End of training
Browse files- README.md +16 -13
- all_results.json +11 -11
- config.json +10 -10
- eval_results.json +6 -6
- model.safetensors +1 -1
- runs/Apr26_13-16-15_60e8964599d3/events.out.tfevents.1714137383.60e8964599d3.35.0 +3 -0
- runs/Apr26_13-16-15_60e8964599d3/events.out.tfevents.1714139135.60e8964599d3.35.1 +3 -0
- train_results.json +6 -6
- trainer_state.json +204 -128
- training_args.bin +1 -1
README.md
CHANGED
@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
17 |
|
18 |
This model is a fine-tuned version of [facebook/hubert-base-ls960](https://huggingface.co/facebook/hubert-base-ls960) on an unknown dataset.
|
19 |
It achieves the following results on the evaluation set:
|
20 |
-
- Loss: 0.
|
21 |
-
- Accuracy: 0.
|
22 |
|
23 |
## Model description
|
24 |
|
@@ -52,17 +52,20 @@ The following hyperparameters were used during training:
|
|
52 |
|
53 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
54 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
55 |
-
|
|
56 |
-
| 0.
|
57 |
-
| 0.
|
58 |
-
| 0.
|
59 |
-
| 0.
|
60 |
-
| 0.
|
61 |
-
| 0.
|
62 |
-
| 0.
|
63 |
-
| 0.
|
64 |
-
| 0.
|
65 |
-
| 0.
|
|
|
|
|
|
|
66 |
|
67 |
|
68 |
### Framework versions
|
|
|
17 |
|
18 |
This model is a fine-tuned version of [facebook/hubert-base-ls960](https://huggingface.co/facebook/hubert-base-ls960) on an unknown dataset.
|
19 |
It achieves the following results on the evaluation set:
|
20 |
+
- Loss: 0.2980
|
21 |
+
- Accuracy: 0.9545
|
22 |
|
23 |
## Model description
|
24 |
|
|
|
52 |
|
53 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
54 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
55 |
+
| 1.1628 | 1.0 | 223 | 0.7126 | 0.7727 |
|
56 |
+
| 0.6562 | 2.0 | 446 | 0.5069 | 0.8485 |
|
57 |
+
| 0.4199 | 3.0 | 669 | 0.3570 | 0.8990 |
|
58 |
+
| 0.325 | 4.0 | 892 | 0.2092 | 0.9394 |
|
59 |
+
| 0.2217 | 5.0 | 1115 | 0.2392 | 0.9444 |
|
60 |
+
| 0.1831 | 6.0 | 1338 | 0.2754 | 0.9293 |
|
61 |
+
| 0.1598 | 7.0 | 1561 | 0.3294 | 0.9343 |
|
62 |
+
| 0.1676 | 8.0 | 1784 | 0.2669 | 0.9495 |
|
63 |
+
| 0.1597 | 9.0 | 2007 | 0.3438 | 0.9293 |
|
64 |
+
| 0.1132 | 10.0 | 2230 | 0.3159 | 0.9444 |
|
65 |
+
| 0.1224 | 11.0 | 2453 | 0.2980 | 0.9545 |
|
66 |
+
| 0.095 | 12.0 | 2676 | 0.2970 | 0.9444 |
|
67 |
+
| 0.1087 | 13.0 | 2899 | 0.3449 | 0.9343 |
|
68 |
+
| 0.1254 | 14.0 | 3122 | 0.3198 | 0.9444 |
|
69 |
|
70 |
|
71 |
### Framework versions
|
all_results.json
CHANGED
@@ -1,14 +1,14 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_loss": 0.
|
5 |
-
"eval_runtime":
|
6 |
-
"eval_samples_per_second":
|
7 |
-
"eval_steps_per_second": 5.
|
8 |
-
"total_flos":
|
9 |
-
"train_loss": 0.
