alex2awesome
commited on
Commit
•
8113ca1
1
Parent(s):
b23d02a
first update
Browse files- all_results.json +21 -0
- config.json +37 -0
- post-training eval_results.json +9 -0
- pre-training eval_results.json +8 -0
- pytorch_model.bin +3 -0
- runs/Jan11_18-58-42_instance-6/1673463633.084183/events.out.tfevents.1673463633.instance-6.15669.1 +3 -0
- runs/Jan11_18-58-42_instance-6/events.out.tfevents.1673463633.instance-6.15669.0 +3 -0
- train_results.json +8 -0
- trainer_state.json +535 -0
- training_args.bin +3 -0
all_results.json
ADDED
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 3.0,
|
3 |
+
"eval_e": 0.0011185682326621924,
|
4 |
+
"eval_f1": 0.0030897725029911207,
|
5 |
+
"eval_loss": 7.352117538452148,
|
6 |
+
"eval_runtime": 82.3342,
|
7 |
+
"eval_samples": 3576,
|
8 |
+
"eval_samples_per_second": 43.433,
|
9 |
+
"eval_steps_per_second": 43.433,
|
10 |
+
"test_e": 0.5970357941834452,
|
11 |
+
"test_f1": 0.5385527992826207,
|
12 |
+
"test_loss": 3.199585199356079,
|
13 |
+
"test_runtime": 81.8465,
|
14 |
+
"test_samples_per_second": 43.692,
|
15 |
+
"test_steps_per_second": 43.692,
|
16 |
+
"train_loss": 1.5040293447896245,
|
17 |
+
"train_runtime": 3366.7923,
|
18 |
+
"train_samples": 10138,
|
19 |
+
"train_samples_per_second": 9.034,
|
20 |
+
"train_steps_per_second": 9.034
|
21 |
+
}
|
config.json
ADDED
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "google/bigbird-roberta-base",
|
3 |
+
"architectures": [
|
4 |
+
"QAModel"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"attention_type": "original_full",
|
8 |
+
"block_size": 64,
|
9 |
+
"bos_token_id": 1,
|
10 |
+
"classifier_dropout": null,
|
11 |
+
"eos_token_id": 2,
|
12 |
+
"frozen_layers": null,
|
13 |
+
"gradient_checkpointing": false,
|
14 |
+
"hidden_act": "gelu_new",
|
15 |
+
"hidden_dropout_prob": 0.1,
|
16 |
+
"hidden_size": 768,
|
17 |
+
"include_nones_as_positives": false,
|
18 |
+
"initializer_range": 0.02,
|
19 |
+
"intermediate_size": 3072,
|
20 |
+
"layer_norm_eps": 1e-12,
|
21 |
+
"max_position_embeddings": 4096,
|
22 |
+
"model_type": "big_bird",
|
23 |
+
"num_attention_heads": 12,
|
24 |
+
"num_hidden_layers": 12,
|
25 |
+
"num_random_blocks": 3,
|
26 |
+
"pad_token_id": 0,
|
27 |
+
"position_embedding_type": "absolute",
|
28 |
+
"qa_head": {},
|
29 |
+
"rescale_embeddings": false,
|
30 |
+
"sep_token_id": 66,
|
31 |
+
"torch_dtype": "float32",
|
32 |
+
"transformers_version": "4.12.2",
|
33 |
+
"type_vocab_size": 2,
|
34 |
+
"use_bias": true,
|
35 |
+
"use_cache": true,
|
36 |
+
"vocab_size": 50358
|
37 |
+
}
|
post-training eval_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"eval_samples": 3576,
|
3 |
+
"test_e": 0.5970357941834452,
|
