Training in progress, epoch 38, checkpoint
Browse files- checkpoint-7125/adapter_config.json +1 -1
- checkpoint-7125/adapter_model.bin +3 -0
- checkpoint-7125/added_tokens.json +3 -3
- checkpoint-7125/optimizer.pt +2 -2
- checkpoint-7125/rng_state_0.pth +1 -1
- checkpoint-7125/rng_state_1.pth +1 -1
- checkpoint-7125/rng_state_2.pth +1 -1
- checkpoint-7125/rng_state_3.pth +1 -1
- checkpoint-7125/special_tokens_map.json +4 -12
- checkpoint-7125/tokenizer.json +0 -0
- checkpoint-7125/tokenizer.model +3 -0
- checkpoint-7125/tokenizer_config.json +17 -20
- checkpoint-7125/trainer_state.json +573 -658
- checkpoint-7125/training_args.bin +2 -2
checkpoint-7125/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-7125/adapter_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:de02516f8d6cc09b1beee9b77c73dff5fd3a3aeb6fc3e7345756e175b5ab1aa1
|
3 |
+
size 143269386
|
checkpoint-7125/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-7125/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 sha256:47008578ab4175f40b5f395f733682f086689ed1a9f7da1619bc55ba1b52d32d
|
3 |
+
size 286585234
|
checkpoint-7125/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 sha256:7c0156db597a550668d1c64e7bd6ffb1d2226cad4c858aec2ba70ba3372e29da
|
3 |
size 15024
|
checkpoint-7125/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 sha256:15760b903dfa4ccdde7cd1671fe1ea2133d61a1e922b2d175a1043c135365f42
|
3 |
size 15024
|
checkpoint-7125/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:b6893b61cfda7f229a37ddc3ac1713d94ab13a1952bcb5d28c7eab45b129bcc9
|
3 |
size 15024
|
checkpoint-7125/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 sha256:64aa89970507afee7a77bdddbe3fdefd4f24cd1dddc904599a35e6b2bf4346ca
|
3 |
size 15024
|
checkpoint-7125/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-7125/tokenizer.json
CHANGED
The diff for this file is too large to render.
See raw diff
|
|
checkpoint-7125/tokenizer.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
3 |
+
size 499723
|
checkpoint-7125/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-7125/trainer_state.json
CHANGED
@@ -9,1244 +9,1159 @@
|
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
-
"epoch": 0.
|
13 |
-
"grad_norm": 0.275871217250824,
|
14 |
"learning_rate": 3e-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":
|
34 |
"eval_recite_exact_match": 0.0,
|
35 |
"step": 187
|
36 |
},
|
37 |
{
|
38 |
-
"epoch": 1.
|
39 |
-
"grad_norm": 0.32765188813209534,
|
40 |
"learning_rate": 3e-05,
|
41 |
-
"loss": 1.
|
42 |
"step": 200
|
43 |
},
|
44 |
{
|
45 |
"epoch": 1.6,
|
46 |
-
"grad_norm": 0.3532409369945526,
|
47 |
"learning_rate": 3e-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": 0.0,
|
68 |
"step": 375
|
69 |
},
|
70 |
{
|
71 |
-
"epoch": 2.
|
72 |
-
"grad_norm": 0.3777780532836914,
|
73 |
"learning_rate": 3e-05,
|
74 |
-
"loss": 1.
|
75 |
"step": 400
|
76 |
},
|
77 |
{
|
78 |
-
"epoch": 2.
|
79 |
-
"grad_norm": 0.3971036374568939,
|
80 |
"learning_rate": 3e-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":
|
100 |
-
"eval_recite_exact_match": 0.
|
101 |
"step": 562
|
102 |
},
|
103 |
{
|
104 |
"epoch": 3.2,
|
105 |
-
"grad_norm": 0.4782978594303131,
|
106 |
"learning_rate": 3e-05,
|
107 |
-
"loss": 1.
|
108 |
"step": 600
|
109 |
},
|
110 |
{
|
111 |
-
"epoch": 3.
|
112 |
-
"grad_norm": 0.52582186460495,
|
113 |
"learning_rate": 3e-05,
|
114 |
-
"loss": 1.
|
115 |
"step": 700
|
116 |
},
|
117 |
{
|
118 |
"epoch": 4.0,
|
119 |
-
"eval_accuracy": 0.
|
120 |
-
"eval_loss": 1.
|
121 |
-
"eval_runtime":
|
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": 0.6418241262435913,
|
139 |
"learning_rate": 3e-05,
|
140 |
-
"loss": 1.
|
141 |
"step": 800
|
142 |
},
|
143 |
{
|
144 |
"epoch": 4.8,
|
145 |
-
"grad_norm": 0.6805307269096375,
|
146 |
"learning_rate": 3e-05,
|
147 |
-
"loss": 1.
|
148 |
"step": 900
|
149 |
},
|
150 |
{
|
151 |
-
"epoch":
|
152 |
-
"eval_accuracy": 0.
|
153 |
-
"eval_loss": 1.
|
154 |
-
"eval_runtime":
|
155 |
-
"eval_samples_per_second":
|
156 |
-
"eval_steps_per_second":
|
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": 0.8362048864364624,
|
172 |
"learning_rate": 3e-05,
|
173 |
-
"loss": 1.
|
174 |
"step": 1000
|
175 |
},
|
176 |
{
|
177 |
-
"epoch": 5.
|
178 |
-
"grad_norm": 0.8521270155906677,
|
179 |
"learning_rate": 3e-05,
|
180 |
-
"loss": 1.
|
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": 0.9372308254241943,
|
205 |
"learning_rate": 3e-05,
|
206 |
-
"loss": 1.
|
207 |
"step": 1200
|
208 |
},
|
209 |
{
|
210 |
-
"epoch": 6.
|
211 |
-
"grad_norm": 1.0236693620681763,
|
212 |
"learning_rate": 3e-05,
|
213 |
-
"loss":
|
214 |
"step": 1300
|
215 |
},
|
216 |
{
|
217 |
-
"epoch":
|
218 |
-
"eval_accuracy": 0.
|
219 |
-
"eval_loss": 1.
|
220 |
-
"eval_runtime": 6.
|
221 |
-
"eval_samples_per_second":
|
222 |
-
"eval_steps_per_second": 9.
|
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 |
-
"epoch": 7.
|
237 |
-
"grad_norm": 1.0773049592971802,
|
238 |
"learning_rate": 3e-05,
|
239 |
-
"loss":
|
240 |
"step": 1400
|
241 |
},
|
242 |
{
|
243 |
"epoch": 8.0,
|
244 |
-
"grad_norm": 1.0705862045288086,
|
245 |
"learning_rate": 3e-05,
|
246 |
-
"loss":
|
247 |
"step": 1500
|
248 |
},
|
249 |
{
|
250 |
"epoch": 8.0,
|
251 |
-
"eval_accuracy": 0.
|
252 |
-
"eval_loss": 1.
|
253 |
-
"eval_runtime":
|
254 |
-
"eval_samples_per_second":
|
255 |
-
"eval_steps_per_second":
|
256 |
"step": 1500
|
257 |
},
|
258 |
{
|
259 |
"epoch": 8.0,
|
260 |
-
"eval_exact_match":
|
261 |
-
"eval_f1":
|
262 |
-
"eval_qa_bleu":
|
263 |
-
"eval_qa_exact_match": 0.
|
264 |
-
"eval_recite_bleu":
|
265 |
-
"eval_recite_exact_match": 0.
|
266 |
"step": 1500
|
267 |
},
|
268 |
{
|
269 |
-
"epoch": 8.
|
270 |
-
"grad_norm": 1.2424529790878296,
|
271 |
"learning_rate": 3e-05,
|
272 |
-
"loss":
|
273 |
"step": 1600
|
274 |
},
|
275 |
{
|
276 |
-
"epoch":
|
277 |
-
"eval_accuracy": 0.
|
278 |
-
"eval_loss": 1.
|
279 |
-
"eval_runtime":
|
280 |
-
"eval_samples_per_second":
|
281 |
-
"eval_steps_per_second":
|
282 |
"step": 1687
|
283 |
},
|
284 |
{
|
285 |
-
"epoch":
|
286 |
-
"eval_exact_match":
|
287 |
-
"eval_f1":
|
288 |
-
"eval_qa_bleu":
|
289 |
-
"eval_qa_exact_match": 0.
|
290 |
-
"eval_recite_bleu":
|
291 |
-
"eval_recite_exact_match": 0.
|
292 |
"step": 1687
|
293 |
},
|
294 |
{
|
295 |
-
"epoch": 9.
|
296 |
-
"grad_norm": 1.535229206085205,
|
297 |
"learning_rate": 3e-05,
|
298 |
-
"loss":
|
299 |
"step": 1700
|
300 |
},
|
301 |
{
|
302 |
"epoch": 9.6,
|
303 |
-
"grad_norm": 1.4936225414276123,
|
304 |
"learning_rate": 3e-05,
|
305 |
-
"loss":
|
306 |
"step": 1800
|
307 |
},
|
308 |
{
|
309 |
"epoch": 10.0,
|
310 |
-
"eval_accuracy": 0.
|
311 |
-
"eval_loss": 1.
|
312 |
-
"eval_runtime": 6.
|
313 |
-
"eval_samples_per_second":
|
314 |
-
"eval_steps_per_second": 9.
|
315 |
"step": 1875
|
316 |
},
|
317 |
{
|
318 |
"epoch": 10.0,
|
319 |
-
"eval_exact_match":
|
320 |
-
"eval_f1": 21.
|
321 |
-
"eval_qa_bleu":
|
322 |
-
"eval_qa_exact_match": 0.
|
323 |
-
"eval_recite_bleu":
|
324 |
-
"eval_recite_exact_match": 0.
|
325 |
"step": 1875
|
326 |
},
|
327 |
{
|
328 |
-
"epoch": 10.
|
329 |
-
"grad_norm": 1.6687729358673096,
|
330 |
"learning_rate": 3e-05,
|
331 |
-
"loss":
|
332 |
"step": 1900
|
333 |
},
|
334 |
{
|
335 |
-
"epoch": 10.
|
336 |
-
"grad_norm": 1.437538504600525,
|
337 |
"learning_rate": 3e-05,
|
338 |
-
"loss":
|
339 |
"step": 2000
|
340 |
},
|
341 |
{
|
342 |
-
"epoch":
|
343 |
-
"eval_accuracy": 0.
|
344 |
-
"eval_loss": 1.
|
345 |
-
"eval_runtime":
|
346 |
-
"eval_samples_per_second":
|
347 |
-
"eval_steps_per_second":
|
348 |
"step": 2062
|
349 |
},
|
350 |
{
|
351 |
-
"epoch":
|
352 |
-
"eval_exact_match":
|
353 |
-
"eval_f1":
|
354 |
-
"eval_qa_bleu":
|
355 |
-
"eval_qa_exact_match": 0.
|
356 |
-
"eval_recite_bleu":
|
357 |
-
"eval_recite_exact_match": 0.
|
358 |
"step": 2062
|
359 |
},
|
360 |
{
|
361 |
"epoch": 11.2,
|
362 |
-
"grad_norm": 1.708260416984558,
|
363 |
"learning_rate": 3e-05,
|
364 |
-
"loss":
|
365 |
"step": 2100
|
366 |
},
|
367 |
{
|
368 |
-
"epoch": 11.
|
369 |
-
"grad_norm": 1.7026830911636353,
|
370 |
"learning_rate": 3e-05,
|
371 |
-
"loss":
|
372 |
"step": 2200
|
373 |
},
|
374 |
{
|
375 |
"epoch": 12.0,
|
376 |
-
"eval_accuracy": 0.
|
377 |
-
"eval_loss": 1.
|
378 |
-
"eval_runtime":
|
379 |
-
"eval_samples_per_second":
|
380 |
-
"eval_steps_per_second":
|
381 |
"step": 2250
|
382 |
},
|
383 |
{
|
384 |
"epoch": 12.0,
|
385 |
-
"eval_exact_match":
|
386 |
-
"eval_f1":
|
387 |
-
"eval_qa_bleu":
|
388 |
-
"eval_qa_exact_match": 0.
|
389 |
-
"eval_recite_bleu":
|
390 |
-
"eval_recite_exact_match": 0.
|
391 |
"step": 2250
|
392 |
},
|
393 |
{
|
394 |
-
"epoch": 12.
