Upload 12 files
Browse files- README.md +54 -3
- adapter_config.json +31 -0
- adapter_model.safetensors +3 -0
- all_results.json +14 -0
- config.json +28 -0
- eval_results.json +8 -0
- special_tokens_map.json +16 -0
- tokenizer.json +0 -0
- tokenizer_config.json +239 -0
- train_results.json +9 -0
- trainer_state.json +1445 -0
- training_args.bin +3 -0
README.md
CHANGED
|
@@ -1,3 +1,54 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
library_name: peft
|
| 4 |
+
tags:
|
| 5 |
+
- alignment-handbook
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
base_model: allenai/OLMo-1B-hf
|
| 8 |
+
datasets:
|
| 9 |
+
- HuggingFaceH4/ultrachat_200k
|
| 10 |
+
model-index:
|
| 11 |
+
- name: zephyr-1b-olmo-sft-qlora
|
| 12 |
+
results: []
|
| 13 |
+
---
|
| 14 |
+
|
| 15 |
+
**Note**: This model card has been generated automatically according to the information the Trainer had access to.
|
| 16 |
+
Visit the [model card](https://ritvik19.github.io/zephyr-mini/) to see the full description.
|
| 17 |
+
|
| 18 |
+
# zephyr-1b-olmo-sft-qlora
|
| 19 |
+
|
| 20 |
+
This model is a fine-tuned version of [allenai/OLMo-1B-hf](https://huggingface.co/allenai/OLMo-1B-hf) on the HuggingFaceH4/ultrachat_200k dataset.
|
| 21 |
+
It achieves the following results on the evaluation set:
|
| 22 |
+
- Loss: 1.3126
|
| 23 |
+
|
| 24 |
+
## Training procedure
|
| 25 |
+
|
| 26 |
+
### Training hyperparameters
|
| 27 |
+
|
| 28 |
+
The following hyperparameters were used during training:
|
| 29 |
+
- learning_rate: 0.0002
|
| 30 |
+
- train_batch_size: 1
|
| 31 |
+
- eval_batch_size: 1
|
| 32 |
+
- seed: 42
|
| 33 |
+
- distributed_type: multi-GPU
|
| 34 |
+
- gradient_accumulation_steps: 128
|
| 35 |
+
- total_train_batch_size: 128
|
| 36 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 37 |
+
- lr_scheduler_type: cosine
|
| 38 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 39 |
+
- num_epochs: 1
|
| 40 |
+
|
| 41 |
+
### Training results
|
| 42 |
+
|
| 43 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 44 |
+
|:-------------:|:-----:|:----:|:---------------:|
|
| 45 |
+
| 1.3184 | 1.0 | 1001 | 1.3126 |
|
| 46 |
+
|
| 47 |
+
|
| 48 |
+
### Framework versions
|
| 49 |
+
|
| 50 |
+
- PEFT 0.7.1
|
| 51 |
+
- Transformers 4.40.1
|
| 52 |
+
- Pytorch 2.1.2+cu121
|
| 53 |
+
- Datasets 2.19.0
|
| 54 |
+
- Tokenizers 0.19.1
|
adapter_config.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "allenai/OLMo-1B-hf",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"fan_in_fan_out": false,
|
| 7 |
+
"inference_mode": true,
|
| 8 |
+
"init_lora_weights": true,
|
| 9 |
+
"layers_pattern": null,
|
| 10 |
+
"layers_to_transform": null,
|
| 11 |
+
"loftq_config": {},
|
| 12 |
+
"lora_alpha": 16,
|
| 13 |
+
"lora_dropout": 0.05,
|
| 14 |
+
"megatron_config": null,
|
| 15 |
+
"megatron_core": "megatron.core",
|
| 16 |
+
"modules_to_save": null,
|
| 17 |
+
"peft_type": "LORA",
|
| 18 |
+
"r": 16,
|
| 19 |
+
"rank_pattern": {},
|
| 20 |
+
"revision": null,
|
| 21 |
+
"target_modules": [
|
| 22 |
+
"o_proj",
|
| 23 |
+
"gate_proj",
|
| 24 |
+
"down_proj",
|
| 25 |
+
"up_proj",
|
| 26 |
+
"k_proj",
|
| 27 |
+
"q_proj",
|
| 28 |
+
"v_proj"
|
| 29 |
+
],
|
| 30 |
+
"task_type": "CAUSAL_LM"
|
| 31 |
+
}
|
adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b302b5488b99687525108213ec95e9d94696fdb107bf44345ea5857bf761eaef
|
| 3 |
+
size 24147064
|
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_loss": 1.3126407861709595,
|
| 4 |
+
"eval_runtime": 1197.5096,
|
| 5 |
+
"eval_samples": 23109,
|
| 6 |
+
"eval_samples_per_second": 11.84,
|
| 7 |
+
"eval_steps_per_second": 11.84,
|
| 8 |
+
"total_flos": 1.709524252278915e+18,
|
| 9 |
+
"train_loss": 0.07992646839473393,
|
| 10 |
+
"train_runtime": 3675.5245,
|
| 11 |
+
"train_samples": 207864,
|
| 12 |
+
"train_samples_per_second": 34.86,
|
| 13 |
+
"train_steps_per_second": 0.272
|
| 14 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "allenai/OLMo-1B-hf",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"OlmoForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": false,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"clip_qkv": null,
|
| 9 |
+
"eos_token_id": 50279,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 2048,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 8192,
|
| 14 |
+
"max_position_embeddings": 2048,
|
| 15 |
+
"model_type": "olmo",
|
| 16 |
+
"num_attention_heads": 16,
|
| 17 |
+
"num_hidden_layers": 16,
|
| 18 |
+
"num_key_value_heads": 16,
|
| 19 |
+
"pad_token_id": 1,
|
| 20 |
+
"pretraining_tp": 1,
|
| 21 |
+
"rope_scaling": null,
|
| 22 |
+
"rope_theta": 10000.0,
|
| 23 |
+
"tie_word_embeddings": true,
|
| 24 |
+
"torch_dtype": "bfloat16",
|
| 25 |
+
"transformers_version": "4.40.1",
|
| 26 |
+
"use_cache": true,
|
| 27 |
+
"vocab_size": 50304
|
| 28 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_loss": 1.3126407861709595,
|
| 4 |
+
"eval_runtime": 1197.5096,
|
| 5 |
+
"eval_samples": 23109,
|
| 6 |
+
"eval_samples_per_second": 11.84,
|
| 7 |
+
"eval_steps_per_second": 11.84
|
| 8 |
+
}
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"eos_token": {
|
| 3 |
+
"content": "<|endoftext|>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"pad_token": {
|
| 10 |
+
"content": "<|padding|>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
}
|
| 16 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,239 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": false,
|
| 5 |
+
"added_tokens_decoder": {
|
| 6 |
+
"0": {
|
| 7 |
+
"content": "|||IP_ADDRESS|||",
|
| 8 |
+
"lstrip": false,
|
| 9 |
+
"normalized": true,
|
| 10 |
+
"rstrip": false,
|
| 11 |
+
"single_word": false,
|
| 12 |
+
"special": false
|
| 13 |
+
},
|
| 14 |
+
"1": {
|
| 15 |
+
"content": "<|padding|>",
|
| 16 |
+
"lstrip": false,
|
| 17 |
+
"normalized": false,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"single_word": false,
|
| 20 |
+
"special": true
|
| 21 |
+
},
|
| 22 |
+
"50254": {
|
| 23 |
+
"content": " ",
|
| 24 |
+
"lstrip": false,
|
| 25 |
+
"normalized": true,
|
| 26 |
+
"rstrip": false,
|
| 27 |
+
"single_word": false,
|
| 28 |
+
"special": false
|
| 29 |
+
},
|
| 30 |
+
"50255": {
|
| 31 |
+
"content": " ",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": true,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false,
|
| 36 |
+
"special": false
|
| 37 |
+
},
|
| 38 |
+
"50256": {
|
| 39 |
+
"content": " ",
|
| 40 |
+
"lstrip": false,
|
| 41 |
+
"normalized": true,
|
| 42 |
+
"rstrip": false,
|
| 43 |
+
"single_word": false,
|
| 44 |
+
"special": false
|
| 45 |
+
},
|
| 46 |
+
"50257": {
|
| 47 |
+
"content": " ",
|
| 48 |
+
"lstrip": false,
|
| 49 |
+
"normalized": true,
|
| 50 |
+
"rstrip": false,
|
| 51 |
+
"single_word": false,
|
| 52 |
+
"special": false
|
| 53 |
+
},
|
| 54 |
+
"50258": {
|
| 55 |
+
"content": " ",
|
| 56 |
+
"lstrip": false,
|
| 57 |
+
"normalized": true,
|
| 58 |
+
"rstrip": false,
|
| 59 |
+
"single_word": false,
|
| 60 |
+
"special": false
|
| 61 |
+
},
|
| 62 |
+
"50259": {
|
| 63 |
+
"content": " ",
|
| 64 |
+
"lstrip": false,
|
| 65 |
+
"normalized": true,
|
| 66 |
+
"rstrip": false,
|
| 67 |
+
"single_word": false,
