Training in progress, epoch 6, checkpoint
Browse files- checkpoint-1312/adapter_config.json +1 -1
- checkpoint-1312/adapter_model.bin +3 -0
- checkpoint-1312/added_tokens.json +3 -3
- checkpoint-1312/optimizer.pt +2 -2
- checkpoint-1312/rng_state_0.pth +1 -1
- checkpoint-1312/rng_state_1.pth +1 -1
- checkpoint-1312/rng_state_2.pth +1 -1
- checkpoint-1312/rng_state_3.pth +1 -1
- checkpoint-1312/special_tokens_map.json +4 -12
- checkpoint-1312/tokenizer.json +0 -0
- checkpoint-1312/tokenizer.model +3 -0
- checkpoint-1312/tokenizer_config.json +17 -20
- checkpoint-1312/trainer_state.json +107 -134
- checkpoint-1312/training_args.bin +2 -2
checkpoint-1312/adapter_config.json
CHANGED
@@ -1,6 +1,6 @@
|
|
1 |
{
|
2 |
"auto_mapping": null,
|
3 |
-
"base_model_name_or_path": "
|
4 |
"bias": "none",
|
5 |
"fan_in_fan_out": false,
|
6 |
"inference_mode": true,
|
|
|
1 |
{
|
2 |
"auto_mapping": null,
|
3 |
+
"base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
|
4 |
"bias": "none",
|
5 |
"fan_in_fan_out": false,
|
6 |
"inference_mode": true,
|
checkpoint-1312/adapter_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:902a0bfaf7f16bae59d35027b3855673fa0763175a368c73c9cc9726679e02ed
|
3 |
+
size 143269386
|
checkpoint-1312/added_tokens.json
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
{
|
2 |
-
"
|
3 |
-
"
|
4 |
-
"
|
5 |
}
|
|
|
1 |
{
|
2 |
+
"</s>": 2,
|
3 |
+
"<s>": 1,
|
4 |
+
"<unk>": 0
|
5 |
}
|
checkpoint-1312/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 sha256:30ab22943e4ae2a31cfbbf1348ee91e9dd188b0902cd6db44531ab76ac47ebe0
|
3 |
+
size 286585234
|
checkpoint-1312/rng_state_0.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 15024
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:673637d6f9be228678a509c7b5666590a8c9d04e7ccab584752427e997bd3d2a
|
3 |
size 15024
|
checkpoint-1312/rng_state_1.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 15024
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fe3f6962ef0507f254af005b5a501a46bec78691385c7e14d0ea34c9ae9f73f3
|
3 |
size 15024
|
checkpoint-1312/rng_state_2.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 15024
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b983138ed2deae7ba7e0e85e2e8440ff0c5187cc02e9aa86639bf98078409666
|
3 |
size 15024
|
checkpoint-1312/rng_state_3.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 15024
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c243574d3fcedd81f1eb18b2a12861db402bdbdbe12e1bdeb8d68bc1d7eefbad
|
3 |
size 15024
|
checkpoint-1312/special_tokens_map.json
CHANGED
@@ -1,14 +1,6 @@
|
|
1 |
{
|
2 |
-
"
|
3 |
-
|
4 |
-
|
5 |
-
|
6 |
-
"eos_token": {
|
7 |
-
"content": "<|endoftext|>",
|
8 |
-
"lstrip": false,
|
9 |
-
"normalized": false,
|
10 |
-
"rstrip": false,
|
11 |
-
"single_word": false
|
12 |
-
},
|
13 |
-
"pad_token": "<|endoftext|>"
|
14 |
}
|
|
|
1 |
{
|
2 |
+
"bos_token": "<s>",
|
3 |
+
"eos_token": "</s>",
|
4 |
+
"pad_token": "</s>",
|
5 |
+
"unk_token": "<unk>"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
6 |
}
|
checkpoint-1312/tokenizer.json
CHANGED
The diff for this file is too large to render.
