DewiBrynJones
commited on
Commit
•
b78f7a3
1
Parent(s):
d20d107
End of training
Browse files- README.md +3 -1
- all_results.json +11 -11
- eval_results.json +6 -6
- runs/May16_17-08-56_2e698a9c250f/events.out.tfevents.1715883146.2e698a9c250f.1150.1 +3 -0
- train_results.json +6 -6
- trainer_state.json +158 -149
README.md
CHANGED
@@ -2,6 +2,8 @@
|
|
2 |
license: apache-2.0
|
3 |
base_model: facebook/wav2vec2-large-xlsr-53
|
4 |
tags:
|
|
|
|
|
5 |
- generated_from_trainer
|
6 |
metrics:
|
7 |
- wer
|
@@ -15,7 +17,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
15 |
|
16 |
# wav2vec2-xlsr-53-ft-btb-cy
|
17 |
|
18 |
-
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on
|
19 |
It achieves the following results on the evaluation set:
|
20 |
- Loss: 0.4086
|
21 |
- Wer: 0.3122
|
|
|
2 |
license: apache-2.0
|
3 |
base_model: facebook/wav2vec2-large-xlsr-53
|
4 |
tags:
|
5 |
+
- automatic-speech-recognition
|
6 |
+
- DewiBrynJones/banc-trawsgrifiadau-bangor-normalized
|
7 |
- generated_from_trainer
|
8 |
metrics:
|
9 |
- wer
|
|
|
17 |
|
18 |
# wav2vec2-xlsr-53-ft-btb-cy
|
19 |
|
20 |
+
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the DEWIBRYNJONES/BANC-TRAWSGRIFIADAU-BANGOR-NORMALIZED - DEFAULT dataset.
|
21 |
It achieves the following results on the evaluation set:
|
22 |
- Loss: 0.4086
|
23 |
- Wer: 0.3122
|
all_results.json
CHANGED
@@ -1,15 +1,15 @@
|
|
1 |
{
|
2 |
-
"epoch": 3.
|
3 |
-
"eval_loss": 0.
|
4 |
-
"eval_runtime":
|
5 |
"eval_samples": 5656,
|
6 |
-
"eval_samples_per_second":
|
7 |
-
"eval_steps_per_second": 4.
|
8 |
-
"eval_wer": 0.
|
9 |
-
"total_flos": 9.
|
10 |
-
"train_loss": 1.
|
11 |
-
"train_runtime":
|
12 |
"train_samples": 22621,
|
13 |
-
"train_samples_per_second":
|
14 |
-
"train_steps_per_second": 0.
|
15 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 3.6775106082036775,
|
3 |
+
"eval_loss": 0.4086272120475769,
|
4 |
+
"eval_runtime": 163.0134,
|
5 |
"eval_samples": 5656,
|
6 |
+
"eval_samples_per_second": 34.697,
|
7 |
+
"eval_steps_per_second": 4.337,
|
8 |
+
"eval_wer": 0.3122344604235807,
|
9 |
+
"total_flos": 9.93185228450562e+18,
|
10 |
+
"train_loss": 1.1287707856985238,
|
11 |
+
"train_runtime": 7180.9983,
|
12 |
"train_samples": 22621,
|
13 |
+
"train_samples_per_second": 11.586,
|
14 |
+
"train_steps_per_second": 0.362
|
15 |
}
|
eval_results.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
-
"epoch": 3.
|
3 |
-
"eval_loss": 0.
|
4 |
-
"eval_runtime":
|
5 |
"eval_samples": 5656,
|
6 |
-
"eval_samples_per_second":
|
7 |
-
"eval_steps_per_second": 4.
|
8 |
-
"eval_wer": 0.
|
9 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 3.6775106082036775,
|
3 |
+
"eval_loss": 0.4086272120475769,
|
4 |
+
"eval_runtime": 163.0134,
|
5 |
"eval_samples": 5656,
|
6 |
+
"eval_samples_per_second": 34.697,
|
7 |
+
"eval_steps_per_second": 4.337,
|
8 |
+
"eval_wer": 0.3122344604235807
|
9 |
}
|
runs/May16_17-08-56_2e698a9c250f/events.out.tfevents.1715883146.2e698a9c250f.1150.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9043e4d9391056e8ffc9fc424a53bff804f24ec4a64e68e541d2b235b8cdccb8
|
3 |
+
size 406
|
train_results.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
-
"epoch": 3.
