Commit ·
0024eab
1
Parent(s): 13ee241
Upload folder using huggingface_hub
Browse files- README.md +21 -0
- adapter_config.json +21 -0
- adapter_model.bin +3 -0
- special_tokens_map.json +24 -0
- tokenizer.json +0 -0
- tokenizer.model +3 -0
- tokenizer_config.json +32 -0
- train_args.json +123 -0
- training_args.bin +3 -0
- training_logs.json +1601 -0
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 sha256:2551a0219c2d8b838b7c8c6d9a2c83d6e152a97ab806cfbebf769fd468ad38de
|
| 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/classify_lora/llama-13b-math/checkpoint-150",
|
| 3 |
+
"distribution_id": "math",
|
| 4 |
+
"date_trained": "10/10/2023 05:33:16",
|
| 5 |
+
"output_dir": "models/classify_lora/llama-13b-math",
|
| 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-math/runs/Oct10_05-31-36_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": 0,
|
| 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-math",
|
| 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_math_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 sha256:6960c4973f74be9a9607bdd958eb74bf6bc53c869043172ac32a0626ba3af2dd
|
| 3 |
+
size 5115
|
training_logs.json
ADDED
|
@@ -0,0 +1,1601 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 0.02,
|
| 4 |
+
"learning_rate": 0.0002,
|
| 5 |
+
"loss": 0.7128,
|
| 6 |
+
"step": 1
|
| 7 |
+
},
|
| 8 |
+
{
|
| 9 |
+
"epoch": 0.04,
|
| 10 |
+
"learning_rate": 0.0002,
|
| 11 |
+
"loss": 0.7061,
|
| 12 |
+
"step": 2
|
| 13 |
+
},
|
| 14 |
+
{
|
| 15 |
+
"epoch": 0.05,
|
| 16 |
+
"learning_rate": 0.0002,
|
| 17 |
+
"loss": 0.686,
|
| 18 |
+
"step": 3
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"epoch": 0.07,
|
| 22 |
+
"learning_rate": 0.0002,
|
| 23 |
+
"loss": 0.6748,
|
| 24 |
+
"step": 4
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.09,
|
| 28 |
+
"learning_rate": 0.0002,
|
| 29 |
+
"loss": 0.6941,
|
| 30 |
+
"step": 5
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.11,
|
| 34 |
+
"learning_rate": 0.0002,
|
| 35 |
+
"loss": 0.6695,
|
| 36 |
+
"step": 6
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 0.12,
|
| 40 |
+
"learning_rate": 0.0002,
|
| 41 |
+
"loss": 0.6602,
|
| 42 |
+
"step": 7
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 0.14,
|
| 46 |
+
"learning_rate": 0.0002,
|
| 47 |
+
"loss": 0.6737,
|
| 48 |
+
"step": 8
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.16,
|
| 52 |
+
"learning_rate": 0.0002,
|
| 53 |
+
"loss": 0.6839,
|
| 54 |
+
"step": 9
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.18,
|
| 58 |
+
"learning_rate": 0.0002,
|
| 59 |
+
"loss": 0.6513,
|
| 60 |
+
"step": 10
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.19,
|
| 64 |
+
"learning_rate": 0.0002,
|
| 65 |
+
"loss": 0.6545,
|
| 66 |
+
"step": 11
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.21,
|
| 70 |
+
"learning_rate": 0.0002,
|
| 71 |
+
"loss": 0.6413,
|
| 72 |
+
"step": 12
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.23,
|
| 76 |
+
"learning_rate": 0.0002,
|
| 77 |
+
"loss": 0.6252,
|
| 78 |
+
"step": 13
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 0.25,
|
| 82 |
+
"learning_rate": 0.0002,
|
| 83 |
+
"loss": 0.6435,
|
| 84 |
+
"step": 14
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.26,
|
| 88 |
+
"learning_rate": 0.0002,
|
| 89 |
+
"loss": 0.6541,
|
| 90 |
+
"step": 15
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.28,
|
| 94 |
+
"learning_rate": 0.0002,
|
| 95 |
+
"loss": 0.5772,
|
| 96 |
+
"step": 16
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.3,
|
| 100 |
+
"learning_rate": 0.0002,
|
| 101 |
+
"loss": 0.6002,
|
| 102 |
+
"step": 17
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.32,
|
| 106 |
+
"learning_rate": 0.0002,
|
| 107 |
+
"loss": 0.6551,
|
| 108 |
+
"step": 18
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.33,
|
| 112 |
+
"learning_rate": 0.0002,
|
| 113 |
+
"loss": 0.5857,
|
| 114 |
+
"step": 19
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.35,
|
| 118 |
+
"learning_rate": 0.0002,
|
| 119 |
+
"loss": 0.6159,
|
| 120 |
+
"step": 20
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 0.37,
|
| 124 |
+
"learning_rate": 0.0002,
|
| 125 |
+
"loss": 0.5753,
|
| 126 |
+
"step": 21
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 0.39,
|
| 130 |
+
"learning_rate": 0.0002,
|
| 131 |
+
"loss": 0.5567,
|
| 132 |
+
"step": 22
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 0.4,
|
| 136 |
+
"learning_rate": 0.0002,
|
| 137 |
+
"loss": 0.5487,
|
| 138 |
+
"step": 23
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.42,
|
| 142 |
+
"learning_rate": 0.0002,
|
| 143 |
+
"loss": 0.5023,
|
| 144 |
+
"step": 24
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.44,
|
| 148 |
+
"learning_rate": 0.0002,
|
| 149 |
+
"loss": 0.4681,
|
| 150 |
+
"step": 25
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.44,
|
| 154 |
+
"eval_math_accuracy": 0.93,
|
| 155 |
+
"eval_math_average_probability": 0.6979916095733643,
|
| 156 |
+
"eval_math_brier_score": 0.11860229820013046,
|
| 157 |
+
"eval_math_loss": 0.3839128017425537,
|
| 158 |
+
"eval_math_probabilities": [
|
| 159 |
+
0.5415274500846863,
|
| 160 |
+
0.5390767455101013,
|
| 161 |
+
0.5601692795753479,
|
| 162 |
+
0.6123054623603821,
|
| 163 |
+
0.6459183096885681,
|
| 164 |
+
0.6191246509552002,
|
| 165 |
+
0.9590169787406921,
|
| 166 |
+
0.932368278503418,
|
| 167 |
+
0.9546666145324707,
|
| 168 |
+
0.6975629925727844,
|
| 169 |
+
0.7943072319030762,
|
| 170 |
+
0.8020127415657043,
|
| 171 |
+
0.7124451994895935,
|
| 172 |
+
0.4208241105079651,
|
| 173 |
+
0.5985718369483948,
|
| 174 |
+
0.5591943860054016,
|
| 175 |
+
0.5160695314407349,
|
| 176 |
+
0.5296780467033386,
|
| 177 |
+
0.5664644837379456,
