tyzhu commited on
Commit
8185a39
1 Parent(s): 62633b6

Training in progress, epoch 22, checkpoint

Browse files
checkpoint-7513/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-7513/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-7513/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:808bdd8a65ac76cb7c3d532415025eec0bb24b2b0c1ea816be86b3e43a26df76
3
+ size 112130416
checkpoint-7513/added_tokens.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "<|endoftext|>": 151643,
3
+ "<|im_end|>": 151645,
4
+ "<|im_start|>": 151644
5
+ }
checkpoint-7513/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-7513/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:112a23aba55aed8934ab4befbb5aa4903e80f786695bf5241873c5ef6fbceaba
3
+ size 224537202
checkpoint-7513/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d791c2b7b62d38f06d4e38a09bbaf1d267f04e00244e8242c466f17fdb483aed
3
+ size 15024
checkpoint-7513/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:741d2106e6c3acabfd37d4a1dde525faa73250d86cd87f4ee5112ef8ed257304
3
+ size 15024
checkpoint-7513/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2234498a7668534c2dd35e516267003fb4a0f4dc873836abb7403428466a4323
3
+ size 15024
checkpoint-7513/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:25b5b38b19e1eabeabfeed4523bc3b14047b45e29045918217478ba337a8afd0
3
+ size 15024
checkpoint-7513/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:91b92200d8aa134e57a089958c8cfe507795fff9b5ca2e58fbbdc0e131eab10f
3
+ size 1064
checkpoint-7513/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-7513/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-7513/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-7513/trainer_state.json ADDED
@@ -0,0 +1,910 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 22.0,
5
+ "eval_steps": 500,
6
+ "global_step": 7513,
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": 0.26115021109580994,
14
+ "learning_rate": 3e-05,
15
+ "loss": 2.0719,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.5856515373352855,
20
+ "grad_norm": 0.2679530680179596,
21
+ "learning_rate": 3e-05,
22
+ "loss": 1.969,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.8784773060029283,
27
+ "grad_norm": 0.2916224002838135,
28
+ "learning_rate": 3e-05,
29
+ "loss": 1.9686,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.9985358711566618,
34
+ "eval_accuracy": 0.43743589743589745,
35
+ "eval_loss": 3.6918885707855225,
36
+ "eval_runtime": 5.2974,
37
+ "eval_samples_per_second": 94.385,
38
+ "eval_steps_per_second": 11.893,
39
+ "step": 341
40
+ },
41
+ {
42
+ "epoch": 0.9985358711566618,
43
+ "eval_bleu": 0.1527908371131025,
44
+ "eval_exact_match": 0.012,
45
+ "eval_prefix_exact_match": 0.026,
46
+ "step": 341
47
+ },
48
+ {
49
+ "epoch": 1.171303074670571,
50
+ "grad_norm": 0.3534347712993622,
51
+ "learning_rate": 3e-05,
52
+ "loss": 1.9603,
53
+ "step": 400
54
+ },
55
+ {
56
+ "epoch": 1.4641288433382138,
57
+ "grad_norm": 0.3338424563407898,
58
