Training in progress, epoch 28, checkpoint
Browse files- checkpoint-5250/adapter_config.json +1 -1
- checkpoint-5250/adapter_model.bin +3 -0
- checkpoint-5250/added_tokens.json +3 -3
- checkpoint-5250/optimizer.pt +2 -2
- checkpoint-5250/rng_state_0.pth +1 -1
- checkpoint-5250/rng_state_1.pth +1 -1
- checkpoint-5250/rng_state_2.pth +1 -1
- checkpoint-5250/rng_state_3.pth +1 -1
- checkpoint-5250/special_tokens_map.json +4 -12
- checkpoint-5250/tokenizer.json +0 -0
- checkpoint-5250/tokenizer.model +3 -0
- checkpoint-5250/tokenizer_config.json +17 -20
- checkpoint-5250/trainer_state.json +420 -486
- checkpoint-5250/training_args.bin +2 -2
checkpoint-5250/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-5250/adapter_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0b4c872be53c6470c1d06d19adac3727705a2b91b3d128b2821720eb31a89ee0
|
3 |
+
size 143269386
|
checkpoint-5250/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-5250/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 sha256:09166786c9d2997511be2bab9c6d9cb4fb4bdba0f2987ad19097542f97484e5f
|
3 |
+
size 286585234
|
checkpoint-5250/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 sha256:8907ea18d4a16d2e8f7d93c4efae2ec675f201e532c3650f99cc2cec3634b6f4
|
3 |
size 15024
|
checkpoint-5250/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 sha256:8ce1f1c51b9c3e9e4db4c4c1078ffe28bbb8b83e4526da1f0fcb701abb05e911
|
3 |
size 15024
|
checkpoint-5250/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 sha256:86d026856e69926931d262b81b9fdccb1a841ba70b12fbe6476412f9d1039991
|
3 |
size 15024
|
checkpoint-5250/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 sha256:3c3672e3885980cf88f8a7ceb40f1e7680bc5c78e0281e08ad6efce1f5aaabd7
|
3 |
size 15024
|
checkpoint-5250/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-5250/tokenizer.json
CHANGED
The diff for this file is too large to render.
See raw diff
|
|
checkpoint-5250/tokenizer.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
3 |
+
size 499723
|
checkpoint-5250/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-5250/trainer_state.json
CHANGED
@@ -9,921 +9,855 @@
|
|
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 |
-
"epoch": 7.
|
237 |
-
"grad_norm": 1.1534159183502197,
|
238 |
"learning_rate": 5e-05,
|
239 |
-
"loss":
|
240 |
"step": 1400
|
241 |
},
|
242 |
{
|
243 |
"epoch": 8.0,
|
244 |
-
"grad_norm": 1.1209707260131836,
|
245 |
"learning_rate": 5e-05,
|
246 |
-
"loss":
|
247 |
"step": 1500
|
248 |
},
|
249 |
{
|
250 |
"epoch": 8.0,
|
251 |
-
"eval_accuracy": 0.
|
252 |
-
"eval_loss": 1.
|
253 |
-
"eval_runtime": 6.
|
254 |
-
"eval_samples_per_second":
|
255 |
-
"eval_steps_per_second": 9.
|
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": 1.3646217584609985,
|
271 |
"learning_rate": 5e-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": 1.6948473453521729,
|
297 |
"learning_rate": 5e-05,
|
298 |
-
"loss":
|
299 |
"step": 1700
|
300 |
},
|
301 |
{
|
302 |
"epoch": 9.6,
|
303 |
-
"grad_norm": 1.6024364233016968,
|
304 |
"learning_rate": 5e-05,
|
305 |
-
"loss":
|
306 |
"step": 1800
|
307 |
},
|
308 |
{
|
309 |
"epoch": 10.0,
|
310 |
-
"eval_accuracy": 0.
|
311 |
-
"eval_loss": 1.
|
312 |
-
"eval_runtime":
|
313 |
-
"eval_samples_per_second":
|
314 |
-
"eval_steps_per_second":
|
315 |
"step": 1875
|
316 |
},
|
317 |
{
|
318 |
"epoch": 10.0,
|
319 |
-
"eval_exact_match":
|
320 |
-
"eval_f1":
|
321 |
-
"eval_qa_bleu":
|
322 |
-
"eval_qa_exact_match": 0.
|
323 |
-
"eval_recite_bleu":
|
324 |
"eval_recite_exact_match": 0.016,
|
325 |
"step": 1875
|
326 |
},
|
327 |
{
|
328 |
-
"epoch": 10.
|
329 |
-
"grad_norm": 1.6970032453536987,
|
330 |
"learning_rate": 5e-05,
|
331 |
-
"loss": 0.
|
332 |
"step": 1900
|
333 |
},
|
334 |
{
|
335 |
-
"epoch": 10.
|
336 |
-
"grad_norm": 1.559332251548767,
|
337 |
"learning_rate": 5e-05,
|
338 |
-
"loss": 0.
|
339 |
"step": 2000
|
340 |
},
|
341 |
{
|
342 |
-
"epoch":
|
343 |
-
"eval_accuracy": 0.
|
344 |
-
"eval_loss":
|
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": 1.667855143547058,
|
363 |
"learning_rate": 5e-05,
|
364 |
-
"loss": 0.
|
365 |
"step": 2100
|
366 |
},
|
367 |
{
|
368 |
-
"epoch": 11.
|
369 |
-
"grad_norm": 1.6040037870407104,
|
370 |
"learning_rate": 5e-05,
|
371 |
-
"loss": 0.
|
372 |
"step": 2200
|
373 |
},
|
374 |
{
|
375 |
"epoch": 12.0,
|
376 |
-
"eval_accuracy": 0.
|
377 |
-
"eval_loss":
|
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": 1.794092059135437,
|
396 |
"learning_rate": 5e-05,
|
397 |
-
"loss": 0.
