joshuaclymer commited on
Commit
8ba91b6
·
1 Parent(s): 2bdd301

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-13b",
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:f201ef5aeeda8d2cfa09ee4ecd1b01c68cafaa6a0d9b4e68eb751326962b06d0
3
+ size 209811921
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-13b",
3
+ "distribution_id": "code_hard",
4
+ "date_trained": "10/10/2023 09:37:05",
5
+ "output_dir": "models/classify_lora/llama-13b-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": 16,
13
+ "per_device_eval_batch_size": 16,
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": 3,
26
+ "max_steps": 150,
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-13b-code_hard/runs/Oct10_05-40-18_compute-permanent-node-978",
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": 1,
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-13b-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:d64d8d4f9e2c1920215256853d6337fd6f69c3e74db1730786d16164e94e8a96
3
+ size 5115
training_logs.json ADDED
@@ -0,0 +1,1595 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "loss": 0.7559,
4
+ "learning_rate": 0.0002,
5
+ "epoch": 0.02,
6
+ "step": 1
7
+ },
8
+ {
9
+ "loss": 0.809,
10
+ "learning_rate": 0.0002,
11
+ "epoch": 0.04,
12
+ "step": 2
13
+ },
14
+ {
15
+ "loss": 0.6615,
16
+ "learning_rate": 0.0002,
17
+ "epoch": 0.05,
18
+ "step": 3
19
+ },
20
+ {
21
+ "loss": 0.6488,
22
+ "learning_rate": 0.0002,
23
+ "epoch": 0.07,
24
+ "step": 4
25
+ },
26
+ {
27
+ "loss": 0.6381,
28
+ "learning_rate": 0.0002,
29
+ "epoch": 0.09,
30
+ "step": 5
31
+ },
32
+ {
33
+ "loss": 0.5315,
34
+ "learning_rate": 0.0002,
35
+ "epoch": 0.11,
36
+ "step": 6
37
+ },
38
+ {
39
+ "loss": 0.5713,
40
+ "learning_rate": 0.0002,
41
+ "epoch": 0.12,
42
+ "step": 7
43
+ },
44
+ {
45
+ "loss": 0.5663,
46
+ "learning_rate": 0.0002,
47
+ "epoch": 0.14,
48
+ "step": 8
49
+ },
50
+ {
51
+ "loss": 0.6267,
52
+ "learning_rate": 0.0002,
53
+ "epoch": 0.16,
54
+ "step": 9
55
+ },
56
+ {
57
+ "loss": 0.5685,
58
+ "learning_rate": 0.0002,
59
+ "epoch": 0.18,
60
+ "step": 10
61
+ },
62
+ {
63
+ "loss": 0.501,
64
+ "learning_rate": 0.0002,
65
+ "epoch": 0.19,
66
+ "step": 11
67
+ },
68
+ {
69
+ "loss": 0.5033,
70
+ "learning_rate": 0.0002,
71
+ "epoch": 0.21,
72
+ "step": 12
73
+ },
74
+ {
75
+ "loss": 0.5903,
76
+ "learning_rate": 0.0002,
77
+ "epoch": 0.23,
78
+ "step": 13
79
+ },
80
+ {
81
+ "loss": 0.4598,
82
+ "learning_rate": 0.0002,
83
+ "epoch": 0.25,
84
+ "step": 14
85
+ },
86
+ {
87
+ "loss": 0.4332,
88
+ "learning_rate": 0.0002,
89
+ "epoch": 0.26,
90
+ "step": 15
91
+ },
92
+ {
93
+ "loss": 0.4749,
94
+ "learning_rate": 0.0002,
95
+ "epoch": 0.28,
96
+ "step": 16
97
+ },
98
+ {
99
+ "loss": 0.3193,
100
+ "learning_rate": 0.0002,
101
+ "epoch": 0.3,
102
+ "step": 17
103
+ },
104
+ {
105
+ "loss": 0.4061,
106
+ "learning_rate": 0.0002,
107
+ "epoch": 0.32,
108
+ "step": 18
109
+ },
110
+ {
111
+ "loss": 0.3541,
112
+ "learning_rate": 0.0002,
113
+ "epoch": 0.33,
114
+ "step": 19
115
+ },
116
+ {
117
+ "loss": 0.3414,
118
+ "learning_rate": 0.0002,
119
+ "epoch": 0.35,
120
+ "step": 20
121
+ },
122
+ {
123
+ "loss": 0.3453,
124
+ "learning_rate": 0.0002,
125
+ "epoch": 0.37,
126
+ "step": 21
127
+ },
128
+ {
129
+ "loss": 0.5322,
130
+ "learning_rate": 0.0002,
131
+ "epoch": 0.39,
132
+ "step": 22
133
+ },
134
+ {
135
+ "loss": 0.3092,
136
+ "learning_rate": 0.0002,
137
+ "epoch": 0.4,
138
+ "step": 23
139
+ },
140
+ {
141
+ "loss": 0.2027,
142
+ "learning_rate": 0.0002,
143
+ "epoch": 0.42,
144
+ "step": 24
145
+ },
146
+ {
147
+ "loss": 0.403,
148
+ "learning_rate": 0.0002,
149
+ "epoch": 0.44,
150
+ "step": 25
151
+ },
152
+ {
153
+ "eval_code_hard_loss": 0.24356038868427277,
154
+ "eval_code_hard_score": -0.08314146846532822,
155
+ "eval_code_hard_brier_score": 0.08314146846532822,
156
+ "eval_code_hard_average_probability": 0.8104196786880493,
157
+ "eval_code_hard_accuracy": 0.88,
158
+ "eval_code_hard_probabilities": [
159
+ 0.7802631855010986,
160
+ 0.5728839039802551,
161
+ 0.601148784160614,
162
+ 0.9993261098861694,
163
+ 0.9960917830467224,
164
+ 0.9883104562759399,
165
+ 0.9998921155929565,
166
+ 0.9998250603675842,
167
+ 0.9999023675918579,
168
+ 0.7125715017318726,
169
+ 0.9519601464271545,
170
