Hugging Face's logo

tyzhu
/
lmind_nq_train6000_eval6489_v1_recite_qa_v3_5e-4_lora2

Safetensors
Model card Files Community
tyzhu commited on
Commit
7bdb440
•
1 Parent(s): 330a676

Training in progress, epoch 15, checkpoint

Browse files
Files changed (17) hide show
  1. checkpoint-7935/README.md +9 -0
  2. checkpoint-7935/adapter_config.json +25 -0
  3. checkpoint-7935/adapter_model.safetensors +3 -0
  4. checkpoint-7935/added_tokens.json +5 -0
  5. checkpoint-7935/merges.txt +0 -0
  6. checkpoint-7935/optimizer.pt +3 -0
  7. checkpoint-7935/rng_state_0.pth +3 -0
  8. checkpoint-7935/rng_state_1.pth +3 -0
  9. checkpoint-7935/rng_state_2.pth +3 -0
  10. checkpoint-7935/rng_state_3.pth +3 -0
  11. checkpoint-7935/scheduler.pt +3 -0
  12. checkpoint-7935/special_tokens_map.json +14 -0
  13. checkpoint-7935/tokenizer.json +0 -0
  14. checkpoint-7935/tokenizer_config.json +43 -0
  15. checkpoint-7935/trainer_state.json +871 -0
  16. checkpoint-7935/training_args.bin +3 -0
  17. checkpoint-7935/vocab.json +0 -0
checkpoint-7935/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-7935/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-7935/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a66895c43a60a7f6ea94723b5284ae989d5065ccf1b4f32134868542f1487d3
3
+ size 112130416
checkpoint-7935/added_tokens.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "<|endoftext|>": 151643,
3
+ "<|im_end|>": 151645,
4
+ "<|im_start|>": 151644
5
+ }
checkpoint-7935/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-7935/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:66de7f53e607858035ed0d1b1acb2540fa599dad7e7e643cbc804db2788f3977
3
+ size 224537202
checkpoint-7935/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0da0902550584b56f599595439af2832c8af43a5d3785d061e17d1cd87883b56
3
+ size 15024
checkpoint-7935/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:78e30c4c3520429b9e5a27826017423df4a0a25797452a0370e587353fb06af5
3
+ size 15024
checkpoint-7935/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1f20678582ec64b634734aec6a3119b12172cfce494a2d20eb15342e0f666ae2
3
+ size 15024
checkpoint-7935/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:847c34bbd64ee54949570f51f0f238df3b1695410987856b793ceca7a24b49a4
3
+ size 15024
checkpoint-7935/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c8ada34ae0766670aa7e502870023cb62f43cd43177f61d3c8aba39d0c547716
3
+ size 1064
checkpoint-7935/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-7935/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-7935/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-7935/trainer_state.json ADDED
@@ -0,0 +1,871 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 15.0,
5
+ "eval_steps": 500,
6
+ "global_step": 7935,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.1890359168241966,
13
+ "grad_norm": 7.269421577453613,
