DewiBrynJones
commited on
Commit
•
d59d9a7
1
Parent(s):
21a680a
End of training
Browse files- README.md +3 -1
- all_results.json +9 -9
- eval_results.json +5 -5
- runs/Jun27_19-17-27_1a548bea00a4/events.out.tfevents.1719531034.1a548bea00a4.30.1 +3 -0
- train_results.json +4 -4
- trainer_state.json +560 -560
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-ccv-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: 0.4122
|
21 |
- Wer: 0.3223
|
|
|
2 |
license: apache-2.0
|
3 |
base_model: facebook/wav2vec2-large-xlsr-53
|
4 |
tags:
|
5 |
+
- automatic-speech-recognition
|
6 |
+
- DewiBrynJones/banc-trawsgrifiadau-bangor-clean-with-ccv
|
7 |
- generated_from_trainer
|
8 |
metrics:
|
9 |
- wer
|
|
|
17 |
|
18 |
# wav2vec2-xlsr-53-ft-btb-ccv-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-CLEAN-WITH-CCV - DEFAULT dataset.
|
21 |
It achieves the following results on the evaluation set:
|
22 |
- Loss: 0.4122
|
23 |
- Wer: 0.3223
|
all_results.json
CHANGED
@@ -1,15 +1,15 @@
|
|
1 |
{
|
2 |
"epoch": 1.9357336430507162,
|
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": 1.1255918428180738e+19,
|
10 |
-
"train_loss": 0.
|
11 |
-
"train_runtime":
|
12 |
"train_samples": 41326,
|
13 |
-
"train_samples_per_second": 4.
|
14 |
-
"train_steps_per_second": 0.
|
15 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.9357336430507162,
|
3 |
+
"eval_loss": 0.4121682941913605,
|
4 |
+
"eval_runtime": 148.2246,
|
5 |
"eval_samples": 5656,
|
6 |
+
"eval_samples_per_second": 38.158,
|
7 |
+
"eval_steps_per_second": 4.77,
|
8 |
+
"eval_wer": 0.3222865946622587,
|
9 |
"total_flos": 1.1255918428180738e+19,
|
10 |
+
"train_loss": 0.7340839981079101,
|
11 |
+
"train_runtime": 18566.2158,
|
12 |
"train_samples": 41326,
|
13 |
+
"train_samples_per_second": 4.309,
|
14 |
+
"train_steps_per_second": 0.539
|
15 |
}
|
eval_results.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
"epoch": 1.9357336430507162,
|
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": 1.9357336430507162,
|
3 |
+
"eval_loss": 0.4121682941913605,
|
4 |
+
"eval_runtime": 148.2246,
|
5 |
"eval_samples": 5656,
|
6 |
+
"eval_samples_per_second": 38.158,
|
7 |
+
"eval_steps_per_second": 4.77,
|
8 |
+
"eval_wer": 0.3222865946622587
|
9 |
}
|
runs/Jun27_19-17-27_1a548bea00a4/events.out.tfevents.1719531034.1a548bea00a4.30.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f8e98e1b76ea283ce2766410db9e3a516095ddf9df759a51ef455d31f605235b
|
3 |
+
size 406
|
train_results.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
"epoch": 1.9357336430507162,
|
3 |
"total_flos": 1.1255918428180738e+19,
|
4 |
-
"train_loss": 0.
|
5 |
-
"train_runtime":
|
6 |
"train_samples": 41326,
|
7 |
-
"train_samples_per_second": 4.
|
8 |
-
"train_steps_per_second": 0.
|
9 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.9357336430507162,
|
3 |
"total_flos": 1.1255918428180738e+19,
|
4 |
+
"train_loss": 0.7340839981079101,
|
5 |
+
"train_runtime": 18566.2158,
|
6 |
"train_samples": 41326,
|
7 |
+
"train_samples_per_second": 4.309,
|
8 |
+
"train_steps_per_second": 0.539
|
9 |
}
|
trainer_state.json
CHANGED
@@ -10,1052 +10,1052 @@
|
|
10 |
"log_history": [
|
11 |
{
|
12 |
"epoch": 0.019357336430507164,
|
13 |
-
"eval_loss": 3.
|
14 |
-
"eval_runtime":
|
15 |
-
"eval_samples_per_second": 38.
|
16 |
-
"eval_steps_per_second": 4.
|
17 |
"eval_wer": 1.0,
|
18 |
"step": 100
|
19 |
},
|
20 |
{
|
21 |
"epoch": 0.03871467286101433,
|
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": 0.05807200929152149,
|
31 |
-
"eval_loss":
|
32 |
-
"eval_runtime":
|
33 |
-
"eval_samples_per_second":
|
34 |
-
"eval_steps_per_second": 4.
|
35 |
"eval_wer": 1.0,
|
36 |
"step": 300
|
37 |
},
|
38 |
{
|
39 |
"epoch": 0.07742934572202866,
|
40 |
-
"eval_loss": 2.
|
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": 0.09678668215253582,
|
49 |
-
"grad_norm":
|
50 |
-
"learning_rate": 0.
|
51 |
-
"loss": 4.
|
52 |
"step": 500
|
53 |
},
|
54 |
{
|
55 |
"epoch": 0.09678668215253582,
|
56 |
-
"eval_loss": 1.
|
57 |
-
"eval_runtime": 145.
|
58 |
-
"eval_samples_per_second": 38.
|
59 |
-
"eval_steps_per_second": 4.
|
60 |
-
"eval_wer": 0.
|
61 |
"step": 500
|
62 |
},
|
63 |
{
|
64 |
"epoch": 0.11614401858304298,
|
65 |
-
"eval_loss": 1.
|
66 |
-
"eval_runtime":
|
67 |
-
"eval_samples_per_second": 38.
|
68 |
-
"eval_steps_per_second": 4.
|
69 |
-
"eval_wer": 0.
|
70 |
"step": 600
|
71 |
},
|
72 |
{
|
73 |
"epoch": 0.13550135501355012,
|
74 |
-
"eval_loss": 1.
|
75 |
-
"eval_runtime":
|
76 |
-
"eval_samples_per_second": 38.
|
77 |
-
"eval_steps_per_second": 4.
|
78 |
-
"eval_wer": 0.
|
79 |
"step": 700
|
80 |
},
|
81 |
{
|
82 |
"epoch": 0.1548586914440573,
|
83 |
-
"eval_loss": 1.
|
84 |
-
"eval_runtime": 146.
|
85 |
-
"eval_samples_per_second": 38.
|
86 |
-
"eval_steps_per_second": 4.
|
87 |
-
"eval_wer": 0.
|
88 |
"step": 800
|
89 |
},
|
90 |
{
|
91 |
"epoch": 0.17421602787456447,
|
92 |
-
"eval_loss": 0.
|
93 |
-
"eval_runtime":
|
94 |
-
"eval_samples_per_second": 38.
|
95 |
-
"eval_steps_per_second": 4.
|
96 |
-
"eval_wer": 0.
|
97 |
"step": 900
|
98 |
},
|
99 |
{
|
100 |
"epoch": 0.19357336430507163,
|
101 |
-
"grad_norm":
|
102 |
-
"learning_rate": 0.
|
103 |
-
"loss": 1.
|
104 |
"step": 1000
|
105 |
},
|
106 |
{
|
107 |
"epoch": 0.19357336430507163,
|
108 |
-
"eval_loss": 0.
|
109 |
-
"eval_runtime": 146.
|
110 |
-
"eval_samples_per_second": 38.
|
111 |
-
"eval_steps_per_second": 4.
|
112 |
-
"eval_wer": 0.
|
113 |
"step": 1000
|
114 |
},
|
115 |
{
|
116 |
"epoch": 0.2129307007355788,
|
117 |
-
"eval_loss": 0.
|
118 |
-
"eval_runtime": 146.
|
119 |
-
"eval_samples_per_second": 38.
|
120 |
"eval_steps_per_second": 4.818,
|
121 |
-
"eval_wer": 0.
|
122 |
"step": 1100
|
123 |
},
|
124 |
{
|
125 |
"epoch": 0.23228803716608595,
|
126 |
-
"eval_loss": 0.
|
127 |
-
"eval_runtime":
|
128 |
-
"eval_samples_per_second": 38.
|
129 |
-
"eval_steps_per_second": 4.
|
130 |
-
"eval_wer": 0.
|
131 |
"step": 1200
|
132 |
},
|
133 |
{
|
134 |
"epoch": 0.2516453735965931,
|
135 |
-
"eval_loss": 0.
|
136 |
-
"eval_runtime":
|
137 |
-
"eval_samples_per_second": 38.
|
138 |
-
"eval_steps_per_second": 4.
|
139 |
-
"eval_wer": 0.
|
140 |
"step": 1300
|
141 |
},
|
142 |
{
|
143 |
"epoch": 0.27100271002710025,
|
144 |
-
"eval_loss": 0.
|
145 |
-
"eval_runtime":
|
146 |
-
"eval_samples_per_second": 38.
|
147 |
-
"eval_steps_per_second": 4.
|
148 |
-
"eval_wer": 0.
|
149 |
"step": 1400
|
150 |
},
|
151 |
{
|
152 |
"epoch": 0.29036004645760743,
|
153 |
-
"grad_norm":
|
154 |
-
"learning_rate": 0.
|
155 |
-
"loss": 0.
|
156 |
"step": 1500
|
157 |
},
|
158 |
{
|
159 |
"epoch": 0.29036004645760743,
|
160 |
-
"eval_loss": 0.
|
161 |
-
"eval_runtime": 151.
|
162 |
-
"eval_samples_per_second": 37.
|
163 |
-
"eval_steps_per_second": 4.
|
164 |
-
"eval_wer": 0.
|
165 |
"step": 1500
|
166 |
},
|
167 |
{
|
168 |
"epoch": 0.3097173828881146,
|
169 |
-
"eval_loss": 0.
|
170 |
-
"eval_runtime": 147.
|
171 |
-
"eval_samples_per_second": 38.
|
172 |
-
"eval_steps_per_second": 4.
|
173 |
-
"eval_wer": 0.
|
174 |
"step": 1600
|
175 |
},
|
176 |
{
|
177 |
"epoch": 0.32907471931862176,
|
178 |
-
"eval_loss": 0.
|
179 |
-
"eval_runtime":
|
180 |
-
"eval_samples_per_second": 38.
|
181 |
-
"eval_steps_per_second": 4.
|
182 |
-
"eval_wer": 0.
|
183 |
"step": 1700
|
184 |
},
|
185 |
{
|
186 |
"epoch": 0.34843205574912894,
|
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": 0.3677893921796361,
|
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": 0.38714672861014326,
|
205 |
-
"grad_norm":
|
206 |
-
"learning_rate": 0.
|
207 |
-
"loss": 0.
|
208 |
"step": 2000
|
209 |
},
|
210 |
{
|
211 |
"epoch": 0.38714672861014326,
|
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": 0.4065040650406504,
|
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": 0.4258614014711576,
|
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": 0.4452187379016647,
|
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": 0.4645760743321719,
|
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": 0.48393341076267904,
|
257 |
-
"grad_norm":
|
258 |
-
"learning_rate": 0.
|
259 |
-
"loss": 0.
|
260 |
"step": 2500
|
261 |
},
|
262 |
{
|
263 |
"epoch": 0.48393341076267904,
|
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": 0.5032907471931862,
|
273 |
-
"eval_loss": 0.
|
274 |
-
"eval_runtime":
|
275 |
-
"eval_samples_per_second": 38.
|
276 |
-
"eval_steps_per_second": 4.
|
277 |
-
"eval_wer": 0.
