Hugging Face's logo

Anacle
/
codellama-13-lora

Model card Files
xet
Community
thientananacle commited on
Commit
4f85170
·
1 Parent(s): 344831e

Upload lora

Browse files
Files changed (18) hide show
  1. lora-out/README.md +47 -0
  2. lora-out/adapter_config.json +26 -0
  3. lora-out/adapter_model.bin +3 -0
  4. lora-out/checkpoint-146/README.md +21 -0
  5. lora-out/checkpoint-146/adapter_config.json +26 -0
  6. lora-out/checkpoint-146/adapter_model.bin +3 -0
  7. lora-out/checkpoint-146/adapter_model/README.md +21 -0
  8. lora-out/checkpoint-146/adapter_model/adapter_config.json +26 -0
  9. lora-out/checkpoint-146/adapter_model/adapter_model.bin +3 -0
  10. lora-out/checkpoint-146/optimizer.pt +3 -0
  11. lora-out/checkpoint-146/rng_state.pth +3 -0
  12. lora-out/checkpoint-146/scheduler.pt +3 -0
  13. lora-out/checkpoint-146/trainer_state.json +951 -0
  14. lora-out/checkpoint-146/training_args.bin +3 -0
  15. lora-out/config.json +38 -0
  16. lora-out/special_tokens_map.json +12 -0
  17. lora-out/tokenizer.model +3 -0
  18. lora-out/tokenizer_config.json +49 -0
lora-out/README.md ADDED
@@ -0,0 +1,47 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: True
10
+ - load_in_4bit: False
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: fp4
16
+ - bnb_4bit_use_double_quant: False
17
+ - bnb_4bit_compute_dtype: float32
18
+
19
+ The following `bitsandbytes` quantization config was used during training:
20
+ - quant_method: bitsandbytes
21
+ - load_in_8bit: False
22
+ - load_in_4bit: True
23
+ - llm_int8_threshold: 6.0
24
+ - llm_int8_skip_modules: None
25
+ - llm_int8_enable_fp32_cpu_offload: False
26
+ - llm_int8_has_fp16_weight: False
27
+ - bnb_4bit_quant_type: fp4
28
+ - bnb_4bit_use_double_quant: False
29
+ - bnb_4bit_compute_dtype: float32
30
+
31
+ The following `bitsandbytes` quantization config was used during training:
32
+ - quant_method: bitsandbytes
33
+ - load_in_8bit: False
34
+ - load_in_4bit: True
35
+ - llm_int8_threshold: 6.0
36
+ - llm_int8_skip_modules: None
37
+ - llm_int8_enable_fp32_cpu_offload: False
38
+ - llm_int8_has_fp16_weight: False
39
+ - bnb_4bit_quant_type: fp4
40
+ - bnb_4bit_use_double_quant: False
41
+ - bnb_4bit_compute_dtype: float32
42
+ ### Framework versions
43
+
44
+ - PEFT 0.6.0.dev0
45
+ - PEFT 0.6.0.dev0
46
+
47
+ - PEFT 0.6.0.dev0
lora-out/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "codellama/CodeLlama-13b-instruct-hf",
4
+ "bias": "none",
5
+ "fan_in_fan_out": null,
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": 32,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "up_proj",
19
+ "k_proj",
20
+ "v_proj",
21
+ "gate_proj",
22
+ "down_proj",
23
+ "o_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
lora-out/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:27f64ff321e93cf187fbf49d66c2476f9195f37980242fd785a75c72e79073f7
3
+ size 500897101
lora-out/checkpoint-146/README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: True
10
+ - load_in_4bit: False
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: fp4
16
+ - bnb_4bit_use_double_quant: False
17
+ - bnb_4bit_compute_dtype: float32
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.6.0.dev0
lora-out/checkpoint-146/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "codellama/CodeLlama-13b-instruct-hf",
4
+ "bias": "none",
5
+ "fan_in_fan_out": null,
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": 32,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "up_proj",