|
10 |
-
"train_runtime":
|
11 |
"train_samples": 1781,
|
12 |
-
"train_samples_per_second":
|
13 |
-
"train_steps_per_second":
|
14 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 14.0,
|
3 |
+
"eval_accuracy": 0.9545454382896423,
|
4 |
+
"eval_loss": 0.29796990752220154,
|
5 |
+
"eval_runtime": 8.9823,
|
6 |
+
"eval_samples_per_second": 22.043,
|
7 |
+
"eval_steps_per_second": 5.566,
|
8 |
+
"total_flos": 7.9842219974856e+17,
|
9 |
+
"train_loss": 0.30023268478028214,
|
10 |
+
"train_runtime": 1742.0026,
|
11 |
"train_samples": 1781,
|
12 |
+
"train_samples_per_second": 15.336,
|
13 |
+
"train_steps_per_second": 1.92
|
14 |
}
|
config.json
CHANGED
@@ -53,20 +53,20 @@
|
|
53 |
"hidden_dropout_prob": 0.1,
|
54 |
"hidden_size": 768,
|
55 |
"id2label": {
|
56 |
-
"0": "
|
57 |
-
"1": "
|
58 |
-
"2": "
|
59 |
-
"3": "
|
60 |
-
"4": "
|
61 |
},
|
62 |
"initializer_range": 0.02,
|
63 |
"intermediate_size": 3072,
|
64 |
"label2id": {
|
65 |
-
"
|
66 |
-
"
|
67 |
-
"
|
68 |
-
"
|
69 |
-
"
|
70 |
},
|
71 |
"layer_norm_eps": 1e-05,
|
72 |
"layerdrop": 0.1,
|
|
|
53 |
"hidden_dropout_prob": 0.1,
|
54 |
"hidden_size": 768,
|
55 |
"id2label": {
|
56 |
+
"0": "01Neutral",
|
57 |
+
"1": "02Fearful",
|
58 |
+
"2": "03Happy",
|
59 |
+
"3": "04Sad",
|
60 |
+
"4": "05Angry"
|
61 |
},
|
62 |
"initializer_range": 0.02,
|
63 |
"intermediate_size": 3072,
|
64 |
"label2id": {
|
65 |
+
"01Neutral": 0,
|
66 |
+
"02Fearful": 1,
|
67 |
+
"03Happy": 2,
|
68 |
+
"04Sad": 3,
|
69 |
+
"05Angry": 4
|
70 |
},
|
71 |
"layer_norm_eps": 1e-05,
|
72 |
"layerdrop": 0.1,
|
eval_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_loss": 0.
|
5 |
-
"eval_runtime":
|
6 |
-
"eval_samples_per_second":
|
7 |
-
"eval_steps_per_second": 5.
|
8 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 14.0,
|
3 |
+
"eval_accuracy": 0.9545454382896423,
|
4 |
+
"eval_loss": 0.29796990752220154,
|
5 |
+
"eval_runtime": 8.9823,
|
6 |
+
"eval_samples_per_second": 22.043,
|
7 |
+
"eval_steps_per_second": 5.566
|
8 |
}
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 379890236
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5823ad5dc453f63bc711ccd08c86437beecb7c407032319c5660cf83beb4b2fd
|
3 |
size 379890236
|
runs/Apr26_13-16-15_60e8964599d3/events.out.tfevents.1714137383.60e8964599d3.35.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4dc75d4da3ec31eec728b4ec220ce2370e52770367a3ff760027c9b8782c8e90
|
3 |
+
size 17226
|
runs/Apr26_13-16-15_60e8964599d3/events.out.tfevents.1714139135.60e8964599d3.35.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6f9a7e227a5da52a6596d6b0f73bbcb80b399ac6e1ebdbe2c906f1682b73f3e8
|
3 |
+
size 734
|
train_results.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"total_flos":
|
4 |
-
"train_loss": 0.
|
5 |
-
"train_runtime":
|
6 |
"train_samples": 1781,
|
7 |
-
"train_samples_per_second":
|
8 |
-
"train_steps_per_second":
|
9 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 14.0,
|
3 |
+
"total_flos": 7.9842219974856e+17,
|
4 |
+
"train_loss": 0.30023268478028214,
|
5 |
+
"train_runtime": 1742.0026,
|
6 |
"train_samples": 1781,
|
7 |
+
"train_samples_per_second": 15.336,
|
8 |
+
"train_steps_per_second": 1.92
|
9 |
}
|
trainer_state.json
CHANGED
@@ -1,297 +1,373 @@
|
|
1 |
{
|
2 |
-
"best_metric": 0.
|
3 |
-
"best_model_checkpoint": "/kaggle/working/ckpts/checkpoint-
|
4 |
-
"epoch":
|
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": 0.45,
|
13 |
-
"grad_norm":
|
14 |
"learning_rate": 9.701046337817639e-06,
|
15 |
-
"loss": 1.
|
16 |
"step": 100
|
17 |
},
|
18 |
{
|
19 |
"epoch": 0.9,
|
20 |
-
"grad_norm": 2.
|
21 |
-
"learning_rate": 9.
|
22 |
-
"loss":
|
23 |
"step": 200
|
24 |
},
|
25 |
{
|
26 |
"epoch": 1.0,
|
27 |
-
"eval_accuracy": 0.
|
28 |
-
"eval_loss": 0.
|
29 |
-
"eval_runtime": 9.