+ 0.8720795512199402,
171
+ 1.0,
172
+ 0.9999998807907104,
173
+ 0.9999998807907104,
174
+ 0.9911274313926697,
175
+ 0.6485933065414429,
176
+ 0.831139326095581,
177
+ 0.9780414700508118,
178
+ 0.9999822378158569,
179
+ 0.9999316930770874,
180
+ 0.9999449253082275,
181
+ 0.9999632835388184,
182
+ 0.9999346733093262,
183
+ 0.9998100399971008,
184
+ 0.9998506307601929,
185
+ 0.999704897403717,
186
+ 0.9949617385864258,
187
+ 0.9998635053634644,
188
+ 0.9998619556427002,
189
+ 0.6420595645904541,
190
+ 0.7948806881904602,
191
+ 0.6511540412902832,
192
+ 0.9534466862678528,
193
+ 0.9455068707466125,
194
+ 0.944064736366272,
195
+ 0.9278716444969177,
196
+ 0.8786247968673706,
197
+ 0.9867132902145386,
198
+ 0.9512923359870911,
199
+ 0.975769579410553,
200
+ 0.9804350137710571,
201
+ 0.9998563528060913,
202
+ 0.9999313354492188,
203
+ 0.9999279975891113,
204
+ 0.8713787198066711,
205
+ 0.5257866382598877,
206
+ 0.5062138438224792,
207
+ 0.7719318270683289,
208
+ 0.42283594608306885,
209
+ 0.43223893642425537,
210
+ 0.5629215836524963,
211
+ 0.9509612321853638,
212
+ 0.6529775857925415,
213
+ 0.998943030834198,
214
+ 0.9988754391670227,
215
+ 0.9993382096290588,
216
+ 0.4989044666290283,
217
+ 0.5084826946258545,
218
+ 0.5146058797836304,
219
+ 0.9999768733978271,
220
+ 0.9999709129333496,
221
+ 0.9999724626541138,
222
+ 0.6513444781303406,
223
+ 0.5117835402488708,
224
+ 0.8250595331192017,
225
+ 0.9398683309555054,
226
+ 0.9772916436195374,
227
+ 0.9429591298103333,
228
+ 0.9575245380401611,
229
+ 0.8619183301925659,
230
+ 0.5832539796829224,
231
+ 0.9970285296440125,
232
+ 0.999936580657959,
233
+ 0.9999572038650513,
234
+ 0.9991602897644043,
235
+ 0.8450718522071838,
236
+ 0.7737744450569153,
237
+ 0.4931277930736542,
238
+ 0.483172208070755,
239
+ 0.5014657974243164,
240
+ 0.5353665947914124,
241
+ 0.434240460395813,
242
+ 0.516578197479248,
243
+ 0.9958915710449219,
244
+ 0.9994507431983948,
245
+ 0.5593408942222595,
246
+ 0.4825429320335388,
247
+ 0.4861961901187897,
248
+ 0.49605879187583923,
249
+ 0.9998632669448853,
250
+ 0.9998589754104614,
251
+ 0.9998570680618286,
252
+ 0.39110851287841797,
253
+ 0.38484665751457214,
254
+ 0.3077014684677124,
255
+ 0.5388085246086121,
256
+ 0.6023823618888855,
257
+ 0.5493093729019165,
258
+ 0.6540540456771851
259
+ ],
260
+ "eval_code_hard_runtime": 110.6525,
261
+ "eval_code_hard_samples_per_second": 0.904,
262
+ "eval_code_hard_steps_per_second": 0.036,
263
+ "epoch": 0.44,
264
+ "step": 25
265
+ },
266
+ {
267
+ "loss": 0.2619,
268
+ "learning_rate": 0.0002,
269
+ "epoch": 0.46,
270
+ "step": 26
271
+ },
272
+ {
273
+ "loss": 0.2734,
274
+ "learning_rate": 0.0002,
275
+ "epoch": 0.47,
276
+ "step": 27
277
+ },
278
+ {
279
+ "loss": 0.4004,
280
+ "learning_rate": 0.0002,
281
+ "epoch": 0.49,
282
+ "step": 28
283
+ },
284
+ {
285
+ "loss": 0.1422,
286
+ "learning_rate": 0.0002,
287
+ "epoch": 0.51,
288
+ "step": 29
289
+ },
290
+ {
291
+ "loss": 0.2156,
292
+ "learning_rate": 0.0002,
293
+ "epoch": 0.53,
294
+ "step": 30
295
+ },
296
+ {
297
+ "loss": 0.2327,
298
+ "learning_rate": 0.0002,
299
+ "epoch": 0.54,
300
+ "step": 31
301
+ },
302
+ {
303
+ "loss": 0.1781,
304
+ "learning_rate": 0.0002,
305
+ "epoch": 0.56,
306
+ "step": 32
307
+ },
308
+ {
309
+ "loss": 0.2628,
310
+ "learning_rate": 0.0002,
311
+ "epoch": 0.58,
312
+ "step": 33
313
+ },
314
+ {
315
+ "loss": 0.3038,
316
+ "learning_rate": 0.0002,
317
+ "epoch": 0.6,
318
+ "step": 34
319
+ },
320
+ {
321
+ "loss": 0.2291,
322
+ "learning_rate": 0.0002,
323
+ "epoch": 0.61,
324
+ "step": 35
325
+ },
326
+ {
327
+ "loss": 0.1074,
328
+ "learning_rate": 0.0002,
329
+ "epoch": 0.63,
330
+ "step": 36
331
+ },
332
+ {
333
+ "loss": 0.2432,
334
+ "learning_rate": 0.0002,
335
+ "epoch": 0.65,
336
+ "step": 37
337
+ },
338
+ {
339
+ "loss": 0.1269,
340
+ "learning_rate": 0.0002,
341
+ "epoch": 0.67,
342
+ "step": 38
343
+ },
344
+ {
345
+ "loss": 0.2828,
346
+ "learning_rate": 0.0002,
347
+ "epoch": 0.68,
348
+ "step": 39
349
+ },
350
+ {
351
+ "loss": 0.2679,
352
+ "learning_rate": 0.0002,
353
+ "epoch": 0.7,
354
+ "step": 40
355
+ },
356
+ {
357
+ "loss": 0.1949,
358
+ "learning_rate": 0.0002,
359
+ "epoch": 0.72,
360
+ "step": 41
361
+ },
362
+ {
363
+ "loss": 0.2657,
364
+ "learning_rate": 0.0002,
365
