Hugging Face's logo

tyzhu
/
lmind_hotpot_train8000_eval7405_v1_qa_5e-5_lora2

Safetensors
Generated from Trainer
Eval Results
Model card Files Community
tyzhu commited on
Commit
c903a11
•
1 Parent(s): b3414d9

Training in progress, epoch 26, checkpoint

Browse files
Files changed (17) hide show
  1. checkpoint-6500/README.md +9 -0
  2. checkpoint-6500/adapter_config.json +25 -0
  3. checkpoint-6500/adapter_model.safetensors +3 -0
  4. checkpoint-6500/added_tokens.json +5 -0
  5. checkpoint-6500/merges.txt +0 -0
  6. checkpoint-6500/optimizer.pt +3 -0
  7. checkpoint-6500/rng_state_0.pth +3 -0
  8. checkpoint-6500/rng_state_1.pth +3 -0
  9. checkpoint-6500/rng_state_2.pth +3 -0
  10. checkpoint-6500/rng_state_3.pth +3 -0
  11. checkpoint-6500/scheduler.pt +3 -0
  12. checkpoint-6500/special_tokens_map.json +14 -0
  13. checkpoint-6500/tokenizer.json +0 -0
  14. checkpoint-6500/tokenizer_config.json +43 -0
  15. checkpoint-6500/trainer_state.json +878 -0
  16. checkpoint-6500/training_args.bin +3 -0
  17. checkpoint-6500/vocab.json +0 -0
checkpoint-6500/README.md ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+ ### Framework versions
7
+
8
+
9
+ - PEFT 0.5.0
checkpoint-6500/adapter_config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "Qwen/Qwen1.5-4B",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 16,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 16,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "k_proj",
19
+ "v_proj",
20
+ "down_proj",
21
+ "gate_proj",
22
+ "up_proj"
23
+ ],
24
+ "task_type": "CAUSAL_LM"
25
+ }
checkpoint-6500/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:edc8f1aa766cc4fb510ef01e72326a0d5685f00e1b5c2b149569344a8457cf6e
3
+ size 112130416
checkpoint-6500/added_tokens.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "<|endoftext|>": 151643,
3
+ "<|im_end|>": 151645,
4
+ "<|im_start|>": 151644
5
+ }
checkpoint-6500/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-6500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a039a1e8ffed8181f984c1c5af72fb29f8705583e377311bcb59f1fe9fcf2581
3
+ size 224537202
checkpoint-6500/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:14f789438082dd784e62ae62d9104695a14511fb5740a70545d87bb353388471
3
+ size 15024
checkpoint-6500/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bf531718dfcefad10c8614534993002983d2affd5b94db0ae1c52c54d5951524
3
+ size 15024
checkpoint-6500/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cb7eddff500d42dff679f55a7da49cfc2538f71f6f2e6ea376862d382026ed47
3
+ size 15024
checkpoint-6500/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:952624b0c12d54e614917bf1e78861b6b6d4014ea6513a8f3b8d421b58151c91
3
+ size 15024
checkpoint-6500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:85e453409c9e3c8042ebf3266b79d069692fb0d2565f89ae9a563257c2aced16
3
+ size 1064
checkpoint-6500/special_tokens_map.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|im_start|>",
4
+ "<|im_end|>"
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
