Hugging Face's logo

tyzhu
/
lmind_nq_train6000_eval6489_v1_qa_3e-5_lora2

Safetensors
Generated from Trainer
Eval Results
Model card Files Community
tyzhu commited on
Commit
7cd0fa5
•
1 Parent(s): db6d0b9

Training in progress, epoch 40, checkpoint

Browse files
Files changed (17) hide show
  1. checkpoint-7500/README.md +9 -0
  2. checkpoint-7500/adapter_config.json +25 -0
  3. checkpoint-7500/adapter_model.safetensors +3 -0
  4. checkpoint-7500/added_tokens.json +5 -0
  5. checkpoint-7500/merges.txt +0 -0
  6. checkpoint-7500/optimizer.pt +3 -0
  7. checkpoint-7500/rng_state_0.pth +3 -0
  8. checkpoint-7500/rng_state_1.pth +3 -0
  9. checkpoint-7500/rng_state_2.pth +3 -0
  10. checkpoint-7500/rng_state_3.pth +3 -0
  11. checkpoint-7500/scheduler.pt +3 -0
  12. checkpoint-7500/special_tokens_map.json +14 -0
  13. checkpoint-7500/tokenizer.json +0 -0
  14. checkpoint-7500/tokenizer_config.json +43 -0
  15. checkpoint-7500/trainer_state.json +1158 -0
  16. checkpoint-7500/training_args.bin +3 -0
  17. checkpoint-7500/vocab.json +0 -0
checkpoint-7500/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-7500/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-7500/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d9ddce2d502160e2333876165eb0bc2de9255fe6ab831af5bd21cc2686308ce2
3
+ size 112130416
checkpoint-7500/added_tokens.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "<|endoftext|>": 151643,
3
+ "<|im_end|>": 151645,
4
+ "<|im_start|>": 151644
5
+ }
checkpoint-7500/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-7500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dd2d4dc0c76468ea76c1af7dbc72d2080b9363ce8b9b27dccbe170c0e201f3a2
3
+ size 224537202
checkpoint-7500/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:df492db3a2199c2a312444e637f1f393fe794d6244762b8f57a92a82c62c6534
3
+ size 15024
checkpoint-7500/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2f2f0f188f93998871498a99ce6ee3e54cf0aa485be22f9d9bf26485e35ee0bd
3
+ size 15024
checkpoint-7500/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:efb27262959a6f3c649d67c16f0f1b4d83475adb03bd2fe3b84ab676e97652bc
3
+ size 15024
checkpoint-7500/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ed7035abe0083be2e38ad1b021a5184ddc2d7a8d998f4095214b4cd828962602
3
+ size 15024
checkpoint-7500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:34540704d450cf32444d5eea1ca425c2adaac252ee0683cfa63ad54015c1de5b
3
+ size 1064
checkpoint-7500/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-7500/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-7500/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-7500/trainer_state.json ADDED
@@ -0,0 +1,1158 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 40.0,
5
+ "eval_steps": 500,
6
+ "global_step": 7500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.5333333333333333,
13
+ "grad_norm": 0.9283596873283386,
14
+ "learning_rate": 3e-05,
15
+ "loss": 1.963,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.9973333333333333,
20
+ "eval_accuracy": 0.5695384615384615,
21
+ "eval_loss": 1.6438876390457153,
22
+ "eval_runtime": 5.0534,
23
+ "eval_samples_per_second": 98.943,
24
+ "eval_steps_per_second": 12.467,
25
+ "step": 187
26
+ },
27
+ {
28
+ "epoch": 0.9973333333333333,
29
+ "eval_exact_match": 16.2,
30
+ "eval_f1": 25.888571428571442,
31
+ "step": 187
32
+ },
33
+ {
34
+ "epoch": 1.0666666666666667,
35
+ "grad_norm": 1.0407154560089111,
36
+ "learning_rate": 3e-05,
37
+ "loss": 1.6668,
38
+ "step": 200
39
+ },
40
+ {
41
+ "epoch": 1.6,
42
+ "grad_norm": 1.1281720399856567,
43
+ "learning_rate": 3e-05,
44
+ "loss": 1.6099,
45
+ "step": 300
46
+ },
47
+ {
48
+ "epoch": 2.0,
49
+ "eval_accuracy": 0.5733333333333334,
50
+ "eval_loss": 1.618270754814148,
51
+ "eval_runtime": 5.1983,
52
+ "eval_samples_per_second": 96.186,
53
+ "eval_steps_per_second": 12.119,
