ILT37 commited on
Commit
777f24f
1 Parent(s): 0269215

Upload 8 files

Browse files
config.json ADDED
@@ -0,0 +1,328 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "google/vit-base-patch16-224-in21k",
3
+ "architectures": [
4
+ "ViTForImageClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.0,
7
+ "encoder_stride": 16,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.0,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "Golbat",
13
+ "1": "Machoke",
14
+ "10": "Raichu",
15
+ "100": "Dragonite",
16
+ "101": "Fearow",
17
+ "102": "Slowpoke",
18
+ "103": "Weezing",
19
+ "104": "Beedrill",
20
+ "105": "Weedle",
21
+ "106": "Cloyster",
22
+ "107": "Vaporeon",
23
+ "108": "Gyarados",
24
+ "109": "Golduck",
25
+ "11": "Zapdos",
26
+ "110": "Machamp",
27
+ "111": "Hitmonlee",
28
+ "112": "Primeape",
29
+ "113": "Cubone",
30
+ "114": "Sandslash",
31
+ "115": "Scyther",
32
+ "116": "Haunter",
33
+ "117": "Metapod",
34
+ "118": "Tentacruel",
35
+ "119": "Aerodactyl",
36
+ "12": "Raticate",
37
+ "120": "Kabutops",
38
+ "121": "Ninetales",
39
+ "122": "Zubat",
40
+ "123": "Rhydon",
41
+ "124": "Mew",
42
+ "125": "Pinsir",
43
+ "126": "Ditto",
44
+ "127": "Victreebel",
45
+ "128": "Omanyte",
46
+ "129": "Horsea",
47
+ "13": "Magnemite",
48
+ "130": "Pikachu",
49
+ "131": "Blastoise",
50
+ "132": "Venomoth",
51
+ "133": "Charizard",
52
+ "134": "Seadra",
53
+ "135": "Muk",
54
+ "136": "Spearow",
55
+ "137": "Bulbasaur",
56
+ "138": "Bellsprout",
57
+ "139": "Electrode",
58
+ "14": "Ivysaur",
59
+ "140": "Gloom",
60
+ "141": "Poliwhirl",
61
+ "142": "Flareon",
62
+ "143": "Seaking",
63
+ "144": "Hypno",
64
+ "145": "Wartortle",
65
+ "146": "Mankey",
66
+ "147": "Tentacool",
67
+ "148": "Exeggcute",
68
+ "149": "Meowth",
69
+ "15": "Growlithe",
70
+ "16": "Tangela",
71
+ "17": "Drowzee",
72
+ "18": "Rapidash",
73
+ "19": "Venonat",
74
+ "2": "Omastar",
75
+ "20": "Pidgeot",
76
+ "21": "Nidorino",
77
+ "22": "Porygon",
78
+ "23": "Lickitung",
79
+ "24": "Rattata",
80
+ "25": "Machop",
81
+ "26": "Charmeleon",
82
+ "27": "Slowbro",
83
+ "28": "Parasect",
84
+ "29": "Eevee",
85
+ "3": "Diglett",
86
+ "30": "Starmie",
87
+ "31": "Staryu",
88
+ "32": "Psyduck",
89
+ "33": "Dragonair",
90
+ "34": "Magikarp",
91
+ "35": "Vileplume",
92
+ "36": "Marowak",
93
+ "37": "Pidgeotto",
94
+ "38": "Shellder",
95
+ "39": "Mewtwo",
96
+ "4": "Lapras",
97
+ "40": "Farfetchd",
98
+ "41": "Kingler",
99
+ "42": "Seel",
100
+ "43": "Kakuna",
101
+ "44": "Doduo",
102
+ "45": "Electabuzz",
103
+ "46": "Charmander",
104
+ "47": "Rhyhorn",
105
+ "48": "Tauros",
106
+ "49": "Dugtrio",
107
+ "5": "Kabuto",
108
+ "50": "Poliwrath",
109
+ "51": "Gengar",
110
+ "52": "Exeggutor",
111
+ "53": "Dewgong",
112
+ "54": "Jigglypuff",
113
+ "55": "Geodude",
114
+ "56": "Kadabra",
115
+ "57": "Nidorina",
116
+ "58": "Sandshrew",
117
+ "59": "Grimer",
118
+ "6": "Persian",
119
+ "60": "MrMime",
120
+ "61": "Pidgey",
121
+ "62": "Koffing",
122
+ "63": "Ekans",
123
+ "64": "Alolan Sandslash",
124
+ "65": "Venusaur",
125
+ "66": "Snorlax",
126
+ "67": "Paras",
127
+ "68": "Jynx",
128
+ "69": "Chansey",
129
+ "7": "Weepinbell",
130
+ "70": "Hitmonchan",
131
+ "71": "Gastly",
132
+ "72": "Kangaskhan",
133
+ "73": "Oddish",
134
+ "74": "Wigglytuff",
135
+ "75": "Graveler",
136
+ "76": "Arcanine",
137
+ "77": "Clefairy",
138
+ "78": "Articuno",
139
+ "79": "Poliwag",
140
+ "8": "Golem",
141
+ "80": "Abra",
142
+ "81": "Squirtle",
143
+ "82": "Voltorb",
144
+ "83": "Ponyta",
145
+ "84": "Moltres",
146
+ "85": "Nidoqueen",
147