|
278 |
"step": 2600
|
279 |
},
|
280 |
{
|
281 |
"epoch": 0.5226480836236934,
|
282 |
-
"eval_loss": 0.
|
283 |
-
"eval_runtime":
|
284 |
-
"eval_samples_per_second": 38.
|
285 |
-
"eval_steps_per_second": 4.
|
286 |
-
"eval_wer": 0.
|
287 |
"step": 2700
|
288 |
},
|
289 |
{
|
290 |
"epoch": 0.5420054200542005,
|
291 |
-
"eval_loss": 0.
|
292 |
-
"eval_runtime":
|
293 |
-
"eval_samples_per_second": 38.
|
294 |
-
"eval_steps_per_second": 4.
|
295 |
-
"eval_wer": 0.
|
296 |
"step": 2800
|
297 |
},
|
298 |
{
|
299 |
"epoch": 0.5613627564847077,
|
300 |
-
"eval_loss": 0.
|
301 |
-
"eval_runtime":
|
302 |
-
"eval_samples_per_second":
|
303 |
-
"eval_steps_per_second": 4.
|
304 |
-
"eval_wer": 0.
|
305 |
"step": 2900
|
306 |
},
|
307 |
{
|
308 |
"epoch": 0.5807200929152149,
|
309 |
-
"grad_norm":
|
310 |
-
"learning_rate": 0.
|
311 |
-
"loss": 0.
|
312 |
"step": 3000
|
313 |
},
|
314 |
{
|
315 |
"epoch": 0.5807200929152149,
|
316 |
-
"eval_loss": 0.
|
317 |
-
"eval_runtime":
|
318 |
-
"eval_samples_per_second":
|
319 |
-
"eval_steps_per_second": 4.
|
320 |
-
"eval_wer": 0.
|
321 |
"step": 3000
|
322 |
},
|
323 |
{
|
324 |
"epoch": 0.6000774293457221,
|
325 |
-
"eval_loss": 0.
|
326 |
-
"eval_runtime":
|
327 |
-
"eval_samples_per_second":
|
328 |
-
"eval_steps_per_second": 4.
|
329 |
-
"eval_wer": 0.
|
330 |
"step": 3100
|
331 |
},
|
332 |
{
|
333 |
"epoch": 0.6194347657762292,
|
334 |
-
"eval_loss": 0.
|
335 |
-
"eval_runtime":
|
336 |
-
"eval_samples_per_second": 38.
|
337 |
-
"eval_steps_per_second": 4.
|
338 |
-
"eval_wer": 0.
|
339 |
"step": 3200
|
340 |
},
|
341 |
{
|
342 |
"epoch": 0.6387921022067363,
|
343 |
-
"eval_loss": 0.
|
344 |
-
"eval_runtime":
|
345 |
-
"eval_samples_per_second":
|
346 |
-
"eval_steps_per_second": 4.
|
347 |
-
"eval_wer": 0.
|
348 |
"step": 3300
|
349 |
},
|
350 |
{
|
351 |
"epoch": 0.6581494386372435,
|
352 |
-
"eval_loss": 0.
|
353 |
-
"eval_runtime":
|
354 |
-
"eval_samples_per_second": 38.
|
355 |
-
"eval_steps_per_second": 4.
|
356 |
-
"eval_wer": 0.
|
357 |
"step": 3400
|
358 |
},
|
359 |
{
|
360 |
"epoch": 0.6775067750677507,
|
361 |
-
"grad_norm":
|
362 |
-
"learning_rate": 0.
|
363 |
-
"loss": 0.
|
364 |
"step": 3500
|
365 |
},
|
366 |
{
|
367 |
"epoch": 0.6775067750677507,
|
368 |
-
"eval_loss": 0.
|
369 |
-
"eval_runtime":
|
370 |
-
"eval_samples_per_second": 38.
|
371 |
-
"eval_steps_per_second": 4.
|
372 |
-
"eval_wer": 0.
|
373 |
"step": 3500
|
374 |
},
|
375 |
{
|
376 |
"epoch": 0.6968641114982579,
|
377 |
-
"eval_loss": 0.
|
378 |
-
"eval_runtime":
|
379 |
-
"eval_samples_per_second": 38.
|
380 |
-
"eval_steps_per_second": 4.
|
381 |
-
"eval_wer": 0.
|
382 |
"step": 3600
|
383 |
},
|
384 |
{
|
385 |
"epoch": 0.716221447928765,
|
386 |
-
"eval_loss": 0.
|
387 |
-
"eval_runtime":
|
388 |
-
"eval_samples_per_second":
|
389 |
-
"eval_steps_per_second": 4.
|
390 |
-
"eval_wer": 0.
|
391 |
"step": 3700
|
392 |
},
|
393 |
{
|
394 |
"epoch": 0.7355787843592722,
|
395 |
-
"eval_loss": 0.
|
396 |
-
"eval_runtime":
|
397 |
-
"eval_samples_per_second": 38.
|
398 |
-
"eval_steps_per_second": 4.
|
399 |
-
"eval_wer": 0.
|
400 |
"step": 3800
|
401 |
},
|
402 |
{
|
403 |
"epoch": 0.7549361207897793,
|
404 |
-
"eval_loss": 0.
|
405 |
-
"eval_runtime":
|
406 |
-
"eval_samples_per_second": 38.
|
407 |
-
"eval_steps_per_second": 4.
|
408 |
-
"eval_wer": 0.
|
409 |
"step": 3900
|
410 |
},
|
411 |
{
|
412 |
"epoch": 0.7742934572202865,
|
413 |
-
"grad_norm":
|
414 |
-
"learning_rate": 0.
|
415 |
-
"loss": 0.
|
416 |
"step": 4000
|
417 |
},
|
418 |
{
|
419 |
"epoch": 0.7742934572202865,
|
420 |
-
"eval_loss": 0.
|
421 |
-
"eval_runtime":
|
422 |
-
"eval_samples_per_second": 38.
|
423 |
-
"eval_steps_per_second": 4.
|
424 |
-
"eval_wer": 0.
|
425 |
"step": 4000
|
426 |
},
|
427 |
{
|
428 |
"epoch": 0.7936507936507936,
|
429 |
-
"eval_loss": 0.
|
430 |
-
"eval_runtime":
|
431 |
-
"eval_samples_per_second":
|
432 |
-
"eval_steps_per_second": 4.
|
433 |
-
"eval_wer": 0.
|
434 |
"step": 4100
|
435 |
},
|
436 |
{
|
437 |
"epoch": 0.8130081300813008,
|
438 |
-
"eval_loss": 0.
|
439 |
-
"eval_runtime":
|
440 |
-
"eval_samples_per_second": 38.
|
441 |
-
"eval_steps_per_second": 4.
|
442 |
-
"eval_wer": 0.
|
443 |
"step": 4200
|
444 |
},
|
445 |
{
|
446 |
"epoch": 0.832365466511808,
|
447 |
-
"eval_loss": 0.
|
448 |
-
"eval_runtime":
|
449 |
-
"eval_samples_per_second":
|
450 |
-
"eval_steps_per_second": 4.
|
451 |
-
"eval_wer": 0.
|
452 |
"step": 4300
|
453 |
},
|
454 |
{
|
455 |
"epoch": 0.8517228029423152,
|
456 |
-
"eval_loss": 0.
|
457 |
-
"eval_runtime":
|
458 |
-
"eval_samples_per_second":
|
459 |
-
"eval_steps_per_second": 4.
|
460 |
-
"eval_wer": 0.
|
461 |
"step": 4400
|
462 |
},
|
463 |
{
|
464 |
"epoch": 0.8710801393728222,
|
465 |
-
"grad_norm": 3.
|
466 |
-
"learning_rate": 0.
|
467 |
-
"loss": 0.
|
468 |
"step": 4500
|
469 |
},
|
470 |
{
|
471 |
"epoch": 0.8710801393728222,
|
472 |
-
"eval_loss": 0.
|
473 |
-
"eval_runtime":
|
474 |
-
"eval_samples_per_second":
|
475 |
-
"eval_steps_per_second": 4.
|
476 |
-
"eval_wer": 0.
|
477 |
"step": 4500
|
478 |
},
|
479 |
{
|
480 |
"epoch": 0.8904374758033294,
|
481 |
-
"eval_loss": 0.
|
482 |
-
"eval_runtime":
|
483 |
-
"eval_samples_per_second":
|
484 |
-
"eval_steps_per_second": 4.
|
485 |
-
"eval_wer": 0.
|
486 |
"step": 4600
|
487 |
},
|
488 |
{
|
489 |
"epoch": 0.9097948122338366,
|
490 |
-
"eval_loss": 0.
|
491 |
-
"eval_runtime":
|
492 |
-
"eval_samples_per_second":
|
493 |
-
"eval_steps_per_second": 4.
|
494 |
-
"eval_wer": 0.
|
495 |
"step": 4700
|
496 |
},
|
497 |
{
|
498 |
"epoch": 0.9291521486643438,
|
499 |
-
"eval_loss": 0.
|
500 |
-
"eval_runtime":
|
501 |
-
"eval_samples_per_second": 38.
|
502 |
-
"eval_steps_per_second": 4.
|
503 |
-
"eval_wer": 0.
|
504 |
"step": 4800
|
505 |
},
|
506 |
{
|
507 |
"epoch": 0.948509485094851,
|
508 |
-
"eval_loss": 0.
|
509 |
-
"eval_runtime":
|
510 |
-
"eval_samples_per_second":
|
511 |
-
"eval_steps_per_second": 4.
|
512 |
-
"eval_wer": 0.
|
513 |
"step": 4900
|
514 |
},
|
515 |
{
|
516 |
"epoch": 0.9678668215253581,
|
517 |
-
"grad_norm":
|
518 |
-
"learning_rate": 0.
|
519 |
-
"loss": 0.
|
520 |
"step": 5000
|
521 |
},
|
522 |
{
|
523 |
"epoch": 0.9678668215253581,
|
524 |
-
"eval_loss": 0.
|
525 |
-
"eval_runtime":
|
526 |
-
"eval_samples_per_second":
|
527 |
-
"eval_steps_per_second": 4.
|
528 |
-
"eval_wer": 0.
|
529 |
"step": 5000
|
530 |
},
|
531 |
{
|
532 |
"epoch": 0.9872241579558653,
|
533 |
-
"eval_loss": 0.
|
534 |
-
"eval_runtime":
|
535 |
-
"eval_samples_per_second":
|
536 |
-
"eval_steps_per_second": 4.
|
537 |
-
"eval_wer": 0.
|
538 |
"step": 5100
|
539 |
},
|
540 |
{
|
541 |
"epoch": 1.0065814943863725,
|
542 |
-
"eval_loss": 0.
|
543 |
-
"eval_runtime":
|
544 |
-
"eval_samples_per_second": 38.
|
545 |
-
"eval_steps_per_second": 4.
|
546 |
-
"eval_wer": 0.
|
547 |
"step": 5200
|
548 |
},
|
549 |
{
|
550 |
"epoch": 1.0259388308168795,
|
551 |
-
"eval_loss": 0.
|
552 |
-
"eval_runtime":
|
553 |
-
"eval_samples_per_second":
|
554 |
-
"eval_steps_per_second": 4.
|
555 |
-
"eval_wer": 0.
|
556 |
"step": 5300
|
557 |
},
|
558 |
{
|
559 |
"epoch": 1.0452961672473868,
|
560 |
-
"eval_loss": 0.
|
561 |
-
"eval_runtime":
|
562 |
-
"eval_samples_per_second": 38.
|
563 |
-
"eval_steps_per_second": 4.
|
564 |
-
"eval_wer": 0.