|
398 |
"step": 2300
|
399 |
},
|
400 |
{
|
401 |
"epoch": 12.8,
|
402 |
-
"grad_norm": 1.8774526119232178,
|
403 |
"learning_rate": 5e-05,
|
404 |
-
"loss": 0.
|
405 |
"step": 2400
|
406 |
},
|
407 |
{
|
408 |
-
"epoch":
|
409 |
-
"eval_accuracy": 0.
|
410 |
-
"eval_loss":
|
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": 13.
|
423 |
-
"eval_recite_exact_match": 0.
|
424 |
"step": 2437
|
425 |
},
|
426 |
{
|
427 |
-
"epoch": 13.
|
428 |
-
"grad_norm": 1.9102168083190918,
|
429 |
"learning_rate": 5e-05,
|
430 |
-
"loss": 0.
|
431 |
"step": 2500
|
432 |
},
|
433 |
{
|
434 |
-
"epoch": 13.
|
435 |
-
"grad_norm": 1.9183365106582642,
|
436 |
"learning_rate": 5e-05,
|
437 |
-
"loss": 0.
|
438 |
"step": 2600
|
439 |
},
|
440 |
{
|
441 |
"epoch": 14.0,
|
442 |
-
"eval_accuracy": 0.
|
443 |
-
"eval_loss":
|
444 |
-
"eval_runtime":
|
445 |
-
"eval_samples_per_second":
|
446 |
-
"eval_steps_per_second":
|
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": 2.326387405395508,
|
462 |
"learning_rate": 5e-05,
|
463 |
-
"loss": 0.
|
464 |
"step": 2700
|
465 |
},
|
466 |
{
|
467 |
-
"epoch": 14.
|
468 |
-
"grad_norm": 2.1555840969085693,
|
469 |
"learning_rate": 5e-05,
|
470 |
-
"loss": 0.
|
471 |
"step": 2800
|
472 |
},
|
473 |
{
|
474 |
-
"epoch":
|
475 |
-
"eval_accuracy": 0.
|
476 |
-
"eval_loss":
|
477 |
-
"eval_runtime":
|
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":
|
489 |
-
"eval_recite_exact_match": 0.
|
490 |
"step": 2812
|
491 |
},
|
492 |
{
|
493 |
-
"epoch": 15.
|
494 |
-
"grad_norm": 1.8364830017089844,
|
495 |
"learning_rate": 5e-05,
|
496 |
-
"loss": 0.
|
497 |
"step": 2900
|
498 |
},
|
499 |
{
|
500 |
"epoch": 16.0,
|
501 |
-
"grad_norm": 1.9223560094833374,
|
502 |
"learning_rate": 5e-05,
|
503 |
-
"loss": 0.
|
504 |
"step": 3000
|
505 |
},
|
506 |
{
|
507 |
"epoch": 16.0,
|
508 |
-
"eval_accuracy": 0.
|
509 |
-
"eval_loss":
|
510 |
-
"eval_runtime": 6.
|
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":
|
520 |
-
"eval_qa_exact_match": 0.
|
521 |
-
"eval_recite_bleu":
|
522 |
-
"eval_recite_exact_match": 0.
|
523 |
"step": 3000
|
524 |
},
|
525 |
{
|
526 |
-
"epoch": 16.
|
527 |
-
"grad_norm": 2.403259515762329,
|
528 |
"learning_rate": 5e-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":
|
548 |
"eval_recite_exact_match": 0.016,
|
549 |
"step": 3187
|
550 |
},
|
551 |
{
|
552 |
-
"epoch": 17.
|
553 |
-
"grad_norm": 2.1279492378234863,
|
554 |
"learning_rate": 5e-05,
|
555 |
-
"loss": 0.
|
556 |
"step": 3200
|
557 |
},
|
558 |
{
|
559 |
"epoch": 17.6,
|
560 |
-
"grad_norm": 2.377486228942871,
|
561 |
"learning_rate": 5e-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":
|
581 |
-
"eval_recite_exact_match": 0.
|
582 |
"step": 3375
|
583 |
},
|
584 |
{
|
585 |
-
"epoch": 18.
|
586 |
-
"grad_norm": 2.425731897354126,
|
587 |
"learning_rate": 5e-05,
|
588 |
-
"loss": 0.
|
589 |
"step": 3400
|
590 |
},
|
591 |
{
|
592 |
-
"epoch": 18.
|
593 |
-
"grad_norm": 2.4585390090942383,
|
594 |
"learning_rate": 5e-05,
|
595 |
-
"loss": 0.
|
596 |
"step": 3500
|
597 |
},
|
598 |
{
|
599 |
-
"epoch":
|
600 |
-
"eval_accuracy": 0.
|
601 |
-
"eval_loss": 2.
|
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":
|
614 |
-
"eval_recite_exact_match": 0.
|
615 |
"step": 3562
|
616 |
},
|
617 |
{
|
618 |
"epoch": 19.2,
|
619 |
-
"grad_norm": 2.00604510307312,
|
620 |
"learning_rate": 5e-05,
|
621 |
-
"loss": 0.
|
622 |
"step": 3600
|
623 |
},
|
624 |
{
|
625 |
-
"epoch": 19.
|
626 |
-
"grad_norm": 2.393627405166626,
|
627 |
"learning_rate": 5e-05,
|
628 |
-
"loss": 0.
|
629 |
"step": 3700
|
630 |
},
|
631 |
{
|
632 |
"epoch": 20.0,
|
633 |
-
"eval_accuracy": 0.
|
634 |
-
"eval_loss": 2.
|
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":
|
647 |
-
"eval_recite_exact_match": 0.
|
648 |
"step": 3750
|
649 |
},
|
650 |
{
|
651 |
-
"epoch": 20.
|
652 |
-
"grad_norm": 2.1227269172668457,
|
653 |
"learning_rate": 5e-05,
|
654 |
-
"loss": 0.