+ }
checkpoint-6500/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-6500/tokenizer_config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "151643": {
5
+ "content": "<|endoftext|>",
6
+ "lstrip": false,
7
+ "normalized": false,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "151644": {
13
+ "content": "<|im_start|>",
14
+ "lstrip": false,
15
+ "normalized": false,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "151645": {
21
+ "content": "<|im_end|>",
22
+ "lstrip": false,
23
+ "normalized": false,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": true
27
+ }
28
+ },
29
+ "additional_special_tokens": [
30
+ "<|im_start|>",
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": "<|endoftext|>",
37
+ "errors": "replace",
38
+ "model_max_length": 32768,
39
+ "pad_token": "<|endoftext|>",
40
+ "split_special_tokens": false,
41
+ "tokenizer_class": "Qwen2Tokenizer",
42
+ "unk_token": null
43
+ }
checkpoint-6500/trainer_state.json ADDED
@@ -0,0 +1,878 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 26.0,
5
+ "eval_steps": 500,
6
+ "global_step": 6500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.4,
13
+ "grad_norm": 0.6251392960548401,
14
+ "learning_rate": 5e-05,
15
+ "loss": 2.427,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.8,
20
+ "grad_norm": 0.6590011715888977,
21
+ "learning_rate": 5e-05,
22
+ "loss": 2.2748,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 1.0,
27
+ "eval_accuracy": 0.5142539682539683,
28
+ "eval_loss": 2.336690902709961,
29
+ "eval_runtime": 5.4733,
30
+ "eval_samples_per_second": 91.353,
31
+ "eval_steps_per_second": 11.51,
32
+ "step": 250
33
+ },
34
+ {
35
+ "epoch": 1.0,
36
+ "eval_exact_match": 16.8,
37
+ "eval_f1": 24.850246583481884,
38
+ "step": 250
39
+ },
40
+ {
41
+ "epoch": 1.2,
42
+ "grad_norm": 0.8194239139556885,
43
+ "learning_rate": 5e-05,
44
+ "loss": 2.2322,
45
+ "step": 300
46
+ },
47
+ {
48
+ "epoch": 1.6,
49
+ "grad_norm": 0.9867294430732727,
50
+ "learning_rate": 5e-05,
51
+ "loss": 2.1904,
52
+ "step": 400
53
+ },
54
+ {
55
+ "epoch": 2.0,
56
+ "grad_norm": 0.9660400152206421,
57
+ "learning_rate": 5e-05,
58
+ "loss": 2.1657,
59
+ "step": 500
60
+ },
61
+ {
62
+ "epoch": 2.0,
63
+ "eval_accuracy": 0.5168253968253969,
64
+ "eval_loss": 2.323688507080078,
65
+ "eval_runtime": 5.1683,
66
+ "eval_samples_per_second": 96.744,
67
+ "eval_steps_per_second": 12.19,
68
+ "step": 500
69
+ },
70
+ {
71
+ "epoch": 2.0,
72
+ "eval_exact_match": 17.2,
73
+ "eval_f1": 24.620036075036083,
74
+ "step": 500
75
+ },
76
+ {
77
+ "epoch": 2.4,
78
+ "grad_norm": 1.281128168106079,
79
+ "learning_rate": 5e-05,
80
+ "loss": 2.0251,
81
+ "step": 600
82
+ },
83
+ {
84
+ "epoch": 2.8,
85
+ "grad_norm": 1.3365116119384766,
86
+ "learning_rate": 5e-05,
87
+ "loss": 2.0158,
88
+ "step": 700
89
+ },
90
+ {
91
+ "epoch": 3.0,
92
+ "eval_accuracy": 0.5156507936507937,
93
+ "eval_loss": 2.3560824394226074,
94
+ "eval_runtime": 5.4675,
95
+ "eval_samples_per_second": 91.449,
96