See raw diff
|
|
checkpoint-1312/tokenizer.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
3 |
+
size 499723
|
checkpoint-1312/tokenizer_config.json
CHANGED
@@ -1,24 +1,23 @@
|
|
1 |
{
|
2 |
-
"add_prefix_space": false,
|
3 |
"added_tokens_decoder": {
|
4 |
-
"
|
5 |
-
"content": "
|
6 |
"lstrip": false,
|
7 |
"normalized": false,
|
8 |
"rstrip": false,
|
9 |
"single_word": false,
|
10 |
"special": true
|
11 |
},
|
12 |
-
"
|
13 |
-
"content": "
|
14 |
"lstrip": false,
|
15 |
"normalized": false,
|
16 |
"rstrip": false,
|
17 |
"single_word": false,
|
18 |
"special": true
|
19 |
},
|
20 |
-
"
|
21 |
-
"content": "
|
22 |
"lstrip": false,
|
23 |
"normalized": false,
|
24 |
"rstrip": false,
|
@@ -26,18 +25,16 @@
|
|
26 |
"special": true
|
27 |
}
|
28 |
},
|
29 |
-
"additional_special_tokens": [
|
30 |
-
|
31 |
-
"<|im_end|>"
|
32 |
-
],
|
33 |
-
"bos_token": null,
|
34 |
-
"chat_template": "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful assistant<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
|
35 |
"clean_up_tokenization_spaces": false,
|
36 |
-
"eos_token": "
|
37 |
-
"
|
38 |
-
"model_max_length":
|
39 |
-
"pad_token": "
|
40 |
-
"
|
41 |
-
"
|
42 |
-
"
|
|
|
|
|
43 |
}
|
|
|
1 |
{
|
|
|
2 |
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "<unk>",
|
5 |
"lstrip": false,
|
6 |
"normalized": false,
|
7 |
"rstrip": false,
|
8 |
"single_word": false,
|
9 |
"special": true
|
10 |
},
|
11 |
+
"1": {
|
12 |
+
"content": "<s>",
|
13 |
"lstrip": false,
|
14 |
"normalized": false,
|
15 |
"rstrip": false,
|
16 |
"single_word": false,
|
17 |
"special": true
|
18 |
},
|
19 |
+
"2": {
|
20 |
+
"content": "</s>",
|
21 |
"lstrip": false,
|
22 |
"normalized": false,
|
23 |
"rstrip": false,
|
|
|
25 |
"special": true
|
26 |
}
|
27 |
},
|
28 |
+
"additional_special_tokens": [],
|
29 |
+
"bos_token": "<s>",
|
|
|
|
|
|
|
|
|
30 |
"clean_up_tokenization_spaces": false,
|
31 |
+
"eos_token": "</s>",
|
32 |
+
"legacy": false,
|
33 |
+
"model_max_length": 1000000000000000019884624838656,
|
34 |
+
"pad_token": "</s>",
|
35 |
+
"padding_side": "left",
|
36 |
+
"sp_model_kwargs": {},
|
37 |
+
"tokenizer_class": "LlamaTokenizer",
|
38 |
+
"unk_token": "<unk>",
|
39 |
+
"use_default_system_prompt": true
|
40 |
}
|
checkpoint-1312/trainer_state.json
CHANGED
@@ -9,249 +9,222 @@
|
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
-
"epoch": 0.
|
13 |
-
"grad_norm": 0.271308958530426,
|
14 |
"learning_rate": 5e-05,
|
15 |
-
"loss": 1.
|
16 |
"step": 100
|
17 |
},
|
18 |
{
|
19 |
-
"epoch": 0
|
20 |
-
"eval_accuracy": 0.
|
21 |
-
"eval_loss": 1.
