Amr Keleg
commited on
Commit
•
5a59f8d
1
Parent(s):
7cb5a13
Track the final version of the model
Browse files- optimizer.pt +1 -1
- pytorch_model.bin +1 -1
- rng_state.pth +2 -2
- scheduler.pt +1 -1
- trainer_state.json +262 -40
- training_args.bin +1 -1
optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1298159621
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:dc0454bac094eb9bcfc5225a1906cb99c102336172ef9e7bafc8054a4c830871
|
3 |
size 1298159621
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 649093613
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7ff85dba579bca3aa2ae6faff5086f9422af9dfca51f08c249c6ddf177bc7771
|
3 |
size 649093613
|
rng_state.pth
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 sha256:4e69904db53ae4ac2e537fe1ecca98d908ffe407b7bf84009f7869243dbb2d8b
|
3 |
+
size 14575
|
scheduler.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 627
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4f7f223e541d8ca8282c09e6f4666d154f61e406961790c043b389c95e9eedcc
|
3 |
size 627
|
trainer_state.json
CHANGED
@@ -2,75 +2,297 @@
|
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
"epoch": 3.0,
|
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.
|
12 |
-
"eval_loss": 0.
|
13 |
-
"eval_runtime":
|
14 |
-
"eval_samples_per_second":
|
15 |
-
"eval_steps_per_second":
|
16 |
"step": 100
|
17 |
},
|
18 |
{
|
19 |
-
"epoch": 0.
|
20 |
-
"eval_loss": 0.
|
21 |
-
"eval_runtime":
|
22 |
-
"eval_samples_per_second":
|
23 |
-
"eval_steps_per_second":
|
24 |
"step": 200
|
25 |
},
|
26 |
{
|
27 |
-
"epoch":
|
28 |
-
"eval_loss": 0.
|
29 |
-
"eval_runtime":
|
30 |
-
"eval_samples_per_second":
|
31 |
-
"eval_steps_per_second":
|
32 |
"step": 300
|
33 |
},
|
34 |
{
|
35 |
-
"epoch":
|
36 |
-
"eval_loss": 0.
|
37 |
-
"eval_runtime":
|
38 |
-
"eval_samples_per_second":
|
39 |
-
"eval_steps_per_second":
|
40 |
"step": 400
|
41 |
},
|
42 |
{
|
43 |
-
"epoch":
|
44 |
-
"learning_rate":
|
45 |
-
"loss": 0.
|
46 |
"step": 500
|
47 |
},
|
48 |
{
|
49 |
-
"epoch":
|
50 |
-
"eval_loss": 0.
|
51 |
-
"eval_runtime":
|
52 |
-
"eval_samples_per_second":
|
53 |
-
"eval_steps_per_second":
|
54 |
"step": 500
|
55 |
},
|
56 |
{
|
57 |
-
"epoch":
|
58 |
-
"eval_loss": 0.
|
59 |
-
"eval_runtime":
|
60 |
-
"eval_samples_per_second":
|
61 |
-
"eval_steps_per_second":
|
62 |
"step": 600
|
63 |
},
|
64 |
{
|
65 |
-
"epoch":
|
66 |
-
"eval_loss": 0.
