Upload 13 files
Browse files- README.md +12 -12
- all_results.json +13 -13
- config.json +2 -2
- eval_results.json +7 -7
- model.safetensors +1 -1
- special_tokens_map.json +6 -42
- tokenizer.json +0 -0
- tokenizer_config.json +1 -5
- train_results.json +6 -6
- trainer_state.json +1059 -233
- training_args.bin +2 -2
README.md
CHANGED
@@ -20,22 +20,22 @@ model-index:
|
|
20 |
type: Zyphra/Zyda-2
|
21 |
metrics:
|
22 |
- type: accuracy
|
23 |
-
value: 0.
|
24 |
name: Accuracy
|
25 |
-
base_model:
|
26 |
---
|
27 |
|
28 |
-
# DeBERTa-v3-xsmall-
|
29 |
|
30 |
## Model Description
|
31 |
|
32 |
-
This model is a fine-tuned version of [
|
33 |
|
34 |
## Performance
|
35 |
|
36 |
The model achieves the following results on the evaluation set:
|
37 |
-
- Loss: 2.
|
38 |
-
- Accuracy: 0.
|
39 |
|
40 |
## Intended Uses & Limitations
|
41 |
|
@@ -50,7 +50,7 @@ This model is designed to be used and finetuned for the following tasks:
|
|
50 |
|
51 |
## Training Data
|
52 |
|
53 |
-
The model was trained on the first
|
54 |
5% of that data was used for validation.
|
55 |
|
56 |
## Training Procedure
|
@@ -66,12 +66,12 @@ The following hyperparameters were used during training:
|
|
66 |
- Learning rate scheduler: Linear
|
67 |
- Number of epochs: 1.0
|
68 |
|
69 |
-
### Framework
|
70 |
|
71 |
-
- Transformers: 4.
|
72 |
-
-
|
73 |
- Datasets: 3.1.0
|
74 |
-
- Tokenizers: 0.
|
75 |
|
76 |
## Usage Examples
|
77 |
|
@@ -110,4 +110,4 @@ As this model is trained on a subset of the Zyda-2 dataset, it may inherit biase
|
|
110 |
|
111 |
## Additional Information
|
112 |
|
113 |
-
For more details about the base model, please refer to [
|
|
|
20 |
type: Zyphra/Zyda-2
|
21 |
metrics:
|
22 |
- type: accuracy
|
23 |
+
value: 0.5607
|
24 |
name: Accuracy
|
25 |
+
base_model: microsoft/deberta-v3-xsmall
|
26 |
---
|
27 |
|
28 |
+
# DeBERTa-v3-xsmall-Zyda-2
|
29 |
|
30 |
## Model Description
|
31 |
|
32 |
+
This model is a fine-tuned version of [microsoft/deberta-v3-xsmall](https://huggingface.co/microsoft/deberta-v3-xsmall) on a subset of the [Zyphra/Zyda-2](https://huggingface.co/datasets/Zyphra/Zyda-2) dataset. It was trained using the Masked Language Modeling (MLM) objective to enhance its understanding of the English language.
|
33 |
|
34 |
## Performance
|
35 |
|
36 |
The model achieves the following results on the evaluation set:
|
37 |
+
- Loss: 2.6347
|
38 |
+
- Accuracy: 0.5607
|
39 |
|
40 |
## Intended Uses & Limitations
|
41 |
|
|
|
50 |
|
51 |
## Training Data
|
52 |
|
53 |
+
The model was trained on the first 300 000 rows of the [Zyphra/Zyda-2](https://huggingface.co/datasets/Zyphra/Zyda-2) dataset.
|
54 |
5% of that data was used for validation.
|
55 |
|
56 |
## Training Procedure
|
|
|
66 |
- Learning rate scheduler: Linear
|
67 |
- Number of epochs: 1.0
|
68 |
|
69 |
+
### Framework versions
|
70 |
|
71 |
+
- Transformers: 4.46.3
|
72 |
+
- Pytorch: 2.5.1+cu124
|
73 |
- Datasets: 3.1.0
|
74 |
+
- Tokenizers: 0.20.3
|
75 |
|
76 |
## Usage Examples
|
77 |
|
|
|
110 |
|
111 |
## Additional Information
|
112 |
|
113 |
+
For more details about the base model, please refer to [microsoft/deberta-v3-xsmall](https://huggingface.co/microsoft/deberta-v3-xsmall).
|
all_results.json
CHANGED
@@ -1,16 +1,16 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_loss": 2.
|
5 |
-
"eval_runtime":
|
6 |
-
"eval_samples":
|
7 |
-
"eval_samples_per_second":
|
8 |
-
"eval_steps_per_second": 11.
|
9 |
-
"perplexity":
|
10 |
-
"total_flos":
|
11 |
-
"train_loss": 3.
|
12 |
-
"train_runtime":
|
13 |
-
"train_samples":
|
14 |
-
"train_samples_per_second": 34.
|
15 |
-
"train_steps_per_second": 4.
|
16 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"eval_accuracy": 0.5606630977618748,
|
4 |
+
"eval_loss": 2.634730815887451,
|
5 |
+
"eval_runtime": 407.3409,
|
6 |
+
"eval_samples": 36612,
|
7 |
+
"eval_samples_per_second": 89.88,
|
8 |
+
"eval_steps_per_second": 11.236,
|
9 |
+
"perplexity": 13.939559650092963,
|
10 |
+
"total_flos": 4.634223291773338e+16,
|
11 |
+
"train_loss": 3.132497888326312,
|
12 |
+
"train_runtime": 20429.0413,
|
13 |
+
"train_samples": 699309,
|
14 |
+
"train_samples_per_second": 34.231,
|
15 |
+
"train_steps_per_second": 4.279
|
16 |
}
|
config.json
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "
|
3 |
"architectures": [
|
4 |
"DebertaV2ForMaskedLM"
|
5 |
],
|
@@ -29,7 +29,7 @@
|
|
29 |
"relative_attention": true,
|
30 |
"share_att_key": true,
|
31 |
"torch_dtype": "float32",
|
32 |
-
"transformers_version": "4.
|
33 |
"type_vocab_size": 0,
|
34 |
"vocab_size": 128100
|
35 |
}
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "microsoft/deberta-v3-xsmall",
|
3 |
"architectures": [
|
4 |
"DebertaV2ForMaskedLM"
|
5 |
],
|
|
|
29 |
"relative_attention": true,
|
30 |
"share_att_key": true,
|
31 |
"torch_dtype": "float32",
|
32 |
+
"transformers_version": "4.46.3",
|
33 |
"type_vocab_size": 0,
|
34 |
"vocab_size": 128100
|
35 |
}
|
eval_results.json
CHANGED
@@ -1,10 +1,10 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_loss": 2.
|
5 |
-
"eval_runtime":
|
6 |
-
"eval_samples":
|
7 |
-
"eval_samples_per_second":
|
8 |
-
"eval_steps_per_second": 11.
|
9 |
-
"perplexity":
|
10 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"eval_accuracy": 0.5606630977618748,
|
4 |
+
"eval_loss": 2.634730815887451,
|
5 |
+
"eval_runtime": 407.3409,
|
6 |
+
"eval_samples": 36612,
|
7 |
+
"eval_samples_per_second": 89.88,
|
8 |
+
"eval_steps_per_second": 11.236,
|
9 |
+
"perplexity": 13.939559650092963
|
10 |
}
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 283860016
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:656a30eef186b57c86bad3c788ba0177fbd70e219eb02d0459ef9eda09df43c2
|
3 |
size 283860016
|
special_tokens_map.json
CHANGED
@@ -1,46 +1,10 @@
|
|
1 |
{
|
2 |
-
"bos_token":
|
3 |
-
|
4 |
-
|
5 |
-
|
6 |
-
|
7 |
-
|
8 |
-
},
|
9 |
-
"cls_token": {
|
10 |
-
"content": "[CLS]",
|
11 |
-
"lstrip": false,
|
12 |
-
"normalized": false,
|
13 |
-
"rstrip": false,
|
14 |
-
"single_word": false
|
15 |
-
},
|
16 |
-
"eos_token": {
|
17 |
-
"content": "[SEP]",
|
18 |
-
"lstrip": false,
|
19 |
-
"normalized": false,
|
20 |
-
"rstrip": false,
|
21 |
-
"single_word": false
|
22 |
-
},
|
23 |
-
"mask_token": {
|
24 |
-
"content": "[MASK]",
|
25 |
-
"lstrip": false,
|
26 |
-
"normalized": false,
|
27 |
-
"rstrip": false,
|
28 |
-
"single_word": false
|
29 |
-
},
|
30 |
-
"pad_token": {
|
31 |
-
"content": "[PAD]",
|
32 |
-
"lstrip": false,
|
33 |
-
"normalized": false,
|
34 |
-
"rstrip": false,
|
35 |
-
"single_word": false
|
36 |
-
},
|
37 |
-
"sep_token": {
|
38 |
-
"content": "[SEP]",
|
39 |
-
"lstrip": false,
|
40 |
-
"normalized": false,
|
41 |
-
"rstrip": false,
|
42 |
-
"single_word": false
|
43 |
-
},
|
44 |
"unk_token": {
|
45 |
"content": "[UNK]",
|
46 |
"lstrip": false,
|
|
|
1 |
{
|
2 |
+
"bos_token": "[CLS]",
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"eos_token": "[SEP]",
|
5 |
+
"mask_token": "[MASK]",
|
6 |
+
"pad_token": "[PAD]",
|
7 |
+
"sep_token": "[SEP]",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
8 |
"unk_token": {
|
9 |
"content": "[UNK]",
|
10 |
"lstrip": false,
|
tokenizer.json
CHANGED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
CHANGED
@@ -42,21 +42,17 @@
|
|
42 |
}
|
43 |
},
|
44 |
"bos_token": "[CLS]",
|
45 |
-
"clean_up_tokenization_spaces":
|
46 |
"cls_token": "[CLS]",
|
47 |
"do_lower_case": false,
|
48 |
"eos_token": "[SEP]",
|
49 |
"mask_token": "[MASK]",
|
50 |
-
"max_length": 1024,
|
51 |
"model_max_length": 1000000000000000019884624838656,
|
52 |
"pad_token": "[PAD]",
|
53 |
"sep_token": "[SEP]",
|
54 |
"sp_model_kwargs": {},
|
55 |
"split_by_punct": false,
|
56 |
-
"stride": 0,
|
57 |
"tokenizer_class": "DebertaV2Tokenizer",
|
58 |
-
"truncation_side": "right",
|
59 |
-
"truncation_strategy": "longest_first",
|
60 |
"unk_token": "[UNK]",
|
61 |
"vocab_type": "spm"
|
62 |
}
|
|
|
42 |
}
|
43 |
},
|
44 |
"bos_token": "[CLS]",
|
45 |
+
"clean_up_tokenization_spaces": false,
|
46 |
"cls_token": "[CLS]",
|
47 |
"do_lower_case": false,
|
48 |
"eos_token": "[SEP]",
|
49 |
"mask_token": "[MASK]",
|
|
|
50 |
"model_max_length": 1000000000000000019884624838656,
|
51 |
"pad_token": "[PAD]",
|
52 |
"sep_token": "[SEP]",
|
53 |
"sp_model_kwargs": {},
|
54 |
"split_by_punct": false,
|
|
|
55 |
"tokenizer_class": "DebertaV2Tokenizer",
|
|
|
|
|
56 |
"unk_token": "[UNK]",
|
57 |
"vocab_type": "spm"
|
58 |
}
|
train_results.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
-
"total_flos":
|
4 |
-
"train_loss": 3.
