Upload folder using huggingface_hub
Browse files- 29_128_e3_3e-5/README.md +63 -0
- 29_128_e3_3e-5/adapter_config.json +39 -0
- 29_128_e3_3e-5/adapter_model.safetensors +3 -0
- 29_128_e3_3e-5/all_results.json +9 -0
- 29_128_e3_3e-5/config.json +32 -0
- 29_128_e3_3e-5/merges.txt +0 -0
- 29_128_e3_3e-5/special_tokens_map.json +45 -0
- 29_128_e3_3e-5/tokenizer.json +0 -0
- 29_128_e3_3e-5/tokenizer_config.json +188 -0
- 29_128_e3_3e-5/train_results.json +9 -0
- 29_128_e3_3e-5/trainer_state.json +1723 -0
- 29_128_e3_3e-5/training_args.bin +3 -0
- 29_128_e3_3e-5/vocab.json +0 -0
29_128_e3_3e-5/README.md
ADDED
|
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
license: apache-2.0
|
| 4 |
+
base_model: ibm-granite/granite-3.3-8b-base
|
| 5 |
+
tags:
|
| 6 |
+
- alignment-handbook
|
| 7 |
+
- generated_from_trainer
|
| 8 |
+
datasets:
|
| 9 |
+
- data/knowledge_lora_training_data_2000
|
| 10 |
+
model-index:
|
| 11 |
+
- name: 29_128_e3_3e-5
|
| 12 |
+
results: []
|
| 13 |
+
---
|
| 14 |
+
|
| 15 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 16 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 17 |
+
|
| 18 |
+
# 29_128_e3_3e-5
|
| 19 |
+
|
| 20 |
+
This model is a fine-tuned version of [ibm-granite/granite-3.3-8b-base](https://huggingface.co/ibm-granite/granite-3.3-8b-base) on the data/knowledge_lora_training_data_2000 dataset.
|
| 21 |
+
|
| 22 |
+
## Model description
|
| 23 |
+
|
| 24 |
+
More information needed
|
| 25 |
+
|
| 26 |
+
## Intended uses & limitations
|
| 27 |
+
|
| 28 |
+
More information needed
|
| 29 |
+
|
| 30 |
+
## Training and evaluation data
|
| 31 |
+
|
| 32 |
+
More information needed
|
| 33 |
+
|
| 34 |
+
## Training procedure
|
| 35 |
+
|
| 36 |
+
### Training hyperparameters
|
| 37 |
+
|
| 38 |
+
The following hyperparameters were used during training:
|
| 39 |
+
- learning_rate: 3e-05
|
| 40 |
+
- train_batch_size: 2
|
| 41 |
+
- eval_batch_size: 8
|
| 42 |
+
- seed: 42
|
| 43 |
+
- distributed_type: multi-GPU
|
| 44 |
+
- num_devices: 8
|
| 45 |
+
- gradient_accumulation_steps: 2
|
| 46 |
+
- total_train_batch_size: 32
|
| 47 |
+
- total_eval_batch_size: 64
|
| 48 |
+
- optimizer: Use adamw_torch with betas=(0.9,0.95) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 49 |
+
- lr_scheduler_type: cosine
|
| 50 |
+
- lr_scheduler_warmup_ratio: 0.05
|
| 51 |
+
- num_epochs: 3.0
|
| 52 |
+
|
| 53 |
+
### Training results
|
| 54 |
+
|
| 55 |
+
|
| 56 |
+
|
| 57 |
+
### Framework versions
|
| 58 |
+
|
| 59 |
+
- PEFT 0.15.2
|
| 60 |
+
- Transformers 4.52.4
|
| 61 |
+
- Pytorch 2.7.0+cu126
|
| 62 |
+
- Datasets 3.6.0
|
| 63 |
+
- Tokenizers 0.21.2
|
29_128_e3_3e-5/adapter_config.json
ADDED
|
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "ibm-granite/granite-3.3-8b-base",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"corda_config": null,
|
| 7 |
+
"eva_config": null,
|
| 8 |
+
"exclude_modules": null,
|
| 9 |
+
"fan_in_fan_out": false,
|
| 10 |
+
"inference_mode": true,
|
| 11 |
+
"init_lora_weights": true,
|
| 12 |
+
"layer_replication": null,
|
| 13 |
+
"layers_pattern": null,
|
| 14 |
+
"layers_to_transform": null,
|
| 15 |
+
"loftq_config": {},
|
| 16 |
+
"lora_alpha": 256,
|
| 17 |
+
"lora_bias": false,
|
| 18 |
+
"lora_dropout": 0.05,
|
| 19 |
+
"megatron_config": null,
|
| 20 |
+
"megatron_core": "megatron.core",
|
| 21 |
+
"modules_to_save": null,
|
| 22 |
+
"peft_type": "LORA",
|
| 23 |
+
"r": 128,
|
| 24 |
+
"rank_pattern": {},
|
| 25 |
+
"revision": null,
|
| 26 |
+
"target_modules": [
|
| 27 |
+
"q_proj",
|
| 28 |
+
"o_proj",
|
| 29 |
+
"v_proj",
|
| 30 |
+
"gate_proj",
|
| 31 |
+
"up_proj",
|
| 32 |
+
"down_proj",
|
| 33 |
+
"k_proj"
|
| 34 |
+
],
|
| 35 |
+
"task_type": "CAUSAL_LM",
|
| 36 |
+
"trainable_token_indices": null,
|
| 37 |
+
"use_dora": false,
|
| 38 |
+
"use_rslora": false
|
| 39 |
+
}
|
29_128_e3_3e-5/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8da9d5bf0be63ed20777c8fa09b03bf962da22a3f40572b4b455765be70603ec
|
| 3 |
+
size 791751704
|
29_128_e3_3e-5/all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 1.64227002403704e+18,
|
| 4 |
+
"train_loss": 0.565493812263012,
|
| 5 |
+
"train_runtime": 754.0979,
|
| 6 |
+
"train_samples": 12789,
|
| 7 |
+
"train_samples_per_second": 50.878,
|
| 8 |
+
"train_steps_per_second": 1.591
|
| 9 |
+
}
|
29_128_e3_3e-5/config.json
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"GraniteForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0,
|
| 7 |
+
"attention_multiplier": 0.0078125,
|
| 8 |
+
"bos_token_id": 0,
|
| 9 |
+
"embedding_multiplier": 12.0,
|
| 10 |
+
"eos_token_id": 0,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 4096,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 12800,
|
| 15 |
+
"logits_scaling": 16.0,
|
| 16 |
+
"max_position_embeddings": 131072,
|
| 17 |
+
"mlp_bias": false,
|
| 18 |
+
"model_type": "granite",
|
| 19 |
+
"num_attention_heads": 32,
|
| 20 |
+
"num_hidden_layers": 40,
|
| 21 |
+
"num_key_value_heads": 8,
|
| 22 |
+
"pad_token_id": 0,
|
| 23 |
+
"residual_multiplier": 0.22,
|
| 24 |
+
"rms_norm_eps": 1e-05,
|
| 25 |
+
"rope_scaling": null,
|
| 26 |
+
"rope_theta": 10000000.0,
|
| 27 |
+
"tie_word_embeddings": true,
|
| 28 |
+
"torch_dtype": "bfloat16",
|
| 29 |
+
"transformers_version": "4.52.4",
|
| 30 |
+
"use_cache": true,
|
| 31 |
+
"vocab_size": 49152
|
| 32 |
+
}
|
29_128_e3_3e-5/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
29_128_e3_3e-5/special_tokens_map.json
ADDED
|
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<|endoftext|>",
|
| 4 |
+
"<fim_prefix>",
|
| 5 |
+
"<fim_middle>",
|
| 6 |
+
"<fim_suffix>",
|
| 7 |
+
"<fim_pad>",
|
| 8 |
+
"<filename>",
|
| 9 |
+
"<gh_stars>",
|
| 10 |
+
"<issue_start>",
|
| 11 |
+
"<issue_comment>",
|
| 12 |
+
"<issue_closed>",
|
| 13 |
+
"<jupyter_start>",
|
| 14 |
+
"<jupyter_text>",
|
| 15 |
+
"<jupyter_code>",
|
| 16 |
+
"<jupyter_output>",
|
| 17 |
+
"<empty_output>",
|
| 18 |
+
"<commit_before>",
|
| 19 |
+
"<commit_msg>",
|
| 20 |
+
"<commit_after>",
|
| 21 |
+
"<reponame>"
|
| 22 |
+
],
|
| 23 |
+
"bos_token": {
|
| 24 |
+
"content": "<|endoftext|>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"eos_token": {
|
| 31 |
+
"content": "<|endoftext|>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"pad_token": "<reponame>",
|
| 38 |
+
"unk_token": {
|
| 39 |
+
"content": "<|endoftext|>",
|
| 40 |
+
"lstrip": false,
|
| 41 |
+
"normalized": false,
|
| 42 |
+
"rstrip": false,
|
| 43 |
+
"single_word": false
|
| 44 |
+
}
|
| 45 |
+
}
|
29_128_e3_3e-5/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
29_128_e3_3e-5/tokenizer_config.json
ADDED
|
@@ -0,0 +1,188 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"0": {
|
| 5 |
+
"content": "<|endoftext|>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": false,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
+
"1": {
|
| 13 |
+
"content": "<fim_prefix>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": false,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false,
|
| 18 |
+
"special": true
|
| 19 |
+
},
|
| 20 |
+
"2": {
|
| 21 |
+
"content": "<fim_middle>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": false,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false,
|
| 26 |
+
"special": true
|
| 27 |
+
},
|
| 28 |
+
"3": {
|
| 29 |
+
"content": "<fim_suffix>",
|
| 30 |
+
"lstrip": false,
|
| 31 |
+
"normalized": false,
|
| 32 |
+
"rstrip": false,
|
| 33 |
+
"single_word": false,
|
| 34 |
+
"special": true
|
| 35 |
+
},
|
| 36 |
+
"4": {
|
| 37 |
+
"content": "<fim_pad>",
|
| 38 |
+
"lstrip": false,
|
| 39 |
+
"normalized": false,
|
| 40 |
+
"rstrip": false,
|
| 41 |
+
"single_word": false,
|
| 42 |
+
"special": true
|
| 43 |
+
},
|
| 44 |
+
"5": {
|
| 45 |
+
"content": "<filename>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false,
|
| 50 |
+
"special": true
|
| 51 |
+
},
|
| 52 |
+
"6": {
|
| 53 |
+
"content": "<gh_stars>",
|
| 54 |
+
"lstrip": false,
|
| 55 |
+
"normalized": false,
|
| 56 |
+
"rstrip": false,
