Chang-Hoo commited on
Commit
d26cb30
·
verified ·
1 Parent(s): e2df44c

Upload folder using huggingface_hub

Browse files
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ checkpoint-1178/tokenizer.json filter=lfs diff=lfs merge=lfs -text
checkpoint-1178/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "<image_soft_token>": 262144
3
+ }
checkpoint-1178/config.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Gemma3ForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "attn_logit_softcapping": null,
8
+ "bos_token_id": 2,
9
+ "cache_implementation": "hybrid",
10
+ "eos_token_id": 1,
11
+ "final_logit_softcapping": null,
12
+ "head_dim": 256,
13
+ "hidden_activation": "gelu_pytorch_tanh",
14
+ "hidden_size": 1152,
15
+ "initializer_range": 0.02,
16
+ "intermediate_size": 6912,
17
+ "max_position_embeddings": 32768,
18
+ "model_type": "gemma3_text",
19
+ "num_attention_heads": 4,
20
+ "num_hidden_layers": 26,
21
+ "num_key_value_heads": 1,
22
+ "pad_token_id": 0,
23
+ "query_pre_attn_scalar": 256,
24
+ "rms_norm_eps": 1e-06,
25
+ "rope_local_base_freq": 10000,
26
+ "rope_scaling": null,
27
+ "rope_theta": 1000000,
28
+ "sliding_window": 512,
29
+ "sliding_window_pattern": 6,
30
+ "torch_dtype": "bfloat16",
31
+ "transformers_version": "4.51.3",
32
+ "use_cache": false,
33
+ "vocab_size": 262144
34
+ }
checkpoint-1178/generation_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "attn_implementation": "eager",
3
+ "bos_token_id": 2,
4
+ "cache_implementation": "hybrid",
5
+ "do_sample": true,
6
+ "eos_token_id": [
7
+ 1,
8
+ 106
9
+ ],
10
+ "pad_token_id": 0,
11
+ "top_k": 64,
12
+ "top_p": 0.95,
13
+ "transformers_version": "4.51.3"
14
+ }
checkpoint-1178/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6ca6ff115d0c87b9828175a3905790418483efb5f3db53890191527dc9cbaca1
3
+ size 1999811208
checkpoint-1178/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa39dfd039eb16d79744ab7a215e22fe8805022b5051464ea78fa762d7476631
3
+ size 3999833742
checkpoint-1178/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e0a6c4ff85c2ea915faa44ce3243aeed88fae9e068f9390ddc786520c0e4c0cc
3
+ size 14512
checkpoint-1178/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3846a90b7dcfe50ad794a55718045d186eeef8318ee50e463651a0a6f4b3b747
3
+ size 14512
checkpoint-1178/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:049642a522f78395382e2dafafcfe8b6c14961cebfb8aa168314e3e9c8f0202b
3
+ size 1064
checkpoint-1178/special_tokens_map.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "boi_token": "<start_of_image>",
3
+ "bos_token": {
4
+ "content": "<bos>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ "eoi_token": "<end_of_image>",
11
+ "eos_token": {
12
+ "content": "<eos>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false
17
+ },
18
+ "image_token": "<image_soft_token>",
19
+ "pad_token": {
20
+ "content": "<pad>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false
25
+ },
26
+ "unk_token": {
27
+ "content": "<unk>",
28
+ "lstrip": false,
29
+ "normalized": false,
30
+ "rstrip": false,
31
+ "single_word": false
32
+ }
33
+ }
checkpoint-1178/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4667f2089529e8e7657cfb6d1c19910ae71ff5f28aa7ab2ff2763330affad795
3
+ size 33384568
checkpoint-1178/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1299c11d7cf632ef3b4e11937501358ada021bbdf7c47638d13c0ee982f2e79c
3
+ size 4689074
checkpoint-1178/tokenizer_config.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1178/trainer_state.json ADDED
@@ -0,0 +1,970 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 1.9990812071524489,
6
+ "eval_steps": 500,
7
+ "global_step": 1178,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.016962329493250407,
