Hugging Face's logo

tyzhu
/
lmind_nq_train6000_eval6489_v1_docidx_v3_1e-4_lora2

Safetensors
Generated from Trainer
Model card Files Community
tyzhu commited on
Commit
7b8f2c3
•
1 Parent(s): b8c78b2

Training in progress, epoch 20, checkpoint

Browse files
Files changed (17) hide show
  1. checkpoint-6830/README.md +9 -0
  2. checkpoint-6830/adapter_config.json +25 -0
  3. checkpoint-6830/adapter_model.safetensors +3 -0
  4. checkpoint-6830/added_tokens.json +5 -0
  5. checkpoint-6830/merges.txt +0 -0
  6. checkpoint-6830/optimizer.pt +3 -0
  7. checkpoint-6830/rng_state_0.pth +3 -0
  8. checkpoint-6830/rng_state_1.pth +3 -0
  9. checkpoint-6830/rng_state_2.pth +3 -0
  10. checkpoint-6830/rng_state_3.pth +3 -0
  11. checkpoint-6830/scheduler.pt +3 -0
  12. checkpoint-6830/special_tokens_map.json +14 -0
  13. checkpoint-6830/tokenizer.json +0 -0
  14. checkpoint-6830/tokenizer_config.json +43 -0
  15. checkpoint-6830/trainer_state.json +829 -0
  16. checkpoint-6830/training_args.bin +3 -0
  17. checkpoint-6830/vocab.json +0 -0
checkpoint-6830/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-6830/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-6830/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:283c12944e45c62445854de373180f50d4a5a7a7aca781055f1ef74a6b2c1dcc
3
+ size 112130416
checkpoint-6830/added_tokens.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "<|endoftext|>": 151643,
3
+ "<|im_end|>": 151645,
4
+ "<|im_start|>": 151644
5
+ }
checkpoint-6830/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-6830/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01afbe4e504fc7d1f9c5a464e26f5905dd20ed7474336f24893f92ba08dd7528
3
+ size 224537202
checkpoint-6830/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f8e535e98b718993e7013e87d3e3318febcbddf3830ec7a25db803b56199f1d4
3
+ size 15024
checkpoint-6830/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:73119044581445fde2b7cf4cb795bb2fc5160cca009e5fc23eee5a92980cdeef
3
+ size 15024
checkpoint-6830/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3ba623e2a4eb29803878ac03ec6cb8910faa6120fccdbd5e9dbeae137fe62ce9
3
+ size 15024
checkpoint-6830/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4415ac1c3399b78b35c91f3f33bcd082c7c8ada91cf8ca6c4e6d5f8a32652365
3
+ size 15024
checkpoint-6830/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:244865848ec3536ddc86a45ec7db9351fb3a73f3cf883822d0e57e21c1748cf3
3
+ size 1064
checkpoint-6830/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-6830/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-6830/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-6830/trainer_state.json ADDED
@@ -0,0 +1,829 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 20.0,
5
+ "eval_steps": 500,
6
+ "global_step": 6830,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.29282576866764276,
13
+ "grad_norm": 0.2411503791809082,
14
