Upload folder using huggingface_hub
Browse files- adapter_config.json +3 -3
- adapter_model.safetensors +1 -1
- optimizer.pt +2 -2
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +655 -55
- training_args.bin +1 -1
adapter_config.json
CHANGED
@@ -19,10 +19,10 @@
|
|
19 |
"rank_pattern": {},
|
20 |
"revision": null,
|
21 |
"target_modules": [
|
22 |
-
"query_key_value",
|
23 |
-
"dense",
|
24 |
"dense_4h_to_h",
|
25 |
-
"dense_h_to_4h"
|
|
|
|
|
26 |
],
|
27 |
"task_type": "CAUSAL_LM"
|
28 |
}
|
|
|
19 |
"rank_pattern": {},
|
20 |
"revision": null,
|
21 |
"target_modules": [
|
|
|
|
|
22 |
"dense_4h_to_h",
|
23 |
+
"dense_h_to_4h",
|
24 |
+
"dense",
|
25 |
+
"query_key_value"
|
26 |
],
|
27 |
"task_type": "CAUSAL_LM"
|
28 |
}
|
adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 134235712
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6e5c5c60ac960231df9f8470ee166ad44f7cbdd42330ccbb02e2b2f266910e08
|
3 |
size 134235712
|
optimizer.pt
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 sha256:42cc40db1cb1d7c66d79f9aad1bb927dbb1efe1c1c72e306e8ad4cca0f604df5
|
3 |
+
size 268515002
|
rng_state.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14244
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3850cffcdc194c123cd16ed51c68d080cca74165fcc78ea052f3750f4f9baac4
|
3 |
size 14244
|
scheduler.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1064
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c79c66b60a0abe3dc1f1792ced2b6c99f10b3ada4ba94ee60000ba5931c603a9
|
3 |
size 1064
|
trainer_state.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
-
"epoch":
|
5 |
"eval_steps": 500,
|
6 |
-
"global_step":
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
@@ -11,310 +11,910 @@
|
|
11 |
{
|
12 |
"epoch": 0.02,
|
13 |
"learning_rate": 0.0002,
|
14 |
-
"loss": 2.
|
15 |
"step": 2
|
16 |
},
|
17 |
{
|
18 |
"epoch": 0.04,
|
19 |
"learning_rate": 0.0002,
|
20 |
-
"loss": 2.
|
21 |
"step": 4
|
22 |
},
|
23 |
{
|
24 |
"epoch": 0.06,
|
25 |
"learning_rate": 0.0002,
|
26 |
-
"loss": 2.
|
27 |
"step": 6
|
28 |
},
|
29 |
{
|
30 |
"epoch": 0.08,
|
31 |
"learning_rate": 0.0002,
|
32 |
-
"loss": 2.
|
33 |
"step": 8
|
34 |
},
|
35 |
{
|
36 |
"epoch": 0.1,
|
37 |
"learning_rate": 0.0002,
|
38 |
-
"loss": 2.
|
39 |
"step": 10
|
40 |
},
|
41 |
{
|
42 |
"epoch": 0.12,
|
43 |
"learning_rate": 0.0002,
|
44 |
-
"loss": 2.
|
45 |
"step": 12
|
46 |
},
|
47 |
{
|
48 |
"epoch": 0.14,
|
49 |
"learning_rate": 0.0002,
|
50 |
-
"loss": 2.
|
51 |
"step": 14
|
52 |
},
|
53 |
{
|
54 |
"epoch": 0.16,
|
55 |
"learning_rate": 0.0002,
|
56 |
-
"loss": 2.
|
57 |
"step": 16
|
58 |
},
|
59 |
{
|
60 |
"epoch": 0.18,
|
61 |
"learning_rate": 0.0002,
|
62 |
-
"loss": 2.
|
63 |
"step": 18
|
64 |
},
|
65 |
{
|
66 |
"epoch": 0.2,
|
67 |
"learning_rate": 0.0002,
|
68 |
-
"loss":
|
69 |
"step": 20
|
70 |
},
|
71 |
{
|
72 |
"epoch": 0.22,
|
73 |
"learning_rate": 0.0002,
|
74 |
-
"loss": 2.
|
75 |
"step": 22
|
76 |
},
|
77 |
{
|
78 |
"epoch": 0.24,
|
79 |
"learning_rate": 0.0002,
|
80 |
-
"loss": 2.
|
81 |
"step": 24
|
82 |
},
|
83 |
{
|
84 |
"epoch": 0.26,
|
85 |
"learning_rate": 0.0002,
|
86 |
-
"loss": 2.
|
87 |
"step": 26
|
88 |
},
|
89 |
{
|
90 |
"epoch": 0.28,
|
91 |
"learning_rate": 0.0002,
|
92 |
-
"loss": 2.
|
93 |
"step": 28
|
94 |
},
|
95 |
{
|
96 |
"epoch": 0.3,
|
97 |
"learning_rate": 0.0002,
|
98 |
-
"loss": 2.
|
99 |
"step": 30
|
100 |
},
|
101 |
{
|
102 |
"epoch": 0.32,
|
103 |
"learning_rate": 0.0002,
|
104 |
-
"loss": 2.
|
105 |
"step": 32
|
106 |
},
|
107 |
{
|
108 |
"epoch": 0.34,
|
109 |
"learning_rate": 0.0002,
|
110 |
-
"loss": 2.
|
111 |
"step": 34
|
112 |
},
|
113 |
{
|
114 |
"epoch": 0.36,
|
115 |
"learning_rate": 0.0002,
|
116 |
-
"loss": 2.
|
117 |
"step": 36
|
118 |
},
|
119 |
{
|
120 |
"epoch": 0.38,
|
121 |
"learning_rate": 0.0002,
|
122 |
-
"loss": 2.