|
67 |
-
"eval_runtime":
|
68 |
-
"eval_samples_per_second":
|
69 |
-
"eval_steps_per_second":
|
70 |
"step": 700
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
71 |
}
|
72 |
],
|
73 |
-
"max_steps":
|
74 |
"num_train_epochs": 3,
|
75 |
"total_flos": 6635087188033536.0,
|
76 |
"trial_name": null,
|
|
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
"epoch": 3.0,
|
5 |
+
"global_step": 3174,
|
6 |
"is_hyper_param_search": false,
|
7 |
"is_local_process_zero": true,
|
8 |
"is_world_process_zero": true,
|
9 |
"log_history": [
|
10 |
{
|
11 |
+
"epoch": 0.09,
|
12 |
+
"eval_loss": 0.291690856218338,
|
13 |
+
"eval_runtime": 11.8699,
|
14 |
+
"eval_samples_per_second": 94.019,
|
15 |
+
"eval_steps_per_second": 11.795,
|
16 |
"step": 100
|
17 |
},
|
18 |
{
|
19 |
+
"epoch": 0.19,
|
20 |
+
"eval_loss": 0.2913290560245514,
|
21 |
+
"eval_runtime": 11.9204,
|
22 |
+
"eval_samples_per_second": 93.621,
|
23 |
+
"eval_steps_per_second": 11.745,
|
24 |
"step": 200
|
25 |
},
|
26 |
{
|
27 |
+
"epoch": 0.28,
|
28 |
+
"eval_loss": 0.27973252534866333,
|
29 |
+
"eval_runtime": 11.9409,
|
30 |
+
"eval_samples_per_second": 93.46,
|
31 |
+
"eval_steps_per_second": 11.724,
|
32 |
"step": 300
|
33 |
},
|
34 |
{
|
35 |
+
"epoch": 0.38,
|
36 |
+
"eval_loss": 0.25020790100097656,
|
37 |
+
"eval_runtime": 11.9448,
|
38 |
+
"eval_samples_per_second": 93.43,
|
39 |
+
"eval_steps_per_second": 11.721,
|
40 |
"step": 400
|
41 |
},
|
42 |
{
|
43 |
+
"epoch": 0.47,
|
44 |
+
"learning_rate": 4.2123503465658476e-05,
|
45 |
+
"loss": 0.3052,
|
46 |
"step": 500
|
47 |
},
|
48 |
{
|
49 |
+
"epoch": 0.47,
|
50 |
+
"eval_loss": 0.2535267770290375,
|
51 |
+
"eval_runtime": 11.9395,
|
52 |
+
"eval_samples_per_second": 93.471,
|
53 |
+
"eval_steps_per_second": 11.726,
|
54 |
"step": 500
|
55 |
},
|
56 |
{
|
57 |
+
"epoch": 0.57,
|
58 |
+
"eval_loss": 0.29139193892478943,
|
59 |
+
"eval_runtime": 11.942,
|
60 |
+
"eval_samples_per_second": 93.452,
|
61 |
+
"eval_steps_per_second": 11.723,
|
62 |
"step": 600
|
63 |
},
|
64 |
{
|
65 |
+
"epoch": 0.66,
|
66 |
+
"eval_loss": 0.2721957862377167,
|
67 |
+
"eval_runtime": 11.936,
|
68 |
+
"eval_samples_per_second": 93.498,
|
69 |
+
"eval_steps_per_second": 11.729,
|
70 |
"step": 700
|
71 |
+
},
|
72 |
+
{
|
73 |
+
"epoch": 0.76,
|
74 |
+
"eval_loss": 0.24325571954250336,
|
75 |
+
"eval_runtime": 11.9084,
|
76 |
+
"eval_samples_per_second": 93.716,
|
77 |
+
"eval_steps_per_second": 11.756,
|
78 |
+
"step": 800
|
79 |
+
},
|
80 |
+
{
|
81 |
+
"epoch": 0.85,
|
82 |
+
"eval_loss": 0.2575836777687073,
|
83 |
+
"eval_runtime": 11.9447,
|
84 |
+
"eval_samples_per_second": 93.43,
|
85 |
+
"eval_steps_per_second": 11.721,
|
86 |
+
"step": 900
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.95,
|
90 |
+
"learning_rate": 3.424700693131695e-05,
|
91 |
+
"loss": 0.2236,
|
92 |
+
"step": 1000
|
93 |
+
},
|
94 |
+
{
|
95 |
+
"epoch": 0.95,
|
96 |
+
"eval_loss": 0.2607925832271576,
|
97 |
+
"eval_runtime": 11.9435,
|
98 |
+
"eval_samples_per_second": 93.44,
|
99 |
+