|
30 |
-
"eval_samples_per_second": 21.
|
31 |
-
"eval_steps_per_second": 5.
|
32 |
"step": 223
|
33 |
},
|
34 |
{
|
35 |
"epoch": 1.35,
|
36 |
-
"grad_norm":
|
37 |
"learning_rate": 9.106128550074738e-06,
|
38 |
-
"loss": 0.
|
39 |
"step": 300
|
40 |
},
|
41 |
{
|
42 |
"epoch": 1.79,
|
43 |
-
"grad_norm":
|
44 |
"learning_rate": 8.807174887892378e-06,
|
45 |
-
"loss": 0.
|
46 |
"step": 400
|
47 |
},
|
48 |
{
|
49 |
"epoch": 2.0,
|
50 |
-
"eval_accuracy": 0.
|
51 |
-
"eval_loss": 0.
|
52 |
-
"eval_runtime": 9.
|
53 |
-
"eval_samples_per_second":
|
54 |
-
"eval_steps_per_second": 5.
|
55 |
"step": 446
|
56 |
},
|
57 |
{
|
58 |
"epoch": 2.24,
|
59 |
-
"grad_norm":
|
60 |
"learning_rate": 8.51121076233184e-06,
|
61 |
-
"loss": 0.
|
62 |
"step": 500
|
63 |
},
|
64 |
{
|
65 |
"epoch": 2.69,
|
66 |
-
"grad_norm":
|
67 |
-
"learning_rate": 8.
|
68 |
-
"loss": 0.
|
69 |
"step": 600
|
70 |
},
|
71 |
{
|
72 |
"epoch": 3.0,
|
73 |
-
"eval_accuracy": 0.
|
74 |
-
"eval_loss": 0.
|
75 |
-
"eval_runtime": 9.
|
76 |
-
"eval_samples_per_second":
|
77 |
-
"eval_steps_per_second": 5.
|
78 |
"step": 669
|
79 |
},
|
80 |
{
|
81 |
"epoch": 3.14,
|
82 |
-
"grad_norm":
|
83 |
"learning_rate": 7.916292974588939e-06,
|
84 |
-
"loss": 0.
|
85 |
"step": 700
|
86 |
},
|
87 |
{
|
88 |
"epoch": 3.59,
|
89 |
-
"grad_norm":
|
90 |
"learning_rate": 7.617339312406578e-06,
|
91 |
-
"loss": 0.
|
92 |
"step": 800
|
93 |
},
|
94 |
{
|
95 |
"epoch": 4.0,
|
96 |
-
"eval_accuracy": 0.
|
97 |
-
"eval_loss": 0.
|
98 |
-
"eval_runtime": 9.
|
99 |
-
"eval_samples_per_second": 21.
|
100 |
-
"eval_steps_per_second": 5.
|
101 |
"step": 892
|
102 |
},
|
103 |
{
|
104 |
"epoch": 4.04,
|
105 |
-
"grad_norm":
|
106 |
"learning_rate": 7.318385650224216e-06,
|
107 |
-
"loss": 0.
|
108 |
"step": 900
|
109 |
},
|
110 |
{
|
111 |
"epoch": 4.48,
|
112 |
-
"grad_norm":
|
113 |
"learning_rate": 7.019431988041854e-06,
|
114 |
-
"loss": 0.
|
115 |
"step": 1000
|
116 |
},
|
117 |
{
|
118 |
"epoch": 4.93,
|
119 |
-
"grad_norm":
|
120 |
-
"learning_rate": 6.
|
121 |
-
"loss": 0.
|
122 |
"step": 1100
|
123 |
},
|
124 |
{
|
125 |
"epoch": 5.0,
|
126 |
-
"eval_accuracy": 0.
|
127 |
-
"eval_loss": 0.
|
128 |
-
"eval_runtime": 9.
|
129 |
-
"eval_samples_per_second": 21.
|
130 |
-
"eval_steps_per_second": 5.
|
131 |
"step": 1115
|
132 |
},
|
133 |
{
|
134 |
"epoch": 5.38,
|
135 |
-
"grad_norm": 0.
|
136 |
-
"learning_rate": 6.
|
137 |
-
"loss": 0.
|
138 |
"step": 1200
|
139 |
},
|
140 |
{
|
141 |
"epoch": 5.83,
|
142 |
-
"grad_norm":
|
143 |
-
"learning_rate": 6.
|
144 |
-
"loss": 0.
|
145 |
"step": 1300
|
146 |
},
|
147 |
{
|
148 |
"epoch": 6.0,
|
149 |
-
"eval_accuracy": 0.