|
655 |
"step": 3800
|
656 |
},
|
657 |
{
|
658 |
"epoch": 20.8,
|
659 |
-
"grad_norm": 2.2578742504119873,
|
660 |
"learning_rate": 5e-05,
|
661 |
-
"loss": 0.
|
662 |
"step": 3900
|
663 |
},
|
664 |
{
|
665 |
-
"epoch":
|
666 |
-
"eval_accuracy": 0.
|
667 |
-
"eval_loss": 2.
|
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": 1.9079784154891968,
|
686 |
"learning_rate": 5e-05,
|
687 |
-
"loss": 0.
|
688 |
"step": 4000
|
689 |
},
|
690 |
{
|
691 |
-
"epoch": 21.
|
692 |
-
"grad_norm": 2.5200765132904053,
|
693 |
"learning_rate": 5e-05,
|
694 |
-
"loss": 0.
|
695 |
"step": 4100
|
696 |
},
|
697 |
{
|
698 |
"epoch": 22.0,
|
699 |
-
"eval_accuracy": 0.
|
700 |
-
"eval_loss": 2.
|
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":
|
713 |
-
"eval_recite_exact_match": 0.
|
714 |
"step": 4125
|
715 |
},
|
716 |
{
|
717 |
"epoch": 22.4,
|
718 |
-
"grad_norm": 1.9616879224777222,
|
719 |
"learning_rate": 5e-05,
|
720 |
-
"loss": 0.
|
721 |
"step": 4200
|
722 |
},
|
723 |
{
|
724 |
-
"epoch": 22.
|
725 |
-
"grad_norm": 1.854027271270752,
|
726 |
"learning_rate": 5e-05,
|
727 |
-
"loss": 0.
|
728 |
"step": 4300
|
729 |
},
|
730 |
{
|
731 |
-
"epoch":
|
732 |
-
"eval_accuracy": 0.
|
733 |
-
"eval_loss": 2.
|
734 |
-
"eval_runtime":
|
735 |
-
"eval_samples_per_second":
|
736 |
-
"eval_steps_per_second":
|
737 |
"step": 4312
|
738 |
},
|
739 |
{
|
740 |
-
"epoch":
|
741 |
-
"eval_exact_match":
|
742 |
-
"eval_f1":
|
743 |
-
"eval_qa_bleu":
|
744 |
-
"eval_qa_exact_match": 0.
|
745 |
-
"eval_recite_bleu":
|
746 |
-
"eval_recite_exact_match": 0.
|
747 |
"step": 4312
|
748 |
},
|
749 |
{
|
750 |
-
"epoch": 23.
|
751 |
-
"grad_norm": 1.8527085781097412,
|
752 |
"learning_rate": 5e-05,
|
753 |
-
"loss": 0.
|
754 |
"step": 4400
|
755 |
},
|
756 |
{
|
757 |
"epoch": 24.0,
|
758 |
-
"grad_norm": 2.258077621459961,
|
759 |
"learning_rate": 5e-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": 2.0298190116882324,
|
785 |
"learning_rate": 5e-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":
|
805 |
-
"eval_recite_exact_match": 0.
|
806 |
"step": 4687
|
807 |
},
|
808 |
{
|
809 |
-
"epoch": 25.
|
810 |
-
"grad_norm": 1.9202630519866943,
|
811 |
"learning_rate": 5e-05,
|
812 |
-
"loss": 0.
|
813 |
"step": 4700
|
814 |
},
|
815 |
{
|
816 |
"epoch": 25.6,
|
817 |
-
"grad_norm": 2.008223295211792,
|
818 |
"learning_rate": 5e-05,
|
819 |
-
"loss": 0.
|
820 |
"step": 4800
|
821 |
},
|
822 |
{
|
823 |
"epoch": 26.0,
|
824 |
-
"eval_accuracy": 0.
|
825 |
-
"eval_loss":
|
826 |
-
"eval_runtime": 6.
|
827 |
-
"eval_samples_per_second": 75.
|
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": 7.
|
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": 2.0603179931640625,
|
844 |
"learning_rate": 5e-05,
|
845 |
-
"loss": 0.
|
846 |
"step": 4900
|
847 |
},
|
848 |
{
|
849 |
-
"epoch": 26.
|
850 |
-
"grad_norm": 2.4074087142944336,
|
851 |
"learning_rate": 5e-05,
|
852 |
-
"loss": 0.
|
853 |
"step": 5000
|
854 |
},
|
855 |
{
|
856 |
-
"epoch":
|
857 |
-
"eval_accuracy": 0.
|
858 |
-
"eval_loss":
|
859 |
-
"eval_runtime": 6.
|
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": 0.022,
|
872 |
"step": 5062
|
873 |
},
|
874 |
{
|
875 |
"epoch": 27.2,
|
876 |
-
"grad_norm": 2.3454041481018066,
|
877 |
"learning_rate": 5e-05,
|
878 |
-
"loss": 0.
|
879 |
"step": 5100
|
880 |
},
|
881 |
{
|
882 |
-
"epoch": 27.
|
883 |
-
"grad_norm": 1.7204937934875488,
|
884 |
"learning_rate": 5e-05,
|
885 |
-
"loss": 0.
|
886 |
"step": 5200
|
887 |
},
|
888 |
{
|
889 |
"epoch": 28.0,
|
890 |
-
"eval_accuracy": 0.
|
891 |
-
"eval_loss":
|
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":
|
900 |
-
"eval_f1":
|
901 |
-
"eval_qa_bleu":
|
902 |
-
"eval_qa_exact_match": 0.
|
903 |
-
"eval_recite_bleu":
|
904 |
-
"eval_recite_exact_match": 0.
