gomgomcode
commited on
Commit
•
2db404b
1
Parent(s):
338f453
Upload 9 files
Browse files- pytorch_model.bin +1 -1
- train_results.txt +3 -3
- trainer_state.json +231 -99
- training_args.bin +1 -1
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 501634165
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2b6196274b15ecaba8ff351fc9738fb49e7483a98ff2033e8c1bca2cb96a90f4
|
3 |
size 501634165
|
train_results.txt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
-
epoch =
|
2 |
-
train_runtime =
|
3 |
-
train_samples_per_second = 17.
|
|
|
1 |
+
epoch = 3.0
|
2 |
+
train_runtime = 1945.6382
|
3 |
+
train_samples_per_second = 17.359
|
trainer_state.json
CHANGED
@@ -1,292 +1,424 @@
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
-
"epoch":
|
5 |
-
"global_step":
|
6 |
"is_hyper_param_search": false,
|
7 |
"is_local_process_zero": true,
|
8 |
"is_world_process_zero": true,
|
9 |
"log_history": [
|
10 |
{
|
11 |
"epoch": 0.04,
|
12 |
-
"learning_rate": 2.
|
13 |
-
"loss": 0.
|
14 |
"step": 500
|
15 |
},
|
16 |
{
|
17 |
"epoch": 0.09,
|
18 |
-
"learning_rate": 2.
|
19 |
-
"loss": 0.
|
20 |
"step": 1000
|
21 |
},
|
22 |
{
|
23 |
"epoch": 0.13,
|
24 |
-
"learning_rate": 2.
|
25 |
-
"loss": 0.
|
26 |
"step": 1500
|
27 |
},
|
28 |
{
|
29 |
"epoch": 0.18,
|
30 |
-
"learning_rate": 2.
|
31 |
-
"loss": 0.
|
32 |
"step": 2000
|
33 |
},
|
34 |
{
|
35 |
"epoch": 0.22,
|
36 |
-
"learning_rate": 2.
|
37 |
-
"loss": 0.
|
38 |
"step": 2500
|
39 |
},
|
40 |
{
|
41 |
"epoch": 0.27,
|
42 |
-
"learning_rate": 2.
|
43 |
-
"loss": 0.
|
44 |
"step": 3000
|
45 |
},
|
46 |
{
|
47 |
"epoch": 0.31,
|
48 |
-
"learning_rate": 2.
|
49 |
-
"loss": 0.
|
50 |
"step": 3500
|
51 |
},
|
52 |
{
|
53 |
"epoch": 0.36,
|
54 |
-
"learning_rate": 2.
|
55 |
-
"loss": 0.
|
56 |
"step": 4000
|
57 |
},
|
58 |
{
|
59 |
"epoch": 0.4,
|
60 |
-
"learning_rate": 2.
|
61 |
-
"loss": 0.
|
62 |
"step": 4500
|
63 |
},
|
64 |
{
|
65 |
"epoch": 0.44,
|
66 |
-
"learning_rate": 2.
|
67 |
-
"loss": 0.
|
68 |
"step": 5000
|
69 |
},
|
70 |
{
|
71 |
"epoch": 0.49,
|
72 |
-
"learning_rate": 2.
|
73 |
-
"loss": 0.
|
74 |
"step": 5500
|
75 |
},
|
76 |
{
|
77 |
"epoch": 0.53,
|
78 |
-
"learning_rate": 2.
|
79 |
-
"loss": 0.
|
80 |
"step": 6000
|
81 |
},
|
82 |
{
|
83 |
"epoch": 0.58,
|
84 |
-
"learning_rate": 2.
|
85 |
-
"loss": 0.
|
86 |
"step": 6500
|
87 |
},
|
88 |
{
|
89 |
"epoch": 0.62,
|
90 |
-
"learning_rate": 2.
|
91 |
-
"loss": 0.