|
| 68 |
+
"special": false
|
| 69 |
+
},
|
| 70 |
+
"50260": {
|
| 71 |
+
"content": " ",
|
| 72 |
+
"lstrip": false,
|
| 73 |
+
"normalized": true,
|
| 74 |
+
"rstrip": false,
|
| 75 |
+
"single_word": false,
|
| 76 |
+
"special": false
|
| 77 |
+
},
|
| 78 |
+
"50261": {
|
| 79 |
+
"content": " ",
|
| 80 |
+
"lstrip": false,
|
| 81 |
+
"normalized": true,
|
| 82 |
+
"rstrip": false,
|
| 83 |
+
"single_word": false,
|
| 84 |
+
"special": false
|
| 85 |
+
},
|
| 86 |
+
"50262": {
|
| 87 |
+
"content": " ",
|
| 88 |
+
"lstrip": false,
|
| 89 |
+
"normalized": true,
|
| 90 |
+
"rstrip": false,
|
| 91 |
+
"single_word": false,
|
| 92 |
+
"special": false
|
| 93 |
+
},
|
| 94 |
+
"50263": {
|
| 95 |
+
"content": " ",
|
| 96 |
+
"lstrip": false,
|
| 97 |
+
"normalized": true,
|
| 98 |
+
"rstrip": false,
|
| 99 |
+
"single_word": false,
|
| 100 |
+
"special": false
|
| 101 |
+
},
|
| 102 |
+
"50264": {
|
| 103 |
+
"content": " ",
|
| 104 |
+
"lstrip": false,
|
| 105 |
+
"normalized": true,
|
| 106 |
+
"rstrip": false,
|
| 107 |
+
"single_word": false,
|
| 108 |
+
"special": false
|
| 109 |
+
},
|
| 110 |
+
"50265": {
|
| 111 |
+
"content": " ",
|
| 112 |
+
"lstrip": false,
|
| 113 |
+
"normalized": true,
|
| 114 |
+
"rstrip": false,
|
| 115 |
+
"single_word": false,
|
| 116 |
+
"special": false
|
| 117 |
+
},
|
| 118 |
+
"50266": {
|
| 119 |
+
"content": " ",
|
| 120 |
+
"lstrip": false,
|
| 121 |
+
"normalized": true,
|
| 122 |
+
"rstrip": false,
|
| 123 |
+
"single_word": false,
|
| 124 |
+
"special": false
|
| 125 |
+
},
|
| 126 |
+
"50267": {
|
| 127 |
+
"content": " ",
|
| 128 |
+
"lstrip": false,
|
| 129 |
+
"normalized": true,
|
| 130 |
+
"rstrip": false,
|
| 131 |
+
"single_word": false,
|
| 132 |
+
"special": false
|
| 133 |
+
},
|
| 134 |
+
"50268": {
|
| 135 |
+
"content": " ",
|
| 136 |
+
"lstrip": false,
|
| 137 |
+
"normalized": true,
|
| 138 |
+
"rstrip": false,
|
| 139 |
+
"single_word": false,
|
| 140 |
+
"special": false
|
| 141 |
+
},
|
| 142 |
+
"50269": {
|
| 143 |
+
"content": " ",
|
| 144 |
+
"lstrip": false,
|
| 145 |
+
"normalized": true,
|
| 146 |
+
"rstrip": false,
|
| 147 |
+
"single_word": false,
|
| 148 |
+
"special": false
|
| 149 |
+
},
|
| 150 |
+
"50270": {
|
| 151 |
+
"content": " ",
|
| 152 |
+
"lstrip": false,
|
| 153 |
+
"normalized": true,
|
| 154 |
+
"rstrip": false,
|
| 155 |
+
"single_word": false,
|
| 156 |
+
"special": false
|
| 157 |
+
},
|
| 158 |
+
"50271": {
|
| 159 |
+
"content": " ",
|
| 160 |
+
"lstrip": false,
|
| 161 |
+
"normalized": true,
|
| 162 |
+
"rstrip": false,
|
| 163 |
+
"single_word": false,
|
| 164 |
+
"special": false
|
| 165 |
+
},
|
| 166 |
+
"50272": {
|
| 167 |
+
"content": " ",
|
| 168 |
+
"lstrip": false,
|
| 169 |
+
"normalized": true,
|
| 170 |
+
"rstrip": false,
|
| 171 |
+
"single_word": false,
|
| 172 |
+
"special": false
|
| 173 |
+
},
|
| 174 |
+
"50273": {
|
| 175 |
+
"content": " ",
|
| 176 |
+
"lstrip": false,
|
| 177 |
+
"normalized": true,
|
| 178 |
+
"rstrip": false,
|
| 179 |
+
"single_word": false,
|
| 180 |
+
"special": false
|
| 181 |
+
},
|
| 182 |
+
"50274": {
|
| 183 |
+
"content": " ",
|
| 184 |
+
"lstrip": false,
|
| 185 |
+
"normalized": true,
|
| 186 |
+
"rstrip": false,
|
| 187 |
+
"single_word": false,
|
| 188 |
+
"special": false
|
| 189 |
+
},
|
| 190 |
+
"50275": {
|
| 191 |
+
"content": " ",
|
| 192 |
+
"lstrip": false,
|
| 193 |
+
"normalized": true,
|
| 194 |
+
"rstrip": false,
|
| 195 |
+
"single_word": false,
|
| 196 |
+
"special": false
|
| 197 |
+
},
|
| 198 |
+
"50276": {
|
| 199 |
+
"content": " ",
|
| 200 |
+
"lstrip": false,
|
| 201 |
+
"normalized": true,
|
| 202 |
+
"rstrip": false,
|
| 203 |
+
"single_word": false,
|
| 204 |
+
"special": false
|
| 205 |
+
},
|
| 206 |
+
"50277": {
|
| 207 |
+
"content": "|||EMAIL_ADDRESS|||",
|
| 208 |
+
"lstrip": false,
|
| 209 |
+
"normalized": true,
|
| 210 |
+
"rstrip": false,
|
| 211 |
+
"single_word": false,
|
| 212 |
+
"special": false
|
| 213 |
+
},
|
| 214 |
+
"50278": {
|
| 215 |
+
"content": "|||PHONE_NUMBER|||",
|
| 216 |
+
"lstrip": false,
|
| 217 |
+
"normalized": true,
|
| 218 |
+
"rstrip": false,
|
| 219 |
+
"single_word": false,
|
| 220 |
+
"special": false
|
| 221 |
+
},
|
| 222 |
+
"50279": {
|
| 223 |
+
"content": "<|endoftext|>",
|
| 224 |
+
"lstrip": false,
|
| 225 |
+
"normalized": false,
|
| 226 |
+
"rstrip": false,
|
| 227 |
+
"single_word": false,
|
| 228 |
+
"special": true
|
| 229 |
+
}
|
| 230 |
+
},
|
| 231 |
+
"bos_token": null,
|
| 232 |
+
"chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '<|user|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'system' %}\n{{ '<|system|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'assistant' %}\n{{ '<|assistant|>\n' + message['content'] + eos_token }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '<|assistant|>' }}\n{% endif %}\n{% endfor %}",
|
| 233 |
+
"clean_up_tokenization_spaces": true,
|
| 234 |
+
"eos_token": "<|endoftext|>",
|
| 235 |
+
"model_max_length": 2048,
|
| 236 |
+
"pad_token": "<|padding|>",
|
| 237 |
+
"tokenizer_class": "GPTNeoXTokenizer",
|
| 238 |
+
"unk_token": null
|
| 239 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"total_flos": 1.709524252278915e+18,
|
| 4 |
+
"train_loss": 0.07992646839473393,
|
| 5 |
+
"train_runtime": 3675.5245,
|
| 6 |
+
"train_samples": 207864,
|
| 7 |
+
"train_samples_per_second": 34.86,
|
| 8 |
+
"train_steps_per_second": 0.272
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1445 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1001,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.000999000999000999,
|
| 13 |
+
"grad_norm": 1.0703125,
|
| 14 |
+
"learning_rate": 1.9801980198019803e-06,
|
| 15 |
+
"loss": 1.9138,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.004995004995004995,
|
| 20 |
+
"grad_norm": 1.0546875,
|
| 21 |
+
"learning_rate": 9.900990099009901e-06,
|
| 22 |
+
"loss": 1.95,
|
| 23 |
+
"step": 5
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.00999000999000999,
|
| 27 |
+
"grad_norm": 0.80078125,
|
| 28 |
+
"learning_rate": 1.9801980198019803e-05,
|
| 29 |
+
"loss": 1.9142,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.014985014985014986,
|
| 34 |
+
"grad_norm": 0.671875,
|
| 35 |
+
"learning_rate": 2.9702970297029702e-05,
|
| 36 |
+
"loss": 1.8437,
|
| 37 |
+
"step": 15
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.01998001998001998,
|
| 41 |
+
"grad_norm": 0.83203125,
|
| 42 |
+
"learning_rate": 3.9603960396039605e-05,
|
| 43 |
+
"loss": 1.7806,
|
| 44 |
+
"step": 20
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.024975024975024976,
|
| 48 |
+
"grad_norm": 0.439453125,
|
| 49 |
+
"learning_rate": 4.950495049504951e-05,
|
| 50 |
+
"loss": 1.645,
|
| 51 |
+
"step": 25
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.029970029970029972,
|
| 55 |
+
"grad_norm": 0.33203125,
|
| 56 |
+
"learning_rate": 5.9405940594059404e-05,
|
| 57 |
+
"loss": 1.5936,
|
| 58 |
+
"step": 30
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.03496503496503497,
|
| 62 |
+
"grad_norm": 0.345703125,
|
| 63 |
+
"learning_rate": 6.93069306930693e-05,
|
| 64 |
+
"loss": 1.5451,
|
| 65 |