|
395 |
-
"grad_norm": 1.6281794309616089,
|
396 |
"learning_rate": 3e-05,
|
397 |
-
"loss":
|
398 |
"step": 2300
|
399 |
},
|
400 |
{
|
401 |
"epoch": 12.8,
|
402 |
-
"grad_norm": 1.686490535736084,
|
403 |
"learning_rate": 3e-05,
|
404 |
-
"loss":
|
405 |
"step": 2400
|
406 |
},
|
407 |
{
|
408 |
-
"epoch":
|
409 |
-
"eval_accuracy": 0.
|
410 |
-
"eval_loss": 1.
|
411 |
-
"eval_runtime":
|
412 |
-
"eval_samples_per_second":
|
413 |
-
"eval_steps_per_second":
|
414 |
"step": 2437
|
415 |
},
|
416 |
{
|
417 |
-
"epoch":
|
418 |
-
"eval_exact_match":
|
419 |
-
"eval_f1":
|
420 |
-
"eval_qa_bleu":
|
421 |
-
"eval_qa_exact_match": 0.
|
422 |
-
"eval_recite_bleu":
|
423 |
-
"eval_recite_exact_match": 0.
|
424 |
"step": 2437
|
425 |
},
|
426 |
{
|
427 |
-
"epoch": 13.
|
428 |
-
"grad_norm": 1.8666125535964966,
|
429 |
"learning_rate": 3e-05,
|
430 |
-
"loss":
|
431 |
"step": 2500
|
432 |
},
|
433 |
{
|
434 |
-
"epoch": 13.
|
435 |
-
"grad_norm": 1.9091932773590088,
|
436 |
"learning_rate": 3e-05,
|
437 |
-
"loss":
|
438 |
"step": 2600
|
439 |
},
|
440 |
{
|
441 |
"epoch": 14.0,
|
442 |
-
"eval_accuracy": 0.
|
443 |
-
"eval_loss": 1.
|
444 |
-
"eval_runtime": 6.
|
445 |
-
"eval_samples_per_second":
|
446 |
-
"eval_steps_per_second": 9.
|
447 |
"step": 2625
|
448 |
},
|
449 |
{
|
450 |
"epoch": 14.0,
|
451 |
-
"eval_exact_match":
|
452 |
-
"eval_f1":
|
453 |
-
"eval_qa_bleu":
|
454 |
-
"eval_qa_exact_match": 0.
|
455 |
-
"eval_recite_bleu":
|
456 |
-
"eval_recite_exact_match": 0.
|
457 |
"step": 2625
|
458 |
},
|
459 |
{
|
460 |
"epoch": 14.4,
|
461 |
-
"grad_norm": 2.042712688446045,
|
462 |
"learning_rate": 3e-05,
|
463 |
-
"loss":
|
464 |
"step": 2700
|
465 |
},
|
466 |
{
|
467 |
-
"epoch": 14.
|
468 |
-
"grad_norm": 1.8875784873962402,
|
469 |
"learning_rate": 3e-05,
|
470 |
-
"loss":
|
471 |
"step": 2800
|
472 |
},
|
473 |
{
|
474 |
-
"epoch":
|
475 |
-
"eval_accuracy": 0.
|
476 |
-
"eval_loss": 1.
|
477 |
-
"eval_runtime": 6.
|
478 |
-
"eval_samples_per_second":
|
479 |
-
"eval_steps_per_second":
|
480 |
"step": 2812
|
481 |
},
|
482 |
{
|
483 |
-
"epoch":
|
484 |
-
"eval_exact_match":
|
485 |
-
"eval_f1":
|
486 |
-
"eval_qa_bleu":
|
487 |
-
"eval_qa_exact_match": 0.
|
488 |
-
"eval_recite_bleu": 12.
|
489 |
-
"eval_recite_exact_match": 0.
|
490 |
"step": 2812
|
491 |
},
|
492 |
{
|
493 |
-
"epoch": 15.
|
494 |
-
"grad_norm": 2.027426242828369,
|
495 |
"learning_rate": 3e-05,
|
496 |
-
"loss": 0.
|
497 |
"step": 2900
|
498 |
},
|
499 |
{
|
500 |
"epoch": 16.0,
|
501 |
-
"grad_norm": 2.0708699226379395,
|
502 |
"learning_rate": 3e-05,
|
503 |
-
"loss": 0.
|
504 |
"step": 3000
|
505 |
},
|
506 |
{
|
507 |
"epoch": 16.0,
|
508 |
-
"eval_accuracy": 0.
|
509 |
-
"eval_loss":
|
510 |
-
"eval_runtime":
|
511 |
-
"eval_samples_per_second":
|
512 |
-
"eval_steps_per_second":
|
513 |
"step": 3000
|
514 |
},
|
515 |
{
|
516 |
"epoch": 16.0,
|
517 |
-
"eval_exact_match":
|
518 |
-
"eval_f1":
|
519 |
-
"eval_qa_bleu": 9.
|
520 |
-
"eval_qa_exact_match": 0.
|
521 |
-
"eval_recite_bleu": 12.
|
522 |
-
"eval_recite_exact_match": 0.
|
523 |
"step": 3000
|
524 |
},
|
525 |
{
|
526 |
-
"epoch": 16.
|
527 |
-
"grad_norm": 1.995875358581543,
|
528 |
"learning_rate": 3e-05,
|
529 |
-
"loss": 0.
|
530 |
"step": 3100
|
531 |
},
|
532 |
{
|
533 |
-
"epoch":
|
534 |
-
"eval_accuracy": 0.
|
535 |
-
"eval_loss":
|
536 |
-
"eval_runtime":
|
537 |
-
"eval_samples_per_second":
|
538 |
-
"eval_steps_per_second":
|
539 |
"step": 3187
|
540 |
},
|
541 |
{
|
542 |
-
"epoch":
|
543 |
-
"eval_exact_match":
|
544 |
-
"eval_f1":
|
545 |
-
"eval_qa_bleu":
|
546 |
-
"eval_qa_exact_match": 0.
|
547 |
-
"eval_recite_bleu": 12.
|
548 |
-
"eval_recite_exact_match": 0.
|
549 |
"step": 3187
|
550 |
},
|
551 |
{
|
552 |
-
"epoch": 17.
|
553 |
-
"grad_norm": 2.623262643814087,
|
554 |
"learning_rate": 3e-05,
|
555 |
-
"loss": 0.
|
556 |
"step": 3200
|
557 |
},
|
558 |
{
|
559 |
"epoch": 17.6,
|
560 |
-
"grad_norm": 2.4125397205352783,
|
561 |
"learning_rate": 3e-05,
|
562 |
-
"loss": 0.
|
563 |
"step": 3300
|
564 |
},
|
565 |
{
|
566 |
"epoch": 18.0,
|
567 |
-
"eval_accuracy": 0.
|
568 |
-
"eval_loss":
|
569 |
-
"eval_runtime":
|
570 |
-
"eval_samples_per_second":
|
571 |
-
"eval_steps_per_second":
|
572 |
"step": 3375
|
573 |
},
|
574 |
{
|
575 |
"epoch": 18.0,
|
576 |
-
"eval_exact_match":
|
577 |
-
"eval_f1":
|
578 |
-
"eval_qa_bleu":
|
579 |
-
"eval_qa_exact_match": 0.
|
580 |
-
"eval_recite_bleu": 12.
|
581 |
-
"eval_recite_exact_match": 0.
|
582 |
"step": 3375
|
583 |
},
|
584 |
{
|
585 |
-
"epoch": 18.
|
586 |
-
"grad_norm": 2.3584065437316895,
|
587 |
"learning_rate": 3e-05,
|
588 |
-
"loss": 0.
|
589 |
"step": 3400
|
590 |
},
|
591 |
{
|
592 |
-
"epoch": 18.
|
593 |
-
"grad_norm": 2.130357265472412,
|
594 |
"learning_rate": 3e-05,
|
595 |
-
"loss": 0.
|
596 |
"step": 3500
|
597 |
},
|
598 |
{
|
599 |
-
"epoch":
|
600 |
-
"eval_accuracy": 0.
|
601 |
-
"eval_loss":
|
602 |
-
"eval_runtime":
|
603 |
-
"eval_samples_per_second":
|
604 |
-
"eval_steps_per_second":
|
605 |
"step": 3562
|
606 |
},
|
607 |
{
|
608 |
-
"epoch":
|
609 |
-
"eval_exact_match":
|
610 |
-
"eval_f1":
|
611 |
-
"eval_qa_bleu":
|
612 |
-
"eval_qa_exact_match": 0.
|
613 |
-
"eval_recite_bleu": 12.
|
614 |
-
"eval_recite_exact_match": 0.
|
615 |
"step": 3562
|
616 |
},
|
617 |
{
|
618 |
"epoch": 19.2,
|
619 |
-
"grad_norm": 2.6397740840911865,
|
620 |
"learning_rate": 3e-05,
|
621 |
-
"loss": 0.
|
622 |
"step": 3600
|
623 |
},
|
624 |
{
|
625 |
-
"epoch": 19.
|
626 |
-
"grad_norm": 3.049431085586548,
|
627 |
"learning_rate": 3e-05,
|
628 |
-
"loss": 0.
|
629 |
"step": 3700
|
630 |
},
|
631 |
{
|
632 |
"epoch": 20.0,
|
633 |
-
"eval_accuracy": 0.
|
634 |
-
"eval_loss":
|
635 |
-
"eval_runtime": 6.
|
636 |
-
"eval_samples_per_second":
|
637 |
-
"eval_steps_per_second": 9.
|
638 |
"step": 3750
|
639 |
},
|
640 |
{
|
641 |
"epoch": 20.0,
|
642 |
-
"eval_exact_match":
|
643 |
-
"eval_f1":
|
644 |
-
"eval_qa_bleu":
|
645 |
-
"eval_qa_exact_match": 0.
|
646 |
-
"eval_recite_bleu": 11.
|
647 |
-
"eval_recite_exact_match": 0.
|
648 |
"step": 3750
|
649 |
},
|
650 |
{
|
651 |
-
"epoch": 20.
|
652 |
-
"grad_norm": 2.647123098373413,
|
653 |
"learning_rate": 3e-05,
|
654 |
-
"loss": 0.
|
655 |
"step": 3800
|
656 |
},
|
657 |
{
|
658 |
"epoch": 20.8,
|
659 |
-
"grad_norm": 2.798877239227295,
|
660 |
"learning_rate": 3e-05,
|
661 |
-
"loss": 0.
|
662 |
"step": 3900
|
663 |
},
|
664 |
{
|
665 |
-
"epoch":
|
666 |
-
"eval_accuracy": 0.
|
667 |
-
"eval_loss":
|
668 |
-
"eval_runtime":
|
669 |
-
"eval_samples_per_second":
|
670 |
-
"eval_steps_per_second":
|
671 |
"step": 3937
|
672 |
},
|
673 |
{
|
674 |
-
"epoch":
|
675 |
-
"eval_exact_match":
|
676 |
-
"eval_f1":
|
677 |
-
"eval_qa_bleu":
|
678 |
-
"eval_qa_exact_match": 0.
|
679 |
-
"eval_recite_bleu":
|
680 |
-
"eval_recite_exact_match": 0.
|
681 |
"step": 3937
|
682 |
},
|
683 |
{
|
684 |
-
"epoch": 21.
|
685 |
-
"grad_norm": 2.7208638191223145,
|
686 |
"learning_rate": 3e-05,
|
687 |
-
"loss": 0.
|
688 |
"step": 4000
|
689 |
},
|
690 |
{
|
691 |
-
"epoch": 21.
|
692 |
-
"grad_norm": 2.9887781143188477,
|
693 |
"learning_rate": 3e-05,
|
694 |
-
"loss": 0.
|
695 |
"step": 4100
|
696 |
},
|
697 |
{
|
698 |
"epoch": 22.0,
|
699 |
-
"eval_accuracy": 0.
|
700 |
-
"eval_loss":
|
701 |
-
"eval_runtime":
|
702 |
-
"eval_samples_per_second":
|
703 |
-
"eval_steps_per_second":
|
704 |
"step": 4125
|
705 |
},
|
706 |
{
|
707 |
"epoch": 22.0,
|
708 |
-
"eval_exact_match":
|
709 |
-
"eval_f1":
|
710 |
-
"eval_qa_bleu":
|
711 |
-
"eval_qa_exact_match": 0.