+ "epoch": 0.74,
366
+ "step": 42
367
+ },
368
+ {
369
+ "loss": 0.1775,
370
+ "learning_rate": 0.0002,
371
+ "epoch": 0.75,
372
+ "step": 43
373
+ },
374
+ {
375
+ "loss": 0.2224,
376
+ "learning_rate": 0.0002,
377
+ "epoch": 0.77,
378
+ "step": 44
379
+ },
380
+ {
381
+ "loss": 0.2793,
382
+ "learning_rate": 0.0002,
383
+ "epoch": 0.79,
384
+ "step": 45
385
+ },
386
+ {
387
+ "loss": 0.0357,
388
+ "learning_rate": 0.0002,
389
+ "epoch": 0.81,
390
+ "step": 46
391
+ },
392
+ {
393
+ "loss": 0.1634,
394
+ "learning_rate": 0.0002,
395
+ "epoch": 0.82,
396
+ "step": 47
397
+ },
398
+ {
399
+ "loss": 0.2606,
400
+ "learning_rate": 0.0002,
401
+ "epoch": 0.84,
402
+ "step": 48
403
+ },
404
+ {
405
+ "loss": 0.2214,
406
+ "learning_rate": 0.0002,
407
+ "epoch": 0.86,
408
+ "step": 49
409
+ },
410
+ {
411
+ "loss": 0.0889,
412
+ "learning_rate": 0.0002,
413
+ "epoch": 0.88,
414
+ "step": 50
415
+ },
416
+ {
417
+ "eval_code_hard_loss": 0.16699519753456116,
418
+ "eval_code_hard_score": -0.05673101916909218,
419
+ "eval_code_hard_brier_score": 0.05673101916909218,
420
+ "eval_code_hard_average_probability": 0.8761882781982422,
421
+ "eval_code_hard_accuracy": 0.93,
422
+ "eval_code_hard_probabilities": [
423
+ 0.9874073266983032,
424
+ 0.7980256080627441,
425
+ 0.9544906616210938,
426
+ 0.9999760389328003,
427
+ 0.9833183288574219,
428
+ 0.9734002947807312,
429
+ 1.0,
430
+ 1.0,
431
+ 1.0,
432
+ 0.8710820078849792,
433
+ 0.9463728070259094,
434
+ 0.9430120587348938,
435
+ 1.0,
436
+ 1.0,
437
+ 1.0,
438
+ 0.9999483823776245,
439
+ 0.9996401071548462,
440
+ 0.9999982118606567,
441
+ 0.6661297082901001,
442
+ 0.9998952150344849,
443
+ 0.9999295473098755,
444
+ 0.9999996423721313,
445
+ 1.0,
446
+ 0.9999994039535522,
447
+ 0.9999994039535522,
448
+ 0.9999171495437622,
449
+ 0.9984815716743469,
450
+ 0.9999889135360718,
451
+ 0.9999997615814209,
452
+ 0.9999967813491821,
453
+ 0.997412383556366,
454
+ 0.9992191791534424,
455
+ 0.9149346947669983,
456
+ 0.9999977350234985,
457
+ 0.9999892711639404,
458
+ 0.9999914169311523,
459
+ 1.0,
460
+ 0.9999998807907104,
461
+ 1.0,
462
+ 0.9768458604812622,
463
+ 0.9713558554649353,
464
+ 0.9916350841522217,
465
+ 0.9999990463256836,
466
+ 1.0,
467
+ 1.0,
468
+ 0.7556970119476318,
469
+ 0.5051810145378113,
470
+ 0.443418949842453,
471
+ 0.9936054944992065,
472
+ 0.3165128529071808,
473
+ 0.3557163178920746,
474
+ 0.5657868385314941,
475
+ 0.9981420040130615,
476
+ 0.5603527426719666,
477
+ 0.9821581840515137,
478
+ 0.9779713749885559,
479
+ 0.9982566237449646,
480
+ 0.5955212712287903,
481
+ 0.5466123819351196,
482
+ 0.5278903245925903,
483
+ 0.999467670917511,
484
+ 0.9992781281471252,
485
+ 0.9993084669113159,
486
+ 0.9714255332946777,
487
+ 0.5806359648704529,
488
+ 0.9231281876564026,
489
+ 0.9999953508377075,
490
+ 1.0,
491
+ 0.9999998807907104,
492
+ 0.996260941028595,
493
+ 0.9989031553268433,
494
+ 0.9639995098114014,
495
+ 0.9999957084655762,
496
+ 1.0,
497
+ 0.9999998807907104,
498
+ 0.9980409741401672,
499
+ 0.8687159419059753,
500
+ 0.821354866027832,
501
+ 0.4796599745750427,
502
+ 0.4628332555294037,
503
+ 0.5007415413856506,
504
+ 0.3400116562843323,
505
+ 0.6277292966842651,
506
+ 0.5096919536590576,
507
+ 0.9999779462814331,
508
+ 0.9994074106216431,
509
+ 0.596235454082489,
510
+ 0.5236307382583618,
511
+ 0.44950583577156067,
512
+ 0.5437645316123962,
513
+ 0.9999998807907104,
514
+ 0.9999997615814209,
515
+ 0.9999998807907104,
516
+ 0.9999169111251831,
517
+ 0.9999459981918335,
518
+ 0.9999337196350098,
519
+ 0.9902818202972412,
520
+ 0.9997207522392273,
521
+ 0.5093064308166504,
522
+ 0.870800793170929
523
+ ],
524
+ "eval_code_hard_runtime": 110.6115,
525
+ "eval_code_hard_samples_per_second": 0.904,
526
+ "eval_code_hard_steps_per_second": 0.036,
527
+ "epoch": 0.88,
528
+ "step": 50
529
+ },
530
+ {
531
+ "loss": 0.3248,
532
+ "learning_rate": 0.0002,
533
+ "epoch": 0.89,
534
+ "step": 51
535
+ },
536
+ {
537
+ "loss": 0.1904,
538
+ "learning_rate": 0.0002,
539
+ "epoch": 0.91,
540
+ "step": 52
541
+ },
542
+ {
543
+ "loss": 0.185,
544
+ "learning_rate": 0.0002,
545
+ "epoch": 0.93,
546
+ "step": 53
547
+ },
548
+ {
549
+ "loss": 0.0932,
550
+ "learning_rate": 0.0002,
551
+ "epoch": 0.95,
552
+ "step": 54
553
+ },
554
+ {
555
+ "loss": 0.1643,