+ "86": "Magmar",
148
+ "87": "Onix",
149
+ "88": "Vulpix",
150
+ "89": "Butterfree",
151
+ "9": "Dodrio",
152
+ "90": "Krabby",
153
+ "91": "Arbok",
154
+ "92": "Clefable",
155
+ "93": "Goldeen",
156
+ "94": "Magneton",
157
+ "95": "Dratini",
158
+ "96": "Caterpie",
159
+ "97": "Jolteon",
160
+ "98": "Nidoking",
161
+ "99": "Alakazam"
162
+ },
163
+ "image_size": 224,
164
+ "initializer_range": 0.02,
165
+ "intermediate_size": 3072,
166
+ "label2id": {
167
+ "Abra": "80",
168
+ "Aerodactyl": "119",
169
+ "Alakazam": "99",
170
+ "Alolan Sandslash": "64",
171
+ "Arbok": "91",
172
+ "Arcanine": "76",
173
+ "Articuno": "78",
174
+ "Beedrill": "104",
175
+ "Bellsprout": "138",
176
+ "Blastoise": "131",
177
+ "Bulbasaur": "137",
178
+ "Butterfree": "89",
179
+ "Caterpie": "96",
180
+ "Chansey": "69",
181
+ "Charizard": "133",
182
+ "Charmander": "46",
183
+ "Charmeleon": "26",
184
+ "Clefable": "92",
185
+ "Clefairy": "77",
186
+ "Cloyster": "106",
187
+ "Cubone": "113",
188
+ "Dewgong": "53",
189
+ "Diglett": "3",
190
+ "Ditto": "126",
191
+ "Dodrio": "9",
192
+ "Doduo": "44",
193
+ "Dragonair": "33",
194
+ "Dragonite": "100",
195
+ "Dratini": "95",
196
+ "Drowzee": "17",
197
+ "Dugtrio": "49",
198
+ "Eevee": "29",
199
+ "Ekans": "63",
200
+ "Electabuzz": "45",
201
+ "Electrode": "139",
202
+ "Exeggcute": "148",
203
+ "Exeggutor": "52",
204
+ "Farfetchd": "40",
205
+ "Fearow": "101",
206
+ "Flareon": "142",
207
+ "Gastly": "71",
208
+ "Gengar": "51",
209
+ "Geodude": "55",
210
+ "Gloom": "140",
211
+ "Golbat": "0",
212
+ "Goldeen": "93",
213
+ "Golduck": "109",
214
+ "Golem": "8",
215
+ "Graveler": "75",
216
+ "Grimer": "59",
217
+ "Growlithe": "15",
218
+ "Gyarados": "108",
219
+ "Haunter": "116",
220
+ "Hitmonchan": "70",
221
+ "Hitmonlee": "111",
222
+ "Horsea": "129",
223
+ "Hypno": "144",
224
+ "Ivysaur": "14",
225
+ "Jigglypuff": "54",
226
+ "Jolteon": "97",
227
+ "Jynx": "68",
228
+ "Kabuto": "5",
229
+ "Kabutops": "120",
230
+ "Kadabra": "56",
231
+ "Kakuna": "43",
232
+ "Kangaskhan": "72",
233
+ "Kingler": "41",
234
+ "Koffing": "62",
235
+ "Krabby": "90",
236
+ "Lapras": "4",
237
+ "Lickitung": "23",
238
+ "Machamp": "110",
239
+ "Machoke": "1",
240
+ "Machop": "25",
241
+ "Magikarp": "34",
242
+ "Magmar": "86",
243
+ "Magnemite": "13",
244
+ "Magneton": "94",
245
+ "Mankey": "146",
246
+ "Marowak": "36",
247
+ "Meowth": "149",
248
+ "Metapod": "117",
249
+ "Mew": "124",
250
+ "Mewtwo": "39",
251
+ "Moltres": "84",
252
+ "MrMime": "60",
253
+ "Muk": "135",
254
+ "Nidoking": "98",
255
+ "Nidoqueen": "85",
256
+ "Nidorina": "57",
257
+ "Nidorino": "21",
258
+ "Ninetales": "121",
259
+ "Oddish": "73",
260
+ "Omanyte": "128",
261
+ "Omastar": "2",
262
+ "Onix": "87",
263
+ "Paras": "67",
264
+ "Parasect": "28",
265
+ "Persian": "6",
266
+ "Pidgeot": "20",
267
+ "Pidgeotto": "37",
268
+ "Pidgey": "61",
269
+ "Pikachu": "130",
270
+ "Pinsir": "125",
271
+ "Poliwag": "79",
272
+ "Poliwhirl": "141",
273
+ "Poliwrath": "50",
274
+ "Ponyta": "83",
275
+ "Porygon": "22",
276
+ "Primeape": "112",
277
+ "Psyduck": "32",
278
+ "Raichu": "10",
279
+ "Rapidash": "18",
280
+ "Raticate": "12",
281
+ "Rattata": "24",
282
+ "Rhydon": "123",
283
+ "Rhyhorn": "47",
284
+ "Sandshrew": "58",
285
+ "Sandslash": "114",
286
+ "Scyther": "115",
287
+ "Seadra": "134",
288
+ "Seaking": "143",
289
+ "Seel": "42",
290
+ "Shellder": "38",
291
+ "Slowbro": "27",
292
+ "Slowpoke": "102",
293
+ "Snorlax": "66",
294
+ "Spearow": "136",
295
+ "Squirtle": "81",
296
+ "Starmie": "30",
297
+ "Staryu": "31",
298
