joshuaclymer commited on
Commit
23ed358
·
1 Parent(s): 721683a

Upload folder using huggingface_hub

Browse files
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: False
10
+ - load_in_4bit: True
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: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: float16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0
adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "models/llama-30b",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 16,
11
+ "lora_dropout": 0.0,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 64,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "SEQ_CLS"
21
+ }
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b6bd84e7f15e95593e083a99751b7d84da632694311e1684b23f16db3b5528d
3
+ size 409081617
special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "</s>",
17
+ "unk_token": {
18
+ "content": "<unk>",
19
+ "lstrip": false,
20
+ "normalized": true,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
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 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "__type": "AddedToken",
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": true,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ "clean_up_tokenization_spaces": false,
11
+ "eos_token": {
12
+ "__type": "AddedToken",
13
+ "content": "</s>",
14
+ "lstrip": false,
15
+ "normalized": true,
16
+ "rstrip": false,
17
+ "single_word": false
18
+ },
19
+ "model_max_length": 2048,
20
+ "pad_token": null,
21
+ "sp_model_kwargs": {},
22
+ "tokenizer_class": "LlamaTokenizer",
23
+ "unk_token": {
24
+ "__type": "AddedToken",
25
+ "content": "<unk>",
26
+ "lstrip": false,
27
+ "normalized": true,
28
+ "rstrip": false,
29
+ "single_word": false
30
+ },
31
+ "use_default_system_prompt": true
32
+ }
train_args.json ADDED
@@ -0,0 +1,123 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "initial_model_dir": "models/llama-30b",
3
+ "distribution_id": "code_hard",
4
+ "date_trained": "11/10/2023 11:46:13",
5
+ "output_dir": "models/classify_lora/llama-30b-code_hard",
6
+ "overwrite_output_dir": false,
7
+ "do_train": false,
8
+ "do_eval": true,
9
+ "do_predict": false,
10
+ "evaluation_strategy": "steps",
11
+ "prediction_loss_only": false,
12
+ "per_device_train_batch_size": 8,
13
+ "per_device_eval_batch_size": 8,
14
+ "per_gpu_train_batch_size": null,
15
+ "per_gpu_eval_batch_size": null,
16
+ "gradient_accumulation_steps": 1,
17
+ "eval_accumulation_steps": null,
18
+ "eval_delay": 0,
19
+ "learning_rate": 0.0002,
20
+ "weight_decay": 0.0,
21
+ "adam_beta1": 0.9,
22
+ "adam_beta2": 0.999,
23
+ "adam_epsilon": 1e-08,
24
+ "max_grad_norm": 0.3,
25
+ "num_train_epochs": 2,
26
+ "max_steps": 100,
27
+ "lr_scheduler_type": "constant",
28
+ "warmup_ratio": 0.03,
29
+ "warmup_steps": 0,
30
+ "log_level": "passive",
31
+ "log_level_replica": "warning",
32
+ "log_on_each_node": true,
33
+ "logging_dir": "models/classify_lora/llama-30b-code_hard/runs/Oct11_08-25-27_compute-permanent-node-975",
34
+ "logging_strategy": "steps",
35
+ "logging_first_step": false,
36
+ "logging_steps": 1,
37
+ "logging_nan_inf_filter": true,
38
+ "save_strategy": "steps",
39
+ "save_steps": 25,
40
+ "save_total_limit": 0,
41
+ "save_safetensors": false,
42
+ "save_on_each_node": false,
43
+ "no_cuda": false,
44
+ "use_cpu": false,
45
+ "use_mps_device": false,
46
+ "seed": 42,
47
+ "data_seed": null,
48
+ "jit_mode_eval": false,