556
+ "learning_rate": 0.0002,
557
+ "epoch": 0.96,
558
+ "step": 55
559
+ },
560
+ {
561
+ "loss": 0.2504,
562
+ "learning_rate": 0.0002,
563
+ "epoch": 0.98,
564
+ "step": 56
565
+ },
566
+ {
567
+ "loss": 0.17,
568
+ "learning_rate": 0.0002,
569
+ "epoch": 1.0,
570
+ "step": 57
571
+ },
572
+ {
573
+ "loss": 0.1939,
574
+ "learning_rate": 0.0002,
575
+ "epoch": 1.02,
576
+ "step": 58
577
+ },
578
+ {
579
+ "loss": 0.1383,
580
+ "learning_rate": 0.0002,
581
+ "epoch": 1.04,
582
+ "step": 59
583
+ },
584
+ {
585
+ "loss": 0.149,
586
+ "learning_rate": 0.0002,
587
+ "epoch": 1.05,
588
+ "step": 60
589
+ },
590
+ {
591
+ "loss": 0.0699,
592
+ "learning_rate": 0.0002,
593
+ "epoch": 1.07,
594
+ "step": 61
595
+ },
596
+ {
597
+ "loss": 0.1258,
598
+ "learning_rate": 0.0002,
599
+ "epoch": 1.09,
600
+ "step": 62
601
+ },
602
+ {
603
+ "loss": 0.0563,
604
+ "learning_rate": 0.0002,
605
+ "epoch": 1.11,
606
+ "step": 63
607
+ },
608
+ {
609
+ "loss": 0.1171,
610
+ "learning_rate": 0.0002,
611
+ "epoch": 1.12,
612
+ "step": 64
613
+ },
614
+ {
615
+ "loss": 0.1197,
616
+ "learning_rate": 0.0002,
617
+ "epoch": 1.14,
618
+ "step": 65
619
+ },
620
+ {
621
+ "loss": 0.0812,
622
+ "learning_rate": 0.0002,
623
+ "epoch": 1.16,
624
+ "step": 66
625
+ },
626
+ {
627
+ "loss": 0.1488,
628
+ "learning_rate": 0.0002,
629
+ "epoch": 1.18,
630
+ "step": 67
631
+ },
632
+ {
633
+ "loss": 0.1459,
634
+ "learning_rate": 0.0002,
635
+ "epoch": 1.19,
636
+ "step": 68
637
+ },
638
+ {
639
+ "loss": 0.0691,
640
+ "learning_rate": 0.0002,
641
+ "epoch": 1.21,
642
+ "step": 69
643
+ },
644
+ {
645
+ "loss": 0.0529,
646
+ "learning_rate": 0.0002,
647
+ "epoch": 1.23,
648
+ "step": 70
649
+ },
650
+ {
651
+ "loss": 0.0443,
652
+ "learning_rate": 0.0002,
653
+ "epoch": 1.25,
654
+ "step": 71
655
+ },
656
+ {
657
+ "loss": 0.214,
658
+ "learning_rate": 0.0002,
659
+ "epoch": 1.26,
660
+ "step": 72
661
+ },
662
+ {
663
+ "loss": 0.0409,
664
+ "learning_rate": 0.0002,
665
+ "epoch": 1.28,
666
+ "step": 73
667
+ },
668
+ {
669
+ "loss": 0.0357,
670
+ "learning_rate": 0.0002,
671
+ "epoch": 1.3,
672
+ "step": 74
673
+ },
674
+ {
675
+ "loss": 0.0676,
676
+ "learning_rate": 0.0002,
677
+ "epoch": 1.32,
678
+ "step": 75
679
+ },
680
+ {
681
+ "eval_code_hard_loss": 0.15125365555286407,
682
+ "eval_code_hard_score": -0.047643136233091354,
683
+ "eval_code_hard_brier_score": 0.047643136233091354,
684
+ "eval_code_hard_average_probability": 0.9072666764259338,
685
+ "eval_code_hard_accuracy": 0.94,
686
+ "eval_code_hard_probabilities": [
687
+ 0.9881106615066528,
688
+ 0.67057204246521,
689
+ 0.9074558615684509,
690
+ 1.0,
691
+ 1.0,
692
+ 1.0,
693
+ 0.9999914169311523,
694
+ 0.9999996423721313,
695
+ 0.9999991655349731,
696
+ 0.8589997291564941,
697
+ 0.9643713235855103,
698
+ 0.9970460534095764,
699
+ 1.0,
700
+ 1.0,
701
+ 1.0,
702
+ 0.9999948740005493,
703
+ 0.9998992681503296,
704
+ 0.9999996423721313,
705
+ 1.0,
706
+ 1.0,
707
+ 1.0,
708
+ 0.9999997615814209,
709
+ 1.0,
710
+ 0.9999988079071045,
711
+ 0.9999991655349731,
712
+ 0.9999972581863403,
713
+ 0.9980295300483704,
714
+ 0.9999752044677734,
715
+ 0.9991430044174194,
716
+ 0.999230146408081,
717
+ 0.9863813519477844,
718
+ 0.9973239898681641,
719
+ 0.8935987949371338,
720
+ 1.0,
721
+ 0.9999996423721313,
722
+ 0.9999998807907104,
723
+ 0.9999157190322876,
724
+ 0.9997393488883972,
725
+ 0.999843955039978,
726
+ 0.8906758427619934,
727
+ 0.7957176566123962,
728
+ 0.8944621682167053,
729
+ 0.9999850988388062,
730
+ 0.9999994039535522,
731
+ 0.9999984502792358,
732
+ 0.8667024970054626,
733
+ 0.5891015529632568,
734
+ 0.5897384881973267,
735
+ 0.9999762773513794,
736
+ 0.920762836933136,
737
+ 0.8804107904434204,
738
+ 0.037825245410203934,
739
+ 0.9987059831619263,
740
+ 0.9998573064804077,
741
+ 1.0,
742
+ 1.0,
743
+ 1.0,
744
+ 0.5703369379043579,
745
+ 0.6154165863990784,
746
+ 0.47148749232292175,
747
+ 1.0,
748
+ 1.0,
749
+ 1.0,
750
+ 1.0,
751
+ 0.9997634291648865,
752
+ 0.9999960660934448,
753
+ 0.9999247789382935,
754
+ 0.9999982118606567,
755
+ 0.9999983310699463,
756
+ 0.7629109025001526,
757
+ 0.9476748108863831,
758
+ 0.7605304718017578,
759
+ 0.999997615814209,
760
+ 0.9999995231628418,
761
+ 0.9999983310699463,
762
+ 0.9999594688415527,
763
+ 0.9989461302757263,