54
+ "step": 375
55
+ },
56
+ {
57
+ "epoch": 2.0,
58
+ "eval_exact_match": 16.6,
59
+ "eval_f1": 26.287619047619057,
60
+ "step": 375
61
+ },
62
+ {
63
+ "epoch": 2.1333333333333333,
64
+ "grad_norm": 1.2508225440979004,
65
+ "learning_rate": 3e-05,
66
+ "loss": 1.5866,
67
+ "step": 400
68
+ },
69
+ {
70
+ "epoch": 2.6666666666666665,
71
+ "grad_norm": 1.3197834491729736,
72
+ "learning_rate": 3e-05,
73
+ "loss": 1.524,
74
+ "step": 500
75
+ },
76
+ {
77
+ "epoch": 2.997333333333333,
78
+ "eval_accuracy": 0.5743589743589743,
79
+ "eval_loss": 1.6164497137069702,
80
+ "eval_runtime": 5.0495,
81
+ "eval_samples_per_second": 99.019,
82
+ "eval_steps_per_second": 12.476,
83
+ "step": 562
84
+ },
85
+ {
86
+ "epoch": 2.997333333333333,
87
+ "eval_exact_match": 17.8,
88
+ "eval_f1": 27.35428571428571,
89
+ "step": 562
90
+ },
91
+ {
92
+ "epoch": 3.2,
93
+ "grad_norm": 1.6320394277572632,
94
+ "learning_rate": 3e-05,
95
+ "loss": 1.4752,
96
+ "step": 600
97
+ },
98
+ {
99
+ "epoch": 3.7333333333333334,
100
+ "grad_norm": 1.7101725339889526,
101
+ "learning_rate": 3e-05,
102
+ "loss": 1.3938,
103
+ "step": 700
104
+ },
105
+ {
106
+ "epoch": 4.0,
107
+ "eval_accuracy": 0.5728717948717948,
108
+ "eval_loss": 1.6376241445541382,
109
+ "eval_runtime": 4.9662,
110
+ "eval_samples_per_second": 100.68,
111
+ "eval_steps_per_second": 12.686,
112
+ "step": 750
113
+ },
114
+ {
115
+ "epoch": 4.0,
116
+ "eval_exact_match": 18.0,
117
+ "eval_f1": 27.498571428571427,
118
+ "step": 750
119
+ },
120
+ {
121
+ "epoch": 4.266666666666667,
122
+ "grad_norm": 2.1727664470672607,
123
+ "learning_rate": 3e-05,
124
+ "loss": 1.342,
125
+ "step": 800
126
+ },
127
+ {
128
+ "epoch": 4.8,
129
+ "grad_norm": 2.341384172439575,
130
+ "learning_rate": 3e-05,
131
+ "loss": 1.2685,
132
+ "step": 900
133
+ },
134
+ {
135
+ "epoch": 4.997333333333334,
136
+ "eval_accuracy": 0.5699487179487179,
137
+ "eval_loss": 1.684618353843689,
138
+ "eval_runtime": 5.5457,
139
+ "eval_samples_per_second": 90.16,
140
+ "eval_steps_per_second": 11.36,
141
+ "step": 937
142
+ },
143
+ {
144
+ "epoch": 4.997333333333334,
145
+ "eval_exact_match": 19.2,
146
+ "eval_f1": 28.190952380952375,
147
+ "step": 937
148
+ },
149
+ {
150
+ "epoch": 5.333333333333333,
151
+ "grad_norm": 2.5624172687530518,
152
+ "learning_rate": 3e-05,
153
+ "loss": 1.1847,
154
+ "step": 1000
155
+ },
156
+ {
157
+ "epoch": 5.866666666666667,
158
+ "grad_norm": 2.8159019947052,
159
+ "learning_rate": 3e-05,
160
+ "loss": 1.1591,
161
+ "step": 1100
162
+ },
163
+ {
164
+ "epoch": 6.0,
165
+ "eval_accuracy": 0.5672820512820513,
166
+ "eval_loss": 1.7546838521957397,
167
+ "eval_runtime": 5.0186,
168
+ "eval_samples_per_second": 99.629,
169
+ "eval_steps_per_second": 12.553,
170
+ "step": 1125
171
+ },
172
+ {
173
+ "epoch": 6.0,
174
+ "eval_exact_match": 16.6,
175
+ "eval_f1": 25.054761904761914,
176
+ "step": 1125
177
+ },
178
+ {
179
+ "epoch": 6.4,
180
+ "grad_norm": 3.056166410446167,
181
+ "learning_rate": 3e-05,
182
+ "loss": 1.0486,
183
+ "step": 1200
184
+ },
185
+ {
186
+ "epoch": 6.933333333333334,
187
+ "grad_norm": 2.873281240463257,
188
+ "learning_rate": 3e-05,
189
+ "loss": 1.0444,
190
+ "step": 1300
191
+ },
192
+ {
193
+ "epoch": 6.997333333333334,
194
+ "eval_accuracy": 0.5643076923076923,
195
+ "eval_loss": 1.8394943475723267,
196
+ "eval_runtime": 5.5931,
197
+ "eval_samples_per_second": 89.396,
198
+ "eval_steps_per_second": 11.264,
199
+ "step": 1312
200
+ },
201
+ {
202
+ "epoch": 6.997333333333334,
203
+ "eval_exact_match": 15.8,
204
+ "eval_f1": 23.925714285714296,
205
+ "step": 1312
206
+ },
207
+ {
208
+ "epoch": 7.466666666666667,
209
+ "grad_norm": 3.069322109222412,
210
+ "learning_rate": 3e-05,
211
+ "loss": 0.9374,
212
+ "step": 1400
213
+ },
214
+ {
215
+ "epoch": 8.0,
216
+ "grad_norm": 3.2732129096984863,
217
+ "learning_rate": 3e-05,
