GaetanMichelet
commited on
Commit
•
f69bf6e
1
Parent(s):
d1c4122
Model save
Browse files- README.md +81 -0
- adapter_model.safetensors +1 -1
- all_results.json +9 -0
- train_results.json +9 -0
- trainer_state.json +698 -0
README.md
ADDED
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
base_model: google/gemma-2-2b-it
|
3 |
+
library_name: peft
|
4 |
+
license: gemma
|
5 |
+
tags:
|
6 |
+
- trl
|
7 |
+
- sft
|
8 |
+
- generated_from_trainer
|
9 |
+
model-index:
|
10 |
+
- name: Gemma-2-2B_task-2_180-samples_config-2_full
|
11 |
+
results: []
|
12 |
+
---
|
13 |
+
|
14 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
15 |
+
should probably proofread and complete it, then remove this comment. -->
|
16 |
+
|
17 |
+
# Gemma-2-2B_task-2_180-samples_config-2_full
|
18 |
+
|
19 |
+
This model is a fine-tuned version of [google/gemma-2-2b-it](https://huggingface.co/google/gemma-2-2b-it) on the None dataset.
|
20 |
+
It achieves the following results on the evaluation set:
|
21 |
+
- Loss: 1.0435
|
22 |
+
|
23 |
+
## Model description
|
24 |
+
|
25 |
+
More information needed
|
26 |
+
|
27 |
+
## Intended uses & limitations
|
28 |
+
|
29 |
+
More information needed
|
30 |
+
|
31 |
+
## Training and evaluation data
|
32 |
+
|
33 |
+
More information needed
|
34 |
+
|
35 |
+
## Training procedure
|
36 |
+
|
37 |
+
### Training hyperparameters
|
38 |
+
|
39 |
+
The following hyperparameters were used during training:
|
40 |
+
- learning_rate: 0.0001
|
41 |
+
- train_batch_size: 1
|
42 |
+
- eval_batch_size: 1
|
43 |
+
- seed: 42
|
44 |
+
- distributed_type: multi-GPU
|
45 |
+
- gradient_accumulation_steps: 16
|
46 |
+
- total_train_batch_size: 16
|
47 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
48 |
+
- lr_scheduler_type: cosine
|
49 |
+
- lr_scheduler_warmup_ratio: 0.1
|
50 |
+
- num_epochs: 50
|
51 |
+
|
52 |
+
### Training results
|
53 |
+
|
54 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
55 |
+
|:-------------:|:-------:|:----:|:---------------:|
|
56 |
+
| 1.3182 | 0.9412 | 8 | 1.3144 |
|
57 |
+
| 1.2253 | 2.0 | 17 | 1.1693 |
|
58 |
+
| 1.1196 | 2.9412 | 25 | 1.0760 |
|
59 |
+
| 0.9289 | 4.0 | 34 | 0.9662 |
|
60 |
+
| 0.8623 | 4.9412 | 42 | 0.9212 |
|
61 |
+
| 0.8786 | 6.0 | 51 | 0.8973 |
|
62 |
+
| 0.861 | 6.9412 | 59 | 0.8845 |
|
63 |
+
| 0.8112 | 8.0 | 68 | 0.8751 |
|
64 |
+
| 0.7491 | 8.9412 | 76 | 0.8714 |
|
65 |
+
| 0.7439 | 10.0 | 85 | 0.8710 |
|
66 |
+
| 0.7136 | 10.9412 | 93 | 0.8754 |
|
67 |
+
| 0.6546 | 12.0 | 102 | 0.8866 |
|
68 |
+
| 0.6539 | 12.9412 | 110 | 0.9093 |
|
69 |
+
| 0.632 | 14.0 | 119 | 0.9347 |
|
70 |
+
| 0.543 | 14.9412 | 127 | 0.9885 |
|
71 |
+
| 0.4869 | 16.0 | 136 | 1.0262 |
|
72 |
+
| 0.4926 | 16.9412 | 144 | 1.0435 |
|
73 |
+
|
74 |
+
|
75 |
+
### Framework versions
|
76 |
+
|
77 |
+
- PEFT 0.12.0
|
78 |
+
- Transformers 4.44.0
|
79 |
+
- Pytorch 2.1.2+cu121
|
80 |
+
- Datasets 2.20.0
|
81 |
+
- 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 83115256
