Hugging Face's logo

frank098
/
freewilly2-juniper-lora

PEFT
Model card Files Community
1
frank098 commited on
Commit
fd8cca8
•
1 Parent(s): f6404d7

First model version

Browse files
Files changed (15) hide show
  1. README.md +20 -0
  2. adapter_config.json +26 -0
  3. adapter_model.bin +3 -0
  4. adapter_model/README.md +20 -0
  5. adapter_model/adapter_config.json +26 -0
  6. adapter_model/adapter_model.bin +3 -0
  7. added_tokens.json +3 -0
  8. optimizer.pt +3 -0
  9. rng_state.pth +3 -0
  10. scheduler.pt +3 -0
  11. special_tokens_map.json +6 -0
  12. tokenizer.model +3 -0
  13. tokenizer_config.json +35 -0
  14. trainer_state.json +589 -0
  15. training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: bfloat16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "stabilityai/FreeWilly2",
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": 16.0,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 64,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "o_proj",
18
+ "down_proj",
19
+ "q_proj",
20
+ "gate_proj",
21
+ "v_proj",
22
+ "k_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:67fb9f03c2e8ceb467423fa6504afe0bd79f408642f9f3560f88cfa701f49cf9
3
+ size 1657155077
adapter_model/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: bfloat16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
adapter_model/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "stabilityai/FreeWilly2",
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": 16.0,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 64,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "o_proj",
18
+ "down_proj",
19
+ "q_proj",
20
+ "gate_proj",
21
+ "v_proj",
22
+ "k_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:67fb9f03c2e8ceb467423fa6504afe0bd79f408642f9f3560f88cfa701f49cf9
3
+ size 1657155077
added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "[PAD]": 32000
3
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1e5607144267115a66bb1c6e6be451bee497d11a45475e3e60516854a442b54c
3
+ size 6627702473
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e1a7a574cb4613a14e32d10fb4c7a9f4aa813f17ad3927491ad424a231fb477
3
+ size 14511
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:039c9142d517867ae9b688df123e8552919ac9857c34af19261775a1a789ad77
3
+ size 627
special_tokens_map.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "eos_token": "</s>",
4
+ "pad_token": "[PAD]",
5
+ "unk_token": "<unk>"
6
+ }
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
tokenizer_config.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "bos_token": {
5
+ "__type": "AddedToken",
6
+ "content": "<s>",
7
+ "lstrip": false,
8
+ "normalized": true,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "clean_up_tokenization_spaces": false,
13
+ "eos_token": {
14
+ "__type": "AddedToken",
15
+ "content": "</s>",
16
+ "lstrip": false,
17
+ "normalized": true,
18
+ "rstrip": false,
19
+ "single_word": false
20
+ },
21
+ "legacy": true,
22
+ "model_max_length": 4096,
23
+ "pad_token": null,
24
+ "padding_side": "right",
25
+ "sp_model_kwargs": {},
26
+ "tokenizer_class": "LlamaTokenizer",
27
+ "unk_token": {
28
+ "__type": "AddedToken",
29
+ "content": "<unk>",
30
+ "lstrip": false,
31
+ "normalized": true,
32
+ "rstrip": false,
33
+ "single_word": false
34
+ }
35
+ }
trainer_state.json ADDED
@@ -0,0 +1,589 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.2972972972972974,
5
+ "global_step": 600,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.02,
12
+ "learning_rate": 0.0001,
13
+ "loss": 0.7874,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.04,