218
+ "loss": 0.9535,
219
+ "step": 1500
220
+ },
221
+ {
222
+ "epoch": 8.0,
223
+ "eval_accuracy": 0.5612820512820513,
224
+ "eval_loss": 1.9008216857910156,
225
+ "eval_runtime": 5.1357,
226
+ "eval_samples_per_second": 97.358,
227
+ "eval_steps_per_second": 12.267,
228
+ "step": 1500
229
+ },
230
+ {
231
+ "epoch": 8.0,
232
+ "eval_exact_match": 15.2,
233
+ "eval_f1": 23.12063492063493,
234
+ "step": 1500
235
+ },
236
+ {
237
+ "epoch": 8.533333333333333,
238
+ "grad_norm": 3.382610321044922,
239
+ "learning_rate": 3e-05,
240
+ "loss": 0.8235,
241
+ "step": 1600
242
+ },
243
+ {
244
+ "epoch": 8.997333333333334,
245
+ "eval_accuracy": 0.5592307692307692,
246
+ "eval_loss": 2.026776075363159,
247
+ "eval_runtime": 5.5056,
248
+ "eval_samples_per_second": 90.817,
249
+ "eval_steps_per_second": 11.443,
250
+ "step": 1687
251
+ },
252
+ {
253
+ "epoch": 8.997333333333334,
254
+ "eval_exact_match": 13.8,
255
+ "eval_f1": 21.42285714285715,
256
+ "step": 1687
257
+ },
258
+ {
259
+ "epoch": 9.066666666666666,
260
+ "grad_norm": 4.363819122314453,
261
+ "learning_rate": 3e-05,
262
+ "loss": 0.8547,
263
+ "step": 1700
264
+ },
265
+ {
266
+ "epoch": 9.6,
267
+ "grad_norm": 3.685805320739746,
268
+ "learning_rate": 3e-05,
269
+ "loss": 0.7635,
270
+ "step": 1800
271
+ },
272
+ {
273
+ "epoch": 10.0,
274
+ "eval_accuracy": 0.5567692307692308,
275
+ "eval_loss": 2.0937435626983643,
276
+ "eval_runtime": 5.235,
277
+ "eval_samples_per_second": 95.512,
278
+ "eval_steps_per_second": 12.034,
279
+ "step": 1875
280
+ },
281
+ {
282
+ "epoch": 10.0,
283
+ "eval_exact_match": 15.2,
284
+ "eval_f1": 22.031904761904762,
285
+ "step": 1875
286
+ },
287
+ {
288
+ "epoch": 10.133333333333333,
289
+ "grad_norm": 3.5318357944488525,
290
+ "learning_rate": 3e-05,
291
+ "loss": 0.7554,
292
+ "step": 1900
293
+ },
294
+ {
295
+ "epoch": 10.666666666666666,
296
+ "grad_norm": 3.6454901695251465,
297
+ "learning_rate": 3e-05,
298
+ "loss": 0.6978,
299
+ "step": 2000
300
+ },
301
+ {
302
+ "epoch": 10.997333333333334,
303
+ "eval_accuracy": 0.556974358974359,
304
+ "eval_loss": 2.1557741165161133,
305
+ "eval_runtime": 5.1556,
306
+ "eval_samples_per_second": 96.981,
307
+ "eval_steps_per_second": 12.22,
308
+ "step": 2062
309
+ },
310
+ {
311
+ "epoch": 10.997333333333334,
312
+ "eval_exact_match": 15.0,
313
+ "eval_f1": 23.010000000000012,
314
+ "step": 2062
315
+ },
316
+ {
317
+ "epoch": 11.2,
318
+ "grad_norm": 3.658707618713379,
319
+ "learning_rate": 3e-05,
320
+ "loss": 0.6969,
321
+ "step": 2100
322
+ },
323
+ {
324
+ "epoch": 11.733333333333333,
325
+ "grad_norm": 3.7350716590881348,
326
+ "learning_rate": 3e-05,
327
+ "loss": 0.6615,
328
+ "step": 2200
329
+ },
330
+ {
331
+ "epoch": 12.0,
332
+ "eval_accuracy": 0.5551794871794872,
333
+ "eval_loss": 2.239973545074463,
334
+ "eval_runtime": 4.9026,
335
+ "eval_samples_per_second": 101.986,
336
+ "eval_steps_per_second": 12.85,
337
+ "step": 2250
338
+ },
339
+ {
340
+ "epoch": 12.0,
341
+ "eval_exact_match": 15.6,
342
+ "eval_f1": 23.333492063492066,
343
+ "step": 2250
344
+ },
345
+ {
346
+ "epoch": 12.266666666666667,
347
+ "grad_norm": 4.4434709548950195,
348
+ "learning_rate": 3e-05,
349
+ "loss": 0.632,
350
+ "step": 2300
351
+ },
352
+ {
353
+ "epoch": 12.8,
354
+ "grad_norm": 3.4198718070983887,
355
+ "learning_rate": 3e-05,
356
+ "loss": 0.6262,
357
+ "step": 2400
358
+ },
359
+ {
360
+ "epoch": 12.997333333333334,
361
+ "eval_accuracy": 0.5556410256410257,
362
+ "eval_loss": 2.268662452697754,
363
+ "eval_runtime": 5.327,
364
+ "eval_samples_per_second": 93.862,
365
+ "eval_steps_per_second": 11.827,
366
+ "step": 2437
367
+ },
368
+ {
369
+ "epoch": 12.997333333333334,
370
+ "eval_exact_match": 14.4,
371
+ "eval_f1": 22.583492063492077,
372
+ "step": 2437
373
+ },
374
+ {
375
+ "epoch": 13.333333333333334,
376
+ "grad_norm": 3.158496618270874,
377