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:20429bbc1fc4821378a59277ddd8c5994a360d4355d934e7cab9190550fff6e1
|
3 |
size 83115256
|
all_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 16.941176470588236,
|
3 |
+
"total_flos": 7.574414567879475e+16,
|
4 |
+
"train_loss": 0.8188091479241848,
|
5 |
+
"train_runtime": 3221.0584,
|
6 |
+
"train_samples": 136,
|
7 |
+
"train_samples_per_second": 2.111,
|
8 |
+
"train_steps_per_second": 0.124
|
9 |
+
}
|
train_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 16.941176470588236,
|
3 |
+
"total_flos": 7.574414567879475e+16,
|
4 |
+
"train_loss": 0.8188091479241848,
|
5 |
+
"train_runtime": 3221.0584,
|
6 |
+
"train_samples": 136,
|
7 |
+
"train_samples_per_second": 2.111,
|
8 |
+
"train_steps_per_second": 0.124
|
9 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,698 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.8710008263587952,
|
3 |
+
"best_model_checkpoint": "data/Gemma-2-2B_task-2_180-samples_config-2_full/checkpoint-85",
|
4 |
+
"epoch": 16.941176470588236,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 144,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.11764705882352941,
|
13 |
+
"grad_norm": 0.4646625816822052,
|
14 |
+
"learning_rate": 2.5e-06,
|
15 |
+
"loss": 1.3348,
|
16 |
+
"step": 1
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.23529411764705882,
|
20 |
+
"grad_norm": 0.4941255748271942,
|
21 |
+
"learning_rate": 5e-06,
|
22 |
+
"loss": 1.3615,
|
23 |
+
"step": 2
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.47058823529411764,
|
27 |
+
"grad_norm": 0.4547555446624756,
|
28 |
+
"learning_rate": 1e-05,
|
29 |
+
"loss": 1.3619,
|
30 |
+
"step": 4
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.7058823529411765,
|
34 |
+
"grad_norm": 0.4238773584365845,
|
35 |
+
"learning_rate": 1.5e-05,
|
36 |
+
"loss": 1.2933,
|
37 |
+
"step": 6
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.9411764705882353,
|
41 |
+
"grad_norm": 0.41738811135292053,
|
42 |
+
"learning_rate": 2e-05,
|
43 |
+
"loss": 1.3182,
|
44 |
+
"step": 8
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.9411764705882353,
|
48 |
+
"eval_loss": 1.314428687095642,
|
49 |
+
"eval_runtime": 13.9271,
|
50 |
+
"eval_samples_per_second": 2.585,
|
51 |
+
"eval_steps_per_second": 2.585,
|
52 |
+
"step": 8
|
53 |
+
},
|
54 |
+
{
|
55 |
+
"epoch": 1.1764705882352942,
|
56 |
+
"grad_norm": 0.46454063057899475,
|
57 |
+
"learning_rate": 2.5e-05,
|
58 |
+
"loss": 1.3438,
|
59 |
+
"step": 10
|
60 |
+
},
|
61 |
+
{
|
62 |
+
"epoch": 1.4117647058823528,
|
63 |
+
"grad_norm": 0.3903467655181885,
|
64 |
+
"learning_rate": 3e-05,
|
65 |
+
"loss": 1.2762,
|
66 |
+
"step": 12
|
67 |
+
},
|
68 |
+
{
|
69 |
+
"epoch": 1.6470588235294117,
|
70 |
+
"grad_norm": 0.35459548234939575,
|
71 |
+
"learning_rate": 3.5e-05,
|
72 |
+
"loss": 1.2376,
|
73 |
+
"step": 14
|
74 |
+
},
|
75 |
+
{
|
76 |
+
"epoch": 1.8823529411764706,
|
77 |
+
"grad_norm": 0.3327643573284149,
|
78 |
+
"learning_rate": 4e-05,
|
79 |
+
"loss": 1.2253,
|
80 |
+
"step": 16