|
3 |
-
"total_flos": 9.
|
4 |
-
"train_loss": 1.
|
5 |
-
"train_runtime":
|
6 |
"train_samples": 22621,
|
7 |
-
"train_samples_per_second":
|
8 |
-
"train_steps_per_second": 0.
|
9 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 3.6775106082036775,
|
3 |
+
"total_flos": 9.93185228450562e+18,
|
4 |
+
"train_loss": 1.1287707856985238,
|
5 |
+
"train_runtime": 7180.9983,
|
6 |
"train_samples": 22621,
|
7 |
+
"train_samples_per_second": 11.586,
|
8 |
+
"train_steps_per_second": 0.362
|
9 |
}
|
trainer_state.json
CHANGED
@@ -1,289 +1,298 @@
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
-
"epoch": 3.
|
5 |
"eval_steps": 100,
|
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": 0.14144271570014144,
|
13 |
-
"eval_loss":
|
14 |
-
"eval_runtime":
|
15 |
-
"eval_samples_per_second":
|
16 |
-
"eval_steps_per_second": 4.
|
17 |
"eval_wer": 1.0,
|
18 |
"step": 100
|
19 |
},
|
20 |
{
|
21 |
"epoch": 0.2828854314002829,
|
22 |
-
"eval_loss": 3.
|
23 |
-
"eval_runtime":
|
24 |
-
"eval_samples_per_second":
|
25 |
-
"eval_steps_per_second": 4.
|
26 |
"eval_wer": 1.0,
|
27 |
"step": 200
|
28 |
},
|
29 |
{
|
30 |
"epoch": 0.4243281471004243,
|
31 |
-
"eval_loss":
|
32 |
-
"eval_runtime":
|
33 |
-
"eval_samples_per_second":
|
34 |
-
"eval_steps_per_second": 4.
|
35 |
-
"eval_wer":
|
36 |
"step": 300
|
37 |
},
|
38 |
{
|
39 |
"epoch": 0.5657708628005658,
|
40 |
-
"eval_loss": 1.
|
41 |
-
"eval_runtime":
|
42 |
-
"eval_samples_per_second":
|
43 |
-
"eval_steps_per_second": 4.
|
44 |
-
"eval_wer": 0.
|
45 |
"step": 400
|
46 |
},
|
47 |
{
|
48 |
"epoch": 0.7072135785007072,
|
49 |
-
"grad_norm":
|
50 |
-
"learning_rate": 0.
|
51 |
-
"loss": 3.
|
52 |
"step": 500
|
53 |
},
|
54 |
{
|
55 |
"epoch": 0.7072135785007072,
|
56 |
-
"eval_loss": 1.
|
57 |
-
"eval_runtime":
|
58 |
-
"eval_samples_per_second":
|
59 |
-
"eval_steps_per_second": 4.
|
60 |
-
"eval_wer": 0.
|
61 |
"step": 500
|
62 |
},
|
63 |
{
|
64 |
"epoch": 0.8486562942008486,
|
65 |
-
"eval_loss": 0.
|
66 |
-
"eval_runtime":
|
67 |
-
"eval_samples_per_second":
|
68 |
-
"eval_steps_per_second": 4.
|
69 |
-
"eval_wer": 0.
|
70 |
"step": 600
|
71 |
},
|
72 |
{
|
73 |
"epoch": 0.9900990099009901,
|
74 |
-
"eval_loss": 0.
|
75 |
-
"eval_runtime":
|
76 |
-
"eval_samples_per_second":
|
77 |
-
"eval_steps_per_second": 4.
|
78 |
-
"eval_wer": 0.
|
79 |
"step": 700
|
80 |
},
|
81 |
{
|
82 |
"epoch": 1.1315417256011315,
|
83 |
-
"eval_loss": 0.
|
84 |
-
"eval_runtime":
|
85 |
-
"eval_samples_per_second":
|
86 |
-
"eval_steps_per_second": 4.
|
87 |
-
"eval_wer": 0.
|
88 |
"step": 800
|
89 |
},
|
90 |
{
|
91 |
"epoch": 1.272984441301273,
|
92 |
-
"eval_loss": 0.
|
93 |
-
"eval_runtime":
|
94 |
-
"eval_samples_per_second":
|
95 |
-
"eval_steps_per_second": 4.