18
+ "learning_rate": 0.0001,
19
+ "loss": 0.7241,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.06,
24
+ "learning_rate": 0.0001,
25
+ "loss": 0.6801,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.09,
30
+ "learning_rate": 0.0001,
31
+ "loss": 0.7074,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.11,
36
+ "learning_rate": 0.0001,
37
+ "loss": 1.2215,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.13,
42
+ "learning_rate": 0.0001,
43
+ "loss": 0.6756,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.15,
48
+ "learning_rate": 0.0001,
49
+ "loss": 0.6333,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.17,
54
+ "learning_rate": 0.0001,
55
+ "loss": 0.6226,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.19,
60
+ "learning_rate": 0.0001,
61
+ "loss": 0.6498,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.22,
66
+ "learning_rate": 0.0001,
67
+ "loss": 1.0335,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.24,
72
+ "learning_rate": 0.0001,
73
+ "loss": 0.6274,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.26,
78
+ "learning_rate": 0.0001,
79
+ "loss": 0.5948,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.28,
84
+ "learning_rate": 0.0001,
85
+ "loss": 0.617,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.3,
90
+ "learning_rate": 0.0001,
91
+ "loss": 0.6596,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.32,
96
+ "learning_rate": 0.0001,
97
+ "loss": 1.1614,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.35,
102
+ "learning_rate": 0.0001,
103
+ "loss": 0.6195,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.37,
108
+ "learning_rate": 0.0001,
109
+ "loss": 0.5798,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.39,
114
+ "learning_rate": 0.0001,
115
+ "loss": 0.5892,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.4,
120
+ "eval_loss": 0.6906093955039978,
121
+ "eval_runtime": 1140.1455,
122
+ "eval_samples_per_second": 0.877,
123
+ "eval_steps_per_second": 0.877,
124
+ "step": 187
125
+ },
126
+ {
127
+ "epoch": 0.4,
128
+ "mmlu_eval_accuracy": 0.6468424709524904,
129
+ "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
130
+ "mmlu_eval_accuracy_anatomy": 0.5,
131
+ "mmlu_eval_accuracy_astronomy": 0.8125,
132
+ "mmlu_eval_accuracy_business_ethics": 0.8181818181818182,
133
+ "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034,
134
+ "mmlu_eval_accuracy_college_biology": 0.6875,
135
+ "mmlu_eval_accuracy_college_chemistry": 0.25,
136
+ "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
137
+ "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
138
+ "mmlu_eval_accuracy_college_medicine": 0.6818181818181818,
139
+ "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
140
+ "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
141
+ "mmlu_eval_accuracy_conceptual_physics": 0.6153846153846154,
142
+ "mmlu_eval_accuracy_econometrics": 0.5,
143
+ "mmlu_eval_accuracy_electrical_engineering": 0.6875,
144
+ "mmlu_eval_accuracy_elementary_mathematics": 0.5121951219512195,
145
+ "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
146
+ "mmlu_eval_accuracy_global_facts": 0.5,
147
+ "mmlu_eval_accuracy_high_school_biology": 0.8125,
148
+ "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
149
+ "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
150
+ "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334,
151
+ "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
152
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571,
153
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.6976744186046512,
154
+ "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862,