|
565 |
"step": 5400
|
566 |
},
|
567 |
{
|
568 |
"epoch": 1.064653503677894,
|
569 |
-
"grad_norm":
|
570 |
-
"learning_rate": 0.
|
571 |
-
"loss": 0.
|
572 |
"step": 5500
|
573 |
},
|
574 |
{
|
575 |
"epoch": 1.064653503677894,
|
576 |
-
"eval_loss": 0.
|
577 |
-
"eval_runtime":
|
578 |
-
"eval_samples_per_second":
|
579 |
-
"eval_steps_per_second": 4.
|
580 |
-
"eval_wer": 0.
|
581 |
"step": 5500
|
582 |
},
|
583 |
{
|
584 |
"epoch": 1.084010840108401,
|
585 |
-
"eval_loss": 0.
|
586 |
-
"eval_runtime":
|
587 |
-
"eval_samples_per_second": 38.
|
588 |
-
"eval_steps_per_second": 4.
|
589 |
-
"eval_wer": 0.
|
590 |
"step": 5600
|
591 |
},
|
592 |
{
|
593 |
"epoch": 1.1033681765389083,
|
594 |
-
"eval_loss": 0.
|
595 |
-
"eval_runtime":
|
596 |
-
"eval_samples_per_second": 38.
|
597 |
-
"eval_steps_per_second": 4.
|
598 |
-
"eval_wer": 0.
|
599 |
"step": 5700
|
600 |
},
|
601 |
{
|
602 |
"epoch": 1.1227255129694154,
|
603 |
-
"eval_loss": 0.
|
604 |
-
"eval_runtime":
|
605 |
-
"eval_samples_per_second": 38.
|
606 |
-
"eval_steps_per_second": 4.
|
607 |
-
"eval_wer": 0.
|
608 |
"step": 5800
|
609 |
},
|
610 |
{
|
611 |
"epoch": 1.1420828493999227,
|
612 |
-
"eval_loss": 0.
|
613 |
-
"eval_runtime":
|
614 |
-
"eval_samples_per_second":
|
615 |
-
"eval_steps_per_second": 4.
|
616 |
-
"eval_wer": 0.
|
617 |
"step": 5900
|
618 |
},
|
619 |
{
|
620 |
"epoch": 1.1614401858304297,
|
621 |
-
"grad_norm":
|
622 |
-
"learning_rate": 0.
|
623 |
-
"loss": 0.
|
624 |
"step": 6000
|
625 |
},
|
626 |
{
|
627 |
"epoch": 1.1614401858304297,
|
628 |
-
"eval_loss": 0.
|
629 |
-
"eval_runtime":
|
630 |
-
"eval_samples_per_second":
|
631 |
-
"eval_steps_per_second": 4.
|
632 |
-
"eval_wer": 0.
|
633 |
"step": 6000
|
634 |
},
|
635 |
{
|
636 |
"epoch": 1.1807975222609368,
|
637 |
-
"eval_loss": 0.
|
638 |
-
"eval_runtime":
|
639 |
-
"eval_samples_per_second":
|
640 |
-
"eval_steps_per_second": 4.
|
641 |
-
"eval_wer": 0.
|
642 |
"step": 6100
|
643 |
},
|
644 |
{
|
645 |
"epoch": 1.2001548586914441,
|
646 |
-
"eval_loss": 0.
|
647 |
-
"eval_runtime":
|
648 |
-
"eval_samples_per_second":
|
649 |
-
"eval_steps_per_second": 4.
|
650 |
-
"eval_wer": 0.
|
651 |
"step": 6200
|
652 |
},
|
653 |
{
|
654 |
"epoch": 1.2195121951219512,
|
655 |
-
"eval_loss": 0.
|
656 |
-
"eval_runtime":
|
657 |
-
"eval_samples_per_second":
|
658 |
-
"eval_steps_per_second": 4.
|
659 |
-
"eval_wer": 0.
|
660 |
"step": 6300
|
661 |
},
|
662 |
{
|
663 |
"epoch": 1.2388695315524583,
|
664 |
-
"eval_loss": 0.
|
665 |
-
"eval_runtime":
|
666 |
-
"eval_samples_per_second":
|
667 |
-
"eval_steps_per_second": 4.
|
668 |
-
"eval_wer": 0.
|
669 |
"step": 6400
|
670 |
},
|
671 |
{
|
672 |
"epoch": 1.2582268679829656,
|
673 |
-
"grad_norm": 1.
|
674 |
-
"learning_rate": 0.
|
675 |
-
"loss": 0.
|
676 |
"step": 6500
|
677 |
},
|
678 |
{
|
679 |
"epoch": 1.2582268679829656,
|
680 |
-
"eval_loss": 0.
|
681 |
-
"eval_runtime":
|
682 |
-
"eval_samples_per_second":
|
683 |
-
"eval_steps_per_second": 4.
|
684 |
-
"eval_wer": 0.
|
685 |
"step": 6500
|
686 |
},
|
687 |
{
|
688 |
"epoch": 1.2775842044134726,
|
689 |
-
"eval_loss": 0.
|
690 |
-
"eval_runtime":
|
691 |
-
"eval_samples_per_second":
|
692 |
-
"eval_steps_per_second": 4.
|
693 |
-
"eval_wer": 0.
|
694 |
"step": 6600
|
695 |
},
|
696 |
{
|
697 |
"epoch": 1.29694154084398,
|
698 |
-
"eval_loss": 0.
|
699 |
-
"eval_runtime":
|
700 |
-
"eval_samples_per_second":
|
701 |
-
"eval_steps_per_second": 4.
|
702 |
-
"eval_wer": 0.
|
703 |
"step": 6700
|
704 |
},
|
705 |
{
|
706 |
"epoch": 1.316298877274487,
|
707 |
-
"eval_loss": 0.
|
708 |
-
"eval_runtime":
|
709 |
-
"eval_samples_per_second": 38.
|
710 |
-
"eval_steps_per_second": 4.
|
711 |
-
"eval_wer": 0.
|
712 |
"step": 6800
|
713 |
},
|
714 |
{
|
715 |
"epoch": 1.3356562137049943,
|
716 |
-
"eval_loss": 0.
|
717 |
-
"eval_runtime":
|
718 |
-
"eval_samples_per_second":
|
719 |
-
"eval_steps_per_second": 4.
|
720 |
-
"eval_wer": 0.
|
721 |
"step": 6900
|
722 |
},
|
723 |
{
|
724 |
"epoch": 1.3550135501355014,
|
725 |
-
"grad_norm": 0.
|
726 |
-
"learning_rate": 9.
|
727 |
-
"loss": 0.
|
728 |
"step": 7000
|
729 |
},
|
730 |
{
|
731 |
"epoch": 1.3550135501355014,
|
732 |
-
"eval_loss": 0.
|
733 |
-
"eval_runtime":
|
734 |
-
"eval_samples_per_second":
|
735 |
-
"eval_steps_per_second": 4.
|
736 |
-
"eval_wer": 0.
|
737 |
"step": 7000
|
738 |
},
|
739 |
{
|
740 |
"epoch": 1.3743708865660085,
|
741 |
-
"eval_loss": 0.
|
742 |
-
"eval_runtime":
|
743 |
-
"eval_samples_per_second": 38.
|
744 |
-
"eval_steps_per_second": 4.
|
745 |
-
"eval_wer": 0.
|
746 |
"step": 7100
|
747 |
},
|
748 |
{
|
749 |
"epoch": 1.3937282229965158,
|
750 |
-
"eval_loss": 0.
|
751 |
-
"eval_runtime":
|
752 |
-
"eval_samples_per_second":
|
753 |
-
"eval_steps_per_second": 4.
|
754 |
-
"eval_wer": 0.
|
755 |
"step": 7200
|
756 |
},
|
757 |
{
|
758 |
"epoch": 1.4130855594270229,
|
759 |
-
"eval_loss": 0.
|
760 |
-
"eval_runtime":
|
761 |
-
"eval_samples_per_second":
|
762 |
-
"eval_steps_per_second": 4.
|
763 |
-
"eval_wer": 0.
|
764 |
"step": 7300
|
765 |
},
|
766 |
{
|
767 |
"epoch": 1.43244289585753,
|
768 |
-
"eval_loss": 0.
|
769 |
-
"eval_runtime":
|
770 |
-
"eval_samples_per_second":
|
771 |
-
"eval_steps_per_second": 4.
|
772 |
-
"eval_wer": 0.
|
773 |
"step": 7400
|
774 |
},
|
775 |
{
|
776 |
"epoch": 1.4518002322880372,
|
777 |
-
"grad_norm":
|
778 |
-
"learning_rate": 7.
|
779 |
-
"loss": 0.
|
780 |
"step": 7500
|
781 |
},
|
782 |
{
|
783 |
"epoch": 1.4518002322880372,
|
784 |
-
"eval_loss": 0.
|
785 |
-
"eval_runtime": 149.
|
786 |
-
"eval_samples_per_second": 37.
|
787 |
-
"eval_steps_per_second": 4.
|
788 |
-
"eval_wer": 0.
|
789 |
"step": 7500
|
790 |
},
|
791 |
{
|
792 |
"epoch": 1.4711575687185443,
|
793 |
-
"eval_loss": 0.
|
794 |
-
"eval_runtime":
|
795 |
-
"eval_samples_per_second":
|
796 |
-
"eval_steps_per_second": 4.
|
797 |
-
"eval_wer": 0.
|
798 |
"step": 7600
|
799 |
},
|
800 |
{
|
801 |
"epoch": 1.4905149051490514,
|
802 |
-
"eval_loss": 0.
|
803 |
-
"eval_runtime":
|
804 |
-
"eval_samples_per_second":
|
805 |
-
"eval_steps_per_second": 4.
|
806 |
-
"eval_wer": 0.
|
807 |
"step": 7700
|
808 |
},
|
809 |
{
|
810 |
"epoch": 1.5098722415795587,
|
811 |
-
"eval_loss": 0.
|
812 |
-
"eval_runtime":
|
813 |
-
"eval_samples_per_second":
|
814 |
-
"eval_steps_per_second": 4.
|
815 |
-
"eval_wer": 0.
|
816 |
"step": 7800
|
817 |
},
|
818 |
{
|
819 |
"epoch": 1.5292295780100658,
|
820 |
-
"eval_loss": 0.
|
821 |
-
"eval_runtime":
|
822 |
-
"eval_samples_per_second":
|
823 |
-
"eval_steps_per_second": 4.
|
824 |
-
"eval_wer": 0.
|
825 |
"step": 7900
|
826 |
},
|
827 |
{
|
828 |
"epoch": 1.5485869144405728,
|
829 |
-
"grad_norm": 1.
|
830 |
-
"learning_rate": 6.
|
831 |
-
"loss": 0.
|
832 |
"step": 8000
|
833 |
},
|
834 |
{
|
835 |
"epoch": 1.5485869144405728,
|
836 |
-
"eval_loss": 0.
|
837 |
-
"eval_runtime":
|
838 |
-
"eval_samples_per_second":
|
839 |
-
"eval_steps_per_second": 4.
|
840 |
-
"eval_wer": 0.
|
841 |
"step": 8000
|
842 |
},
|
843 |
{
|
844 |
"epoch": 1.5679442508710801,
|
845 |
-
"eval_loss": 0.
|
846 |
-
"eval_runtime":
|
847 |
-
"eval_samples_per_second":
|
848 |
-
"eval_steps_per_second": 4.
|
849 |
-
"eval_wer": 0.
|
850 |
"step": 8100
|
851 |
},
|
852 |
{
|
853 |
"epoch": 1.5873015873015874,
|
854 |
-
"eval_loss": 0.
|
855 |
-
"eval_runtime":
|
856 |
-
"eval_samples_per_second":
|
857 |
-
"eval_steps_per_second": 4.