|
905 |
"step": 5250
|
906 |
}
|
907 |
],
|
908 |
"logging_steps": 100,
|
909 |
"max_steps": 9350,
|
910 |
-
"num_input_tokens_seen": 0,
|
911 |
"num_train_epochs": 50,
|
912 |
"save_steps": 500,
|
913 |
-
"
|
914 |
-
"TrainerControl": {
|
915 |
-
"args": {
|
916 |
-
"should_epoch_stop": false,
|
917 |
-
"should_evaluate": false,
|
918 |
-
"should_log": false,
|
919 |
-
"should_save": true,
|
920 |
-
"should_training_stop": false
|
921 |
-
},
|
922 |
-
"attributes": {}
|
923 |
-
}
|
924 |
-
},
|
925 |
-
"total_flos": 5.28823278232404e+17,
|
926 |
-
"train_batch_size": 1,
|
927 |
"trial_name": null,
|
928 |
"trial_params": null
|
929 |
}
|
|
|
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 |
+
"epoch": 7.47,
|
|
|
224 |
"learning_rate": 5e-05,
|
225 |
+
"loss": 0.8,
|
226 |
"step": 1400
|
227 |
},
|
228 |
{
|
229 |
"epoch": 8.0,
|
|
|
230 |
"learning_rate": 5e-05,
|
231 |
+
"loss": 0.7933,
|
232 |
"step": 1500
|
233 |
},
|
234 |
{
|
235 |
"epoch": 8.0,
|
236 |
+
"eval_accuracy": 0.659607843137255,
|
237 |
+
"eval_loss": 1.3466852903366089,
|
238 |
+
"eval_runtime": 6.6563,
|
239 |
+
"eval_samples_per_second": 75.117,
|
240 |
+
"eval_steps_per_second": 9.465,
|
241 |
"step": 1500
|
242 |
},
|
243 |
{
|
244 |
"epoch": 8.0,
|
245 |
+
"eval_exact_match": 22.6,
|
246 |
+
"eval_f1": 31.16493506493508,
|
247 |
+
"eval_qa_bleu": 21.45561748665761,
|
248 |
+
"eval_qa_exact_match": 0.17,
|
249 |
+
"eval_recite_bleu": 14.23612921374107,
|
250 |
+
"eval_recite_exact_match": 0.012,
|
251 |
"step": 1500
|
252 |
},
|
253 |
{
|
254 |
+
"epoch": 8.53,
|
|
|
255 |
"learning_rate": 5e-05,
|
256 |
+
"loss": 0.7185,
|
257 |
"step": 1600
|
258 |
},
|
259 |
{
|
260 |
+
"epoch": 9.0,
|
261 |
+
"eval_accuracy": 0.6574980392156863,
|
262 |
+
"eval_loss": 1.391683578491211,
|
263 |
+
"eval_runtime": 7.7056,
|
264 |
+
"eval_samples_per_second": 64.888,
|
265 |
+
"eval_steps_per_second": 8.176,
|
266 |
"step": 1687
|
267 |
},
|
268 |
{
|
269 |
+
"epoch": 9.0,
|
270 |
+
"eval_exact_match": 21.8,
|
271 |
+
"eval_f1": 31.150317460317464,
|
272 |
+
"eval_qa_bleu": 18.411626756160196,
|
273 |
+
"eval_qa_exact_match": 0.17,
|
274 |
+
"eval_recite_bleu": 15.144735177405446,
|
275 |
+
"eval_recite_exact_match": 0.014,
|
276 |
"step": 1687
|
277 |
},
|
278 |
{
|
279 |
+
"epoch": 9.07,
|
|
|
280 |
"learning_rate": 5e-05,
|
281 |
+
"loss": 0.7176,
|
282 |
"step": 1700
|
283 |
},
|
284 |
{
|
285 |
"epoch": 9.6,
|
|
|
286 |
"learning_rate": 5e-05,
|
287 |
+
"loss": 0.6489,
|
288 |
"step": 1800
|
289 |
},
|
290 |
{
|
291 |
"epoch": 10.0,
|
292 |
+
"eval_accuracy": 0.6549647058823529,
|
293 |
+
"eval_loss": 1.4609253406524658,
|
294 |
+
"eval_runtime": 6.6599,
|
295 |
+
"eval_samples_per_second": 75.076,
|
296 |
+
"eval_steps_per_second": 9.46,
|
297 |
"step": 1875
|
298 |
},
|
299 |
{
|
300 |
"epoch": 10.0,
|
301 |
+
"eval_exact_match": 24.8,
|
302 |
+
"eval_f1": 32.80031746031747,
|
303 |
+
"eval_qa_bleu": 19.203751437503612,
|
304 |
+
"eval_qa_exact_match": 0.204,
|
305 |
+
"eval_recite_bleu": 14.497298759377461,
|
306 |
"eval_recite_exact_match": 0.016,
|
307 |
"step": 1875
|
308 |
},
|
309 |
{
|
310 |
+
"epoch": 10.13,
|
|
|
311 |
"learning_rate": 5e-05,
|
312 |
+
"loss": 0.6426,
|
313 |
"step": 1900
|
314 |
},
|
315 |
{
|
316 |
+
"epoch": 10.67,
|
|
|
317 |
"learning_rate": 5e-05,
|
318 |
+
"loss": 0.586,
|
319 |
"step": 2000
|
320 |
},
|
321 |
{
|
322 |
+
"epoch": 11.0,
|
323 |
+
"eval_accuracy": 0.6531372549019608,
|
324 |
+
"eval_loss": 1.5205447673797607,
|
325 |
+
"eval_runtime": 6.6744,
|
326 |
+
"eval_samples_per_second": 74.913,
|
327 |
+
"eval_steps_per_second": 9.439,
|
328 |
"step": 2062
|
329 |
},
|
330 |
{
|
331 |
+
"epoch": 11.0,
|
332 |
+
"eval_exact_match": 24.4,
|
333 |