|
5 |
-
"train_runtime":
|
6 |
-
"train_samples":
|
7 |
-
"train_samples_per_second": 34.
|
8 |
-
"train_steps_per_second": 4.
|
9 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"total_flos": 4.634223291773338e+16,
|
4 |
+
"train_loss": 3.132497888326312,
|
5 |
+
"train_runtime": 20429.0413,
|
6 |
+
"train_samples": 699309,
|
7 |
+
"train_samples_per_second": 34.231,
|
8 |
+
"train_steps_per_second": 4.279
|
9 |
}
|
trainer_state.json
CHANGED
@@ -3,415 +3,1241 @@
|
|
3 |
"best_model_checkpoint": null,
|
4 |
"epoch": 1.0,
|
5 |
"eval_steps": 500,
|
6 |
-
"global_step":
|
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.
|
13 |
-
"grad_norm":
|
14 |
-
"learning_rate": 4.
|
15 |
-
"loss":
|
16 |
"step": 500
|
17 |
},
|
18 |
{
|
19 |
-
"epoch": 0.
|
20 |
-
"grad_norm":
|
21 |
-
"learning_rate": 4.
|
22 |
-
"loss":
|
23 |
"step": 1000
|
24 |
},
|
25 |
{
|
26 |
-
"epoch": 0.
|
27 |
-
"grad_norm":
|
28 |
-
"learning_rate": 4.
|
29 |
-
"loss":
|
30 |
"step": 1500
|
31 |
},
|
32 |
{
|
33 |
-
"epoch": 0.
|
34 |
-
"grad_norm":
|
35 |
-
"learning_rate": 4.
|
36 |
-
"loss":
|
37 |
"step": 2000
|
38 |
},
|
39 |
{
|
40 |
-
"epoch": 0.
|
41 |
-
"grad_norm":
|
42 |
-
"learning_rate": 4.
|
43 |
-
"loss":
|
44 |
"step": 2500
|
45 |
},
|
46 |
{
|
47 |
-
"epoch": 0.
|
48 |
-
"grad_norm":
|
49 |
-
"learning_rate": 4.
|
50 |
-
"loss":
|
51 |
"step": 3000
|
52 |
},
|
53 |
{
|
54 |
-
"epoch": 0.
|
55 |
-
"grad_norm": 5.
|
56 |
-
"learning_rate": 4.
|
57 |
-
"loss":
|
58 |
"step": 3500
|
59 |
},
|
60 |
{
|
61 |
-
"epoch": 0.
|
62 |
-
"grad_norm":
|
63 |
-
"learning_rate": 4.
|
64 |
-
"loss":
|
65 |
"step": 4000
|
66 |
},
|
67 |
{
|
68 |
-
"epoch": 0.
|
69 |
-
"grad_norm": 5.
|
70 |
-
"learning_rate": 4.
|
71 |
-
"loss": 3.
|
72 |
"step": 4500
|
73 |
},
|
74 |
{
|
75 |
-
"epoch": 0.
|
76 |
-
"grad_norm":
|
77 |
-
"learning_rate": 4.
|
78 |
-
"loss": 3.
|
79 |
"step": 5000
|
80 |
},
|
81 |
{
|
82 |
-
"epoch": 0.
|
83 |
-
"grad_norm":
|
84 |
-
"learning_rate": 4.
|
85 |
-
"loss": 3.
|
86 |
"step": 5500
|
87 |
},
|
88 |
{
|
89 |
-
"epoch": 0.
|
90 |
-
"grad_norm": 5.
|
91 |
-
"learning_rate":
|
92 |
-
"loss": 3.
|
93 |
"step": 6000
|
94 |
},
|
95 |
{
|
96 |
-
"epoch": 0.
|
97 |
-
"grad_norm": 5.
|
98 |
-
"learning_rate":
|
99 |
-
"loss": 3.
|
100 |
"step": 6500
|
101 |
},
|
102 |
{
|
103 |
-
"epoch": 0.
|
104 |
-
"grad_norm":
|
105 |
-
"learning_rate":
|
106 |
-
"loss": 3.
|
107 |
"step": 7000
|
108 |
},
|
109 |
{
|
110 |
-
"epoch": 0.
|
111 |
-
"grad_norm":
|
112 |
-
"learning_rate":
|
113 |
-
"loss": 3.
|
114 |
"step": 7500
|
115 |
},
|
116 |
{
|
117 |
-
"epoch": 0.
|
118 |
-
"grad_norm": 5.
|
119 |
-
"learning_rate":
|
120 |
-
"loss": 3.
|
121 |
"step": 8000
|
122 |
},
|
123 |
{
|
124 |
-
"epoch": 0.
|
125 |
-
"grad_norm":
|
126 |
-
"learning_rate":
|
127 |
-
"loss": 3.
|
128 |
"step": 8500
|
129 |
},
|
130 |
{
|
131 |
-
"epoch": 0.
|
132 |
-
"grad_norm": 5.
|
133 |
-
"learning_rate":
|
134 |
-
"loss": 3.
|
135 |
"step": 9000
|
136 |
},
|
137 |
{
|
138 |
-
"epoch": 0.
|
139 |
-
"grad_norm": 5.
|
140 |
-
"learning_rate":
|
141 |
-
"loss": 3.
|
142 |
"step": 9500
|
143 |
},
|
144 |
{
|
145 |
-
"epoch": 0.
|
146 |
-
"grad_norm":
|
147 |
-
"learning_rate":
|
148 |
-
"loss": 3.
|
149 |
"step": 10000
|
150 |
},
|
151 |
{
|
152 |
-
"epoch": 0.
|
153 |
-
"grad_norm": 5.
|
154 |
-
"learning_rate":
|
155 |
-
"loss": 3.
|
156 |
"step": 10500
|
157 |
},
|
158 |
{
|
159 |
-
"epoch": 0.
|
160 |
-
"grad_norm": 5.
|
161 |
-
"learning_rate":
|
162 |
-
"loss": 3.
|
163 |
"step": 11000
|
164 |
},
|
165 |
{
|
166 |
-
"epoch": 0.
|
167 |
-
"grad_norm":
|
168 |
-
"learning_rate":
|
169 |
-
"loss": 3.
|
170 |
"step": 11500
|
171 |
},
|
172 |
{
|
173 |
-
"epoch": 0.
|
174 |
-
"grad_norm":
|
175 |
-
"learning_rate":
|
176 |
-
"loss": 3.
|
177 |
"step": 12000
|
178 |
},
|
179 |
{
|
180 |
-
"epoch": 0.
|
181 |
-
"grad_norm":
|
182 |
-
"learning_rate":
|
183 |
-
"loss": 3.
|
184 |
"step": 12500
|
185 |
},
|
186 |
{
|
187 |
-
"epoch": 0.
|
188 |
-
"grad_norm": 5.
|
189 |
-
"learning_rate":
|
190 |
-
"loss": 3.
|
191 |
"step": 13000
|
192 |
},
|
193 |
{
|
194 |
-
"epoch": 0.
|
195 |
-
"grad_norm": 5.