|
92 |
"step": 7000
|
93 |
},
|
94 |
{
|
95 |
"epoch": 0.67,
|
96 |
-
"learning_rate": 2.
|
97 |
-
"loss": 0.
|
98 |
"step": 7500
|
99 |
},
|
100 |
{
|
101 |
"epoch": 0.71,
|
102 |
-
"learning_rate":
|
103 |
-
"loss": 0.
|
104 |
"step": 8000
|
105 |
},
|
106 |
{
|
107 |
"epoch": 0.76,
|
108 |
-
"learning_rate":
|
109 |
-
"loss": 0.
|
110 |
"step": 8500
|
111 |
},
|
112 |
{
|
113 |
"epoch": 0.8,
|
114 |
-
"learning_rate":
|
115 |
-
"loss": 0.
|
116 |
"step": 9000
|
117 |
},
|
118 |
{
|
119 |
"epoch": 0.84,
|
120 |
-
"learning_rate":
|
121 |
-
"loss": 0.
|
122 |
"step": 9500
|
123 |
},
|
124 |
{
|
125 |
"epoch": 0.89,
|
126 |
-
"learning_rate":
|
127 |
-
"loss": 0.
|
128 |
"step": 10000
|
129 |
},
|
130 |
{
|
131 |
"epoch": 0.93,
|
132 |
-
"learning_rate":
|
133 |
-
"loss": 0.
|
134 |
"step": 10500
|
135 |
},
|
136 |
{
|
137 |
"epoch": 0.98,
|
138 |
-
"learning_rate":
|
139 |
-
"loss": 0.
|
140 |
"step": 11000
|
141 |
},
|
142 |
{
|
143 |
"epoch": 1.02,
|
144 |
-
"learning_rate": 1.
|
145 |
-
"loss": 0.
|
146 |
"step": 11500
|
147 |
},
|
148 |
{
|
149 |
"epoch": 1.07,
|
150 |
-
"learning_rate": 1.
|
151 |
-
"loss": 0.
|
152 |
"step": 12000
|
153 |
},
|
154 |
{
|
155 |
"epoch": 1.11,
|
156 |
-
"learning_rate": 1.
|
157 |
-
"loss": 0.
|
158 |
"step": 12500
|
159 |
},
|
160 |
{
|
161 |
"epoch": 1.15,
|
162 |
-
"learning_rate": 1.
|
163 |
-
"loss": 0.
|
164 |
"step": 13000
|
165 |
},
|
166 |
{
|
167 |
"epoch": 1.2,
|
168 |
-
"learning_rate": 1.
|
169 |
-
"loss": 0.
|
170 |
"step": 13500
|
171 |
},
|
172 |
{
|
173 |
"epoch": 1.24,
|
174 |
-
"learning_rate": 1.
|
175 |
-
"loss": 0.
|
176 |
"step": 14000
|
177 |
},
|
178 |
{
|
179 |
"epoch": 1.29,
|
180 |
-
"learning_rate": 1.
|
181 |
-
"loss": 0.
|
182 |
"step": 14500
|
183 |
},
|
184 |
{
|
185 |
"epoch": 1.33,
|
186 |
-
"learning_rate": 1.
|
187 |
-
"loss": 0.
|
188 |
"step": 15000
|
189 |
},
|
190 |
{
|
191 |
"epoch": 1.38,
|
192 |
-
"learning_rate":
|
193 |
-
"loss": 0.
|
194 |
"step": 15500
|
195 |
},
|
196 |
{
|
197 |
"epoch": 1.42,
|
198 |
-
"learning_rate":
|
199 |
-
"loss": 0.
|
200 |
"step": 16000
|
201 |
},
|
202 |
{
|
203 |
"epoch": 1.47,
|
204 |
-
"learning_rate":
|
205 |
-
"loss": 0.
|
206 |
"step": 16500
|
207 |
},
|
208 |
{
|
209 |
"epoch": 1.51,
|
210 |
-
"learning_rate":
|
211 |
-
"loss": 0.