764
+ 0.99898761510849,
765
+ 0.5305535793304443,
766
+ 0.7419759631156921,
767
+ 0.9985262155532837,
768
+ 0.9876091480255127,
769
+ 0.9818170070648193,
770
+ 0.8375343084335327,
771
+ 0.9999997615814209,
772
+ 0.9999995231628418,
773
+ 0.07670325040817261,
774
+ 0.6875349283218384,
775
+ 0.35709840059280396,
776
+ 0.3490428626537323,
777
+ 0.99744713306427,
778
+ 0.9997475743293762,
779
+ 0.9991705417633057,
780
+ 0.9985999464988708,
781
+ 0.9992005228996277,
782
+ 0.9987180233001709,
783
+ 0.9907884001731873,
784
+ 0.9931132197380066,
785
+ 0.35269367694854736,
786
+ 0.9993346333503723
787
+ ],
788
+ "eval_code_hard_runtime": 110.7336,
789
+ "eval_code_hard_samples_per_second": 0.903,
790
+ "eval_code_hard_steps_per_second": 0.036,
791
+ "epoch": 1.32,
792
+ "step": 75
793
+ },
794
+ {
795
+ "loss": 0.0635,
796
+ "learning_rate": 0.0002,
797
+ "epoch": 1.33,
798
+ "step": 76
799
+ },
800
+ {
801
+ "loss": 0.0923,
802
+ "learning_rate": 0.0002,
803
+ "epoch": 1.35,
804
+ "step": 77
805
+ },
806
+ {
807
+ "loss": 0.1026,
808
+ "learning_rate": 0.0002,
809
+ "epoch": 1.37,
810
+ "step": 78
811
+ },
812
+ {
813
+ "loss": 0.0674,
814
+ "learning_rate": 0.0002,
815
+ "epoch": 1.39,
816
+ "step": 79
817
+ },
818
+ {
819
+ "loss": 0.0493,
820
+ "learning_rate": 0.0002,
821
+ "epoch": 1.4,
822
+ "step": 80
823
+ },
824
+ {
825
+ "loss": 0.1209,
826
+ "learning_rate": 0.0002,
827
+ "epoch": 1.42,
828
+ "step": 81
829
+ },
830
+ {
831
+ "loss": 0.5175,
832
+ "learning_rate": 0.0002,
833
+ "epoch": 1.44,
834
+ "step": 82
835
+ },
836
+ {
837
+ "loss": 0.1547,
838
+ "learning_rate": 0.0002,
839
+ "epoch": 1.46,
840
+ "step": 83
841
+ },
842
+ {
843
+ "loss": 0.1185,
844
+ "learning_rate": 0.0002,
845
+ "epoch": 1.47,
846
+ "step": 84
847
+ },
848
+ {
849
+ "loss": 0.0621,
850
+ "learning_rate": 0.0002,
851
+ "epoch": 1.49,
852
+ "step": 85
853
+ },
854
+ {
855
+ "loss": 0.07,
856
+ "learning_rate": 0.0002,
857
+ "epoch": 1.51,
858
+ "step": 86
859
+ },
860
+ {
861
+ "loss": 0.092,
862
+ "learning_rate": 0.0002,
863
+ "epoch": 1.53,
864
+ "step": 87
865
+ },
866
+ {
867
+ "loss": 0.0422,
868
+ "learning_rate": 0.0002,
869
+ "epoch": 1.54,
870
+ "step": 88
871
+ },
872
+ {
873
+ "loss": 0.1117,
874
+ "learning_rate": 0.0002,
875
+ "epoch": 1.56,
876
+ "step": 89
877
+ },
878
+ {
879
+ "loss": 0.0654,
880
+ "learning_rate": 0.0002,
881
+ "epoch": 1.58,
882
+ "step": 90
883
+ },
884
+ {
885
+ "loss": 0.1185,
886
+ "learning_rate": 0.0002,
887
+ "epoch": 1.6,
888
+ "step": 91
889
+ },
890
+ {
891
+ "loss": 0.1308,
892
+ "learning_rate": 0.0002,
893
+ "epoch": 1.61,
894
+ "step": 92
895
+ },
896
+ {
897
+ "loss": 0.1293,
898
+ "learning_rate": 0.0002,
899
+ "epoch": 1.63,
900
+ "step": 93
901
+ },
902
+ {
903
+ "loss": 0.1071,
904
+ "learning_rate": 0.0002,
905
+ "epoch": 1.65,
906
+ "step": 94
907
+ },
908
+ {
909
+ "loss": 0.0593,
910
+ "learning_rate": 0.0002,
911
+ "epoch": 1.67,
912
+ "step": 95
913
+ },
914
+ {
915
+ "loss": 0.0494,
916
+ "learning_rate": 0.0002,
917
+ "epoch": 1.68,
918
+ "step": 96
919
+ },
920
+ {
921
+ "loss": 0.0641,
922
+ "learning_rate": 0.0002,
923
+ "epoch": 1.7,
924
+ "step": 97
925
+ },
926
+ {
927
+ "loss": 0.0815,
928
+ "learning_rate": 0.0002,
929
+ "epoch": 1.72,
930
+ "step": 98
931
+ },
932
+ {
933
+ "loss": 0.0747,
934
+ "learning_rate": 0.0002,
935
+ "epoch": 1.74,
936
+ "step": 99
937
+ },
938
+ {
939
+ "loss": 0.0776,
940
+ "learning_rate": 0.0002,
941
+ "epoch": 1.75,
942
+ "step": 100
943
+ },
944
+ {
945
+ "eval_code_hard_loss": 0.09847890585660934,
946
+ "eval_code_hard_score": -0.035213593393564224,
947
+ "eval_code_hard_brier_score": 0.035213593393564224,
948
+ "eval_code_hard_average_probability": 0.9248704314231873,
949
+ "eval_code_hard_accuracy": 0.94,
950
+ "eval_code_hard_probabilities": [
951
+ 0.9995959401130676,
952
+ 0.823664665222168,
953
+ 0.9978653788566589,
954
+ 1.0,
955
+ 1.0,
956
+ 1.0,
957
+ 1.0,
958
+ 1.0,
959
+ 1.0,
960
+ 0.7222961187362671,
961
+ 0.9800193309783936,
962
+ 0.9987350106239319,
963
+ 1.0,
964
+ 1.0,
965
+ 1.0,
966
+ 1.0,
967
+ 0.999994158744812,
968
+ 1.0,
969
+ 0.303742915391922,
970
+ 0.9999943971633911,
971
+ 0.9999933242797852,
972
+ 1.0,
973
+ 1.0,
974
+ 1.0,
975
+ 1.0,
976
+ 0.9999995231628418,
977
+ 0.9999909400939941,