+ "Tangela": "16",
299
+ "Tauros": "48",
300
+ "Tentacool": "147",
301
+ "Tentacruel": "118",
302
+ "Vaporeon": "107",
303
+ "Venomoth": "132",
304
+ "Venonat": "19",
305
+ "Venusaur": "65",
306
+ "Victreebel": "127",
307
+ "Vileplume": "35",
308
+ "Voltorb": "82",
309
+ "Vulpix": "88",
310
+ "Wartortle": "145",
311
+ "Weedle": "105",
312
+ "Weepinbell": "7",
313
+ "Weezing": "103",
314
+ "Wigglytuff": "74",
315
+ "Zapdos": "11",
316
+ "Zubat": "122"
317
+ },
318
+ "layer_norm_eps": 1e-12,
319
+ "model_type": "vit",
320
+ "num_attention_heads": 12,
321
+ "num_channels": 3,
322
+ "num_hidden_layers": 12,
323
+ "patch_size": 16,
324
+ "problem_type": "single_label_classification",
325
+ "qkv_bias": true,
326
+ "torch_dtype": "float32",
327
+ "transformers_version": "4.36.0"
328
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c28dcaa2d2be8d4704cab165c677f25b4270b624587e7633ac36c7a5f8e892a8
3
+ size 343679232
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc1d73444a304e94e5a3a83c6813d1158e4e90f52fc0bf7358b8a2e88e8b8664
3
+ size 687473786
preprocessor_config.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "do_rescale": true,
4
+ "do_resize": true,
5
+ "image_mean": [
6
+ 0.5,
7
+ 0.5,
8
+ 0.5
9
+ ],
10
+ "image_processor_type": "ViTImageProcessor",
11
+ "image_std": [
12
+ 0.5,
13
+ 0.5,
14
+ 0.5
15
+ ],
16
+ "resample": 2,
17
+ "rescale_factor": 0.00392156862745098,
18
+ "size": {
19
+ "height": 224,
20
+ "width": 224
21
+ }
22
+ }
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:871a2b0b3d0be54fee74343292af9b7281a3f77417a97c8b9e2b20f80e215aa5
3
+ size 13990
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c855437d80306b6e4683c73fbc3bbe2978ddcaaa00039979af9c1a9e50f7f25
3
+ size 1064
trainer_state.json ADDED
@@ -0,0 +1,861 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9240393208221627,
3
+ "best_model_checkpoint": "pokemon_models\\checkpoint-1750",
4
+ "epoch": 28.0,
5
+ "eval_steps": 500,
6
+ "global_step": 1960,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.29,
13
+ "learning_rate": 5e-06,
14
+ "loss": 5.0145,
15
+ "step": 20
16
+ },
17
+ {
18
+ "epoch": 0.57,
19
+ "learning_rate": 1e-05,
20
+ "loss": 5.0039,
21
+ "step": 40
22
+ },
23
+ {
24
+ "epoch": 0.86,
25
+ "learning_rate": 1.5e-05,
26
+ "loss": 4.9942,
27
+ "step": 60
28
+ },
29
+ {
30
+ "epoch": 1.0,
31
+ "eval_accuracy": 0.01876675603217158,
32
+ "eval_loss": 4.973499298095703,
33
+ "eval_runtime": 102.0829,
34
+ "eval_samples_per_second": 10.962,
35
+ "eval_steps_per_second": 0.686,
36
+ "step": 70
37
+ },
38
+ {
39
+ "epoch": 1.14,
40
+ "learning_rate": 2e-05,
41
+ "loss": 4.97,
42
+ "step": 80
43
+ },
44
+ {
45
+ "epoch": 1.43,
46
+ "learning_rate": 2.5e-05,
47
+ "loss": 4.9313,
48
+ "step": 100
49
+ },
50
+ {
51
+ "epoch": 1.71,
52
+ "learning_rate": 3e-05,
53
+ "loss": 4.893,
54
+ "step": 120
55
+ },
56
+ {
57
+ "epoch": 2.0,
58
+ "learning_rate": 3.5e-05,
59
+ "loss": 4.8374,
60
+ "step": 140
61
+ },
62
+ {
63
+ "epoch": 2.0,
64
+ "eval_accuracy": 0.20196604110813227,
65
+ "eval_loss": 4.816006660461426,
66
+ "eval_runtime": 124.2897,
67
+ "eval_samples_per_second": 9.003,
68
+ "eval_steps_per_second": 0.563,
69
+ "step": 140
70
+ },
71
+ {
72
+ "epoch": 2.29,
73
+ "learning_rate": 4e-05,
74
+ "loss": 4.7329,
75
+ "step": 160
76
+ },
77
+ {
78
+ "epoch": 2.57,
79
+ "learning_rate": 4.5e-05,
80
+ "loss": 4.6472,
81
+ "step": 180
82
+ },
83
+ {
84
+ "epoch": 2.86,
85
+ "learning_rate": 5e-05,
86
+ "loss": 4.541,
87
+ "step": 200
88
+ },
89
+ {
90
+ "epoch": 3.0,
91