14
+ "grad_norm": 3.53125,
15
+ "learning_rate": 1.6666666666666667e-05,
16
+ "loss": 2.1453,
17
+ "mean_token_accuracy": 0.5800132380177577,
18
+ "step": 10
19
+ },
20
+ {
21
+ "epoch": 0.033924658986500815,
22
+ "grad_norm": 2.53125,
23
+ "learning_rate": 3.518518518518519e-05,
24
+ "loss": 1.713,
25
+ "mean_token_accuracy": 0.6173386509219806,
26
+ "step": 20
27
+ },
28
+ {
29
+ "epoch": 0.05088698847975122,
30
+ "grad_norm": 1.671875,
31
+ "learning_rate": 5.370370370370371e-05,
32
+ "loss": 1.5855,
33
+ "mean_token_accuracy": 0.6299800969660282,
34
+ "step": 30
35
+ },
36
+ {
37
+ "epoch": 0.06784931797300163,
38
+ "grad_norm": 1.90625,
39
+ "learning_rate": 7.222222222222222e-05,
40
+ "loss": 1.5199,
41
+ "mean_token_accuracy": 0.6394588612020016,
42
+ "step": 40
43
+ },
44
+ {
45
+ "epoch": 0.08481164746625203,
46
+ "grad_norm": 1.5546875,
47
+ "learning_rate": 9.074074074074075e-05,
48
+ "loss": 1.5192,
49
+ "mean_token_accuracy": 0.6397132301082213,
50
+ "step": 50
51
+ },
52
+ {
53
+ "epoch": 0.10177397695950244,
54
+ "grad_norm": 1.6875,
55
+ "learning_rate": 9.999789785826972e-05,
56
+ "loss": 1.5276,
57
+ "mean_token_accuracy": 0.6391662692030271,
58
+ "step": 60
59
+ },
60
+ {
61
+ "epoch": 0.11873630645275285,
62
+ "grad_norm": 1.7734375,
63
+ "learning_rate": 9.998108178497258e-05,
64
+ "loss": 1.5183,
65
+ "mean_token_accuracy": 0.6404511784513791,
66
+ "step": 70
67
+ },
68
+ {
69
+ "epoch": 0.13569863594600326,
70
+ "grad_norm": 6.3125,
71
+ "learning_rate": 9.99474552942225e-05,
72
+ "loss": 1.5112,
73
+ "mean_token_accuracy": 0.6418155938386917,
74
+ "step": 80
75
+ },
76
+ {
77
+ "epoch": 0.15266096543925367,
78
+ "grad_norm": 1.4765625,
79
+ "learning_rate": 9.989702969580565e-05,
80
+ "loss": 1.5161,
81
+ "mean_token_accuracy": 0.6407449401915073,
82
+ "step": 90
83
+ },
84
+ {
85
+ "epoch": 0.16962329493250405,
86
+ "grad_norm": 1.421875,
87
+ "learning_rate": 9.982982194964625e-05,
88
+ "loss": 1.4945,
89
+ "mean_token_accuracy": 0.6452280322710673,
90
+ "step": 100
91
+ },
92
+ {
93
+ "epoch": 0.18658562442575446,
94
+ "grad_norm": 1.3203125,
95
+ "learning_rate": 9.974585466010236e-05,
96
+ "loss": 1.4927,
97
+ "mean_token_accuracy": 0.6472473913182815,
98
+ "step": 110
99
+ },
100
+ {
101
+ "epoch": 0.20354795391900488,
102
+ "grad_norm": 1.390625,
103
+ "learning_rate": 9.964515606836321e-05,
104
+ "loss": 1.4876,
105
+ "mean_token_accuracy": 0.646234019100666,
106
+ "step": 120
107
+ },
108
+ {
109
+ "epoch": 0.2205102834122553,
110
+ "grad_norm": 1.390625,
111
+ "learning_rate": 9.952776004295077e-05,
112
+ "loss": 1.4768,
113
+ "mean_token_accuracy": 0.6493511145313581,
114
+ "step": 130
115
+ },
116
+ {
117
+ "epoch": 0.2374726129055057,
118
+ "grad_norm": 1.4375,
119
+ "learning_rate": 9.939370606832841e-05,
120
+ "loss": 1.4627,
121
+ "mean_token_accuracy": 0.6511956502993902,
122
+ "step": 140
123
+ },
124
+ {
125
+ "epoch": 0.2544349423987561,
126
+ "grad_norm": 1.3515625,
127
+ "learning_rate": 9.924303923162097e-05,
128
+ "loss": 1.4855,
129
+ "mean_token_accuracy": 0.650639555354913,
130
+ "step": 150
131
+ },
132
+ {
133
+ "epoch": 0.2713972718920065,
134
+ "grad_norm": 1.1875,
135
+ "learning_rate": 9.907581020745037e-05,
136
+ "loss": 1.4492,
137
+ "mean_token_accuracy": 0.6547389343380928,
138
+ "step": 160
139
+ },
140
+ {
141
+ "epoch": 0.2883596013852569,
142
+ "grad_norm": 1.2734375,
143
+ "learning_rate": 9.889207524089187e-05,
144
+ "loss": 1.4298,
145
+ "mean_token_accuracy": 0.65942026724418,
146
+ "step": 170
147
+ },
148
+ {
149
+ "epoch": 0.30532193087850734,
150
+ "grad_norm": 1.2109375,
151
+ "learning_rate": 9.869189612855685e-05,
152
+ "loss": 1.4457,