+
"step": 35
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.03996003996003996,
|
| 69 |
+
"grad_norm": 0.2138671875,
|
| 70 |
+
"learning_rate": 7.920792079207921e-05,
|
| 71 |
+
"loss": 1.5002,
|
| 72 |
+
"step": 40
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.04495504495504495,
|
| 76 |
+
"grad_norm": 0.15625,
|
| 77 |
+
"learning_rate": 8.910891089108912e-05,
|
| 78 |
+
"loss": 1.4846,
|
| 79 |
+
"step": 45
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.04995004995004995,
|
| 83 |
+
"grad_norm": 0.16015625,
|
| 84 |
+
"learning_rate": 9.900990099009902e-05,
|
| 85 |
+
"loss": 1.4525,
|
| 86 |
+
"step": 50
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.054945054945054944,
|
| 90 |
+
"grad_norm": 0.1328125,
|
| 91 |
+
"learning_rate": 0.00010891089108910893,
|
| 92 |
+
"loss": 1.4746,
|
| 93 |
+
"step": 55
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.059940059940059943,
|
| 97 |
+
"grad_norm": 0.177734375,
|
| 98 |
+
"learning_rate": 0.00011881188118811881,
|
| 99 |
+
"loss": 1.439,
|
| 100 |
+
"step": 60
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.06493506493506493,
|
| 104 |
+
"grad_norm": 0.1044921875,
|
| 105 |
+
"learning_rate": 0.00012871287128712872,
|
| 106 |
+
"loss": 1.4421,
|
| 107 |
+
"step": 65
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.06993006993006994,
|
| 111 |
+
"grad_norm": 0.09326171875,
|
| 112 |
+
"learning_rate": 0.0001386138613861386,
|
| 113 |
+
"loss": 1.4223,
|
| 114 |
+
"step": 70
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.07492507492507493,
|
| 118 |
+
"grad_norm": 0.091796875,
|
| 119 |
+
"learning_rate": 0.0001485148514851485,
|
| 120 |
+
"loss": 1.4025,
|
| 121 |
+
"step": 75
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.07992007992007992,
|
| 125 |
+
"grad_norm": 0.1083984375,
|
| 126 |
+
"learning_rate": 0.00015841584158415842,
|
| 127 |
+
"loss": 1.4102,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.08491508491508491,
|
| 132 |
+
"grad_norm": 0.0869140625,
|
| 133 |
+
"learning_rate": 0.00016831683168316833,
|
| 134 |
+
"loss": 1.3908,
|
| 135 |
+
"step": 85
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.0899100899100899,
|
| 139 |
+
"grad_norm": 0.09814453125,
|
| 140 |
+
"learning_rate": 0.00017821782178217824,
|
| 141 |
+
"loss": 1.4007,
|
| 142 |
+
"step": 90
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.09490509490509491,
|
| 146 |
+
"grad_norm": 0.0859375,
|
| 147 |
+
"learning_rate": 0.00018811881188118812,
|
| 148 |
+
"loss": 1.3824,
|
| 149 |
+
"step": 95
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.0999000999000999,
|
| 153 |
+
"grad_norm": 0.099609375,
|
| 154 |
+
"learning_rate": 0.00019801980198019803,
|
| 155 |
+
"loss": 1.3897,
|
| 156 |
+
"step": 100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.1048951048951049,
|
| 160 |
+
"grad_norm": 0.08935546875,
|
| 161 |
+
"learning_rate": 0.00019999025240093044,
|
| 162 |
+
"loss": 1.3614,
|
| 163 |
+
"step": 105
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.10989010989010989,
|
| 167 |
+
"grad_norm": 0.087890625,
|
| 168 |
+
"learning_rate": 0.00019995065603657316,
|
| 169 |
+
"loss": 1.3792,
|
| 170 |
+
"step": 110
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.11488511488511488,
|
| 174 |
+
"grad_norm": 0.09423828125,
|
| 175 |
+
"learning_rate": 0.0001998806137341434,
|
| 176 |
+
"loss": 1.37,
|
| 177 |
+
"step": 115
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.11988011988011989,
|
| 181 |
+
"grad_norm": 0.09619140625,
|
| 182 |
+
"learning_rate": 0.000199780146829205,
|
| 183 |
+
"loss": 1.3638,
|
| 184 |
+
"step": 120
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.12487512487512488,
|
| 188 |
+
"grad_norm": 0.10107421875,
|
| 189 |
+
"learning_rate": 0.00019964928592495045,
|
| 190 |
+
"loss": 1.3911,
|
| 191 |
+
"step": 125
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.12987012987012986,
|
| 195 |
+
"grad_norm": 0.10107421875,
|
| 196 |
+
"learning_rate": 0.00019948807088287883,
|
| 197 |
+
"loss": 1.3534,
|
| 198 |
+
"step": 130
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.13486513486513488,
|
| 202 |
+
"grad_norm": 0.095703125,
|
| 203 |
+
"learning_rate": 0.0001992965508106537,
|
| 204 |
+
"loss": 1.3619,
|
| 205 |
+
"step": 135
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.13986013986013987,
|
| 209 |
+
"grad_norm": 0.1025390625,
|
| 210 |
+
"learning_rate": 0.00019907478404714436,
|
| 211 |
+
"loss": 1.3612,
|
| 212 |
+
"step": 140
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.14485514485514486,
|
| 216 |
+
"grad_norm": 0.10302734375,
|
| 217 |
+
"learning_rate": 0.0001988228381446553,
|
| 218 |
+
"loss": 1.3531,
|
| 219 |
+
"step": 145
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.14985014985014986,
|
| 223 |
+
"grad_norm": 0.09130859375,
|
| 224 |
+
"learning_rate": 0.00019854078984834903,
|
| 225 |
+
"loss": 1.3589,
|
| 226 |
+
"step": 150
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.15484515484515485,
|
| 230 |
+
"grad_norm": 0.1044921875,
|
| 231 |
+
"learning_rate": 0.0001982287250728689,
|
| 232 |
+
"loss": 1.3393,
|
| 233 |
+
"step": 155
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.15984015984015984,
|
| 237 |
+
"grad_norm": 0.09765625,
|
| 238 |
+
"learning_rate": 0.0001978867388761685,
|
| 239 |
+
"loss": 1.3378,
|
| 240 |
+
"step": 160
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.16483516483516483,
|
| 244 |
+
"grad_norm": 0.09521484375,
|
| 245 |
+
"learning_rate": 0.00019751493543055632,
|
| 246 |
+
"loss": 1.3435,
|
| 247 |
+
"step": 165
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.16983016983016982,
|
| 251 |
+
"grad_norm": 0.10595703125,
|
| 252 |
+
"learning_rate": 0.00019711342799096361,
|
| 253 |
+
"loss": 1.3241,
|
| 254 |
+
"step": 170
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.17482517482517482,
|
| 258 |
+
"grad_norm": 0.11376953125,
|
| 259 |
+
"learning_rate": 0.00019668233886044597,
|
| 260 |
+
"loss": 1.3394,
|
| 261 |
+
"step": 175
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.1798201798201798,
|
| 265 |
+
"grad_norm": 0.09765625,
|
| 266 |
+
"learning_rate": 0.00019622179935292855,
|
| 267 |
+
"loss": 1.3357,
|
| 268 |
+
"step": 180
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.1848151848151848,
|
| 272 |
+
"grad_norm": 0.09814453125,
|
| 273 |
+
"learning_rate": 0.00019573194975320673,
|
| 274 |
+
"loss": 1.3377,
|
| 275 |
+
"step": 185
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.18981018981018982,
|
| 279 |
+
"grad_norm": 0.103515625,
|
| 280 |
+
"learning_rate": 0.00019521293927421388,
|
| 281 |
+
"loss": 1.3461,
|
| 282 |
+
"step": 190
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.19480519480519481,
|
| 286 |
+
"grad_norm": 0.09765625,
|
| 287 |
+
"learning_rate": 0.00019466492601156966,
|
| 288 |
+
"loss": 1.3366,
|
| 289 |
+
"step": 195
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.1998001998001998,
|
| 293 |
+
"grad_norm": 0.11083984375,
|
| 294 |
+
"learning_rate": 0.00019408807689542257,
|
| 295 |
+
"loss": 1.3307,
|
| 296 |
+
"step": 200
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.2047952047952048,
