,
|
408 |
+
"learning_rate": 0.000756007393715342,
|
409 |
+
"loss": 0.0303,
|
410 |
+
"step": 1375
|
411 |
+
},
|
412 |
+
{
|
413 |
+
"epoch": 2.56,
|
414 |
+
"grad_norm": 0.29885298013687134,
|
415 |
+
"learning_rate": 0.0007513863216266173,
|
416 |
+
"loss": 0.0302,
|
417 |
+
"step": 1400
|
418 |
+
},
|
419 |
+
{
|
420 |
+
"epoch": 2.61,
|
421 |
+
"grad_norm": 0.30983462929725647,
|
422 |
+
"learning_rate": 0.0007467652495378928,
|
423 |
+
"loss": 0.0306,
|
424 |
+
"step": 1425
|
425 |
+
},
|
426 |
+
{
|
427 |
+
"epoch": 2.66,
|
428 |
+
"grad_norm": 0.3211610019207001,
|
429 |
+
"learning_rate": 0.0007421441774491681,
|
430 |
+
"loss": 0.0223,
|
431 |
+
"step": 1450
|
432 |
+
},
|
433 |
+
{
|
434 |
+
"epoch": 2.7,
|
435 |
+
"grad_norm": 0.020555827766656876,
|
436 |
+
"learning_rate": 0.0007375231053604437,
|
437 |
+
"loss": 0.0236,
|
438 |
+
"step": 1475
|
439 |
+
},
|
440 |
+
{
|
441 |
+
"epoch": 2.75,
|
442 |
+
"grad_norm": 0.3003218472003937,
|
443 |
+
"learning_rate": 0.000732902033271719,
|
444 |
+
"loss": 0.0292,
|
445 |
+
"step": 1500
|
446 |
+
},
|
447 |
+
{
|
448 |
+
"epoch": 2.79,
|
449 |
+
"grad_norm": 0.23368410766124725,
|
450 |
+
"learning_rate": 0.0007282809611829945,
|
451 |
+
"loss": 0.035,
|
452 |
+
"step": 1525
|
453 |
+
},
|
454 |
+
{
|
455 |
+
"epoch": 2.84,
|
456 |
+
"grad_norm": 0.24805304408073425,
|
457 |
+
"learning_rate": 0.00072365988909427,
|
458 |
+
"loss": 0.0168,
|
459 |
+
"step": 1550
|
460 |
+
},
|
461 |
+
{
|
462 |
+
"epoch": 2.88,
|
463 |
+
"grad_norm": 0.1878635734319687,
|
464 |
+
"learning_rate": 0.0007190388170055453,
|
465 |
+
"loss": 0.0277,
|
466 |
+
"step": 1575
|
467 |
+
},
|
468 |
+
{
|
469 |
+
"epoch": 2.93,
|
470 |
+
"grad_norm": 0.3910519480705261,
|
471 |
+
"learning_rate": 0.0007144177449168208,
|
472 |
+
"loss": 0.0317,
|
473 |
+
"step": 1600
|
474 |
+
},
|
475 |
+
{
|
476 |
+
"epoch": 2.98,
|
477 |
+
"grad_norm": 0.12426433712244034,
|
478 |
+
"learning_rate": 0.0007097966728280961,
|
479 |
+
"loss": 0.0287,
|
480 |
+
"step": 1625
|
481 |
+
},
|
482 |
+
{
|
483 |
+
"epoch": 3.0,
|
484 |
+
"eval_loss": 0.16812074184417725,
|
485 |
+
"eval_runtime": 461.1656,
|
486 |
+
"eval_samples_per_second": 1.763,
|
487 |
+
"eval_steps_per_second": 0.295,
|
488 |
+
"step": 1638
|
489 |
+
},
|
490 |
+
{
|
491 |
+
"epoch": 3.02,
|
492 |
+
"grad_norm": 0.009334838949143887,
|
493 |
+
"learning_rate": 0.0007051756007393716,
|
494 |
+
"loss": 0.0203,
|
495 |
+
"step": 1650
|
496 |
+
},
|
497 |
+
{
|
498 |
+
"epoch": 3.07,
|
499 |
+
"grad_norm": 0.15624983608722687,
|
500 |
+
"learning_rate": 0.000700554528650647,
|
501 |
+
"loss": 0.0222,
|
502 |
+
"step": 1675
|
503 |
+
},
|
504 |
+
{
|
505 |
+
"epoch": 3.11,
|
506 |
+
"grad_norm": 0.016200415790081024,
|
507 |
+
"learning_rate": 0.0006959334565619224,
|
508 |
+
"loss": 0.0157,
|
509 |
+
"step": 1700
|
510 |
+
},
|
511 |
+
{
|
512 |
+
"epoch": 3.16,
|
513 |
+
"grad_norm": 0.23733393847942352,
|
514 |
+
"learning_rate": 0.0006913123844731978,
|
515 |
+
"loss": 0.0291,
|
516 |
+
"step": 1725
|
517 |
+
},
|
518 |
+
{
|
519 |
+
"epoch": 3.21,
|
520 |
+
"grad_norm": 0.3168778419494629,
|
521 |
+
"learning_rate": 0.0006866913123844732,
|
522 |
+
"loss": 0.0256,
|
523 |
+
"step": 1750
|
524 |
+
},
|
525 |
+
{
|
526 |
+
"epoch": 3.25,
|
527 |
+
"grad_norm": 0.22910478711128235,
|
528 |
+
"learning_rate": 0.0006820702402957486,
|
529 |
+
"loss": 0.0211,
|
530 |
+
"step": 1775
|
531 |
+
},
|
532 |
+
{
|
533 |
+
"epoch": 3.3,
|
534 |
+
"grad_norm": 0.17075732350349426,
|
535 |
+
"learning_rate": 0.000677449168207024,
|
536 |
+
"loss": 0.0267,
|
537 |
+
"step": 1800
|
538 |
+
},
|
539 |
+
{
|
540 |
+
"epoch": 3.34,
|
541 |
+
"grad_norm": 0.17666079103946686,
|
542 |
+
"learning_rate": 0.0006728280961182994,
|
543 |
+
"loss": 0.0187,
|
544 |
+
"step": 1825
|
545 |
+
},
|
546 |
+
{
|
547 |
+
"epoch": 3.39,
|
548 |
+
"grad_norm": 0.020759843289852142,
|
549 |
+
"learning_rate": 0.0006682070240295749,
|
550 |
+
"loss": 0.0265,
|
551 |
+
"step": 1850
|
552 |
+
},
|
553 |
+
{
|
554 |
+
"epoch": 3.43,
|
555 |
+
"grad_norm": 0.04912843555212021,
|
556 |
+
"learning_rate": 0.0006635859519408502,
|
557 |
+
"loss": 0.0359,
|
558 |
+
"step": 1875
|
559 |
+
},
|
560 |
+
{
|
561 |
+
"epoch": 3.48,
|
562 |
+
"grad_norm": 0.32245051860809326,
|
563 |
+
"learning_rate": 0.0006589648798521257,
|
564 |
+
"loss": 0.0175,
|
565 |
+
"step": 1900
|
566 |
+