|
150 |
-
"eval_loss": 0.
|
151 |
-
"eval_runtime": 9.
|
152 |
-
"eval_samples_per_second": 21.
|
153 |
-
"eval_steps_per_second": 5.
|
154 |
"step": 1338
|
155 |
},
|
156 |
{
|
157 |
"epoch": 6.28,
|
158 |
-
"grad_norm": 0.
|
159 |
-
"learning_rate": 5.
|
160 |
-
"loss": 0.
|
161 |
"step": 1400
|
162 |
},
|
163 |
{
|
164 |
"epoch": 6.73,
|
165 |
-
"grad_norm":
|
166 |
-
"learning_rate": 5.
|
167 |
-
"loss": 0.
|
168 |
"step": 1500
|
169 |
},
|
170 |
{
|
171 |
"epoch": 7.0,
|
172 |
-
"eval_accuracy": 0.
|
173 |
-
"eval_loss": 0.
|
174 |
-
"eval_runtime": 9.
|
175 |
-
"eval_samples_per_second": 21.
|
176 |
-
"eval_steps_per_second": 5.
|
177 |
"step": 1561
|
178 |
},
|
179 |
{
|
180 |
"epoch": 7.17,
|
181 |
-
"grad_norm": 0.
|
182 |
"learning_rate": 5.228699551569507e-06,
|
183 |
-
"loss": 0.
|
184 |
"step": 1600
|
185 |
},
|
186 |
{
|
187 |
"epoch": 7.62,
|
188 |
-
"grad_norm":
|
189 |
"learning_rate": 4.929745889387145e-06,
|
190 |
-
"loss": 0.
|
191 |
"step": 1700
|
192 |
},
|
193 |
{
|
194 |
"epoch": 8.0,
|
195 |
-
"eval_accuracy": 0.
|
196 |
-
"eval_loss": 0.
|
197 |
-
"eval_runtime": 9.
|
198 |
-
"eval_samples_per_second": 21.
|
199 |
-
"eval_steps_per_second": 5.
|
200 |
"step": 1784
|
201 |
},
|
202 |
{
|
203 |
"epoch": 8.07,
|
204 |
-
"grad_norm":
|
205 |
"learning_rate": 4.630792227204783e-06,
|
206 |
-
"loss": 0.
|
207 |
"step": 1800
|
208 |
},
|
209 |
{
|
210 |
"epoch": 8.52,
|
211 |
-
"grad_norm": 0.
|
212 |
"learning_rate": 4.3318385650224224e-06,
|
213 |
-
"loss": 0.
|
214 |
"step": 1900
|
215 |
},
|
216 |
{
|
217 |
"epoch": 8.97,
|
218 |
-
"grad_norm":
|
219 |
"learning_rate": 4.03288490284006e-06,
|
220 |
-
"loss": 0.
|
221 |
"step": 2000
|
222 |
},
|
223 |
{
|
224 |
"epoch": 9.0,
|
225 |
-
"eval_accuracy": 0.
|
226 |
-
"eval_loss": 0.
|
227 |
-
"eval_runtime": 9.
|
228 |
-
"eval_samples_per_second": 21.
|
229 |
-
"eval_steps_per_second": 5.
|
230 |
"step": 2007
|
231 |
},
|
232 |
{
|
233 |
"epoch": 9.42,
|
234 |
-
"grad_norm": 0.
|
235 |
"learning_rate": 3.7339312406576984e-06,
|
236 |
-
"loss": 0.
|
237 |
"step": 2100
|
238 |
},
|
239 |
{
|
240 |
"epoch": 9.87,
|
241 |
-
"grad_norm":
|
242 |
"learning_rate": 3.4349775784753366e-06,
|
243 |
-
"loss": 0.
|
244 |
"step": 2200
|
245 |
},
|
246 |
{
|
247 |
"epoch": 10.0,
|
248 |
-
"eval_accuracy": 0.
|
249 |
-
"eval_loss": 0.
|
250 |
-
"eval_runtime": 9.
|
251 |
-
"eval_samples_per_second": 21.
|
252 |
-
"eval_steps_per_second": 5.
|
253 |
"step": 2230
|
254 |
},
|
255 |
{
|
256 |
"epoch": 10.31,
|
257 |
-
"grad_norm": 0.
|
258 |
"learning_rate": 3.136023916292975e-06,
|
259 |
-
"loss": 0.
|
260 |
"step": 2300
|
261 |
},
|
262 |
{
|
263 |
"epoch": 10.76,
|
264 |
-
"grad_norm": 0.
|
265 |
"learning_rate": 2.8370702541106134e-06,
|
266 |
-
"loss": 0.