978
+ 0.9999997615814209,
979
+ 1.0,
980
+ 1.0,
981
+ 0.999997615814209,
982
+ 1.0,
983
+ 0.9949355721473694,
984
+ 1.0,
985
+ 0.9999887943267822,
986
+ 0.999993085861206,
987
+ 1.0,
988
+ 1.0,
989
+ 1.0,
990
+ 0.9999736547470093,
991
+ 0.9997861981391907,
992
+ 0.9999978542327881,
993
+ 1.0,
994
+ 1.0,
995
+ 1.0,
996
+ 0.9989193677902222,
997
+ 0.6210824847221375,
998
+ 0.8116233944892883,
999
+ 0.9997373223304749,
1000
+ 0.8275205492973328,
1001
+ 0.7434841394424438,
1002
+ 0.27361589670181274,
1003
+ 0.9993396401405334,
1004
+ 0.9998682737350464,
1005
+ 0.999985933303833,
1006
+ 0.9999828338623047,
1007
+ 0.9999997615814209,
1008
+ 0.39626461267471313,
1009
+ 0.5240136384963989,
1010
+ 0.6546356081962585,
1011
+ 0.9999992847442627,
1012
+ 0.9999991655349731,
1013
+ 0.9999990463256836,
1014
+ 0.9999998807907104,
1015
+ 0.9612408876419067,
1016
+ 0.9947011470794678,
1017
+ 0.9999994039535522,
1018
+ 1.0,
1019
+ 1.0,
1020
+ 0.9999735355377197,
1021
+ 0.9999927282333374,
1022
+ 0.9999716281890869,
1023
+ 1.0,
1024
+ 1.0,
1025
+ 1.0,
1026
+ 0.9999995231628418,
1027
+ 0.9988247752189636,
1028
+ 0.9937974810600281,
1029
+ 0.49393144249916077,
1030
+ 0.4931151568889618,
1031
+ 0.7545955777168274,
1032
+ 0.918387770652771,
1033
+ 0.9984613656997681,
1034
+ 0.9579277634620667,
1035
+ 1.0,
1036
+ 0.9999995231628418,
1037
+ 0.7278538942337036,
1038
+ 0.5116866827011108,
1039
+ 0.9212220311164856,
1040
+ 0.7694430947303772,
1041
+ 1.0,
1042
+ 1.0,
1043
+ 1.0,
1044
+ 0.9999998807907104,
1045
+ 1.0,
1046
+ 1.0,
1047
+ 0.9999598264694214,
1048
+ 0.9999998807907104,
1049
+ 0.3216085433959961,
1050
+ 0.999714195728302
1051
+ ],
1052
+ "eval_code_hard_runtime": 110.6096,
1053
+ "eval_code_hard_samples_per_second": 0.904,
1054
+ "eval_code_hard_steps_per_second": 0.036,
1055
+ "epoch": 1.75,
1056
+ "step": 100
1057
+ },
1058
+ {
1059
+ "loss": 0.0619,
1060
+ "learning_rate": 0.0002,
1061
+ "epoch": 1.77,
1062
+ "step": 101
1063
+ },
1064
+ {
1065
+ "loss": 0.0526,
1066
+ "learning_rate": 0.0002,
1067
+ "epoch": 1.79,
1068
+ "step": 102
1069
+ },
1070
+ {
1071
+ "loss": 0.1902,
1072
+ "learning_rate": 0.0002,
1073
+ "epoch": 1.81,
1074
+ "step": 103
1075
+ },
1076
+ {
1077
+ "loss": 0.2676,
1078
+ "learning_rate": 0.0002,
1079
+ "epoch": 1.82,
1080
+ "step": 104
1081
+ },
1082
+ {
1083
+ "loss": 0.1422,
1084
+ "learning_rate": 0.0002,
1085
+ "epoch": 1.84,
1086
+ "step": 105
1087
+ },
1088
+ {
1089
+ "loss": 0.0869,
1090
+ "learning_rate": 0.0002,
1091
+ "epoch": 1.86,
1092
+ "step": 106
1093
+ },
1094
+ {
1095
+ "loss": 0.1517,
1096
+ "learning_rate": 0.0002,
1097
+ "epoch": 1.88,
1098
+ "step": 107
1099
+ },
1100
+ {
1101
+ "loss": 0.0282,
1102
+ "learning_rate": 0.0002,
1103
+ "epoch": 1.89,
1104
+ "step": 108
1105
+ },
1106
+ {
1107
+ "loss": 0.0516,
1108
+ "learning_rate": 0.0002,
1109
+ "epoch": 1.91,
1110
+ "step": 109
1111
+ },
1112
+ {
1113
+ "loss": 0.0479,
1114
+ "learning_rate": 0.0002,
1115
+ "epoch": 1.93,
1116
+ "step": 110
1117
+ },
1118
+ {
1119
+ "loss": 0.108,
1120
+ "learning_rate": 0.0002,
1121
+ "epoch": 1.95,
1122
+ "step": 111
1123
+ },
1124
+ {
1125
+ "loss": 0.0833,
1126
+ "learning_rate": 0.0002,
1127
+ "epoch": 1.96,
1128
+ "step": 112
1129
+ },
1130
+ {
1131
+ "loss": 0.1826,
1132
+ "learning_rate": 0.0002,
1133
+ "epoch": 1.98,
1134
+ "step": 113
1135
+ },
1136
+ {
1137
+ "loss": 0.071,
1138
+ "learning_rate": 0.0002,
1139
+ "epoch": 2.0,
1140
+ "step": 114
1141
+ },
1142
+ {
1143
+ "loss": 0.0619,
1144
+ "learning_rate": 0.0002,
1145
+ "epoch": 2.02,
1146
+ "step": 115
1147
+ },
1148
+ {
1149
+ "loss": 0.0764,
1150
+ "learning_rate": 0.0002,
1151
+ "epoch": 2.04,
1152
+ "step": 116
1153
+ },
1154
+ {
1155
+ "loss": 0.0009,
1156
+ "learning_rate": 0.0002,
1157
+ "epoch": 2.05,
1158
+ "step": 117
1159
+ },
1160
+ {
1161
+ "loss": 0.0175,
1162
+ "learning_rate": 0.0002,
1163
+ "epoch": 2.07,
1164
+ "step": 118
1165
+ },
1166
+ {
1167
+ "loss": 0.0216,
1168
+ "learning_rate": 0.0002,
1169
+ "epoch": 2.09,
1170
+ "step": 119
1171
+ },
1172
+ {
1173
+ "loss": 0.0278,
1174
+ "learning_rate": 0.0002,
1175
+ "epoch": 2.11,
1176
+ "step": 120
1177
+ },
1178
+ {
1179
+ "loss": 0.0665,
1180
+ "learning_rate": 0.0002,
1181
+ "epoch": 2.12,
1182
+ "step": 121
1183
+ },
1184
+ {
1185