+ "learning_rate": 3e-05,
59
+ "loss": 1.9235,
60
+ "step": 500
61
+ },
62
+ {
63
+ "epoch": 1.7569546120058566,
64
+ "grad_norm": 0.4033390283584595,
65
+ "learning_rate": 3e-05,
66
+ "loss": 1.9337,
67
+ "step": 600
68
+ },
69
+ {
70
+ "epoch": 2.0,
71
+ "eval_accuracy": 0.44764102564102565,
72
+ "eval_loss": 3.7484843730926514,
73
+ "eval_runtime": 5.0649,
74
+ "eval_samples_per_second": 98.718,
75
+ "eval_steps_per_second": 12.438,
76
+ "step": 683
77
+ },
78
+ {
79
+ "epoch": 2.0,
80
+ "eval_bleu": 0.21311912714932216,
81
+ "eval_exact_match": 0.02,
82
+ "eval_prefix_exact_match": 0.032,
83
+ "step": 683
84
+ },
85
+ {
86
+ "epoch": 2.049780380673499,
87
+ "grad_norm": 0.38882532715797424,
88
+ "learning_rate": 3e-05,
89
+ "loss": 1.9267,
90
+ "step": 700
91
+ },
92
+ {
93
+ "epoch": 2.342606149341142,
94
+ "grad_norm": 0.44310709834098816,
95
+ "learning_rate": 3e-05,
96
+ "loss": 1.8823,
97
+ "step": 800
98
+ },
99
+ {
100
+ "epoch": 2.6354319180087846,
101
+ "grad_norm": 0.5196680426597595,
102
+ "learning_rate": 3e-05,
103
+ "loss": 1.901,
104
+ "step": 900
105
+ },
106
+ {
107
+ "epoch": 2.9282576866764276,
108
+ "grad_norm": 0.5293437242507935,
109
+ "learning_rate": 3e-05,
110
+ "loss": 1.9033,
111
+ "step": 1000
112
+ },
113
+ {
114
+ "epoch": 2.998535871156662,
115
+ "eval_accuracy": 0.4495897435897436,
116
+ "eval_loss": 3.882551670074463,
117
+ "eval_runtime": 5.1314,
118
+ "eval_samples_per_second": 97.439,
119
+ "eval_steps_per_second": 12.277,
120
+ "step": 1024
121
+ },
122
+ {
123
+ "epoch": 2.998535871156662,
124
+ "eval_bleu": 0.24307488023362148,
125
+ "eval_exact_match": 0.016,
126
+ "eval_prefix_exact_match": 0.032,
127
+ "step": 1024
128
+ },
129
+ {
130
+ "epoch": 3.22108345534407,
131
+ "grad_norm": 0.552943229675293,
132
+ "learning_rate": 3e-05,
133
+ "loss": 1.8529,
134
+ "step": 1100
135
+ },
136
+ {
137
+ "epoch": 3.513909224011713,
138
+ "grad_norm": 0.6421984434127808,
139
+ "learning_rate": 3e-05,
140
+ "loss": 1.8342,
141
+ "step": 1200
142
+ },
143
+ {
144
+ "epoch": 3.8067349926793557,
145
+ "grad_norm": 0.7155198454856873,
146
+ "learning_rate": 3e-05,
147
+ "loss": 1.857,
148
+ "step": 1300
149
+ },
150
+ {
151
+ "epoch": 4.0,
152
+ "eval_accuracy": 0.4481025641025641,
153
+ "eval_loss": 3.970107316970825,
154
+ "eval_runtime": 5.1563,
155
+ "eval_samples_per_second": 96.969,
156
+ "eval_steps_per_second": 12.218,
157
+ "step": 1366
158
+ },
159
+ {
160
+ "epoch": 4.0,
161
+ "eval_bleu": 0.25309406645138827,
162
+ "eval_exact_match": 0.016,
163
+ "eval_prefix_exact_match": 0.044,
164
+ "step": 1366
165
+ },
166
+ {
167
+ "epoch": 4.099560761346998,
168
+ "grad_norm": 0.7837833166122437,
169
+ "learning_rate": 3e-05,
170
+ "loss": 1.8284,
171
+ "step": 1400
172
+ },
173
+ {
174
+ "epoch": 4.392386530014641,
175
+ "grad_norm": 0.784609854221344,
176
+ "learning_rate": 3e-05,
177
+ "loss": 1.7884,
178
+ "step": 1500
179
+ },
180
+ {
181
+ "epoch": 4.685212298682284,
182
+ "grad_norm": 0.834195613861084,
183
+ "learning_rate": 3e-05,
