End of training
Browse files- all_results.json +9 -9
- eval_results.json +5 -5
- pytorch_model.bin +1 -1
- runs/Jan28_06-54-16_job-e3562c4a-f6a3-47dc-8131-7e578f1139a3/events.out.tfevents.1643352878.job-e3562c4a-f6a3-47dc-8131-7e578f1139a3.778557.0 +2 -2
- runs/Jan28_06-54-16_job-e3562c4a-f6a3-47dc-8131-7e578f1139a3/events.out.tfevents.1643383627.job-e3562c4a-f6a3-47dc-8131-7e578f1139a3.778557.2 +3 -0
- train_results.json +4 -4
- trainer_state.json +158 -158
all_results.json
CHANGED
@@ -1,14 +1,14 @@
|
|
1 |
{
|
2 |
"epoch": 100.0,
|
3 |
-
"eval_loss":
|
4 |
-
"eval_runtime":
|
5 |
"eval_samples": 1112,
|
6 |
-
"eval_samples_per_second":
|
7 |
-
"eval_steps_per_second": 4.
|
8 |
-
"eval_wer": 0.
|
9 |
-
"train_loss":
|
10 |
-
"train_runtime":
|
11 |
"train_samples": 3336,
|
12 |
-
"train_samples_per_second": 10.
|
13 |
-
"train_steps_per_second": 0.
|
14 |
}
|
|
|
1 |
{
|
2 |
"epoch": 100.0,
|
3 |
+
"eval_loss": 3.384493350982666,
|
4 |
+
"eval_runtime": 65.2378,
|
5 |
"eval_samples": 1112,
|
6 |
+
"eval_samples_per_second": 17.045,
|
7 |
+
"eval_steps_per_second": 4.261,
|
8 |
+
"eval_wer": 0.9869058295964126,
|
9 |
+
"train_loss": 2.424026096050556,
|
10 |
+
"train_runtime": 30676.4279,
|
11 |
"train_samples": 3336,
|
12 |
+
"train_samples_per_second": 10.875,
|
13 |
+
"train_steps_per_second": 0.17
|
14 |
}
|
eval_results.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
"epoch": 100.0,
|
3 |
-
"eval_loss":
|
4 |
-
"eval_runtime":
|
5 |
"eval_samples": 1112,
|
6 |
-
"eval_samples_per_second":
|
7 |
-
"eval_steps_per_second": 4.
|
8 |
-
"eval_wer": 0.
|
9 |
}
|
|
|
1 |
{
|
2 |
"epoch": 100.0,
|
3 |
+
"eval_loss": 3.384493350982666,
|
4 |
+
"eval_runtime": 65.2378,
|
5 |
"eval_samples": 1112,
|
6 |
+
"eval_samples_per_second": 17.045,
|
7 |
+
"eval_steps_per_second": 4.261,
|
8 |
+
"eval_wer": 0.9869058295964126
|
9 |
}
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 3850717489
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:06fe919eab98bc12ffc7b9ac5f1bb776f71d1502f7036df84461f078d7f4357d
|
3 |
size 3850717489
|
runs/Jan28_06-54-16_job-e3562c4a-f6a3-47dc-8131-7e578f1139a3/events.out.tfevents.1643352878.job-e3562c4a-f6a3-47dc-8131-7e578f1139a3.778557.0
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:79086fce91a69cadecf971dbba0baaa955ef5be4371bf2a9647257c72a27483b
|
3 |
+
size 16374
|
runs/Jan28_06-54-16_job-e3562c4a-f6a3-47dc-8131-7e578f1139a3/events.out.tfevents.1643383627.job-e3562c4a-f6a3-47dc-8131-7e578f1139a3.778557.2
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:33e0fa954a0844e51d914b007036c4780bb75a9cc5b86a1de1b2859a91eafb04
|
3 |
+
size 358
|
train_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 100.0,
|
3 |
-
"train_loss":
|
4 |
-
"train_runtime":
|
5 |
"train_samples": 3336,
|
6 |
-
"train_samples_per_second": 10.
|
7 |
-
"train_steps_per_second": 0.