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 2.0,
|
84 |
+
"eval_loss": 1.1693224906921387,
|
85 |
+
"eval_runtime": 13.922,
|
86 |
+
"eval_samples_per_second": 2.586,
|
87 |
+
"eval_steps_per_second": 2.586,
|
88 |
+
"step": 17
|
89 |
+
},
|
90 |
+
{
|
91 |
+
"epoch": 2.1176470588235294,
|
92 |
+
"grad_norm": 0.26664429903030396,
|
93 |
+
"learning_rate": 4.5e-05,
|
94 |
+
"loss": 1.1858,
|
95 |
+
"step": 18
|
96 |
+
},
|
97 |
+
{
|
98 |
+
"epoch": 2.3529411764705883,
|
99 |
+
"grad_norm": 0.22536298632621765,
|
100 |
+
"learning_rate": 5e-05,
|
101 |
+
"loss": 1.1146,
|
102 |
+
"step": 20
|
103 |
+
},
|
104 |
+
{
|
105 |
+
"epoch": 2.588235294117647,
|
106 |
+
"grad_norm": 0.29759112000465393,
|
107 |
+
"learning_rate": 5.500000000000001e-05,
|
108 |
+
"loss": 1.1207,
|
109 |
+
"step": 22
|
110 |
+
},
|
111 |
+
{
|
112 |
+
"epoch": 2.8235294117647056,
|
113 |
+
"grad_norm": 0.2725456655025482,
|
114 |
+
"learning_rate": 6e-05,
|
115 |
+
"loss": 1.1196,
|
116 |
+
"step": 24
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"epoch": 2.9411764705882355,
|
120 |
+
"eval_loss": 1.0760149955749512,
|
121 |
+
"eval_runtime": 13.9219,
|
122 |
+
"eval_samples_per_second": 2.586,
|
123 |
+
"eval_steps_per_second": 2.586,
|
124 |
+
"step": 25
|
125 |
+
},
|
126 |
+
{
|
127 |
+
"epoch": 3.0588235294117645,
|
128 |
+
"grad_norm": 0.2560402452945709,
|
129 |
+
"learning_rate": 6.500000000000001e-05,
|
130 |
+
"loss": 1.052,
|
131 |
+
"step": 26
|
132 |
+
},
|
133 |
+
{
|
134 |
+
"epoch": 3.2941176470588234,
|
135 |
+
"grad_norm": 0.21541963517665863,
|
136 |
+
"learning_rate": 7e-05,
|
137 |
+
"loss": 1.023,
|
138 |
+
"step": 28
|
139 |
+
},
|
140 |
+
{
|
141 |
+
"epoch": 3.5294117647058822,
|
142 |
+
"grad_norm": 0.23459596931934357,
|
143 |
+
"learning_rate": 7.500000000000001e-05,
|
144 |
+
"loss": 1.0138,
|
145 |
+
"step": 30
|
146 |
+
},
|
147 |
+
{
|
148 |
+
"epoch": 3.764705882352941,
|
149 |
+
"grad_norm": 0.2933233678340912,
|
150 |
+
"learning_rate": 8e-05,
|
151 |
+
"loss": 1.0071,
|
152 |
+
"step": 32
|
153 |
+
},
|
154 |
+
{
|
155 |
+
"epoch": 4.0,
|
156 |
+
"grad_norm": 0.2300327569246292,
|
157 |
+
"learning_rate": 8.5e-05,
|
158 |
+
"loss": 0.9289,
|
159 |
+
"step": 34
|
160 |
+
},
|
161 |
+
{
|
162 |
+
"epoch": 4.0,
|
163 |
+
"eval_loss": 0.9661836624145508,
|
164 |
+
"eval_runtime": 13.9237,
|
165 |
+
"eval_samples_per_second": 2.586,
|
166 |
+
"eval_steps_per_second": 2.586,
|
167 |
+
"step": 34
|
168 |
+
},
|
169 |
+
{
|
170 |
+
"epoch": 4.235294117647059,
|
171 |
+
"grad_norm": 0.209079310297966,
|
172 |
+
"learning_rate": 9e-05,
|
173 |
+
"loss": 0.9432,
|
174 |
+
"step": 36
|
175 |
+
},
|
176 |
+
{
|
177 |
+
"epoch": 4.470588235294118,
|
178 |
+
"grad_norm": 0.168331116437912,
|
179 |
+
"learning_rate": 9.5e-05,
|
180 |
+
"loss": 0.9028,
|
181 |
+
"step": 38
|
182 |
+
},
|
183 |
+
{
|
184 |
+
"epoch": 4.705882352941177,
|
185 |
+
"grad_norm": 0.14146201312541962,
|
186 |
+
"learning_rate": 0.0001,
|
187 |
+
"loss": 0.8951,
|
188 |
+
"step": 40
|
189 |
+