|
858 |
-
"eval_wer": 0.
|
859 |
"step": 8200
|
860 |
},
|
861 |
{
|
862 |
"epoch": 1.6066589237320945,
|
863 |
-
"eval_loss": 0.
|
864 |
-
"eval_runtime":
|
865 |
-
"eval_samples_per_second":
|
866 |
-
"eval_steps_per_second": 4.
|
867 |
-
"eval_wer": 0.
|
868 |
"step": 8300
|
869 |
},
|
870 |
{
|
871 |
"epoch": 1.6260162601626016,
|
872 |
-
"eval_loss": 0.
|
873 |
-
"eval_runtime":
|
874 |
-
"eval_samples_per_second":
|
875 |
-
"eval_steps_per_second": 4.
|
876 |
-
"eval_wer": 0.
|
877 |
"step": 8400
|
878 |
},
|
879 |
{
|
880 |
"epoch": 1.645373596593109,
|
881 |
-
"grad_norm":
|
882 |
-
"learning_rate": 4.
|
883 |
-
"loss": 0.
|
884 |
"step": 8500
|
885 |
},
|
886 |
{
|
887 |
"epoch": 1.645373596593109,
|
888 |
-
"eval_loss": 0.
|
889 |
-
"eval_runtime":
|
890 |
-
"eval_samples_per_second":
|
891 |
-
"eval_steps_per_second": 4.
|
892 |
-
"eval_wer": 0.
|
893 |
"step": 8500
|
894 |
},
|
895 |
{
|
896 |
"epoch": 1.664730933023616,
|
897 |
-
"eval_loss": 0.
|
898 |
-
"eval_runtime":
|
899 |
-
"eval_samples_per_second":
|
900 |
-
"eval_steps_per_second": 4.
|
901 |
-
"eval_wer": 0.
|
902 |
"step": 8600
|
903 |
},
|
904 |
{
|
905 |
"epoch": 1.684088269454123,
|
906 |
-
"eval_loss": 0.
|
907 |
-
"eval_runtime":
|
908 |
-
"eval_samples_per_second":
|
909 |
-
"eval_steps_per_second": 4.
|
910 |
-
"eval_wer": 0.
|
911 |
"step": 8700
|
912 |
},
|
913 |
{
|
914 |
"epoch": 1.7034456058846303,
|
915 |
-
"eval_loss": 0.
|
916 |
-
"eval_runtime":
|
917 |
-
"eval_samples_per_second":
|
918 |
-
"eval_steps_per_second": 4.
|
919 |
-
"eval_wer": 0.
|
920 |
"step": 8800
|
921 |
},
|
922 |
{
|
923 |
"epoch": 1.7228029423151374,
|
924 |
-
"eval_loss": 0.
|
925 |
-
"eval_runtime":
|
926 |
-
"eval_samples_per_second":
|
927 |
-
"eval_steps_per_second": 4.
|
928 |
-
"eval_wer": 0.
|
929 |
"step": 8900
|
930 |
},
|
931 |
{
|
932 |
"epoch": 1.7421602787456445,
|
933 |
-
"grad_norm": 1.
|
934 |
-
"learning_rate": 3.
|
935 |
-
"loss": 0.
|
936 |
"step": 9000
|
937 |
},
|
938 |
{
|
939 |
"epoch": 1.7421602787456445,
|
940 |
-
"eval_loss": 0.
|
941 |
-
"eval_runtime":
|
942 |
-
"eval_samples_per_second":
|
943 |
-
"eval_steps_per_second": 4.
|
944 |
-
"eval_wer": 0.
|
945 |
"step": 9000
|
946 |
},
|
947 |
{
|
948 |
"epoch": 1.7615176151761518,
|
949 |
-
"eval_loss": 0.
|
950 |
-
"eval_runtime":
|
951 |
-
"eval_samples_per_second":
|
952 |
-
"eval_steps_per_second": 4.
|
953 |
-
"eval_wer": 0.
|
954 |
"step": 9100
|
955 |
},
|
956 |
{
|
957 |
"epoch": 1.7808749516066589,
|
958 |
-
"eval_loss": 0.
|
959 |
-
"eval_runtime":
|
960 |
-
"eval_samples_per_second":
|
961 |
-
"eval_steps_per_second": 4.
|
962 |
-
"eval_wer": 0.
|
963 |
"step": 9200
|
964 |
},
|
965 |
{
|
966 |
"epoch": 1.800232288037166,
|
967 |
-
"eval_loss": 0.
|
968 |
-
"eval_runtime":
|
969 |
-
"eval_samples_per_second":
|
970 |
-
"eval_steps_per_second": 4.
|
971 |
-
"eval_wer": 0.
|
972 |
"step": 9300
|
973 |
},
|
974 |
{
|
975 |
"epoch": 1.8195896244676733,
|
976 |
-
"eval_loss": 0.
|
977 |
-
"eval_runtime": 150.
|
978 |
-
"eval_samples_per_second": 37.
|
979 |
-
"eval_steps_per_second": 4.
|
980 |
-
"eval_wer": 0.
|
981 |
"step": 9400
|
982 |
},
|
983 |
{
|
984 |
"epoch": 1.8389469608981805,
|
985 |
-
"grad_norm":
|
986 |
-
"learning_rate": 1.
|
987 |
-
"loss": 0.
|
988 |
"step": 9500
|
989 |
},
|
990 |
{
|
991 |
"epoch": 1.8389469608981805,
|
992 |
-
"eval_loss": 0.
|
993 |
-
"eval_runtime":
|
994 |
-
"eval_samples_per_second":
|
995 |
-
"eval_steps_per_second": 4.
|
996 |
-
"eval_wer": 0.
|
997 |
"step": 9500
|
998 |
},
|
999 |
{
|
1000 |
"epoch": 1.8583042973286876,
|
1001 |
-
"eval_loss": 0.
|
1002 |
-
"eval_runtime":
|
1003 |
-
"eval_samples_per_second":
|
1004 |
-
"eval_steps_per_second": 4.
|
1005 |
-
"eval_wer": 0.
|
1006 |
"step": 9600
|
1007 |
},
|
1008 |
{
|
1009 |
"epoch": 1.8776616337591947,
|
1010 |
-
"eval_loss": 0.
|
1011 |
-
"eval_runtime":
|
1012 |
-
"eval_samples_per_second":
|
1013 |
-
"eval_steps_per_second": 4.
|
1014 |
-
"eval_wer": 0.
|
1015 |
"step": 9700
|
1016 |
},
|
1017 |
{
|
1018 |
"epoch": 1.897018970189702,
|
1019 |
-
"eval_loss": 0.
|
1020 |
-
"eval_runtime":
|
1021 |
-
"eval_samples_per_second":
|
1022 |
-
"eval_steps_per_second": 4.
|
1023 |
-
"eval_wer": 0.
|
1024 |
"step": 9800
|
1025 |
},
|
1026 |
{
|
1027 |
"epoch": 1.916376306620209,
|
1028 |
-
"eval_loss": 0.
|
1029 |
-
"eval_runtime":
|
1030 |
-
"eval_samples_per_second":
|
1031 |
-
"eval_steps_per_second": 4.
|
1032 |
-
"eval_wer": 0.
|
1033 |
"step": 9900
|
1034 |
},
|
1035 |
{
|
1036 |
"epoch": 1.9357336430507162,
|
1037 |
-
"grad_norm":
|
1038 |
-
"learning_rate":
|
1039 |
-
"loss": 0.
|
1040 |
"step": 10000
|
1041 |
},
|
1042 |
{
|
1043 |
"epoch": 1.9357336430507162,
|
1044 |
-
"eval_loss": 0.
|
1045 |
-
"eval_runtime":
|
1046 |
-
"eval_samples_per_second":
|
1047 |
-
"eval_steps_per_second": 4.
|
1048 |
-
"eval_wer": 0.
|
1049 |
"step": 10000
|
1050 |
},
|
1051 |
{
|
1052 |
"epoch": 1.9357336430507162,
|
1053 |
"step": 10000,
|
1054 |
"total_flos": 1.1255918428180738e+19,
|
1055 |
-
"train_loss": 0.
|
1056 |
-
"train_runtime":
|
1057 |
-
"train_samples_per_second": 4.
|
1058 |
-
"train_steps_per_second": 0.