},
|
567 |
+
{
|
568 |
+
"epoch": 3.53,
|
569 |
+
"grad_norm": 0.11074571311473846,
|
570 |
+
"learning_rate": 0.000654343807763401,
|
571 |
+
"loss": 0.0178,
|
572 |
+
"step": 1925
|
573 |
+
},
|
574 |
+
{
|
575 |
+
"epoch": 3.57,
|
576 |
+
"grad_norm": 0.03214950114488602,
|
577 |
+
"learning_rate": 0.0006497227356746766,
|
578 |
+
"loss": 0.0192,
|
579 |
+
"step": 1950
|
580 |
+
},
|
581 |
+
{
|
582 |
+
"epoch": 3.62,
|
583 |
+
"grad_norm": 0.008081772364675999,
|
584 |
+
"learning_rate": 0.000645101663585952,
|
585 |
+
"loss": 0.0317,
|
586 |
+
"step": 1975
|
587 |
+
},
|
588 |
+
{
|
589 |
+
"epoch": 3.66,
|
590 |
+
"grad_norm": 0.1419508308172226,
|
591 |
+
"learning_rate": 0.0006404805914972274,
|
592 |
+
"loss": 0.0297,
|
593 |
+
"step": 2000
|
594 |
+
},
|
595 |
+
{
|
596 |
+
"epoch": 3.71,
|
597 |
+
"grad_norm": 0.01866302080452442,
|
598 |
+
"learning_rate": 0.0006358595194085029,
|
599 |
+
"loss": 0.0264,
|
600 |
+
"step": 2025
|
601 |
+
},
|
602 |
+
{
|
603 |
+
"epoch": 3.75,
|
604 |
+
"grad_norm": 0.0233779214322567,
|
605 |
+
"learning_rate": 0.0006312384473197782,
|
606 |
+
"loss": 0.0285,
|
607 |
+
"step": 2050
|
608 |
+
},
|
609 |
+
{
|
610 |
+
"epoch": 3.8,
|
611 |
+
"grad_norm": 0.023354342207312584,
|
612 |
+
"learning_rate": 0.0006266173752310537,
|
613 |
+
"loss": 0.021,
|
614 |
+
"step": 2075
|
615 |
+
},
|
616 |
+
{
|
617 |
+
"epoch": 3.85,
|
618 |
+
"grad_norm": 0.2687942087650299,
|
619 |
+
"learning_rate": 0.000621996303142329,
|
620 |
+
"loss": 0.0201,
|
621 |
+
"step": 2100
|
622 |
+
},
|
623 |
+
{
|
624 |
+
"epoch": 3.89,
|
625 |
+
"grad_norm": 0.15729880332946777,
|
626 |
+
"learning_rate": 0.0006173752310536045,
|
627 |
+
"loss": 0.0433,
|
628 |
+
"step": 2125
|
629 |
+
},
|
630 |
+
{
|
631 |
+
"epoch": 3.94,
|
632 |
+
"grad_norm": 0.13736210763454437,
|
633 |
+
"learning_rate": 0.0006127541589648799,
|
634 |
+
"loss": 0.0269,
|
635 |
+
"step": 2150
|
636 |
+
},
|
637 |
+
{
|
638 |
+
"epoch": 3.98,
|
639 |
+
"grad_norm": 0.17346683144569397,
|
640 |
+
"learning_rate": 0.0006081330868761553,
|
641 |
+
"loss": 0.0148,
|
642 |
+
"step": 2175
|
643 |
+
},
|
644 |
+
{
|
645 |
+
"epoch": 4.0,
|
646 |
+
"eval_loss": 0.1938144713640213,
|
647 |
+
"eval_runtime": 460.937,
|
648 |
+
"eval_samples_per_second": 1.764,
|
649 |
+
"eval_steps_per_second": 0.295,
|
650 |
+
"step": 2184
|
651 |
+
},
|
652 |
+
{
|
653 |
+
"epoch": 4.03,
|
654 |
+
"grad_norm": 0.1308988779783249,
|
655 |
+
"learning_rate": 0.0006035120147874307,
|
656 |
+
"loss": 0.0195,
|
657 |
+
"step": 2200
|
658 |
+
},
|
659 |
+
{
|
660 |
+
"epoch": 4.08,
|
661 |
+
"grad_norm": 0.21798363327980042,
|
662 |
+
"learning_rate": 0.0005988909426987061,
|
663 |
+
"loss": 0.018,
|
664 |
+
"step": 2225
|
665 |
+
},
|
666 |
+
{
|
667 |
+
"epoch": 4.12,
|
668 |
+
"grad_norm": 0.059930045157670975,
|
669 |
+
"learning_rate": 0.0005942698706099815,
|
670 |
+
"loss": 0.0157,
|
671 |
+
"step": 2250
|
672 |
+
},
|
673 |
+
{
|
674 |
+
"epoch": 4.17,
|
675 |
+
"grad_norm": 1.1142582893371582,
|
676 |
+
"learning_rate": 0.000589648798521257,
|
677 |
+
"loss": 0.0252,
|
678 |
+
"step": 2275
|
679 |
+
},
|
680 |
+
{
|
681 |
+
"epoch": 4.21,
|
682 |
+
"grad_norm": 0.17724983394145966,
|
683 |
+
"learning_rate": 0.0005850277264325323,
|
684 |
+
"loss": 0.0251,
|
685 |
+
"step": 2300
|
686 |
+
},
|
687 |
+
{
|
688 |
+
"epoch": 4.26,
|
689 |
+
"grad_norm": 0.7539493441581726,
|
690 |
+
"learning_rate": 0.0005804066543438078,
|
691 |
+
"loss": 0.0226,
|
692 |
+
"step": 2325
|
693 |
+
},
|
694 |
+
{
|
695 |
+
"epoch": 4.3,
|
696 |
+
"grad_norm": 0.15379472076892853,
|
697 |
+
"learning_rate": 0.0005757855822550831,
|
698 |
+
"loss": 0.0154,
|
699 |
+
"step": 2350
|
700 |
+
},
|
701 |
+
{
|
702 |
+
"epoch": 4.35,
|
703 |
+
"grad_norm": 0.11480142921209335,
|
704 |
+
"learning_rate": 0.0005711645101663586,
|
705 |
+
"loss": 0.0302,
|
706 |
+
"step": 2375
|
707 |
+
},
|
708 |
+
{
|
709 |
+
"epoch": 4.4,
|
710 |
+
"grad_norm": 0.29920852184295654,
|
711 |
+
"learning_rate": 0.0005665434380776339,
|
712 |
+
"loss": 0.0223,
|
713 |
+
"step": 2400
|
714 |
+
},
|
715 |
+
{
|
716 |
+
"epoch": 4.44,
|
717 |
+
"grad_norm": 0.2625471353530884,
|
718 |
+
"learning_rate": 0.0005619223659889095,
|
719 |
+
"loss": 0.0126,
|
720 |
+
"step": 2425
|
721 |
+
},
|
722 |
+
{
|
723 |
+
"epoch": 4.49,
|
724 |
+
"grad_norm": 0.2014468014240265,
|
725 |
+
"learning_rate": 0.0005573012939001849,
