Model save
Browse files- README.md +78 -0
- adapter_model.safetensors +1 -1
- all_results.json +9 -0
- runs/Jun05_05-05-56_7bdd13775218/events.out.tfevents.1717563989.7bdd13775218.74978.0 +2 -2
- train_results.json +9 -0
- trainer_state.json +845 -0
README.md
ADDED
@@ -0,0 +1,78 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: gemma
|
3 |
+
library_name: peft
|
4 |
+
tags:
|
5 |
+
- trl
|
6 |
+
- sft
|
7 |
+
- generated_from_trainer
|
8 |
+
base_model: google/gemma-2b
|
9 |
+
datasets:
|
10 |
+
- generator
|
11 |
+
model-index:
|
12 |
+
- name: gemma2b-summarize-gemini1_5flash-32k
|
13 |
+
results: []
|
14 |
+
---
|
15 |
+
|
16 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
17 |
+
should probably proofread and complete it, then remove this comment. -->
|
18 |
+
|
19 |
+
# gemma2b-summarize-gemini1_5flash-32k
|
20 |
+
|
21 |
+
This model is a fine-tuned version of [google/gemma-2b](https://huggingface.co/google/gemma-2b) on the generator dataset.
|
22 |
+
It achieves the following results on the evaluation set:
|
23 |
+
- Loss: 2.5309
|
24 |
+
|
25 |
+
## Model description
|
26 |
+
|
27 |
+
More information needed
|
28 |
+
|
29 |
+
## Intended uses & limitations
|
30 |
+
|
31 |
+
More information needed
|
32 |
+
|
33 |
+
## Training and evaluation data
|
34 |
+
|
35 |
+
More information needed
|
36 |
+
|
37 |
+
## Training procedure
|
38 |
+
|
39 |
+
### Training hyperparameters
|
40 |
+
|
41 |
+
The following hyperparameters were used during training:
|
42 |
+
- learning_rate: 0.0002
|
43 |
+
- train_batch_size: 8
|
44 |
+
- eval_batch_size: 8
|
45 |
+
- seed: 42
|
46 |
+
- distributed_type: multi-GPU
|
47 |
+
- num_devices: 4
|
48 |
+
- gradient_accumulation_steps: 2
|
49 |
+
- total_train_batch_size: 64
|
50 |
+
- total_eval_batch_size: 32
|
51 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
52 |
+
- lr_scheduler_type: cosine
|
53 |
+
- lr_scheduler_warmup_ratio: 0.1
|
54 |
+
- num_epochs: 10
|
55 |
+
|
56 |
+
### Training results
|
57 |
+
|
58 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
59 |
+
|:-------------:|:-----:|:----:|:---------------:|
|
60 |
+
| 1.4893 | 1.0 | 52 | 2.5957 |
|
61 |
+
| 1.143 | 2.0 | 104 | 2.5094 |
|
62 |
+
| 1.0743 | 3.0 | 156 | 2.4955 |
|
63 |
+
| 1.0412 | 4.0 | 208 | 2.4910 |
|
64 |
+
| 1.0117 | 5.0 | 260 | 2.5050 |
|
65 |
+
| 0.9886 | 6.0 | 312 | 2.5159 |
|
66 |
+
| 0.9711 | 7.0 | 364 | 2.5234 |
|
67 |
+
| 0.9637 | 8.0 | 416 | 2.5294 |
|
68 |
+
| 0.9604 | 9.0 | 468 | 2.5299 |
|
69 |
+
| 0.9523 | 10.0 | 520 | 2.5309 |
|
70 |
+
|
71 |
+
|
72 |
+
### Framework versions
|
73 |
+
|
74 |
+
- PEFT 0.11.1
|
75 |
+
- Transformers 4.40.1
|
76 |
+
- Pytorch 2.2.0+cu121
|
77 |
+
- Datasets 2.19.2
|
78 |
+
- Tokenizers 0.19.1
|
adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 19644912
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:29c4b9695e07bfc8793431250e0d50f697c4f2a08946ed1aa0105002b6673179
|
3 |
size 19644912
|
all_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 10.0,
|
3 |
+
"total_flos": 4.072455240941568e+17,
|
4 |
+
"train_loss": 1.1447187717144305,
|
5 |
+
"train_runtime": 1794.4362,
|
6 |
+
"train_samples": 31676,
|
7 |
+
"train_samples_per_second": 18.541,
|
8 |
+
"train_steps_per_second": 0.29
|
9 |
+
}
|
runs/Jun05_05-05-56_7bdd13775218/events.out.tfevents.1717563989.7bdd13775218.74978.0
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:b02d877c4327e91933acd30dac7edae7ca338a2c56ac844b5d416e3f1a35a485
|