+ "eval_steps_per_second": 11.523,
97
+ "step": 750
98
+ },
99
+ {
100
+ "epoch": 3.0,
101
+ "eval_exact_match": 19.0,
102
+ "eval_f1": 27.242790688843332,
103
+ "step": 750
104
+ },
105
+ {
106
+ "epoch": 3.2,
107
+ "grad_norm": 1.7840828895568848,
108
+ "learning_rate": 5e-05,
109
+ "loss": 1.9394,
110
+ "step": 800
111
+ },
112
+ {
113
+ "epoch": 3.6,
114
+ "grad_norm": 1.8030190467834473,
115
+ "learning_rate": 5e-05,
116
+ "loss": 1.8502,
117
+ "step": 900
118
+ },
119
+ {
120
+ "epoch": 4.0,
121
+ "grad_norm": 1.9682222604751587,
122
+ "learning_rate": 5e-05,
123
+ "loss": 1.8682,
124
+ "step": 1000
125
+ },
126
+ {
127
+ "epoch": 4.0,
128
+ "eval_accuracy": 0.5131746031746032,
129
+ "eval_loss": 2.4071946144104004,
130
+ "eval_runtime": 5.2825,
131
+ "eval_samples_per_second": 94.653,
132
+ "eval_steps_per_second": 11.926,
133
+ "step": 1000
134
+ },
135
+ {
136
+ "epoch": 4.0,
137
+ "eval_exact_match": 17.6,
138
+ "eval_f1": 26.010988455988468,
139
+ "step": 1000
140
+ },
141
+ {
142
+ "epoch": 4.4,
143
+ "grad_norm": 2.138505220413208,
144
+ "learning_rate": 5e-05,
145
+ "loss": 1.6795,
146
+ "step": 1100
147
+ },
148
+ {
149
+ "epoch": 4.8,
150
+ "grad_norm": 2.3212926387786865,
151
+ "learning_rate": 5e-05,
152
+ "loss": 1.7036,
153
+ "step": 1200
154
+ },
155
+ {
156
+ "epoch": 5.0,
157
+ "eval_accuracy": 0.5099365079365079,
158
+ "eval_loss": 2.494770050048828,
159
+ "eval_runtime": 5.1799,
160
+ "eval_samples_per_second": 96.528,
161
+ "eval_steps_per_second": 12.162,
162
+ "step": 1250
163
+ },
164
+ {
165
+ "epoch": 5.0,
166
+ "eval_exact_match": 17.2,
167
+ "eval_f1": 25.84752525252526,
168
+ "step": 1250
169
+ },
170
+ {
171
+ "epoch": 5.2,
172
+ "grad_norm": 2.4878485202789307,
173
+ "learning_rate": 5e-05,
174
+ "loss": 1.6305,
175
+ "step": 1300
176
+ },
177
+ {
178
+ "epoch": 5.6,
179
+ "grad_norm": 2.599329710006714,
180
+ "learning_rate": 5e-05,
181
+ "loss": 1.5477,
182
+ "step": 1400
183
+ },
184
+ {
185
+ "epoch": 6.0,
186
+ "grad_norm": 2.5542759895324707,
187
+ "learning_rate": 5e-05,
188
+ "loss": 1.5708,
189
+ "step": 1500
190
+ },
191
+ {
192
+ "epoch": 6.0,
193
+ "eval_accuracy": 0.5065079365079365,
194
+ "eval_loss": 2.5797038078308105,
195
+ "eval_runtime": 5.6827,
196
+ "eval_samples_per_second": 87.986,
197
+ "eval_steps_per_second": 11.086,
198
+ "step": 1500
199
+ },
200
+ {
201
+ "epoch": 6.0,
202
+ "eval_exact_match": 17.0,
203
+ "eval_f1": 25.310858585858586,
204
+ "step": 1500
205
+ },
206
+ {
207
+ "epoch": 6.4,
208
+ "grad_norm": 3.1968846321105957,
209
+ "learning_rate": 5e-05,
210
+ "loss": 1.3896,
211
+ "step": 1600
212
+ },
213
+ {
214
+ "epoch": 6.8,
215
+ "grad_norm": 3.8637375831604004,
216
+ "learning_rate": 5e-05,
217
+ "loss": 1.421,
218
+ "step": 1700
219
+ },
220
+ {
221
+ "epoch": 7.0,
222
+ "eval_accuracy": 0.5038412698412699,
223
+ "eval_loss": 2.6977317333221436,
224
+ "eval_runtime": 5.3312,
225
+ "eval_samples_per_second": 93.787,
226
+ "eval_steps_per_second": 11.817,
227
+ "step": 1750
228
+ },
229
+ {
230