49
+ "use_ipex": false,
50
+ "bf16": false,
51
+ "fp16": false,
52
+ "fp16_opt_level": "O1",
53
+ "half_precision_backend": "auto",
54
+ "bf16_full_eval": false,
55
+ "fp16_full_eval": false,
56
+ "tf32": null,
57
+ "local_rank": 2,
58
+ "ddp_backend": null,
59
+ "tpu_num_cores": null,
60
+ "tpu_metrics_debug": false,
61
+ "debug": [],
62
+ "dataloader_drop_last": false,
63
+ "eval_steps": 25,
64
+ "dataloader_num_workers": 0,
65
+ "past_index": -1,
66
+ "run_name": "train|models-classify_lora-llama-30b-code_hard",
67
+ "disable_tqdm": false,
68
+ "remove_unused_columns": false,
69
+ "label_names": null,
70
+ "load_best_model_at_end": false,
71
+ "metric_for_best_model": "eval_code_hard_score",
72
+ "greater_is_better": true,
73
+ "ignore_data_skip": false,
74
+ "sharded_ddp": [],
75
+ "fsdp": [],
76
+ "fsdp_min_num_params": 0,
77
+ "fsdp_config": {
78
+ "min_num_params": 0,
79
+ "xla": false,
80
+ "xla_fsdp_grad_ckpt": false
81
+ },
82
+ "fsdp_transformer_layer_cls_to_wrap": null,
83
+ "deepspeed": "configs/ds_zero_1.json",
84
+ "label_smoothing_factor": 0.0,
85
+ "optim": "paged_adamw_32bit",
86
+ "optim_args": null,
87
+ "adafactor": false,
88
+ "group_by_length": false,
89
+ "length_column_name": "length",
90
+ "report_to": [
91
+ "wandb"
92
+ ],
93
+ "ddp_find_unused_parameters": false,
94
+ "ddp_bucket_cap_mb": null,
95
+ "ddp_broadcast_buffers": null,
96
+ "dataloader_pin_memory": true,
97
+ "skip_memory_metrics": true,
98
+ "use_legacy_prediction_loop": false,
99
+ "push_to_hub": false,
100
+ "resume_from_checkpoint": null,
101
+ "hub_model_id": null,
102
+ "hub_strategy": "every_save",
103
+ "hub_token": null,
104
+ "hub_private_repo": false,
105
+ "hub_always_push": false,
106
+ "gradient_checkpointing": false,
107
+ "include_inputs_for_metrics": false,
108
+ "fp16_backend": "auto",
109
+ "push_to_hub_model_id": null,
110
+ "push_to_hub_organization": null,
111
+ "push_to_hub_token": null,
112
+ "_n_gpu": 1,
113
+ "mp_parameters": "",
114
+ "auto_find_batch_size": false,
115
+ "full_determinism": false,
116
+ "torchdynamo": null,
117
+ "ray_scope": "last",
118
+ "ddp_timeout": 1800,
119
+ "torch_compile": false,
120
+ "torch_compile_backend": null,
121
+ "torch_compile_mode": null,
122
+ "dispatch_batches": null
123
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:13ff15e8442ca298720e5bc9114f5c2aed778ffc82d39477f5ba867feaf2d8ee
3
+ size 5115
training_logs.json ADDED
@@ -0,0 +1,1067 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "loss": 0.7451,
4
+ "learning_rate": 0.0002,
5
+ "epoch": 0.02,
6
+ "step": 1
7
+ },
8
+ {
9
+ "loss": 0.7931,
10
+ "learning_rate": 0.0002,
11
+ "epoch": 0.04,
12
+ "step": 2
13
+ },
14
+ {
15
+ "loss": 0.7147,
16
+ "learning_rate": 0.0002,
17
+ "epoch": 0.05,
18
+ "step": 3
19
+ },
20
+ {
21
+ "loss": 0.6208,
22
+ "learning_rate": 0.0002,
23
+ "epoch": 0.07,
24
+ "step": 4
25
+ },
26
+ {
27
+ "loss": 0.6318,
28
+ "learning_rate": 0.0002,
29
+ "epoch": 0.09,
30
+ "step": 5
31
+ },
32
+ {
33
+ "loss": 0.5347,
34
+ "learning_rate": 0.0002,
35
+ "epoch": 0.11,
36
+ "step": 6
37
+ },
38
+ {
39
+ "loss": 0.5445,
40
+ "learning_rate": 0.0002,
41
+ "epoch": 0.12,
42
+ "step": 7
43
+ },
44
+ {
45
+ "loss": 0.5323,
46
+ "learning_rate": 0.0002,
47
+ "epoch": 0.14,
48
+ "step": 8
49
+ },
50
+ {
51
+ "loss": 0.5599,
52
+ "learning_rate": 0.0002,
53
+ "epoch": 0.16,
54
+ "step": 9
55
+ },
56
+ {
57
+ "loss": 0.544,
58
+ "learning_rate": 0.0002,