3 |
+
size 30556
|
train_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 10.0,
|
3 |
+
"total_flos": 4.072455240941568e+17,
|
4 |
+
"train_loss": 1.1447187717144305,
|
5 |
+
"train_runtime": 1794.4362,
|
6 |
+
"train_samples": 31676,
|
7 |
+
"train_samples_per_second": 18.541,
|
8 |
+
"train_steps_per_second": 0.29
|
9 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,845 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 10.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 520,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.019230769230769232,
|
13 |
+
"grad_norm": 2.859375,
|
14 |
+
"learning_rate": 3.846153846153847e-06,
|
15 |
+
"loss": 3.0191,
|
16 |
+
"step": 1
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.09615384615384616,
|
20 |
+
"grad_norm": 2.25,
|
21 |
+
"learning_rate": 1.923076923076923e-05,
|
22 |
+
"loss": 3.0172,
|
23 |
+
"step": 5
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.19230769230769232,
|
27 |
+
"grad_norm": 1.9453125,
|
28 |
+
"learning_rate": 3.846153846153846e-05,
|
29 |
+
"loss": 2.9954,
|
30 |
+
"step": 10
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.28846153846153844,
|
34 |
+
"grad_norm": 3.28125,
|
35 |
+
"learning_rate": 5.769230769230769e-05,
|
36 |
+
"loss": 2.8517,
|
37 |
+
"step": 15
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.38461538461538464,
|
41 |
+
"grad_norm": 2.328125,
|
42 |
+
"learning_rate": 7.692307692307693e-05,
|
43 |
+
"loss": 2.6083,
|
44 |
+
"step": 20
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.4807692307692308,
|
48 |
+
"grad_norm": 1.296875,
|
49 |
+
"learning_rate": 9.615384615384617e-05,
|
50 |
+
"loss": 2.3838,
|
51 |
+
"step": 25
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.5769230769230769,
|
55 |
+
"grad_norm": 8.6875,
|
56 |
+
"learning_rate": 0.00011538461538461538,
|
57 |
+
"loss": 2.1596,
|
58 |
+
"step": 30
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.6730769230769231,
|
62 |
+
"grad_norm": 0.90234375,
|
63 |
+
"learning_rate": 0.00013461538461538464,
|
64 |
+
"loss": 1.9352,
|
65 |
+
"step": 35
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.7692307692307693,
|
69 |
+
"grad_norm": 1.9375,
|
70 |
+
"learning_rate": 0.00015384615384615385,
|
71 |
+
"loss": 1.7534,
|
72 |
+
"step": 40
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.8653846153846154,
|
76 |
+
"grad_norm": 0.65625,
|
77 |
+
"learning_rate": 0.0001730769230769231,
|
78 |
+
"loss": 1.6127,
|
79 |
+
"step": 45
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.9615384615384616,
|
83 |
+
"grad_norm": 1.0,
|
84 |
+
"learning_rate": 0.00019230769230769233,
|
85 |
+
"loss": 1.4893,
|
86 |
+
"step": 50
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 1.0,
|
90 |
+
"eval_loss": 2.595663547515869,
|
91 |
+
"eval_runtime": 0.5487,
|
92 |
+
"eval_samples_per_second": 18.225,
|
93 |
+
"eval_steps_per_second": 1.823,
|
94 |
+
"step": 52
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"epoch": 1.0576923076923077,
|
98 |
+
"grad_norm": 0.37109375,
|
99 |
+
"learning_rate": 0.00019997972289848503,
|
100 |
+
"loss": 1.4008,
|
101 |
+
"step": 55
|
102 |
+
},
|
103 |
+
{
|
104 |
+
"epoch": 1.1538461538461537,
|
105 |
+
"grad_norm": 0.73828125,
|
106 |
+
"learning_rate": 0.00019985583705641418,
|
107 |
+
"loss": 1.3355,
|
108 |
+
"step": 60
|
109 |
+
},
|
110 |
+
{
|
111 |
+
"epoch": 1.25,
|
112 |
+
"grad_norm": 0.232421875,
|
113 |
+
"learning_rate": 0.00019961946980917456,