"eval_steps_per_second": 11.722,
|
100 |
+
"step": 1000
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 1.04,
|
104 |
+
"eval_loss": 0.2817261815071106,
|
105 |
+
"eval_runtime": 11.9296,
|
106 |
+
"eval_samples_per_second": 93.549,
|
107 |
+
"eval_steps_per_second": 11.735,
|
108 |
+
"step": 1100
|
109 |
+
},
|
110 |
+
{
|
111 |
+
"epoch": 1.13,
|
112 |
+
"eval_loss": 0.24339333176612854,
|
113 |
+
"eval_runtime": 11.9423,
|
114 |
+
"eval_samples_per_second": 93.449,
|
115 |
+
"eval_steps_per_second": 11.723,
|
116 |
+
"step": 1200
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"epoch": 1.23,
|
120 |
+
"eval_loss": 0.24544629454612732,
|
121 |
+
"eval_runtime": 11.9445,
|
122 |
+
"eval_samples_per_second": 93.432,
|
123 |
+
"eval_steps_per_second": 11.721,
|
124 |
+
"step": 1300
|
125 |
+
},
|
126 |
+
{
|
127 |
+
"epoch": 1.32,
|
128 |
+
"eval_loss": 0.23857346177101135,
|
129 |
+
"eval_runtime": 11.9479,
|
130 |
+
"eval_samples_per_second": 93.406,
|
131 |
+
"eval_steps_per_second": 11.718,
|
132 |
+
"step": 1400
|
133 |
+
},
|
134 |
+
{
|
135 |
+
"epoch": 1.42,
|
136 |
+
"learning_rate": 2.637051039697543e-05,
|
137 |
+
"loss": 0.1449,
|
138 |
+
"step": 1500
|
139 |
+
},
|
140 |
+
{
|
141 |
+
"epoch": 1.42,
|
142 |
+
"eval_loss": 0.2612239718437195,
|
143 |
+
"eval_runtime": 11.947,
|
144 |
+
"eval_samples_per_second": 93.413,
|
145 |
+
"eval_steps_per_second": 11.718,
|
146 |
+
"step": 1500
|
147 |
+
},
|
148 |
+
{
|
149 |
+
"epoch": 1.51,
|
150 |
+
"eval_loss": 0.264009565114975,
|
151 |
+
"eval_runtime": 11.9446,
|
152 |
+
"eval_samples_per_second": 93.431,
|
153 |
+
"eval_steps_per_second": 11.721,
|
154 |
+
"step": 1600
|
155 |
+
},
|
156 |
+
{
|
157 |
+
"epoch": 1.61,
|
158 |
+
"eval_loss": 0.23957186937332153,
|
159 |
+
"eval_runtime": 11.9459,
|
160 |
+
"eval_samples_per_second": 93.421,
|
161 |
+
"eval_steps_per_second": 11.72,
|
162 |
+
"step": 1700
|
163 |
+
},
|
164 |
+
{
|
165 |
+
"epoch": 1.7,
|
166 |
+
"eval_loss": 0.23903459310531616,
|
167 |
+
"eval_runtime": 11.9432,
|
168 |
+
"eval_samples_per_second": 93.443,
|
169 |
+
"eval_steps_per_second": 11.722,
|
170 |
+
"step": 1800
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 1.8,
|
174 |
+
"eval_loss": 0.22416594624519348,
|
175 |
+
"eval_runtime": 11.944,
|
176 |
+
"eval_samples_per_second": 93.436,
|
177 |
+
"eval_steps_per_second": 11.721,
|
178 |
+
"step": 1900
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 1.89,
|
182 |
+
"learning_rate": 1.8494013862633903e-05,
|
183 |
+
"loss": 0.1408,
|
184 |
+
"step": 2000
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 1.89,
|
188 |
+
"eval_loss": 0.2341969758272171,
|
189 |
+
"eval_runtime": 11.9399,
|
190 |
+
"eval_samples_per_second": 93.468,
|
191 |
+
"eval_steps_per_second": 11.725,
|
192 |
+
"step": 2000
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 1.98,
|
196 |
+
"eval_loss": 0.23950397968292236,
|
197 |
+
"eval_runtime": 11.9383,
|
198 |
+
"eval_samples_per_second": 93.481,
|
199 |
+
"eval_steps_per_second": 11.727,
|
200 |
+
"step": 2100
|