184
+ "loss": 1.7841,
185
+ "step": 1600
186
+ },
187
+ {
188
+ "epoch": 4.978038067349927,
189
+ "grad_norm": 0.8902062177658081,
190
+ "learning_rate": 3e-05,
191
+ "loss": 1.8042,
192
+ "step": 1700
193
+ },
194
+ {
195
+ "epoch": 4.998535871156662,
196
+ "eval_accuracy": 0.4472820512820513,
197
+ "eval_loss": 4.117065906524658,
198
+ "eval_runtime": 5.2376,
199
+ "eval_samples_per_second": 95.464,
200
+ "eval_steps_per_second": 12.028,
201
+ "step": 1707
202
+ },
203
+ {
204
+ "epoch": 4.998535871156662,
205
+ "eval_bleu": 0.2531049955279983,
206
+ "eval_exact_match": 0.014,
207
+ "eval_prefix_exact_match": 0.042,
208
+ "step": 1707
209
+ },
210
+ {
211
+ "epoch": 5.270863836017569,
212
+ "grad_norm": 0.9495383501052856,
213
+ "learning_rate": 3e-05,
214
+ "loss": 1.7493,
215
+ "step": 1800
216
+ },
217
+ {
218
+ "epoch": 5.563689604685212,
219
+ "grad_norm": 0.9647482633590698,
220
+ "learning_rate": 3e-05,
221
+ "loss": 1.7334,
222
+ "step": 1900
223
+ },
224
+ {
225
+ "epoch": 5.856515373352855,
226
+ "grad_norm": 0.9825767874717712,
227
+ "learning_rate": 3e-05,
228
+ "loss": 1.7443,
229
+ "step": 2000
230
+ },
231
+ {
232
+ "epoch": 6.0,
233
+ "eval_accuracy": 0.446974358974359,
234
+ "eval_loss": 4.183700084686279,
235
+ "eval_runtime": 5.2674,
236
+ "eval_samples_per_second": 94.923,
237
+ "eval_steps_per_second": 11.96,
238
+ "step": 2049
239
+ },
240
+ {
241
+ "epoch": 6.0,
242
+ "eval_bleu": 0.23587359955369594,
243
+ "eval_exact_match": 0.012,
244
+ "eval_prefix_exact_match": 0.04,
245
+ "step": 2049
246
+ },
247
+ {
248
+ "epoch": 6.149341142020498,
249
+ "grad_norm": 1.0430296659469604,
250
+ "learning_rate": 3e-05,
251
+ "loss": 1.6959,
252
+ "step": 2100
253
+ },
254
+ {
255
+ "epoch": 6.44216691068814,
256
+ "grad_norm": 1.0777732133865356,
257
+ "learning_rate": 3e-05,
258
+ "loss": 1.6795,
259
+ "step": 2200
260
+ },
261
+ {
262
+ "epoch": 6.734992679355783,
263
+ "grad_norm": 1.0959677696228027,
264
+ "learning_rate": 3e-05,
265
+ "loss": 1.7019,
266
+ "step": 2300
267
+ },
268
+ {
269
+ "epoch": 6.998535871156662,
270
+ "eval_accuracy": 0.4461538461538462,
271
+ "eval_loss": 4.260408401489258,
272
+ "eval_runtime": 5.178,
273
+ "eval_samples_per_second": 96.563,
274
+ "eval_steps_per_second": 12.167,
275
+ "step": 2390
276
+ },
277
+ {
278
+ "epoch": 6.998535871156662,
279
+ "eval_bleu": 0.2008158466316808,
280
+ "eval_exact_match": 0.006,
281
+ "eval_prefix_exact_match": 0.042,
282
+ "step": 2390
283
+ },
284
+ {
285
+ "epoch": 7.027818448023426,
286
+ "grad_norm": 1.0649549961090088,
287
+ "learning_rate": 3e-05,
288
+ "loss": 1.6812,
289
+ "step": 2400
290
+ },
291
+ {
292
+ "epoch": 7.320644216691068,
293
+ "grad_norm": 1.294122576713562,
294
+ "learning_rate": 3e-05,
295
+ "loss": 1.6226,
296
+ "step": 2500
297
+ },
298
+ {
299
+ "epoch": 7.613469985358711,
300
+ "grad_norm": 1.2657020092010498,
301
+ "learning_rate": 3e-05,
302
+ "loss": 1.6404,
303
+ "step": 2600
304
+ },
305
+ {
306
+ "epoch": 7.906295754026354,