+
"eval_f1": 32.14761904761905,
|
334 |
+
"eval_qa_bleu": 17.66314588893528,
|
335 |
+
"eval_qa_exact_match": 0.198,
|
336 |
+
"eval_recite_bleu": 14.285546168231928,
|
337 |
+
"eval_recite_exact_match": 0.016,
|
338 |
"step": 2062
|
339 |
},
|
340 |
{
|
341 |
"epoch": 11.2,
|
|
|
342 |
"learning_rate": 5e-05,
|
343 |
+
"loss": 0.5625,
|
344 |
"step": 2100
|
345 |
},
|
346 |
{
|
347 |
+
"epoch": 11.73,
|
|
|
348 |
"learning_rate": 5e-05,
|
349 |
+
"loss": 0.5267,
|
350 |
"step": 2200
|
351 |
},
|
352 |
{
|
353 |
"epoch": 12.0,
|
354 |
+
"eval_accuracy": 0.6518117647058823,
|
355 |
+
"eval_loss": 1.5615981817245483,
|
356 |
+
"eval_runtime": 7.0623,
|
357 |
+
"eval_samples_per_second": 70.799,
|
358 |
+
"eval_steps_per_second": 8.921,
|
359 |
"step": 2250
|
360 |
},
|
361 |
{
|
362 |
"epoch": 12.0,
|
363 |
+
"eval_exact_match": 22.0,
|
364 |
+
"eval_f1": 29.764505494505517,
|
365 |
+
"eval_qa_bleu": 14.595863052742681,
|
366 |
+
"eval_qa_exact_match": 0.174,
|
367 |
+
"eval_recite_bleu": 15.269859538584459,
|
368 |
+
"eval_recite_exact_match": 0.02,
|
369 |
"step": 2250
|
370 |
},
|
371 |
{
|
372 |
+
"epoch": 12.27,
|
|
|
373 |
"learning_rate": 5e-05,
|
374 |
+
"loss": 0.5074,
|
375 |
"step": 2300
|
376 |
},
|
377 |
{
|
378 |
"epoch": 12.8,
|
|
|
379 |
"learning_rate": 5e-05,
|
380 |
+
"loss": 0.4702,
|
381 |
"step": 2400
|
382 |
},
|
383 |
{
|
384 |
+
"epoch": 13.0,
|
385 |
+
"eval_accuracy": 0.6498588235294117,
|
386 |
+
"eval_loss": 1.6518086194992065,
|
387 |
+
"eval_runtime": 7.4952,
|
388 |
+
"eval_samples_per_second": 66.71,
|
389 |
+
"eval_steps_per_second": 8.405,
|
390 |
"step": 2437
|
391 |
},
|
392 |
{
|
393 |
+
"epoch": 13.0,
|
394 |
+
"eval_exact_match": 19.8,
|
395 |
+
"eval_f1": 29.199523809523825,
|
396 |
+
"eval_qa_bleu": 15.26272043877601,
|
397 |
+
"eval_qa_exact_match": 0.156,
|
398 |
+
"eval_recite_bleu": 13.712487638236569,
|
399 |
+
"eval_recite_exact_match": 0.012,
|
400 |
"step": 2437
|
401 |
},
|
402 |
{
|
403 |
+
"epoch": 13.33,
|
|
|
404 |
"learning_rate": 5e-05,
|
405 |
+
"loss": 0.4335,
|
406 |
"step": 2500
|
407 |
},
|
408 |
{
|
409 |
+
"epoch": 13.87,
|
|
|
410 |
"learning_rate": 5e-05,
|
411 |
+
"loss": 0.42,
|
412 |
"step": 2600
|
413 |
},
|
414 |
{
|
415 |
"epoch": 14.0,
|
416 |
+
"eval_accuracy": 0.6481411764705882,
|
417 |
+
"eval_loss": 1.7193351984024048,
|
418 |
+
"eval_runtime": 7.6621,
|
419 |
+
"eval_samples_per_second": 65.256,
|
420 |
+
"eval_steps_per_second": 8.222,
|
421 |
"step": 2625
|
422 |
},
|
423 |
{
|
424 |
"epoch": 14.0,
|
425 |
+
"eval_exact_match": 18.4,
|
426 |
+
"eval_f1": 27.123376623376636,
|
427 |
+
"eval_qa_bleu": 13.906554273237205,
|
428 |
+
"eval_qa_exact_match": 0.15,
|
429 |
+
"eval_recite_bleu": 14.177162892556794,
|
430 |
+
"eval_recite_exact_match": 0.016,
|
431 |
"step": 2625
|
432 |
},
|
433 |
{
|
434 |
"epoch": 14.4,
|
|
|
435 |
"learning_rate": 5e-05,
|
436 |
+
"loss": 0.3784,
|
437 |
"step": 2700
|
438 |
},
|
439 |
{
|
440 |
+
"epoch": 14.93,
|
|
|
441 |
"learning_rate": 5e-05,
|
442 |
+
"loss": 0.3811,
|
443 |
"step": 2800
|
444 |
},
|
445 |
{
|
446 |
+
"epoch": 15.0,
|
447 |
+
"eval_accuracy": 0.6458509803921568,
|
448 |
+
"eval_loss": 1.8015950918197632,
|
449 |
+
"eval_runtime": 7.7113,
|
450 |
+
"eval_samples_per_second": 64.84,
|
451 |
+
"eval_steps_per_second": 8.17,
|
452 |
"step": 2812
|
453 |
},
|
454 |
{
|
455 |
+
"epoch": 15.0,
|
456 |
+
"eval_exact_match": 19.0,
|
457 |
+
"eval_f1": 28.891111111111115,
|
458 |
+
"eval_qa_bleu": 16.83679026235492,
|
459 |
+
"eval_qa_exact_match": 0.154,
|
460 |
+
"eval_recite_bleu": 13.04215589870331,
|
461 |
+
"eval_recite_exact_match": 0.012,
|
462 |
"step": 2812
|
463 |
},
|
464 |
{
|
465 |
+
"epoch": 15.47,
|
|
|
466 |