|
267 |
"step": 2400
|
268 |
},
|
269 |
{
|
270 |
"epoch": 11.0,
|
271 |
-
"eval_accuracy": 0.
|
272 |
-
"eval_loss": 0.
|
273 |
-
"eval_runtime": 9.
|
274 |
-
"eval_samples_per_second": 21.
|
275 |
-
"eval_steps_per_second": 5.
|
276 |
"step": 2453
|
277 |
},
|
278 |
{
|
279 |
-
"epoch": 11.
|
280 |
-
"
|
281 |
-
"
|
282 |
-
"
|
283 |
-
"
|
284 |
-
"train_samples_per_second": 20.276,
|
285 |
-
"train_steps_per_second": 2.539
|
286 |
},
|
287 |
{
|
288 |
-
"epoch": 11.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
289 |
"eval_accuracy": 0.9444444179534912,
|
290 |
-
"eval_loss": 0.
|
291 |
-
"eval_runtime": 9.
|
292 |
-
"eval_samples_per_second": 21.
|
293 |
-
"eval_steps_per_second": 5.
|
294 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
295 |
}
|
296 |
],
|
297 |
"logging_steps": 100,
|
@@ -299,7 +375,7 @@
|
|
299 |
"num_input_tokens_seen": 0,
|
300 |
"num_train_epochs": 15,
|
301 |
"save_steps": 500,
|
302 |
-
"total_flos":
|
303 |
"train_batch_size": 4,
|
304 |
"trial_name": null,
|
305 |
"trial_params": null
|
|
|
1 |
{
|
2 |
+
"best_metric": 0.9545454382896423,
|
3 |
+
"best_model_checkpoint": "/kaggle/working/ckpts/checkpoint-2453",
|
4 |
+
"epoch": 14.0,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 3122,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
"epoch": 0.45,
|
13 |
+
"grad_norm": 2.3771965503692627,
|
14 |
"learning_rate": 9.701046337817639e-06,
|
15 |
+
"loss": 1.5285,
|
16 |
"step": 100
|
17 |
},
|
18 |
{
|
19 |
"epoch": 0.9,
|
20 |
+
"grad_norm": 2.858564615249634,
|
21 |
+
"learning_rate": 9.402092675635277e-06,
|
22 |
+
"loss": 1.1628,
|
23 |
"step": 200
|
24 |
},
|
25 |
{
|
26 |
"epoch": 1.0,
|
27 |
+
"eval_accuracy": 0.7727272510528564,
|
28 |
+
"eval_loss": 0.712559700012207,
|
29 |
+
"eval_runtime": 9.2054,
|
30 |
+
"eval_samples_per_second": 21.509,
|
31 |
+
"eval_steps_per_second": 5.432,
|
32 |
"step": 223
|
33 |
},
|
34 |
{
|
35 |
"epoch": 1.35,
|
36 |
+
"grad_norm": 12.418754577636719,
|
37 |
"learning_rate": 9.106128550074738e-06,
|
38 |
+
"loss": 0.8286,
|
39 |
"step": 300
|
40 |
},
|
41 |
{
|
42 |
"epoch": 1.79,
|
43 |
+
"grad_norm": 7.768007755279541,
|
44 |
"learning_rate": 8.807174887892378e-06,
|
45 |
+
"loss": 0.6562,
|
46 |
"step": 400
|
47 |
},
|
48 |
{
|
49 |
"epoch": 2.0,
|
50 |
+
"eval_accuracy": 0.8484848737716675,
|
51 |
+
"eval_loss": 0.5068599581718445,
|
52 |
+
"eval_runtime": 9.5278,
|
53 |
+
"eval_samples_per_second": 20.781,
|
54 |
+
"eval_steps_per_second": 5.248,
|
55 |
"step": 446
|
56 |
},
|
57 |
{
|
58 |
"epoch": 2.24,
|
59 |
+
"grad_norm": 9.381482124328613,
|
60 |
"learning_rate": 8.51121076233184e-06,
|
61 |
+
"loss": 0.5053,
|
62 |
"step": 500
|
63 |
},
|
64 |
{
|
65 |
"epoch": 2.69,
|
66 |
+
"grad_norm": 2.293752908706665,
|
67 |
+
"learning_rate": 8.212257100149478e-06,
|
68 |
+
"loss": 0.4199,
|
69 |
"step": 600
|
70 |
},
|
71 |
{
|
72 |
"epoch": 3.0,
|
73 |
+
"eval_accuracy": 0.8989899158477783,
|
74 |
+