14
+ "learning_rate": 0.0005,
15
+ "loss": 1.9195,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.3780718336483932,
20
+ "grad_norm": 1.071650505065918,
21
+ "learning_rate": 0.0005,
22
+ "loss": 1.9396,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.5671077504725898,
27
+ "grad_norm": 0.5662499070167542,
28
+ "learning_rate": 0.0005,
29
+ "loss": 2.3641,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.7561436672967864,
34
+ "grad_norm": 0.47088223695755005,
35
+ "learning_rate": 0.0005,
36
+ "loss": 1.8969,
37
+ "step": 400
38
+ },
39
+ {
40
+ "epoch": 0.945179584120983,
41
+ "grad_norm": 0.3653932511806488,
42
+ "learning_rate": 0.0005,
43
+ "loss": 1.8531,
44
+ "step": 500
45
+ },
46
+ {
47
+ "epoch": 1.0,
48
+ "eval_accuracy": 0.6214529147982063,
49
+ "eval_loss": 1.5580899715423584,
50
+ "eval_runtime": 5.56,
51
+ "eval_samples_per_second": 89.928,
52
+ "eval_steps_per_second": 11.331,
53
+ "step": 529
54
+ },
55
+ {
56
+ "epoch": 1.0,
57
+ "eval_exact_match": 14.4,
58
+ "eval_f1": 20.983333333333345,
59
+ "eval_qa_bleu": 12.514681390986086,
60
+ "eval_qa_exact_match": 0.102,
61
+ "eval_recite_bleu": 12.61307158352263,
62
+ "eval_recite_exact_match": 0.0,
63
+ "step": 529
64
+ },
65
+ {
66
+ "epoch": 1.1342155009451795,
67
+ "grad_norm": 0.5318541526794434,
68
+ "learning_rate": 0.0005,
69
+ "loss": 1.6622,
70
+ "step": 600
71
+ },
72
+ {
73
+ "epoch": 1.3232514177693762,
74
+ "grad_norm": 235.6529998779297,
75
+ "learning_rate": 0.0005,
76
+ "loss": 1.5596,
77
+ "step": 700
78
+ },
79
+ {
80
+ "epoch": 1.5122873345935728,
81
+ "grad_norm": 0.5057249069213867,
82
+ "learning_rate": 0.0005,
83
+ "loss": 1.5296,
84
+ "step": 800
85
+ },
86
+ {
87
+ "epoch": 1.7013232514177694,
88
+ "grad_norm": 0.5339187383651733,
89
+ "learning_rate": 0.0005,
90
+ "loss": 1.5036,
91
+ "step": 900
92
+ },
93
+ {
94
+ "epoch": 1.8903591682419658,
95
+ "grad_norm": 0.5649825930595398,
96
+ "learning_rate": 0.0005,
97
+ "loss": 1.4884,
98
+ "step": 1000
99
+ },
100
+ {
101
+ "epoch": 2.0,
102
+ "eval_accuracy": 0.6643587443946188,
103
+ "eval_loss": 1.2189537286758423,
104
+ "eval_runtime": 6.192,
105
+ "eval_samples_per_second": 80.749,
106
+ "eval_steps_per_second": 10.174,
107
+ "step": 1058
108
+ },
109
+ {
110
+ "epoch": 2.0,
111
+ "eval_exact_match": 11.2,
112
+ "eval_f1": 18.31238095238096,
113
+ "eval_qa_bleu": 10.766593815425745,
114
+ "eval_qa_exact_match": 0.096,
115
+ "eval_recite_bleu": 13.257993783037875,
116
+ "eval_recite_exact_match": 0.002,
117
+ "step": 1058
118
+ },
119
+ {
120
+ "epoch": 2.0793950850661624,
121
+ "grad_norm": 0.6042512655258179,
122
+ "learning_rate": 0.0005,
123
+ "loss": 1.2966,
124
+ "step": 1100
125
+ },
126
+ {
127
+ "epoch": 2.268431001890359,
128
+ "grad_norm": 965.2459106445312,
129
+ "learning_rate": 0.0005,
130
+ "loss": 8.8153,
131
+ "step": 1200
132