|
123 |
"step": 38
|
124 |
},
|
125 |
{
|
126 |
"epoch": 0.4,
|
127 |
"learning_rate": 0.0002,
|
128 |
-
"loss": 2.
|
129 |
"step": 40
|
130 |
},
|
131 |
{
|
132 |
"epoch": 0.42,
|
133 |
"learning_rate": 0.0002,
|
134 |
-
"loss": 2.
|
135 |
"step": 42
|
136 |
},
|
137 |
{
|
138 |
"epoch": 0.44,
|
139 |
"learning_rate": 0.0002,
|
140 |
-
"loss": 2.
|
141 |
"step": 44
|
142 |
},
|
143 |
{
|
144 |
"epoch": 0.46,
|
145 |
"learning_rate": 0.0002,
|
146 |
-
"loss": 2.
|
147 |
"step": 46
|
148 |
},
|
149 |
{
|
150 |
"epoch": 0.48,
|
151 |
"learning_rate": 0.0002,
|
152 |
-
"loss": 2.
|
153 |
"step": 48
|
154 |
},
|
155 |
{
|
156 |
"epoch": 0.5,
|
157 |
"learning_rate": 0.0002,
|
158 |
-
"loss":
|
159 |
"step": 50
|
160 |
},
|
161 |
{
|
162 |
"epoch": 0.52,
|
163 |
"learning_rate": 0.0002,
|
164 |
-
"loss": 2.
|
165 |
"step": 52
|
166 |
},
|
167 |
{
|
168 |
"epoch": 0.54,
|
169 |
"learning_rate": 0.0002,
|
170 |
-
"loss": 2.
|
171 |
"step": 54
|
172 |
},
|
173 |
{
|
174 |
"epoch": 0.56,
|
175 |
"learning_rate": 0.0002,
|
176 |
-
"loss": 2.
|
177 |
"step": 56
|
178 |
},
|
179 |
{
|
180 |
"epoch": 0.58,
|
181 |
"learning_rate": 0.0002,
|
182 |
-
"loss": 2.
|
183 |
"step": 58
|
184 |
},
|
185 |
{
|
186 |
"epoch": 0.6,
|
187 |
"learning_rate": 0.0002,
|
188 |
-
"loss": 2.
|
189 |
"step": 60
|
190 |
},
|
191 |
{
|
192 |
"epoch": 0.62,
|
193 |
"learning_rate": 0.0002,
|
194 |
-
"loss": 2.
|
195 |
"step": 62
|
196 |
},
|
197 |
{
|
198 |
"epoch": 0.64,
|
199 |
"learning_rate": 0.0002,
|
200 |
-
"loss": 2.
|
201 |
"step": 64
|
202 |
},
|
203 |
{
|
204 |
"epoch": 0.66,
|
205 |
"learning_rate": 0.0002,
|
206 |
-
"loss": 2.
|
207 |
"step": 66
|
208 |
},
|
209 |
{
|
210 |
"epoch": 0.68,
|
211 |
"learning_rate": 0.0002,
|
212 |
-
"loss": 2.
|
213 |
"step": 68
|
214 |
},
|
215 |
{
|
216 |
"epoch": 0.7,
|
217 |
"learning_rate": 0.0002,
|
218 |
-
"loss": 2.
|
219 |
"step": 70
|
220 |
},
|
221 |
{
|
222 |
"epoch": 0.72,
|
223 |
"learning_rate": 0.0002,
|
224 |
-
"loss": 2.
|
225 |
"step": 72
|
226 |
},
|
227 |
{
|
228 |
"epoch": 0.74,
|
229 |
"learning_rate": 0.0002,
|
230 |
-
"loss":
|
231 |
"step": 74
|
232 |
},
|
233 |
{
|
234 |
"epoch": 0.76,
|
235 |
"learning_rate": 0.0002,
|
236 |
-
"loss": 2.
|
237 |
"step": 76
|
238 |
},
|
239 |
{
|
240 |
"epoch": 0.78,
|
241 |
"learning_rate": 0.0002,
|
242 |
-
"loss": 2.
|
243 |
"step": 78
|
244 |
},
|
245 |
{
|
246 |
"epoch": 0.8,
|
247 |
"learning_rate": 0.0002,
|
248 |
-
"loss": 2.
|
249 |
"step": 80
|
250 |
},
|
251 |
{
|
252 |
"epoch": 0.82,
|
253 |
"learning_rate": 0.0002,
|
254 |
-
"loss": 2.
|
255 |
"step": 82
|
256 |
},
|
257 |
{
|
258 |
"epoch": 0.84,
|
259 |
"learning_rate": 0.0002,
|
260 |
-
"loss": 2.
|
261 |
"step": 84
|
262 |
},
|
263 |
{
|
264 |
"epoch": 0.86,
|
265 |
"learning_rate": 0.0002,
|
266 |
-
"loss": 2.
|
267 |
"step": 86
|
268 |
},
|
269 |
{
|
270 |
"epoch": 0.88,
|
271 |
"learning_rate": 0.0002,
|
272 |
-
"loss": 2.
|
273 |
"step": 88
|
274 |
},
|
275 |
{
|
276 |
"epoch": 0.89,
|
277 |
"learning_rate": 0.0002,
|
278 |
-
"loss": 2.
|
279 |
"step": 90
|
280 |
},
|
281 |
{
|
282 |
"epoch": 0.91,
|
283 |
"learning_rate": 0.0002,
|
284 |
-
"loss": 2.
|
285 |
"step": 92
|
286 |
},
|
287 |
{
|
288 |
"epoch": 0.93,
|
289 |
"learning_rate": 0.0002,
|
290 |
-
"loss":
|
291 |
"step": 94
|
292 |
},
|
293 |
{
|
294 |
"epoch": 0.95,
|
295 |
"learning_rate": 0.0002,
|
296 |
-
"loss": 2.