|
96 |
-
"eval_wer": 0.
|
97 |
"step": 900
|
98 |
},
|
99 |
{
|
100 |
"epoch": 1.4144271570014144,
|
101 |
-
"grad_norm":
|
102 |
-
"learning_rate": 0.
|
103 |
-
"loss": 0.
|
104 |
"step": 1000
|
105 |
},
|
106 |
{
|
107 |
"epoch": 1.4144271570014144,
|
108 |
-
"eval_loss": 0.
|
109 |
-
"eval_runtime":
|
110 |
-
"eval_samples_per_second":
|
111 |
-
"eval_steps_per_second": 4.
|
112 |
-
"eval_wer": 0.
|
113 |
"step": 1000
|
114 |
},
|
115 |
{
|
116 |
"epoch": 1.5558698727015559,
|
117 |
-
"eval_loss": 0.
|
118 |
-
"eval_runtime":
|
119 |
-
"eval_samples_per_second":
|
120 |
-
"eval_steps_per_second": 4.
|
121 |
-
"eval_wer": 0.
|
122 |
"step": 1100
|
123 |
},
|
124 |
{
|
125 |
"epoch": 1.6973125884016973,
|
126 |
-
"eval_loss": 0.
|
127 |
-
"eval_runtime":
|
128 |
-
"eval_samples_per_second":
|
129 |
-
"eval_steps_per_second": 4.
|
130 |
-
"eval_wer": 0.
|
131 |
"step": 1200
|
132 |
},
|
133 |
{
|
134 |
"epoch": 1.8387553041018387,
|
135 |
-
"eval_loss": 0.
|
136 |
-
"eval_runtime":
|
137 |
-
"eval_samples_per_second":
|
138 |
-
"eval_steps_per_second": 4.
|
139 |
-
"eval_wer": 0.
|
140 |
"step": 1300
|
141 |
},
|
142 |
{
|
143 |
"epoch": 1.9801980198019802,
|
144 |
-
"eval_loss": 0.
|
145 |
-
"eval_runtime":
|
146 |
-
"eval_samples_per_second":
|
147 |
-
"eval_steps_per_second": 4.
|
148 |
-
"eval_wer": 0.
|
149 |
"step": 1400
|
150 |
},
|
151 |
{
|
152 |
"epoch": 2.1216407355021216,
|
153 |
-
"grad_norm":
|
154 |
-
"learning_rate": 0.
|
155 |
-
"loss": 0.
|
156 |
"step": 1500
|
157 |
},
|
158 |
{
|
159 |
"epoch": 2.1216407355021216,
|
160 |
-
"eval_loss": 0.
|
161 |
-
"eval_runtime":
|
162 |
-
"eval_samples_per_second":
|
163 |
-
"eval_steps_per_second": 4.
|
164 |
-
"eval_wer": 0.
|
165 |
"step": 1500
|
166 |
},
|
167 |
{
|
168 |
"epoch": 2.263083451202263,
|
169 |
-
"eval_loss": 0.
|
170 |
-
"eval_runtime":
|
171 |
-
"eval_samples_per_second":
|
172 |
-
"eval_steps_per_second": 4.
|
173 |
-
"eval_wer": 0.
|
174 |
"step": 1600
|
175 |
},
|
176 |
{
|
177 |
"epoch": 2.4045261669024045,
|
178 |
-
"eval_loss": 0.
|
179 |
-
"eval_runtime":
|
180 |
-
"eval_samples_per_second":
|
181 |
-
"eval_steps_per_second": 4.
|
182 |
-
"eval_wer": 0.
|
183 |
"step": 1700
|
184 |
},
|
185 |
{
|
186 |
"epoch": 2.545968882602546,
|
187 |
-
"eval_loss": 0.
|
188 |
-
"eval_runtime":
|
189 |
-
"eval_samples_per_second":
|
190 |
-
"eval_steps_per_second": 4.
|
191 |
-
"eval_wer": 0.
|
192 |
"step": 1800
|
193 |
},
|
194 |
{
|
195 |
"epoch": 2.6874115983026874,
|
196 |
-
"eval_loss": 0.
|
197 |
-
"eval_runtime":
|
198 |
-
"eval_samples_per_second":
|
199 |
-
"eval_steps_per_second": 4.
|
200 |
-
"eval_wer": 0.