|
712 |
-
"eval_recite_bleu": 11.
|
713 |
-
"eval_recite_exact_match": 0.
|
714 |
"step": 4125
|
715 |
},
|
716 |
{
|
717 |
"epoch": 22.4,
|
718 |
-
"grad_norm": 2.706368923187256,
|
719 |
"learning_rate": 3e-05,
|
720 |
-
"loss": 0.
|
721 |
"step": 4200
|
722 |
},
|
723 |
{
|
724 |
-
"epoch": 22.
|
725 |
-
"grad_norm": 2.850799798965454,
|
726 |
"learning_rate": 3e-05,
|
727 |
-
"loss": 0.
|
728 |
"step": 4300
|
729 |
},
|
730 |
{
|
731 |
-
"epoch":
|
732 |
-
"eval_accuracy": 0.
|
733 |
-
"eval_loss":
|
734 |
-
"eval_runtime":
|
735 |
-
"eval_samples_per_second":
|
736 |
-
"eval_steps_per_second":
|
737 |
"step": 4312
|
738 |
},
|
739 |
{
|
740 |
-
"epoch":
|
741 |
-
"eval_exact_match": 8
|
742 |
-
"eval_f1":
|
743 |
-
"eval_qa_bleu":
|
744 |
-
"eval_qa_exact_match": 0.
|
745 |
-
"eval_recite_bleu": 11.
|
746 |
-
"eval_recite_exact_match": 0.
|
747 |
"step": 4312
|
748 |
},
|
749 |
{
|
750 |
-
"epoch": 23.
|
751 |
-
"grad_norm": 2.95532488822937,
|
752 |
"learning_rate": 3e-05,
|
753 |
-
"loss": 0.
|
754 |
"step": 4400
|
755 |
},
|
756 |
{
|
757 |
"epoch": 24.0,
|
758 |
-
"grad_norm": 3.36663556098938,
|
759 |
"learning_rate": 3e-05,
|
760 |
-
"loss": 0.
|
761 |
"step": 4500
|
762 |
},
|
763 |
{
|
764 |
"epoch": 24.0,
|
765 |
-
"eval_accuracy": 0.
|
766 |
-
"eval_loss":
|
767 |
-
"eval_runtime":
|
768 |
-
"eval_samples_per_second":
|
769 |
-
"eval_steps_per_second":
|
770 |
"step": 4500
|
771 |
},
|
772 |
{
|
773 |
"epoch": 24.0,
|
774 |
-
"eval_exact_match":
|
775 |
-
"eval_f1":
|
776 |
-
"eval_qa_bleu":
|
777 |
-
"eval_qa_exact_match": 0.
|
778 |
-
"eval_recite_bleu":
|
779 |
-
"eval_recite_exact_match": 0.
|
780 |
"step": 4500
|
781 |
},
|
782 |
{
|
783 |
-
"epoch": 24.
|
784 |
-
"grad_norm": 3.074622631072998,
|
785 |
"learning_rate": 3e-05,
|
786 |
-
"loss": 0.
|
787 |
"step": 4600
|
788 |
},
|
789 |
{
|
790 |
-
"epoch":
|
791 |
-
"eval_accuracy": 0.
|
792 |
-
"eval_loss":
|
793 |
-
"eval_runtime":
|
794 |
-
"eval_samples_per_second":
|
795 |
-
"eval_steps_per_second":
|
796 |
"step": 4687
|
797 |
},
|
798 |
{
|
799 |
-
"epoch":
|
800 |
-
"eval_exact_match":
|
801 |
-
"eval_f1":
|
802 |
-
"eval_qa_bleu":
|
803 |
-
"eval_qa_exact_match": 0.
|
804 |
-
"eval_recite_bleu": 11.
|
805 |
-
"eval_recite_exact_match": 0.
|
806 |
"step": 4687
|
807 |
},
|
808 |
{
|
809 |
-
"epoch": 25.
|
810 |
-
"grad_norm": 2.702502727508545,
|
811 |
"learning_rate": 3e-05,
|
812 |
-
"loss": 0.
|
813 |
"step": 4700
|
814 |
},
|
815 |
{
|
816 |
"epoch": 25.6,
|
817 |
-
"grad_norm": 3.233320951461792,
|
818 |
"learning_rate": 3e-05,
|
819 |
-
"loss": 0.
|
820 |
"step": 4800
|
821 |
},
|
822 |
{
|
823 |
"epoch": 26.0,
|
824 |
-
"eval_accuracy": 0.
|
825 |
-
"eval_loss": 2.
|
826 |
-
"eval_runtime": 6.
|
827 |
-
"eval_samples_per_second":
|
828 |
-
"eval_steps_per_second": 9.
|
829 |
"step": 4875
|
830 |
},
|
831 |
{
|
832 |
"epoch": 26.0,
|
833 |
-
"eval_exact_match":
|
834 |
-
"eval_f1":
|
835 |
-
"eval_qa_bleu":
|
836 |
-
"eval_qa_exact_match": 0.
|
837 |
-
"eval_recite_bleu":
|
838 |
-
"eval_recite_exact_match": 0.
|
839 |
"step": 4875
|
840 |
},
|
841 |
{
|
842 |
-
"epoch": 26.
|
843 |
-
"grad_norm": 3.280097246170044,
|
844 |
"learning_rate": 3e-05,
|
845 |
-
"loss": 0.
|
846 |
"step": 4900
|
847 |
},
|
848 |
{
|
849 |
-
"epoch": 26.
|
850 |
-
"grad_norm": 3.1984827518463135,
|
851 |
"learning_rate": 3e-05,
|
852 |
-
"loss": 0.
|
853 |
"step": 5000
|
854 |
},
|
855 |
{
|
856 |
-
"epoch":
|
857 |
-
"eval_accuracy": 0.
|
858 |
-
"eval_loss": 2.
|
859 |
-
"eval_runtime":
|
860 |
-
"eval_samples_per_second":
|
861 |
-
"eval_steps_per_second":
|
862 |
"step": 5062
|
863 |
},
|
864 |
{
|
865 |
-
"epoch":
|
866 |
-
"eval_exact_match":
|
867 |
-
"eval_f1":
|
868 |
-
"eval_qa_bleu":
|
869 |
-
"eval_qa_exact_match": 0.
|
870 |
-
"eval_recite_bleu":
|
871 |
-
"eval_recite_exact_match": 0.
|
872 |
"step": 5062
|
873 |
},
|
874 |
{
|
875 |
"epoch": 27.2,
|
876 |
-
"grad_norm": 3.5599937438964844,
|
877 |
"learning_rate": 3e-05,
|
878 |
-
"loss": 0.
|
879 |
"step": 5100
|
880 |
},
|
881 |
{
|
882 |
-
"epoch": 27.
|
883 |
-
"grad_norm": 2.9382145404815674,
|
884 |
"learning_rate": 3e-05,
|
885 |
-
"loss": 0.
|
886 |
"step": 5200
|
887 |
},
|
888 |
{
|
889 |
"epoch": 28.0,
|
890 |
-
"eval_accuracy": 0.
|
891 |
-
"eval_loss": 2.
|
892 |
-
"eval_runtime":
|
893 |
-
"eval_samples_per_second":
|
894 |
-
"eval_steps_per_second":
|
895 |
"step": 5250
|
896 |
},
|
897 |
{
|
898 |
"epoch": 28.0,
|
899 |
-
"eval_exact_match": 8
|
900 |
-
"eval_f1":
|
901 |
-
"eval_qa_bleu":
|
902 |
-
"eval_qa_exact_match": 0.
|
903 |
-
"eval_recite_bleu": 11.
|
904 |
-
"eval_recite_exact_match": 0.
|
905 |
"step": 5250
|
906 |
},
|
907 |
{
|
908 |
-
"epoch": 28.
|
909 |
-
"grad_norm": 4.097602367401123,
|
910 |
"learning_rate": 3e-05,
|
911 |
-
"loss": 0.
|
912 |
"step": 5300
|
913 |
},
|
914 |
{
|
915 |
"epoch": 28.8,
|
916 |
-
"grad_norm": 3.9295685291290283,
|
917 |
"learning_rate": 3e-05,
|
918 |
-
"loss": 0.
|
919 |
"step": 5400
|
920 |
},
|
921 |
{
|
922 |
-
"epoch":
|
923 |
-
"eval_accuracy": 0.
|
924 |
-
"eval_loss": 2.
|
925 |
-
"eval_runtime":
|
926 |
-
"eval_samples_per_second":
|
927 |
-
"eval_steps_per_second":
|
928 |
"step": 5437
|
929 |
},
|
930 |
{
|
931 |
-
"epoch":
|
932 |
-
"eval_exact_match":
|
933 |
-
"eval_f1":
|
934 |
-
"eval_qa_bleu": 4.
|
935 |
-
"eval_qa_exact_match": 0.
|
936 |
-
"eval_recite_bleu": 11.
|
937 |
-
"eval_recite_exact_match": 0.
|
938 |
"step": 5437
|
939 |
},
|
940 |
{
|
941 |
-
"epoch": 29.
|
942 |
-
"grad_norm": 3.4413976669311523,
|
943 |
"learning_rate": 3e-05,
|
944 |
-
"loss": 0.
|
945 |
"step": 5500
|
946 |
},
|
947 |
{
|
948 |
-
"epoch": 29.
|
949 |
-
"grad_norm": 2.88114857673645,
|
950 |
"learning_rate": 3e-05,
|
951 |
-
"loss": 0.
|
952 |
"step": 5600
|
953 |
},
|
954 |
{
|
955 |
"epoch": 30.0,
|
956 |
-
"eval_accuracy": 0.
|
957 |
-
"eval_loss": 2.
|
958 |
-
"eval_runtime":
|
959 |
-
"eval_samples_per_second":
|
960 |
-
"eval_steps_per_second":
|
961 |
"step": 5625
|
962 |
},
|
963 |
{
|
964 |
"epoch": 30.0,
|
965 |
-
"eval_exact_match": 8
|
966 |
-
"eval_f1":
|
967 |
-
"eval_qa_bleu":
|
968 |
-
"eval_qa_exact_match": 0.
|
969 |
-
"eval_recite_bleu": 11.
|
970 |
-
"eval_recite_exact_match": 0.
|
971 |
"step": 5625
|
972 |
},
|
973 |
{
|
974 |
"epoch": 30.4,
|
975 |
-
"grad_norm": 4.096495628356934,
|
976 |
"learning_rate": 3e-05,
|
977 |
-
"loss": 0.
|
978 |
"step": 5700
|
979 |
},
|
980 |
{
|
981 |
-
"epoch": 30.
|
982 |
-
"grad_norm": 3.567011833190918,
|
983 |
"learning_rate": 3e-05,
|
984 |
-
"loss": 0.
|
985 |
"step": 5800
|
986 |
},
|
987 |
{
|
988 |
-
"epoch":
|
989 |
-
"eval_accuracy": 0.
|
990 |
-
"eval_loss": 2.
|
991 |
-
"eval_runtime":
|
992 |
-
"eval_samples_per_second":
|
993 |
-
"eval_steps_per_second":
|
994 |
"step": 5812
|
995 |
},
|
996 |
{
|
997 |
-
"epoch":
|
998 |
-
"eval_exact_match":
|
999 |
-
"eval_f1":
|
1000 |
-
"eval_qa_bleu":
|
1001 |
-
"eval_qa_exact_match": 0.
|
1002 |
-
"eval_recite_bleu":
|
1003 |
-
"eval_recite_exact_match": 0.
|
1004 |
"step": 5812
|
1005 |
},
|
1006 |
{
|
1007 |
-
"epoch": 31.
|
1008 |
-
"grad_norm": 3.1184980869293213,
|
1009 |
"learning_rate": 3e-05,
|
1010 |
-
"loss": 0.
|
1011 |
"step": 5900
|
1012 |
},
|
1013 |
{
|
1014 |
"epoch": 32.0,
|
1015 |
-
"grad_norm": 3.125758647918701,
|
1016 |
"learning_rate": 3e-05,
|
1017 |
-
"loss": 0.
|
1018 |
"step": 6000
|
1019 |
},
|
1020 |
{
|
1021 |
"epoch": 32.0,
|
1022 |
-
"eval_accuracy": 0.