|
22 |
-
"eval_runtime":
|
23 |
-
"eval_samples_per_second":
|
24 |
-
"eval_steps_per_second":
|
25 |
"step": 187
|
26 |
},
|
27 |
{
|
28 |
-
"epoch": 0
|
29 |
-
"eval_exact_match":
|
30 |
-
"eval_f1":
|
31 |
-
"eval_qa_bleu":
|
32 |
-
"eval_qa_exact_match": 0.
|
33 |
-
"eval_recite_bleu": 12.
|
34 |
"eval_recite_exact_match": 0.0,
|
35 |
"step": 187
|
36 |
},
|
37 |
{
|
38 |
-
"epoch": 1.
|
39 |
-
"grad_norm": 0.301651269197464,
|
40 |
"learning_rate": 5e-05,
|
41 |
-
"loss": 1.
|
42 |
"step": 200
|
43 |
},
|
44 |
{
|
45 |
"epoch": 1.6,
|
46 |
-
"grad_norm": 0.35949045419692993,
|
47 |
"learning_rate": 5e-05,
|
48 |
-
"loss": 1.
|
49 |
"step": 300
|
50 |
},
|
51 |
{
|
52 |
"epoch": 2.0,
|
53 |
-
"eval_accuracy": 0.
|
54 |
-
"eval_loss": 1.
|
55 |
-
"eval_runtime":
|
56 |
-
"eval_samples_per_second":
|
57 |
-
"eval_steps_per_second":
|
58 |
"step": 375
|
59 |
},
|
60 |
{
|
61 |
"epoch": 2.0,
|
62 |
-
"eval_exact_match":
|
63 |
-
"eval_f1":
|
64 |
-
"eval_qa_bleu":
|
65 |
-
"eval_qa_exact_match": 0.
|
66 |
-
"eval_recite_bleu":
|
67 |
-
"eval_recite_exact_match": 0.
|
68 |
"step": 375
|
69 |
},
|
70 |
{
|
71 |
-
"epoch": 2.
|
72 |
-
"grad_norm": 0.41301965713500977,
|
73 |
"learning_rate": 5e-05,
|
74 |
-
"loss": 1.
|
75 |
"step": 400
|
76 |
},
|
77 |
{
|
78 |
-
"epoch": 2.
|
79 |
-
"grad_norm": 0.5057485699653625,
|
80 |
"learning_rate": 5e-05,
|
81 |
-
"loss": 1.
|
82 |
"step": 500
|
83 |
},
|
84 |
{
|
85 |
-
"epoch":
|
86 |
-
"eval_accuracy": 0.
|
87 |
-
"eval_loss": 1.
|
88 |
-
"eval_runtime":
|
89 |
-
"eval_samples_per_second":
|
90 |
-
"eval_steps_per_second":
|
91 |
"step": 562
|
92 |
},
|
93 |
{
|
94 |
-
"epoch":
|
95 |
-
"eval_exact_match":
|
96 |
-
"eval_f1":
|
97 |
-
"eval_qa_bleu":
|
98 |
-
"eval_qa_exact_match": 0.
|
99 |
-
"eval_recite_bleu": 12.
|
100 |
-
"eval_recite_exact_match": 0.
|
101 |
"step": 562
|
102 |
},
|
103 |
{
|
104 |
"epoch": 3.2,
|
105 |
-
"grad_norm": 0.6363334655761719,
|
106 |
"learning_rate": 5e-05,
|
107 |
-
"loss": 1.
|
108 |
"step": 600
|
109 |
},
|
110 |
{
|
111 |
-
"epoch": 3.
|
112 |
-
"grad_norm": 0.6466453671455383,
|
113 |
"learning_rate": 5e-05,
|
114 |
-
"loss": 1.
|
115 |
"step": 700
|
116 |
},
|
117 |
{
|
118 |
"epoch": 4.0,
|
119 |
-
"eval_accuracy": 0.
|
120 |
-
"eval_loss": 1.
|
121 |
-
"eval_runtime": 6.