+ "learning_rate": 0.0001,
15
+ "loss": 2.0009,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.5856515373352855,
20
+ "grad_norm": 0.22286373376846313,
21
+ "learning_rate": 0.0001,
22
+ "loss": 1.9538,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.8784773060029283,
27
+ "grad_norm": 0.24079065024852753,
28
+ "learning_rate": 0.0001,
29
+ "loss": 1.9571,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.9985358711566618,
34
+ "eval_accuracy": 0.45379487179487177,
35
+ "eval_loss": 3.9512319564819336,
36
+ "eval_runtime": 5.0647,
37
+ "eval_samples_per_second": 98.722,
38
+ "eval_steps_per_second": 12.439,
39
+ "step": 341
40
+ },
41
+ {
42
+ "epoch": 0.9985358711566618,
43
+ "eval_bleu": 0.28767194245703276,
44
+ "eval_exact_match": 0.03,
45
+ "eval_prefix_exact_match": 0.042,
46
+ "step": 341
47
+ },
48
+ {
49
+ "epoch": 1.171303074670571,
50
+ "grad_norm": 0.33546799421310425,
51
+ "learning_rate": 0.0001,
52
+ "loss": 1.9255,
53
+ "step": 400
54
+ },
55
+ {
56
+ "epoch": 1.4641288433382138,
57
+ "grad_norm": 0.3515666723251343,
58
+ "learning_rate": 0.0001,
59
+ "loss": 1.8684,
60
+ "step": 500
61
+ },
62
+ {
63
+ "epoch": 1.7569546120058566,
64
+ "grad_norm": 0.35733896493911743,
65
+ "learning_rate": 0.0001,
66
+ "loss": 1.8819,
67
+ "step": 600
68
+ },
69
+ {
70
+ "epoch": 2.0,
71
+ "eval_accuracy": 0.4482564102564103,
72
+ "eval_loss": 4.112804412841797,
73
+ "eval_runtime": 5.0848,
74
+ "eval_samples_per_second": 98.333,
75
+ "eval_steps_per_second": 12.39,
76
+ "step": 683
77
+ },
78
+ {
79
+ "epoch": 2.0,
80
+ "eval_bleu": 0.24521038067633488,
81
+ "eval_exact_match": 0.02,
82
+ "eval_prefix_exact_match": 0.044,
83
+ "step": 683
84
+ },
85
+ {
86
+ "epoch": 2.049780380673499,
87
+ "grad_norm": 0.3784458339214325,
88
+ "learning_rate": 0.0001,
89
+ "loss": 1.8613,
90
+ "step": 700
91
+ },
92
+ {
93
+ "epoch": 2.342606149341142,
94
+ "grad_norm": 0.46702849864959717,
95
+ "learning_rate": 0.0001,
96
+ "loss": 1.743,
97
+ "step": 800
98
+ },
99
+ {
100
+ "epoch": 2.6354319180087846,
101
+ "grad_norm": 0.5156407356262207,
102
+ "learning_rate": 0.0001,
103
+ "loss": 1.7653,
104
+ "step": 900
105
+ },
106
+ {
107
+ "epoch": 2.9282576866764276,
108
+ "grad_norm": 0.4755038619041443,
109
+ "learning_rate": 0.0001,
110
+ "loss": 1.7702,
111
+ "step": 1000
112
+ },
113
+ {
114
+ "epoch": 2.998535871156662,
115
+ "eval_accuracy": 0.446051282051282,
116
+ "eval_loss": 4.327652454376221,
117
+ "eval_runtime": 5.1641,
118
+ "eval_samples_per_second": 96.822,
119
+ "eval_steps_per_second": 12.2,
120
+ "step": 1024
121
+ },
122
+ {
123
+ "epoch": 2.998535871156662,
124
+ "eval_bleu": 0.2598440457703149,
125
+ "eval_exact_match": 0.018,
126
+ "eval_prefix_exact_match": 0.034,
127
+ "step": 1024
128
+ },
129
+ {
130
+ "epoch": 3.22108345534407,
131
+ "grad_norm": 0.5288150906562805,
132
+ "learning_rate": 0.0001,
133
+ "loss": 1.6325,