|
726 |
+
"loss": 0.0262,
|
727 |
+
"step": 2450
|
728 |
+
},
|
729 |
+
{
|
730 |
+
"epoch": 4.53,
|
731 |
+
"grad_norm": 0.20631028711795807,
|
732 |
+
"learning_rate": 0.0005526802218114603,
|
733 |
+
"loss": 0.0128,
|
734 |
+
"step": 2475
|
735 |
+
},
|
736 |
+
{
|
737 |
+
"epoch": 4.58,
|
738 |
+
"grad_norm": 0.1370575875043869,
|
739 |
+
"learning_rate": 0.0005480591497227358,
|
740 |
+
"loss": 0.0156,
|
741 |
+
"step": 2500
|
742 |
+
},
|
743 |
+
{
|
744 |
+
"epoch": 4.62,
|
745 |
+
"grad_norm": 0.058717742562294006,
|
746 |
+
"learning_rate": 0.0005434380776340111,
|
747 |
+
"loss": 0.0162,
|
748 |
+
"step": 2525
|
749 |
+
},
|
750 |
+
{
|
751 |
+
"epoch": 4.67,
|
752 |
+
"grad_norm": 0.010219058953225613,
|
753 |
+
"learning_rate": 0.0005388170055452866,
|
754 |
+
"loss": 0.0169,
|
755 |
+
"step": 2550
|
756 |
+
},
|
757 |
+
{
|
758 |
+
"epoch": 4.72,
|
759 |
+
"grad_norm": 0.1966046839952469,
|
760 |
+
"learning_rate": 0.0005341959334565619,
|
761 |
+
"loss": 0.0138,
|
762 |
+
"step": 2575
|
763 |
+
},
|
764 |
+
{
|
765 |
+
"epoch": 4.76,
|
766 |
+
"grad_norm": 0.4009633958339691,
|
767 |
+
"learning_rate": 0.0005295748613678374,
|
768 |
+
"loss": 0.0219,
|
769 |
+
"step": 2600
|
770 |
+
},
|
771 |
+
{
|
772 |
+
"epoch": 4.81,
|
773 |
+
"grad_norm": 0.002215989399701357,
|
774 |
+
"learning_rate": 0.0005249537892791128,
|
775 |
+
"loss": 0.0141,
|
776 |
+
"step": 2625
|
777 |
+
},
|
778 |
+
{
|
779 |
+
"epoch": 4.85,
|
780 |
+
"grad_norm": 0.31632617115974426,
|
781 |
+
"learning_rate": 0.0005203327171903882,
|
782 |
+
"loss": 0.0121,
|
783 |
+
"step": 2650
|
784 |
+
},
|
785 |
+
{
|
786 |
+
"epoch": 4.9,
|
787 |
+
"grad_norm": 0.003695544321089983,
|
788 |
+
"learning_rate": 0.0005157116451016636,
|
789 |
+
"loss": 0.0152,
|
790 |
+
"step": 2675
|
791 |
+
},
|
792 |
+
{
|
793 |
+
"epoch": 4.95,
|
794 |
+
"grad_norm": 0.03792522847652435,
|
795 |
+
"learning_rate": 0.000511090573012939,
|
796 |
+
"loss": 0.0248,
|
797 |
+
"step": 2700
|
798 |
+
},
|
799 |
+
{
|
800 |
+
"epoch": 4.99,
|
801 |
+
"grad_norm": 0.22871936857700348,
|
802 |
+
"learning_rate": 0.0005064695009242144,
|
803 |
+
"loss": 0.0263,
|
804 |
+
"step": 2725
|
805 |
+
},
|
806 |
+
{
|
807 |
+
"epoch": 5.0,
|
808 |
+
"eval_loss": 0.1845603585243225,
|
809 |
+
"eval_runtime": 463.3179,
|
810 |
+
"eval_samples_per_second": 1.755,
|
811 |
+
"eval_steps_per_second": 0.294,
|
812 |
+
"step": 2730
|
813 |
+
},
|
814 |
+
{
|
815 |
+
"epoch": 5.04,
|
816 |
+
"grad_norm": 0.2171017825603485,
|
817 |
+
"learning_rate": 0.0005018484288354898,
|
818 |
+
"loss": 0.0141,
|
819 |
+
"step": 2750
|
820 |
+
},
|
821 |
+
{
|
822 |
+
"epoch": 5.08,
|
823 |
+
"grad_norm": 0.0038620266132056713,
|
824 |
+
"learning_rate": 0.0004972273567467653,
|
825 |
+
"loss": 0.0058,
|
826 |
+
"step": 2775
|
827 |
+
},
|
828 |
+
{
|
829 |
+
"epoch": 5.13,
|
830 |
+
"grad_norm": 0.2298993319272995,
|
831 |
+
"learning_rate": 0.0004926062846580407,
|
832 |
+
"loss": 0.0109,
|
833 |
+
"step": 2800
|
834 |
+
},
|
835 |
+
{
|
836 |
+
"epoch": 5.17,
|
837 |
+
"grad_norm": 0.11024118214845657,
|
838 |
+
"learning_rate": 0.0004879852125693161,
|
839 |
+
"loss": 0.0189,
|
840 |
+
"step": 2825
|
841 |
+
},
|
842 |
+
{
|
843 |
+
"epoch": 5.22,
|
844 |
+
"grad_norm": 0.0567074678838253,
|
845 |
+
"learning_rate": 0.00048336414048059153,
|
846 |
+
"loss": 0.0089,
|
847 |
+
"step": 2850
|
848 |
+
},
|
849 |
+
{
|
850 |
+
"epoch": 5.27,
|
851 |
+
"grad_norm": 0.004749275743961334,
|
852 |
+
"learning_rate": 0.00047874306839186694,
|
853 |
+
"loss": 0.0114,
|
854 |
+
"step": 2875
|
855 |
+
},
|
856 |
+
{
|
857 |
+
"epoch": 5.31,
|
858 |
+
"grad_norm": 0.14577604830265045,
|
859 |
+
"learning_rate": 0.00047412199630314235,
|
860 |
+
"loss": 0.0129,
|
861 |
+
"step": 2900
|
862 |
+
},
|
863 |
+
{
|
864 |
+
"epoch": 5.36,
|
865 |
+
"grad_norm": 0.11412041634321213,
|
866 |
+
"learning_rate": 0.00046950092421441775,
|
867 |
+
"loss": 0.0127,
|
868 |
+
"step": 2925
|
869 |
+
},
|
870 |
+
{
|
871 |
+
"epoch": 5.4,
|
872 |
+
"grad_norm": 0.29708778858184814,
|
873 |
+
"learning_rate": 0.00046487985212569316,
|
874 |
+
"loss": 0.0334,
|
875 |
+
"step": 2950
|
876 |
+
},
|
877 |
+
{
|
878 |
+
"epoch": 5.45,
|
879 |
+
"grad_norm": 0.12024960666894913,
|
880 |
+
"learning_rate": 0.00046025878003696857,
|
881 |
+
"loss": 0.0153,
|
882 |
+
"step": 2975
|
883 |
+
},
|
884 |
+
{
|