+ },
133
+ {
134
+ "epoch": 2.4574669187145557,
135
+ "grad_norm": 3.0128896236419678,
136
+ "learning_rate": 0.0005,
137
+ "loss": 9.1241,
138
+ "step": 1300
139
+ },
140
+ {
141
+ "epoch": 2.6465028355387523,
142
+ "grad_norm": 108062179328.0,
143
+ "learning_rate": 0.0005,
144
+ "loss": 10.5947,
145
+ "step": 1400
146
+ },
147
+ {
148
+ "epoch": 2.835538752362949,
149
+ "grad_norm": 5002999889920.0,
150
+ "learning_rate": 0.0005,
151
+ "loss": 12.5972,
152
+ "step": 1500
153
+ },
154
+ {
155
+ "epoch": 3.0,
156
+ "eval_accuracy": 0.21054708520179372,
157
+ "eval_loss": 19.724363327026367,
158
+ "eval_runtime": 6.5285,
159
+ "eval_samples_per_second": 76.587,
160
+ "eval_steps_per_second": 9.65,
161
+ "step": 1587
162
+ },
163
+ {
164
+ "epoch": 3.0,
165
+ "eval_exact_match": 0.0,
166
+ "eval_f1": 0.0,
167
+ "eval_qa_bleu": 0.0,
168
+ "eval_qa_exact_match": 0.0,
169
+ "eval_recite_bleu": 0.0,
170
+ "eval_recite_exact_match": 0.0,
171
+ "step": 1587
172
+ },
173
+ {
174
+ "epoch": 3.0245746691871456,
175
+ "grad_norm": 62.19309616088867,
176
+ "learning_rate": 0.0005,
177
+ "loss": 13.3192,
178
+ "step": 1600
179
+ },
180
+ {
181
+ "epoch": 3.213610586011342,
182
+ "grad_norm": 4.109750747680664,
183
+ "learning_rate": 0.0005,
184
+ "loss": 8.4804,
185
+ "step": 1700
186
+ },
187
+ {
188
+ "epoch": 3.402646502835539,
189
+ "grad_norm": 7.620689392089844,
190
+ "learning_rate": 0.0005,
191
+ "loss": 8.347,
192
+ "step": 1800
193
+ },
194
+ {
195
+ "epoch": 3.5916824196597354,
196
+ "grad_norm": 33.092830657958984,
197
+ "learning_rate": 0.0005,
198
+ "loss": 7.4667,
199
+ "step": 1900
200
+ },
201
+ {
202
+ "epoch": 3.780718336483932,
203
+ "grad_norm": 204828.203125,
204
+ "learning_rate": 0.0005,
205
+ "loss": 7.885,
206
+ "step": 2000
207
+ },
208
+ {
209
+ "epoch": 3.9697542533081287,
210
+ "grad_norm": 229958729728.0,
211
+ "learning_rate": 0.0005,
212
+ "loss": 8.4174,
213
+ "step": 2100
214
+ },
215
+ {
216
+ "epoch": 4.0,
217
+ "eval_accuracy": 0.22234977578475337,
218
+ "eval_loss": 9.6956148147583,
219
+ "eval_runtime": 5.4245,
220
+ "eval_samples_per_second": 92.174,
221
+ "eval_steps_per_second": 11.614,
222
+ "step": 2116
223
+ },
224
+ {
225
+ "epoch": 4.0,
226
+ "eval_exact_match": 0.0,
227
+ "eval_f1": 0.0,
228
+ "eval_qa_bleu": 0.0,
229
+ "eval_qa_exact_match": 0.0,
230
+ "eval_recite_bleu": 0.0,
231
+ "eval_recite_exact_match": 0.0,
232
+ "step": 2116
233
+ },
234
+ {
235
+ "epoch": 4.158790170132325,
236
+ "grad_norm": 17.098289489746094,
237
+ "learning_rate": 0.0005,
238
+ "loss": 10.4359,
239
+ "step": 2200
240
+ },
241
+ {
242
+ "epoch": 4.3478260869565215,
243
+ "grad_norm": 1069.9039306640625,
244
+ "learning_rate": 0.0005,
245
+ "loss": 9.5369,
246
+ "step": 2300
247
+ },
248
+ {
249
+ "epoch": 4.536862003780718,
250
+ "grad_norm": 7.694222927093506,
251
+ "learning_rate": 0.0005,