+ "epoch": 7.0,
231
+ "eval_exact_match": 16.0,
232
+ "eval_f1": 23.99441558441559,
233
+ "step": 1750
234
+ },
235
+ {
236
+ "epoch": 7.2,
237
+ "grad_norm": 3.468459367752075,
238
+ "learning_rate": 5e-05,
239
+ "loss": 1.3366,
240
+ "step": 1800
241
+ },
242
+ {
243
+ "epoch": 7.6,
244
+ "grad_norm": 3.5727667808532715,
245
+ "learning_rate": 5e-05,
246
+ "loss": 1.2585,
247
+ "step": 1900
248
+ },
249
+ {
250
+ "epoch": 8.0,
251
+ "grad_norm": 3.508258819580078,
252
+ "learning_rate": 5e-05,
253
+ "loss": 1.296,
254
+ "step": 2000
255
+ },
256
+ {
257
+ "epoch": 8.0,
258
+ "eval_accuracy": 0.5013333333333333,
259
+ "eval_loss": 2.786928653717041,
260
+ "eval_runtime": 5.8398,
261
+ "eval_samples_per_second": 85.62,
262
+ "eval_steps_per_second": 10.788,
263
+ "step": 2000
264
+ },
265
+ {
266
+ "epoch": 8.0,
267
+ "eval_exact_match": 15.4,
268
+ "eval_f1": 22.542871572871572,
269
+ "step": 2000
270
+ },
271
+ {
272
+ "epoch": 8.4,
273
+ "grad_norm": 4.042112350463867,
274
+ "learning_rate": 5e-05,
275
+ "loss": 1.1084,
276
+ "step": 2100
277
+ },
278
+ {
279
+ "epoch": 8.8,
280
+ "grad_norm": 3.7548155784606934,
281
+ "learning_rate": 5e-05,
282
+ "loss": 1.1457,
283
+ "step": 2200
284
+ },
285
+ {
286
+ "epoch": 9.0,
287
+ "eval_accuracy": 0.49692063492063493,
288
+ "eval_loss": 2.9215197563171387,
289
+ "eval_runtime": 5.1412,
290
+ "eval_samples_per_second": 97.254,
291
+ "eval_steps_per_second": 12.254,
292
+ "step": 2250
293
+ },
294
+ {
295
+ "epoch": 9.0,
296
+ "eval_exact_match": 15.0,
297
+ "eval_f1": 21.59087301587302,
298
+ "step": 2250
299
+ },
300
+ {
301
+ "epoch": 9.2,
302
+ "grad_norm": 4.0875773429870605,
303
+ "learning_rate": 5e-05,
304
+ "loss": 1.0677,
305
+ "step": 2300
306
+ },
307
+ {
308
+ "epoch": 9.6,
309
+ "grad_norm": 4.143901348114014,
310
+ "learning_rate": 5e-05,
311
+ "loss": 1.0028,
312
+ "step": 2400
313
+ },
314
+ {
315
+ "epoch": 10.0,
316
+ "grad_norm": 5.0514445304870605,
317
+ "learning_rate": 5e-05,
318
+ "loss": 1.0401,
319
+ "step": 2500
320
+ },
321
+ {
322
+ "epoch": 10.0,
323
+ "eval_accuracy": 0.4953015873015873,
324
+ "eval_loss": 3.054328203201294,
325
+ "eval_runtime": 5.4195,
326
+ "eval_samples_per_second": 92.259,
327
+ "eval_steps_per_second": 11.625,
328
+ "step": 2500
329
+ },
330
+ {
331
+ "epoch": 10.0,
332
+ "eval_exact_match": 15.4,
333
+ "eval_f1": 21.505119880119878,
334
+ "step": 2500
335
+ },
336
+ {
337
+ "epoch": 10.4,
338
+ "grad_norm": 3.8740394115448,
339
+ "learning_rate": 5e-05,
340
+ "loss": 0.8595,
341
+ "step": 2600
342
+ },
343
+ {
344
+ "epoch": 10.8,
345
+ "grad_norm": 4.170004844665527,
346
+ "learning_rate": 5e-05,
347
+ "loss": 0.9186,
348
+ "step": 2700
349
+ },
350
+ {
351
+ "epoch": 11.0,
352
+ "eval_accuracy": 0.49298412698412697,
353
+ "eval_loss": 3.177030324935913,
354
+ "eval_runtime": 5.9501,
355
+ "eval_samples_per_second": 84.033,
356
+ "eval_steps_per_second": 10.588,
357
+ "step": 2750
358
+ },
359
+ {
360
+ "epoch": 11.0,
361
+ "eval_exact_match": 15.6,
362