155
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.8076923076923077,
156
+ "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
157
+ "mmlu_eval_accuracy_high_school_psychology": 0.9166666666666666,
158
+ "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
159
+ "mmlu_eval_accuracy_high_school_us_history": 0.9090909090909091,
160
+ "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693,
161
+ "mmlu_eval_accuracy_human_aging": 0.782608695652174,
162
+ "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
163
+ "mmlu_eval_accuracy_international_law": 0.9230769230769231,
164
+ "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
165
+ "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
166
+ "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
167
+ "mmlu_eval_accuracy_management": 0.9090909090909091,
168
+ "mmlu_eval_accuracy_marketing": 0.88,
169
+ "mmlu_eval_accuracy_medical_genetics": 1.0,
170
+ "mmlu_eval_accuracy_miscellaneous": 0.8023255813953488,
171
+ "mmlu_eval_accuracy_moral_disputes": 0.6842105263157895,
172
+ "mmlu_eval_accuracy_moral_scenarios": 0.47,
173
+ "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
174
+ "mmlu_eval_accuracy_philosophy": 0.7647058823529411,
175
+ "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
176
+ "mmlu_eval_accuracy_professional_accounting": 0.6129032258064516,
177
+ "mmlu_eval_accuracy_professional_law": 0.5470588235294118,
178
+ "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806,
179
+ "mmlu_eval_accuracy_professional_psychology": 0.7101449275362319,
180
+ "mmlu_eval_accuracy_public_relations": 0.5,
181
+ "mmlu_eval_accuracy_security_studies": 0.7037037037037037,
182
+ "mmlu_eval_accuracy_sociology": 0.9545454545454546,
183
+ "mmlu_eval_accuracy_us_foreign_policy": 1.0,
184
+ "mmlu_eval_accuracy_virology": 0.5555555555555556,
185
+ "mmlu_eval_accuracy_world_religions": 0.8947368421052632,
186
+ "mmlu_loss": 1.1459644647311495,
187
+ "step": 187
188
+ },
189
+ {
190
+ "epoch": 0.41,
191
+ "learning_rate": 0.0001,
192
+ "loss": 0.63,
193
+ "step": 190
194
+ },
195
+ {
196
+ "epoch": 0.43,
197
+ "learning_rate": 0.0001,
198
+ "loss": 1.0713,
199
+ "step": 200
200
+ },
201
+ {
202
+ "epoch": 0.45,
203
+ "learning_rate": 0.0001,
204
+ "loss": 0.5899,
205
+ "step": 210
206
+ },
207
+ {
208
+ "epoch": 0.48,
209
+ "learning_rate": 0.0001,
210
+ "loss": 0.5853,
211
+ "step": 220
212
+ },
213
+ {
214
+ "epoch": 0.5,
215
+ "learning_rate": 0.0001,
216
+ "loss": 0.5808,
217
+ "step": 230
218
+ },
219
+ {
220
+ "epoch": 0.52,
221
+ "learning_rate": 0.0001,
222
+ "loss": 0.6217,
223
+ "step": 240
224
+ },
225
+ {
226
+ "epoch": 0.54,
227
+ "learning_rate": 0.0001,
228
+ "loss": 0.9323,
229
+ "step": 250
230
+ },
231
+ {
232
+ "epoch": 0.56,
233
+ "learning_rate": 0.0001,
234
+ "loss": 0.5872,
235
+ "step": 260
236
+ },
237
+ {
238
+ "epoch": 0.58,
239
+ "learning_rate": 0.0001,
240
+ "loss": 0.547,
241
+ "step": 270
242
+ },
243
+ {
244
+ "epoch": 0.61,
245
+ "learning_rate": 0.0001,
246
+ "loss": 0.5627,
247
+ "step": 280
248
+ },
249
+ {
250
+ "epoch": 0.63,
251
+ "learning_rate": 0.0001,
252
+ "loss": 0.6126,
253
+ "step": 290
254
+ },
255
+ {
256
+ "epoch": 0.65,
257
+ "learning_rate": 0.0001,
258
+ "loss": 0.9871,
259
+ "step": 300
260
+ },
261
+ {
262
+ "epoch": 0.67,
263
+ "learning_rate": 0.0001,
264
+ "loss": 0.568,
265
+ "step": 310
266
+ },
267
+ {
268
+ "epoch": 0.69,
269
+ "learning_rate": 0.0001,
270
+ "loss": 0.561,
271
+ "step": 320
272
+ },
273
+ {
274
+ "epoch": 0.71,
275
+ "learning_rate": 0.0001,
276
+ "loss": 0.5675,
277
+ "step": 330