|
196 |
-
"learning_rate":
|
197 |
-
"loss": 3.
|
198 |
"step": 13500
|
199 |
},
|
200 |
{
|
201 |
-
"epoch": 0.
|
202 |
-
"grad_norm":
|
203 |
-
"learning_rate":
|
204 |
-
"loss": 3.
|
205 |
"step": 14000
|
206 |
},
|
207 |
{
|
208 |
-
"epoch": 0.
|
209 |
-
"grad_norm": 5.
|
210 |
-
"learning_rate":
|
211 |
-
"loss": 3.
|
212 |
"step": 14500
|
213 |
},
|
214 |
{
|
215 |
-
"epoch": 0.
|
216 |
-
"grad_norm":
|
217 |
-
"learning_rate":
|
218 |
-
"loss": 3.
|
219 |
"step": 15000
|
220 |
},
|
221 |
{
|
222 |
-
"epoch": 0.
|
223 |
-
"grad_norm":
|
224 |
-
"learning_rate":
|
225 |
-
"loss": 3.
|
226 |
"step": 15500
|
227 |
},
|
228 |
{
|
229 |
-
"epoch": 0.
|
230 |
-
"grad_norm":
|
231 |
-
"learning_rate":
|
232 |
-
"loss": 3.
|
233 |
"step": 16000
|
234 |
},
|
235 |
{
|
236 |
-
"epoch": 0.
|
237 |
-
"grad_norm": 5.
|
238 |
-
"learning_rate":
|
239 |
-
"loss": 3.
|
240 |
"step": 16500
|
241 |
},
|
242 |
{
|
243 |
-
"epoch": 0.
|
244 |
-
"grad_norm":
|
245 |
-
"learning_rate":
|
246 |
-
"loss": 3.
|
247 |
"step": 17000
|
248 |
},
|
249 |
{
|
250 |
-
"epoch": 0.
|
251 |
-
"grad_norm":
|
252 |
-
"learning_rate":
|
253 |
-
"loss": 3.
|
254 |
"step": 17500
|
255 |
},
|
256 |
{
|
257 |
-
"epoch": 0.
|
258 |
-
"grad_norm": 5.
|
259 |
-
"learning_rate":
|
260 |
-
"loss": 3.
|
261 |
"step": 18000
|
262 |
},
|
263 |
{
|
264 |
-
"epoch": 0.
|
265 |
-
"grad_norm": 5.
|
266 |
-
"learning_rate":
|
267 |
-
"loss": 3.
|
268 |
"step": 18500
|
269 |
},
|
270 |
{
|
271 |
-
"epoch": 0.
|
272 |
-
"grad_norm": 5.
|
273 |
-
"learning_rate":
|
274 |
-
"loss": 3.
|
275 |
"step": 19000
|
276 |
},
|
277 |
{
|
278 |
-
"epoch": 0.
|
279 |
-
"grad_norm": 5.
|
280 |
-
"learning_rate":
|
281 |
-
"loss": 3.
|
282 |
"step": 19500
|
283 |
},
|
284 |
{
|
285 |
-
"epoch": 0.
|
286 |
-
"grad_norm":
|
287 |
-
"learning_rate":
|
288 |
-
"loss": 3.
|
289 |
"step": 20000
|
290 |
},
|
291 |
{
|
292 |
-
"epoch": 0.
|
293 |
-
"grad_norm":
|
294 |
-
"learning_rate":
|
295 |
-
"loss": 3.
|
296 |
"step": 20500
|
297 |
},
|
298 |
{
|
299 |
-
"epoch": 0.
|
300 |
-
"grad_norm":
|
301 |
-
"learning_rate":
|
302 |
-
"loss": 3.
|
303 |
"step": 21000
|
304 |
},
|
305 |
{
|
306 |
-
"epoch": 0.
|
307 |
-
"grad_norm":
|
308 |
-
"learning_rate":
|
309 |
-
"loss": 3.
|
310 |
"step": 21500
|
311 |
},
|
312 |
{
|
313 |
-
"epoch": 0.
|
314 |
-
"grad_norm": 5.
|
315 |
-
"learning_rate":
|
316 |
-
"loss": 3.
|
317 |
"step": 22000
|
318 |
},
|
319 |
{
|
320 |
-
"epoch": 0.
|
321 |
-
"grad_norm":
|
322 |
-
"learning_rate":
|
323 |
-
"loss": 3.
|
324 |
"step": 22500
|
325 |
},
|
326 |
{
|
327 |
-
"epoch": 0.
|
328 |
-
"grad_norm": 5.
|
329 |
-
"learning_rate":
|
330 |
-
"loss": 3.
|
331 |
"step": 23000
|
332 |
},
|
333 |
{
|
334 |
-
"epoch": 0.
|
335 |
-
"grad_norm": 5.
|
336 |
-
"learning_rate":
|
337 |
-
"loss": 3.
|
338 |
"step": 23500
|
339 |
},
|
340 |
{
|
341 |
-
"epoch": 0.
|
342 |
-
"grad_norm":
|
343 |
-
"learning_rate":
|
344 |
-
"loss": 3.
|
345 |
"step": 24000
|
346 |
},
|
347 |
{
|
348 |
-
"epoch": 0.
|
349 |
-
"grad_norm":
|
350 |
-
"learning_rate":
|
351 |
-
"loss": 3.
|
352 |
"step": 24500
|
353 |
},
|
354 |
{
|
355 |
-
"epoch": 0.
|
356 |
-
"grad_norm": 5.
|
357 |
-
"learning_rate":
|
358 |
-
"loss": 3.
|
359 |
"step": 25000
|
360 |
},
|
361 |
{
|
362 |
-
"epoch": 0.
|
363 |
-
"grad_norm": 5.
|
364 |
-
"learning_rate":
|
365 |
-
"loss": 3.
|
366 |
"step": 25500
|
367 |
},
|
368 |
{
|
369 |
-
"epoch": 0.
|
370 |
-
"grad_norm":
|
371 |
-
"learning_rate":
|
372 |
-
"loss": 3.
|
373 |
"step": 26000
|
374 |
},
|
375 |
{
|
376 |
-
"epoch": 0.
|
377 |
-
"grad_norm":
|
378 |
-
"learning_rate": 3.
|
379 |
-
"loss": 3.
|
380 |
"step": 26500
|
381 |
},
|
382 |
{
|
383 |
-
"epoch": 0.
|
384 |
-
"grad_norm":
|
385 |
-
"learning_rate":
|
386 |
-
"loss": 3.
|
387 |
"step": 27000
|
388 |
},
|
389 |
{
|
390 |
-
"epoch": 0.
|
391 |
-
"grad_norm":
|
392 |
-
"learning_rate":
|
393 |
-
"loss": 3.
|
394 |
"step": 27500
|
395 |
},
|
396 |
{
|
397 |
-
"epoch": 0.
|
398 |
-
"grad_norm":
|
399 |
-
"learning_rate":
|
400 |
-
"loss": 3.
|
401 |
"step": 28000
|
402 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
403 |
{
|
404 |
"epoch": 1.0,
|
405 |
-
"step":
|
406 |
-
"total_flos":
|
407 |
-
"train_loss": 3.
|
408 |
-
"train_runtime":
|
409 |
-
"train_samples_per_second": 34.
|
410 |
-
"train_steps_per_second": 4.