|
| 178 |
+
0.6503645777702332,
|
| 179 |
+
0.6152858734130859,
|
| 180 |
+
0.5448443293571472,
|
| 181 |
+
0.4932275414466858,
|
| 182 |
+
0.5111082196235657,
|
| 183 |
+
0.6046796441078186,
|
| 184 |
+
0.7649803161621094,
|
| 185 |
+
0.5845022201538086,
|
| 186 |
+
0.6277315020561218,
|
| 187 |
+
0.8466326594352722,
|
| 188 |
+
0.7372856736183167,
|
| 189 |
+
0.47781234979629517,
|
| 190 |
+
0.4657546877861023,
|
| 191 |
+
0.45426762104034424,
|
| 192 |
+
0.9421902298927307,
|
| 193 |
+
0.9553028345108032,
|
| 194 |
+
0.95943284034729,
|
| 195 |
+
0.81365966796875,
|
| 196 |
+
0.8718599081039429,
|
| 197 |
+
0.8876155018806458,
|
| 198 |
+
0.5101025700569153,
|
| 199 |
+
0.424481064081192,
|
| 200 |
+
0.4407404065132141,
|
| 201 |
+
0.8702844381332397,
|
| 202 |
+
0.9398714303970337,
|
| 203 |
+
0.9255726933479309,
|
| 204 |
+
0.5184469223022461,
|
| 205 |
+
0.6204009652137756,
|
| 206 |
+
0.6979585289955139,
|
| 207 |
+
0.7835100293159485,
|
| 208 |
+
0.9228054881095886,
|
| 209 |
+
0.5449931621551514,
|
| 210 |
+
0.6148080229759216,
|
| 211 |
+
0.6196415424346924,
|
| 212 |
+
0.6206129193305969,
|
| 213 |
+
0.5946707725524902,
|
| 214 |
+
0.5806065797805786,
|
| 215 |
+
0.69861900806427,
|
| 216 |
+
0.9202491044998169,
|
| 217 |
+
0.8827500939369202,
|
| 218 |
+
0.9218969941139221,
|
| 219 |
+
0.827384889125824,
|
| 220 |
+
0.6205788254737854,
|
| 221 |
+
0.6759516596794128,
|
| 222 |
+
0.6175774931907654,
|
| 223 |
+
0.6537926197052002,
|
| 224 |
+
0.790869951248169,
|
| 225 |
+
0.5193009376525879,
|
| 226 |
+
0.5676986575126648,
|
| 227 |
+
0.6558927893638611,
|
| 228 |
+
0.7213096022605896,
|
| 229 |
+
0.8478593230247498,
|
| 230 |
+
0.7222801446914673,
|
| 231 |
+
0.6536849737167358,
|
| 232 |
+
0.6680243015289307,
|
| 233 |
+
0.6657971739768982,
|
| 234 |
+
0.9944479465484619,
|
| 235 |
+
0.9916308522224426,
|
| 236 |
+
0.9944629073143005,
|
| 237 |
+
0.9825392961502075,
|
| 238 |
+
0.9810364246368408,
|
| 239 |
+
0.979194700717926,
|
| 240 |
+
0.5630966424942017,
|
| 241 |
+
0.558582603931427,
|
| 242 |
+
0.5793667435646057,
|
| 243 |
+
0.5963504910469055,
|
| 244 |
+
0.7267915606498718,
|
| 245 |
+
0.6603485345840454,
|
| 246 |
+
0.6573605537414551,
|
| 247 |
+
0.6383423209190369,
|
| 248 |
+
0.598394513130188,
|
| 249 |
+
0.9555463790893555,
|
| 250 |
+
0.9942784309387207,
|
| 251 |
+
0.9970775842666626,
|
| 252 |
+
0.5390485525131226,
|
| 253 |
+
0.6269036531448364,
|
| 254 |
+
0.5502526760101318,
|
| 255 |
+
0.5693672895431519,
|
| 256 |
+
0.6798619627952576,
|
| 257 |
+
0.6312905550003052,
|
| 258 |
+
0.5227184891700745
|
| 259 |
+
],
|
| 260 |
+
"eval_math_runtime": 71.4058,
|
| 261 |
+
"eval_math_samples_per_second": 1.4,
|
| 262 |
+
"eval_math_score": -0.11860229820013046,
|
| 263 |
+
"eval_math_steps_per_second": 0.056,
|
| 264 |
+
"step": 25
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.46,
|
| 268 |
+
"learning_rate": 0.0002,
|
| 269 |
+
"loss": 0.3186,
|
| 270 |
+
"step": 26
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.47,
|
| 274 |
+
"learning_rate": 0.0002,
|
| 275 |
+
"loss": 0.3506,
|
| 276 |
+
"step": 27
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.49,
|
| 280 |
+
"learning_rate": 0.0002,
|
| 281 |
+
"loss": 0.2576,
|
| 282 |
+
"step": 28
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.51,
|
| 286 |
+
"learning_rate": 0.0002,
|
| 287 |
+
"loss": 0.2805,
|
| 288 |
+
"step": 29
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.53,
|
| 292 |
+
"learning_rate": 0.0002,
|
| 293 |
+
"loss": 0.1683,
|
| 294 |
+
"step": 30
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.54,
|
| 298 |
+
"learning_rate": 0.0002,
|
| 299 |
+
"loss": 0.3277,
|
| 300 |
+
"step": 31
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 0.56,
|
| 304 |
+
"learning_rate": 0.0002,
|
| 305 |
+
"loss": 0.4455,
|
| 306 |
+
"step": 32
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.58,
|
| 310 |
+
"learning_rate": 0.0002,
|
| 311 |
+
"loss": 0.2145,
|
| 312 |
+
"step": 33
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.6,
|
| 316 |
+
"learning_rate": 0.0002,
|
| 317 |
+
"loss": 0.2788,
|
| 318 |
+
"step": 34
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.61,
|
| 322 |
+
"learning_rate": 0.0002,
|
| 323 |
+
"loss": 0.2119,
|
| 324 |
+
"step": 35
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.63,
|
| 328 |
+
"learning_rate": 0.0002,
|
| 329 |
+
"loss": 0.2962,
|
| 330 |
+
"step": 36
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 0.65,
|
| 334 |
+
"learning_rate": 0.0002,
|
| 335 |
+
"loss": 0.4291,
|
| 336 |
+
"step": 37
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 0.67,
|
| 340 |
+
"learning_rate": 0.0002,
|
| 341 |
+
"loss": 0.275,
|
| 342 |
+
"step": 38
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.68,
|
| 346 |
+
"learning_rate": 0.0002,
|
| 347 |
+
"loss": 0.311,
|
| 348 |
+
"step": 39
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 0.7,
|
| 352 |
+
"learning_rate": 0.0002,
|
| 353 |
+
"loss": 0.3002,
|
| 354 |
+
"step": 40
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.72,
|
| 358 |
+
"learning_rate": 0.0002,
|
| 359 |
+
"loss": 0.4225,
|
| 360 |
+
"step": 41
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.74,
|
| 364 |
+
"learning_rate": 0.0002,
|
| 365 |
+
"loss": 0.4027,
|
| 366 |
+
"step": 42
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.75,
|
| 370 |
+
"learning_rate": 0.0002,
|
| 371 |
+
"loss": 0.301,
|
| 372 |
+
"step": 43
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 0.77,
|
| 376 |
+
"learning_rate": 0.0002,
|
| 377 |
+
"loss": 0.1088,
|
| 378 |
+
"step": 44
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 0.79,
|
| 382 |
+
"learning_rate": 0.0002,
|
| 383 |
+