+ "eval_accuracy": 0.5495978552278821,
92
+ "eval_loss": 4.4448018074035645,
93
+ "eval_runtime": 101.0357,
94
+ "eval_samples_per_second": 11.075,
95
+ "eval_steps_per_second": 0.693,
96
+ "step": 210
97
+ },
98
+ {
99
+ "epoch": 3.14,
100
+ "learning_rate": 4.9444444444444446e-05,
101
+ "loss": 4.4117,
102
+ "step": 220
103
+ },
104
+ {
105
+ "epoch": 3.43,
106
+ "learning_rate": 4.888888888888889e-05,
107
+ "loss": 4.2454,
108
+ "step": 240
109
+ },
110
+ {
111
+ "epoch": 3.71,
112
+ "learning_rate": 4.8333333333333334e-05,
113
+ "loss": 4.1227,
114
+ "step": 260
115
+ },
116
+ {
117
+ "epoch": 4.0,
118
+ "learning_rate": 4.7777777777777784e-05,
119
+ "loss": 4.0198,
120
+ "step": 280
121
+ },
122
+ {
123
+ "epoch": 4.0,
124
+ "eval_accuracy": 0.7042001787310098,
125
+ "eval_loss": 4.0061211585998535,
126
+ "eval_runtime": 100.1956,
127
+ "eval_samples_per_second": 11.168,
128
+ "eval_steps_per_second": 0.699,
129
+ "step": 280
130
+ },
131
+ {
132
+ "epoch": 4.29,
133
+ "learning_rate": 4.722222222222222e-05,
134
+ "loss": 3.84,
135
+ "step": 300
136
+ },
137
+ {
138
+ "epoch": 4.57,
139
+ "learning_rate": 4.666666666666667e-05,
140
+ "loss": 3.757,
141
+ "step": 320
142
+ },
143
+ {
144
+ "epoch": 4.86,
145
+ "learning_rate": 4.6111111111111115e-05,
146
+ "loss": 3.6626,
147
+ "step": 340
148
+ },
149
+ {
150
+ "epoch": 5.0,
151
+ "eval_accuracy": 0.7605004468275246,
152
+ "eval_loss": 3.630556106567383,
153
+ "eval_runtime": 100.0509,
154
+ "eval_samples_per_second": 11.184,
155
+ "eval_steps_per_second": 0.7,
156
+ "step": 350
157
+ },
158
+ {
159
+ "epoch": 5.14,
160
+ "learning_rate": 4.555555555555556e-05,
161
+ "loss": 3.5477,
162
+ "step": 360
163
+ },
164
+ {
165
+ "epoch": 5.43,
166
+ "learning_rate": 4.5e-05,
167
+ "loss": 3.3914,
168
+ "step": 380
169
+ },
170
+ {
171
+ "epoch": 5.71,
172
+ "learning_rate": 4.4444444444444447e-05,
173
+ "loss": 3.3164,
174
+ "step": 400
175
+ },
176
+ {
177
+ "epoch": 6.0,
178
+ "learning_rate": 4.388888888888889e-05,
179
+ "loss": 3.2654,
180
+ "step": 420
181
+ },
182
+ {
183
+ "epoch": 6.0,
184
+ "eval_accuracy": 0.7971403038427167,
185
+ "eval_loss": 3.3061511516571045,
186
+ "eval_runtime": 99.8013,
187
+ "eval_samples_per_second": 11.212,
188
+ "eval_steps_per_second": 0.701,
189
+ "step": 420
190
+ },
191
+ {
192
+ "epoch": 6.29,
193
+ "learning_rate": 4.3333333333333334e-05,
194
+ "loss": 3.1041,
195
+ "step": 440
196
+ },
197
+ {
198
+ "epoch": 6.57,
199
+ "learning_rate": 4.277777777777778e-05,
200
+ "loss": 3.0193,
201
+ "step": 460
202
+ },
203
+ {
204
+ "epoch": 6.86,
205
+ "learning_rate": 4.222222222222222e-05,
206
+ "loss": 2.9314,
207
+ "step": 480
208
+ },
209
+ {
210
+ "epoch": 7.0,
211
+ "eval_accuracy": 0.8310991957104558,
212
+ "eval_loss": 2.994609832763672,
213
+ "eval_runtime": 106.5638,
214
+ "eval_samples_per_second": 10.501,
215
+ "eval_steps_per_second": 0.657,
216
+ "step": 490
217
+ },
218
+ {
219
+ "epoch": 7.14,
220
+ "learning_rate": 4.166666666666667e-05,
221
+ "loss": 2.871,
222
+ "step": 500
223
+ },
224
+ {
225
+ "epoch": 7.43,
226
+ "learning_rate": 4.111111111111111e-05,
227
+ "loss": 2.7418,
228
+ "step": 520
229
+ },
230
+ {
231
+ "epoch": 7.71,
232
+ "learning_rate": 4.055555555555556e-05,
233
+ "loss": 2.6542,
234
+ "step": 540
235
+ },
236
+ {
237
+ "epoch": 8.0,
238
+ "learning_rate": 4e-05,
239
+ "loss": 2.5893,
240
+ "step": 560
241
+ },
242
+ {
243
+ "epoch": 8.0,
244
+ "eval_accuracy": 0.8507596067917784,
245
+ "eval_loss": 2.7318336963653564,
246
+ "eval_runtime": 125.4233,
247