+ "loss": 0.0205,
1186
+ "learning_rate": 0.0002,
1187
+ "epoch": 2.14,
1188
+ "step": 122
1189
+ },
1190
+ {
1191
+ "loss": 0.0052,
1192
+ "learning_rate": 0.0002,
1193
+ "epoch": 2.16,
1194
+ "step": 123
1195
+ },
1196
+ {
1197
+ "loss": 0.0635,
1198
+ "learning_rate": 0.0002,
1199
+ "epoch": 2.18,
1200
+ "step": 124
1201
+ },
1202
+ {
1203
+ "loss": 0.0132,
1204
+ "learning_rate": 0.0002,
1205
+ "epoch": 2.19,
1206
+ "step": 125
1207
+ },
1208
+ {
1209
+ "eval_code_hard_loss": 0.05377120152115822,
1210
+ "eval_code_hard_score": -0.015216915868222713,
1211
+ "eval_code_hard_brier_score": 0.015216915868222713,
1212
+ "eval_code_hard_average_probability": 0.9605898261070251,
1213
+ "eval_code_hard_accuracy": 0.99,
1214
+ "eval_code_hard_probabilities": [
1215
+ 0.9999129772186279,
1216
+ 0.9065911173820496,
1217
+ 0.9998986721038818,
1218
+ 1.0,
1219
+ 1.0,
1220
+ 1.0,
1221
+ 1.0,
1222
+ 1.0,
1223
+ 1.0,
1224
+ 0.8372133374214172,
1225
+ 0.9947729706764221,
1226
+ 0.999957799911499,
1227
+ 1.0,
1228
+ 1.0,
1229
+ 1.0,
1230
+ 1.0,
1231
+ 0.9999998807907104,
1232
+ 1.0,
1233
+ 0.9999995231628418,
1234
+ 1.0,
1235
+ 1.0,
1236
+ 1.0,
1237
+ 1.0,
1238
+ 1.0,
1239
+ 1.0,
1240
+ 1.0,
1241
+ 0.9999923706054688,
1242
+ 0.9862022995948792,
1243
+ 1.0,
1244
+ 1.0,
1245
+ 0.9995848536491394,
1246
+ 0.9999656677246094,
1247
+ 0.9963919520378113,
1248
+ 1.0,
1249
+ 0.9999942779541016,
1250
+ 0.999997615814209,
1251
+ 0.9999997615814209,
1252
+ 0.9999994039535522,
1253
+ 0.9999994039535522,
1254
+ 0.9998449087142944,
1255
+ 0.996722400188446,
1256
+ 0.9999366998672485,
1257
+ 0.9999986886978149,
1258
+ 0.9999994039535522,
1259
+ 0.9999995231628418,
1260
+ 0.9986975193023682,
1261
+ 0.6720824837684631,
1262
+ 0.8636956810951233,
1263
+ 0.9999996423721313,
1264
+ 0.9801883697509766,
1265
+ 0.9772735238075256,
1266
+ 0.22563716769218445,
1267
+ 0.9999665021896362,
1268
+ 0.9957283139228821,
1269
+ 1.0,
1270
+ 1.0,
1271
+ 1.0,
1272
+ 0.928858757019043,
1273
+ 0.7347054481506348,
1274
+ 0.8363421559333801,
1275
+ 1.0,
1276
+ 1.0,
1277
+ 1.0,
1278
+ 1.0,
1279
+ 0.9992239475250244,
1280
+ 0.9996882677078247,
1281
+ 0.9999969005584717,
1282
+ 1.0,
1283
+ 1.0,
1284
+ 0.9928505420684814,
1285
+ 0.9979680180549622,
1286
+ 0.9968575239181519,
1287
+ 1.0,
1288
+ 1.0,
1289
+ 1.0,
1290
+ 0.9999997615814209,
1291
+ 0.9999778270721436,
1292
+ 0.9998857975006104,
1293
+ 0.5365374684333801,
1294
+ 0.6125951409339905,
1295
+ 0.9207331538200378,
1296
+ 0.9939044117927551,
1297
+ 0.9994903802871704,
1298
+ 0.9971593618392944,
1299
+ 1.0,
1300
+ 1.0,
1301
+ 0.8098579049110413,
1302
+ 0.5238723158836365,
1303
+ 0.9532552361488342,
1304
+ 0.8978452682495117,
1305
+ 1.0,
1306
+ 1.0,
1307
+ 1.0,
1308
+ 0.9999998807907104,
1309
+ 1.0,
1310
+ 1.0,
1311
+ 0.9999985694885254,
1312
+ 1.0,
1313
+ 0.8971565961837769,
1314
+ 0.9999786615371704
1315
+ ],
1316
+ "eval_code_hard_runtime": 110.7219,
1317
+ "eval_code_hard_samples_per_second": 0.903,
1318
+ "eval_code_hard_steps_per_second": 0.036,
1319
+ "epoch": 2.19,
1320
+ "step": 125
1321
+ },
1322
+ {
1323
+ "loss": 0.0251,
1324
+ "learning_rate": 0.0002,
1325
+ "epoch": 2.21,
1326
+ "step": 126
1327
+ },
1328
+ {
1329
+ "loss": 0.0102,
1330
+ "learning_rate": 0.0002,
1331
+ "epoch": 2.23,
1332
+ "step": 127
1333
+ },
1334
+ {
1335
+ "loss": 0.0179,
1336
+ "learning_rate": 0.0002,
1337
+ "epoch": 2.25,
1338
+ "step": 128
1339
+ },
1340
+ {
1341
+ "loss": 0.1329,
1342
+ "learning_rate": 0.0002,
1343
+ "epoch": 2.26,
1344
+ "step": 129
1345
+ },
1346
+ {
1347
+ "loss": 0.0082,
1348
+ "learning_rate": 0.0002,
1349
+ "epoch": 2.28,
1350
+ "step": 130
1351
+ },
1352
+ {
1353
+ "loss": 0.0325,
1354
+ "learning_rate": 0.0002,
1355
+ "epoch": 2.3,
1356
+ "step": 131
1357
+ },
1358
+ {
1359
+ "loss": 0.0094,
1360
+ "learning_rate": 0.0002,
1361
+ "epoch": 2.32,
1362
+ "step": 132
1363
+ },
1364
+ {
1365
+ "loss": 0.004,
1366
+ "learning_rate": 0.0002,
1367
+ "epoch": 2.33,
1368
+ "step": 133
1369
+ },
1370
+ {
1371
+ "loss": 0.0099,
1372
+ "learning_rate": 0.0002,
1373
+ "epoch": 2.35,
1374
+ "step": 134
1375
+ },
1376
+ {
1377
+ "loss": 0.0454,
1378
+ "learning_rate": 0.0002,
1379
+ "epoch": 2.37,
1380
+ "step": 135
1381
+ },
1382
+ {
1383
+ "loss": 0.0342,
1384
+ "learning_rate": 0.0002,
1385
+ "epoch": 2.39,