4 |
+
"test_f1": 0.5385527992826207,
|
5 |
+
"test_loss": 3.199585199356079,
|
6 |
+
"test_runtime": 81.8465,
|
7 |
+
"test_samples_per_second": 43.692,
|
8 |
+
"test_steps_per_second": 43.692
|
9 |
+
}
|
pre-training eval_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"eval_e": 0.0011185682326621924,
|
3 |
+
"eval_f1": 0.0030897725029911207,
|
4 |
+
"eval_loss": 7.352117538452148,
|
5 |
+
"eval_runtime": 82.3342,
|
6 |
+
"eval_samples_per_second": 43.433,
|
7 |
+
"eval_steps_per_second": 43.433
|
8 |
+
}
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c932054221f36a93a52d76b45cda9ca37027b61c6728690a07e334802b7c5642
|
3 |
+
size 510000619
|
runs/Jan11_18-58-42_instance-6/1673463633.084183/events.out.tfevents.1673463633.instance-6.15669.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b3e685b10496bac1c731d82540116797e0a14a2bbe357f748e6c1f71c1a239b3
|
3 |
+
size 4624
|
runs/Jan11_18-58-42_instance-6/events.out.tfevents.1673463633.instance-6.15669.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:63171de535a9241b40fa9a46576d9189cfedd77e49f3968dd86ad6f0f489d180
|
3 |
+
size 19212
|
train_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 3.0,
|
3 |
+
"train_loss": 1.5040293447896245,
|
4 |
+
"train_runtime": 3366.7923,
|
5 |
+
"train_samples": 10138,
|
6 |
+
"train_samples_per_second": 9.034,
|
7 |
+
"train_steps_per_second": 9.034
|
8 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,535 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 3.0,
|
5 |
+
"global_step": 30414,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.05,
|
12 |
+
"learning_rate": 4.917801012691524e-05,
|
13 |
+
"loss": 4.2387,
|
14 |
+
"step": 500
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 0.1,
|
18 |
+
"learning_rate": 4.8356020253830476e-05,
|
19 |
+
"loss": 3.1858,
|
20 |
+
"step": 1000
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"epoch": 0.15,
|
24 |
+
"learning_rate": 4.7534030380745706e-05,
|
25 |
+
"loss": 2.9,
|
26 |
+
"step": 1500
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"epoch": 0.2,
|
30 |
+
"learning_rate": 4.671204050766095e-05,
|
31 |
+
"loss": 2.8336,
|
32 |
+
"step": 2000
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"epoch": 0.2,
|
36 |
+
"eval_e": 0.3347315436241611,
|
37 |
+
"eval_f1": 0.3097366216626532,
|
38 |
+
"eval_loss": 3.770604372024536,
|
39 |
+
"eval_runtime": 82.6518,
|
40 |
+
"eval_samples_per_second": 43.266,
|
41 |
+
"eval_steps_per_second": 43.266,
|
42 |
+
"step": 2000
|
43 |
+
},
|
44 |
+
{
|
45 |
+
"epoch": 0.25,
|
46 |
+
"learning_rate": 4.589005063457618e-05,
|
47 |
+
"loss": 2.6593,
|
48 |
+
"step": 2500
|
49 |
+
},
|
50 |
+
{
|
51 |
+
"epoch": 0.3,
|
52 |
+
