Upload folder using huggingface_hub
Browse files- .ipynb_checkpoints/config-checkpoint.json +61 -0
- .ipynb_checkpoints/generation_config-checkpoint.json +15 -0
- model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +83 -257
- training_args.bin +2 -2
.ipynb_checkpoints/config-checkpoint.json
ADDED
@@ -0,0 +1,61 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "Sunbird/sunbird-en-mul",
|
3 |
+
"activation_dropout": 0.0,
|
4 |
+
"activation_function": "swish",
|
5 |
+
"add_bias_logits": false,
|
6 |
+
"add_final_layer_norm": false,
|
7 |
+
"architectures": [
|
8 |
+
"MarianMTModel"
|
9 |
+
],
|
10 |
+
"attention_dropout": 0.0,
|
11 |
+
"bad_words_ids": [
|
12 |
+
[
|
13 |
+
64109
|
14 |
+
]
|
15 |
+
],
|
16 |
+
"bos_token_id": 0,
|
17 |
+
"classif_dropout": 0.0,
|
18 |
+
"classifier_dropout": 0.0,
|
19 |
+
"d_model": 512,
|
20 |
+
"decoder_attention_heads": 8,
|
21 |
+
"decoder_ffn_dim": 2048,
|
22 |
+
"decoder_layerdrop": 0.0,
|
23 |
+
"decoder_layers": 6,
|
24 |
+
"decoder_start_token_id": 64109,
|
25 |
+
"decoder_vocab_size": 64110,
|
26 |
+
"dropout": 0.1,
|
27 |
+
"encoder_attention_heads": 8,
|
28 |
+
"encoder_ffn_dim": 2048,
|
29 |
+
"encoder_layerdrop": 0.0,
|
30 |
+
"encoder_layers": 6,
|
31 |
+
"eos_token_id": 0,
|
32 |
+
"extra_pos_embeddings": 64110,
|
33 |
+
"forced_eos_token_id": 0,
|
34 |
+
"id2label": {
|
35 |
+
"0": "LABEL_0",
|
36 |
+
"1": "LABEL_1",
|
37 |
+
"2": "LABEL_2"
|
38 |
+
},
|
39 |
+
"init_std": 0.02,
|
40 |
+
"is_encoder_decoder": true,
|
41 |
+
"label2id": {
|
42 |
+
"LABEL_0": 0,
|
43 |
+
"LABEL_1": 1,
|
44 |
+
"LABEL_2": 2
|
45 |
+
},
|
46 |
+
"max_length": 512,
|
47 |
+
"max_position_embeddings": 512,
|
48 |
+
"model_type": "marian",
|
49 |
+
"normalize_before": false,
|
50 |
+
"normalize_embedding": false,
|
51 |
+
"num_beams": 4,
|
52 |
+
"num_hidden_layers": 6,
|
53 |
+
"pad_token_id": 64109,
|
54 |
+
"scale_embedding": true,
|
55 |
+
"share_encoder_decoder_embeddings": true,
|
56 |
+
"static_position_embeddings": true,
|
57 |
+
"torch_dtype": "float32",
|
58 |
+
"transformers_version": "4.42.3",
|
59 |
+
"use_cache": true,
|
60 |
+
"vocab_size": 64110
|
61 |
+
}
|
.ipynb_checkpoints/generation_config-checkpoint.json
ADDED
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bad_words_ids": [
|
3 |
+
[
|
4 |
+
64109
|
5 |
+
]
|
6 |
+
],
|
7 |
+
"bos_token_id": 0,
|
8 |
+
"decoder_start_token_id": 64109,
|
9 |
+
"eos_token_id": 0,
|
10 |
+
"forced_eos_token_id": 0,
|
11 |
+
"max_length": 512,
|
12 |
+
"num_beams": 4,
|
13 |
+
"pad_token_id": 64109,
|
14 |
+
"transformers_version": "4.42.3"
|
15 |
+
}
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 308136760
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8a671b44e845c38b4c922359787150f67fe827da60cb8013c57f1e95b8979bda
|
3 |
size 308136760
|
optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 615914106
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:314c29c585a428a949c96d0ad30940975e42977868f743cb743eca8a7ef11d5e
|
3 |
size 615914106
|
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 sha256:944e670087748feb93d27d6f02cd19e3a0d45a2067db8c91e6ba9d7d1d9530ee
|
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 sha256:06585fe70176c3147fca57438a795025afd40446eb83631c3d5f29dcfab8b5fd
|
3 |
size 1064
|
trainer_state.json
CHANGED
@@ -1,292 +1,118 @@
|
|
1 |
{
|
2 |
-
"best_metric":
|
3 |
-
"best_model_checkpoint": "
|
4 |
-
"epoch":
|
5 |
"eval_steps": 50,
|
6 |
-
"global_step":
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
-
"epoch": 0
|
13 |
-
"eval_bleu": 0.