|
| 57 |
+
"single_word": false,
|
| 58 |
+
"special": true
|
| 59 |
+
},
|
| 60 |
+
"7": {
|
| 61 |
+
"content": "<issue_start>",
|
| 62 |
+
"lstrip": false,
|
| 63 |
+
"normalized": false,
|
| 64 |
+
"rstrip": false,
|
| 65 |
+
"single_word": false,
|
| 66 |
+
"special": true
|
| 67 |
+
},
|
| 68 |
+
"8": {
|
| 69 |
+
"content": "<issue_comment>",
|
| 70 |
+
"lstrip": false,
|
| 71 |
+
"normalized": false,
|
| 72 |
+
"rstrip": false,
|
| 73 |
+
"single_word": false,
|
| 74 |
+
"special": true
|
| 75 |
+
},
|
| 76 |
+
"9": {
|
| 77 |
+
"content": "<issue_closed>",
|
| 78 |
+
"lstrip": false,
|
| 79 |
+
"normalized": false,
|
| 80 |
+
"rstrip": false,
|
| 81 |
+
"single_word": false,
|
| 82 |
+
"special": true
|
| 83 |
+
},
|
| 84 |
+
"10": {
|
| 85 |
+
"content": "<jupyter_start>",
|
| 86 |
+
"lstrip": false,
|
| 87 |
+
"normalized": false,
|
| 88 |
+
"rstrip": false,
|
| 89 |
+
"single_word": false,
|
| 90 |
+
"special": true
|
| 91 |
+
},
|
| 92 |
+
"11": {
|
| 93 |
+
"content": "<jupyter_text>",
|
| 94 |
+
"lstrip": false,
|
| 95 |
+
"normalized": false,
|
| 96 |
+
"rstrip": false,
|
| 97 |
+
"single_word": false,
|
| 98 |
+
"special": true
|
| 99 |
+
},
|
| 100 |
+
"12": {
|
| 101 |
+
"content": "<jupyter_code>",
|
| 102 |
+
"lstrip": false,
|
| 103 |
+
"normalized": false,
|
| 104 |
+
"rstrip": false,
|
| 105 |
+
"single_word": false,
|
| 106 |
+
"special": true
|
| 107 |
+
},
|
| 108 |
+
"13": {
|
| 109 |
+
"content": "<jupyter_output>",
|
| 110 |
+
"lstrip": false,
|
| 111 |
+
"normalized": false,
|
| 112 |
+
"rstrip": false,
|
| 113 |
+
"single_word": false,
|
| 114 |
+
"special": true
|
| 115 |
+
},
|
| 116 |
+
"14": {
|
| 117 |
+
"content": "<empty_output>",
|
| 118 |
+
"lstrip": false,
|
| 119 |
+
"normalized": false,
|
| 120 |
+
"rstrip": false,
|
| 121 |
+
"single_word": false,
|
| 122 |
+
"special": true
|
| 123 |
+
},
|
| 124 |
+
"15": {
|
| 125 |
+
"content": "<commit_before>",
|
| 126 |
+
"lstrip": false,
|
| 127 |
+
"normalized": false,
|
| 128 |
+
"rstrip": false,
|
| 129 |
+
"single_word": false,
|
| 130 |
+
"special": true
|
| 131 |
+
},
|
| 132 |
+
"16": {
|
| 133 |
+
"content": "<commit_msg>",
|
| 134 |
+
"lstrip": false,
|
| 135 |
+
"normalized": false,
|
| 136 |
+
"rstrip": false,
|
| 137 |
+
"single_word": false,
|
| 138 |
+
"special": true
|
| 139 |
+
},
|
| 140 |
+
"17": {
|
| 141 |
+
"content": "<commit_after>",
|
| 142 |
+
"lstrip": false,
|
| 143 |
+
"normalized": false,
|
| 144 |
+
"rstrip": false,
|
| 145 |
+
"single_word": false,
|
| 146 |
+
"special": true
|
| 147 |
+
},
|
| 148 |
+
"18": {
|
| 149 |
+
"content": "<reponame>",
|
| 150 |
+
"lstrip": false,
|
| 151 |
+
"normalized": false,
|
| 152 |
+
"rstrip": false,
|
| 153 |
+
"single_word": false,
|
| 154 |
+
"special": true
|
| 155 |
+
}
|
| 156 |
+
},
|
| 157 |
+
"additional_special_tokens": [
|
| 158 |
+
"<|endoftext|>",
|
| 159 |
+
"<fim_prefix>",
|
| 160 |
+
"<fim_middle>",
|
| 161 |
+
"<fim_suffix>",
|
| 162 |
+
"<fim_pad>",
|
| 163 |
+
"<filename>",
|
| 164 |
+
"<gh_stars>",
|
| 165 |
+
"<issue_start>",
|
| 166 |
+
"<issue_comment>",
|
| 167 |
+
"<issue_closed>",
|
| 168 |
+
"<jupyter_start>",
|
| 169 |
+
"<jupyter_text>",
|
| 170 |
+
"<jupyter_code>",
|
| 171 |
+
"<jupyter_output>",
|
| 172 |
+
"<empty_output>",
|
| 173 |
+
"<commit_before>",
|
| 174 |
+
"<commit_msg>",
|
| 175 |
+
"<commit_after>",
|
| 176 |
+
"<reponame>"
|
| 177 |
+
],
|
| 178 |
+
"bos_token": "<|endoftext|>",
|
| 179 |
+
"clean_up_tokenization_spaces": true,
|
| 180 |
+
"eos_token": "<|endoftext|>",
|
| 181 |
+
"extra_special_tokens": {},
|
| 182 |
+
"model_max_length": 8192,
|
| 183 |
+
"pad_token": "<reponame>",
|
| 184 |
+
"padding_side": "left",
|
| 185 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 186 |
+
"unk_token": "<|endoftext|>",
|
| 187 |
+
"vocab_size": 49152
|
| 188 |
+
}
|
29_128_e3_3e-5/train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 1.64227002403704e+18,
|
| 4 |
+
"train_loss": 0.565493812263012,
|
| 5 |
+
"train_runtime": 754.0979,
|
| 6 |
+
"train_samples": 12789,
|
| 7 |
+
"train_samples_per_second": 50.878,
|
| 8 |
+
"train_steps_per_second": 1.591
|
| 9 |
+
}
|
29_128_e3_3e-5/trainer_state.json
ADDED
|
@@ -0,0 +1,1723 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1200,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0125,
|
| 14 |
+
"grad_norm": 1.096501350402832,
|
| 15 |
+
"learning_rate": 2e-06,
|
| 16 |
+
"loss": 1.3059,
|
| 17 |
+
"step": 5
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.025,
|
| 21 |
+
"grad_norm": 0.8720690608024597,
|
| 22 |
+
"learning_rate": 4.5e-06,
|
| 23 |
+
"loss": 1.3236,
|
| 24 |
+
"step": 10
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0375,
|
| 28 |
+
"grad_norm": 0.6325503587722778,
|
| 29 |
+
"learning_rate": 7e-06,
|
| 30 |
+
"loss": 1.3225,
|
| 31 |
+
"step": 15
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.05,
|
| 35 |
+
"grad_norm": 0.6217526793479919,
|
| 36 |
+
"learning_rate": 9.5e-06,
|
| 37 |
+
"loss": 1.2885,
|
| 38 |
+
"step": 20
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.0625,
|
| 42 |
+
"grad_norm": 0.47595125436782837,
|
| 43 |
+
"learning_rate": 1.2e-05,
|
| 44 |
+
"loss": 1.316,
|
| 45 |
+
"step": 25
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.075,
|
| 49 |
+
"grad_norm": 0.5025820136070251,
|
| 50 |
+
"learning_rate": 1.45e-05,
|
| 51 |
+
"loss": 1.2431,
|
| 52 |
+
"step": 30
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.0875,
|
| 56 |
+
"grad_norm": 0.5048433542251587,
|
| 57 |
+
"learning_rate": 1.7e-05,
|
| 58 |
+
"loss": 1.2902,
|
| 59 |
+
"step": 35
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.1,
|
| 63 |
+
"grad_norm": 0.5279613137245178,
|
| 64 |
+
"learning_rate": 1.95e-05,
|
| 65 |
+
"loss": 1.2261,
|
| 66 |
+
"step": 40
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.1125,
|
| 70 |
+
"grad_norm": 1.3911240100860596,
|
| 71 |
+
"learning_rate": 2.2e-05,
|
| 72 |
+
"loss": 1.2389,
|
| 73 |
+
"step": 45
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.125,
|
| 77 |
+
"grad_norm": 0.45133835077285767,
|
| 78 |
+
"learning_rate": 2.45e-05,
|
| 79 |
+
"loss": 1.1965,
|
| 80 |
+
"step": 50
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.1375,
|
| 84 |
+
"grad_norm": 0.4962674677371979,
|
| 85 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 86 |
+
"loss": 1.2192,
|
| 87 |
+
"step": 55
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.15,
|
| 91 |
+
"grad_norm": 0.48562711477279663,
|
| 92 |
+
"learning_rate": 2.95e-05,
|
| 93 |
+
"loss": 1.1988,
|
| 94 |
+
"step": 60
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.1625,
|
| 98 |
+
"grad_norm": 0.4591135084629059,
|
| 99 |
+
"learning_rate": 2.9999088688415145e-05,
|
| 100 |
+
"loss": 1.1593,
|
| 101 |
+
"step": 65
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.175,
|
| 105 |
+
"grad_norm": 0.4925994575023651,
|
| 106 |
+
"learning_rate": 2.9995386674880524e-05,
|
| 107 |
+
"loss": 1.1978,
|
| 108 |
+
"step": 70
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.1875,
|
| 112 |
+
"grad_norm": 0.5008227229118347,
|
| 113 |
+
"learning_rate": 2.9988837704729014e-05,
|
| 114 |
+
"loss": 1.1277,
|
| 115 |
+
"step": 75
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.2,
|
| 119 |
+
"grad_norm": 0.5128725171089172,
|
| 120 |
+
"learning_rate": 2.9979443021318607e-05,
|
| 121 |
+
"loss": 1.1022,
|
| 122 |
+
"step": 80
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.2125,
|
| 126 |
+
"grad_norm": 0.49049896001815796,
|
| 127 |
+
"learning_rate": 2.9967204408281618e-05,
|
| 128 |
+
"loss": 1.1422,
|
| 129 |
+
"step": 85
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.225,
|
| 133 |
+
"grad_norm": 0.647089958190918,
|
| 134 |
+
"learning_rate": 2.9952124189186028e-05,
|
| 135 |
+
"loss": 1.1323,
|
| 136 |
+
"step": 90
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.2375,
|
| 140 |
+
"grad_norm": 0.5251462459564209,
|
| 141 |
+
"learning_rate": 2.9934205227094347e-05,
|
| 142 |
+
"loss": 1.1207,
|