153
+ "mean_token_accuracy": 0.6557608020802339,
154
+ "step": 180
155
+ },
156
+ {
157
+ "epoch": 0.3222842603717577,
158
+ "grad_norm": 1.1328125,
159
+ "learning_rate": 9.847534019780848e-05,
160
+ "loss": 1.4136,
161
+ "mean_token_accuracy": 0.6603276548286279,
162
+ "step": 190
163
+ },
164
+ {
165
+ "epoch": 0.3392465898650081,
166
+ "grad_norm": 1.2421875,
167
+ "learning_rate": 9.824248028411703e-05,
168
+ "loss": 1.4263,
169
+ "mean_token_accuracy": 0.6607817115883032,
170
+ "step": 200
171
+ },
172
+ {
173
+ "epoch": 0.35620891935825855,
174
+ "grad_norm": 1.1015625,
175
+ "learning_rate": 9.79933947065628e-05,
176
+ "loss": 1.4148,
177
+ "mean_token_accuracy": 0.6610926086703937,
178
+ "step": 210
179
+ },
180
+ {
181
+ "epoch": 0.37317124885150893,
182
+ "grad_norm": 1.046875,
183
+ "learning_rate": 9.772816724149459e-05,
184
+ "loss": 1.3812,
185
+ "mean_token_accuracy": 0.6686675310134887,
186
+ "step": 220
187
+ },
188
+ {
189
+ "epoch": 0.39013357834475937,
190
+ "grad_norm": 1.1328125,
191
+ "learning_rate": 9.744688709435268e-05,
192
+ "loss": 1.4174,
193
+ "mean_token_accuracy": 0.6620935648679733,
194
+ "step": 230
195
+ },
196
+ {
197
+ "epoch": 0.40709590783800975,
198
+ "grad_norm": 1.125,
199
+ "learning_rate": 9.714964886966598e-05,
200
+ "loss": 1.4102,
201
+ "mean_token_accuracy": 0.6628513303895791,
202
+ "step": 240
203
+ },
204
+ {
205
+ "epoch": 0.42405823733126013,
206
+ "grad_norm": 1.1484375,
207
+ "learning_rate": 9.6836552539233e-05,
208
+ "loss": 1.3909,
209
+ "mean_token_accuracy": 0.6663404104610284,
210
+ "step": 250
211
+ },
212
+ {
213
+ "epoch": 0.4410205668245106,
214
+ "grad_norm": 1.03125,
215
+ "learning_rate": 9.650770340849796e-05,
216
+ "loss": 1.4062,
217
+ "mean_token_accuracy": 0.6653375633060932,
218
+ "step": 260
219
+ },
220
+ {
221
+ "epoch": 0.45798289631776096,
222
+ "grad_norm": 1.265625,
223
+ "learning_rate": 9.616321208113262e-05,
224
+ "loss": 1.4091,
225
+ "mean_token_accuracy": 0.6641857360800107,
226
+ "step": 270
227
+ },
228
+ {
229
+ "epoch": 0.4749452258110114,
230
+ "grad_norm": 1.078125,
231
+ "learning_rate": 9.580319442183654e-05,
232
+ "loss": 1.3828,
233
+ "mean_token_accuracy": 0.6680692491432031,
234
+ "step": 280
235
+ },
236
+ {
237
+ "epoch": 0.4919075553042618,
238
+ "grad_norm": 1.1484375,
239
+ "learning_rate": 9.542777151736746e-05,
240
+ "loss": 1.3675,
241
+ "mean_token_accuracy": 0.6707451656460762,
242
+ "step": 290
243
+ },
244
+ {
245
+ "epoch": 0.5088698847975122,
246
+ "grad_norm": 1.0625,
247
+ "learning_rate": 9.503706963581562e-05,
248
+ "loss": 1.3782,
249
+ "mean_token_accuracy": 0.6665912042061488,
250
+ "step": 300
251
+ },
252
+ {
253
+ "epoch": 0.5258322142907625,
254
+ "grad_norm": 1.171875,
255
+ "learning_rate": 9.463122018413532e-05,
256
+ "loss": 1.3536,
257
+ "mean_token_accuracy": 0.6734739691019058,
258
+ "step": 310
259
+ },
260
+ {
261
+ "epoch": 0.542794543784013,
262
+ "grad_norm": 1.015625,
263
+ "learning_rate": 9.4210359663948e-05,
264
+ "loss": 1.3887,
265
+ "mean_token_accuracy": 0.668233826259772,
266
+ "step": 320
267
+ },
268
+ {
269
+ "epoch": 0.5597568732772634,
270
+ "grad_norm": 1.0078125,
271
+ "learning_rate": 9.377462962563195e-05,
272
+ "loss": 1.3819,
273
+ "mean_token_accuracy": 0.6673479390641054,
274
+ "step": 330
275
+ },
276
+ {
277
+ "epoch": 0.5767192027705138,
278
+ "grad_norm": 1.015625,
279
+ "learning_rate": 9.332417662071385e-05,
280
+ "loss": 1.3679,
281
+ "mean_token_accuracy": 0.6706842251121998,
282
+ "step": 340
283
+ },
284
+ {
285
+ "epoch": 0.5936815322637642,
286
+ "grad_norm": 0.9609375,
287
+ "learning_rate": 9.285915215257828e-05,
288
+ "loss": 1.3615,
289
