End of training
Browse files
all_results.json
ADDED
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 10.0,
|
3 |
+
"eval_exact_match": 66.4375,
|
4 |
+
"eval_f1": 80.63612438300382,
|
5 |
+
"eval_runtime": 58.1435,
|
6 |
+
"eval_samples": 5111,
|
7 |
+
"eval_samples_per_second": 87.903,
|
8 |
+
"eval_steps_per_second": 3.147,
|
9 |
+
"total_flos": 4.451320899376128e+16,
|
10 |
+
"train_loss": 0.6037781062384544,
|
11 |
+
"train_runtime": 5220.6695,
|
12 |
+
"train_samples": 22714,
|
13 |
+
"train_samples_per_second": 43.508,
|
14 |
+
"train_steps_per_second": 1.555
|
15 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 10.0,
|
3 |
+
"eval_exact_match": 66.4375,
|
4 |
+
"eval_f1": 80.63612438300382,
|
5 |
+
"eval_runtime": 58.1435,
|
6 |
+
"eval_samples": 5111,
|
7 |
+
"eval_samples_per_second": 87.903,
|
8 |
+
"eval_steps_per_second": 3.147
|
9 |
+
}
|
runs/May11_10-13-08_ml-machine/events.out.tfevents.1715427773.ml-machine.260784.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b27821dc6c3d9d0d4a7d89c21ee1b7e8c31720a6f8f7d8b7a69e759f8f21c189
|
3 |
+
size 412
|
train_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 10.0,
|
3 |
+
"total_flos": 4.451320899376128e+16,
|
4 |
+
"train_loss": 0.6037781062384544,
|
5 |
+
"train_runtime": 5220.6695,
|
6 |
+
"train_samples": 22714,
|
7 |
+
"train_samples_per_second": 43.508,
|
8 |
+
"train_steps_per_second": 1.555
|
9 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,280 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 80.63612438300382,
|
3 |
+
"best_model_checkpoint": "/root/turkic_qa/tr_uzn_models/tr_uzn_xlm_roberta_base_squad_model/checkpoint-8120",
|
4 |
+
"epoch": 10.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 8120,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 1.0,
|
13 |
+
"step": 812,
|
14 |
+
"train_exact_match": 56.743256743256744,
|
15 |
+
"train_f1": 75.16234224627293,
|
16 |
+
"train_runtime": 18.93,
|
17 |
+
"train_samples_per_second": 87.692,
|
18 |
+
"train_steps_per_second": 3.17
|
19 |
+
},
|
20 |
+
{
|
21 |
+
"epoch": 1.0,
|
22 |
+
"grad_norm": 33.5258674621582,
|
23 |
+
"learning_rate": 5e-06,
|
24 |
+
"loss": 1.3302,
|
25 |
+
"step": 812
|
26 |
+
},
|
27 |
+
{
|
28 |
+
"epoch": 1.0,
|
29 |
+
"eval_exact_match": 56.78125,
|
30 |
+
"eval_f1": 74.92855180045365,
|
31 |
+
"eval_runtime": 58.0842,
|
32 |
+
"eval_samples_per_second": 87.993,
|
33 |
+
"eval_steps_per_second": 3.151,
|
34 |
+
"step": 812
|
35 |
+
},
|
36 |
+
{
|
37 |
+
"epoch": 2.0,
|
38 |
+
"step": 1624,
|
39 |
+
"train_exact_match": 66.53346653346654,
|
40 |
+
"train_f1": 81.01468766554503,
|
41 |
+
"train_runtime": 19.1594,
|
42 |
+
"train_samples_per_second": 84.919,
|
43 |
+
"train_steps_per_second": 3.079
|
44 |
+
},
|
45 |
+
{
|
46 |
+
"epoch": 2.0,
|
47 |
+
"grad_norm": 39.375850677490234,
|
48 |
+
"learning_rate": 1e-05,
|
49 |
+
"loss": 0.9867,
|
50 |
+
"step": 1624
|
51 |
+
},
|
52 |
+
{
|
53 |
+
"epoch": 2.0,
|
54 |
+
"eval_exact_match": 60.71875,
|
55 |
+
"eval_f1": 77.48042477059983,
|
56 |
+
"eval_runtime": 59.4097,
|
57 |
+
"eval_samples_per_second": 86.03,
|
58 |
+
"eval_steps_per_second": 3.08,
|
59 |
+
"step": 1624
|
60 |
+
},
|
61 |
+
{
|
62 |
+
"epoch": 3.0,
|
63 |
+
"step": 2436,
|
64 |
+