307
+ "grad_norm": 1.1963133811950684,
308
+ "learning_rate": 3e-05,
309
+ "loss": 1.6305,
310
+ "step": 2700
311
+ },
312
+ {
313
+ "epoch": 8.0,
314
+ "eval_accuracy": 0.44148717948717947,
315
+ "eval_loss": 4.406544208526611,
316
+ "eval_runtime": 5.3355,
317
+ "eval_samples_per_second": 93.711,
318
+ "eval_steps_per_second": 11.808,
319
+ "step": 2732
320
+ },
321
+ {
322
+ "epoch": 8.0,
323
+ "eval_bleu": 0.24435500637461924,
324
+ "eval_exact_match": 0.008,
325
+ "eval_prefix_exact_match": 0.038,
326
+ "step": 2732
327
+ },
328
+ {
329
+ "epoch": 8.199121522693996,
330
+ "grad_norm": 1.3801360130310059,
331
+ "learning_rate": 3e-05,
332
+ "loss": 1.5734,
333
+ "step": 2800
334
+ },
335
+ {
336
+ "epoch": 8.49194729136164,
337
+ "grad_norm": 1.3306772708892822,
338
+ "learning_rate": 3e-05,
339
+ "loss": 1.5651,
340
+ "step": 2900
341
+ },
342
+ {
343
+ "epoch": 8.784773060029282,
344
+ "grad_norm": 1.5165746212005615,
345
+ "learning_rate": 3e-05,
346
+ "loss": 1.6056,
347
+ "step": 3000
348
+ },
349
+ {
350
+ "epoch": 8.998535871156662,
351
+ "eval_accuracy": 0.4397948717948718,
352
+ "eval_loss": 4.448728561401367,
353
+ "eval_runtime": 5.187,
354
+ "eval_samples_per_second": 96.395,
355
+ "eval_steps_per_second": 12.146,
356
+ "step": 3073
357
+ },
358
+ {
359
+ "epoch": 8.998535871156662,
360
+ "eval_bleu": 0.20120283162176952,
361
+ "eval_exact_match": 0.008,
362
+ "eval_prefix_exact_match": 0.036,
363
+ "step": 3073
364
+ },
365
+ {
366
+ "epoch": 9.077598828696924,
367
+ "grad_norm": 1.4077060222625732,
368
+ "learning_rate": 3e-05,
369
+ "loss": 1.5612,
370
+ "step": 3100
371
+ },
372
+ {
373
+ "epoch": 9.370424597364568,
374
+ "grad_norm": 1.475984811782837,
375
+ "learning_rate": 3e-05,
376
+ "loss": 1.5158,
377
+ "step": 3200
378
+ },
379
+ {
380
+ "epoch": 9.66325036603221,
381
+ "grad_norm": 1.586300015449524,
382
+ "learning_rate": 3e-05,
383
+ "loss": 1.5145,
384
+ "step": 3300
385
+ },
386
+ {
387
+ "epoch": 9.956076134699854,
388
+ "grad_norm": 1.510048270225525,
389
+ "learning_rate": 3e-05,
390
+ "loss": 1.5521,
391
+ "step": 3400
392
+ },
393
+ {
394
+ "epoch": 10.0,
395
+ "eval_accuracy": 0.4388717948717949,
396
+ "eval_loss": 4.5474138259887695,
397
+ "eval_runtime": 5.169,
398
+ "eval_samples_per_second": 96.731,
399
+ "eval_steps_per_second": 12.188,
400
+ "step": 3415
401
+ },
402
+ {
403
+ "epoch": 10.0,
404
+ "eval_bleu": 0.2258317707850195,
405
+ "eval_exact_match": 0.004,
406
+ "eval_prefix_exact_match": 0.032,
407
+ "step": 3415
408
+ },
409
+ {
410
+ "epoch": 10.248901903367496,
411
+ "grad_norm": 1.6384536027908325,
412
+ "learning_rate": 3e-05,
413
+ "loss": 1.4631,
414
+ "step": 3500
415
+ },
416
+ {
417
+ "epoch": 10.541727672035138,
418
+ "grad_norm": 1.7626007795333862,
419
+ "learning_rate": 3e-05,
420
+ "loss": 1.4623,
421
+ "step": 3600
422
+ },
423
+ {
424
+ "epoch": 10.834553440702782,
425
+ "grad_norm": 1.6483694314956665,
426
+ "learning_rate": 3e-05,
427
+ "loss": 1.4934,
428
+ "step": 3700