"eval_loss": 0.356963574886322,
|
75 |
+
"eval_runtime": 9.2472,
|
76 |
+
"eval_samples_per_second": 21.412,
|
77 |
+
"eval_steps_per_second": 5.407,
|
78 |
"step": 669
|
79 |
},
|
80 |
{
|
81 |
"epoch": 3.14,
|
82 |
+
"grad_norm": 23.293209075927734,
|
83 |
"learning_rate": 7.916292974588939e-06,
|
84 |
+
"loss": 0.3121,
|
85 |
"step": 700
|
86 |
},
|
87 |
{
|
88 |
"epoch": 3.59,
|
89 |
+
"grad_norm": 3.8754687309265137,
|
90 |
"learning_rate": 7.617339312406578e-06,
|
91 |
+
"loss": 0.325,
|
92 |
"step": 800
|
93 |
},
|
94 |
{
|
95 |
"epoch": 4.0,
|
96 |
+
"eval_accuracy": 0.939393937587738,
|
97 |
+
"eval_loss": 0.20920716226100922,
|
98 |
+
"eval_runtime": 9.2568,
|
99 |
+
"eval_samples_per_second": 21.39,
|
100 |
+
"eval_steps_per_second": 5.401,
|
101 |
"step": 892
|
102 |
},
|
103 |
{
|
104 |
"epoch": 4.04,
|
105 |
+
"grad_norm": 62.81392288208008,
|
106 |
"learning_rate": 7.318385650224216e-06,
|
107 |
+
"loss": 0.2896,
|
108 |
"step": 900
|
109 |
},
|
110 |
{
|
111 |
"epoch": 4.48,
|
112 |
+
"grad_norm": 35.08163833618164,
|
113 |
"learning_rate": 7.019431988041854e-06,
|
114 |
+
"loss": 0.2535,
|
115 |
"step": 1000
|
116 |
},
|
117 |
{
|
118 |
"epoch": 4.93,
|
119 |
+
"grad_norm": 14.269490242004395,
|
120 |
+
"learning_rate": 6.720478325859492e-06,
|
121 |
+
"loss": 0.2217,
|
122 |
"step": 1100
|
123 |
},
|
124 |
{
|
125 |
"epoch": 5.0,
|
126 |
+
"eval_accuracy": 0.9444444179534912,
|
127 |
+
"eval_loss": 0.23924072086811066,
|
128 |
+
"eval_runtime": 9.2044,
|
129 |
+
"eval_samples_per_second": 21.511,
|
130 |
+
"eval_steps_per_second": 5.432,
|
131 |
"step": 1115
|
132 |
},
|
133 |
{
|
134 |
"epoch": 5.38,
|
135 |
+
"grad_norm": 0.41719043254852295,
|
136 |
+
"learning_rate": 6.421524663677131e-06,
|
137 |
+
"loss": 0.2165,
|
138 |
"step": 1200
|
139 |
},
|
140 |
{
|
141 |
"epoch": 5.83,
|
142 |
+
"grad_norm": 1.484471321105957,
|
143 |
+
"learning_rate": 6.1225710014947695e-06,
|
144 |
+
"loss": 0.1831,
|
145 |
"step": 1300
|
146 |
},
|
147 |
{
|
148 |
"epoch": 6.0,
|
149 |
+
"eval_accuracy": 0.9292929172515869,
|
150 |
+
"eval_loss": 0.27538299560546875,
|
151 |
+
"eval_runtime": 9.1435,
|
152 |
+
"eval_samples_per_second": 21.655,
|
153 |
+
"eval_steps_per_second": 5.468,
|
154 |
"step": 1338
|
155 |
},
|
156 |
{
|
157 |
"epoch": 6.28,
|
158 |
+
"grad_norm": 0.09743738174438477,
|
159 |
+
"learning_rate": 5.823617339312408e-06,
|
160 |
+
"loss": 0.2059,
|
161 |
"step": 1400
|
162 |
},
|
163 |
{
|
164 |
"epoch": 6.73,
|
165 |
+
"grad_norm": 0.3065042793750763,
|
166 |
+
"learning_rate": 5.524663677130046e-06,
|
167 |
+
"loss": 0.1598,
|
168 |
"step": 1500
|
169 |
},
|
170 |
{
|
171 |
"epoch": 7.0,
|
172 |
+
"eval_accuracy": 0.9343434572219849,
|
173 |
+
"eval_loss": 0.3294394910335541,
|
174 |
+
"eval_runtime": 9.1064,
|
175 |
+
"eval_samples_per_second": 21.743,
|
176 |
+
"eval_steps_per_second": 5.491,
|
177 |
"step": 1561
|
178 |
},
|
179 |
{
|
180 |
"epoch": 7.17,
|
181 |
+
"grad_norm": 0.05342373996973038,