|
14 |
-
"eval_gen_len": 20.
|
15 |
-
"eval_loss":
|
16 |
-
"eval_runtime":
|
17 |
-
"eval_samples_per_second": 25.
|
18 |
-
"eval_steps_per_second": 0.
|
19 |
-
"step":
|
20 |
},
|
21 |
{
|
22 |
"epoch": 2.0,
|
23 |
-
"eval_bleu": 0.
|
24 |
-
"eval_gen_len": 20.
|
25 |
-
"eval_loss":
|
26 |
-
"eval_runtime":
|
27 |
-
"eval_samples_per_second":
|
28 |
-
"eval_steps_per_second": 0.
|
29 |
-
"step":
|
30 |
},
|
31 |
{
|
32 |
-
"epoch":
|
33 |
-
"eval_bleu": 0.
|
34 |
-
"eval_gen_len": 20.
|
35 |
-
"eval_loss":
|
36 |
-
"eval_runtime":
|
37 |
-
"eval_samples_per_second":
|
38 |
-
"eval_steps_per_second": 0.
|
39 |
-
"step":
|
40 |
},
|
41 |
{
|
42 |
"epoch": 4.0,
|
43 |
-
"eval_bleu": 0.
|
44 |
-
"eval_gen_len": 20.
|
45 |
-
"eval_loss":
|
46 |
-
"eval_runtime":
|
47 |
-
"eval_samples_per_second":
|
48 |
-
"eval_steps_per_second": 0.
|
49 |
-
"step":
|
50 |
},
|
51 |
{
|
52 |
-
"epoch":
|
53 |
-
"eval_bleu": 0.
|
54 |
-
"eval_gen_len": 20.
|
55 |
-
"eval_loss":
|
56 |
-
"eval_runtime":
|
57 |
-
"eval_samples_per_second":
|
58 |
-
"eval_steps_per_second": 0.
|
59 |
-
"step":
|
60 |
},
|
61 |
{
|
62 |
"epoch": 6.0,
|
63 |
-
"eval_bleu": 0.
|
64 |
-
"eval_gen_len": 20.
|
65 |
-
"eval_loss":
|
66 |
-
"eval_runtime":
|
67 |
-
"eval_samples_per_second":
|
68 |
-
"eval_steps_per_second": 0.
|
69 |
-
"step":
|
70 |
},
|
71 |
{
|
72 |
-
"epoch":
|
73 |
-
"eval_bleu": 0.
|
74 |
-
"eval_gen_len": 20.
|
75 |
-
"eval_loss":
|
76 |
-
"eval_runtime":
|
77 |
-
"eval_samples_per_second":
|
78 |
-
"eval_steps_per_second": 0.
|
79 |
-
"step":
|
80 |
},
|
81 |
{
|
82 |
"epoch": 8.0,
|
83 |
-
"eval_bleu": 0.
|
84 |
-
"eval_gen_len": 20.
|
85 |
-
"eval_loss": 4.
|
86 |
-
"eval_runtime":
|
87 |
-
"eval_samples_per_second":
|
88 |
-
"eval_steps_per_second": 0.
|
89 |
-
"step":
|
90 |
},
|
91 |
{
|
92 |
-
"epoch":
|
93 |
-
"eval_bleu": 0.