+ "mean_token_accuracy": 0.6692039887110393,
290
+ "step": 350
291
+ },
292
+ {
293
+ "epoch": 0.6106438617570147,
294
+ "grad_norm": 1.078125,
295
+ "learning_rate": 9.237971262551175e-05,
296
+ "loss": 1.3566,
297
+ "mean_token_accuracy": 0.6739495868484179,
298
+ "step": 360
299
+ },
300
+ {
301
+ "epoch": 0.6276061912502651,
302
+ "grad_norm": 1.1875,
303
+ "learning_rate": 9.188601929209835e-05,
304
+ "loss": 1.3103,
305
+ "mean_token_accuracy": 0.6818199207385381,
306
+ "step": 370
307
+ },
308
+ {
309
+ "epoch": 0.6445685207435154,
310
+ "grad_norm": 1.0,
311
+ "learning_rate": 9.137823819898477e-05,
312
+ "loss": 1.3527,
313
+ "mean_token_accuracy": 0.6732211743791898,
314
+ "step": 380
315
+ },
316
+ {
317
+ "epoch": 0.6615308502367658,
318
+ "grad_norm": 0.9453125,
319
+ "learning_rate": 9.08565401310329e-05,
320
+ "loss": 1.3332,
321
+ "mean_token_accuracy": 0.6759632855653763,
322
+ "step": 390
323
+ },
324
+ {
325
+ "epoch": 0.6784931797300162,
326
+ "grad_norm": 0.98046875,
327
+ "learning_rate": 9.03211005538788e-05,
328
+ "loss": 1.3457,
329
+ "mean_token_accuracy": 0.6737116026381652,
330
+ "step": 400
331
+ },
332
+ {
333
+ "epoch": 0.6954555092232667,
334
+ "grad_norm": 1.078125,
335
+ "learning_rate": 8.977209955491739e-05,
336
+ "loss": 1.3713,
337
+ "mean_token_accuracy": 0.6700817617277305,
338
+ "step": 410
339
+ },
340
+ {
341
+ "epoch": 0.7124178387165171,
342
+ "grad_norm": 1.03125,
343
+ "learning_rate": 8.920972178273257e-05,
344
+ "loss": 1.3243,
345
+ "mean_token_accuracy": 0.6765588760375977,
346
+ "step": 420
347
+ },
348
+ {
349
+ "epoch": 0.7293801682097675,
350
+ "grad_norm": 0.98046875,
351
+ "learning_rate": 8.863415638499341e-05,
352
+ "loss": 1.3272,
353
+ "mean_token_accuracy": 0.6768454472223918,
354
+ "step": 430
355
+ },
356
+ {
357
+ "epoch": 0.7463424977030179,
358
+ "grad_norm": 1.015625,
359
+ "learning_rate": 8.8045596944837e-05,
360
+ "loss": 1.3411,
361
+ "mean_token_accuracy": 0.6759745722015699,
362
+ "step": 440
363
+ },
364
+ {
365
+ "epoch": 0.7633048271962682,
366
+ "grad_norm": 1.0,
367
+ "learning_rate": 8.744424141575959e-05,
368
+ "loss": 1.3138,
369
+ "mean_token_accuracy": 0.6793065622448922,
370
+ "step": 450
371
+ },
372
+ {
373
+ "epoch": 0.7802671566895187,
374
+ "grad_norm": 0.9765625,
375
+ "learning_rate": 8.683029205503773e-05,
376
+ "loss": 1.3562,
377
+ "mean_token_accuracy": 0.6727576293051243,
378
+ "step": 460
379
+ },
380
+ {
381
+ "epoch": 0.7972294861827691,
382
+ "grad_norm": 1.0546875,
383
+ "learning_rate": 8.620395535570198e-05,
384
+ "loss": 1.347,
385
+ "mean_token_accuracy": 0.6732823781669139,
386
+ "step": 470
387
+ },
388
+ {
389
+ "epoch": 0.8141918156760195,
390
+ "grad_norm": 0.93359375,
391
+ "learning_rate": 8.556544197708596e-05,
392
+ "loss": 1.3011,
393
+ "mean_token_accuracy": 0.6819205803175767,
394
+ "step": 480
395
+ },
396
+ {
397
+ "epoch": 0.8311541451692699,
398
+ "grad_norm": 0.96484375,
399
+ "learning_rate": 8.491496667397408e-05,
400
+ "loss": 1.3051,
401
+ "mean_token_accuracy": 0.6827595402797063,
402
+ "step": 490
403
+ },
404
+ {
405
+ "epoch": 0.8481164746625203,
406
+ "grad_norm": 1.1015625,
407
+ "learning_rate": 8.42527482243719e-05,
408
+ "loss": 1.2991,
409
+ "mean_token_accuracy": 0.6841210166613261,
410
+ "step": 500
411
+ },
412
+ {
413
+ "epoch": 0.8650788041557708,
414
+ "grad_norm": 0.9375,
415
+ "learning_rate": 8.357900935592327e-05,
416
+ "loss": 1.3055,
417
+ "mean_token_accuracy": 0.679613892485698,
418
+ "step": 510
419
+ },
420
+ {
421
+ "epoch": 0.8820411336490211,
422
+ "grad_norm": 0.95703125,
423
+ "learning_rate": 8.289397667099909e-05,
424
+ "loss": 1.3155,
425
+ "mean_token_accuracy": 0.6812887417773406,
426