|
8 |
}
|
|
|
1 |
{
|
2 |
"epoch": 100.0,
|
3 |
+
"train_loss": 2.424026096050556,
|
4 |
+
"train_runtime": 30676.4279,
|
5 |
"train_samples": 3336,
|
6 |
+
"train_samples_per_second": 10.875,
|
7 |
+
"train_steps_per_second": 0.17
|
8 |
}
|
trainer_state.json
CHANGED
@@ -9,414 +9,414 @@
|
|
9 |
"log_history": [
|
10 |
{
|
11 |
"epoch": 1.92,
|
12 |
-
"learning_rate":
|
13 |
-
"loss":
|
14 |
"step": 100
|
15 |
},
|
16 |
{
|
17 |
"epoch": 3.84,
|
18 |
-
"learning_rate":
|
19 |
-
"loss":
|
20 |
"step": 200
|
21 |
},
|
22 |
{
|
23 |
"epoch": 5.77,
|
24 |
-
"learning_rate":
|
25 |
-
"loss":
|
26 |
"step": 300
|
27 |
},
|
28 |
{
|
29 |
"epoch": 7.69,
|
30 |
-
"learning_rate":
|
31 |
-
"loss":
|
32 |
"step": 400
|
33 |
},
|
34 |
{
|
35 |
"epoch": 9.61,
|
36 |
-
"learning_rate":
|
37 |
-
"loss":
|
38 |
"step": 500
|
39 |
},
|
40 |
{
|
41 |
"epoch": 9.61,
|
42 |
-
"eval_loss":
|
43 |
-
"eval_runtime":
|
44 |
-
"eval_samples_per_second":
|
45 |
-
"eval_steps_per_second": 4.
|
46 |
-
"eval_wer": 0
|
47 |
"step": 500
|
48 |
},
|
49 |
{
|
50 |
"epoch": 11.54,
|
51 |
-
"learning_rate":
|
52 |
-
"loss":
|
53 |
"step": 600
|
54 |
},
|
55 |
{
|
56 |
"epoch": 13.46,
|
57 |
-
"learning_rate":
|
58 |
-
"loss":
|
59 |
"step": 700
|
60 |
},
|
61 |
{
|
62 |
"epoch": 15.38,
|
63 |
-
"learning_rate":
|
64 |
-
"loss":
|
65 |
"step": 800
|
66 |
},
|
67 |
{
|
68 |
"epoch": 17.31,
|
69 |
-
"learning_rate":
|
70 |
-
"loss":
|
71 |
"step": 900
|
72 |
},
|
73 |
{
|
74 |
"epoch": 19.23,
|
75 |
-
"learning_rate":
|
76 |
-
"loss":
|
77 |
"step": 1000
|
78 |
},
|
79 |
{
|
80 |
"epoch": 19.23,
|
81 |
-
"eval_loss":
|
82 |
-
"eval_runtime":
|
83 |
-
"eval_samples_per_second":
|
84 |
-
"eval_steps_per_second": 4.
|
85 |
-
"eval_wer": 0
|
86 |
"step": 1000
|
87 |
},
|
88 |
{
|
89 |
"epoch": 21.15,
|
90 |
-
"learning_rate":
|
91 |
-
"loss":
|
92 |
"step": 1100
|
93 |
},
|
94 |
{
|
95 |
"epoch": 23.08,
|
96 |
-
"learning_rate":
|
97 |
-
"loss":
|
98 |
"step": 1200
|
99 |
},
|
100 |
{
|
101 |
"epoch": 25.0,
|
102 |
-
"learning_rate":
|
103 |
-
"loss":
|
104 |
"step": 1300
|
105 |
},
|
106 |
{
|
107 |
"epoch": 26.92,
|
108 |
-
"learning_rate":
|
109 |
-
"loss":
|
110 |
"step": 1400
|
111 |
},
|
112 |
{
|
113 |
"epoch": 28.84,
|
114 |
-
"learning_rate":
|
115 |
-
"loss":
|
116 |
"step": 1500
|
117 |
},
|
118 |
{
|
119 |
"epoch": 28.84,
|
120 |
-
"eval_loss":
|
121 |
-
"eval_runtime":
|
122 |
-
"eval_samples_per_second": 17.
|
123 |
-
"eval_steps_per_second": 4.