+ "learning_rate": 3e-05,
378
+ "loss": 0.5973,
379
+ "step": 2500
380
+ },
381
+ {
382
+ "epoch": 13.866666666666667,
383
+ "grad_norm": 3.5720713138580322,
384
+ "learning_rate": 3e-05,
385
+ "loss": 0.5958,
386
+ "step": 2600
387
+ },
388
+ {
389
+ "epoch": 14.0,
390
+ "eval_accuracy": 0.5536923076923077,
391
+ "eval_loss": 2.358220100402832,
392
+ "eval_runtime": 5.236,
393
+ "eval_samples_per_second": 95.493,
394
+ "eval_steps_per_second": 12.032,
395
+ "step": 2625
396
+ },
397
+ {
398
+ "epoch": 14.0,
399
+ "eval_exact_match": 14.6,
400
+ "eval_f1": 21.79904761904763,
401
+ "step": 2625
402
+ },
403
+ {
404
+ "epoch": 14.4,
405
+ "grad_norm": 5.126227855682373,
406
+ "learning_rate": 3e-05,
407
+ "loss": 0.5692,
408
+ "step": 2700
409
+ },
410
+ {
411
+ "epoch": 14.933333333333334,
412
+ "grad_norm": 3.3391811847686768,
413
+ "learning_rate": 3e-05,
414
+ "loss": 0.5778,
415
+ "step": 2800
416
+ },
417
+ {
418
+ "epoch": 14.997333333333334,
419
+ "eval_accuracy": 0.5533846153846154,
420
+ "eval_loss": 2.395972728729248,
421
+ "eval_runtime": 5.1135,
422
+ "eval_samples_per_second": 97.78,
423
+ "eval_steps_per_second": 12.32,
424
+ "step": 2812
425
+ },
426
+ {
427
+ "epoch": 14.997333333333334,
428
+ "eval_exact_match": 14.6,
429
+ "eval_f1": 22.537301587301595,
430
+ "step": 2812
431
+ },
432
+ {
433
+ "epoch": 15.466666666666667,
434
+ "grad_norm": 2.8773677349090576,
435
+ "learning_rate": 3e-05,
436
+ "loss": 0.5461,
437
+ "step": 2900
438
+ },
439
+ {
440
+ "epoch": 16.0,
441
+ "grad_norm": 3.580195903778076,
442
+ "learning_rate": 3e-05,
443
+ "loss": 0.5661,
444
+ "step": 3000
445
+ },
446
+ {
447
+ "epoch": 16.0,
448
+ "eval_accuracy": 0.5533846153846154,
449
+ "eval_loss": 2.4321775436401367,
450
+ "eval_runtime": 5.0238,
451
+ "eval_samples_per_second": 99.527,
452
+ "eval_steps_per_second": 12.54,
453
+ "step": 3000
454
+ },
455
+ {
456
+ "epoch": 16.0,
457
+ "eval_exact_match": 14.0,
458
+ "eval_f1": 22.300808080808093,
459
+ "step": 3000
460
+ },
461
+ {
462
+ "epoch": 16.533333333333335,
463
+ "grad_norm": 3.6829302310943604,
464
+ "learning_rate": 3e-05,
465
+ "loss": 0.5277,
466
+ "step": 3100
467
+ },
468
+ {
469
+ "epoch": 16.997333333333334,
470
+ "eval_accuracy": 0.5515384615384615,
471
+ "eval_loss": 2.482792615890503,
472
+ "eval_runtime": 5.3006,
473
+ "eval_samples_per_second": 94.328,
474
+ "eval_steps_per_second": 11.885,
475
+ "step": 3187
476
+ },
477
+ {
478
+ "epoch": 16.997333333333334,
479
+ "eval_exact_match": 13.4,
480
+ "eval_f1": 21.439682539682543,
481
+ "step": 3187
482
+ },
483
+ {
484
+ "epoch": 17.066666666666666,
485
+ "grad_norm": 3.1766672134399414,
486
+ "learning_rate": 3e-05,
487
+ "loss": 0.5475,
488
+ "step": 3200
489
+ },
490
+ {
491
+ "epoch": 17.6,
492
+ "grad_norm": 3.070977210998535,
493
+ "learning_rate": 3e-05,
494
+ "loss": 0.5211,
495
+ "step": 3300
496
+ },
497
+ {
498
+ "epoch": 18.0,
499
+ "eval_accuracy": 0.5516410256410257,
500
+ "eval_loss": 2.510619640350342,
501
+ "eval_runtime": 5.0519,
502
+ "eval_samples_per_second": 98.973,
503
+ "eval_steps_per_second": 12.471,
504
+ "step": 3375
505
+ },
506
+ {
507
+ "epoch": 18.0,
508
+ "eval_exact_match": 12.6,
509
+ "eval_f1": 20.136349206349216,
510
+ "step": 3375
511
+ },
512
+ {
513
+ "epoch": 18.133333333333333,
514
+ "grad_norm": 4.728194713592529,
515
+ "learning_rate": 3e-05,
516
+ "loss": 0.5344,
517
+ "step": 3400
518
+ },
519
+ {
520
+ "epoch": 18.666666666666668,
521
+ "grad_norm": 3.1506965160369873,
522
+ "learning_rate": 3e-05,
523
+ "loss": 0.5189,
524
+ "step": 3500
525
+ },
526
+ {
527
+ "epoch": 18.997333333333334,
528
+ "eval_accuracy": 0.5514871794871795,
529
+ "eval_loss": 2.5705809593200684,
530
+ "eval_runtime": 4.9081,
531
+ "eval_samples_per_second": 101.872,
532
+ "eval_steps_per_second": 12.836,
533
+ "step": 3562
534
+ },
535
+ {
536