+ "step": 520
427
+ },
428
+ {
429
+ "epoch": 0.8990034631422715,
430
+ "grad_norm": 0.9453125,
431
+ "learning_rate": 8.219788057048286e-05,
432
+ "loss": 1.3154,
433
+ "mean_token_accuracy": 0.6792417210837205,
434
+ "step": 530
435
+ },
436
+ {
437
+ "epoch": 0.9159657926355219,
438
+ "grad_norm": 0.9375,
439
+ "learning_rate": 8.149095517627871e-05,
440
+ "loss": 1.3034,
441
+ "mean_token_accuracy": 0.6806264075140158,
442
+ "step": 540
443
+ },
444
+ {
445
+ "epoch": 0.9329281221287723,
446
+ "grad_norm": 0.8984375,
447
+ "learning_rate": 8.077343825256783e-05,
448
+ "loss": 1.3126,
449
+ "mean_token_accuracy": 0.6810529338816802,
450
+ "step": 550
451
+ },
452
+ {
453
+ "epoch": 0.9498904516220228,
454
+ "grad_norm": 0.9609375,
455
+ "learning_rate": 8.004557112583986e-05,
456
+ "loss": 1.3134,
457
+ "mean_token_accuracy": 0.6798362337052822,
458
+ "step": 560
459
+ },
460
+ {
461
+ "epoch": 0.9668527811152732,
462
+ "grad_norm": 0.921875,
463
+ "learning_rate": 7.930759860372628e-05,
464
+ "loss": 1.2856,
465
+ "mean_token_accuracy": 0.683349988112847,
466
+ "step": 570
467
+ },
468
+ {
469
+ "epoch": 0.9838151106085236,
470
+ "grad_norm": 0.88671875,
471
+ "learning_rate": 7.855976889266288e-05,
472
+ "loss": 1.2901,
473
+ "mean_token_accuracy": 0.6849393486976624,
474
+ "step": 580
475
+ },
476
+ {
477
+ "epoch": 1.001696232949325,
478
+ "grad_norm": 2.609375,
479
+ "learning_rate": 7.780233351440903e-05,
480
+ "loss": 1.395,
481
+ "mean_token_accuracy": 0.6899135421733467,
482
+ "step": 590
483
+ },
484
+ {
485
+ "epoch": 1.0186585624425755,
486
+ "grad_norm": 0.93359375,
487
+ "learning_rate": 7.703554722145201e-05,
488
+ "loss": 1.1017,
489
+ "mean_token_accuracy": 0.7215727421144644,
490
+ "step": 600
491
+ },
492
+ {
493
+ "epoch": 1.0356208919358258,
494
+ "grad_norm": 0.9609375,
495
+ "learning_rate": 7.625966791132468e-05,
496
+ "loss": 1.0907,
497
+ "mean_token_accuracy": 0.7208627772827944,
498
+ "step": 610
499
+ },
500
+ {
501
+ "epoch": 1.0525832214290762,
502
+ "grad_norm": 0.99609375,
503
+ "learning_rate": 7.547495653986536e-05,
504
+ "loss": 1.0893,
505
+ "mean_token_accuracy": 0.7245491112271945,
506
+ "step": 620
507
+ },
508
+ {
509
+ "epoch": 1.0695455509223266,
510
+ "grad_norm": 1.015625,
511
+ "learning_rate": 7.468167703344902e-05,
512
+ "loss": 1.0853,
513
+ "mean_token_accuracy": 0.7234922610223293,
514
+ "step": 630
515
+ },
516
+ {
517
+ "epoch": 1.086507880415577,
518
+ "grad_norm": 0.92578125,
519
+ "learning_rate": 7.388009620021959e-05,
520
+ "loss": 1.1004,
521
+ "mean_token_accuracy": 0.7199652560055256,
522
+ "step": 640
523
+ },
524
+ {
525
+ "epoch": 1.1034702099088274,
526
+ "grad_norm": 0.94921875,
527
+ "learning_rate": 7.307048364035266e-05,
528
+ "loss": 1.1206,
529
+ "mean_token_accuracy": 0.719177692135175,
530
+ "step": 650
531
+ },
532
+ {
533
+ "epoch": 1.120432539402078,
534
+ "grad_norm": 0.9375,
535
+ "learning_rate": 7.225311165537956e-05,
536
+ "loss": 1.0905,
537
+ "mean_token_accuracy": 0.7229047452410062,
538
+ "step": 660
539
+ },
540
+ {
541
+ "epoch": 1.1373948688953284,
542
+ "grad_norm": 0.90234375,
543
+ "learning_rate": 7.142825515660259e-05,
544
+ "loss": 1.1184,
545
+ "mean_token_accuracy": 0.7195753792921702,
546
+ "step": 670
547
+ },
548
+ {
549
+ "epoch": 1.1543571983885788,
550
+ "grad_norm": 0.9453125,
551
+ "learning_rate": 7.059619157263245e-05,
552
+ "loss": 1.1152,
553
+ "mean_token_accuracy": 0.7191205089290936,
554
+ "step": 680
555
+ },
556
+ {
557
+ "epoch": 1.1713195278818291,
558
+ "grad_norm": 0.984375,
559
+ "learning_rate": 6.975720075607927e-05,
560
+ "loss": 1.1029,
561
+ "mean_token_accuracy": 0.7206906000773112,
562
+ "step": 690
563
+ },
564