"learning_rate": 5e-05,
|
467 |
+
"loss": 0.3287,
|
468 |
"step": 2900
|
469 |
},
|
470 |
{
|
471 |
"epoch": 16.0,
|
|
|
472 |
"learning_rate": 5e-05,
|
473 |
+
"loss": 0.3373,
|
474 |
"step": 3000
|
475 |
},
|
476 |
{
|
477 |
"epoch": 16.0,
|
478 |
+
"eval_accuracy": 0.6464627450980392,
|
479 |
+
"eval_loss": 1.8555551767349243,
|
480 |
+
"eval_runtime": 6.6548,
|
481 |
+
"eval_samples_per_second": 75.133,
|
482 |
+
"eval_steps_per_second": 9.467,
|
483 |
"step": 3000
|
484 |
},
|
485 |
{
|
486 |
"epoch": 16.0,
|
487 |
+
"eval_exact_match": 21.0,
|
488 |
+
"eval_f1": 30.546060606060603,
|
489 |
+
"eval_qa_bleu": 15.51818073374785,
|
490 |
+
"eval_qa_exact_match": 0.178,
|
491 |
+
"eval_recite_bleu": 14.611971996497491,
|
492 |
+
"eval_recite_exact_match": 0.016,
|
493 |
"step": 3000
|
494 |
},
|
495 |
{
|
496 |
+
"epoch": 16.53,
|
|
|
497 |
"learning_rate": 5e-05,
|
498 |
+
"loss": 0.284,
|
499 |
"step": 3100
|
500 |
},
|
501 |
{
|
502 |
+
"epoch": 17.0,
|
503 |
+
"eval_accuracy": 0.644313725490196,
|
504 |
+
"eval_loss": 1.9406696557998657,
|
505 |
+
"eval_runtime": 6.7553,
|
506 |
+
"eval_samples_per_second": 74.016,
|
507 |
+
"eval_steps_per_second": 9.326,
|
508 |
"step": 3187
|
509 |
},
|
510 |
{
|
511 |
+
"epoch": 17.0,
|
512 |
+
"eval_exact_match": 19.2,
|
513 |
+
"eval_f1": 26.95666666666667,
|
514 |
+
"eval_qa_bleu": 13.067462248071651,
|
515 |
+
"eval_qa_exact_match": 0.148,
|
516 |
+
"eval_recite_bleu": 13.516002177466241,
|
517 |
"eval_recite_exact_match": 0.016,
|
518 |
"step": 3187
|
519 |
},
|
520 |
{
|
521 |
+
"epoch": 17.07,
|
|
|
522 |
"learning_rate": 5e-05,
|
523 |
+
"loss": 0.2933,
|
524 |
"step": 3200
|
525 |
},
|
526 |
{
|
527 |
"epoch": 17.6,
|
|
|
528 |
"learning_rate": 5e-05,
|
529 |
+
"loss": 0.25,
|
530 |
"step": 3300
|
531 |
},
|
532 |
{
|
533 |
"epoch": 18.0,
|
534 |
+
"eval_accuracy": 0.6435764705882353,
|
535 |
+
"eval_loss": 1.9759635925292969,
|
536 |
+
"eval_runtime": 7.1655,
|
537 |
+
"eval_samples_per_second": 69.779,
|
538 |
+
"eval_steps_per_second": 8.792,
|
539 |
"step": 3375
|
540 |
},
|
541 |
{
|
542 |
"epoch": 18.0,
|
543 |
+
"eval_exact_match": 18.6,
|
544 |
+
"eval_f1": 27.20079365079366,
|
545 |
+
"eval_qa_bleu": 12.454425793037345,
|
546 |
+
"eval_qa_exact_match": 0.158,
|
547 |
+
"eval_recite_bleu": 13.86644732825742,
|
548 |
+
"eval_recite_exact_match": 0.014,
|
549 |
"step": 3375
|
550 |
},
|
551 |
{
|
552 |
+
"epoch": 18.13,
|
|
|
553 |
"learning_rate": 5e-05,
|
554 |
+
"loss": 0.2587,
|
555 |
"step": 3400
|
556 |
},
|
557 |
{
|
558 |
+
"epoch": 18.67,
|
|
|
559 |
"learning_rate": 5e-05,
|
560 |
+
"loss": 0.2274,
|
561 |
"step": 3500
|
562 |
},
|
563 |
{
|
564 |
+
"epoch": 19.0,
|
565 |
+
"eval_accuracy": 0.6424313725490196,
|
566 |
+
"eval_loss": 2.100283145904541,
|
567 |
+
"eval_runtime": 7.4607,
|
568 |
+
"eval_samples_per_second": 67.018,
|
569 |
+
"eval_steps_per_second": 8.444,
|
570 |
"step": 3562
|
571 |
},
|
572 |
{
|
573 |
+
"epoch": 19.0,
|
574 |
+
"eval_exact_match": 18.8,
|
575 |
+
"eval_f1": 26.78857142857144,
|
576 |
+
"eval_qa_bleu": 6.755652921468526,
|
577 |
+
"eval_qa_exact_match": 0.16,
|
578 |
+
"eval_recite_bleu": 13.96083887839662,
|
579 |
+
"eval_recite_exact_match": 0.018,
|
580 |
"step": 3562
|
581 |
},
|
582 |
{
|
583 |
"epoch": 19.2,
|
|
|
584 |
"learning_rate": 5e-05,
|
585 |
+
"loss": 0.2209,
|
586 |
"step": 3600
|
587 |
},
|
588 |
{
|
589 |
+
"epoch": 19.73,
|
|
|
590 |
"learning_rate": 5e-05,
|
591 |
+
"loss": 0.2033,
|
592 |
"step": 3700
|
593 |
},
|
594 |
{
|
595 |
"epoch": 20.0,
|
596 |
+
"eval_accuracy": 0.6424627450980392,
|
597 |
+
"eval_loss": 2.141883373260498,
|
598 |
+
"eval_runtime": 6.6558,
|
599 |
+
"eval_samples_per_second": 75.122,
|
600 |
+
"eval_steps_per_second": 9.465,