|
| 300 |
+
"grad_norm": 0.1005859375,
|
| 301 |
+
"learning_rate": 0.00019348256763960145,
|
| 302 |
+
"loss": 1.336,
|
| 303 |
+
"step": 205
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.2097902097902098,
|
| 307 |
+
"grad_norm": 0.10205078125,
|
| 308 |
+
"learning_rate": 0.00019284858268809137,
|
| 309 |
+
"loss": 1.3395,
|
| 310 |
+
"step": 210
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.21478521478521478,
|
| 314 |
+
"grad_norm": 0.09814453125,
|
| 315 |
+
"learning_rate": 0.00019218631515885006,
|
| 316 |
+
"loss": 1.3457,
|
| 317 |
+
"step": 215
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.21978021978021978,
|
| 321 |
+
"grad_norm": 0.09912109375,
|
| 322 |
+
"learning_rate": 0.0001914959667849825,
|
| 323 |
+
"loss": 1.3472,
|
| 324 |
+
"step": 220
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.22477522477522477,
|
| 328 |
+
"grad_norm": 0.10791015625,
|
| 329 |
+
"learning_rate": 0.00019077774785329087,
|
| 330 |
+
"loss": 1.3418,
|
| 331 |
+
"step": 225
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.22977022977022976,
|
| 335 |
+
"grad_norm": 0.1044921875,
|
| 336 |
+
"learning_rate": 0.00019003187714021938,
|
| 337 |
+
"loss": 1.3217,
|
| 338 |
+
"step": 230
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.23476523476523475,
|
| 342 |
+
"grad_norm": 0.09716796875,
|
| 343 |
+
"learning_rate": 0.00018925858184521256,
|
| 344 |
+
"loss": 1.3256,
|
| 345 |
+
"step": 235
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.23976023976023977,
|
| 349 |
+
"grad_norm": 0.099609375,
|
| 350 |
+
"learning_rate": 0.0001884580975215084,
|
| 351 |
+
"loss": 1.3417,
|
| 352 |
+
"step": 240
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.24475524475524477,
|
| 356 |
+
"grad_norm": 0.09619140625,
|
| 357 |
+
"learning_rate": 0.00018763066800438636,
|
| 358 |
+
"loss": 1.3261,
|
| 359 |
+
"step": 245
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.24975024975024976,
|
| 363 |
+
"grad_norm": 0.09912109375,
|
| 364 |
+
"learning_rate": 0.00018677654533689287,
|
| 365 |
+
"loss": 1.3252,
|
| 366 |
+
"step": 250
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.2547452547452547,
|
| 370 |
+
"grad_norm": 0.1025390625,
|
| 371 |
+
"learning_rate": 0.00018589598969306645,
|
| 372 |
+
"loss": 1.3387,
|
| 373 |
+
"step": 255
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.2597402597402597,
|
| 377 |
+
"grad_norm": 0.1025390625,
|
| 378 |
+
"learning_rate": 0.00018498926929868642,
|
| 379 |
+
"loss": 1.3577,
|
| 380 |
+
"step": 260
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.2647352647352647,
|
| 384 |
+
"grad_norm": 0.10302734375,
|
| 385 |
+
"learning_rate": 0.00018405666034956844,
|
| 386 |
+
"loss": 1.3321,
|
| 387 |
+
"step": 265
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.26973026973026976,
|
| 391 |
+
"grad_norm": 0.10400390625,
|
| 392 |
+
"learning_rate": 0.00018309844692743283,
|
| 393 |
+
"loss": 1.3067,
|
| 394 |
+
"step": 270
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.27472527472527475,
|
| 398 |
+
"grad_norm": 0.1103515625,
|
| 399 |
+
"learning_rate": 0.00018211492091337042,
|
| 400 |
+
"loss": 1.3489,
|
| 401 |
+
"step": 275
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.27972027972027974,
|
| 405 |
+
"grad_norm": 0.10400390625,
|
| 406 |
+
"learning_rate": 0.00018110638189893267,
|
| 407 |
+
"loss": 1.3276,
|
| 408 |
+
"step": 280
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.28471528471528473,
|
| 412 |
+
"grad_norm": 0.09814453125,
|
| 413 |
+
"learning_rate": 0.00018007313709487334,
|
| 414 |
+
"loss": 1.3258,
|
| 415 |
+
"step": 285
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.2897102897102897,
|
| 419 |
+
"grad_norm": 0.09521484375,
|
| 420 |
+
"learning_rate": 0.00017901550123756906,
|
| 421 |
+
"loss": 1.332,
|
| 422 |
+
"step": 290
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.2947052947052947,
|
| 426 |
+
"grad_norm": 0.10009765625,
|
| 427 |
+
"learning_rate": 0.00017793379649314744,
|
| 428 |
+
"loss": 1.335,
|
| 429 |
+
"step": 295
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.2997002997002997,
|
| 433 |
+
"grad_norm": 0.09912109375,
|
| 434 |
+
"learning_rate": 0.00017682835235935236,
|
| 435 |
+
"loss": 1.3266,
|
| 436 |
+
"step": 300
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.3046953046953047,
|
| 440 |
+
"grad_norm": 0.1083984375,
|
| 441 |
+
"learning_rate": 0.00017569950556517566,
|
| 442 |
+
"loss": 1.3174,
|
| 443 |
+
"step": 305
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.3096903096903097,
|
| 447 |
+
"grad_norm": 0.09326171875,
|
| 448 |
+
"learning_rate": 0.00017454759996828623,
|
| 449 |
+
"loss": 1.3212,
|
| 450 |
+
"step": 310
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.3146853146853147,
|
| 454 |
+
"grad_norm": 0.10107421875,
|
| 455 |
+
"learning_rate": 0.00017337298645028764,
|
| 456 |
+
"loss": 1.326,
|
| 457 |
+
"step": 315
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.3196803196803197,
|
| 461 |
+
"grad_norm": 0.10400390625,
|
| 462 |
+
"learning_rate": 0.00017217602280983623,
|
| 463 |
+
"loss": 1.3364,
|
| 464 |
+
"step": 320
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.3246753246753247,
|
| 468 |
+
"grad_norm": 0.09814453125,
|
| 469 |
+
"learning_rate": 0.0001709570736536521,
|
| 470 |
+
"loss": 1.3168,
|
| 471 |
+
"step": 325
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.32967032967032966,
|
| 475 |
+
"grad_norm": 0.10498046875,
|
| 476 |
+
"learning_rate": 0.00016971651028545648,
|
| 477 |
+
"loss": 1.3166,
|
| 478 |
+
"step": 330
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.33466533466533466,
|
| 482 |
+
"grad_norm": 0.1005859375,
|
| 483 |
+
"learning_rate": 0.00016845471059286887,
|
| 484 |
+
"loss": 1.3228,
|
| 485 |
+
"step": 335
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.33966033966033965,
|
| 489 |
+
"grad_norm": 0.1015625,
|
| 490 |
+
"learning_rate": 0.00016717205893229903,
|
| 491 |
+
"loss": 1.3277,
|
| 492 |
+
"step": 340
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.34465534465534464,
|
| 496 |
+
"grad_norm": 0.0986328125,
|
| 497 |
+
"learning_rate": 0.00016586894601186805,
|
| 498 |
+
"loss": 1.3197,
|
| 499 |
+
"step": 345
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.34965034965034963,
|
| 503 |
+
"grad_norm": 0.09765625,
|
| 504 |
+
"learning_rate": 0.00016454576877239507,
|
| 505 |
+
"loss": 1.3276,
|
| 506 |
+
"step": 350
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.3546453546453546,
|
| 510 |
+
"grad_norm": 0.09912109375,
|
| 511 |
+
"learning_rate": 0.0001632029302664851,
|
| 512 |
+
"loss": 1.309,
|
| 513 |
+
"step": 355
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.3596403596403596,
|
| 517 |
+
"grad_norm": 0.095703125,
|
| 518 |
+
"learning_rate": 0.0001618408395357554,
|
| 519 |
+
"loss": 1.334,
|
| 520 |
+
"step": 360
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.3646353646353646,
|
| 524 |
+
"grad_norm": 0.10400390625,
|
| 525 |
+
"learning_rate": 0.0001604599114862375,
|
| 526 |
+
"loss": 1.3287,
|
| 527 |
+
"step": 365
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.3696303696303696,