"loss": 0.1684,
|
| 384 |
+
"step": 45
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.81,
|
| 388 |
+
"learning_rate": 0.0002,
|
| 389 |
+
"loss": 0.2787,
|
| 390 |
+
"step": 46
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 0.82,
|
| 394 |
+
"learning_rate": 0.0002,
|
| 395 |
+
"loss": 0.104,
|
| 396 |
+
"step": 47
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 0.84,
|
| 400 |
+
"learning_rate": 0.0002,
|
| 401 |
+
"loss": 0.191,
|
| 402 |
+
"step": 48
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.86,
|
| 406 |
+
"learning_rate": 0.0002,
|
| 407 |
+
"loss": 0.197,
|
| 408 |
+
"step": 49
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.88,
|
| 412 |
+
"learning_rate": 0.0002,
|
| 413 |
+
"loss": 0.1931,
|
| 414 |
+
"step": 50
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 0.88,
|
| 418 |
+
"eval_math_accuracy": 0.93,
|
| 419 |
+
"eval_math_average_probability": 0.862511396408081,
|
| 420 |
+
"eval_math_brier_score": 0.07176309823989868,
|
| 421 |
+
"eval_math_loss": 0.2247738242149353,
|
| 422 |
+
"eval_math_probabilities": [
|
| 423 |
+
0.9903278946876526,
|
| 424 |
+
0.9961929321289062,
|
| 425 |
+
0.9953399896621704,
|
| 426 |
+
0.9755529165267944,
|
| 427 |
+
0.9934984445571899,
|
| 428 |
+
0.9834696054458618,
|
| 429 |
+
0.9999315738677979,
|
| 430 |
+
0.9999486207962036,
|
| 431 |
+
0.9999667406082153,
|
| 432 |
+
0.9027199149131775,
|
| 433 |
+
0.8275994658470154,
|
| 434 |
+
0.8796305060386658,
|
| 435 |
+
0.7624842524528503,
|
| 436 |
+
0.10068129003047943,
|
| 437 |
+
0.6096217632293701,
|
| 438 |
+
0.9473002552986145,
|
| 439 |
+
0.8130220770835876,
|
| 440 |
+
0.9198105931282043,
|
| 441 |
+
0.8807535767555237,
|
| 442 |
+
0.8420571088790894,
|
| 443 |
+
0.8424850702285767,
|
| 444 |
+
0.6525168418884277,
|
| 445 |
+
0.5038415789604187,
|
| 446 |
+
0.5285898447036743,
|
| 447 |
+
0.573636531829834,
|
| 448 |
+
0.9978896975517273,
|
| 449 |
+
0.9572906494140625,
|
| 450 |
+
0.9913204312324524,
|
| 451 |
+
0.9998912811279297,
|
| 452 |
+
0.9779029488563538,
|
| 453 |
+
0.20692378282546997,
|
| 454 |
+
0.4642832577228546,
|
| 455 |
+
0.08425822854042053,
|
| 456 |
+
0.9890357851982117,
|
| 457 |
+
0.9975990653038025,
|
| 458 |
+
0.9984292387962341,
|
| 459 |
+
0.9994285702705383,
|
| 460 |
+
0.9998854398727417,
|
| 461 |
+
0.9994675517082214,
|
| 462 |
+
0.8004253506660461,
|
| 463 |
+
0.11325926333665848,
|
| 464 |
+
0.07679656893014908,
|
| 465 |
+
0.9997519850730896,
|
| 466 |
+
0.9999046325683594,
|
| 467 |
+
0.9996449947357178,
|
| 468 |
+
0.9345480799674988,
|
| 469 |
+
0.9894824028015137,
|
| 470 |
+
0.9927423000335693,
|
| 471 |
+
0.999702513217926,
|
| 472 |
+
0.999909520149231,
|
| 473 |
+
0.9942907094955444,
|
| 474 |
+
0.9694045782089233,
|
| 475 |
+
0.9404439926147461,
|
| 476 |
+
0.9774153828620911,
|
| 477 |
+
0.7379315495491028,
|
| 478 |
+
0.7462725639343262,
|
| 479 |
+
0.9905346035957336,
|
| 480 |
+
0.9998317956924438,
|
| 481 |
+
0.9993352293968201,
|
| 482 |
+
0.9998403787612915,
|
| 483 |
+
0.9997817873954773,
|
| 484 |
+
0.9951385259628296,
|
| 485 |
+
0.9980565905570984,
|
| 486 |
+
0.5971735715866089,
|
| 487 |
+
0.6330729722976685,
|
| 488 |
+
0.9122549891471863,
|
| 489 |
+
0.5430846810340881,
|
| 490 |
+
0.7226282954216003,
|
| 491 |
+
0.9123758673667908,
|
| 492 |
+
0.9973883032798767,
|
| 493 |
+
0.9989885687828064,
|
| 494 |
+
0.9983044862747192,
|
| 495 |
+
0.9972793459892273,
|
| 496 |
+
0.9993230104446411,
|
| 497 |
+
0.9994377493858337,
|
| 498 |
+
0.9999897480010986,
|
| 499 |
+
0.9999876022338867,
|
| 500 |
+
0.9999881982803345,
|
| 501 |
+
0.9974867105484009,
|
| 502 |
+
0.9922911524772644,
|
| 503 |
+
0.9865967631340027,
|
| 504 |
+
0.9349151253700256,
|
| 505 |
+
0.9713960289955139,
|
| 506 |
+
0.9607037901878357,
|
| 507 |
+
0.9824057817459106,
|
| 508 |
+
0.9970927238464355,
|
| 509 |
+
0.9967008233070374,
|
| 510 |
+
0.9534872174263,
|
| 511 |
+
0.9496076703071594,
|
| 512 |
+
0.9766536355018616,
|
| 513 |
+
0.9999955892562866,
|
| 514 |
+
0.9999985694885254,
|
| 515 |
+
0.9999978542327881,
|
| 516 |
+
0.3180721402168274,
|
| 517 |
+
0.5712988376617432,
|
| 518 |
+
0.54411780834198,
|
| 519 |
+
0.7374041676521301,
|
| 520 |
+
0.9847201108932495,
|
| 521 |
+
0.9775702953338623,
|
| 522 |
+
0.6683135628700256
|
| 523 |
+
],
|
| 524 |
+
"eval_math_runtime": 71.3916,
|
| 525 |
+
"eval_math_samples_per_second": 1.401,
|
| 526 |
+
"eval_math_score": -0.07176309823989868,
|
| 527 |
+
"eval_math_steps_per_second": 0.056,
|
| 528 |
+
"step": 50
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.89,
|
| 532 |
+
"learning_rate": 0.0002,
|
| 533 |
+
"loss": 0.3329,
|
| 534 |
+
"step": 51
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.91,
|
| 538 |
+
"learning_rate": 0.0002,
|
| 539 |
+
"loss": 0.1205,
|
| 540 |
+
"step": 52
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.93,
|
| 544 |
+
"learning_rate": 0.0002,
|
| 545 |
+
"loss": 0.3012,
|
| 546 |
+
"step": 53
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 0.95,
|
| 550 |
+
"learning_rate": 0.0002,
|
| 551 |
+
"loss": 0.174,
|
| 552 |
+
"step": 54
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 0.96,
|
| 556 |
+
"learning_rate": 0.0002,
|
| 557 |
+
"loss": 0.1347,
|
| 558 |
+
"step": 55
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 0.98,
|
| 562 |
+
"learning_rate": 0.0002,
|
| 563 |
+
"loss": 0.145,
|
| 564 |
+
"step": 56
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.0,
|
| 568 |
+
"learning_rate": 0.0002,
|
| 569 |
+
"loss": 0.112,
|
| 570 |
+
"step": 57
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.02,
|
| 574 |
+
"learning_rate": 0.0002,
|
| 575 |
+
"loss": 0.159,
|
| 576 |
+
"step": 58