+ "epoch": 18.997333333333334,
537
+ "eval_exact_match": 13.8,
538
+ "eval_f1": 21.51634920634921,
539
+ "step": 3562
540
+ },
541
+ {
542
+ "epoch": 19.2,
543
+ "grad_norm": 2.7202513217926025,
544
+ "learning_rate": 3e-05,
545
+ "loss": 0.5198,
546
+ "step": 3600
547
+ },
548
+ {
549
+ "epoch": 19.733333333333334,
550
+ "grad_norm": 2.581169366836548,
551
+ "learning_rate": 3e-05,
552
+ "loss": 0.5166,
553
+ "step": 3700
554
+ },
555
+ {
556
+ "epoch": 20.0,
557
+ "eval_accuracy": 0.5526153846153846,
558
+ "eval_loss": 2.542161226272583,
559
+ "eval_runtime": 5.0196,
560
+ "eval_samples_per_second": 99.609,
561
+ "eval_steps_per_second": 12.551,
562
+ "step": 3750
563
+ },
564
+ {
565
+ "epoch": 20.0,
566
+ "eval_exact_match": 14.4,
567
+ "eval_f1": 21.563492063492074,
568
+ "step": 3750
569
+ },
570
+ {
571
+ "epoch": 20.266666666666666,
572
+ "grad_norm": 2.6741623878479004,
573
+ "learning_rate": 3e-05,
574
+ "loss": 0.5112,
575
+ "step": 3800
576
+ },
577
+ {
578
+ "epoch": 20.8,
579
+ "grad_norm": 3.241607666015625,
580
+ "learning_rate": 3e-05,
581
+ "loss": 0.5132,
582
+ "step": 3900
583
+ },
584
+ {
585
+ "epoch": 20.997333333333334,
586
+ "eval_accuracy": 0.550923076923077,
587
+ "eval_loss": 2.594764232635498,
588
+ "eval_runtime": 5.2553,
589
+ "eval_samples_per_second": 95.142,
590
+ "eval_steps_per_second": 11.988,
591
+ "step": 3937
592
+ },
593
+ {
594
+ "epoch": 20.997333333333334,
595
+ "eval_exact_match": 13.4,
596
+ "eval_f1": 21.25015873015874,
597
+ "step": 3937
598
+ },
599
+ {
600
+ "epoch": 21.333333333333332,
601
+ "grad_norm": 2.6324737071990967,
602
+ "learning_rate": 3e-05,
603
+ "loss": 0.5011,
604
+ "step": 4000
605
+ },
606
+ {
607
+ "epoch": 21.866666666666667,
608
+ "grad_norm": 3.572909116744995,
609
+ "learning_rate": 3e-05,
610
+ "loss": 0.5115,
611
+ "step": 4100
612
+ },
613
+ {
614
+ "epoch": 22.0,
615
+ "eval_accuracy": 0.5512307692307692,
616
+ "eval_loss": 2.604835033416748,
617
+ "eval_runtime": 5.1325,
618
+ "eval_samples_per_second": 97.419,
619
+ "eval_steps_per_second": 12.275,
620
+ "step": 4125
621
+ },
622
+ {
623
+ "epoch": 22.0,
624
+ "eval_exact_match": 14.2,
625
+ "eval_f1": 21.673968253968265,
626
+ "step": 4125
627
+ },
628
+ {
629
+ "epoch": 22.4,
630
+ "grad_norm": 2.303550958633423,
631
+ "learning_rate": 3e-05,
632
+ "loss": 0.4973,
633
+ "step": 4200
634
+ },
635
+ {
636
+ "epoch": 22.933333333333334,
637
+ "grad_norm": 2.5230953693389893,
638
+ "learning_rate": 3e-05,
639
+ "loss": 0.5083,
640
+ "step": 4300
641
+ },
642
+ {
643
+ "epoch": 22.997333333333334,
644
+ "eval_accuracy": 0.5520512820512821,
645
+ "eval_loss": 2.5811188220977783,
646
+ "eval_runtime": 5.2251,
647
+ "eval_samples_per_second": 95.692,
648
+ "eval_steps_per_second": 12.057,
649
+ "step": 4312
650
+ },
651
+ {
652
+ "epoch": 22.997333333333334,
653
+ "eval_exact_match": 14.6,
654
+ "eval_f1": 22.511111111111124,
655
+ "step": 4312
656
+ },
657
+ {
658
+ "epoch": 23.466666666666665,
659
+ "grad_norm": 2.1872711181640625,
660
+ "learning_rate": 3e-05,
661
+ "loss": 0.49,
662
+ "step": 4400
663
+ },
664
+ {
665
+ "epoch": 24.0,
666
+ "grad_norm": 2.1276657581329346,
667
+ "learning_rate": 3e-05,
668
+ "loss": 0.5081,
669
+ "step": 4500
670
+ },
671
+ {
672
+ "epoch": 24.0,
673
+ "eval_accuracy": 0.5513333333333333,
674
+ "eval_loss": 2.5661673545837402,
675
+ "eval_runtime": 5.1065,
676
+ "eval_samples_per_second": 97.915,
677
+ "eval_steps_per_second": 12.337,
678
+ "step": 4500
679
+ },
680
+ {
681
+ "epoch": 24.0,
682
+ "eval_exact_match": 15.2,
683
+ "eval_f1": 23.028547008547015,
684
+ "step": 4500
685
+ },
686
+ {
687
+ "epoch": 24.533333333333335,
688
+ "grad_norm": 3.943235158920288,
689
+ "learning_rate": 3e-05,
690
+ "loss": 0.4862,
691
+ "step": 4600
692
+ },
693
+ {
694
+ "epoch": 24.997333333333334,
695
+ "eval_accuracy": 0.5522051282051282,
696