|
212 |
"step": 17000
|
213 |
},
|
214 |
{
|
215 |
"epoch": 1.55,
|
216 |
-
"learning_rate":
|
217 |
-
"loss": 0.
|
218 |
"step": 17500
|
219 |
},
|
220 |
{
|
221 |
"epoch": 1.6,
|
222 |
-
"learning_rate":
|
223 |
-
"loss": 0.
|
224 |
"step": 18000
|
225 |
},
|
226 |
{
|
227 |
"epoch": 1.64,
|
228 |
-
"learning_rate":
|
229 |
-
"loss": 0.
|
230 |
"step": 18500
|
231 |
},
|
232 |
{
|
233 |
"epoch": 1.69,
|
234 |
-
"learning_rate":
|
235 |
-
"loss": 0.
|
236 |
"step": 19000
|
237 |
},
|
238 |
{
|
239 |
"epoch": 1.73,
|
240 |
-
"learning_rate":
|
241 |
-
"loss": 0.
|
242 |
"step": 19500
|
243 |
},
|
244 |
{
|
245 |
"epoch": 1.78,
|
246 |
-
"learning_rate":
|
247 |
-
"loss": 0.
|
248 |
"step": 20000
|
249 |
},
|
250 |
{
|
251 |
"epoch": 1.82,
|
252 |
-
"learning_rate":
|
253 |
-
"loss": 0.
|
254 |
"step": 20500
|
255 |
},
|
256 |
{
|
257 |
"epoch": 1.87,
|
258 |
-
"learning_rate":
|
259 |
-
"loss": 0.
|
260 |
"step": 21000
|
261 |
},
|
262 |
{
|
263 |
"epoch": 1.91,
|
264 |
-
"learning_rate": 1.
|
265 |
-
"loss": 0.
|
266 |
"step": 21500
|
267 |
},
|
268 |
{
|
269 |
"epoch": 1.95,
|
270 |
-
"learning_rate":
|
271 |
-
"loss": 0.
|
272 |
"step": 22000
|
273 |
},
|
274 |
{
|
275 |
"epoch": 2.0,
|
276 |
-
"learning_rate":
|
277 |
-
"loss": 0.
|
278 |
"step": 22500
|
279 |
},
|
280 |
{
|
281 |
-
"epoch": 2.
|
282 |
-
"
|
283 |
-
"
|
284 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
285 |
}
|
286 |
],
|
287 |
-
"max_steps":
|
288 |
-
"num_train_epochs":
|
289 |
-
"total_flos":
|
290 |
"trial_name": null,
|
291 |
"trial_params": null
|
292 |
}
|
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
+
"epoch": 3.0,
|
5 |
+
"global_step": 33774,
|
6 |
"is_hyper_param_search": false,
|
7 |
"is_local_process_zero": true,
|
8 |
"is_world_process_zero": true,
|
9 |
"log_history": [
|
10 |
{
|
11 |
"epoch": 0.04,
|
12 |
+
"learning_rate": 2.9555871380351753e-05,
|
13 |
+
"loss": 0.5739,
|
14 |
"step": 500
|
15 |
},
|
16 |
{
|
17 |
"epoch": 0.09,
|
18 |
+
"learning_rate": 2.91117427607035e-05,
|
19 |
+
"loss": 0.4701,
|
20 |
"step": 1000
|
21 |
},
|
22 |
{
|
23 |
"epoch": 0.13,
|
24 |
+
"learning_rate": 2.8667614141055253e-05,
|
25 |
+
"loss": 0.4401,
|
26 |
"step": 1500
|
27 |
},
|
28 |
{
|
29 |
"epoch": 0.18,
|
30 |
+
"learning_rate": 2.8223485521406998e-05,
|
31 |
+
"loss": 0.4007,
|
32 |
"step": 2000
|
33 |
},
|
34 |
{
|
35 |
"epoch": 0.22,
|
36 |
+