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.04,
|
| 580 |
+
"learning_rate": 0.0002,
|
| 581 |
+
"loss": 0.1406,
|
| 582 |
+
"step": 59
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 1.05,
|
| 586 |
+
"learning_rate": 0.0002,
|
| 587 |
+
"loss": 0.3111,
|
| 588 |
+
"step": 60
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 1.07,
|
| 592 |
+
"learning_rate": 0.0002,
|
| 593 |
+
"loss": 0.2551,
|
| 594 |
+
"step": 61
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 1.09,
|
| 598 |
+
"learning_rate": 0.0002,
|
| 599 |
+
"loss": 0.0922,
|
| 600 |
+
"step": 62
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 1.11,
|
| 604 |
+
"learning_rate": 0.0002,
|
| 605 |
+
"loss": 0.233,
|
| 606 |
+
"step": 63
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.12,
|
| 610 |
+
"learning_rate": 0.0002,
|
| 611 |
+
"loss": 0.0813,
|
| 612 |
+
"step": 64
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.14,
|
| 616 |
+
"learning_rate": 0.0002,
|
| 617 |
+
"loss": 0.2089,
|
| 618 |
+
"step": 65
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.16,
|
| 622 |
+
"learning_rate": 0.0002,
|
| 623 |
+
"loss": 0.0368,
|
| 624 |
+
"step": 66
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 1.18,
|
| 628 |
+
"learning_rate": 0.0002,
|
| 629 |
+
"loss": 0.2031,
|
| 630 |
+
"step": 67
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 1.19,
|
| 634 |
+
"learning_rate": 0.0002,
|
| 635 |
+
"loss": 0.145,
|
| 636 |
+
"step": 68
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 1.21,
|
| 640 |
+
"learning_rate": 0.0002,
|
| 641 |
+
"loss": 0.1448,
|
| 642 |
+
"step": 69
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 1.23,
|
| 646 |
+
"learning_rate": 0.0002,
|
| 647 |
+
"loss": 0.4538,
|
| 648 |
+
"step": 70
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.25,
|
| 652 |
+
"learning_rate": 0.0002,
|
| 653 |
+
"loss": 0.1073,
|
| 654 |
+
"step": 71
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.26,
|
| 658 |
+
"learning_rate": 0.0002,
|
| 659 |
+
"loss": 0.1287,
|
| 660 |
+
"step": 72
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.28,
|
| 664 |
+
"learning_rate": 0.0002,
|
| 665 |
+
"loss": 0.0717,
|
| 666 |
+
"step": 73
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 1.3,
|
| 670 |
+
"learning_rate": 0.0002,
|
| 671 |
+
"loss": 0.2391,
|
| 672 |
+
"step": 74
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 1.32,
|
| 676 |
+
"learning_rate": 0.0002,
|
| 677 |
+
"loss": 0.063,
|
| 678 |
+
"step": 75
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 1.32,
|
| 682 |
+
"eval_math_accuracy": 0.93,
|
| 683 |
+
"eval_math_average_probability": 0.886993408203125,
|
| 684 |
+
"eval_math_brier_score": 0.06159758195281029,
|
| 685 |
+
"eval_math_loss": 0.21683579683303833,
|
| 686 |
+
"eval_math_probabilities": [
|
| 687 |
+
0.9708817005157471,
|
| 688 |
+
0.9872758984565735,
|
| 689 |
+
0.9881438612937927,
|
| 690 |
+
0.999626874923706,
|
| 691 |
+
0.9998425245285034,
|
| 692 |
+
0.9997356534004211,
|
| 693 |
+
0.9999996423721313,
|
| 694 |
+
0.9999995231628418,
|
| 695 |
+
0.9999996423721313,
|
| 696 |
+
0.9958369731903076,
|
| 697 |
+
0.979812741279602,
|
| 698 |
+
0.9877761602401733,
|
| 699 |
+
0.9479932188987732,
|
| 700 |
+
0.609644889831543,
|
| 701 |
+
0.9154527187347412,
|
| 702 |
+
0.8812452554702759,
|
| 703 |
+
0.6881961226463318,
|
| 704 |
+
0.8928657174110413,
|
| 705 |
+
0.9649090766906738,
|
| 706 |
+
0.9266186356544495,
|
| 707 |
+
0.9296641945838928,
|
| 708 |
+
0.6537356972694397,
|
| 709 |
+
0.5258322954177856,
|
| 710 |
+
0.5929333567619324,
|
| 711 |
+
0.88760906457901,
|
| 712 |
+
0.9985673427581787,
|
| 713 |
+
0.9800016283988953,
|
| 714 |
+
0.9999486207962036,
|
| 715 |
+
0.999967098236084,
|
| 716 |
+
0.9999837875366211,
|
| 717 |
+
0.15832553803920746,
|
| 718 |
+
0.29652073979377747,
|
| 719 |
+
0.0025304730515927076,
|
| 720 |
+
0.8137555718421936,
|
| 721 |
+
0.9484632015228271,
|
| 722 |
+
0.9539691805839539,
|
| 723 |
+
0.9994146823883057,
|
| 724 |
+
0.9998959302902222,
|
| 725 |
+
0.9994522929191589,
|
| 726 |
+
0.3250795304775238,
|
| 727 |
+
0.12795241177082062,
|
| 728 |
+
0.1360873132944107,
|
| 729 |
+
0.9998359680175781,
|
| 730 |
+
0.9999209642410278,
|
| 731 |
+
0.999779999256134,
|
| 732 |
+
0.9624126553535461,
|
| 733 |
+
0.9971975088119507,
|
| 734 |
+
0.9986832737922668,
|
| 735 |
+
0.9997674822807312,
|
| 736 |
+
0.9999611377716064,
|
| 737 |
+
0.9986693859100342,
|
| 738 |
+
0.9999688863754272,
|
| 739 |
+
0.999846339225769,
|
| 740 |
+
0.9999701976776123,
|
| 741 |
+
0.9358548521995544,
|
| 742 |
+
0.8916678428649902,
|
| 743 |
+
0.992230236530304,
|
| 744 |
+
0.9996994733810425,
|
| 745 |
+
0.9995322227478027,
|
| 746 |
+
0.9997522234916687,
|
| 747 |
+
0.9999985694885254,
|
| 748 |
+
0.9999802112579346,
|
| 749 |
+
0.9999529123306274,
|
| 750 |
+
0.783003032207489,
|
| 751 |
+
0.8378167152404785,
|
| 752 |
+
0.9600664973258972,
|
| 753 |
+
0.7785875201225281,
|
| 754 |
+
0.6786913871765137,
|
| 755 |
+
0.9715420603752136,
|
| 756 |
+
0.9995951056480408,
|
| 757 |
+
0.9998225569725037,
|
| 758 |
+
0.9997715353965759,
|
| 759 |
+
0.9968163371086121,
|
| 760 |
+
0.9992282390594482,
|
| 761 |
+
0.9992213249206543,
|
| 762 |
+
0.9999867677688599,
|
| 763 |
+
0.9999890327453613,
|
| 764 |
+
0.9999853372573853,
|
| 765 |
+
0.9856051802635193,
|
| 766 |
+
0.9555090665817261,
|
| 767 |
+
0.9302185773849487,
|
| 768 |
+
0.9948233366012573,
|
| 769 |
+
0.9981561303138733,
|
| 770 |
+
0.9972609281539917,
|
| 771 |
+
0.9951260685920715,
|
| 772 |
+
0.9982566237449646,
|
| 773 |
+
0.9970597624778748,
|
| 774 |
+
0.9945375323295593,
|
| 775 |
+
0.9840699434280396,
|
| 776 |
+
0.9972852468490601,
|
| 777 |
+
0.9999970197677612,
|
| 778 |