+ "eval_loss": 2.6429293155670166,
697
+ "eval_runtime": 4.8982,
698
+ "eval_samples_per_second": 102.078,
699
+ "eval_steps_per_second": 12.862,
700
+ "step": 4687
701
+ },
702
+ {
703
+ "epoch": 24.997333333333334,
704
+ "eval_exact_match": 13.8,
705
+ "eval_f1": 21.402380952380966,
706
+ "step": 4687
707
+ },
708
+ {
709
+ "epoch": 25.066666666666666,
710
+ "grad_norm": 4.37394380569458,
711
+ "learning_rate": 3e-05,
712
+ "loss": 0.4998,
713
+ "step": 4700
714
+ },
715
+ {
716
+ "epoch": 25.6,
717
+ "grad_norm": 2.133417844772339,
718
+ "learning_rate": 3e-05,
719
+ "loss": 0.4845,
720
+ "step": 4800
721
+ },
722
+ {
723
+ "epoch": 26.0,
724
+ "eval_accuracy": 0.5533846153846154,
725
+ "eval_loss": 2.6020326614379883,
726
+ "eval_runtime": 5.0467,
727
+ "eval_samples_per_second": 99.075,
728
+ "eval_steps_per_second": 12.483,
729
+ "step": 4875
730
+ },
731
+ {
732
+ "epoch": 26.0,
733
+ "eval_exact_match": 13.8,
734
+ "eval_f1": 22.005714285714294,
735
+ "step": 4875
736
+ },
737
+ {
738
+ "epoch": 26.133333333333333,
739
+ "grad_norm": 2.307835102081299,
740
+ "learning_rate": 3e-05,
741
+ "loss": 0.4967,
742
+ "step": 4900
743
+ },
744
+ {
745
+ "epoch": 26.666666666666668,
746
+ "grad_norm": 2.352783203125,
747
+ "learning_rate": 3e-05,
748
+ "loss": 0.4869,
749
+ "step": 5000
750
+ },
751
+ {
752
+ "epoch": 26.997333333333334,
753
+ "eval_accuracy": 0.5521538461538461,
754
+ "eval_loss": 2.633915424346924,
755
+ "eval_runtime": 5.2178,
756
+ "eval_samples_per_second": 95.826,
757
+ "eval_steps_per_second": 12.074,
758
+ "step": 5062
759
+ },
760
+ {
761
+ "epoch": 26.997333333333334,
762
+ "eval_exact_match": 13.6,
763
+ "eval_f1": 21.69952380952382,
764
+ "step": 5062
765
+ },
766
+ {
767
+ "epoch": 27.2,
768
+ "grad_norm": 3.5862042903900146,
769
+ "learning_rate": 3e-05,
770
+ "loss": 0.4863,
771
+ "step": 5100
772
+ },
773
+ {
774
+ "epoch": 27.733333333333334,
775
+ "grad_norm": 2.1928396224975586,
776
+ "learning_rate": 3e-05,
777
+ "loss": 0.4862,
778
+ "step": 5200
779
+ },
780
+ {
781
+ "epoch": 28.0,
782
+ "eval_accuracy": 0.5524102564102564,
783
+ "eval_loss": 2.6161999702453613,
784
+ "eval_runtime": 5.0462,
785
+ "eval_samples_per_second": 99.085,
786
+ "eval_steps_per_second": 12.485,
787
+ "step": 5250
788
+ },
789
+ {
790
+ "epoch": 28.0,
791
+ "eval_exact_match": 14.8,
792
+ "eval_f1": 22.126190476190487,
793
+ "step": 5250
794
+ },
795
+ {
796
+ "epoch": 28.266666666666666,
797
+ "grad_norm": 2.0402398109436035,
798
+ "learning_rate": 3e-05,
799
+ "loss": 0.4836,
800
+ "step": 5300
801
+ },
802
+ {
803
+ "epoch": 28.8,
804
+ "grad_norm": 2.1713831424713135,
805
+ "learning_rate": 3e-05,
806
+ "loss": 0.4856,
807
+ "step": 5400
808
+ },
809
+ {
810
+ "epoch": 28.997333333333334,
811
+ "eval_accuracy": 0.5525641025641026,
812
+ "eval_loss": 2.6764330863952637,
813
+ "eval_runtime": 5.9146,
814
+ "eval_samples_per_second": 84.537,
815
+ "eval_steps_per_second": 10.652,
816
+ "step": 5437
817
+ },
818
+ {
819
+ "epoch": 28.997333333333334,
820
+ "eval_exact_match": 14.2,
821
+ "eval_f1": 21.448412698412714,
822
+ "step": 5437
823
+ },
824
+ {
825
+ "epoch": 29.333333333333332,
826
+ "grad_norm": 2.2155110836029053,
827
+ "learning_rate": 3e-05,
828
+ "loss": 0.4786,
829
+ "step": 5500
830
+ },
831
+ {
832
+ "epoch": 29.866666666666667,
833
+ "grad_norm": 1.7778940200805664,
834
+ "learning_rate": 3e-05,
835
+ "loss": 0.4871,
836
+ "step": 5600
837
+ },
838
+ {
839
+ "epoch": 30.0,
840
+ "eval_accuracy": 0.5525641025641026,
841
+ "eval_loss": 2.6703062057495117,
842
+ "eval_runtime": 4.8938,
843
+ "eval_samples_per_second": 102.171,
844
+ "eval_steps_per_second": 12.874,
845
+ "step": 5625
846
+ },
847
+ {
848
+ "epoch": 30.0,
849
+ "eval_exact_match": 14.2,
850
+ "eval_f1": 20.916825396825406,
851
+ "step": 5625
852
+ },
853
+ {
854
+ "epoch": 30.4,
855
+ "grad_norm": 2.07055401802063,