|
411 |
}
|
412 |
],
|
413 |
"logging_steps": 500,
|
414 |
-
"max_steps":
|
415 |
"num_input_tokens_seen": 0,
|
416 |
"num_train_epochs": 1,
|
417 |
"save_steps": 500,
|
@@ -427,7 +1253,7 @@
|
|
427 |
"attributes": {}
|
428 |
}
|
429 |
},
|
430 |
-
"total_flos":
|
431 |
"train_batch_size": 8,
|
432 |
"trial_name": null,
|
433 |
"trial_params": null
|
|
|
3 |
"best_model_checkpoint": null,
|
4 |
"epoch": 1.0,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 87414,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
+
"epoch": 0.005719907566293729,
|
13 |
+
"grad_norm": 7.0386528968811035,
|
14 |
+
"learning_rate": 4.971400462168532e-05,
|
15 |
+
"loss": 7.2889,
|
16 |
"step": 500
|
17 |
},
|
18 |
{
|
19 |
+
"epoch": 0.011439815132587457,
|
20 |
+
"grad_norm": 6.592615604400635,
|
21 |
+
"learning_rate": 4.942800924337063e-05,
|
22 |
+
"loss": 5.389,
|
23 |
"step": 1000
|
24 |
},
|
25 |
{
|
26 |
+
"epoch": 0.017159722698881188,
|
27 |
+
"grad_norm": 6.088059425354004,
|
28 |
+
"learning_rate": 4.914201386505594e-05,
|
29 |
+
"loss": 4.8994,
|
30 |
"step": 1500
|
31 |
},
|
32 |
{
|
33 |
+
"epoch": 0.022879630265174915,
|
34 |
+
"grad_norm": 6.355687618255615,
|
35 |
+
"learning_rate": 4.885601848674126e-05,
|
36 |
+
"loss": 4.6219,
|
37 |
"step": 2000
|
38 |
},
|
39 |
{
|
40 |
+
"epoch": 0.028599537831468645,
|
41 |
+
"grad_norm": 6.1159186363220215,
|
42 |
+
"learning_rate": 4.8570023108426574e-05,
|
43 |
+
"loss": 4.4233,
|
44 |
"step": 2500
|
45 |
},
|
46 |
{
|
47 |
+
"epoch": 0.034319445397762376,
|
48 |
+
"grad_norm": 6.636866092681885,
|
49 |
+
"learning_rate": 4.828402773011189e-05,
|
50 |
+
"loss": 4.2802,
|
51 |
"step": 3000
|
52 |
},
|
53 |
{
|
54 |
+
"epoch": 0.0400393529640561,
|
55 |
+
"grad_norm": 5.87605619430542,
|
56 |
+
"learning_rate": 4.7998032351797196e-05,
|
57 |
+
"loss": 4.1453,
|
58 |
"step": 3500
|
59 |
},
|
60 |
{
|
61 |
+
"epoch": 0.04575926053034983,
|
62 |
+
"grad_norm": 6.0966691970825195,
|
63 |
+
"learning_rate": 4.771203697348251e-05,
|
64 |
+
"loss": 4.0822,
|
65 |
"step": 4000
|
66 |
},
|
67 |
{
|
68 |
+
"epoch": 0.05147916809664356,
|
69 |
+
"grad_norm": 5.99679708480835,
|
70 |
+
"learning_rate": 4.7426041595167824e-05,
|
71 |
+
"loss": 3.9941,
|
72 |
"step": 4500
|
73 |
},
|
74 |
{
|
75 |
+
"epoch": 0.05719907566293729,
|
76 |
+
"grad_norm": 5.738614082336426,
|
77 |
+
"learning_rate": 4.714004621685314e-05,
|
78 |
+
"loss": 3.9104,
|
79 |
"step": 5000
|
80 |
},
|
81 |
{
|
82 |
+
"epoch": 0.06291898322923102,
|
83 |
+
"grad_norm": 6.038717746734619,
|
84 |
+
"learning_rate": 4.6854050838538446e-05,
|
85 |
+
"loss": 3.8528,
|
86 |
"step": 5500
|
87 |
},
|
88 |
{
|
89 |
+
"epoch": 0.06863889079552475,
|
90 |
+
"grad_norm": 5.92971134185791,
|
91 |
+
"learning_rate": 4.656805546022377e-05,
|
92 |
+
"loss": 3.78,
|
93 |
"step": 6000
|
94 |
},
|
95 |
{
|
96 |
+
"epoch": 0.07435879836181847,
|
97 |
+
"grad_norm": 5.791173458099365,
|
98 |
+
"learning_rate": 4.628206008190908e-05,
|
99 |
+
"loss": 3.7407,
|
100 |
"step": 6500
|
101 |
},
|
102 |
{
|
103 |
+
"epoch": 0.0800787059281122,
|
104 |
+
"grad_norm": 6.2113566398620605,
|
105 |
+
"learning_rate": 4.5996064703594396e-05,
|
106 |
+
"loss": 3.6834,
|
107 |
"step": 7000
|
108 |
},
|
109 |
{
|
110 |
+
"epoch": 0.08579861349440593,
|
111 |
+
"grad_norm": 5.970207214355469,
|
112 |
+
"learning_rate": 4.57100693252797e-05,
|
113 |
+
"loss": 3.663,
|
114 |
"step": 7500
|
115 |
},
|
116 |
{
|
117 |
+
"epoch": 0.09151852106069966,
|
118 |
+
"grad_norm": 5.8883514404296875,
|
119 |
+
"learning_rate": 4.542407394696502e-05,
|
120 |
+
"loss": 3.6161,
|
121 |
"step": 8000
|
122 |
},
|
123 |
{
|
124 |
+
"epoch": 0.09723842862699339,
|
125 |
+
"grad_norm": 6.351154804229736,
|
126 |
+
"learning_rate": 4.513807856865033e-05,
|
127 |
+
"loss": 3.6193,
|
128 |
"step": 8500
|
129 |
},
|
130 |
{
|
131 |
+
"epoch": 0.10295833619328712,
|
132 |
+
"grad_norm": 5.69071102142334,
|
133 |
+
"learning_rate": 4.4852083190335646e-05,
|
134 |
+
"loss": 3.545,
|
135 |
"step": 9000
|
136 |
},
|
137 |
{
|
138 |
+
"epoch": 0.10867824375958085,
|
139 |
+
"grad_norm": 5.988426208496094,
|
140 |
+
"learning_rate": 4.456608781202096e-05,
|
141 |
+
"loss": 3.5088,
|
142 |
"step": 9500
|
143 |
},
|
144 |
{
|
145 |
+
"epoch": 0.11439815132587458,
|
146 |
+
"grad_norm": 6.251087665557861,
|
147 |
+
"learning_rate": 4.4280092433706274e-05,
|
148 |
+
"loss": 3.5158,
|
149 |
"step": 10000
|
150 |
},
|
151 |
{
|
152 |
+
"epoch": 0.1201180588921683,
|
153 |
+
"grad_norm": 5.840632438659668,
|
154 |
+
"learning_rate": 4.399409705539159e-05,
|
155 |
+
"loss": 3.5045,
|
156 |
"step": 10500
|
157 |
},
|
158 |
{
|
159 |
+
"epoch": 0.12583796645846204,
|
160 |
+
"grad_norm": 5.9697723388671875,
|
161 |
+
"learning_rate": 4.37081016770769e-05,
|
162 |
+
"loss": 3.4466,
|
163 |
"step": 11000
|
164 |
},
|
165 |
{
|
166 |
+
"epoch": 0.13155787402475577,
|
167 |
+
"grad_norm": 6.149275779724121,
|
168 |
+
"learning_rate": 4.342210629876222e-05,
|
169 |
+
"loss": 3.4544,
|
170 |
"step": 11500
|
171 |
},
|
172 |
{
|
173 |
+
"epoch": 0.1372777815910495,
|
174 |
+
"grad_norm": 5.605130672454834,
|
175 |
+
"learning_rate": 4.3136110920447525e-05,
|
176 |
+
"loss": 3.4116,
|
177 |
"step": 12000
|
178 |
},
|
179 |
{
|
180 |
+
"epoch": 0.1429976891573432,
|
181 |
+
"grad_norm": 5.5137152671813965,
|
182 |
+
"learning_rate": 4.285011554213284e-05,
|
183 |
+
"loss": 3.3919,
|
184 |
"step": 12500
|
185 |
},
|
186 |
{
|
187 |
+
"epoch": 0.14871759672363694,
|
188 |
+
"grad_norm": 5.848428726196289,
|
189 |
+
"learning_rate": 4.256412016381815e-05,
|
190 |
+
"loss": 3.3695,
|
191 |
"step": 13000
|
192 |
},
|
193 |
{
|
194 |
+
"epoch": 0.15443750428993067,
|
195 |
+
"grad_norm": 5.794093132019043,
|
196 |
+
"learning_rate": 4.227812478550347e-05,
|
197 |
+
"loss": 3.3509,
|
198 |
"step": 13500
|
199 |
},
|
200 |
{
|
201 |
+
"epoch": 0.1601574118562244,
|
202 |
+
"grad_norm": 5.879131317138672,
|
203 |
+
"learning_rate": 4.199212940718878e-05,
|
204 |
+
"loss": 3.3551,
|
205 |
"step": 14000
|
206 |
},
|
207 |
{
|
208 |
+
"epoch": 0.16587731942251813,
|
209 |
+
"grad_norm": 5.5501179695129395,
|
210 |
+
"learning_rate": 4.1706134028874096e-05,
|
211 |
+
"loss": 3.3327,
|
212 |
"step": 14500
|
213 |
},
|
214 |
{
|
215 |
+
"epoch": 0.17159722698881186,
|
216 |
+
"grad_norm": 5.30273962020874,
|
217 |
+
"learning_rate": 4.142013865055941e-05,
|
218 |
+
"loss": 3.3045,
|
219 |
"step": 15000
|
220 |
},
|
221 |
{
|
222 |
+
"epoch": 0.1773171345551056,
|
223 |
+
"grad_norm": 6.049214839935303,
|
224 |
+
"learning_rate": 4.1134143272244725e-05,
|
225 |
+
"loss": 3.2868,
|
226 |
"step": 15500
|
227 |
},
|
228 |
{
|
229 |
+
"epoch": 0.18303704212139932,
|
230 |
+
"grad_norm": 5.499197483062744,
|
231 |
+
"learning_rate": 4.084814789393003e-05,
|
232 |
+
"loss": 3.2837,
|
233 |
"step": 16000
|
234 |
},
|
235 |
{
|
236 |
+
"epoch": 0.18875694968769305,
|
237 |
+
"grad_norm": 5.448641777038574,
|
238 |
+
"learning_rate": 4.0562152515615347e-05,
|
239 |
+
"loss": 3.2751,
|
240 |
"step": 16500
|
241 |
},
|
242 |
{
|
243 |
+
"epoch": 0.19447685725398678,
|
244 |
+
"grad_norm": 5.422529697418213,
|
245 |
+
"learning_rate": 4.027615713730066e-05,
|
246 |
+
"loss": 3.2585,
|
247 |
"step": 17000
|
248 |
},
|
249 |
{
|
250 |
+
"epoch": 0.2001967648202805,
|
251 |
+
"grad_norm": 5.639166831970215,
|
252 |
+
"learning_rate": 3.9990161758985975e-05,
|
253 |
+
"loss": 3.2587,
|
254 |
"step": 17500
|
255 |
},
|
256 |
{
|
257 |
+
"epoch": 0.20591667238657424,
|
258 |
+
"grad_norm": 5.255904674530029,
|
259 |
+
"learning_rate": 3.970416638067129e-05,
|
260 |
+
"loss": 3.2505,
|
261 |