+ "eval_f1": 20.912753357753356,
363
+ "step": 2750
364
+ },
365
+ {
366
+ "epoch": 11.2,
367
+ "grad_norm": 4.2533392906188965,
368
+ "learning_rate": 5e-05,
369
+ "loss": 0.8411,
370
+ "step": 2800
371
+ },
372
+ {
373
+ "epoch": 11.6,
374
+ "grad_norm": 4.165424346923828,
375
+ "learning_rate": 5e-05,
376
+ "loss": 0.7934,
377
+ "step": 2900
378
+ },
379
+ {
380
+ "epoch": 12.0,
381
+ "grad_norm": 4.517414093017578,
382
+ "learning_rate": 5e-05,
383
+ "loss": 0.8386,
384
+ "step": 3000
385
+ },
386
+ {
387
+ "epoch": 12.0,
388
+ "eval_accuracy": 0.4916190476190476,
389
+ "eval_loss": 3.2663722038269043,
390
+ "eval_runtime": 5.1518,
391
+ "eval_samples_per_second": 97.053,
392
+ "eval_steps_per_second": 12.229,
393
+ "step": 3000
394
+ },
395
+ {
396
+ "epoch": 12.0,
397
+ "eval_exact_match": 15.2,
398
+ "eval_f1": 20.465463425463433,
399
+ "step": 3000
400
+ },
401
+ {
402
+ "epoch": 12.4,
403
+ "grad_norm": 4.059298515319824,
404
+ "learning_rate": 5e-05,
405
+ "loss": 0.6908,
406
+ "step": 3100
407
+ },
408
+ {
409
+ "epoch": 12.8,
410
+ "grad_norm": 3.4847490787506104,
411
+ "learning_rate": 5e-05,
412
+ "loss": 0.7326,
413
+ "step": 3200
414
+ },
415
+ {
416
+ "epoch": 13.0,
417
+ "eval_accuracy": 0.49047619047619045,
418
+ "eval_loss": 3.38826584815979,
419
+ "eval_runtime": 5.2085,
420
+ "eval_samples_per_second": 95.997,
421
+ "eval_steps_per_second": 12.096,
422
+ "step": 3250
423
+ },
424
+ {
425
+ "epoch": 13.0,
426
+ "eval_exact_match": 14.0,
427
+ "eval_f1": 20.144243697478988,
428
+ "step": 3250
429
+ },
430
+ {
431
+ "epoch": 13.2,
432
+ "grad_norm": 4.07555627822876,
433
+ "learning_rate": 5e-05,
434
+ "loss": 0.6901,
435
+ "step": 3300
436
+ },
437
+ {
438
+ "epoch": 13.6,
439
+ "grad_norm": 3.918684482574463,
440
+ "learning_rate": 5e-05,
441
+ "loss": 0.6432,
442
+ "step": 3400
443
+ },
444
+ {
445
+ "epoch": 14.0,
446
+ "grad_norm": 4.3187384605407715,
447
+ "learning_rate": 5e-05,
448
+ "loss": 0.6762,
449
+ "step": 3500
450
+ },
451
+ {
452
+ "epoch": 14.0,
453
+ "eval_accuracy": 0.4887936507936508,
454
+ "eval_loss": 3.466782808303833,
455
+ "eval_runtime": 5.5309,
456
+ "eval_samples_per_second": 90.402,
457
+ "eval_steps_per_second": 11.391,
458
+ "step": 3500
459
+ },
460
+ {
461
+ "epoch": 14.0,
462
+ "eval_exact_match": 14.6,
463
+ "eval_f1": 20.548629148629157,
464
+ "step": 3500
465
+ },
466
+ {
467
+ "epoch": 14.4,
468
+ "grad_norm": 3.270817995071411,
469
+ "learning_rate": 5e-05,
470
+ "loss": 0.5785,
471
+ "step": 3600
472
+ },
473
+ {
474
+ "epoch": 14.8,
475
+ "grad_norm": 3.882906436920166,
476
+ "learning_rate": 5e-05,
477
+ "loss": 0.6117,
478
+ "step": 3700
479
+ },
480
+ {
481
+ "epoch": 15.0,
482
+ "eval_accuracy": 0.48904761904761906,
483
+ "eval_loss": 3.5602962970733643,
484
+ "eval_runtime": 5.4424,
485
+ "eval_samples_per_second": 91.871,
486
+ "eval_steps_per_second": 11.576,
487
+ "step": 3750
488
+ },
489
+ {
490
+ "epoch": 15.0,
491
+ "eval_exact_match": 14.8,
492
+ "eval_f1": 20.941592997645635,
493
+ "step": 3750
494