|
1023 |
-
"eval_loss": 2.
|
1024 |
-
"eval_runtime":
|
1025 |
-
"eval_samples_per_second":
|
1026 |
-
"eval_steps_per_second":
|
1027 |
"step": 6000
|
1028 |
},
|
1029 |
{
|
1030 |
"epoch": 32.0,
|
1031 |
-
"eval_exact_match":
|
1032 |
-
"eval_f1":
|
1033 |
-
"eval_qa_bleu":
|
1034 |
-
"eval_qa_exact_match": 0.
|
1035 |
-
"eval_recite_bleu":
|
1036 |
"eval_recite_exact_match": 0.008,
|
1037 |
"step": 6000
|
1038 |
},
|
1039 |
{
|
1040 |
-
"epoch": 32.
|
1041 |
-
"grad_norm": 2.6737849712371826,
|
1042 |
"learning_rate": 3e-05,
|
1043 |
-
"loss": 0.
|
1044 |
"step": 6100
|
1045 |
},
|
1046 |
{
|
1047 |
-
"epoch":
|
1048 |
-
"eval_accuracy": 0.
|
1049 |
-
"eval_loss": 2.
|
1050 |
-
"eval_runtime":
|
1051 |
-
"eval_samples_per_second":
|
1052 |
-
"eval_steps_per_second":
|
1053 |
"step": 6187
|
1054 |
},
|
1055 |
{
|
1056 |
-
"epoch":
|
1057 |
-
"eval_exact_match":
|
1058 |
-
"eval_f1":
|
1059 |
-
"eval_qa_bleu":
|
1060 |
-
"eval_qa_exact_match": 0.
|
1061 |
-
"eval_recite_bleu":
|
1062 |
-
"eval_recite_exact_match": 0.
|
1063 |
"step": 6187
|
1064 |
},
|
1065 |
{
|
1066 |
-
"epoch": 33.
|
1067 |
-
"grad_norm": 2.490464448928833,
|
1068 |
"learning_rate": 3e-05,
|
1069 |
-
"loss": 0.
|
1070 |
"step": 6200
|
1071 |
},
|
1072 |
{
|
1073 |
"epoch": 33.6,
|
1074 |
-
"grad_norm": 3.0587589740753174,
|
1075 |
"learning_rate": 3e-05,
|
1076 |
-
"loss": 0.
|
1077 |
"step": 6300
|
1078 |
},
|
1079 |
{
|
1080 |
"epoch": 34.0,
|
1081 |
-
"eval_accuracy": 0.
|
1082 |
-
"eval_loss": 2.
|
1083 |
-
"eval_runtime":
|
1084 |
-
"eval_samples_per_second":
|
1085 |
-
"eval_steps_per_second":
|
1086 |
"step": 6375
|
1087 |
},
|
1088 |
{
|
1089 |
"epoch": 34.0,
|
1090 |
-
"eval_exact_match":
|
1091 |
-
"eval_f1":
|
1092 |
-
"eval_qa_bleu":
|
1093 |
-
"eval_qa_exact_match": 0.
|
1094 |
-
"eval_recite_bleu": 11.
|
1095 |
-
"eval_recite_exact_match": 0.
|
1096 |
"step": 6375
|
1097 |
},
|
1098 |
{
|
1099 |
-
"epoch": 34.
|
1100 |
-
"grad_norm": 3.1941940784454346,
|
1101 |
"learning_rate": 3e-05,
|
1102 |
-
"loss": 0.
|
1103 |
"step": 6400
|
1104 |
},
|
1105 |
{
|
1106 |
-
"epoch": 34.
|
1107 |
-
"grad_norm": 3.2870864868164062,
|
1108 |
"learning_rate": 3e-05,
|
1109 |
-
"loss": 0.
|
1110 |
"step": 6500
|
1111 |
},
|
1112 |
{
|
1113 |
-
"epoch":
|
1114 |
-
"eval_accuracy": 0.
|
1115 |
-
"eval_loss":
|
1116 |
-
"eval_runtime": 6.
|
1117 |
-
"eval_samples_per_second":
|
1118 |
-
"eval_steps_per_second": 9.
|
1119 |
"step": 6562
|
1120 |
},
|
1121 |
{
|
1122 |
-
"epoch":
|
1123 |
-
"eval_exact_match":
|
1124 |
-
"eval_f1":
|
1125 |
-
"eval_qa_bleu": 6.
|
1126 |
-
"eval_qa_exact_match": 0.
|
1127 |
-
"eval_recite_bleu":
|
1128 |
-
"eval_recite_exact_match": 0.
|
1129 |
"step": 6562
|
1130 |
},
|
1131 |
{
|
1132 |
"epoch": 35.2,
|
1133 |
-
"grad_norm": 3.2354586124420166,
|
1134 |
"learning_rate": 3e-05,
|
1135 |
-
"loss": 0.
|
1136 |
"step": 6600
|
1137 |
},
|
1138 |
{
|
1139 |
-
"epoch": 35.
|
1140 |
-
"grad_norm": 3.262101173400879,
|
1141 |
"learning_rate": 3e-05,
|
1142 |
-
"loss": 0.
|
1143 |
"step": 6700
|
1144 |
},
|
1145 |
{
|
1146 |
"epoch": 36.0,
|
1147 |
-
"eval_accuracy": 0.
|
1148 |
-
"eval_loss":
|
1149 |
-
"eval_runtime":
|
1150 |
-
"eval_samples_per_second":
|
1151 |
-
"eval_steps_per_second":
|
1152 |
"step": 6750
|
1153 |
},
|
1154 |
{
|
1155 |
"epoch": 36.0,
|
1156 |
-
"eval_exact_match":
|
1157 |
-
"eval_f1":
|
1158 |
-
"eval_qa_bleu":
|
1159 |
-
"eval_qa_exact_match": 0.
|
1160 |
-
"eval_recite_bleu":
|
1161 |
-
"eval_recite_exact_match": 0.
|
1162 |
"step": 6750
|
1163 |
},
|
1164 |
{
|
1165 |
-
"epoch": 36.
|
1166 |
-
"grad_norm": 2.951231002807617,
|
1167 |
"learning_rate": 3e-05,
|
1168 |
-
"loss": 0.
|
1169 |
"step": 6800
|
1170 |
},
|
1171 |
{
|
1172 |
"epoch": 36.8,
|
1173 |
-
"grad_norm": 3.770897626876831,
|
1174 |
"learning_rate": 3e-05,
|
1175 |
-
"loss": 0.
|
1176 |
"step": 6900
|
1177 |
},
|
1178 |
{
|
1179 |
-
"epoch":
|
1180 |
-
"eval_accuracy": 0.
|
1181 |
-
"eval_loss":
|
1182 |
-
"eval_runtime":
|
1183 |
-
"eval_samples_per_second":
|
1184 |
-
"eval_steps_per_second":
|
1185 |
"step": 6937
|
1186 |
},
|
1187 |
{
|
1188 |
-
"epoch":
|
1189 |
-
"eval_exact_match":
|
1190 |
-
"eval_f1":
|
1191 |
-
"eval_qa_bleu":
|
1192 |
-
"eval_qa_exact_match": 0.
|
1193 |
-
"eval_recite_bleu":
|
1194 |
-
"eval_recite_exact_match": 0.
|
1195 |
"step": 6937
|
1196 |
},
|
1197 |
{
|
1198 |
-
"epoch": 37.
|
1199 |
-
"grad_norm": 3.664700984954834,
|
1200 |
"learning_rate": 3e-05,
|
1201 |
-
"loss": 0.
|
1202 |
"step": 7000
|
1203 |
},
|
1204 |
{
|
1205 |
-
"epoch": 37.
|
1206 |
-
"grad_norm": 3.442922830581665,
|
1207 |
"learning_rate": 3e-05,
|
1208 |
-
"loss": 0.
|
1209 |
"step": 7100
|
1210 |
},
|
1211 |
{
|
1212 |
"epoch": 38.0,
|
1213 |
-
"eval_accuracy": 0.
|
1214 |
-
"eval_loss":
|
1215 |
-
"eval_runtime": 6.
|
1216 |
-
"eval_samples_per_second": 74.
|
1217 |
-
"eval_steps_per_second": 9.
|
1218 |
"step": 7125
|
1219 |
},
|
1220 |
{
|
1221 |
"epoch": 38.0,
|
1222 |
-
"eval_exact_match":
|
1223 |
-
"eval_f1":
|
1224 |
-
"eval_qa_bleu": 6.
|
1225 |
-
"eval_qa_exact_match": 0.
|
1226 |
-
"eval_recite_bleu":
|
1227 |
-
"eval_recite_exact_match": 0.
|
1228 |
"step": 7125
|
1229 |
}
|
1230 |
],
|
1231 |
"logging_steps": 100,
|
1232 |
"max_steps": 9350,
|
1233 |
-
"num_input_tokens_seen": 0,
|
1234 |
"num_train_epochs": 50,
|
1235 |
"save_steps": 500,
|
1236 |
-
"
|
1237 |
-
"TrainerControl": {
|
1238 |
-
"args": {
|
1239 |
-
"should_epoch_stop": false,
|
1240 |
-
"should_evaluate": false,
|
1241 |
-
"should_log": false,
|
1242 |
-
"should_save": true,
|
1243 |
-
"should_training_stop": false
|
1244 |
-
},
|
1245 |
-
"attributes": {}
|
1246 |
-
}
|
1247 |
-
},
|
1248 |
-
"total_flos": 7.862973798872515e+17,
|
1249 |
-
"train_batch_size": 2,
|
1250 |
"trial_name": null,
|
1251 |
"trial_params": null
|
1252 |
}
|
|
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
+
"epoch": 0.53,
|
|
|
13 |
"learning_rate": 3e-05,
|
14 |
+
"loss": 1.4597,
|
15 |
"step": 100
|
16 |
},
|
17 |
{
|
18 |
+
"epoch": 1.0,
|
19 |
+
"eval_accuracy": 0.6618980392156862,
|
20 |
+
"eval_loss": 1.2718806266784668,
|
21 |
+
"eval_runtime": 7.6194,
|
22 |
+
"eval_samples_per_second": 65.622,
|
23 |
+
"eval_steps_per_second": 8.268,
|
24 |
"step": 187
|
25 |
},
|
26 |
{
|
27 |
+
"epoch": 1.0,
|
28 |
+
"eval_exact_match": 22.4,
|
29 |
+
"eval_f1": 30.941064425770318,
|
30 |
+
"eval_qa_bleu": 7.097403515221558,
|
31 |
+
"eval_qa_exact_match": 0.196,
|
32 |
+
"eval_recite_bleu": 13.258373740102225,
|
33 |
"eval_recite_exact_match": 0.0,
|
34 |
"step": 187
|
35 |
},
|
36 |
{
|
37 |
+
"epoch": 1.07,
|
|
|
38 |
"learning_rate": 3e-05,
|
39 |
+
"loss": 1.2905,
|
40 |
"step": 200
|
41 |
},
|
42 |
{
|
43 |
"epoch": 1.6,
|
|
|
44 |
"learning_rate": 3e-05,
|
45 |
+
"loss": 1.2157,
|
46 |
"step": 300
|
47 |
},
|
48 |
{
|
49 |
"epoch": 2.0,
|
50 |
+
"eval_accuracy": 0.6670039215686274,
|
51 |
+
"eval_loss": 1.206207036972046,
|
52 |
+
"eval_runtime": 7.6793,
|
53 |
+
"eval_samples_per_second": 65.11,
|
54 |
+
"eval_steps_per_second": 8.204,
|
55 |
"step": 375
|
56 |
},
|
57 |
{
|
58 |
"epoch": 2.0,
|
59 |
+
"eval_exact_match": 3.2,
|
60 |
+
"eval_f1": 3.96976236704394,
|
61 |
+
"eval_qa_bleu": 0.20484366294597936,
|
62 |
+
"eval_qa_exact_match": 0.026,
|
63 |
+
"eval_recite_bleu": 2.5778714769111824e-05,
|
64 |
"eval_recite_exact_match": 0.0,
|
65 |
"step": 375
|
66 |
},
|
67 |
{
|
68 |
+
"epoch": 2.13,
|
|
|
69 |
"learning_rate": 3e-05,
|
70 |
+
"loss": 1.1852,
|
71 |
"step": 400
|
72 |
},
|
73 |
{
|
74 |
+
"epoch": 2.67,
|
|
|
75 |
"learning_rate": 3e-05,
|
76 |
+
"loss": 1.1861,
|
77 |
"step": 500
|
78 |
},
|
79 |
{
|
80 |
+
"epoch": 3.0,