"step": 18000
|
262 |
},
|
263 |
{
|
264 |
+
"epoch": 0.21163657995286797,
|
265 |
+
"grad_norm": 5.453869819641113,
|
266 |
+
"learning_rate": 3.9418171002356604e-05,
|
267 |
+
"loss": 3.2382,
|
268 |
"step": 18500
|
269 |
},
|
270 |
{
|
271 |
+
"epoch": 0.2173564875191617,
|
272 |
+
"grad_norm": 5.905749797821045,
|
273 |
+
"learning_rate": 3.913217562404192e-05,
|
274 |
+
"loss": 3.2309,
|
275 |
"step": 19000
|
276 |
},
|
277 |
{
|
278 |
+
"epoch": 0.22307639508545543,
|
279 |
+
"grad_norm": 5.28553581237793,
|
280 |
+
"learning_rate": 3.884618024572723e-05,
|
281 |
+
"loss": 3.2308,
|
282 |
"step": 19500
|
283 |
},
|
284 |
{
|
285 |
+
"epoch": 0.22879630265174916,
|
286 |
+
"grad_norm": 5.1942830085754395,
|
287 |
+
"learning_rate": 3.8560184867412547e-05,
|
288 |
+
"loss": 3.2163,
|
289 |
"step": 20000
|
290 |
},
|
291 |
{
|
292 |
+
"epoch": 0.23451621021804286,
|
293 |
+
"grad_norm": 6.12723970413208,
|
294 |
+
"learning_rate": 3.8274189489097854e-05,
|
295 |
+
"loss": 3.2082,
|
296 |
"step": 20500
|
297 |
},
|
298 |
{
|
299 |
+
"epoch": 0.2402361177843366,
|
300 |
+
"grad_norm": 5.664548873901367,
|
301 |
+
"learning_rate": 3.798819411078317e-05,
|
302 |
+
"loss": 3.1802,
|
303 |
"step": 21000
|
304 |
},
|
305 |
{
|
306 |
+
"epoch": 0.24595602535063033,
|
307 |
+
"grad_norm": 5.903208255767822,
|
308 |
+
"learning_rate": 3.770219873246848e-05,
|
309 |
+
"loss": 3.1946,
|
310 |
"step": 21500
|
311 |
},
|
312 |
{
|
313 |
+
"epoch": 0.2516759329169241,
|
314 |
+
"grad_norm": 5.729937553405762,
|
315 |
+
"learning_rate": 3.7416203354153804e-05,
|
316 |
+
"loss": 3.1646,
|
317 |
"step": 22000
|
318 |
},
|
319 |
{
|
320 |
+
"epoch": 0.2573958404832178,
|
321 |
+
"grad_norm": 6.068752765655518,
|
322 |
+
"learning_rate": 3.713020797583911e-05,
|
323 |
+
"loss": 3.2006,
|
324 |
"step": 22500
|
325 |
},
|
326 |
{
|
327 |
+
"epoch": 0.26311574804951154,
|
328 |
+
"grad_norm": 5.97099494934082,
|
329 |
+
"learning_rate": 3.6844212597524425e-05,
|
330 |
+
"loss": 3.1437,
|
331 |
"step": 23000
|
332 |
},
|
333 |
{
|
334 |
+
"epoch": 0.26883565561580525,
|
335 |
+
"grad_norm": 5.777164936065674,
|
336 |
+
"learning_rate": 3.655821721920974e-05,
|
337 |
+
"loss": 3.1407,
|
338 |
"step": 23500
|
339 |
},
|
340 |
{
|
341 |
+
"epoch": 0.274555563182099,
|
342 |
+
"grad_norm": 4.982606887817383,
|
343 |
+
"learning_rate": 3.6272221840895054e-05,
|
344 |
+
"loss": 3.1282,
|
345 |
"step": 24000
|
346 |
},
|
347 |
{
|
348 |
+
"epoch": 0.2802754707483927,
|
349 |
+
"grad_norm": 5.495816230773926,
|
350 |
+
"learning_rate": 3.598622646258037e-05,
|
351 |
+
"loss": 3.1434,
|
352 |
"step": 24500
|
353 |
},
|
354 |
{
|
355 |
+
"epoch": 0.2859953783146864,
|
356 |
+
"grad_norm": 5.898298263549805,
|
357 |
+
"learning_rate": 3.5700231084265676e-05,
|
358 |
+
"loss": 3.1446,
|
359 |
"step": 25000
|
360 |
},
|
361 |
{
|
362 |
+
"epoch": 0.29171528588098017,
|
363 |
+
"grad_norm": 5.729229927062988,
|
364 |
+
"learning_rate": 3.541423570595099e-05,
|
365 |
+
"loss": 3.1254,
|
366 |
"step": 25500
|
367 |
},
|
368 |
{
|
369 |
+
"epoch": 0.29743519344727387,
|
370 |
+
"grad_norm": 6.3333821296691895,
|
371 |
+
"learning_rate": 3.512824032763631e-05,
|
372 |
+
"loss": 3.1543,
|
373 |
"step": 26000
|
374 |
},
|
375 |
{
|
376 |
+
"epoch": 0.30315510101356763,
|
377 |
+
"grad_norm": 6.0027756690979,
|
378 |
+
"learning_rate": 3.4842244949321625e-05,
|
379 |
+
"loss": 3.1379,
|
380 |
"step": 26500
|
381 |
},
|
382 |
{
|
383 |
+
"epoch": 0.30887500857986133,
|
384 |
+
"grad_norm": 5.95717716217041,
|
385 |
+
"learning_rate": 3.455624957100693e-05,
|
386 |
+
"loss": 3.1188,
|
387 |
"step": 27000
|
388 |
},
|
389 |
{
|
390 |
+
"epoch": 0.3145949161461551,
|
391 |
+
"grad_norm": 6.262216567993164,
|
392 |
+
"learning_rate": 3.427025419269225e-05,
|
393 |
+
"loss": 3.1096,
|
394 |
"step": 27500
|
395 |
},
|
396 |
{
|
397 |
+
"epoch": 0.3203148237124488,
|
398 |
+
"grad_norm": 6.436416149139404,
|
399 |
+
"learning_rate": 3.398425881437756e-05,
|
400 |
+
"loss": 3.1294,
|
401 |
"step": 28000
|
402 |
},
|
403 |
+
{
|
404 |
+
"epoch": 0.32603473127874255,
|
405 |
+
"grad_norm": 5.524046421051025,
|
406 |
+
"learning_rate": 3.3698263436062876e-05,
|
407 |
+
"loss": 3.0965,
|
408 |
+
"step": 28500
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 0.33175463884503625,
|
412 |
+
"grad_norm": 5.237400531768799,
|
413 |
+
"learning_rate": 3.341226805774818e-05,
|
414 |
+
"loss": 3.0826,
|
415 |
+
"step": 29000
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 0.33747454641133,
|
419 |
+
"grad_norm": 5.551352500915527,
|
420 |
+
"learning_rate": 3.3126272679433504e-05,
|
421 |
+
"loss": 3.0666,
|
422 |
+
"step": 29500
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 0.3431944539776237,
|
426 |
+
"grad_norm": 5.407064914703369,
|
427 |
+
"learning_rate": 3.284027730111882e-05,
|
428 |
+
"loss": 3.0738,
|
429 |
+
"step": 30000
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 0.3489143615439175,
|
433 |
+
"grad_norm": 5.428358554840088,
|
434 |
+
"learning_rate": 3.255428192280413e-05,
|
435 |
+
"loss": 3.0628,
|
436 |
+
"step": 30500
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 0.3546342691102112,
|
440 |
+
"grad_norm": 6.178744792938232,
|
441 |
+
"learning_rate": 3.226828654448944e-05,
|
442 |
+
"loss": 3.0683,
|
443 |
+
"step": 31000
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 0.36035417667650493,
|
447 |
+
"grad_norm": 5.695249080657959,
|
448 |
+
"learning_rate": 3.1982291166174755e-05,
|
449 |
+
"loss": 3.0732,
|
450 |
+
"step": 31500
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 0.36607408424279864,
|
454 |
+
"grad_norm": 5.676379203796387,
|
455 |
+
"learning_rate": 3.169629578786007e-05,
|
456 |
+
"loss": 3.0531,
|
457 |
+
"step": 32000
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 0.37179399180909234,
|
461 |
+
"grad_norm": 5.420720100402832,
|
462 |
+
"learning_rate": 3.141030040954538e-05,
|
463 |
+
"loss": 3.0472,
|
464 |
+
"step": 32500
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 0.3775138993753861,
|
468 |
+
"grad_norm": 5.6645379066467285,
|
469 |
+
"learning_rate": 3.11243050312307e-05,
|
470 |
+
"loss": 3.0391,
|
471 |
+
"step": 33000
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 0.3832338069416798,
|
475 |
+
"grad_norm": 6.123884201049805,
|
476 |
+
"learning_rate": 3.083830965291601e-05,
|
477 |
+
"loss": 3.0527,
|
478 |
+
"step": 33500
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 0.38895371450797356,
|
482 |
+
"grad_norm": 5.331460475921631,
|
483 |
+
"learning_rate": 3.0552314274601326e-05,
|
484 |
+
"loss": 3.049,
|
485 |
+
"step": 34000
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 0.39467362207426726,
|
489 |
+
"grad_norm": 6.356675624847412,
|
490 |
+
"learning_rate": 3.0266318896286637e-05,
|
491 |
+
"loss": 3.0416,
|
492 |
+
"step": 34500
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 0.400393529640561,
|
496 |
+
"grad_norm": 6.067440986633301,
|
497 |
+
"learning_rate": 2.9980323517971955e-05,
|
498 |
+
"loss": 3.0303,
|
499 |
+
"step": 35000
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 0.4061134372068547,
|
503 |
+
"grad_norm": 5.3762030601501465,
|
504 |
+
"learning_rate": 2.9694328139657262e-05,
|
505 |
+
"loss": 3.0363,
|
506 |
+
"step": 35500
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 0.4118333447731485,
|
510 |
+
"grad_norm": 5.837817192077637,
|
511 |
+
"learning_rate": 2.9408332761342576e-05,
|
512 |
+
"loss": 3.0284,
|
513 |
+
"step": 36000
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 0.4175532523394422,
|
517 |
+
"grad_norm": 5.484166622161865,
|
518 |
+