"learning_rate": 4.506806076149142e-05,
|
53 |
+
"loss": 2.6292,
|
54 |
+
"step": 3000
|
55 |
+
},
|
56 |
+
{
|
57 |
+
"epoch": 0.35,
|
58 |
+
"learning_rate": 4.424607088840665e-05,
|
59 |
+
"loss": 2.5001,
|
60 |
+
"step": 3500
|
61 |
+
},
|
62 |
+
{
|
63 |
+
"epoch": 0.39,
|
64 |
+
"learning_rate": 4.3424081015321896e-05,
|
65 |
+
"loss": 2.779,
|
66 |
+
"step": 4000
|
67 |
+
},
|
68 |
+
{
|
69 |
+
"epoch": 0.39,
|
70 |
+
"eval_e": 0.4082774049217002,
|
71 |
+
"eval_f1": 0.3790107109138272,
|
72 |
+
"eval_loss": 3.6473984718322754,
|
73 |
+
"eval_runtime": 82.1465,
|
74 |
+
"eval_samples_per_second": 43.532,
|
75 |
+
"eval_steps_per_second": 43.532,
|
76 |
+
"step": 4000
|
77 |
+
},
|
78 |
+
{
|
79 |
+
"epoch": 0.44,
|
80 |
+
"learning_rate": 4.2602091142237125e-05,
|
81 |
+
"loss": 2.4774,
|
82 |
+
"step": 4500
|
83 |
+
},
|
84 |
+
{
|
85 |
+
"epoch": 0.49,
|
86 |
+
"learning_rate": 4.178010126915237e-05,
|
87 |
+
"loss": 2.2356,
|
88 |
+
"step": 5000
|
89 |
+
},
|
90 |
+
{
|
91 |
+
"epoch": 0.54,
|
92 |
+
"learning_rate": 4.09581113960676e-05,
|
93 |
+
"loss": 2.312,
|
94 |
+
"step": 5500
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"epoch": 0.59,
|
98 |
+
"learning_rate": 4.013612152298284e-05,
|
99 |
+
"loss": 2.1882,
|
100 |
+
"step": 6000
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.59,
|
104 |
+
"eval_e": 0.47511185682326623,
|
105 |
+
"eval_f1": 0.42880102269324966,
|
106 |
+
"eval_loss": 3.5800905227661133,
|
107 |
+
"eval_runtime": 82.0337,
|
108 |
+
"eval_samples_per_second": 43.592,
|
109 |
+
"eval_steps_per_second": 43.592,
|
110 |
+
"step": 6000
|
111 |
+
},
|
112 |
+
{
|
113 |
+
"epoch": 0.64,
|
114 |
+
"learning_rate": 3.931413164989807e-05,
|
115 |
+
"loss": 2.2187,
|
116 |
+
"step": 6500
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"epoch": 0.69,
|
120 |
+
"learning_rate": 3.849214177681331e-05,
|
121 |
+
"loss": 1.9332,
|
122 |
+
"step": 7000
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 0.74,
|
126 |
+
"learning_rate": 3.7670151903728545e-05,
|
127 |
+
"loss": 2.0221,
|
128 |
+
"step": 7500
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.79,
|
132 |
+
"learning_rate": 3.684816203064378e-05,
|
133 |
+
"loss": 2.1572,
|
134 |
+
"step": 8000
|
135 |
+
},
|
136 |
+
{
|
137 |
+
"epoch": 0.79,
|
138 |
+
"eval_e": 0.465324384787472,
|
139 |
+
"eval_f1": 0.43269356216847144,
|
140 |
+
"eval_loss": 3.3040895462036133,
|
141 |
+
"eval_runtime": 82.0712,
|
142 |
+
"eval_samples_per_second": 43.572,
|
143 |
+
"eval_steps_per_second": 43.572,
|
144 |
+
"step": 8000
|
145 |
+
},
|
146 |
+
{
|
147 |
+
"epoch": 0.84,
|
148 |
+
"learning_rate": 3.602617215755902e-05,
|
149 |
+
"loss": 1.8899,
|
150 |