|
1059 |
}
|
1060 |
],
|
1061 |
"logging_steps": 500,
|
|
|
10 |
"log_history": [
|
11 |
{
|
12 |
"epoch": 0.019357336430507164,
|
13 |
+
"eval_loss": 3.554515838623047,
|
14 |
+
"eval_runtime": 148.6327,
|
15 |
+
"eval_samples_per_second": 38.054,
|
16 |
+
"eval_steps_per_second": 4.757,
|
17 |
"eval_wer": 1.0,
|
18 |
"step": 100
|
19 |
},
|
20 |
{
|
21 |
"epoch": 0.03871467286101433,
|
22 |
+
"eval_loss": 3.02604603767395,
|
23 |
+
"eval_runtime": 146.1639,
|
24 |
+
"eval_samples_per_second": 38.696,
|
25 |
+
"eval_steps_per_second": 4.837,
|
26 |
"eval_wer": 1.0,
|
27 |
"step": 200
|
28 |
},
|
29 |
{
|
30 |
"epoch": 0.05807200929152149,
|
31 |
+
"eval_loss": 2.906620979309082,
|
32 |
+
"eval_runtime": 146.0171,
|
33 |
+
"eval_samples_per_second": 38.735,
|
34 |
+
"eval_steps_per_second": 4.842,
|
35 |
"eval_wer": 1.0,
|
36 |
"step": 300
|
37 |
},
|
38 |
{
|
39 |
"epoch": 0.07742934572202866,
|
40 |
+
"eval_loss": 2.013317823410034,
|
41 |
+
"eval_runtime": 145.6964,
|
42 |
+
"eval_samples_per_second": 38.82,
|
43 |
+
"eval_steps_per_second": 4.853,
|
44 |
+
"eval_wer": 0.9847217987193272,
|
45 |
"step": 400
|
46 |
},
|
47 |
{
|
48 |
"epoch": 0.09678668215253582,
|
49 |
+
"grad_norm": 4.750732421875,
|
50 |
+
"learning_rate": 0.0002982,
|
51 |
+
"loss": 4.0489,
|
52 |
"step": 500
|
53 |
},
|
54 |
{
|
55 |
"epoch": 0.09678668215253582,
|
56 |
+
"eval_loss": 1.459844708442688,
|
57 |
+
"eval_runtime": 145.9454,
|
58 |
+
"eval_samples_per_second": 38.754,
|
59 |
+
"eval_steps_per_second": 4.844,
|
60 |
+
"eval_wer": 0.9003546725297299,
|
61 |
"step": 500
|
62 |
},
|
63 |
{
|
64 |
"epoch": 0.11614401858304298,
|
65 |
+
"eval_loss": 1.1772356033325195,
|
66 |
+
"eval_runtime": 145.9941,
|
67 |
+
"eval_samples_per_second": 38.741,
|
68 |
+
"eval_steps_per_second": 4.843,
|
69 |
+
"eval_wer": 0.8042079247644878,
|
70 |
"step": 600
|
71 |
},
|
72 |
{
|
73 |
"epoch": 0.13550135501355012,
|
74 |
+
"eval_loss": 1.07865571975708,
|
75 |
+
"eval_runtime": 146.4442,
|
76 |
+
"eval_samples_per_second": 38.622,
|
77 |
+
"eval_steps_per_second": 4.828,
|
78 |
+
"eval_wer": 0.7589510680297219,
|
79 |
"step": 700
|
80 |
},
|
81 |
{
|
82 |
"epoch": 0.1548586914440573,
|
83 |
+
"eval_loss": 1.0144398212432861,
|
84 |
+
"eval_runtime": 146.6444,
|
85 |
+
"eval_samples_per_second": 38.569,
|
86 |
+
"eval_steps_per_second": 4.821,
|
87 |
+
"eval_wer": 0.7211567780969652,
|
88 |
"step": 800
|
89 |
},
|
90 |
{
|
91 |
"epoch": 0.17421602787456447,
|
92 |
+
"eval_loss": 0.9338767528533936,
|
93 |
+
"eval_runtime": 147.2439,
|
94 |
+
"eval_samples_per_second": 38.412,
|
95 |
+
"eval_steps_per_second": 4.802,
|
96 |
+
"eval_wer": 0.6931520919259841,
|
97 |
"step": 900
|
98 |
},
|
99 |
{
|
100 |
"epoch": 0.19357336430507163,
|
101 |
+
"grad_norm": 3.122023105621338,
|
102 |
+
"learning_rate": 0.0002843684210526315,
|
103 |
+
"loss": 1.0454,
|
104 |
"step": 1000
|
105 |
},
|
106 |
{
|
107 |
"epoch": 0.19357336430507163,
|
108 |
+
"eval_loss": 0.8805950284004211,
|
109 |
+
"eval_runtime": 146.7917,
|
110 |
+
"eval_samples_per_second": 38.531,
|
111 |
+
"eval_steps_per_second": 4.816,
|
112 |
+
"eval_wer": 0.659739050889891,
|
113 |
"step": 1000
|
114 |
},
|
115 |
{
|
116 |
"epoch": 0.2129307007355788,
|
117 |
+
"eval_loss": 0.864371657371521,
|
118 |
+
"eval_runtime": 146.75,
|
119 |
+
"eval_samples_per_second": 38.542,
|
120 |
"eval_steps_per_second": 4.818,
|
121 |
+
"eval_wer": 0.6553738505239846,
|
122 |
"step": 1100
|
123 |
},
|
124 |
{
|
125 |
"epoch": 0.23228803716608595,
|
126 |
+
"eval_loss": 0.845355749130249,
|
127 |
+
"eval_runtime": 147.3826,
|
128 |
+
"eval_samples_per_second": 38.376,
|
129 |
+
"eval_steps_per_second": 4.797,
|
130 |
+
"eval_wer": 0.6313652485114988,
|
131 |
"step": 1200
|
132 |
},
|
133 |
{
|
134 |
"epoch": 0.2516453735965931,
|
135 |
+
"eval_loss": 0.8092750906944275,
|
136 |
+
"eval_runtime": 147.8588,
|
137 |
+
"eval_samples_per_second": 38.253,
|
138 |
+
"eval_steps_per_second": 4.782,
|
139 |
+
"eval_wer": 0.5919179599107701,
|
140 |
"step": 1300
|
141 |
},
|
142 |
{
|
143 |
"epoch": 0.27100271002710025,
|
144 |
+
"eval_loss": 0.8075847029685974,
|
145 |
+
"eval_runtime": 147.3238,
|
146 |
+
"eval_samples_per_second": 38.392,
|
147 |
+
"eval_steps_per_second": 4.799,
|
148 |
+
"eval_wer": 0.6072282582529569,
|
149 |
"step": 1400
|
150 |
},
|
151 |
{
|
152 |
"epoch": 0.29036004645760743,
|
153 |
+
"grad_norm": 3.26877498626709,
|
154 |
+
"learning_rate": 0.000268578947368421,
|
155 |
+
"loss": 0.842,
|
156 |
"step": 1500
|
157 |
},
|
158 |
{
|
159 |
"epoch": 0.29036004645760743,
|
160 |
+
"eval_loss": 0.7783301472663879,
|
161 |
+
"eval_runtime": 151.3584,
|
162 |
+
"eval_samples_per_second": 37.368,
|
163 |
+
"eval_steps_per_second": 4.671,
|
164 |
+
"eval_wer": 0.5857232270385646,
|
165 |
"step": 1500
|
166 |
},
|
167 |
{
|
168 |
"epoch": 0.3097173828881146,
|
169 |
+
"eval_loss": 0.7964575290679932,
|
170 |
+
"eval_runtime": 147.3418,
|
171 |
+
"eval_samples_per_second": 38.387,
|
172 |
+
"eval_steps_per_second": 4.798,
|
173 |
+
"eval_wer": 0.5940524145014524,
|
174 |
"step": 1600
|
175 |
},
|
176 |
{
|
177 |
"epoch": 0.32907471931862176,
|
178 |
+
"eval_loss": 0.7414730191230774,
|
179 |
+
"eval_runtime": 145.4997,
|
180 |
+
"eval_samples_per_second": 38.873,
|
181 |
+
"eval_steps_per_second": 4.859,
|
182 |
+
"eval_wer": 0.5505127505576864,
|
183 |
"step": 1700
|
184 |
},
|
185 |
{
|
186 |
"epoch": 0.34843205574912894,
|
187 |
+
"eval_loss": 0.7440277934074402,
|
188 |
+
"eval_runtime": 144.5466,
|
189 |
+
"eval_samples_per_second": 39.129,
|
190 |
+
"eval_steps_per_second": 4.891,
|
191 |
+
"eval_wer": 0.563656497247677,
|
192 |
"step": 1800
|
193 |
},
|
194 |
{
|
195 |
"epoch": 0.3677893921796361,
|
196 |
+
"eval_loss": 0.7360929846763611,
|
197 |
+
"eval_runtime": 144.6106,
|
198 |
+
"eval_samples_per_second": 39.112,
|
199 |
+
"eval_steps_per_second": 4.889,
|
200 |
+
"eval_wer": 0.5865256535764151,
|
201 |
"step": 1900
|
202 |
},
|
203 |
{
|
204 |
"epoch": 0.38714672861014326,
|
205 |
+
"grad_norm": 1.8907363414764404,
|
206 |
+
"learning_rate": 0.0002527894736842105,
|
207 |
+
"loss": 0.755,
|
208 |
"step": 2000
|
209 |
},
|
210 |
{
|
211 |
"epoch": 0.38714672861014326,
|
212 |
+
"eval_loss": 0.7313781380653381,
|
213 |
+
"eval_runtime": 144.7461,
|
214 |
+
"eval_samples_per_second": 39.075,
|
215 |
+
"eval_steps_per_second": 4.884,
|
216 |
+
"eval_wer": 0.542729213140537,
|
217 |
"step": 2000
|
218 |
},
|
219 |
{
|
220 |
"epoch": 0.4065040650406504,
|
221 |
+
"eval_loss": 0.6866404414176941,
|
222 |
+
"eval_runtime": 144.8178,
|
223 |
+
"eval_samples_per_second": 39.056,
|
224 |
+
"eval_steps_per_second": 4.882,
|
225 |
+
"eval_wer": 0.5181428640207989,
|
226 |
"step": 2100
|
227 |
},
|
228 |
{
|
229 |
"epoch": 0.4258614014711576,
|
230 |
+
"eval_loss": 0.6947888731956482,
|
231 |
+
"eval_runtime": 144.3916,
|
232 |
+
"eval_samples_per_second": 39.171,
|
233 |
+
"eval_steps_per_second": 4.896,
|
234 |
+
"eval_wer": 0.5425847763637239,
|
235 |
"step": 2200
|
236 |
},
|
237 |
{
|
238 |
"epoch": 0.4452187379016647,
|
239 |
+
"eval_loss": 0.6796152591705322,
|
240 |
+
"eval_runtime": 144.4575,
|
241 |
+
"eval_samples_per_second": 39.153,
|
242 |
+
"eval_steps_per_second": 4.894,
|
243 |
+
"eval_wer": 0.5158960697148176,
|
244 |
"step": 2300
|
245 |
},
|
246 |
{
|
247 |
"epoch": 0.4645760743321719,
|
248 |
+
"eval_loss": 0.6898629069328308,
|
249 |
+
"eval_runtime": 144.5493,
|
250 |
+
"eval_samples_per_second": 39.129,
|
251 |
+
"eval_steps_per_second": 4.891,
|
252 |
+
"eval_wer": 0.5305162812344529,
|
253 |
"step": 2400
|
254 |
},
|
255 |
{
|
256 |
"epoch": 0.48393341076267904,
|
257 |
+
"grad_norm": 2.9580163955688477,
|
258 |
+
"learning_rate": 0.000237,
|
259 |
+
"loss": 0.6884,
|
260 |
"step": 2500
|
261 |
},
|
262 |
{
|
263 |
"epoch": 0.48393341076267904,
|
264 |
+
"eval_loss": 0.6736070513725281,
|
265 |
+
"eval_runtime": 144.9197,
|
266 |
+
"eval_samples_per_second": 39.029,
|
267 |
+
"eval_steps_per_second": 4.879,
|
268 |
+
"eval_wer": 0.5103272295421354,
|
269 |
"step": 2500
|
270 |
},
|
271 |
{
|
272 |
"epoch": 0.5032907471931862,
|
273 |
+
"eval_loss": 0.6728157997131348,
|
274 |
+
"eval_runtime": 145.4712,
|
275 |
+
"eval_samples_per_second": 38.881,
|