},
|
190 |
+
{
|
191 |
+
"epoch": 4.9411764705882355,
|
192 |
+
"grad_norm": 0.15306784212589264,
|
193 |
+
"learning_rate": 9.999238475781957e-05,
|
194 |
+
"loss": 0.8623,
|
195 |
+
"step": 42
|
196 |
+
},
|
197 |
+
{
|
198 |
+
"epoch": 4.9411764705882355,
|
199 |
+
"eval_loss": 0.9211580753326416,
|
200 |
+
"eval_runtime": 13.9243,
|
201 |
+
"eval_samples_per_second": 2.585,
|
202 |
+
"eval_steps_per_second": 2.585,
|
203 |
+
"step": 42
|
204 |
+
},
|
205 |
+
{
|
206 |
+
"epoch": 5.176470588235294,
|
207 |
+
"grad_norm": 0.1463106870651245,
|
208 |
+
"learning_rate": 9.99695413509548e-05,
|
209 |
+
"loss": 0.8416,
|
210 |
+
"step": 44
|
211 |
+
},
|
212 |
+
{
|
213 |
+
"epoch": 5.411764705882353,
|
214 |
+
"grad_norm": 0.14740553498268127,
|
215 |
+
"learning_rate": 9.99314767377287e-05,
|
216 |
+
"loss": 0.8785,
|
217 |
+
"step": 46
|
218 |
+
},
|
219 |
+
{
|
220 |
+
"epoch": 5.647058823529412,
|
221 |
+
"grad_norm": 0.1291574090719223,
|
222 |
+
"learning_rate": 9.987820251299122e-05,
|
223 |
+
"loss": 0.8335,
|
224 |
+
"step": 48
|
225 |
+
},
|
226 |
+
{
|
227 |
+
"epoch": 5.882352941176471,
|
228 |
+
"grad_norm": 0.14685429632663727,
|
229 |
+
"learning_rate": 9.980973490458728e-05,
|
230 |
+
"loss": 0.8786,
|
231 |
+
"step": 50
|
232 |
+
},
|
233 |
+
{
|
234 |
+
"epoch": 6.0,
|
235 |
+
"eval_loss": 0.8972748517990112,
|
236 |
+
"eval_runtime": 13.9294,
|
237 |
+
"eval_samples_per_second": 2.584,
|
238 |
+
"eval_steps_per_second": 2.584,
|
239 |
+
"step": 51
|
240 |
+
},
|
241 |
+
{
|
242 |
+
"epoch": 6.117647058823529,
|
243 |
+
"grad_norm": 0.13178876042366028,
|
244 |
+
"learning_rate": 9.972609476841367e-05,
|
245 |
+
"loss": 0.8167,
|
246 |
+
"step": 52
|
247 |
+
},
|
248 |
+
{
|
249 |
+
"epoch": 6.352941176470588,
|
250 |
+
"grad_norm": 0.12248564511537552,
|
251 |
+
"learning_rate": 9.962730758206611e-05,
|
252 |
+
"loss": 0.8085,
|
253 |
+
"step": 54
|
254 |
+
},
|
255 |
+
{
|
256 |
+
"epoch": 6.588235294117647,
|
257 |
+
"grad_norm": 0.12658923864364624,
|
258 |
+
"learning_rate": 9.951340343707852e-05,
|
259 |
+
"loss": 0.8299,
|
260 |
+
"step": 56
|
261 |
+
},
|
262 |
+
{
|
263 |
+
"epoch": 6.823529411764706,
|
264 |
+
"grad_norm": 0.12179239839315414,
|
265 |
+
"learning_rate": 9.938441702975689e-05,
|
266 |
+
"loss": 0.861,
|
267 |
+
"step": 58
|
268 |
+
},
|
269 |
+
{
|
270 |
+
"epoch": 6.9411764705882355,
|
271 |
+
"eval_loss": 0.8844776749610901,
|
272 |
+
"eval_runtime": 13.9268,
|
273 |
+
"eval_samples_per_second": 2.585,
|
274 |
+
"eval_steps_per_second": 2.585,
|
275 |
+
"step": 59
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 7.0588235294117645,
|
279 |
+
"grad_norm": 0.13848571479320526,
|
280 |
+
"learning_rate": 9.924038765061042e-05,
|
281 |
+
"loss": 0.8375,
|
282 |
+
"step": 60
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 7.294117647058823,
|
286 |
+
"grad_norm": 0.13029764592647552,
|
287 |
+
"learning_rate": 9.908135917238321e-05,
|
288 |
+
"loss": 0.8097,
|
289 |
+
"step": 62