+
"step": 8500
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 0.89,
|
154 |
+
"learning_rate": 3.5204182284474255e-05,
|
155 |
+
"loss": 2.0571,
|
156 |
+
"step": 9000
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.94,
|
160 |
+
"learning_rate": 3.438219241138949e-05,
|
161 |
+
"loss": 1.8117,
|
162 |
+
"step": 9500
|
163 |
+
},
|
164 |
+
{
|
165 |
+
"epoch": 0.99,
|
166 |
+
"learning_rate": 3.356020253830473e-05,
|
167 |
+
"loss": 1.9861,
|
168 |
+
"step": 10000
|
169 |
+
},
|
170 |
+
{
|
171 |
+
"epoch": 0.99,
|
172 |
+
"eval_e": 0.5022371364653244,
|
173 |
+
"eval_f1": 0.4571106421513448,
|
174 |
+
"eval_loss": 3.3798608779907227,
|
175 |
+
"eval_runtime": 82.3014,
|
176 |
+
"eval_samples_per_second": 43.45,
|
177 |
+
"eval_steps_per_second": 43.45,
|
178 |
+
"step": 10000
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 1.04,
|
182 |
+
"learning_rate": 3.2738212665219965e-05,
|
183 |
+
"loss": 1.5352,
|
184 |
+
"step": 10500
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 1.09,
|
188 |
+
"learning_rate": 3.19162227921352e-05,
|
189 |
+
"loss": 1.4703,
|
190 |
+
"step": 11000
|
191 |
+
},
|
192 |
+
{
|
193 |
+
"epoch": 1.13,
|
194 |
+
"learning_rate": 3.109423291905044e-05,
|
195 |
+
"loss": 1.5819,
|
196 |
+
"step": 11500
|
197 |
+
},
|
198 |
+
{
|
199 |
+
"epoch": 1.18,
|
200 |
+
"learning_rate": 3.0272243045965675e-05,
|
201 |
+
"loss": 1.3955,
|
202 |
+
"step": 12000
|
203 |
+
},
|
204 |
+
{
|
205 |
+
"epoch": 1.18,
|
206 |
+
"eval_e": 0.5125838926174496,
|
207 |
+
"eval_f1": 0.47356235492642834,
|
208 |
+
"eval_loss": 3.511906147003174,
|
209 |
+
"eval_runtime": 81.9656,
|
210 |
+
"eval_samples_per_second": 43.628,
|
211 |
+
"eval_steps_per_second": 43.628,
|
212 |
+
"step": 12000
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 1.23,
|
216 |
+
"learning_rate": 2.9450253172880908e-05,
|
217 |
+
"loss": 1.4744,
|
218 |
+
"step": 12500
|
219 |
+
},
|
220 |
+
{
|
221 |
+
"epoch": 1.28,
|
222 |
+
"learning_rate": 2.8628263299796148e-05,
|
223 |
+
"loss": 1.2676,
|
224 |
+
"step": 13000
|
225 |
+
},
|
226 |
+
{
|
227 |
+
"epoch": 1.33,
|
228 |
+
"learning_rate": 2.780627342671138e-05,
|
229 |
+
"loss": 1.3808,
|
230 |
+
"step": 13500
|
231 |
+
},
|
232 |
+
{
|
233 |
+
"epoch": 1.38,
|
234 |
+
"learning_rate": 2.698428355362662e-05,
|
235 |
+
"loss": 1.4232,
|
236 |
+
"step": 14000
|
237 |
+
},
|
238 |
+
{
|
239 |
+
"epoch": 1.38,
|
240 |
+
"eval_e": 0.47231543624161076,
|
241 |
+
"eval_f1": 0.4436310133766935,
|
242 |
+
"eval_loss": 4.007138252258301,
|
243 |
+
"eval_runtime": 82.0736,
|
244 |
+
"eval_samples_per_second": 43.571,
|
245 |
+
"eval_steps_per_second": 43.571,
|
246 |
+
"step": 14000
|
247 |
+
},