276 |
+
"eval_steps_per_second": 4.86,
|
277 |
+
"eval_wer": 0.5257338190688643,
|
278 |
"step": 2600
|
279 |
},
|
280 |
{
|
281 |
"epoch": 0.5226480836236934,
|
282 |
+
"eval_loss": 0.6537250876426697,
|
283 |
+
"eval_runtime": 145.152,
|
284 |
+
"eval_samples_per_second": 38.966,
|
285 |
+
"eval_steps_per_second": 4.871,
|
286 |
+
"eval_wer": 0.5026560318402851,
|
287 |
"step": 2700
|
288 |
},
|
289 |
{
|
290 |
"epoch": 0.5420054200542005,
|
291 |
+
"eval_loss": 0.631415843963623,
|
292 |
+
"eval_runtime": 145.09,
|
293 |
+
"eval_samples_per_second": 38.983,
|
294 |
+
"eval_steps_per_second": 4.873,
|
295 |
+
"eval_wer": 0.48227439777888337,
|
296 |
"step": 2800
|
297 |
},
|
298 |
{
|
299 |
"epoch": 0.5613627564847077,
|
300 |
+
"eval_loss": 0.6316519379615784,
|
301 |
+
"eval_runtime": 144.7345,
|
302 |
+
"eval_samples_per_second": 39.078,
|
303 |
+
"eval_steps_per_second": 4.885,
|
304 |
+
"eval_wer": 0.48296448460143476,
|
305 |
"step": 2900
|
306 |
},
|
307 |
{
|
308 |
"epoch": 0.5807200929152149,
|
309 |
+
"grad_norm": 8.703415870666504,
|
310 |
+
"learning_rate": 0.00022121052631578946,
|
311 |
+
"loss": 0.6756,
|
312 |
"step": 3000
|
313 |
},
|
314 |
{
|
315 |
"epoch": 0.5807200929152149,
|
316 |
+
"eval_loss": 0.6204401254653931,
|
317 |
+
"eval_runtime": 144.9451,
|
318 |
+
"eval_samples_per_second": 39.022,
|
319 |
+
"eval_steps_per_second": 4.878,
|
320 |
+
"eval_wer": 0.4761117619681918,
|
321 |
"step": 3000
|
322 |
},
|
323 |
{
|
324 |
"epoch": 0.6000774293457221,
|
325 |
+
"eval_loss": 0.6310548782348633,
|
326 |
+
"eval_runtime": 144.6818,
|
327 |
+
"eval_samples_per_second": 39.093,
|
328 |
+
"eval_steps_per_second": 4.887,
|
329 |
+
"eval_wer": 0.4810868065028647,
|
330 |
"step": 3100
|
331 |
},
|
332 |
{
|
333 |
"epoch": 0.6194347657762292,
|
334 |
+
"eval_loss": 0.623622477054596,
|
335 |
+
"eval_runtime": 145.2962,
|
336 |
+
"eval_samples_per_second": 38.927,
|
337 |
+
"eval_steps_per_second": 4.866,
|
338 |
+
"eval_wer": 0.48631862752964966,
|
339 |
"step": 3200
|
340 |
},
|
341 |
{
|
342 |
"epoch": 0.6387921022067363,
|
343 |
+
"eval_loss": 0.6224333047866821,
|
344 |
+
"eval_runtime": 144.731,
|
345 |
+
"eval_samples_per_second": 39.079,
|
346 |
+
"eval_steps_per_second": 4.885,
|
347 |
+
"eval_wer": 0.46287172409365923,
|
348 |
"step": 3300
|
349 |
},
|
350 |
{
|
351 |
"epoch": 0.6581494386372435,
|
352 |
+
"eval_loss": 0.597332239151001,
|
353 |
+
"eval_runtime": 145.3577,
|
354 |
+
"eval_samples_per_second": 38.911,
|
355 |
+
"eval_steps_per_second": 4.864,
|
356 |
+
"eval_wer": 0.4622939769864069,
|
357 |
"step": 3400
|
358 |
},
|
359 |
{
|
360 |
"epoch": 0.6775067750677507,
|
361 |
+
"grad_norm": 4.671802997589111,
|
362 |
+
"learning_rate": 0.00020542105263157893,
|
363 |
+
"loss": 0.6435,
|
364 |
"step": 3500
|
365 |
},
|
366 |
{
|
367 |
"epoch": 0.6775067750677507,
|
368 |
+
"eval_loss": 0.5913041830062866,
|
369 |
+
"eval_runtime": 145.2068,
|
370 |
+
"eval_samples_per_second": 38.951,
|
371 |
+
"eval_steps_per_second": 4.869,
|
372 |
+
"eval_wer": 0.47079969828762175,
|
373 |
"step": 3500
|
374 |
},
|
375 |
{
|
376 |
"epoch": 0.6968641114982579,
|
377 |
+
"eval_loss": 0.6086587905883789,
|
378 |
+
"eval_runtime": 145.0548,
|
379 |
+
"eval_samples_per_second": 38.992,
|
380 |
+
"eval_steps_per_second": 4.874,
|
381 |
+
"eval_wer": 0.4743945691771918,
|
382 |
"step": 3600
|
383 |
},
|
384 |
{
|
385 |
"epoch": 0.716221447928765,
|
386 |
+
"eval_loss": 0.5827310681343079,
|
387 |
+
"eval_runtime": 145.6494,
|
388 |
+
"eval_samples_per_second": 38.833,
|
389 |
+
"eval_steps_per_second": 4.854,
|
390 |
+
"eval_wer": 0.45213525701722007,
|
391 |
"step": 3700
|
392 |
},
|
393 |
{
|
394 |
"epoch": 0.7355787843592722,
|
395 |
+
"eval_loss": 0.5875205397605896,
|
396 |
+
"eval_runtime": 145.7866,
|
397 |
+
"eval_samples_per_second": 38.796,
|
398 |
+
"eval_steps_per_second": 4.85,
|
399 |
+
"eval_wer": 0.460785415095248,
|
400 |
"step": 3800
|
401 |
},
|
402 |
{
|
403 |
"epoch": 0.7549361207897793,
|
404 |
+
"eval_loss": 0.592467725276947,
|
405 |
+
"eval_runtime": 145.4885,
|
406 |
+
"eval_samples_per_second": 38.876,
|
407 |
+
"eval_steps_per_second": 4.859,
|
408 |
+
"eval_wer": 0.45573012790679013,
|
409 |
"step": 3900
|
410 |
},
|
411 |
{
|
412 |
"epoch": 0.7742934572202865,
|
413 |
+
"grad_norm": 3.854473114013672,
|
414 |
+
"learning_rate": 0.0001896315789473684,
|
415 |
+
"loss": 0.6282,
|
416 |
"step": 4000
|
417 |
},
|
418 |
{
|
419 |
"epoch": 0.7742934572202865,
|
420 |
+
"eval_loss": 0.5799296498298645,
|
421 |
+
"eval_runtime": 145.6866,
|
422 |
+
"eval_samples_per_second": 38.823,
|
423 |
+
"eval_steps_per_second": 4.853,
|
424 |
+
"eval_wer": 0.4494230553192855,
|
425 |
"step": 4000
|
426 |
},
|
427 |
{
|
428 |
"epoch": 0.7936507936507936,
|
429 |
+
"eval_loss": 0.567882239818573,
|
430 |
+
"eval_runtime": 146.7561,
|
431 |
+
"eval_samples_per_second": 38.54,
|
432 |
+
"eval_steps_per_second": 4.818,
|
433 |
+
"eval_wer": 0.45260066440917335,
|
434 |
"step": 4100
|
435 |
},
|
436 |
{
|
437 |
"epoch": 0.8130081300813008,
|
438 |
+
"eval_loss": 0.5699547529220581,
|
439 |
+
"eval_runtime": 145.6353,
|
440 |
+
"eval_samples_per_second": 38.837,
|
441 |
+
"eval_steps_per_second": 4.855,
|
442 |
+
"eval_wer": 0.4549597984304537,
|
443 |
"step": 4200
|
444 |
},
|
445 |
{
|
446 |
"epoch": 0.832365466511808,
|
447 |
+
"eval_loss": 0.5610417127609253,
|
448 |
+
"eval_runtime": 145.0921,
|
449 |
+
"eval_samples_per_second": 38.982,
|
450 |
+
"eval_steps_per_second": 4.873,
|
451 |
+
"eval_wer": 0.43432138787693986,
|
452 |
"step": 4300
|
453 |
},
|
454 |
{
|
455 |
"epoch": 0.8517228029423152,
|
456 |
+
"eval_loss": 0.5616321563720703,
|
457 |
+
"eval_runtime": 145.7565,
|
458 |
+
"eval_samples_per_second": 38.804,
|
459 |
+
"eval_steps_per_second": 4.851,
|
460 |
+
"eval_wer": 0.42727608287461283,
|
461 |
"step": 4400
|
462 |
},
|
463 |
{
|
464 |
"epoch": 0.8710801393728222,
|
465 |
+
"grad_norm": 3.4542479515075684,
|
466 |
+
"learning_rate": 0.0001738421052631579,
|
467 |
+
"loss": 0.5937,
|
468 |
"step": 4500
|
469 |
},
|
470 |
{
|
471 |
"epoch": 0.8710801393728222,
|
472 |
+
"eval_loss": 0.5463823080062866,
|
473 |
+
"eval_runtime": 145.3415,
|
474 |
+
"eval_samples_per_second": 38.915,
|
475 |
+
"eval_steps_per_second": 4.864,
|
476 |
+
"eval_wer": 0.42207635890934186,
|
477 |
"step": 4500
|
478 |
},
|
479 |
{
|
480 |
"epoch": 0.8904374758033294,
|
481 |
+
"eval_loss": 0.5485692620277405,
|
482 |
+
"eval_runtime": 145.5247,
|
483 |
+
"eval_samples_per_second": 38.866,
|
484 |
+
"eval_steps_per_second": 4.858,
|
485 |
+
"eval_wer": 0.4287685962350147,
|
486 |
"step": 4600
|
487 |
},
|
488 |
{
|
489 |
"epoch": 0.9097948122338366,
|
490 |
+
"eval_loss": 0.5307685136795044,
|
491 |
+
"eval_runtime": 145.423,
|
492 |
+
"eval_samples_per_second": 38.893,
|
493 |
+
"eval_steps_per_second": 4.862,
|
494 |
+
"eval_wer": 0.41673219816725776,
|
495 |
"step": 4700
|
496 |
},
|
497 |
{
|
498 |
"epoch": 0.9291521486643438,
|
499 |
+
"eval_loss": 0.5520421862602234,
|
500 |
+
"eval_runtime": 145.5542,
|
501 |
+
"eval_samples_per_second": 38.858,
|
502 |
+
"eval_steps_per_second": 4.857,
|
503 |
+
"eval_wer": 0.41997400138017366,
|
504 |
"step": 4800
|
505 |
},
|
506 |
{
|
507 |
"epoch": 0.948509485094851,
|
508 |
+
"eval_loss": 0.5321463346481323,
|
509 |
+
"eval_runtime": 146.3191,
|
510 |
+
"eval_samples_per_second": 38.655,
|
511 |
+
"eval_steps_per_second": 4.832,
|
512 |
+
"eval_wer": 0.41804817768933256,
|
513 |
"step": 4900
|
514 |
},
|
515 |
{
|
516 |
"epoch": 0.9678668215253581,
|
517 |
+
"grad_norm": 5.257072925567627,
|
518 |
+
"learning_rate": 0.00015808421052631577,
|
519 |
+
"loss": 0.5659,
|
520 |
"step": 5000
|
521 |
},
|
522 |
{
|
523 |
"epoch": 0.9678668215253581,
|
524 |
+
"eval_loss": 0.5333205461502075,
|
525 |
+
"eval_runtime": 145.4188,
|
526 |
+
"eval_samples_per_second": 38.895,
|
527 |
+
"eval_steps_per_second": 4.862,
|
528 |
+
"eval_wer": 0.4176148673588933,
|
529 |
"step": 5000
|
530 |
},
|
531 |
{
|
532 |
"epoch": 0.9872241579558653,
|
533 |
+
"eval_loss": 0.5260410904884338,
|
534 |
+
"eval_runtime": 146.9543,
|
535 |
+
"eval_samples_per_second": 38.488,
|
536 |
+
"eval_steps_per_second": 4.811,
|
537 |
+
"eval_wer": 0.41113126093306157,