134
+ "step": 1100
135
+ },
136
+ {
137
+ "epoch": 3.513909224011713,
138
+ "grad_norm": 0.5784010887145996,
139
+ "learning_rate": 0.0001,
140
+ "loss": 1.5896,
141
+ "step": 1200
142
+ },
143
+ {
144
+ "epoch": 3.8067349926793557,
145
+ "grad_norm": 0.6458429098129272,
146
+ "learning_rate": 0.0001,
147
+ "loss": 1.6163,
148
+ "step": 1300
149
+ },
150
+ {
151
+ "epoch": 4.0,
152
+ "eval_accuracy": 0.44235897435897437,
153
+ "eval_loss": 4.584905624389648,
154
+ "eval_runtime": 4.9243,
155
+ "eval_samples_per_second": 101.538,
156
+ "eval_steps_per_second": 12.794,
157
+ "step": 1366
158
+ },
159
+ {
160
+ "epoch": 4.0,
161
+ "eval_bleu": 0.27967222841565076,
162
+ "eval_exact_match": 0.01,
163
+ "eval_prefix_exact_match": 0.038,
164
+ "step": 1366
165
+ },
166
+ {
167
+ "epoch": 4.099560761346998,
168
+ "grad_norm": 0.732841432094574,
169
+ "learning_rate": 0.0001,
170
+ "loss": 1.5441,
171
+ "step": 1400
172
+ },
173
+ {
174
+ "epoch": 4.392386530014641,
175
+ "grad_norm": 0.7372816205024719,
176
+ "learning_rate": 0.0001,
177
+ "loss": 1.4142,
178
+ "step": 1500
179
+ },
180
+ {
181
+ "epoch": 4.685212298682284,
182
+ "grad_norm": 0.8235049247741699,
183
+ "learning_rate": 0.0001,
184
+ "loss": 1.4198,
185
+ "step": 1600
186
+ },
187
+ {
188
+ "epoch": 4.978038067349927,
189
+ "grad_norm": 0.7765743732452393,
190
+ "learning_rate": 0.0001,
191
+ "loss": 1.4427,
192
+ "step": 1700
193
+ },
194
+ {
195
+ "epoch": 4.998535871156662,
196
+ "eval_accuracy": 0.43861538461538463,
197
+ "eval_loss": 4.850330829620361,
198
+ "eval_runtime": 4.9395,
199
+ "eval_samples_per_second": 101.224,
200
+ "eval_steps_per_second": 12.754,
201
+ "step": 1707
202
+ },
203
+ {
204
+ "epoch": 4.998535871156662,
205
+ "eval_bleu": 0.2170008640509292,
206
+ "eval_exact_match": 0.006,
207
+ "eval_prefix_exact_match": 0.032,
208
+ "step": 1707
209
+ },
210
+ {
211
+ "epoch": 5.270863836017569,
212
+ "grad_norm": 0.9119400978088379,
213
+ "learning_rate": 0.0001,
214
+ "loss": 1.2469,
215
+ "step": 1800
216
+ },
217
+ {
218
+ "epoch": 5.563689604685212,
219
+ "grad_norm": 0.8966185450553894,
220
+ "learning_rate": 0.0001,
221
+ "loss": 1.2294,
222
+ "step": 1900
223
+ },
224
+ {
225
+ "epoch": 5.856515373352855,
226
+ "grad_norm": 0.9629669189453125,
227
+ "learning_rate": 0.0001,
228
+ "loss": 1.2498,
229
+ "step": 2000
230
+ },
231
+ {
232
+ "epoch": 6.0,
233
+ "eval_accuracy": 0.4348717948717949,
234
+ "eval_loss": 5.09260892868042,
235
+ "eval_runtime": 4.9572,
236
+ "eval_samples_per_second": 100.863,
237
+ "eval_steps_per_second": 12.709,
238
+ "step": 2049
239
+ },
240
+ {
241
+ "epoch": 6.0,
242
+ "eval_bleu": 0.20196479624851718,
243
+ "eval_exact_match": 0.002,
244
+ "eval_prefix_exact_match": 0.03,
245
+ "step": 2049
246
+ },
247
+ {
248
+ "epoch": 6.149341142020498,
249
+ "grad_norm": 1.0808535814285278,
250
+ "learning_rate": 0.0001,
251
+ "loss": 1.1238,