59
+ "epoch": 0.18,
60
+ "step": 10
61
+ },
62
+ {
63
+ "loss": 0.495,
64
+ "learning_rate": 0.0002,
65
+ "epoch": 0.19,
66
+ "step": 11
67
+ },
68
+ {
69
+ "loss": 0.4648,
70
+ "learning_rate": 0.0002,
71
+ "epoch": 0.21,
72
+ "step": 12
73
+ },
74
+ {
75
+ "loss": 0.5144,
76
+ "learning_rate": 0.0002,
77
+ "epoch": 0.23,
78
+ "step": 13
79
+ },
80
+ {
81
+ "loss": 0.3049,
82
+ "learning_rate": 0.0002,
83
+ "epoch": 0.25,
84
+ "step": 14
85
+ },
86
+ {
87
+ "loss": 0.2755,
88
+ "learning_rate": 0.0002,
89
+ "epoch": 0.26,
90
+ "step": 15
91
+ },
92
+ {
93
+ "loss": 0.3953,
94
+ "learning_rate": 0.0002,
95
+ "epoch": 0.28,
96
+ "step": 16
97
+ },
98
+ {
99
+ "loss": 0.293,
100
+ "learning_rate": 0.0002,
101
+ "epoch": 0.3,
102
+ "step": 17
103
+ },
104
+ {
105
+ "loss": 0.3035,
106
+ "learning_rate": 0.0002,
107
+ "epoch": 0.32,
108
+ "step": 18
109
+ },
110
+ {
111
+ "loss": 0.2193,
112
+ "learning_rate": 0.0002,
113
+ "epoch": 0.33,
114
+ "step": 19
115
+ },
116
+ {
117
+ "loss": 0.306,
118
+ "learning_rate": 0.0002,
119
+ "epoch": 0.35,
120
+ "step": 20
121
+ },
122
+ {
123
+ "loss": 0.3844,
124
+ "learning_rate": 0.0002,
125
+ "epoch": 0.37,
126
+ "step": 21
127
+ },
128
+ {
129
+ "loss": 0.4754,
130
+ "learning_rate": 0.0002,
131
+ "epoch": 0.39,
132
+ "step": 22
133
+ },
134
+ {
135
+ "loss": 0.247,
136
+ "learning_rate": 0.0002,
137
+ "epoch": 0.4,
138
+ "step": 23
139
+ },
140
+ {
141
+ "loss": 0.2831,
142
+ "learning_rate": 0.0002,
143
+ "epoch": 0.42,
144
+ "step": 24
145
+ },
146
+ {
147
+ "loss": 0.2395,
148
+ "learning_rate": 0.0002,
149
+ "epoch": 0.44,
150
+ "step": 25
151
+ },
152
+ {
153
+ "eval_code_hard_loss": 0.18990269303321838,
154
+ "eval_code_hard_score": -0.06281973421573639,
155
+ "eval_code_hard_brier_score": 0.06281973421573639,
156
+ "eval_code_hard_average_probability": 0.8535524010658264,
157
+ "eval_code_hard_accuracy": 0.91,
158
+ "eval_code_hard_probabilities": [
159
+ 0.9951574206352234,
160
+ 0.6952859163284302,
161
+ 0.695438027381897,
162
+ 1.0,
163
+ 0.9999997615814209,
164
+ 0.9999988079071045,
165
+ 1.0,
166
+ 0.9999998807907104,
167
+ 0.9999998807907104,
168
+ 0.9022552371025085,
169
+ 0.9998458623886108,
170
+ 0.9992383718490601,
171
+ 1.0,
172
+ 1.0,
173
+ 1.0,
174
+ 0.9705260396003723,
175
+ 0.9934076070785522,
176
+ 0.9893102645874023,
177
+ 0.9892656803131104,
178
+ 1.0,
179
+ 1.0,
180
+ 1.0,
181
+ 0.9999995231628418,
182
+ 0.9999997615814209,
183
+ 0.999992847442627,
184
+ 0.9999940395355225,
185
+ 0.9999849796295166,
186
+ 0.9993022680282593,
187
+ 0.9998225569725037,
188
+ 0.9997448325157166,
189
+ 0.5516940355300903,
190
+ 0.6634960770606995,
191
+ 0.5038003325462341,
192
+ 0.9990589022636414,
193
+ 0.9990129470825195,
194
+ 0.9992672801017761,
195
+ 0.9999462366104126,
196
+ 0.9998952150344849,
197
+ 0.9999340772628784,
198
+ 0.9476701021194458,
199
+ 0.9273682832717896,
200
+ 0.9702255725860596,
201
+ 0.9999784231185913,
202
+ 0.9999960660934448,
203
+ 0.999988317489624,
204
+ 0.6653634309768677,
205
+ 0.4677712023258209,
206
+ 0.5697581768035889,
207
+ 0.8528237342834473,
208
+ 0.8113780617713928,
209
+ 0.7631744742393494,
210
+ 0.4795503616333008,
211
+ 0.4501705467700958,
212
+ 0.45226800441741943,
213
+ 0.9999998807907104,
214
+ 0.9999998807907104,
215
+ 1.0,
216
+ 0.5940175652503967,
217
+ 0.5912690162658691,
218
+ 0.5215416550636292,
219
+ 1.0,
220
+ 1.0,
221
+ 1.0,
222
+ 0.5378963351249695,
223
+ 0.4695456922054291,