19
+ "k_proj",
20
+ "v_proj",
21
+ "gate_proj",
22
+ "down_proj",
23
+ "o_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
lora-out/checkpoint-146/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:27f64ff321e93cf187fbf49d66c2476f9195f37980242fd785a75c72e79073f7
3
+ size 500897101
lora-out/checkpoint-146/adapter_model/README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: True
10
+ - load_in_4bit: False
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: fp4
16
+ - bnb_4bit_use_double_quant: False
17
+ - bnb_4bit_compute_dtype: float32
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.6.0.dev0
lora-out/checkpoint-146/adapter_model/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "codellama/CodeLlama-13b-instruct-hf",
4
+ "bias": "none",
5
+ "fan_in_fan_out": null,
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": 32,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "up_proj",
19
+ "k_proj",
20
+ "v_proj",
21
+ "gate_proj",
22
+ "down_proj",
23
+ "o_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
lora-out/checkpoint-146/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:27f64ff321e93cf187fbf49d66c2476f9195f37980242fd785a75c72e79073f7
3
+ size 500897101
lora-out/checkpoint-146/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:51002e844b44b582daf83870375f3bad166e83684618cb46238c85cbd1915e15
3
+ size 251484639
lora-out/checkpoint-146/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b7c85f72790be31c8aacf8fb119456b193f1d588861d6c785047eff11446558
3
+ size 14575
lora-out/checkpoint-146/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aec3c659b188995d40fb2bc37652b4347ec6468d71c5abb38d3b7edd8ac59ca5
3
+ size 627
lora-out/checkpoint-146/trainer_state.json ADDED
@@ -0,0 +1,951 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9965870307167235,
5
+ "eval_steps": 20,
6
+ "global_step": 146,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 2e-05,
14
+ "loss": 0.5157,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.01,
19
+ "learning_rate": 4e-05,
20
+ "loss": 0.4922,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.02,
25
+ "learning_rate": 6e-05,
26
+ "loss": 0.5314,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.03,
31
+ "learning_rate": 8e-05,
32
+ "loss": 0.4443,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.03,
37
+ "learning_rate": 0.0001,
38
+ "loss": 0.5018,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.04,
43
+ "learning_rate": 0.00012,
44
+ "loss": 0.459,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.05,
49
+ "learning_rate": 0.00014,
50
+ "loss": 0.4406,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.05,
55
+ "learning_rate": 0.00016,
56
+ "loss": 0.4045,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.06,
61
+ "learning_rate": 0.00018,
62
+ "loss": 0.4072,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.07,
67
+ "learning_rate": 0.0002,
68
+ "loss": 0.2343,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.08,
73
+ "learning_rate": 0.00019997332081116373,
74
+ "loss": 0.2335,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.08,
79
+ "learning_rate": 0.00019989329748023725,
80
+ "loss": 0.2649,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.09,
85
+ "learning_rate": 0.0001997599727063717,
86
+ "loss": 0.1558,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.1,
91
+ "learning_rate": 0.00019957341762950344,
92
+ "loss": 0.2704,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.1,