|
122 |
-
"eval_samples_per_second":
|
123 |
-
"eval_steps_per_second":
|
124 |
"step": 750
|
125 |
},
|
126 |
{
|
127 |
"epoch": 4.0,
|
128 |
-
"eval_exact_match":
|
129 |
-
"eval_f1":
|
130 |
-
"eval_qa_bleu":
|
131 |
-
"eval_qa_exact_match": 0.
|
132 |
-
"eval_recite_bleu":
|
133 |
-
"eval_recite_exact_match": 0.
|
134 |
"step": 750
|
135 |
},
|
136 |
{
|
137 |
-
"epoch": 4.
|
138 |
-
"grad_norm": 0.8162885904312134,
|
139 |
"learning_rate": 5e-05,
|
140 |
-
"loss": 1.
|
141 |
"step": 800
|
142 |
},
|
143 |
{
|
144 |
"epoch": 4.8,
|
145 |
-
"grad_norm": 0.8799439072608948,
|
146 |
"learning_rate": 5e-05,
|
147 |
-
"loss": 1.
|
148 |
"step": 900
|
149 |
},
|
150 |
{
|
151 |
-
"epoch":
|
152 |
-
"eval_accuracy": 0.
|
153 |
-
"eval_loss": 1.
|
154 |
-
"eval_runtime": 6.
|
155 |
-
"eval_samples_per_second":
|
156 |
-
"eval_steps_per_second": 9.
|
157 |
"step": 937
|
158 |
},
|
159 |
{
|
160 |
-
"epoch":
|
161 |
-
"eval_exact_match":
|
162 |
-
"eval_f1":
|
163 |
-
"eval_qa_bleu":
|
164 |
-
"eval_qa_exact_match": 0.
|
165 |
-
"eval_recite_bleu":
|
166 |
-
"eval_recite_exact_match": 0.
|
167 |
"step": 937
|
168 |
},
|
169 |
{
|
170 |
-
"epoch": 5.
|
171 |
-
"grad_norm": 0.9834449291229248,
|
172 |
"learning_rate": 5e-05,
|
173 |
-
"loss":
|
174 |
"step": 1000
|
175 |
},
|
176 |
{
|
177 |
-
"epoch": 5.
|
178 |
-
"grad_norm": 0.9377098679542542,
|
179 |
"learning_rate": 5e-05,
|
180 |
-
"loss":
|
181 |
"step": 1100
|
182 |
},
|
183 |
{
|
184 |
"epoch": 6.0,
|
185 |
-
"eval_accuracy": 0.
|
186 |
-
"eval_loss": 1.
|
187 |
-
"eval_runtime":
|
188 |
-
"eval_samples_per_second":
|
189 |
-
"eval_steps_per_second":
|
190 |
"step": 1125
|
191 |
},
|
192 |
{
|
193 |
"epoch": 6.0,
|
194 |
-
"eval_exact_match":
|
195 |
-
"eval_f1":
|
196 |
-
"eval_qa_bleu":
|
197 |
-
"eval_qa_exact_match": 0.
|
198 |
-
"eval_recite_bleu":
|
199 |
-
"eval_recite_exact_match": 0.
|
200 |
"step": 1125
|
201 |
},
|
202 |
{
|
203 |
"epoch": 6.4,
|
204 |
-
"grad_norm": 1.1172072887420654,
|
205 |
"learning_rate": 5e-05,
|
206 |
-
"loss":
|
207 |
"step": 1200
|
208 |
},
|
209 |
{
|
210 |
-
"epoch": 6.
|
211 |
-
"grad_norm": 1.0980473756790161,
|
212 |
"learning_rate": 5e-05,
|
213 |
-
"loss":
|
214 |
"step": 1300
|
215 |
},
|
216 |
{
|
217 |
-
"epoch":
|
218 |
-
"eval_accuracy": 0.
|
219 |
-
"eval_loss": 1.