| 143 |
+
"step": 95
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.25,
|
| 147 |
+
"grad_norm": 0.5728532075881958,
|
| 148 |
+
"learning_rate": 2.9913450924020067e-05,
|
| 149 |
+
"loss": 1.0544,
|
| 150 |
+
"step": 100
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.2625,
|
| 154 |
+
"grad_norm": 0.5652995109558105,
|
| 155 |
+
"learning_rate": 2.9889865220281747e-05,
|
| 156 |
+
"loss": 1.1441,
|
| 157 |
+
"step": 105
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.275,
|
| 161 |
+
"grad_norm": 0.6100318431854248,
|
| 162 |
+
"learning_rate": 2.9863452593754943e-05,
|
| 163 |
+
"loss": 1.0696,
|
| 164 |
+
"step": 110
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.2875,
|
| 168 |
+
"grad_norm": 0.5077236890792847,
|
| 169 |
+
"learning_rate": 2.9834218059022027e-05,
|
| 170 |
+
"loss": 1.074,
|
| 171 |
+
"step": 115
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.3,
|
| 175 |
+
"grad_norm": 0.5974183678627014,
|
| 176 |
+
"learning_rate": 2.9802167166420182e-05,
|
| 177 |
+
"loss": 1.0579,
|
| 178 |
+
"step": 120
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.3125,
|
| 182 |
+
"grad_norm": 0.5726377367973328,
|
| 183 |
+
"learning_rate": 2.976730600098761e-05,
|
| 184 |
+
"loss": 1.0062,
|
| 185 |
+
"step": 125
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.325,
|
| 189 |
+
"grad_norm": 0.5457852482795715,
|
| 190 |
+
"learning_rate": 2.972964118130826e-05,
|
| 191 |
+
"loss": 1.0449,
|
| 192 |
+
"step": 130
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.3375,
|
| 196 |
+
"grad_norm": 0.633044958114624,
|
| 197 |
+
"learning_rate": 2.9689179858255252e-05,
|
| 198 |
+
"loss": 1.0299,
|
| 199 |
+
"step": 135
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.35,
|
| 203 |
+
"grad_norm": 0.7141392827033997,
|
| 204 |
+
"learning_rate": 2.964592971363327e-05,
|
| 205 |
+
"loss": 1.02,
|
| 206 |
+
"step": 140
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.3625,
|
| 210 |
+
"grad_norm": 0.675027072429657,
|
| 211 |
+
"learning_rate": 2.9599898958720088e-05,
|
| 212 |
+
"loss": 1.0242,
|
| 213 |
+
"step": 145
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.375,
|
| 217 |
+
"grad_norm": 0.6282708644866943,
|
| 218 |
+
"learning_rate": 2.955109633270764e-05,
|
| 219 |
+
"loss": 0.9516,
|
| 220 |
+
"step": 150
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.3875,
|
| 224 |
+
"grad_norm": 0.784489631652832,
|
| 225 |
+
"learning_rate": 2.9499531101042834e-05,
|
| 226 |
+
"loss": 0.9881,
|
| 227 |
+
"step": 155
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.4,
|
| 231 |
+
"grad_norm": 0.6914710998535156,
|
| 232 |
+
"learning_rate": 2.9445213053668444e-05,
|
| 233 |
+
"loss": 0.9733,
|
| 234 |
+
"step": 160
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.4125,
|
| 238 |
+
"grad_norm": 0.687872052192688,
|
| 239 |
+
"learning_rate": 2.938815250316445e-05,
|
| 240 |
+
"loss": 0.8953,
|
| 241 |
+
"step": 165
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.425,
|
| 245 |
+
"grad_norm": 0.83106929063797,
|
| 246 |
+
"learning_rate": 2.932836028279013e-05,
|
| 247 |
+
"loss": 0.9592,
|
| 248 |
+
"step": 170
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.4375,
|
| 252 |
+
"grad_norm": 0.6633298993110657,
|
| 253 |
+
"learning_rate": 2.9265847744427305e-05,
|
| 254 |
+
"loss": 0.9416,
|
| 255 |
+
"step": 175
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.45,
|
| 259 |
+
"grad_norm": 0.7028163075447083,
|
| 260 |
+
"learning_rate": 2.9200626756425132e-05,
|
| 261 |
+
"loss": 0.9324,
|
| 262 |
+
"step": 180
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.4625,
|
| 266 |
+
"grad_norm": 0.7386612296104431,
|
| 267 |
+
"learning_rate": 2.9132709701346815e-05,
|
| 268 |
+
"loss": 0.9592,
|
| 269 |
+
"step": 185
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.475,
|
| 273 |
+
"grad_norm": 0.6585894823074341,
|
| 274 |
+
"learning_rate": 2.9062109473618732e-05,
|
| 275 |
+
"loss": 0.955,
|
| 276 |
+
"step": 190
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.4875,
|
| 280 |
+
"grad_norm": 0.7104294896125793,
|
| 281 |
+
"learning_rate": 2.898883947708233e-05,
|
| 282 |
+
"loss": 0.9462,
|
| 283 |
+
"step": 195
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.5,
|
| 287 |
+
"grad_norm": 0.7890173196792603,
|
| 288 |
+
"learning_rate": 2.8912913622449335e-05,
|
| 289 |
+
"loss": 0.9005,
|
| 290 |
+
"step": 200
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.5125,
|
| 294 |
+
"grad_norm": 0.7631654143333435,
|
| 295 |
+
"learning_rate": 2.883434632466077e-05,
|
| 296 |
+
"loss": 0.9035,
|
| 297 |
+
"step": 205
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.525,
|
| 301 |
+
"grad_norm": 0.7017130851745605,
|
| 302 |
+
"learning_rate": 2.875315250015014e-05,
|
| 303 |
+
"loss": 0.8929,
|
| 304 |
+
"step": 210
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.5375,
|
| 308 |
+
"grad_norm": 0.8477493524551392,
|
| 309 |
+
"learning_rate": 2.8669347564011493e-05,
|
| 310 |
+
"loss": 0.8661,
|
| 311 |
+
"step": 215
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.55,
|
| 315 |
+
"grad_norm": 0.8296931385993958,
|
| 316 |
+
"learning_rate": 2.8582947427072784e-05,
|
| 317 |
+
"loss": 0.859,
|
| 318 |
+
"step": 220
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.5625,
|
| 322 |
+
"grad_norm": 0.7606105804443359,
|
| 323 |
+
"learning_rate": 2.8493968492875104e-05,
|
| 324 |
+
"loss": 0.878,
|
| 325 |
+
"step": 225
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.575,
|
| 329 |
+
"grad_norm": 0.7497696280479431,
|
| 330 |
+
"learning_rate": 2.84024276545584e-05,
|
| 331 |
+
"loss": 0.9029,
|
| 332 |
+
"step": 230
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.5875,
|
| 336 |
+
"grad_norm": 0.8201003670692444,
|
| 337 |
+
"learning_rate": 2.830834229165418e-05,
|
| 338 |
+
"loss": 0.8247,
|
| 339 |
+
"step": 235
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.6,
|
| 343 |
+
"grad_norm": 0.7867859601974487,
|
| 344 |
+
"learning_rate": 2.821173026678595e-05,
|
| 345 |
+
"loss": 0.8574,
|
| 346 |
+
"step": 240
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.6125,
|
| 350 |
+
"grad_norm": 0.8410252332687378,
|
| 351 |
+
"learning_rate": 2.8112609922277886e-05,
|
| 352 |
+
"loss": 0.85,
|
| 353 |
+
"step": 245
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.625,
|
| 357 |
+
"grad_norm": 0.8968583345413208,
|
| 358 |
+
"learning_rate": 2.801100007667243e-05,
|
| 359 |
+
"loss": 0.8596,
|
| 360 |
+
"step": 250
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.6375,
|
| 364 |
+
"grad_norm": 0.8730499744415283,
|
| 365 |
+
"learning_rate": 2.7906920021157508e-05,
|
| 366 |
+
"loss": 0.8076,
|
| 367 |
+
"step": 255
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.65,
|
| 371 |
+
"grad_norm": 0.942690908908844,
|
| 372 |
+
"learning_rate": 2.780038951590397e-05,
|
| 373 |
+
"loss": 0.8235,
|
| 374 |
+
"step": 260
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.6625,
|
| 378 |
+
"grad_norm": 0.9312007427215576,
|
| 379 |
+
"learning_rate": 2.769142878631403e-05,
|
| 380 |
+
"loss": 0.8294,
|
| 381 |
+
"step": 265
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.675,
|
| 385 |
+
"grad_norm": 0.9603184461593628,
|
| 386 |
+
"learning_rate": 2.7580058519181363e-05,
|
| 387 |
+
"loss": 0.7691,
|
| 388 |
+
"step": 270
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.6875,
|
| 392 |
+
"grad_norm": 0.9270436763763428,
|
| 393 |
+
"learning_rate": 2.7466299858763573e-05,
|
| 394 |
+
"loss": 0.7912,
|
| 395 |
+
"step": 275
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.7,
|
| 399 |
+
"grad_norm": 0.9059388637542725,
|
| 400 |
+
"learning_rate": 2.7350174402767887e-05,
|
| 401 |
+
"loss": 0.8148,
|
| 402 |
+
"step": 280
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.7125,
|
| 406 |
+
"grad_norm": 0.8338863849639893,
|
| 407 |
+
"learning_rate": 2.723170419825067e-05,
|
| 408 |
+
"loss": 0.8079,
|
| 409 |
+
"step": 285
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.725,