"learning_rate": 2.777935690175875e-05,
|
37 |
+
"loss": 0.3886,
|
38 |
"step": 2500
|
39 |
},
|
40 |
{
|
41 |
"epoch": 0.27,
|
42 |
+
"learning_rate": 2.73352282821105e-05,
|
43 |
+
"loss": 0.3668,
|
44 |
"step": 3000
|
45 |
},
|
46 |
{
|
47 |
"epoch": 0.31,
|
48 |
+
"learning_rate": 2.689109966246225e-05,
|
49 |
+
"loss": 0.3622,
|
50 |
"step": 3500
|
51 |
},
|
52 |
{
|
53 |
"epoch": 0.36,
|
54 |
+
"learning_rate": 2.6446971042814e-05,
|
55 |
+
"loss": 0.3513,
|
56 |
"step": 4000
|
57 |
},
|
58 |
{
|
59 |
"epoch": 0.4,
|
60 |
+
"learning_rate": 2.600284242316575e-05,
|
61 |
+
"loss": 0.3233,
|
62 |
"step": 4500
|
63 |
},
|
64 |
{
|
65 |
"epoch": 0.44,
|
66 |
+
"learning_rate": 2.55587138035175e-05,
|
67 |
+
"loss": 0.3341,
|
68 |
"step": 5000
|
69 |
},
|
70 |
{
|
71 |
"epoch": 0.49,
|
72 |
+
"learning_rate": 2.5114585183869247e-05,
|
73 |
+
"loss": 0.325,
|
74 |
"step": 5500
|
75 |
},
|
76 |
{
|
77 |
"epoch": 0.53,
|
78 |
+
"learning_rate": 2.4670456564221e-05,
|
79 |
+
"loss": 0.3164,
|
80 |
"step": 6000
|
81 |
},
|
82 |
{
|
83 |
"epoch": 0.58,
|
84 |
+
"learning_rate": 2.4226327944572748e-05,
|
85 |
+
"loss": 0.2953,
|
86 |
"step": 6500
|
87 |
},
|
88 |
{
|
89 |
"epoch": 0.62,
|
90 |
+
"learning_rate": 2.37821993249245e-05,
|
91 |
+
"loss": 0.3033,
|
92 |
"step": 7000
|
93 |
},
|
94 |
{
|
95 |
"epoch": 0.67,
|
96 |
+
"learning_rate": 2.3338070705276248e-05,
|
97 |
+
"loss": 0.2875,
|
98 |
"step": 7500
|
99 |
},
|
100 |
{
|
101 |
"epoch": 0.71,
|
102 |
+
"learning_rate": 2.2893942085628e-05,
|
103 |
+
"loss": 0.2848,
|
104 |
"step": 8000
|
105 |
},
|
106 |
{
|
107 |
"epoch": 0.76,
|
108 |
+
"learning_rate": 2.244981346597975e-05,
|
109 |
+
"loss": 0.27,
|
110 |
"step": 8500
|
111 |
},
|
112 |
{
|
113 |
"epoch": 0.8,
|
114 |
+
"learning_rate": 2.20056848463315e-05,
|
115 |
+
"loss": 0.2822,
|
116 |
"step": 9000
|
117 |
},
|
118 |
{
|
119 |
"epoch": 0.84,
|
120 |
+
"learning_rate": 2.1561556226683245e-05,
|
121 |
+
"loss": 0.2595,
|
122 |
"step": 9500
|
123 |
},
|
124 |
{
|
125 |
"epoch": 0.89,
|
126 |
+
"learning_rate": 2.1117427607034997e-05,
|
127 |
+
"loss": 0.2592,
|
128 |
"step": 10000
|
129 |
},
|
130 |
{
|
131 |
"epoch": 0.93,
|
132 |
+
"learning_rate": 2.0673298987386746e-05,
|
133 |
+
"loss": 0.2493,
|
134 |
"step": 10500
|
135 |
},
|
136 |
{
|
137 |
"epoch": 0.98,
|
138 |
+
"learning_rate": 2.0229170367738498e-05,
|
139 |
+
"loss": 0.2574,
|
140 |