"train_exact_match": 73.32667332667333,
|
65 |
+
"train_f1": 86.12936978278546,
|
66 |
+
"train_runtime": 18.141,
|
67 |
+
"train_samples_per_second": 87.04,
|
68 |
+
"train_steps_per_second": 3.142
|
69 |
+
},
|
70 |
+
{
|
71 |
+
"epoch": 3.0,
|
72 |
+
"grad_norm": 43.98362350463867,
|
73 |
+
"learning_rate": 8.750000000000001e-06,
|
74 |
+
"loss": 0.7954,
|
75 |
+
"step": 2436
|
76 |
+
},
|
77 |
+
{
|
78 |
+
"epoch": 3.0,
|
79 |
+
"eval_exact_match": 63.53125,
|
80 |
+
"eval_f1": 79.36767777349903,
|
81 |
+
"eval_runtime": 58.582,
|
82 |
+
"eval_samples_per_second": 87.245,
|
83 |
+
"eval_steps_per_second": 3.124,
|
84 |
+
"step": 2436
|
85 |
+
},
|
86 |
+
{
|
87 |
+
"epoch": 4.0,
|
88 |
+
"step": 3248,
|
89 |
+
"train_exact_match": 74.72527472527473,
|
90 |
+
"train_f1": 88.40189950318806,
|
91 |
+
"train_runtime": 18.0666,
|
92 |
+
"train_samples_per_second": 88.174,
|
93 |
+
"train_steps_per_second": 3.155
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 4.0,
|
97 |
+
"grad_norm": 28.4106502532959,
|
98 |
+
"learning_rate": 7.500000000000001e-06,
|
99 |
+
"loss": 0.6427,
|
100 |
+
"step": 3248
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 4.0,
|
104 |
+
"eval_exact_match": 65.21875,
|
105 |
+
"eval_f1": 80.19939907948,
|
106 |
+
"eval_runtime": 57.9356,
|
107 |
+
"eval_samples_per_second": 88.219,
|
108 |
+
"eval_steps_per_second": 3.159,
|
109 |
+
"step": 3248
|
110 |
+
},
|
111 |
+
{
|
112 |
+
"epoch": 5.0,
|
113 |
+
"step": 4060,
|
114 |
+
"train_exact_match": 79.72027972027972,
|
115 |
+
"train_f1": 90.39293523910672,
|
116 |
+
"train_runtime": 18.0742,
|
117 |
+
"train_samples_per_second": 88.303,
|
118 |
+
"train_steps_per_second": 3.154
|
119 |
+
},
|
120 |
+
{
|
121 |
+
"epoch": 5.0,
|
122 |
+
"grad_norm": 58.56404495239258,
|
123 |
+
"learning_rate": 6.25e-06,
|
124 |
+
"loss": 0.5287,
|
125 |
+
"step": 4060
|
126 |
+
},
|
127 |
+
{
|
128 |
+
"epoch": 5.0,
|
129 |
+
"eval_exact_match": 64.53125,
|
130 |
+
"eval_f1": 79.76982128782555,
|
131 |
+
"eval_runtime": 57.84,
|
132 |
+
"eval_samples_per_second": 88.365,
|
133 |
+
"eval_steps_per_second": 3.164,
|
134 |
+
"step": 4060
|
135 |
+
},
|
136 |
+
{
|
137 |
+
"epoch": 6.0,
|
138 |
+
"step": 4872,
|
139 |
+
"train_exact_match": 79.42057942057941,
|
140 |
+
"train_f1": 91.11407011087398,
|
141 |
+
"train_runtime": 18.4811,
|
142 |
+
"train_samples_per_second": 88.252,
|
143 |
+
"train_steps_per_second": 3.192
|
144 |
+
},
|
145 |
+
{
|
146 |
+
"epoch": 6.0,
|
147 |
+
"grad_norm": 45.58940124511719,
|
148 |
+
"learning_rate": 5e-06,
|
149 |
+
"loss": 0.4506,
|
150 |
+
"step": 4872
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 6.0,
|
154 |
+
"eval_exact_match": 65.46875,
|
155 |
+
"eval_f1": 80.18984733785271,
|
156 |
+
"eval_runtime": 58.0085,
|
157 |
+
"eval_samples_per_second": 88.108,
|
158 |
+
"eval_steps_per_second": 3.155,
|
159 |
+
"step": 4872
|
160 |
+
},
|
161 |
+
{
|
162 |
+
"epoch": 7.0,
|
163 |
+
"step": 5684,
|
164 |
+
"train_exact_match": 80.41958041958041,
|
165 |
+
"train_f1": 91.05656126706158,
|
166 |
+
"train_runtime": 17.8755,
|
167 |
+
"train_samples_per_second": 89.172,
|
168 |
+
"train_steps_per_second": 3.189
|
169 |
+
},
|
170 |
+
{
|
171 |
+
"epoch": 7.0,
|
172 |
+
"grad_norm": 26.53660774230957,