|
81 |
+
"eval_accuracy": 0.667513725490196,
|
82 |
+
"eval_loss": 1.19828462600708,
|
83 |
+
"eval_runtime": 7.1425,
|
84 |
+
"eval_samples_per_second": 70.004,
|
85 |
+
"eval_steps_per_second": 8.82,
|
86 |
"step": 562
|
87 |
},
|
88 |
{
|
89 |
+
"epoch": 3.0,
|
90 |
+
"eval_exact_match": 4.4,
|
91 |
+
"eval_f1": 6.333660342551587,
|
92 |
+
"eval_qa_bleu": 0.522391008398391,
|
93 |
+
"eval_qa_exact_match": 0.028,
|
94 |
+
"eval_recite_bleu": 0.027569975851374308,
|
95 |
+
"eval_recite_exact_match": 0.0,
|
96 |
"step": 562
|
97 |
},
|
98 |
{
|
99 |
"epoch": 3.2,
|
|
|
100 |
"learning_rate": 3e-05,
|
101 |
+
"loss": 1.1722,
|
102 |
"step": 600
|
103 |
},
|
104 |
{
|
105 |
+
"epoch": 3.73,
|
|
|
106 |
"learning_rate": 3e-05,
|
107 |
+
"loss": 1.1445,
|
108 |
"step": 700
|
109 |
},
|
110 |
{
|
111 |
"epoch": 4.0,
|
112 |
+
"eval_accuracy": 0.6672,
|
113 |
+
"eval_loss": 1.197709560394287,
|
114 |
+
"eval_runtime": 7.7427,
|
115 |
+
"eval_samples_per_second": 64.577,
|
116 |
+
"eval_steps_per_second": 8.137,
|
117 |
"step": 750
|
118 |
},
|
119 |
{
|
120 |
"epoch": 4.0,
|
121 |
+
"eval_exact_match": 7.4,
|
122 |
+
"eval_f1": 10.950016966795278,
|
123 |
+
"eval_qa_bleu": 0.6700083846206139,
|
124 |
+
"eval_qa_exact_match": 0.052,
|
125 |
+
"eval_recite_bleu": 1.259776140465531,
|
126 |
+
"eval_recite_exact_match": 0.0,
|
127 |
"step": 750
|
128 |
},
|
129 |
{
|
130 |
+
"epoch": 4.27,
|
|
|
131 |
"learning_rate": 3e-05,
|
132 |
+
"loss": 1.1018,
|
133 |
"step": 800
|
134 |
},
|
135 |
{
|
136 |
"epoch": 4.8,
|
|
|
137 |
"learning_rate": 3e-05,
|
138 |
+
"loss": 1.1,
|
139 |
"step": 900
|
140 |
},
|
141 |
{
|
142 |
+
"epoch": 5.0,
|
143 |
+
"eval_accuracy": 0.6664941176470588,
|
144 |
+
"eval_loss": 1.2078713178634644,
|
145 |
+
"eval_runtime": 7.7721,
|
146 |
+
"eval_samples_per_second": 64.333,
|
147 |
+
"eval_steps_per_second": 8.106,
|
148 |
"step": 937
|
149 |
},
|
150 |
{
|
151 |
+
"epoch": 5.0,
|
152 |
+
"eval_exact_match": 6.6,
|
153 |
+
"eval_f1": 10.12048028049545,
|
154 |
+
"eval_qa_bleu": 0.8314625442591895,
|
155 |
+
"eval_qa_exact_match": 0.05,
|
156 |
+
"eval_recite_bleu": 0.8030246920772521,
|
157 |
+
"eval_recite_exact_match": 0.0,
|
158 |
"step": 937
|
159 |
},
|
160 |
{
|
161 |
+
"epoch": 5.33,
|
|
|
162 |
"learning_rate": 3e-05,
|
163 |
+
"loss": 1.0666,
|
164 |
"step": 1000
|
165 |
},
|
166 |
{
|
167 |
+
"epoch": 5.87,
|
|
|
168 |
"learning_rate": 3e-05,
|
169 |
+
"loss": 1.0439,
|
170 |
"step": 1100
|
171 |
},
|
172 |
{
|
173 |
"epoch": 6.0,
|
174 |
+
"eval_accuracy": 0.6649725490196079,
|
175 |
+
"eval_loss": 1.22383713722229,
|
176 |
+
"eval_runtime": 7.6639,
|
177 |
+
"eval_samples_per_second": 65.241,
|
178 |
+
"eval_steps_per_second": 8.22,
|
179 |
"step": 1125
|
180 |
},
|
181 |
{
|
182 |
"epoch": 6.0,
|
183 |
+
"eval_exact_match": 9.0,
|
184 |
+
"eval_f1": 14.19791549030679,
|
185 |
+
"eval_qa_bleu": 2.7960558519683327,
|
186 |
+
"eval_qa_exact_match": 0.07,
|
187 |
+
"eval_recite_bleu": 2.082318255561628,
|
188 |
+
"eval_recite_exact_match": 0.0,
|
189 |
"step": 1125
|
190 |
},
|
191 |
{
|
192 |
"epoch": 6.4,
|
|
|
193 |
"learning_rate": 3e-05,
|
194 |
+
"loss": 1.0029,
|
195 |
"step": 1200
|
196 |
},
|
197 |
{
|
198 |
+
"epoch": 6.93,
|
|
|
199 |
"learning_rate": 3e-05,
|
200 |
+
"loss": 0.9888,
|
201 |
"step": 1300
|
202 |
},
|
203 |
{
|
204 |
+
"epoch": 7.0,
|
205 |
+
"eval_accuracy": 0.6638117647058823,
|
206 |
+
"eval_loss": 1.2457345724105835,
|
207 |
+
"eval_runtime": 6.7577,
|
208 |
+
"eval_samples_per_second": 73.99,
|
209 |
+
"eval_steps_per_second": 9.323,
|
210 |
"step": 1312
|
211 |
},
|
212 |
{
|
213 |
+
"epoch": 7.0,
|
214 |
+
"eval_exact_match": 10.6,
|
215 |
+
"eval_f1": 15.30586844512479,
|
216 |
+
"eval_qa_bleu": 0.94556412718174,
|
217 |
+
"eval_qa_exact_match": 0.086,
|
218 |
+
"eval_recite_bleu": 3.1558565534273777,
|
219 |
+
"eval_recite_exact_match": 0.0,
|
220 |
"step": 1312
|
221 |
},
|
222 |
{
|
223 |
+
"epoch": 7.47,
|
|
|
224 |
"learning_rate": 3e-05,
|
225 |
+
"loss": 0.9453,
|
226 |
"step": 1400
|
227 |
},
|
228 |
{
|
229 |
"epoch": 8.0,
|
|
|
230 |
"learning_rate": 3e-05,
|
231 |
+
"loss": 0.9364,
|
232 |
"step": 1500
|
233 |
},
|
234 |
{
|
235 |
"epoch": 8.0,
|
236 |
+
"eval_accuracy": 0.6615921568627451,
|
237 |
+
"eval_loss": 1.2816376686096191,
|
238 |
+
"eval_runtime": 7.5902,
|
239 |
+
"eval_samples_per_second": 65.875,
|
240 |
+
"eval_steps_per_second": 8.3,
|
241 |
"step": 1500
|
242 |
},
|
243 |
{
|
244 |
"epoch": 8.0,
|
245 |
+
"eval_exact_match": 9.2,
|
246 |
+
"eval_f1": 12.989213707726288,
|
247 |
+
"eval_qa_bleu": 1.9797940311992752,
|
248 |
+
"eval_qa_exact_match": 0.074,
|
249 |
+
"eval_recite_bleu": 2.551621610315755,
|
250 |
+
"eval_recite_exact_match": 0.0,
|
251 |
"step": 1500
|
252 |
},
|
253 |
{
|
254 |
+
"epoch": 8.53,
|
|
|
255 |
"learning_rate": 3e-05,
|
256 |
+
"loss": 0.8889,
|
257 |
"step": 1600
|
258 |
},
|
259 |
{
|
260 |
+
"epoch": 9.0,
|
261 |
+
"eval_accuracy": 0.6606196078431372,
|
262 |
+
"eval_loss": 1.3036144971847534,
|
263 |
+
"eval_runtime": 7.4845,
|
264 |
+
"eval_samples_per_second": 66.805,
|
265 |
+
"eval_steps_per_second": 8.417,
|
266 |
"step": 1687
|
267 |
},
|
268 |
{
|
269 |
+
"epoch": 9.0,
|
270 |
+
"eval_exact_match": 14.2,
|
271 |
+
"eval_f1": 19.66897045942666,
|
272 |
+
"eval_qa_bleu": 3.3530752528407004,
|
273 |
+
"eval_qa_exact_match": 0.112,
|
274 |
+
"eval_recite_bleu": 5.3293362575452425,
|
275 |
+
"eval_recite_exact_match": 0.002,
|
276 |
"step": 1687
|
277 |
},
|
278 |
{
|
279 |
+
"epoch": 9.07,
|
|
|
280 |
"learning_rate": 3e-05,
|
281 |
+
"loss": 0.8853,
|
282 |
"step": 1700
|
283 |
},
|
284 |
{
|
285 |
"epoch": 9.6,
|
|
|
286 |
"learning_rate": 3e-05,
|
287 |
+
"loss": 0.8373,
|
288 |
"step": 1800
|
289 |
},
|
290 |
{
|
291 |
"epoch": 10.0,
|
292 |
+
"eval_accuracy": 0.6587450980392157,
|
293 |
+
"eval_loss": 1.3337457180023193,
|
294 |
+
"eval_runtime": 6.6361,
|
295 |
+
"eval_samples_per_second": 75.345,
|
296 |
+
"eval_steps_per_second": 9.493,
|
297 |
"step": 1875
|
298 |
},
|
299 |
{
|
300 |
"epoch": 10.0,
|
301 |
+
"eval_exact_match": 14.4,
|
302 |
+
"eval_f1": 21.20886352105531,
|
303 |
+
"eval_qa_bleu": 2.50339041197013,
|
304 |
+
"eval_qa_exact_match": 0.106,
|
305 |
+
"eval_recite_bleu": 6.648771889251858,
|
306 |
+
"eval_recite_exact_match": 0.0,
|
307 |
"step": 1875
|
308 |
},
|
309 |
{
|
310 |
+
"epoch": 10.13,
|
|
|
311 |
"learning_rate": 3e-05,
|
312 |
+
"loss": 0.8353,
|
313 |
"step": 1900
|
314 |
},
|
315 |
{
|
316 |
+
"epoch": 10.67,
|
|
|
317 |
"learning_rate": 3e-05,
|
318 |
+
"loss": 0.7949,
|
319 |
"step": 2000
|
320 |
},
|
321 |
{
|
322 |
+
"epoch": 11.0,
|
323 |
+
"eval_accuracy": 0.6575058823529412,
|
324 |
+
"eval_loss": 1.367775797843933,
|
325 |
+
"eval_runtime": 7.4976,
|
326 |
+
"eval_samples_per_second": 66.688,
|
327 |
+
"eval_steps_per_second": 8.403,
|
328 |
"step": 2062
|
329 |
},
|
330 |
{
|
331 |
+
"epoch": 11.0,
|
332 |
+
"eval_exact_match": 15.4,
|
333 |
+
"eval_f1": 21.670913900913916,
|
334 |
+
"eval_qa_bleu": 4.131625289761948,
|
335 |
+
"eval_qa_exact_match": 0.104,
|
336 |
+
"eval_recite_bleu": 8.111027123751208,
|
337 |
+
"eval_recite_exact_match": 0.0,
|
338 |
"step": 2062
|
339 |
},
|
340 |
{
|
341 |
"epoch": 11.2,
|
|
|
342 |
"learning_rate": 3e-05,
|
343 |
+
"loss": 0.7758,
|
344 |
"step": 2100
|
345 |
},
|
346 |
{
|
347 |
+
"epoch": 11.73,
|
|
|
348 |
"learning_rate": 3e-05,
|
349 |
+
"loss": 0.7539,
|
350 |
"step": 2200
|
351 |
},
|
352 |
{
|
353 |
"epoch": 12.0,
|
354 |
+
"eval_accuracy": 0.6553725490196078,
|
355 |
+
"eval_loss": 1.3984336853027344,
|
356 |
+
"eval_runtime": 7.6892,
|
357 |
+