|
248 |
+
{
|
249 |
+
"epoch": 1.43,
|
250 |
+
"learning_rate": 2.6162293680541855e-05,
|
251 |
+
"loss": 1.3547,
|
252 |
+
"step": 14500
|
253 |
+
},
|
254 |
+
{
|
255 |
+
"epoch": 1.48,
|
256 |
+
"learning_rate": 2.5340303807457095e-05,
|
257 |
+
"loss": 1.4022,
|
258 |
+
"step": 15000
|
259 |
+
},
|
260 |
+
{
|
261 |
+
"epoch": 1.53,
|
262 |
+
"learning_rate": 2.451831393437233e-05,
|
263 |
+
"loss": 1.2848,
|
264 |
+
"step": 15500
|
265 |
+
},
|
266 |
+
{
|
267 |
+
"epoch": 1.58,
|
268 |
+
"learning_rate": 2.3696324061287568e-05,
|
269 |
+
"loss": 1.3669,
|
270 |
+
"step": 16000
|
271 |
+
},
|
272 |
+
{
|
273 |
+
"epoch": 1.58,
|
274 |
+
"eval_e": 0.5131431767337807,
|
275 |
+
"eval_f1": 0.47548209797210494,
|
276 |
+
"eval_loss": 2.720799207687378,
|
277 |
+
"eval_runtime": 82.4071,
|
278 |
+
"eval_samples_per_second": 43.394,
|
279 |
+
"eval_steps_per_second": 43.394,
|
280 |
+
"step": 16000
|
281 |
+
},
|
282 |
+
{
|
283 |
+
"epoch": 1.63,
|
284 |
+
"learning_rate": 2.2874334188202805e-05,
|
285 |
+
"loss": 1.1994,
|
286 |
+
"step": 16500
|
287 |
+
},
|
288 |
+
{
|
289 |
+
"epoch": 1.68,
|
290 |
+
"learning_rate": 2.205234431511804e-05,
|
291 |
+
"loss": 1.2361,
|
292 |
+
"step": 17000
|
293 |
+
},
|
294 |
+
{
|
295 |
+
"epoch": 1.73,
|
296 |
+
"learning_rate": 2.1230354442033275e-05,
|
297 |
+
"loss": 1.3868,
|
298 |
+
"step": 17500
|
299 |
+
},
|
300 |
+
{
|
301 |
+
"epoch": 1.78,
|
302 |
+
"learning_rate": 2.040836456894851e-05,
|
303 |
+
"loss": 1.2285,
|
304 |
+
"step": 18000
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 1.78,
|
308 |
+
"eval_e": 0.5497762863534675,
|
309 |
+
"eval_f1": 0.5054523815466766,
|
310 |
+
"eval_loss": 3.2848546504974365,
|
311 |
+
"eval_runtime": 82.2488,
|
312 |
+
"eval_samples_per_second": 43.478,
|
313 |
+
"eval_steps_per_second": 43.478,
|
314 |
+
"step": 18000
|
315 |
+
},
|
316 |
+
{
|
317 |
+
"epoch": 1.82,
|
318 |
+
"learning_rate": 1.9586374695863748e-05,
|
319 |
+
"loss": 1.0095,
|
320 |
+
"step": 18500
|
321 |
+
},
|
322 |
+
{
|
323 |
+
"epoch": 1.87,
|
324 |
+
"learning_rate": 1.8764384822778985e-05,
|
325 |
+
"loss": 1.3675,
|
326 |
+
"step": 19000
|
327 |
+
},
|
328 |
+
{
|
329 |
+
"epoch": 1.92,
|
330 |
+
"learning_rate": 1.794239494969422e-05,
|
331 |
+
"loss": 1.2251,
|
332 |
+
"step": 19500
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 1.97,
|
336 |
+
"learning_rate": 1.7120405076609458e-05,
|
337 |
+
"loss": 1.1866,
|
338 |
+
"step": 20000
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 1.97,
|
342 |
+
"eval_e": 0.5528523489932886,
|
343 |
+
"eval_f1": 0.5108568928693717,
|
344 |
+
"eval_loss": 3.511319398880005,
|
345 |
+
"eval_runtime": 82.0354,
|
346 |
+