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 7.529411764705882,
|
293 |
+
"grad_norm": 0.1324029564857483,
|
294 |
+
"learning_rate": 9.890738003669029e-05,
|
295 |
+
"loss": 0.7823,
|
296 |
+
"step": 64
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 7.764705882352941,
|
300 |
+
"grad_norm": 0.13409382104873657,
|
301 |
+
"learning_rate": 9.871850323926177e-05,
|
302 |
+
"loss": 0.774,
|
303 |
+
"step": 66
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 8.0,
|
307 |
+
"grad_norm": 0.1507219523191452,
|
308 |
+
"learning_rate": 9.851478631379982e-05,
|
309 |
+
"loss": 0.8112,
|
310 |
+
"step": 68
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 8.0,
|
314 |
+
"eval_loss": 0.8751078248023987,
|
315 |
+
"eval_runtime": 13.9264,
|
316 |
+
"eval_samples_per_second": 2.585,
|
317 |
+
"eval_steps_per_second": 2.585,
|
318 |
+
"step": 68
|
319 |
+
},
|
320 |
+
{
|
321 |
+
"epoch": 8.235294117647058,
|
322 |
+
"grad_norm": 0.13311730325222015,
|
323 |
+
"learning_rate": 9.829629131445342e-05,
|
324 |
+
"loss": 0.7713,
|
325 |
+
"step": 70
|
326 |
+
},
|
327 |
+
{
|
328 |
+
"epoch": 8.470588235294118,
|
329 |
+
"grad_norm": 0.1371852308511734,
|
330 |
+
"learning_rate": 9.806308479691595e-05,
|
331 |
+
"loss": 0.7967,
|
332 |
+
"step": 72
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 8.705882352941176,
|
336 |
+
"grad_norm": 0.14112600684165955,
|
337 |
+
"learning_rate": 9.781523779815179e-05,
|
338 |
+
"loss": 0.7909,
|
339 |
+
"step": 74
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 8.941176470588236,
|
343 |
+
"grad_norm": 0.16504345834255219,
|
344 |
+
"learning_rate": 9.755282581475769e-05,
|
345 |
+
"loss": 0.7491,
|
346 |
+
"step": 76
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"epoch": 8.941176470588236,
|
350 |
+
"eval_loss": 0.8714024424552917,
|
351 |
+
"eval_runtime": 13.9243,
|
352 |
+
"eval_samples_per_second": 2.585,
|
353 |
+
"eval_steps_per_second": 2.585,
|
354 |
+
"step": 76
|
355 |
+
},
|
356 |
+
{
|
357 |
+
"epoch": 9.176470588235293,
|
358 |
+
"grad_norm": 0.17292000353336334,
|
359 |
+
"learning_rate": 9.727592877996585e-05,
|
360 |
+
"loss": 0.7597,
|
361 |
+
"step": 78
|
362 |
+
},
|
363 |
+
{
|
364 |
+
"epoch": 9.411764705882353,
|
365 |
+
"grad_norm": 0.1559344381093979,
|
366 |
+
"learning_rate": 9.698463103929542e-05,
|
367 |
+
"loss": 0.7566,
|
368 |
+
"step": 80
|
369 |
+
},
|
370 |
+
{
|
371 |
+
"epoch": 9.647058823529411,
|
372 |
+
"grad_norm": 0.17069514095783234,
|
373 |
+
"learning_rate": 9.667902132486009e-05,
|
374 |
+
"loss": 0.7534,
|
375 |
+
"step": 82
|
376 |
+
},
|
377 |
+
{
|
378 |
+
"epoch": 9.882352941176471,
|
379 |
+
"grad_norm": 0.1825936883687973,
|
380 |
+
"learning_rate": 9.635919272833938e-05,
|
381 |
+
"loss": 0.7439,
|
382 |
+
"step": 84
|
383 |
+
},
|
384 |
+
{
|
385 |
+
"epoch": 10.0,
|
386 |
+
"eval_loss": 0.8710008263587952,
|
387 |
+
"eval_runtime": 13.9329,
|
388 |
+
"eval_samples_per_second": 2.584,
|
389 |
+
"eval_steps_per_second": 2.584,
|
390 |
+
"step": 85
|
391 |
+
},
|
392 |
+
{
|
393 |