252
+ "step": 2100
253
+ },
254
+ {
255
+ "epoch": 6.44216691068814,
256
+ "grad_norm": 1.1639196872711182,
257
+ "learning_rate": 0.0001,
258
+ "loss": 1.0301,
259
+ "step": 2200
260
+ },
261
+ {
262
+ "epoch": 6.734992679355783,
263
+ "grad_norm": 1.0662291049957275,
264
+ "learning_rate": 0.0001,
265
+ "loss": 1.0655,
266
+ "step": 2300
267
+ },
268
+ {
269
+ "epoch": 6.998535871156662,
270
+ "eval_accuracy": 0.4326153846153846,
271
+ "eval_loss": 5.270845890045166,
272
+ "eval_runtime": 5.089,
273
+ "eval_samples_per_second": 98.251,
274
+ "eval_steps_per_second": 12.38,
275
+ "step": 2390
276
+ },
277
+ {
278
+ "epoch": 6.998535871156662,
279
+ "eval_bleu": 0.17780932376070374,
280
+ "eval_exact_match": 0.004,
281
+ "eval_prefix_exact_match": 0.024,
282
+ "step": 2390
283
+ },
284
+ {
285
+ "epoch": 7.027818448023426,
286
+ "grad_norm": 1.6462695598602295,
287
+ "learning_rate": 0.0001,
288
+ "loss": 1.0358,
289
+ "step": 2400
290
+ },
291
+ {
292
+ "epoch": 7.320644216691068,
293
+ "grad_norm": 1.2542779445648193,
294
+ "learning_rate": 0.0001,
295
+ "loss": 0.8413,
296
+ "step": 2500
297
+ },
298
+ {
299
+ "epoch": 7.613469985358711,
300
+ "grad_norm": 1.334762454032898,
301
+ "learning_rate": 0.0001,
302
+ "loss": 0.8654,
303
+ "step": 2600
304
+ },
305
+ {
306
+ "epoch": 7.906295754026354,
307
+ "grad_norm": 1.2401787042617798,
308
+ "learning_rate": 0.0001,
309
+ "loss": 0.8733,
310
+ "step": 2700
311
+ },
312
+ {
313
+ "epoch": 8.0,
314
+ "eval_accuracy": 0.4316923076923077,
315
+ "eval_loss": 5.402440547943115,
316
+ "eval_runtime": 4.9383,
317
+ "eval_samples_per_second": 101.25,
318
+ "eval_steps_per_second": 12.758,
319
+ "step": 2732
320
+ },
321
+ {
322
+ "epoch": 8.0,
323
+ "eval_bleu": 0.21418356891762502,
324
+ "eval_exact_match": 0.002,
325
+ "eval_prefix_exact_match": 0.026,
326
+ "step": 2732
327
+ },
328
+ {
329
+ "epoch": 8.199121522693996,
330
+ "grad_norm": 1.255996584892273,
331
+ "learning_rate": 0.0001,
332
+ "loss": 0.7302,
333
+ "step": 2800
334
+ },
335
+ {
336
+ "epoch": 8.49194729136164,
337
+ "grad_norm": 1.2836525440216064,
338
+ "learning_rate": 0.0001,
339
+ "loss": 0.6875,
340
+ "step": 2900
341
+ },
342
+ {
343
+ "epoch": 8.784773060029282,
344
+ "grad_norm": 1.375715732574463,
345
+ "learning_rate": 0.0001,
346
+ "loss": 0.7219,
347
+ "step": 3000
348
+ },
349
+ {
350
+ "epoch": 8.998535871156662,
351
+ "eval_accuracy": 0.42938461538461536,
352
+ "eval_loss": 5.534832000732422,
353
+ "eval_runtime": 5.2363,
354
+ "eval_samples_per_second": 95.487,
355
+ "eval_steps_per_second": 12.031,
356
+ "step": 3073
357
+ },
358
+ {
359
+ "epoch": 8.998535871156662,
360
+ "eval_bleu": 0.20459866130314297,
361
+ "eval_exact_match": 0.0,
362
+ "eval_prefix_exact_match": 0.024,
363
+ "step": 3073
364
+ },
365
+ {
366
+ "epoch": 9.077598828696924,
367
+ "grad_norm": 1.481998324394226,
368
+ "learning_rate": 0.0001,