|
538 |
"step": 5100
|
539 |
},
|
540 |
{
|
541 |
"epoch": 1.0065814943863725,
|
542 |
+
"eval_loss": 0.5185408592224121,
|
543 |
+
"eval_runtime": 145.8818,
|
544 |
+
"eval_samples_per_second": 38.771,
|
545 |
+
"eval_steps_per_second": 4.846,
|
546 |
+
"eval_wer": 0.3973616215435477,
|
547 |
"step": 5200
|
548 |
},
|
549 |
{
|
550 |
"epoch": 1.0259388308168795,
|
551 |
+
"eval_loss": 0.5147408843040466,
|
552 |
+
"eval_runtime": 150.3669,
|
553 |
+
"eval_samples_per_second": 37.615,
|
554 |
+
"eval_steps_per_second": 4.702,
|
555 |
+
"eval_wer": 0.3917927813708655,
|
556 |
"step": 5300
|
557 |
},
|
558 |
{
|
559 |
"epoch": 1.0452961672473868,
|
560 |
+
"eval_loss": 0.5154542326927185,
|
561 |
+
"eval_runtime": 145.571,
|
562 |
+
"eval_samples_per_second": 38.854,
|
563 |
+
"eval_steps_per_second": 4.857,
|
564 |
+
"eval_wer": 0.3975863009741458,
|
565 |
"step": 5400
|
566 |
},
|
567 |
{
|
568 |
"epoch": 1.064653503677894,
|
569 |
+
"grad_norm": 1.1321543455123901,
|
570 |
+
"learning_rate": 0.00014232631578947366,
|
571 |
+
"loss": 0.4928,
|
572 |
"step": 5500
|
573 |
},
|
574 |
{
|
575 |
"epoch": 1.064653503677894,
|
576 |
+
"eval_loss": 0.5057936906814575,
|
577 |
+
"eval_runtime": 144.9813,
|
578 |
+
"eval_samples_per_second": 39.012,
|
579 |
+
"eval_steps_per_second": 4.876,
|
580 |
+
"eval_wer": 0.3936062653464075,
|
581 |
"step": 5500
|
582 |
},
|
583 |
{
|
584 |
"epoch": 1.084010840108401,
|
585 |
+
"eval_loss": 0.504751443862915,
|
586 |
+
"eval_runtime": 145.1866,
|
587 |
+
"eval_samples_per_second": 38.957,
|
588 |
+
"eval_steps_per_second": 4.87,
|
589 |
+
"eval_wer": 0.3964629038211552,
|
590 |
"step": 5600
|
591 |
},
|
592 |
{
|
593 |
"epoch": 1.1033681765389083,
|
594 |
+
"eval_loss": 0.5011361241340637,
|
595 |
+
"eval_runtime": 146.2522,
|
596 |
+
"eval_samples_per_second": 38.673,
|
597 |
+
"eval_steps_per_second": 4.834,
|
598 |
+
"eval_wer": 0.3818266437707628,
|
599 |
"step": 5700
|
600 |
},
|
601 |
{
|
602 |
"epoch": 1.1227255129694154,
|
603 |
+
"eval_loss": 0.4964805543422699,
|
604 |
+
"eval_runtime": 146.0597,
|
605 |
+
"eval_samples_per_second": 38.724,
|
606 |
+
"eval_steps_per_second": 4.84,
|
607 |
+
"eval_wer": 0.3830463321082955,
|
608 |
"step": 5800
|
609 |
},
|
610 |
{
|
611 |
"epoch": 1.1420828493999227,
|
612 |
+
"eval_loss": 0.4969277083873749,
|
613 |
+
"eval_runtime": 145.1404,
|
614 |
+
"eval_samples_per_second": 38.969,
|
615 |
+
"eval_steps_per_second": 4.871,
|
616 |
+
"eval_wer": 0.383993195422959,
|
617 |
"step": 5900
|
618 |
},
|
619 |
{
|
620 |
"epoch": 1.1614401858304297,
|
621 |
+
"grad_norm": 1.1488393545150757,
|
622 |
+
"learning_rate": 0.00012653684210526316,
|
623 |
+
"loss": 0.4619,
|
624 |
"step": 6000
|
625 |
},
|
626 |
{
|
627 |
"epoch": 1.1614401858304297,
|
628 |
+
"eval_loss": 0.4863020181655884,
|
629 |
+
"eval_runtime": 146.181,
|
630 |
+
"eval_samples_per_second": 38.692,
|
631 |
+
"eval_steps_per_second": 4.836,
|
632 |
+
"eval_wer": 0.3799650142029497,
|
633 |
"step": 6000
|
634 |
},
|
635 |
{
|
636 |
"epoch": 1.1807975222609368,
|
637 |
+
"eval_loss": 0.49075642228126526,
|
638 |
+
"eval_runtime": 145.2768,
|
639 |
+
"eval_samples_per_second": 38.933,
|
640 |
+
"eval_steps_per_second": 4.867,
|
641 |
+
"eval_wer": 0.37998106273370674,
|
642 |
"step": 6100
|
643 |
},
|
644 |
{
|
645 |
"epoch": 1.2001548586914441,
|
646 |
+
"eval_loss": 0.4835449457168579,
|
647 |
+
"eval_runtime": 145.0604,
|
648 |
+
"eval_samples_per_second": 38.991,
|
649 |
+
"eval_steps_per_second": 4.874,
|
650 |
+
"eval_wer": 0.3712185649403797,
|
651 |
"step": 6200
|
652 |
},
|
653 |
{
|
654 |
"epoch": 1.2195121951219512,
|
655 |
+
"eval_loss": 0.4926937520503998,
|
656 |
+
"eval_runtime": 145.5282,
|
657 |
+
"eval_samples_per_second": 38.865,
|
658 |
+
"eval_steps_per_second": 4.858,
|
659 |
+
"eval_wer": 0.3766590168670058,
|
660 |
"step": 6300
|
661 |
},
|
662 |
{
|
663 |
"epoch": 1.2388695315524583,
|
664 |
+
"eval_loss": 0.4942048490047455,
|
665 |
+
"eval_runtime": 146.7525,
|
666 |
+
"eval_samples_per_second": 38.541,
|
667 |
+
"eval_steps_per_second": 4.818,
|
668 |
+
"eval_wer": 0.368281683811847,
|
669 |
"step": 6400
|
670 |
},
|
671 |
{
|
672 |
"epoch": 1.2582268679829656,
|
673 |
+
"grad_norm": 1.8319953680038452,
|
674 |
+
"learning_rate": 0.00011074736842105263,
|
675 |
+
"loss": 0.4421,
|
676 |
"step": 6500
|
677 |
},
|
678 |
{
|
679 |
"epoch": 1.2582268679829656,
|
680 |
+
"eval_loss": 0.4833586513996124,
|
681 |
+
"eval_runtime": 147.4272,
|
682 |
+
"eval_samples_per_second": 38.365,
|
683 |
+
"eval_steps_per_second": 4.796,
|
684 |
+
"eval_wer": 0.37393076663831426,
|
685 |
"step": 6500
|
686 |
},
|
687 |
{
|
688 |
"epoch": 1.2775842044134726,
|
689 |
+
"eval_loss": 0.47513511776924133,
|
690 |
+
"eval_runtime": 146.0617,
|
691 |
+
"eval_samples_per_second": 38.723,
|
692 |
+
"eval_steps_per_second": 4.84,
|
693 |
+
"eval_wer": 0.3633868819309592,
|
694 |
"step": 6600
|
695 |
},
|
696 |
{
|
697 |
"epoch": 1.29694154084398,
|
698 |
+
"eval_loss": 0.4733775854110718,
|
699 |
+
"eval_runtime": 145.526,
|
700 |
+
"eval_samples_per_second": 38.866,
|
701 |
+
"eval_steps_per_second": 4.858,
|
702 |
+
"eval_wer": 0.36327454221566013,
|
703 |
"step": 6700
|
704 |
},
|
705 |
{
|
706 |
"epoch": 1.316298877274487,
|
707 |
+
"eval_loss": 0.4685443937778473,
|
708 |
+
"eval_runtime": 145.3437,
|
709 |
+
"eval_samples_per_second": 38.915,
|
710 |
+
"eval_steps_per_second": 4.864,
|
711 |
+
"eval_wer": 0.36447818202243587,
|
712 |
"step": 6800
|
713 |
},
|
714 |
{
|
715 |
"epoch": 1.3356562137049943,
|
716 |
+
"eval_loss": 0.4654460847377777,
|
717 |
+
"eval_runtime": 145.3646,
|
718 |
+
"eval_samples_per_second": 38.909,
|
719 |
+
"eval_steps_per_second": 4.864,
|
720 |
+
"eval_wer": 0.3624560671470527,
|
721 |
"step": 6900
|
722 |
},
|
723 |
{
|
724 |
"epoch": 1.3550135501355014,
|
725 |
+
"grad_norm": 0.9300881624221802,
|
726 |
+
"learning_rate": 9.49578947368421e-05,
|
727 |
+
"loss": 0.4304,
|
728 |
"step": 7000
|
729 |
},
|
730 |
{
|
731 |
"epoch": 1.3550135501355014,
|
732 |
+
"eval_loss": 0.47420966625213623,
|
733 |
+
"eval_runtime": 145.5058,
|
734 |
+
"eval_samples_per_second": 38.871,
|
735 |
+
"eval_steps_per_second": 4.859,
|
736 |
+
"eval_wer": 0.3615413008939032,
|
737 |
"step": 7000
|
738 |
},
|
739 |
{
|
740 |
"epoch": 1.3743708865660085,
|
741 |
+
"eval_loss": 0.46446511149406433,
|
742 |
+
"eval_runtime": 145.4472,
|
743 |
+
"eval_samples_per_second": 38.887,
|
744 |
+
"eval_steps_per_second": 4.861,
|
745 |
+
"eval_wer": 0.359567331610791,
|
746 |
"step": 7100
|
747 |
},
|
748 |
{
|
749 |
"epoch": 1.3937282229965158,
|
750 |
+
"eval_loss": 0.45991310477256775,
|
751 |
+
"eval_runtime": 145.7318,
|
752 |
+
"eval_samples_per_second": 38.811,
|
753 |
+
"eval_steps_per_second": 4.851,
|
754 |
+
"eval_wer": 0.3593587007109499,
|
755 |
"step": 7200
|
756 |
},
|
757 |
{
|
758 |
"epoch": 1.4130855594270229,
|
759 |
+
"eval_loss": 0.4554171562194824,
|
760 |
+
"eval_runtime": 145.4195,
|
761 |
+
"eval_samples_per_second": 38.894,
|
762 |
+
"eval_steps_per_second": 4.862,
|
763 |
+
"eval_wer": 0.3555391503907817,
|
764 |
"step": 7300
|
765 |
},
|
766 |
{
|
767 |
"epoch": 1.43244289585753,
|
768 |
+
"eval_loss": 0.457757830619812,
|
769 |
+
"eval_runtime": 145.3536,
|
770 |
+
"eval_samples_per_second": 38.912,
|
771 |
+
"eval_steps_per_second": 4.864,
|
772 |
+
"eval_wer": 0.35780199322752004,
|
773 |
"step": 7400
|
774 |
},
|
775 |
{
|
776 |
"epoch": 1.4518002322880372,
|
777 |
+
"grad_norm": 1.012635588645935,
|
778 |
+
"learning_rate": 7.916842105263156e-05,
|
779 |
+
"loss": 0.4275,
|
780 |
"step": 7500
|
781 |
},
|
782 |
{
|
783 |
"epoch": 1.4518002322880372,
|
784 |
+
"eval_loss": 0.45184171199798584,
|
785 |
+
"eval_runtime": 149.8003,
|
786 |
+
"eval_samples_per_second": 37.757,
|
787 |
+
"eval_steps_per_second": 4.72,
|
788 |
+
"eval_wer": 0.3521850074625668,
|
789 |
"step": 7500
|
790 |
},
|
791 |
{
|
792 |
"epoch": 1.4711575687185443,
|
793 |
+
"eval_loss": 0.44799017906188965,
|
794 |
+
"eval_runtime": 145.3878,
|
795 |
+
"eval_samples_per_second": 38.903,
|
796 |
+
"eval_steps_per_second": 4.863,
|
797 |
+
"eval_wer": 0.35107765884033315,
|
798 |
"step": 7600
|
799 |
},
|
800 |
{
|
801 |
"epoch": 1.4905149051490514,