885 |
+
"epoch": 5.49,
|
886 |
+
"grad_norm": 0.12928707897663116,
|
887 |
+
"learning_rate": 0.000455637707948244,
|
888 |
+
"loss": 0.0153,
|
889 |
+
"step": 3000
|
890 |
+
},
|
891 |
+
{
|
892 |
+
"epoch": 5.54,
|
893 |
+
"grad_norm": 0.1562725156545639,
|
894 |
+
"learning_rate": 0.0004510166358595194,
|
895 |
+
"loss": 0.0174,
|
896 |
+
"step": 3025
|
897 |
+
},
|
898 |
+
{
|
899 |
+
"epoch": 5.59,
|
900 |
+
"grad_norm": 0.270773321390152,
|
901 |
+
"learning_rate": 0.00044639556377079484,
|
902 |
+
"loss": 0.0221,
|
903 |
+
"step": 3050
|
904 |
+
},
|
905 |
+
{
|
906 |
+
"epoch": 5.63,
|
907 |
+
"grad_norm": 0.36300143599510193,
|
908 |
+
"learning_rate": 0.00044177449168207025,
|
909 |
+
"loss": 0.0219,
|
910 |
+
"step": 3075
|
911 |
+
},
|
912 |
+
{
|
913 |
+
"epoch": 5.68,
|
914 |
+
"grad_norm": 0.1869664192199707,
|
915 |
+
"learning_rate": 0.00043715341959334566,
|
916 |
+
"loss": 0.0127,
|
917 |
+
"step": 3100
|
918 |
+
},
|
919 |
+
{
|
920 |
+
"epoch": 5.72,
|
921 |
+
"grad_norm": 0.23992718756198883,
|
922 |
+
"learning_rate": 0.00043253234750462107,
|
923 |
+
"loss": 0.0144,
|
924 |
+
"step": 3125
|
925 |
+
},
|
926 |
+
{
|
927 |
+
"epoch": 5.77,
|
928 |
+
"grad_norm": 0.0021614329889416695,
|
929 |
+
"learning_rate": 0.00042791127541589647,
|
930 |
+
"loss": 0.0115,
|
931 |
+
"step": 3150
|
932 |
+
},
|
933 |
+
{
|
934 |
+
"epoch": 5.82,
|
935 |
+
"grad_norm": 0.10629579424858093,
|
936 |
+
"learning_rate": 0.0004232902033271719,
|
937 |
+
"loss": 0.0097,
|
938 |
+
"step": 3175
|
939 |
+
},
|
940 |
+
{
|
941 |
+
"epoch": 5.86,
|
942 |
+
"grad_norm": 0.15990929305553436,
|
943 |
+
"learning_rate": 0.00041866913123844734,
|
944 |
+
"loss": 0.0118,
|
945 |
+
"step": 3200
|
946 |
+
},
|
947 |
+
{
|
948 |
+
"epoch": 5.91,
|
949 |
+
"grad_norm": 0.19984115660190582,
|
950 |
+
"learning_rate": 0.00041404805914972275,
|
951 |
+
"loss": 0.014,
|
952 |
+
"step": 3225
|
953 |
+
},
|
954 |
+
{
|
955 |
+
"epoch": 5.95,
|
956 |
+
"grad_norm": 0.06720598042011261,
|
957 |
+
"learning_rate": 0.00040942698706099816,
|
958 |
+
"loss": 0.0182,
|
959 |
+
"step": 3250
|
960 |
+
},
|
961 |
+
{
|
962 |
+
"epoch": 6.0,
|
963 |
+
"grad_norm": 0.006452410947531462,
|
964 |
+
"learning_rate": 0.0004048059149722736,
|
965 |
+
"loss": 0.0168,
|
966 |
+
"step": 3275
|
967 |
+
},
|
968 |
+
{
|
969 |
+
"epoch": 6.0,
|
970 |
+
"eval_loss": 0.18991737067699432,
|
971 |
+
"eval_runtime": 463.6922,
|
972 |
+
"eval_samples_per_second": 1.753,
|
973 |
+
"eval_steps_per_second": 0.293,
|
974 |
+
"step": 3276
|
975 |
+
},
|
976 |
+
{
|
977 |
+
"epoch": 6.04,
|
978 |
+
"grad_norm": 0.003035791451111436,
|
979 |
+
"learning_rate": 0.000400184842883549,
|
980 |
+
"loss": 0.0103,
|
981 |
+
"step": 3300
|
982 |
+
},
|
983 |
+
{
|
984 |
+
"epoch": 6.09,
|
985 |
+
"grad_norm": 0.20400433242321014,
|
986 |
+
"learning_rate": 0.00039556377079482443,
|
987 |
+
"loss": 0.0094,
|
988 |
+
"step": 3325
|
989 |
+
},
|
990 |
+
{
|
991 |
+
"epoch": 6.14,
|
992 |
+
"grad_norm": 0.05333884805440903,
|
993 |
+
"learning_rate": 0.00039094269870609984,
|
994 |
+
"loss": 0.0135,
|
995 |
+
"step": 3350
|
996 |
+
},
|
997 |
+
{
|
998 |
+
"epoch": 6.18,
|
999 |
+
"grad_norm": 0.0005341291544027627,
|
1000 |
+
"learning_rate": 0.00038632162661737525,
|
1001 |
+
"loss": 0.0042,
|
1002 |
+
"step": 3375
|
1003 |
+
},
|
1004 |
+
{
|
1005 |
+
"epoch": 6.23,
|
1006 |
+
"grad_norm": 0.008850090205669403,
|
1007 |
+
"learning_rate": 0.00038170055452865065,
|
1008 |
+
"loss": 0.012,
|
1009 |
+
"step": 3400
|
1010 |
+
},
|
1011 |
+
{
|
1012 |
+
"epoch": 6.27,
|
1013 |
+
"grad_norm": 0.013096541166305542,
|
1014 |
+
"learning_rate": 0.00037707948243992606,
|
1015 |
+
"loss": 0.0121,
|
1016 |
+
"step": 3425
|
1017 |
+
},
|
1018 |
+
{
|
1019 |
+
"epoch": 6.32,
|
1020 |
+
"grad_norm": 0.04941894859075546,
|
1021 |
+
"learning_rate": 0.00037245841035120147,
|
1022 |
+
"loss": 0.0086,
|
1023 |
+
"step": 3450
|
1024 |
+
},
|
1025 |
+
{
|
1026 |
+
"epoch": 6.36,
|
1027 |
+
"grad_norm": 0.02113133668899536,
|
1028 |
+
"learning_rate": 0.0003678373382624769,
|
1029 |
+
"loss": 0.0109,
|
1030 |
+
"step": 3475
|
1031 |
+
},
|
1032 |
+
{
|
1033 |
+
"epoch": 6.41,
|
1034 |
+
"grad_norm": 0.03568890690803528,
|
1035 |
+
"learning_rate": 0.0003632162661737523,
|
1036 |
+
"loss": 0.0072,
|
1037 |
+
"step": 3500
|
1038 |
+
},
|
1039 |
+
{
|
1040 |
+
"epoch": 6.46,
|
1041 |
+