+
0.9999997615814209,
|
| 779 |
+
0.9999991655349731,
|
| 780 |
+
0.39414581656455994,
|
| 781 |
+
0.5715559720993042,
|
| 782 |
+
0.5152676105499268,
|
| 783 |
+
0.7351731657981873,
|
| 784 |
+
0.9919071197509766,
|
| 785 |
+
0.9899667501449585,
|
| 786 |
+
0.8290728330612183
|
| 787 |
+
],
|
| 788 |
+
"eval_math_runtime": 71.3623,
|
| 789 |
+
"eval_math_samples_per_second": 1.401,
|
| 790 |
+
"eval_math_score": -0.06159758195281029,
|
| 791 |
+
"eval_math_steps_per_second": 0.056,
|
| 792 |
+
"step": 75
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 1.33,
|
| 796 |
+
"learning_rate": 0.0002,
|
| 797 |
+
"loss": 0.2009,
|
| 798 |
+
"step": 76
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 1.35,
|
| 802 |
+
"learning_rate": 0.0002,
|
| 803 |
+
"loss": 0.0804,
|
| 804 |
+
"step": 77
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 1.37,
|
| 808 |
+
"learning_rate": 0.0002,
|
| 809 |
+
"loss": 0.1053,
|
| 810 |
+
"step": 78
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 1.39,
|
| 814 |
+
"learning_rate": 0.0002,
|
| 815 |
+
"loss": 0.0679,
|
| 816 |
+
"step": 79
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.4,
|
| 820 |
+
"learning_rate": 0.0002,
|
| 821 |
+
"loss": 0.2037,
|
| 822 |
+
"step": 80
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.42,
|
| 826 |
+
"learning_rate": 0.0002,
|
| 827 |
+
"loss": 0.1627,
|
| 828 |
+
"step": 81
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.44,
|
| 832 |
+
"learning_rate": 0.0002,
|
| 833 |
+
"loss": 0.1514,
|
| 834 |
+
"step": 82
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 1.46,
|
| 838 |
+
"learning_rate": 0.0002,
|
| 839 |
+
"loss": 0.0252,
|
| 840 |
+
"step": 83
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 1.47,
|
| 844 |
+
"learning_rate": 0.0002,
|
| 845 |
+
"loss": 0.1374,
|
| 846 |
+
"step": 84
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 1.49,
|
| 850 |
+
"learning_rate": 0.0002,
|
| 851 |
+
"loss": 0.2176,
|
| 852 |
+
"step": 85
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 1.51,
|
| 856 |
+
"learning_rate": 0.0002,
|
| 857 |
+
"loss": 0.1134,
|
| 858 |
+
"step": 86
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 1.53,
|
| 862 |
+
"learning_rate": 0.0002,
|
| 863 |
+
"loss": 0.0623,
|
| 864 |
+
"step": 87
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.54,
|
| 868 |
+
"learning_rate": 0.0002,
|
| 869 |
+
"loss": 0.0578,
|
| 870 |
+
"step": 88
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.56,
|
| 874 |
+
"learning_rate": 0.0002,
|
| 875 |
+
"loss": 0.2308,
|
| 876 |
+
"step": 89
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 1.58,
|
| 880 |
+
"learning_rate": 0.0002,
|
| 881 |
+
"loss": 0.1051,
|
| 882 |
+
"step": 90
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 1.6,
|
| 886 |
+
"learning_rate": 0.0002,
|
| 887 |
+
"loss": 0.0932,
|
| 888 |
+
"step": 91
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 1.61,
|
| 892 |
+
"learning_rate": 0.0002,
|
| 893 |
+
"loss": 0.1234,
|
| 894 |
+
"step": 92
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 1.63,
|
| 898 |
+
"learning_rate": 0.0002,
|
| 899 |
+
"loss": 0.1145,
|
| 900 |
+
"step": 93
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.65,
|
| 904 |
+
"learning_rate": 0.0002,
|
| 905 |
+
"loss": 0.0803,
|
| 906 |
+
"step": 94
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.67,
|
| 910 |
+
"learning_rate": 0.0002,
|
| 911 |
+
"loss": 0.0799,
|
| 912 |
+
"step": 95
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.68,
|
| 916 |
+
"learning_rate": 0.0002,
|
| 917 |
+
"loss": 0.1906,
|
| 918 |
+
"step": 96
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 1.7,
|
| 922 |
+
"learning_rate": 0.0002,
|
| 923 |
+
"loss": 0.1056,
|
| 924 |
+
"step": 97
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 1.72,
|
| 928 |
+
"learning_rate": 0.0002,
|
| 929 |
+
"loss": 0.0837,
|
| 930 |
+
"step": 98
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 1.74,
|
| 934 |
+
"learning_rate": 0.0002,
|
| 935 |
+
"loss": 0.2687,
|
| 936 |
+
"step": 99
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 1.75,
|
| 940 |
+
"learning_rate": 0.0002,
|
| 941 |
+
"loss": 0.0419,
|
| 942 |
+
"step": 100
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.75,
|
| 946 |
+
"eval_math_accuracy": 0.92,
|
| 947 |
+
"eval_math_average_probability": 0.897860586643219,
|
| 948 |
+
"eval_math_brier_score": 0.06293144077062607,
|
| 949 |
+
"eval_math_loss": 0.2073674350976944,
|
| 950 |
+
"eval_math_probabilities": [
|
| 951 |
+
0.9997542500495911,
|
| 952 |
+
0.9999017715454102,
|
| 953 |
+
0.9999037981033325,
|
| 954 |
+
0.9999974966049194,
|
| 955 |
+
0.9999986886978149,
|
| 956 |
+
0.9999978542327881,
|
| 957 |
+
1.0,
|
| 958 |
+
0.9999998807907104,
|
| 959 |
+
1.0,
|
| 960 |
+
0.9984958171844482,
|
| 961 |
+
0.9447609186172485,
|
| 962 |
+
0.9886316657066345,
|
| 963 |
+
0.9511963129043579,
|
| 964 |
+
0.3284871280193329,
|
| 965 |
+
0.9281124472618103,
|
| 966 |
+
0.9668608903884888,
|
| 967 |
+
0.7682732343673706,
|
| 968 |
+
0.9800459146499634,
|
| 969 |
+
0.8894421458244324,
|
| 970 |
+
0.7420941591262817,
|
| 971 |
+
0.751176118850708,
|
| 972 |
+
0.8641813397407532,
|
| 973 |
+
0.5344230532646179,
|
| 974 |
+
0.5974614024162292,
|
| 975 |
+
0.9974793791770935,
|
| 976 |
+
0.9999979734420776,
|
| 977 |
+
0.9988064765930176,
|
| 978 |
+
0.99997878074646,
|
| 979 |
+
0.9999936819076538,
|
| 980 |
+
0.9996273517608643,
|
| 981 |
+
0.15381988883018494,
|
| 982 |
+
0.18481139838695526,
|
| 983 |
+
0.009875369258224964,
|
| 984 |
+
0.9291513562202454,
|
| 985 |
+
0.9974228143692017,
|
| 986 |
+
0.9989088773727417,
|
| 987 |
+
0.9999997615814209,
|
| 988 |
+
1.0,
|
| 989 |
+
0.9999991655349731,