369
+ "loss": 0.6722,
370
+ "step": 3100
371
+ },
372
+ {
373
+ "epoch": 9.370424597364568,
374
+ "grad_norm": 1.4086748361587524,
375
+ "learning_rate": 0.0001,
376
+ "loss": 0.5489,
377
+ "step": 3200
378
+ },
379
+ {
380
+ "epoch": 9.66325036603221,
381
+ "grad_norm": 1.4381046295166016,
382
+ "learning_rate": 0.0001,
383
+ "loss": 0.565,
384
+ "step": 3300
385
+ },
386
+ {
387
+ "epoch": 9.956076134699854,
388
+ "grad_norm": 1.7364269495010376,
389
+ "learning_rate": 0.0001,
390
+ "loss": 0.5932,
391
+ "step": 3400
392
+ },
393
+ {
394
+ "epoch": 10.0,
395
+ "eval_accuracy": 0.42605128205128207,
396
+ "eval_loss": 5.768980503082275,
397
+ "eval_runtime": 5.0821,
398
+ "eval_samples_per_second": 98.384,
399
+ "eval_steps_per_second": 12.396,
400
+ "step": 3415
401
+ },
402
+ {
403
+ "epoch": 10.0,
404
+ "eval_bleu": 0.19515580420261575,
405
+ "eval_exact_match": 0.002,
406
+ "eval_prefix_exact_match": 0.026,
407
+ "step": 3415
408
+ },
409
+ {
410
+ "epoch": 10.248901903367496,
411
+ "grad_norm": 1.599753737449646,
412
+ "learning_rate": 0.0001,
413
+ "loss": 0.4581,
414
+ "step": 3500
415
+ },
416
+ {
417
+ "epoch": 10.541727672035138,
418
+ "grad_norm": 1.5595277547836304,
419
+ "learning_rate": 0.0001,
420
+ "loss": 0.4519,
421
+ "step": 3600
422
+ },
423
+ {
424
+ "epoch": 10.834553440702782,
425
+ "grad_norm": 1.6201436519622803,
426
+ "learning_rate": 0.0001,
427
+ "loss": 0.4719,
428
+ "step": 3700
429
+ },
430
+ {
431
+ "epoch": 10.998535871156662,
432
+ "eval_accuracy": 0.42538461538461536,
433
+ "eval_loss": 5.89434289932251,
434
+ "eval_runtime": 5.0717,
435
+ "eval_samples_per_second": 98.586,
436
+ "eval_steps_per_second": 12.422,
437
+ "step": 3756
438
+ },
439
+ {
440
+ "epoch": 10.998535871156662,
441
+ "eval_bleu": 0.2074466521823303,
442
+ "eval_exact_match": 0.0,
443
+ "eval_prefix_exact_match": 0.02,
444
+ "step": 3756
445
+ },
446
+ {
447
+ "epoch": 11.127379209370424,
448
+ "grad_norm": 1.6209241151809692,
449
+ "learning_rate": 0.0001,
450
+ "loss": 0.4263,
451
+ "step": 3800
452
+ },
453
+ {
454
+ "epoch": 11.420204978038067,
455
+ "grad_norm": 1.5148394107818604,
456
+ "learning_rate": 0.0001,
457
+ "loss": 0.3625,
458
+ "step": 3900
459
+ },
460
+ {
461
+ "epoch": 11.71303074670571,
462
+ "grad_norm": 1.6862399578094482,
463
+ "learning_rate": 0.0001,
464
+ "loss": 0.3838,
465
+ "step": 4000
466
+ },
467
+ {
468
+ "epoch": 12.0,
469
+ "eval_accuracy": 0.4247179487179487,
470
+ "eval_loss": 6.019149303436279,
471
+ "eval_runtime": 5.3028,
472
+ "eval_samples_per_second": 94.291,
473
+ "eval_steps_per_second": 11.881,
474
+ "step": 4098
475
+ },
476
+ {
477
+ "epoch": 12.0,
478
+ "eval_bleu": 0.16189430124724263,
479
+ "eval_exact_match": 0.002,
480
+ "eval_prefix_exact_match": 0.028,
481
+ "step": 4098
482
+ },
483
+ {
484
+ "epoch": 12.005856515373353,
485
+ "grad_norm": 1.3296407461166382,
486
+ "learning_rate": 0.0001,