|
114 |
+
"loss": 1.2736,
|
115 |
+
"step": 65
|
116 |
+
},
|
117 |
+
{
|
118 |
+
"epoch": 1.3461538461538463,
|
119 |
+
"grad_norm": 0.443359375,
|
120 |
+
"learning_rate": 0.0001992708874098054,
|
121 |
+
"loss": 1.2402,
|
122 |
+
"step": 70
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 1.4423076923076923,
|
126 |
+
"grad_norm": 0.5,
|
127 |
+
"learning_rate": 0.0001988104825147528,
|
128 |
+
"loss": 1.2168,
|
129 |
+
"step": 75
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 1.5384615384615383,
|
133 |
+
"grad_norm": 0.58984375,
|
134 |
+
"learning_rate": 0.00019823877374156647,
|
135 |
+
"loss": 1.2052,
|
136 |
+
"step": 80
|
137 |
+
},
|
138 |
+
{
|
139 |
+
"epoch": 1.6346153846153846,
|
140 |
+
"grad_norm": 0.921875,
|
141 |
+
"learning_rate": 0.00019755640508470942,
|
142 |
+
"loss": 1.1812,
|
143 |
+
"step": 85
|
144 |
+
},
|
145 |
+
{
|
146 |
+
"epoch": 1.7307692307692308,
|
147 |
+
"grad_norm": 0.2578125,
|
148 |
+
"learning_rate": 0.00019676414519013781,
|
149 |
+
"loss": 1.1707,
|
150 |
+
"step": 90
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 1.8269230769230769,
|
154 |
+
"grad_norm": 0.82421875,
|
155 |
+
"learning_rate": 0.00019586288648946947,
|
156 |
+
"loss": 1.1671,
|
157 |
+
"step": 95
|
158 |
+
},
|
159 |
+
{
|
160 |
+
"epoch": 1.9230769230769231,
|
161 |
+
"grad_norm": 0.271484375,
|
162 |
+
"learning_rate": 0.00019485364419471454,
|
163 |
+
"loss": 1.143,
|
164 |
+
"step": 100
|
165 |
+
},
|
166 |
+
{
|
167 |
+
"epoch": 2.0,
|
168 |
+
"eval_loss": 2.5093514919281006,
|
169 |
+
"eval_runtime": 0.551,
|
170 |
+
"eval_samples_per_second": 18.149,
|
171 |
+
"eval_steps_per_second": 1.815,
|
172 |
+
"step": 104
|
173 |
+
},
|
174 |
+
{
|
175 |
+
"epoch": 2.019230769230769,
|
176 |
+
"grad_norm": 0.30078125,
|
177 |
+
"learning_rate": 0.00019373755515470254,
|
178 |
+
"loss": 1.1342,
|
179 |
+
"step": 105
|
180 |
+
},
|
181 |
+
{
|
182 |
+
"epoch": 2.1153846153846154,
|
183 |
+
"grad_norm": 0.416015625,
|
184 |
+
"learning_rate": 0.00019251587657449236,
|
185 |
+
"loss": 1.111,
|
186 |
+
"step": 110
|
187 |
+
},
|
188 |
+
{
|
189 |
+
"epoch": 2.2115384615384617,
|
190 |
+
"grad_norm": 0.3515625,
|
191 |
+
"learning_rate": 0.00019118998459920902,
|
192 |
+
"loss": 1.104,
|
193 |
+
"step": 115
|
194 |
+
},
|
195 |
+
{
|
196 |
+
"epoch": 2.3076923076923075,
|
197 |
+
"grad_norm": 0.36328125,
|
198 |
+
"learning_rate": 0.0001897613727639014,
|
199 |
+
"loss": 1.0995,
|
200 |
+
"step": 120
|
201 |
+
},
|
202 |
+
{
|
203 |
+
"epoch": 2.4038461538461537,
|
204 |
+
"grad_norm": 0.5625,
|
205 |
+
"learning_rate": 0.0001882316503111678,
|
206 |
+
"loss": 1.0938,
|
207 |
+
"step": 125
|
208 |
+
},
|
209 |
+
{
|
210 |
+
"epoch": 2.5,
|
211 |
+
"grad_norm": 0.86328125,
|
212 |
+
"learning_rate": 0.00018660254037844388,
|
213 |
+
"loss": 1.098,
|
214 |
+
"step": 130
|
215 |
+
},
|
216 |
+
{
|
217 |
+
"epoch": 2.5961538461538463,
|
218 |
+
"grad_norm": 0.54296875,
|
219 |
+
"learning_rate": 0.00018487587805699526,
|
220 |
+
"loss": 1.0819,
|
221 |
+
"step": 135
|
222 |
+
},
|
223 |
+
{
|
224 |
+
"epoch": 2.6923076923076925,
|
225 |
+
"grad_norm": 0.419921875,
|
226 |
+
"learning_rate": 0.00018305360832480117,
|
227 |
+
"loss": 1.0878,
|
228 |
+
"step": 140
|
229 |
+
},
|
230 |
+
{
|
231 |
+
"epoch": 2.7884615384615383,
|
232 |
+
"grad_norm": 0.416015625,
|
233 |
+
"learning_rate": 0.00018113778385565733,
|
234 |
+
"loss": 1.0805,
|
235 |
+
"step": 145
|
236 |