"eval_samples_per_second": 65.026,
|
358 |
+
"eval_steps_per_second": 8.193,
|
359 |
"step": 2250
|
360 |
},
|
361 |
{
|
362 |
"epoch": 12.0,
|
363 |
+
"eval_exact_match": 16.2,
|
364 |
+
"eval_f1": 23.074922854922868,
|
365 |
+
"eval_qa_bleu": 10.80394366395148,
|
366 |
+
"eval_qa_exact_match": 0.132,
|
367 |
+
"eval_recite_bleu": 9.158207459098465,
|
368 |
+
"eval_recite_exact_match": 0.002,
|
369 |
"step": 2250
|
370 |
},
|
371 |
{
|
372 |
+
"epoch": 12.27,
|
|
|
373 |
"learning_rate": 3e-05,
|
374 |
+
"loss": 0.7415,
|
375 |
"step": 2300
|
376 |
},
|
377 |
{
|
378 |
"epoch": 12.8,
|
|
|
379 |
"learning_rate": 3e-05,
|
380 |
+
"loss": 0.7133,
|
381 |
"step": 2400
|
382 |
},
|
383 |
{
|
384 |
+
"epoch": 13.0,
|
385 |
+
"eval_accuracy": 0.6538196078431373,
|
386 |
+
"eval_loss": 1.4471032619476318,
|
387 |
+
"eval_runtime": 6.7177,
|
388 |
+
"eval_samples_per_second": 74.431,
|
389 |
+
"eval_steps_per_second": 9.378,
|
390 |
"step": 2437
|
391 |
},
|
392 |
{
|
393 |
+
"epoch": 13.0,
|
394 |
+
"eval_exact_match": 16.4,
|
395 |
+
"eval_f1": 23.343965796845062,
|
396 |
+
"eval_qa_bleu": 4.473837470175788,
|
397 |
+
"eval_qa_exact_match": 0.136,
|
398 |
+
"eval_recite_bleu": 10.58390244263022,
|
399 |
+
"eval_recite_exact_match": 0.002,
|
400 |
"step": 2437
|
401 |
},
|
402 |
{
|
403 |
+
"epoch": 13.33,
|
|
|
404 |
"learning_rate": 3e-05,
|
405 |
+
"loss": 0.682,
|
406 |
"step": 2500
|
407 |
},
|
408 |
{
|
409 |
+
"epoch": 13.87,
|
|
|
410 |
"learning_rate": 3e-05,
|
411 |
+
"loss": 0.6704,
|
412 |
"step": 2600
|
413 |
},
|
414 |
{
|
415 |
"epoch": 14.0,
|
416 |
+
"eval_accuracy": 0.6524862745098039,
|
417 |
+
"eval_loss": 1.4829775094985962,
|
418 |
+
"eval_runtime": 6.6339,
|
419 |
+
"eval_samples_per_second": 75.37,
|
420 |
+
"eval_steps_per_second": 9.497,
|
421 |
"step": 2625
|
422 |
},
|
423 |
{
|
424 |
"epoch": 14.0,
|
425 |
+
"eval_exact_match": 15.6,
|
426 |
+
"eval_f1": 23.615572621393074,
|
427 |
+
"eval_qa_bleu": 7.563435997390864,
|
428 |
+
"eval_qa_exact_match": 0.13,
|
429 |
+
"eval_recite_bleu": 11.120038652898813,
|
430 |
+
"eval_recite_exact_match": 0.008,
|
431 |
"step": 2625
|
432 |
},
|
433 |
{
|
434 |
"epoch": 14.4,
|
|
|
435 |
"learning_rate": 3e-05,
|
436 |
+
"loss": 0.6412,
|
437 |
"step": 2700
|
438 |
},
|
439 |
{
|
440 |
+
"epoch": 14.93,
|
|
|
441 |
"learning_rate": 3e-05,
|
442 |
+
"loss": 0.6436,
|
443 |
"step": 2800
|
444 |
},
|
445 |
{
|
446 |
+
"epoch": 15.0,
|
447 |
+
"eval_accuracy": 0.6507529411764705,
|
448 |
+
"eval_loss": 1.524308681488037,
|
449 |
+
"eval_runtime": 6.6481,
|
450 |
+
"eval_samples_per_second": 75.209,
|
451 |
+
"eval_steps_per_second": 9.476,
|
452 |
"step": 2812
|
453 |
},
|
454 |
{
|
455 |
+
"epoch": 15.0,
|
456 |
+
"eval_exact_match": 17.8,
|
457 |
+
"eval_f1": 27.00021164021165,
|
458 |
+
"eval_qa_bleu": 7.90349748901623,
|
459 |
+
"eval_qa_exact_match": 0.15,
|
460 |
+
"eval_recite_bleu": 12.277483381668418,
|
461 |
+
"eval_recite_exact_match": 0.008,
|
462 |
"step": 2812
|
463 |
},
|
464 |
{
|
465 |
+
"epoch": 15.47,
|
|
|
466 |
"learning_rate": 3e-05,
|
467 |
+
"loss": 0.5969,
|
468 |
"step": 2900
|
469 |
},
|
470 |
{
|
471 |
"epoch": 16.0,
|
|
|
472 |
"learning_rate": 3e-05,
|
473 |
+
"loss": 0.6073,
|
474 |
"step": 3000
|
475 |
},
|
476 |
{
|
477 |
"epoch": 16.0,
|
478 |
+
"eval_accuracy": 0.6499764705882353,
|
479 |
+
"eval_loss": 1.5640995502471924,
|
480 |
+
"eval_runtime": 7.7119,
|
481 |
+
"eval_samples_per_second": 64.835,
|
482 |
+
"eval_steps_per_second": 8.169,
|
483 |
"step": 3000
|
484 |
},
|
485 |
{
|
486 |
"epoch": 16.0,
|
487 |
+
"eval_exact_match": 19.2,
|
488 |
+
"eval_f1": 27.747719298245627,
|
489 |
+
"eval_qa_bleu": 9.793788404463436,
|
490 |
+
"eval_qa_exact_match": 0.14,
|
491 |
+
"eval_recite_bleu": 12.835117221544726,
|
492 |
+
"eval_recite_exact_match": 0.01,
|
493 |
"step": 3000
|
494 |
},
|
495 |
{
|
496 |
+
"epoch": 16.53,
|
|
|
497 |
"learning_rate": 3e-05,
|
498 |
+
"loss": 0.5567,
|
499 |
"step": 3100
|
500 |
},
|
501 |
{
|
502 |
+
"epoch": 17.0,
|
503 |
+
"eval_accuracy": 0.6486588235294117,
|
504 |
+
"eval_loss": 1.6058622598648071,
|
505 |
+
"eval_runtime": 7.7727,
|
506 |
+
"eval_samples_per_second": 64.328,
|
507 |
+
"eval_steps_per_second": 8.105,
|
508 |
"step": 3187
|
509 |
},
|
510 |
{
|
511 |
+
"epoch": 17.0,
|
512 |
+
"eval_exact_match": 19.6,
|
513 |
+
"eval_f1": 26.446776949066113,
|
514 |
+
"eval_qa_bleu": 10.294098070146385,
|
515 |
+
"eval_qa_exact_match": 0.156,
|
516 |
+
"eval_recite_bleu": 12.34815708077889,
|
517 |
+
"eval_recite_exact_match": 0.004,
|
518 |
"step": 3187
|
519 |
},
|
520 |
{
|
521 |
+
"epoch": 17.07,
|
|
|
522 |
"learning_rate": 3e-05,
|
523 |
+
"loss": 0.5654,
|
524 |
"step": 3200
|
525 |
},
|
526 |
{
|
527 |
"epoch": 17.6,
|
|
|
528 |
"learning_rate": 3e-05,
|
529 |
+
"loss": 0.5204,
|
530 |
"step": 3300
|
531 |
},
|
532 |
{
|
533 |
"epoch": 18.0,
|
534 |
+
"eval_accuracy": 0.6472705882352942,
|
535 |
+
"eval_loss": 1.6425700187683105,
|
536 |
+
"eval_runtime": 7.1347,
|
537 |
+
"eval_samples_per_second": 70.08,
|
538 |
+
"eval_steps_per_second": 8.83,
|
539 |
"step": 3375
|
540 |
},
|
541 |
{
|
542 |
"epoch": 18.0,
|
543 |
+
"eval_exact_match": 18.4,
|
544 |
+
"eval_f1": 26.083665223665236,
|
545 |
+
"eval_qa_bleu": 12.978225559731756,
|
546 |
+
"eval_qa_exact_match": 0.142,
|
547 |
+
"eval_recite_bleu": 12.629003802964625,
|
548 |
+
"eval_recite_exact_match": 0.006,
|
549 |
"step": 3375
|
550 |
},
|
551 |
{
|
552 |
+
"epoch": 18.13,
|
|
|
553 |
"learning_rate": 3e-05,
|
554 |
+
"loss": 0.5348,
|
555 |
"step": 3400
|
556 |
},
|
557 |
{
|
558 |
+
"epoch": 18.67,
|
|
|
559 |
"learning_rate": 3e-05,
|
560 |
+
"loss": 0.5019,
|
561 |
"step": 3500
|
562 |
},
|
563 |
{
|
564 |
+
"epoch": 19.0,
|
565 |
+
"eval_accuracy": 0.6459843137254901,
|
566 |
+
"eval_loss": 1.6983567476272583,
|
567 |
+
"eval_runtime": 7.0372,
|
568 |
+
"eval_samples_per_second": 71.051,
|
569 |
+
"eval_steps_per_second": 8.952,
|
570 |
"step": 3562
|
571 |
},
|
572 |
{
|
573 |
+
"epoch": 19.0,
|
574 |
+
"eval_exact_match": 19.4,
|
575 |
+
"eval_f1": 27.658198757763987,
|
576 |
+
"eval_qa_bleu": 10.765170656210177,
|
577 |
+
"eval_qa_exact_match": 0.146,
|
578 |
+
"eval_recite_bleu": 12.874816956011239,
|
579 |
+
"eval_recite_exact_match": 0.006,
|
580 |
"step": 3562
|
581 |
},
|
582 |
{
|
583 |
"epoch": 19.2,
|
|
|
584 |
"learning_rate": 3e-05,
|
585 |
+
"loss": 0.4888,
|
586 |
"step": 3600
|
587 |
},
|
588 |
{
|
589 |
+
"epoch": 19.73,
|
|
|
590 |
"learning_rate": 3e-05,
|
591 |
+
"loss": 0.4676,
|
592 |
"step": 3700
|
593 |
},
|
594 |
{
|
595 |
"epoch": 20.0,
|
596 |
+
"eval_accuracy": 0.6450431372549019,
|
597 |
+
"eval_loss": 1.74748957157135,
|
598 |
+
"eval_runtime": 6.6386,
|
599 |
+
"eval_samples_per_second": 75.317,
|
600 |
+
"eval_steps_per_second": 9.49,
|
601 |
"step": 3750
|
602 |
},
|
603 |
{
|
604 |
"epoch": 20.0,
|
605 |
+
"eval_exact_match": 17.4,
|
606 |
+
"eval_f1": 25.5869051909052,
|
607 |
+
"eval_qa_bleu": 7.975401244418332,
|
608 |
+
"eval_qa_exact_match": 0.144,
|
609 |
+
"eval_recite_bleu": 11.978547399136012,
|
610 |
+
"eval_recite_exact_match": 0.004,
|
611 |
"step": 3750
|
612 |
},
|
613 |
{
|
614 |
+
"epoch": 20.27,
|
|
|
615 |
"learning_rate": 3e-05,
|
616 |
+
"loss": 0.4528,
|
617 |
"step": 3800
|
618 |
},
|
619 |
{
|
620 |
"epoch": 20.8,
|
|
|
621 |
"learning_rate": 3e-05,
|
622 |
+
"loss": 0.4408,
|
623 |
"step": 3900
|
624 |
},
|
625 |
{
|
626 |
+
"epoch": 21.0,
|
627 |
+
"eval_accuracy": 0.6450274509803922,
|
628 |
+
"eval_loss": 1.7759969234466553,
|
629 |
+
"eval_runtime": 7.7092,
|
630 |
+
"eval_samples_per_second": 64.857,
|
631 |
+
"eval_steps_per_second": 8.172,
|