|
124 |
-
"eval_wer":
|
125 |
"step": 1500
|
126 |
},
|
127 |
{
|
128 |
"epoch": 30.77,
|
129 |
-
"learning_rate":
|
130 |
-
"loss":
|
131 |
"step": 1600
|
132 |
},
|
133 |
{
|
134 |
"epoch": 32.69,
|
135 |
-
"learning_rate":
|
136 |
-
"loss":
|
137 |
"step": 1700
|
138 |
},
|
139 |
{
|
140 |
"epoch": 34.61,
|
141 |
-
"learning_rate":
|
142 |
-
"loss":
|
143 |
"step": 1800
|
144 |
},
|
145 |
{
|
146 |
"epoch": 36.54,
|
147 |
-
"learning_rate":
|
148 |
-
"loss":
|
149 |
"step": 1900
|
150 |
},
|
151 |
{
|
152 |
"epoch": 38.46,
|
153 |
-
"learning_rate":
|
154 |
-
"loss":
|
155 |
"step": 2000
|
156 |
},
|
157 |
{
|
158 |
"epoch": 38.46,
|
159 |
-
"eval_loss":
|
160 |
-
"eval_runtime":
|
161 |
-
"eval_samples_per_second":
|
162 |
-
"eval_steps_per_second": 4.
|
163 |
-
"eval_wer": 0.
|
164 |
"step": 2000
|
165 |
},
|
166 |
{
|
167 |
"epoch": 40.38,
|
168 |
-
"learning_rate":
|
169 |
-
"loss":
|
170 |
"step": 2100
|
171 |
},
|
172 |
{
|
173 |
"epoch": 42.31,
|
174 |
-
"learning_rate":
|
175 |
-
"loss":
|
176 |
"step": 2200
|
177 |
},
|
178 |
{
|
179 |
"epoch": 44.23,
|
180 |
-
"learning_rate":
|
181 |
-
"loss":
|
182 |
"step": 2300
|
183 |
},
|
184 |
{
|
185 |
"epoch": 46.15,
|
186 |
-
"learning_rate":
|
187 |
-
"loss":
|
188 |
"step": 2400
|
189 |
},
|
190 |
{
|
191 |
"epoch": 48.08,
|
192 |
-
"learning_rate":
|
193 |
-
"loss":
|
194 |
"step": 2500
|
195 |
},
|
196 |
{
|
197 |
"epoch": 48.08,
|
198 |
-
"eval_loss":
|
199 |
-
"eval_runtime":
|
200 |
-
"eval_samples_per_second": 17.
|
201 |
-
"eval_steps_per_second": 4.
|
202 |
-
"eval_wer": 0.
|
203 |
"step": 2500
|
204 |
},
|
205 |
{
|
206 |
"epoch": 50.0,
|
207 |
-
"learning_rate":
|
208 |
-
"loss":
|
209 |
"step": 2600
|
210 |
},
|
211 |
{
|
212 |
"epoch": 51.92,
|
213 |
-
"learning_rate":
|
214 |
-
"loss":
|
215 |
"step": 2700
|
216 |
},
|
217 |
{
|
218 |
"epoch": 53.84,
|
219 |
-
"learning_rate":
|
220 |
-
"loss":
|
221 |
"step": 2800
|
222 |
},
|
223 |
{
|
224 |
"epoch": 55.77,
|
225 |
-
"learning_rate":
|
226 |
-
"loss":
|
227 |
"step": 2900
|
228 |
},
|
229 |
{
|
230 |
"epoch": 57.69,
|
231 |
-
"learning_rate":
|
232 |
-
"loss":
|
233 |
"step": 3000
|
234 |
},
|
235 |
{
|
236 |
"epoch": 57.69,
|
237 |
-
"eval_loss":
|
238 |
-
"eval_runtime":
|
239 |
-
"eval_samples_per_second": 17.
|
240 |
-
"eval_steps_per_second": 4.
|
241 |
-
"eval_wer": 0.
|
242 |
"step": 3000
|
243 |
},
|
244 |
{
|
245 |
"epoch": 59.61,
|
246 |
-
"learning_rate":
|
247 |
-
"loss":
|
248 |
"step": 3100
|
249 |
},
|
250 |
{
|
251 |
"epoch": 61.54,
|
252 |
-
"learning_rate":
|
253 |
-
"loss":
|
254 |
"step": 3200
|
255 |
},
|
256 |
{
|
257 |
"epoch": 63.46,
|
258 |
-
"learning_rate":
|
259 |
-
"loss":
|
260 |
"step": 3300
|
261 |
},
|
262 |
{
|
263 |
"epoch": 65.38,
|
264 |
-
"learning_rate":
|
265 |
-
"loss":
|
266 |
"step": 3400
|
267 |
},
|
268 |
{
|
269 |
"epoch": 67.31,
|
270 |
-
"learning_rate":
|
271 |
-
"loss":
|
272 |
"step": 3500
|
273 |
},
|
274 |
{
|
275 |
"epoch": 67.31,
|
276 |
-
"eval_loss":
|
277 |
-
"eval_runtime":
|
278 |
-
"eval_samples_per_second":
|
279 |
-
"eval_steps_per_second":
|
280 |
-
"eval_wer": 0.