+ },
495
+ {
496
+ "epoch": 15.2,
497
+ "grad_norm": 4.002102851867676,
498
+ "learning_rate": 5e-05,
499
+ "loss": 0.5725,
500
+ "step": 3800
501
+ },
502
+ {
503
+ "epoch": 15.6,
504
+ "grad_norm": 2.8693177700042725,
505
+ "learning_rate": 5e-05,
506
+ "loss": 0.5581,
507
+ "step": 3900
508
+ },
509
+ {
510
+ "epoch": 16.0,
511
+ "grad_norm": 4.3405256271362305,
512
+ "learning_rate": 5e-05,
513
+ "loss": 0.5729,
514
+ "step": 4000
515
+ },
516
+ {
517
+ "epoch": 16.0,
518
+ "eval_accuracy": 0.48742857142857143,
519
+ "eval_loss": 3.6583478450775146,
520
+ "eval_runtime": 5.1711,
521
+ "eval_samples_per_second": 96.692,
522
+ "eval_steps_per_second": 12.183,
523
+ "step": 4000
524
+ },
525
+ {
526
+ "epoch": 16.0,
527
+ "eval_exact_match": 14.8,
528
+ "eval_f1": 20.59385852709383,
529
+ "step": 4000
530
+ },
531
+ {
532
+ "epoch": 16.4,
533
+ "grad_norm": 4.715936183929443,
534
+ "learning_rate": 5e-05,
535
+ "loss": 0.5028,
536
+ "step": 4100
537
+ },
538
+ {
539
+ "epoch": 16.8,
540
+ "grad_norm": 3.7355644702911377,
541
+ "learning_rate": 5e-05,
542
+ "loss": 0.5317,
543
+ "step": 4200
544
+ },
545
+ {
546
+ "epoch": 17.0,
547
+ "eval_accuracy": 0.4873968253968254,
548
+ "eval_loss": 3.6984667778015137,
549
+ "eval_runtime": 5.4593,
550
+ "eval_samples_per_second": 91.587,
551
+ "eval_steps_per_second": 11.54,
552
+ "step": 4250
553
+ },
554
+ {
555
+ "epoch": 17.0,
556
+ "eval_exact_match": 13.8,
557
+ "eval_f1": 20.517505228159344,
558
+ "step": 4250
559
+ },
560
+ {
561
+ "epoch": 17.2,
562
+ "grad_norm": 2.7771668434143066,
563
+ "learning_rate": 5e-05,
564
+ "loss": 0.505,
565
+ "step": 4300
566
+ },
567
+ {
568
+ "epoch": 17.6,
569
+ "grad_norm": 3.6721012592315674,
570
+ "learning_rate": 5e-05,
571
+ "loss": 0.496,
572
+ "step": 4400
573
+ },
574
+ {
575
+ "epoch": 18.0,
576
+ "grad_norm": 4.477765083312988,
577
+ "learning_rate": 5e-05,
578
+ "loss": 0.5109,
579
+ "step": 4500
580
+ },
581
+ {
582
+ "epoch": 18.0,
583
+ "eval_accuracy": 0.4871111111111111,
584
+ "eval_loss": 3.7672173976898193,
585
+ "eval_runtime": 5.1624,
586
+ "eval_samples_per_second": 96.855,
587
+ "eval_steps_per_second": 12.204,
588
+ "step": 4500
589
+ },
590
+ {
591
+ "epoch": 18.0,
592
+ "eval_exact_match": 13.2,
593
+ "eval_f1": 19.935008488243795,
594
+ "step": 4500
595
+ },
596
+ {
597
+ "epoch": 18.4,
598
+ "grad_norm": 2.88653302192688,
599
+ "learning_rate": 5e-05,
600
+ "loss": 0.4564,
601
+ "step": 4600
602
+ },
603
+ {
604
+ "epoch": 18.8,
605
+ "grad_norm": 2.8045480251312256,
606
+ "learning_rate": 5e-05,
607
+ "loss": 0.479,
608
+ "step": 4700
609
+ },
610
+ {
611
+ "epoch": 19.0,
612
+ "eval_accuracy": 0.4865714285714286,
613
+ "eval_loss": 3.847074270248413,
614
+ "eval_runtime": 5.4643,
615
+ "eval_samples_per_second": 91.503,
616
+ "eval_steps_per_second": 11.529,
617
+ "step": 4750
618
+ },
619
+ {
620
+ "epoch": 19.0,
621
+ "eval_exact_match": 14.2,
622
+ "eval_f1": 20.744217407979033,
623
+ "step": 4750
624
+ },
625
+ {
626
+ "epoch": 19.2,
627