252
+ "loss": 7.7028,
253
+ "step": 2400
254
+ },
255
+ {
256
+ "epoch": 4.725897920604915,
257
+ "grad_norm": Infinity,
258
+ "learning_rate": 0.0005,
259
+ "loss": 8.4852,
260
+ "step": 2500
261
+ },
262
+ {
263
+ "epoch": 4.914933837429111,
264
+ "grad_norm": 396939770200064.0,
265
+ "learning_rate": 0.0005,
266
+ "loss": 9.5996,
267
+ "step": 2600
268
+ },
269
+ {
270
+ "epoch": 5.0,
271
+ "eval_accuracy": 0.2166188340807175,
272
+ "eval_loss": 10.412571907043457,
273
+ "eval_runtime": 5.8473,
274
+ "eval_samples_per_second": 85.51,
275
+ "eval_steps_per_second": 10.774,
276
+ "step": 2645
277
+ },
278
+ {
279
+ "epoch": 5.0,
280
+ "eval_exact_match": 0.0,
281
+ "eval_f1": 0.0,
282
+ "eval_qa_bleu": 0.0,
283
+ "eval_qa_exact_match": 0.0,
284
+ "eval_recite_bleu": 0.0,
285
+ "eval_recite_exact_match": 0.0,
286
+ "step": 2645
287
+ },
288
+ {
289
+ "epoch": 5.103969754253308,
290
+ "grad_norm": 3.166937635134374e+18,
291
+ "learning_rate": 0.0005,
292
+ "loss": 10.3832,
293
+ "step": 2700
294
+ },
295
+ {
296
+ "epoch": 5.293005671077505,
297
+ "grad_norm": 1.8363503072027607e+18,
298
+ "learning_rate": 0.0005,
299
+ "loss": 9.7717,
300
+ "step": 2800
301
+ },
302
+ {
303
+ "epoch": 5.482041587901701,
304
+ "grad_norm": 113262112800768.0,
305
+ "learning_rate": 0.0005,
306
+ "loss": 8.3066,
307
+ "step": 2900
308
+ },
309
+ {
310
+ "epoch": 5.671077504725898,
311
+ "grad_norm": 1283434880.0,
312
+ "learning_rate": 0.0005,
313
+ "loss": 8.6003,
314
+ "step": 3000
315
+ },
316
+ {
317
+ "epoch": 5.8601134215500945,
318
+ "grad_norm": 1221250.5,
319
+ "learning_rate": 0.0005,
320
+ "loss": 9.1302,
321
+ "step": 3100
322
+ },
323
+ {
324
+ "epoch": 6.0,
325
+ "eval_accuracy": 0.21948878923766815,
326
+ "eval_loss": 9.434032440185547,
327
+ "eval_runtime": 6.1208,
328
+ "eval_samples_per_second": 81.689,
329
+ "eval_steps_per_second": 10.293,
330
+ "step": 3174
331
+ },
332
+ {
333
+ "epoch": 6.0,
334
+ "eval_exact_match": 0.0,
335
+ "eval_f1": 0.0,
336
+ "eval_qa_bleu": 0.0,
337
+ "eval_qa_exact_match": 0.0,
338
+ "eval_recite_bleu": 0.0,
339
+ "eval_recite_exact_match": 0.0,
340
+ "step": 3174
341
+ },
342
+ {
343
+ "epoch": 6.049149338374291,
344
+ "grad_norm": 2019.499267578125,
345
+ "learning_rate": 0.0005,
346
+ "loss": 9.3999,
347
+ "step": 3200
348
+ },
349
+ {
350
+ "epoch": 6.238185255198488,
351
+ "grad_norm": 164.9034881591797,
352
+ "learning_rate": 0.0005,
353
+ "loss": 8.1768,
354
+ "step": 3300
355
+ },
356
+ {
357
+ "epoch": 6.427221172022684,
358
+ "grad_norm": 434.93157958984375,
359
+ "learning_rate": 0.0005,
360
+ "loss": 7.6329,
361
+ "step": 3400
362
+ },
363
+ {
364
+ "epoch": 6.616257088846881,
365
+ "grad_norm": 1324.7332763671875,
366
+ "learning_rate": 0.0005,
367
+ "loss": 7.7765,
368
+ "step": 3500
369
+ },
370
+ {
371
+ "epoch": 6.805293005671078,
372
+ "grad_norm": 52531002605568.0,