97
+ "learning_rate": 0.00019933373179239502,
98
+ "loss": 0.2089,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.11,
103
+ "learning_rate": 0.0001990410430875205,
104
+ "loss": 0.2987,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.12,
109
+ "learning_rate": 0.00019869550768882455,
110
+ "loss": 0.2292,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 0.12,
115
+ "learning_rate": 0.0001982973099683902,
116
+ "loss": 0.4453,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 0.13,
121
+ "learning_rate": 0.0001978466623980609,
122
+ "loss": 0.3501,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 0.14,
127
+ "learning_rate": 0.0001973438054360693,
128
+ "loss": 0.1394,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 0.14,
133
+ "eval_loss": 0.21920710802078247,
134
+ "eval_runtime": 3.7615,
135
+ "eval_samples_per_second": 2.659,
136
+ "eval_steps_per_second": 2.659,
137
+ "step": 20
138
+ },
139
+ {
140
+ "epoch": 0.14,
141
+ "learning_rate": 0.00019678900739873226,
142
+ "loss": 0.1803,
143
+ "step": 21
144
+ },
145
+ {
146
+ "epoch": 0.15,
147
+ "learning_rate": 0.00019618256431728194,
148
+ "loss": 0.4044,
149
+ "step": 22
150
+ },
151
+ {
152
+ "epoch": 0.16,
153
+ "learning_rate": 0.000195524799779908,
154
+ "loss": 0.213,
155
+ "step": 23
156
+ },
157
+ {
158
+ "epoch": 0.16,
159
+ "learning_rate": 0.0001948160647590966,
160
+ "loss": 0.1978,
161
+ "step": 24
162
+ },
163
+ {
164
+ "epoch": 0.17,
165
+ "learning_rate": 0.00019405673742435678,
166
+ "loss": 0.179,
167
+ "step": 25
168
+ },
169
+ {
170
+ "epoch": 0.18,
171
+ "learning_rate": 0.00019324722294043558,
172
+ "loss": 0.1895,
173
+ "step": 26
174
+ },
175
+ {
176
+ "epoch": 0.18,
177
+ "learning_rate": 0.0001923879532511287,
178
+ "loss": 0.2622,
179
+ "step": 27
180
+ },
181
+ {
182
+ "epoch": 0.19,
183
+ "learning_rate": 0.0001914793868488021,
184
+ "loss": 0.2377,
185
+ "step": 28
186
+ },
187
+ {
188
+ "epoch": 0.2,
189
+ "learning_rate": 0.00019052200852974819,
190
+ "loss": 0.1352,
191
+ "step": 29
192
+ },
193
+ {
194
+ "epoch": 0.2,
195
+ "learning_rate": 0.00018951632913550626,
196
+ "loss": 0.2677,
197
+ "step": 30
198
+ },
199
+ {
200
+ "epoch": 0.21,
201
+ "learning_rate": 0.00018846288528028555,
202
+ "loss": 0.2494,
203
+ "step": 31
204
+ },
205
+ {
206
+ "epoch": 0.22,
207
+ "learning_rate": 0.00018736223906463696,
208
+ "loss": 0.322,
209
+ "step": 32
210
+ },
211
+ {
212
+ "epoch": 0.23,
213
+ "learning_rate": 0.00018621497777552507,
214
+ "loss": 0.1475,
215
+ "step": 33
216
+ },
217
+ {
218
+ "epoch": 0.23,
219
+ "learning_rate": 0.00018502171357296144,
220
+ "loss": 0.3654,
221
+ "step": 34
222
+ },
223
+ {
224
+ "epoch": 0.24,
225
+ "learning_rate": 0.00018378308316336584,
226
+ "loss": 0.3612,
227
+ "step": 35
228
+ },
229
+ {
230
+ "epoch": 0.25,
231
+ "learning_rate": 0.00018249974745983023,
232
+ "loss": 0.2789,
233
+ "step": 36
234
+ },
235
+ {
236
+ "epoch": 0.25,
237
+ "learning_rate": 0.00018117239122946615,
238
+ "loss": 0.2432,
239
+ "step": 37
240
+ },
241
+ {
242
+ "epoch": 0.26,
243
+ "learning_rate": 0.000179801722728024,
244
+ "loss": 0.1477,
245
+ "step": 38
246
+ },
247
+ {
248
+ "epoch": 0.27,
249
+ "learning_rate": 0.00017838847332197938,
250
+ "loss": 0.2012,
251
+ "step": 39
252
+ },
253
+ {
254
+ "epoch": 0.27,
255
+ "learning_rate": 0.00017693339709828792,
256
+ "loss": 0.3422,
257
+ "step": 40
258
+ },
259
+ {
260
+ "epoch": 0.27,
261
+ "eval_loss": 0.17625783383846283,