487
+ "loss": 0.3952,
488
+ "step": 4100
489
+ },
490
+ {
491
+ "epoch": 12.298682284040996,
492
+ "grad_norm": 1.3439353704452515,
493
+ "learning_rate": 0.0001,
494
+ "loss": 0.2901,
495
+ "step": 4200
496
+ },
497
+ {
498
+ "epoch": 12.591508052708638,
499
+ "grad_norm": 1.5329118967056274,
500
+ "learning_rate": 0.0001,
501
+ "loss": 0.3114,
502
+ "step": 4300
503
+ },
504
+ {
505
+ "epoch": 12.88433382137628,
506
+ "grad_norm": 1.649775743484497,
507
+ "learning_rate": 0.0001,
508
+ "loss": 0.329,
509
+ "step": 4400
510
+ },
511
+ {
512
+ "epoch": 12.998535871156662,
513
+ "eval_accuracy": 0.42456410256410254,
514
+ "eval_loss": 6.104373455047607,
515
+ "eval_runtime": 5.9944,
516
+ "eval_samples_per_second": 83.411,
517
+ "eval_steps_per_second": 10.51,
518
+ "step": 4439
519
+ },
520
+ {
521
+ "epoch": 12.998535871156662,
522
+ "eval_bleu": 0.17170793542335236,
523
+ "eval_exact_match": 0.002,
524
+ "eval_prefix_exact_match": 0.028,
525
+ "step": 4439
526
+ },
527
+ {
528
+ "epoch": 13.177159590043924,
529
+ "grad_norm": 1.5262281894683838,
530
+ "learning_rate": 0.0001,
531
+ "loss": 0.2758,
532
+ "step": 4500
533
+ },
534
+ {
535
+ "epoch": 13.469985358711567,
536
+ "grad_norm": 1.5962131023406982,
537
+ "learning_rate": 0.0001,
538
+ "loss": 0.2594,
539
+ "step": 4600
540
+ },
541
+ {
542
+ "epoch": 13.762811127379209,
543
+ "grad_norm": 1.4258382320404053,
544
+ "learning_rate": 0.0001,
545
+ "loss": 0.2742,
546
+ "step": 4700
547
+ },
548
+ {
549
+ "epoch": 14.0,
550
+ "eval_accuracy": 0.4216410256410256,
551
+ "eval_loss": 6.146524906158447,
552
+ "eval_runtime": 4.9576,
553
+ "eval_samples_per_second": 100.855,
554
+ "eval_steps_per_second": 12.708,
555
+ "step": 4781
556
+ },
557
+ {
558
+ "epoch": 14.0,
559
+ "eval_bleu": 0.20004000747911055,
560
+ "eval_exact_match": 0.002,
561
+ "eval_prefix_exact_match": 0.026,
562
+ "step": 4781
563
+ },
564
+ {
565
+ "epoch": 14.055636896046853,
566
+ "grad_norm": 1.388717532157898,
567
+ "learning_rate": 0.0001,
568
+ "loss": 0.269,
569
+ "step": 4800
570
+ },
571
+ {
572
+ "epoch": 14.348462664714495,
573
+ "grad_norm": 1.2370668649673462,
574
+ "learning_rate": 0.0001,
575
+ "loss": 0.2226,
576
+ "step": 4900
577
+ },
578
+ {
579
+ "epoch": 14.641288433382137,
580
+ "grad_norm": 1.3778610229492188,
581
+ "learning_rate": 0.0001,
582
+ "loss": 0.2334,
583
+ "step": 5000
584
+ },
585
+ {
586
+ "epoch": 14.93411420204978,
587
+ "grad_norm": 1.6371277570724487,
588
+ "learning_rate": 0.0001,
589
+ "loss": 0.2432,
590
+ "step": 5100
591
+ },
592
+ {
593
+ "epoch": 14.998535871156662,
594
+ "eval_accuracy": 0.4226666666666667,
595
+ "eval_loss": 6.3254475593566895,
596
+ "eval_runtime": 5.2196,
597
+ "eval_samples_per_second": 95.794,
598
+ "eval_steps_per_second": 12.07,
599
+ "step": 5122
600
+ },
601
+ {
602
+ "epoch": 14.998535871156662,
603
+ "eval_bleu": 0.16083749853030369,
604