|
281 |
"step": 3500
|
282 |
},
|
283 |
{
|
284 |
"epoch": 69.23,
|
285 |
-
"learning_rate":
|
286 |
-
"loss":
|
287 |
"step": 3600
|
288 |
},
|
289 |
{
|
290 |
"epoch": 71.15,
|
291 |
-
"learning_rate":
|
292 |
-
"loss":
|
293 |
"step": 3700
|
294 |
},
|
295 |
{
|
296 |
"epoch": 73.08,
|
297 |
-
"learning_rate":
|
298 |
-
"loss":
|
299 |
"step": 3800
|
300 |
},
|
301 |
{
|
302 |
"epoch": 75.0,
|
303 |
-
"learning_rate":
|
304 |
-
"loss":
|
305 |
"step": 3900
|
306 |
},
|
307 |
{
|
308 |
"epoch": 76.92,
|
309 |
-
"learning_rate":
|
310 |
-
"loss":
|
311 |
"step": 4000
|
312 |
},
|
313 |
{
|
314 |
"epoch": 76.92,
|
315 |
-
"eval_loss":
|
316 |
-
"eval_runtime":
|
317 |
-
"eval_samples_per_second": 17.
|
318 |
-
"eval_steps_per_second": 4.
|
319 |
-
"eval_wer": 0.
|
320 |
"step": 4000
|
321 |
},
|
322 |
{
|
323 |
"epoch": 78.84,
|
324 |
-
"learning_rate":
|
325 |
-
"loss": 0.
|
326 |
"step": 4100
|
327 |
},
|
328 |
{
|
329 |
"epoch": 80.77,
|
330 |
-
"learning_rate":
|
331 |
-
"loss": 0.
|
332 |
"step": 4200
|
333 |
},
|
334 |
{
|
335 |
"epoch": 82.69,
|
336 |
-
"learning_rate":
|
337 |
-
"loss": 0.
|
338 |
"step": 4300
|
339 |
},
|
340 |
{
|
341 |
"epoch": 84.61,
|
342 |
-
"learning_rate":
|
343 |
-
"loss": 0.
|
344 |
"step": 4400
|
345 |
},
|
346 |
{
|
347 |
"epoch": 86.54,
|
348 |
-
"learning_rate":
|
349 |
-
"loss": 0.
|
350 |
"step": 4500
|
351 |
},
|
352 |
{
|
353 |
"epoch": 86.54,
|
354 |
-
"eval_loss":
|
355 |
-
"eval_runtime":
|
356 |
-
"eval_samples_per_second": 17.
|
357 |
-
"eval_steps_per_second": 4.
|
358 |
-
"eval_wer": 0.
|
359 |
"step": 4500
|
360 |
},
|
361 |
{
|
362 |
"epoch": 88.46,
|
363 |
-
"learning_rate":
|
364 |
-
"loss": 0.
|
365 |
"step": 4600
|
366 |
},
|
367 |
{
|
368 |
"epoch": 90.38,
|
369 |
-
"learning_rate":
|
370 |
-
"loss": 0.
|
371 |
"step": 4700
|
372 |
},
|
373 |
{
|
374 |
"epoch": 92.31,
|
375 |
-
"learning_rate":
|
376 |
-
"loss": 0.
|
377 |
"step": 4800
|
378 |
},
|
379 |
{
|
380 |
"epoch": 94.23,
|
381 |
-
"learning_rate":
|
382 |
-
"loss": 0.
|
383 |
"step": 4900
|
384 |
},
|
385 |
{
|
386 |
"epoch": 96.15,
|
387 |
-
"learning_rate":
|
388 |
-
"loss": 0.
|
389 |
"step": 5000
|
390 |
},
|
391 |
{
|
392 |
"epoch": 96.15,
|
393 |
-
"eval_loss":
|
394 |
-
"eval_runtime":
|
395 |
-
"eval_samples_per_second":
|
396 |
-
"eval_steps_per_second": 4.
|
397 |
-
"eval_wer": 0.
|
398 |
"step": 5000
|
399 |
},
|
400 |
{
|
401 |
"epoch": 98.08,
|
402 |
-
"learning_rate":
|
403 |
-
"loss": 0.