632 |
"step": 3937
|
633 |
},
|
634 |
{
|
635 |
+
"epoch": 21.0,
|
636 |
+
"eval_exact_match": 19.0,
|
637 |
+
"eval_f1": 27.044218949435447,
|
638 |
+
"eval_qa_bleu": 5.961857744132503,
|
639 |
+
"eval_qa_exact_match": 0.158,
|
640 |
+
"eval_recite_bleu": 12.078600703350125,
|
641 |
+
"eval_recite_exact_match": 0.004,
|
642 |
"step": 3937
|
643 |
},
|
644 |
{
|
645 |
+
"epoch": 21.33,
|
|
|
646 |
"learning_rate": 3e-05,
|
647 |
+
"loss": 0.4191,
|
648 |
"step": 4000
|
649 |
},
|
650 |
{
|
651 |
+
"epoch": 21.87,
|
|
|
652 |
"learning_rate": 3e-05,
|
653 |
+
"loss": 0.4217,
|
654 |
"step": 4100
|
655 |
},
|
656 |
{
|
657 |
"epoch": 22.0,
|
658 |
+
"eval_accuracy": 0.644078431372549,
|
659 |
+
"eval_loss": 1.7986315488815308,
|
660 |
+
"eval_runtime": 7.6507,
|
661 |
+
"eval_samples_per_second": 65.353,
|
662 |
+
"eval_steps_per_second": 8.235,
|
663 |
"step": 4125
|
664 |
},
|
665 |
{
|
666 |
"epoch": 22.0,
|
667 |
+
"eval_exact_match": 18.2,
|
668 |
+
"eval_f1": 25.124761904761915,
|
669 |
+
"eval_qa_bleu": 15.392825669562168,
|
670 |
+
"eval_qa_exact_match": 0.14,
|
671 |
+
"eval_recite_bleu": 11.50341909728548,
|
672 |
+
"eval_recite_exact_match": 0.006,
|
673 |
"step": 4125
|
674 |
},
|
675 |
{
|
676 |
"epoch": 22.4,
|
|
|
677 |
"learning_rate": 3e-05,
|
678 |
+
"loss": 0.393,
|
679 |
"step": 4200
|
680 |
},
|
681 |
{
|
682 |
+
"epoch": 22.93,
|
|
|
683 |
"learning_rate": 3e-05,
|
684 |
+
"loss": 0.3886,
|
685 |
"step": 4300
|
686 |
},
|
687 |
{
|
688 |
+
"epoch": 23.0,
|
689 |
+
"eval_accuracy": 0.6439529411764706,
|
690 |
+
"eval_loss": 1.8378911018371582,
|
691 |
+
"eval_runtime": 7.4779,
|
692 |
+
"eval_samples_per_second": 66.864,
|
693 |
+
"eval_steps_per_second": 8.425,
|
694 |
"step": 4312
|
695 |
},
|
696 |
{
|
697 |
+
"epoch": 23.0,
|
698 |
+
"eval_exact_match": 14.8,
|
699 |
+
"eval_f1": 21.914444444444456,
|
700 |
+
"eval_qa_bleu": 10.979013053975578,
|
701 |
+
"eval_qa_exact_match": 0.128,
|
702 |
+
"eval_recite_bleu": 11.265801595707984,
|
703 |
+
"eval_recite_exact_match": 0.004,
|
704 |
"step": 4312
|
705 |
},
|
706 |
{
|
707 |
+
"epoch": 23.47,
|
|
|
708 |
"learning_rate": 3e-05,
|
709 |
+
"loss": 0.3605,
|
710 |
"step": 4400
|
711 |
},
|
712 |
{
|
713 |
"epoch": 24.0,
|
|
|
714 |
"learning_rate": 3e-05,
|
715 |
+
"loss": 0.3729,
|
716 |
"step": 4500
|
717 |
},
|
718 |
{
|
719 |
"epoch": 24.0,
|
720 |
+
"eval_accuracy": 0.6420470588235294,
|
721 |
+
"eval_loss": 1.927443504333496,
|
722 |
+
"eval_runtime": 7.4721,
|
723 |
+
"eval_samples_per_second": 66.915,
|
724 |
+
"eval_steps_per_second": 8.431,
|
725 |
"step": 4500
|
726 |
},
|
727 |
{
|
728 |
"epoch": 24.0,
|
729 |
+
"eval_exact_match": 15.6,
|
730 |
+
"eval_f1": 22.679302325581407,
|
731 |
+
"eval_qa_bleu": 7.33716058981409,
|
732 |
+
"eval_qa_exact_match": 0.132,
|
733 |
+
"eval_recite_bleu": 11.564503717897392,
|
734 |
+
"eval_recite_exact_match": 0.002,
|
735 |
"step": 4500
|
736 |
},
|
737 |
{
|
738 |
+
"epoch": 24.53,
|
|
|
739 |
"learning_rate": 3e-05,
|
740 |
+
"loss": 0.3316,
|
741 |
"step": 4600
|
742 |
},
|
743 |
{
|
744 |
+
"epoch": 25.0,
|
745 |
+
"eval_accuracy": 0.641921568627451,
|
746 |
+
"eval_loss": 1.9378185272216797,
|
747 |
+
"eval_runtime": 7.6806,
|
748 |
+
"eval_samples_per_second": 65.099,
|
749 |
+
"eval_steps_per_second": 8.203,
|
750 |
"step": 4687
|
751 |
},
|
752 |
{
|
753 |
+
"epoch": 25.0,
|
754 |
+
"eval_exact_match": 16.8,
|
755 |
+
"eval_f1": 23.502857142857145,
|
756 |
+
"eval_qa_bleu": 8.430315553769542,
|
757 |
+
"eval_qa_exact_match": 0.13,
|
758 |
+
"eval_recite_bleu": 11.821782294825612,
|
759 |
+
"eval_recite_exact_match": 0.006,
|
760 |
"step": 4687
|
761 |
},
|
762 |
{
|
763 |
+
"epoch": 25.07,
|
|
|
764 |
"learning_rate": 3e-05,
|
765 |
+
"loss": 0.3454,
|
766 |
"step": 4700
|
767 |
},
|
768 |
{
|
769 |
"epoch": 25.6,
|
|
|
770 |
"learning_rate": 3e-05,
|
771 |
+
"loss": 0.313,
|
772 |
"step": 4800
|
773 |
},
|
774 |
{
|
775 |
"epoch": 26.0,
|
776 |
+
"eval_accuracy": 0.6410274509803922,
|
777 |
+
"eval_loss": 2.0155365467071533,
|
778 |
+
"eval_runtime": 6.6459,
|
779 |
+
"eval_samples_per_second": 75.235,
|
780 |
+
"eval_steps_per_second": 9.48,
|
781 |
"step": 4875
|
782 |
},
|
783 |
{
|
784 |
"epoch": 26.0,
|
785 |
+
"eval_exact_match": 15.4,
|
786 |
+
"eval_f1": 23.056190476190483,
|
787 |
+
"eval_qa_bleu": 10.828743073480153,
|
788 |
+
"eval_qa_exact_match": 0.132,
|
789 |
+
"eval_recite_bleu": 11.70611308380729,
|
790 |
+
"eval_recite_exact_match": 0.004,
|
791 |
"step": 4875
|
792 |
},
|
793 |
{
|
794 |
+
"epoch": 26.13,
|
|
|
795 |
"learning_rate": 3e-05,
|
796 |
+
"loss": 0.316,
|
797 |
"step": 4900
|
798 |
},
|
799 |
{
|
800 |
+
"epoch": 26.67,
|
|
|
801 |
"learning_rate": 3e-05,
|
802 |
+
"loss": 0.2891,
|
803 |
"step": 5000
|
804 |
},
|
805 |
{
|
806 |
+
"epoch": 27.0,
|
807 |
+
"eval_accuracy": 0.6408313725490196,
|
808 |
+
"eval_loss": 2.0192761421203613,
|
809 |
+
"eval_runtime": 7.4551,
|
810 |
+
"eval_samples_per_second": 67.068,
|
811 |
+
"eval_steps_per_second": 8.451,
|
812 |
"step": 5062
|
813 |
},
|
814 |
{
|
815 |
+
"epoch": 27.0,
|
816 |
+
"eval_exact_match": 14.0,
|
817 |
+
"eval_f1": 22.766833842561553,
|
818 |
+
"eval_qa_bleu": 3.5222468357048804,
|
819 |
+
"eval_qa_exact_match": 0.108,
|
820 |
+
"eval_recite_bleu": 10.474195118057567,
|
821 |
+
"eval_recite_exact_match": 0.002,
|
822 |
"step": 5062
|
823 |
},
|
824 |
{
|
825 |
"epoch": 27.2,
|
|
|
826 |
"learning_rate": 3e-05,
|
827 |
+
"loss": 0.299,
|
828 |
"step": 5100
|
829 |
},
|
830 |
{
|
831 |
+
"epoch": 27.73,
|
|
|
832 |
"learning_rate": 3e-05,
|
833 |
+
"loss": 0.281,
|
834 |
"step": 5200
|
835 |
},
|
836 |
{
|
837 |
"epoch": 28.0,
|
838 |
+
"eval_accuracy": 0.640243137254902,
|
839 |
+
"eval_loss": 2.074819803237915,
|
840 |
+
"eval_runtime": 7.0598,
|
841 |
+
"eval_samples_per_second": 70.823,
|
842 |
+
"eval_steps_per_second": 8.924,
|
843 |
"step": 5250
|
844 |
},
|
845 |
{
|
846 |
"epoch": 28.0,
|
847 |
+
"eval_exact_match": 14.8,
|
848 |
+
"eval_f1": 22.03527508090615,
|
849 |
+
"eval_qa_bleu": 4.8943988922035935,
|
850 |
+
"eval_qa_exact_match": 0.118,
|
851 |
+
"eval_recite_bleu": 11.435320168344504,
|
852 |
+
"eval_recite_exact_match": 0.008,
|
853 |
"step": 5250
|
854 |
},
|
855 |
{
|
856 |
+
"epoch": 28.27,
|
|
|
857 |
"learning_rate": 3e-05,
|
858 |
+
"loss": 0.2637,
|
859 |
"step": 5300
|
860 |
},
|
861 |
{
|
862 |
"epoch": 28.8,
|
|
|
863 |
"learning_rate": 3e-05,
|
864 |
+
"loss": 0.2639,
|
865 |
"step": 5400
|
866 |
},
|
867 |
{
|
868 |
+
"epoch": 29.0,
|
869 |
+
"eval_accuracy": 0.6389333333333334,
|
870 |
+
"eval_loss": 2.1383912563323975,
|
871 |
+
"eval_runtime": 7.7452,
|
872 |
+
"eval_samples_per_second": 64.556,
|
873 |
+
"eval_steps_per_second": 8.134,
|
874 |
"step": 5437
|
875 |
},
|
876 |
{
|
877 |
+
"epoch": 29.0,
|
878 |
+
"eval_exact_match": 16.2,
|
879 |
+
"eval_f1": 24.30754076973256,
|
880 |
+
"eval_qa_bleu": 4.486401752816634,
|
881 |
+
"eval_qa_exact_match": 0.128,
|
882 |
+
"eval_recite_bleu": 11.730902648782326,
|
883 |
+
"eval_recite_exact_match": 0.004,
|
884 |
"step": 5437
|
885 |
},
|
886 |
{
|
887 |
+
"epoch": 29.33,
|
|
|
888 |
"learning_rate": 3e-05,
|
889 |
+
"loss": 0.2473,
|
890 |
"step": 5500
|
891 |
},
|
892 |
{
|
893 |
+
"epoch": 29.87,
|
|
|
894 |
"learning_rate": 3e-05,
|
895 |
+
"loss": 0.249,
|
896 |
"step": 5600
|
897 |
},
|
898 |
{
|
899 |
"epoch": 30.0,
|
900 |
+
"eval_accuracy": 0.6398509803921568,
|
901 |
+
"eval_loss": 2.125810384750366,
|
902 |
+
"eval_runtime": 7.6954,
|
903 |
+
"eval_samples_per_second": 64.974,
|
904 |
+
"eval_steps_per_second": 8.187,
|
905 |
"step": 5625
|
906 |
},
|
907 |
{
|
908 |
"epoch": 30.0,