|
182 |
"learning_rate": 5.228699551569507e-06,
|
183 |
+
"loss": 0.1455,
|
184 |
"step": 1600
|
185 |
},
|
186 |
{
|
187 |
"epoch": 7.62,
|
188 |
+
"grad_norm": 1.5460679531097412,
|
189 |
"learning_rate": 4.929745889387145e-06,
|
190 |
+
"loss": 0.1676,
|
191 |
"step": 1700
|
192 |
},
|
193 |
{
|
194 |
"epoch": 8.0,
|
195 |
+
"eval_accuracy": 0.9494949579238892,
|
196 |
+
"eval_loss": 0.2668905258178711,
|
197 |
+
"eval_runtime": 9.2118,
|
198 |
+
"eval_samples_per_second": 21.494,
|
199 |
+
"eval_steps_per_second": 5.428,
|
200 |
"step": 1784
|
201 |
},
|
202 |
{
|
203 |
"epoch": 8.07,
|
204 |
+
"grad_norm": 17.537992477416992,
|
205 |
"learning_rate": 4.630792227204783e-06,
|
206 |
+
"loss": 0.1762,
|
207 |
"step": 1800
|
208 |
},
|
209 |
{
|
210 |
"epoch": 8.52,
|
211 |
+
"grad_norm": 0.20349286496639252,
|
212 |
"learning_rate": 4.3318385650224224e-06,
|
213 |
+
"loss": 0.1566,
|
214 |
"step": 1900
|
215 |
},
|
216 |
{
|
217 |
"epoch": 8.97,
|
218 |
+
"grad_norm": 15.300110816955566,
|
219 |
"learning_rate": 4.03288490284006e-06,
|
220 |
+
"loss": 0.1597,
|
221 |
"step": 2000
|
222 |
},
|
223 |
{
|
224 |
"epoch": 9.0,
|
225 |
+
"eval_accuracy": 0.9292929172515869,
|
226 |
+
"eval_loss": 0.34383586049079895,
|
227 |
+
"eval_runtime": 9.179,
|
228 |
+
"eval_samples_per_second": 21.571,
|
229 |
+
"eval_steps_per_second": 5.447,
|
230 |
"step": 2007
|
231 |
},
|
232 |
{
|
233 |
"epoch": 9.42,
|
234 |
+
"grad_norm": 0.4512959420681,
|
235 |
"learning_rate": 3.7339312406576984e-06,
|
236 |
+
"loss": 0.1416,
|
237 |
"step": 2100
|
238 |
},
|
239 |
{
|
240 |
"epoch": 9.87,
|
241 |
+
"grad_norm": 0.7455862760543823,
|
242 |
"learning_rate": 3.4349775784753366e-06,
|
243 |
+
"loss": 0.1132,
|
244 |
"step": 2200
|
245 |
},
|
246 |
{
|
247 |
"epoch": 10.0,
|
248 |
+
"eval_accuracy": 0.9444444179534912,
|
249 |
+
"eval_loss": 0.31586208939552307,
|
250 |
+
"eval_runtime": 9.1631,
|
251 |
+
"eval_samples_per_second": 21.608,
|
252 |
+
"eval_steps_per_second": 5.457,
|
253 |
"step": 2230
|
254 |
},
|
255 |
{
|
256 |
"epoch": 10.31,
|
257 |
+
"grad_norm": 0.25966259837150574,
|
258 |
"learning_rate": 3.136023916292975e-06,
|
259 |
+
"loss": 0.1654,
|
260 |
"step": 2300
|
261 |
},
|
262 |
{
|
263 |
"epoch": 10.76,
|
264 |
+
"grad_norm": 0.45347365736961365,
|
265 |
"learning_rate": 2.8370702541106134e-06,
|
266 |
+
"loss": 0.1224,
|
267 |
"step": 2400
|
268 |
},
|
269 |
{
|
270 |
"epoch": 11.0,
|
271 |
+
"eval_accuracy": 0.9545454382896423,
|
272 |
+
"eval_loss": 0.29796990752220154,
|
273 |
+
"eval_runtime": 9.1354,
|
274 |
+
"eval_samples_per_second": 21.674,
|
275 |
+
"eval_steps_per_second": 5.473,
|
276 |
"step": 2453
|
277 |
},
|
278 |
{
|
279 |
+
"epoch": 11.21,
|
280 |
+
"grad_norm": 27.043094635009766,
|
281 |
+
"learning_rate": 2.538116591928251e-06,
|
282 |
+
"loss": 0.1021,
|
283 |
+
"step": 2500
|
|
|
|
|
284 |
},
|
285 |
{
|
286 |
+
"epoch": 11.66,
|
287 |
+
"grad_norm": 72.37726593017578,
|