+
},
|
237 |
+
{
|
238 |
+
"epoch": 2.8846153846153846,
|
239 |
+
"grad_norm": 0.306640625,
|
240 |
+
"learning_rate": 0.0001791305627069662,
|
241 |
+
"loss": 1.0708,
|
242 |
+
"step": 150
|
243 |
+
},
|
244 |
+
{
|
245 |
+
"epoch": 2.980769230769231,
|
246 |
+
"grad_norm": 0.267578125,
|
247 |
+
"learning_rate": 0.00017703420588881946,
|
248 |
+
"loss": 1.0743,
|
249 |
+
"step": 155
|
250 |
+
},
|
251 |
+
{
|
252 |
+
"epoch": 3.0,
|
253 |
+
"eval_loss": 2.4954726696014404,
|
254 |
+
"eval_runtime": 0.693,
|
255 |
+
"eval_samples_per_second": 14.43,
|
256 |
+
"eval_steps_per_second": 1.443,
|
257 |
+
"step": 156
|
258 |
+
},
|
259 |
+
{
|
260 |
+
"epoch": 3.076923076923077,
|
261 |
+
"grad_norm": 0.3203125,
|
262 |
+
"learning_rate": 0.00017485107481711012,
|
263 |
+
"loss": 1.0613,
|
264 |
+
"step": 160
|
265 |
+
},
|
266 |
+
{
|
267 |
+
"epoch": 3.173076923076923,
|
268 |
+
"grad_norm": 0.267578125,
|
269 |
+
"learning_rate": 0.00017258362865354426,
|
270 |
+
"loss": 1.0387,
|
271 |
+
"step": 165
|
272 |
+
},
|
273 |
+
{
|
274 |
+
"epoch": 3.269230769230769,
|
275 |
+
"grad_norm": 0.287109375,
|
276 |
+
"learning_rate": 0.00017023442153554777,
|
277 |
+
"loss": 1.0335,
|
278 |
+
"step": 170
|
279 |
+
},
|
280 |
+
{
|
281 |
+
"epoch": 3.3653846153846154,
|
282 |
+
"grad_norm": 0.361328125,
|
283 |
+
"learning_rate": 0.0001678060996991891,
|
284 |
+
"loss": 1.0387,
|
285 |
+
"step": 175
|
286 |
+
},
|
287 |
+
{
|
288 |
+
"epoch": 3.4615384615384617,
|
289 |
+
"grad_norm": 0.318359375,
|
290 |
+
"learning_rate": 0.0001653013984983585,
|
291 |
+
"loss": 1.0532,
|
292 |
+
"step": 180
|
293 |
+
},
|
294 |
+
{
|
295 |
+
"epoch": 3.5576923076923075,
|
296 |
+
"grad_norm": 0.5,
|
297 |
+
"learning_rate": 0.00016272313932356162,
|
298 |
+
"loss": 1.0507,
|
299 |
+
"step": 185
|
300 |
+
},
|
301 |
+
{
|
302 |
+
"epoch": 3.6538461538461537,
|
303 |
+
"grad_norm": 0.376953125,
|
304 |
+
"learning_rate": 0.0001600742264237979,
|
305 |
+
"loss": 1.0327,
|
306 |
+
"step": 190
|
307 |
+
},
|
308 |
+
{
|
309 |
+
"epoch": 3.75,
|
310 |
+
"grad_norm": 0.30859375,
|
311 |
+
"learning_rate": 0.0001573576436351046,
|
312 |
+
"loss": 1.0336,
|
313 |
+
"step": 195
|
314 |
+
},
|
315 |
+
{
|
316 |
+
"epoch": 3.8461538461538463,
|
317 |
+
"grad_norm": 0.63671875,
|
318 |
+
"learning_rate": 0.00015457645101945046,
|
319 |
+
"loss": 1.0366,
|
320 |
+
"step": 200
|
321 |
+
},
|
322 |
+
{
|
323 |
+
"epoch": 3.9423076923076925,
|
324 |
+
"grad_norm": 0.40625,
|
325 |
+
"learning_rate": 0.00015173378141776568,
|
326 |
+
"loss": 1.0412,
|
327 |
+
"step": 205
|
328 |
+
},
|
329 |
+
{
|
330 |
+
"epoch": 4.0,
|
331 |
+
"eval_loss": 2.4909958839416504,
|
332 |
+
"eval_runtime": 0.5333,
|
333 |
+
"eval_samples_per_second": 18.75,
|
334 |
+
"eval_steps_per_second": 1.875,
|
335 |
+
"step": 208
|
336 |
+
},
|
337 |
+
{
|
338 |
+
"epoch": 4.038461538461538,
|
339 |
+
"grad_norm": 0.322265625,
|
340 |
+
"learning_rate": 0.00014883283692099112,
|
341 |
+
"loss": 1.0274,
|
342 |
+
"step": 210
|
343 |
+
},
|
344 |
+
{
|
345 |
+
"epoch": 4.134615384615385,
|
346 |
+
"grad_norm": 0.490234375,
|
347 |
+
"learning_rate": 0.00014587688526312143,
|
348 |
+
"loss": 1.0145,
|
349 |
+
"step": 215
|
350 |
+
},
|
351 |
+
{
|
352 |
+
"epoch": 4.230769230769231,
|
353 |
+
"grad_norm": 0.345703125,
|
354 |
+
"learning_rate": 0.00014286925614030542,
|
355 |
+
"loss": 1.0097,
|
356 |
+
"step": 220
|
357 |
+
},
|
358 |
+
{
|
359 |
+