856
+ "learning_rate": 3e-05,
857
+ "loss": 0.475,
858
+ "step": 5700
859
+ },
860
+ {
861
+ "epoch": 30.933333333333334,
862
+ "grad_norm": 1.788883924484253,
863
+ "learning_rate": 3e-05,
864
+ "loss": 0.4863,
865
+ "step": 5800
866
+ },
867
+ {
868
+ "epoch": 30.997333333333334,
869
+ "eval_accuracy": 0.5533333333333333,
870
+ "eval_loss": 2.6787095069885254,
871
+ "eval_runtime": 5.0298,
872
+ "eval_samples_per_second": 99.407,
873
+ "eval_steps_per_second": 12.525,
874
+ "step": 5812
875
+ },
876
+ {
877
+ "epoch": 30.997333333333334,
878
+ "eval_exact_match": 14.0,
879
+ "eval_f1": 21.74063492063493,
880
+ "step": 5812
881
+ },
882
+ {
883
+ "epoch": 31.466666666666665,
884
+ "grad_norm": 1.6789844036102295,
885
+ "learning_rate": 3e-05,
886
+ "loss": 0.4693,
887
+ "step": 5900
888
+ },
889
+ {
890
+ "epoch": 32.0,
891
+ "grad_norm": 1.622942566871643,
892
+ "learning_rate": 3e-05,
893
+ "loss": 0.4884,
894
+ "step": 6000
895
+ },
896
+ {
897
+ "epoch": 32.0,
898
+ "eval_accuracy": 0.5528205128205128,
899
+ "eval_loss": 2.6847567558288574,
900
+ "eval_runtime": 5.2277,
901
+ "eval_samples_per_second": 95.644,
902
+ "eval_steps_per_second": 12.051,
903
+ "step": 6000
904
+ },
905
+ {
906
+ "epoch": 32.0,
907
+ "eval_exact_match": 14.2,
908
+ "eval_f1": 21.513968253968265,
909
+ "step": 6000
910
+ },
911
+ {
912
+ "epoch": 32.53333333333333,
913
+ "grad_norm": 1.7828670740127563,
914
+ "learning_rate": 3e-05,
915
+ "loss": 0.467,
916
+ "step": 6100
917
+ },
918
+ {
919
+ "epoch": 32.99733333333333,
920
+ "eval_accuracy": 0.5531282051282052,
921
+ "eval_loss": 2.668905735015869,
922
+ "eval_runtime": 5.7429,
923
+ "eval_samples_per_second": 87.064,
924
+ "eval_steps_per_second": 10.97,
925
+ "step": 6187
926
+ },
927
+ {
928
+ "epoch": 32.99733333333333,
929
+ "eval_exact_match": 14.8,
930
+ "eval_f1": 22.606190476190484,
931
+ "step": 6187
932
+ },
933
+ {
934
+ "epoch": 33.06666666666667,
935
+ "grad_norm": 1.8942480087280273,
936
+ "learning_rate": 3e-05,
937
+ "loss": 0.4805,
938
+ "step": 6200
939
+ },
940
+ {
941
+ "epoch": 33.6,
942
+ "grad_norm": 1.859069585800171,
943
+ "learning_rate": 3e-05,
944
+ "loss": 0.4694,
945
+ "step": 6300
946
+ },
947
+ {
948
+ "epoch": 34.0,
949
+ "eval_accuracy": 0.5524615384615384,
950
+ "eval_loss": 2.701338768005371,
951
+ "eval_runtime": 5.1485,
952
+ "eval_samples_per_second": 97.115,
953
+ "eval_steps_per_second": 12.237,
954
+ "step": 6375
955
+ },
956
+ {
957
+ "epoch": 34.0,
958
+ "eval_exact_match": 13.4,
959
+ "eval_f1": 21.174920634920642,
960
+ "step": 6375
961
+ },
962
+ {
963
+ "epoch": 34.13333333333333,
964
+ "grad_norm": 1.8238064050674438,
965
+ "learning_rate": 3e-05,
966
+ "loss": 0.476,
967
+ "step": 6400
968
+ },
969
+ {
970
+ "epoch": 34.666666666666664,
971
+ "grad_norm": 1.784118413925171,
972
+ "learning_rate": 3e-05,
973
+ "loss": 0.4712,
974
+ "step": 6500
975
+ },
976
+ {
977
+ "epoch": 34.99733333333333,
978
+ "eval_accuracy": 0.5521025641025641,
979
+ "eval_loss": 2.706479072570801,
980
+ "eval_runtime": 5.7012,
981
+ "eval_samples_per_second": 87.7,
982
+ "eval_steps_per_second": 11.05,
983
+ "step": 6562
984
+ },
985
+ {
986
+ "epoch": 34.99733333333333,
987
+ "eval_exact_match": 13.2,
988
+ "eval_f1": 21.548253968253984,
989
+ "step": 6562
990
+ },
991
+ {
992
+ "epoch": 35.2,
993
+ "grad_norm": 1.6702320575714111,
994
+ "learning_rate": 3e-05,
995
+ "loss": 0.4711,
996
+ "step": 6600
997
+ },
998
+ {
999
+ "epoch": 35.733333333333334,
1000
+ "grad_norm": 2.1222314834594727,
1001
+ "learning_rate": 3e-05,
1002
+ "loss": 0.4733,
1003
+ "step": 6700
1004
+ },
1005
+ {
1006
+ "epoch": 36.0,
1007
+ "eval_accuracy": 0.5522564102564103,
1008
+ "eval_loss": 2.670701742172241,
1009
+ "eval_runtime": 5.2238,
1010
+ "eval_samples_per_second": 95.716,
1011
+ "eval_steps_per_second": 12.06,
1012
+ "step": 6750
1013
+ },