262
+ "eval_runtime": 3.6753,
263
+ "eval_samples_per_second": 2.721,
264
+ "eval_steps_per_second": 2.721,
265
+ "step": 40
266
+ },
267
+ {
268
+ "epoch": 0.28,
269
+ "learning_rate": 0.0001754372704620164,
270
+ "loss": 0.325,
271
+ "step": 41
272
+ },
273
+ {
274
+ "epoch": 0.29,
275
+ "learning_rate": 0.00017390089172206592,
276
+ "loss": 0.2051,
277
+ "step": 42
278
+ },
279
+ {
280
+ "epoch": 0.29,
281
+ "learning_rate": 0.00017232508066520702,
282
+ "loss": 0.3078,
283
+ "step": 43
284
+ },
285
+ {
286
+ "epoch": 0.3,
287
+ "learning_rate": 0.00017071067811865476,
288
+ "loss": 0.2758,
289
+ "step": 44
290
+ },
291
+ {
292
+ "epoch": 0.31,
293
+ "learning_rate": 0.00016905854550141716,
294
+ "loss": 0.2596,
295
+ "step": 45
296
+ },
297
+ {
298
+ "epoch": 0.31,
299
+ "learning_rate": 0.00016736956436465573,
300
+ "loss": 0.299,
301
+ "step": 46
302
+ },
303
+ {
304
+ "epoch": 0.32,
305
+ "learning_rate": 0.00016564463592130428,
306
+ "loss": 0.3817,
307
+ "step": 47
308
+ },
309
+ {
310
+ "epoch": 0.33,
311
+ "learning_rate": 0.00016388468056519612,
312
+ "loss": 0.1732,
313
+ "step": 48
314
+ },
315
+ {
316
+ "epoch": 0.33,
317
+ "learning_rate": 0.00016209063737995715,
318
+ "loss": 0.3681,
319
+ "step": 49
320
+ },
321
+ {
322
+ "epoch": 0.34,
323
+ "learning_rate": 0.00016026346363792567,
324
+ "loss": 0.1973,
325
+ "step": 50
326
+ },
327
+ {
328
+ "epoch": 0.35,
329
+ "learning_rate": 0.00015840413428936767,
330
+ "loss": 0.2199,
331
+ "step": 51
332
+ },
333
+ {
334
+ "epoch": 0.35,
335
+ "learning_rate": 0.0001565136414422592,
336
+ "loss": 0.1982,
337
+ "step": 52
338
+ },
339
+ {
340
+ "epoch": 0.36,
341
+ "learning_rate": 0.00015459299383291345,
342
+ "loss": 0.574,
343
+ "step": 53
344
+ },
345
+ {
346
+ "epoch": 0.37,
347
+ "learning_rate": 0.0001526432162877356,
348
+ "loss": 0.1955,
349
+ "step": 54
350
+ },
351
+ {
352
+ "epoch": 0.38,
353
+ "learning_rate": 0.00015066534917639195,
354
+ "loss": 0.2212,
355
+ "step": 55
356
+ },
357
+ {
358
+ "epoch": 0.38,
359
+ "learning_rate": 0.00014866044785668563,
360
+ "loss": 0.1718,
361
+ "step": 56
362
+ },
363
+ {
364
+ "epoch": 0.39,
365
+ "learning_rate": 0.0001466295821114348,
366
+ "loss": 0.2307,
367
+ "step": 57
368
+ },
369
+ {
370
+ "epoch": 0.4,
371
+ "learning_rate": 0.00014457383557765386,
372
+ "loss": 0.2641,
373
+ "step": 58
374
+ },
375
+ {
376
+ "epoch": 0.4,
377
+ "learning_rate": 0.0001424943051683422,
378
+ "loss": 0.1767,
379
+ "step": 59
380
+ },
381
+ {
382
+ "epoch": 0.41,
383
+ "learning_rate": 0.00014039210048718949,
384
+ "loss": 0.1949,
385
+ "step": 60
386
+ },
387
+ {
388
+ "epoch": 0.41,
389
+ "eval_loss": 0.1635545790195465,
390
+ "eval_runtime": 3.6815,
391
+ "eval_samples_per_second": 2.716,
392
+ "eval_steps_per_second": 2.716,
393
+ "step": 60
394
+ },
395
+ {
396
+ "epoch": 0.42,
397
+ "learning_rate": 0.000138268343236509,
398
+ "loss": 0.1599,
399
+ "step": 61
400
+ },
401
+ {
402
+ "epoch": 0.42,
403
+ "learning_rate": 0.00013612416661871533,
404
+ "loss": 0.3705,
405
+ "step": 62
406
+ },
407
+ {
408
+ "epoch": 0.43,
409
+ "learning_rate": 0.00013396071473166613,
410
+ "loss": 0.2493,
411
+ "step": 63
412
+ },
413
+ {
414
+ "epoch": 0.44,
415
+ "learning_rate": 0.00013177914195819016,
416
+ "loss": 0.1596,
417
+ "step": 64
418
+ },
419
+ {
420
+ "epoch": 0.44,
421
+ "learning_rate": 0.00012958061235012706,
422
+ "loss": 0.2262,
423
+ "step": 65
424
+ },
425
+ {
426
+ "epoch": 0.45,
427