|
404 |
"step": 5100
|
405 |
},
|
406 |
{
|
407 |
"epoch": 100.0,
|
408 |
-
"learning_rate":
|
409 |
-
"loss": 0.
|
410 |
"step": 5200
|
411 |
},
|
412 |
{
|
413 |
"epoch": 100.0,
|
414 |
"step": 5200,
|
415 |
"total_flos": 1.4653197993726655e+20,
|
416 |
-
"train_loss":
|
417 |
-
"train_runtime":
|
418 |
-
"train_samples_per_second": 10.
|
419 |
-
"train_steps_per_second": 0.
|
420 |
}
|
421 |
],
|
422 |
"max_steps": 5200,
|
|
|
9 |
"log_history": [
|
10 |
{
|
11 |
"epoch": 1.92,
|
12 |
+
"learning_rate": 1.6166666666666665e-07,
|
13 |
+
"loss": 24.0656,
|
14 |
"step": 100
|
15 |
},
|
16 |
{
|
17 |
"epoch": 3.84,
|
18 |
+
"learning_rate": 3.283333333333333e-07,
|
19 |
+
"loss": 19.0268,
|
20 |
"step": 200
|
21 |
},
|
22 |
{
|
23 |
"epoch": 5.77,
|
24 |
+
"learning_rate": 4.95e-07,
|
25 |
+
"loss": 6.1995,
|
26 |
"step": 300
|
27 |
},
|
28 |
{
|
29 |
"epoch": 7.69,
|
30 |
+
"learning_rate": 4.901020408163265e-07,
|
31 |
+
"loss": 3.6484,
|
32 |
"step": 400
|
33 |
},
|
34 |
{
|
35 |
"epoch": 9.61,
|
36 |
+
"learning_rate": 4.798979591836734e-07,
|
37 |
+
"loss": 3.5131,
|
38 |
"step": 500
|
39 |
},
|
40 |
{
|
41 |
"epoch": 9.61,
|
42 |
+
"eval_loss": 3.5294487476348877,
|
43 |
+
"eval_runtime": 64.5425,
|
44 |
+
"eval_samples_per_second": 17.229,
|
45 |
+
"eval_steps_per_second": 4.307,
|
46 |
+
"eval_wer": 1.0,
|
47 |
"step": 500
|
48 |
},
|
49 |
{
|
50 |
"epoch": 11.54,
|
51 |
+
"learning_rate": 4.696938775510204e-07,
|
52 |
+
"loss": 3.4699,
|
53 |
"step": 600
|
54 |
},
|
55 |
{
|
56 |
"epoch": 13.46,
|
57 |
+
"learning_rate": 4.5948979591836735e-07,
|
58 |
+
"loss": 3.3973,
|
59 |
"step": 700
|
60 |
},
|
61 |
{
|
62 |
"epoch": 15.38,
|
63 |
+
"learning_rate": 4.4928571428571426e-07,
|
64 |
+
"loss": 3.2876,
|
65 |
"step": 800
|
66 |
},
|
67 |
{
|
68 |
"epoch": 17.31,
|
69 |
+
"learning_rate": 4.390816326530612e-07,
|
70 |
+
"loss": 3.0902,
|
71 |
"step": 900
|
72 |
},
|
73 |
{
|
74 |
"epoch": 19.23,
|
75 |
+
"learning_rate": 4.288775510204081e-07,
|
76 |
+
"loss": 2.8596,
|
77 |
"step": 1000
|
78 |
},
|
79 |
{
|
80 |
"epoch": 19.23,
|
81 |
+
"eval_loss": 3.5708465576171875,
|
82 |
+
"eval_runtime": 66.2433,
|
83 |
+
"eval_samples_per_second": 16.787,
|
84 |
+
"eval_steps_per_second": 4.197,
|
85 |
+
"eval_wer": 1.0,
|
86 |
"step": 1000
|
87 |
},
|
88 |
{
|
89 |
"epoch": 21.15,
|
90 |
+
"learning_rate": 4.1867346938775513e-07,
|
91 |
+
"loss": 2.6397,
|
92 |
"step": 1100
|
93 |
},
|
94 |
{
|
95 |
"epoch": 23.08,
|
96 |
+
"learning_rate": 4.0846938775510203e-07,
|
97 |
+
"loss": 2.4302,
|
98 |
"step": 1200
|
99 |
},
|
100 |
{
|
101 |
"epoch": 25.0,
|
102 |
+
"learning_rate": 3.98265306122449e-07,
|
103 |
+
"loss": 2.2341,
|
104 |
"step": 1300