|
220 |
-
"eval_runtime":
|
221 |
-
"eval_samples_per_second":
|
222 |
-
"eval_steps_per_second":
|
223 |
"step": 1312
|
224 |
},
|
225 |
{
|
226 |
-
"epoch":
|
227 |
-
"eval_exact_match":
|
228 |
-
"eval_f1":
|
229 |
-
"eval_qa_bleu":
|
230 |
-
"eval_qa_exact_match": 0.
|
231 |
-
"eval_recite_bleu":
|
232 |
-
"eval_recite_exact_match": 0.
|
233 |
"step": 1312
|
234 |
}
|
235 |
],
|
236 |
"logging_steps": 100,
|
237 |
"max_steps": 9350,
|
238 |
-
"num_input_tokens_seen": 0,
|
239 |
"num_train_epochs": 50,
|
240 |
"save_steps": 500,
|
241 |
-
"
|
242 |
-
"TrainerControl": {
|
243 |
-
"args": {
|
244 |
-
"should_epoch_stop": false,
|
245 |
-
"should_evaluate": false,
|
246 |
-
"should_log": false,
|
247 |
-
"should_save": true,
|
248 |
-
"should_training_stop": false
|
249 |
-
},
|
250 |
-
"attributes": {}
|
251 |
-
}
|
252 |
-
},
|
253 |
-
"total_flos": 1.3220581835014144e+17,
|
254 |
-
"train_batch_size": 1,
|
255 |
"trial_name": null,
|
256 |
"trial_params": null
|
257 |
}
|
|
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
+
"epoch": 0.53,
|
|
|
13 |
"learning_rate": 5e-05,
|
14 |
+
"loss": 1.3988,
|
15 |
"step": 100
|
16 |
},
|
17 |
{
|
18 |
+
"epoch": 1.0,
|
19 |
+
"eval_accuracy": 0.6660784313725491,
|
20 |
+
"eval_loss": 1.2107256650924683,
|
21 |
+
"eval_runtime": 7.728,
|
22 |
+
"eval_samples_per_second": 64.7,
|
23 |
+
"eval_steps_per_second": 8.152,
|
24 |
"step": 187
|
25 |
},
|
26 |
{
|
27 |
+
"epoch": 1.0,
|
28 |
+
"eval_exact_match": 19.4,
|
29 |
+
"eval_f1": 27.911959489312448,
|
30 |
+
"eval_qa_bleu": 2.842200543476061,
|
31 |
+
"eval_qa_exact_match": 0.152,
|
32 |
+
"eval_recite_bleu": 12.149078895446083,
|
33 |
"eval_recite_exact_match": 0.0,
|
34 |
"step": 187
|
35 |
},
|
36 |
{
|
37 |
+
"epoch": 1.07,
|
|
|
38 |
"learning_rate": 5e-05,
|
39 |
+
"loss": 1.238,
|
40 |
"step": 200
|
41 |
},
|
42 |
{
|
43 |
"epoch": 1.6,
|
|
|
44 |
"learning_rate": 5e-05,
|
45 |
+
"loss": 1.1977,
|
46 |
"step": 300
|
47 |
},
|
48 |
{
|
49 |
"epoch": 2.0,
|
50 |
+
"eval_accuracy": 0.6676078431372549,
|
51 |
+
"eval_loss": 1.1987448930740356,
|
52 |
+
"eval_runtime": 7.4628,
|
53 |
+
"eval_samples_per_second": 66.999,
|
54 |
+
"eval_steps_per_second": 8.442,
|
55 |
"step": 375
|
56 |
},
|
57 |
{
|
58 |
"epoch": 2.0,
|
59 |
+
"eval_exact_match": 21.4,
|
60 |
+
"eval_f1": 27.979969946667268,
|
61 |
+
"eval_qa_bleu": 3.4273838915653356,
|
62 |
+