+ "learning_rate": 0.0001273662990072083,
428
+ "loss": 0.2358,
429
+ "step": 66
430
+ },
431
+ {
432
+ "epoch": 0.46,
433
+ "learning_rate": 0.0001251373834511103,
434
+ "loss": 0.2056,
435
+ "step": 67
436
+ },
437
+ {
438
+ "epoch": 0.46,
439
+ "learning_rate": 0.0001228950549950134,
440
+ "loss": 0.1974,
441
+ "step": 68
442
+ },
443
+ {
444
+ "epoch": 0.47,
445
+ "learning_rate": 0.00012064051010900397,
446
+ "loss": 0.1966,
447
+ "step": 69
448
+ },
449
+ {
450
+ "epoch": 0.48,
451
+ "learning_rate": 0.00011837495178165706,
452
+ "loss": 0.2422,
453
+ "step": 70
454
+ },
455
+ {
456
+ "epoch": 0.48,
457
+ "learning_rate": 0.00011609958887814129,
458
+ "loss": 0.1438,
459
+ "step": 71
460
+ },
461
+ {
462
+ "epoch": 0.49,
463
+ "learning_rate": 0.00011381563549518823,
464
+ "loss": 0.2491,
465
+ "step": 72
466
+ },
467
+ {
468
+ "epoch": 0.5,
469
+ "learning_rate": 0.00011152431031326978,
470
+ "loss": 0.2552,
471
+ "step": 73
472
+ },
473
+ {
474
+ "epoch": 0.51,
475
+ "learning_rate": 0.00010922683594633021,
476
+ "loss": 0.2239,
477
+ "step": 74
478
+ },
479
+ {
480
+ "epoch": 0.51,
481
+ "learning_rate": 0.00010692443828941918,
482
+ "loss": 0.2898,
483
+ "step": 75
484
+ },
485
+ {
486
+ "epoch": 0.52,
487
+ "learning_rate": 0.00010461834586457398,
488
+ "loss": 0.2131,
489
+ "step": 76
490
+ },
491
+ {
492
+ "epoch": 0.53,
493
+ "learning_rate": 0.00010230978916530012,
494
+ "loss": 0.1795,
495
+ "step": 77
496
+ },
497
+ {
498
+ "epoch": 0.53,
499
+ "learning_rate": 0.0001,
500
+ "loss": 0.19,
501
+ "step": 78
502
+ },
503
+ {
504
+ "epoch": 0.54,
505
+ "learning_rate": 9.76902108346999e-05,
506
+ "loss": 0.2988,
507
+ "step": 79
508
+ },
509
+ {
510
+ "epoch": 0.55,
511
+ "learning_rate": 9.538165413542607e-05,
512
+ "loss": 0.2314,
513
+ "step": 80
514
+ },
515
+ {
516
+ "epoch": 0.55,
517
+ "eval_loss": 0.15595673024654388,
518
+ "eval_runtime": 3.5994,
519
+ "eval_samples_per_second": 2.778,
520
+ "eval_steps_per_second": 2.778,
521
+ "step": 80
522
+ },
523
+ {
524
+ "epoch": 0.55,
525
+ "learning_rate": 9.307556171058085e-05,
526
+ "loss": 0.1515,
527
+ "step": 81
528
+ },
529
+ {
530
+ "epoch": 0.56,
531
+ "learning_rate": 9.077316405366981e-05,
532
+ "loss": 0.1399,
533
+ "step": 82
534
+ },
535
+ {
536
+ "epoch": 0.57,
537
+ "learning_rate": 8.847568968673026e-05,
538
+ "loss": 0.1124,
539
+ "step": 83
540
+ },
541
+ {
542
+ "epoch": 0.57,
543
+ "learning_rate": 8.61843645048118e-05,
544
+ "loss": 0.3192,
545
+ "step": 84
546
+ },
547
+ {
548
+ "epoch": 0.58,
549
+ "learning_rate": 8.39004111218587e-05,
550
+ "loss": 0.1489,
551
+ "step": 85
552
+ },
553
+ {
554
+ "epoch": 0.59,
555
+ "learning_rate": 8.162504821834295e-05,
556
+ "loss": 0.1377,
557
+ "step": 86
558
+ },
559
+ {
560
+ "epoch": 0.59,
561
+ "learning_rate": 7.935948989099605e-05,
562
+ "loss": 0.2076,
563
+ "step": 87
564
+ },
565
+ {
566
+ "epoch": 0.6,
567
+ "learning_rate": 7.710494500498662e-05,
568
+ "loss": 0.1829,
569
+ "step": 88
570
+ },
571
+ {
572
+ "epoch": 0.61,
573
+ "learning_rate": 7.486261654888973e-05,
574
+ "loss": 0.224,
575
+ "step": 89
576
+ },
577
+ {
578
+ "epoch": 0.61,
579
+ "learning_rate": 7.263370099279172e-05,
580
+ "loss": 0.1991,
581
+ "step": 90
582
+ },
583
+ {
584
+ "epoch": 0.62,
585
+ "learning_rate": 7.041938764987297e-05,
586
+ "loss": 0.1634,
587
+ "step": 91
588
+ },
589
+ {
590
+ "epoch": 0.63,
591
+ "learning_rate": 6.822085804180984e-05,
592
+ "loss": 0.1486,
593
+ "step": 92
594
+ },
595