|
105 |
},
|
106 |
{
|
107 |
"epoch": 26.92,
|
108 |
+
"learning_rate": 3.880612244897959e-07,
|
109 |
+
"loss": 2.071,
|
110 |
"step": 1400
|
111 |
},
|
112 |
{
|
113 |
"epoch": 28.84,
|
114 |
+
"learning_rate": 3.778571428571428e-07,
|
115 |
+
"loss": 1.9055,
|
116 |
"step": 1500
|
117 |
},
|
118 |
{
|
119 |
"epoch": 28.84,
|
120 |
+
"eval_loss": 3.6432793140411377,
|
121 |
+
"eval_runtime": 65.1639,
|
122 |
+
"eval_samples_per_second": 17.065,
|
123 |
+
"eval_steps_per_second": 4.266,
|
124 |
+
"eval_wer": 1.0007174887892376,
|
125 |
"step": 1500
|
126 |
},
|
127 |
{
|
128 |
"epoch": 30.77,
|
129 |
+
"learning_rate": 3.676530612244898e-07,
|
130 |
+
"loss": 1.7787,
|
131 |
"step": 1600
|
132 |
},
|
133 |
{
|
134 |
"epoch": 32.69,
|
135 |
+
"learning_rate": 3.574489795918367e-07,
|
136 |
+
"loss": 1.6655,
|
137 |
"step": 1700
|
138 |
},
|
139 |
{
|
140 |
"epoch": 34.61,
|
141 |
+
"learning_rate": 3.4724489795918366e-07,
|
142 |
+
"loss": 1.5696,
|
143 |
"step": 1800
|
144 |
},
|
145 |
{
|
146 |
"epoch": 36.54,
|
147 |
+
"learning_rate": 3.3704081632653057e-07,
|
148 |
+
"loss": 1.4902,
|
149 |
"step": 1900
|
150 |
},
|
151 |
{
|
152 |
"epoch": 38.46,
|
153 |
+
"learning_rate": 3.268367346938775e-07,
|
154 |
+
"loss": 1.4239,
|
155 |
"step": 2000
|
156 |
},
|
157 |
{
|
158 |
"epoch": 38.46,
|
159 |
+
"eval_loss": 3.6568963527679443,
|
160 |
+
"eval_runtime": 66.0028,
|
161 |
+
"eval_samples_per_second": 16.848,
|
162 |
+
"eval_steps_per_second": 4.212,
|
163 |
+
"eval_wer": 0.9994618834080717,
|
164 |
"step": 2000
|
165 |
},
|
166 |
{
|
167 |
"epoch": 40.38,
|
168 |
+
"learning_rate": 3.166326530612245e-07,
|
169 |
+
"loss": 1.3735,
|
170 |
"step": 2100
|
171 |
},
|
172 |
{
|
173 |
"epoch": 42.31,
|
174 |
+
"learning_rate": 3.0642857142857144e-07,
|
175 |
+
"loss": 1.3228,
|
176 |
"step": 2200
|
177 |
},
|
178 |
{
|
179 |
"epoch": 44.23,
|
180 |
+
"learning_rate": 2.9622448979591834e-07,
|
181 |
+
"loss": 1.2834,
|
182 |
"step": 2300
|
183 |
},
|
184 |
{
|
185 |
"epoch": 46.15,
|
186 |
+
"learning_rate": 2.860204081632653e-07,
|
187 |
+
"loss": 1.2438,
|
188 |
"step": 2400
|
189 |
},
|
190 |
{
|
191 |
"epoch": 48.08,
|
192 |
+
"learning_rate": 2.758163265306122e-07,
|
193 |
+
"loss": 1.2168,
|
194 |
"step": 2500
|
195 |
},
|
196 |
{
|
197 |
"epoch": 48.08,
|
198 |
+
"eval_loss": 3.6079351902008057,
|
199 |
+
"eval_runtime": 65.3166,
|
200 |
+
"eval_samples_per_second": 17.025,
|
201 |
+
"eval_steps_per_second": 4.256,
|
202 |
+
"eval_wer": 0.995695067264574,
|
203 |
"step": 2500
|
204 |
},
|
205 |
{
|
206 |
"epoch": 50.0,
|
207 |
+
"learning_rate": 2.656122448979592e-07,
|
208 |
+
"loss": 1.1792,
|
209 |
"step": 2600
|
210 |
},
|
211 |