"epoch": 4.326923076923077,
|
360 |
+
"grad_norm": 0.306640625,
|
361 |
+
"learning_rate": 0.0001398133374601501,
|
362 |
+
"loss": 1.0137,
|
363 |
+
"step": 225
|
364 |
+
},
|
365 |
+
{
|
366 |
+
"epoch": 4.423076923076923,
|
367 |
+
"grad_norm": 0.427734375,
|
368 |
+
"learning_rate": 0.00013671257152545277,
|
369 |
+
"loss": 1.0119,
|
370 |
+
"step": 230
|
371 |
+
},
|
372 |
+
{
|
373 |
+
"epoch": 4.519230769230769,
|
374 |
+
"grad_norm": 1.0625,
|
375 |
+
"learning_rate": 0.0001335704511566605,
|
376 |
+
"loss": 1.0112,
|
377 |
+
"step": 235
|
378 |
+
},
|
379 |
+
{
|
380 |
+
"epoch": 4.615384615384615,
|
381 |
+
"grad_norm": 0.53125,
|
382 |
+
"learning_rate": 0.0001303905157574247,
|
383 |
+
"loss": 1.0102,
|
384 |
+
"step": 240
|
385 |
+
},
|
386 |
+
{
|
387 |
+
"epoch": 4.711538461538462,
|
388 |
+
"grad_norm": 0.373046875,
|
389 |
+
"learning_rate": 0.00012717634732768243,
|
390 |
+
"loss": 1.0055,
|
391 |
+
"step": 245
|
392 |
+
},
|
393 |
+
{
|
394 |
+
"epoch": 4.8076923076923075,
|
395 |
+
"grad_norm": 0.36328125,
|
396 |
+
"learning_rate": 0.0001239315664287558,
|
397 |
+
"loss": 1.0024,
|
398 |
+
"step": 250
|
399 |
+
},
|
400 |
+
{
|
401 |
+
"epoch": 4.903846153846154,
|
402 |
+
"grad_norm": 0.326171875,
|
403 |
+
"learning_rate": 0.00012065982810501404,
|
404 |
+
"loss": 1.0077,
|
405 |
+
"step": 255
|
406 |
+
},
|
407 |
+
{
|
408 |
+
"epoch": 5.0,
|
409 |
+
"grad_norm": 0.328125,
|
410 |
+
"learning_rate": 0.00011736481776669306,
|
411 |
+
"loss": 1.0117,
|
412 |
+
"step": 260
|
413 |
+
},
|
414 |
+
{
|
415 |
+
"epoch": 5.0,
|
416 |
+
"eval_loss": 2.5050010681152344,
|
417 |
+
"eval_runtime": 0.5358,
|
418 |
+
"eval_samples_per_second": 18.663,
|
419 |
+
"eval_steps_per_second": 1.866,
|
420 |
+
"step": 260
|
421 |
+
},
|
422 |
+
{
|
423 |
+
"epoch": 5.096153846153846,
|
424 |
+
"grad_norm": 0.44921875,
|
425 |
+
"learning_rate": 0.00011405024703850929,
|
426 |
+
"loss": 0.9883,
|
427 |
+
"step": 265
|
428 |
+
},
|
429 |
+
{
|
430 |
+
"epoch": 5.1923076923076925,
|
431 |
+
"grad_norm": 0.478515625,
|
432 |
+
"learning_rate": 0.00011071984957874479,
|
433 |
+
"loss": 0.9915,
|
434 |
+
"step": 270
|
435 |
+
},
|
436 |
+
{
|
437 |
+
"epoch": 5.288461538461538,
|
438 |
+
"grad_norm": 0.45703125,
|
439 |
+
"learning_rate": 0.00010737737687351284,
|
440 |
+
"loss": 0.9871,
|
441 |
+
"step": 275
|
442 |
+
},
|
443 |
+
{
|
444 |
+
"epoch": 5.384615384615385,
|
445 |
+
"grad_norm": 0.3828125,
|
446 |
+
"learning_rate": 0.00010402659401094152,
|
447 |
+
"loss": 0.9854,
|
448 |
+
"step": 280
|
449 |
+
},
|
450 |
+
{
|
451 |
+
"epoch": 5.480769230769231,
|
452 |
+
"grad_norm": 0.36328125,
|
453 |
+
"learning_rate": 0.00010067127544003563,
|
454 |
+
"loss": 0.9884,
|
455 |
+
"step": 285
|
456 |
+
},
|
457 |
+
{
|
458 |
+
"epoch": 5.576923076923077,
|
459 |
+
"grad_norm": 0.380859375,
|
460 |
+
"learning_rate": 9.73152007189939e-05,
|
461 |
+
"loss": 0.9908,
|
462 |
+
"step": 290
|
463 |
+
},
|
464 |
+
{
|
465 |
+
"epoch": 5.673076923076923,
|
466 |
+
"grad_norm": 0.33203125,
|
467 |
+
"learning_rate": 9.396215025777139e-05,
|
468 |
+
"loss": 0.9891,
|
469 |
+
"step": 295
|
470 |
+
},
|
471 |
+
{
|
472 |
+
"epoch": 5.769230769230769,
|
473 |
+
"grad_norm": 0.30078125,
|
474 |
+
"learning_rate": 9.061590105968208e-05,
|
475 |
+
"loss": 0.982,
|
476 |
+
"step": 300
|
477 |
+
},
|
478 |
+
{
|
479 |
+
"epoch": 5.865384615384615,
|
480 |
+
"grad_norm": 0.404296875,
|
481 |
+
"learning_rate": 8.728022246683894e-05,