|
| 990 |
+
0.8100789189338684,
|
| 991 |
+
0.07439813762903214,
|
| 992 |
+
0.05608365684747696,
|
| 993 |
+
0.9999886751174927,
|
| 994 |
+
0.9999966621398926,
|
| 995 |
+
0.9999871253967285,
|
| 996 |
+
0.9790005087852478,
|
| 997 |
+
0.995776355266571,
|
| 998 |
+
0.997306227684021,
|
| 999 |
+
0.9999855756759644,
|
| 1000 |
+
0.9999990463256836,
|
| 1001 |
+
0.9990392923355103,
|
| 1002 |
+
0.9999954700469971,
|
| 1003 |
+
0.9998663663864136,
|
| 1004 |
+
0.9999920129776001,
|
| 1005 |
+
0.9822255373001099,
|
| 1006 |
+
0.9193406105041504,
|
| 1007 |
+
0.999954342842102,
|
| 1008 |
+
0.9999996423721313,
|
| 1009 |
+
0.9999985694885254,
|
| 1010 |
+
0.9999997615814209,
|
| 1011 |
+
0.9999997615814209,
|
| 1012 |
+
0.9999094009399414,
|
| 1013 |
+
0.999971866607666,
|
| 1014 |
+
0.9311190843582153,
|
| 1015 |
+
0.9527549147605896,
|
| 1016 |
+
0.9944831728935242,
|
| 1017 |
+
0.932115375995636,
|
| 1018 |
+
0.689570426940918,
|
| 1019 |
+
0.9954724907875061,
|
| 1020 |
+
0.9999904632568359,
|
| 1021 |
+
0.999997615814209,
|
| 1022 |
+
0.9999985694885254,
|
| 1023 |
+
0.9999154806137085,
|
| 1024 |
+
0.9999428987503052,
|
| 1025 |
+
0.9999589920043945,
|
| 1026 |
+
0.9999998807907104,
|
| 1027 |
+
0.9999998807907104,
|
| 1028 |
+
0.9999997615814209,
|
| 1029 |
+
0.9999972581863403,
|
| 1030 |
+
0.9999542236328125,
|
| 1031 |
+
0.9998880624771118,
|
| 1032 |
+
0.9994339346885681,
|
| 1033 |
+
0.9999072551727295,
|
| 1034 |
+
0.9999172687530518,
|
| 1035 |
+
0.9996486902236938,
|
| 1036 |
+
0.9999575614929199,
|
| 1037 |
+
0.9999730587005615,
|
| 1038 |
+
0.9997407793998718,
|
| 1039 |
+
0.9387125968933105,
|
| 1040 |
+
0.998235821723938,
|
| 1041 |
+
0.9999998807907104,
|
| 1042 |
+
1.0,
|
| 1043 |
+
1.0,
|
| 1044 |
+
0.42074549198150635,
|
| 1045 |
+
0.4002394378185272,
|
| 1046 |
+
0.6501257419586182,
|
| 1047 |
+
0.6124982237815857,
|
| 1048 |
+
0.9996664524078369,
|
| 1049 |
+
0.9985406398773193,
|
| 1050 |
+
0.957599401473999
|
| 1051 |
+
],
|
| 1052 |
+
"eval_math_runtime": 71.3769,
|
| 1053 |
+
"eval_math_samples_per_second": 1.401,
|
| 1054 |
+
"eval_math_score": -0.06293144077062607,
|
| 1055 |
+
"eval_math_steps_per_second": 0.056,
|
| 1056 |
+
"step": 100
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 1.77,
|
| 1060 |
+
"learning_rate": 0.0002,
|
| 1061 |
+
"loss": 0.0741,
|
| 1062 |
+
"step": 101
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 1.79,
|
| 1066 |
+
"learning_rate": 0.0002,
|
| 1067 |
+
"loss": 0.1895,
|
| 1068 |
+
"step": 102
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 1.81,
|
| 1072 |
+
"learning_rate": 0.0002,
|
| 1073 |
+
"loss": 0.1329,
|
| 1074 |
+
"step": 103
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.82,
|
| 1078 |
+
"learning_rate": 0.0002,
|
| 1079 |
+
"loss": 0.0341,
|
| 1080 |
+
"step": 104
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.84,
|
| 1084 |
+
"learning_rate": 0.0002,
|
| 1085 |
+
"loss": 0.0519,
|
| 1086 |
+
"step": 105
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 1.86,
|
| 1090 |
+
"learning_rate": 0.0002,
|
| 1091 |
+
"loss": 0.086,
|
| 1092 |
+
"step": 106
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 1.88,
|
| 1096 |
+
"learning_rate": 0.0002,
|
| 1097 |
+
"loss": 0.1358,
|
| 1098 |
+
"step": 107
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 1.89,
|
| 1102 |
+
"learning_rate": 0.0002,
|
| 1103 |
+
"loss": 0.0155,
|
| 1104 |
+
"step": 108
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 1.91,
|
| 1108 |
+
"learning_rate": 0.0002,
|
| 1109 |
+
"loss": 0.1158,
|
| 1110 |
+
"step": 109
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 1.93,
|
| 1114 |
+
"learning_rate": 0.0002,
|
| 1115 |
+
"loss": 0.1017,
|
| 1116 |
+
"step": 110
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.95,
|
| 1120 |
+
"learning_rate": 0.0002,
|
| 1121 |
+
"loss": 0.1276,
|
| 1122 |
+
"step": 111
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.96,
|
| 1126 |
+
"learning_rate": 0.0002,
|
| 1127 |
+
"loss": 0.0458,
|
| 1128 |
+
"step": 112
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 1.98,
|
| 1132 |
+
"learning_rate": 0.0002,
|
| 1133 |
+
"loss": 0.1282,
|
| 1134 |
+
"step": 113
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 2.0,
|
| 1138 |
+
"learning_rate": 0.0002,
|
| 1139 |
+
"loss": 0.047,
|
| 1140 |
+
"step": 114
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 2.02,
|
| 1144 |
+
"learning_rate": 0.0002,
|
| 1145 |
+
"loss": 0.0111,
|
| 1146 |
+
"step": 115
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 2.04,
|
| 1150 |
+
"learning_rate": 0.0002,
|
| 1151 |
+
"loss": 0.0176,
|
| 1152 |
+
"step": 116
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 2.05,
|
| 1156 |
+
"learning_rate": 0.0002,
|
| 1157 |
+
"loss": 0.0145,
|
| 1158 |
+
"step": 117
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.07,
|
| 1162 |
+
"learning_rate": 0.0002,
|
| 1163 |
+
"loss": 0.0281,
|
| 1164 |
+
"step": 118
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 2.09,
|
| 1168 |
+
"learning_rate": 0.0002,
|
| 1169 |
+
"loss": 0.0303,
|
| 1170 |
+
"step": 119
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 2.11,
|
| 1174 |
+
"learning_rate": 0.0002,
|
| 1175 |
+
"loss": 0.0136,
|
| 1176 |
+
"step": 120
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 2.12,
|
| 1180 |
+
"learning_rate": 0.0002,
|
| 1181 |
+
"loss": 0.0062,
|
| 1182 |
+
"step": 121
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.14,
|
| 1186 |
+
"learning_rate": 0.0002,
|
| 1187 |
+
"loss": 0.0234,
|
| 1188 |
+
"step": 122
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 2.16,
|
| 1192 |
+
"learning_rate": 0.0002,
|
| 1193 |
+
"loss": 0.0269,
|
| 1194 |
+
"step": 123