429
+ },
430
+ {
431
+ "epoch": 10.998535871156662,
432
+ "eval_accuracy": 0.4367179487179487,
433
+ "eval_loss": 4.589761734008789,
434
+ "eval_runtime": 5.853,
435
+ "eval_samples_per_second": 85.426,
436
+ "eval_steps_per_second": 10.764,
437
+ "step": 3756
438
+ },
439
+ {
440
+ "epoch": 10.998535871156662,
441
+ "eval_bleu": 0.20792247275869374,
442
+ "eval_exact_match": 0.004,
443
+ "eval_prefix_exact_match": 0.026,
444
+ "step": 3756
445
+ },
446
+ {
447
+ "epoch": 11.127379209370424,
448
+ "grad_norm": 1.8367844820022583,
449
+ "learning_rate": 3e-05,
450
+ "loss": 1.4521,
451
+ "step": 3800
452
+ },
453
+ {
454
+ "epoch": 11.420204978038067,
455
+ "grad_norm": 1.8123779296875,
456
+ "learning_rate": 3e-05,
457
+ "loss": 1.4085,
458
+ "step": 3900
459
+ },
460
+ {
461
+ "epoch": 11.71303074670571,
462
+ "grad_norm": 1.9437752962112427,
463
+ "learning_rate": 3e-05,
464
+ "loss": 1.4287,
465
+ "step": 4000
466
+ },
467
+ {
468
+ "epoch": 12.0,
469
+ "eval_accuracy": 0.43553846153846154,
470
+ "eval_loss": 4.691086769104004,
471
+ "eval_runtime": 5.3738,
472
+ "eval_samples_per_second": 93.043,
473
+ "eval_steps_per_second": 11.723,
474
+ "step": 4098
475
+ },
476
+ {
477
+ "epoch": 12.0,
478
+ "eval_bleu": 0.21672573731424186,
479
+ "eval_exact_match": 0.006,
480
+ "eval_prefix_exact_match": 0.028,
481
+ "step": 4098
482
+ },
483
+ {
484
+ "epoch": 12.005856515373353,
485
+ "grad_norm": 1.5838289260864258,
486
+ "learning_rate": 3e-05,
487
+ "loss": 1.4361,
488
+ "step": 4100
489
+ },
490
+ {
491
+ "epoch": 12.298682284040996,
492
+ "grad_norm": 1.8575290441513062,
493
+ "learning_rate": 3e-05,
494
+ "loss": 1.3433,
495
+ "step": 4200
496
+ },
497
+ {
498
+ "epoch": 12.591508052708638,
499
+ "grad_norm": 1.9598041772842407,
500
+ "learning_rate": 3e-05,
501
+ "loss": 1.3661,
502
+ "step": 4300
503
+ },
504
+ {
505
+ "epoch": 12.88433382137628,
506
+ "grad_norm": 1.972550868988037,
507
+ "learning_rate": 3e-05,
508
+ "loss": 1.3846,
509
+ "step": 4400
510
+ },
511
+ {
512
+ "epoch": 12.998535871156662,
513
+ "eval_accuracy": 0.43553846153846154,
514
+ "eval_loss": 4.762892246246338,
515
+ "eval_runtime": 5.106,
516
+ "eval_samples_per_second": 97.925,
517
+ "eval_steps_per_second": 12.339,
518
+ "step": 4439
519
+ },
520
+ {
521
+ "epoch": 12.998535871156662,
522
+ "eval_bleu": 0.2416046504676966,
523
+ "eval_exact_match": 0.006,
524
+ "eval_prefix_exact_match": 0.028,
525
+ "step": 4439
526
+ },
527
+ {
528
+ "epoch": 13.177159590043924,
529
+ "grad_norm": 1.9469107389450073,
530
+ "learning_rate": 3e-05,
531
+ "loss": 1.3229,
532
+ "step": 4500
533
+ },
534
+ {
535
+ "epoch": 13.469985358711567,
536
+ "grad_norm": 2.1680469512939453,
537
+ "learning_rate": 3e-05,
538
+ "loss": 1.3172,
539
+ "step": 4600
540
+ },
541
+ {
542
+ "epoch": 13.762811127379209,
543
+ "grad_norm": 2.1960484981536865,
544
+ "learning_rate": 3e-05,
545
+ "loss": 1.3185,
546
+ "step": 4700
547
+ },
548
+ {
549
+ "epoch": 14.0,
550
+ "eval_accuracy": 0.4327692307692308,