"step": 11000
|
141 |
},
|
142 |
{
|
143 |
"epoch": 1.02,
|
144 |
+
"learning_rate": 1.978504174809025e-05,
|
145 |
+
"loss": 0.2059,
|
146 |
"step": 11500
|
147 |
},
|
148 |
{
|
149 |
"epoch": 1.07,
|
150 |
+
"learning_rate": 1.9340913128441998e-05,
|
151 |
+
"loss": 0.1537,
|
152 |
"step": 12000
|
153 |
},
|
154 |
{
|
155 |
"epoch": 1.11,
|
156 |
+
"learning_rate": 1.889678450879375e-05,
|
157 |
+
"loss": 0.1588,
|
158 |
"step": 12500
|
159 |
},
|
160 |
{
|
161 |
"epoch": 1.15,
|
162 |
+
"learning_rate": 1.8452655889145495e-05,
|
163 |
+
"loss": 0.1487,
|
164 |
"step": 13000
|
165 |
},
|
166 |
{
|
167 |
"epoch": 1.2,
|
168 |
+
"learning_rate": 1.8008527269497247e-05,
|
169 |
+
"loss": 0.1608,
|
170 |
"step": 13500
|
171 |
},
|
172 |
{
|
173 |
"epoch": 1.24,
|
174 |
+
"learning_rate": 1.7564398649848995e-05,
|
175 |
+
"loss": 0.1619,
|
176 |
"step": 14000
|
177 |
},
|
178 |
{
|
179 |
"epoch": 1.29,
|
180 |
+
"learning_rate": 1.7120270030200747e-05,
|
181 |
+
"loss": 0.155,
|
182 |
"step": 14500
|
183 |
},
|
184 |
{
|
185 |
"epoch": 1.33,
|
186 |
+
"learning_rate": 1.6676141410552495e-05,
|
187 |
+
"loss": 0.1418,
|
188 |
"step": 15000
|
189 |
},
|
190 |
{
|
191 |
"epoch": 1.38,
|
192 |
+
"learning_rate": 1.6232012790904247e-05,
|
193 |
+
"loss": 0.1558,
|
194 |
"step": 15500
|
195 |
},
|
196 |
{
|
197 |
"epoch": 1.42,
|
198 |
+
"learning_rate": 1.5787884171255996e-05,
|
199 |
+
"loss": 0.155,
|
200 |
"step": 16000
|
201 |
},
|
202 |
{
|
203 |
"epoch": 1.47,
|
204 |
+
"learning_rate": 1.5343755551607744e-05,
|
205 |
+
"loss": 0.1522,
|
206 |
"step": 16500
|
207 |
},
|
208 |
{
|
209 |
"epoch": 1.51,
|
210 |
+
"learning_rate": 1.4899626931959496e-05,
|
211 |
+
"loss": 0.1386,
|
212 |
"step": 17000
|
213 |
},
|
214 |
{
|
215 |
"epoch": 1.55,
|
216 |
+
"learning_rate": 1.4455498312311245e-05,
|
217 |
+
"loss": 0.1442,
|
218 |
"step": 17500
|
219 |
},
|
220 |
{
|
221 |
"epoch": 1.6,
|
222 |
+
"learning_rate": 1.4011369692662995e-05,
|
223 |
+
"loss": 0.1382,
|
224 |
"step": 18000
|
225 |
},
|
226 |
{
|
227 |
"epoch": 1.64,
|
228 |
+
"learning_rate": 1.3567241073014745e-05,
|
229 |
+
"loss": 0.1437,
|
230 |
"step": 18500
|
231 |
},
|
232 |
{
|
233 |
"epoch": 1.69,
|
234 |
+
"learning_rate": 1.3123112453366497e-05,
|
235 |
+
"loss": 0.1405,
|
236 |
"step": 19000
|
237 |
},
|
238 |
{
|
239 |
"epoch": 1.73,
|
240 |
+
"learning_rate": 1.2678983833718245e-05,
|
241 |