373
+ "learning_rate": 0.0005,
374
+ "loss": 8.006,
375
+ "step": 3600
376
+ },
377
+ {
378
+ "epoch": 6.994328922495274,
379
+ "grad_norm": Infinity,
380
+ "learning_rate": 0.0005,
381
+ "loss": 15.3926,
382
+ "step": 3700
383
+ },
384
+ {
385
+ "epoch": 7.0,
386
+ "eval_accuracy": 0.23054708520179373,
387
+ "eval_loss": 15.052428245544434,
388
+ "eval_runtime": 5.4671,
389
+ "eval_samples_per_second": 91.457,
390
+ "eval_steps_per_second": 11.524,
391
+ "step": 3703
392
+ },
393
+ {
394
+ "epoch": 7.0,
395
+ "eval_exact_match": 0.0,
396
+ "eval_f1": 0.0,
397
+ "eval_qa_bleu": 0.0,
398
+ "eval_qa_exact_match": 0.0,
399
+ "eval_recite_bleu": 0.0,
400
+ "eval_recite_exact_match": 0.0,
401
+ "step": 3703
402
+ },
403
+ {
404
+ "epoch": 7.183364839319471,
405
+ "grad_norm": Infinity,
406
+ "learning_rate": 0.0005,
407
+ "loss": 17.6263,
408
+ "step": 3800
409
+ },
410
+ {
411
+ "epoch": 7.3724007561436675,
412
+ "grad_norm": Infinity,
413
+ "learning_rate": 0.0005,
414
+ "loss": 17.7669,
415
+ "step": 3900
416
+ },
417
+ {
418
+ "epoch": 7.561436672967864,
419
+ "grad_norm": Infinity,
420
+ "learning_rate": 0.0005,
421
+ "loss": 17.7462,
422
+ "step": 4000
423
+ },
424
+ {
425
+ "epoch": 7.750472589792061,
426
+ "grad_norm": Infinity,
427
+ "learning_rate": 0.0005,
428
+ "loss": 17.6186,
429
+ "step": 4100
430
+ },
431
+ {
432
+ "epoch": 7.939508506616257,
433
+ "grad_norm": NaN,
434
+ "learning_rate": 0.0005,
435
+ "loss": 1574.3041,
436
+ "step": 4200
437
+ },
438
+ {
439
+ "epoch": 8.0,
440
+ "eval_accuracy": 0.21060089686098654,
441
+ "eval_loss": NaN,
442
+ "eval_runtime": 5.4376,
443
+ "eval_samples_per_second": 91.952,
444
+ "eval_steps_per_second": 11.586,
445
+ "step": 4232
446
+ },
447
+ {
448
+ "epoch": 8.0,
449
+ "eval_exact_match": 0.0,
450
+ "eval_f1": 0.0,
451
+ "eval_qa_bleu": 0.0,
452
+ "eval_qa_exact_match": 0.0,
453
+ "eval_recite_bleu": 0.0,
454
+ "eval_recite_exact_match": 0.0,
455
+ "step": 4232
456
+ },
457
+ {
458
+ "epoch": 8.128544423440454,
459
+ "grad_norm": NaN,
460
+ "learning_rate": 0.0005,
461
+ "loss": 0.0,
462
+ "step": 4300
463
+ },
464
+ {
465
+ "epoch": 8.31758034026465,
466
+ "grad_norm": NaN,
467
+ "learning_rate": 0.0005,
468
+ "loss": 0.0,
469
+ "step": 4400
470
+ },
471
+ {
472
+ "epoch": 8.506616257088847,
473
+ "grad_norm": NaN,
474
+ "learning_rate": 0.0005,
475
+ "loss": 0.0,
476
+ "step": 4500
477
+ },
478
+ {
479
+ "epoch": 8.695652173913043,
480
+ "grad_norm": NaN,
481
+ "learning_rate": 0.0005,
482
+ "loss": 0.0,
483
+ "step": 4600
484
+ },
485
+ {
486
+ "epoch": 8.88468809073724,
487
+ "grad_norm": NaN,
488
+ "learning_rate": 0.0005,
489
+ "loss": 0.0,
490
+ "step": 4700
491
+ },
492
+ {
493
+ "epoch": 9.0,
494
+ "eval_accuracy": 0.21060089686098654,
495
+ "eval_loss": NaN,
496
+ "eval_runtime": 6.1031,
497
+ "eval_samples_per_second": 81.925,
498