|
| 531 |
+
"grad_norm": 0.09765625,
|
| 532 |
+
"learning_rate": 0.00015906056676199255,
|
| 533 |
+
"loss": 1.3165,
|
| 534 |
+
"step": 370
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.37462537462537465,
|
| 538 |
+
"grad_norm": 0.10107421875,
|
| 539 |
+
"learning_rate": 0.00015764323161697935,
|
| 540 |
+
"loss": 1.3127,
|
| 541 |
+
"step": 375
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.37962037962037964,
|
| 545 |
+
"grad_norm": 0.09814453125,
|
| 546 |
+
"learning_rate": 0.00015620833778521307,
|
| 547 |
+
"loss": 1.3297,
|
| 548 |
+
"step": 380
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.38461538461538464,
|
| 552 |
+
"grad_norm": 0.10400390625,
|
| 553 |
+
"learning_rate": 0.00015475632234925504,
|
| 554 |
+
"loss": 1.3184,
|
| 555 |
+
"step": 385
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.38961038961038963,
|
| 559 |
+
"grad_norm": 0.0986328125,
|
| 560 |
+
"learning_rate": 0.000153287627607073,
|
| 561 |
+
"loss": 1.3004,
|
| 562 |
+
"step": 390
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.3946053946053946,
|
| 566 |
+
"grad_norm": 0.09912109375,
|
| 567 |
+
"learning_rate": 0.00015180270093731303,
|
| 568 |
+
"loss": 1.3179,
|
| 569 |
+
"step": 395
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.3996003996003996,
|
| 573 |
+
"grad_norm": 0.09619140625,
|
| 574 |
+
"learning_rate": 0.00015030199466302353,
|
| 575 |
+
"loss": 1.3294,
|
| 576 |
+
"step": 400
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.4045954045954046,
|
| 580 |
+
"grad_norm": 0.1044921875,
|
| 581 |
+
"learning_rate": 0.0001487859659138733,
|
| 582 |
+
"loss": 1.3469,
|
| 583 |
+
"step": 405
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.4095904095904096,
|
| 587 |
+
"grad_norm": 0.10009765625,
|
| 588 |
+
"learning_rate": 0.00014725507648690543,
|
| 589 |
+
"loss": 1.3166,
|
| 590 |
+
"step": 410
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.4145854145854146,
|
| 594 |
+
"grad_norm": 0.10205078125,
|
| 595 |
+
"learning_rate": 0.00014570979270586945,
|
| 596 |
+
"loss": 1.3149,
|
| 597 |
+
"step": 415
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.4195804195804196,
|
| 601 |
+
"grad_norm": 0.09716796875,
|
| 602 |
+
"learning_rate": 0.00014415058527917452,
|
| 603 |
+
"loss": 1.3365,
|
| 604 |
+
"step": 420
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.4245754245754246,
|
| 608 |
+
"grad_norm": 0.10693359375,
|
| 609 |
+
"learning_rate": 0.00014257792915650728,
|
| 610 |
+
"loss": 1.3289,
|
| 611 |
+
"step": 425
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.42957042957042957,
|
| 615 |
+
"grad_norm": 0.103515625,
|
| 616 |
+
"learning_rate": 0.00014099230338415728,
|
| 617 |
+
"loss": 1.2932,
|
| 618 |
+
"step": 430
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.43456543456543456,
|
| 622 |
+
"grad_norm": 0.11865234375,
|
| 623 |
+
"learning_rate": 0.00013939419095909512,
|
| 624 |
+
"loss": 1.318,
|
| 625 |
+
"step": 435
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.43956043956043955,
|
| 629 |
+
"grad_norm": 0.10400390625,
|
| 630 |
+
"learning_rate": 0.00013778407868184672,
|
| 631 |
+
"loss": 1.2892,
|
| 632 |
+
"step": 440
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.44455544455544455,
|
| 636 |
+
"grad_norm": 0.09619140625,
|
| 637 |
+
"learning_rate": 0.00013616245700820922,
|
| 638 |
+
"loss": 1.3114,
|
| 639 |
+
"step": 445
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.44955044955044954,
|
| 643 |
+
"grad_norm": 0.1015625,
|
| 644 |
+
"learning_rate": 0.00013452981989985348,
|
| 645 |
+
"loss": 1.3143,
|
| 646 |
+
"step": 450
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.45454545454545453,
|
| 650 |
+
"grad_norm": 0.10498046875,
|
| 651 |
+
"learning_rate": 0.00013288666467385833,
|
| 652 |
+
"loss": 1.3231,
|
| 653 |
+
"step": 455
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.4595404595404595,
|
| 657 |
+
"grad_norm": 0.095703125,
|
| 658 |
+
"learning_rate": 0.00013123349185122327,
|
| 659 |
+
"loss": 1.3077,
|
| 660 |
+
"step": 460
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.4645354645354645,
|
| 664 |
+
"grad_norm": 0.0966796875,
|
| 665 |
+
"learning_rate": 0.00012957080500440468,
|
| 666 |
+
"loss": 1.3247,
|
| 667 |
+
"step": 465
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.4695304695304695,
|
| 671 |
+
"grad_norm": 0.0947265625,
|
| 672 |
+
"learning_rate": 0.00012789911060392294,
|
| 673 |
+
"loss": 1.3097,
|
| 674 |
+
"step": 470
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.4745254745254745,
|
| 678 |
+
"grad_norm": 0.09765625,
|
| 679 |
+
"learning_rate": 0.00012621891786408648,
|
| 680 |
+
"loss": 1.3099,
|
| 681 |
+
"step": 475
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.47952047952047955,
|
| 685 |
+
"grad_norm": 0.09912109375,
|
| 686 |
+
"learning_rate": 0.00012453073858788026,
|
| 687 |
+
"loss": 1.3267,
|
| 688 |
+
"step": 480
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.48451548451548454,
|
| 692 |
+
"grad_norm": 0.0986328125,
|
| 693 |
+
"learning_rate": 0.00012283508701106557,
|
| 694 |
+
"loss": 1.3114,
|
| 695 |
+
"step": 485
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.48951048951048953,
|
| 699 |
+
"grad_norm": 0.09521484375,
|
| 700 |
+
"learning_rate": 0.00012113247964553888,
|
| 701 |
+
"loss": 1.2904,
|
| 702 |
+
"step": 490
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.4945054945054945,
|
| 706 |
+
"grad_norm": 0.103515625,
|
| 707 |
+
"learning_rate": 0.0001194234351219972,
|
| 708 |
+
"loss": 1.3151,
|
| 709 |
+
"step": 495
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.4995004995004995,
|
| 713 |
+
"grad_norm": 0.10693359375,
|
| 714 |
+
"learning_rate": 0.00011770847403195834,
|
| 715 |
+
"loss": 1.3126,
|
| 716 |
+
"step": 500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.5044955044955045,
|
| 720 |
+
"grad_norm": 0.099609375,
|
| 721 |
+
"learning_rate": 0.0001159881187691835,
|
| 722 |
+
"loss": 1.3108,
|
| 723 |
+
"step": 505
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.5094905094905094,
|
| 727 |
+
"grad_norm": 0.10302734375,
|
| 728 |
+
"learning_rate": 0.00011426289337055119,
|
| 729 |
+
"loss": 1.3032,
|
| 730 |
+
"step": 510
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.5144855144855145,
|
| 734 |
+
"grad_norm": 0.09912109375,
|
| 735 |
+
"learning_rate": 0.00011253332335643043,
|
| 736 |
+
"loss": 1.3043,
|
| 737 |
+
"step": 515
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.5194805194805194,
|
| 741 |
+
"grad_norm": 0.099609375,
|
| 742 |
+
"learning_rate": 0.0001107999355706023,
|
| 743 |
+
"loss": 1.3188,
|
| 744 |
+
"step": 520
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.5244755244755245,
|
| 748 |
+
"grad_norm": 0.09619140625,
|
| 749 |
+
"learning_rate": 0.00010906325801977804,
|
| 750 |
+
"loss": 1.3266,
|
| 751 |
+
"step": 525
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.5294705294705294,
|
| 755 |
+
"grad_norm": 0.10498046875,
|
| 756 |
+
"learning_rate": 0.00010732381971276318,
|
| 757 |
+
"loss": 1.3156,
|
| 758 |
+
"step": 530
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.5344655344655345,
|