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 2.18,
|
| 1198 |
+
"learning_rate": 0.0002,
|
| 1199 |
+
"loss": 0.0453,
|
| 1200 |
+
"step": 124
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 2.19,
|
| 1204 |
+
"learning_rate": 0.0002,
|
| 1205 |
+
"loss": 0.0736,
|
| 1206 |
+
"step": 125
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 2.19,
|
| 1210 |
+
"eval_math_accuracy": 0.94,
|
| 1211 |
+
"eval_math_average_probability": 0.9156234264373779,
|
| 1212 |
+
"eval_math_brier_score": 0.05634632706642151,
|
| 1213 |
+
"eval_math_loss": 0.26747652888298035,
|
| 1214 |
+
"eval_math_probabilities": [
|
| 1215 |
+
0.9998124241828918,
|
| 1216 |
+
0.9998958110809326,
|
| 1217 |
+
0.9998985528945923,
|
| 1218 |
+
1.0,
|
| 1219 |
+
1.0,
|
| 1220 |
+
1.0,
|
| 1221 |
+
1.0,
|
| 1222 |
+
1.0,
|
| 1223 |
+
1.0,
|
| 1224 |
+
0.9999892711639404,
|
| 1225 |
+
0.9945842623710632,
|
| 1226 |
+
0.9986479878425598,
|
| 1227 |
+
0.8236495852470398,
|
| 1228 |
+
0.11610127985477448,
|
| 1229 |
+
0.7955105900764465,
|
| 1230 |
+
0.7467267513275146,
|
| 1231 |
+
0.6469255089759827,
|
| 1232 |
+
0.965029776096344,
|
| 1233 |
+
0.9882959723472595,
|
| 1234 |
+
0.9713000059127808,
|
| 1235 |
+
0.9623332619667053,
|
| 1236 |
+
0.9890497922897339,
|
| 1237 |
+
0.6750863790512085,
|
| 1238 |
+
0.8250890970230103,
|
| 1239 |
+
0.983614981174469,
|
| 1240 |
+
0.9999977350234985,
|
| 1241 |
+
0.9976145029067993,
|
| 1242 |
+
0.9999997615814209,
|
| 1243 |
+
0.9999997615814209,
|
| 1244 |
+
0.9999943971633911,
|
| 1245 |
+
0.08161313086748123,
|
| 1246 |
+
0.07091999799013138,
|
| 1247 |
+
1.767081266734749e-05,
|
| 1248 |
+
0.991096019744873,
|
| 1249 |
+
0.9984982013702393,
|
| 1250 |
+
0.9995262622833252,
|
| 1251 |
+
1.0,
|
| 1252 |
+
1.0,
|
| 1253 |
+
0.9999998807907104,
|
| 1254 |
+
0.6652071475982666,
|
| 1255 |
+
0.3877899646759033,
|
| 1256 |
+
0.020432589575648308,
|
| 1257 |
+
0.9999984502792358,
|
| 1258 |
+
0.9999997615814209,
|
| 1259 |
+
0.9999995231628418,
|
| 1260 |
+
0.9990777969360352,
|
| 1261 |
+
0.9999066591262817,
|
| 1262 |
+
0.9999903440475464,
|
| 1263 |
+
0.9999997615814209,
|
| 1264 |
+
1.0,
|
| 1265 |
+
0.9999938011169434,
|
| 1266 |
+
0.9999997615814209,
|
| 1267 |
+
0.9999713897705078,
|
| 1268 |
+
0.9999994039535522,
|
| 1269 |
+
0.998889148235321,
|
| 1270 |
+
0.9776034355163574,
|
| 1271 |
+
0.999996542930603,
|
| 1272 |
+
0.9999998807907104,
|
| 1273 |
+
0.9999996423721313,
|
| 1274 |
+
0.9999997615814209,
|
| 1275 |
+
1.0,
|
| 1276 |
+
0.9999998807907104,
|
| 1277 |
+
0.9999998807907104,
|
| 1278 |
+
0.926174521446228,
|
| 1279 |
+
0.9989579916000366,
|
| 1280 |
+
0.9986373782157898,
|
| 1281 |
+
0.9930163621902466,
|
| 1282 |
+
0.7508591413497925,
|
| 1283 |
+
0.9998103976249695,
|
| 1284 |
+
0.9999997615814209,
|
| 1285 |
+
0.9999998807907104,
|
| 1286 |
+
1.0,
|
| 1287 |
+
0.999890923500061,
|
| 1288 |
+
0.9999988079071045,
|
| 1289 |
+
0.9999992847442627,
|
| 1290 |
+
1.0,
|
| 1291 |
+
1.0,
|
| 1292 |
+
1.0,
|
| 1293 |
+
0.9999998807907104,
|
| 1294 |
+
0.9999982118606567,
|
| 1295 |
+
0.999996542930603,
|
| 1296 |
+
0.999996542930603,
|
| 1297 |
+
1.0,
|
| 1298 |
+
0.9999998807907104,
|
| 1299 |
+
0.9999873638153076,
|
| 1300 |
+
0.9999982118606567,
|
| 1301 |
+
0.9999979734420776,
|
| 1302 |
+
0.9999932050704956,
|
| 1303 |
+
0.9997509121894836,
|
| 1304 |
+
0.9999990463256836,
|
| 1305 |
+
1.0,
|
| 1306 |
+
1.0,
|
| 1307 |
+
1.0,
|
| 1308 |
+
0.8197342753410339,
|
| 1309 |
+
0.5860379934310913,
|
| 1310 |
+
0.8383031487464905,
|
| 1311 |
+
0.9970198273658752,
|
| 1312 |
+
0.9999985694885254,
|
| 1313 |
+
0.9999600648880005,
|
| 1314 |
+
0.9845534563064575
|
| 1315 |
+
],
|
| 1316 |
+
"eval_math_runtime": 71.3514,
|
| 1317 |
+
"eval_math_samples_per_second": 1.402,
|
| 1318 |
+
"eval_math_score": -0.05634632706642151,
|
| 1319 |
+
"eval_math_steps_per_second": 0.056,
|
| 1320 |
+
"step": 125
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 2.21,
|
| 1324 |
+
"learning_rate": 0.0002,
|
| 1325 |
+
"loss": 0.1389,
|
| 1326 |
+
"step": 126
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 2.23,
|
| 1330 |
+
"learning_rate": 0.0002,
|
| 1331 |
+
"loss": 0.0058,
|
| 1332 |
+
"step": 127
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 2.25,
|
| 1336 |
+
"learning_rate": 0.0002,
|
| 1337 |
+
"loss": 0.0119,
|
| 1338 |
+
"step": 128
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 2.26,
|
| 1342 |
+
"learning_rate": 0.0002,
|
| 1343 |
+
"loss": 0.0911,
|
| 1344 |
+
"step": 129
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 2.28,
|
| 1348 |
+
"learning_rate": 0.0002,
|
| 1349 |
+
"loss": 0.0169,
|
| 1350 |
+
"step": 130
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 2.3,
|
| 1354 |
+
"learning_rate": 0.0002,
|
| 1355 |
+
"loss": 0.0027,
|
| 1356 |
+
"step": 131
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 2.32,
|
| 1360 |
+
"learning_rate": 0.0002,
|
| 1361 |
+
"loss": 0.0058,
|
| 1362 |
+
"step": 132
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 2.33,
|
| 1366 |
+
"learning_rate": 0.0002,
|
| 1367 |
+
"loss": 0.0437,
|
| 1368 |
+
"step": 133
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 2.35,
|
| 1372 |
+
"learning_rate": 0.0002,
|
| 1373 |
+
"loss": 0.0005,
|
| 1374 |
+
"step": 134
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 2.37,
|
| 1378 |
+
"learning_rate": 0.0002,
|
| 1379 |
+
"loss": 0.004,
|
| 1380 |
+
"step": 135
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 2.39,
|
| 1384 |
+
"learning_rate": 0.0002,
|
| 1385 |
+
"loss": 0.0317,
|
| 1386 |
+
"step": 136
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 2.4,
|
| 1390 |
+
"learning_rate": 0.0002,
|
| 1391 |
+
"loss": 0.0114,
|
| 1392 |
+
"step": 137