{
|
212 |
"epoch": 51.92,
|
213 |
+
"learning_rate": 2.554081632653061e-07,
|
214 |
+
"loss": 1.1706,
|
215 |
"step": 2700
|
216 |
},
|
217 |
{
|
218 |
"epoch": 53.84,
|
219 |
+
"learning_rate": 2.4520408163265307e-07,
|
220 |
+
"loss": 1.1429,
|
221 |
"step": 2800
|
222 |
},
|
223 |
{
|
224 |
"epoch": 55.77,
|
225 |
+
"learning_rate": 2.3499999999999997e-07,
|
226 |
+
"loss": 1.1318,
|
227 |
"step": 2900
|
228 |
},
|
229 |
{
|
230 |
"epoch": 57.69,
|
231 |
+
"learning_rate": 2.2479591836734693e-07,
|
232 |
+
"loss": 1.1063,
|
233 |
"step": 3000
|
234 |
},
|
235 |
{
|
236 |
"epoch": 57.69,
|
237 |
+
"eval_loss": 3.5737504959106445,
|
238 |
+
"eval_runtime": 64.9336,
|
239 |
+
"eval_samples_per_second": 17.125,
|
240 |
+
"eval_steps_per_second": 4.281,
|
241 |
+
"eval_wer": 0.9924663677130044,
|
242 |
"step": 3000
|
243 |
},
|
244 |
{
|
245 |
"epoch": 59.61,
|
246 |
+
"learning_rate": 2.1459183673469386e-07,
|
247 |
+
"loss": 1.0893,
|
248 |
"step": 3100
|
249 |
},
|
250 |
{
|
251 |
"epoch": 61.54,
|
252 |
+
"learning_rate": 2.0438775510204082e-07,
|
253 |
+
"loss": 1.0744,
|
254 |
"step": 3200
|
255 |
},
|
256 |
{
|
257 |
"epoch": 63.46,
|
258 |
+
"learning_rate": 1.9418367346938775e-07,
|
259 |
+
"loss": 1.0591,
|
260 |
"step": 3300
|
261 |
},
|
262 |
{
|
263 |
"epoch": 65.38,
|
264 |
+
"learning_rate": 1.839795918367347e-07,
|
265 |
+
"loss": 1.05,
|
266 |
"step": 3400
|
267 |
},
|
268 |
{
|
269 |
"epoch": 67.31,
|
270 |
+
"learning_rate": 1.7377551020408163e-07,
|
271 |
+
"loss": 1.0404,
|
272 |
"step": 3500
|
273 |
},
|
274 |
{
|
275 |
"epoch": 67.31,
|
276 |
+
"eval_loss": 3.4857470989227295,
|
277 |
+
"eval_runtime": 64.9602,
|
278 |
+
"eval_samples_per_second": 17.118,
|
279 |
+
"eval_steps_per_second": 4.28,
|
280 |
+
"eval_wer": 0.9888789237668162,
|
281 |
"step": 3500
|
282 |
},
|
283 |
{
|
284 |
"epoch": 69.23,
|
285 |
+
"learning_rate": 1.6357142857142856e-07,
|
286 |
+
"loss": 1.0325,
|
287 |
"step": 3600
|
288 |
},
|
289 |
{
|
290 |
"epoch": 71.15,
|
291 |
+
"learning_rate": 1.5336734693877552e-07,
|
292 |
+
"loss": 1.0266,
|
293 |
"step": 3700
|
294 |
},
|
295 |
{
|
296 |
"epoch": 73.08,
|
297 |
+
"learning_rate": 1.4316326530612245e-07,
|
298 |
+
"loss": 1.0164,
|
299 |
"step": 3800
|
300 |
},
|
301 |
{
|
302 |
"epoch": 75.0,
|
303 |
+
"learning_rate": 1.329591836734694e-07,
|
304 |
+
"loss": 1.0002,
|
305 |
"step": 3900
|
306 |
},
|
307 |
{
|
308 |
"epoch": 76.92,
|
309 |
+
"learning_rate": 1.227551020408163e-07,
|
310 |
+
"loss": 1.001,
|
311 |
"step": 4000
|
312 |
},
|
313 |
{
|
314 |
"epoch": 76.92,
|
315 |
+
"eval_loss": 3.4881510734558105,
|
316 |
+
"eval_runtime": 64.7811,
|
317 |
+
"eval_samples_per_second": 17.165,