+ {
596
+ "epoch": 0.63,
597
+ "learning_rate": 6.603928526833387e-05,
598
+ "loss": 0.2025,
599
+ "step": 93
600
+ },
601
+ {
602
+ "epoch": 0.64,
603
+ "learning_rate": 6.387583338128471e-05,
604
+ "loss": 0.1692,
605
+ "step": 94
606
+ },
607
+ {
608
+ "epoch": 0.65,
609
+ "learning_rate": 6.173165676349103e-05,
610
+ "loss": 0.1231,
611
+ "step": 95
612
+ },
613
+ {
614
+ "epoch": 0.66,
615
+ "learning_rate": 5.960789951281052e-05,
616
+ "loss": 0.4388,
617
+ "step": 96
618
+ },
619
+ {
620
+ "epoch": 0.66,
621
+ "learning_rate": 5.750569483165784e-05,
622
+ "loss": 0.3304,
623
+ "step": 97
624
+ },
625
+ {
626
+ "epoch": 0.67,
627
+ "learning_rate": 5.542616442234618e-05,
628
+ "loss": 0.1453,
629
+ "step": 98
630
+ },
631
+ {
632
+ "epoch": 0.68,
633
+ "learning_rate": 5.337041788856518e-05,
634
+ "loss": 0.1316,
635
+ "step": 99
636
+ },
637
+ {
638
+ "epoch": 0.68,
639
+ "learning_rate": 5.1339552143314384e-05,
640
+ "loss": 0.2603,
641
+ "step": 100
642
+ },
643
+ {
644
+ "epoch": 0.68,
645
+ "eval_loss": 0.15206387639045715,
646
+ "eval_runtime": 3.6265,
647
+ "eval_samples_per_second": 2.757,
648
+ "eval_steps_per_second": 2.757,
649
+ "step": 100
650
+ },
651
+ {
652
+ "epoch": 0.69,
653
+ "learning_rate": 4.933465082360807e-05,
654
+ "loss": 0.1104,
655
+ "step": 101
656
+ },
657
+ {
658
+ "epoch": 0.7,
659
+ "learning_rate": 4.735678371226441e-05,
660
+ "loss": 0.2666,
661
+ "step": 102
662
+ },
663
+ {
664
+ "epoch": 0.7,
665
+ "learning_rate": 4.540700616708658e-05,
666
+ "loss": 0.2607,
667
+ "step": 103
668
+ },
669
+ {
670
+ "epoch": 0.71,
671
+ "learning_rate": 4.3486358557740814e-05,
672
+ "loss": 0.1935,
673
+ "step": 104
674
+ },
675
+ {
676
+ "epoch": 0.72,
677
+ "learning_rate": 4.159586571063236e-05,
678
+ "loss": 0.1436,
679
+ "step": 105
680
+ },
681
+ {
682
+ "epoch": 0.72,
683
+ "learning_rate": 3.973653636207437e-05,
684
+ "loss": 0.2068,
685
+ "step": 106
686
+ },
687
+ {
688
+ "epoch": 0.73,
689
+ "learning_rate": 3.7909362620042865e-05,
690
+ "loss": 0.1742,
691
+ "step": 107
692
+ },
693
+ {
694
+ "epoch": 0.74,
695
+ "learning_rate": 3.6115319434803894e-05,
696
+ "loss": 0.2428,
697
+ "step": 108
698
+ },
699
+ {
700
+ "epoch": 0.74,
701
+ "learning_rate": 3.435536407869575e-05,
702
+ "loss": 0.2295,
703
+ "step": 109
704
+ },
705
+ {
706
+ "epoch": 0.75,
707
+ "learning_rate": 3.263043563534428e-05,
708
+ "loss": 0.2146,
709
+ "step": 110
710
+ },
711
+ {
712
+ "epoch": 0.76,
713
+ "learning_rate": 3.094145449858285e-05,
714
+ "loss": 0.2282,
715
+ "step": 111
716
+ },
717
+ {
718
+ "epoch": 0.76,
719
+ "learning_rate": 2.9289321881345254e-05,
720
+ "loss": 0.2215,
721
+ "step": 112
722
+ },
723
+ {
724
+ "epoch": 0.77,
725
+ "learning_rate": 2.7674919334793035e-05,
726
+ "loss": 0.2211,
727
+ "step": 113
728
+ },
729
+ {
730
+ "epoch": 0.78,
731
+ "learning_rate": 2.6099108277934103e-05,
732
+ "loss": 0.3357,
733
+ "step": 114
734
+ },
735
+ {
736
+ "epoch": 0.78,
737
+ "learning_rate": 2.456272953798361e-05,
738
+ "loss": 0.1609,
739
+ "step": 115
740
+ },
741
+ {
742
+ "epoch": 0.79,
743
+ "learning_rate": 2.3066602901712108e-05,
744
+ "loss": 0.1864,
745
+ "step": 116
746
+ },
747
+ {
748
+ "epoch": 0.8,
749
+ "learning_rate": 2.1611526678020654e-05,
750
+ "loss": 0.2003,
751
+ "step": 117
752
+ },
753
+ {
754
+ "epoch": 0.81,
755
+ "learning_rate": 2.0198277271976052e-05,
756
+ "loss": 0.1861,
757
+ "step": 118
758
+ },
759
+ {
760
+ "epoch": 0.81,
761