|
201 |
"step": 1900
|
202 |
},
|
203 |
{
|
204 |
"epoch": 2.828854314002829,
|
205 |
-
"grad_norm": 0.
|
206 |
-
"learning_rate":
|
207 |
-
"loss": 0.
|
208 |
"step": 2000
|
209 |
},
|
210 |
{
|
211 |
"epoch": 2.828854314002829,
|
212 |
-
"eval_loss": 0.
|
213 |
-
"eval_runtime":
|
214 |
-
"eval_samples_per_second":
|
215 |
-
"eval_steps_per_second": 4.
|
216 |
-
"eval_wer": 0.
|
217 |
"step": 2000
|
218 |
},
|
219 |
{
|
220 |
"epoch": 2.9702970297029703,
|
221 |
-
"eval_loss": 0.
|
222 |
-
"eval_runtime":
|
223 |
-
"eval_samples_per_second":
|
224 |
-
"eval_steps_per_second": 4.
|
225 |
-
"eval_wer": 0.
|
226 |
"step": 2100
|
227 |
},
|
228 |
{
|
229 |
"epoch": 3.1117397454031117,
|
230 |
-
"eval_loss": 0.
|
231 |
-
"eval_runtime":
|
232 |
-
"eval_samples_per_second":
|
233 |
-
"eval_steps_per_second": 4.
|
234 |
-
"eval_wer": 0.
|
235 |
"step": 2200
|
236 |
},
|
237 |
{
|
238 |
"epoch": 3.253182461103253,
|
239 |
-
"eval_loss": 0.
|
240 |
-
"eval_runtime":
|
241 |
-
"eval_samples_per_second":
|
242 |
-
"eval_steps_per_second": 4.
|
243 |
-
"eval_wer": 0.
|
244 |
"step": 2300
|
245 |
},
|
246 |
{
|
247 |
"epoch": 3.3946251768033946,
|
248 |
-
"eval_loss": 0.
|
249 |
-
"eval_runtime":
|
250 |
-
"eval_samples_per_second":
|
251 |
-
"eval_steps_per_second": 4.
|
252 |
-
"eval_wer": 0.
|
253 |
"step": 2400
|
254 |
},
|
255 |
{
|
256 |
"epoch": 3.536067892503536,
|
257 |
-
"grad_norm": 0.
|
258 |
-
"learning_rate":
|
259 |
-
"loss": 0.
|
260 |
"step": 2500
|
261 |
},
|
262 |
{
|
263 |
"epoch": 3.536067892503536,
|
264 |
-
"eval_loss": 0.
|
265 |
-
"eval_runtime":
|
266 |
-
"eval_samples_per_second":
|
267 |
-
"eval_steps_per_second": 4.
|
268 |
-
"eval_wer": 0.
|
269 |
"step": 2500
|
270 |
},
|
271 |
{
|
272 |
-
"epoch": 3.
|
273 |
-
"
|
274 |
-
"
|
275 |
-
"
|
276 |
-
"
|
277 |
-
"
|
278 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
279 |
}
|
280 |
],
|
281 |
"logging_steps": 500,
|
282 |
-
"max_steps":
|
283 |
"num_input_tokens_seen": 0,
|
284 |
"num_train_epochs": 4,
|
285 |
"save_steps": 400,
|
286 |
-
"total_flos": 9.