|
297 |
"step": 96
|
298 |
},
|
299 |
{
|
300 |
"epoch": 0.97,
|
301 |
"learning_rate": 0.0002,
|
302 |
-
"loss": 2.
|
303 |
"step": 98
|
304 |
},
|
305 |
{
|
306 |
"epoch": 0.99,
|
307 |
"learning_rate": 0.0002,
|
308 |
-
"loss":
|
309 |
"step": 100
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
310 |
}
|
311 |
],
|
312 |
"logging_steps": 2,
|
313 |
-
"max_steps":
|
314 |
"num_input_tokens_seen": 0,
|
315 |
-
"num_train_epochs":
|
316 |
"save_steps": 500,
|
317 |
-
"total_flos":
|
318 |
"train_batch_size": 1,
|
319 |
"trial_name": null,
|
320 |
"trial_params": null
|
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
+
"epoch": 2.983219390926041,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 300,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
|
|
11 |
{
|
12 |
"epoch": 0.02,
|
13 |
"learning_rate": 0.0002,
|
14 |
+
"loss": 2.5828,
|
15 |
"step": 2
|
16 |
},
|
17 |
{
|
18 |
"epoch": 0.04,
|
19 |
"learning_rate": 0.0002,
|
20 |
+
"loss": 2.6581,
|
21 |
"step": 4
|
22 |
},
|
23 |
{
|
24 |
"epoch": 0.06,
|
25 |
"learning_rate": 0.0002,
|
26 |
+
"loss": 2.5371,
|
27 |
"step": 6
|
28 |
},
|
29 |
{
|
30 |
"epoch": 0.08,
|
31 |
"learning_rate": 0.0002,
|
32 |
+
"loss": 2.4502,
|
33 |
"step": 8
|
34 |
},
|
35 |
{
|
36 |
"epoch": 0.1,
|
37 |
"learning_rate": 0.0002,
|
38 |
+
"loss": 2.3872,
|
39 |
"step": 10
|
40 |
},
|
41 |
{
|
42 |
"epoch": 0.12,
|
43 |
"learning_rate": 0.0002,
|
44 |
+
"loss": 2.3808,
|
45 |
"step": 12
|
46 |
},
|
47 |
{
|
48 |
"epoch": 0.14,
|
49 |
"learning_rate": 0.0002,
|
50 |
+
"loss": 2.3424,
|
51 |
"step": 14
|
52 |
},
|
53 |
{
|
54 |
"epoch": 0.16,
|
55 |
"learning_rate": 0.0002,
|
56 |
+
"loss": 2.2899,
|
57 |
"step": 16
|
58 |
},
|
59 |
{
|
60 |
"epoch": 0.18,
|
61 |
"learning_rate": 0.0002,
|
62 |
+
"loss": 2.3867,
|
63 |
"step": 18
|
64 |
},
|
65 |
{
|
66 |
"epoch": 0.2,
|
67 |
"learning_rate": 0.0002,
|
68 |
+
"loss": 1.9899,
|
69 |
"step": 20
|
70 |
},
|
71 |
{
|
72 |
"epoch": 0.22,
|
73 |
"learning_rate": 0.0002,
|
74 |
+
"loss": 2.0737,
|
75 |
"step": 22
|
76 |
},
|
77 |
{
|
78 |
"epoch": 0.24,
|
79 |
"learning_rate": 0.0002,
|
80 |
+
"loss": 2.2409,
|
81 |
"step": 24
|
82 |
},
|
83 |
{
|
84 |
"epoch": 0.26,
|
85 |
"learning_rate": 0.0002,
|
86 |
+
"loss": 2.2095,
|
87 |
"step": 26
|
88 |
},
|
89 |
{
|
90 |
"epoch": 0.28,
|
91 |
"learning_rate": 0.0002,
|
92 |
+
"loss": 2.3744,
|
93 |
"step": 28
|
94 |
},
|
95 |
{
|
96 |
"epoch": 0.3,
|
97 |
"learning_rate": 0.0002,
|
98 |
+
"loss": 2.2856,
|
99 |
"step": 30
|
100 |
},
|
101 |
{
|
102 |
"epoch": 0.32,
|
103 |
"learning_rate": 0.0002,
|
104 |
+
"loss": 2.2101,
|
105 |
"step": 32
|
106 |
},
|
107 |
{
|
108 |
"epoch": 0.34,
|
109 |
"learning_rate": 0.0002,
|
110 |
+
"loss": 2.2994,
|
111 |
"step": 34
|
112 |
},
|
113 |
{
|
114 |
"epoch": 0.36,
|
115 |
"learning_rate": 0.0002,
|
116 |
+
"loss": 2.1992,
|
117 |
"step": 36
|
118 |
},
|
119 |
{
|
120 |
"epoch": 0.38,
|
121 |
"learning_rate": 0.0002,
|
122 |
+
"loss": 2.1496,
|
123 |
"step": 38
|
124 |
},
|
125 |
{
|
126 |
"epoch": 0.4,
|
127 |
"learning_rate": 0.0002,
|
128 |
+
"loss": 2.1543,
|
129 |
"step": 40
|
130 |
},
|
131 |
{
|
132 |
"epoch": 0.42,
|
133 |
"learning_rate": 0.0002,
|
134 |
+
"loss": 2.0936,
|
135 |
"step": 42
|
136 |
},
|
137 |
{
|
138 |
"epoch": 0.44,
|
139 |
"learning_rate": 0.0002,
|
140 |
+
"loss": 2.0874,
|
141 |
"step": 44
|
142 |
},
|
143 |
{
|
144 |
"epoch": 0.46,
|
145 |
"learning_rate": 0.0002,
|
146 |
+
"loss": 2.0742,