|
173 |
+
"learning_rate": 3.7500000000000005e-06,
|
174 |
+
"loss": 0.3805,
|
175 |
+
"step": 5684
|
176 |
+
},
|
177 |
+
{
|
178 |
+
"epoch": 7.0,
|
179 |
+
"eval_exact_match": 65.125,
|
180 |
+
"eval_f1": 80.04186683186089,
|
181 |
+
"eval_runtime": 57.4933,
|
182 |
+
"eval_samples_per_second": 88.897,
|
183 |
+
"eval_steps_per_second": 3.183,
|
184 |
+
"step": 5684
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 8.0,
|
188 |
+
"step": 6496,
|
189 |
+
"train_exact_match": 86.31368631368632,
|
190 |
+
"train_f1": 93.75812824286764,
|
191 |
+
"train_runtime": 17.4124,
|
192 |
+
"train_samples_per_second": 88.328,
|
193 |
+
"train_steps_per_second": 3.159
|
194 |
+
},
|
195 |
+
{
|
196 |
+
"epoch": 8.0,
|
197 |
+
"grad_norm": 34.22230911254883,
|
198 |
+
"learning_rate": 2.5e-06,
|
199 |
+
"loss": 0.3381,
|
200 |
+
"step": 6496
|
201 |
+
},
|
202 |
+
{
|
203 |
+
"epoch": 8.0,
|
204 |
+
"eval_exact_match": 65.90625,
|
205 |
+
"eval_f1": 80.14333737032665,
|
206 |
+
"eval_runtime": 57.8402,
|
207 |
+
"eval_samples_per_second": 88.364,
|
208 |
+
"eval_steps_per_second": 3.164,
|
209 |
+
"step": 6496
|
210 |
+
},
|
211 |
+
{
|
212 |
+
"epoch": 9.0,
|
213 |
+
"step": 7308,
|
214 |
+
"train_exact_match": 86.01398601398601,
|
215 |
+
"train_f1": 94.05045602201064,
|
216 |
+
"train_runtime": 18.8091,
|
217 |
+
"train_samples_per_second": 86.767,
|
218 |
+
"train_steps_per_second": 3.137
|
219 |
+
},
|
220 |
+
{
|
221 |
+
"epoch": 9.0,
|
222 |
+
"grad_norm": 87.17084503173828,
|
223 |
+
"learning_rate": 1.25e-06,
|
224 |
+
"loss": 0.3032,
|
225 |
+
"step": 7308
|
226 |
+
},
|
227 |
+
{
|
228 |
+
"epoch": 9.0,
|
229 |
+
"eval_exact_match": 66.21875,
|
230 |
+
"eval_f1": 80.52702562432783,
|
231 |
+
"eval_runtime": 58.1783,
|
232 |
+
"eval_samples_per_second": 87.851,
|
233 |
+
"eval_steps_per_second": 3.146,
|
234 |
+
"step": 7308
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 10.0,
|
238 |
+
"step": 8120,
|
239 |
+
"train_exact_match": 86.41358641358642,
|
240 |
+
"train_f1": 94.52312811018126,
|
241 |
+
"train_runtime": 17.2785,
|
242 |
+
"train_samples_per_second": 87.276,
|
243 |
+
"train_steps_per_second": 3.125
|
244 |
+
},
|
245 |
+
{
|
246 |
+
"epoch": 10.0,
|
247 |
+
"grad_norm": 31.117448806762695,
|
248 |
+
"learning_rate": 0.0,
|
249 |
+
"loss": 0.2815,
|
250 |
+
"step": 8120
|
251 |
+
},
|
252 |
+
{
|
253 |
+
"epoch": 10.0,
|
254 |
+
"eval_exact_match": 66.4375,
|
255 |
+
"eval_f1": 80.63612438300382,
|
256 |
+
"eval_runtime": 58.3935,
|
257 |
+
"eval_samples_per_second": 87.527,
|
258 |
+
"eval_steps_per_second": 3.134,
|
259 |
+
"step": 8120
|
260 |
+
},
|
261 |
+
{
|
262 |
+
"epoch": 10.0,
|
263 |
+
"step": 8120,
|
264 |
+
"total_flos": 4.451320899376128e+16,
|
265 |
+
"train_loss": 0.6037781062384544,
|
266 |
+
"train_runtime": 5220.6695,
|
267 |
+
"train_samples_per_second": 43.508,
|
268 |
+
"train_steps_per_second": 1.555
|
269 |
+
}
|
270 |
+
],
|
271 |
+
"logging_steps": 500,
|
272 |
+
"max_steps": 8120,
|
273 |
+
"num_input_tokens_seen": 0,
|
274 |
+
"num_train_epochs": 10,
|
275 |
+
"save_steps": 500,
|
276 |
+
"total_flos": 4.451320899376128e+16,
|
277 |
+
"train_batch_size": 28,
|
278 |
+
"trial_name": null,
|
279 |
+
"trial_params": null
|
280 |
+
}
|