"grad_norm": 0.0471993163228035,
|
1042 |
+
"learning_rate": 0.00035859519408502774,
|
1043 |
+
"loss": 0.0074,
|
1044 |
+
"step": 3525
|
1045 |
+
},
|
1046 |
+
{
|
1047 |
+
"epoch": 6.5,
|
1048 |
+
"grad_norm": 0.13575506210327148,
|
1049 |
+
"learning_rate": 0.00035397412199630315,
|
1050 |
+
"loss": 0.0084,
|
1051 |
+
"step": 3550
|
1052 |
+
},
|
1053 |
+
{
|
1054 |
+
"epoch": 6.55,
|
1055 |
+
"grad_norm": 0.14595919847488403,
|
1056 |
+
"learning_rate": 0.00034935304990757856,
|
1057 |
+
"loss": 0.0063,
|
1058 |
+
"step": 3575
|
1059 |
+
},
|
1060 |
+
{
|
1061 |
+
"epoch": 6.59,
|
1062 |
+
"grad_norm": 0.012155482545495033,
|
1063 |
+
"learning_rate": 0.00034473197781885397,
|
1064 |
+
"loss": 0.0092,
|
1065 |
+
"step": 3600
|
1066 |
+
},
|
1067 |
+
{
|
1068 |
+
"epoch": 6.64,
|
1069 |
+
"grad_norm": 0.010629130527377129,
|
1070 |
+
"learning_rate": 0.0003401109057301294,
|
1071 |
+
"loss": 0.0134,
|
1072 |
+
"step": 3625
|
1073 |
+
},
|
1074 |
+
{
|
1075 |
+
"epoch": 6.68,
|
1076 |
+
"grad_norm": 0.19480323791503906,
|
1077 |
+
"learning_rate": 0.00033548983364140483,
|
1078 |
+
"loss": 0.0141,
|
1079 |
+
"step": 3650
|
1080 |
+
},
|
1081 |
+
{
|
1082 |
+
"epoch": 6.73,
|
1083 |
+
"grad_norm": 0.0660039409995079,
|
1084 |
+
"learning_rate": 0.00033086876155268024,
|
1085 |
+
"loss": 0.0077,
|
1086 |
+
"step": 3675
|
1087 |
+
},
|
1088 |
+
{
|
1089 |
+
"epoch": 6.78,
|
1090 |
+
"grad_norm": 0.0346703939139843,
|
1091 |
+
"learning_rate": 0.00032624768946395565,
|
1092 |
+
"loss": 0.0117,
|
1093 |
+
"step": 3700
|
1094 |
+
},
|
1095 |
+
{
|
1096 |
+
"epoch": 6.82,
|
1097 |
+
"grad_norm": 0.22081167995929718,
|
1098 |
+
"learning_rate": 0.00032162661737523106,
|
1099 |
+
"loss": 0.0099,
|
1100 |
+
"step": 3725
|
1101 |
+
},
|
1102 |
+
{
|
1103 |
+
"epoch": 6.87,
|
1104 |
+
"grad_norm": 0.06806311756372452,
|
1105 |
+
"learning_rate": 0.0003170055452865065,
|
1106 |
+
"loss": 0.0115,
|
1107 |
+
"step": 3750
|
1108 |
+
},
|
1109 |
+
{
|
1110 |
+
"epoch": 6.91,
|
1111 |
+
"grad_norm": 0.0030449284240603447,
|
1112 |
+
"learning_rate": 0.0003123844731977819,
|
1113 |
+
"loss": 0.0136,
|
1114 |
+
"step": 3775
|
1115 |
+
},
|
1116 |
+
{
|
1117 |
+
"epoch": 6.96,
|
1118 |
+
"grad_norm": 0.07848404347896576,
|
1119 |
+
"learning_rate": 0.00030776340110905733,
|
1120 |
+
"loss": 0.0086,
|
1121 |
+
"step": 3800
|
1122 |
+
},
|
1123 |
+
{
|
1124 |
+
"epoch": 7.0,
|
1125 |
+
"eval_loss": 0.19748586416244507,
|
1126 |
+
"eval_runtime": 457.6513,
|
1127 |
+
"eval_samples_per_second": 1.776,
|
1128 |
+
"eval_steps_per_second": 0.297,
|
1129 |
+
"step": 3822
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 7.01,
|
1133 |
+
"grad_norm": 0.04702220484614372,
|
1134 |
+
"learning_rate": 0.00030314232902033274,
|
1135 |
+
"loss": 0.009,
|
1136 |
+
"step": 3825
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 7.05,
|
1140 |
+
"grad_norm": 0.026063207536935806,
|
1141 |
+
"learning_rate": 0.00029852125693160815,
|
1142 |
+
"loss": 0.0079,
|
1143 |
+
"step": 3850
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 7.1,
|
1147 |
+
"grad_norm": 0.021074611693620682,
|
1148 |
+
"learning_rate": 0.00029390018484288355,
|
1149 |
+
"loss": 0.0042,
|
1150 |
+
"step": 3875
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 7.14,
|
1154 |
+
"grad_norm": 0.15062950551509857,
|
1155 |
+
"learning_rate": 0.00028927911275415896,
|
1156 |
+
"loss": 0.0073,
|
1157 |
+
"step": 3900
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 7.19,
|
1161 |
+
"grad_norm": 0.12703749537467957,
|
1162 |
+
"learning_rate": 0.00028465804066543437,
|
1163 |
+
"loss": 0.0047,
|
1164 |
+
"step": 3925
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 7.23,
|
1168 |
+
"grad_norm": 0.0032193493098020554,
|
1169 |
+
"learning_rate": 0.0002800369685767098,
|
1170 |
+
"loss": 0.0071,
|
1171 |
+
"step": 3950
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 7.28,
|
1175 |
+
"grad_norm": 0.046458516269922256,
|
1176 |
+
"learning_rate": 0.0002754158964879852,
|
1177 |
+
"loss": 0.0074,
|
1178 |
+
"step": 3975
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 7.33,
|
1182 |
+
"grad_norm": 0.0037984629161655903,
|
1183 |
+
"learning_rate": 0.00027079482439926065,
|
1184 |
+
"loss": 0.0059,
|
1185 |
+
"step": 4000
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 7.37,
|
1189 |
+
"grad_norm": 0.14948821067810059,
|
1190 |
+
"learning_rate": 0.00026617375231053605,