+ {
565
+ "epoch": 1.1882818573750795,
566
+ "grad_norm": 0.953125,
567
+ "learning_rate": 6.891156488942811e-05,
568
+ "loss": 1.0929,
569
+ "mean_token_accuracy": 0.7221428496142228,
570
+ "step": 700
571
+ },
572
+ {
573
+ "epoch": 1.20524418686833,
574
+ "grad_norm": 1.03125,
575
+ "learning_rate": 6.805956839013107e-05,
576
+ "loss": 1.1047,
577
+ "mean_token_accuracy": 0.7189865835011006,
578
+ "step": 710
579
+ },
580
+ {
581
+ "epoch": 1.2222065163615803,
582
+ "grad_norm": 0.97265625,
583
+ "learning_rate": 6.720149781494738e-05,
584
+ "loss": 1.1151,
585
+ "mean_token_accuracy": 0.7195135744909446,
586
+ "step": 720
587
+ },
588
+ {
589
+ "epoch": 1.2391688458548307,
590
+ "grad_norm": 1.0,
591
+ "learning_rate": 6.633764176356434e-05,
592
+ "loss": 1.0717,
593
+ "mean_token_accuracy": 0.7264206613103549,
594
+ "step": 730
595
+ },
596
+ {
597
+ "epoch": 1.2561311753480813,
598
+ "grad_norm": 1.0625,
599
+ "learning_rate": 6.546829078153086e-05,
600
+ "loss": 1.109,
601
+ "mean_token_accuracy": 0.718661529570818,
602
+ "step": 740
603
+ },
604
+ {
605
+ "epoch": 1.2730935048413317,
606
+ "grad_norm": 0.94140625,
607
+ "learning_rate": 6.459373726253672e-05,
608
+ "loss": 1.0936,
609
+ "mean_token_accuracy": 0.7220857585469882,
610
+ "step": 750
611
+ },
612
+ {
613
+ "epoch": 1.290055834334582,
614
+ "grad_norm": 0.9453125,
615
+ "learning_rate": 6.371427535007008e-05,
616
+ "loss": 1.0909,
617
+ "mean_token_accuracy": 0.7229609449704488,
618
+ "step": 760
619
+ },
620
+ {
621
+ "epoch": 1.3070181638278324,
622
+ "grad_norm": 0.8984375,
623
+ "learning_rate": 6.283020083848661e-05,
624
+ "loss": 1.1011,
625
+ "mean_token_accuracy": 0.7228824739654859,
626
+ "step": 770
627
+ },
628
+ {
629
+ "epoch": 1.3239804933210828,
630
+ "grad_norm": 0.91015625,
631
+ "learning_rate": 6.194181107352331e-05,
632
+ "loss": 1.0762,
633
+ "mean_token_accuracy": 0.7260710549851258,
634
+ "step": 780
635
+ },
636
+ {
637
+ "epoch": 1.3409428228143332,
638
+ "grad_norm": 0.88671875,
639
+ "learning_rate": 6.104940485229054e-05,
640
+ "loss": 1.097,
641
+ "mean_token_accuracy": 0.7203553736209869,
642
+ "step": 790
643
+ },
644
+ {
645
+ "epoch": 1.3579051523075836,
646
+ "grad_norm": 1.0234375,
647
+ "learning_rate": 6.015328232277593e-05,
648
+ "loss": 1.1041,
649
+ "mean_token_accuracy": 0.722166525820891,
650
+ "step": 800
651
+ },
652
+ {
653
+ "epoch": 1.374867481800834,
654
+ "grad_norm": 0.97265625,
655
+ "learning_rate": 5.925374488289388e-05,
656
+ "loss": 1.096,
657
+ "mean_token_accuracy": 0.7233567799131075,
658
+ "step": 810
659
+ },
660
+ {
661
+ "epoch": 1.3918298112940843,
662
+ "grad_norm": 0.96875,
663
+ "learning_rate": 5.8351095079114745e-05,
664
+ "loss": 1.103,
665
+ "mean_token_accuracy": 0.721570813159148,
666
+ "step": 820
667
+ },
668
+ {
669
+ "epoch": 1.4087921407873347,
670
+ "grad_norm": 0.95703125,
671
+ "learning_rate": 5.74456365047077e-05,
672
+ "loss": 1.1058,
673
+ "mean_token_accuracy": 0.7204289863506953,
674
+ "step": 830
675
+ },
676
+ {
677
+ "epoch": 1.425754470280585,
678
+ "grad_norm": 1.0859375,
679
+ "learning_rate": 5.653767369763148e-05,
680
+ "loss": 1.0895,
681
+ "mean_token_accuracy": 0.7256171715756258,
682
+ "step": 840
683
+ },
684
+ {
685
+ "epoch": 1.4427167997738355,
686
+ "grad_norm": 0.953125,
687
+ "learning_rate": 5.562751203810742e-05,
688
+ "loss": 1.0696,
689
+ "mean_token_accuracy": 0.72752467567722,
690
+ "step": 850
691
+ },
692
+ {
693
+ "epoch": 1.459679129267086,
694
+ "grad_norm": 0.95703125,
695
+ "learning_rate": 5.471545764590924e-05,
696
+ "loss": 1.0586,
697
+ "mean_token_accuracy": 0.7292891172071297,
698
+ "step": 860
699
+ },
700
+ {
701
+ "epoch": 1.4766414587603365,