224
+ 0.8803982734680176,
225
+ 0.9970523118972778,
226
+ 0.9994822144508362,
227
+ 0.9994255304336548,
228
+ 0.7984318733215332,
229
+ 0.6354186534881592,
230
+ 0.8742010593414307,
231
+ 0.9999979734420776,
232
+ 0.9999983310699463,
233
+ 0.999997615814209,
234
+ 0.9906328916549683,
235
+ 0.9663383364677429,
236
+ 0.9692559838294983,
237
+ 0.46825510263442993,
238
+ 0.6134918928146362,
239
+ 0.5529954433441162,
240
+ 0.6660357713699341,
241
+ 0.9829654097557068,
242
+ 0.47695887088775635,
243
+ 0.9999682903289795,
244
+ 0.9999746084213257,
245
+ 0.5424057841300964,
246
+ 0.518418550491333,
247
+ 0.4060291051864624,
248
+ 0.5124395489692688,
249
+ 0.9932987689971924,
250
+ 0.9838415384292603,
251
+ 0.975809633731842,
252
+ 0.9949588179588318,
253
+ 0.9947303533554077,
254
+ 0.9956455230712891,
255
+ 0.6032052636146545,
256
+ 0.830151379108429,
257
+ 0.38333284854888916,
258
+ 0.7332690954208374
259
+ ],
260
+ "eval_code_hard_runtime": 140.0539,
261
+ "eval_code_hard_samples_per_second": 0.714,
262
+ "eval_code_hard_steps_per_second": 0.029,
263
+ "epoch": 0.44,
264
+ "step": 25
265
+ },
266
+ {
267
+ "loss": 0.217,
268
+ "learning_rate": 0.0002,
269
+ "epoch": 0.46,
270
+ "step": 26
271
+ },
272
+ {
273
+ "loss": 0.2851,
274
+ "learning_rate": 0.0002,
275
+ "epoch": 0.47,
276
+ "step": 27
277
+ },
278
+ {
279
+ "loss": 0.2988,
280
+ "learning_rate": 0.0002,
281
+ "epoch": 0.49,
282
+ "step": 28
283
+ },
284
+ {
285
+ "loss": 0.1119,
286
+ "learning_rate": 0.0002,
287
+ "epoch": 0.51,
288
+ "step": 29
289
+ },
290
+ {
291
+ "loss": 0.1369,
292
+ "learning_rate": 0.0002,
293
+ "epoch": 0.53,
294
+ "step": 30
295
+ },
296
+ {
297
+ "loss": 0.2311,
298
+ "learning_rate": 0.0002,
299
+ "epoch": 0.54,
300
+ "step": 31
301
+ },
302
+ {
303
+ "loss": 0.2201,
304
+ "learning_rate": 0.0002,
305
+ "epoch": 0.56,
306
+ "step": 32
307
+ },
308
+ {
309
+ "loss": 0.1824,
310
+ "learning_rate": 0.0002,
311
+ "epoch": 0.58,
312
+ "step": 33
313
+ },
314
+ {
315
+ "loss": 0.1181,
316
+ "learning_rate": 0.0002,
317
+ "epoch": 0.6,
318
+ "step": 34
319
+ },
320
+ {
321
+ "loss": 0.0781,
322
+ "learning_rate": 0.0002,
323
+ "epoch": 0.61,
324
+ "step": 35
325
+ },
326
+ {
327
+ "loss": 0.1475,
328
+ "learning_rate": 0.0002,
329
+ "epoch": 0.63,
330
+ "step": 36
331
+ },
332
+ {
333
+ "loss": 0.198,
334
+ "learning_rate": 0.0002,
335
+ "epoch": 0.65,
336
+ "step": 37
337
+ },
338
+ {
339
+ "loss": 0.1074,
340
+ "learning_rate": 0.0002,
341
+ "epoch": 0.67,
342
+ "step": 38
343
+ },
344
+ {
345
+ "loss": 0.3816,
346
+ "learning_rate": 0.0002,
347
+ "epoch": 0.68,
348
+ "step": 39
349
+ },
350
+ {
351
+ "loss": 0.2657,
352
+ "learning_rate": 0.0002,
353
+ "epoch": 0.7,
354
+ "step": 40
355
+ },
356
+ {
357
+ "loss": 0.1525,
358
+ "learning_rate": 0.0002,
359
+ "epoch": 0.72,
360
+ "step": 41
361
+ },
362
+ {
363
+ "loss": 0.1426,
364
+ "learning_rate": 0.0002,
365
+ "epoch": 0.74,
366
+ "step": 42
367
+ },
368
+ {
369
+ "loss": 0.1578,
370
+ "learning_rate": 0.0002,
371
+ "epoch": 0.75,
372
+ "step": 43
373
+ },
374
+ {
375
+ "loss": 0.1234,
376
+ "learning_rate": 0.0002,
377
+ "epoch": 0.77,
378
+ "step": 44
379
+ },
380
+ {
381
+ "loss": 0.1591,
382
+ "learning_rate": 0.0002,
383
+ "epoch": 0.79,
384
+ "step": 45
385
+ },
386
+ {
387
+ "loss": 0.0388,
388
+ "learning_rate": 0.0002,
389
+ "epoch": 0.81,
390
+ "step": 46
391
+ },
392
+ {
393
+ "loss": 0.1186,
394
+ "learning_rate": 0.0002,
395