+ "grad_norm": 2.717168092727661,
628
+ "learning_rate": 5e-05,
629
+ "loss": 0.4612,
630
+ "step": 4800
631
+ },
632
+ {
633
+ "epoch": 19.6,
634
+ "grad_norm": 3.0051259994506836,
635
+ "learning_rate": 5e-05,
636
+ "loss": 0.453,
637
+ "step": 4900
638
+ },
639
+ {
640
+ "epoch": 20.0,
641
+ "grad_norm": 2.753119468688965,
642
+ "learning_rate": 5e-05,
643
+ "loss": 0.4675,
644
+ "step": 5000
645
+ },
646
+ {
647
+ "epoch": 20.0,
648
+ "eval_accuracy": 0.4867936507936508,
649
+ "eval_loss": 3.914645195007324,
650
+ "eval_runtime": 5.9419,
651
+ "eval_samples_per_second": 84.148,
652
+ "eval_steps_per_second": 10.603,
653
+ "step": 5000
654
+ },
655
+ {
656
+ "epoch": 20.0,
657
+ "eval_exact_match": 13.2,
658
+ "eval_f1": 20.16965067612128,
659
+ "step": 5000
660
+ },
661
+ {
662
+ "epoch": 20.4,
663
+ "grad_norm": 2.736196517944336,
664
+ "learning_rate": 5e-05,
665
+ "loss": 0.425,
666
+ "step": 5100
667
+ },
668
+ {
669
+ "epoch": 20.8,
670
+ "grad_norm": 3.410940647125244,
671
+ "learning_rate": 5e-05,
672
+ "loss": 0.448,
673
+ "step": 5200
674
+ },
675
+ {
676
+ "epoch": 21.0,
677
+ "eval_accuracy": 0.485968253968254,
678
+ "eval_loss": 3.955155849456787,
679
+ "eval_runtime": 5.1881,
680
+ "eval_samples_per_second": 96.375,
681
+ "eval_steps_per_second": 12.143,
682
+ "step": 5250
683
+ },
684
+ {
685
+ "epoch": 21.0,
686
+ "eval_exact_match": 13.4,
687
+ "eval_f1": 19.219269782505087,
688
+ "step": 5250
689
+ },
690
+ {
691
+ "epoch": 21.2,
692
+ "grad_norm": 3.2622647285461426,
693
+ "learning_rate": 5e-05,
694
+ "loss": 0.4306,
695
+ "step": 5300
696
+ },
697
+ {
698
+ "epoch": 21.6,
699
+ "grad_norm": 2.9486541748046875,
700
+ "learning_rate": 5e-05,
701
+ "loss": 0.4291,
702
+ "step": 5400
703
+ },
704
+ {
705
+ "epoch": 22.0,
706
+ "grad_norm": 3.2296924591064453,
707
+ "learning_rate": 5e-05,
708
+ "loss": 0.4389,
709
+ "step": 5500
710
+ },
711
+ {
712
+ "epoch": 22.0,
713
+ "eval_accuracy": 0.4855238095238095,
714
+ "eval_loss": 3.967400312423706,
715
+ "eval_runtime": 5.4481,
716
+ "eval_samples_per_second": 91.776,
717
+ "eval_steps_per_second": 11.564,
718
+ "step": 5500
719
+ },
720
+ {
721
+ "epoch": 22.0,
722
+ "eval_exact_match": 14.6,
723
+ "eval_f1": 20.702058565191567,
724
+ "step": 5500
725
+ },
726
+ {
727
+ "epoch": 22.4,
728
+ "grad_norm": 2.210437059402466,
729
+ "learning_rate": 5e-05,
730
+ "loss": 0.4047,
731
+ "step": 5600
732
+ },
733
+ {
734
+ "epoch": 22.8,
735
+ "grad_norm": 2.7681708335876465,
736
+ "learning_rate": 5e-05,
737
+ "loss": 0.4227,
738
+ "step": 5700
739
+ },
740
+ {
741
+ "epoch": 23.0,
742
+ "eval_accuracy": 0.48682539682539683,
743
+ "eval_loss": 4.001584053039551,
744
+ "eval_runtime": 5.9051,
745
+ "eval_samples_per_second": 84.673,
746
+ "eval_steps_per_second": 10.669,
747
+ "step": 5750
748
+ },
749
+ {
750
+ "epoch": 23.0,
751
+ "eval_exact_match": 14.6,
752
+ "eval_f1": 22.022229437229445,
753
+ "step": 5750
754
+ },
755
+ {
756
+ "epoch": 23.2,
757
+ "grad_norm": 2.4847607612609863,
758
+ "learning_rate": 5e-05,