+ "eval_steps_per_second": 10.323,
499
+ "step": 4761
500
+ },
501
+ {
502
+ "epoch": 9.0,
503
+ "eval_exact_match": 0.0,
504
+ "eval_f1": 0.0,
505
+ "eval_qa_bleu": 0.0,
506
+ "eval_qa_exact_match": 0.0,
507
+ "eval_recite_bleu": 0.0,
508
+ "eval_recite_exact_match": 0.0,
509
+ "step": 4761
510
+ },
511
+ {
512
+ "epoch": 9.073724007561436,
513
+ "grad_norm": NaN,
514
+ "learning_rate": 0.0005,
515
+ "loss": 0.0,
516
+ "step": 4800
517
+ },
518
+ {
519
+ "epoch": 9.262759924385634,
520
+ "grad_norm": NaN,
521
+ "learning_rate": 0.0005,
522
+ "loss": 0.0,
523
+ "step": 4900
524
+ },
525
+ {
526
+ "epoch": 9.45179584120983,
527
+ "grad_norm": NaN,
528
+ "learning_rate": 0.0005,
529
+ "loss": 0.0,
530
+ "step": 5000
531
+ },
532
+ {
533
+ "epoch": 9.640831758034027,
534
+ "grad_norm": NaN,
535
+ "learning_rate": 0.0005,
536
+ "loss": 0.0,
537
+ "step": 5100
538
+ },
539
+ {
540
+ "epoch": 9.829867674858223,
541
+ "grad_norm": NaN,
542
+ "learning_rate": 0.0005,
543
+ "loss": 0.0,
544
+ "step": 5200
545
+ },
546
+ {
547
+ "epoch": 10.0,
548
+ "eval_accuracy": 0.21060089686098654,
549
+ "eval_loss": NaN,
550
+ "eval_runtime": 5.5143,
551
+ "eval_samples_per_second": 90.673,
552
+ "eval_steps_per_second": 11.425,
553
+ "step": 5290
554
+ },
555
+ {
556
+ "epoch": 10.0,
557
+ "eval_exact_match": 0.0,
558
+ "eval_f1": 0.0,
559
+ "eval_qa_bleu": 0.0,
560
+ "eval_qa_exact_match": 0.0,
561
+ "eval_recite_bleu": 0.0,
562
+ "eval_recite_exact_match": 0.0,
563
+ "step": 5290
564
+ },
565
+ {
566
+ "epoch": 10.01890359168242,
567
+ "grad_norm": NaN,
568
+ "learning_rate": 0.0005,
569
+ "loss": 0.0,
570
+ "step": 5300
571
+ },
572
+ {
573
+ "epoch": 10.207939508506616,
574
+ "grad_norm": NaN,
575
+ "learning_rate": 0.0005,
576
+ "loss": 0.0,
577
+ "step": 5400
578
+ },
579
+ {
580
+ "epoch": 10.396975425330814,
581
+ "grad_norm": NaN,
582
+ "learning_rate": 0.0005,
583
+ "loss": 0.0,
584
+ "step": 5500
585
+ },
586
+ {
587
+ "epoch": 10.58601134215501,
588
+ "grad_norm": NaN,
589
+ "learning_rate": 0.0005,
590
+ "loss": 0.0,
591
+ "step": 5600
592
+ },
593
+ {
594
+ "epoch": 10.775047258979207,
595
+ "grad_norm": NaN,
596
+ "learning_rate": 0.0005,
597
+ "loss": 0.0,
598
+ "step": 5700
599
+ },
600
+ {
601
+ "epoch": 10.964083175803403,
602
+ "grad_norm": NaN,
603
+ "learning_rate": 0.0005,
604
+ "loss": 0.0,
605
+ "step": 5800
606
+ },
607
+ {
608
+ "epoch": 11.0,
609
+ "eval_accuracy": 0.21060089686098654,
610
+ "eval_loss": NaN,
611
+ "eval_runtime": 6.5846,
612
+ "eval_samples_per_second": 75.934,
613
+ "eval_steps_per_second": 9.568,
614
+ "step": 5819
615
+ },
616
+ {
617
+ "epoch": 11.0,
618
+ "eval_exact_match": 0.0,
619
+ "eval_f1": 0.0,
620
+ "eval_qa_bleu": 0.0,
621
+ "eval_qa_exact_match": 0.0,
622
+ "eval_recite_bleu": 0.0,
623
+ "eval_recite_exact_match": 0.0,
624
+ "step": 5819
625
+ },
626
+ {
627