|
94 |
-
"eval_gen_len": 20.
|
95 |
-
"eval_loss": 4.
|
96 |
-
"eval_runtime":
|
97 |
-
"eval_samples_per_second":
|
98 |
-
"eval_steps_per_second": 0.
|
99 |
-
"step":
|
100 |
},
|
101 |
{
|
102 |
"epoch": 10.0,
|
103 |
-
"eval_bleu": 0.
|
104 |
-
"eval_gen_len": 20.
|
105 |
-
"eval_loss": 4.
|
106 |
-
"eval_runtime":
|
107 |
-
"eval_samples_per_second":
|
108 |
-
"eval_steps_per_second": 0.
|
109 |
-
"step":
|
110 |
-
},
|
111 |
-
{
|
112 |
-
"epoch": 10.987341772151899,
|
113 |
-
"eval_bleu": 0.7619,
|
114 |
-
"eval_gen_len": 20.4121,
|
115 |
-
"eval_loss": 4.748700141906738,
|
116 |
-
"eval_runtime": 479.5702,
|
117 |
-
"eval_samples_per_second": 25.35,
|
118 |
-
"eval_steps_per_second": 0.792,
|
119 |
-
"step": 434
|
120 |
-
},
|
121 |
-
{
|
122 |
-
"epoch": 12.0,
|
123 |
-
"eval_bleu": 0.7865,
|
124 |
-
"eval_gen_len": 20.3477,
|
125 |
-
"eval_loss": 4.832670211791992,
|
126 |
-
"eval_runtime": 473.4727,
|
127 |
-
"eval_samples_per_second": 25.676,
|
128 |
-
"eval_steps_per_second": 0.803,
|
129 |
-
"step": 474
|
130 |
-
},
|
131 |
-
{
|
132 |
-
"epoch": 12.658227848101266,
|
133 |
-
"grad_norm": 1.2026784420013428,
|
134 |
-
"learning_rate": 1.1452991452991454e-05,
|
135 |
-
"loss": 0.1585,
|
136 |
-
"step": 500
|
137 |
-
},
|
138 |
-
{
|
139 |
-
"epoch": 12.987341772151899,
|
140 |
-
"eval_bleu": 0.7909,
|
141 |
-
"eval_gen_len": 20.307,
|
142 |
-
"eval_loss": 4.844118595123291,
|
143 |
-
"eval_runtime": 475.2982,
|
144 |
-
"eval_samples_per_second": 25.578,
|
145 |
-
"eval_steps_per_second": 0.799,
|
146 |
-
"step": 513
|
147 |
-
},
|
148 |
-
{
|
149 |
-
"epoch": 14.0,
|
150 |
-
"eval_bleu": 0.8004,
|
151 |
-
"eval_gen_len": 20.3038,
|
152 |
-
"eval_loss": 4.876136302947998,
|
153 |
-
"eval_runtime": 475.2478,
|
154 |
-
"eval_samples_per_second": 25.58,
|
155 |
-
"eval_steps_per_second": 0.8,
|
156 |
-
"step": 553
|
157 |
-
},
|
158 |
-
{
|
159 |
-
"epoch": 14.987341772151899,
|
160 |
-
"eval_bleu": 0.7823,
|
161 |
-
"eval_gen_len": 20.2747,
|
162 |
-
"eval_loss": 4.92393684387207,
|
163 |
-
"eval_runtime": 476.8033,
|
164 |
-
"eval_samples_per_second": 25.497,
|
165 |
-
"eval_steps_per_second": 0.797,
|
166 |
-
"step": 592
|
167 |
-
},
|
168 |
-
{
|
169 |
-
"epoch": 16.0,
|
170 |
-
"eval_bleu": 0.79,
|
171 |
-
"eval_gen_len": 20.226,
|
172 |
-
"eval_loss": 4.880424976348877,
|
173 |
-
"eval_runtime": 474.5946,
|
174 |
-
"eval_samples_per_second": 25.616,
|
175 |
-
"eval_steps_per_second": 0.801,
|
176 |
-
"step": 632
|
177 |
-
},
|
178 |
-
{
|
179 |
-
"epoch": 16.9873417721519,
|
180 |