+ "eval_samples_per_second": 8.922,
248
+ "eval_steps_per_second": 0.558,
249
+ "step": 560
250
+ },
251
+ {
252
+ "epoch": 8.29,
253
+ "learning_rate": 3.944444444444445e-05,
254
+ "loss": 2.5106,
255
+ "step": 580
256
+ },
257
+ {
258
+ "epoch": 8.57,
259
+ "learning_rate": 3.888888888888889e-05,
260
+ "loss": 2.4358,
261
+ "step": 600
262
+ },
263
+ {
264
+ "epoch": 8.86,
265
+ "learning_rate": 3.8333333333333334e-05,
266
+ "loss": 2.3645,
267
+ "step": 620
268
+ },
269
+ {
270
+ "epoch": 9.0,
271
+ "eval_accuracy": 0.8579088471849866,
272
+ "eval_loss": 2.4826338291168213,
273
+ "eval_runtime": 121.4568,
274
+ "eval_samples_per_second": 9.213,
275
+ "eval_steps_per_second": 0.576,
276
+ "step": 630
277
+ },
278
+ {
279
+ "epoch": 9.14,
280
+ "learning_rate": 3.777777777777778e-05,
281
+ "loss": 2.2831,
282
+ "step": 640
283
+ },
284
+ {
285
+ "epoch": 9.43,
286
+ "learning_rate": 3.722222222222222e-05,
287
+ "loss": 2.2297,
288
+ "step": 660
289
+ },
290
+ {
291
+ "epoch": 9.71,
292
+ "learning_rate": 3.6666666666666666e-05,
293
+ "loss": 2.1367,
294
+ "step": 680
295
+ },
296
+ {
297
+ "epoch": 10.0,
298
+ "learning_rate": 3.611111111111111e-05,
299
+ "loss": 2.0793,
300
+ "step": 700
301
+ },
302
+ {
303
+ "epoch": 10.0,
304
+ "eval_accuracy": 0.871313672922252,
305
+ "eval_loss": 2.245124578475952,
306
+ "eval_runtime": 122.6079,
307
+ "eval_samples_per_second": 9.127,
308
+ "eval_steps_per_second": 0.571,
309
+ "step": 700
310
+ },
311
+ {
312
+ "epoch": 10.29,
313
+ "learning_rate": 3.555555555555556e-05,
314
+ "loss": 1.9796,
315
+ "step": 720
316
+ },
317
+ {
318
+ "epoch": 10.57,
319
+ "learning_rate": 3.5e-05,
320
+ "loss": 1.9471,
321
+ "step": 740
322
+ },
323
+ {
324
+ "epoch": 10.86,
325
+ "learning_rate": 3.444444444444445e-05,
326
+ "loss": 1.8754,
327
+ "step": 760
328
+ },
329
+ {
330
+ "epoch": 11.0,
331
+ "eval_accuracy": 0.871313672922252,
332
+ "eval_loss": 2.060222625732422,
333
+ "eval_runtime": 122.2722,
334
+ "eval_samples_per_second": 9.152,
335
+ "eval_steps_per_second": 0.572,
336
+ "step": 770
337
+ },
338
+ {
339
+ "epoch": 11.14,
340
+ "learning_rate": 3.388888888888889e-05,
341
+ "loss": 1.8259,
342
+ "step": 780
343
+ },
344
+ {
345
+ "epoch": 11.43,
346
+ "learning_rate": 3.3333333333333335e-05,
347
+ "loss": 1.7872,
348
+ "step": 800
349
+ },
350
+ {
351
+ "epoch": 11.71,
352
+ "learning_rate": 3.277777777777778e-05,
353
+ "loss": 1.6884,
354
+ "step": 820
355
+ },
356
+ {
357
+ "epoch": 12.0,
358
+ "learning_rate": 3.222222222222223e-05,
359
+ "loss": 1.6703,
360
+ "step": 840
361
+ },
362
+ {
363
+ "epoch": 12.0,
364
+ "eval_accuracy": 0.8811438784629133,
365
+ "eval_loss": 1.872039556503296,
366
+ "eval_runtime": 98.0421,
367
+ "eval_samples_per_second": 11.413,
368
+ "eval_steps_per_second": 0.714,
369
+ "step": 840
370
+ },
371
+ {
372
+ "epoch": 12.29,
373
+ "learning_rate": 3.1666666666666666e-05,
374
+ "loss": 1.6003,
375
+ "step": 860
376
+ },
377
+ {
378
+ "epoch": 12.57,
379
+ "learning_rate": 3.111111111111111e-05,
380
+ "loss": 1.5433,
381
+ "step": 880
382
+ },
383
+ {
384
+ "epoch": 12.86,
385
+ "learning_rate": 3.055555555555556e-05,
386
+ "loss": 1.5198,
387
+ "step": 900
388
+ },
389
+ {
390
+ "epoch": 13.0,
391
+ "eval_accuracy": 0.8900804289544236,
392
+ "eval_loss": 1.7361352443695068,
393
+ "eval_runtime": 97.0673,
394
+ "eval_samples_per_second": 11.528,
395
+ "eval_steps_per_second": 0.721,
396
+ "step": 910
397
+ },
398
+ {
399
+ "epoch": 13.14,
400
+ "learning_rate": 3e-05,
401
+ "loss": 1.4742,
402
+ "step": 920
403
+ },
404
+ {