+ "epoch": 0.82,
396
+ "step": 47
397
+ },
398
+ {
399
+ "loss": 0.2242,
400
+ "learning_rate": 0.0002,
401
+ "epoch": 0.84,
402
+ "step": 48
403
+ },
404
+ {
405
+ "loss": 0.2245,
406
+ "learning_rate": 0.0002,
407
+ "epoch": 0.86,
408
+ "step": 49
409
+ },
410
+ {
411
+ "loss": 0.0825,
412
+ "learning_rate": 0.0002,
413
+ "epoch": 0.88,
414
+ "step": 50
415
+ },
416
+ {
417
+ "eval_code_hard_loss": 0.1537313610315323,
418
+ "eval_code_hard_score": -0.04667011648416519,
419
+ "eval_code_hard_brier_score": 0.04667011648416519,
420
+ "eval_code_hard_average_probability": 0.8784838318824768,
421
+ "eval_code_hard_accuracy": 0.94,
422
+ "eval_code_hard_probabilities": [
423
+ 0.9725497364997864,
424
+ 0.7011394500732422,
425
+ 0.6710378527641296,
426
+ 1.0,
427
+ 1.0,
428
+ 1.0,
429
+ 0.9999716281890869,
430
+ 0.9999884366989136,
431
+ 0.999969482421875,
432
+ 0.48456111550331116,
433
+ 0.9981953501701355,
434
+ 0.9864223003387451,
435
+ 1.0,
436
+ 1.0,
437
+ 1.0,
438
+ 0.9983637928962708,
439
+ 0.9989921450614929,
440
+ 0.9978587031364441,
441
+ 1.0,
442
+ 1.0,
443
+ 1.0,
444
+ 0.9999933242797852,
445
+ 0.9999672174453735,
446
+ 0.9999806880950928,
447
+ 0.9999942779541016,
448
+ 0.9999948740005493,
449
+ 0.9999951124191284,
450
+ 0.9994611144065857,
451
+ 0.9993894100189209,
452
+ 0.999222993850708,
453
+ 0.8560943007469177,
454
+ 0.8921459317207336,
455
+ 0.7734678387641907,
456
+ 0.9890369772911072,
457
+ 0.9659588932991028,
458
+ 0.9716930985450745,
459
+ 0.9913560152053833,
460
+ 0.9888952970504761,
461
+ 0.9655161499977112,
462
+ 0.882901132106781,
463
+ 0.8454601168632507,
464
+ 0.8470443487167358,
465
+ 0.9999980926513672,
466
+ 0.9999996423721313,
467
+ 0.9999994039535522,
468
+ 0.6798665523529053,
469
+ 0.49168092012405396,
470
+ 0.5368497967720032,
471
+ 0.9873051047325134,
472
+ 0.9402137994766235,
473
+ 0.928394615650177,
474
+ 0.35502684116363525,
475
+ 0.9124428629875183,
476
+ 0.946331799030304,
477
+ 1.0,
478
+ 1.0,
479
+ 1.0,
480
+ 0.6016813516616821,
481
+ 0.49267151951789856,
482
+ 0.6437432169914246,
483
+ 1.0,
484
+ 1.0,
485
+ 1.0,
486
+ 0.4486885964870453,
487
+ 0.640812337398529,
488
+ 0.879755437374115,
489
+ 0.9916812777519226,
490
+ 0.9911965131759644,
491
+ 0.9993059635162354,
492
+ 0.589796245098114,
493
+ 0.7118774652481079,
494
+ 0.9268589019775391,
495
+ 0.9999767541885376,
496
+ 0.9999942779541016,
497
+ 0.9999902248382568,
498
+ 0.9828012585639954,
499
+ 0.9357141256332397,
500
+ 0.9222304224967957,
501
+ 0.5056607723236084,
502
+ 0.7614033818244934,
503
+ 0.7656963467597961,
504
+ 0.6500656008720398,
505
+ 0.8527267575263977,
506
+ 0.5424817800521851,
507
+ 0.9999986886978149,
508
+ 0.999997615814209,
509
+ 0.8840383291244507,
510
+ 0.5736863613128662,
511
+ 0.5833655595779419,
512
+ 0.69774329662323,
513
+ 0.9973189234733582,
514
+ 0.9964283108711243,
515
+ 0.9958376884460449,
516
+ 0.9838255047798157,
517
+ 0.9848774671554565,
518
+ 0.9828516244888306,
519
+ 0.7398984432220459,
520
+ 0.9901463389396667,
521
+ 0.3069886267185211,
522
+ 0.7398353219032288
523
+ ],
524
+ "eval_code_hard_runtime": 140.0621,
525
+ "eval_code_hard_samples_per_second": 0.714,
526
+ "eval_code_hard_steps_per_second": 0.029,
527
+ "epoch": 0.88,
528
+ "step": 50
529
+ },
530
+ {
531
+ "loss": 0.1986,
532
+ "learning_rate": 0.0002,
533
+ "epoch": 0.89,
534
+ "step": 51
535
+ },
536
+ {
537
+ "loss": 0.1768,
538
+ "learning_rate": 0.0002,
539
+ "epoch": 0.91,
540
+ "step": 52