551
+ "eval_loss": 4.758542537689209,
552
+ "eval_runtime": 5.778,
553
+ "eval_samples_per_second": 86.535,
554
+ "eval_steps_per_second": 10.903,
555
+ "step": 4781
556
+ },
557
+ {
558
+ "epoch": 14.0,
559
+ "eval_bleu": 0.19085834097154403,
560
+ "eval_exact_match": 0.004,
561
+ "eval_prefix_exact_match": 0.022,
562
+ "step": 4781
563
+ },
564
+ {
565
+ "epoch": 14.055636896046853,
566
+ "grad_norm": 2.4733147621154785,
567
+ "learning_rate": 3e-05,
568
+ "loss": 1.3027,
569
+ "step": 4800
570
+ },
571
+ {
572
+ "epoch": 14.348462664714495,
573
+ "grad_norm": 2.2826030254364014,
574
+ "learning_rate": 3e-05,
575
+ "loss": 1.2513,
576
+ "step": 4900
577
+ },
578
+ {
579
+ "epoch": 14.641288433382137,
580
+ "grad_norm": 2.041346788406372,
581
+ "learning_rate": 3e-05,
582
+ "loss": 1.2637,
583
+ "step": 5000
584
+ },
585
+ {
586
+ "epoch": 14.93411420204978,
587
+ "grad_norm": 2.3494467735290527,
588
+ "learning_rate": 3e-05,
589
+ "loss": 1.2667,
590
+ "step": 5100
591
+ },
592
+ {
593
+ "epoch": 14.998535871156662,
594
+ "eval_accuracy": 0.4308717948717949,
595
+ "eval_loss": 4.938868522644043,
596
+ "eval_runtime": 5.9024,
597
+ "eval_samples_per_second": 84.711,
598
+ "eval_steps_per_second": 10.674,
599
+ "step": 5122
600
+ },
601
+ {
602
+ "epoch": 14.998535871156662,
603
+ "eval_bleu": 0.18549760176612626,
604
+ "eval_exact_match": 0.002,
605
+ "eval_prefix_exact_match": 0.024,
606
+ "step": 5122
607
+ },
608
+ {
609
+ "epoch": 15.226939970717423,
610
+ "grad_norm": 2.411022663116455,
611
+ "learning_rate": 3e-05,
612
+ "loss": 1.2158,
613
+ "step": 5200
614
+ },
615
+ {
616
+ "epoch": 15.519765739385067,
617
+ "grad_norm": 2.3778491020202637,
618
+ "learning_rate": 3e-05,
619
+ "loss": 1.193,
620
+ "step": 5300
621
+ },
622
+ {
623
+ "epoch": 15.812591508052709,
624
+ "grad_norm": 2.2003743648529053,
625
+ "learning_rate": 3e-05,
626
+ "loss": 1.2144,
627
+ "step": 5400
628
+ },
629
+ {
630
+ "epoch": 16.0,
631
+ "eval_accuracy": 0.43025641025641026,
632
+ "eval_loss": 4.898688316345215,
633
+ "eval_runtime": 5.1834,
634
+ "eval_samples_per_second": 96.463,
635
+ "eval_steps_per_second": 12.154,
636
+ "step": 5464
637
+ },
638
+ {
639
+ "epoch": 16.0,
640
+ "eval_bleu": 0.20104450766148693,
641
+ "eval_exact_match": 0.004,
642
+ "eval_prefix_exact_match": 0.026,
643
+ "step": 5464
644
+ },
645
+ {
646
+ "epoch": 16.105417276720353,
647
+ "grad_norm": 2.1213016510009766,
648
+ "learning_rate": 3e-05,
649
+ "loss": 1.1842,
650
+ "step": 5500
651
+ },
652
+ {
653
+ "epoch": 16.398243045387993,
654
+ "grad_norm": 2.4457600116729736,
655
+ "learning_rate": 3e-05,
656
+ "loss": 1.1527,
657
+ "step": 5600
658
+ },
659
+ {
660
+ "epoch": 16.691068814055637,
661
+ "grad_norm": 2.5844104290008545,
662
+ "learning_rate": 3e-05,
663
+ "loss": 1.1536,
664
+ "step": 5700
665
+ },
666
+ {
667
+ "epoch": 16.98389458272328,
668
+ "grad_norm": 2.7657299041748047,
669
+ "learning_rate": 3e-05,
670
+ "loss": 1.1708,
671
+ "step": 5800
672
+ },
673
+ {
674