|
482 |
+
"loss": 0.9928,
|
483 |
+
"step": 305
|
484 |
+
},
|
485 |
+
{
|
486 |
+
"epoch": 5.961538461538462,
|
487 |
+
"grad_norm": 0.369140625,
|
488 |
+
"learning_rate": 8.395887191422397e-05,
|
489 |
+
"loss": 0.9886,
|
490 |
+
"step": 310
|
491 |
+
},
|
492 |
+
{
|
493 |
+
"epoch": 6.0,
|
494 |
+
"eval_loss": 2.5159404277801514,
|
495 |
+
"eval_runtime": 0.5352,
|
496 |
+
"eval_samples_per_second": 18.685,
|
497 |
+
"eval_steps_per_second": 1.869,
|
498 |
+
"step": 312
|
499 |
+
},
|
500 |
+
{
|
501 |
+
"epoch": 6.0576923076923075,
|
502 |
+
"grad_norm": 0.3046875,
|
503 |
+
"learning_rate": 8.065559069717088e-05,
|
504 |
+
"loss": 0.9754,
|
505 |
+
"step": 315
|
506 |
+
},
|
507 |
+
{
|
508 |
+
"epoch": 6.153846153846154,
|
509 |
+
"grad_norm": 0.345703125,
|
510 |
+
"learning_rate": 7.73740997570278e-05,
|
511 |
+
"loss": 0.9761,
|
512 |
+
"step": 320
|
513 |
+
},
|
514 |
+
{
|
515 |
+
"epoch": 6.25,
|
516 |
+
"grad_norm": 0.41796875,
|
517 |
+
"learning_rate": 7.411809548974792e-05,
|
518 |
+
"loss": 0.9699,
|
519 |
+
"step": 325
|
520 |
+
},
|
521 |
+
{
|
522 |
+
"epoch": 6.346153846153846,
|
523 |
+
"grad_norm": 0.421875,
|
524 |
+
"learning_rate": 7.089124558212871e-05,
|
525 |
+
"loss": 0.9879,
|
526 |
+
"step": 330
|
527 |
+
},
|
528 |
+
{
|
529 |
+
"epoch": 6.4423076923076925,
|
530 |
+
"grad_norm": 0.48046875,
|
531 |
+
"learning_rate": 6.769718488039023e-05,
|
532 |
+
"loss": 0.9686,
|
533 |
+
"step": 335
|
534 |
+
},
|
535 |
+
{
|
536 |
+
"epoch": 6.538461538461538,
|
537 |
+
"grad_norm": 0.4296875,
|
538 |
+
"learning_rate": 6.453951129574644e-05,
|
539 |
+
"loss": 0.9671,
|
540 |
+
"step": 340
|
541 |
+
},
|
542 |
+
{
|
543 |
+
"epoch": 6.634615384615385,
|
544 |
+
"grad_norm": 0.3359375,
|
545 |
+
"learning_rate": 6.142178175158149e-05,
|
546 |
+
"loss": 0.9746,
|
547 |
+
"step": 345
|
548 |
+
},
|
549 |
+
{
|
550 |
+
"epoch": 6.730769230769231,
|
551 |
+
"grad_norm": 0.322265625,
|
552 |
+
"learning_rate": 5.834750817679606e-05,
|
553 |
+
"loss": 0.9724,
|
554 |
+
"step": 350
|
555 |
+
},
|
556 |
+
{
|
557 |
+
"epoch": 6.826923076923077,
|
558 |
+
"grad_norm": 0.31640625,
|
559 |
+
"learning_rate": 5.5320153549837415e-05,
|
560 |
+
"loss": 0.9788,
|
561 |
+
"step": 355
|
562 |
+
},
|
563 |
+
{
|
564 |
+
"epoch": 6.923076923076923,
|
565 |
+
"grad_norm": 0.330078125,
|
566 |
+
"learning_rate": 5.234312799786921e-05,
|
567 |
+
"loss": 0.9711,
|
568 |
+
"step": 360
|
569 |
+
},
|
570 |
+
{
|
571 |
+
"epoch": 7.0,
|
572 |
+
"eval_loss": 2.5233545303344727,
|
573 |
+
"eval_runtime": 0.5403,
|
574 |
+
"eval_samples_per_second": 18.508,
|
575 |
+
"eval_steps_per_second": 1.851,
|
576 |
+
"step": 364
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 7.019230769230769,
|
580 |
+
"grad_norm": 0.353515625,
|
581 |
+
"learning_rate": 4.9419784955474524e-05,
|
582 |
+
"loss": 0.9709,
|
583 |
+
"step": 365
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 7.115384615384615,
|
587 |
+
"grad_norm": 0.3203125,
|
588 |
+
"learning_rate": 4.6553417387219886e-05,
|
589 |
+
"loss": 0.9553,
|
590 |
+
"step": 370
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 7.211538461538462,
|
594 |
+
"grad_norm": 0.310546875,
|
595 |
+
"learning_rate": 4.374725407833532e-05,
|
596 |
+
"loss": 0.9601,
|
597 |
+
"step": 375
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 7.3076923076923075,
|
601 |
+
"grad_norm": 0.30078125,
|
602 |
+
"learning_rate": 4.100445599768774e-05,