"eval_samples_per_second": 43.591,
|
347 |
+
"eval_steps_per_second": 43.591,
|
348 |
+
"step": 20000
|
349 |
+
},
|
350 |
+
{
|
351 |
+
"epoch": 2.02,
|
352 |
+
"learning_rate": 1.6298415203524694e-05,
|
353 |
+
"loss": 1.1769,
|
354 |
+
"step": 20500
|
355 |
+
},
|
356 |
+
{
|
357 |
+
"epoch": 2.07,
|
358 |
+
"learning_rate": 1.547642533043993e-05,
|
359 |
+
"loss": 0.8723,
|
360 |
+
"step": 21000
|
361 |
+
},
|
362 |
+
{
|
363 |
+
"epoch": 2.12,
|
364 |
+
"learning_rate": 1.4654435457355168e-05,
|
365 |
+
"loss": 0.6951,
|
366 |
+
"step": 21500
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 2.17,
|
370 |
+
"learning_rate": 1.3832445584270404e-05,
|
371 |
+
"loss": 0.7215,
|
372 |
+
"step": 22000
|
373 |
+
},
|
374 |
+
{
|
375 |
+
"epoch": 2.17,
|
376 |
+
"eval_e": 0.5564876957494407,
|
377 |
+
"eval_f1": 0.49972487210398886,
|
378 |
+
"eval_loss": 3.31522798538208,
|
379 |
+
"eval_runtime": 82.1318,
|
380 |
+
"eval_samples_per_second": 43.54,
|
381 |
+
"eval_steps_per_second": 43.54,
|
382 |
+
"step": 22000
|
383 |
+
},
|
384 |
+
{
|
385 |
+
"epoch": 2.22,
|
386 |
+
"learning_rate": 1.301045571118564e-05,
|
387 |
+
"loss": 0.8564,
|
388 |
+
"step": 22500
|
389 |
+
},
|
390 |
+
{
|
391 |
+
"epoch": 2.27,
|
392 |
+
"learning_rate": 1.2188465838100876e-05,
|
393 |
+
"loss": 0.7706,
|
394 |
+
"step": 23000
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 2.32,
|
398 |
+
"learning_rate": 1.1366475965016111e-05,
|
399 |
+
"loss": 0.7556,
|
400 |
+
"step": 23500
|
401 |
+
},
|
402 |
+
{
|
403 |
+
"epoch": 2.37,
|
404 |
+
"learning_rate": 1.0544486091931348e-05,
|
405 |
+
"loss": 0.8466,
|
406 |
+
"step": 24000
|
407 |
+
},
|
408 |
+
{
|
409 |
+
"epoch": 2.37,
|
410 |
+
"eval_e": 0.581096196868009,
|
411 |
+
"eval_f1": 0.5250151109087391,
|
412 |
+
"eval_loss": 3.5018973350524902,
|
413 |
+
"eval_runtime": 82.0779,
|
414 |
+
"eval_samples_per_second": 43.568,
|
415 |
+
"eval_steps_per_second": 43.568,
|
416 |
+
"step": 24000
|
417 |
+
},
|
418 |
+
{
|
419 |
+
"epoch": 2.42,
|
420 |
+
"learning_rate": 9.722496218846584e-06,
|
421 |
+
"loss": 0.8433,
|
422 |
+
"step": 24500
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 2.47,
|
426 |
+
"learning_rate": 8.90050634576182e-06,
|
427 |
+
"loss": 0.7424,
|
428 |
+
"step": 25000
|
429 |
+
},
|
430 |
+
{
|
431 |
+
"epoch": 2.52,
|
432 |
+
"learning_rate": 8.078516472677057e-06,
|
433 |
+
"loss": 0.7016,
|
434 |
+
"step": 25500
|
435 |
+
},
|
436 |
+
{
|
437 |
+
"epoch": 2.56,
|
438 |
+
"learning_rate": 7.256526599592293e-06,
|
439 |
+
"loss": 0.7638,
|
440 |
+
"step": 26000
|
441 |
+
},
|
442 |
+
{
|
443 |
+
"epoch": 2.56,
|
444 |
+
"eval_e": 0.5746644295302014,
|
445 |
+