|
287 |
"train_batch_size": 16,
|
288 |
"trial_name": null,
|
289 |
"trial_params": null
|
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
+
"epoch": 3.6775106082036775,
|
5 |
"eval_steps": 100,
|
6 |
+
"global_step": 2600,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
"epoch": 0.14144271570014144,
|
13 |
+
"eval_loss": 3.626366138458252,
|
14 |
+
"eval_runtime": 164.728,
|
15 |
+
"eval_samples_per_second": 34.335,
|
16 |
+
"eval_steps_per_second": 4.292,
|
17 |
"eval_wer": 1.0,
|
18 |
"step": 100
|
19 |
},
|
20 |
{
|
21 |
"epoch": 0.2828854314002829,
|
22 |
+
"eval_loss": 3.1079769134521484,
|
23 |
+
"eval_runtime": 161.0986,
|
24 |
+
"eval_samples_per_second": 35.109,
|
25 |
+
"eval_steps_per_second": 4.389,
|
26 |
"eval_wer": 1.0,
|
27 |
"step": 200
|
28 |
},
|
29 |
{
|
30 |
"epoch": 0.4243281471004243,
|
31 |
+
"eval_loss": 2.5440056324005127,
|
32 |
+
"eval_runtime": 161.942,
|
33 |
+
"eval_samples_per_second": 34.926,
|
34 |
+
"eval_steps_per_second": 4.366,
|
35 |
+
"eval_wer": 0.9862761130617415,
|
36 |
"step": 300
|
37 |
},
|
38 |
{
|
39 |
"epoch": 0.5657708628005658,
|
40 |
+
"eval_loss": 1.2381267547607422,
|
41 |
+
"eval_runtime": 162.1273,
|
42 |
+
"eval_samples_per_second": 34.886,
|
43 |
+
"eval_steps_per_second": 4.361,
|
44 |
+
"eval_wer": 0.8110079681913649,
|
45 |
"step": 400
|
46 |
},
|
47 |
{
|
48 |
"epoch": 0.7072135785007072,
|
49 |
+
"grad_norm": 3.555830955505371,
|
50 |
+
"learning_rate": 0.00029699999999999996,
|
51 |
+
"loss": 3.4586,
|
52 |
"step": 500
|
53 |
},
|
54 |
{
|
55 |
"epoch": 0.7072135785007072,
|
56 |
+
"eval_loss": 1.0035638809204102,
|
57 |
+
"eval_runtime": 161.1941,
|
58 |
+
"eval_samples_per_second": 35.088,
|
59 |
+
"eval_steps_per_second": 4.386,
|
60 |
+
"eval_wer": 0.7184358616709153,
|
61 |
"step": 500
|
62 |
},
|
63 |
{
|
64 |
"epoch": 0.8486562942008486,
|
65 |
+
"eval_loss": 0.8524216413497925,
|
66 |
+
"eval_runtime": 162.0442,
|
67 |
+
"eval_samples_per_second": 34.904,
|
68 |
+
"eval_steps_per_second": 4.363,
|
69 |
+
"eval_wer": 0.6379042213778398,
|
70 |
"step": 600
|
71 |
},
|
72 |
{
|
73 |
"epoch": 0.9900990099009901,
|
74 |
+
"eval_loss": 0.7411457300186157,
|
75 |
+
"eval_runtime": 162.8526,
|
76 |
+
"eval_samples_per_second": 34.731,
|
77 |
+
"eval_steps_per_second": 4.341,
|
78 |
+
"eval_wer": 0.5673288121462813,
|
79 |
"step": 700
|
80 |
},
|
81 |
{
|
82 |
"epoch": 1.1315417256011315,
|
83 |
+
"eval_loss": 0.6202757954597473,
|
84 |
+
"eval_runtime": 162.5052,
|
85 |
+
"eval_samples_per_second": 34.805,
|
86 |
+
"eval_steps_per_second": 4.351,
|
87 |
+
"eval_wer": 0.48383114488640916,
|
88 |
"step": 800
|
89 |
},
|
90 |
{
|
91 |
"epoch": 1.272984441301273,
|
92 |
+
"eval_loss": 0.6016646027565002,
|
93 |
+
"eval_runtime": 162.6225,
|
94 |
+
"eval_samples_per_second": 34.78,
|
95 |
+
"eval_steps_per_second": 4.347,
|
96 |
+
"eval_wer": 0.48198739839353566,
|
97 |
"step": 900
|
98 |
},
|
99 |
{
|
100 |
"epoch": 1.4144271570014144,
|
101 |
+
"grad_norm": 1.029418706893921,
|
102 |
+
"learning_rate": 0.00022928571428571426,
|
103 |
+