+ "learning_rate": 1.882760877053388e-05,
762
+ "loss": 0.2372,
763
+ "step": 119
764
+ },
765
+ {
766
+ "epoch": 0.82,
767
+ "learning_rate": 1.750025254016978e-05,
768
+ "loss": 0.169,
769
+ "step": 120
770
+ },
771
+ {
772
+ "epoch": 0.82,
773
+ "eval_loss": 0.14629685878753662,
774
+ "eval_runtime": 3.7136,
775
+ "eval_samples_per_second": 2.693,
776
+ "eval_steps_per_second": 2.693,
777
+ "step": 120
778
+ },
779
+ {
780
+ "epoch": 0.83,
781
+ "learning_rate": 1.621691683663418e-05,
782
+ "loss": 0.1814,
783
+ "step": 121
784
+ },
785
+ {
786
+ "epoch": 0.83,
787
+ "learning_rate": 1.4978286427038601e-05,
788
+ "loss": 0.2888,
789
+ "step": 122
790
+ },
791
+ {
792
+ "epoch": 0.84,
793
+ "learning_rate": 1.3785022224474942e-05,
794
+ "loss": 0.1635,
795
+ "step": 123
796
+ },
797
+ {
798
+ "epoch": 0.85,
799
+ "learning_rate": 1.2637760935363053e-05,
800
+ "loss": 0.2197,
801
+ "step": 124
802
+ },
803
+ {
804
+ "epoch": 0.85,
805
+ "learning_rate": 1.1537114719714481e-05,
806
+ "loss": 0.1754,
807
+ "step": 125
808
+ },
809
+ {
810
+ "epoch": 0.86,
811
+ "learning_rate": 1.0483670864493778e-05,
812
+ "loss": 0.2,
813
+ "step": 126
814
+ },
815
+ {
816
+ "epoch": 0.87,
817
+ "learning_rate": 9.477991470251791e-06,
818
+ "loss": 0.2091,
819
+ "step": 127
820
+ },
821
+ {
822
+ "epoch": 0.87,
823
+ "learning_rate": 8.520613151197898e-06,
824
+ "loss": 0.1672,
825
+ "step": 128
826
+ },
827
+ {
828
+ "epoch": 0.88,
829
+ "learning_rate": 7.612046748871327e-06,
830
+ "loss": 0.1932,
831
+ "step": 129
832
+ },
833
+ {
834
+ "epoch": 0.89,
835
+ "learning_rate": 6.75277705956443e-06,
836
+ "loss": 0.1514,
837
+ "step": 130
838
+ },
839
+ {
840
+ "epoch": 0.89,
841
+ "learning_rate": 5.943262575643238e-06,
842
+ "loss": 0.1766,
843
+ "step": 131
844
+ },
845
+ {
846
+ "epoch": 0.9,
847
+ "learning_rate": 5.183935240903414e-06,
848
+ "loss": 0.2772,
849
+ "step": 132
850
+ },
851
+ {
852
+ "epoch": 0.91,
853
+ "learning_rate": 4.475200220092002e-06,
854
+ "loss": 0.2011,
855
+ "step": 133
856
+ },
857
+ {
858
+ "epoch": 0.91,
859
+ "learning_rate": 3.817435682718096e-06,
860
+ "loss": 0.1981,
861
+ "step": 134
862
+ },
863
+ {
864
+ "epoch": 0.92,
865
+ "learning_rate": 3.2109926012677484e-06,
866
+ "loss": 0.2357,
867
+ "step": 135
868
+ },
869
+ {
870
+ "epoch": 0.93,
871
+ "learning_rate": 2.656194563930714e-06,
872
+ "loss": 0.2335,
873
+ "step": 136
874
+ },
875
+ {
876
+ "epoch": 0.94,
877
+ "learning_rate": 2.1533376019391094e-06,
878
+ "loss": 0.205,
879
+ "step": 137
880
+ },
881
+ {
882
+ "epoch": 0.94,
883
+ "learning_rate": 1.7026900316098215e-06,
884
+ "loss": 0.159,
885
+ "step": 138
886
+ },
887
+ {
888
+ "epoch": 0.95,
889
+ "learning_rate": 1.3044923111754427e-06,
890
+ "loss": 0.4785,
891
+ "step": 139
892
+ },
893
+ {
894
+ "epoch": 0.96,
895
+ "learning_rate": 9.589569124794916e-07,
896
+ "loss": 0.1521,
897
+ "step": 140
898
+ },
899
+ {
900
+ "epoch": 0.96,
901
+ "eval_loss": 0.1433035135269165,
902
+ "eval_runtime": 3.6678,
903
+ "eval_samples_per_second": 2.726,
904
+ "eval_steps_per_second": 2.726,
905
+ "step": 140
906
+ },
907
+ {
908
+ "epoch": 0.96,
909
+ "learning_rate": 6.662682076050031e-07,
910
+ "loss": 0.3414,
911
+ "step": 141
912
+ },
913
+ {
914
+ "epoch": 0.97,
915
+ "learning_rate": 4.2658237049655323e-07,
916
+ "loss": 0.149,
917
+ "step": 142
918
+ },
919
+ {
920
+ "epoch": 0.98,
921
+ "learning_rate": 2.400272936283088e-07,
922
+ "loss": 0.2175,
923
+ "step": 143
924
+ },
925
+ {