-
"eval_bleu": 0.7821,
|
181 |
-
"eval_gen_len": 20.3653,
|
182 |
-
"eval_loss": 4.9470534324646,
|
183 |
-
"eval_runtime": 477.2511,
|
184 |
-
"eval_samples_per_second": 25.473,
|
185 |
-
"eval_steps_per_second": 0.796,
|
186 |
-
"step": 671
|
187 |
-
},
|
188 |
-
{
|
189 |
-
"epoch": 18.0,
|
190 |
-
"eval_bleu": 0.7876,
|
191 |
-
"eval_gen_len": 20.3285,
|
192 |
-
"eval_loss": 4.9639482498168945,
|
193 |
-
"eval_runtime": 477.3388,
|
194 |
-
"eval_samples_per_second": 25.468,
|
195 |
-
"eval_steps_per_second": 0.796,
|
196 |
-
"step": 711
|
197 |
-
},
|
198 |
-
{
|
199 |
-
"epoch": 18.9873417721519,
|
200 |
-
"eval_bleu": 0.7895,
|
201 |
-
"eval_gen_len": 20.3364,
|
202 |
-
"eval_loss": 4.970686912536621,
|
203 |
-
"eval_runtime": 476.4474,
|
204 |
-
"eval_samples_per_second": 25.516,
|
205 |
-
"eval_steps_per_second": 0.798,
|
206 |
-
"step": 750
|
207 |
-
},
|
208 |
-
{
|
209 |
-
"epoch": 20.0,
|
210 |
-
"eval_bleu": 0.784,
|
211 |
-
"eval_gen_len": 20.3641,
|
212 |
-
"eval_loss": 4.984948635101318,
|
213 |
-
"eval_runtime": 477.378,
|
214 |
-
"eval_samples_per_second": 25.466,
|
215 |
-
"eval_steps_per_second": 0.796,
|
216 |
-
"step": 790
|
217 |
-
},
|
218 |
-
{
|
219 |
-
"epoch": 20.9873417721519,
|
220 |
-
"eval_bleu": 0.7838,
|
221 |
-
"eval_gen_len": 20.3852,
|
222 |
-
"eval_loss": 5.011585712432861,
|
223 |
-
"eval_runtime": 477.6336,
|
224 |
-
"eval_samples_per_second": 25.453,
|
225 |
-
"eval_steps_per_second": 0.796,
|
226 |
-
"step": 829
|
227 |
-
},
|
228 |
-
{
|
229 |
-
"epoch": 22.0,
|
230 |
-
"eval_bleu": 0.7874,
|
231 |
-
"eval_gen_len": 20.3104,
|
232 |
-
"eval_loss": 5.026498794555664,
|
233 |
-
"eval_runtime": 477.8381,
|
234 |
-
"eval_samples_per_second": 25.442,
|
235 |
-
"eval_steps_per_second": 0.795,
|
236 |
-
"step": 869
|
237 |
-
},
|
238 |
-
{
|
239 |
-
"epoch": 22.9873417721519,
|
240 |
-
"eval_bleu": 0.7886,
|
241 |
-
"eval_gen_len": 20.3484,
|
242 |
-
"eval_loss": 5.026541233062744,
|
243 |
-
"eval_runtime": 478.7469,
|
244 |
-
"eval_samples_per_second": 25.393,
|
245 |
-
"eval_steps_per_second": 0.794,
|
246 |
-
"step": 908
|
247 |
-
},
|
248 |
-
{
|
249 |
-
"epoch": 24.0,
|
250 |
-
"eval_bleu": 0.7833,
|
251 |
-
"eval_gen_len": 20.3558,
|
252 |
-
"eval_loss": 5.028345584869385,
|
253 |
-
"eval_runtime": 476.6444,
|
254 |
-
"eval_samples_per_second": 25.505,
|
255 |
-
"eval_steps_per_second": 0.797,
|
256 |
-
"step": 948
|
257 |
-
},
|
258 |
-
{
|
259 |
-
"epoch": 24.9873417721519,
|
260 |
-
"eval_bleu": 0.7835,
|
261 |
-
"eval_gen_len": 20.313,
|
262 |
-