278
+ },
279
+ {
280
+ "epoch": 0.74,
281
+ "learning_rate": 0.0001,
282
+ "loss": 0.5864,
283
+ "step": 340
284
+ },
285
+ {
286
+ "epoch": 0.76,
287
+ "learning_rate": 0.0001,
288
+ "loss": 1.0407,
289
+ "step": 350
290
+ },
291
+ {
292
+ "epoch": 0.78,
293
+ "learning_rate": 0.0001,
294
+ "loss": 0.5774,
295
+ "step": 360
296
+ },
297
+ {
298
+ "epoch": 0.8,
299
+ "learning_rate": 0.0001,
300
+ "loss": 0.5696,
301
+ "step": 370
302
+ },
303
+ {
304
+ "epoch": 0.81,
305
+ "eval_loss": 0.6527115702629089,
306
+ "eval_runtime": 1139.7605,
307
+ "eval_samples_per_second": 0.877,
308
+ "eval_steps_per_second": 0.877,
309
+ "step": 374
310
+ },
311
+ {
312
+ "epoch": 0.81,
313
+ "mmlu_eval_accuracy": 0.651057156408483,
314
+ "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
315
+ "mmlu_eval_accuracy_anatomy": 0.5,
316
+ "mmlu_eval_accuracy_astronomy": 0.8125,
317
+ "mmlu_eval_accuracy_business_ethics": 0.8181818181818182,
318
+ "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034,
319
+ "mmlu_eval_accuracy_college_biology": 0.625,
320
+ "mmlu_eval_accuracy_college_chemistry": 0.25,
321
+ "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
322
+ "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
323
+ "mmlu_eval_accuracy_college_medicine": 0.6818181818181818,
324
+ "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
325
+ "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
326
+ "mmlu_eval_accuracy_conceptual_physics": 0.6153846153846154,
327
+ "mmlu_eval_accuracy_econometrics": 0.5,
328
+ "mmlu_eval_accuracy_electrical_engineering": 0.6875,
329
+ "mmlu_eval_accuracy_elementary_mathematics": 0.5121951219512195,
330
+ "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
331
+ "mmlu_eval_accuracy_global_facts": 0.5,
332
+ "mmlu_eval_accuracy_high_school_biology": 0.78125,
333
+ "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
334
+ "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
335
+ "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334,
336
+ "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
337
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571,
338
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.6976744186046512,
339
+ "mmlu_eval_accuracy_high_school_mathematics": 0.4482758620689655,
340
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.8076923076923077,
341
+ "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
342
+ "mmlu_eval_accuracy_high_school_psychology": 0.95,
343
+ "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
344
+ "mmlu_eval_accuracy_high_school_us_history": 0.9090909090909091,
345
+ "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693,
346
+ "mmlu_eval_accuracy_human_aging": 0.782608695652174,
347
+ "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
348
+ "mmlu_eval_accuracy_international_law": 0.9230769230769231,
349
+ "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
350
+ "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
351
+ "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
352
+ "mmlu_eval_accuracy_management": 0.9090909090909091,
353
+ "mmlu_eval_accuracy_marketing": 0.88,
354
+ "mmlu_eval_accuracy_medical_genetics": 1.0,
355
+ "mmlu_eval_accuracy_miscellaneous": 0.8023255813953488,
356
+ "mmlu_eval_accuracy_moral_disputes": 0.6842105263157895,
357
+ "mmlu_eval_accuracy_moral_scenarios": 0.49,
358
+ "mmlu_eval_accuracy_nutrition": 0.696969696969697,
359
+ "mmlu_eval_accuracy_philosophy": 0.7647058823529411,
360
+ "mmlu_eval_accuracy_prehistory": 0.6,
361
+ "mmlu_eval_accuracy_professional_accounting": 0.6451612903225806,
362