|
| 413 |
+
"grad_norm": 0.9500786066055298,
|
| 414 |
+
"learning_rate": 2.7110911737431697e-05,
|
| 415 |
+
"loss": 0.7996,
|
| 416 |
+
"step": 290
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.7375,
|
| 420 |
+
"grad_norm": 0.951601505279541,
|
| 421 |
+
"learning_rate": 2.698781995342387e-05,
|
| 422 |
+
"loss": 0.8073,
|
| 423 |
+
"step": 295
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.75,
|
| 427 |
+
"grad_norm": 0.9220702052116394,
|
| 428 |
+
"learning_rate": 2.686245221587924e-05,
|
| 429 |
+
"loss": 0.7983,
|
| 430 |
+
"step": 300
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.7625,
|
| 434 |
+
"grad_norm": 0.9541069865226746,
|
| 435 |
+
"learning_rate": 2.6734832326552148e-05,
|
| 436 |
+
"loss": 0.7416,
|
| 437 |
+
"step": 305
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.775,
|
| 441 |
+
"grad_norm": 0.9559035897254944,
|
| 442 |
+
"learning_rate": 2.6604984514780343e-05,
|
| 443 |
+
"loss": 0.8257,
|
| 444 |
+
"step": 310
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.7875,
|
| 448 |
+
"grad_norm": 1.0292361974716187,
|
| 449 |
+
"learning_rate": 2.6472933432884897e-05,
|
| 450 |
+
"loss": 0.7343,
|
| 451 |
+
"step": 315
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.8,
|
| 455 |
+
"grad_norm": 0.9453174471855164,
|
| 456 |
+
"learning_rate": 2.6338704151489827e-05,
|
| 457 |
+
"loss": 0.7561,
|
| 458 |
+
"step": 320
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.8125,
|
| 462 |
+
"grad_norm": 0.9862915873527527,
|
| 463 |
+
"learning_rate": 2.620232215476231e-05,
|
| 464 |
+
"loss": 0.7356,
|
| 465 |
+
"step": 325
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.825,
|
| 469 |
+
"grad_norm": 0.9760687351226807,
|
| 470 |
+
"learning_rate": 2.6063813335574357e-05,
|
| 471 |
+
"loss": 0.7632,
|
| 472 |
+
"step": 330
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.8375,
|
| 476 |
+
"grad_norm": 0.9775648713111877,
|
| 477 |
+
"learning_rate": 2.5923203990586933e-05,
|
| 478 |
+
"loss": 0.73,
|
| 479 |
+
"step": 335
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.85,
|
| 483 |
+
"grad_norm": 1.1117346286773682,
|
| 484 |
+
"learning_rate": 2.578052081525736e-05,
|
| 485 |
+
"loss": 0.7226,
|
| 486 |
+
"step": 340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.8625,
|
| 490 |
+
"grad_norm": 0.9384631514549255,
|
| 491 |
+
"learning_rate": 2.563579089877106e-05,
|
| 492 |
+
"loss": 0.6991,
|
| 493 |
+
"step": 345
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.875,
|
| 497 |
+
"grad_norm": 1.0402817726135254,
|
| 498 |
+
"learning_rate": 2.548904171889852e-05,
|
| 499 |
+
"loss": 0.712,
|
| 500 |
+
"step": 350
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.8875,
|
| 504 |
+
"grad_norm": 0.9653849005699158,
|
| 505 |
+
"learning_rate": 2.534030113677849e-05,
|
| 506 |
+
"loss": 0.7059,
|
| 507 |
+
"step": 355
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.9,
|
| 511 |
+
"grad_norm": 1.017627239227295,
|
| 512 |
+
"learning_rate": 2.518959739162837e-05,
|
| 513 |
+
"loss": 0.6552,
|
| 514 |
+
"step": 360
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.9125,
|
| 518 |
+
"grad_norm": 0.8845812082290649,
|
| 519 |
+
"learning_rate": 2.5036959095382875e-05,
|
| 520 |
+
"loss": 0.723,
|
| 521 |
+
"step": 365
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.925,
|
| 525 |
+
"grad_norm": 1.0637366771697998,
|
| 526 |
+
"learning_rate": 2.488241522726187e-05,
|
| 527 |
+
"loss": 0.6525,
|
| 528 |
+
"step": 370
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.9375,
|
| 532 |
+
"grad_norm": 0.9697546362876892,
|
| 533 |
+
"learning_rate": 2.4725995128268523e-05,
|
| 534 |
+
"loss": 0.6579,
|
| 535 |
+
"step": 375
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.95,
|
| 539 |
+
"grad_norm": 0.962817907333374,
|
| 540 |
+
"learning_rate": 2.4567728495618763e-05,
|
| 541 |
+
"loss": 0.6756,
|
| 542 |
+
"step": 380
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.9625,
|
| 546 |
+
"grad_norm": 1.0425231456756592,
|
| 547 |
+
"learning_rate": 2.4407645377103056e-05,
|
| 548 |
+
"loss": 0.7168,
|
| 549 |
+
"step": 385
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.975,
|
| 553 |
+
"grad_norm": 0.9876998066902161,
|
| 554 |
+
"learning_rate": 2.424577616538173e-05,
|
| 555 |
+
"loss": 0.6452,
|
| 556 |
+
"step": 390
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.9875,
|
| 560 |
+
"grad_norm": 0.9951984286308289,
|
| 561 |
+
"learning_rate": 2.4082151592214717e-05,
|
| 562 |
+
"loss": 0.6772,
|
| 563 |
+
"step": 395
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.0,
|
| 567 |
+
"grad_norm": 0.9820576906204224,
|
| 568 |
+
"learning_rate": 2.3916802722626972e-05,
|
| 569 |
+
"loss": 0.6617,
|
| 570 |
+
"step": 400
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.0125,
|
| 574 |
+
"grad_norm": 0.9402483105659485,
|
| 575 |
+
"learning_rate": 2.37497609490106e-05,
|
| 576 |
+
"loss": 0.5926,
|
| 577 |
+
"step": 405
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.025,
|
| 581 |
+
"grad_norm": 1.0088366270065308,
|
| 582 |
+
"learning_rate": 2.3581057985164857e-05,
|
| 583 |
+
"loss": 0.5814,
|
| 584 |
+
"step": 410
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.0375,
|
| 588 |
+
"grad_norm": 1.1835291385650635,
|
| 589 |
+
"learning_rate": 2.3410725860275092e-05,
|
| 590 |
+
"loss": 0.5888,
|
| 591 |
+
"step": 415
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.05,
|
| 595 |
+
"grad_norm": 1.0397685766220093,
|
| 596 |
+
"learning_rate": 2.323879691283184e-05,
|
| 597 |
+
"loss": 0.6096,
|
| 598 |
+
"step": 420
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.0625,
|
| 602 |
+
"grad_norm": 1.2611539363861084,
|
| 603 |
+
"learning_rate": 2.3065303784491174e-05,
|
| 604 |
+
"loss": 0.5664,
|
| 605 |
+
"step": 425
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.075,
|
| 609 |
+
"grad_norm": 1.0762382745742798,
|
| 610 |
+
"learning_rate": 2.2890279413877512e-05,
|
| 611 |
+
"loss": 0.5758,
|
| 612 |
+
"step": 430
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.0875,
|
| 616 |
+
"grad_norm": 1.06105375289917,
|
| 617 |
+
"learning_rate": 2.2713757030330046e-05,
|
| 618 |
+
"loss": 0.554,
|
| 619 |
+
"step": 435
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.1,
|
| 623 |
+
"grad_norm": 1.3117653131484985,
|
| 624 |
+
"learning_rate": 2.253577014759395e-05,
|
| 625 |
+
"loss": 0.5716,
|
| 626 |
+
"step": 440
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.1125,
|
| 630 |
+
"grad_norm": 1.1829787492752075,
|
| 631 |
+
"learning_rate": 2.2356352557457624e-05,
|
| 632 |
+
"loss": 0.5253,
|
| 633 |
+
"step": 445
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.125,
|
| 637 |
+
"grad_norm": 1.2821694612503052,
|
| 638 |
+
"learning_rate": 2.217553832333714e-05,
|
| 639 |
+
"loss": 0.5658,
|
| 640 |
+
"step": 450
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.1375,
|
| 644 |
+
"grad_norm": 1.1681571006774902,
|
| 645 |
+
"learning_rate": 2.1993361773809102e-05,
|
| 646 |
+
"loss": 0.5154,
|
| 647 |
+
"step": 455
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.15,
|
| 651 |
+
"grad_norm": 1.218019723892212,
|
| 652 |
+
"learning_rate": 2.1809857496093203e-05,
|
| 653 |
+
"loss": 0.5681,
|
| 654 |
+
"step": 460
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.1625,
|
| 658 |
+
"grad_norm": 1.1740199327468872,
|
| 659 |
+
"learning_rate": 2.162506032948561e-05,
|
| 660 |
+
"loss": 0.4932,
|
| 661 |
+
"step": 465
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.175,
|
| 665 |
+
"grad_norm": 1.0119123458862305,
|
| 666 |
+
"learning_rate": 2.143900535874457e-05,
|
| 667 |
+
"loss": 0.5498,
|
| 668 |
+
"step": 470
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.1875,
|
| 672 |
+
"grad_norm": 1.17697012424469,
|
| 673 |
+
"learning_rate": 2.1251727907429357e-05,
|
| 674 |
+
"loss": 0.5778,
|
| 675 |
+
"step": 475
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.2,
|
| 679 |
+
"grad_norm": 1.0151280164718628,
|
| 680 |
+