|
601 |
"step": 3750
|
602 |
},
|
603 |
{
|
604 |
"epoch": 20.0,
|
605 |
+
"eval_exact_match": 20.4,
|
606 |
+
"eval_f1": 28.975054945054936,
|
607 |
+
"eval_qa_bleu": 14.791681538972547,
|
608 |
+
"eval_qa_exact_match": 0.168,
|
609 |
+
"eval_recite_bleu": 14.920846894015328,
|
610 |
+
"eval_recite_exact_match": 0.016,
|
611 |
"step": 3750
|
612 |
},
|
613 |
{
|
614 |
+
"epoch": 20.27,
|
|
|
615 |
"learning_rate": 5e-05,
|
616 |
+
"loss": 0.1928,
|
617 |
"step": 3800
|
618 |
},
|
619 |
{
|
620 |
"epoch": 20.8,
|
|
|
621 |
"learning_rate": 5e-05,
|
622 |
+
"loss": 0.1832,
|
623 |
"step": 3900
|
624 |
},
|
625 |
{
|
626 |
+
"epoch": 21.0,
|
627 |
+
"eval_accuracy": 0.6405333333333333,
|
628 |
+
"eval_loss": 2.2004899978637695,
|
629 |
+
"eval_runtime": 7.0682,
|
630 |
+
"eval_samples_per_second": 70.739,
|
631 |
+
"eval_steps_per_second": 8.913,
|
632 |
"step": 3937
|
633 |
},
|
634 |
{
|
635 |
+
"epoch": 21.0,
|
636 |
+
"eval_exact_match": 19.8,
|
637 |
+
"eval_f1": 27.03857142857144,
|
638 |
+
"eval_qa_bleu": 14.16698111965369,
|
639 |
+
"eval_qa_exact_match": 0.152,
|
640 |
+
"eval_recite_bleu": 14.34411744472555,
|
641 |
+
"eval_recite_exact_match": 0.016,
|
642 |
"step": 3937
|
643 |
},
|
644 |
{
|
645 |
+
"epoch": 21.33,
|
|
|
646 |
"learning_rate": 5e-05,
|
647 |
+
"loss": 0.1674,
|
648 |
"step": 4000
|
649 |
},
|
650 |
{
|
651 |
+
"epoch": 21.87,
|
|
|
652 |
"learning_rate": 5e-05,
|
653 |
+
"loss": 0.1678,
|
654 |
"step": 4100
|
655 |
},
|
656 |
{
|
657 |
"epoch": 22.0,
|
658 |
+
"eval_accuracy": 0.6396627450980392,
|
659 |
+
"eval_loss": 2.2465262413024902,
|
660 |
+
"eval_runtime": 7.6611,
|
661 |
+
"eval_samples_per_second": 65.264,
|
662 |
+
"eval_steps_per_second": 8.223,
|
663 |
"step": 4125
|
664 |
},
|
665 |
{
|
666 |
"epoch": 22.0,
|
667 |
+
"eval_exact_match": 18.8,
|
668 |
+
"eval_f1": 26.868253968253967,
|
669 |
+
"eval_qa_bleu": 16.683806863576216,
|
670 |
+
"eval_qa_exact_match": 0.15,
|
671 |
+
"eval_recite_bleu": 15.201480663492395,
|
672 |
+
"eval_recite_exact_match": 0.018,
|
673 |
"step": 4125
|
674 |
},
|
675 |
{
|
676 |
"epoch": 22.4,
|
|
|
677 |
"learning_rate": 5e-05,
|
678 |
+
"loss": 0.1516,
|
679 |
"step": 4200
|
680 |
},
|
681 |
{
|
682 |
+
"epoch": 22.93,
|
|
|
683 |
"learning_rate": 5e-05,
|
684 |
+
"loss": 0.1505,
|
685 |
"step": 4300
|
686 |
},
|
687 |
{
|
688 |
+
"epoch": 23.0,
|
689 |
+
"eval_accuracy": 0.6399607843137255,
|
690 |
+
"eval_loss": 2.292829990386963,
|
691 |
+
"eval_runtime": 7.6966,
|
692 |
+
"eval_samples_per_second": 64.963,
|
693 |
+
"eval_steps_per_second": 8.185,
|
694 |
"step": 4312
|
695 |
},
|
696 |
{
|
697 |
+
"epoch": 23.0,
|
698 |
+
"eval_exact_match": 20.4,
|
699 |
+
"eval_f1": 28.723809523809532,
|
700 |
+
"eval_qa_bleu": 14.649664836489828,
|
701 |
+
"eval_qa_exact_match": 0.16,
|
702 |
+
"eval_recite_bleu": 14.44141809683148,
|
703 |
+
"eval_recite_exact_match": 0.016,
|
704 |
"step": 4312
|
705 |
},
|
706 |
{
|
707 |
+
"epoch": 23.47,
|
|
|
708 |
"learning_rate": 5e-05,
|
709 |
+
"loss": 0.1345,
|
710 |
"step": 4400
|
711 |
},
|
712 |
{
|
713 |
"epoch": 24.0,
|
|
|
714 |
"learning_rate": 5e-05,
|
715 |
+
"loss": 0.1406,
|
716 |
"step": 4500
|
717 |
},
|
718 |
{
|
719 |
"epoch": 24.0,
|
720 |
+
"eval_accuracy": 0.6399686274509804,
|
721 |
+
"eval_loss": 2.369138240814209,
|
722 |
+
"eval_runtime": 7.0721,
|
723 |
+
"eval_samples_per_second": 70.701,
|
724 |
+
"eval_steps_per_second": 8.908,
|
725 |
"step": 4500
|
726 |
},
|
727 |
{
|
728 |
"epoch": 24.0,
|
729 |
+
"eval_exact_match": 17.8,
|
730 |
+
"eval_f1": 25.43047619047621,
|
731 |
+
"eval_qa_bleu": 13.339685450002685,
|
732 |
+
"eval_qa_exact_match": 0.148,
|
733 |
+
"eval_recite_bleu": 13.754679861863655,
|
734 |