|
603 |
+
"loss": 0.9674,
|
604 |
+
"step": 380
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 7.403846153846154,
|
608 |
+
"grad_norm": 0.4140625,
|
609 |
+
"learning_rate": 3.832811273714569e-05,
|
610 |
+
"loss": 0.9669,
|
611 |
+
"step": 385
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 7.5,
|
615 |
+
"grad_norm": 0.37109375,
|
616 |
+
"learning_rate": 3.5721239031346066e-05,
|
617 |
+
"loss": 0.9633,
|
618 |
+
"step": 390
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 7.596153846153846,
|
622 |
+
"grad_norm": 0.328125,
|
623 |
+
"learning_rate": 3.318677136178228e-05,
|
624 |
+
"loss": 0.9557,
|
625 |
+
"step": 395
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 7.6923076923076925,
|
629 |
+
"grad_norm": 0.37109375,
|
630 |
+
"learning_rate": 3.072756464904006e-05,
|
631 |
+
"loss": 0.966,
|
632 |
+
"step": 400
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 7.788461538461538,
|
636 |
+
"grad_norm": 0.341796875,
|
637 |
+
"learning_rate": 2.8346389036906828e-05,
|
638 |
+
"loss": 0.9638,
|
639 |
+
"step": 405
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 7.884615384615385,
|
643 |
+
"grad_norm": 0.33984375,
|
644 |
+
"learning_rate": 2.6045926771976303e-05,
|
645 |
+
"loss": 0.9654,
|
646 |
+
"step": 410
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 7.980769230769231,
|
650 |
+
"grad_norm": 0.310546875,
|
651 |
+
"learning_rate": 2.382876918226409e-05,
|
652 |
+
"loss": 0.9637,
|
653 |
+
"step": 415
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 8.0,
|
657 |
+
"eval_loss": 2.5294220447540283,
|
658 |
+
"eval_runtime": 0.536,
|
659 |
+
"eval_samples_per_second": 18.656,
|
660 |
+
"eval_steps_per_second": 1.866,
|
661 |
+
"step": 416
|
662 |
+
},
|
663 |
+
{
|
664 |
+
"epoch": 8.076923076923077,
|
665 |
+
"grad_norm": 0.310546875,
|
666 |
+
"learning_rate": 2.1697413758237784e-05,
|
667 |
+
"loss": 0.9582,
|
668 |
+
"step": 420
|
669 |
+
},
|
670 |
+
{
|
671 |
+
"epoch": 8.173076923076923,
|
672 |
+
"grad_norm": 0.30859375,
|
673 |
+
"learning_rate": 1.965426133954854e-05,
|
674 |
+
"loss": 0.9599,
|
675 |
+
"step": 425
|
676 |
+
},
|
677 |
+
{
|
678 |
+
"epoch": 8.26923076923077,
|
679 |
+
"grad_norm": 0.30078125,
|
680 |
+
"learning_rate": 1.7701613410634365e-05,
|
681 |
+
"loss": 0.959,
|
682 |
+
"step": 430
|
683 |
+
},
|
684 |
+
{
|
685 |
+
"epoch": 8.365384615384615,
|
686 |
+
"grad_norm": 0.314453125,
|
687 |
+
"learning_rate": 1.584166950824061e-05,
|
688 |
+
"loss": 0.9586,
|
689 |
+
"step": 435
|
690 |
+
},
|
691 |
+
{
|
692 |
+
"epoch": 8.461538461538462,
|
693 |
+
"grad_norm": 0.30859375,
|
694 |
+
"learning_rate": 1.4076524743778319e-05,
|
695 |
+
"loss": 0.9512,
|
696 |
+
"step": 440
|
697 |
+
},
|
698 |
+
{
|
699 |
+
"epoch": 8.557692307692308,
|
700 |
+
"grad_norm": 0.29296875,
|
701 |
+
"learning_rate": 1.2408167443311214e-05,
|
702 |
+
"loss": 0.9549,
|
703 |
+
"step": 445
|
704 |
+
},
|
705 |
+
{
|
706 |
+
"epoch": 8.653846153846153,
|
707 |
+
"grad_norm": 0.302734375,
|
708 |
+
"learning_rate": 1.083847690782972e-05,
|
709 |
+
"loss": 0.955,
|
710 |
+
"step": 450
|
711 |
+
},
|
712 |
+
{
|
713 |
+
"epoch": 8.75,
|
714 |
+
"grad_norm": 0.326171875,
|
715 |
+
"learning_rate": 9.369221296335006e-06,
|
716 |
+
"loss": 0.9558,
|
717 |
+
"step": 455
|
718 |
+
},
|
719 |
+
{
|
720 |
+
"epoch": 8.846153846153847,
|
721 |
+
"grad_norm": 0.310546875,
|
722 |
+
"learning_rate": 8.002055634117578e-06,
|
723 |
+