201 |
+
},
|
202 |
+
{
|
203 |
+
"epoch": 2.08,
|
204 |
+
"eval_loss": 0.285567045211792,
|
205 |
+
"eval_runtime": 11.9279,
|
206 |
+
"eval_samples_per_second": 93.562,
|
207 |
+
"eval_steps_per_second": 11.737,
|
208 |
+
"step": 2200
|
209 |
+
},
|
210 |
+
{
|
211 |
+
"epoch": 2.17,
|
212 |
+
"eval_loss": 0.2656622529029846,
|
213 |
+
"eval_runtime": 11.9385,
|
214 |
+
"eval_samples_per_second": 93.479,
|
215 |
+
"eval_steps_per_second": 11.727,
|
216 |
+
"step": 2300
|
217 |
+
},
|
218 |
+
{
|
219 |
+
"epoch": 2.27,
|
220 |
+
"eval_loss": 0.25764182209968567,
|
221 |
+
"eval_runtime": 11.9434,
|
222 |
+
"eval_samples_per_second": 93.44,
|
223 |
+
"eval_steps_per_second": 11.722,
|
224 |
+
"step": 2400
|
225 |
+
},
|
226 |
+
{
|
227 |
+
"epoch": 2.36,
|
228 |
+
"learning_rate": 1.0617517328292375e-05,
|
229 |
+
"loss": 0.0893,
|
230 |
+
"step": 2500
|
231 |
+
},
|
232 |
+
{
|
233 |
+
"epoch": 2.36,
|
234 |
+
"eval_loss": 0.26884153485298157,
|
235 |
+
"eval_runtime": 11.9443,
|
236 |
+
"eval_samples_per_second": 93.434,
|
237 |
+
"eval_steps_per_second": 11.721,
|
238 |
+
"step": 2500
|
239 |
+
},
|
240 |
+
{
|
241 |
+
"epoch": 2.46,
|
242 |
+
"eval_loss": 0.2657739520072937,
|
243 |
+
"eval_runtime": 11.9455,
|
244 |
+
"eval_samples_per_second": 93.425,
|
245 |
+
"eval_steps_per_second": 11.72,
|
246 |
+
"step": 2600
|
247 |
+
},
|
248 |
+
{
|
249 |
+
"epoch": 2.55,
|
250 |
+
"eval_loss": 0.2614665925502777,
|
251 |
+
"eval_runtime": 11.9422,
|
252 |
+
"eval_samples_per_second": 93.45,
|
253 |
+
"eval_steps_per_second": 11.723,
|
254 |
+
"step": 2700
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 2.65,
|
258 |
+
"eval_loss": 0.27043506503105164,
|
259 |
+
"eval_runtime": 11.9429,
|
260 |
+
"eval_samples_per_second": 93.445,
|
261 |
+
"eval_steps_per_second": 11.722,
|
262 |
+
"step": 2800
|
263 |
+
},
|
264 |
+
{
|
265 |
+
"epoch": 2.74,
|
266 |
+
"eval_loss": 0.2741823196411133,
|
267 |
+
"eval_runtime": 11.9422,
|
268 |
+
"eval_samples_per_second": 93.45,
|
269 |
+
"eval_steps_per_second": 11.723,
|
270 |
+
"step": 2900
|
271 |
+
},
|
272 |
+
{
|
273 |
+
"epoch": 2.84,
|
274 |
+
"learning_rate": 2.741020793950851e-06,
|
275 |
+
"loss": 0.0668,
|
276 |
+
"step": 3000
|
277 |
+
},
|
278 |
+
{
|
279 |
+
"epoch": 2.84,
|
280 |
+
"eval_loss": 0.2574303448200226,
|
281 |
+
"eval_runtime": 11.9428,
|
282 |
+
"eval_samples_per_second": 93.445,
|
283 |
+
"eval_steps_per_second": 11.723,
|
284 |
+
"step": 3000
|
285 |
+
},
|
286 |
+
{
|
287 |
+
"epoch": 2.93,
|
288 |
+
"eval_loss": 0.2549898028373718,
|
289 |
+
"eval_runtime": 11.9436,
|
290 |
+
"eval_samples_per_second": 93.439,
|
291 |
+
"eval_steps_per_second": 11.722,
|
292 |
+
"step": 3100
|
293 |
}
|
294 |
],
|
295 |
+
"max_steps": 3174,
|
296 |
"num_train_epochs": 3,
|
297 |
"total_flos": 6635087188033536.0,
|
298 |
"trial_name": null,
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 3515
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4bb3243e713170949466998e8a474a592be4c2bf750567ed9f5e8fd61388dc92
|
3 |
size 3515
|