+ "epoch": 16.998535871156662,
675
+ "eval_accuracy": 0.4297435897435897,
676
+ "eval_loss": 5.00166654586792,
677
+ "eval_runtime": 5.4251,
678
+ "eval_samples_per_second": 92.164,
679
+ "eval_steps_per_second": 11.613,
680
+ "step": 5805
681
+ },
682
+ {
683
+ "epoch": 16.998535871156662,
684
+ "eval_bleu": 0.2059810081312132,
685
+ "eval_exact_match": 0.004,
686
+ "eval_prefix_exact_match": 0.028,
687
+ "step": 5805
688
+ },
689
+ {
690
+ "epoch": 17.27672035139092,
691
+ "grad_norm": 2.4050705432891846,
692
+ "learning_rate": 3e-05,
693
+ "loss": 1.083,
694
+ "step": 5900
695
+ },
696
+ {
697
+ "epoch": 17.569546120058565,
698
+ "grad_norm": 2.584331750869751,
699
+ "learning_rate": 3e-05,
700
+ "loss": 1.0993,
701
+ "step": 6000
702
+ },
703
+ {
704
+ "epoch": 17.86237188872621,
705
+ "grad_norm": 3.257659912109375,
706
+ "learning_rate": 3e-05,
707
+ "loss": 1.1146,
708
+ "step": 6100
709
+ },
710
+ {
711
+ "epoch": 18.0,
712
+ "eval_accuracy": 0.43066666666666664,
713
+ "eval_loss": 4.977816581726074,
714
+ "eval_runtime": 5.2952,
715
+ "eval_samples_per_second": 94.425,
716
+ "eval_steps_per_second": 11.898,
717
+ "step": 6147
718
+ },
719
+ {
720
+ "epoch": 18.0,
721
+ "eval_bleu": 0.1948234970665932,
722
+ "eval_exact_match": 0.004,
723
+ "eval_prefix_exact_match": 0.032,
724
+ "step": 6147
725
+ },
726
+ {
727
+ "epoch": 18.15519765739385,
728
+ "grad_norm": 3.068880319595337,
729
+ "learning_rate": 3e-05,
730
+ "loss": 1.0673,
731
+ "step": 6200
732
+ },
733
+ {
734
+ "epoch": 18.448023426061493,
735
+ "grad_norm": 2.737912893295288,
736
+ "learning_rate": 3e-05,
737
+ "loss": 1.0381,
738
+ "step": 6300
739
+ },
740
+ {
741
+ "epoch": 18.740849194729137,
742
+ "grad_norm": 2.878462076187134,
743
+ "learning_rate": 3e-05,
744
+ "loss": 1.0531,
745
+ "step": 6400
746
+ },
747
+ {
748
+ "epoch": 18.998535871156662,
749
+ "eval_accuracy": 0.42866666666666664,
750
+ "eval_loss": 5.121572494506836,
751
+ "eval_runtime": 5.3996,
752
+ "eval_samples_per_second": 92.599,
753
+ "eval_steps_per_second": 11.667,
754
+ "step": 6488
755
+ },
756
+ {
757
+ "epoch": 18.998535871156662,
758
+ "eval_bleu": 0.19788734504730915,
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": 2.6934640407562256,
766
+ "learning_rate": 3e-05,
767
+ "loss": 1.0564,
768
+ "step": 6500
769
+ },
770
+ {
771
+ "epoch": 19.32650073206442,
772
+ "grad_norm": 2.822169542312622,
773
+ "learning_rate": 3e-05,
774
+ "loss": 0.99,
775
+ "step": 6600
776
+ },
777
+ {
778
+ "epoch": 19.619326500732065,
779
+ "grad_norm": 2.520814895629883,
780
+ "learning_rate": 3e-05,
781
+ "loss": 0.991,
782
+ "step": 6700
783
+ },
784
+ {
785
+ "epoch": 19.91215226939971,
786
+ "grad_norm": 3.1566810607910156,
787
+ "learning_rate": 3e-05,
788
+ "loss": 1.0158,
789
+ "step": 6800
790
+ },
791
+ {
792
+ "epoch": 20.0,
793
+ "eval_accuracy": 0.42728205128205127,
794
+ "eval_loss": 5.120981216430664,
795
+ "eval_runtime": 5.394,
796
+ "eval_samples_per_second": 92.696,
797
+ "eval_steps_per_second": 11.68,