| 762 |
+
"grad_norm": 0.10400390625,
|
| 763 |
+
"learning_rate": 0.00010558215049931638,
|
| 764 |
+
"loss": 1.304,
|
| 765 |
+
"step": 535
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.5394605394605395,
|
| 769 |
+
"grad_norm": 0.095703125,
|
| 770 |
+
"learning_rate": 0.00010383878090875201,
|
| 771 |
+
"loss": 1.3089,
|
| 772 |
+
"step": 540
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.5444555444555444,
|
| 776 |
+
"grad_norm": 0.09765625,
|
| 777 |
+
"learning_rate": 0.0001020942419883357,
|
| 778 |
+
"loss": 1.3298,
|
| 779 |
+
"step": 545
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.5494505494505495,
|
| 783 |
+
"grad_norm": 0.09521484375,
|
| 784 |
+
"learning_rate": 0.00010034906514152238,
|
| 785 |
+
"loss": 1.3127,
|
| 786 |
+
"step": 550
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.5544455544455544,
|
| 790 |
+
"grad_norm": 0.10107421875,
|
| 791 |
+
"learning_rate": 9.860378196608549e-05,
|
| 792 |
+
"loss": 1.3248,
|
| 793 |
+
"step": 555
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.5594405594405595,
|
| 797 |
+
"grad_norm": 0.10107421875,
|
| 798 |
+
"learning_rate": 9.685892409218717e-05,
|
| 799 |
+
"loss": 1.3116,
|
| 800 |
+
"step": 560
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.5644355644355644,
|
| 804 |
+
"grad_norm": 0.09619140625,
|
| 805 |
+
"learning_rate": 9.511502302043868e-05,
|
| 806 |
+
"loss": 1.3196,
|
| 807 |
+
"step": 565
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.5694305694305695,
|
| 811 |
+
"grad_norm": 0.1005859375,
|
| 812 |
+
"learning_rate": 9.337260996000002e-05,
|
| 813 |
+
"loss": 1.3183,
|
| 814 |
+
"step": 570
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.5744255744255744,
|
| 818 |
+
"grad_norm": 0.10107421875,
|
| 819 |
+
"learning_rate": 9.163221566676847e-05,
|
| 820 |
+
"loss": 1.3165,
|
| 821 |
+
"step": 575
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.5794205794205795,
|
| 825 |
+
"grad_norm": 0.09716796875,
|
| 826 |
+
"learning_rate": 8.989437028170537e-05,
|
| 827 |
+
"loss": 1.3125,
|
| 828 |
+
"step": 580
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.5844155844155844,
|
| 832 |
+
"grad_norm": 0.0986328125,
|
| 833 |
+
"learning_rate": 8.81596031693499e-05,
|
| 834 |
+
"loss": 1.325,
|
| 835 |
+
"step": 585
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.5894105894105894,
|
| 839 |
+
"grad_norm": 0.09716796875,
|
| 840 |
+
"learning_rate": 8.642844275656957e-05,
|
| 841 |
+
"loss": 1.3308,
|
| 842 |
+
"step": 590
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.5944055944055944,
|
| 846 |
+
"grad_norm": 0.10107421875,
|
| 847 |
+
"learning_rate": 8.47014163715962e-05,
|
| 848 |
+
"loss": 1.3099,
|
| 849 |
+
"step": 595
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.5994005994005994,
|
| 853 |
+
"grad_norm": 0.10009765625,
|
| 854 |
+
"learning_rate": 8.297905008339677e-05,
|
| 855 |
+
"loss": 1.2791,
|
| 856 |
+
"step": 600
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.6043956043956044,
|
| 860 |
+
"grad_norm": 0.09619140625,
|
| 861 |
+
"learning_rate": 8.126186854142752e-05,
|
| 862 |
+
"loss": 1.3162,
|
| 863 |
+
"step": 605
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.6093906093906094,
|
| 867 |
+
"grad_norm": 0.095703125,
|
| 868 |
+
"learning_rate": 7.955039481582097e-05,
|
| 869 |
+
"loss": 1.3093,
|
| 870 |
+
"step": 610
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.6143856143856143,
|
| 874 |
+
"grad_norm": 0.09716796875,
|
| 875 |
+
"learning_rate": 7.784515023805328e-05,
|
| 876 |
+
"loss": 1.309,
|
| 877 |
+
"step": 615
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.6193806193806194,
|
| 881 |
+
"grad_norm": 0.095703125,
|
| 882 |
+
"learning_rate": 7.614665424214193e-05,
|
| 883 |
+
"loss": 1.285,
|
| 884 |
+
"step": 620
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.6243756243756243,
|
| 888 |
+
"grad_norm": 0.095703125,
|
| 889 |
+
"learning_rate": 7.445542420642097e-05,
|
| 890 |
+
"loss": 1.3155,
|
| 891 |
+
"step": 625
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.6293706293706294,
|
| 895 |
+
"grad_norm": 0.1005859375,
|
| 896 |
+
"learning_rate": 7.277197529594257e-05,
|
| 897 |
+
"loss": 1.301,
|
| 898 |
+
"step": 630
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.6343656343656343,
|
| 902 |
+
"grad_norm": 0.09912109375,
|
| 903 |
+
"learning_rate": 7.109682030555283e-05,
|
| 904 |
+
"loss": 1.295,
|
| 905 |
+
"step": 635
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.6393606393606394,
|
| 909 |
+
"grad_norm": 0.0966796875,
|
| 910 |
+
"learning_rate": 6.943046950368944e-05,
|
| 911 |
+
"loss": 1.315,
|
| 912 |
+
"step": 640
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.6443556443556444,
|
| 916 |
+
"grad_norm": 0.09765625,
|
| 917 |
+
"learning_rate": 6.77734304769489e-05,
|
| 918 |
+
"loss": 1.2878,
|
| 919 |
+
"step": 645
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.6493506493506493,
|
| 923 |
+
"grad_norm": 0.1005859375,
|
| 924 |
+
"learning_rate": 6.612620797547087e-05,
|
| 925 |
+
"loss": 1.3135,
|
| 926 |
+
"step": 650
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.6543456543456544,
|
| 930 |
+
"grad_norm": 0.09912109375,
|
| 931 |
+
"learning_rate": 6.448930375918631e-05,
|
| 932 |
+
"loss": 1.2958,
|
| 933 |
+
"step": 655
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.6593406593406593,
|
| 937 |
+
"grad_norm": 0.10107421875,
|
| 938 |
+
"learning_rate": 6.286321644497655e-05,
|
| 939 |
+
"loss": 1.3286,
|
| 940 |
+
"step": 660
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.6643356643356644,
|
| 944 |
+
"grad_norm": 0.09765625,
|
| 945 |
+
"learning_rate": 6.12484413547897e-05,
|
| 946 |
+
"loss": 1.2988,
|
| 947 |
+
"step": 665
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.6693306693306693,
|
| 951 |
+
"grad_norm": 0.0966796875,
|
| 952 |
+
"learning_rate": 5.964547036476099e-05,
|
| 953 |
+
"loss": 1.3086,
|
| 954 |
+
"step": 670
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.6743256743256744,
|
| 958 |
+
"grad_norm": 0.09765625,
|
| 959 |
+
"learning_rate": 5.805479175538229e-05,
|
| 960 |
+
"loss": 1.3056,
|
| 961 |
+
"step": 675
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.6793206793206793,
|
| 965 |
+
"grad_norm": 0.0966796875,
|
| 966 |
+
"learning_rate": 5.647689006276726e-05,
|
| 967 |
+
"loss": 1.3259,
|
| 968 |
+
"step": 680
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.6843156843156843,
|
| 972 |
+
"grad_norm": 0.09619140625,
|
| 973 |
+
"learning_rate": 5.491224593105695e-05,
|
| 974 |
+
"loss": 1.3105,
|
| 975 |
+
"step": 685
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.6893106893106893,
|
| 979 |
+
"grad_norm": 0.095703125,
|
| 980 |
+
"learning_rate": 5.33613359660109e-05,
|
| 981 |
+
"loss": 1.3113,
|
| 982 |
+
"step": 690
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.6943056943056943,
|
| 986 |
+
"grad_norm": 0.095703125,
|
| 987 |
+
"learning_rate": 5.182463258982846e-05,
|
| 988 |
+
"loss": 1.3144,
|
| 989 |
+
"step": 695
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.6993006993006993,
|
| 993 |
+
"grad_norm": 0.09912109375,
|
| 994 |
+