926
+ "epoch": 0.98,
927
+ "learning_rate": 1.0670251976275803e-07,
928
+ "loss": 0.141,
929
+ "step": 144
930
+ },
931
+ {
932
+ "epoch": 0.99,
933
+ "learning_rate": 2.667918883627607e-08,
934
+ "loss": 0.1842,
935
+ "step": 145
936
+ },
937
+ {
938
+ "epoch": 1.0,
939
+ "learning_rate": 0.0,
940
+ "loss": 0.1173,
941
+ "step": 146
942
+ }
943
+ ],
944
+ "logging_steps": 1,
945
+ "max_steps": 146,
946
+ "num_train_epochs": 1,
947
+ "save_steps": 500,
948
+ "total_flos": 9.312745291776e+16,
949
+ "trial_name": null,
950
+ "trial_params": null
951
+ }
lora-out/checkpoint-146/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c8c9d593d67af119f8ec4a89dc82888ac41f6183599a972d26cc187d65d77669
3
+ size 4411
lora-out/config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "codellama/CodeLlama-13b-instruct-hf",
3
+ "architectures": [
4
+ "LlamaForCausalLM"
5
+ ],
6
+ "bos_token_id": 1,
7
+ "eos_token_id": 2,
8
+ "hidden_act": "silu",
9
+ "hidden_size": 5120,
10
+ "initializer_range": 0.02,
11
+ "intermediate_size": 13824,
12
+ "max_position_embeddings": 16384,
13
+ "model_type": "llama",
14
+ "num_attention_heads": 40,
15
+ "num_hidden_layers": 40,
16
+ "num_key_value_heads": 40,
17
+ "pretraining_tp": 1,
18
+ "quantization_config": {
19
+ "bnb_4bit_compute_dtype": "float32",
20
+ "bnb_4bit_quant_type": "fp4",
21
+ "bnb_4bit_use_double_quant": false,
22
+ "llm_int8_enable_fp32_cpu_offload": false,
23
+ "llm_int8_has_fp16_weight": false,
24
+ "llm_int8_skip_modules": null,
25
+ "llm_int8_threshold": 6.0,
26
+ "load_in_4bit": false,
27
+ "load_in_8bit": true,
28
+ "quant_method": "bitsandbytes"
29
+ },
30
+ "rms_norm_eps": 1e-05,
31
+ "rope_scaling": null,
32
+ "rope_theta": 1000000,
33
+ "tie_word_embeddings": false,
34
+ "torch_dtype": "bfloat16",
35
+ "transformers_version": "4.34.0.dev0",
36
+ "use_cache": false,
37
+ "vocab_size": 32016
38
+ }
lora-out/special_tokens_map.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "▁<PRE>",
4
+ "▁<MID>",
5
+ "▁<SUF>",
6
+ "▁<EOT>"
7
+ ],
8
+ "bos_token": "<s>",
9
+ "eos_token": "</s>",
10
+ "pad_token": "</s>",
11
+ "unk_token": "<unk>"
12
+ }
lora-out/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45ccb9c8b6b561889acea59191d66986d314e7cbd6a78abc6e49b139ca91c1e6
3
+ size 500058
lora-out/tokenizer_config.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "additional_special_tokens": [
5
+ "▁<PRE>",
6
+ "▁<MID>",
7
+ "▁<SUF>",
8
+ "▁<EOT>"
9
+ ],
10
+ "bos_token": {
11
+ "__type": "AddedToken",
12
+ "content": "<s>",
13
+ "lstrip": false,
14
+ "normalized": true,
15
+ "rstrip": false,
16
+ "single_word": false
17
+ },
18
+ "clean_up_tokenization_spaces": false,
19
+ "eos_token": {
20
+ "__type": "AddedToken",
21
+ "content": "</s>",
22
+ "lstrip": false,
23
+ "normalized": true,
24
+ "rstrip": false,
25
+ "single_word": false
26
+ },
27
+ "eot_token": "▁<EOT>",
28
+ "fill_token": "<FILL_ME>",
29
+ "legacy": null,
30
+ "middle_token": "▁<MID>",
31
+ "model_max_length": 1000000000000000019884624838656,
32
+ "pad_token": null,
33
+ "prefix_token": "▁<PRE>",
34
+ "sp_model_kwargs": {},
35
+ "suffix_first": false,
36
+ "suffix_token": "▁<SUF>",
37
+ "tokenizer_class": "CodeLlamaTokenizer",
38
+ "trust_remote_code": false,
39
+ "unk_token": {
40
+ "__type": "AddedToken",
41
+ "content": "<unk>",
42
+ "lstrip": false,
43
+ "normalized": true,
44
+ "rstrip": false,
45
+ "single_word": false
46
+ },
47
+ "use_default_system_prompt": false,
48
+ "use_fast": true
49
+ }