|
1191 |
+
"loss": 0.0058,
|
1192 |
+
"step": 4025
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 7.42,
|
1196 |
+
"grad_norm": 0.07740973681211472,
|
1197 |
+
"learning_rate": 0.00026155268022181146,
|
1198 |
+
"loss": 0.0069,
|
1199 |
+
"step": 4050
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 7.46,
|
1203 |
+
"grad_norm": 0.0008731162524782121,
|
1204 |
+
"learning_rate": 0.00025693160813308687,
|
1205 |
+
"loss": 0.0077,
|
1206 |
+
"step": 4075
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 7.51,
|
1210 |
+
"grad_norm": 0.001257123309187591,
|
1211 |
+
"learning_rate": 0.0002523105360443623,
|
1212 |
+
"loss": 0.0046,
|
1213 |
+
"step": 4100
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 7.55,
|
1217 |
+
"grad_norm": 0.042853593826293945,
|
1218 |
+
"learning_rate": 0.00024768946395563774,
|
1219 |
+
"loss": 0.0065,
|
1220 |
+
"step": 4125
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 7.6,
|
1224 |
+
"grad_norm": 0.0009361489792354405,
|
1225 |
+
"learning_rate": 0.00024306839186691312,
|
1226 |
+
"loss": 0.0073,
|
1227 |
+
"step": 4150
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 7.65,
|
1231 |
+
"grad_norm": 0.04179251566529274,
|
1232 |
+
"learning_rate": 0.00023844731977818855,
|
1233 |
+
"loss": 0.0043,
|
1234 |
+
"step": 4175
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 7.69,
|
1238 |
+
"grad_norm": 0.008879727683961391,
|
1239 |
+
"learning_rate": 0.00023382624768946396,
|
1240 |
+
"loss": 0.0095,
|
1241 |
+
"step": 4200
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 7.74,
|
1245 |
+
"grad_norm": 0.12861858308315277,
|
1246 |
+
"learning_rate": 0.00022920517560073937,
|
1247 |
+
"loss": 0.008,
|
1248 |
+
"step": 4225
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 7.78,
|
1252 |
+
"grad_norm": 0.01530044712126255,
|
1253 |
+
"learning_rate": 0.0002245841035120148,
|
1254 |
+
"loss": 0.0032,
|
1255 |
+
"step": 4250
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 7.83,
|
1259 |
+
"grad_norm": 0.02794441021978855,
|
1260 |
+
"learning_rate": 0.0002199630314232902,
|
1261 |
+
"loss": 0.0087,
|
1262 |
+
"step": 4275
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 7.88,
|
1266 |
+
"grad_norm": 0.16127441823482513,
|
1267 |
+
"learning_rate": 0.00021534195933456564,
|
1268 |
+
"loss": 0.0063,
|
1269 |
+
"step": 4300
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 7.92,
|
1273 |
+
"grad_norm": 0.09033193439245224,
|
1274 |
+
"learning_rate": 0.00021072088724584105,
|
1275 |
+
"loss": 0.0086,
|
1276 |
+
"step": 4325
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 7.97,
|
1280 |
+
"grad_norm": 0.016767054796218872,
|
1281 |
+
"learning_rate": 0.00020609981515711646,
|
1282 |
+
"loss": 0.0102,
|
1283 |
+
"step": 4350
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 8.0,
|
1287 |
+
"eval_loss": 0.21702823042869568,
|
1288 |
+
"eval_runtime": 460.637,
|
1289 |
+
"eval_samples_per_second": 1.765,
|
1290 |
+
"eval_steps_per_second": 0.295,
|
1291 |
+
"step": 4368
|
1292 |
+
},
|
1293 |
+
{
|
1294 |
+
"epoch": 8.01,
|
1295 |
+
"grad_norm": 0.10574544966220856,
|
1296 |
+
"learning_rate": 0.00020147874306839186,
|
1297 |
+
"loss": 0.0047,
|
1298 |
+
"step": 4375
|
1299 |
+
},
|
1300 |
+
{
|
1301 |
+
"epoch": 8.06,
|
1302 |
+
"grad_norm": 0.0006509744562208652,
|
1303 |
+
"learning_rate": 0.00019685767097966727,
|
1304 |
+
"loss": 0.005,
|
1305 |
+
"step": 4400
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"epoch": 8.1,
|
1309 |
+
"grad_norm": 0.0009672873420640826,
|
1310 |
+
"learning_rate": 0.0001922365988909427,
|
1311 |
+
"loss": 0.0046,
|
1312 |
+
"step": 4425
|
1313 |
+
},
|
1314 |
+
{
|
1315 |
+
"epoch": 8.15,
|
1316 |
+
"grad_norm": 0.07224971055984497,
|
1317 |
+
"learning_rate": 0.0001876155268022181,
|
1318 |
+
"loss": 0.0043,
|
1319 |
+
"step": 4450
|
1320 |
+
},
|
1321 |
+
{
|
1322 |
+
"epoch": 8.2,
|
1323 |
+
"grad_norm": 0.12703950703144073,
|
1324 |
+
"learning_rate": 0.00018299445471349355,
|
1325 |
+
"loss": 0.0058,
|
1326 |
+
"step": 4475
|
1327 |
+
},
|
1328 |
+
{
|
1329 |
+
"epoch": 8.24,
|
1330 |
+
"grad_norm": 0.0013393750414252281,
|
1331 |
+
"learning_rate": 0.00017837338262476895,
|
1332 |
+
"loss": 0.0041,
|
1333 |
+
"step": 4500
|
1334 |
+
},
|
1335 |
+
{
|
1336 |
+
"epoch": 8.29,
|
1337 |
+
"grad_norm": 0.03772176802158356,
|
1338 |
+
"learning_rate": 0.00017375231053604436,