"learning_rate": 5.0302603897244474e-05,
|
| 995 |
+
"loss": 1.3059,
|
| 996 |
+
"step": 700
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.7042957042957043,
|
| 1000 |
+
"grad_norm": 0.0966796875,
|
| 1001 |
+
"learning_rate": 4.8795713512942865e-05,
|
| 1002 |
+
"loss": 1.2977,
|
| 1003 |
+
"step": 705
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.7092907092907093,
|
| 1007 |
+
"grad_norm": 0.10009765625,
|
| 1008 |
+
"learning_rate": 4.7304420450332244e-05,
|
| 1009 |
+
"loss": 1.3051,
|
| 1010 |
+
"step": 710
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.7142857142857143,
|
| 1014 |
+
"grad_norm": 0.09912109375,
|
| 1015 |
+
"learning_rate": 4.582917897172603e-05,
|
| 1016 |
+
"loss": 1.3229,
|
| 1017 |
+
"step": 715
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.7192807192807192,
|
| 1021 |
+
"grad_norm": 0.09423828125,
|
| 1022 |
+
"learning_rate": 4.437043844996952e-05,
|
| 1023 |
+
"loss": 1.3063,
|
| 1024 |
+
"step": 720
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.7242757242757243,
|
| 1028 |
+
"grad_norm": 0.099609375,
|
| 1029 |
+
"learning_rate": 4.2928643231556844e-05,
|
| 1030 |
+
"loss": 1.2956,
|
| 1031 |
+
"step": 725
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.7292707292707292,
|
| 1035 |
+
"grad_norm": 0.0947265625,
|
| 1036 |
+
"learning_rate": 4.150423250127845e-05,
|
| 1037 |
+
"loss": 1.3036,
|
| 1038 |
+
"step": 730
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.7342657342657343,
|
| 1042 |
+
"grad_norm": 0.0947265625,
|
| 1043 |
+
"learning_rate": 4.009764014844143e-05,
|
| 1044 |
+
"loss": 1.3252,
|
| 1045 |
+
"step": 735
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.7392607392607392,
|
| 1049 |
+
"grad_norm": 0.09375,
|
| 1050 |
+
"learning_rate": 3.8709294634702376e-05,
|
| 1051 |
+
"loss": 1.3207,
|
| 1052 |
+
"step": 740
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.7442557442557443,
|
| 1056 |
+
"grad_norm": 0.09423828125,
|
| 1057 |
+
"learning_rate": 3.733961886355398e-05,
|
| 1058 |
+
"loss": 1.2915,
|
| 1059 |
+
"step": 745
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.7492507492507493,
|
| 1063 |
+
"grad_norm": 0.0986328125,
|
| 1064 |
+
"learning_rate": 3.5989030051504434e-05,
|
| 1065 |
+
"loss": 1.3117,
|
| 1066 |
+
"step": 750
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.7542457542457542,
|
| 1070 |
+
"grad_norm": 0.09814453125,
|
| 1071 |
+
"learning_rate": 3.465793960098945e-05,
|
| 1072 |
+
"loss": 1.3022,
|
| 1073 |
+
"step": 755
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.7592407592407593,
|
| 1077 |
+
"grad_norm": 0.09716796875,
|
| 1078 |
+
"learning_rate": 3.334675297505476e-05,
|
| 1079 |
+
"loss": 1.3298,
|
| 1080 |
+
"step": 760
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.7642357642357642,
|
| 1084 |
+
"grad_norm": 0.09521484375,
|
| 1085 |
+
"learning_rate": 3.205586957384838e-05,
|
| 1086 |
+
"loss": 1.3085,
|
| 1087 |
+
"step": 765
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.7692307692307693,
|
| 1091 |
+
"grad_norm": 0.09326171875,
|
| 1092 |
+
"learning_rate": 3.078568261295933e-05,
|
| 1093 |
+
"loss": 1.2917,
|
| 1094 |
+
"step": 770
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.7742257742257742,
|
| 1098 |
+
"grad_norm": 0.09814453125,
|
| 1099 |
+
"learning_rate": 2.953657900364053e-05,
|
| 1100 |
+
"loss": 1.3111,
|
| 1101 |
+
"step": 775
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.7792207792207793,
|
| 1105 |
+
"grad_norm": 0.09326171875,
|
| 1106 |
+
"learning_rate": 2.8308939234951726e-05,
|
| 1107 |
+
"loss": 1.3108,
|
| 1108 |
+
"step": 780
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.7842157842157842,
|
| 1112 |
+
"grad_norm": 0.09326171875,
|
| 1113 |
+
"learning_rate": 2.7103137257858868e-05,
|
| 1114 |
+
"loss": 1.3106,
|
| 1115 |
+
"step": 785
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.7892107892107892,
|
| 1119 |
+
"grad_norm": 0.0947265625,
|
| 1120 |
+
"learning_rate": 2.5919540371325e-05,
|
| 1121 |
+
"loss": 1.292,
|
| 1122 |
+
"step": 790
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.7942057942057942,
|
| 1126 |
+
"grad_norm": 0.0947265625,
|
| 1127 |
+
"learning_rate": 2.4758509110427575e-05,
|
| 1128 |
+
"loss": 1.3089,
|
| 1129 |
+
"step": 795
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.7992007992007992,
|
| 1133 |
+
"grad_norm": 0.09375,
|
| 1134 |
+
"learning_rate": 2.362039713653581e-05,
|
| 1135 |
+
"loss": 1.3201,
|
| 1136 |
+
"step": 800
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.8041958041958042,
|
| 1140 |
+
"grad_norm": 0.09326171875,
|
| 1141 |
+
"learning_rate": 2.2505551129582047e-05,
|
| 1142 |
+
"loss": 1.2985,
|
| 1143 |
+
"step": 805
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.8091908091908092,
|
| 1147 |
+
"grad_norm": 0.095703125,
|
| 1148 |
+
"learning_rate": 2.1414310682459802e-05,
|
| 1149 |
+
"loss": 1.309,
|
| 1150 |
+
"step": 810
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.8141858141858141,
|
| 1154 |
+
"grad_norm": 0.0927734375,
|
| 1155 |
+
"learning_rate": 2.0347008197580374e-05,
|
| 1156 |
+
"loss": 1.3097,
|
| 1157 |
+
"step": 815
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.8191808191808192,
|
| 1161 |
+
"grad_norm": 0.0927734375,
|
| 1162 |
+
"learning_rate": 1.930396878561983e-05,
|
| 1163 |
+
"loss": 1.3215,
|
| 1164 |
+
"step": 820
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.8241758241758241,
|
| 1168 |
+
"grad_norm": 0.09326171875,
|
| 1169 |
+
"learning_rate": 1.8285510166487152e-05,
|
| 1170 |
+
"loss": 1.3002,
|
| 1171 |
+
"step": 825
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.8291708291708292,
|
| 1175 |
+
"grad_norm": 0.0927734375,
|
| 1176 |
+
"learning_rate": 1.7291942572543807e-05,
|
| 1177 |
+
"loss": 1.3233,
|
| 1178 |
+
"step": 830
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.8341658341658341,
|
| 1182 |
+
"grad_norm": 0.09521484375,
|
| 1183 |
+
"learning_rate": 1.632356865410384e-05,
|
| 1184 |
+
"loss": 1.3163,
|
| 1185 |
+
"step": 835
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.8391608391608392,
|
| 1189 |
+
"grad_norm": 0.0947265625,
|
| 1190 |
+
"learning_rate": 1.538068338724361e-05,
|
| 1191 |
+
"loss": 1.2986,
|
| 1192 |
+
"step": 840
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.8441558441558441,
|
| 1196 |
+
"grad_norm": 0.09375,
|
| 1197 |
+
"learning_rate": 1.4463573983949341e-05,
|
| 1198 |
+
"loss": 1.319,
|
| 1199 |
+
"step": 845
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.8491508491508492,
|
| 1203 |
+
"grad_norm": 0.09326171875,
|
| 1204 |
+
"learning_rate": 1.3572519804629536e-05,
|
| 1205 |
+
"loss": 1.3119,
|
| 1206 |
+
"step": 850
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.8541458541458542,
|
| 1210 |
+
"grad_norm": 0.09326171875,
|
| 1211 |
+
"learning_rate": 1.2707792273019048e-05,
|
| 1212 |
+
"loss": 1.3047,
|
| 1213 |
+
"step": 855
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.8591408591408591,
|
| 1217 |
+
"grad_norm": 0.09375,
|
| 1218 |
+
"learning_rate": 1.1869654793500784e-05,
|
| 1219 |
+
"loss": 1.3001,
|
| 1220 |
+
"step": 860
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.8641358641358642,