|
147 |
"step": 46
|
148 |
},
|
149 |
{
|
150 |
"epoch": 0.48,
|
151 |
"learning_rate": 0.0002,
|
152 |
+
"loss": 2.0173,
|
153 |
"step": 48
|
154 |
},
|
155 |
{
|
156 |
"epoch": 0.5,
|
157 |
"learning_rate": 0.0002,
|
158 |
+
"loss": 1.9083,
|
159 |
"step": 50
|
160 |
},
|
161 |
{
|
162 |
"epoch": 0.52,
|
163 |
"learning_rate": 0.0002,
|
164 |
+
"loss": 2.16,
|
165 |
"step": 52
|
166 |
},
|
167 |
{
|
168 |
"epoch": 0.54,
|
169 |
"learning_rate": 0.0002,
|
170 |
+
"loss": 2.176,
|
171 |
"step": 54
|
172 |
},
|
173 |
{
|
174 |
"epoch": 0.56,
|
175 |
"learning_rate": 0.0002,
|
176 |
+
"loss": 2.2623,
|
177 |
"step": 56
|
178 |
},
|
179 |
{
|
180 |
"epoch": 0.58,
|
181 |
"learning_rate": 0.0002,
|
182 |
+
"loss": 2.1539,
|
183 |
"step": 58
|
184 |
},
|
185 |
{
|
186 |
"epoch": 0.6,
|
187 |
"learning_rate": 0.0002,
|
188 |
+
"loss": 2.1608,
|
189 |
"step": 60
|
190 |
},
|
191 |
{
|
192 |
"epoch": 0.62,
|
193 |
"learning_rate": 0.0002,
|
194 |
+
"loss": 2.2897,
|
195 |
"step": 62
|
196 |
},
|
197 |
{
|
198 |
"epoch": 0.64,
|
199 |
"learning_rate": 0.0002,
|
200 |
+
"loss": 2.1425,
|
201 |
"step": 64
|
202 |
},
|
203 |
{
|
204 |
"epoch": 0.66,
|
205 |
"learning_rate": 0.0002,
|
206 |
+
"loss": 2.2373,
|
207 |
"step": 66
|
208 |
},
|
209 |
{
|
210 |
"epoch": 0.68,
|
211 |
"learning_rate": 0.0002,
|
212 |
+
"loss": 2.0494,
|
213 |
"step": 68
|
214 |
},
|
215 |
{
|
216 |
"epoch": 0.7,
|
217 |
"learning_rate": 0.0002,
|
218 |
+
"loss": 2.0225,
|
219 |
"step": 70
|
220 |
},
|
221 |
{
|
222 |
"epoch": 0.72,
|
223 |
"learning_rate": 0.0002,
|
224 |
+
"loss": 2.1813,
|
225 |
"step": 72
|
226 |
},
|
227 |
{
|
228 |
"epoch": 0.74,
|
229 |
"learning_rate": 0.0002,
|
230 |
+
"loss": 2.2155,
|
231 |
"step": 74
|
232 |
},
|
233 |
{
|
234 |
"epoch": 0.76,
|
235 |
"learning_rate": 0.0002,
|
236 |
+
"loss": 2.0284,
|
237 |
"step": 76
|
238 |
},
|
239 |
{
|
240 |
"epoch": 0.78,
|
241 |
"learning_rate": 0.0002,
|
242 |
+
"loss": 2.1858,
|
243 |
"step": 78
|
244 |
},
|
245 |
{
|
246 |
"epoch": 0.8,
|
247 |
"learning_rate": 0.0002,
|
248 |
+
"loss": 2.2214,
|
249 |
"step": 80
|
250 |
},
|
251 |
{
|
252 |
"epoch": 0.82,
|
253 |
"learning_rate": 0.0002,
|
254 |
+
"loss": 2.1439,
|
255 |
"step": 82
|
256 |
},
|
257 |
{
|
258 |
"epoch": 0.84,
|
259 |
"learning_rate": 0.0002,
|
260 |
+
"loss": 2.372,
|
261 |
"step": 84
|
262 |
},
|
263 |
{
|
264 |
"epoch": 0.86,
|
265 |
"learning_rate": 0.0002,
|
266 |
+
"loss": 2.2334,
|
267 |
"step": 86
|
268 |
},
|
269 |
{
|
270 |
"epoch": 0.88,
|
271 |
"learning_rate": 0.0002,
|
272 |
+
"loss": 2.1863,
|
273 |
"step": 88
|
274 |
},
|
275 |
{
|
276 |
"epoch": 0.89,
|
277 |
"learning_rate": 0.0002,
|
278 |
+
"loss": 2.3142,
|
279 |
"step": 90
|
280 |
},
|
281 |
{
|
282 |
"epoch": 0.91,
|
283 |
"learning_rate": 0.0002,
|
284 |
+
"loss": 2.0363,
|
285 |
"step": 92
|
286 |
},
|
287 |
{
|
288 |
"epoch": 0.93,
|
289 |
"learning_rate": 0.0002,
|
290 |
+
"loss": 2.0347,
|
291 |
"step": 94
|
292 |
},
|
293 |
{
|
294 |
"epoch": 0.95,
|
295 |
"learning_rate": 0.0002,
|
296 |
+
"loss": 2.1288,
|
297 |
"step": 96
|
298 |
},
|
299 |
{
|
300 |
"epoch": 0.97,
|
301 |
"learning_rate": 0.0002,
|
302 |
+
"loss": 2.0572,
|
303 |
"step": 98
|
304 |
},
|
305 |
{
|
306 |
"epoch": 0.99,
|
307 |
"learning_rate": 0.0002,
|
308 |
+
"loss": 2.083,
|
309 |
"step": 100
|
310 |
+
},
|
311 |
+
{
|
312 |
+
"epoch": 1.01,
|
313 |
+
"learning_rate": 0.0002,
|
314 |
+
"loss": 2.2397,
|
315 |
+
"step": 102
|
316 |