|
1339 |
+
"loss": 0.0043,
|
1340 |
+
"step": 4525
|
1341 |
+
},
|
1342 |
+
{
|
1343 |
+
"epoch": 8.33,
|
1344 |
+
"grad_norm": 0.0009993729181587696,
|
1345 |
+
"learning_rate": 0.0001691312384473198,
|
1346 |
+
"loss": 0.0051,
|
1347 |
+
"step": 4550
|
1348 |
+
},
|
1349 |
+
{
|
1350 |
+
"epoch": 8.38,
|
1351 |
+
"grad_norm": 0.0068801455199718475,
|
1352 |
+
"learning_rate": 0.0001645101663585952,
|
1353 |
+
"loss": 0.0042,
|
1354 |
+
"step": 4575
|
1355 |
+
},
|
1356 |
+
{
|
1357 |
+
"epoch": 8.42,
|
1358 |
+
"grad_norm": 0.05337873101234436,
|
1359 |
+
"learning_rate": 0.0001598890942698706,
|
1360 |
+
"loss": 0.0048,
|
1361 |
+
"step": 4600
|
1362 |
+
},
|
1363 |
+
{
|
1364 |
+
"epoch": 8.47,
|
1365 |
+
"grad_norm": 0.022360146045684814,
|
1366 |
+
"learning_rate": 0.00015526802218114602,
|
1367 |
+
"loss": 0.0058,
|
1368 |
+
"step": 4625
|
1369 |
+
},
|
1370 |
+
{
|
1371 |
+
"epoch": 8.52,
|
1372 |
+
"grad_norm": 0.13477857410907745,
|
1373 |
+
"learning_rate": 0.00015064695009242142,
|
1374 |
+
"loss": 0.0029,
|
1375 |
+
"step": 4650
|
1376 |
+
},
|
1377 |
+
{
|
1378 |
+
"epoch": 8.56,
|
1379 |
+
"grad_norm": 0.23147088289260864,
|
1380 |
+
"learning_rate": 0.00014602587800369686,
|
1381 |
+
"loss": 0.0057,
|
1382 |
+
"step": 4675
|
1383 |
+
},
|
1384 |
+
{
|
1385 |
+
"epoch": 8.61,
|
1386 |
+
"grad_norm": 0.0034095763694494963,
|
1387 |
+
"learning_rate": 0.0001414048059149723,
|
1388 |
+
"loss": 0.0043,
|
1389 |
+
"step": 4700
|
1390 |
+
},
|
1391 |
+
{
|
1392 |
+
"epoch": 8.65,
|
1393 |
+
"grad_norm": 0.024832753464579582,
|
1394 |
+
"learning_rate": 0.0001367837338262477,
|
1395 |
+
"loss": 0.0043,
|
1396 |
+
"step": 4725
|
1397 |
+
},
|
1398 |
+
{
|
1399 |
+
"epoch": 8.7,
|
1400 |
+
"grad_norm": 0.0007142982794903219,
|
1401 |
+
"learning_rate": 0.0001321626617375231,
|
1402 |
+
"loss": 0.0052,
|
1403 |
+
"step": 4750
|
1404 |
+
},
|
1405 |
+
{
|
1406 |
+
"epoch": 8.75,
|
1407 |
+
"grad_norm": 0.02869781292974949,
|
1408 |
+
"learning_rate": 0.00012754158964879852,
|
1409 |
+
"loss": 0.0027,
|
1410 |
+
"step": 4775
|
1411 |
+
},
|
1412 |
+
{
|
1413 |
+
"epoch": 8.79,
|
1414 |
+
"grad_norm": 0.0004513516614679247,
|
1415 |
+
"learning_rate": 0.00012292051756007395,
|
1416 |
+
"loss": 0.0036,
|
1417 |
+
"step": 4800
|
1418 |
+
},
|
1419 |
+
{
|
1420 |
+
"epoch": 8.84,
|
1421 |
+
"grad_norm": 0.0008141061407513916,
|
1422 |
+
"learning_rate": 0.00011829944547134936,
|
1423 |
+
"loss": 0.0063,
|
1424 |
+
"step": 4825
|
1425 |
+
},
|
1426 |
+
{
|
1427 |
+
"epoch": 8.88,
|
1428 |
+
"grad_norm": 0.054795410484075546,
|
1429 |
+
"learning_rate": 0.00011367837338262476,
|
1430 |
+
"loss": 0.0021,
|
1431 |
+
"step": 4850
|
1432 |
+
},
|
1433 |
+
{
|
1434 |
+
"epoch": 8.93,
|
1435 |
+
"grad_norm": 0.02664073184132576,
|
1436 |
+
"learning_rate": 0.0001090573012939002,
|
1437 |
+
"loss": 0.0047,
|
1438 |
+
"step": 4875
|
1439 |
+
},
|
1440 |
+
{
|
1441 |
+
"epoch": 8.97,
|
1442 |
+
"grad_norm": 0.0008485654252581298,
|
1443 |
+
"learning_rate": 0.0001044362292051756,
|
1444 |
+
"loss": 0.0023,
|
1445 |
+
"step": 4900
|
1446 |
+
},
|
1447 |
+
{
|
1448 |
+
"epoch": 9.0,
|
1449 |
+
"eval_loss": 0.22940348088741302,
|
1450 |
+
"eval_runtime": 460.0411,
|
1451 |
+
"eval_samples_per_second": 1.767,
|
1452 |
+
"eval_steps_per_second": 0.296,
|
1453 |
+
"step": 4914
|
1454 |
+
},
|
1455 |
+
{
|
1456 |
+
"epoch": 9.02,
|
1457 |
+
"grad_norm": 0.03195321932435036,
|
1458 |
+
"learning_rate": 9.981515711645101e-05,
|
1459 |
+
"loss": 0.0034,
|
1460 |
+
"step": 4925
|
1461 |
+
},
|
1462 |
+
{
|
1463 |
+
"epoch": 9.07,
|
1464 |
+
"grad_norm": 0.0012715512420982122,
|
1465 |
+
"learning_rate": 9.519408502772643e-05,
|
1466 |
+
"loss": 0.0022,
|
1467 |
+
"step": 4950
|
1468 |
+
},
|
1469 |
+
{
|
1470 |
+
"epoch": 9.11,
|
1471 |
+
"grad_norm": 0.010319654829800129,
|
1472 |
+
"learning_rate": 9.057301293900184e-05,
|
1473 |
+
"loss": 0.0024,
|
1474 |
+
"step": 4975
|
1475 |
+
},
|
1476 |
+
{
|
1477 |
+
"epoch": 9.16,
|
1478 |
+
"grad_norm": 0.003004108089953661,
|
1479 |
+
"learning_rate": 8.595194085027728e-05,
|
1480 |
+
"loss": 0.0043,
|
1481 |
+
"step": 5000
|
1482 |
+
},
|
1483 |
+
{
|
1484 |
+
"epoch": 9.2,
|
1485 |
+
"grad_norm": 0.0019885245710611343,
|
1486 |
+
"learning_rate": 8.133086876155268e-05,
|
1487 |
+
"loss": 0.0022,
|
1488 |
+
"step": 5025
|
1489 |
+
},
|
1490 |
+
{
|
1491 |
+
"epoch": 9.25,
|
1492 |
+