541
+ },
542
+ {
543
+ "loss": 0.1257,
544
+ "learning_rate": 0.0002,
545
+ "epoch": 0.93,
546
+ "step": 53
547
+ },
548
+ {
549
+ "loss": 0.1255,
550
+ "learning_rate": 0.0002,
551
+ "epoch": 0.95,
552
+ "step": 54
553
+ },
554
+ {
555
+ "loss": 0.067,
556
+ "learning_rate": 0.0002,
557
+ "epoch": 0.96,
558
+ "step": 55
559
+ },
560
+ {
561
+ "loss": 0.2208,
562
+ "learning_rate": 0.0002,
563
+ "epoch": 0.98,
564
+ "step": 56
565
+ },
566
+ {
567
+ "loss": 0.1187,
568
+ "learning_rate": 0.0002,
569
+ "epoch": 1.0,
570
+ "step": 57
571
+ },
572
+ {
573
+ "loss": 0.159,
574
+ "learning_rate": 0.0002,
575
+ "epoch": 1.02,
576
+ "step": 58
577
+ },
578
+ {
579
+ "loss": 0.0512,
580
+ "learning_rate": 0.0002,
581
+ "epoch": 1.04,
582
+ "step": 59
583
+ },
584
+ {
585
+ "loss": 0.0608,
586
+ "learning_rate": 0.0002,
587
+ "epoch": 1.05,
588
+ "step": 60
589
+ },
590
+ {
591
+ "loss": 0.1049,
592
+ "learning_rate": 0.0002,
593
+ "epoch": 1.07,
594
+ "step": 61
595
+ },
596
+ {
597
+ "loss": 0.0634,
598
+ "learning_rate": 0.0002,
599
+ "epoch": 1.09,
600
+ "step": 62
601
+ },
602
+ {
603
+ "loss": 0.0609,
604
+ "learning_rate": 0.0002,
605
+ "epoch": 1.11,
606
+ "step": 63
607
+ },
608
+ {
609
+ "loss": 0.0888,
610
+ "learning_rate": 0.0002,
611
+ "epoch": 1.12,
612
+ "step": 64
613
+ },
614
+ {
615
+ "loss": 0.1165,
616
+ "learning_rate": 0.0002,
617
+ "epoch": 1.14,
618
+ "step": 65
619
+ },
620
+ {
621
+ "loss": 0.017,
622
+ "learning_rate": 0.0002,
623
+ "epoch": 1.16,
624
+ "step": 66
625
+ },
626
+ {
627
+ "loss": 0.0504,
628
+ "learning_rate": 0.0002,
629
+ "epoch": 1.18,
630
+ "step": 67
631
+ },
632
+ {
633
+ "loss": 0.0958,
634
+ "learning_rate": 0.0002,
635
+ "epoch": 1.19,
636
+ "step": 68
637
+ },
638
+ {
639
+ "loss": 0.0276,
640
+ "learning_rate": 0.0002,
641
+ "epoch": 1.21,
642
+ "step": 69
643
+ },
644
+ {
645
+ "loss": 0.0394,
646
+ "learning_rate": 0.0002,
647
+ "epoch": 1.23,
648
+ "step": 70
649
+ },
650
+ {
651
+ "loss": 0.0398,
652
+ "learning_rate": 0.0002,
653
+ "epoch": 1.25,
654
+ "step": 71
655
+ },
656
+ {
657
+ "loss": 0.108,
658
+ "learning_rate": 0.0002,
659
+ "epoch": 1.26,
660
+ "step": 72
661
+ },
662
+ {
663
+ "loss": 0.0392,
664
+ "learning_rate": 0.0002,
665
+ "epoch": 1.28,
666
+ "step": 73
667
+ },
668
+ {
669
+ "loss": 0.0233,
670
+ "learning_rate": 0.0002,
671
+ "epoch": 1.3,
672
+ "step": 74
673
+ },
674
+ {
675
+ "loss": 0.1554,
676
+ "learning_rate": 0.0002,
677
+ "epoch": 1.32,
678
+ "step": 75
679
+ },
680
+ {
681
+ "eval_code_hard_loss": 0.13956719636917114,
682
+ "eval_code_hard_score": -0.042688366025686264,
683
+ "eval_code_hard_brier_score": 0.042688366025686264,
684
+ "eval_code_hard_average_probability": 0.9274539947509766,
685
+ "eval_code_hard_accuracy": 0.93,
686
+ "eval_code_hard_probabilities": [
687
+ 0.9999963045120239,
688
+ 0.9699520468711853,
689
+ 0.9745141267776489,
690
+ 1.0,
691
+ 1.0,
692
+ 1.0,
693
+ 1.0,
694
+ 1.0,
695
+ 1.0,
696
+ 0.35654664039611816,
697
+ 0.886887788772583,
698
+ 0.9559746980667114,
699
+ 1.0,
700
+ 1.0,
701
+ 1.0,
702
+ 0.9999935626983643,
703
+ 0.9999960660934448,
704
+ 0.9999923706054688,
705
+ 0.9999997615814209,
706
+ 1.0,
707
+ 0.9999996423721313,
708
+ 1.0,
709
+ 1.0,
710
+ 1.0,
711
+ 1.0,
712
+ 0.9999995231628418,
713
+ 0.9999997615814209,
714
+ 0.9998072981834412,
715
+ 0.9999998807907104,
716
+ 0.999997615814209,
717
+ 0.9905921816825867,
718
+ 0.9987447261810303,
719