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 2.42,
|
| 1396 |
+
"learning_rate": 0.0002,
|
| 1397 |
+
"loss": 0.0021,
|
| 1398 |
+
"step": 138
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 2.44,
|
| 1402 |
+
"learning_rate": 0.0002,
|
| 1403 |
+
"loss": 0.0081,
|
| 1404 |
+
"step": 139
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 2.46,
|
| 1408 |
+
"learning_rate": 0.0002,
|
| 1409 |
+
"loss": 0.0302,
|
| 1410 |
+
"step": 140
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 2.47,
|
| 1414 |
+
"learning_rate": 0.0002,
|
| 1415 |
+
"loss": 0.0612,
|
| 1416 |
+
"step": 141
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 2.49,
|
| 1420 |
+
"learning_rate": 0.0002,
|
| 1421 |
+
"loss": 0.0256,
|
| 1422 |
+
"step": 142
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 2.51,
|
| 1426 |
+
"learning_rate": 0.0002,
|
| 1427 |
+
"loss": 0.0016,
|
| 1428 |
+
"step": 143
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 2.53,
|
| 1432 |
+
"learning_rate": 0.0002,
|
| 1433 |
+
"loss": 0.0009,
|
| 1434 |
+
"step": 144
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 2.54,
|
| 1438 |
+
"learning_rate": 0.0002,
|
| 1439 |
+
"loss": 0.0276,
|
| 1440 |
+
"step": 145
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 2.56,
|
| 1444 |
+
"learning_rate": 0.0002,
|
| 1445 |
+
"loss": 0.0229,
|
| 1446 |
+
"step": 146
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 2.58,
|
| 1450 |
+
"learning_rate": 0.0002,
|
| 1451 |
+
"loss": 0.0093,
|
| 1452 |
+
"step": 147
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 2.6,
|
| 1456 |
+
"learning_rate": 0.0002,
|
| 1457 |
+
"loss": 0.0041,
|
| 1458 |
+
"step": 148
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 2.61,
|
| 1462 |
+
"learning_rate": 0.0002,
|
| 1463 |
+
"loss": 0.0034,
|
| 1464 |
+
"step": 149
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 2.63,
|
| 1468 |
+
"learning_rate": 0.0002,
|
| 1469 |
+
"loss": 0.0019,
|
| 1470 |
+
"step": 150
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 2.63,
|
| 1474 |
+
"eval_math_accuracy": 0.93,
|
| 1475 |
+
"eval_math_average_probability": 0.9121327996253967,
|
| 1476 |
+
"eval_math_brier_score": 0.06956099718809128,
|
| 1477 |
+
"eval_math_loss": 0.4282163381576538,
|
| 1478 |
+
"eval_math_probabilities": [
|
| 1479 |
+
0.9996767044067383,
|
| 1480 |
+
0.9998788833618164,
|
| 1481 |
+
0.9999418258666992,
|
| 1482 |
+
1.0,
|
| 1483 |
+
1.0,
|
| 1484 |
+
1.0,
|
| 1485 |
+
1.0,
|
| 1486 |
+
1.0,
|
| 1487 |
+
1.0,
|
| 1488 |
+
0.9999996423721313,
|
| 1489 |
+
0.9997425675392151,
|
| 1490 |
+
0.9999622106552124,
|
| 1491 |
+
0.9822726249694824,
|
| 1492 |
+
0.10904553532600403,
|
| 1493 |
+
0.8348792791366577,
|
| 1494 |
+
0.9605673551559448,
|
| 1495 |
+
0.9648340940475464,
|
| 1496 |
+
0.9984903335571289,
|
| 1497 |
+
0.9981614947319031,
|
| 1498 |
+
0.9723353385925293,
|
| 1499 |
+
0.9089441895484924,
|
| 1500 |
+
0.9888822436332703,
|
| 1501 |
+
0.6690593361854553,
|
| 1502 |
+
0.8258305788040161,
|
| 1503 |
+
0.9999411106109619,
|
| 1504 |
+
1.0,
|
| 1505 |
+
0.9999996423721313,
|
| 1506 |
+
0.9920840263366699,
|
| 1507 |
+
1.0,
|
| 1508 |
+
0.9999914169311523,
|
| 1509 |
+
0.00976644828915596,
|
| 1510 |
+
0.1135169118642807,
|
| 1511 |
+
4.027563349495722e-08,
|
| 1512 |
+
0.8849645256996155,
|
| 1513 |
+
0.9890426993370056,
|
| 1514 |
+
0.9991269707679749,
|
| 1515 |
+
1.0,
|
| 1516 |
+
1.0,
|
| 1517 |
+
1.0,
|
| 1518 |
+
0.8971572518348694,
|
| 1519 |
+
0.04284805804491043,
|
| 1520 |
+
0.00048620859161019325,
|
| 1521 |
+
1.0,
|
| 1522 |
+
1.0,
|
| 1523 |
+
1.0,
|
| 1524 |
+
0.9994839429855347,
|
| 1525 |
+
0.9999998807907104,
|
| 1526 |
+
0.9999997615814209,
|
| 1527 |
+
1.0,
|
| 1528 |
+
1.0,
|
| 1529 |
+
0.9999953508377075,
|
| 1530 |
+
1.0,
|
| 1531 |
+
1.0,
|
| 1532 |
+
1.0,
|
| 1533 |
+
0.9995635151863098,
|
| 1534 |
+
0.9855570197105408,
|
| 1535 |
+
0.9999997615814209,
|
| 1536 |
+
1.0,
|
| 1537 |
+
1.0,
|
| 1538 |
+
1.0,
|
| 1539 |
+
1.0,
|
| 1540 |
+
1.0,
|
| 1541 |
+
1.0,
|
| 1542 |
+
0.9868708848953247,
|
| 1543 |
+
0.9964863061904907,
|
| 1544 |
+
0.9999667406082153,
|
| 1545 |
+
0.9630876183509827,
|
| 1546 |
+
0.9137640595436096,
|
| 1547 |
+
0.9995352029800415,
|
| 1548 |
+
1.0,
|
| 1549 |
+
1.0,
|
| 1550 |
+
1.0,
|
| 1551 |
+
0.9999998807907104,
|
| 1552 |
+
0.9999997615814209,
|
| 1553 |
+
1.0,
|
| 1554 |
+
1.0,
|
| 1555 |
+
1.0,
|
| 1556 |
+
1.0,
|
| 1557 |
+
1.0,
|
| 1558 |
+
0.9999997615814209,
|
| 1559 |
+
0.999997615814209,
|
| 1560 |
+
1.0,
|
| 1561 |
+
1.0,
|
| 1562 |
+
1.0,
|
| 1563 |
+
1.0,
|
| 1564 |
+
1.0,
|
| 1565 |
+
1.0,
|
| 1566 |
+
0.9999831914901733,
|
| 1567 |
+
0.9999042749404907,
|
| 1568 |
+
0.9999977350234985,
|
| 1569 |
+
1.0,
|
| 1570 |
+
1.0,
|
| 1571 |
+
1.0,
|
| 1572 |
+
0.030157605186104774,
|
| 1573 |
+
0.5301457047462463,
|
| 1574 |
+
0.6733043789863586,
|
| 1575 |
+
0.9995803236961365,
|
| 1576 |
+
0.9999997615814209,
|
| 1577 |
+
0.9999979734420776,
|
| 1578 |
+
0.9944654107093811
|
| 1579 |
+
],
|
| 1580 |
+
"eval_math_runtime": 71.4375,
|
| 1581 |
+
"eval_math_samples_per_second": 1.4,
|
| 1582 |
+
"eval_math_score": -0.06956099718809128,
|
| 1583 |
+
"eval_math_steps_per_second": 0.056,
|
| 1584 |
+
"step": 150
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"loss": 0.0076,
|
| 1588 |
+
"learning_rate": 0.0002,
|
| 1589 |
+
"epoch": 2.65,
|
| 1590 |
+
"step": 151
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"train_runtime": 79.1835,
|
| 1594 |
+
"train_samples_per_second": 60.619,
|
| 1595 |
+
"train_steps_per_second": 1.894,
|
| 1596 |
+
"total_flos": 0.0,
|
| 1597 |
+
"train_loss": 5.016510146639205e-05,
|
| 1598 |
+
"epoch": 2.65,
|
| 1599 |
+
"step": 151
|
| 1600 |
+
}
|
| 1601 |
+
]
|