+ "mmlu_eval_accuracy_professional_law": 0.5529411764705883,
363
+ "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806,
364
+ "mmlu_eval_accuracy_professional_psychology": 0.7101449275362319,
365
+ "mmlu_eval_accuracy_public_relations": 0.5,
366
+ "mmlu_eval_accuracy_security_studies": 0.6666666666666666,
367
+ "mmlu_eval_accuracy_sociology": 0.9545454545454546,
368
+ "mmlu_eval_accuracy_us_foreign_policy": 1.0,
369
+ "mmlu_eval_accuracy_virology": 0.5555555555555556,
370
+ "mmlu_eval_accuracy_world_religions": 0.8947368421052632,
371
+ "mmlu_loss": 1.1366865825197574,
372
+ "step": 374
373
+ },
374
+ {
375
+ "epoch": 0.82,
376
+ "learning_rate": 0.0001,
377
+ "loss": 0.5626,
378
+ "step": 380
379
+ },
380
+ {
381
+ "epoch": 0.84,
382
+ "learning_rate": 0.0001,
383
+ "loss": 0.597,
384
+ "step": 390
385
+ },
386
+ {
387
+ "epoch": 0.86,
388
+ "learning_rate": 0.0001,
389
+ "loss": 0.9613,
390
+ "step": 400
391
+ },
392
+ {
393
+ "epoch": 0.89,
394
+ "learning_rate": 0.0001,
395
+ "loss": 0.5486,
396
+ "step": 410
397
+ },
398
+ {
399
+ "epoch": 0.91,
400
+ "learning_rate": 0.0001,
401
+ "loss": 0.5496,
402
+ "step": 420
403
+ },
404
+ {
405
+ "epoch": 0.93,
406
+ "learning_rate": 0.0001,
407
+ "loss": 0.5468,
408
+ "step": 430
409
+ },
410
+ {
411
+ "epoch": 0.95,
412
+ "learning_rate": 0.0001,
413
+ "loss": 0.5914,
414
+ "step": 440
415
+ },
416
+ {
417
+ "epoch": 0.97,
418
+ "learning_rate": 0.0001,
419
+ "loss": 0.9408,
420
+ "step": 450
421
+ },
422
+ {
423
+ "epoch": 0.99,
424
+ "learning_rate": 0.0001,
425
+ "loss": 0.5851,
426
+ "step": 460
427
+ },
428
+ {
429
+ "epoch": 1.02,
430
+ "learning_rate": 0.0001,
431
+ "loss": 0.6053,
432
+ "step": 470
433
+ },
434
+ {
435
+ "epoch": 1.04,
436
+ "learning_rate": 0.0001,
437
+ "loss": 0.4955,
438
+ "step": 480
439
+ },
440
+ {
441
+ "epoch": 1.06,
442
+ "learning_rate": 0.0001,
443
+ "loss": 0.5032,
444
+ "step": 490
445
+ },
446
+ {
447
+ "epoch": 1.08,
448
+ "learning_rate": 0.0001,
449
+ "loss": 0.5175,
450
+ "step": 500
451
+ },
452
+ {
453
+ "epoch": 1.1,
454
+ "learning_rate": 0.0001,
455
+ "loss": 0.6447,
456
+ "step": 510
457
+ },
458
+ {
459
+ "epoch": 1.12,
460
+ "learning_rate": 0.0001,
461
+ "loss": 0.6054,
462
+ "step": 520
463
+ },
464
+ {
465
+ "epoch": 1.15,
466
+ "learning_rate": 0.0001,
467
+ "loss": 0.5034,
468
+ "step": 530
469
+ },
470
+ {
471
+ "epoch": 1.17,
472
+ "learning_rate": 0.0001,
473
+ "loss": 0.5105,
474
+ "step": 540
475
+ },
476
+ {
477
+ "epoch": 1.19,
478
+ "learning_rate": 0.0001,
479
+ "loss": 0.5272,
480
+ "step": 550
481
+ },
482
+ {
483
+ "epoch": 1.21,
484
+ "learning_rate": 0.0001,
485
+ "loss": 0.6821,
486
+ "step": 560
487
+ },
488
+ {
489
+ "epoch": 1.21,
490
+ "eval_loss": 0.6524714827537537,
491
+ "eval_runtime": 1139.5177,
492
+ "eval_samples_per_second": 0.878,
493
+ "eval_steps_per_second": 0.878,
494
+ "step": 561
495
+ },
496
+ {
497
+ "epoch": 1.21,
498
+ "mmlu_eval_accuracy": 0.657632739983987,
499
+ "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
500
+ "mmlu_eval_accuracy_anatomy": 0.5,
501
+ "mmlu_eval_accuracy_astronomy": 0.8125,
502
+ "mmlu_eval_accuracy_business_ethics": 0.8181818181818182,
503
+ "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034,
504
+ "mmlu_eval_accuracy_college_biology": 0.6875,
505
+ "mmlu_eval_accuracy_college_chemistry": 0.25,
506
+ "mmlu_eval_accuracy_college_computer_science": 0.5454545454545454,
507
+ "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
508
+ "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
509
+ "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
510