+
},
|
317 |
+
{
|
318 |
+
"epoch": 1.03,
|
319 |
+
"learning_rate": 0.0002,
|
320 |
+
"loss": 2.1792,
|
321 |
+
"step": 104
|
322 |
+
},
|
323 |
+
{
|
324 |
+
"epoch": 1.05,
|
325 |
+
"learning_rate": 0.0002,
|
326 |
+
"loss": 2.2482,
|
327 |
+
"step": 106
|
328 |
+
},
|
329 |
+
{
|
330 |
+
"epoch": 1.07,
|
331 |
+
"learning_rate": 0.0002,
|
332 |
+
"loss": 2.0806,
|
333 |
+
"step": 108
|
334 |
+
},
|
335 |
+
{
|
336 |
+
"epoch": 1.09,
|
337 |
+
"learning_rate": 0.0002,
|
338 |
+
"loss": 2.16,
|
339 |
+
"step": 110
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 1.11,
|
343 |
+
"learning_rate": 0.0002,
|
344 |
+
"loss": 2.292,
|
345 |
+
"step": 112
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 1.13,
|
349 |
+
"learning_rate": 0.0002,
|
350 |
+
"loss": 2.0676,
|
351 |
+
"step": 114
|
352 |
+
},
|
353 |
+
{
|
354 |
+
"epoch": 1.15,
|
355 |
+
"learning_rate": 0.0002,
|
356 |
+
"loss": 2.2526,
|
357 |
+
"step": 116
|
358 |
+
},
|
359 |
+
{
|
360 |
+
"epoch": 1.17,
|
361 |
+
"learning_rate": 0.0002,
|
362 |
+
"loss": 1.922,
|
363 |
+
"step": 118
|
364 |
+
},
|
365 |
+
{
|
366 |
+
"epoch": 1.19,
|
367 |
+
"learning_rate": 0.0002,
|
368 |
+
"loss": 1.9116,
|
369 |
+
"step": 120
|
370 |
+
},
|
371 |
+
{
|
372 |
+
"epoch": 1.21,
|
373 |
+
"learning_rate": 0.0002,
|
374 |
+
"loss": 1.8892,
|
375 |
+
"step": 122
|
376 |
+
},
|
377 |
+
{
|
378 |
+
"epoch": 1.23,
|
379 |
+
"learning_rate": 0.0002,
|
380 |
+
"loss": 1.9477,
|
381 |
+
"step": 124
|
382 |
+
},
|
383 |
+
{
|
384 |
+
"epoch": 1.25,
|
385 |
+
"learning_rate": 0.0002,
|
386 |
+
"loss": 2.0279,
|
387 |
+
"step": 126
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 1.27,
|
391 |
+
"learning_rate": 0.0002,
|
392 |
+
"loss": 2.2849,
|
393 |
+
"step": 128
|
394 |
+
},
|
395 |
+
{
|
396 |
+
"epoch": 1.29,
|
397 |
+
"learning_rate": 0.0002,
|
398 |
+
"loss": 2.125,
|
399 |
+
"step": 130
|
400 |
+
},
|
401 |
+
{
|
402 |
+
"epoch": 1.31,
|
403 |
+
"learning_rate": 0.0002,
|
404 |
+
"loss": 2.1052,
|
405 |
+
"step": 132
|
406 |
+
},
|
407 |
+
{
|
408 |
+
"epoch": 1.33,
|
409 |
+
"learning_rate": 0.0002,
|
410 |
+
"loss": 2.1967,
|
411 |
+
"step": 134
|
412 |
+
},
|
413 |
+
{
|
414 |
+
"epoch": 1.35,
|
415 |
+
"learning_rate": 0.0002,
|
416 |
+
"loss": 2.1822,
|
417 |
+
"step": 136
|
418 |
+
},
|
419 |
+
{
|
420 |
+
"epoch": 1.37,
|
421 |
+
"learning_rate": 0.0002,
|
422 |
+
"loss": 2.1636,
|
423 |
+
"step": 138
|
424 |
+
},
|
425 |
+
{
|
426 |
+
"epoch": 1.39,
|
427 |
+
"learning_rate": 0.0002,
|
428 |
+
"loss": 2.1071,
|
429 |
+
"step": 140
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 1.41,
|
433 |
+
"learning_rate": 0.0002,
|
434 |
+
"loss": 2.0495,
|
435 |
+
"step": 142
|
436 |
+
},
|
437 |
+
{
|
438 |
+
"epoch": 1.43,
|
439 |
+
"learning_rate": 0.0002,
|
440 |
+
"loss": 2.0297,
|
441 |
+
"step": 144
|
442 |
+
},
|
443 |
+
{
|
444 |
+
"epoch": 1.45,
|
445 |
+
"learning_rate": 0.0002,
|
446 |
+
"loss": 2.029,
|
447 |
+
"step": 146
|
448 |
+
},
|
449 |
+
{
|
450 |
+
"epoch": 1.47,
|
451 |
+
"learning_rate": 0.0002,
|
452 |
+
"loss": 2.0418,
|
453 |
+
"step": 148
|
454 |
+
},
|
455 |
+
{
|
456 |
+
"epoch": 1.49,
|
457 |
+
"learning_rate": 0.0002,
|
458 |
+
"loss": 1.9593,
|
459 |
+
"step": 150
|
460 |
+
},
|
461 |
+
{
|
462 |
+
"epoch": 1.51,
|
463 |
+
"learning_rate": 0.0002,
|
464 |
+
"loss": 2.0313,
|
465 |
+
"step": 152
|
466 |
+
},
|
467 |
+
{
|
468 |
+
"epoch": 1.53,
|
469 |
+
"learning_rate": 0.0002,
|
470 |
+