+ "epoch": 11.1531190926276,
628
+ "grad_norm": NaN,
629
+ "learning_rate": 0.0005,
630
+ "loss": 0.0,
631
+ "step": 5900
632
+ },
633
+ {
634
+ "epoch": 11.342155009451796,
635
+ "grad_norm": NaN,
636
+ "learning_rate": 0.0005,
637
+ "loss": 0.0,
638
+ "step": 6000
639
+ },
640
+ {
641
+ "epoch": 11.531190926275993,
642
+ "grad_norm": NaN,
643
+ "learning_rate": 0.0005,
644
+ "loss": 0.0,
645
+ "step": 6100
646
+ },
647
+ {
648
+ "epoch": 11.720226843100189,
649
+ "grad_norm": NaN,
650
+ "learning_rate": 0.0005,
651
+ "loss": 0.0,
652
+ "step": 6200
653
+ },
654
+ {
655
+ "epoch": 11.909262759924385,
656
+ "grad_norm": NaN,
657
+ "learning_rate": 0.0005,
658
+ "loss": 0.0,
659
+ "step": 6300
660
+ },
661
+ {
662
+ "epoch": 12.0,
663
+ "eval_accuracy": 0.21060089686098654,
664
+ "eval_loss": NaN,
665
+ "eval_runtime": 5.4929,
666
+ "eval_samples_per_second": 91.027,
667
+ "eval_steps_per_second": 11.469,
668
+ "step": 6348
669
+ },
670
+ {
671
+ "epoch": 12.0,
672
+ "eval_exact_match": 0.0,
673
+ "eval_f1": 0.0,
674
+ "eval_qa_bleu": 0.0,
675
+ "eval_qa_exact_match": 0.0,
676
+ "eval_recite_bleu": 0.0,
677
+ "eval_recite_exact_match": 0.0,
678
+ "step": 6348
679
+ },
680
+ {
681
+ "epoch": 12.098298676748582,
682
+ "grad_norm": NaN,
683
+ "learning_rate": 0.0005,
684
+ "loss": 0.0,
685
+ "step": 6400
686
+ },
687
+ {
688
+ "epoch": 12.287334593572778,
689
+ "grad_norm": NaN,
690
+ "learning_rate": 0.0005,
691
+ "loss": 0.0,
692
+ "step": 6500
693
+ },
694
+ {
695
+ "epoch": 12.476370510396976,
696
+ "grad_norm": NaN,
697
+ "learning_rate": 0.0005,
698
+ "loss": 0.0,
699
+ "step": 6600
700
+ },
701
+ {
702
+ "epoch": 12.665406427221171,
703
+ "grad_norm": NaN,
704
+ "learning_rate": 0.0005,
705
+ "loss": 0.0,
706
+ "step": 6700
707
+ },
708
+ {
709
+ "epoch": 12.854442344045369,
710
+ "grad_norm": NaN,
711
+ "learning_rate": 0.0005,
712
+ "loss": 0.0,
713
+ "step": 6800
714
+ },
715
+ {
716
+ "epoch": 13.0,
717
+ "eval_accuracy": 0.21060089686098654,
718
+ "eval_loss": NaN,
719
+ "eval_runtime": 5.7926,
720
+ "eval_samples_per_second": 86.317,
721
+ "eval_steps_per_second": 10.876,
722
+ "step": 6877
723
+ },
724
+ {
725
+ "epoch": 13.0,
726
+ "eval_exact_match": 0.0,
727
+ "eval_f1": 0.0,
728
+ "eval_qa_bleu": 0.0,
729
+ "eval_qa_exact_match": 0.0,
730
+ "eval_recite_bleu": 0.0,
731
+ "eval_recite_exact_match": 0.0,
732
+ "step": 6877
733
+ },
734
+ {
735
+ "epoch": 13.043478260869565,
736
+ "grad_norm": NaN,
737
+ "learning_rate": 0.0005,
738
+ "loss": 0.0,
739
+ "step": 6900
740
+ },
741
+ {
742
+ "epoch": 13.232514177693762,
743
+ "grad_norm": NaN,
744
+ "learning_rate": 0.0005,
745
+ "loss": 0.0,
746
+ "step": 7000
747
+ },
748
+ {
749
+ "epoch": 13.421550094517958,
750
+ "grad_norm": NaN,
751
+ "learning_rate": 0.0005,
752
+ "loss": 0.0,
753
+ "step": 7100
754
+ },
755
+ {
756
+ "epoch": 13.610586011342155,
757