405
+ "epoch": 13.43,
406
+ "learning_rate": 2.9444444444444448e-05,
407
+ "loss": 1.3876,
408
+ "step": 940
409
+ },
410
+ {
411
+ "epoch": 13.71,
412
+ "learning_rate": 2.8888888888888888e-05,
413
+ "loss": 1.3603,
414
+ "step": 960
415
+ },
416
+ {
417
+ "epoch": 14.0,
418
+ "learning_rate": 2.8333333333333335e-05,
419
+ "loss": 1.329,
420
+ "step": 980
421
+ },
422
+ {
423
+ "epoch": 14.0,
424
+ "eval_accuracy": 0.900804289544236,
425
+ "eval_loss": 1.563855528831482,
426
+ "eval_runtime": 97.4399,
427
+ "eval_samples_per_second": 11.484,
428
+ "eval_steps_per_second": 0.718,
429
+ "step": 980
430
+ },
431
+ {
432
+ "epoch": 14.29,
433
+ "learning_rate": 2.777777777777778e-05,
434
+ "loss": 1.2523,
435
+ "step": 1000
436
+ },
437
+ {
438
+ "epoch": 14.57,
439
+ "learning_rate": 2.7222222222222223e-05,
440
+ "loss": 1.2747,
441
+ "step": 1020
442
+ },
443
+ {
444
+ "epoch": 14.86,
445
+ "learning_rate": 2.6666666666666667e-05,
446
+ "loss": 1.203,
447
+ "step": 1040
448
+ },
449
+ {
450
+ "epoch": 15.0,
451
+ "eval_accuracy": 0.8927613941018767,
452
+ "eval_loss": 1.4685680866241455,
453
+ "eval_runtime": 96.9819,
454
+ "eval_samples_per_second": 11.538,
455
+ "eval_steps_per_second": 0.722,
456
+ "step": 1050
457
+ },
458
+ {
459
+ "epoch": 15.14,
460
+ "learning_rate": 2.6111111111111114e-05,
461
+ "loss": 1.1697,
462
+ "step": 1060
463
+ },
464
+ {
465
+ "epoch": 15.43,
466
+ "learning_rate": 2.5555555555555554e-05,
467
+ "loss": 1.0943,
468
+ "step": 1080
469
+ },
470
+ {
471
+ "epoch": 15.71,
472
+ "learning_rate": 2.5e-05,
473
+ "loss": 1.0947,
474
+ "step": 1100
475
+ },
476
+ {
477
+ "epoch": 16.0,
478
+ "learning_rate": 2.4444444444444445e-05,
479
+ "loss": 1.104,
480
+ "step": 1120
481
+ },
482
+ {
483
+ "epoch": 16.0,
484
+ "eval_accuracy": 0.8981233243967829,
485
+ "eval_loss": 1.3596620559692383,
486
+ "eval_runtime": 97.1177,
487
+ "eval_samples_per_second": 11.522,
488
+ "eval_steps_per_second": 0.721,
489
+ "step": 1120
490
+ },
491
+ {
492
+ "epoch": 16.29,
493
+ "learning_rate": 2.3888888888888892e-05,
494
+ "loss": 1.0113,
495
+ "step": 1140
496
+ },
497
+ {
498
+ "epoch": 16.57,
499
+ "learning_rate": 2.3333333333333336e-05,
500
+ "loss": 1.0285,
501
+ "step": 1160
502
+ },
503
+ {
504
+ "epoch": 16.86,
505
+ "learning_rate": 2.277777777777778e-05,
506
+ "loss": 0.9682,
507
+ "step": 1180
508
+ },
509
+ {
510
+ "epoch": 17.0,
511
+ "eval_accuracy": 0.8990169794459338,
512
+ "eval_loss": 1.2199994325637817,
513
+ "eval_runtime": 486.7671,
514
+ "eval_samples_per_second": 2.299,
515
+ "eval_steps_per_second": 0.144,
516
+ "step": 1190
517
+ },
518
+ {
519
+ "epoch": 17.14,
520
+ "learning_rate": 2.2222222222222223e-05,
521
+ "loss": 0.9578,
522
+ "step": 1200
523
+ },
524
+ {
525
+ "epoch": 17.43,
526
+ "learning_rate": 2.1666666666666667e-05,
527
+ "loss": 0.9403,
528
+ "step": 1220
529
+ },
530
+ {
531
+ "epoch": 17.71,
532
+ "learning_rate": 2.111111111111111e-05,
533
+ "loss": 0.8924,
534
+ "step": 1240
535
+ },
536
+ {
537
+ "epoch": 18.0,
538
+ "learning_rate": 2.0555555555555555e-05,
539
+ "loss": 0.872,
540
+ "step": 1260
541
+ },
542
+ {
543
+ "epoch": 18.0,
544
+ "eval_accuracy": 0.903485254691689,
545
+ "eval_loss": 1.1389293670654297,
546
+ "eval_runtime": 110.8112,
547
+ "eval_samples_per_second": 10.098,
548
+ "eval_steps_per_second": 0.632,
549
+ "step": 1260
550
+ },
551
+ {
552
+ "epoch": 18.29,
553
+ "learning_rate": 2e-05,
554
+ "loss": 0.8312,
555
+ "step": 1280
556
+ },
557
+ {
558
+ "epoch": 18.57,
559
+ "learning_rate": 1.9444444444444445e-05,