+
"epoch": 10.117647058823529,
|
394 |
+
"grad_norm": 0.17276407778263092,
|
395 |
+
"learning_rate": 9.602524267262203e-05,
|
396 |
+
"loss": 0.7186,
|
397 |
+
"step": 86
|
398 |
+
},
|
399 |
+
{
|
400 |
+
"epoch": 10.352941176470589,
|
401 |
+
"grad_norm": 0.20144543051719666,
|
402 |
+
"learning_rate": 9.567727288213005e-05,
|
403 |
+
"loss": 0.736,
|
404 |
+
"step": 88
|
405 |
+
},
|
406 |
+
{
|
407 |
+
"epoch": 10.588235294117647,
|
408 |
+
"grad_norm": 0.19740185141563416,
|
409 |
+
"learning_rate": 9.53153893518325e-05,
|
410 |
+
"loss": 0.7088,
|
411 |
+
"step": 90
|
412 |
+
},
|
413 |
+
{
|
414 |
+
"epoch": 10.823529411764707,
|
415 |
+
"grad_norm": 0.22668859362602234,
|
416 |
+
"learning_rate": 9.493970231495835e-05,
|
417 |
+
"loss": 0.7136,
|
418 |
+
"step": 92
|
419 |
+
},
|
420 |
+
{
|
421 |
+
"epoch": 10.941176470588236,
|
422 |
+
"eval_loss": 0.8754350543022156,
|
423 |
+
"eval_runtime": 13.9268,
|
424 |
+
"eval_samples_per_second": 2.585,
|
425 |
+
"eval_steps_per_second": 2.585,
|
426 |
+
"step": 93
|
427 |
+
},
|
428 |
+
{
|
429 |
+
"epoch": 11.058823529411764,
|
430 |
+
"grad_norm": 0.20473405718803406,
|
431 |
+
"learning_rate": 9.45503262094184e-05,
|
432 |
+
"loss": 0.7447,
|
433 |
+
"step": 94
|
434 |
+
},
|
435 |
+
{
|
436 |
+
"epoch": 11.294117647058824,
|
437 |
+
"grad_norm": 0.21799054741859436,
|
438 |
+
"learning_rate": 9.414737964294636e-05,
|
439 |
+
"loss": 0.7287,
|
440 |
+
"step": 96
|
441 |
+
},
|
442 |
+
{
|
443 |
+
"epoch": 11.529411764705882,
|
444 |
+
"grad_norm": 0.2471453845500946,
|
445 |
+
"learning_rate": 9.373098535696979e-05,
|
446 |
+
"loss": 0.6675,
|
447 |
+
"step": 98
|
448 |
+
},
|
449 |
+
{
|
450 |
+
"epoch": 11.764705882352942,
|
451 |
+
"grad_norm": 0.2599719762802124,
|
452 |
+
"learning_rate": 9.330127018922194e-05,
|
453 |
+
"loss": 0.6934,
|
454 |
+
"step": 100
|
455 |
+
},
|
456 |
+
{
|
457 |
+
"epoch": 12.0,
|
458 |
+
"grad_norm": 0.2612287104129791,
|
459 |
+
"learning_rate": 9.285836503510562e-05,
|
460 |
+
"loss": 0.6546,
|
461 |
+
"step": 102
|
462 |
+
},
|
463 |
+
{
|
464 |
+
"epoch": 12.0,
|
465 |
+
"eval_loss": 0.8866230845451355,
|
466 |
+
"eval_runtime": 13.9305,
|
467 |
+
"eval_samples_per_second": 2.584,
|
468 |
+
"eval_steps_per_second": 2.584,
|
469 |
+
"step": 102
|
470 |
+
},
|
471 |
+
{
|
472 |
+
"epoch": 12.235294117647058,
|
473 |
+
"grad_norm": 0.3070856034755707,
|
474 |
+
"learning_rate": 9.24024048078213e-05,
|
475 |
+
"loss": 0.6527,
|
476 |
+
"step": 104
|
477 |
+
},
|
478 |
+
{
|
479 |
+
"epoch": 12.470588235294118,
|
480 |
+
"grad_norm": 0.3168947100639343,
|
481 |
+
"learning_rate": 9.193352839727121e-05,
|
482 |
+
"loss": 0.6387,
|
483 |
+
"step": 106
|
484 |
+
},
|
485 |
+
{
|
486 |
+
"epoch": 12.705882352941176,
|
487 |
+
"grad_norm": 0.3805824816226959,
|
488 |
+
"learning_rate": 9.145187862775209e-05,
|
489 |
+
"loss": 0.666,
|
490 |
+
"step": 108
|
491 |
+
},
|
492 |
+
{
|
493 |
+
"epoch": 12.941176470588236,
|
494 |
+
"grad_norm": 0.42959412932395935,
|
495 |
+