+ "eval_exact_match": 0.0,
605
+ "eval_prefix_exact_match": 0.024,
606
+ "step": 5122
607
+ },
608
+ {
609
+ "epoch": 15.226939970717423,
610
+ "grad_norm": 1.2761411666870117,
611
+ "learning_rate": 0.0001,
612
+ "loss": 0.2066,
613
+ "step": 5200
614
+ },
615
+ {
616
+ "epoch": 15.519765739385067,
617
+ "grad_norm": 1.6878156661987305,
618
+ "learning_rate": 0.0001,
619
+ "loss": 0.2044,
620
+ "step": 5300
621
+ },
622
+ {
623
+ "epoch": 15.812591508052709,
624
+ "grad_norm": 1.4964863061904907,
625
+ "learning_rate": 0.0001,
626
+ "loss": 0.2158,
627
+ "step": 5400
628
+ },
629
+ {
630
+ "epoch": 16.0,
631
+ "eval_accuracy": 0.4227692307692308,
632
+ "eval_loss": 6.441042423248291,
633
+ "eval_runtime": 4.9987,
634
+ "eval_samples_per_second": 100.025,
635
+ "eval_steps_per_second": 12.603,
636
+ "step": 5464
637
+ },
638
+ {
639
+ "epoch": 16.0,
640
+ "eval_bleu": 0.1349628182313855,
641
+ "eval_exact_match": 0.002,
642
+ "eval_prefix_exact_match": 0.026,
643
+ "step": 5464
644
+ },
645
+ {
646
+ "epoch": 16.105417276720353,
647
+ "grad_norm": 0.9965784549713135,
648
+ "learning_rate": 0.0001,
649
+ "loss": 0.2038,
650
+ "step": 5500
651
+ },
652
+ {
653
+ "epoch": 16.398243045387993,
654
+ "grad_norm": 1.2694511413574219,
655
+ "learning_rate": 0.0001,
656
+ "loss": 0.1857,
657
+ "step": 5600
658
+ },
659
+ {
660
+ "epoch": 16.691068814055637,
661
+ "grad_norm": 1.436286211013794,
662
+ "learning_rate": 0.0001,
663
+ "loss": 0.1944,
664
+ "step": 5700
665
+ },
666
+ {
667
+ "epoch": 16.98389458272328,
668
+ "grad_norm": 1.526081919670105,
669
+ "learning_rate": 0.0001,
670
+ "loss": 0.2013,
671
+ "step": 5800
672
+ },
673
+ {
674
+ "epoch": 16.998535871156662,
675
+ "eval_accuracy": 0.4214871794871795,
676
+ "eval_loss": 6.39243745803833,
677
+ "eval_runtime": 5.1033,
678
+ "eval_samples_per_second": 97.975,
679
+ "eval_steps_per_second": 12.345,
680
+ "step": 5805
681
+ },
682
+ {
683
+ "epoch": 16.998535871156662,
684
+ "eval_bleu": 0.18366001556911046,
685
+ "eval_exact_match": 0.0,
686
+ "eval_prefix_exact_match": 0.018,
687
+ "step": 5805
688
+ },
689
+ {
690
+ "epoch": 17.27672035139092,
691
+ "grad_norm": 1.1292760372161865,
692
+ "learning_rate": 0.0001,
693
+ "loss": 0.1694,
694
+ "step": 5900
695
+ },
696
+ {
697
+ "epoch": 17.569546120058565,
698
+ "grad_norm": 1.440766453742981,
699
+ "learning_rate": 0.0001,
700
+ "loss": 0.1782,
701
+ "step": 6000
702
+ },
703
+ {
704
+ "epoch": 17.86237188872621,
705
+ "grad_norm": 1.2640533447265625,
706
+ "learning_rate": 0.0001,
707
+ "loss": 0.1851,
708
+ "step": 6100
709
+ },
710
+ {
711
+ "epoch": 18.0,
712
+ "eval_accuracy": 0.42005128205128206,
713
+ "eval_loss": 6.5217366218566895,
714
+ "eval_runtime": 5.0655,
715
+ "eval_samples_per_second": 98.707,
716
+ "eval_steps_per_second": 12.437,
717
+ "step": 6147
718
+ },
719
+ {
720
+ "epoch": 18.0,
721
+ "eval_bleu": 0.14616910828479238,