|
909 |
+
"eval_exact_match": 15.8,
|
910 |
+
"eval_f1": 22.123535353535363,
|
911 |
+
"eval_qa_bleu": 3.2440134844965627,
|
912 |
+
"eval_qa_exact_match": 0.126,
|
913 |
+
"eval_recite_bleu": 11.76325197849696,
|
914 |
+
"eval_recite_exact_match": 0.004,
|
915 |
"step": 5625
|
916 |
},
|
917 |
{
|
918 |
"epoch": 30.4,
|
|
|
919 |
"learning_rate": 3e-05,
|
920 |
+
"loss": 0.2263,
|
921 |
"step": 5700
|
922 |
},
|
923 |
{
|
924 |
+
"epoch": 30.93,
|
|
|
925 |
"learning_rate": 3e-05,
|
926 |
+
"loss": 0.236,
|
927 |
"step": 5800
|
928 |
},
|
929 |
{
|
930 |
+
"epoch": 31.0,
|
931 |
+
"eval_accuracy": 0.6378666666666667,
|
932 |
+
"eval_loss": 2.2261879444122314,
|
933 |
+
"eval_runtime": 7.6462,
|
934 |
+
"eval_samples_per_second": 65.392,
|
935 |
+
"eval_steps_per_second": 8.239,
|
936 |
"step": 5812
|
937 |
},
|
938 |
{
|
939 |
+
"epoch": 31.0,
|
940 |
+
"eval_exact_match": 17.6,
|
941 |
+
"eval_f1": 25.120476190476193,
|
942 |
+
"eval_qa_bleu": 5.240120013213656,
|
943 |
+
"eval_qa_exact_match": 0.13,
|
944 |
+
"eval_recite_bleu": 11.888730171325669,
|
945 |
+
"eval_recite_exact_match": 0.008,
|
946 |
"step": 5812
|
947 |
},
|
948 |
{
|
949 |
+
"epoch": 31.47,
|
|
|
950 |
"learning_rate": 3e-05,
|
951 |
+
"loss": 0.2062,
|
952 |
"step": 5900
|
953 |
},
|
954 |
{
|
955 |
"epoch": 32.0,
|
|
|
956 |
"learning_rate": 3e-05,
|
957 |
+
"loss": 0.2235,
|
958 |
"step": 6000
|
959 |
},
|
960 |
{
|
961 |
"epoch": 32.0,
|
962 |
+
"eval_accuracy": 0.6381254901960784,
|
963 |
+
"eval_loss": 2.2489171028137207,
|
964 |
+
"eval_runtime": 7.0486,
|
965 |
+
"eval_samples_per_second": 70.936,
|
966 |
+
"eval_steps_per_second": 8.938,
|
967 |
"step": 6000
|
968 |
},
|
969 |
{
|
970 |
"epoch": 32.0,
|
971 |
+
"eval_exact_match": 16.6,
|
972 |
+
"eval_f1": 25.16174603174605,
|
973 |
+
"eval_qa_bleu": 12.336439673952102,
|
974 |
+
"eval_qa_exact_match": 0.134,
|
975 |
+
"eval_recite_bleu": 11.520322065980684,
|
976 |
"eval_recite_exact_match": 0.008,
|
977 |
"step": 6000
|
978 |
},
|
979 |
{
|
980 |
+
"epoch": 32.53,
|
|
|
981 |
"learning_rate": 3e-05,
|
982 |
+
"loss": 0.1952,
|
983 |
"step": 6100
|
984 |
},
|
985 |
{
|
986 |
+
"epoch": 33.0,
|
987 |
+
"eval_accuracy": 0.6373098039215687,
|
988 |
+
"eval_loss": 2.2877373695373535,
|
989 |
+
"eval_runtime": 7.7488,
|
990 |
+
"eval_samples_per_second": 64.526,
|
991 |
+
"eval_steps_per_second": 8.13,
|
992 |
"step": 6187
|
993 |
},
|
994 |
{
|
995 |
+
"epoch": 33.0,
|
996 |
+
"eval_exact_match": 15.2,
|
997 |
+
"eval_f1": 23.447550852439747,
|
998 |
+
"eval_qa_bleu": 3.977016577169375,
|
999 |
+
"eval_qa_exact_match": 0.126,
|
1000 |
+
"eval_recite_bleu": 13.076208653752236,
|
1001 |
+
"eval_recite_exact_match": 0.016,
|
1002 |
"step": 6187
|
1003 |
},
|
1004 |
{
|
1005 |
+
"epoch": 33.07,
|
|
|
1006 |
"learning_rate": 3e-05,
|
1007 |
+
"loss": 0.2023,
|
1008 |
"step": 6200
|
1009 |
},
|
1010 |
{
|
1011 |
"epoch": 33.6,
|
|
|
1012 |
"learning_rate": 3e-05,
|
1013 |
+
"loss": 0.1857,
|
1014 |
"step": 6300
|
1015 |
},
|
1016 |
{
|
1017 |
"epoch": 34.0,
|
1018 |
+
"eval_accuracy": 0.6367921568627452,
|
1019 |
+
"eval_loss": 2.309786319732666,
|
1020 |
+
"eval_runtime": 6.6393,
|
1021 |
+
"eval_samples_per_second": 75.309,
|
1022 |
+
"eval_steps_per_second": 9.489,
|
1023 |
"step": 6375
|
1024 |
},
|
1025 |
{
|
1026 |
"epoch": 34.0,
|
1027 |
+
"eval_exact_match": 14.6,
|
1028 |
+
"eval_f1": 21.406193118687113,
|
1029 |
+
"eval_qa_bleu": 3.892800891482867,
|
1030 |
+
"eval_qa_exact_match": 0.116,
|
1031 |
+
"eval_recite_bleu": 11.365741641301113,
|
1032 |
+
"eval_recite_exact_match": 0.008,
|
1033 |
"step": 6375
|
1034 |
},
|
1035 |
{
|
1036 |
+
"epoch": 34.13,
|
|
|
1037 |
"learning_rate": 3e-05,
|
1038 |
+
"loss": 0.1888,
|
1039 |
"step": 6400
|
1040 |
},
|
1041 |
{
|
1042 |
+
"epoch": 34.67,
|
|
|
1043 |
"learning_rate": 3e-05,
|
1044 |
+
"loss": 0.1739,
|
1045 |
"step": 6500
|
1046 |
},
|
1047 |
{
|
1048 |
+
"epoch": 35.0,
|
1049 |
+
"eval_accuracy": 0.6370196078431373,
|
1050 |
+
"eval_loss": 2.356412172317505,
|
1051 |
+
"eval_runtime": 6.7022,
|
1052 |
+
"eval_samples_per_second": 74.602,
|
1053 |
+
"eval_steps_per_second": 9.4,
|
1054 |
"step": 6562
|
1055 |
},
|
1056 |
{
|
1057 |
+
"epoch": 35.0,
|
1058 |
+
"eval_exact_match": 16.0,
|
1059 |
+
"eval_f1": 24.90098936662045,
|
1060 |
+
"eval_qa_bleu": 6.669415710869239,
|
1061 |
+
"eval_qa_exact_match": 0.13,
|
1062 |
+
"eval_recite_bleu": 12.340892651236608,
|
1063 |
+
"eval_recite_exact_match": 0.01,
|
1064 |
"step": 6562
|
1065 |
},
|
1066 |
{
|
1067 |
"epoch": 35.2,
|
|
|
1068 |
"learning_rate": 3e-05,
|
1069 |
+
"loss": 0.1771,
|
1070 |
"step": 6600
|
1071 |
},
|
1072 |
{
|
1073 |
+
"epoch": 35.73,
|
|
|
1074 |
"learning_rate": 3e-05,
|
1075 |
+
"loss": 0.1675,
|
1076 |
"step": 6700
|
1077 |
},
|
1078 |
{
|
1079 |
"epoch": 36.0,
|
1080 |
+
"eval_accuracy": 0.6365647058823529,
|
1081 |
+
"eval_loss": 2.413015127182007,
|
1082 |
+
"eval_runtime": 7.498,
|
1083 |
+
"eval_samples_per_second": 66.684,
|
1084 |
+
"eval_steps_per_second": 8.402,
|
1085 |
"step": 6750
|
1086 |
},
|
1087 |
{
|
1088 |
"epoch": 36.0,
|
1089 |
+
"eval_exact_match": 16.6,
|
1090 |
+
"eval_f1": 24.060404040404055,
|
1091 |
+
"eval_qa_bleu": 9.329376121653498,
|
1092 |
+
"eval_qa_exact_match": 0.134,
|
1093 |
+
"eval_recite_bleu": 12.441487241994658,
|
1094 |
+
"eval_recite_exact_match": 0.012,
|
1095 |
"step": 6750
|
1096 |
},
|
1097 |
{
|
1098 |
+
"epoch": 36.27,
|
|
|
1099 |
"learning_rate": 3e-05,
|
1100 |
+
"loss": 0.1623,
|
1101 |
"step": 6800
|
1102 |
},
|
1103 |
{
|
1104 |
"epoch": 36.8,
|
|
|
1105 |
"learning_rate": 3e-05,
|
1106 |
+
"loss": 0.158,
|
1107 |
"step": 6900
|
1108 |
},
|
1109 |
{
|
1110 |
+
"epoch": 37.0,
|
1111 |
+
"eval_accuracy": 0.6361333333333333,
|
1112 |
+
"eval_loss": 2.4566919803619385,
|
1113 |
+
"eval_runtime": 7.1162,
|
1114 |
+
"eval_samples_per_second": 70.263,
|
1115 |
+
"eval_steps_per_second": 8.853,
|
1116 |
"step": 6937
|
1117 |
},
|
1118 |
{
|
1119 |
+
"epoch": 37.0,
|
1120 |
+
"eval_exact_match": 16.0,
|
1121 |
+
"eval_f1": 22.78879744333159,
|
1122 |
+
"eval_qa_bleu": 4.048618436225478,
|
1123 |
+
"eval_qa_exact_match": 0.128,
|
1124 |
+
"eval_recite_bleu": 12.710989230469986,
|
1125 |
+
"eval_recite_exact_match": 0.014,
|
1126 |
"step": 6937
|
1127 |
},
|
1128 |
{
|
1129 |
+
"epoch": 37.33,
|
|
|
1130 |
"learning_rate": 3e-05,
|
1131 |
+
"loss": 0.1525,
|
1132 |
"step": 7000
|
1133 |
},
|
1134 |
{
|
1135 |
+
"epoch": 37.87,
|
|
|
1136 |
"learning_rate": 3e-05,
|
1137 |
+
"loss": 0.1509,
|
1138 |
"step": 7100
|
1139 |
},
|
1140 |
{
|
1141 |
"epoch": 38.0,
|
1142 |
+
"eval_accuracy": 0.6366117647058823,
|
1143 |
+
"eval_loss": 2.466296672821045,
|
1144 |
+
"eval_runtime": 6.7087,
|
1145 |
+
"eval_samples_per_second": 74.53,
|
1146 |
+
"eval_steps_per_second": 9.391,
|
1147 |
"step": 7125
|
1148 |
},
|
1149 |
{
|
1150 |
"epoch": 38.0,
|
1151 |
+
"eval_exact_match": 17.8,
|
1152 |
+
"eval_f1": 24.827619047619052,
|
1153 |
+
"eval_qa_bleu": 6.898280882518603,
|
1154 |
+
"eval_qa_exact_match": 0.142,
|
1155 |
+
"eval_recite_bleu": 13.041079892055818,
|
1156 |
+
"eval_recite_exact_match": 0.012,
|
1157 |
"step": 7125
|
1158 |
}
|
1159 |
],
|
1160 |
"logging_steps": 100,
|
1161 |
"max_steps": 9350,
|
|
|
1162 |
"num_train_epochs": 50,
|
1163 |
"save_steps": 500,
|
1164 |
+
"total_flos": 1.651433148321366e+18,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1165 |
"trial_name": null,
|
1166 |
"trial_params": null
|
1167 |
}
|
checkpoint-7125/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 sha256:9347580f7d2a7b3957c29c0f67ee2342d50aa99cfc14c2f59eba6127909f6932
|
3 |
+
size 4728
|