"learning_rate": 2.912233738302789e-05,
|
519 |
+
"loss": 2.9946,
|
520 |
+
"step": 36500
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 0.42327315990573594,
|
524 |
+
"grad_norm": 5.776547908782959,
|
525 |
+
"learning_rate": 2.883634200471321e-05,
|
526 |
+
"loss": 3.037,
|
527 |
+
"step": 37000
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 0.42899306747202964,
|
531 |
+
"grad_norm": 5.481433868408203,
|
532 |
+
"learning_rate": 2.8550346626398516e-05,
|
533 |
+
"loss": 3.0289,
|
534 |
+
"step": 37500
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 0.4347129750383234,
|
538 |
+
"grad_norm": 5.784084320068359,
|
539 |
+
"learning_rate": 2.826435124808383e-05,
|
540 |
+
"loss": 2.9928,
|
541 |
+
"step": 38000
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 0.4404328826046171,
|
545 |
+
"grad_norm": 5.899621486663818,
|
546 |
+
"learning_rate": 2.7978355869769148e-05,
|
547 |
+
"loss": 3.0009,
|
548 |
+
"step": 38500
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 0.44615279017091086,
|
552 |
+
"grad_norm": 5.488452434539795,
|
553 |
+
"learning_rate": 2.7692360491454462e-05,
|
554 |
+
"loss": 3.0099,
|
555 |
+
"step": 39000
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 0.45187269773720457,
|
559 |
+
"grad_norm": 5.848759174346924,
|
560 |
+
"learning_rate": 2.740636511313977e-05,
|
561 |
+
"loss": 3.0009,
|
562 |
+
"step": 39500
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 0.4575926053034983,
|
566 |
+
"grad_norm": 5.612068176269531,
|
567 |
+
"learning_rate": 2.7120369734825084e-05,
|
568 |
+
"loss": 2.9828,
|
569 |
+
"step": 40000
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 0.463312512869792,
|
573 |
+
"grad_norm": 5.79826021194458,
|
574 |
+
"learning_rate": 2.68343743565104e-05,
|
575 |
+
"loss": 3.0152,
|
576 |
+
"step": 40500
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 0.46903242043608573,
|
580 |
+
"grad_norm": 5.842123508453369,
|
581 |
+
"learning_rate": 2.6548378978195716e-05,
|
582 |
+
"loss": 2.9719,
|
583 |
+
"step": 41000
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 0.4747523280023795,
|
587 |
+
"grad_norm": 5.69782018661499,
|
588 |
+
"learning_rate": 2.626238359988103e-05,
|
589 |
+
"loss": 2.975,
|
590 |
+
"step": 41500
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 0.4804722355686732,
|
594 |
+
"grad_norm": 6.189919948577881,
|
595 |
+
"learning_rate": 2.5976388221566338e-05,
|
596 |
+
"loss": 2.9709,
|
597 |
+
"step": 42000
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 0.48619214313496695,
|
601 |
+
"grad_norm": 5.761579513549805,
|
602 |
+
"learning_rate": 2.5690392843251655e-05,
|
603 |
+
"loss": 2.9583,
|
604 |
+
"step": 42500
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 0.49191205070126065,
|
608 |
+
"grad_norm": 6.164900779724121,
|
609 |
+
"learning_rate": 2.540439746493697e-05,
|
610 |
+
"loss": 2.9742,
|
611 |
+
"step": 43000
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 0.4976319582675544,
|
615 |
+
"grad_norm": 5.3809285163879395,
|
616 |
+
"learning_rate": 2.5118402086622284e-05,
|
617 |
+
"loss": 2.9849,
|
618 |
+
"step": 43500
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 0.5033518658338482,
|
622 |
+
"grad_norm": 5.787545680999756,
|
623 |
+
"learning_rate": 2.4832406708307595e-05,
|
624 |
+
"loss": 2.963,
|
625 |
+
"step": 44000
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 0.5090717734001419,
|
629 |
+
"grad_norm": 5.825649261474609,
|
630 |
+
"learning_rate": 2.454641132999291e-05,
|
631 |
+
"loss": 2.9634,
|
632 |
+
"step": 44500
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 0.5147916809664356,
|
636 |
+
"grad_norm": 5.936666488647461,
|
637 |
+
"learning_rate": 2.4260415951678223e-05,
|
638 |
+
"loss": 2.9678,
|
639 |
+
"step": 45000
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 0.5205115885327293,
|
643 |
+
"grad_norm": 5.980503082275391,
|
644 |
+
"learning_rate": 2.3974420573363534e-05,
|
645 |
+
"loss": 2.9675,
|
646 |
+
"step": 45500
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 0.5262314960990231,
|
650 |
+
"grad_norm": 5.755555629730225,
|
651 |
+
"learning_rate": 2.368842519504885e-05,
|
652 |
+
"loss": 2.9597,
|
653 |
+
"step": 46000
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 0.5319514036653168,
|
657 |
+
"grad_norm": 5.1978936195373535,
|
658 |
+
"learning_rate": 2.3402429816734163e-05,
|
659 |
+
"loss": 2.947,
|
660 |
+
"step": 46500
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 0.5376713112316105,
|
664 |
+
"grad_norm": 5.265974521636963,
|
665 |
+
"learning_rate": 2.3116434438419477e-05,
|
666 |
+
"loss": 2.9521,
|
667 |
+
"step": 47000
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 0.5433912187979042,
|
671 |
+
"grad_norm": 6.028165340423584,
|
672 |
+
"learning_rate": 2.2830439060104788e-05,
|
673 |
+
"loss": 2.9579,
|
674 |
+
"step": 47500
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 0.549111126364198,
|
678 |
+
"grad_norm": 5.533000946044922,
|
679 |
+
"learning_rate": 2.2544443681790102e-05,
|
680 |
+
"loss": 2.9541,
|
681 |
+
"step": 48000
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 0.5548310339304917,
|
685 |
+
"grad_norm": 5.428481101989746,
|
686 |
+
"learning_rate": 2.225844830347542e-05,
|
687 |
+
"loss": 2.9532,
|
688 |
+
"step": 48500
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 0.5605509414967854,
|
692 |
+
"grad_norm": 5.905336856842041,
|
693 |
+
"learning_rate": 2.197245292516073e-05,
|
694 |
+
"loss": 2.94,
|
695 |
+
"step": 49000
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 0.5662708490630791,
|
699 |
+
"grad_norm": 6.032477855682373,
|
700 |
+
"learning_rate": 2.1686457546846045e-05,
|
701 |
+
"loss": 2.9435,
|
702 |
+
"step": 49500
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 0.5719907566293728,
|
706 |
+
"grad_norm": 5.996410369873047,
|
707 |
+
"learning_rate": 2.1400462168531356e-05,
|
708 |
+
"loss": 2.9358,
|
709 |
+
"step": 50000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 0.5777106641956666,
|
713 |
+
"grad_norm": 5.634001731872559,
|
714 |
+
"learning_rate": 2.1114466790216674e-05,
|
715 |
+
"loss": 2.9361,
|
716 |
+
"step": 50500
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 0.5834305717619603,
|
720 |
+
"grad_norm": 5.509332656860352,
|
721 |
+
"learning_rate": 2.0828471411901985e-05,
|
722 |
+
"loss": 2.9314,
|
723 |
+
"step": 51000
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 0.589150479328254,
|
727 |
+
"grad_norm": 6.294771194458008,
|
728 |
+
"learning_rate": 2.05424760335873e-05,
|
729 |
+
"loss": 2.9238,
|
730 |
+
"step": 51500
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 0.5948703868945477,
|
734 |
+
"grad_norm": 5.542776107788086,
|
735 |
+
"learning_rate": 2.025648065527261e-05,
|
736 |
+
"loss": 2.9226,
|
737 |
+
"step": 52000
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 0.6005902944608416,
|
741 |
+
"grad_norm": 5.870414733886719,
|
742 |
+
"learning_rate": 1.9970485276957927e-05,
|
743 |
+
"loss": 2.9202,
|
744 |
+
"step": 52500
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 0.6063102020271353,
|
748 |
+
"grad_norm": 6.047051429748535,
|
749 |
+
"learning_rate": 1.9684489898643238e-05,
|
750 |
+
"loss": 2.9285,
|
751 |
+
"step": 53000
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 0.612030109593429,
|
755 |
+
"grad_norm": 5.594234943389893,
|
756 |
+
"learning_rate": 1.9398494520328553e-05,
|
757 |
+
"loss": 2.9307,
|
758 |
+
"step": 53500
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 0.6177500171597227,
|
762 |
+
"grad_norm": 5.4298295974731445,
|
763 |
+
"learning_rate": 1.9112499142013863e-05,
|
764 |
+
"loss": 2.9382,