"eval_qa_exact_match": 0.158,
|
63 |
+
"eval_recite_bleu": 11.474877690710539,
|
64 |
+
"eval_recite_exact_match": 0.0,
|
65 |
"step": 375
|
66 |
},
|
67 |
{
|
68 |
+
"epoch": 2.13,
|
|
|
69 |
"learning_rate": 5e-05,
|
70 |
+
"loss": 1.1628,
|
71 |
"step": 400
|
72 |
},
|
73 |
{
|
74 |
+
"epoch": 2.67,
|
|
|
75 |
"learning_rate": 5e-05,
|
76 |
+
"loss": 1.1458,
|
77 |
"step": 500
|
78 |
},
|
79 |
{
|
80 |
+
"epoch": 3.0,
|
81 |
+
"eval_accuracy": 0.6675764705882353,
|
82 |
+
"eval_loss": 1.1964411735534668,
|
83 |
+
"eval_runtime": 6.6677,
|
84 |
+
"eval_samples_per_second": 74.988,
|
85 |
+
"eval_steps_per_second": 9.448,
|
86 |
"step": 562
|
87 |
},
|
88 |
{
|
89 |
+
"epoch": 3.0,
|
90 |
+
"eval_exact_match": 23.4,
|
91 |
+
"eval_f1": 31.525761083856903,
|
92 |
+
"eval_qa_bleu": 2.0606924929592334,
|
93 |
+
"eval_qa_exact_match": 0.17,
|
94 |
+
"eval_recite_bleu": 12.027903845545522,
|
95 |
+
"eval_recite_exact_match": 0.0,
|
96 |
"step": 562
|
97 |
},
|
98 |
{
|
99 |
"epoch": 3.2,
|
|
|
100 |
"learning_rate": 5e-05,
|
101 |
+
"loss": 1.1179,
|
102 |
"step": 600
|
103 |
},
|
104 |
{
|
105 |
+
"epoch": 3.73,
|
|
|
106 |
"learning_rate": 5e-05,
|
107 |
+
"loss": 1.0707,
|
108 |
"step": 700
|
109 |
},
|
110 |
{
|
111 |
"epoch": 4.0,
|
112 |
+
"eval_accuracy": 0.6665960784313726,
|
113 |
+
"eval_loss": 1.208784580230713,
|
114 |
+
"eval_runtime": 6.6546,
|
115 |
+
"eval_samples_per_second": 75.136,
|
116 |
+
"eval_steps_per_second": 9.467,
|
117 |
"step": 750
|
118 |
},
|
119 |
{
|
120 |
"epoch": 4.0,
|
121 |
+
"eval_exact_match": 23.4,
|
122 |
+
"eval_f1": 33.18893657873498,
|
123 |
+
"eval_qa_bleu": 3.8183981024520506,
|
124 |
+
"eval_qa_exact_match": 0.176,
|
125 |
+
"eval_recite_bleu": 14.071857337313705,
|
126 |
+
"eval_recite_exact_match": 0.0,
|
127 |
"step": 750
|
128 |
},
|
129 |
{
|
130 |
+
"epoch": 4.27,
|
|
|
131 |
"learning_rate": 5e-05,
|
132 |
+
"loss": 1.0179,
|
133 |
"step": 800
|
134 |
},
|
135 |
{
|
136 |
"epoch": 4.8,
|
|
|
137 |
"learning_rate": 5e-05,
|
138 |
+
"loss": 1.0066,
|
139 |
"step": 900
|
140 |
},
|
141 |
{
|
142 |
+
"epoch": 5.0,
|
143 |
+
"eval_accuracy": 0.6651764705882353,
|
144 |
+
"eval_loss": 1.2342489957809448,
|
145 |
+
"eval_runtime": 6.699,
|
146 |
+
"eval_samples_per_second": 74.638,
|
147 |
+
"eval_steps_per_second": 9.404,
|
148 |
"step": 937
|
149 |
},
|
150 |
{
|
151 |
+
"epoch": 5.0,
|
152 |
+
"eval_exact_match": 24.0,
|
153 |