"learning_rate": 2.1063263531193905e-05,
|
| 681 |
+
"loss": 0.5157,
|
| 682 |
+
"step": 480
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.2125,
|
| 686 |
+
"grad_norm": 1.1586745977401733,
|
| 687 |
+
"learning_rate": 2.0873648011036373e-05,
|
| 688 |
+
"loss": 0.5108,
|
| 689 |
+
"step": 485
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.225,
|
| 693 |
+
"grad_norm": 1.2907254695892334,
|
| 694 |
+
"learning_rate": 2.0682917346505917e-05,
|
| 695 |
+
"loss": 0.5131,
|
| 696 |
+
"step": 490
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.2375,
|
| 700 |
+
"grad_norm": 1.1124744415283203,
|
| 701 |
+
"learning_rate": 2.0491107748867985e-05,
|
| 702 |
+
"loss": 0.5237,
|
| 703 |
+
"step": 495
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.25,
|
| 707 |
+
"grad_norm": 1.2544411420822144,
|
| 708 |
+
"learning_rate": 2.0298255634229426e-05,
|
| 709 |
+
"loss": 0.5386,
|
| 710 |
+
"step": 500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.2625,
|
| 714 |
+
"grad_norm": 1.0245000123977661,
|
| 715 |
+
"learning_rate": 2.0104397616624646e-05,
|
| 716 |
+
"loss": 0.5461,
|
| 717 |
+
"step": 505
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.275,
|
| 721 |
+
"grad_norm": 1.1500096321105957,
|
| 722 |
+
"learning_rate": 1.990957050106428e-05,
|
| 723 |
+
"loss": 0.5202,
|
| 724 |
+
"step": 510
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.2875,
|
| 728 |
+
"grad_norm": 0.9860815405845642,
|
| 729 |
+
"learning_rate": 1.9713811276547527e-05,
|
| 730 |
+
"loss": 0.505,
|
| 731 |
+
"step": 515
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.3,
|
| 735 |
+
"grad_norm": 1.0547618865966797,
|
| 736 |
+
"learning_rate": 1.9517157109039614e-05,
|
| 737 |
+
"loss": 0.4947,
|
| 738 |
+
"step": 520
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.3125,
|
| 742 |
+
"grad_norm": 1.0779049396514893,
|
| 743 |
+
"learning_rate": 1.9319645334415594e-05,
|
| 744 |
+
"loss": 0.5056,
|
| 745 |
+
"step": 525
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.325,
|
| 749 |
+
"grad_norm": 1.2494122982025146,
|
| 750 |
+
"learning_rate": 1.9121313451371957e-05,
|
| 751 |
+
"loss": 0.5309,
|
| 752 |
+
"step": 530
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.3375,
|
| 756 |
+
"grad_norm": 1.2767302989959717,
|
| 757 |
+
"learning_rate": 1.8922199114307297e-05,
|
| 758 |
+
"loss": 0.5095,
|
| 759 |
+
"step": 535
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.35,
|
| 763 |
+
"grad_norm": 1.0909416675567627,
|
| 764 |
+
"learning_rate": 1.872234012617339e-05,
|
| 765 |
+
"loss": 0.456,
|
| 766 |
+
"step": 540
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.3625,
|
| 770 |
+
"grad_norm": 1.008742332458496,
|
| 771 |
+
"learning_rate": 1.8521774431298116e-05,
|
| 772 |
+
"loss": 0.4669,
|
| 773 |
+
"step": 545
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.375,
|
| 777 |
+
"grad_norm": 1.1449681520462036,
|
| 778 |
+
"learning_rate": 1.8320540108181518e-05,
|
| 779 |
+
"loss": 0.5324,
|
| 780 |
+
"step": 550
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.3875,
|
| 784 |
+
"grad_norm": 1.0932765007019043,
|
| 785 |
+
"learning_rate": 1.8118675362266388e-05,
|
| 786 |
+
"loss": 0.4762,
|
| 787 |
+
"step": 555
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.4,
|
| 791 |
+
"grad_norm": 1.0657200813293457,
|
| 792 |
+
"learning_rate": 1.791621851868476e-05,
|
| 793 |
+
"loss": 0.4846,
|
| 794 |
+
"step": 560
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.4125,
|
| 798 |
+
"grad_norm": 1.1053435802459717,
|
| 799 |
+
"learning_rate": 1.771320801498165e-05,
|
| 800 |
+
"loss": 0.5211,
|
| 801 |
+
"step": 565
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.425,
|
| 805 |
+
"grad_norm": 1.0770117044448853,
|
| 806 |
+
"learning_rate": 1.75096823938175e-05,
|
| 807 |
+
"loss": 0.5032,
|
| 808 |
+
"step": 570
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.4375,
|
| 812 |
+
"grad_norm": 1.2533434629440308,
|
| 813 |
+
"learning_rate": 1.7305680295650638e-05,
|
| 814 |
+
"loss": 0.4753,
|
| 815 |
+
"step": 575
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.45,
|
| 819 |
+
"grad_norm": 1.1737210750579834,
|
| 820 |
+
"learning_rate": 1.710124045140117e-05,
|
| 821 |
+
"loss": 0.4317,
|
| 822 |
+
"step": 580
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.4625,
|
| 826 |
+
"grad_norm": 1.1556503772735596,
|
| 827 |
+
"learning_rate": 1.6896401675097696e-05,
|
| 828 |
+
"loss": 0.4803,
|
| 829 |
+
"step": 585
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.475,
|
| 833 |
+
"grad_norm": 1.074533224105835,
|
| 834 |
+
"learning_rate": 1.6691202856508267e-05,
|
| 835 |
+
"loss": 0.459,
|
| 836 |
+
"step": 590
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.4875,
|
| 840 |
+
"grad_norm": 1.2261135578155518,
|
| 841 |
+
"learning_rate": 1.6485682953756945e-05,
|
| 842 |
+
"loss": 0.4701,
|
| 843 |
+
"step": 595
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.5,
|
| 847 |
+
"grad_norm": 1.1798652410507202,
|
| 848 |
+
"learning_rate": 1.627988098592737e-05,
|
| 849 |
+
"loss": 0.5234,
|
| 850 |
+
"step": 600
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.5125,
|
| 854 |
+
"grad_norm": 1.1400408744812012,
|
| 855 |
+
"learning_rate": 1.6073836025654805e-05,
|
| 856 |
+
"loss": 0.44,
|
| 857 |
+
"step": 605
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.525,
|
| 861 |
+
"grad_norm": 1.1641569137573242,
|
| 862 |
+
"learning_rate": 1.5867587191707962e-05,
|
| 863 |
+
"loss": 0.4535,
|
| 864 |
+
"step": 610
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.5375,
|
| 868 |
+
"grad_norm": 1.1707454919815063,
|
| 869 |
+
"learning_rate": 1.5661173641562085e-05,
|
| 870 |
+
"loss": 0.4404,
|
| 871 |
+
"step": 615
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.55,
|
| 875 |
+
"grad_norm": 1.2437214851379395,
|
| 876 |
+
"learning_rate": 1.5454634563964686e-05,
|
| 877 |
+
"loss": 0.4599,
|
| 878 |
+
"step": 620
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.5625,
|
| 882 |
+
"grad_norm": 1.313765287399292,
|
| 883 |
+
"learning_rate": 1.524800917149538e-05,
|
| 884 |
+
"loss": 0.4303,
|
| 885 |
+
"step": 625
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.575,
|
| 889 |
+
"grad_norm": 1.2089906930923462,
|
| 890 |
+
"learning_rate": 1.5041336693121106e-05,
|
| 891 |
+
"loss": 0.4079,
|
| 892 |
+
"step": 630
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.5875,
|
| 896 |
+
"grad_norm": 1.2107923030853271,
|
| 897 |
+
"learning_rate": 1.4834656366748348e-05,
|
| 898 |
+
"loss": 0.4414,
|
| 899 |
+
"step": 635
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.6,
|
| 903 |
+
"grad_norm": 1.0747815370559692,
|
| 904 |
+
"learning_rate": 1.462800743177355e-05,
|
| 905 |
+
"loss": 0.4232,
|
| 906 |
+
"step": 640
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.6125,
|
| 910 |
+
"grad_norm": 1.1553806066513062,
|
| 911 |
+
"learning_rate": 1.4421429121633342e-05,
|
| 912 |
+
"loss": 0.4199,
|
| 913 |
+
"step": 645
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.625,
|
| 917 |
+
"grad_norm": 1.0519508123397827,
|
| 918 |
+
"learning_rate": 1.4214960656355843e-05,
|
| 919 |
+
"loss": 0.4449,
|
| 920 |
+
"step": 650
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.6375,
|
| 924 |
+
"grad_norm": 1.206945776939392,
|
| 925 |
+
"learning_rate": 1.400864123511451e-05,
|
| 926 |
+
"loss": 0.4428,
|
| 927 |
+
"step": 655
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.65,
|
| 931 |
+
"grad_norm": 1.1311391592025757,
|
| 932 |
+
"learning_rate": 1.3802510028785977e-05,
|
| 933 |
+
"loss": 0.4549,
|
| 934 |
+
"step": 660
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.6625,
|
| 938 |
+
"grad_norm": 1.1363489627838135,
|
| 939 |
+
"learning_rate": 1.3596606172513235e-05,
|
| 940 |
+
"loss": 0.4562,
|
| 941 |
+
"step": 665
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.675,
|
| 945 |
+
"grad_norm": 1.1651484966278076,
|
| 946 |
+
"learning_rate": 1.3390968758275628e-05,
|
| 947 |
+
"loss": 0.4394,