+ 0.9492596983909607,
720
+ 1.0,
721
+ 0.9999996423721313,
722
+ 0.9999969005584717,
723
+ 1.0,
724
+ 0.9999995231628418,
725
+ 0.9999992847442627,
726
+ 0.9940958023071289,
727
+ 0.9881225228309631,
728
+ 0.9979448914527893,
729
+ 1.0,
730
+ 1.0,
731
+ 1.0,
732
+ 0.9875559210777283,
733
+ 0.578984797000885,
734
+ 0.643305242061615,
735
+ 0.9999995231628418,
736
+ 0.9999939203262329,
737
+ 0.9999977350234985,
738
+ 0.005097602494060993,
739
+ 0.9997228980064392,
740
+ 0.9975023865699768,
741
+ 1.0,
742
+ 1.0,
743
+ 1.0,
744
+ 0.3311256468296051,
745
+ 0.45201003551483154,
746
+ 0.7562749981880188,
747
+ 1.0,
748
+ 1.0,
749
+ 1.0,
750
+ 0.9942985773086548,
751
+ 0.9975347518920898,
752
+ 0.9981868863105774,
753
+ 0.9999996423721313,
754
+ 0.9999785423278809,
755
+ 1.0,
756
+ 0.46586939692497253,
757
+ 0.9903834462165833,
758
+ 0.9999986886978149,
759
+ 1.0,
760
+ 1.0,
761
+ 1.0,
762
+ 0.9817127585411072,
763
+ 0.2876265347003937,
764
+ 0.8499608635902405,
765
+ 0.61001056432724,
766
+ 0.996311604976654,
767
+ 0.999902606010437,
768
+ 0.9834624528884888,
769
+ 0.9992788434028625,
770
+ 0.9218541383743286,
771
+ 1.0,
772
+ 1.0,
773
+ 0.9609642028808594,
774
+ 0.7825002074241638,
775
+ 0.9717795252799988,
776
+ 0.9860764741897583,
777
+ 0.9999998807907104,
778
+ 0.9999990463256836,
779
+ 0.9999997615814209,
780
+ 0.9999983310699463,
781
+ 0.9999971389770508,
782
+ 0.9999949932098389,
783
+ 0.9969584941864014,
784
+ 0.9999991655349731,
785
+ 0.15736030042171478,
786
+ 0.9994120597839355
787
+ ],
788
+ "eval_code_hard_runtime": 140.0578,
789
+ "eval_code_hard_samples_per_second": 0.714,
790
+ "eval_code_hard_steps_per_second": 0.029,
791
+ "epoch": 1.32,
792
+ "step": 75
793
+ },
794
+ {
795
+ "loss": 0.0227,
796
+ "learning_rate": 0.0002,
797
+ "epoch": 1.33,
798
+ "step": 76
799
+ },
800
+ {
801
+ "loss": 0.0849,
802
+ "learning_rate": 0.0002,
803
+ "epoch": 1.35,
804
+ "step": 77
805
+ },
806
+ {
807
+ "loss": 0.1094,
808
+ "learning_rate": 0.0002,
809
+ "epoch": 1.37,
810
+ "step": 78
811
+ },
812
+ {
813
+ "loss": 0.0689,
814
+ "learning_rate": 0.0002,
815
+ "epoch": 1.39,
816
+ "step": 79
817
+ },
818
+ {
819
+ "loss": 0.0418,
820
+ "learning_rate": 0.0002,
821
+ "epoch": 1.4,
822
+ "step": 80
823
+ },
824
+ {
825
+ "loss": 0.0886,
826
+ "learning_rate": 0.0002,
827
+ "epoch": 1.42,
828
+ "step": 81
829
+ },
830
+ {
831
+ "loss": 0.1833,
832
+ "learning_rate": 0.0002,
833
+ "epoch": 1.44,
834
+ "step": 82
835
+ },
836
+ {
837
+ "loss": 0.1213,
838
+ "learning_rate": 0.0002,
839
+ "epoch": 1.46,
840
+ "step": 83
841
+ },
842
+ {
843
+ "loss": 0.0424,
844
+ "learning_rate": 0.0002,
845
+ "epoch": 1.47,
846
+ "step": 84
847
+ },
848
+ {
849
+ "loss": 0.0194,
850
+ "learning_rate": 0.0002,
851
+ "epoch": 1.49,
852
+ "step": 85
853
+ },
854
+ {
855
+ "loss": 0.0229,
856
+ "learning_rate": 0.0002,
857
+ "epoch": 1.51,
858
+ "step": 86
859
+ },
860
+ {
861
+ "loss": 0.0231,
862
+ "learning_rate": 0.0002,
863
+ "epoch": 1.53,
864
+ "step": 87
865
+ },
866
+ {
867
+ "loss": 0.0263,
868
+ "learning_rate": 0.0002,
869
+ "epoch": 1.54,
870
+ "step": 88
871
+ },
872
+ {
873
+ "loss": 0.0619,
874
+ "learning_rate": 0.0002,
875
+ "epoch": 1.56,
876
+ "step": 89
877
+ },
878
+ {
879
+ "loss": 0.0365,
880
+ "learning_rate": 0.0002,
881
+ "epoch": 1.58,
882
+ "step": 90
883
+ },
884
+ {
885
+ "loss": 0.0523,
886
+ "learning_rate": 0.0002,
887
+ "epoch": 1.6,
888
+ "step": 91
889
+ },
890
+ {
891