"eval_f1": 0.5266203379718399,
|
446 |
+
"eval_loss": 3.3142940998077393,
|
447 |
+
"eval_runtime": 82.0427,
|
448 |
+
"eval_samples_per_second": 43.587,
|
449 |
+
"eval_steps_per_second": 43.587,
|
450 |
+
"step": 26000
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 2.61,
|
454 |
+
"learning_rate": 6.43453672650753e-06,
|
455 |
+
"loss": 0.6414,
|
456 |
+
"step": 26500
|
457 |
+
},
|
458 |
+
{
|
459 |
+
"epoch": 2.66,
|
460 |
+
"learning_rate": 5.6125468534227665e-06,
|
461 |
+
"loss": 0.7472,
|
462 |
+
"step": 27000
|
463 |
+
},
|
464 |
+
{
|
465 |
+
"epoch": 2.71,
|
466 |
+
"learning_rate": 4.790556980338002e-06,
|
467 |
+
"loss": 0.6423,
|
468 |
+
"step": 27500
|
469 |
+
},
|
470 |
+
{
|
471 |
+
"epoch": 2.76,
|
472 |
+
"learning_rate": 3.968567107253239e-06,
|
473 |
+
"loss": 0.78,
|
474 |
+
"step": 28000
|
475 |
+
},
|
476 |
+
{
|
477 |
+
"epoch": 2.76,
|
478 |
+
"eval_e": 0.5953579418344519,
|
479 |
+
"eval_f1": 0.5431688990260519,
|
480 |
+
"eval_loss": 3.2026774883270264,
|
481 |
+
"eval_runtime": 82.1365,
|
482 |
+
"eval_samples_per_second": 43.537,
|
483 |
+
"eval_steps_per_second": 43.537,
|
484 |
+
"step": 28000
|
485 |
+
},
|
486 |
+
{
|
487 |
+
"epoch": 2.81,
|
488 |
+
"learning_rate": 3.1465772341684756e-06,
|
489 |
+
"loss": 0.6853,
|
490 |
+
"step": 28500
|
491 |
+
},
|
492 |
+
{
|
493 |
+
"epoch": 2.86,
|
494 |
+
"learning_rate": 2.3245873610837114e-06,
|
495 |
+
"loss": 0.6939,
|
496 |
+
"step": 29000
|
497 |
+
},
|
498 |
+
{
|
499 |
+
"epoch": 2.91,
|
500 |
+
"learning_rate": 1.5025974879989478e-06,
|
501 |
+
"loss": 0.6681,
|
502 |
+
"step": 29500
|
503 |
+
},
|
504 |
+
{
|
505 |
+
"epoch": 2.96,
|
506 |
+
"learning_rate": 6.806076149141843e-07,
|
507 |
+
"loss": 0.6097,
|
508 |
+
"step": 30000
|
509 |
+
},
|
510 |
+
{
|
511 |
+
"epoch": 2.96,
|
512 |
+
"eval_e": 0.5953579418344519,
|
513 |
+
"eval_f1": 0.5371973709058047,
|
514 |
+
"eval_loss": 3.181272029876709,
|
515 |
+
"eval_runtime": 82.1847,
|
516 |
+
"eval_samples_per_second": 43.512,
|
517 |
+
"eval_steps_per_second": 43.512,
|
518 |
+
"step": 30000
|
519 |
+
},
|
520 |
+
{
|
521 |
+
"epoch": 3.0,
|
522 |
+
"step": 30414,
|
523 |
+
"total_flos": 1.541864598291324e+16,
|
524 |
+
"train_loss": 1.5040293447896245,
|
525 |
+
"train_runtime": 3366.7923,
|
526 |
+
"train_samples_per_second": 9.034,
|
527 |
+
"train_steps_per_second": 9.034
|
528 |
+
}
|
529 |
+
],
|
530 |
+
"max_steps": 30414,
|
531 |
+
"num_train_epochs": 3,
|
532 |
+
"total_flos": 1.541864598291324e+16,
|
533 |
+
"trial_name": null,
|
534 |
+
"trial_params": null
|
535 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:95c6b1cab31dad3e2c21c0d6f1d830b38b27036a42d8b1277afc0e4f599cbe05
|
3 |
+
size 2927
|