288 |
+
"learning_rate": 2.2391629297458894e-06,
|
289 |
+
"loss": 0.095,
|
290 |
+
"step": 2600
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 12.0,
|
294 |
"eval_accuracy": 0.9444444179534912,
|
295 |
+
"eval_loss": 0.2970119118690491,
|
296 |
+
"eval_runtime": 9.1388,
|
297 |
+
"eval_samples_per_second": 21.666,
|
298 |
+
"eval_steps_per_second": 5.471,
|
299 |
+
"step": 2676
|
300 |
+
},
|
301 |
+
{
|
302 |
+
"epoch": 12.11,
|
303 |
+
"grad_norm": 0.6068007946014404,
|
304 |
+
"learning_rate": 1.940209267563528e-06,
|
305 |
+
"loss": 0.1307,
|
306 |
+
"step": 2700
|
307 |
+
},
|
308 |
+
{
|
309 |
+
"epoch": 12.56,
|
310 |
+
"grad_norm": 4.567564964294434,
|
311 |
+
"learning_rate": 1.641255605381166e-06,
|
312 |
+
"loss": 0.1087,
|
313 |
+
"step": 2800
|
314 |
+
},
|
315 |
+
{
|
316 |
+
"epoch": 13.0,
|
317 |
+
"eval_accuracy": 0.9343434572219849,
|
318 |
+
"eval_loss": 0.34486597776412964,
|
319 |
+
"eval_runtime": 9.3094,
|
320 |
+
"eval_samples_per_second": 21.269,
|
321 |
+
"eval_steps_per_second": 5.371,
|
322 |
+
"step": 2899
|
323 |
+
},
|
324 |
+
{
|
325 |
+
"epoch": 13.0,
|
326 |
+
"grad_norm": 41.62958908081055,
|
327 |
+
"learning_rate": 1.3423019431988044e-06,
|
328 |
+
"loss": 0.0917,
|
329 |
+
"step": 2900
|
330 |
+
},
|
331 |
+
{
|
332 |
+
"epoch": 13.45,
|
333 |
+
"grad_norm": 0.026164406910538673,
|
334 |
+
"learning_rate": 1.0433482810164425e-06,
|
335 |
+
"loss": 0.0904,
|
336 |
+
"step": 3000
|
337 |
+
},
|
338 |
+
{
|
339 |
+
"epoch": 13.9,
|
340 |
+
"grad_norm": 52.47389221191406,
|
341 |
+
"learning_rate": 7.443946188340807e-07,
|
342 |
+
"loss": 0.1254,
|
343 |
+
"step": 3100
|
344 |
+
},
|
345 |
+
{
|
346 |
+
"epoch": 14.0,
|
347 |
+
"eval_accuracy": 0.9444444179534912,
|
348 |
+
"eval_loss": 0.31978654861450195,
|
349 |
+
"eval_runtime": 9.2595,
|
350 |
+
"eval_samples_per_second": 21.384,
|
351 |
+
"eval_steps_per_second": 5.4,
|
352 |
+
"step": 3122
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 14.0,
|
356 |
+
"step": 3122,
|
357 |
+
"total_flos": 7.9842219974856e+17,
|
358 |
+
"train_loss": 0.30023268478028214,
|
359 |
+
"train_runtime": 1742.0026,
|
360 |
+
"train_samples_per_second": 15.336,
|
361 |
+
"train_steps_per_second": 1.92
|
362 |
+
},
|
363 |
+
{
|
364 |
+
"epoch": 14.0,
|
365 |
+
"eval_accuracy": 0.9545454382896423,
|
366 |
+
"eval_loss": 0.29796990752220154,
|
367 |
+
"eval_runtime": 8.9823,
|
368 |
+
"eval_samples_per_second": 22.043,
|
369 |
+
"eval_steps_per_second": 5.566,
|
370 |
+
"step": 3122
|
371 |
}
|
372 |
],
|
373 |
"logging_steps": 100,
|
|
|
375 |
"num_input_tokens_seen": 0,
|
376 |
"num_train_epochs": 15,
|
377 |
"save_steps": 500,
|
378 |
+
"total_flos": 7.9842219974856e+17,
|
379 |
"train_batch_size": 4,
|
380 |
"trial_name": null,
|
381 |
"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 4984
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d4f11d524efea23fe3851521cb1d052cadfa01563a005deccc39e6b7563628ad
|
3 |
size 4984
|