+
"eval_f1": 32.88692081962977,
|
154 |
+
"eval_qa_bleu": 3.6747877275980905,
|
155 |
+
"eval_qa_exact_match": 0.174,
|
156 |
+
"eval_recite_bleu": 14.36158051041567,
|
157 |
+
"eval_recite_exact_match": 0.006,
|
158 |
"step": 937
|
159 |
},
|
160 |
{
|
161 |
+
"epoch": 5.33,
|
|
|
162 |
"learning_rate": 5e-05,
|
163 |
+
"loss": 0.9595,
|
164 |
"step": 1000
|
165 |
},
|
166 |
{
|
167 |
+
"epoch": 5.87,
|
|
|
168 |
"learning_rate": 5e-05,
|
169 |
+
"loss": 0.9353,
|
170 |
"step": 1100
|
171 |
},
|
172 |
{
|
173 |
"epoch": 6.0,
|
174 |
+
"eval_accuracy": 0.6628392156862745,
|
175 |
+
"eval_loss": 1.2645816802978516,
|
176 |
+
"eval_runtime": 7.4539,
|
177 |
+
"eval_samples_per_second": 67.079,
|
178 |
+
"eval_steps_per_second": 8.452,
|
179 |
"step": 1125
|
180 |
},
|
181 |
{
|
182 |
"epoch": 6.0,
|
183 |
+
"eval_exact_match": 23.4,
|
184 |
+
"eval_f1": 33.50380952380953,
|
185 |
+
"eval_qa_bleu": 13.968612067397348,
|
186 |
+
"eval_qa_exact_match": 0.168,
|
187 |
+
"eval_recite_bleu": 15.026821124315997,
|
188 |
+
"eval_recite_exact_match": 0.012,
|
189 |
"step": 1125
|
190 |
},
|
191 |
{
|
192 |
"epoch": 6.4,
|
|
|
193 |
"learning_rate": 5e-05,
|
194 |
+
"loss": 0.8774,
|
195 |
"step": 1200
|
196 |
},
|
197 |
{
|
198 |
+
"epoch": 6.93,
|
|
|
199 |
"learning_rate": 5e-05,
|
200 |
+
"loss": 0.8629,
|
201 |
"step": 1300
|
202 |
},
|
203 |
{
|
204 |
+
"epoch": 7.0,
|
205 |
+
"eval_accuracy": 0.6617333333333333,
|
206 |
+
"eval_loss": 1.2990769147872925,
|
207 |
+
"eval_runtime": 6.6963,
|
208 |
+
"eval_samples_per_second": 74.668,
|
209 |
+
"eval_steps_per_second": 9.408,
|
210 |
"step": 1312
|
211 |
},
|
212 |
{
|
213 |
+
"epoch": 7.0,
|
214 |
+
"eval_exact_match": 22.0,
|
215 |
+
"eval_f1": 31.8904761904762,
|
216 |
+
"eval_qa_bleu": 15.632357578213835,
|
217 |
+
"eval_qa_exact_match": 0.166,
|
218 |
+
"eval_recite_bleu": 14.318212267740977,
|
219 |
+
"eval_recite_exact_match": 0.016,
|
220 |
"step": 1312
|
221 |
}
|
222 |
],
|
223 |
"logging_steps": 100,
|
224 |
"max_steps": 9350,
|
|
|
225 |
"num_train_epochs": 50,
|
226 |
"save_steps": 500,
|
227 |
+
"total_flos": 3.041325183349555e+17,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
228 |
"trial_name": null,
|
229 |
"trial_params": null
|
230 |
}
|
checkpoint-1312/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:ae9143de0ad8fd0f027f30f757d262a6e02edc4ceeaea2c21697f4f44b5b1eeb
|
3 |
+
size 4728
|