1386
+ "step": 136
1387
+ },
1388
+ {
1389
+ "loss": 0.0161,
1390
+ "learning_rate": 0.0002,
1391
+ "epoch": 2.4,
1392
+ "step": 137
1393
+ },
1394
+ {
1395
+ "loss": 0.0686,
1396
+ "learning_rate": 0.0002,
1397
+ "epoch": 2.42,
1398
+ "step": 138
1399
+ },
1400
+ {
1401
+ "loss": 0.0404,
1402
+ "learning_rate": 0.0002,
1403
+ "epoch": 2.44,
1404
+ "step": 139
1405
+ },
1406
+ {
1407
+ "loss": 0.0311,
1408
+ "learning_rate": 0.0002,
1409
+ "epoch": 2.46,
1410
+ "step": 140
1411
+ },
1412
+ {
1413
+ "loss": 0.0426,
1414
+ "learning_rate": 0.0002,
1415
+ "epoch": 2.47,
1416
+ "step": 141
1417
+ },
1418
+ {
1419
+ "loss": 0.013,
1420
+ "learning_rate": 0.0002,
1421
+ "epoch": 2.49,
1422
+ "step": 142
1423
+ },
1424
+ {
1425
+ "loss": 0.0209,
1426
+ "learning_rate": 0.0002,
1427
+ "epoch": 2.51,
1428
+ "step": 143
1429
+ },
1430
+ {
1431
+ "loss": 0.0441,
1432
+ "learning_rate": 0.0002,
1433
+ "epoch": 2.53,
1434
+ "step": 144
1435
+ },
1436
+ {
1437
+ "loss": 0.0414,
1438
+ "learning_rate": 0.0002,
1439
+ "epoch": 2.54,
1440
+ "step": 145
1441
+ },
1442
+ {
1443
+ "loss": 0.0022,
1444
+ "learning_rate": 0.0002,
1445
+ "epoch": 2.56,
1446
+ "step": 146
1447
+ },
1448
+ {
1449
+ "loss": 0.0098,
1450
+ "learning_rate": 0.0002,
1451
+ "epoch": 2.58,
1452
+ "step": 147
1453
+ },
1454
+ {
1455
+ "loss": 0.0076,
1456
+ "learning_rate": 0.0002,
1457
+ "epoch": 2.6,
1458
+ "step": 148
1459
+ },
1460
+ {
1461
+ "loss": 0.0242,
1462
+ "learning_rate": 0.0002,
1463
+ "epoch": 2.61,
1464
+ "step": 149
1465
+ },
1466
+ {
1467
+ "loss": 0.002,
1468
+ "learning_rate": 0.0002,
1469
+ "epoch": 2.63,
1470
+ "step": 150
1471
+ },
1472
+ {
1473
+ "eval_code_hard_loss": 0.06451280415058136,
1474
+ "eval_code_hard_score": -0.019366176798939705,
1475
+ "eval_code_hard_brier_score": 0.019366176798939705,
1476
+ "eval_code_hard_average_probability": 0.957671582698822,
1477
+ "eval_code_hard_accuracy": 0.97,
1478
+ "eval_code_hard_probabilities": [
1479
+ 0.9999499320983887,
1480
+ 0.8330075144767761,
1481
+ 0.9997716546058655,
1482
+ 1.0,
1483
+ 1.0,
1484
+ 1.0,
1485
+ 1.0,
1486
+ 1.0,
1487
+ 1.0,
1488
+ 0.4579625427722931,
1489
+ 0.9998301267623901,
1490
+ 1.0,
1491
+ 1.0,
1492
+ 1.0,
1493
+ 1.0,
1494
+ 1.0,
1495
+ 1.0,
1496
+ 1.0,
1497
+ 0.9999998807907104,
1498
+ 1.0,
1499
+ 1.0,
1500
+ 1.0,
1501
+ 1.0,
1502
+ 1.0,
1503
+ 1.0,
1504
+ 1.0,
1505
+ 1.0,
1506
+ 0.9999909400939941,
1507
+ 1.0,
1508
+ 1.0,
1509
+ 0.9999996423721313,
1510
+ 1.0,
1511
+ 0.9973661303520203,
1512
+ 1.0,
1513
+ 0.9999998807907104,
1514
+ 1.0,
1515
+ 0.9999984502792358,
1516
+ 0.9999953508377075,
1517
+ 0.9999986886978149,
1518
+ 1.0,
1519
+ 0.9999997615814209,
1520
+ 1.0,
1521
+ 1.0,
1522
+ 1.0,
1523
+ 1.0,
1524
+ 0.9999028444290161,
1525
+ 0.7898390293121338,
1526
+ 0.8942784070968628,
1527
+ 1.0,
1528
+ 0.9987273812294006,
1529
+ 0.9987801909446716,
1530
+ 0.14106149971485138,
1531
+ 1.0,
1532
+ 0.9999027252197266,
1533
+ 1.0,
1534
+ 1.0,
1535
+ 1.0,
1536
+ 0.6841090321540833,
1537
+ 0.7871870398521423,
1538
+ 0.896818220615387,
1539
+ 1.0,
1540
+ 1.0,
1541
+ 1.0,
1542
+ 1.0,
1543
+ 0.999897837638855,
1544
+ 0.9999984502792358,
1545
+ 0.9999983310699463,
1546
+ 1.0,
1547
+ 1.0,
1548
+ 0.9256799817085266,
1549
+ 0.9996222257614136,
1550
+ 0.9997231364250183,
1551
+ 1.0,
1552
+ 1.0,
1553
+ 1.0,
1554
+ 1.0,
1555
+ 0.9999997615814209,
1556
+ 0.9999966621398926,
1557
+ 0.5312544107437134,
1558
+ 0.583741307258606,
1559
+ 0.9588684439659119,
1560
+ 0.9209092259407043,
1561
+ 0.999733030796051,
1562
+ 0.9926868677139282,
1563
+ 1.0,
1564
+ 1.0,
1565
+ 0.9422075748443604,
1566
+ 0.4956146478652954,
1567
+ 0.9603027105331421,
1568
+ 0.9903588891029358,
1569
+ 1.0,
1570
+ 1.0,
1571
+ 1.0,
1572
+ 1.0,
1573
+ 1.0,
1574
+ 1.0,
1575
+ 1.0,
1576
+ 1.0,
1577
+ 0.988089382648468,
1578
+ 0.9999991655349731
1579
+ ],
1580
+ "eval_code_hard_runtime": 110.671,
1581
+ "eval_code_hard_samples_per_second": 0.904,
1582
+ "eval_code_hard_steps_per_second": 0.036,
1583
+ "epoch": 2.63,
1584
+ "step": 150
1585
+ },
1586
+ {
1587
+ "train_runtime": 14190.0736,
1588
+ "train_samples_per_second": 0.338,
1589
+ "train_steps_per_second": 0.011,
1590
+ "total_flos": 0.0,
1591
+ "train_loss": 0.17637818212776135,
1592
+ "epoch": 2.63,
1593
+ "step": 150
1594
+ }
1595
+ ]]