1014
+ {
1015
+ "epoch": 36.0,
1016
+ "eval_exact_match": 12.8,
1017
+ "eval_f1": 20.31333333333334,
1018
+ "step": 6750
1019
+ },
1020
+ {
1021
+ "epoch": 36.266666666666666,
1022
+ "grad_norm": 1.5269900560379028,
1023
+ "learning_rate": 3e-05,
1024
+ "loss": 0.4678,
1025
+ "step": 6800
1026
+ },
1027
+ {
1028
+ "epoch": 36.8,
1029
+ "grad_norm": 1.7864423990249634,
1030
+ "learning_rate": 3e-05,
1031
+ "loss": 0.4752,
1032
+ "step": 6900
1033
+ },
1034
+ {
1035
+ "epoch": 36.99733333333333,
1036
+ "eval_accuracy": 0.5531794871794872,
1037
+ "eval_loss": 2.675729513168335,
1038
+ "eval_runtime": 6.0665,
1039
+ "eval_samples_per_second": 82.42,
1040
+ "eval_steps_per_second": 10.385,
1041
+ "step": 6937
1042
+ },
1043
+ {
1044
+ "epoch": 36.99733333333333,
1045
+ "eval_exact_match": 14.4,
1046
+ "eval_f1": 22.22396825396826,
1047
+ "step": 6937
1048
+ },
1049
+ {
1050
+ "epoch": 37.333333333333336,
1051
+ "grad_norm": 2.6186816692352295,
1052
+ "learning_rate": 3e-05,
1053
+ "loss": 0.4671,
1054
+ "step": 7000
1055
+ },
1056
+ {
1057
+ "epoch": 37.86666666666667,
1058
+ "grad_norm": 2.1805689334869385,
1059
+ "learning_rate": 3e-05,
1060
+ "loss": 0.4744,
1061
+ "step": 7100
1062
+ },
1063
+ {
1064
+ "epoch": 38.0,
1065
+ "eval_accuracy": 0.5534358974358974,
1066
+ "eval_loss": 2.701582431793213,
1067
+ "eval_runtime": 4.964,
1068
+ "eval_samples_per_second": 100.725,
1069
+ "eval_steps_per_second": 12.691,
1070
+ "step": 7125
1071
+ },
1072
+ {
1073
+ "epoch": 38.0,
1074
+ "eval_exact_match": 13.6,
1075
+ "eval_f1": 21.16952380952382,
1076
+ "step": 7125
1077
+ },
1078
+ {
1079
+ "epoch": 38.4,
1080
+ "grad_norm": 1.9267176389694214,
1081
+ "learning_rate": 3e-05,
1082
+ "loss": 0.4611,
1083
+ "step": 7200
1084
+ },
1085
+ {
1086
+ "epoch": 38.93333333333333,
1087
+ "grad_norm": 1.8288675546646118,
1088
+ "learning_rate": 3e-05,
1089
+ "loss": 0.4759,
1090
+ "step": 7300
1091
+ },
1092
+ {
1093
+ "epoch": 38.99733333333333,
1094
+ "eval_accuracy": 0.5525641025641026,
1095
+ "eval_loss": 2.726337432861328,
1096
+ "eval_runtime": 5.6276,
1097
+ "eval_samples_per_second": 88.847,
1098
+ "eval_steps_per_second": 11.195,
1099
+ "step": 7312
1100
+ },
1101
+ {
1102
+ "epoch": 38.99733333333333,
1103
+ "eval_exact_match": 12.2,
1104
+ "eval_f1": 19.783968253968265,
1105
+ "step": 7312
1106
+ },
1107
+ {
1108
+ "epoch": 39.46666666666667,
1109
+ "grad_norm": 1.7793669700622559,
1110
+ "learning_rate": 3e-05,
1111
+ "loss": 0.4589,
1112
+ "step": 7400
1113
+ },
1114
+ {
1115
+ "epoch": 40.0,
1116
+ "grad_norm": 1.6059751510620117,
1117
+ "learning_rate": 3e-05,
1118
+ "loss": 0.4759,
1119
+ "step": 7500
1120
+ },
1121
+ {
1122
+ "epoch": 40.0,
1123
+ "eval_accuracy": 0.5525128205128205,
1124
+ "eval_loss": 2.736008405685425,
1125
+ "eval_runtime": 5.6736,
1126
+ "eval_samples_per_second": 88.128,
1127
+ "eval_steps_per_second": 11.104,
1128
+ "step": 7500
1129
+ },
1130
+ {
1131
+ "epoch": 40.0,
1132
+ "eval_exact_match": 12.4,
1133
+ "eval_f1": 20.80158730158731,
1134
+ "step": 7500
1135
+ }
1136
+ ],
1137
+ "logging_steps": 100,
1138
+ "max_steps": 9350,
1139
+ "num_input_tokens_seen": 0,
1140
+ "num_train_epochs": 50,
1141
+ "save_steps": 500,
1142
+ "stateful_callbacks": {
1143
+ "TrainerControl": {
1144
+ "args": {
1145
+ "should_epoch_stop": false,
1146
+ "should_evaluate": false,
1147
+ "should_log": false,
1148
+ "should_save": true,
1149
+ "should_training_stop": false
1150
+ },
1151
+ "attributes": {}
1152
+ }
1153
+ },
1154
+ "total_flos": 1.3738407464481587e+17,
1155
+ "train_batch_size": 2,
1156
+ "trial_name": null,
1157
+ "trial_params": null
1158
+ }
checkpoint-7500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a175de730fa6917b3a611bb9e53766640b7a5e4a426f825f2a602cf98661e47
3
+ size 5304
checkpoint-7500/vocab.json ADDED
The diff for this file is too large to render. See raw diff