|
| 948 |
+
"step": 670
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.6875,
|
| 952 |
+
"grad_norm": 1.0412386655807495,
|
| 953 |
+
"learning_rate": 1.3185636827467006e-05,
|
| 954 |
+
"loss": 0.4657,
|
| 955 |
+
"step": 675
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.7,
|
| 959 |
+
"grad_norm": 1.0228976011276245,
|
| 960 |
+
"learning_rate": 1.2980649363483511e-05,
|
| 961 |
+
"loss": 0.4393,
|
| 962 |
+
"step": 680
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.7125,
|
| 966 |
+
"grad_norm": 1.1123242378234863,
|
| 967 |
+
"learning_rate": 1.277604528432237e-05,
|
| 968 |
+
"loss": 0.4142,
|
| 969 |
+
"step": 685
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.725,
|
| 973 |
+
"grad_norm": 1.1547119617462158,
|
| 974 |
+
"learning_rate": 1.2571863435193088e-05,
|
| 975 |
+
"loss": 0.4662,
|
| 976 |
+
"step": 690
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.7375,
|
| 980 |
+
"grad_norm": 1.1411031484603882,
|
| 981 |
+
"learning_rate": 1.2368142581142479e-05,
|
| 982 |
+
"loss": 0.4317,
|
| 983 |
+
"step": 695
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.75,
|
| 987 |
+
"grad_norm": 1.362444281578064,
|
| 988 |
+
"learning_rate": 1.2164921399694899e-05,
|
| 989 |
+
"loss": 0.4444,
|
| 990 |
+
"step": 700
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.7625,
|
| 994 |
+
"grad_norm": 1.3785865306854248,
|
| 995 |
+
"learning_rate": 1.1962238473509122e-05,
|
| 996 |
+
"loss": 0.4037,
|
| 997 |
+
"step": 705
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.775,
|
| 1001 |
+
"grad_norm": 1.1370779275894165,
|
| 1002 |
+
"learning_rate": 1.1760132283053199e-05,
|
| 1003 |
+
"loss": 0.4188,
|
| 1004 |
+
"step": 710
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.7875,
|
| 1008 |
+
"grad_norm": 1.2741678953170776,
|
| 1009 |
+
"learning_rate": 1.1558641199298728e-05,
|
| 1010 |
+
"loss": 0.4074,
|
| 1011 |
+
"step": 715
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.8,
|
| 1015 |
+
"grad_norm": 1.2674919366836548,
|
| 1016 |
+
"learning_rate": 1.1357803476435919e-05,
|
| 1017 |
+
"loss": 0.3899,
|
| 1018 |
+
"step": 720
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.8125,
|
| 1022 |
+
"grad_norm": 1.211328148841858,
|
| 1023 |
+
"learning_rate": 1.1157657244610825e-05,
|
| 1024 |
+
"loss": 0.3987,
|
| 1025 |
+
"step": 725
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.825,
|
| 1029 |
+
"grad_norm": 1.5080320835113525,
|
| 1030 |
+
"learning_rate": 1.095824050268613e-05,
|
| 1031 |
+
"loss": 0.4008,
|
| 1032 |
+
"step": 730
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.8375,
|
| 1036 |
+
"grad_norm": 1.1555554866790771,
|
| 1037 |
+
"learning_rate": 1.0759591111026837e-05,
|
| 1038 |
+
"loss": 0.4042,
|
| 1039 |
+
"step": 735
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.85,
|
| 1043 |
+
"grad_norm": 1.4236955642700195,
|
| 1044 |
+
"learning_rate": 1.0561746784312277e-05,
|
| 1045 |
+
"loss": 0.4372,
|
| 1046 |
+
"step": 740
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.8625,
|
| 1050 |
+
"grad_norm": 1.066102147102356,
|
| 1051 |
+
"learning_rate": 1.036474508437579e-05,
|
| 1052 |
+
"loss": 0.3984,
|
| 1053 |
+
"step": 745
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.875,
|
| 1057 |
+
"grad_norm": 1.2895863056182861,
|
| 1058 |
+
"learning_rate": 1.0168623413073387e-05,
|
| 1059 |
+
"loss": 0.393,
|
| 1060 |
+
"step": 750
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.8875,
|
| 1064 |
+
"grad_norm": 1.2252490520477295,
|
| 1065 |
+
"learning_rate": 9.973419005182792e-06,
|
| 1066 |
+
"loss": 0.391,
|
| 1067 |
+
"step": 755
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.9,
|
| 1071 |
+
"grad_norm": 1.3331528902053833,
|
| 1072 |
+
"learning_rate": 9.779168921334232e-06,
|
| 1073 |
+
"loss": 0.4127,
|
| 1074 |
+
"step": 760
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.9125,
|
| 1078 |
+
"grad_norm": 1.3168723583221436,
|
| 1079 |
+
"learning_rate": 9.585910040974282e-06,
|
| 1080 |
+
"loss": 0.3986,
|
| 1081 |
+
"step": 765
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.925,
|
| 1085 |
+
"grad_norm": 1.1572790145874023,
|
| 1086 |
+
"learning_rate": 9.393679055364054e-06,
|
| 1087 |
+
"loss": 0.3686,
|
| 1088 |
+
"step": 770
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.9375,
|
| 1092 |
+
"grad_norm": 1.1793782711029053,
|
| 1093 |
+
"learning_rate": 9.20251246061322e-06,
|
| 1094 |
+
"loss": 0.3819,
|
| 1095 |
+
"step": 775
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.95,
|
| 1099 |
+
"grad_norm": 1.2285219430923462,
|
| 1100 |
+
"learning_rate": 9.01244655075097e-06,
|
| 1101 |
+
"loss": 0.4054,
|
| 1102 |
+
"step": 780
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.9625,
|
| 1106 |
+
"grad_norm": 1.2546799182891846,
|
| 1107 |
+
"learning_rate": 8.823517410835427e-06,
|
| 1108 |
+
"loss": 0.3962,
|
| 1109 |
+
"step": 785
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.975,
|
| 1113 |
+
"grad_norm": 1.2645176649093628,
|
| 1114 |
+
"learning_rate": 8.635760910102697e-06,
|
| 1115 |
+
"loss": 0.3819,
|
| 1116 |
+
"step": 790
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.9875,
|
| 1120 |
+
"grad_norm": 1.1737213134765625,
|
| 1121 |
+
"learning_rate": 8.449212695156873e-06,
|
| 1122 |
+
"loss": 0.3666,
|
| 1123 |
+
"step": 795
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 2.0,
|
| 1127 |
+
"grad_norm": 1.5198034048080444,
|
| 1128 |
+
"learning_rate": 8.263908183202348e-06,
|
| 1129 |
+
"loss": 0.3473,
|
| 1130 |
+
"step": 800
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.0125,
|
| 1134 |
+
"grad_norm": 1.1322892904281616,
|
| 1135 |
+
"learning_rate": 8.079882555319685e-06,
|
| 1136 |
+
"loss": 0.3098,
|
| 1137 |
+
"step": 805
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 2.025,
|
| 1141 |
+
"grad_norm": 1.3172906637191772,
|
| 1142 |
+
"learning_rate": 7.897170749786242e-06,
|
| 1143 |
+
"loss": 0.3046,
|
| 1144 |
+
"step": 810
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 2.0375,
|
| 1148 |
+
"grad_norm": 1.1878982782363892,
|
| 1149 |
+
"learning_rate": 7.715807455443015e-06,
|
| 1150 |
+
"loss": 0.3438,
|
| 1151 |
+
"step": 815
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 2.05,
|
| 1155 |
+
"grad_norm": 1.3433789014816284,
|
| 1156 |
+
"learning_rate": 7.5358271051086975e-06,
|
| 1157 |
+
"loss": 0.2994,
|
| 1158 |
+
"step": 820
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.0625,
|
| 1162 |
+
"grad_norm": 1.3891838788986206,
|
| 1163 |
+
"learning_rate": 7.357263869042479e-06,
|
| 1164 |
+
"loss": 0.3286,
|
| 1165 |
+
"step": 825
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.075,
|
| 1169 |
+
"grad_norm": 1.303449034690857,
|
| 1170 |
+
"learning_rate": 7.18015164845662e-06,
|
| 1171 |
+
"loss": 0.3307,
|
| 1172 |
+
"step": 830
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.0875,
|
| 1176 |
+
"grad_norm": 1.2112329006195068,
|
| 1177 |
+
"learning_rate": 7.0045240690800975e-06,
|
| 1178 |
+
"loss": 0.3302,
|
| 1179 |
+
"step": 835
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 2.1,
|
| 1183 |
+
"grad_norm": 1.2413291931152344,
|
| 1184 |
+
"learning_rate": 6.830414474774594e-06,
|
| 1185 |
+
"loss": 0.3125,
|
| 1186 |
+
"step": 840
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 2.1125,
|
| 1190 |
+
"grad_norm": 1.4404971599578857,
|
| 1191 |
+
"learning_rate": 6.657855921203991e-06,
|
| 1192 |
+
"loss": 0.3063,
|
| 1193 |
+
"step": 845
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 2.125,
|
| 1197 |
+
"grad_norm": 1.0838373899459839,
|
| 1198 |
+
"learning_rate": 6.4868811695585465e-06,
|
| 1199 |
+
"loss": 0.327,
|
| 1200 |
+
"step": 850
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 2.1375,
|
| 1204 |
+
"grad_norm": 1.1617887020111084,
|
| 1205 |
+
"learning_rate": 6.317522680335015e-06,
|
| 1206 |
+
"loss": 0.2965,
|
| 1207 |
+
"step": 855
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 2.15,
|
| 1211 |
+
"grad_norm": 1.256178617477417,
|
| 1212 |