|
765 |
+
"step": 54000
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 0.6234699247260165,
|
769 |
+
"grad_norm": 6.184563636779785,
|
770 |
+
"learning_rate": 1.882650376369918e-05,
|
771 |
+
"loss": 2.9341,
|
772 |
+
"step": 54500
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 0.6291898322923102,
|
776 |
+
"grad_norm": 5.776815414428711,
|
777 |
+
"learning_rate": 1.8540508385384492e-05,
|
778 |
+
"loss": 2.9287,
|
779 |
+
"step": 55000
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 0.6349097398586039,
|
783 |
+
"grad_norm": 5.83139181137085,
|
784 |
+
"learning_rate": 1.8254513007069806e-05,
|
785 |
+
"loss": 2.9007,
|
786 |
+
"step": 55500
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 0.6406296474248976,
|
790 |
+
"grad_norm": 5.469008922576904,
|
791 |
+
"learning_rate": 1.7968517628755117e-05,
|
792 |
+
"loss": 2.904,
|
793 |
+
"step": 56000
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 0.6463495549911913,
|
797 |
+
"grad_norm": 6.898833751678467,
|
798 |
+
"learning_rate": 1.7682522250440435e-05,
|
799 |
+
"loss": 2.9105,
|
800 |
+
"step": 56500
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 0.6520694625574851,
|
804 |
+
"grad_norm": 5.798022747039795,
|
805 |
+
"learning_rate": 1.739652687212575e-05,
|
806 |
+
"loss": 2.8995,
|
807 |
+
"step": 57000
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 0.6577893701237788,
|
811 |
+
"grad_norm": 5.57025146484375,
|
812 |
+
"learning_rate": 1.711053149381106e-05,
|
813 |
+
"loss": 2.9093,
|
814 |
+
"step": 57500
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 0.6635092776900725,
|
818 |
+
"grad_norm": 5.779621124267578,
|
819 |
+
"learning_rate": 1.6824536115496374e-05,
|
820 |
+
"loss": 2.9118,
|
821 |
+
"step": 58000
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 0.6692291852563662,
|
825 |
+
"grad_norm": 5.683529853820801,
|
826 |
+
"learning_rate": 1.653854073718169e-05,
|
827 |
+
"loss": 2.9102,
|
828 |
+
"step": 58500
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 0.67494909282266,
|
832 |
+
"grad_norm": 6.1109538078308105,
|
833 |
+
"learning_rate": 1.6252545358867003e-05,
|
834 |
+
"loss": 2.9102,
|
835 |
+
"step": 59000
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 0.6806690003889537,
|
839 |
+
"grad_norm": 5.536868095397949,
|
840 |
+
"learning_rate": 1.5966549980552314e-05,
|
841 |
+
"loss": 2.8966,
|
842 |
+
"step": 59500
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 0.6863889079552474,
|
846 |
+
"grad_norm": 6.560556888580322,
|
847 |
+
"learning_rate": 1.5680554602237628e-05,
|
848 |
+
"loss": 2.8896,
|
849 |
+
"step": 60000
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 0.6921088155215411,
|
853 |
+
"grad_norm": 5.969814300537109,
|
854 |
+
"learning_rate": 1.5394559223922942e-05,
|
855 |
+
"loss": 2.8961,
|
856 |
+
"step": 60500
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 0.697828723087835,
|
860 |
+
"grad_norm": 5.238883018493652,
|
861 |
+
"learning_rate": 1.5108563845608257e-05,
|
862 |
+
"loss": 2.882,
|
863 |
+
"step": 61000
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 0.7035486306541286,
|
867 |
+
"grad_norm": 5.538156509399414,
|
868 |
+
"learning_rate": 1.4822568467293567e-05,
|
869 |
+
"loss": 2.8955,
|
870 |
+
"step": 61500
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 0.7092685382204224,
|
874 |
+
"grad_norm": 6.287049770355225,
|
875 |
+
"learning_rate": 1.4536573088978883e-05,
|
876 |
+
"loss": 2.8932,
|
877 |
+
"step": 62000
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 0.714988445786716,
|
881 |
+
"grad_norm": 6.05150032043457,
|
882 |
+
"learning_rate": 1.4250577710664196e-05,
|
883 |
+
"loss": 2.8868,
|
884 |
+
"step": 62500
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 0.7207083533530099,
|
888 |
+
"grad_norm": 5.857907295227051,
|
889 |
+
"learning_rate": 1.396458233234951e-05,
|
890 |
+
"loss": 2.8872,
|
891 |
+
"step": 63000
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 0.7264282609193036,
|
895 |
+
"grad_norm": 5.748091697692871,
|
896 |
+
"learning_rate": 1.3678586954034823e-05,
|
897 |
+
"loss": 2.8964,
|
898 |
+
"step": 63500
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 0.7321481684855973,
|
902 |
+
"grad_norm": 5.55509614944458,
|
903 |
+
"learning_rate": 1.3392591575720137e-05,
|
904 |
+
"loss": 2.8758,
|
905 |
+
"step": 64000
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 0.737868076051891,
|
909 |
+
"grad_norm": 6.135568618774414,
|
910 |
+
"learning_rate": 1.3106596197405451e-05,
|
911 |
+
"loss": 2.8685,
|
912 |
+
"step": 64500
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 0.7435879836181847,
|
916 |
+
"grad_norm": 5.816187381744385,
|
917 |
+
"learning_rate": 1.2820600819090764e-05,
|
918 |
+
"loss": 2.8802,
|
919 |
+
"step": 65000
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 0.7493078911844785,
|
923 |
+
"grad_norm": 6.309732437133789,
|
924 |
+
"learning_rate": 1.2534605440776078e-05,
|
925 |
+
"loss": 2.8859,
|
926 |
+
"step": 65500
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 0.7550277987507722,
|
930 |
+
"grad_norm": 5.721366882324219,
|
931 |
+
"learning_rate": 1.2248610062461391e-05,
|
932 |
+
"loss": 2.8702,
|
933 |
+
"step": 66000
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 0.7607477063170659,
|
937 |
+
"grad_norm": 5.648174285888672,
|
938 |
+
"learning_rate": 1.1962614684146704e-05,
|
939 |
+
"loss": 2.8603,
|
940 |
+
"step": 66500
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 0.7664676138833596,
|
944 |
+
"grad_norm": 6.4422688484191895,
|
945 |
+
"learning_rate": 1.1676619305832018e-05,
|
946 |
+
"loss": 2.875,
|
947 |
+
"step": 67000
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 0.7721875214496534,
|
951 |
+
"grad_norm": 5.673862934112549,
|
952 |
+
"learning_rate": 1.1390623927517332e-05,
|
953 |
+
"loss": 2.8743,
|
954 |
+
"step": 67500
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 0.7779074290159471,
|
958 |
+
"grad_norm": 5.87379789352417,
|
959 |
+
"learning_rate": 1.1104628549202645e-05,
|
960 |
+
"loss": 2.8678,
|
961 |
+
"step": 68000
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 0.7836273365822408,
|
965 |
+
"grad_norm": 5.968353271484375,
|
966 |
+
"learning_rate": 1.0818633170887959e-05,
|
967 |
+
"loss": 2.8797,
|
968 |
+
"step": 68500
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 0.7893472441485345,
|
972 |
+
"grad_norm": 5.516451835632324,
|
973 |
+
"learning_rate": 1.0532637792573273e-05,
|
974 |
+
"loss": 2.8716,
|
975 |
+
"step": 69000
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 0.7950671517148283,
|
979 |
+
"grad_norm": 6.277103900909424,
|
980 |
+
"learning_rate": 1.0246642414258586e-05,
|
981 |
+
"loss": 2.8483,
|
982 |
+
"step": 69500
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 0.800787059281122,
|
986 |
+
"grad_norm": 5.54793643951416,
|
987 |
+
"learning_rate": 9.9606470359439e-06,
|
988 |
+
"loss": 2.8462,
|
989 |
+
"step": 70000
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 0.8065069668474157,
|
993 |
+
"grad_norm": 5.989738464355469,
|
994 |
+
"learning_rate": 9.674651657629213e-06,
|
995 |
+
"loss": 2.8672,
|
996 |
+
"step": 70500
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 0.8122268744137094,
|
1000 |
+
"grad_norm": 5.7795844078063965,
|
1001 |
+
"learning_rate": 9.388656279314527e-06,
|
1002 |
+
"loss": 2.8566,
|
1003 |
+
"step": 71000
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 0.8179467819800033,
|
1007 |
+
"grad_norm": 5.732882976531982,
|
1008 |
+
"learning_rate": 9.10266090099984e-06,
|
1009 |
+
"loss": 2.8497,
|
1010 |
+
"step": 71500