+ "grad_norm": NaN,
758
+ "learning_rate": 0.0005,
759
+ "loss": 0.0,
760
+ "step": 7200
761
+ },
762
+ {
763
+ "epoch": 13.799621928166351,
764
+ "grad_norm": NaN,
765
+ "learning_rate": 0.0005,
766
+ "loss": 0.0,
767
+ "step": 7300
768
+ },
769
+ {
770
+ "epoch": 13.988657844990549,
771
+ "grad_norm": NaN,
772
+ "learning_rate": 0.0005,
773
+ "loss": 0.0,
774
+ "step": 7400
775
+ },
776
+ {
777
+ "epoch": 14.0,
778
+ "eval_accuracy": 0.21060089686098654,
779
+ "eval_loss": NaN,
780
+ "eval_runtime": 5.8325,
781
+ "eval_samples_per_second": 85.726,
782
+ "eval_steps_per_second": 10.802,
783
+ "step": 7406
784
+ },
785
+ {
786
+ "epoch": 14.0,
787
+ "eval_exact_match": 0.0,
788
+ "eval_f1": 0.0,
789
+ "eval_qa_bleu": 0.0,
790
+ "eval_qa_exact_match": 0.0,
791
+ "eval_recite_bleu": 0.0,
792
+ "eval_recite_exact_match": 0.0,
793
+ "step": 7406
794
+ },
795
+ {
796
+ "epoch": 14.177693761814744,
797
+ "grad_norm": NaN,
798
+ "learning_rate": 0.0005,
799
+ "loss": 0.0,
800
+ "step": 7500
801
+ },
802
+ {
803
+ "epoch": 14.366729678638942,
804
+ "grad_norm": NaN,
805
+ "learning_rate": 0.0005,
806
+ "loss": 0.0,
807
+ "step": 7600
808
+ },
809
+ {
810
+ "epoch": 14.555765595463138,
811
+ "grad_norm": NaN,
812
+ "learning_rate": 0.0005,
813
+ "loss": 0.0,
814
+ "step": 7700
815
+ },
816
+ {
817
+ "epoch": 14.744801512287335,
818
+ "grad_norm": NaN,
819
+ "learning_rate": 0.0005,
820
+ "loss": 0.0,
821
+ "step": 7800
822
+ },
823
+ {
824
+ "epoch": 14.93383742911153,
825
+ "grad_norm": NaN,
826
+ "learning_rate": 0.0005,
827
+ "loss": 0.0,
828
+ "step": 7900
829
+ },
830
+ {
831
+ "epoch": 15.0,
832
+ "eval_accuracy": 0.21060089686098654,
833
+ "eval_loss": NaN,
834
+ "eval_runtime": 5.8758,
835
+ "eval_samples_per_second": 85.095,
836
+ "eval_steps_per_second": 10.722,
837
+ "step": 7935
838
+ },
839
+ {
840
+ "epoch": 15.0,
841
+ "eval_exact_match": 0.0,
842
+ "eval_f1": 0.0,
843
+ "eval_qa_bleu": 0.0,
844
+ "eval_qa_exact_match": 0.0,
845
+ "eval_recite_bleu": 0.0,
846
+ "eval_recite_exact_match": 0.0,
847
+ "step": 7935
848
+ }
849
+ ],
850
+ "logging_steps": 100,
851
+ "max_steps": 26450,
852
+ "num_input_tokens_seen": 0,
853
+ "num_train_epochs": 50,
854
+ "save_steps": 500,
855
+ "stateful_callbacks": {
856
+ "TrainerControl": {
857
+ "args": {
858
+ "should_epoch_stop": false,
859
+ "should_evaluate": false,
860
+ "should_log": false,
861
+ "should_save": true,
862
+ "should_training_stop": false
863
+ },
864
+ "attributes": {}
865
+ }
866
+ },
867
+ "total_flos": 7.231829752267407e+17,
868
+ "train_batch_size": 1,
869
+ "trial_name": null,
870
+ "trial_params": null
871
+ }
checkpoint-7935/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1820a0b512c2fa8b3950f986e6f44420dacfa98792c1f255a5842682ec51f235
3
+ size 5304
checkpoint-7935/vocab.json ADDED
The diff for this file is too large to render. See raw diff