|
| 1224 |
+
"grad_norm": 0.09228515625,
|
| 1225 |
+
"learning_rate": 1.1058362670870249e-05,
|
| 1226 |
+
"loss": 1.3192,
|
| 1227 |
+
"step": 865
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.8691308691308691,
|
| 1231 |
+
"grad_norm": 0.0947265625,
|
| 1232 |
+
"learning_rate": 1.0274163032567163e-05,
|
| 1233 |
+
"loss": 1.3049,
|
| 1234 |
+
"step": 870
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.8741258741258742,
|
| 1238 |
+
"grad_norm": 0.09130859375,
|
| 1239 |
+
"learning_rate": 9.517294753398064e-06,
|
| 1240 |
+
"loss": 1.3049,
|
| 1241 |
+
"step": 875
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.8791208791208791,
|
| 1245 |
+
"grad_norm": 0.091796875,
|
| 1246 |
+
"learning_rate": 8.787988382772705e-06,
|
| 1247 |
+
"loss": 1.3111,
|
| 1248 |
+
"step": 880
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.8841158841158842,
|
| 1252 |
+
"grad_norm": 0.09375,
|
| 1253 |
+
"learning_rate": 8.086466074476563e-06,
|
| 1254 |
+
"loss": 1.2984,
|
| 1255 |
+
"step": 885
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.8891108891108891,
|
| 1259 |
+
"grad_norm": 0.09228515625,
|
| 1260 |
+
"learning_rate": 7.412941519000527e-06,
|
| 1261 |
+
"loss": 1.3023,
|
| 1262 |
+
"step": 890
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.8941058941058941,
|
| 1266 |
+
"grad_norm": 0.09326171875,
|
| 1267 |
+
"learning_rate": 6.767619878448783e-06,
|
| 1268 |
+
"loss": 1.3173,
|
| 1269 |
+
"step": 895
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.8991008991008991,
|
| 1273 |
+
"grad_norm": 0.0927734375,
|
| 1274 |
+
"learning_rate": 6.1506977240444074e-06,
|
| 1275 |
+
"loss": 1.3053,
|
| 1276 |
+
"step": 900
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.9040959040959041,
|
| 1280 |
+
"grad_norm": 0.09375,
|
| 1281 |
+
"learning_rate": 5.562362976251901e-06,
|
| 1282 |
+
"loss": 1.312,
|
| 1283 |
+
"step": 905
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.9090909090909091,
|
| 1287 |
+
"grad_norm": 0.09326171875,
|
| 1288 |
+
"learning_rate": 5.002794847534764e-06,
|
| 1289 |
+
"loss": 1.2995,
|
| 1290 |
+
"step": 910
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.9140859140859141,
|
| 1294 |
+
"grad_norm": 0.0947265625,
|
| 1295 |
+
"learning_rate": 4.4721637877656375e-06,
|
| 1296 |
+
"loss": 1.332,
|
| 1297 |
+
"step": 915
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.919080919080919,
|
| 1301 |
+
"grad_norm": 0.091796875,
|
| 1302 |
+
"learning_rate": 3.970631432305694e-06,
|
| 1303 |
+
"loss": 1.3225,
|
| 1304 |
+
"step": 920
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.9240759240759241,
|
| 1308 |
+
"grad_norm": 0.0927734375,
|
| 1309 |
+
"learning_rate": 3.4983505527688586e-06,
|
| 1310 |
+
"loss": 1.2955,
|
| 1311 |
+
"step": 925
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.929070929070929,
|
| 1315 |
+
"grad_norm": 0.095703125,
|
| 1316 |
+
"learning_rate": 3.0554650104861136e-06,
|
| 1317 |
+
"loss": 1.3007,
|
| 1318 |
+
"step": 930
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.9340659340659341,
|
| 1322 |
+
"grad_norm": 0.09326171875,
|
| 1323 |
+
"learning_rate": 2.6421097126839712e-06,
|
| 1324 |
+
"loss": 1.3259,
|
| 1325 |
+
"step": 935
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.939060939060939,
|
| 1329 |
+
"grad_norm": 0.095703125,
|
| 1330 |
+
"learning_rate": 2.2584105713904125e-06,
|
| 1331 |
+
"loss": 1.3091,
|
| 1332 |
+
"step": 940
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.9440559440559441,
|
| 1336 |
+
"grad_norm": 0.09423828125,
|
| 1337 |
+
"learning_rate": 1.904484465080847e-06,
|
| 1338 |
+
"loss": 1.3195,
|
| 1339 |
+
"step": 945
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.949050949050949,
|
| 1343 |
+
"grad_norm": 0.09423828125,
|
| 1344 |
+
"learning_rate": 1.580439203075812e-06,
|
| 1345 |
+
"loss": 1.3043,
|
| 1346 |
+
"step": 950
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.954045954045954,
|
| 1350 |
+
"grad_norm": 0.09375,
|
| 1351 |
+
"learning_rate": 1.2863734927012095e-06,
|
| 1352 |
+
"loss": 1.3167,
|
| 1353 |
+
"step": 955
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.9590409590409591,
|
| 1357 |
+
"grad_norm": 0.09326171875,
|
| 1358 |
+
"learning_rate": 1.0223769092211012e-06,
|
| 1359 |
+
"loss": 1.3064,
|
| 1360 |
+
"step": 960
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.964035964035964,
|
| 1364 |
+
"grad_norm": 0.091796875,
|
| 1365 |
+
"learning_rate": 7.885298685522235e-07,
|
| 1366 |
+
"loss": 1.3197,
|
| 1367 |
+
"step": 965
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.9690309690309691,
|
| 1371 |
+
"grad_norm": 0.09228515625,
|
| 1372 |
+
"learning_rate": 5.849036027684606e-07,
|
| 1373 |
+
"loss": 1.3074,
|
| 1374 |
+
"step": 970
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.974025974025974,
|
| 1378 |
+
"grad_norm": 0.09912109375,
|
| 1379 |
+
"learning_rate": 4.115601384029666e-07,
|
| 1380 |
+
"loss": 1.3183,
|
| 1381 |
+
"step": 975
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.9790209790209791,
|
| 1385 |
+
"grad_norm": 0.09326171875,
|
| 1386 |
+
"learning_rate": 2.685522775541904e-07,
|
| 1387 |
+
"loss": 1.3165,
|
| 1388 |
+
"step": 980
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.984015984015984,
|
| 1392 |
+
"grad_norm": 0.09130859375,
|
| 1393 |
+
"learning_rate": 1.5592358180189782e-07,
|
| 1394 |
+
"loss": 1.3114,
|
| 1395 |
+
"step": 985
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.989010989010989,
|
| 1399 |
+
"grad_norm": 0.09326171875,
|
| 1400 |
+
"learning_rate": 7.370835893788508e-08,
|
| 1401 |
+
"loss": 1.2917,
|
| 1402 |
+
"step": 990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.994005994005994,
|
| 1406 |
+
"grad_norm": 0.095703125,
|
| 1407 |
+
"learning_rate": 2.193165251545004e-08,
|
| 1408 |
+
"loss": 1.3122,
|
| 1409 |
+
"step": 995
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.999000999000999,
|
| 1413 |
+
"grad_norm": 0.09326171875,
|
| 1414 |
+
"learning_rate": 6.092342209607083e-10,
|
| 1415 |
+
"loss": 1.3184,
|
| 1416 |
+
"step": 1000
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.0,
|
| 1420 |
+
"eval_loss": 1.3126407861709595,
|
| 1421 |
+
"eval_runtime": 1177.4856,
|
| 1422 |
+
"eval_samples_per_second": 12.041,
|
| 1423 |
+
"eval_steps_per_second": 12.041,
|
| 1424 |
+
"step": 1001
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.0,
|
| 1428 |
+
"step": 1001,
|
| 1429 |
+
"total_flos": 1.709524252278915e+18,
|
| 1430 |
+
"train_loss": 0.07992646839473393,
|
| 1431 |
+
"train_runtime": 3675.5245,
|
| 1432 |
+
"train_samples_per_second": 34.86,
|
| 1433 |
+
"train_steps_per_second": 0.272
|
| 1434 |
+
}
|
| 1435 |
+
],
|
| 1436 |
+
"logging_steps": 5,
|
| 1437 |
+
"max_steps": 1001,
|
| 1438 |
+
"num_input_tokens_seen": 0,
|
| 1439 |
+
"num_train_epochs": 1,
|
| 1440 |
+
"save_steps": 20,
|
| 1441 |
+
"total_flos": 1.709524252278915e+18,
|
| 1442 |
+
"train_batch_size": 1,
|
| 1443 |
+
"trial_name": null,
|
| 1444 |
+
"trial_params": null
|
| 1445 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9c395792916924ca30fb54300ada798edb11ba992b2e01331927487d9f2c6ce3
|
| 3 |
+
size 5112
|