"loss": 2.1371,
|
471 |
+
"step": 154
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 1.55,
|
475 |
+
"learning_rate": 0.0002,
|
476 |
+
"loss": 2.0588,
|
477 |
+
"step": 156
|
478 |
+
},
|
479 |
+
{
|
480 |
+
"epoch": 1.57,
|
481 |
+
"learning_rate": 0.0002,
|
482 |
+
"loss": 2.1388,
|
483 |
+
"step": 158
|
484 |
+
},
|
485 |
+
{
|
486 |
+
"epoch": 1.59,
|
487 |
+
"learning_rate": 0.0002,
|
488 |
+
"loss": 2.1441,
|
489 |
+
"step": 160
|
490 |
+
},
|
491 |
+
{
|
492 |
+
"epoch": 1.61,
|
493 |
+
"learning_rate": 0.0002,
|
494 |
+
"loss": 2.2533,
|
495 |
+
"step": 162
|
496 |
+
},
|
497 |
+
{
|
498 |
+
"epoch": 1.63,
|
499 |
+
"learning_rate": 0.0002,
|
500 |
+
"loss": 2.2119,
|
501 |
+
"step": 164
|
502 |
+
},
|
503 |
+
{
|
504 |
+
"epoch": 1.65,
|
505 |
+
"learning_rate": 0.0002,
|
506 |
+
"loss": 2.1546,
|
507 |
+
"step": 166
|
508 |
+
},
|
509 |
+
{
|
510 |
+
"epoch": 1.67,
|
511 |
+
"learning_rate": 0.0002,
|
512 |
+
"loss": 1.9094,
|
513 |
+
"step": 168
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 1.69,
|
517 |
+
"learning_rate": 0.0002,
|
518 |
+
"loss": 1.8923,
|
519 |
+
"step": 170
|
520 |
+
},
|
521 |
+
{
|
522 |
+
"epoch": 1.71,
|
523 |
+
"learning_rate": 0.0002,
|
524 |
+
"loss": 2.1945,
|
525 |
+
"step": 172
|
526 |
+
},
|
527 |
+
{
|
528 |
+
"epoch": 1.73,
|
529 |
+
"learning_rate": 0.0002,
|
530 |
+
"loss": 1.9362,
|
531 |
+
"step": 174
|
532 |
+
},
|
533 |
+
{
|
534 |
+
"epoch": 1.75,
|
535 |
+
"learning_rate": 0.0002,
|
536 |
+
"loss": 1.9729,
|
537 |
+
"step": 176
|
538 |
+
},
|
539 |
+
{
|
540 |
+
"epoch": 1.77,
|
541 |
+
"learning_rate": 0.0002,
|
542 |
+
"loss": 2.1719,
|
543 |
+
"step": 178
|
544 |
+
},
|
545 |
+
{
|
546 |
+
"epoch": 1.79,
|
547 |
+
"learning_rate": 0.0002,
|
548 |
+
"loss": 2.1011,
|
549 |
+
"step": 180
|
550 |
+
},
|
551 |
+
{
|
552 |
+
"epoch": 1.81,
|
553 |
+
"learning_rate": 0.0002,
|
554 |
+
"loss": 1.9506,
|
555 |
+
"step": 182
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 1.83,
|
559 |
+
"learning_rate": 0.0002,
|
560 |
+
"loss": 2.1621,
|
561 |
+
"step": 184
|
562 |
+
},
|
563 |
+
{
|
564 |
+
"epoch": 1.85,
|
565 |
+
"learning_rate": 0.0002,
|
566 |
+
"loss": 2.1376,
|
567 |
+
"step": 186
|
568 |
+
},
|
569 |
+
{
|
570 |
+
"epoch": 1.87,
|
571 |
+
"learning_rate": 0.0002,
|
572 |
+
"loss": 2.0654,
|
573 |
+
"step": 188
|
574 |
+
},
|
575 |
+
{
|
576 |
+
"epoch": 1.89,
|
577 |
+
"learning_rate": 0.0002,
|
578 |
+
"loss": 2.13,
|
579 |
+
"step": 190
|
580 |
+
},
|
581 |
+
{
|
582 |
+
"epoch": 1.91,
|
583 |
+
"learning_rate": 0.0002,
|
584 |
+
"loss": 2.2511,
|
585 |
+
"step": 192
|
586 |
+
},
|
587 |
+
{
|
588 |
+
"epoch": 1.93,
|
589 |
+
"learning_rate": 0.0002,
|
590 |
+
"loss": 2.0819,
|
591 |
+
"step": 194
|
592 |
+
},
|
593 |
+
{
|
594 |
+
"epoch": 1.95,
|
595 |
+
"learning_rate": 0.0002,
|
596 |
+
"loss": 1.9898,
|
597 |
+
"step": 196
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 1.97,
|
601 |
+
"learning_rate": 0.0002,
|
602 |
+
"loss": 2.0087,
|
603 |
+
"step": 198
|
604 |
+
},
|
605 |
+
{
|
606 |
+
"epoch": 1.99,
|
607 |
+
"learning_rate": 0.0002,
|
608 |
+
"loss": 2.0657,
|
609 |
+
"step": 200
|
610 |
+
},
|
611 |
+
{
|
612 |
+
"epoch": 2.01,
|
613 |
+
"learning_rate": 0.0002,
|
614 |
+
"loss": 1.9649,
|
615 |
+
"step": 202
|
616 |
+
},
|
617 |
+
{
|
618 |
+
"epoch": 2.03,
|
619 |
+
"learning_rate": 0.0002,
|
620 |
+
"loss": 2.2439,
|
621 |
+
"step": 204
|
622 |
+
},
|
623 |
+
{
|
624 |
+
"epoch": 2.05,
|
625 |