"loss": 0.8146,
|
104 |
"step": 1000
|
105 |
},
|
106 |
{
|
107 |
"epoch": 1.4144271570014144,
|
108 |
+
"eval_loss": 0.5637466311454773,
|
109 |
+
"eval_runtime": 163.5441,
|
110 |
+
"eval_samples_per_second": 34.584,
|
111 |
+
"eval_steps_per_second": 4.323,
|
112 |
+
"eval_wer": 0.4461706186971927,
|
113 |
"step": 1000
|
114 |
},
|
115 |
{
|
116 |
"epoch": 1.5558698727015559,
|
117 |
+
"eval_loss": 0.5380699038505554,
|
118 |
+
"eval_runtime": 163.2757,
|
119 |
+
"eval_samples_per_second": 34.641,
|
120 |
+
"eval_steps_per_second": 4.33,
|
121 |
+
"eval_wer": 0.4207589822519359,
|
122 |
"step": 1100
|
123 |
},
|
124 |
{
|
125 |
"epoch": 1.6973125884016973,
|
126 |
+
"eval_loss": 0.5155439972877502,
|
127 |
+
"eval_runtime": 163.2795,
|
128 |
+
"eval_samples_per_second": 34.64,
|
129 |
+
"eval_steps_per_second": 4.33,
|
130 |
+
"eval_wer": 0.4043416221762622,
|
131 |
"step": 1200
|
132 |
},
|
133 |
{
|
134 |
"epoch": 1.8387553041018387,
|
135 |
+
"eval_loss": 0.4857912063598633,
|
136 |
+
"eval_runtime": 163.399,
|
137 |
+
"eval_samples_per_second": 34.615,
|
138 |
+
"eval_steps_per_second": 4.327,
|
139 |
+
"eval_wer": 0.39034518140862234,
|
140 |
"step": 1300
|
141 |
},
|
142 |
{
|
143 |
"epoch": 1.9801980198019802,
|
144 |
+
"eval_loss": 0.47579917311668396,
|
145 |
+
"eval_runtime": 162.8853,
|
146 |
+
"eval_samples_per_second": 34.724,
|
147 |
+
"eval_steps_per_second": 4.34,
|
148 |
+
"eval_wer": 0.38234492488737115,
|
149 |
"step": 1400
|
150 |
},
|
151 |
{
|
152 |
"epoch": 2.1216407355021216,
|
153 |
+
"grad_norm": 1.043891429901123,
|
154 |
+
"learning_rate": 0.0001582857142857143,
|
155 |
+
"loss": 0.6294,
|
156 |
"step": 1500
|
157 |
},
|
158 |
{
|
159 |
"epoch": 2.1216407355021216,
|
160 |
+
"eval_loss": 0.46035128831863403,
|
161 |
+
"eval_runtime": 163.2756,
|
162 |
+
"eval_samples_per_second": 34.641,
|
163 |
+
"eval_steps_per_second": 4.33,
|
164 |
+
"eval_wer": 0.36421207894441504,
|
165 |
"step": 1500
|
166 |
},
|
167 |
{
|
168 |
"epoch": 2.263083451202263,
|
169 |
+
"eval_loss": 0.45305728912353516,
|
170 |
+
"eval_runtime": 164.2898,
|
171 |
+
"eval_samples_per_second": 34.427,
|
172 |
+
"eval_steps_per_second": 4.303,
|
173 |
+
"eval_wer": 0.35355041444214647,
|
174 |
"step": 1600
|
175 |
},
|
176 |
{
|
177 |
"epoch": 2.4045261669024045,
|
178 |
+
"eval_loss": 0.44349414110183716,
|
179 |
+
"eval_runtime": 164.2629,
|
180 |
+
"eval_samples_per_second": 34.433,
|
181 |
+
"eval_steps_per_second": 4.304,
|
182 |
+
"eval_wer": 0.3510813973995158,
|
183 |
"step": 1700
|
184 |
},
|
185 |
{
|
186 |
"epoch": 2.545968882602546,
|
187 |
+
"eval_loss": 0.4366327226161957,
|
188 |
+
"eval_runtime": 164.0684,
|
189 |
+
"eval_samples_per_second": 34.473,
|
190 |
+
"eval_steps_per_second": 4.309,
|
191 |
+
"eval_wer": 0.3498949866127972,
|
192 |
"step": 1800
|
193 |
},
|
194 |
{
|
195 |
"epoch": 2.6874115983026874,
|
196 |
+
"eval_loss": 0.43093863129615784,
|
197 |
+
"eval_runtime": 164.1859,
|
198 |
+
"eval_samples_per_second": 34.449,
|
199 |
+
"eval_steps_per_second": 4.306,
|
200 |
+