722
+ "eval_exact_match": 0.0,
723
+ "eval_prefix_exact_match": 0.026,
724
+ "step": 6147
725
+ },
726
+ {
727
+ "epoch": 18.15519765739385,
728
+ "grad_norm": 1.430617094039917,
729
+ "learning_rate": 0.0001,
730
+ "loss": 0.1717,
731
+ "step": 6200
732
+ },
733
+ {
734
+ "epoch": 18.448023426061493,
735
+ "grad_norm": 1.1885381937026978,
736
+ "learning_rate": 0.0001,
737
+ "loss": 0.1656,
738
+ "step": 6300
739
+ },
740
+ {
741
+ "epoch": 18.740849194729137,
742
+ "grad_norm": 1.64545738697052,
743
+ "learning_rate": 0.0001,
744
+ "loss": 0.1721,
745
+ "step": 6400
746
+ },
747
+ {
748
+ "epoch": 18.998535871156662,
749
+ "eval_accuracy": 0.4208717948717949,
750
+ "eval_loss": 6.557307720184326,
751
+ "eval_runtime": 5.092,
752
+ "eval_samples_per_second": 98.193,
753
+ "eval_steps_per_second": 12.372,
754
+ "step": 6488
755
+ },
756
+ {
757
+ "epoch": 18.998535871156662,
758
+ "eval_bleu": 0.16916503510298256,
759
+ "eval_exact_match": 0.0,
760
+ "eval_prefix_exact_match": 0.02,
761
+ "step": 6488
762
+ },
763
+ {
764
+ "epoch": 19.03367496339678,
765
+ "grad_norm": 1.2168452739715576,
766
+ "learning_rate": 0.0001,
767
+ "loss": 0.1743,
768
+ "step": 6500
769
+ },
770
+ {
771
+ "epoch": 19.32650073206442,
772
+ "grad_norm": 1.2320917844772339,
773
+ "learning_rate": 0.0001,
774
+ "loss": 0.152,
775
+ "step": 6600
776
+ },
777
+ {
778
+ "epoch": 19.619326500732065,
779
+ "grad_norm": 1.1541388034820557,
780
+ "learning_rate": 0.0001,
781
+ "loss": 0.1619,
782
+ "step": 6700
783
+ },
784
+ {
785
+ "epoch": 19.91215226939971,
786
+ "grad_norm": 1.2821346521377563,
787
+ "learning_rate": 0.0001,
788
+ "loss": 0.1676,
789
+ "step": 6800
790
+ },
791
+ {
792
+ "epoch": 20.0,
793
+ "eval_accuracy": 0.42138461538461536,
794
+ "eval_loss": 6.566097736358643,
795
+ "eval_runtime": 5.0056,
796
+ "eval_samples_per_second": 99.888,
797
+ "eval_steps_per_second": 12.586,
798
+ "step": 6830
799
+ },
800
+ {
801
+ "epoch": 20.0,
802
+ "eval_bleu": 0.21213643830050385,
803
+ "eval_exact_match": 0.0,
804
+ "eval_prefix_exact_match": 0.026,
805
+ "step": 6830
806
+ }
807
+ ],
808
+ "logging_steps": 100,
809
+ "max_steps": 17050,
810
+ "num_input_tokens_seen": 0,
811
+ "num_train_epochs": 50,
812
+ "save_steps": 500,
813
+ "stateful_callbacks": {
814
+ "TrainerControl": {
815
+ "args": {
816
+ "should_epoch_stop": false,
817
+ "should_evaluate": false,
818
+ "should_log": false,
819
+ "should_save": true,
820
+ "should_training_stop": false
821
+ },
822
+ "attributes": {}
823
+ }
824
+ },
825
+ "total_flos": 5.86504958073897e+17,
826
+ "train_batch_size": 1,
827
+ "trial_name": null,
828
+ "trial_params": null
829
+ }
checkpoint-6830/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7e63e74298c414f7364874e60281a380506bddfac466373e2426a740b4eb5594
3
+ size 5304
checkpoint-6830/vocab.json ADDED
The diff for this file is too large to render. See raw diff