798
+ "step": 6830
799
+ },
800
+ {
801
+ "epoch": 20.0,
802
+ "eval_bleu": 0.15806516253752823,
803
+ "eval_exact_match": 0.006,
804
+ "eval_prefix_exact_match": 0.028,
805
+ "step": 6830
806
+ },
807
+ {
808
+ "epoch": 20.20497803806735,
809
+ "grad_norm": 2.937910556793213,
810
+ "learning_rate": 3e-05,
811
+ "loss": 0.9615,
812
+ "step": 6900
813
+ },
814
+ {
815
+ "epoch": 20.497803806734993,
816
+ "grad_norm": 3.2016685009002686,
817
+ "learning_rate": 3e-05,
818
+ "loss": 0.9404,
819
+ "step": 7000
820
+ },
821
+ {
822
+ "epoch": 20.790629575402637,
823
+ "grad_norm": 3.2996864318847656,
824
+ "learning_rate": 3e-05,
825
+ "loss": 0.9555,
826
+ "step": 7100
827
+ },
828
+ {
829
+ "epoch": 20.998535871156662,
830
+ "eval_accuracy": 0.42928205128205127,
831
+ "eval_loss": 5.198832988739014,
832
+ "eval_runtime": 5.8905,
833
+ "eval_samples_per_second": 84.882,
834
+ "eval_steps_per_second": 10.695,
835
+ "step": 7171
836
+ },
837
+ {
838
+ "epoch": 20.998535871156662,
839
+ "eval_bleu": 0.18303907416923465,
840
+ "eval_exact_match": 0.004,
841
+ "eval_prefix_exact_match": 0.02,
842
+ "step": 7171
843
+ },
844
+ {
845
+ "epoch": 21.083455344070277,
846
+ "grad_norm": 3.4803571701049805,
847
+ "learning_rate": 3e-05,
848
+ "loss": 0.9454,
849
+ "step": 7200
850
+ },
851
+ {
852
+ "epoch": 21.37628111273792,
853
+ "grad_norm": 2.6315698623657227,
854
+ "learning_rate": 3e-05,
855
+ "loss": 0.8835,
856
+ "step": 7300
857
+ },
858
+ {
859
+ "epoch": 21.669106881405565,
860
+ "grad_norm": 3.0926690101623535,
861
+ "learning_rate": 3e-05,
862
+ "loss": 0.9004,
863
+ "step": 7400
864
+ },
865
+ {
866
+ "epoch": 21.961932650073205,
867
+ "grad_norm": 2.9990713596343994,
868
+ "learning_rate": 3e-05,
869
+ "loss": 0.9205,
870
+ "step": 7500
871
+ },
872
+ {
873
+ "epoch": 22.0,
874
+ "eval_accuracy": 0.427025641025641,
875
+ "eval_loss": 5.2239508628845215,
876
+ "eval_runtime": 5.4025,
877
+ "eval_samples_per_second": 92.55,
878
+ "eval_steps_per_second": 11.661,
879
+ "step": 7513
880
+ },
881
+ {
882
+ "epoch": 22.0,
883
+ "eval_bleu": 0.1854279974336752,
884
+ "eval_exact_match": 0.004,
885
+ "eval_prefix_exact_match": 0.02,
886
+ "step": 7513
887
+ }
888
+ ],
889
+ "logging_steps": 100,
890
+ "max_steps": 17050,
891
+ "num_input_tokens_seen": 0,
892
+ "num_train_epochs": 50,
893
+ "save_steps": 500,
894
+ "stateful_callbacks": {
895
+ "TrainerControl": {
896
+ "args": {
897
+ "should_epoch_stop": false,
898
+ "should_evaluate": false,
899
+ "should_log": false,
900
+ "should_save": true,
901
+ "should_training_stop": false
902
+ },
903
+ "attributes": {}
904
+ }
905
+ },
906
+ "total_flos": 6.451542290019123e+17,
907
+ "train_batch_size": 1,
908
+ "trial_name": null,
909
+ "trial_params": null
910
+ }
checkpoint-7513/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f9cef08c7f10bf396c269dfc26104fd5a94756f5431d6448312805ef8e298380
3
+ size 5304
checkpoint-7513/vocab.json ADDED
The diff for this file is too large to render. See raw diff