"eval_wer": 0.34513331088772387,
|
201 |
"step": 1900
|
202 |
},
|
203 |
{
|
204 |
"epoch": 2.828854314002829,
|
205 |
+
"grad_norm": 0.763507604598999,
|
206 |
+
"learning_rate": 8.728571428571428e-05,
|
207 |
+
"loss": 0.4914,
|
208 |
"step": 2000
|
209 |
},
|
210 |
{
|
211 |
"epoch": 2.828854314002829,
|
212 |
+
"eval_loss": 0.4252130389213562,
|
213 |
+
"eval_runtime": 164.7353,
|
214 |
+
"eval_samples_per_second": 34.334,
|
215 |
+
"eval_steps_per_second": 4.292,
|
216 |
+
"eval_wer": 0.3367803376460969,
|
217 |
"step": 2000
|
218 |
},
|
219 |
{
|
220 |
"epoch": 2.9702970297029703,
|
221 |
+
"eval_loss": 0.42176735401153564,
|
222 |
+
"eval_runtime": 163.889,
|
223 |
+
"eval_samples_per_second": 34.511,
|
224 |
+
"eval_steps_per_second": 4.314,
|
225 |
+
"eval_wer": 0.3285075272954644,
|
226 |
"step": 2100
|
227 |
},
|
228 |
{
|
229 |
"epoch": 3.1117397454031117,
|
230 |
+
"eval_loss": 0.42082250118255615,
|
231 |
+
"eval_runtime": 165.1043,
|
232 |
+
"eval_samples_per_second": 34.257,
|
233 |
+
"eval_steps_per_second": 4.282,
|
234 |
+
"eval_wer": 0.32510862071729757,
|
235 |
"step": 2200
|
236 |
},
|
237 |
{
|
238 |
"epoch": 3.253182461103253,
|
239 |
+
"eval_loss": 0.41444018483161926,
|
240 |
+
"eval_runtime": 164.0969,
|
241 |
+
"eval_samples_per_second": 34.467,
|
242 |
+
"eval_steps_per_second": 4.308,
|
243 |
+
"eval_wer": 0.32363362352299874,
|
244 |
"step": 2300
|
245 |
},
|
246 |
{
|
247 |
"epoch": 3.3946251768033946,
|
248 |
+
"eval_loss": 0.41402822732925415,
|
249 |
+
"eval_runtime": 164.4996,
|
250 |
+
"eval_samples_per_second": 34.383,
|
251 |
+
"eval_steps_per_second": 4.298,
|
252 |
+
"eval_wer": 0.31654722395908486,
|
253 |
"step": 2400
|
254 |
},
|
255 |
{
|
256 |
"epoch": 3.536067892503536,
|
257 |
+
"grad_norm": 0.8644358515739441,
|
258 |
+
"learning_rate": 1.614285714285714e-05,
|
259 |
+
"loss": 0.4011,
|
260 |
"step": 2500
|
261 |
},
|
262 |
{
|
263 |
"epoch": 3.536067892503536,
|
264 |
+
"eval_loss": 0.4133088290691376,
|
265 |
+
"eval_runtime": 164.6412,
|
266 |
+
"eval_samples_per_second": 34.353,
|
267 |
+
"eval_steps_per_second": 4.294,
|
268 |
+
"eval_wer": 0.31574559504913985,
|
269 |
"step": 2500
|
270 |
},
|
271 |
{
|
272 |
+
"epoch": 3.6775106082036775,
|
273 |
+
"eval_loss": 0.4086272120475769,
|
274 |
+
"eval_runtime": 163.5841,
|
275 |
+
"eval_samples_per_second": 34.575,
|
276 |
+
"eval_steps_per_second": 4.322,
|
277 |
+
"eval_wer": 0.3122344604235807,
|
278 |
+
"step": 2600
|
279 |
+
},
|
280 |
+
{
|
281 |
+
"epoch": 3.6775106082036775,
|
282 |
+
"step": 2600,
|
283 |
+
"total_flos": 9.93185228450562e+18,
|
284 |
+
"train_loss": 1.1287707856985238,
|
285 |
+
"train_runtime": 7180.9983,
|
286 |
+
"train_samples_per_second": 11.586,
|
287 |
+
"train_steps_per_second": 0.362
|
288 |
}
|
289 |
],
|
290 |
"logging_steps": 500,
|
291 |
+
"max_steps": 2600,
|
292 |
"num_input_tokens_seen": 0,
|
293 |
"num_train_epochs": 4,
|
294 |
"save_steps": 400,
|
295 |
+
"total_flos": 9.93185228450562e+18,
|
296 |
"train_batch_size": 16,
|
297 |
"trial_name": null,
|
298 |
"trial_params": null
|