"eval_loss": 5.029706001281738,
|
263 |
-
"eval_runtime": 478.6376,
|
264 |
-
"eval_samples_per_second": 25.399,
|
265 |
-
"eval_steps_per_second": 0.794,
|
266 |
-
"step": 987
|
267 |
-
},
|
268 |
-
{
|
269 |
-
"epoch": 25.31645569620253,
|
270 |
-
"grad_norm": 0.5030060410499573,
|
271 |
-
"learning_rate": 2.9059829059829063e-06,
|
272 |
-
"loss": 0.0263,
|
273 |
-
"step": 1000
|
274 |
-
},
|
275 |
-
{
|
276 |
-
"epoch": 26.0,
|
277 |
-
"eval_bleu": 0.7812,
|
278 |
-
"eval_gen_len": 20.3503,
|
279 |
-
"eval_loss": 5.040848731994629,
|
280 |
-
"eval_runtime": 478.615,
|
281 |
-
"eval_samples_per_second": 25.4,
|
282 |
-
"eval_steps_per_second": 0.794,
|
283 |
-
"step": 1027
|
284 |
}
|
285 |
],
|
286 |
"logging_steps": 500,
|
287 |
-
"max_steps":
|
288 |
"num_input_tokens_seen": 0,
|
289 |
-
"num_train_epochs":
|
290 |
"save_steps": 1000,
|
291 |
"stateful_callbacks": {
|
292 |
"TrainerControl": {
|
@@ -295,12 +121,12 @@
|
|
295 |
"should_evaluate": false,
|
296 |
"should_log": false,
|
297 |
"should_save": true,
|
298 |
-
"should_training_stop":
|
299 |
},
|
300 |
"attributes": {}
|
301 |
}
|
302 |
},
|
303 |
-
"total_flos":
|
304 |
"train_batch_size": 32,
|
305 |
"trial_name": null,
|
306 |
"trial_params": null
|
|
|
1 |
{
|
2 |
+
"best_metric": 3.7421457767486572,
|
3 |
+
"best_model_checkpoint": "en-to-lg/checkpoint-16",
|
4 |
+
"epoch": 10.0,
|
5 |
"eval_steps": 50,
|
6 |
+
"global_step": 160,
|
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 |
+
"eval_bleu": 0.8596,
|
14 |
+
"eval_gen_len": 20.4303,
|
15 |
+
"eval_loss": 3.7421457767486572,
|
16 |
+
"eval_runtime": 486.1664,
|
17 |
+
"eval_samples_per_second": 25.006,
|
18 |
+
"eval_steps_per_second": 0.782,
|
19 |
+
"step": 16
|
20 |
},
|
21 |
{
|
22 |
"epoch": 2.0,
|
23 |
+
"eval_bleu": 0.855,
|
24 |
+
"eval_gen_len": 20.5137,
|
25 |
+
"eval_loss": 3.800355911254883,
|
26 |
+
"eval_runtime": 489.6809,
|
27 |
+
"eval_samples_per_second": 24.826,
|
28 |
+
"eval_steps_per_second": 0.776,
|
29 |
+
"step": 32
|
30 |
},
|
31 |
{
|
32 |
+
"epoch": 3.0,
|
33 |
+
"eval_bleu": 0.8413,
|
34 |
+
"eval_gen_len": 20.4,
|
35 |
+
"eval_loss": 3.806553602218628,
|
36 |
+
"eval_runtime": 491.5388,
|
37 |
+
"eval_samples_per_second": 24.733,
|
38 |
+
"eval_steps_per_second": 0.773,
|
39 |
+
"step": 48
|
40 |
},
|
41 |
{
|
42 |
"epoch": 4.0,
|
43 |
+
"eval_bleu": 0.8455,
|
44 |
+
"eval_gen_len": 20.3859,
|
45 |
+
"eval_loss": 3.864736557006836,
|
46 |
+
"eval_runtime": 490.916,
|
47 |
+
"eval_samples_per_second": 24.764,
|
48 |
+
"eval_steps_per_second": 0.774,