702
+ "grad_norm": 0.92578125,
703
+ "learning_rate": 5.3801817277404066e-05,
704
+ "loss": 1.0876,
705
+ "mean_token_accuracy": 0.7224949277937412,
706
+ "step": 870
707
+ },
708
+ {
709
+ "epoch": 1.4936037882535869,
710
+ "grad_norm": 0.921875,
711
+ "learning_rate": 5.28868982223793e-05,
712
+ "loss": 1.0675,
713
+ "mean_token_accuracy": 0.7286781263848146,
714
+ "step": 880
715
+ },
716
+ {
717
+ "epoch": 1.5105661177468372,
718
+ "grad_norm": 0.94140625,
719
+ "learning_rate": 5.197100820069016e-05,
720
+ "loss": 1.0845,
721
+ "mean_token_accuracy": 0.7255125172436238,
722
+ "step": 890
723
+ },
724
+ {
725
+ "epoch": 1.5275284472400876,
726
+ "grad_norm": 0.94140625,
727
+ "learning_rate": 5.1054455258762535e-05,
728
+ "loss": 1.0998,
729
+ "mean_token_accuracy": 0.7214603280027707,
730
+ "step": 900
731
+ },
732
+ {
733
+ "epoch": 1.544490776733338,
734
+ "grad_norm": 0.99609375,
735
+ "learning_rate": 5.0137547665985985e-05,
736
+ "loss": 1.1018,
737
+ "mean_token_accuracy": 0.7225766807794571,
738
+ "step": 910
739
+ },
740
+ {
741
+ "epoch": 1.5614531062265886,
742
+ "grad_norm": 0.9453125,
743
+ "learning_rate": 4.9220593811031786e-05,
744
+ "loss": 1.0482,
745
+ "mean_token_accuracy": 0.7321401789784432,
746
+ "step": 920
747
+ },
748
+ {
749
+ "epoch": 1.578415435719839,
750
+ "grad_norm": 0.9921875,
751
+ "learning_rate": 4.83039020981308e-05,
752
+ "loss": 1.0618,
753
+ "mean_token_accuracy": 0.7308160757025083,
754
+ "step": 930
755
+ },
756
+ {
757
+ "epoch": 1.5953777652130894,
758
+ "grad_norm": 0.9453125,
759
+ "learning_rate": 4.738778084334625e-05,
760
+ "loss": 1.079,
761
+ "mean_token_accuracy": 0.7259436552723249,
762
+ "step": 940
763
+ },
764
+ {
765
+ "epoch": 1.6123400947063398,
766
+ "grad_norm": 0.9921875,
767
+ "learning_rate": 4.6472538170875924e-05,
768
+ "loss": 1.054,
769
+ "mean_token_accuracy": 0.7302558933695157,
770
+ "step": 950
771
+ },
772
+ {
773
+ "epoch": 1.6293024241995901,
774
+ "grad_norm": 0.984375,
775
+ "learning_rate": 4.5558481909419095e-05,
776
+ "loss": 1.0522,
777
+ "mean_token_accuracy": 0.7308388692637284,
778
+ "step": 960
779
+ },
780
+ {
781
+ "epoch": 1.6462647536928405,
782
+ "grad_norm": 0.93359375,
783
+ "learning_rate": 4.46459194886428e-05,
784
+ "loss": 1.0886,
785
+ "mean_token_accuracy": 0.7265370438496271,
786
+ "step": 970
787
+ },
788
+ {
789
+ "epoch": 1.663227083186091,
790
+ "grad_norm": 0.9375,
791
+ "learning_rate": 4.373515783578226e-05,
792
+ "loss": 1.0657,
793
+ "mean_token_accuracy": 0.7278412433962027,
794
+ "step": 980
795
+ },
796
+ {
797
+ "epoch": 1.6801894126793413,
798
+ "grad_norm": 0.90234375,
799
+ "learning_rate": 4.2826503272410304e-05,
800
+ "loss": 1.0636,
801
+ "mean_token_accuracy": 0.7297646810611089,
802
+ "step": 990
803
+ },
804
+ {
805
+ "epoch": 1.6971517421725917,
806
+ "grad_norm": 0.90625,
807
+ "learning_rate": 4.1920261411410536e-05,
808
+ "loss": 1.0802,
809
+ "mean_token_accuracy": 0.7267571208377679,
810
+ "step": 1000
811
+ },
812
+ {
813
+ "epoch": 1.714114071665842,
814
+ "grad_norm": 0.9921875,
815
+ "learning_rate": 4.101673705418888e-05,
816
+ "loss": 1.0609,
817
+ "mean_token_accuracy": 0.7287999058763186,
818
+ "step": 1010
819
+ },
820
+ {
821
+ "epoch": 1.7310764011590924,
822
+ "grad_norm": 1.046875,
823
+ "learning_rate": 4.011623408815799e-05,
824
+ "loss": 1.0515,
825
+ "mean_token_accuracy": 0.7311748243868351,
826
+ "step": 1020
827
+ },
828
+ {
829
+ "epoch": 1.7480387306523428,
830
+ "grad_norm": 0.8984375,
831
+ "learning_rate": 3.9219055384529e-05,
832
+ "loss": 1.0815,
833
+ "mean_token_accuracy": 0.7260699895521004,
834
+ "step": 1030
835
+ },
836
+ {
837
+ "epoch": 1.7650010601455932,
838