+
"learning_rate": 6.149812607173844e-06,
|
| 1213 |
+
"loss": 0.2983,
|
| 1214 |
+
"step": 860
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.1625,
|
| 1218 |
+
"grad_norm": 1.2694519758224487,
|
| 1219 |
+
"learning_rate": 5.983782790754624e-06,
|
| 1220 |
+
"loss": 0.2831,
|
| 1221 |
+
"step": 865
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 2.175,
|
| 1225 |
+
"grad_norm": 1.1719343662261963,
|
| 1226 |
+
"learning_rate": 5.819464752750978e-06,
|
| 1227 |
+
"loss": 0.2873,
|
| 1228 |
+
"step": 870
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 2.1875,
|
| 1232 |
+
"grad_norm": 1.22670578956604,
|
| 1233 |
+
"learning_rate": 5.656889689845964e-06,
|
| 1234 |
+
"loss": 0.2979,
|
| 1235 |
+
"step": 875
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 2.2,
|
| 1239 |
+
"grad_norm": 1.1985101699829102,
|
| 1240 |
+
"learning_rate": 5.496088467809243e-06,
|
| 1241 |
+
"loss": 0.3081,
|
| 1242 |
+
"step": 880
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 2.2125,
|
| 1246 |
+
"grad_norm": 1.339827060699463,
|
| 1247 |
+
"learning_rate": 5.337091615637039e-06,
|
| 1248 |
+
"loss": 0.2944,
|
| 1249 |
+
"step": 885
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 2.225,
|
| 1253 |
+
"grad_norm": 1.3325270414352417,
|
| 1254 |
+
"learning_rate": 5.179929319756025e-06,
|
| 1255 |
+
"loss": 0.2841,
|
| 1256 |
+
"step": 890
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.2375,
|
| 1260 |
+
"grad_norm": 1.170397162437439,
|
| 1261 |
+
"learning_rate": 5.024631418292275e-06,
|
| 1262 |
+
"loss": 0.3372,
|
| 1263 |
+
"step": 895
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 2.25,
|
| 1267 |
+
"grad_norm": 1.2334169149398804,
|
| 1268 |
+
"learning_rate": 4.871227395406347e-06,
|
| 1269 |
+
"loss": 0.2963,
|
| 1270 |
+
"step": 900
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 2.2625,
|
| 1274 |
+
"grad_norm": 1.1435058116912842,
|
| 1275 |
+
"learning_rate": 4.719746375695505e-06,
|
| 1276 |
+
"loss": 0.2856,
|
| 1277 |
+
"step": 905
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 2.275,
|
| 1281 |
+
"grad_norm": 1.2726664543151855,
|
| 1282 |
+
"learning_rate": 4.570217118664321e-06,
|
| 1283 |
+
"loss": 0.285,
|
| 1284 |
+
"step": 910
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 2.2875,
|
| 1288 |
+
"grad_norm": 1.1388657093048096,
|
| 1289 |
+
"learning_rate": 4.422668013264467e-06,
|
| 1290 |
+
"loss": 0.2911,
|
| 1291 |
+
"step": 915
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 2.3,
|
| 1295 |
+
"grad_norm": 1.2628424167633057,
|
| 1296 |
+
"learning_rate": 4.277127072504945e-06,
|
| 1297 |
+
"loss": 0.3115,
|
| 1298 |
+
"step": 920
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.3125,
|
| 1302 |
+
"grad_norm": 1.4541245698928833,
|
| 1303 |
+
"learning_rate": 4.133621928133666e-06,
|
| 1304 |
+
"loss": 0.2879,
|
| 1305 |
+
"step": 925
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 2.325,
|
| 1309 |
+
"grad_norm": 1.2590599060058594,
|
| 1310 |
+
"learning_rate": 3.992179825391391e-06,
|
| 1311 |
+
"loss": 0.2664,
|
| 1312 |
+
"step": 930
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 2.3375,
|
| 1316 |
+
"grad_norm": 1.1543575525283813,
|
| 1317 |
+
"learning_rate": 3.852827617839085e-06,
|
| 1318 |
+
"loss": 0.2709,
|
| 1319 |
+
"step": 935
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 2.35,
|
| 1323 |
+
"grad_norm": 1.2745298147201538,
|
| 1324 |
+
"learning_rate": 3.715591762259633e-06,
|
| 1325 |
+
"loss": 0.3077,
|
| 1326 |
+
"step": 940
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 2.3625,
|
| 1330 |
+
"grad_norm": 1.2039170265197754,
|
| 1331 |
+
"learning_rate": 3.5804983136348313e-06,
|
| 1332 |
+
"loss": 0.299,
|
| 1333 |
+
"step": 945
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 2.375,
|
| 1337 |
+
"grad_norm": 1.2109931707382202,
|
| 1338 |
+
"learning_rate": 3.4475729201987693e-06,
|
| 1339 |
+
"loss": 0.2897,
|
| 1340 |
+
"step": 950
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 2.3875,
|
| 1344 |
+
"grad_norm": 1.2224966287612915,
|
| 1345 |
+
"learning_rate": 3.3168408185683153e-06,
|
| 1346 |
+
"loss": 0.2615,
|
| 1347 |
+
"step": 955
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 2.4,
|
| 1351 |
+
"grad_norm": 1.1242839097976685,
|
| 1352 |
+
"learning_rate": 3.1883268289518308e-06,
|
| 1353 |
+
"loss": 0.2877,
|
| 1354 |
+
"step": 960
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 2.4125,
|
| 1358 |
+
"grad_norm": 1.2263603210449219,
|
| 1359 |
+
"learning_rate": 3.062055350436923e-06,
|
| 1360 |
+
"loss": 0.2895,
|
| 1361 |
+
"step": 965
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 2.425,
|
| 1365 |
+
"grad_norm": 1.2495533227920532,
|
| 1366 |
+
"learning_rate": 2.9380503563581295e-06,
|
| 1367 |
+
"loss": 0.3306,
|
| 1368 |
+
"step": 970
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 2.4375,
|
| 1372 |
+
"grad_norm": 1.2463157176971436,
|
| 1373 |
+
"learning_rate": 2.8163353897454474e-06,
|
| 1374 |
+
"loss": 0.2893,
|
| 1375 |
+
"step": 975
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 2.45,
|
| 1379 |
+
"grad_norm": 1.2636357545852661,
|
| 1380 |
+
"learning_rate": 2.696933558854589e-06,
|
| 1381 |
+
"loss": 0.2875,
|
| 1382 |
+
"step": 980
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 2.4625,
|
| 1386 |
+
"grad_norm": 1.3296539783477783,
|
| 1387 |
+
"learning_rate": 2.5798675327796994e-06,
|
| 1388 |
+
"loss": 0.2906,
|
| 1389 |
+
"step": 985
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 2.475,
|
| 1393 |
+
"grad_norm": 1.4464213848114014,
|
| 1394 |
+
"learning_rate": 2.465159537149563e-06,
|
| 1395 |
+
"loss": 0.3093,
|
| 1396 |
+
"step": 990
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 2.4875,
|
| 1400 |
+
"grad_norm": 1.2355504035949707,
|
| 1401 |
+
"learning_rate": 2.352831349907904e-06,
|
| 1402 |
+
"loss": 0.2875,
|
| 1403 |
+
"step": 995
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 2.5,
|
| 1407 |
+
"grad_norm": 1.3024587631225586,
|
| 1408 |
+
"learning_rate": 2.2429042971787366e-06,
|
| 1409 |
+
"loss": 0.2882,
|
| 1410 |
+
"step": 1000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 2.5125,
|
| 1414 |
+
"grad_norm": 1.4043537378311157,
|
| 1415 |
+
"learning_rate": 2.1353992492174944e-06,
|
| 1416 |
+
"loss": 0.2705,
|
| 1417 |
+
"step": 1005
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 2.525,
|
| 1421 |
+
"grad_norm": 1.4674831628799438,
|
| 1422 |
+
"learning_rate": 2.03033661644868e-06,
|
| 1423 |
+
"loss": 0.2741,
|
| 1424 |
+
"step": 1010
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 2.5375,
|
| 1428 |
+
"grad_norm": 1.1488456726074219,
|
| 1429 |
+
"learning_rate": 1.927736345590839e-06,
|
| 1430 |
+
"loss": 0.3123,
|
| 1431 |
+
"step": 1015
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 2.55,
|
| 1435 |
+
"grad_norm": 1.2895066738128662,
|
| 1436 |
+
"learning_rate": 1.8276179158695822e-06,
|
| 1437 |
+
"loss": 0.2657,
|
| 1438 |
+
"step": 1020
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 2.5625,
|
| 1442 |
+
"grad_norm": 1.3622872829437256,
|
| 1443 |
+
"learning_rate": 1.7300003353193117e-06,
|
| 1444 |
+
"loss": 0.2768,
|
| 1445 |
+
"step": 1025
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 2.575,
|
| 1449 |
+
"grad_norm": 1.2856098413467407,
|
| 1450 |
+
"learning_rate": 1.6349021371744833e-06,
|
| 1451 |
+
"loss": 0.2654,
|
| 1452 |
+
"step": 1030
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 2.5875,
|
| 1456 |
+
"grad_norm": 1.2219773530960083,
|
| 1457 |
+
"learning_rate": 1.5423413763509376e-06,
|
| 1458 |
+
"loss": 0.2553,
|
| 1459 |
+
"step": 1035
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 2.6,
|
| 1463 |
+
"grad_norm": 1.32290780544281,
|
| 1464 |
+
"learning_rate": 1.452335626018086e-06,
|
| 1465 |
+
"loss": 0.2954,
|
| 1466 |
+
"step": 1040
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 2.6125,
|
| 1470 |
+
"grad_norm": 1.2187912464141846,
|
| 1471 |
+
"learning_rate": 1.3649019742625624e-06,
|
| 1472 |