+
"loss": 0.1357,
|
242 |
"step": 19500
|
243 |
},
|
244 |
{
|
245 |
"epoch": 1.78,
|
246 |
+
"learning_rate": 1.2234855214069995e-05,
|
247 |
+
"loss": 0.125,
|
248 |
"step": 20000
|
249 |
},
|
250 |
{
|
251 |
"epoch": 1.82,
|
252 |
+
"learning_rate": 1.1790726594421745e-05,
|
253 |
+
"loss": 0.1376,
|
254 |
"step": 20500
|
255 |
},
|
256 |
{
|
257 |
"epoch": 1.87,
|
258 |
+
"learning_rate": 1.1346597974773496e-05,
|
259 |
+
"loss": 0.136,
|
260 |
"step": 21000
|
261 |
},
|
262 |
{
|
263 |
"epoch": 1.91,
|
264 |
+
"learning_rate": 1.0902469355125244e-05,
|
265 |
+
"loss": 0.1251,
|
266 |
"step": 21500
|
267 |
},
|
268 |
{
|
269 |
"epoch": 1.95,
|
270 |
+
"learning_rate": 1.0458340735476994e-05,
|
271 |
+
"loss": 0.1258,
|
272 |
"step": 22000
|
273 |
},
|
274 |
{
|
275 |
"epoch": 2.0,
|
276 |
+
"learning_rate": 1.0014212115828744e-05,
|
277 |
+
"loss": 0.1303,
|
278 |
"step": 22500
|
279 |
},
|
280 |
{
|
281 |
+
"epoch": 2.04,
|
282 |
+
"learning_rate": 9.570083496180493e-06,
|
283 |
+
"loss": 0.0792,
|
284 |
+
"step": 23000
|
285 |
+
},
|
286 |
+
{
|
287 |
+
"epoch": 2.09,
|
288 |
+
"learning_rate": 9.125954876532243e-06,
|
289 |
+
"loss": 0.0794,
|
290 |
+
"step": 23500
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 2.13,
|
294 |
+
"learning_rate": 8.681826256883993e-06,
|
295 |
+
"loss": 0.0781,
|
296 |
+
"step": 24000
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 2.18,
|
300 |
+
"learning_rate": 8.237697637235745e-06,
|
301 |
+
"loss": 0.0777,
|
302 |
+
"step": 24500
|
303 |
+
},
|
304 |
+
{
|
305 |
+
"epoch": 2.22,
|
306 |
+
"learning_rate": 7.793569017587494e-06,
|
307 |
+
"loss": 0.077,
|
308 |
+
"step": 25000
|
309 |
+
},
|
310 |
+
{
|
311 |
+
"epoch": 2.27,
|
312 |
+
"learning_rate": 7.349440397939243e-06,
|
313 |
+
"loss": 0.0792,
|
314 |
+
"step": 25500
|
315 |
+
},
|
316 |
+
{
|
317 |
+
"epoch": 2.31,
|
318 |
+
"learning_rate": 6.905311778290994e-06,
|
319 |
+
"loss": 0.0774,
|
320 |
+
"step": 26000
|
321 |
+
},
|
322 |
+
{
|
323 |
+
"epoch": 2.35,
|
324 |
+
"learning_rate": 6.461183158642743e-06,
|
325 |
+
"loss": 0.0762,
|
326 |
+
"step": 26500
|
327 |
+
},
|
328 |
+
{
|
329 |
+
"epoch": 2.4,
|
330 |
+
"learning_rate": 6.017054538994493e-06,
|
331 |
+
"loss": 0.0744,
|
332 |
+
"step": 27000
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 2.44,
|
336 |
+
"learning_rate": 5.572925919346243e-06,
|
337 |
+
"loss": 0.0725,