+
"eval_recite_exact_match": 0.018,
|
735 |
"step": 4500
|
736 |
},
|
737 |
{
|
738 |
+
"epoch": 24.53,
|
|
|
739 |
"learning_rate": 5e-05,
|
740 |
+
"loss": 0.1214,
|
741 |
"step": 4600
|
742 |
},
|
743 |
{
|
744 |
+
"epoch": 25.0,
|
745 |
+
"eval_accuracy": 0.6389882352941176,
|
746 |
+
"eval_loss": 2.41001033782959,
|
747 |
+
"eval_runtime": 7.2032,
|
748 |
+
"eval_samples_per_second": 69.414,
|
749 |
+
"eval_steps_per_second": 8.746,
|
750 |
"step": 4687
|
751 |
},
|
752 |
{
|
753 |
+
"epoch": 25.0,
|
754 |
+
"eval_exact_match": 18.0,
|
755 |
+
"eval_f1": 26.35142857142858,
|
756 |
+
"eval_qa_bleu": 10.51643947107139,
|
757 |
+
"eval_qa_exact_match": 0.146,
|
758 |
+
"eval_recite_bleu": 13.73909597941028,
|
759 |
+
"eval_recite_exact_match": 0.02,
|
760 |
"step": 4687
|
761 |
},
|
762 |
{
|
763 |
+
"epoch": 25.07,
|
|
|
764 |
"learning_rate": 5e-05,
|
765 |
+
"loss": 0.1266,
|
766 |
"step": 4700
|
767 |
},
|
768 |
{
|
769 |
"epoch": 25.6,
|
|
|
770 |
"learning_rate": 5e-05,
|
771 |
+
"loss": 0.1128,
|
772 |
"step": 4800
|
773 |
},
|
774 |
{
|
775 |
"epoch": 26.0,
|
776 |
+
"eval_accuracy": 0.6381725490196078,
|
777 |
+
"eval_loss": 2.484750986099243,
|
778 |
+
"eval_runtime": 6.6571,
|
779 |
+
"eval_samples_per_second": 75.107,
|
780 |
+
"eval_steps_per_second": 9.464,
|
781 |
"step": 4875
|
782 |
},
|
783 |
{
|
784 |
"epoch": 26.0,
|
785 |
+
"eval_exact_match": 19.0,
|
786 |
+
"eval_f1": 28.15619047619049,
|
787 |
+
"eval_qa_bleu": 7.883655295314112,
|
788 |
+
"eval_qa_exact_match": 0.158,
|
789 |
+
"eval_recite_bleu": 14.563137659888193,
|
790 |
+
"eval_recite_exact_match": 0.02,
|
791 |
"step": 4875
|
792 |
},
|
793 |
{
|
794 |
+
"epoch": 26.13,
|
|
|
795 |
"learning_rate": 5e-05,
|
796 |
+
"loss": 0.1179,
|
797 |
"step": 4900
|
798 |
},
|
799 |
{
|
800 |
+
"epoch": 26.67,
|
|
|
801 |
"learning_rate": 5e-05,
|
802 |
+
"loss": 0.1076,
|
803 |
"step": 5000
|
804 |
},
|
805 |
{
|
806 |
+
"epoch": 27.0,
|
807 |
+
"eval_accuracy": 0.638956862745098,
|
808 |
+
"eval_loss": 2.546912908554077,
|
809 |
+
"eval_runtime": 6.6822,
|
810 |
+
"eval_samples_per_second": 74.826,
|
811 |
+
"eval_steps_per_second": 9.428,
|
812 |
"step": 5062
|
813 |
},
|
814 |
{
|
815 |
+
"epoch": 27.0,
|
816 |
+
"eval_exact_match": 18.8,
|
817 |
+
"eval_f1": 25.95916971916973,
|
818 |
+
"eval_qa_bleu": 6.531854323689431,
|
819 |
+
"eval_qa_exact_match": 0.162,
|
820 |
+
"eval_recite_bleu": 13.16436178935274,
|
821 |
"eval_recite_exact_match": 0.022,
|
822 |
"step": 5062
|
823 |
},
|
824 |
{
|
825 |
"epoch": 27.2,
|
|
|
826 |
"learning_rate": 5e-05,
|
827 |
+
"loss": 0.1095,
|
828 |
"step": 5100
|
829 |
},
|
830 |
{
|
831 |
+
"epoch": 27.73,
|
|
|
832 |
"learning_rate": 5e-05,
|
833 |
+
"loss": 0.1046,
|
834 |
"step": 5200
|
835 |
},
|
836 |
{
|
837 |
"epoch": 28.0,
|
838 |
+
"eval_accuracy": 0.6391529411764706,
|
839 |
+
"eval_loss": 2.5205068588256836,
|
840 |
+
"eval_runtime": 7.4592,
|
841 |
+
"eval_samples_per_second": 67.031,
|
842 |
+
"eval_steps_per_second": 8.446,
|
843 |
"step": 5250
|
844 |
},
|
845 |
{
|
846 |
"epoch": 28.0,
|
847 |
+
"eval_exact_match": 17.2,
|
848 |
+
"eval_f1": 25.731298701298712,
|
849 |
+
"eval_qa_bleu": 8.206606144769799,
|
850 |
+
"eval_qa_exact_match": 0.14,
|
851 |
+
"eval_recite_bleu": 14.236424434530772,
|
852 |
+
"eval_recite_exact_match": 0.022,
|
853 |
"step": 5250
|
854 |
}
|
855 |
],
|
856 |
"logging_steps": 100,
|
857 |
"max_steps": 9350,
|
|
|
858 |
"num_train_epochs": 50,
|
859 |
"save_steps": 500,
|
860 |
+
"total_flos": 1.216838762839933e+18,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
861 |
"trial_name": null,
|
862 |
"trial_params": null
|
863 |
}
|
checkpoint-5250/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
|