560
+ "loss": 0.8201,
561
+ "step": 1300
562
+ },
563
+ {
564
+ "epoch": 18.86,
565
+ "learning_rate": 1.888888888888889e-05,
566
+ "loss": 0.844,
567
+ "step": 1320
568
+ },
569
+ {
570
+ "epoch": 19.0,
571
+ "eval_accuracy": 0.9124218051831993,
572
+ "eval_loss": 1.0643764734268188,
573
+ "eval_runtime": 109.2391,
574
+ "eval_samples_per_second": 10.244,
575
+ "eval_steps_per_second": 0.641,
576
+ "step": 1330
577
+ },
578
+ {
579
+ "epoch": 19.14,
580
+ "learning_rate": 1.8333333333333333e-05,
581
+ "loss": 0.8116,
582
+ "step": 1340
583
+ },
584
+ {
585
+ "epoch": 19.43,
586
+ "learning_rate": 1.777777777777778e-05,
587
+ "loss": 0.7649,
588
+ "step": 1360
589
+ },
590
+ {
591
+ "epoch": 19.71,
592
+ "learning_rate": 1.7222222222222224e-05,
593
+ "loss": 0.7402,
594
+ "step": 1380
595
+ },
596
+ {
597
+ "epoch": 20.0,
598
+ "learning_rate": 1.6666666666666667e-05,
599
+ "loss": 0.7605,
600
+ "step": 1400
601
+ },
602
+ {
603
+ "epoch": 20.0,
604
+ "eval_accuracy": 0.9088471849865952,
605
+ "eval_loss": 1.0364218950271606,
606
+ "eval_runtime": 108.8495,
607
+ "eval_samples_per_second": 10.28,
608
+ "eval_steps_per_second": 0.643,
609
+ "step": 1400
610
+ },
611
+ {
612
+ "epoch": 20.29,
613
+ "learning_rate": 1.6111111111111115e-05,
614
+ "loss": 0.7156,
615
+ "step": 1420
616
+ },
617
+ {
618
+ "epoch": 20.57,
619
+ "learning_rate": 1.5555555555555555e-05,
620
+ "loss": 0.7109,
621
+ "step": 1440
622
+ },
623
+ {
624
+ "epoch": 20.86,
625
+ "learning_rate": 1.5e-05,
626
+ "loss": 0.7244,
627
+ "step": 1460
628
+ },
629
+ {
630
+ "epoch": 21.0,
631
+ "eval_accuracy": 0.902591599642538,
632
+ "eval_loss": 0.9655722379684448,
633
+ "eval_runtime": 106.989,
634
+ "eval_samples_per_second": 10.459,
635
+ "eval_steps_per_second": 0.654,
636
+ "step": 1470
637
+ },
638
+ {
639
+ "epoch": 21.14,
640
+ "learning_rate": 1.4444444444444444e-05,
641
+ "loss": 0.6925,
642
+ "step": 1480
643
+ },
644
+ {
645
+ "epoch": 21.43,
646
+ "learning_rate": 1.388888888888889e-05,
647
+ "loss": 0.6687,
648
+ "step": 1500
649
+ },
650
+ {
651
+ "epoch": 21.71,
652
+ "learning_rate": 1.3333333333333333e-05,
653
+ "loss": 0.658,
654
+ "step": 1520
655
+ },
656
+ {
657
+ "epoch": 22.0,
658
+ "learning_rate": 1.2777777777777777e-05,
659
+ "loss": 0.6595,
660
+ "step": 1540
661
+ },
662
+ {
663
+ "epoch": 22.0,
664
+ "eval_accuracy": 0.9133154602323503,
665
+ "eval_loss": 0.9125866889953613,
666
+ "eval_runtime": 106.9609,
667
+ "eval_samples_per_second": 10.462,
668
+ "eval_steps_per_second": 0.654,
669
+ "step": 1540
670
+ },
671
+ {
672
+ "epoch": 22.29,
673
+ "learning_rate": 1.2222222222222222e-05,
674
+ "loss": 0.6489,
675
+ "step": 1560
676
+ },
677
+ {
678
+ "epoch": 22.57,
679
+ "learning_rate": 1.1666666666666668e-05,
680
+ "loss": 0.6666,
681
+ "step": 1580
682
+ },
683
+ {
684
+ "epoch": 22.86,
685
+ "learning_rate": 1.1111111111111112e-05,
686
+ "loss": 0.6188,
687
+ "step": 1600
688
+ },
689
+ {
690
+ "epoch": 23.0,
691
+ "eval_accuracy": 0.9142091152815014,
692
+ "eval_loss": 0.8716733455657959,
693
+ "eval_runtime": 107.8489,
694
+ "eval_samples_per_second": 10.376,
695
+ "eval_steps_per_second": 0.649,
696
+ "step": 1610
697
+ },
698
+ {
699
+ "epoch": 23.14,
700
+ "learning_rate": 1.0555555555555555e-05,
701
+ "loss": 0.5869,
702
+ "step": 1620
703
+ },
704
+ {
705
+ "epoch": 23.43,
706
+ "learning_rate": 1e-05,
707
+ "loss": 0.5809,
708
+ "step": 1640
709
+ },
710
+ {
711
+ "epoch": 23.71,
712
+ "learning_rate": 9.444444444444445e-06,
713
+ "loss": 0.5972,
714
+ "step": 1660