|
318 |
+
"eval_steps_per_second": 4.291,
|
319 |
+
"eval_wer": 0.9858295964125561,
|
320 |
"step": 4000
|
321 |
},
|
322 |
{
|
323 |
"epoch": 78.84,
|
324 |
+
"learning_rate": 1.1255102040816327e-07,
|
325 |
+
"loss": 0.997,
|
326 |
"step": 4100
|
327 |
},
|
328 |
{
|
329 |
"epoch": 80.77,
|
330 |
+
"learning_rate": 1.0234693877551021e-07,
|
331 |
+
"loss": 0.9896,
|
332 |
"step": 4200
|
333 |
},
|
334 |
{
|
335 |
"epoch": 82.69,
|
336 |
+
"learning_rate": 9.214285714285714e-08,
|
337 |
+
"loss": 0.986,
|
338 |
"step": 4300
|
339 |
},
|
340 |
{
|
341 |
"epoch": 84.61,
|
342 |
+
"learning_rate": 8.193877551020407e-08,
|
343 |
+
"loss": 0.9823,
|
344 |
"step": 4400
|
345 |
},
|
346 |
{
|
347 |
"epoch": 86.54,
|
348 |
+
"learning_rate": 7.173469387755101e-08,
|
349 |
+
"loss": 0.982,
|
350 |
"step": 4500
|
351 |
},
|
352 |
{
|
353 |
"epoch": 86.54,
|
354 |
+
"eval_loss": 3.3850555419921875,
|
355 |
+
"eval_runtime": 65.2436,
|
356 |
+
"eval_samples_per_second": 17.044,
|
357 |
+
"eval_steps_per_second": 4.261,
|
358 |
+
"eval_wer": 0.987085201793722,
|
359 |
"step": 4500
|
360 |
},
|
361 |
{
|
362 |
"epoch": 88.46,
|
363 |
+
"learning_rate": 6.153061224489796e-08,
|
364 |
+
"loss": 0.9675,
|
365 |
"step": 4600
|
366 |
},
|
367 |
{
|
368 |
"epoch": 90.38,
|
369 |
+
"learning_rate": 5.132653061224489e-08,
|
370 |
+
"loss": 0.9774,
|
371 |
"step": 4700
|
372 |
},
|
373 |
{
|
374 |
"epoch": 92.31,
|
375 |
+
"learning_rate": 4.1122448979591836e-08,
|
376 |
+
"loss": 0.9687,
|
377 |
"step": 4800
|
378 |
},
|
379 |
{
|
380 |
"epoch": 94.23,
|
381 |
+
"learning_rate": 3.091836734693877e-08,
|
382 |
+
"loss": 0.9771,
|
383 |
"step": 4900
|
384 |
},
|
385 |
{
|
386 |
"epoch": 96.15,
|
387 |
+
"learning_rate": 2.0714285714285713e-08,
|
388 |
+
"loss": 0.9612,
|
389 |
"step": 5000
|
390 |
},
|
391 |
{
|
392 |
"epoch": 96.15,
|
393 |
+
"eval_loss": 3.386908531188965,
|
394 |
+
"eval_runtime": 65.5128,
|
395 |
+
"eval_samples_per_second": 16.974,
|
396 |
+
"eval_steps_per_second": 4.243,
|
397 |
+
"eval_wer": 0.9872645739910314,
|
398 |
"step": 5000
|
399 |
},
|
400 |
{
|
401 |
"epoch": 98.08,
|
402 |
+
"learning_rate": 1.0510204081632651e-08,
|
403 |
+
"loss": 0.9667,
|
404 |
"step": 5100
|
405 |
},
|
406 |
{
|
407 |
"epoch": 100.0,
|
408 |
+
"learning_rate": 3.061224489795918e-10,
|
409 |
+
"loss": 0.9664,
|
410 |
"step": 5200
|
411 |
},
|
412 |
{
|
413 |
"epoch": 100.0,
|
414 |
"step": 5200,
|
415 |
"total_flos": 1.4653197993726655e+20,
|
416 |
+
"train_loss": 2.424026096050556,
|
417 |
+
"train_runtime": 30676.4279,
|
418 |
+
"train_samples_per_second": 10.875,
|
419 |
+
"train_steps_per_second": 0.17
|
420 |
}
|
421 |
],
|
422 |
"max_steps": 5200,
|