+ "grad_norm": 0.95703125,
839
+ "learning_rate": 3.83255026964453e-05,
840
+ "loss": 1.0436,
841
+ "mean_token_accuracy": 0.7335223399102688,
842
+ "step": 1040
843
+ },
844
+ {
845
+ "epoch": 1.7819633896388436,
846
+ "grad_norm": 0.89453125,
847
+ "learning_rate": 3.7435876557492156e-05,
848
+ "loss": 1.0713,
849
+ "mean_token_accuracy": 0.7298086928824584,
850
+ "step": 1050
851
+ },
852
+ {
853
+ "epoch": 1.798925719132094,
854
+ "grad_norm": 1.0,
855
+ "learning_rate": 3.655047618061648e-05,
856
+ "loss": 1.0633,
857
+ "mean_token_accuracy": 0.7291242313881715,
858
+ "step": 1060
859
+ },
860
+ {
861
+ "epoch": 1.8158880486253446,
862
+ "grad_norm": 0.96875,
863
+ "learning_rate": 3.566959935749101e-05,
864
+ "loss": 1.062,
865
+ "mean_token_accuracy": 0.7298740123709043,
866
+ "step": 1070
867
+ },
868
+ {
869
+ "epoch": 1.832850378118595,
870
+ "grad_norm": 0.94921875,
871
+ "learning_rate": 3.479354235835622e-05,
872
+ "loss": 1.041,
873
+ "mean_token_accuracy": 0.7348680111269156,
874
+ "step": 1080
875
+ },
876
+ {
877
+ "epoch": 1.8498127076118454,
878
+ "grad_norm": 0.9375,
879
+ "learning_rate": 3.3922599832374226e-05,
880
+ "loss": 1.0601,
881
+ "mean_token_accuracy": 0.7297768058876196,
882
+ "step": 1090
883
+ },
884
+ {
885
+ "epoch": 1.8667750371050957,
886
+ "grad_norm": 0.9296875,
887
+ "learning_rate": 3.3057064708527686e-05,
888
+ "loss": 1.0516,
889
+ "mean_token_accuracy": 0.7326766779025395,
890
+ "step": 1100
891
+ },
892
+ {
893
+ "epoch": 1.8837373665983461,
894
+ "grad_norm": 0.90234375,
895
+ "learning_rate": 3.2197228097097346e-05,
896
+ "loss": 1.0737,
897
+ "mean_token_accuracy": 0.7286487720906735,
898
+ "step": 1110
899
+ },
900
+ {
901
+ "epoch": 1.9006996960915967,
902
+ "grad_norm": 0.90625,
903
+ "learning_rate": 3.1343379191751364e-05,
904
+ "loss": 1.0685,
905
+ "mean_token_accuracy": 0.7287176544467608,
906
+ "step": 1120
907
+ },
908
+ {
909
+ "epoch": 1.917662025584847,
910
+ "grad_norm": 0.83203125,
911
+ "learning_rate": 3.0495805172279167e-05,
912
+ "loss": 1.0279,
913
+ "mean_token_accuracy": 0.7362240366637707,
914
+ "step": 1130
915
+ },
916
+ {
917
+ "epoch": 1.9346243550780975,
918
+ "grad_norm": 0.9765625,
919
+ "learning_rate": 2.9654791108002567e-05,
920
+ "loss": 1.0658,
921
+ "mean_token_accuracy": 0.7271889204780261,
922
+ "step": 1140
923
+ },
924
+ {
925
+ "epoch": 1.9515866845713479,
926
+ "grad_norm": 0.96875,
927
+ "learning_rate": 2.8820619861896907e-05,
928
+ "loss": 1.0521,
929
+ "mean_token_accuracy": 0.7311085325976213,
930
+ "step": 1150
931
+ },
932
+ {
933
+ "epoch": 1.9685490140645983,
934
+ "grad_norm": 0.875,
935
+ "learning_rate": 2.7993571995454126e-05,
936
+ "loss": 1.0457,
937
+ "mean_token_accuracy": 0.7344075481096903,
938
+ "step": 1160
939
+ },
940
+ {
941
+ "epoch": 1.9855113435578486,
942
+ "grad_norm": 0.92578125,
943
+ "learning_rate": 2.7173925674319957e-05,
944
+ "loss": 1.0498,
945
+ "mean_token_accuracy": 0.7325132201115291,
946
+ "step": 1170
947
+ }
948
+ ],
949
+ "logging_steps": 10,
950
+ "max_steps": 1767,
951
+ "num_input_tokens_seen": 0,
952
+ "num_train_epochs": 3,
953
+ "save_steps": 500,
954
+ "stateful_callbacks": {
955
+ "TrainerControl": {
956
+ "args": {
957
+ "should_epoch_stop": false,
958
+ "should_evaluate": false,
959
+ "should_log": false,
960
+ "should_save": true,
961
+ "should_training_stop": false
962
+ },
963
+ "attributes": {}
964
+ }
965
+ },
966
+ "total_flos": 3.9882636273700045e+17,
967
+ "train_batch_size": 1,
968
+ "trial_name": null,
969
+ "trial_params": null
970
+ }
checkpoint-1178/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b6f91b55a648dec5ac9a0e5602cfae2ecc39937955253e07f1e669899c924bdd
3
+ size 6136