"grad_norm": 0.06616940349340439,
|
1493 |
+
"learning_rate": 7.67097966728281e-05,
|
1494 |
+
"loss": 0.0022,
|
1495 |
+
"step": 5050
|
1496 |
+
},
|
1497 |
+
{
|
1498 |
+
"epoch": 9.29,
|
1499 |
+
"grad_norm": 0.044188376516103745,
|
1500 |
+
"learning_rate": 7.208872458410351e-05,
|
1501 |
+
"loss": 0.0021,
|
1502 |
+
"step": 5075
|
1503 |
+
},
|
1504 |
+
{
|
1505 |
+
"epoch": 9.34,
|
1506 |
+
"grad_norm": 0.07102042436599731,
|
1507 |
+
"learning_rate": 6.746765249537892e-05,
|
1508 |
+
"loss": 0.0023,
|
1509 |
+
"step": 5100
|
1510 |
+
},
|
1511 |
+
{
|
1512 |
+
"epoch": 9.39,
|
1513 |
+
"grad_norm": 0.018956031650304794,
|
1514 |
+
"learning_rate": 6.284658040665435e-05,
|
1515 |
+
"loss": 0.002,
|
1516 |
+
"step": 5125
|
1517 |
+
},
|
1518 |
+
{
|
1519 |
+
"epoch": 9.43,
|
1520 |
+
"grad_norm": 0.007081813644617796,
|
1521 |
+
"learning_rate": 5.822550831792976e-05,
|
1522 |
+
"loss": 0.0012,
|
1523 |
+
"step": 5150
|
1524 |
+
},
|
1525 |
+
{
|
1526 |
+
"epoch": 9.48,
|
1527 |
+
"grad_norm": 0.004433480557054281,
|
1528 |
+
"learning_rate": 5.3604436229205174e-05,
|
1529 |
+
"loss": 0.002,
|
1530 |
+
"step": 5175
|
1531 |
+
},
|
1532 |
+
{
|
1533 |
+
"epoch": 9.52,
|
1534 |
+
"grad_norm": 0.0015681196236982942,
|
1535 |
+
"learning_rate": 4.8983364140480595e-05,
|
1536 |
+
"loss": 0.002,
|
1537 |
+
"step": 5200
|
1538 |
+
},
|
1539 |
+
{
|
1540 |
+
"epoch": 9.57,
|
1541 |
+
"grad_norm": 0.003421030705794692,
|
1542 |
+
"learning_rate": 4.436229205175601e-05,
|
1543 |
+
"loss": 0.0018,
|
1544 |
+
"step": 5225
|
1545 |
+
},
|
1546 |
+
{
|
1547 |
+
"epoch": 9.62,
|
1548 |
+
"grad_norm": 0.13036024570465088,
|
1549 |
+
"learning_rate": 3.974121996303143e-05,
|
1550 |
+
"loss": 0.0016,
|
1551 |
+
"step": 5250
|
1552 |
+
},
|
1553 |
+
{
|
1554 |
+
"epoch": 9.66,
|
1555 |
+
"grad_norm": 0.05646170675754547,
|
1556 |
+
"learning_rate": 3.5120147874306844e-05,
|
1557 |
+
"loss": 0.0017,
|
1558 |
+
"step": 5275
|
1559 |
+
},
|
1560 |
+
{
|
1561 |
+
"epoch": 9.71,
|
1562 |
+
"grad_norm": 0.11519595235586166,
|
1563 |
+
"learning_rate": 3.0499075785582258e-05,
|
1564 |
+
"loss": 0.0041,
|
1565 |
+
"step": 5300
|
1566 |
+
},
|
1567 |
+
{
|
1568 |
+
"epoch": 9.75,
|
1569 |
+
"grad_norm": 0.08969979733228683,
|
1570 |
+
"learning_rate": 2.5878003696857672e-05,
|
1571 |
+
"loss": 0.0022,
|
1572 |
+
"step": 5325
|
1573 |
+
},
|
1574 |
+
{
|
1575 |
+
"epoch": 9.8,
|
1576 |
+
"grad_norm": 0.13669085502624512,
|
1577 |
+
"learning_rate": 2.1256931608133086e-05,
|
1578 |
+
"loss": 0.0014,
|
1579 |
+
"step": 5350
|
1580 |
+
},
|
1581 |
+
{
|
1582 |
+
"epoch": 9.84,
|
1583 |
+
"grad_norm": 0.13668496906757355,
|
1584 |
+
"learning_rate": 1.6635859519408503e-05,
|
1585 |
+
"loss": 0.0021,
|
1586 |
+
"step": 5375
|
1587 |
+
},
|
1588 |
+
{
|
1589 |
+
"epoch": 9.89,
|
1590 |
+
"grad_norm": 0.018298327922821045,
|
1591 |
+
"learning_rate": 1.2014787430683919e-05,
|
1592 |
+
"loss": 0.0023,
|
1593 |
+
"step": 5400
|
1594 |
+
},
|
1595 |
+
{
|
1596 |
+
"epoch": 9.94,
|
1597 |
+
"grad_norm": 0.007483182940632105,
|
1598 |
+
"learning_rate": 7.393715341959335e-06,
|
1599 |
+
"loss": 0.0024,
|
1600 |
+
"step": 5425
|
1601 |
+
},
|
1602 |
+
{
|
1603 |
+
"epoch": 9.98,
|
1604 |
+
"grad_norm": 0.03663533180952072,
|
1605 |
+
"learning_rate": 2.7726432532347505e-06,
|
1606 |
+
"loss": 0.0024,
|
1607 |
+
"step": 5450
|
1608 |
+
},
|
1609 |
+
{
|
1610 |
+
"epoch": 10.0,
|
1611 |
+
"eval_loss": 0.23311161994934082,
|
1612 |
+
"eval_runtime": 467.6697,
|
1613 |
+
"eval_samples_per_second": 1.738,
|
1614 |
+
"eval_steps_per_second": 0.291,
|
1615 |
+
"step": 5460
|
1616 |
+
},
|
1617 |
+
{
|
1618 |
+
"epoch": 10.0,
|
1619 |
+
"step": 5460,
|
1620 |
+
"total_flos": 1.135723105419264e+20,
|
1621 |
+
"train_loss": 0.04421832180674096,
|
1622 |
+
"train_runtime": 28971.5279,
|
1623 |
+
"train_samples_per_second": 1.13,
|
1624 |
+
"train_steps_per_second": 0.188
|
1625 |
+
}
|
1626 |
+
],
|
1627 |
+
"logging_steps": 25,
|
1628 |
+
"max_steps": 5460,
|
1629 |
+
"num_input_tokens_seen": 0,
|
1630 |
+
"num_train_epochs": 10,
|
1631 |
+
"save_steps": 500,
|
1632 |
+
"total_flos": 1.135723105419264e+20,
|
1633 |
+
"train_batch_size": 6,
|
1634 |
+
"trial_name": null,
|
1635 |
+
"trial_params": null
|
1636 |
+
}
|
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 sha256:44d94fd419ab4b304255d4c0a82980578e4fa72b83b3f051c5ca0af108a4e5d1
|
3 |
+
size 5112
|