759
+ "loss": 0.4088,
760
+ "step": 5800
761
+ },
762
+ {
763
+ "epoch": 23.6,
764
+ "grad_norm": 2.022418260574341,
765
+ "learning_rate": 5e-05,
766
+ "loss": 0.4045,
767
+ "step": 5900
768
+ },
769
+ {
770
+ "epoch": 24.0,
771
+ "grad_norm": 1.8691329956054688,
772
+ "learning_rate": 5e-05,
773
+ "loss": 0.4234,
774
+ "step": 6000
775
+ },
776
+ {
777
+ "epoch": 24.0,
778
+ "eval_accuracy": 0.4853650793650794,
779
+ "eval_loss": 4.042004585266113,
780
+ "eval_runtime": 5.5095,
781
+ "eval_samples_per_second": 90.752,
782
+ "eval_steps_per_second": 11.435,
783
+ "step": 6000
784
+ },
785
+ {
786
+ "epoch": 24.0,
787
+ "eval_exact_match": 13.8,
788
+ "eval_f1": 21.028606654783133,
789
+ "step": 6000
790
+ },
791
+ {
792
+ "epoch": 24.4,
793
+ "grad_norm": 2.199568748474121,
794
+ "learning_rate": 5e-05,
795
+ "loss": 0.3915,
796
+ "step": 6100
797
+ },
798
+ {
799
+ "epoch": 24.8,
800
+ "grad_norm": 2.636310338973999,
801
+ "learning_rate": 5e-05,
802
+ "loss": 0.4051,
803
+ "step": 6200
804
+ },
805
+ {
806
+ "epoch": 25.0,
807
+ "eval_accuracy": 0.4865714285714286,
808
+ "eval_loss": 4.044601917266846,
809
+ "eval_runtime": 5.1578,
810
+ "eval_samples_per_second": 96.941,
811
+ "eval_steps_per_second": 12.215,
812
+ "step": 6250
813
+ },
814
+ {
815
+ "epoch": 25.0,
816
+ "eval_exact_match": 12.8,
817
+ "eval_f1": 20.273670088081865,
818
+ "step": 6250
819
+ },
820
+ {
821
+ "epoch": 25.2,
822
+ "grad_norm": 2.0600600242614746,
823
+ "learning_rate": 5e-05,
824
+ "loss": 0.3952,
825
+ "step": 6300
826
+ },
827
+ {
828
+ "epoch": 25.6,
829
+ "grad_norm": 2.2660067081451416,
830
+ "learning_rate": 5e-05,
831
+ "loss": 0.3939,
832
+ "step": 6400
833
+ },
834
+ {
835
+ "epoch": 26.0,
836
+ "grad_norm": 2.225553512573242,
837
+ "learning_rate": 5e-05,
838
+ "loss": 0.4043,
839
+ "step": 6500
840
+ },
841
+ {
842
+ "epoch": 26.0,
843
+ "eval_accuracy": 0.4855238095238095,
844
+ "eval_loss": 4.121946334838867,
845
+ "eval_runtime": 6.1465,
846
+ "eval_samples_per_second": 81.347,
847
+ "eval_steps_per_second": 10.25,
848
+ "step": 6500
849
+ },
850
+ {
851
+ "epoch": 26.0,
852
+ "eval_exact_match": 12.8,
853
+ "eval_f1": 18.98795730413378,
854
+ "step": 6500
855
+ }
856
+ ],
857
+ "logging_steps": 100,
858
+ "max_steps": 12500,
859
+ "num_input_tokens_seen": 0,
860
+ "num_train_epochs": 50,
861
+ "save_steps": 500,
862
+ "stateful_callbacks": {
863
+ "TrainerControl": {
864
+ "args": {
865
+ "should_epoch_stop": false,
866
+ "should_evaluate": false,
867
+ "should_log": false,
868
+ "should_save": true,
869
+ "should_training_stop": false
870
+ },
871
+ "attributes": {}
872
+ }
873
+ },
874
+ "total_flos": 1.6816001702166528e+17,
875
+ "train_batch_size": 1,
876
+ "trial_name": null,
877
+ "trial_params": null
878
+ }
checkpoint-6500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ede3565a5658e9b8743d6efdd527fd23709623a130469c81f1191d07c33fbcb1
3
+ size 5304
checkpoint-6500/vocab.json ADDED
The diff for this file is too large to render. See raw diff