|
802 |
+
"eval_loss": 0.4465474486351013,
|
803 |
+
"eval_runtime": 145.7643,
|
804 |
+
"eval_samples_per_second": 38.802,
|
805 |
+
"eval_steps_per_second": 4.85,
|
806 |
+
"eval_wer": 0.3500986984641556,
|
807 |
"step": 7700
|
808 |
},
|
809 |
{
|
810 |
"epoch": 1.5098722415795587,
|
811 |
+
"eval_loss": 0.44539782404899597,
|
812 |
+
"eval_runtime": 145.5321,
|
813 |
+
"eval_samples_per_second": 38.864,
|
814 |
+
"eval_steps_per_second": 4.858,
|
815 |
+
"eval_wer": 0.34281266550047346,
|
816 |
"step": 7800
|
817 |
},
|
818 |
{
|
819 |
"epoch": 1.5292295780100658,
|
820 |
+
"eval_loss": 0.44271060824394226,
|
821 |
+
"eval_runtime": 145.0706,
|
822 |
+
"eval_samples_per_second": 38.988,
|
823 |
+
"eval_steps_per_second": 4.873,
|
824 |
+
"eval_wer": 0.34387186853043605,
|
825 |
"step": 7900
|
826 |
},
|
827 |
{
|
828 |
"epoch": 1.5485869144405728,
|
829 |
+
"grad_norm": 1.030893087387085,
|
830 |
+
"learning_rate": 6.341052631578946e-05,
|
831 |
+
"loss": 0.4089,
|
832 |
"step": 8000
|
833 |
},
|
834 |
{
|
835 |
"epoch": 1.5485869144405728,
|
836 |
+
"eval_loss": 0.437569797039032,
|
837 |
+
"eval_runtime": 144.8474,
|
838 |
+
"eval_samples_per_second": 39.048,
|
839 |
+
"eval_steps_per_second": 4.881,
|
840 |
+
"eval_wer": 0.3406942594405482,
|
841 |
"step": 8000
|
842 |
},
|
843 |
{
|
844 |
"epoch": 1.5679442508710801,
|
845 |
+
"eval_loss": 0.4396042823791504,
|
846 |
+
"eval_runtime": 145.4051,
|
847 |
+
"eval_samples_per_second": 38.898,
|
848 |
+
"eval_steps_per_second": 4.862,
|
849 |
+
"eval_wer": 0.3415127345091557,
|
850 |
"step": 8100
|
851 |
},
|
852 |
{
|
853 |
"epoch": 1.5873015873015874,
|
854 |
+
"eval_loss": 0.4342670738697052,
|
855 |
+
"eval_runtime": 144.5678,
|
856 |
+
"eval_samples_per_second": 39.124,
|
857 |
+
"eval_steps_per_second": 4.89,
|
858 |
+
"eval_wer": 0.34217072427019307,
|
859 |
"step": 8200
|
860 |
},
|
861 |
{
|
862 |
"epoch": 1.6066589237320945,
|
863 |
+
"eval_loss": 0.4358905851840973,
|
864 |
+
"eval_runtime": 144.9749,
|
865 |
+
"eval_samples_per_second": 39.014,
|
866 |
+
"eval_steps_per_second": 4.877,
|
867 |
+
"eval_wer": 0.3406461138482772,
|
868 |
"step": 8300
|
869 |
},
|
870 |
{
|
871 |
"epoch": 1.6260162601626016,
|
872 |
+
"eval_loss": 0.43577995896339417,
|
873 |
+
"eval_runtime": 144.8242,
|
874 |
+
"eval_samples_per_second": 39.054,
|
875 |
+
"eval_steps_per_second": 4.882,
|
876 |
+
"eval_wer": 0.3373401165123333,
|
877 |
"step": 8400
|
878 |
},
|
879 |
{
|
880 |
"epoch": 1.645373596593109,
|
881 |
+
"grad_norm": 0.9704302549362183,
|
882 |
+
"learning_rate": 4.762105263157894e-05,
|
883 |
+
"loss": 0.4005,
|
884 |
"step": 8500
|
885 |
},
|
886 |
{
|
887 |
"epoch": 1.645373596593109,
|
888 |
+
"eval_loss": 0.4330734312534332,
|
889 |
+
"eval_runtime": 145.4279,
|
890 |
+
"eval_samples_per_second": 38.892,
|
891 |
+
"eval_steps_per_second": 4.862,
|
892 |
+
"eval_wer": 0.3364895443822118,
|
893 |
"step": 8500
|
894 |
},
|
895 |
{
|
896 |
"epoch": 1.664730933023616,
|
897 |
+
"eval_loss": 0.4302414059638977,
|
898 |
+
"eval_runtime": 144.5881,
|
899 |
+
"eval_samples_per_second": 39.118,
|
900 |
+
"eval_steps_per_second": 4.89,
|
901 |
+
"eval_wer": 0.3352859045754361,
|
902 |
"step": 8600
|
903 |
},
|
904 |
{
|
905 |
"epoch": 1.684088269454123,
|
906 |
+
"eval_loss": 0.43084925413131714,
|
907 |
+
"eval_runtime": 145.3335,
|
908 |
+
"eval_samples_per_second": 38.917,
|
909 |
+
"eval_steps_per_second": 4.865,
|
910 |
+
"eval_wer": 0.33549453547527724,
|
911 |
"step": 8700
|
912 |
},
|
913 |
{
|
914 |
"epoch": 1.7034456058846303,
|
915 |
+
"eval_loss": 0.4258446991443634,
|
916 |
+
"eval_runtime": 145.2324,
|
917 |
+
"eval_samples_per_second": 38.944,
|
918 |
+
"eval_steps_per_second": 4.868,
|
919 |
+
"eval_wer": 0.335109370737109,
|
920 |
"step": 8800
|
921 |
},
|
922 |
{
|
923 |
"epoch": 1.7228029423151374,
|
924 |
+
"eval_loss": 0.42223912477493286,
|
925 |
+
"eval_runtime": 145.1268,
|
926 |
+
"eval_samples_per_second": 38.973,
|
927 |
+
"eval_steps_per_second": 4.872,
|
928 |
+
"eval_wer": 0.33530195310619315,
|
929 |
"step": 8900
|
930 |
},
|
931 |
{
|
932 |
"epoch": 1.7421602787456445,
|
933 |
+
"grad_norm": 1.2587796449661255,
|
934 |
+
"learning_rate": 3.186315789473684e-05,
|
935 |
+
"loss": 0.3879,
|
936 |
"step": 9000
|
937 |
},
|
938 |
{
|
939 |
"epoch": 1.7421602787456445,
|
940 |
+
"eval_loss": 0.4238055944442749,
|
941 |
+
"eval_runtime": 145.2053,
|
942 |
+
"eval_samples_per_second": 38.952,
|
943 |
+
"eval_steps_per_second": 4.869,
|
944 |
+
"eval_wer": 0.33119352923239875,
|
945 |
"step": 9000
|
946 |
},
|
947 |
{
|
948 |
"epoch": 1.7615176151761518,
|
949 |
+
"eval_loss": 0.4244863986968994,
|
950 |
+
"eval_runtime": 145.3408,
|
951 |
+
"eval_samples_per_second": 38.915,
|
952 |
+
"eval_steps_per_second": 4.864,
|
953 |
+
"eval_wer": 0.3288022981496044,
|
954 |
"step": 9100
|
955 |
},
|
956 |
{
|
957 |
"epoch": 1.7808749516066589,
|
958 |
+
"eval_loss": 0.42055588960647583,
|
959 |
+
"eval_runtime": 145.738,
|
960 |
+
"eval_samples_per_second": 38.809,
|
961 |
+
"eval_steps_per_second": 4.851,
|
962 |
+
"eval_wer": 0.32641106706681006,
|
963 |
"step": 9200
|
964 |
},
|
965 |
{
|
966 |
"epoch": 1.800232288037166,
|
967 |
+
"eval_loss": 0.4201093018054962,
|
968 |
+
"eval_runtime": 145.1603,
|
969 |
+
"eval_samples_per_second": 38.964,
|
970 |
+
"eval_steps_per_second": 4.87,
|
971 |
+
"eval_wer": 0.3284492304729502,
|
972 |
"step": 9300
|
973 |
},
|
974 |
{
|
975 |
"epoch": 1.8195896244676733,
|
976 |
+
"eval_loss": 0.41892749071121216,
|
977 |
+
"eval_runtime": 150.4269,
|
978 |
+
"eval_samples_per_second": 37.6,
|
979 |
+
"eval_steps_per_second": 4.7,
|
980 |
+
"eval_wer": 0.324597583091268,
|
981 |
"step": 9400
|
982 |
},
|
983 |
{
|
984 |
"epoch": 1.8389469608981805,
|
985 |
+
"grad_norm": 1.2362818717956543,
|
986 |
+
"learning_rate": 1.6073684210526313e-05,
|
987 |
+
"loss": 0.369,
|
988 |
"step": 9500
|
989 |
},
|
990 |
{
|
991 |
"epoch": 1.8389469608981805,
|
992 |
+
"eval_loss": 0.4160093665122986,
|
993 |
+
"eval_runtime": 145.5064,
|
994 |
+
"eval_samples_per_second": 38.871,
|
995 |
+
"eval_steps_per_second": 4.859,
|
996 |
+
"eval_wer": 0.32576912583652967,
|
997 |
"step": 9500
|
998 |
},
|
999 |
{
|
1000 |
"epoch": 1.8583042973286876,
|
1001 |
+
"eval_loss": 0.4141674041748047,
|
1002 |
+
"eval_runtime": 145.8993,
|
1003 |
+
"eval_samples_per_second": 38.766,
|
1004 |
+
"eval_steps_per_second": 4.846,
|
1005 |
+
"eval_wer": 0.3248062139911091,
|
1006 |
"step": 9600
|
1007 |
},
|
1008 |
{
|
1009 |
"epoch": 1.8776616337591947,
|
1010 |
+
"eval_loss": 0.41305750608444214,
|
1011 |
+
"eval_runtime": 146.4678,
|
1012 |
+
"eval_samples_per_second": 38.616,
|
1013 |
+
"eval_steps_per_second": 4.827,
|
1014 |
+
"eval_wer": 0.32517533019852035,
|
1015 |
"step": 9700
|
1016 |
},
|
1017 |
{
|
1018 |
"epoch": 1.897018970189702,
|
1019 |
+
"eval_loss": 0.4127916693687439,
|
1020 |
+
"eval_runtime": 145.2897,
|
1021 |
+
"eval_samples_per_second": 38.929,
|
1022 |
+
"eval_steps_per_second": 4.866,
|
1023 |
+
"eval_wer": 0.32276805058496894,
|
1024 |
"step": 9800
|
1025 |
},
|
1026 |
{
|
1027 |
"epoch": 1.916376306620209,
|
1028 |
+
"eval_loss": 0.41222211718559265,
|
1029 |
+
"eval_runtime": 145.7491,
|
1030 |
+
"eval_samples_per_second": 38.806,
|
1031 |
+
"eval_steps_per_second": 4.851,
|
1032 |
+
"eval_wer": 0.3221421578854456,
|
1033 |
"step": 9900
|
1034 |
},
|
1035 |
{
|
1036 |
"epoch": 1.9357336430507162,
|
1037 |
+
"grad_norm": 1.1011325120925903,
|
1038 |
+
"learning_rate": 2.842105263157894e-07,
|
1039 |
+
"loss": 0.3738,
|
1040 |
"step": 10000
|
1041 |
},
|
1042 |
{
|
1043 |
"epoch": 1.9357336430507162,
|
1044 |
+
"eval_loss": 0.4121682941913605,
|
1045 |
+
"eval_runtime": 145.6809,
|
1046 |
+
"eval_samples_per_second": 38.825,
|
1047 |
+
"eval_steps_per_second": 4.853,
|
1048 |
+
"eval_wer": 0.3222865946622587,
|
1049 |
"step": 10000
|
1050 |
},
|
1051 |
{
|
1052 |
"epoch": 1.9357336430507162,
|
1053 |
"step": 10000,
|
1054 |
"total_flos": 1.1255918428180738e+19,
|
1055 |
+
"train_loss": 0.7340839981079101,
|
1056 |
+
"train_runtime": 18566.2158,
|
1057 |
+
"train_samples_per_second": 4.309,
|
1058 |
+
"train_steps_per_second": 0.539
|
1059 |
}
|
1060 |
],
|
1061 |
"logging_steps": 500,
|