+
"learning_rate": 0.0002,
|
626 |
+
"loss": 2.0759,
|
627 |
+
"step": 206
|
628 |
+
},
|
629 |
+
{
|
630 |
+
"epoch": 2.07,
|
631 |
+
"learning_rate": 0.0002,
|
632 |
+
"loss": 2.0714,
|
633 |
+
"step": 208
|
634 |
+
},
|
635 |
+
{
|
636 |
+
"epoch": 2.09,
|
637 |
+
"learning_rate": 0.0002,
|
638 |
+
"loss": 2.0442,
|
639 |
+
"step": 210
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 2.11,
|
643 |
+
"learning_rate": 0.0002,
|
644 |
+
"loss": 2.0177,
|
645 |
+
"step": 212
|
646 |
+
},
|
647 |
+
{
|
648 |
+
"epoch": 2.13,
|
649 |
+
"learning_rate": 0.0002,
|
650 |
+
"loss": 2.1249,
|
651 |
+
"step": 214
|
652 |
+
},
|
653 |
+
{
|
654 |
+
"epoch": 2.15,
|
655 |
+
"learning_rate": 0.0002,
|
656 |
+
"loss": 2.0028,
|
657 |
+
"step": 216
|
658 |
+
},
|
659 |
+
{
|
660 |
+
"epoch": 2.17,
|
661 |
+
"learning_rate": 0.0002,
|
662 |
+
"loss": 2.0526,
|
663 |
+
"step": 218
|
664 |
+
},
|
665 |
+
{
|
666 |
+
"epoch": 2.19,
|
667 |
+
"learning_rate": 0.0002,
|
668 |
+
"loss": 1.9671,
|
669 |
+
"step": 220
|
670 |
+
},
|
671 |
+
{
|
672 |
+
"epoch": 2.21,
|
673 |
+
"learning_rate": 0.0002,
|
674 |
+
"loss": 1.9116,
|
675 |
+
"step": 222
|
676 |
+
},
|
677 |
+
{
|
678 |
+
"epoch": 2.23,
|
679 |
+
"learning_rate": 0.0002,
|
680 |
+
"loss": 1.9053,
|
681 |
+
"step": 224
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 2.25,
|
685 |
+
"learning_rate": 0.0002,
|
686 |
+
"loss": 1.8237,
|
687 |
+
"step": 226
|
688 |
+
},
|
689 |
+
{
|
690 |
+
"epoch": 2.27,
|
691 |
+
"learning_rate": 0.0002,
|
692 |
+
"loss": 2.0919,
|
693 |
+
"step": 228
|
694 |
+
},
|
695 |
+
{
|
696 |
+
"epoch": 2.29,
|
697 |
+
"learning_rate": 0.0002,
|
698 |
+
"loss": 1.9859,
|
699 |
+
"step": 230
|
700 |
+
},
|
701 |
+
{
|
702 |
+
"epoch": 2.31,
|
703 |
+
"learning_rate": 0.0002,
|
704 |
+
"loss": 2.0955,
|
705 |
+
"step": 232
|
706 |
+
},
|
707 |
+
{
|
708 |
+
"epoch": 2.33,
|
709 |
+
"learning_rate": 0.0002,
|
710 |
+
"loss": 2.0789,
|
711 |
+
"step": 234
|
712 |
+
},
|
713 |
+
{
|
714 |
+
"epoch": 2.35,
|
715 |
+
"learning_rate": 0.0002,
|
716 |
+
"loss": 2.1218,
|
717 |
+
"step": 236
|
718 |
+
},
|
719 |
+
{
|
720 |
+
"epoch": 2.37,
|
721 |
+
"learning_rate": 0.0002,
|
722 |
+
"loss": 2.1313,
|
723 |
+
"step": 238
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 2.39,
|
727 |
+
"learning_rate": 0.0002,
|
728 |
+
"loss": 2.1044,
|
729 |
+
"step": 240
|
730 |
+
},
|
731 |
+
{
|
732 |
+
"epoch": 2.41,
|
733 |
+
"learning_rate": 0.0002,
|
734 |
+
"loss": 2.0011,
|
735 |
+
"step": 242
|
736 |
+
},
|
737 |
+
{
|
738 |
+
"epoch": 2.43,
|
739 |
+
"learning_rate": 0.0002,
|
740 |
+
"loss": 1.8662,
|
741 |
+
"step": 244
|
742 |
+
},
|
743 |
+
{
|
744 |
+
"epoch": 2.45,
|
745 |
+
"learning_rate": 0.0002,
|
746 |
+
"loss": 1.8631,
|
747 |
+
"step": 246
|
748 |
+
},
|
749 |
+
{
|
750 |
+
"epoch": 2.47,
|
751 |
+
"learning_rate": 0.0002,
|
752 |
+
"loss": 2.0623,
|
753 |
+
"step": 248
|
754 |
+
},
|
755 |
+
{
|
756 |
+
"epoch": 2.49,
|
757 |
+
"learning_rate": 0.0002,
|
758 |
+
"loss": 1.8474,
|
759 |
+
"step": 250
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 2.51,
|
763 |
+
"learning_rate": 0.0002,
|
764 |
+
"loss": 2.0302,
|
765 |
+
"step": 252
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 2.53,
|
769 |
+
"learning_rate": 0.0002,
|
770 |
+
"loss": 2.0382,
|
771 |
+
"step": 254
|
772 |
+
},
|
773 |
+
{
|
774 |
+
"epoch": 2.55,
|
775 |
+
"learning_rate": 0.0002,
|
776 |
+
"loss": 2.0559,
|
777 |
+
"step": 256