+ "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
511
+ "mmlu_eval_accuracy_conceptual_physics": 0.6153846153846154,
512
+ "mmlu_eval_accuracy_econometrics": 0.5,
513
+ "mmlu_eval_accuracy_electrical_engineering": 0.6875,
514
+ "mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805,
515
+ "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
516
+ "mmlu_eval_accuracy_global_facts": 0.5,
517
+ "mmlu_eval_accuracy_high_school_biology": 0.8125,
518
+ "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
519
+ "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778,
520
+ "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334,
521
+ "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
522
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.9047619047619048,
523
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.6976744186046512,
524
+ "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862,
525
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.8076923076923077,
526
+ "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
527
+ "mmlu_eval_accuracy_high_school_psychology": 0.9333333333333333,
528
+ "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
529
+ "mmlu_eval_accuracy_high_school_us_history": 0.9090909090909091,
530
+ "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693,
531
+ "mmlu_eval_accuracy_human_aging": 0.782608695652174,
532
+ "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
533
+ "mmlu_eval_accuracy_international_law": 0.9230769230769231,
534
+ "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
535
+ "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
536
+ "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
537
+ "mmlu_eval_accuracy_management": 0.9090909090909091,
538
+ "mmlu_eval_accuracy_marketing": 0.88,
539
+ "mmlu_eval_accuracy_medical_genetics": 1.0,
540
+ "mmlu_eval_accuracy_miscellaneous": 0.7906976744186046,
541
+ "mmlu_eval_accuracy_moral_disputes": 0.6842105263157895,
542
+ "mmlu_eval_accuracy_moral_scenarios": 0.5,
543
+ "mmlu_eval_accuracy_nutrition": 0.696969696969697,
544
+ "mmlu_eval_accuracy_philosophy": 0.7647058823529411,
545
+ "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
546
+ "mmlu_eval_accuracy_professional_accounting": 0.6451612903225806,
547
+ "mmlu_eval_accuracy_professional_law": 0.5411764705882353,
548
+ "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806,
549
+ "mmlu_eval_accuracy_professional_psychology": 0.7101449275362319,
550
+ "mmlu_eval_accuracy_public_relations": 0.5,
551
+ "mmlu_eval_accuracy_security_studies": 0.6666666666666666,
552
+ "mmlu_eval_accuracy_sociology": 0.9545454545454546,
553
+ "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
554
+ "mmlu_eval_accuracy_virology": 0.5555555555555556,
555
+ "mmlu_eval_accuracy_world_religions": 0.8947368421052632,
556
+ "mmlu_loss": 1.0653009439443233,
557
+ "step": 561
558
+ },
559
+ {
560
+ "epoch": 1.23,
561
+ "learning_rate": 0.0001,
562
+ "loss": 0.6131,
563
+ "step": 570
564
+ },
565
+ {
566
+ "epoch": 1.25,
567
+ "learning_rate": 0.0001,
568
+ "loss": 0.5156,
569
+ "step": 580
570
+ },
571
+ {
572
+ "epoch": 1.28,
573
+ "learning_rate": 0.0001,
574
+ "loss": 0.5091,
575
+ "step": 590
576
+ },
577
+ {
578
+ "epoch": 1.3,
579
+ "learning_rate": 0.0001,
580
+ "loss": 0.5251,
581
+ "step": 600
582
+ }
583
+ ],
584
+ "max_steps": 1875,
585
+ "num_train_epochs": 5,
586
+ "total_flos": 8.191898542915584e+17,
587
+ "trial_name": null,
588
+ "trial_params": null
589
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:18da09b5796fc6ca4550f26773744da36b0aca63c7ae8ac96f04778bfc6a6db1
3
+ size 5755