+ "loss": 0.0172,
892
+ "learning_rate": 0.0002,
893
+ "epoch": 1.61,
894
+ "step": 92
895
+ },
896
+ {
897
+ "loss": 0.0455,
898
+ "learning_rate": 0.0002,
899
+ "epoch": 1.63,
900
+ "step": 93
901
+ },
902
+ {
903
+ "loss": 0.2764,
904
+ "learning_rate": 0.0002,
905
+ "epoch": 1.65,
906
+ "step": 94
907
+ },
908
+ {
909
+ "loss": 0.0225,
910
+ "learning_rate": 0.0002,
911
+ "epoch": 1.67,
912
+ "step": 95
913
+ },
914
+ {
915
+ "loss": 0.0133,
916
+ "learning_rate": 0.0002,
917
+ "epoch": 1.68,
918
+ "step": 96
919
+ },
920
+ {
921
+ "loss": 0.0188,
922
+ "learning_rate": 0.0002,
923
+ "epoch": 1.7,
924
+ "step": 97
925
+ },
926
+ {
927
+ "loss": 0.0207,
928
+ "learning_rate": 0.0002,
929
+ "epoch": 1.72,
930
+ "step": 98
931
+ },
932
+ {
933
+ "loss": 0.012,
934
+ "learning_rate": 0.0002,
935
+ "epoch": 1.74,
936
+ "step": 99
937
+ },
938
+ {
939
+ "loss": 0.0628,
940
+ "learning_rate": 0.0002,
941
+ "epoch": 1.75,
942
+ "step": 100
943
+ },
944
+ {
945
+ "eval_code_hard_loss": 0.1615547239780426,
946
+ "eval_code_hard_score": -0.03736421465873718,
947
+ "eval_code_hard_brier_score": 0.03736421465873718,
948
+ "eval_code_hard_average_probability": 0.9438884854316711,
949
+ "eval_code_hard_accuracy": 0.95,
950
+ "eval_code_hard_probabilities": [
951
+ 1.0,
952
+ 0.9725049734115601,
953
+ 0.9244245886802673,
954
+ 1.0,
955
+ 1.0,
956
+ 1.0,
957
+ 1.0,
958
+ 1.0,
959
+ 1.0,
960
+ 0.07730857282876968,
961
+ 0.9794886112213135,
962
+ 0.999785840511322,
963
+ 1.0,
964
+ 1.0,
965
+ 1.0,
966
+ 1.0,
967
+ 1.0,
968
+ 1.0,
969
+ 1.0,
970
+ 1.0,
971
+ 1.0,
972
+ 1.0,
973
+ 1.0,
974
+ 1.0,
975
+ 1.0,
976
+ 1.0,
977
+ 1.0,
978
+ 0.9999511241912842,
979
+ 1.0,
980
+ 1.0,
981
+ 0.9999926090240479,
982
+ 1.0,
983
+ 0.9999614953994751,
984
+ 1.0,
985
+ 1.0,
986
+ 1.0,
987
+ 1.0,
988
+ 0.9999998807907104,
989
+ 1.0,
990
+ 0.9999260902404785,
991
+ 0.9998406171798706,
992
+ 0.9999992847442627,
993
+ 1.0,
994
+ 1.0,
995
+ 1.0,
996
+ 0.9999581575393677,
997
+ 0.6513886451721191,
998
+ 0.7881560921669006,
999
+ 1.0,
1000
+ 1.0,
1001
+ 1.0,
1002
+ 0.00015713961329311132,
1003
+ 1.0,
1004
+ 0.9999954700469971,
1005
+ 1.0,
1006
+ 1.0,
1007
+ 1.0,
1008
+ 0.4968324303627014,
1009
+ 0.40699660778045654,
1010
+ 0.9514529705047607,
1011
+ 1.0,
1012
+ 1.0,
1013
+ 1.0,
1014
+ 1.0,
1015
+ 0.9999997615814209,
1016
+ 1.0,
1017
+ 1.0,
1018
+ 1.0,
1019
+ 1.0,
1020
+ 0.9981179237365723,
1021
+ 0.999592125415802,
1022
+ 0.9999996423721313,
1023
+ 1.0,
1024
+ 1.0,
1025
+ 1.0,
1026
+ 0.9999996423721313,
1027
+ 0.9520946741104126,
1028
+ 0.9975154399871826,
1029
+ 0.5100582242012024,
1030
+ 0.9827112555503845,
1031
+ 0.9999972581863403,
1032
+ 0.9999854564666748,
1033
+ 1.0,
1034
+ 0.9896236062049866,
1035
+ 1.0,
1036
+ 1.0,
1037
+ 0.8571937680244446,
1038
+ 0.796758770942688,
1039
+ 0.9510358572006226,
1040
+ 0.9994876384735107,
1041
+ 1.0,
1042
+ 1.0,
1043
+ 1.0,
1044
+ 1.0,
1045
+ 1.0,
1046
+ 1.0,
1047
+ 0.9998961687088013,
1048
+ 1.0,
1049
+ 0.10666914284229279,
1050
+ 1.0
1051
+ ],
1052
+ "eval_code_hard_runtime": 140.0373,
1053
+ "eval_code_hard_samples_per_second": 0.714,
1054
+ "eval_code_hard_steps_per_second": 0.029,
1055
+ "epoch": 1.75,
1056
+ "step": 100
1057
+ },
1058
+ {
1059
+ "train_runtime": 12026.9778,
1060
+ "train_samples_per_second": 0.266,
1061
+ "train_steps_per_second": 0.008,
1062
+ "total_flos": 0.0,
1063
+ "train_loss": 0.19655362625606357,
1064
+ "epoch": 1.75,
1065
+ "step": 100
1066
+ }
1067
+ ]