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 0.823666689546297,
|
1014 |
+
"grad_norm": 6.427890777587891,
|
1015 |
+
"learning_rate": 8.816665522685154e-06,
|
1016 |
+
"loss": 2.8656,
|
1017 |
+
"step": 72000
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 0.8293865971125907,
|
1021 |
+
"grad_norm": 6.1445746421813965,
|
1022 |
+
"learning_rate": 8.530670144370468e-06,
|
1023 |
+
"loss": 2.8425,
|
1024 |
+
"step": 72500
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 0.8351065046788844,
|
1028 |
+
"grad_norm": 6.342021465301514,
|
1029 |
+
"learning_rate": 8.24467476605578e-06,
|
1030 |
+
"loss": 2.859,
|
1031 |
+
"step": 73000
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 0.8408264122451781,
|
1035 |
+
"grad_norm": 6.117573261260986,
|
1036 |
+
"learning_rate": 7.958679387741095e-06,
|
1037 |
+
"loss": 2.8631,
|
1038 |
+
"step": 73500
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 0.8465463198114719,
|
1042 |
+
"grad_norm": 6.145172595977783,
|
1043 |
+
"learning_rate": 7.672684009426408e-06,
|
1044 |
+
"loss": 2.8536,
|
1045 |
+
"step": 74000
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 0.8522662273777656,
|
1049 |
+
"grad_norm": 5.709812641143799,
|
1050 |
+
"learning_rate": 7.386688631111721e-06,
|
1051 |
+
"loss": 2.8546,
|
1052 |
+
"step": 74500
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 0.8579861349440593,
|
1056 |
+
"grad_norm": 6.244381904602051,
|
1057 |
+
"learning_rate": 7.100693252797034e-06,
|
1058 |
+
"loss": 2.8277,
|
1059 |
+
"step": 75000
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 0.863706042510353,
|
1063 |
+
"grad_norm": 6.229698181152344,
|
1064 |
+
"learning_rate": 6.814697874482348e-06,
|
1065 |
+
"loss": 2.85,
|
1066 |
+
"step": 75500
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 0.8694259500766468,
|
1070 |
+
"grad_norm": 5.480854511260986,
|
1071 |
+
"learning_rate": 6.528702496167661e-06,
|
1072 |
+
"loss": 2.8576,
|
1073 |
+
"step": 76000
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 0.8751458576429405,
|
1077 |
+
"grad_norm": 6.088413715362549,
|
1078 |
+
"learning_rate": 6.2427071178529756e-06,
|
1079 |
+
"loss": 2.8618,
|
1080 |
+
"step": 76500
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 0.8808657652092342,
|
1084 |
+
"grad_norm": 6.159554481506348,
|
1085 |
+
"learning_rate": 5.956711739538289e-06,
|
1086 |
+
"loss": 2.8319,
|
1087 |
+
"step": 77000
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 0.8865856727755279,
|
1091 |
+
"grad_norm": 6.287491798400879,
|
1092 |
+
"learning_rate": 5.6707163612236024e-06,
|
1093 |
+
"loss": 2.8622,
|
1094 |
+
"step": 77500
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 0.8923055803418217,
|
1098 |
+
"grad_norm": 6.237947940826416,
|
1099 |
+
"learning_rate": 5.384720982908916e-06,
|
1100 |
+
"loss": 2.8539,
|
1101 |
+
"step": 78000
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 0.8980254879081154,
|
1105 |
+
"grad_norm": 6.096075057983398,
|
1106 |
+
"learning_rate": 5.09872560459423e-06,
|
1107 |
+
"loss": 2.8256,
|
1108 |
+
"step": 78500
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 0.9037453954744091,
|
1112 |
+
"grad_norm": 6.158285140991211,
|
1113 |
+
"learning_rate": 4.812730226279544e-06,
|
1114 |
+
"loss": 2.8476,
|
1115 |
+
"step": 79000
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 0.9094653030407028,
|
1119 |
+
"grad_norm": 6.239354133605957,
|
1120 |
+
"learning_rate": 4.526734847964857e-06,
|
1121 |
+
"loss": 2.8538,
|
1122 |
+
"step": 79500
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 0.9151852106069966,
|
1126 |
+
"grad_norm": 5.273146629333496,
|
1127 |
+
"learning_rate": 4.2407394696501705e-06,
|
1128 |
+
"loss": 2.849,
|
1129 |
+
"step": 80000
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 0.9209051181732903,
|
1133 |
+
"grad_norm": 6.121274948120117,
|
1134 |
+
"learning_rate": 3.954744091335484e-06,
|
1135 |
+
"loss": 2.8454,
|
1136 |
+
"step": 80500
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 0.926625025739584,
|
1140 |
+
"grad_norm": 6.213692665100098,
|
1141 |
+
"learning_rate": 3.6687487130207977e-06,
|
1142 |
+
"loss": 2.8431,
|
1143 |
+
"step": 81000
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 0.9323449333058778,
|
1147 |
+
"grad_norm": 6.369662761688232,
|
1148 |
+
"learning_rate": 3.382753334706111e-06,
|
1149 |
+
"loss": 2.8589,
|
1150 |
+
"step": 81500
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 0.9380648408721715,
|
1154 |
+
"grad_norm": 5.934210300445557,
|
1155 |
+
"learning_rate": 3.096757956391425e-06,
|
1156 |
+
"loss": 2.8371,
|
1157 |
+
"step": 82000
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 0.9437847484384653,
|
1161 |
+
"grad_norm": 6.598156452178955,
|
1162 |
+
"learning_rate": 2.8107625780767385e-06,
|
1163 |
+
"loss": 2.8282,
|
1164 |
+
"step": 82500
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 0.949504656004759,
|
1168 |
+
"grad_norm": 5.81234073638916,
|
1169 |
+
"learning_rate": 2.524767199762052e-06,
|
1170 |
+
"loss": 2.8244,
|
1171 |
+
"step": 83000
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 0.9552245635710527,
|
1175 |
+
"grad_norm": 6.145198822021484,
|
1176 |
+
"learning_rate": 2.2387718214473658e-06,
|
1177 |
+
"loss": 2.8294,
|
1178 |
+
"step": 83500
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 0.9609444711373464,
|
1182 |
+
"grad_norm": 5.321970462799072,
|
1183 |
+
"learning_rate": 1.952776443132679e-06,
|
1184 |
+
"loss": 2.8425,
|
1185 |
+
"step": 84000
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 0.9666643787036402,
|
1189 |
+
"grad_norm": 5.8561248779296875,
|
1190 |
+
"learning_rate": 1.6667810648179926e-06,
|
1191 |
+
"loss": 2.8618,
|
1192 |
+
"step": 84500
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 0.9723842862699339,
|
1196 |
+
"grad_norm": 6.001429080963135,
|
1197 |
+
"learning_rate": 1.3807856865033063e-06,
|
1198 |
+
"loss": 2.8456,
|
1199 |
+
"step": 85000
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 0.9781041938362276,
|
1203 |
+
"grad_norm": 6.074833393096924,
|
1204 |
+
"learning_rate": 1.0947903081886197e-06,
|
1205 |
+
"loss": 2.8408,
|
1206 |
+
"step": 85500
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 0.9838241014025213,
|
1210 |
+
"grad_norm": 5.579460144042969,
|
1211 |
+
"learning_rate": 8.087949298739332e-07,
|
1212 |
+
"loss": 2.8427,
|
1213 |
+
"step": 86000
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 0.9895440089688151,
|
1217 |
+
"grad_norm": 5.968313217163086,
|
1218 |
+
"learning_rate": 5.227995515592468e-07,
|
1219 |
+
"loss": 2.8345,
|
1220 |
+
"step": 86500
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 0.9952639165351088,
|
1224 |
+
"grad_norm": 5.920719146728516,
|
1225 |
+
"learning_rate": 2.3680417324456038e-07,
|
1226 |
+
"loss": 2.8337,
|
1227 |
+
"step": 87000
|
1228 |
+
},
|
1229 |
{
|
1230 |
"epoch": 1.0,
|
1231 |
+
"step": 87414,
|
1232 |
+
"total_flos": 4.634223291773338e+16,
|
1233 |
+
"train_loss": 3.132497888326312,
|
1234 |
+
"train_runtime": 20429.0413,
|
1235 |
+
"train_samples_per_second": 34.231,
|
1236 |
+
"train_steps_per_second": 4.279
|
1237 |
}
|
1238 |
],
|
1239 |
"logging_steps": 500,
|
1240 |
+
"max_steps": 87414,
|
1241 |
"num_input_tokens_seen": 0,
|
1242 |
"num_train_epochs": 1,
|
1243 |
"save_steps": 500,
|
|
|
1253 |
"attributes": {}
|
1254 |
}
|
1255 |
},
|
1256 |
+
"total_flos": 4.634223291773338e+16,
|
1257 |
"train_batch_size": 8,
|
1258 |
"trial_name": null,
|
1259 |
"trial_params": null
|
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 sha256:8c62c2584fefcb587403e55ceb246ccd105b535709a1ff9735da2465de367912
|
3 |
+
size 5368
|