715
+ },
716
+ {
717
+ "epoch": 24.0,
718
+ "learning_rate": 8.88888888888889e-06,
719
+ "loss": 0.5726,
720
+ "step": 1680
721
+ },
722
+ {
723
+ "epoch": 24.0,
724
+ "eval_accuracy": 0.9186773905272565,
725
+ "eval_loss": 0.8351905345916748,
726
+ "eval_runtime": 109.1081,
727
+ "eval_samples_per_second": 10.256,
728
+ "eval_steps_per_second": 0.642,
729
+ "step": 1680
730
+ },
731
+ {
732
+ "epoch": 24.29,
733
+ "learning_rate": 8.333333333333334e-06,
734
+ "loss": 0.5562,
735
+ "step": 1700
736
+ },
737
+ {
738
+ "epoch": 24.57,
739
+ "learning_rate": 7.777777777777777e-06,
740
+ "loss": 0.5687,
741
+ "step": 1720
742
+ },
743
+ {
744
+ "epoch": 24.86,
745
+ "learning_rate": 7.222222222222222e-06,
746
+ "loss": 0.5418,
747
+ "step": 1740
748
+ },
749
+ {
750
+ "epoch": 25.0,
751
+ "eval_accuracy": 0.9240393208221627,
752
+ "eval_loss": 0.8053392767906189,
753
+ "eval_runtime": 111.5634,
754
+ "eval_samples_per_second": 10.03,
755
+ "eval_steps_per_second": 0.627,
756
+ "step": 1750
757
+ },
758
+ {
759
+ "epoch": 25.14,
760
+ "learning_rate": 6.666666666666667e-06,
761
+ "loss": 0.5496,
762
+ "step": 1760
763
+ },
764
+ {
765
+ "epoch": 25.43,
766
+ "learning_rate": 6.111111111111111e-06,
767
+ "loss": 0.5354,
768
+ "step": 1780
769
+ },
770
+ {
771
+ "epoch": 25.71,
772
+ "learning_rate": 5.555555555555556e-06,
773
+ "loss": 0.5564,
774
+ "step": 1800
775
+ },
776
+ {
777
+ "epoch": 26.0,
778
+ "learning_rate": 5e-06,
779
+ "loss": 0.579,
780
+ "step": 1820
781
+ },
782
+ {
783
+ "epoch": 26.0,
784
+ "eval_accuracy": 0.9151027703306523,
785
+ "eval_loss": 0.8136078119277954,
786
+ "eval_runtime": 105.9237,
787
+ "eval_samples_per_second": 10.564,
788
+ "eval_steps_per_second": 0.661,
789
+ "step": 1820
790
+ },
791
+ {
792
+ "epoch": 26.29,
793
+ "learning_rate": 4.444444444444445e-06,
794
+ "loss": 0.5322,
795
+ "step": 1840
796
+ },
797
+ {
798
+ "epoch": 26.57,
799
+ "learning_rate": 3.888888888888889e-06,
800
+ "loss": 0.5529,
801
+ "step": 1860
802
+ },
803
+ {
804
+ "epoch": 26.86,
805
+ "learning_rate": 3.3333333333333333e-06,
806
+ "loss": 0.5077,
807
+ "step": 1880
808
+ },
809
+ {
810
+ "epoch": 27.0,
811
+ "eval_accuracy": 0.9168900804289544,
812
+ "eval_loss": 0.7921976447105408,
813
+ "eval_runtime": 105.3647,
814
+ "eval_samples_per_second": 10.62,
815
+ "eval_steps_per_second": 0.664,
816
+ "step": 1890
817
+ },
818
+ {
819
+ "epoch": 27.14,
820
+ "learning_rate": 2.777777777777778e-06,
821
+ "loss": 0.5443,
822
+ "step": 1900
823
+ },
824
+ {
825
+ "epoch": 27.43,
826
+ "learning_rate": 2.2222222222222225e-06,
827
+ "loss": 0.5181,
828
+ "step": 1920
829
+ },
830
+ {
831
+ "epoch": 27.71,
832
+ "learning_rate": 1.6666666666666667e-06,
833
+ "loss": 0.5318,
834
+ "step": 1940
835
+ },
836
+ {
837
+ "epoch": 28.0,
838
+ "learning_rate": 1.1111111111111112e-06,
839
+ "loss": 0.5138,
840
+ "step": 1960
841
+ },
842
+ {
843
+ "epoch": 28.0,
844
+ "eval_accuracy": 0.9133154602323503,
845
+ "eval_loss": 0.7894989252090454,
846
+ "eval_runtime": 107.1689,
847
+ "eval_samples_per_second": 10.441,
848
+ "eval_steps_per_second": 0.653,
849
+ "step": 1960
850
+ }
851
+ ],
852
+ "logging_steps": 20,
853
+ "max_steps": 2000,
854
+ "num_input_tokens_seen": 0,
855
+ "num_train_epochs": 29,
856
+ "save_steps": 500,
857
+ "total_flos": 9.718280747932926e+18,
858
+ "train_batch_size": 16,
859
+ "trial_name": null,
860
+ "trial_params": null
861
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e14f0f49e329a54d9c09dad07e5dc67202bf2d77c30c5b085b61482dc7ee342b
3
+ size 4728