|
49 |
+
"step": 64
|
50 |
},
|
51 |
{
|
52 |
+
"epoch": 5.0,
|
53 |
+
"eval_bleu": 0.8527,
|
54 |
+
"eval_gen_len": 20.3211,
|
55 |
+
"eval_loss": 3.916616916656494,
|
56 |
+
"eval_runtime": 490.8429,
|
57 |
+
"eval_samples_per_second": 24.768,
|
58 |
+
"eval_steps_per_second": 0.774,
|
59 |
+
"step": 80
|
60 |
},
|
61 |
{
|
62 |
"epoch": 6.0,
|
63 |
+
"eval_bleu": 0.8539,
|
64 |
+
"eval_gen_len": 20.3088,
|
65 |
+
"eval_loss": 3.9686362743377686,
|
66 |
+
"eval_runtime": 492.3847,
|
67 |
+
"eval_samples_per_second": 24.69,
|
68 |
+
"eval_steps_per_second": 0.772,
|
69 |
+
"step": 96
|
70 |
},
|
71 |
{
|
72 |
+
"epoch": 7.0,
|
73 |
+
"eval_bleu": 0.8492,
|
74 |
+
"eval_gen_len": 20.3507,
|
75 |
+
"eval_loss": 3.9955193996429443,
|
76 |
+
"eval_runtime": 494.2016,
|
77 |
+
"eval_samples_per_second": 24.599,
|
78 |
+
"eval_steps_per_second": 0.769,
|
79 |
+
"step": 112
|
80 |
},
|
81 |
{
|
82 |
"epoch": 8.0,
|
83 |
+
"eval_bleu": 0.8387,
|
84 |
+
"eval_gen_len": 20.3404,
|
85 |
+
"eval_loss": 4.008735179901123,
|
86 |
+
"eval_runtime": 491.9673,
|
87 |
+
"eval_samples_per_second": 24.711,
|
88 |
+
"eval_steps_per_second": 0.772,
|
89 |
+
"step": 128
|
90 |
},
|
91 |
{
|
92 |
+
"epoch": 9.0,
|
93 |
+
"eval_bleu": 0.8345,
|
94 |
+
"eval_gen_len": 20.3244,
|
95 |
+
"eval_loss": 4.021594047546387,
|
96 |
+
"eval_runtime": 491.5119,
|
97 |
+
"eval_samples_per_second": 24.734,
|
98 |
+
"eval_steps_per_second": 0.773,
|
99 |
+
"step": 144
|
100 |
},
|
101 |
{
|
102 |
"epoch": 10.0,
|
103 |
+
"eval_bleu": 0.8431,
|
104 |
+
"eval_gen_len": 20.3462,
|
105 |
+
"eval_loss": 4.02959680557251,
|
106 |
+
"eval_runtime": 491.997,
|
107 |
+
"eval_samples_per_second": 24.71,
|
108 |
+
"eval_steps_per_second": 0.772,
|
109 |
+
"step": 160
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
110 |
}
|
111 |
],
|
112 |
"logging_steps": 500,
|
113 |
+
"max_steps": 160,
|
114 |
"num_input_tokens_seen": 0,
|
115 |
+
"num_train_epochs": 10,
|
116 |
"save_steps": 1000,
|
117 |
"stateful_callbacks": {
|
118 |
"TrainerControl": {
|
|
|
121 |
"should_evaluate": false,
|
122 |
"should_log": false,
|
123 |
"should_save": true,
|
124 |
+
"should_training_stop": true
|
125 |
},
|
126 |
"attributes": {}
|
127 |
}
|
128 |
},
|
129 |
+
"total_flos": 53351789101056.0,
|
130 |
"train_batch_size": 32,
|
131 |
"trial_name": null,
|
132 |
"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
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:dd9395e59bed2f6384e0c693e2944381623a36a75b946cce5bfdab8eebcecec7
|
3 |
+
size 5240
|