"learning_rate": 9.09576022144496e-05,
|
496 |
+
"loss": 0.6539,
|
497 |
+
"step": 110
|
498 |
+
},
|
499 |
+
{
|
500 |
+
"epoch": 12.941176470588236,
|
501 |
+
"eval_loss": 0.9093228578567505,
|
502 |
+
"eval_runtime": 13.9213,
|
503 |
+
"eval_samples_per_second": 2.586,
|
504 |
+
"eval_steps_per_second": 2.586,
|
505 |
+
"step": 110
|
506 |
+
},
|
507 |
+
{
|
508 |
+
"epoch": 13.176470588235293,
|
509 |
+
"grad_norm": 0.3179619312286377,
|
510 |
+
"learning_rate": 9.045084971874738e-05,
|
511 |
+
"loss": 0.6358,
|
512 |
+
"step": 112
|
513 |
+
},
|
514 |
+
{
|
515 |
+
"epoch": 13.411764705882353,
|
516 |
+
"grad_norm": 0.37940555810928345,
|
517 |
+
"learning_rate": 8.993177550236464e-05,
|
518 |
+
"loss": 0.5786,
|
519 |
+
"step": 114
|
520 |
+
},
|
521 |
+
{
|
522 |
+
"epoch": 13.647058823529411,
|
523 |
+
"grad_norm": 0.3830829858779907,
|
524 |
+
"learning_rate": 8.940053768033609e-05,
|
525 |
+
"loss": 0.6127,
|
526 |
+
"step": 116
|
527 |
+
},
|
528 |
+
{
|
529 |
+
"epoch": 13.882352941176471,
|
530 |
+
"grad_norm": 0.42292138934135437,
|
531 |
+
"learning_rate": 8.885729807284856e-05,
|
532 |
+
"loss": 0.632,
|
533 |
+
"step": 118
|
534 |
+
},
|
535 |
+
{
|
536 |
+
"epoch": 14.0,
|
537 |
+
"eval_loss": 0.9347273707389832,
|
538 |
+
"eval_runtime": 13.9419,
|
539 |
+
"eval_samples_per_second": 2.582,
|
540 |
+
"eval_steps_per_second": 2.582,
|
541 |
+
"step": 119
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 14.117647058823529,
|
545 |
+
"grad_norm": 0.41069409251213074,
|
546 |
+
"learning_rate": 8.83022221559489e-05,
|
547 |
+
"loss": 0.626,
|
548 |
+
"step": 120
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 14.352941176470589,
|
552 |
+
"grad_norm": 0.528698742389679,
|
553 |
+
"learning_rate": 8.773547901113862e-05,
|
554 |
+
"loss": 0.5267,
|
555 |
+
"step": 122
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 14.588235294117647,
|
559 |
+
"grad_norm": 0.4924652874469757,
|
560 |
+
"learning_rate": 8.715724127386972e-05,
|
561 |
+
"loss": 0.5981,
|
562 |
+
"step": 124
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 14.823529411764707,
|
566 |
+
"grad_norm": 0.5208938121795654,
|
567 |
+
"learning_rate": 8.656768508095853e-05,
|
568 |
+
"loss": 0.543,
|
569 |
+
"step": 126
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 14.941176470588236,
|
573 |
+
"eval_loss": 0.9885026216506958,
|
574 |
+
"eval_runtime": 13.9217,
|
575 |
+
"eval_samples_per_second": 2.586,
|
576 |
+
"eval_steps_per_second": 2.586,
|
577 |
+
"step": 127
|
578 |
+
},
|
579 |
+
{
|
580 |
+
"epoch": 15.058823529411764,
|
581 |
+
"grad_norm": 0.5211675763130188,
|
582 |
+
"learning_rate": 8.596699001693255e-05,
|
583 |
+
"loss": 0.5916,
|
584 |
+
"step": 128
|
585 |
+
},
|
586 |
+
{
|
587 |
+
"epoch": 15.294117647058824,
|
588 |
+
"grad_norm": 0.5175803899765015,
|
589 |
+
"learning_rate": 8.535533905932738e-05,
|
590 |
+
"loss": 0.5308,
|
591 |
+
"step": 130
|
592 |
+
},
|
593 |
+
{
|
594 |
+
"epoch": 15.529411764705882,
|
595 |
+
"grad_norm": 0.55540931224823,