|
778 |
+
},
|
779 |
+
{
|
780 |
+
"epoch": 2.57,
|
781 |
+
"learning_rate": 0.0002,
|
782 |
+
"loss": 2.0959,
|
783 |
+
"step": 258
|
784 |
+
},
|
785 |
+
{
|
786 |
+
"epoch": 2.59,
|
787 |
+
"learning_rate": 0.0002,
|
788 |
+
"loss": 2.0263,
|
789 |
+
"step": 260
|
790 |
+
},
|
791 |
+
{
|
792 |
+
"epoch": 2.61,
|
793 |
+
"learning_rate": 0.0002,
|
794 |
+
"loss": 2.1544,
|
795 |
+
"step": 262
|
796 |
+
},
|
797 |
+
{
|
798 |
+
"epoch": 2.63,
|
799 |
+
"learning_rate": 0.0002,
|
800 |
+
"loss": 2.2014,
|
801 |
+
"step": 264
|
802 |
+
},
|
803 |
+
{
|
804 |
+
"epoch": 2.65,
|
805 |
+
"learning_rate": 0.0002,
|
806 |
+
"loss": 2.1269,
|
807 |
+
"step": 266
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 2.67,
|
811 |
+
"learning_rate": 0.0002,
|
812 |
+
"loss": 2.0373,
|
813 |
+
"step": 268
|
814 |
+
},
|
815 |
+
{
|
816 |
+
"epoch": 2.68,
|
817 |
+
"learning_rate": 0.0002,
|
818 |
+
"loss": 1.9107,
|
819 |
+
"step": 270
|
820 |
+
},
|
821 |
+
{
|
822 |
+
"epoch": 2.7,
|
823 |
+
"learning_rate": 0.0002,
|
824 |
+
"loss": 1.9081,
|
825 |
+
"step": 272
|
826 |
+
},
|
827 |
+
{
|
828 |
+
"epoch": 2.72,
|
829 |
+
"learning_rate": 0.0002,
|
830 |
+
"loss": 1.9295,
|
831 |
+
"step": 274
|
832 |
+
},
|
833 |
+
{
|
834 |
+
"epoch": 2.74,
|
835 |
+
"learning_rate": 0.0002,
|
836 |
+
"loss": 1.7845,
|
837 |
+
"step": 276
|
838 |
+
},
|
839 |
+
{
|
840 |
+
"epoch": 2.76,
|
841 |
+
"learning_rate": 0.0002,
|
842 |
+
"loss": 2.0504,
|
843 |
+
"step": 278
|
844 |
+
},
|
845 |
+
{
|
846 |
+
"epoch": 2.78,
|
847 |
+
"learning_rate": 0.0002,
|
848 |
+
"loss": 2.2585,
|
849 |
+
"step": 280
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 2.8,
|
853 |
+
"learning_rate": 0.0002,
|
854 |
+
"loss": 2.1938,
|
855 |
+
"step": 282
|
856 |
+
},
|
857 |
+
{
|
858 |
+
"epoch": 2.82,
|
859 |
+
"learning_rate": 0.0002,
|
860 |
+
"loss": 2.0515,
|
861 |
+
"step": 284
|
862 |
+
},
|
863 |
+
{
|
864 |
+
"epoch": 2.84,
|
865 |
+
"learning_rate": 0.0002,
|
866 |
+
"loss": 2.076,
|
867 |
+
"step": 286
|
868 |
+
},
|
869 |
+
{
|
870 |
+
"epoch": 2.86,
|
871 |
+
"learning_rate": 0.0002,
|
872 |
+
"loss": 2.105,
|
873 |
+
"step": 288
|
874 |
+
},
|
875 |
+
{
|
876 |
+
"epoch": 2.88,
|
877 |
+
"learning_rate": 0.0002,
|
878 |
+
"loss": 2.165,
|
879 |
+
"step": 290
|
880 |
+
},
|
881 |
+
{
|
882 |
+
"epoch": 2.9,
|
883 |
+
"learning_rate": 0.0002,
|
884 |
+
"loss": 2.1052,
|
885 |
+
"step": 292
|
886 |
+
},
|
887 |
+
{
|
888 |
+
"epoch": 2.92,
|
889 |
+
"learning_rate": 0.0002,
|
890 |
+
"loss": 1.9451,
|
891 |
+
"step": 294
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 2.94,
|
895 |
+
"learning_rate": 0.0002,
|
896 |
+
"loss": 1.8612,
|
897 |
+
"step": 296
|
898 |
+
},
|
899 |
+
{
|
900 |
+
"epoch": 2.96,
|
901 |
+
"learning_rate": 0.0002,
|
902 |
+
"loss": 1.8693,
|
903 |
+
"step": 298
|
904 |
+
},
|
905 |
+
{
|
906 |
+
"epoch": 2.98,
|
907 |
+
"learning_rate": 0.0002,
|
908 |
+
"loss": 2.0716,
|
909 |
+
"step": 300
|
910 |
}
|
911 |
],
|
912 |
"logging_steps": 2,
|
913 |
+
"max_steps": 300,
|
914 |
"num_input_tokens_seen": 0,
|
915 |
+
"num_train_epochs": 3,
|
916 |
"save_steps": 500,
|
917 |
+
"total_flos": 2.583887994465485e+16,
|
918 |
"train_batch_size": 1,
|
919 |
"trial_name": null,
|
920 |
"trial_params": null
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 4728
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:be6cbfc3a3bc96845390c55d1ca1cfa1ff08278577668e45edff0ecb1a4f119c
|
3 |
size 4728
|