|
338 |
+
"step": 27500
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 2.49,
|
342 |
+
"learning_rate": 5.128797299697992e-06,
|
343 |
+
"loss": 0.076,
|
344 |
+
"step": 28000
|
345 |
+
},
|
346 |
+
{
|
347 |
+
"epoch": 2.53,
|
348 |
+
"learning_rate": 4.684668680049743e-06,
|
349 |
+
"loss": 0.0682,
|
350 |
+
"step": 28500
|
351 |
+
},
|
352 |
+
{
|
353 |
+
"epoch": 2.58,
|
354 |
+
"learning_rate": 4.240540060401492e-06,
|
355 |
+
"loss": 0.074,
|
356 |
+
"step": 29000
|
357 |
+
},
|
358 |
+
{
|
359 |
+
"epoch": 2.62,
|
360 |
+
"learning_rate": 3.7964114407532424e-06,
|
361 |
+
"loss": 0.0738,
|
362 |
+
"step": 29500
|
363 |
+
},
|
364 |
+
{
|
365 |
+
"epoch": 2.66,
|
366 |
+
"learning_rate": 3.3522828211049918e-06,
|
367 |
+
"loss": 0.0678,
|
368 |
+
"step": 30000
|
369 |
+
},
|
370 |
+
{
|
371 |
+
"epoch": 2.71,
|
372 |
+
"learning_rate": 2.908154201456742e-06,
|
373 |
+
"loss": 0.0668,
|
374 |
+
"step": 30500
|
375 |
+
},
|
376 |
+
{
|
377 |
+
"epoch": 2.75,
|
378 |
+
"learning_rate": 2.4640255818084917e-06,
|
379 |
+
"loss": 0.0685,
|
380 |
+
"step": 31000
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 2.8,
|
384 |
+
"learning_rate": 2.0198969621602414e-06,
|
385 |
+
"loss": 0.0668,
|
386 |
+
"step": 31500
|
387 |
+
},
|
388 |
+
{
|
389 |
+
"epoch": 2.84,
|
390 |
+
"learning_rate": 1.5757683425119916e-06,
|
391 |
+
"loss": 0.068,
|
392 |
+
"step": 32000
|
393 |
+
},
|
394 |
+
{
|
395 |
+
"epoch": 2.89,
|
396 |
+
"learning_rate": 1.1316397228637413e-06,
|
397 |
+
"loss": 0.0663,
|
398 |
+
"step": 32500
|
399 |
+
},
|
400 |
+
{
|
401 |
+
"epoch": 2.93,
|
402 |
+
"learning_rate": 6.875111032154912e-07,
|
403 |
+
"loss": 0.0629,
|
404 |
+
"step": 33000
|
405 |
+
},
|
406 |
+
{
|
407 |
+
"epoch": 2.98,
|
408 |
+
"learning_rate": 2.4338248356724103e-07,
|
409 |
+
"loss": 0.0676,
|
410 |
+
"step": 33500
|
411 |
+
},
|
412 |
+
{
|
413 |
+
"epoch": 3.0,
|
414 |
+
"step": 33774,
|
415 |
+
"train_runtime": 1945.6382,
|
416 |
+
"train_samples_per_second": 17.359
|
417 |
}
|
418 |
],
|
419 |
+
"max_steps": 33774,
|
420 |
+
"num_train_epochs": 3,
|
421 |
+
"total_flos": 52036968346583040,
|
422 |
"trial_name": null,
|
423 |
"trial_params": null
|
424 |
}
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 2031
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5ea6da2311f8ad0e07c33c998a50bd07f1b7dba25e67f9bbe18dcb1cd7760aad
|
3 |
size 2031
|