|
596 |
+
"learning_rate": 8.473291852294987e-05,
|
597 |
+
"loss": 0.551,
|
598 |
+
"step": 132
|
599 |
+
},
|
600 |
+
{
|
601 |
+
"epoch": 15.764705882352942,
|
602 |
+
"grad_norm": 0.527597188949585,
|
603 |
+
"learning_rate": 8.409991800312493e-05,
|
604 |
+
"loss": 0.5514,
|
605 |
+
"step": 134
|
606 |
+
},
|
607 |
+
{
|
608 |
+
"epoch": 16.0,
|
609 |
+
"grad_norm": 0.5720416307449341,
|
610 |
+
"learning_rate": 8.345653031794292e-05,
|
611 |
+
"loss": 0.4869,
|
612 |
+
"step": 136
|
613 |
+
},
|
614 |
+
{
|
615 |
+
"epoch": 16.0,
|
616 |
+
"eval_loss": 1.0261539220809937,
|
617 |
+
"eval_runtime": 13.9314,
|
618 |
+
"eval_samples_per_second": 2.584,
|
619 |
+
"eval_steps_per_second": 2.584,
|
620 |
+
"step": 136
|
621 |
+
},
|
622 |
+
{
|
623 |
+
"epoch": 16.235294117647058,
|
624 |
+
"grad_norm": 0.7312718033790588,
|
625 |
+
"learning_rate": 8.280295144952536e-05,
|
626 |
+
"loss": 0.4512,
|
627 |
+
"step": 138
|
628 |
+
},
|
629 |
+
{
|
630 |
+
"epoch": 16.470588235294116,
|
631 |
+
"grad_norm": 0.7295913100242615,
|
632 |
+
"learning_rate": 8.213938048432697e-05,
|
633 |
+
"loss": 0.5245,
|
634 |
+
"step": 140
|
635 |
+
},
|
636 |
+
{
|
637 |
+
"epoch": 16.705882352941178,
|
638 |
+
"grad_norm": 0.7654659152030945,
|
639 |
+
"learning_rate": 8.146601955249188e-05,
|
640 |
+
"loss": 0.4661,
|
641 |
+
"step": 142
|
642 |
+
},
|
643 |
+
{
|
644 |
+
"epoch": 16.941176470588236,
|
645 |
+
"grad_norm": 0.6638887524604797,
|
646 |
+
"learning_rate": 8.07830737662829e-05,
|
647 |
+
"loss": 0.4926,
|
648 |
+
"step": 144
|
649 |
+
},
|
650 |
+
{
|
651 |
+
"epoch": 16.941176470588236,
|
652 |
+
"eval_loss": 1.043527603149414,
|
653 |
+
"eval_runtime": 13.9302,
|
654 |
+
"eval_samples_per_second": 2.584,
|
655 |
+
"eval_steps_per_second": 2.584,
|
656 |
+
"step": 144
|
657 |
+
},
|
658 |
+
{
|
659 |
+
"epoch": 16.941176470588236,
|
660 |
+
"step": 144,
|
661 |
+
"total_flos": 7.574414567879475e+16,
|
662 |
+
"train_loss": 0.8188091479241848,
|
663 |
+
"train_runtime": 3221.0584,
|
664 |
+
"train_samples_per_second": 2.111,
|
665 |
+
"train_steps_per_second": 0.124
|
666 |
+
}
|
667 |
+
],
|
668 |
+
"logging_steps": 2,
|
669 |
+
"max_steps": 400,
|
670 |
+
"num_input_tokens_seen": 0,
|
671 |
+
"num_train_epochs": 50,
|
672 |
+
"save_steps": 25,
|
673 |
+
"stateful_callbacks": {
|
674 |
+
"EarlyStoppingCallback": {
|
675 |
+
"args": {
|
676 |
+
"early_stopping_patience": 7,
|
677 |
+
"early_stopping_threshold": 0.0
|
678 |
+
},
|
679 |
+
"attributes": {
|
680 |
+
"early_stopping_patience_counter": 0
|
681 |
+
}
|
682 |
+
},
|
683 |
+
"TrainerControl": {
|
684 |
+
"args": {
|
685 |
+
"should_epoch_stop": false,
|
686 |
+
"should_evaluate": false,
|
687 |
+
"should_log": false,
|
688 |
+
"should_save": true,
|
689 |
+
"should_training_stop": true
|
690 |
+
},
|
691 |
+
"attributes": {}
|
692 |
+
}
|
693 |
+
},
|
694 |
+
"total_flos": 7.574414567879475e+16,
|
695 |
+
"train_batch_size": 1,
|
696 |
+
"trial_name": null,
|
697 |
+
"trial_params": null
|
698 |
+
}
|