+
"loss": 0.2776,
|
| 1473 |
+
"step": 1045
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 2.625,
|
| 1477 |
+
"grad_norm": 1.0754672288894653,
|
| 1478 |
+
"learning_rate": 1.280057020843927e-06,
|
| 1479 |
+
"loss": 0.3043,
|
| 1480 |
+
"step": 1050
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 2.6375,
|
| 1484 |
+
"grad_norm": 1.2335586547851562,
|
| 1485 |
+
"learning_rate": 1.1978168740431311e-06,
|
| 1486 |
+
"loss": 0.2971,
|
| 1487 |
+
"step": 1055
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 2.65,
|
| 1491 |
+
"grad_norm": 1.2342262268066406,
|
| 1492 |
+
"learning_rate": 1.1181971476042736e-06,
|
| 1493 |
+
"loss": 0.293,
|
| 1494 |
+
"step": 1060
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 2.6625,
|
| 1498 |
+
"grad_norm": 1.2400048971176147,
|
| 1499 |
+
"learning_rate": 1.041212957770225e-06,
|
| 1500 |
+
"loss": 0.2714,
|
| 1501 |
+
"step": 1065
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 2.675,
|
| 1505 |
+
"grad_norm": 1.2360520362854004,
|
| 1506 |
+
"learning_rate": 9.668789204127454e-07,
|
| 1507 |
+
"loss": 0.3051,
|
| 1508 |
+
"step": 1070
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 2.6875,
|
| 1512 |
+
"grad_norm": 1.135083556175232,
|
| 1513 |
+
"learning_rate": 8.952091482575825e-07,
|
| 1514 |
+
"loss": 0.2639,
|
| 1515 |
+
"step": 1075
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 2.7,
|
| 1519 |
+
"grad_norm": 1.126283884048462,
|
| 1520 |
+
"learning_rate": 8.262172482050711e-07,
|
| 1521 |
+
"loss": 0.2521,
|
| 1522 |
+
"step": 1080
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 2.7125,
|
| 1526 |
+
"grad_norm": 1.2299840450286865,
|
| 1527 |
+
"learning_rate": 7.599163187468111e-07,
|
| 1528 |
+
"loss": 0.2674,
|
| 1529 |
+
"step": 1085
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 2.725,
|
| 1533 |
+
"grad_norm": 1.176446795463562,
|
| 1534 |
+
"learning_rate": 6.963189474788378e-07,
|
| 1535 |
+
"loss": 0.2721,
|
| 1536 |
+
"step": 1090
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 2.7375,
|
| 1540 |
+
"grad_norm": 1.374687910079956,
|
| 1541 |
+
"learning_rate": 6.354372087117927e-07,
|
| 1542 |
+
"loss": 0.2971,
|
| 1543 |
+
"step": 1095
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 2.75,
|
| 1547 |
+
"grad_norm": 1.1717859506607056,
|
| 1548 |
+
"learning_rate": 5.772826611785631e-07,
|
| 1549 |
+
"loss": 0.3002,
|
| 1550 |
+
"step": 1100
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 2.7625,
|
| 1554 |
+
"grad_norm": 1.1102797985076904,
|
| 1555 |
+
"learning_rate": 5.218663458397716e-07,
|
| 1556 |
+
"loss": 0.2612,
|
| 1557 |
+
"step": 1105
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 2.775,
|
| 1561 |
+
"grad_norm": 1.257702350616455,
|
| 1562 |
+
"learning_rate": 4.6919878378760885e-07,
|
| 1563 |
+
"loss": 0.2504,
|
| 1564 |
+
"step": 1110
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 2.7875,
|
| 1568 |
+
"grad_norm": 1.372108817100525,
|
| 1569 |
+
"learning_rate": 4.1928997424834525e-07,
|
| 1570 |
+
"loss": 0.2392,
|
| 1571 |
+
"step": 1115
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 2.8,
|
| 1575 |
+
"grad_norm": 1.1753861904144287,
|
| 1576 |
+
"learning_rate": 3.721493926839042e-07,
|
| 1577 |
+
"loss": 0.2725,
|
| 1578 |
+
"step": 1120
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 2.8125,
|
| 1582 |
+
"grad_norm": 1.2952775955200195,
|
| 1583 |
+
"learning_rate": 3.277859889929147e-07,
|
| 1584 |
+
"loss": 0.2951,
|
| 1585 |
+
"step": 1125
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 2.825,
|
| 1589 |
+
"grad_norm": 1.161136507987976,
|
| 1590 |
+
"learning_rate": 2.862081858115151e-07,
|
| 1591 |
+
"loss": 0.2452,
|
| 1592 |
+
"step": 1130
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 2.8375,
|
| 1596 |
+
"grad_norm": 1.0940366983413696,
|
| 1597 |
+
"learning_rate": 2.474238769142645e-07,
|
| 1598 |
+
"loss": 0.2863,
|
| 1599 |
+
"step": 1135
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 2.85,
|
| 1603 |
+
"grad_norm": 1.2279645204544067,
|
| 1604 |
+
"learning_rate": 2.1144042571547618e-07,
|
| 1605 |
+
"loss": 0.281,
|
| 1606 |
+
"step": 1140
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 2.8625,
|
| 1610 |
+
"grad_norm": 1.3091001510620117,
|
| 1611 |
+
"learning_rate": 1.7826466387122597e-07,
|
| 1612 |
+
"loss": 0.2737,
|
| 1613 |
+
"step": 1145
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 2.875,
|
| 1617 |
+
"grad_norm": 1.1495012044906616,
|
| 1618 |
+
"learning_rate": 1.4790288998232316e-07,
|
| 1619 |
+
"loss": 0.2485,
|
| 1620 |
+
"step": 1150
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 2.8875,
|
| 1624 |
+
"grad_norm": 1.218850016593933,
|
| 1625 |
+
"learning_rate": 1.2036086839850025e-07,
|
| 1626 |
+
"loss": 0.2892,
|
| 1627 |
+
"step": 1155
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 2.9,
|
| 1631 |
+
"grad_norm": 1.2607446908950806,
|
| 1632 |
+
"learning_rate": 9.564382812399852e-08,
|
| 1633 |
+
"loss": 0.2453,
|
| 1634 |
+
"step": 1160
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 2.9125,
|
| 1638 |
+
"grad_norm": 1.2456682920455933,
|
| 1639 |
+
"learning_rate": 7.375646182482875e-08,
|
| 1640 |
+
"loss": 0.2859,
|
| 1641 |
+
"step": 1165
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 2.925,
|
| 1645 |
+
"grad_norm": 1.2477439641952515,
|
| 1646 |
+
"learning_rate": 5.470292493783891e-08,
|
| 1647 |
+
"loss": 0.2686,
|
| 1648 |
+
"step": 1170
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 2.9375,
|
| 1652 |
+
"grad_norm": 1.1869429349899292,
|
| 1653 |
+
"learning_rate": 3.848683488177851e-08,
|
| 1654 |
+
"loss": 0.3132,
|
| 1655 |
+
"step": 1175
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 2.95,
|
| 1659 |
+
"grad_norm": 1.2138971090316772,
|
| 1660 |
+
"learning_rate": 2.5111270370511417e-08,
|
| 1661 |
+
"loss": 0.2843,
|
| 1662 |
+
"step": 1180
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 2.9625,
|
| 1666 |
+
"grad_norm": 1.2493516206741333,
|
| 1667 |
+
"learning_rate": 1.4578770828511667e-08,
|
| 1668 |
+
"loss": 0.3035,
|
| 1669 |
+
"step": 1185
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 2.975,
|
| 1673 |
+
"grad_norm": 1.255171298980713,
|
| 1674 |
+
"learning_rate": 6.891335908732477e-09,
|
| 1675 |
+
"loss": 0.2631,
|
| 1676 |
+
"step": 1190
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 2.9875,
|
| 1680 |
+
"grad_norm": 1.3329716920852661,
|
| 1681 |
+
"learning_rate": 2.0504251129649378e-09,
|
| 1682 |
+
"loss": 0.2917,
|
| 1683 |
+
"step": 1195
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 3.0,
|
| 1687 |
+
"grad_norm": 1.5410183668136597,
|
| 1688 |
+
"learning_rate": 5.695751474465993e-11,
|
| 1689 |
+
"loss": 0.25,
|
| 1690 |
+
"step": 1200
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 3.0,
|
| 1694 |
+
"step": 1200,
|
| 1695 |
+
"total_flos": 1.64227002403704e+18,
|
| 1696 |
+
"train_loss": 0.565493812263012,
|
| 1697 |
+
"train_runtime": 754.0979,
|
| 1698 |
+
"train_samples_per_second": 50.878,
|
| 1699 |
+
"train_steps_per_second": 1.591
|
| 1700 |
+
}
|
| 1701 |
+
],
|
| 1702 |
+
"logging_steps": 5,
|
| 1703 |
+
"max_steps": 1200,
|
| 1704 |
+
"num_input_tokens_seen": 0,
|
| 1705 |
+
"num_train_epochs": 3,
|
| 1706 |
+
"save_steps": 2000,
|
| 1707 |
+
"stateful_callbacks": {
|
| 1708 |
+
"TrainerControl": {
|
| 1709 |
+
"args": {
|
| 1710 |
+
"should_epoch_stop": false,
|
| 1711 |
+
"should_evaluate": false,
|
| 1712 |
+
"should_log": false,
|
| 1713 |
+
"should_save": false,
|
| 1714 |
+
"should_training_stop": false
|
| 1715 |
+
},
|
| 1716 |
+
"attributes": {}
|
| 1717 |
+
}
|
| 1718 |
+
},
|
| 1719 |
+
"total_flos": 1.64227002403704e+18,
|
| 1720 |
+
"train_batch_size": 2,
|
| 1721 |
+
"trial_name": null,
|
| 1722 |
+
"trial_params": null
|
| 1723 |
+
}
|
29_128_e3_3e-5/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b779de9e0c2c622ab93d15602e493780d3d693c9e13c98d311ea921dfcaee6db
|
| 3 |
+
size 8145
|
29_128_e3_3e-5/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|