"loss": 0.9628,
|
724 |
+
"step": 460
|
725 |
+
},
|
726 |
+
{
|
727 |
+
"epoch": 8.942307692307692,
|
728 |
+
"grad_norm": 0.298828125,
|
729 |
+
"learning_rate": 6.738519948473976e-06,
|
730 |
+
"loss": 0.9604,
|
731 |
+
"step": 465
|
732 |
+
},
|
733 |
+
{
|
734 |
+
"epoch": 9.0,
|
735 |
+
"eval_loss": 2.5298619270324707,
|
736 |
+
"eval_runtime": 0.5355,
|
737 |
+
"eval_samples_per_second": 18.673,
|
738 |
+
"eval_steps_per_second": 1.867,
|
739 |
+
"step": 468
|
740 |
+
},
|
741 |
+
{
|
742 |
+
"epoch": 9.038461538461538,
|
743 |
+
"grad_norm": 0.2890625,
|
744 |
+
"learning_rate": 5.580037533961546e-06,
|
745 |
+
"loss": 0.9639,
|
746 |
+
"step": 470
|
747 |
+
},
|
748 |
+
{
|
749 |
+
"epoch": 9.134615384615385,
|
750 |
+
"grad_norm": 0.294921875,
|
751 |
+
"learning_rate": 4.527913349145441e-06,
|
752 |
+
"loss": 0.9503,
|
753 |
+
"step": 475
|
754 |
+
},
|
755 |
+
{
|
756 |
+
"epoch": 9.23076923076923,
|
757 |
+
"grad_norm": 0.296875,
|
758 |
+
"learning_rate": 3.5833325466437694e-06,
|
759 |
+
"loss": 0.9569,
|
760 |
+
"step": 480
|
761 |
+
},
|
762 |
+
{
|
763 |
+
"epoch": 9.326923076923077,
|
764 |
+
"grad_norm": 0.3046875,
|
765 |
+
"learning_rate": 2.7473591381266708e-06,
|
766 |
+
"loss": 0.9571,
|
767 |
+
"step": 485
|
768 |
+
},
|
769 |
+
{
|
770 |
+
"epoch": 9.423076923076923,
|
771 |
+
"grad_norm": 0.302734375,
|
772 |
+
"learning_rate": 2.0209347957732328e-06,
|
773 |
+
"loss": 0.9528,
|
774 |
+
"step": 490
|
775 |
+
},
|
776 |
+
{
|
777 |
+
"epoch": 9.51923076923077,
|
778 |
+
"grad_norm": 0.30859375,
|
779 |
+
"learning_rate": 1.404877791536141e-06,
|
780 |
+
"loss": 0.9599,
|
781 |
+
"step": 495
|
782 |
+
},
|
783 |
+
{
|
784 |
+
"epoch": 9.615384615384615,
|
785 |
+
"grad_norm": 0.3125,
|
786 |
+
"learning_rate": 8.998820754091531e-07,
|
787 |
+
"loss": 0.956,
|
788 |
+
"step": 500
|
789 |
+
},
|
790 |
+
{
|
791 |
+
"epoch": 9.711538461538462,
|
792 |
+
"grad_norm": 0.30078125,
|
793 |
+
"learning_rate": 5.065164937354428e-07,
|
794 |
+
"loss": 0.9578,
|
795 |
+
"step": 505
|
796 |
+
},
|
797 |
+
{
|
798 |
+
"epoch": 9.807692307692308,
|
799 |
+
"grad_norm": 0.314453125,
|
800 |
+
"learning_rate": 2.2522414843748618e-07,
|
801 |
+
"loss": 0.9612,
|
802 |
+
"step": 510
|
803 |
+
},
|
804 |
+
{
|
805 |
+
"epoch": 9.903846153846153,
|
806 |
+
"grad_norm": 0.29296875,
|
807 |
+
"learning_rate": 5.632189789027687e-08,
|
808 |
+
"loss": 0.9593,
|
809 |
+
"step": 515
|
810 |
+
},
|
811 |
+
{
|
812 |
+
"epoch": 10.0,
|
813 |
+
"grad_norm": 0.298828125,
|
814 |
+
"learning_rate": 0.0,
|
815 |
+
"loss": 0.9523,
|
816 |
+
"step": 520
|
817 |
+
},
|
818 |
+
{
|
819 |
+
"epoch": 10.0,
|
820 |
+
"eval_loss": 2.5308985710144043,
|
821 |
+
"eval_runtime": 0.5343,
|
822 |
+
"eval_samples_per_second": 18.716,
|
823 |
+
"eval_steps_per_second": 1.872,
|
824 |
+
"step": 520
|
825 |
+
},
|
826 |
+
{
|
827 |
+
"epoch": 10.0,
|
828 |
+
"step": 520,
|
829 |
+
"total_flos": 4.072455240941568e+17,
|
830 |
+
"train_loss": 1.1447187717144305,
|
831 |
+
"train_runtime": 1794.4362,
|
832 |
+
"train_samples_per_second": 18.541,
|
833 |
+
"train_steps_per_second": 0.29
|
834 |
+
}
|
835 |
+
],
|
836 |
+
"logging_steps": 5,
|
837 |
+
"max_steps": 520,
|
838 |
+
"num_input_tokens_seen": 0,
|
839 |
+
"num_train_epochs": 10,
|
840 |
+
"save_steps": 100,
|
841 |
+
"total_flos": 4.072455240941568e+17,
|
842 |
+
"train_batch_size": 8,
|
843 |
+
"trial_name": null,
|
844 |
+
"trial_params": null
|
845 |
+
}
|